Nanomaterial Interfaces in Biology: Methods and Protocols

Methods in
Molecular Biology 1025
Paolo Bergese
Kimberly Hamad-Schifferli
Editors
Nanomaterial
Interfaces in
Biology
Methods and Protocols
METHODS IN M O L E C U L A R B I O LO G Y ™
Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For further volumes:

http://www.springer.com/series/7651
Nanomaterial Interfaces
in Biology
Methods and Protocols
Edited by
Paolo Bergese
Dipartimento di Ingegneria Meccanicae Industriale, Università degli Studi di Brescia, Brescia, Italy
Kimberly Hamad-Schifferli
Department of Mechanical Engineering, Massachusetts Institute of Technology
Cambridge, MA, USA
Editors
Paolo Bergese Kimberly Hamad-Schifferli
Dipartimento di Ingegneria Meccanicae Department of Mechanical Engineering
Industriale Massachusetts Institute of Technology
Università degli Studi di Brescia Cambridge, MA, USA
Brescia, Italy
ISSN 1064-3745 ISSN 1940-6029 (electronic)

ISBN 978-1-62703-461-6 ISBN 978-1-62703-462-3 (eBook)
DOI 10.1007/978-1-62703-462-3
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2013943590
© Springer Science+Business Media New York 2013

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for
the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be
made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Printed on acid-free paper
Humana Press is a brand of Springer

Springer is part of Springer Science+Business Media (www.springer.com)
Preface
The intersection of nanotechnology with biology has given rise to numerous ideas for new
ways to use nanotechnology for biological applications. Nanomaterials possess unique size-
and material-dependent properties, which make them attractive for improving regular bio-
medical fields, such as drug delivery, imaging, therapy, and diagnostics as well as next-level
developments, comprising activation/deactivation, mimicking, and implementation of bio-
molecular systems and functions. Consequently nanotechnology has held great potential
for novel and unique capabilities, and stirred the imagination of many scientists.
While nanotechnology and nanoscience has held great promise for revolutionizing
biology, it has been hindered by the fact that when nanomaterials are interfaced to biomol-
ecules or put into biological environments, many undesirable side effects result, such as
aggregation and nonspecific adsorption, which can compromise biological function or give
rise to negative biological responses. This can be largely attributed to a range of interface
and intermolecular interactions between the nanomaterials with the biomolecules as well as
the solvent that mediates their interactions. These interactions are difficult to control, pre-
dict, and prevent. Interface effects of inorganic surfaces have been a major issue historically,
manifesting as surface fouling of medical device implants and stents. Unfortunately, these
effects worsen or give rise to new and unexpected complications for nanoscale materials
because surface volume ratios are exponentially higher, and nanoscale surfaces have differ-
ent physical and chemical properties.
Therefore, despite the fact that we now have a high degree of control over the synthetic
properties of the nanomaterials, similar control over their interfaces to biology has yet to be
achieved. This is crucial as their biological interface ultimately determines their biological
identity and fate. In the last 10 years, the biological–nanomaterial interface has created
unprecedented challenges not only for finding useful ways to exploit nanomaterials in biol-
ogy but also in their unintentional consequences, such as environmental and toxicological
effects. For example surface fouling, nonspecific adsorption, or unexpected aggregation
and instability have prevented many of the exciting and early ideas of nanobiotechnology
from reaching fruition.
We believe a key pitfall that plagues this area is the difficulty in reproducing results,
where a huge amount of variability exists not only between different labs but also from day
to day in the same lab. This variability is almost never reported in peer-reviewed journal
papers, despite its criticality. Also, members of the nanotechnology community typically
come from chemistry, physics, and materials science, and historically are not accustomed to
providing stepwise protocols. Therefore, a handbook of detailed protocols and best prac-
tices for this field is critically needed.
While there have been some examples of individual protocols in protocol literature
such as Nature Methods, Current Protocols, Molecular Cloning, and on the National Cancer
Institute’s Nanotechnology Characterization Laboratory website, we believe that it is time
to provide a consolidated volume of step-by-step protocols in the tradition of Methods in
Molecular Biology. Methods in Molecular Biology has played a major role in providing
vi Preface
biological recipes and protocols, and in doing so has advanced biology in immeasurable
ways. By creating standards for everyone and making experimental techniques more acces-
sible, Methods in Molecular Biology has accelerated the discovery process. What used to be
difficult experimentally now can be done by nearly anyone. These new tools have benefited
the entire biology community, enabling huge leaps and bounds in scientific discovery and
applications.
Thus, we present this volume with the hopes of improving the utility of nanotechnol-
ogy as a tool to advance biological and medical sciences. While this volume is far from
comprehensive, we hope that it will serve the new and emerging community well, and
enable new capabilities and technologies that were not previously possible. The proposed
protocols predominantly deal with nanomaterials, covering many of the now classic sub-
jects, such as conjugation of nanoparticles to biomolecules and their applications in drug
delivery or imaging. Additionally, some chapters are dedicated to flat surfaces because most
of the methods for manipulating nanomaterials stem from those of flat surfaces, facilitated
by fabrication advances to shrink the dimensions of materials.
The volume is organized into three parts: (1) protocols describing synthesis, fabrica-
tion, and construction of bio-nanomaterial interfaces, (2) characterization protocols of bio-
nanomaterial interfaces, and (3) applications which utilize the bio-nanomaterial interfaces.
We would like to note that we are not including significant coverage of toxicology and the
unintended effects of nanomaterials. Due to the evolving scope of the toxicological studies
of nanomaterials, the current state of the art is still developing, and thus is difficult to cover
adequately in this volume of Methods in Molecular Biology. However, we hope that this col-
lection will aid this growing field by serving as a guide.
We gratefully acknowledge all of the authors contributing to this volume, as well as our
home institutions of the Department of Mechanical Engineering at MIT and the Department
of Mechanical and Industrial Engineering of Università degli Studi di Brescia. This work
was made possible by the UniBS-MIT-MechE faculty exchange program cosponsored by
the CARIPLO Foundation, Italy under grant 2008-2290.
Brescia, Italy Paolo Bergese

Cambridge, MA, USA Kimberly Hamad-Schifferli
Contents
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
PART I MAKING BIO–NANO INTERFACES

1 Preparation of 2 nm Gold Nanoparticles for In Vitro
and In Vivo Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Daniel F. Moyano, Bradley Duncan, and Vincent M. Rotello
2 DNA Conjugation to Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Sunho Park
3 Conjugation of Nanoparticles to Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Marie-Eve Aubin-Tam
4 Water-Solubilization and Functionalization of Semiconductor
Quantum Dots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Christina M. Tyrakowski, Adela Isovic, and Preston T. Snee
5 Synthesizing and Modifying Peptides for Chemoselective Ligation
and Assembly into Quantum Dot-Peptide Bioconjugates. . . . . . . . . . . . . . . . . 47
W. Russ Algar, Juan B. Blanco-Canosa, Rachel L. Manthe, Kimihiro Susumu,
Michael H. Stewart, Philip E. Dawson, and Igor L. Medintz
6 Reliable Methods for Silica Coating of Au Nanoparticles . . . . . . . . . . . . . . . . . 75
Isabel Pastoriza-Santos and Luis M. Liz-Marzán
7 Surface Modifications by Polymers for Biomolecule Conjugation. . . . . . . . . . . 95
Laura Sola, Marina Cretich, Francesco Damin, and Marcella Chiari
8 Functionalization Protocols of Silicon Micro/Nano-mechanical Biosensors . . . 109
Francesca Frascella and Carlo Ricciardi
PART II CHARACTERIZING BIO–NANO INTERFACES

9 Stability and Aggregation Assays of Nanoparticles in Biological Media. . . . . . . 119
James Chen Yong Kah
10 Electrochemical Measurements of DNA Melting on Surfaces. . . . . . . . . . . . . . 127
Irina Belozerova, Dongbiao Ge, and Rastislav Levicky
11 Formation and Characterization of the Nanoparticle–Protein Corona . . . . . . . 137
Marco P. Monopoli, Andrzej S. Pitek, Iseult Lynch, and Kenneth A. Dawson
12 Electrophoretic Implementation of the Solution-Depletion Method
for Measuring Protein Adsorption, Adsorption Kinetics,
and Adsorption Competition Among Multiple Proteins in Solution. . . . . . . . . 157
Hyeran Noh, Naris Barnthip, Purnendu Parhi, and Erwin A. Vogler
viii Contents
13 Hyperspectral Microscopy for Characterization of Gold Nanoparticles

in Biological Media and Cells for Toxicity Assessment . . . . . . . . . . . . . . . . . . . 167
Christin Grabinski, John Schlager, and Saber Hussain
14 Immunocytochemistry, Electron Tomography, and Energy Dispersive X-ray
Spectroscopy (EDXS) on Cryosections of Human Cancer Cells Doped
with Stimuli Responsive Polymeric Nanogels Loaded with Iron
Oxide Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Roberto Marotta, A. Falqui, A. Curcio, A. Quarta, and Teresa Pellegrino
PART III IMPLEMENTING BIO–NANO INTERFACES

15 Zwitterion Siloxane to Passivate Silica Against Nonspecific
Protein Adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Zaki G. Estephan and Joseph B. Schlenoff
16 Preparation and Characterization of DNA Block Copolymer Assemblies
Loaded with Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Xi-Jun Chen, Robert J. Hickey, and So-Jung Park
17 Polyaspartic Acid Coated Iron Oxide Nanoprobes for PET/MRI Imaging . . . 225
Taku Cowger and Jin Xie
18 Ligand Synthesis and Passivation for Silver and Large
Gold Nanoparticles for Single-Particle-Based Sensing and Spectroscopy . . . . . 237
Daniel Montiel, Emma V. Yates, Li Sun, Marissa M. Sampias, John Malona,
Erik J. Sorensen, and Haw Yang
19 Noncovalent Intracellular Drug Delivery of Hydrophobic Drugs on Au NPs. . . . . 251
Tennyson Doane and Clemens Burda
20 Modification of Carbon Nanotubes for Gene Delivery Vectors . . . . . . . . . . . . 261
Victor Ramos-Perez, Anna Cifuentes, Núria Coronas,
Ana de Pablo, and Salvador Borrós
21 Lipid-Based Nanoparticles as Nonviral Gene Delivery Vectors . . . . . . . . . . . . . 269
Daniele Pezzoli, Anna Kajaste-Rudnitski, Roberto Chiesa,
and Gabriele Candiani
22 Stabilizing Gold Nanoparticle Bioconjugates in Physiological Conditions
by PEGylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Joan Comenge and Víctor F. Puntes
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Contributors
W. RUSS ALGAR • Naval Research Laboratory, Washington, DC, USA

MARIE-EVE AUBIN-TAM • Delft University of Technology, Delft, The Netherlands
NARIS BARNTHIP • Division of Physics, Rajamangala University of Technology, Pathum
Thani, Thailand
IRINA BELOZEROVA • Polytechnic Institute of New York University, New York, NY, USA
JUAN B. BLANCO-CANOSA • Scripps Research Institute, La Jolla, CA, USA
SALVADOR BORRÓS • Institut Quimic de Sarria, Universitat Ramon Llull, Barcelona, Spain
CLEMENS BURDA • Case Western Reserve University, Cleveland, OH, USA
GABRIELE CANDIANI • INSTM (National Interuniversity Consortium of Materials Science
and Technology), Research Unit Milano Politecnico, Milan, Italy; Department of
Chemistry, Materials and Chemical Engineering “Giulio Natta”, Politecnico di Milano,
Milan, Italy
XI-JUN CHEN • Department of Chemistry, University of Pennsylvania,
Philadelphia, PA, USA
MARCELLA CHIARI • Consiglio Nazionale delle Ricerche, Istituto di Chimica del
Riconoscimento Molecolare, Milan, Italy
ROBERTO CHIESA • Department of Chemistry, Materials and Chemical Engineering
“Giulio Natta”, Politecnico di Milano, Milan, Italy
ANNA CIFUENTES • Institut Quimic de Sarria, Universitat Ramon Llull, Barcelona, Spain
JOAN COMENGE • Catalan Institute of Nanotechnology, Universitat Autònoma de
Barcelona, Barcelona, Spain
NÚRIA CORONAS • Institut Quimic de Sarria, Universitat Ramon Llull, Barcelona, Spain
TAKU COWGER • University of Georgia, Athens, GA, USA
MARINA CRETICH • Consiglio Nazionale delle Ricerche, Istituto di Chimica del
A. CURCIO • Istituto Italiano di Tecnologia, Genoa, Italy
FRANCESCO DAMIN • Consiglio Nazionale delle Ricerche, Istituto di Chimica del
KENNETH A. DAWSON • Centre for BioNano Interactions, School of Chemistry and Chemical
Biology, University College Dublin, Dublin, Ireland
PHILIP E. DAWSON • Scripps Research Institute, La Jolla, CA, USA
TENNYSON DOANE • Case Western Reserve University, Cleveland, OH, USA
BRADLEY DUNCAN • University of Massachusetts, Amherst, MA, USA
ZAKI G. ESTEPHAN • Florida State University, Tallahassee, FL, USA
A. FALQUI • Istituto Italiano di Tecnologia, Genoa, Italy
FRANCESCA FRASCELLA • Politecnico di Torino, Torino, Italy
DONGBIAO GE • Polytechnic Institute of New York University, New York, NY, USA
CHRISTIN GRABINSKI • Air Force Research Laboratory, Wright-Patterson Air Force Base,
Dayton, OH, USA
x Contributors
ROBERT J. HICKEY • Department of Chemistry, University of Pennsylvania, Philadelphia,

PA, USA
SABER HUSSAIN • Air Force Research Laboratory, Wright-Patterson Air Force Base, Dayton,
OH, USA
ADELA ISOVIC • University of Illinois at Chicago, Chicago, IL, USA
JAMES CHEN YONG KAH • National University Singapore, Singapore
ANNA KAJASTE-RUDNITSKI • Division of Regenerative Medicine, Stem Cells and Gene
Therapy, San Raffaele Telethon Institute for Gene Therapy, San Raffaele Scientific
Institute, Milan, Italy
RASTISLAV LEVICKY • Polytechnic Institute of New York University, New York, NY, USA
LUIS M. LIZ-MARZÁN • CIC biomaGUNE, Donostia - San, Sebastián, Spain
ISEULT LYNCH • Center for BioNano Interactions, School of Chemistry and Chemical
JOHN MALONA • Princeton University, Princeton, NJ, USA
RACHEL L. MANTHE • Sotera Defense Solutions, University of Maryland,
College Park, MD, USA
ROBERTO MAROTTA • Istituto Italiano di Tecnologia, Genoa, Italy
IGOR L. MEDINTZ • U.S. Naval Research Laboratory, Center for Bio/Molecular Science and
Engineering, Washington, DC, USA
MARCO P. MONOPOLI • Center for BioNano Interactions, School of Chemistry and Chemical
DANIEL MONTIEL • Princeton University, Princeton, NJ, USA
DANIEL F. MOYANO • University of Massachusetts, Amherst, MA, USA
HYERAN NOH • Department of Optometry and Vision Science, Seoul National University
and Technology, Seoul, South Korea
ANA dE PABLO • Institut Quimic de Sarria, Universitat Ramon Llull, Barcelona, Spain
PURNENDU PARHI • Department of Chemistry, Ravenshaw University, Cuttack, Orissa,
India
SO-JUNG PARK • Department of Chemistry, University of Pennsylvania,
Philadelphia, PA, USA
SUNHO PARK • Dankook University, Yongin-si, Gyeonggi-do, South Korea
ISABEL PASTORIZA-SANTOS • Departamento de Química Física, Campus Universitario,
Universidade de Vigo, Bilbao, Spain
TERESA PELLEGRINO • Istituto Italiano di Tecnologia, Genoa, Italy; National Nanotechnology
Laboratory of CNR-NANO, Lecce, Italy
DANIELE PEZZOLI • INSTM (National Interuniversity Consortium of Materials Science
and Technology), Research Unit Milano Politecnico, Milan, Italy
ANDRZEJ S. PITEK • Center for BioNano Interactions, School of Chemistry and Chemical
VÍCTOR F. PUNTES • Catalan Institute of Nanotechnology, Universitat Autònoma de
Barcelona, Barcelona, Spain
A. QUARTA • Istituto Italiano di Tecnologia, Genoa, Italy; National Nanotechnology
Laboratory of CNR-NANO, Lecce, Italy
VICTOR RAMOS-PEREZ • Institut Quimic de Sarria, Universitat Ramon Llull,
Barcelona, Spain
CARLO RICCIARDI • Politecnico di Torino, Torino, Italy
VINCENT M. ROTELLO • University of Massachusetts, Amherst, MA, USA
MARISSA M. SAMPIAS • Princeton University, Princeton, NJ, USA
Contributors xi
JOHN SCHLAGER • Air Force Research Laboratory, Wright-Patterson Air Force Base,
Dayton, OH, USA
JOSEPH B. SCHLENOFF • Department of Chemistry and Biochemistry, Florida State University,
Tallahassee, FL, USA
PRESTON T. SNEE • University of Illinois at Chicago, Chicago, IL, USA
LAURA SOLA • Consiglio Nazionale delle Ricerche, Istituto di Chimica del Riconoscimento
Molecolare, Milan, Italy
ERIK J. SORENSEN • Princeton University, Princeton, NJ, USA
MICHAEL H. STEWART • Division of Optical Sciences, U.S. Naval Research Laboratory,
Washington, DC, USA
LI SUN • Princeton University, Princeton, NJ, USA
KIMIHIRO SUSUMU • Naval Research Laboratory, Washington, DC, USA
CHRISTINA M. TYRAKOWSKI • University of Illinois at Chicago, Chicago, IL, USA
ERWIN A. VOGLER • Department of Materials Science and Engineering, The Pennsylvania State
University, University Park, PA, USA; Department of Bioengineering, The Pennsylvania
State University, University Park, PA, USA
JIN XIE • Department of Chemistry, University of Georgia, Athens, GA, USA
HAW YANG • Princeton University, Princeton, NJ, USA
EMMA V. YATES • Princeton University, Princeton, NJ, USA
Part I
Making Bio–Nano Interfaces

Chapter 1
Preparation of 2 nm Gold Nanoparticles for In Vitro

and In Vivo Applications
Daniel F. Moyano, Bradley Duncan, and Vincent M. Rotello
Abstract
Gold nanoparticles have been a versatile tool in recent years for the exploration of biological systems.
However, challenges with purification and adequate surface coverage limit the biocompatibility of gold
nanoparticles. Here, we describe a detailed procedure for the synthesis, purification, and functionalization
of biologically compatible gold nanoparticles for in vitro and in vivo studies.
Key words Gold nanoparticle, Biocompatibility, Purification
1 Introduction
Gold nanoparticle (AuNP) properties and applications continue to

be the focus of a variety of fields since their discovery in 1857 by
Michael Faraday [1]. Among these areas, avid research has been
pointed towards the creation of biocompatible nanoparticles [2].
The precise control introduced by the two-phase Brust–Schiffrin
method [3, 4] allowed more systematic studies of nanoparticle
properties to be performed. However, the use of tetra-n-
octylammonium bromide as the phase transfer catalyst limits bio-
logical applications due to the difficultly in separating this highly
cytotoxic chemical from the nanoparticles [5, 6]. Additionally, the
functionality of nanoparticle surfaces has been a prominent con-
cern due to the direct correlation of cytotoxicity with the function-
ality of the surface monolayer [7]. To further improve nanoparticle
utility, a fine balance between hydrophobic functionalization near
the metallic core to confer monolayer stability [8] and non-fouling
character provided by poly(ethylene glycol) moieties must be
achieved [9]. An excellent ligand base for biological applications is
23-mercapto-3,6,9,12-tetraoxatricosan-1-ol (Fig. 1) as its compo-
nents confer the desired stability and non-fouling properties [10].
This structure can be easily modified to provide selective binding
Paolo Bergese and Kimberly Hamad-Schifferli (eds.), Nanomaterial Interfaces in Biology: Methods and Protocols,
Methods in Molecular Biology, vol. 1025, DOI 10.1007/978-1-62703-462-3_1, © Springer Science+Business Media New York 2013
3
4 Daniel F. Moyano et al.
Fig. 1 Gold nanoparticle chemical structure. The ligand is composed of three parts, namely, the sulfur atom
(binding with gold), an alkyl chain (nanoparticle stability), and a tetraethylene glycol spacer (biocompatibility
and solubility)
making these nanoparticles a promising platform for systematic

biological studies [11]. Employing these concepts, this protocol
elaborates on the synthesis, purification, and functionalization of
gold nanoparticles with the biocompatibility required for in vivo
and in vitro studies.
2 Materials
1. Hydrogen tetrachloroaurate (III) hydrate.

2. Tetra-n-octylammonium bromide (TOAB).
3. 1-Pentanethiol.
4. Sodium borohydride.
5. 23-mercapto-3,6,9,12-tetraoxatricosan-1-ol, synthetized accord-
ing to the reported procedure [12].
6. 200 proof ethanol, acetone, hexanes, and acetonitrile.
7. 10,000 MWCO snakeskin pleated dialysis tubing membrane.
8. 3.0 cm Grade 1 filter paper circles.
3 Methods
All procedures are done at room temperature unless otherwise

specified.
3.1 Synthesis of Gold 1. Pour 1.0 g (2.5 mmol, 1 Eq) of hydrogen tetrachloroaurate
Nanoparticles (from (III) hydrate and 150 mL of Type I ultrapure water into a
the Brust–Schiffrin 1,000 mL round bottom flask. Stir slowly for 5 min using a
Method) [3] 2 in. egg-shaped magnetic stir bar until complete dissolution.
Solution should be clear yellow (see Note 1).
Preparation of Biocompatible Gold Nanoparticles 5
2. Premix 2.1 g (3.8 mmol, 1.5 Eq) of tetra-n-octylammonium

bromide in 150 mL of toluene and add it to the gold solution
in the round bottom flask. Stir at maximum speed for 15 min
until the solution presents is a cloudy dark orange color
(see Note 2).
3. While continuing to rapidly stir, cover the round bottom flask
with a septum stopper and using a syringe add dropwise 0.7 mL
(5 mmol, 2 Eq) of 1-pentanethiol over the course of 15 min.
Stir until the solution presents a cloudy deep white color
(see Note 3).
4. Remove the stopper and quickly add a freshly prepared solu-
tion of 2.0 g (50 mmol, 20 Eq) of sodium borohydride in
10 mL of Type I ultrapure water. Stir the solution for 5 h
(see Note 4).
5. Using a separatory funnel, remove the water phase (colorless)
from the toluene solution (black). Evaporate the majority of
the solvent under reduced pressure at 40 °C (see Note 5).
6. Add 900 mL of 200 proof ethanol and mix until the solid is
completely dispersed. Store the flask at −20 °C.
3.2 Purification 1. Two days after the synthesis, the nanoparticles precipitate and
of Gold Nanoparticles a black solid is observed at the bottom of the flask. Carefully
remove the ethanol solution (brown color) without dispersing
the precipitate (see Note 6).
2. Add fresh 200 proof ethanol and mix the solution until the
nanoparticles are dispersed. Leave the solution at −20 °C until
precipitation is complete (see Note 7).
3. Repeat steps 1 and 2 until the ethanol solution is colorless
after precipitation (approximately five times) (see Note 8).
4. Remove the solvent and redisperse the precipitate in 10 mL of
200 proof ethanol. Sonicate the solution for 10 min.
5. Using a 3.0 cm grade 1 filter paper circle over a 25 mm filter
holder (No. 5 stopper) attached to a vacuum line, wash the
nanoparticles several times with 200 proof ethanol redispers-
ing the precipitate in each wash (see Note 8).
6. Recover the solid from the filter paper.
7. Dissolve the nanoparticles in a minimal amount of toluene
and analyze the pentanethiol-capped AuNPs using mass
spectrometry (LDI) paying particular attention for the pres-
ence of a peak at 466 m/z (TOAB). If the peak is present
repeat steps 5 and 6 until it is no longer observed (see Notes
9 and 10) [13].
8. Use transmission electron microscopy (TEM) to obtain the
average particle diameter and use nuclear magnetic resonance
(solvent—CDCl3) to confirm purity.
3.3 Functionalization 1. Dissolve 30 mg of the purified pentanethiol-capped AuNPs in

of Gold Nanoparticles 10 mL of dry dichloromethane and bubble the solution with
(from the Murray nitrogen for 5 min.
Place Exchange) [14] 2. Separately dissolve 150 mg of 23-mercapto-3,6,9,12-
tetraoxatricosan-1-ol (or the ligand of interest) in 10 mL of
dichloromethane and bubble with nitrogen for 5 min.
3. Mix the two solutions in a 50 mL round bottom flask, cap with
a plastic stopper, and stir (0.5″ octagonal magnetic stir bar) for
2 days.
4. Remove the solvent under reduced pressure and redissolve the
now functionalized nanoparticles in 10 mL Type I ultrapure
water, obtaining a brownish colored solution.
5. Purify the solution in 10,000 MWCO snakeskin pleated dialysis
tubing membrane by dialyzing the system for 5 days using deion-
ized water in a 5,000 mL polypropylene beaker (see Note 11).
6. Recover the AuNP solution from the dialysis bag and lyophi-
lize the nanoparticles. A black solid is obtained.
7. Dissolve a small amount of the solid in D2O (for NMR) and
the rest in 2 mL Type I ultrapure water.
8. Calculate the concentration according to the reported extinc-
tion coefficient methodology using the size value obtained
from TEM [15].
9. Store at 4 °C.
4 Notes
1. Weigh the gold salt using a plastic spatula, not a metallic one
that can potentially reduce the gold salt. Make sure that all the
gold salt is dissolved.
2. Stirring is one of the key components of the process. Maintain
fast stirring for the rest of the synthesis, using the recom-
mended stir bar.
3. Do not add more 1-pentanethiol to achieve the white color.
Wait a prudent amount of time (20 min) and the color should
appear, otherwise some of the components are impure and the
process should be restarted with new materials.
4. A black color should appear as soon as the sodium borohydride
is added. If it takes more than 3 s to appear, the process should
be started again and the solution discarded as the slow appear-
ance of the black color will indicate polydisperse product.
5. No more than 40 °C should be used, otherwise nanoparticles
will grow larger and polydispersity will be observed.
Preparation of Biocompatible Gold Nanoparticles 7
6. If there is no observable precipitation (solution too dark), check

with a glass pipet to observe if the solution is clear or still cloudy.
In case of a cloudy solution, let sit at −20 °C for another day.
7. No sonication should be used in the first cleanings.
8. Acetone can be used for faster precipitation/filtration.
9. If TOAB is still present after several cleanings, pour the solid
nanoparticles into a 1,000 mL round bottom flask, add 5 mL
of toluene to dissolve (brownish color appears), and add
900 mL of ethanol [5]. Repeat the process from step 1 (this
procedure can be repeated until TOAB is removed).
10. If TOAB is still present, an alternative procedure for cleaning
is the use of acetonitrile. Dissolve the AuNPs in hexanes and
extract multiple times with acetonitrile (AuNPs remain in
upper hexanes phase and the TOAB transfers to the lower ace-
tonitrile layer). Collect the hexanes phase and evaporate the
solvent at reduce pressure (<40 °C).
11. During the first day water should be changed every 2 h, the
second and third day every 6 h, and during the last 3 days
every 12 h.
Acknowledgment
This work was supported by the NIH (GM077173 and EB014277).
References
1. Faraday M (1857) The Bakerian lecture: integrity using a proteolytic biomarker. Toxicol
experimental relations of gold (and other met- In Vitro 22:1099–1106
als) to light. Philos Trans R Soc Lond 6. Waters CA, Mills AJ, Johnson KA, Schiffrin DJ
147:145–181 (2003) Purification of dodecanethiol deriva-
2. Saha K, Bajaj A, Duncan B, Rotello VM tised gold nanoparticles. Chem Commun
(2011) Beauty is skin deep: a surface mono- 540–541
layer perspective on nanoparticle interactions 7. Murphy CJ, Gole AM, Stone JW, Sisco PN,
with cells and biomacromolecules. Small Alkilany AM, Goldsmith EC, Baxter SC
14:1903–1918 (2008) Gold nanoparticles in biology: beyond
3. Brust M, Walker M, Bethell D, Schiffrin DJ, toxicity to cellular imaging. Acc Chem Res
Whyman R (1994) Synthesis of thiol- 41:1721–1730
derivatised gold nanoparticles in a two-phase 8. Zhu Z-J, Tang R, Yeh Y-C, Miranda OR,
liquid-liquid system. J Chem Soc Chem Rotello VM, Vachet RW (2012) Determination
Commun 7:801–802 of intracellular stability of gold nanoparticle
4. Li Y, Zaluzhna O, Xu B, Gao Y, Modest JM, monolayers using mass spectrometry. Anal
Tong YJ (2011) Mechanistic insights into the Chem 84:4321–4326
Brust-Schffrin two-phase synthesis of organo- 9. Hoffman AS (2008) The origins and evolution
chalcogenate-protected metal nanoparticles. of “controlled” drug delivery systems.
J Am Chem Soc 133:2092–2095 J Control Release 132:153–163
5. Cho M-H, Niles A, Huang R, Inglese J, Austin 10. Kanaras AG, Kamounah FS, Schaumburg K,
CP, Riss T, Xia M (2008) A bioluminescent Kiely CJ, Brust M (2002) Thioalkylated tetra-
cytotoxicity assay for assessment of membrane ethylene glycol: a new ligand for water soluble
monolayer protected gold clusters. Chem tion mass spectrometry analysis of monolayer-
Commun 2294–2295 protected gold nanoparticles. Anal Bioanal
11. Moyano DF, Rotello VM (2011) Nano meets Chem 396:1025–1035
biology: structure and function at the nanopar- 14. Hostetler MJ, Green SJ, Stokes JJ, Murray RW
ticle interface. Langmuir 27:10376–10385 (1996) Monolayers in three dimensions: syn-
12. Miranda OR, Chen H-T, You C-C, Mortenson thesis and electrochemistry of ω-functionalized
DE, Yang X-C, Bunz UHF, Rotello VM alkanethiolate-stabilized gold cluster com-
(2010) Enzyme-amplified array sensing of pounds. J Am Chem Soc 118:4212–4213
proteins in solution and in biofluids. J Am 15. Liu X, Atwater M, Wang J, Huo Q (2007)
Chem Soc 132:5285–5289 Extinction coefficient of gold nanoparticles
13. Yan B, Zhu Z-J, Miranda OR, Chompoosor A, with different sizes and different capping
Rotello VM (2010) Laser desorption/ioniza- ligands. Colloid Surf B 58:3–7
Chapter 2
DNA Conjugation to Nanoparticles

Sunho Park
Abstract
Nanoparticle–DNA (NP–DNA) conjugates have been highlighted due to their versatility in diverse science
and engineering fields. The protocol of DNA conjugation to gold nanoparticles (AuNPs), which are
among the most popular NPs in bio-applications, is thus described here. This protocol also includes ligand
exchange of AuNP to make AuNPs suitable for conjugation process and a fluorescence technique to evalu-
ate the average number of DNA strands attached to single AuNP.
Key words Nanoparticle, Gold nanoparticle, DNA, Conjugation, Ligand exchange, Gel electropho-
resis, Lyophilization, DNA staining, Fluorescence, DNA coverage ratio
1 Introduction
DNA–nanoparticle conjugates have attracted great interest as there

are numerous opportunities to utilize the properties of the nano-
material and the DNA synergistically on small length scales [1–5].
In particular, gold nanoparticles (AuNPs) are widely used in
biomedical applications such as delivery, sensing, and imaging due
to their exceptional biocompatibility [6–9]. AuNPs have a rela-
tively inert surface but can be easily conjugated to biomolecules by
covalent linkage via the sulfur atom in a thiol. Therefore, thiol-
capped DNA oligos have been popular for constructing AuNP–
DNA conjugates.
Thiol-aided conjugation is considered feasible for the labora-
tory; however, some important issues are often ignored in practice.
The first is the stability of the nanoparticles during the conjugation
process. Conjugation via a thiol group is typically performed in con-
centrated ionic buffers which screen the surface charge of the parti-
cles and DNA, so that electrostatic repulsion is diminished, and
therefore increasing the collision rate between the two species.
Using concentrated samples also raises this collision rate.
Lyophilization is a method that can elevate both the ionic strength
of solution and the concentration of each species extremely high as
9
10 Sunho Park
the water content of the solution dries out. Citrate-stabilized AuNPs

[10] are commonly used in laboratories but they irreversibly aggre-
gate during lyophilization due to their low stability. A chemical
ligand bis(p-sulfonatophenyl) phenylphosphine dihydrate dipotas-
sium salt is thus used to replace the citrate molecules on the AuNP
surface, and allows AuNPs to sustain stability under more severe
ionic conditions.
Another issue is the evaluation of the exact coverage ratio of
AuNP–DNA conjugates. The average number of conjugated
DNA strands per each particle is lower than the number from
DNA:AuNP incubation ratio in the reaction due to <100 % effi-
cient conjugation. To quantify the coverage, the sample must be
cleaned up by removing the free DNA strands from the solution,
and then the conjugated DNA strands are completely liberated
from the AuNPs to count their population. In this method,
6-mercapto-1-hexanol is used for the DNA displacement [11].
Fluorescence measurement of the stained DNA strands is
performed afterwards to quantify the DNA concentrations.
AuNPs should be excluded from the solution before the measure-
ment to avoid fluorescence quenching [12].
This chapter includes the protocols of ligand exchange of
AuNP, AuNP–DNA conjugation, and evaluation of DNA coverage
in detail (Fig. 1). Purification of samples relies only on typical bio-
chemical lab techniques such as centrifugation and agarose gel
electrophoresis. The protocols provide most biochemists and
nanoparticle researchers with an easier and more accurate AuNP–
DNA conjugate preparation that does not require extensive effort
or specialized skills.
2 Materials
All solutions are prepared with 0.5× tris-borate-EDTA buffer

(0.5× TBE, pH = 8.3) unless specified otherwise.
2.1 Ligand Exchange 1. 0.5× Tris-borate-EDTA buffer (TBE): 45 mM tris, 45 mM

boric acid, and 1 mM EDTA in ultrapure deionized water
(resistivity of 18 MΩ cm at room temperature) (see Note 1).
2. Gold nanoparticle (AuNP): Synthesized or commercially
achieved citrate-stabilized aqueous solution (see Note 2).
3. Bis(p-sulfonatophenyl) phenylphosphine dihydrate dipotas-
sium salt (BPS) (see Note 3).
4. Sodium Chloride (NaCl).
5. 1 % Agarose gel: 0.5 g of agarose powder in 50 mL 0.5× TBE.
Cast in an appropriate agarose gel electrophoresis device.
Use 0.5× TBE as gel running buffer (see Note 4).
DNA Conjugation to Nanoparticles 11
Fig. 1 An overview of the protocol of nanoparticle–DNA (NP–DNA) conjugation

and coverage ratio quantification
2.2 Conjugation and 1. DNA oligonucleotide: commercially available thiol-capped

Evaluation of Coverage DNA strands (this protocol works best for 10–50 mer DNA
Ratio oligos). Adjust the concentration to 1 μg/μL (see Note 5).
2. 100 mM Tris(2-carboxyethyl)phosphine hydrochloride
(TCEP): 0.0287 g of TCEP in 1 mL 0.5× TBE (see Note 6).
3. 10 mM 6-Mercapto-1-hexanol (MCH): 1 μL of pure MCH
(~7.3 M) into 729 μL of 0.5× TBE (see Note 7).
4. DNA staining solution: 100× SYBR® Gold (Invitrogen). Dilute
1 μL of stock solution (10,000×) with 99 μL of 0.5× TBE
(see Note 8).
3 Methods
3.1 Ligand Exchange 1. Put a pinch amount (~0.1 g) of BPS into ~100 mL of aqueous
and Purification citrate stabilized AuNP solution, which is typically at ~10−9 M
of AuNPs concentration and pale red. Place the solution on an orbital
shaker and gently mix it for an overnight (see Note 9).
12 Sunho Park
2. Put 1 g of NaCl into the AuNP–BPS solution. Stir and mix

well with a disposable pipette. The color of the solution will
turn dark gray in a few seconds. Separate the solution evenly
into two 50 mL centrifuge tubes or eight 15 mL tubes. Place
the tubes in a centrifuge and spin them for ~5 min at rcf
~1,000 × g (see Note 10).
3. Remove the clear supernatant from all of the tubes with a
pipette and add ~100 μL of 0.5× TBE with 200 μL pipette and
tip right onto the sediment at the bottom of one of the tubes.
Re-disperse the particles well by pressing and releasing the
pipette, take out the dark red solution, and put it right into the
sediment in another one of the tubes. Repeat this step until all
of the AuNP aggregates are recovered (see Note 11).
4. Place the thick red AuNP solution into a long single well in 1 %
agarose gel. Apply an electric field of ~4 V/cm to the gel. The
AuNP band will move from the anode (−) to cathode (+). Stop
running the gel after the band shifts by ~3–4 cm (taking ~1 h)
(see Note 12). Cut out the narrow AuNP band using a clean
razor blade, chop it into many small pieces, and put the sliced
pieces into a 15 mL tube. Pour several milliliters of 0.5× TBE
into the tube until the agarose gel pieces are fully soaked.
Tighten the cap and place the tubes in a 4 °C refrigerator for
~2 days (see Note 13).
5. Take out the red solution and discard the solid pieces left in the
tube. Distribute the solution into 1.5 mL tubes (300 μL each)
and place the tubes in a microcentrifuge. Spin the tubes for
~10 min at rcf ~10,000 × g. Gently remove the supernatant
(pale pink) and collect the red oily layer at the bottom of the
tubes (see Note 14).
6. Use a UV–VIS spectrometer to collect an absorbance spec-
trum. Confirm the peak absorbance is at ~520 nm and record
the peak absorbance value. Determine the concentration of the
stock solution using Beer–Lambert law (see Note 15).
3.2 Conjugation 1. Determine a target coverage ratio (average number of conju-

of DNA Oligos gated DNA strands per AuNP) and the amount of AuNP stock
with AuNPs solution to be used. Calculate the necessary amount of DNA
solution (see Note 16). Mix the same volume of 100 mM
TCEP solution with the DNA solution. Leave the mixture on
the bench for ~1 h (see Note 17).
2. Mix the TCEP + DNA solution well with AuNP solution.
Separate evenly the mixture into 1.5 mL tubes (~100 μL per
tube) and lyophilize the samples until they are dried out
(see Note 18).
3. Re-disperse the dried sample in one of the tubes by adding ~50 μL

0.5× TBE into it, and use this solution to recover all the other
samples. Place the concentrated solution in a 4 °C refrigerator
for ~1 day to allow further conjugation (see Note 19).
4. Add ~1 mL of 0.5× TBE to the sample and separate the diluted
solution evenly into 1.5 mL tubes (~200 μL per each tube).
Spin the tubes with a microcentrifuge (rcf ~10,000 × g) for
10 min to collect AuNP–DNA conjugates at the bottom and
leave unconjugated DNA strands in supernatant. Discard the
supernatant and collect the red layer at the bottom of the tubes.
Repeat this whole washing step at least three times. The red
solution from the final washing is the stock solution of AuNP–
DNA conjugates.
5. Take UV–VIS absorbance spectrum to achieve the concentra-
tion of the stock AuNP–DNA conjugate solution (see Note 20).
3.3 Determination 1. Dilute the AuNP–DNA stock solution to 0.1 μM in 1 mM MCH

of Coverage Ratio and leave the sample undisturbed for ~1 day (see Note 21).
2. Spin the AuNP–DNA + MCH solution with a microcentrifuge
(rcf ~10,000 × g, 10 min). Carefully transfer the clear superna-
tant to a new 1.5 mL tube and discard the original tube con-
taining dark gray residue at the bottom (see Note 22). Put
10 μL of this solution into 89 μL of 0.5× TBE, and add 1 μL
100× SYBR® Gold to the mixture. Make three or four identical
samples for better accuracy of coverage measurement. Do the
following step (staining of reference DNA) immediately (see
Note 23).
3. Dilute the stock DNA solution to 1 μM (see Note 24). Make
reference DNA solutions of 1 × 10−8, 5 × 10−8, 1 × 10−7, 2 × 10−7,
and 4 × 10−7 M, in 0.1 mM MCH and 1× SYBR® Gold (see
Note 25). These reference samples and the samples from the
previous step should be kept in the dark for ~1 h.
4. Take fluorescence spectrum of the samples from step 2 and
reference solutions from step 3. The excitation wavelength is
at 495 nm and the emission peak is around 537 nm for SYBR®
Gold (Fig. 2a). Collect the peak emission intensity of each
solution. Fit the data from DNA references into a linear func-
tion, I = a·C + b, where I is the intensity of fluorescence [cps;
counts per second], C is the concentration of DNA solution
[M], and a, b are constants (Fig. 2b). Calculate the DNA con-
centration of the samples from step 2 by use of the measured
intensity and the fitted equation. Divide this value by
1 × 10−8 M (see Note 26) to get the coverage ratio of the stock
AuNP–DNA.
14 Sunho Park
Fig. 2 Intensity of fluorescence emission from 1× SYBR® Gold-stained DNA oligo (HS-5′-TTTTT TTTTT TTTTT
TTTTT TTTTT-3′) in 0.5× TBE and 0.1 mM MCH was measured. Excitation wavelength was 495 nm. (a) An
example of emission spectrum, (b) A linear relationship between peak emission intensity and concentration of
stained DNA
4 Notes
1. 0.5× TBE is made by either diluting commercially available

concentrated TBE buffer or dissolving powdered TBE in
water.
2. This protocol works for AuNP of diameter in 5–20 nm, while
it is most suitable for ~10 nm AuNPs. Aqueous AuNP colloids
are synthesized mostly by reducing Au3+ ions in sodium citrate
solution; however, citrate-capped AuNPs are not stable enough
to endure the conjugation process suggested here. Most fail-
ures, including irreversible aggregation of AuNPs, happen dur-
ing the centrifugation or lyophilization steps.
3. BPS molecules form dipolar bonds between their own oxygen
atoms and gold atoms on AuNP’s surface. They are negatively
charged and generate sufficiently high zeta-potential in ionized
buffer.
4. Agarose gel electrophoresis is used to exclude unbound BPS
molecules and residues from AuNP synthesis. Add 0.5 g aga-
rose powder to 50 mL 0.5× TBE (or 1 g agarose to 100 mL)
in 250 mL glass beaker and mix well using a disposable pipette.
Pull up and push out the mixture with the pipette several
times to make the agarose solution homogenized. A typical
gel comb has multiple teeth to make separate wells in a gel.
For this purification purpose, put some electrical tape on the
teeth to cover the empty spaces between the teeth so that the
comb will form a long, single well in the gel. After the gel cast-
ing device is assembled, heat up the agarose solution in the
beaker with a microwave oven. Interrupt the heating process

a few times to take out and shake the beaker to re-disperse any
possible sediment at the bottom. The solution is finished heat-
ing when it boils and turns clear. Take the beaker carefully out
of the microwave oven, pour the hot solution quickly into the
gel tray, and place the gel comb at a right position. Remove
small bubbles floating on the surface level of the hot gel solu-
tion before the gel starts hardening. A 10 μL or 20 μL pipette
and tip works well for pulling up the bubbles. It takes 30 min
to 1 h for full hardening of the gel. Remove the gel comb and
pour into the tank 0.5× TBE as gel running buffer. The gel
must be fully immersed.
5. Thiol-capped DNA strands typically form dimers by making
disulfide bonds, or are capped with an additional protecting
group by the manufacturer. Thus, they need to be reduced
before conjugation with AuNPs.
6. It is preferred to use fresh solution. Dithiothreitol (DTT) is
often used for the same purpose; however, TCEP is preferable
over DTT in that it is an irreversible and more powerful reduc-
ing agent.
7. It is preferred to use fresh solution. Handle the stock MCH
bottle in a fume hood as it has a strong odor.
8. It is preferred to use fresh solution. This product is a DNA
staining dye and stored at −20 °C. Thaw the stock in a dark
place at room temperature. Put the stock bottle back into the
freezer right after taking out the necessary amount.
9. The concentration of 0.1 g of BPS [Mw = 534.6] in 100 mL
solution is ~1.9 mM, which is much higher than a typical con-
centration of aqueous AuNP solution. An overnight reaction is
sufficient to ensure the BPS content on AuNP surface has
reached an equilibrium state.
10. Excessive amounts of NaCl can make the ionic strength of the
solution very high. This fully screens the surface charge of the
AuNPs in solution and can result in precipitation of the AuNPs.
Tune the centrifugal force and centrifugation time depending
on the size and surface charge density of AuNP as the precipi-
tated AuNPs may aggregate permanently under too strong
centrifugation conditions.
11. Limit the final volume of the re-dispersed AuNP solution
within a few hundred microliters so that the volume does not
exceed the capacity of the gel loading well.
12. These gel running times and electric field strengths are based
on lab experience. You may need to run the gel for longer if the
particles are larger, ionic strength of running buffer is greater,
or the electric field strength is low.
16 Sunho Park
13. AuNPs in gel band will diffuse into 0.5× TBE buffer and the
concentration distribution of AuNPs will reach an equilibrium
state inside and outside gel. Slicing of the gel band facilitates
this because it increases the surface area for diffusion flux.
More AuNPs are recovered by transferring the remaining solid
gel pieces into another tube filled with clean 0.5× TBE. It takes
less time for diffusion for smaller AuNPs.
14. Spin the tubes for an extended time if not much of the AuNPs
collect at the bottom. This is the stock solution of AuNP to be
used for DNA conjugation, and the concentration is typically
in the order of ~1 μM. You may adjust the concentration
exactly to 1 μM using 0.5× TBE for convenience of later use.
15. A = εcl, where A [OD, optical density] is the peak value of the
absorbance spectrum, ε [M−1 cm−1] is the extinction coefficient,
l [cm] is the path length, and c [M] is the concentration of the
solution. ε of AuNPs depends on their size and ligand types,
and values can be found in literature (~108 M−1 cm−1 for
~10 nm AuNP) [13]. Dilute the stock AuNP solution to make
the measured absorbance peak lie within a typical working
range of most UV–VIS spectrometers, 0.1–10 O.D.
16. If 50 μL of 1 × 10−6 M AuNP stock solution is used to make
1:10 AuNP:DNA conjugates, for example, the exact amount
of DNA strands necessary is 10 × (50 μL) × (1 × 10−6 M) = 500 ×
10−6 μmol. When Mw [μg/μmol] is the molecular weight of
DNA oligo, this amount is equivalent to (500 × 10−6 μmol) × Mw
[μL] of 1 μg/μL stock DNA solution. Due to the nature of
conjugation process, however, use two or three times as much
of the exact amount. Not all the DNA used will be
conjugated.
17. Keep the mixture in a dark place if the DNA strands have fluo-
rescent markers.
18. An easy way of lyophilization is to put the samples in 1.5 mL
tubes and place them in a vacuum chamber with mild centrifu-
gation. Use a pushpin to make an air hole on the lid of the
tubes. Centrifugation helps the samples stay at the bottom of
the tube during the drying process. Spinning the samples too
strongly sometimes results in irreversible aggregation. Higher
success rates are usually associated with running weak centrifu-
gation for the first ~10 min only while under vacuum.
19. This refrigeration step for further conjugation may be skipped
if a higher amount of extra DNA (see Note 16) is used during
the TCEP treatment step.
20. The concentration of AuNP–DNA conjugates is equivalent to
the concentration of core AuNPs. DNA absorbance spectra
have a peak at 260 nm and vanish in most visible light wave-
length range; therefore, the AuNP absorbance peak value at
520 nm is rarely affected by conjugated DNA.
21. The sulfur atom of MCH forms a covalent bond with a gold
atom on the AuNP surface, the same way of DNA conjugates
to the AuNP. Due to the excessive population of MCH mole-
cules in solution, the surface of AuNP becomes densely cov-
ered by MCH, replacing the DNA strands. The AuNP
eventually becomes neutral as the negatively charged DNA and
BPS are displaced in this process. If the concentration of the
stock AuNP–DNA solution is 1 μM, for example, add 10 μL
AuNP–DNA and 10 μL 10 mM MCH to 80 μL 0.5× TBE in
a 1.5 mL tube (100 μL in total).
22. AuNPs have been aggregated and discarded by centrifugation.
Remember that the DNA strands in the supernatant are in
1 mM MCH and have been detached from 0.1 μM
AuNP–DNA.
23. The sample is now in 0.1 mM MCH, 1× SYBR® Gold. DNA
strands in the solution were released from 1 × 10−8 M AuNP–
DNA. SYBR® Gold, a staining agent of DNA, is added after all
other substances have been mixed. DNA staining is fully satu-
rated in about half an hour; however, it starts degrading after a
few hours. Thus, perform the fluorescence measurement
quickly in the following steps.
24. 1 μg/μL DNA stock is at molar concentration of 1/Mw. Dilute
properly with 0.5× TBE.
25. Mix together 1 μL of 1 μM DNA, 1 μL of 10 mM MCH, and
97 μL of 0.5× TBE, then put 1 μL of 100× SYBR® Gold. This
makes 1 × 10−8 M DNA in 0.1 mM MCH, 1× SYBR® Gold,
and 0.5× TBE. Change the amount of 1 μM DNA appropri-
ately for each target DNA concentration, and reduce the
amount of 0.5× TBE by the increased amount of 1 μM DNA
to equalize the volumes of DNA reference solutions at 100 μL.
26. Note that the DNA strands have been detached from 1 × 10−8 M
AuNP–DNA.
References
1. Alivisatos P (2004) The use of nanocrystals in 5. Choi HS, Liu W, Liu F, Nasr K, Misra P,
biological detection. Nat Biotechnol 22: Bawendi MG, Frangioni JV (2010) Design
47–52 considerations for tumor-targeted nanoparti-
2. De M, Ghosh PS, Rotello VM (2008) cles. Nat Nanotechnol 5:42–47
Applications of nanoparticles in biology. Adv 6. Chen C-C, Lin Y-P, Wang C-W, Tzeng H-C,
Mater 20:4225–4241 Wu C-H, Chen Y-C, Chen C-P, Chen L-C,
3. Medintz IL, Uyeda HT, Goldman ER, Wu Y-C (2006) DNA-gold nanorod conju-
Mattoussi H (2005) Quantum dot bioconjugates for remote control of localized gene
gates for imaging, labelling and sensing. Nat expression by near infrared irradiation. J Am
Mater 4:435–446 Chem Soc 128:3709–3715
4. Nel AE, Madler L, Velegol D, Xia T, Hoek 7. Ghosh PS, Kim C-K, Han G, Forbes NS,
EMV, Somasundaran P, Klaessig F, Castranova Rotello VM (2008) Efficient gene delivery
V, Thompson M (2009) Understanding bio- vectors by tuning the surface charge density of
physicochemical interactions at the nano-bio amino acid-functionalized gold nanoparticles.
interface. Nat Mater 8:543–557 ACS Nano 2:2213–2218
18 Sunho Park
8. Thomas M, Klibanov AM (2003) Conjugation 11. Park S, Brown KA, Hamad-Schifferli K (2004)
to gold nanoparticles enhances polyethyleni- Changes in oligonucleotide conformation on
mine’s transfer of plasmid DNA into mamma- nanoparticle surfaces by modification with
lian cells. Proc Natl Acad Sci U S A 100: mercaptohexanol. Nano Lett 4:1925–1929
9138–9143 12. Dulkeith E, Ringler M, Klar TA, Feldmann J
9. Macfarlane RJ, Lee B, Jones MR, Harris N, (2005) Gold nanoparticles quench fluores-
Schatz GC, Mirkin CA (2011) Nanoparticle cence by phase induced radiative rate suppres-
superlattice engineering with DNA. Science sion. Nano Lett 5:585–589
334:204–208 13. Liu X, Atwater M, Wang J, Qun H (2007)
10. Frens G (1973) Controlled nucleation for the Extinction coefficient of gold nanoparticles
regulation of the particle size in monodisperse with different sizes and different capping
gold suspensions. Nat Phys Sci 241:20–22 ligands. Colloids Surf B 58:3–7
Chapter 3
Conjugation of Nanoparticles to Proteins

Marie-Eve Aubin-Tam
Abstract
Nanoparticle–protein conjugates hold great promise in biomedical applications. Diverse strategies have
been developed to link nanoparticles to proteins. This chapter describes a method to assemble and purify
nanoparticle–protein conjugates. First, stable and biocompatible 1.5 nm gold nanoparticles are synthe-
sized. Conjugation of the nanoparticle to the protein is then achieved via two different approaches that do
not require heavy chemical modifications or cloning: cysteine–gold covalent bonding, or electrostatic
attachment of the nanoparticle to charged groups of the protein. Co-functionalization of the nanoparticle
with PEG thiols is recommended to help protein folding. Finally, structural characterization is performed
with circular dichroism, as this spectroscopy technique has proven to be effective at examining protein
secondary structure in nanoparticle–protein conjugates.
Key words Nanoparticles, Proteins, Bioconjugates, Agarose gel electrophoresis, Circular dichroism
1 Introduction
The conjugation of nanoparticles (NPs) to proteins is of increasing

interest in biomedicine. NP–protein conjugates hold great promise
in sensing/diagnostics [1–3], in targeted delivery [4], in the con-
trol of protein activity [5, 6] and in imaging [7]. To this end,
efforts have been made to synthesize biocompatible NPs that are
nontoxic and stable at physiological pH [8].
Several strategies have been developed to assemble NP–protein
conjugates. A commonly used approach is to attach the NP on the
protein via non-covalent interactions, mostly through electrostatic
adsorption. In the 1970s, negatively charged gold NPs were already
linked non-covalently to various proteins, such as antibodies,
horseradish peroxidase, and bovine serum albumin, to serve as
electron dense labels for electron microscopy (EM) of tissue sec-
tions [9, 10]. Although this approach is relatively straightforward
and does not require cloning or chemical cross-linkage, the NP can
dissociate from the protein, especially at pH values above the pro-
tein isoelectric point (pI) [10] or at high salt concentrations [11].
19
20 Marie-Eve Aubin-Tam
In contrast, the formation of a covalent bond with either the NP

core atoms or the NP ligand molecules creates a more stable NP–
protein complex and allows to control the site of attachment on the
protein more easily. Such site-specific labeling is especially advanta-
geous in imaging and sensing applications. In EM imaging applica-
tions, for instance, site-specific NP attachment on the protein
enables one to go beyond the traditional histology applications and
to localize specific structures within large proteins [12, 13].
Various purification schemes are reported for isolating NP–pro-
tein conjugates. Size exclusion chromatography [7, 14] and gel elec-
trophoresis [14–17] have been used to assess conjugation yields and
to purify NP–DNA and NP–protein conjugates of desired stoichi-
ometry. Agarose gel electrophoresis is particularly suitable because it
allows sample to migrate in the gel towards both electrodes. Both
positively charged and negatively charged species can be viewed
simultaneously. This is especially useful for conjugating proteins to
NPs of opposite charge. Furthermore, since no denaturant such as
sodium dodecyl sulfate (SDS) is used, the conjugate can be directly
extracted from the agarose gel with spin centrifugation columns.
Unfortunately, NP conjugation may result in protein unfold-
ing, as previously reported for a number of proteins [11, 17–20].
Therefore, prior to any application, it is advised to verify that the
protein maintains its native fold and activity when conjugated to
the NP. In particular, proteins are more prone to being denatured
or destabilized if covalently attached to the NP [17].
Co-functionalization of the NP with polyethylene glycol (PEG)
molecules can help to circumvent this issue by preventing
unfolding. As thiolated PEG molecules replace BPS ligands on the
surface of the gold NP, the detrimental electrostatic interactions
between the NP and its associated protein can be reduced [17].
This chapter describes the process of purifying and character-
izing proteins labeled with gold NPs. In the first section
(Subheading 3.1), a method of synthesizing soluble biocompatible
1.5 nm gold NPs is presented. Then, labeling (Subheading 3.2),
purification (Subheading 3.3), and structural characterization
(Subheading 3.4) of NP–protein conjugates is described—these
protocols are easily adapted to protein labeling with NPs of differ-
ent sizes, materials, or surface ligands.
2 Materials
All solutions should be prepared with deionized (DI) water.

A Millipore water purification system can be used to reach a resis-
tivity of 18 MΩ·cm at 25 °C.
2.1 Nanoparticle 1. 125 mL conical flask and Teflon stir bar.

Synthesis 2. Separatory funnel of 50 mL capacity.
Conjugation of Nanoparticles to Proteins 21
3. 250 mg hydrogen tetrachloroaurate (HAuCl4) and 400 mg of

tetraoctylammonium bromide dissolved in 12.5 mL DI water
and 16.25 mL toluene.
4. 580 mg triphenylphosphine.
5. 352.5 mg of sodium borohydride.
6. 2.14 g of bis(p-sulphonatophenyl) phenylphosphine dihydrate
(BPS) dissolved in 20 mL of DI water.
7. High-performance liquid chromatography (HPLC) system.
8. HPLC gel filtration column (e.g.: TSK-gel G4000PW,
TOSOH bioscience).
2.2 NP Conjugation 1. PBS 1× or 10 mM NaPi buffer (pH 7.4).

2. Dithiothreitol (DTT).
3. Hexa(ethylene glycol) thiol (Polypure).
4. Dialysis cassette (e.g., Slide-A-Lyzer 2K MWCO, Pierce).
2.3 Agarose Gel 1. 0.5× TBE buffer (Tris-Borate-EDTA): 44.5 mM Tris Base,
Purification 44.5 mM boric acid, 1 mM EDTA. Dissolve 10× from 5× TBE
buffer stock (see Note 1).
2. 3 g ultrapure low melting point agarose dissolved in 100 mL
0.5× TBE buffer.
3. 500 mL beaker or conical flask.
4. Horizontal gel electrophoresis system.
5. SimplyBlue SafeStain (Invitrogen).
6. Centrifugal filter with 0.2 μm pore size membrane (Nanosep
MF Centrifugal Devices with Bio-Inert Membrane, PALL).
2.4 Structural 1. UV–Vis spectrophotometer.

Characterization 2. Circular dichroism spectrometer.
Equipment
3 Methods
All procedures should be carried at room temperature unless speci-

fied otherwise.
3.1 Synthesis and The first eight steps are adapted from Weare et al. [21].
Purification of Water
1. Mix 12.50 mL DI water with 16.25 mL toluene in a 125 mL
Soluble 1.5 nm NPs
conical flask.
2. Weigh 250 mg of HAuCl4 and 400 mg of tetraoctylammo-
nium bromide. Add to the water–toluene mixture and mix
with magnetic stir bar until the yellow color is all transferred in
the toluene phase.
3. Weigh 580 mg of triphenylphosphine and add to the solution.

Stir vigorously for at least 10 min or until the toluene phase is
white and cloudy.
4. Weigh 352.5 mg of sodium borohydride. Dissolve in 2.5 mL
of DI water and immediately add to the HAuCl4 and triphe-
nylphosphine solution in one step as quickly as possible while
stirring.
5. Stir for 2 h.
6. To wash the NPs, separate the organic phase from the water
phase using a separatory funnel. Discard the water phase.
7. Add 25 mL of DI water to the solution of NPs in toluene.
Separate the organic phase from the water with the separatory
funnel. Discard the water phase.
8. Repeat step 7 two additional times.
9. Weigh 2.14 g of bis(p-sulphonatophenyl) phenylphosphine
dihydrate (BPS) and dissolve in 20 mL of DI water to make a
0.2 M solution.
10. Mix the solution of NPs in toluene with the BPS aqueous solu-
tion in a closed bottle. Stir gently with a magnetic bar for 48 h.
The NP will transfer to the water phase upon ligand exchange
(see Note 2). Using a separatory funnel, separate the organic
phase from the aqueous solution of BPS NPs. Dispose properly
of the organic phase. Store the NP solution at 4 °C.
11. (Optional) Purify the water soluble NPs by size exclusion
HPLC to obtain a narrower NP size distribution and to remove
excess BPS ligand. A size exclusion HPLC column such as the
TSK-gel G4000PW (TOSOH bioscience) can be used with DI
water flowing at a rate of 0.5 mL/min. Collect 500 μL frac-
tions. Keep the fractions that show broad UV–Vis absorbance
in the 200–800 nm range with no plasmon peak at ~520 nm
(see Fig. 1). Store at 4 °C. After HPLC purification, it is pos-
sible to measure the stoichiometry of ligands on the surface of
the NPs with analytical techniques like ICP-OES (see Note 3).
12. Determine the NP concentration from the absorbance at
420 nm using an extinction coefficient of 110 mM−1 cm−1.
3.2 NP Conjugation This protocol can be used either with 1.5 nm gold NPs covered
to Protein with negatively charged BPS ligands synthesized as previously
described, or with NPs of other sizes, materials, or surface ligands
(see Note 4 for protocols to synthesize larger NPs).
NPs can be attached via electrostatic interactions. For example,
a protein with a net positive charge (e.g., cytochrome c, ribonuclease
A) would readily attach to a negatively charged BPS NP or citrate
NP. Alternatively, a gold NP can also be attached to an exposed
cysteine on the surface of a protein via gold-thiol chemistry.
Fig. 1 Size exclusion HPLC purification of gold NPs. The picture above shows fractions collected consecutively.
The fractions that contain solutions of NPs of ~1.5 nm diameter (encircled by dashed line) do not show any
plasmon absorption peak at ~520 nm in the UV–Vis spectra. Larger NPs elute from the column earlier and are
pink/purple because of the plasmon absorption. Free BPS molecules elute later
1. (Optional) If the protein attaches to the NP via a surface

cysteine using gold-thiol chemistry, incubate the protein with
10 mM DTT for 15 min. Remove excess DTT by dialyzing
against conjugation buffer (see Note 5) in dialysis cassettes, or
by four successive concentration–dilution cycles with 5 kDa
centrifugal filters using a refrigerated microcentrifuge at 4 °C.
2. Mix NPs and protein solution (both in conjugation buffer) in
the desired NP–protein molar incubation ratio. If using 1.5 nm
gold NPs, NP concentration should be ~0.1 mM to permit gel
visualization.
3. (Optional) Simultaneous incubation of the gold NPs and pro-
teins with poly(ethylene glycol) thiol can prevent protein
unfolding (Fig. 2). If working with 1.5 nm BPS gold NPs, a
5×–25× molar excess of hexa(ethylene glycol) thiol (Polypure)
relative to NP concentration can be used.
4. Incubate overnight at 4 °C.
3.3 Agarose Gel 1. Prepare a 3 % agarose gel by mixing 3 g of ultrapure agarose

Electrophoresis with 100 mL of 0.5× TBE in a 500 mL beaker or conical flask.
Purification Microwave the solution 20 s at HIGH intensity. Swirl the solu-
tion gently. Continue microwaving in 20 s increments and
swirling the solution until it boils and the agarose is fully dis-
solved. Pour the gel into a horizontal gel electrophoresis sys-
tem using a pipettor to prevent bubbles to set into the gel.
Insert a comb in the middle of the gel with wells at least 1 cm
wide. When the gel is set, fill the gel box with 0.5× TBE and
remove comb.
Fig. 2 Schematics of BPS ligand (a) and PEG ligand (b) at the surface of a NP. (c) Agarose gel electrophoresis
of NP covered with BPS ligand (left lane) and co-functionalized with mPEG-thiol (right lane) in a 25× molar
excess relative to the NP concentration
Fig. 3 (a) NP conjugation is confirmed by agarose gel electrophoresis. The 1.5 nm gold NPs with BPS ligand
run as a single band towards the positive electrode (left lane). The NPs migrate more slowly in the agarose gel
when bound covalently to Saccharomyces cerevisiae cytochrome c via cysteine C102. (b) Blue staining for
protein confirms the presence of protein in the slower band. Cytochrome c is positively charged and would run
in the opposite direction towards the negative electrode. (c) UV–Vis absorption spectra of solutions of NPs
(grey line) and NP–protein conjugates (black line)
2. Load the NP, protein and NP–protein solutions into successive

wells. Run the gel with field strength of 10 V/cm until the
NPs have entered into the gel matrix and you get the desired
separation between the bands (Fig. 3).
3. To blue stain the gel for proteins, immerse the gel in SimplyBlue
SafeStain (Invitrogen). Rock the gel gently for 2 h. Discard
staining solution. Rinse with DI water. Immerse the gel in DI
water. Rock the gel gently overnight.
4. If the protein or the NP band position overlaps with the NP–
protein band, run the gel for longer times and/or repeat the
gel electrophoresis with different concentrations of agarose

(in the range of 1–4 %) until the bands run at distinct positions
as shown in Fig. 3.
5. When appropriate gel electrophoresis conditions have been
identified, run another agarose gel without blue staining.
6. Using a clean razor blade, cut the NP–protein band and place
gel into a centrifugal filter with 0.2 μm pore size membrane
(PALL). Cap the filter.
7. Spin at 14,000 × g at 4 °C for 30 min or until the NP–protein
solution has been extracted from the agarose gel.
3.4 Structural Circular dichroism (CD) spectroscopy in the far-UV is an effective

Characterization and widely used technique of probing protein structure in NP–
of Conjugate protein conjugates [11, 17–20, 22]. CD measurements are per-
formed in solution. Clean quartz CD cuvettes with 5 M nitric acid
and rinse them thoroughly with DI water.
Buffers that absorb strongly at low wavelengths (195–250 nm)
should be avoided if possible when performing CD in the far-UV.
NP ligands that strongly absorb in the UV can make CD measure-
ments noisier (see Note 6). If large NPs or aggregates of NPs are
present, correction for absorption flattening might be necessary
(see Note 7).
1. Measure UV–Vis absorption spectra of a 200 μL solution of
NPs and a 200 μL solution of NP–protein (there can be excess
NPs, see Note 8). Both should be in the same buffer. Using the
NP extinction coefficient, calculate the NP concentration. If
needed, add some buffer to have same NP concentration in
both solutions. If the protein absorbs in the visible (e.g., Heme
containing proteins), deconvolute the NPs and proteins con-
tribution to the UV–Vis absorption spectra to calculate NP
and protein concentration.
2. Measure CD spectra of the NPs and NP–protein solutions
(with same NP concentrations) in the 195–250 nm. Subtract
the CD spectrum of NP solution from the CD spectrum of
NP–protein solution.
3. Secondary structure algorithms such as CDSSTR [23] can be
used to deconvolute the proportion of α-helices, β-strands,
turns, and unordered structures.
4 Notes
1. To make 5× TBE (445 mM Tris Base, 445 mM boric acid,

10 mM EDTA), weight 54 g of Tris base and 27.5 g of boric
acid. Add 20 mL of 0.5 M EDTA (pH 8.0) and complete to
1 L with DI water. The pH of this stock solution will be around
pH 8.3.
2. During ligand exchange, the two phases (organic and aqueous)

might not be distinguishable at some point. Eventually, the
NPs will transfer to the aqueous solution of BPS and the
organic phase will become clear.
3. The stochiometry of ligand on the NPs can be measured with
inductively coupled plasma optical emission spectrometry
(ICP-OES). For that purpose, use a solution of HPLC purified
NPs in DI water, with aqua regia if needed.
4. Larger gold NPs can be synthesized as follow. 30 mg HAuCl4
is dissolved in 240 mL DI water and heated to 60 °C. A 60 mL
aqueous solution that contains 120 mg sodium citrate with
tannic acid and sodium carbonate is also heated to 60 °C. NPs
size can be tuned by the amount of tannic acid and sodium
carbonate. For 10 nm NPs, use 250 mg/L tannic acid and
22 mg/L sodium carbonate. For 16 nm NPs, use 8.33 mg/L
tannic acid and 0.73 mg/L sodium carbonate. Once the tem-
perature in both solutions reaches 60 °C, they are rapidly
mixed together and stirred for 10 min. Let cool to room tem-
perature. NPs are functionalized with BPS by adding 0.1 g of
BPS to the solution and incubating overnight. The NPs are
then precipitated by adding NaCl to the solution. The super-
natant is discarded and the NPs are resuspended in a low salt
buffer. The final step is to run the NP sample in a 2 % agarose
gel and extract the NPs from the gel with centrifugal filters as
described in (Subheading 3.3).
5. Several buffers are appropriate for conjugation. Examples of
suitable buffers include 10 mM NaPi buffer (pH 7.4) or 1×
PBS. In high salt concentration, NPs tend to precipitate. DTT
in buffers can bind to gold atoms on the NP surface. In high
enough concentration, DTT can induce NP aggregation.
6. Some NP ligands, like BPS, might have a strong optical absorp-
tion at short wavelengths (~200 nm), which renders the CD
measurements very noisy unless averaging over long times.
7. Due to their small diameter, 1.5 nm NPs do not cause consid-
erable scattering of the UV light. However, if large aggregates
are formed or large NPs are used, correction for absorption
flattening might be necessary [19, 24].
8. If conjugation conditions are such that all proteins are labeled
with NPs (as in Fig. 3a,b right lane), it is not necessary to
purify the protein–NP conjugates with agarose gel electropho-
resis. It is possible to directly measure the CD signal of the
solution of NPs and proteins, since it does not contain any
unlabelled proteins.
References
1. Chah S, Hammond MR, Zare RN (2005) ribosome by difference mapping and Nanogold
Gold nanoparticles as a colorimetric sensor for labeling. J Biol Chem 276:14117–14123
protein conformational changes. Chem Biol 13. Buechel C et al (2001) Localisation of the
12:323–328 PsbH subunit in photosystem II: a new
2. De M et al (2009) Sensing of proteins in approach using labelling of His-tags with a
human serum using conjugates of nanoparti- Ni2+-NTA gold cluster and single particle anal-
cles and green fluorescent protein. Nat Chem ysis. J Mol Biol 312:371–379
1:461–465 14. Ackerson CJ et al (2006) Rigid, specific, and
3. Willner I, Baron R, Willner B (2007) discrete gold nanoparticle/antibody conju-
Integrated nanoparticle-biomolecule systems gates. J Am Chem Soc 128:2635–2640
for biosensing and bioelectronics. Biosens 15. Zanchet D et al (2001) Electrophoretic isola-
Bioelectron 22:1841–1852 tion of discrete Au nanocrystal/DNA conju-
4. Bhattacharyya S et al (2011) Efficient delivery gates. Nano Lett 1:32–35
of gold nanoparticles by dual receptor target- 16. Mamedova NN et al (2001) Albumin-CdTe
ing. Adv Mater 23:5034–5038 nanoparticle bioconjugates: preparation, struc-
5. Park S, Kimberly H-S (2010) Enhancement of ture, and interunit energy transfer with
in vitro translation by gold nanoparticle-DNA antenna effect. Nano Lett 1:281–286
conjugates. ACS Nano 4:2555–2560 17. Aubin-Tam M-E, Hamad-Schifferli K (2005)
6. Aubin M-E, Morales DG, Hamad-Schifferli K Gold nanoparticle-cytochrome c complexes:
(2005) Labeling Ribonuclease S with a 3 nm the effect of nanoparticle ligand charge on
Au nanoparticle by two-step assembly. Nano protein structure. Langmuir 21:
Lett 5:519–522 12080–12084
7. Hu M et al (2008) Gold nanoparticle-protein 18. Tellechea E et al (2012) Engineering the inter-
arrays improve resolution for cryo-electron face between glucose oxidase and nanoparti-
microscopy. J Struct Biol 161:83–91 cles. Langmuir 28:5190–5200
8. Zheng M, Li Z, Huang X (2004) Ethylene gly- 19. Vertegel A, Siegel RW, Dordick JS (2004)
col monolayer protected nanoparticles: synthe- Silica nanoparticle size influences the structure
sis, characterization, and interactions with and enzymatic activity of adsorbed lysozyme.
biological molecules. Langmuir 20:4226–4235 Langmuir 20:6800–6807
9. Faulk WP, Taylor GM (1971) An immunocol- 20. Srivastava S et al (2005) Controlled assembly
loid method for the electron microscope. of protein-nanoparticle composites through
Immunochemistry 8:1081–1083 protein surface recognition. Adv Mater
10. Geoghegan WD, Ackerman GA (1977) 17:617–621
Adsorption of horseradish peroxidase, ovomu- 21. Weare WW et al (2000) Improved synthesis of
coid and anti-immunoglobulin to colloidal small (dcore 1.5 nm) phosphine-stabilized
gold for the indirect detection of concanavalin gold nanoparticles. J Am Chem Soc
A, wheat germ agglutinin and goat anti-human 122:12890–12891
immunoglobulin G on cell surfaces at the elec- 22. Lundqvist M et al (2006) Induction of struc-
tron microscopic level: a new method, theory ture and function in a designed peptide upon
and application. J Histochem Cytochem adsorption on a silica nanoparticle. Angew
25:1187–1200 Chem Int Ed 45:8169–8173
11. Aubin-Tam M-E, Hwang W, Hamad-Schifferl 23. Sreerama N, Woody RW (2004) Computation
K (2009) Site-directed nanoparticle labeling of and analysis of protein circular dichroism spec-
cytochrome c. Proc Natl Acad Sci USA tra. Methods Enzymol 383:318–351
106:4095–4100 24. Gerdova A, Kelly SM, Halling P (2011)
12. Montesano-Roditis L et al (2001) Cryo- Experimental test of absorption flattening
electron microscopic localization of protein correction for circular dichroism of particle
L7/L12 within the Escherichia coli 70 S suspensions. Chirality 23:574–579
Chapter 4
Water-Solubilization and Functionalization

of Semiconductor Quantum Dots
Christina M. Tyrakowski, Adela Isovic, and Preston T. Snee
Abstract
Semiconductor quantum dots (QDs) are highly fluorescent nanocrystals that have abundant potential for
uses in biological imaging and sensing. However, the best materials are synthesized in hydrophobic surfac-
tants that prevent direct aqueous solubilization. While several methods have been developed to impart
water-solubility, an aqueous QD dispersion has no inherent useful purpose and must be functionalized
further. Due to the colloidal nature of QD dispersions, traditional methods of chemical conjugation in
water either have low yields or cause irreversible precipitation of the sample. Here, we describe several
methods to water-solubilize QDs and further functionalize the materials with chemical and/or biological
vectors.
Key words Quantum dot, Semiconductor nanocrystal, Cap-exchange, Encapsulation, Functionalization,

Carbodiimide coupling
1 Introduction
Semiconductor quantum dots (QDs, or nanocrystals) have gained

a significant level of interest since their discovery in the mid-1980s
due to their size-dependent optical properties [1–3]. Over the next
~25 years of study, many fundamental physical properties have
been characterized; applications were also demonstrated for the
use of QDs in lasers [4], indoor lighting [5], and energy genera-
tion [6] to name a few. Although several physical phenomena are
still the subject of investigation, the development of very high
quantum yield core/shell water-soluble QD systems [7, 8] created
inroads for biological imaging applications. These initial reports
also demonstrated dramatic enhancement of photostability and
brightness of QD-based bioimaging agents compared to organic
29
30 Christina M. Tyrakowski et al.
fluorescent dyes. Chemical and biological sensing has also been

reported using mixed QD-organic coupled fluorophores [9, 10].
Significant challenges still exist concerning the use of colloidal
QD systems for imaging and sensing applications. Biological tar-
geting requires the use of functionalized fluorophores, and chem-
ical and biological sensing capability can be imparted by
conjugating QDs with energy-accepting organic dyes or other
compounds. Furthermore, the research scientist must have con-
trol over all the steps in the water-solubilization and subsequent
functionalization processes to guarantee reproducibility and the
development of transferable technologies. We have written this
chapter with the intention of aiding the development of QD-based
bioimaging technology starting with as-prepared CdSe/CdZnS
nanocrystals dispersed in hydrophobic solvents. We detail three
water-solubilization methods to control size and chemical com-
patibility, as well as strategies, that depend on the method of
water-solubilization employed, for functionalizing the disper-
sions with chemical and biological vectors.
Described in this chapter are procedures for purifying as-
prepared QDs as well as the recently developed zinc-metathesis
cap-exchange method that creates small water-soluble QDs with
enhanced aqueous stability [11]. We also detail the highly robust
method of polymer encapsulation with 40 % octylamine-modified
poly(acrylic acid) [12]. As is well-known, carboxylic-acid coated
cap-exchanged or polymer-encapsulated QDs cannot be easily
functionalized with amine-functional chemical or biological vec-
tors using the common reagent 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide hydrochloride (EDC) [13], which we have recently
demonstrated is due to the cationic character of EDC [14].
Neutral reagents such as methoxypolyethylene glycol carbodi-
imide (MPEG CD) are highly effective for QD functionalization
[14], yet the synthesis is challenging due to the use of highly
specific chemical transformation techniques. As such, we have
detailed the synthesis and usage of this reagent as a part of this
chapter. A method for functionalizing silica-coated QDs is also
detailed, which relies on the presence of thiols on the surface of
the QD arising from the thiol-silica precursor used in the water-
solubilization process. The effect of these water-solubilization
methods on the emission of CdSe/CdZnS QDs is shown in
Fig. 1. A sample of hydrophobic QDs was water-solubilized using
the three methods described in this chapter. The quantum yields,
relative to Rhodamine 6G in ethanol, are 81 % for hydrophobic,
50 % for DHLA-capped, 50 % for silica-coated, and 36 % for
polymer-encapsulated QDs.
Water-Solubilization and Functionalization of Semiconductor Quantum Dots 31
Hydrophobic
DHLA
Silica
Polymer
Absorbance
Emission
420 460 500 540 580 620
Wavelength (nm)
Fig. 1 The absorption spectrum of hydrophobic QDs and the emission spectra
relative to quantum yield of hydrophobic, DHLA-capped, silica-coated, and
polymer-encapsulated QDs
2 Materials
There are several bench-top solvents and equipment that are used
in most of the procedures; they are listed below, with the specific
solvents and equipment needed in an individual procedure listed in
that procedure’s materials section. We store most precursors under
ambient conditions grouped by chemical compatibility unless
specified otherwise by the manufacturer. Note that some proce-
dures use toxic and explosive sodium azide; not only is the use of
this chemical an inherent hazard, extreme caution must also be
taken in the disposal of excess azide reagent.
Common Solvents
1. Chloroform (≥99.5 %), dichloromethane (≥99.5 %), dimeth-
ylformamide (99.8 %, dried over molecular sieves), ethyl ace-
tate (≥99.5 %), ethyl ether (99.9 %), hexanes (≥98.5 %, mixture
of isomers), methanol (≥99.8 %), 2-propanol (≥99.5 %), tetra-
hydrofuran, toluene (99.9 %), silicone oil for a heat bath, and
10 MΩ-cm deionized water (D.I. water) (Milli-Q Water
System, Millipore, Billerica, MA); however, we find that house
deionized water is sufficient.
2. 0.1 M sodium hydroxide solution (0.1 M NaOH): Take an alu-
minum weigh boat or a glass vial and record its mass. Add ~4.0 g
of NaOH (97 %) to the vessel, dry in an oven to a constant
weight, and transfer with small portions of D.I. water to a 1 L
glass container on a kitchen scale. Next, add an additional amount
of D.I. water (~1,000 g) as measured using the kitchen scale to
make a 0.1 M solution; store under ambient conditions.
Common Equipment
1. Vacuum gas manifold (Schlenk line) with a 5.9 cfm pump that
generally delivers 80–200 mTorr of low pressure as measured
with a vacuum gauge and nitrogen gas that is purified by pass-
ing over activated BASF catalyst (R3-11G) and Drierite
(see Note 1).
2. Centrifuge with rotor capable of achieving an RCF (relative
centrifugal force) of 2,000 × g.
3. Rotary evaporator with a cold-water condenser and hot water
bath.
4. Electronic scale (110 g capacity, 0.0001 g accuracy) and
kitchen scale (3 kg capacity, 1 g accuracy).
2.1 Precipitation 1. Hydrophobic QDs: We use emissive CdSe/CdZnS QDs pre-

and Purification pared according to the procedures outlined in ref. 15. We
of As-Prepared CdSe/ believe that the following procedures will also be useful when
CdZnS QDs handling commercially available QD materials when following
the manufacturer’s guidelines for purification.
2. Solvents: Hexane, 2-propanol, and methanol.
3. Equipment: 7 mL glass vials with caps, turnover septum stop-
per, 21 gauge needle, Schlenk line, sonicator, and centrifuge.
2.2 Water-Soluble 1. Precipitated and purified CdSe/CdZnS QDs, prepared as

Silica-Coated QDs described in Subheading 3.1.
2. (3-Mercaptopropyl)trimethoxysilane (95 %).
3. Cesium carbonate (99.5 %).
4. Either zinc chloride (≥98.0 %) or zinc nitrate hexahydrate
(98.0 %).
5. Solvents: dry dichloromethane, D.I. water, 0.1 M NaOH, and
hexane.
6. Equipment: 7 mL glass vial with cap, Teflon flea micro stir bar
(10 × 3 mm), Teflon polygon stir bar (38 × 9.5 mm), turnover
septum stopper, 21 gauge needle, 10 mL disposable syringe
fitted with a 100 or 200 nm filter, 10 mL dialysis tube (catalog
number G235071, Float-A-Lyzer, Type: CE, MWCO:
100 kDa, Spectrum Labs, Rancho Dominguez, CA, USA),
large 1 or 2 L flask, pH indicator strips, Schlenk line, stir plate,
and centrifuge.

Dihydrolipoic Acid described in Subheading 3.1.
Cap-Exchanged QDs 2. Dihydrolipoic acid (DHLA) (see Note 2).
3. Sodium hydroxide (99.5 %).
4. Zinc nitrate hexahydrate (98.0 %).
5. Solvents: D.I. water, and 0.1 M NaOH, methanol.

(10 × 3 mm), Teflon polygon stir bar (38 × 9.5 mm), turnover
septum stopper, 21 gauge needle, centrifuge, Schlenk line,
10 mL disposable syringe fitted with a 100 or 200 nm filter,
10 mL dialysis tube (catalog number G235071, Float-A-Lyzer,
Type: CE, MWCO: 100 kDa, Spectrum Labs, Rancho
Dominguez, CA, USA), large 1 or 2 L flask, stirring hot plate,
and oil bath.

40 % Octylamine- described in Subheading 3.1.
Modified Poly(acrylic 2. Poly(acrylic acid), 1,800 MW.
acid) Polymer-
3. Octylamine (98 %).
Encapsulated QDs
4. 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride.
5. Concentrated HCl (37 %) and diluted HCl: Take an appropri-
ately sized glass flask and fill to 100 mL with D.I. water. Add
4 mL of concentrated HCl and mix with a glass stir rod or
pipet.
6. 0.33 M sodium hydroxide solution (0.33 M NaOH): Take an
aluminum weigh boat or a glass vial and record its mass. Add
~1.33 g of NaOH (97 %), dry in an oven to a constant weight,
and transfer with small portions of D.I. water to a glass con-
tainer. Next, add an appropriate amount of D.I. water (~100 g)
using a kitchen scale to make a 0.33 M solution.
7. Solvents: chloroform, ethyl acetate, D.I. water, 0.1 M NaOH,
dimethylformamide, and methanol.
8. Equipment: Glass 1-neck 150 and 200 mL round bottom
flasks fitted with glass hose adapters, glass 500 mL separatory
funnel with a Teflon stopcock and glass stopper, Teflon egg-
shaped stir bars (19 × 9.5 mm), 5 mL disposable syringe,
Schlenk line, stir plate, sonicator, 50 mL polypropylene centri-
fuge tubes, turnover septum stoppers, 21 gauge needle, 10 mL
disposable syringe fitted with a 100 or 200 nm filter, Amicon
Ultra-15 100K Centrifugal Filter (Millipore, Billerica, MA,
USA), pH indicator strips, and centrifuge.
2.5 Methoxypoly- 1. Methoxypolyethylene glycol 350, Average Molecular Weight:

ethylene Glycol 350.
Carbodiimide Reagent 2. Thionyl chloride (≥99 %).
Preparation
3. Sodium azide (99 %).
4. Triphenylphosphine (99 %).
5. Ethyl isothiocyanate (≥97.0 %).
6. Mercury (II) oxide (HgO), yellow (98 %): Add 6.5 g of HgO
to a 50 mL polypropylene centrifuge tube, and then add
30 mL of D.I. water. Sonicate for 30 min, centrifuge at
2,000 × g, and discard the supernatant. Repeat this D.I. water
wash three times. Transfer the wet HgO to a 20 mL glass vial
(see Note 3) fitted with a screw cap with septa (see Note 1),
connect to a Schlenk line, and remove the D.I. water by vac-
uum, which may take upwards of 2 days.
7. Solvents: dichloromethane, dimethylformamide, D.I. water,
ethyl ether, tetrahydrofuran, and toluene.
8. Equipment: Glass 3-neck 50 mL round bottom flasks fitted
with glass hose adapters, glass stoppers, and turnover septum
stoppers. Glass 500 and 250 mL 1-neck round bottom flasks
fitted with glass hose adapters, Büchner funnel with a fritted
glass filter, glass 500 mL separatory funnel with a Teflon stop-
cock and glass stopper, Teflon egg-shaped stir bars
(19 × 9.5 mm), 3 mL disposable syringe, aluminum foil, 50 mL
polypropylene centrifuge tube, small glass flask, grade 1 filter
paper, Schlenk line, sonicator, stir plate, and centrifuge.
2.6 Functionalizing 1. Aqueous dihydrolipoic acid cap-exchanged QDs or 40 % octy-

Aqueous Carboxylic- lamine-modified PAA polymer-encapsulated QDs, prepared as
Acid Functional QDs described in Subheadings 3.3 or 3.4.
Using MPEG CD 2. Rhodamine B piperazine dye, as prepared by ref. 16 (or other
amine-functional compounds such as a protein).
3. Methoxypolyethylene glycol carbodiimide (MPEG CD), pre-
pared as described in Subheading 3.5.
4. pH 8 phosphate buffer: To a 1 L glass bottle on a kitchen bal-
ance, add 6.81 g of potassium phosphate monobasic (99.5 %),
467 mL of 0.1 M NaOH, and dilute to 1 L with D.I. water.
Cap bottle and dissolve by stirring or sonication.
5. Solvents: D.I. water, and 0.1 M NaOH.
6. Equipment: Teflon flea micro stir bar (10 × 3 mm), Teflon
polygon stir bar (38 × 9.5 mm), 5 mL disposable syringe fitted
with a 100 or 200 nm filter, Amicon Ultra-15 100K Centrifugal
Filter (Millipore, Billerica, MA, USA) or 10 mL dialysis tube
(catalog number G235071, Float-A-Lyzer, Type: CE, MWCO:
100 kDa, Spectrum Labs, Rancho Dominguez, CA, USA), pH
indicator strips, stir plate, and centrifuge.
2.7 Functionalizing 1. Water-soluble silica-coated QDs, prepared as described in

Aqueous Silica-Coated Subheading 3.2.
QDs Using Sulfo-SMCC 2. Rhodamine B piperazine dye, as prepared by ref. 16 (or other
amine-functional compound, such as a protein).
3. Sulfosuccinimidyl-4-( N -maleimidomethyl)cyclohexane-1-
carboxylate (Sulfo-SMCC).
4. pH 6 phosphate buffer: To a 1 L glass bottle on a kitchen

balance, add 6.81 g of potassium phosphate monobasic
(99.5 %), 56 mL of 0.1 M NaOH, and dilute to 1 L with D.I.
water. Cap bottle, and dissolve by stirring or sonication.
5. pH 8 phosphate buffer: To a 1 L glass bottle on a kitchen bal-
ance, add 6.81 g of potassium phosphate monobasic (99.5 %),
467 mL of 0.1 M NaOH, and dilute to 1 L with D.I. water.
Cap bottle, and dissolve by stirring or sonication.
6. Solvents: D. I. water, and 0.1 M NaOH.
(10 × 3 mm), Teflon polygon stir bar (38 × 9.5 mm), 50 mL
polypropylene centrifuge tube, 5 mL disposable syringe fitted
with a 100 or 200 nm filter, spin desalting column (catalog
number 89893, Zeba™, MWCO: 7 kDa, 10 mL, Thermo
Scientific, Rockford, IL, USA), 1 mL dialysis tube (catalog
number G235035, Float-A-Lyzer, Type: CE, MWCO:
100 kDa, Spectrum Labs, Rancho Dominguez, CA, USA),
large flask, pH indicator strips, stir plate, and centrifuge.
3 Methods
All reactions should be performed at room temperature in a fume

hood unless stated otherwise.
3.1 Precipitation 1. Add ~1 × 10−8 mol of QDs in growth solution to a 7 mL glass

and Purification vial; this is typically 0.5 g of growth solution for QDs synthe-
of As-Prepared CdSe/ sized according to the published protocols in ref. 15. Next,
CdZnS QDs add 1 mL of hexane, 0.5 mL of 2-propanol, and fill to the top
with methanol. Vigorously shake the vial to induce floccula-
tion. Centrifuge the sample until ~1 mL of a gelatinous paste
is coating the bottom of the flask, typically ~4 min at 1,700 × g.
Discard the clear, noncolored supernatant.
2. Add 1–2 mL of hexane to the gel; strongly shake the vial to
break up the gel to wash it thoroughly with hexane. Sonication
may be necessary. Centrifuge the cloudy mixture at 1,700 × g
until the hexane supernatant appears to be a clear, colored
solution that is free of any particulates (typically 10 min).
Carefully transfer the QD-hexane extract to another pre-
weighed 7 mL glass vial using a pipet as it is essential that the
gel material is removed at this step. As some QDs may remain
in the gel layer, repeat the hexane wash and extraction if desired
and combine all extracts into the 7 mL glass vial, but do not fill
over half the volume of the vial. Add 0.5 mL of 2-propanol to
the QD-hexane extract and fill to the top with methanol.
Vigorously shake the vial to induce flocculation and then cen-
trifuge the cloudy mixture at 1,700 × g for a minimum of
10 min until the QDs have coated the walls of the glass vial.
Discard the supernatant.
3. Remove the cap, attach a turnover septum stopper, and pierce
the top with a 21 gauge needle. Dry under vacuum with a
Schlenk line (see Note 1). The sample is now ready for use in
any of the water-solubilization procedures described below;
however, processed samples should be water-solubilized on the
same day as the phase-transfer efficiency suffers with time
significantly.
3.2 Water-Soluble 1. Purified, dried CdSe/CdZnS QDs in a 7 mL glass vial from

Silica-Coated QDs Subheading 3.1 are solubilized by addition of ~2 mL dichloro-
methane. The solution should be optically clear; if not, the
sample should be discarded and a new QD sample should be
prepared.
2. 50 μL of (3-mercaptopropyl)trimethoxysilane (0.27 mmol),
88 mg of cesium carbonate (0.27 mmol), and either 9 mg of
zinc chloride (0.067 mmol) or 20 mg of zinc nitrate hexahy-
drate (0.067 mmol) are added with a flea micro stir bar to the
QD dispersion in dichloromethane.
3. Stir the solution 1–2 days. Next, remove the stir bar and cen-
trifuge the vial at 1,700 × g until the cesium carbonate is pel-
leted to the bottom of the vial and the solution is optically
clear, typically 5 min. Transfer the clear supernatant using a
glass pipet into a new 7 mL glass vial. Add hexane to the top of
the vial to induce precipitation (see Note 4).
4. Once the QDs have flocculated, centrifuge the vial at 1,700 × g
until the QDs become pelleted to the bottom of the vial.
Uncap, remove the supernatant, and let dry under ambient
conditions for ~1 h. The gel of QDs should noticeably shrink
and crack during this time.
5. Add two to three drops of methanol and a Teflon flea micro
stir bar, and break up the dried gel until most of the material
appears “wet” with methanol, normally after ~1 min stir. Next,
fill the vial with 0.1 M NaOH and stir (see Note 5).
6. Filter the sample with a 100 or 200 nm filter into a 10 mL
dialysis tube; fill the dialysis tube to the top with D.I. water
and cap tightly. Fill a 1 L or 2 L glass flask containing a poly-
gon stir bar with D.I. water to allow the dialysis tube to float
fully extended. Stir 24 h while replacing the D.I. water regu-
larly (see Note 6) or until the pH of the sample matches the
same of the D.I. water. Transfer to a capped glass vial. Samples
have been found to be stable for months under ambient condi-
tions; regardless, storage at 4 °C is advised.
3.3 Water-Soluble 1. Add 60–80 mg of sodium hydroxide and a flea micro stir bar
Dihydrolipoic Acid to a 7 mL glass vial. Dissolve the sodium hydroxide in ~2 mL
Cap-Exchanged QDs of methanol by stirring.
2. Add 100 mg of DHLA and 70 mg of zinc nitrate hexahydrate
to the sodium hydroxide solution. Place a turnover septum
stopper on the vial, pierce the top with a 21 gauge needle, and
put the solution under N2 via a Schlenk line. Using an oil bath
to heat the solution to ~50 °C (see Note 7), stir until the solu-
tion is clear (see Note 8) and then cool to room temperature.
3. Once cool, add this solution and its stir bar to purified and
dried CdSe/CdZnS QDs. Add a turnover septum stopper and
stir overnight under N2 using a Schlenk line.
4. The next day, remove the stir bar and centrifuge the vial at
1,700 × g for 10 min. Drain the clear supernatant. Place a turn-
over septum stopper on the vial, pierce the top with a 21 gauge
needle and dry the sample very briefly (1–2 min) by vacuum
via a Schlenk line. Add ~6 mL D.I. water and a few drops of
0.1 M NaOH to the sample and shake vigorously until the
QDs have dissolved completely.
5. Filter the samples with a 200 nm or 100 nm filter into a 10 mL
dialysis tube. Fill the dialysis tube to the top with D.I. water
and cap tightly. Fill a 1 L or 2 L glass flask containing a poly-
gon stir bar with D.I. water and place the dialysis tube in the
water, allowing the tube to float fully extended. Stir 24 h while
replacing the D.I. water regularly (see Note 6). Transfer to a
capped glass vial and store at 4 °C.
3.4 Water-Soluble 1. Add 5.00 g of poly(acrylic acid) (PAA) with a Teflon egg-
40 % Octylamine- shaped stir bar to a glass 1-neck 150 mL round bottom flask
Modified Poly(acrylic followed by addition of 95 mL of dimethylformamide (see
acid) Polymer- Note 9). Cap with a hose adapter and connect to a Schlenk
Encapsulated QDs line under N2 while stirring PAA until completely dissolved to
yield a clear solution, typically 5 min.
2. Add 5.33 g EDC to the polymer solution and stir under N2
until the EDC has completely dissolved, typically 30 min. Add
additional solvent if the EDC doesn’t fully dissolve. Next, load
a 5 mL syringe with 4.6 mL of octylamine. Briefly remove the
hose adapter and quickly inject the octylamine while the solu-
tion is being stirred vigorously. Reattach the host adapter and
stir the solution under N2 overnight.
3. The next day, the system is placed under vacuum to reduce the
solvent volume, typically for 5 h; this yields a very viscous,
slightly discolored solution that is transferred in equal parts to
three 50 mL centrifuge tubes (see Note 10). Next, add D.I.
water to the centrifuge tubes to capacity to precipitate the
polymer. Shake the tubes vigorously, and then centrifuge them

for 10 min at 2,000 × g. There should be a white or near-white
precipitate at the bottom of the centrifuge tube and an aque-
ous top layer. Discard the supernatant.
4. Add 33 mL of 0.33 M NaOH to each centrifuge tube. Shake
and sonicate until all polymer is dissolved, or add additional
quantities of 0.1 M NaOH dropwise if the solids do not appear
to be dissolving beyond a certain point (see Note 11).
5. Transfer the basic solution of octylamine-modified PAA to a
separatory funnel and wash three times with ~75 mL aliquots
of ethyl acetate. Pour the bottom aqueous layer in equal parts
into three 50 mL centrifuge tubes and add dilute HCl drop-
wise, vigorously shaking the tubes every few drops, during
which time the polymer precipitates out as a white solid. Dilute
HCl should be added until the pH of the solution is below 5.
6. Centrifuge the samples for 5 min at 2,000 × g to collect all the
40 % octylamine-modified PAA at the bottom of the centrifuge
tube. Discard the supernatant.
7. Add D.I. water to the centrifuge tubes and shake them vigor-
ously to break up the polymer solids. Centrifuge the tubes, and
decant the supernatant after measuring its pH. This process is
repeated until the pH of the supernatant is the same as D.I.
water, typically three times.
8. Transfer the 40 % octylamine-modified PAA samples to a
200 mL round bottom flask (see Note 12) and dry with a
rotary evaporator. Transfer to a Schlenk line and dry further
with vacuum. The sample should be broken up into a coarse
powder and put in a centrifuge tube or capped bottle and
stored at 4 °C. These polymers appear stable on the order of
several years.
9. Calculate the mass of the purified and dried CdSe/CdZnS
QDs from Subheading 3.1. Multiply this mass by 5× and add
at least this weight of 40 % octylamine-modified PAA to the
vial containing QDs.
10. Add ~3 mL chloroform to the vial, shake vigorously, and soni-
cate. If the solution is not optically clear, add two to three
drops of methanol to dissolve the polymer; the sample should
become clear after sonicating for several minutes.
11. Uncap the vial and add a turnover septum stopper. Pierce the
stopper with a 21 gauge needle, and remove the solvent under
vacuum with a Schlenk line. It is imperative that the sample be
thoroughly dried at this step.
12. Fill the vial with 0.1 M NaOH and stir or sonicate until com-
pletely dissolved. Filter with a 100 or 200 nm filter into a
15 mL 100K MWCO concentrating centrifugal filter; fill to
the top via dilution with D.I. water, and centrifuge at the low-
est speed possible to concentrate the sample to ~3 mL. Discard
the rinse in the bottom of the tube (see Note 13), re-dilute the
sample with D.I. water, and repeat as above a minimum of five
times to remove the excess polymer. Transfer to a capped glass
vial and store under ambient conditions. Samples are stable on
the order of years; however, bacterial infections may cause the
formation of a cloudy “plume” that is removable with filtra-
tion. This can be suppressed with storage at 4 °C and/or by
addition of sodium azide.
3.5 Methoxypolye- 1. Add 10 g (28.6 mmol) of methoxypolyethylene glycol 350

thylene Glycol (MPEG) and a Teflon egg-shaped stir bar to a 3-neck 50 mL
Carbodiimide Reagent round bottom flask fitted with a glass stopper, a turnover sep-
Preparation tum stopper, and a hose adapter. Degas and dry the MPEG by
stirring at 80 °C in an oil bath under vacuum via a Schlenk line
for several hours (see Notes 1 and 7). Flush the flask with dry
N2 gas from the Schlenk line.
2. Under N2, cool the solution in an ice bath, and then slowly
drip in 3 mL (41 mmol) of thionyl chloride using a syringe,
equipped with a 21 gauge needle, through the septum while
stirring. Allow the solution to warm to room temperature, and
let the solution stir overnight under N2.
3. Remove the thionyl chloride from the solution by vacuum via
a Schlenk line, which may take several hours. Add 10 mL of
dimethylformamide (DMF) using a graduated cylinder to
dilute the solution in the round bottom, and subsequently
remove the DMF by vacuum. Repeat this addition and removal
of DMF for a total of three times to ensure removal of all thio-
nyl chloride; this may take several days.
4. Add 75 mL of DMF to the chlorinated MPEG and transfer the
solution with the stir bar to a 1-neck 250 mL round bottom
flask. Add 2.83 g (43.5 mmol) of sodium azide to the solution,
and cover the flask with aluminum foil to protect from light.
Stir the solution overnight under N2 at 85 °C using an oil bath
(see Note 7). The next day, cool the solution to room tempera-
ture, and reduce the DMF by vacuum, which may take about
2 days.
5. Add 60 mL of dichloromethane (DCM), and remove undis-
solved solids using a Büchner funnel with a fritted glass filter
(see Note 14) while collecting the clear solution in a second
1-neck 250 mL round bottom flask. Add two 4 mL portions of
DCM to rinse out the first round bottom, pour over the pre-
cipitate in funnel, and collect the wash in the second round
bottom flask.
6. Add a Teflon egg-shaped stir bar to the 250 mL round bottom,

and remove the DCM by vacuum.
7. Add 90 mL of tetrahydrofuran (THF) to the solution, and
then add 8.26 g (31.5 mmol) of triphenylphosphine. Stir this
for 4 h under N2, and then add 1.2 mL of D.I. water by gradu-
ated cylinder and stir overnight under N2.
8. Remove the THF by vacuum. Next, add 200 mL of D.I. water
to dilute the solution in the round bottom and transfer to a
separatory funnel. Rinse the round bottom with two 4 mL
portions of D.I. water and transfer to the separatory funnel.
Add 150 mL of toluene to the separatory funnel, shake, and let
the layers separate. Remove the top organic layer, and repeat
the toluene wash of the bottom aqueous layer two more times.
9. Transfer the bottom aqueous layer to a 1-neck 500 mL round
bottom flask, and remove the D.I. water by rotary evaporation.
Dry further using a Schlenk line while stirring using a Teflon
egg-shaped stir bar until the vacuum pressure stabilizes. Store
the light yellow oil product MPEG 350 amine in an airtight
glass vial at 4 °C.
10. Add 3.0 g (8.6 mmol) of MPEG 350 amine with a Teflon egg-
shaped stir bar to a 3-neck round bottom flask fitted with a
glass stopper, a turnover septum stopper, and a hose adapter.
Degas and dry by stirring under vacuum via a Schlenk line at
70 °C using an oil bath for several hours (see Note 7). After
cooling to room temperature under vacuum, purge the round
bottom with N2, add 5 mL of DMF by graduated cylinder, and
re-purge the round bottom with N2. While stirring under N2,
slowly drip in 1.1 mL (12.6 mmol) of ethyl isothiocyanate by
syringe through the septum, and then let the solution stir over-
night under N2.
11. Remove the DMF under vacuum, and then add 10 mL of
DCM. Next, add 4.65 g (21.5 mmol) of washed and dried
HgO, and stir the solution under N2 overnight. The next day,
the solution should be a dark brown-green color due to the
formation of HgS. Add 1.0 g more of washed and dried HgO
and stir overnight under N2 (see Note 15).
12. Transfer the solution to a 50 mL centrifuge tube with three
3 mL portions of dichloromethane (DCM) to rinse the prod-
uct out of the round bottom flask. Centrifuge the tube at
2,000 × g for 10 min and filter the supernatant through a filter
paper-lined glass funnel into a small glass flask. Add 10 mL
DCM to the precipitate in the tube and sonicate for 15 min.
Centrifuge the tube at 2,000 × g for 10 min and again filter the
supernatant through the filter paper. Rinse the filter paper
with a 5 mL portion of DCM. Using new filter paper, refilter
the collected supernatant through a filter paper-lined glass
funnel into a 1-neck 50 mL round bottom (see Note 16).
Rinse filter paper with a 5 mL portion of DCM into the round

bottom, add a stir bar to the round bottom, and remove the
DCM by vacuum.
13. Put 50 mL of ethyl ether in a small glass bottle, cap loosely,
and chill in a container of ice in a ventilation hood. Once cold,
add 20 mL ether to the round bottom to dissolve the product
and filter through new filter paper in a glass funnel into a new
1-neck 50 mL round bottom flask. Rinse the round bottom
with three 5 mL portions of cold ether and filter into the new
round bottom. Remove the ether by vacuum until the pressure
gauge is stable at a low pressure. Store the dark yellow oil
product MPEG CD in an airtight glass vial at 4 °C or at −80 °C
for a longer shelf life.
3.6 Functionalizing 1. Add 1 mL of aqueous DHLA cap-exchanged or 40 %

Aqueous Carboxylic- octylamine-modified PAA polymer-encapsulated QDs (typi-
Acid Functional QDs cally 1 μM concentration) to a 7 mL glass vial with a cap (see
Using MPEG CD Note 17). Add 1–10 mg of MPEG CD (see Note 18) and a
flea micro stir bar. Cap the vial and stir for 10 min.
2. Prepare a solution (see Note 19) of rhodamine B piperazine
(RBpip) dye or other amine-functional compound in pH 8
phosphate buffer.
3. Once the QD solution has stirred for 10 min, drip the desired
amount of prepared RBpip dye solution or other amine-functional
compound solution into the activated QD dispersion. Check the
pH of the QD solution; if the solution is not at pH 8, add more
pH 8 phosphate buffer. Recap and stir the solution overnight.
4. The next day, if necessary, filter the QDs with a 5 mL disposable
syringe fitted with a 100 or 200 nm filter. In the case of func-
tionalized 40 % octylamine-modified PPA polymer-encapsulated
QDs, transfer the sample to a 100K MWCO concentrating cen-
trifugal filter. Fill to the top via dilution with D.I. water, and
centrifuge at the lowest speed possible to concentrate the sample
to ~3 mL. Discard the rinse liquid (see Note 13) in the bottom
portion of the centrifuge tube. Repeat dilution with D.I. water
and concentration of sample for a minimum of five times to
ensure that all unreacted RBpip dye (or other amine-functional
compound) is removed. Functionalized DHLA cap-exchanged
QDs should be transferred to a 1 mL or 10 mL dialysis tube
(see Note 20); fill tube to the top with D.I. water, and cap
tightly. Fill a 1 L or 2 L glass flask containing a polygon stir bar
with D.I. water to allow the dialysis tube to float fully extended.
Stir 24 h while replacing the D.I. water regularly (see Note 6).
Transfer the dialyzed solution to a glass vial and store under
ambient conditions in the case of functionalized polymer-
encapsulated QDs or at 4 °C in the case of functionalized DHLA
cap-exchanged QDs.
3.7 Functionalizing 1. Add 1 mg of sulfo-SMCC to a 7 mL glass vial. Add 0.5 mL of

Aqueous Silica-Coated pH 6 phosphate buffer to dissolve the reagent. If the pH is
QDs Using Sulfo-SMCC lower than pH 6, add more pH 6 phosphate buffer.
2. Add 1 mL of silica-coated water-solubilized QDs to the sulfo-
SMCC solution. Add 0.1 M NaOH dropwise until the pH is
close to or at pH 7 (see Note 21). Add a flea micro stir bar to
the vial, cap, and stir for 30 min.
3. Prepare a spin desalting column by opening the bottom, plac-
ing in a 50 mL centrifuge tube, and removing the storage solu-
tion by centrifugation at 1,000 × g for 2 min. Drain the liquid
in the bottom of the centrifuge tube. Add ~8 mL of pH 8
phosphate buffer solution to the top of the desalting column,
centrifuge at 1,000 × g for 2 min, and drain the liquid from the
bottom of the tube. Repeat this addition and centrifugation of
pH 8 phosphate buffer through the desalting column two
more times (see Note 22).
4. Once the QD solution has stirred for 30 min, add the QD
solution to the top of the desalting column followed by
~0.2 mL of pH 8 phosphate buffer. Centrifuge at 1,000 × g for
2 min. Transfer the QD solution at the bottom of the centri-
fuge tube to a new 7 mL glass vial and add a flea micro stir bar.
5. Prepare a solution (see Note 19) of rhodamine B piperazine
(RBpip) dye or other amine-functional compound in pH 8
phosphate buffer. Drip desired amount of prepared RBpip dye
solution or other amine-functional compound solution into
the activated QD dispersion. Cap and stir the QD solution
overnight.
6. The next day, filter the QD solution if necessary with a 5 mL
disposable syringe fitted with a 100 or 200 nm filter, and add
to a 1 mL dialysis tube (see Note 20). Fill dialysis tube to the
top with D.I. water, and cap tightly. Fill a large glass flask con-
taining a polygon stir bar with D.I. water to allow the dialysis
tube to float fully extended. Stir 24 h while replacing the D.I.
water regularly (see Note 6). Transfer the solution to a glass
vial and store at 4 °C.
4 Notes
1. When drying a chemical to remove a solvent using a Schlenk

line, monitor the pressure using a vacuum gauge to determine
when the solute is dry. To dry materials in a vial with a turn-
over septum stopper, insert a 1 mL plastic syringe with a luer
slip fitting into the end of the Schlenk line vacuum rubber
tubing connected to the vacuum manifold. The syringe is held
in place with a ring clamp. A needle can be attached to the luer

slip end to preserve the integrity of the vacuum while drying.
2. While DHLA is commercially available, due to the expense we
prepared the material using the methods outlined in either ref.
17 or 18.
3. The wet HgO is very hard to transfer.
4. Sometimes the cap-exchanged QDs do not precipitate vigor-
ously upon addition of hexane. In this case, transfer the con-
tents of the vial to a larger vessel and add more hexane. Chilling
the solution and waiting for a few hours will increase the effi-
ciency of precipitation. Repeating the process using a greater
amount of silica precursors (zinc and carbonate as well) can
prevent problems with poor precipitation yield.
5. In some samples, the QDs will become instantly solubilized in
the base water while others may require overnight stirring.
Regardless, it is recommended that the sample stir overnight.
Low transfer yields may be the result of water in the dichloro-
methane solvent.
6. The rinse water in the flask should be monitored for QD emis-
sion to ensure that the dialysis tube has not broken. Transfer
the remaining sample to a new dialysis tube if there is QD
emission in the rinse.
7. A stirring hot plate with oil bath must have the oil stirred (a
paper clip is sufficient); the temperature of the oil must be
monitored with a thermometer.
8. The solution may have a very slight yellow tint. Sometimes
very fine white particles are present in the solution even after
stirring with heat for a very long time. This does not seem to
affect the cap-exchange of the QDs.
9. The solution might have a slight yellow-brown color; this does
not affect the polymer effectiveness or yield of the product.
10. The round bottom flask should be turned over to allow the
product on the walls of the round bottom flask to drip into the
third centrifuge tube.
11. The 40 % octylamine-modified PAA polymer precipitates due
to the acidic pH of the water. The subsequent addition of
0.08 Eq of NaOH is added to bring the pH over 7.
12. The sample will be very viscous; scrape as much as possible into
the round bottom.
13. Monitor the rinse liquid for QD emission to ensure that the
centrifugal filter has not broken; if rinse liquid has strong QD
emission, switch to a new centrifugal filter. Very minor QD
leakage sometimes occurs when the solution is basic and gen-

erally cannot be avoided.
14. While using a fritted filter is best to remove the excess sodium
azide and sodium chloride, the solid precipitate can also be
removed by centrifugation in a 50 mL centrifuge tube fol-
lowed by transferring the supernatant to a round bottom. This
needs to be performed in portions.
15. Some samples may need a second 1.0 g of HgO. The solution
should be centrifuged and the supernatant filtered (as in
step 12 of Subheading 3.5) before this addition. Add washed
and dried HgO until no more green byproduct is produced.
16. The filter paper typically tears and leaks some of the green HgS
byproduct when filtering the first time. After the second filter-
ing, some green color may still remain in the solution.
17. The pH of the QD solution needs to be ~pH 6 for activation
with MPEG CD. The QD solution should be at this pH if it
was dialyzed in D.I. water after water solubilization.
18. 10 mg of MPEG CD will be less than one drop by pipet. The
amount of MPEG CD necessary to observe a good conjuga-
tion efficiency depends on the nature of the QDs and amine-
functional substrate. We have found yields as high as 95 %
using the procedures reported here.
19. If a certain ratio of amine-functional compound to QD is
desired, first calculate how many QDs are in the 1 mL solu-
tion. Then, determine how many amine-functional compounds
are needed to achieve the ratio. Prepare a solution of the
amine-functional compound, and then determine how much
of this solution should be added to the QD solution to obtain
the desired ratio.
20. If the sample is under 2 mL, the entire sample will fit in the
1 mL dialysis tube.
21. Sulfo-SMCC needs to be at pH 6.5–7.5 in order for the
maleimide to react with the thiols on the QDs.
22. The pH of the liquid in the bottom of the centrifuge tube
should be monitored each time. Once it has reached pH 8, no
further additions of pH 8 phosphate buffer are needed.
Acknowledgments
This work was supported by the ACS PRF (50859-ND10) and by

funds from the University of Illinois at Chicago. We would like to
thank Prof. Hedi Mattoussi for helpful discussions concerning the
synthesis of cap-exchanged CdSe/CdZnS QDs.
References
1. Brus LE (1983) A simple model for the ioniza- 10. Medintz IL et al (2003) Self-assembled
tion potential, electron affinity, and aqueous nanoscale biosensors based on quantum dot
redox potentials of small semiconductor crys- FRET donors. Nat Mater 2:630–638
tallites. J Chem Phys 79:5566–5571 11. Liu D, Snee PT (2011) Water soluble semi-
2. Rossetti R, Nakahara S, Brus LE (1983) conductor nanocrystals cap exchanged with
Quantum size effects in the redox potentials, metallated ligands. ACS Nano 5:546–550
resonance Raman-spectra, and electronic- 12. Wu XY et al (2003) Immunofluorescent label-
spectra of CdS crystallites in aqueous-solution. ing of cancer marker Her2 and other cellular
J Chem Phys 79:1086–1088 targets with semiconductor quantum dots. Nat
3. Ekimov AI, Onushchenko AA (1984) Size Biotechnol 21:41–46
quantization of the electron energy spectrum 13. Mattoussi H et al (2000) Self-assembly of
in a microscopic semiconductor crystal. JETP CdSe-ZnS quantum dot bioconjugates using
Lett 40:1136 an engineered recombinant protein. J Am
4. Klimov VI et al (2000) Optical gain and stimu- Chem Soc 122:12142–12150
lated emission in nanocrystal quantum dots. 14. Shen HY, Jawaid AM, Snee PT (2009)
Science 290:314–317 Poly(ethylene glycol) carbodiimide coupling
5. Colvin VL, Schlamp MC, Alivisatos AP (1994) reagents for the biological and chemical func-
Light-emitting-diodes made from cadmium tionalization of water-soluble nanoparticles.
selenide nanocrystals and a semiconducting ACS Nano 3:915–923
polymer. Nature 370:354–357 15. Snee PT et al (2011) Quantifying quantum
6. Huynh WU, Dittmer JJ, Alivisatos AP (2002) dots with Förster resonant energy transfer.
Hybrid nanorod-polymer solar cells. Science J Phys Chem C 115:19578–19582
295:2425–2427 16. Nguyen T, Francis MB (2003) Practical syn-
7. Hines MA, Guyot-Sionnest P (1996) Synthesis thetic route to functionalized rhodamine dyes.
and characterization of strongly luminescing Org Lett 5:3245–3248
ZnS-capped CdSe nanocrystals. J Phys Chem 17. Gunsalus IC, Barton LS, Gruber W (1956)
100:468–471 Biosynthesis and structure of lipoic acid deriv-
8. Michalet X et al (2005) Quantum dots for live atives. J Am Chem Soc 78:1763–1768
cells, in vivo imaging, and diagnostics. Science 18. Susumu K, Mei BC, Mattoussi H (2009)
307:538–544 Multifunctional ligands based on dihydrolipoic
9. Snee PT et al (2006) A ratiometric CdSe/ZnS acid and polyethylene glycol to promote bio-
nanocrystal pH sensor. J Am Chem Soc compatibility of quantum dots. Nat Protoc
128:13320–13321 4:424–436
Chapter 5
Synthesizing and Modifying Peptides for Chemoselective

Ligation and Assembly into Quantum Dot-Peptide
Bioconjugates
W. Russ Algar, Juan B. Blanco-Canosa, Rachel L. Manthe,
Kimihiro Susumu, Michael H. Stewart, Philip E. Dawson,
and Igor L. Medintz
Abstract
Quantum dots (QDs) are well-established as photoluminescent nanoparticle probes for in vitro or in vivo
imaging, sensing, and even drug delivery. A critical component of this research is the need to reliably con-
jugate peptides, proteins, oligonucleotides, and other biomolecules to QDs in a controlled manner. In this
chapter, we describe the conjugation of peptides to CdSe/ZnS QDs using a combination of polyhistidine
self-assembly and hydrazone ligation. The former is a high-affinity interaction with the inorganic surface
of the QD; the latter is a highly efficient and chemoselective reaction that occurs between 4-formylbenzoyl
(4FB) and 2-hydrazinonicotinoyl (HYNIC) moieties. Two methods are presented for modifying peptides
with these functional groups: (1) solid phase peptide synthesis; and (2) solution phase modification of pre-
synthesized, commercial peptides. We further describe the aniline-catalyzed ligation of 4FB- and HYNIC-
modified peptides, in the presence of a fluorescent label on the latter peptide, as well as subsequent
assembly of the ligated peptide to water-soluble QDs. Many technical elements of these protocols can be
extended to labeling peptides with other small molecule reagents. Overall, the bioconjugate chemistry is
robust, selective, and modular, thereby potentiating the controlled conjugation of QDs with a diverse
array of biomolecules for various applications.
Key words Quantum dot, Peptide, Bioconjugation, Polyhistidine, Self-assembly, Chemoselective

ligation, Hydrazone
1 Introduction
1.1 Quantum Dot Nanotechnology continues to provide new tools for studying
Bioconjugates biochemical and biological systems. In particular, nanoparticle
(NP) materials have much to offer due to their small size, large
surface area-to-volume ratio, and interesting electronic/optical/
magnetic properties that are not accessible at the bulk scale. NPs
are being widely explored as in vitro and in vivo probes for imaging
and sensing, as well as platforms for diagnostics and drug delivery
47
48 W. Russ Algar et al.
[1–3]. Colloidal semiconductor nanocrystals, or quantum dots

(QDs), are currently among the most widely utilized NPs for these
applications. The unique, size-dependent electronic properties of
QDs result in bright luminescence that offers many advantages
over conventional molecular fluorophores: strong, broad absorp-
tion characterized by molar absorption coefficients in the range of
105–106 M−1 cm−1; large two-photon absorption cross-sections,
typically 103–104 GM; narrow photoluminescence (PL) character-
ized by full-widths-at-half-maxima in the range 25–40 nm; peak
PL wavelengths that can be continuously tuned across a broad
spectral range through control of nanocrystal size and semicon-
ductor composition; and superior resistance to photochemical
degradation [4, 5]. As a consequence of these properties, QDs
greatly facilitate multiplexed/multicolor experiments, are excellent
emitters for single molecule (particle) fluorescence imaging/track-
ing, and are ideal donors in Förster resonance energy transfer
(FRET) [6–8]. The latter has been exploited to develop a diverse
array of biosensing configurations based on QDs [9, 10].
High quality QDs are typically synthesized in hydrophobic media
as a type I core/shell structure [11]. The core size and material deter-
mines the optical properties of the QD, while the growth of a thin
shell of a higher band-gap semiconductor protects and enhances those
optical properties. The most well-established example of a core/shell
QD is CdSe/ZnS, which provides bright, size-tunable PL across the
visible spectrum. However, CdSe/ZnS and other QD materials are
not biocompatible or biofunctional as synthesized. Utility in biologi-
cal applications is predicated on applying a hydrophilic coating to the
QD (e.g., a monolayer of bifunctional ligands, or an amphiphilic poly-
mer), which imbues biocompatibility, as well as conjugating biomol-
ecules (e.g., proteins, peptides, oligonucleotides) to provide
biofunctionality. The bioconjugation of QDs enables targeting of tis-
sues and biomarkers, cellular uptake, and selective recognition for
sensing, among other capabilities [1, 4]. The challenge is not neces-
sarily in identifying functional biomolecules, but rather in reliably and
controllably conjugating them to the QDs. The notion of control
includes the number of biomolecules per QD, their attachment point
and orientation, and the homogeneity of those properties—all of
which depend on the bioconjugate chemistry utilized and critically
impact the activity of the final QD-bioconjugate [1].
1.2 Self-Assembly One of the most effective means through which to conjugate bio-
of Polyhistidine molecules to QDs is via metal-affinity coordination between poly-
Appended histidine motifs and the ZnS shell of CdSe/ZnS QDs. This
Biomolecules association occurs spontaneously, rapidly (kon ~ 104 M−1 s−1) and
with high-affinity (Kd ~ 1 nM) [12]. The conjugate valence (i.e.,
number of biomolecules per QD) can also be controlled on the
basis of mixing stoichiometry, albeit subject to a Poisson distribu-
tion [13]. We, along with several other groups, have used
Chemoselective Ligation of Peptides 49
polyhistidine motifs to assemble recombinantly expressed proteins,

synthetic peptides, and even chemically modified oligonucleotides
to QDs [12, 14–19]. Up to 50 ± 10 peptides (<3 kDa) or 12 ± 2
proteins (~44 kDa) can be assembled per QD [20]. We have also
developed the concept of a “starter peptide,” which has both a
polyhistidine motif and a chemical handle for ligation to an oligo-
nucleotide, another peptide, or a protein modified with a cognate
chemical handle. The starter peptide strategy has the advantage of
being modular to allow ligation with functionally diverse biomol-
ecules [14–16], and/or introducing a chemical bond that, while
generally stable, is selectively labile under certain conditions (e.g.,
disulfide, hydrazone) [21–24].
1.3 Chemoselective The reaction between 4-formylbenzoyl and 2-hydrazinonicotinoyl

Hydrazone Ligation moieties to yield a hydrazone bond is chemically orthogonal to
most biological functionalities, and has been employed for the che-
moselective modification of proteins and peptides [25–29]. Other
forms of imine chemistry have been widely used for protein and
peptide modification [30–42]; however, the above reaction is par-
ticularly promising. At slightly acidic or neutral pH, conjugation
rates are typically 101–102 M−1 s−1 with 10 μM reactants and
100 mM aniline as a catalyst, going to completion in ca. 30 min
[25]. The aniline reacts with the 4-formylbenzoyl group to form a
highly reactive iminium intermediate that will react with the
2-hydrazinonicotinoyl group to yield a hydrazone. This reaction
has an equilibrium constant of Keq = 2.3 × 106 M−1 at pH 7.0, and
the resulting hydrazone bond has slow hydrolytic kinetics (khydrolysis =
0.08 × 10−3 s−1) in the absence of aniline [25, 43, 44]. Although
slow, the hydrolysis of similar hydrazone bonds is sufficiently accel-
erated under the moderately acidic conditions of the endolysomal
system or cancerous tissue (e.g., pH 5) to be useful for drug deliv-
ery [21, 22]. Recently, we have shown that this chemistry can be
used to ligate a starter polyhistidine peptide to another peptide or
oligonucleotide, both in bulk solution and when the starter pep-
tide is preassembled to QDs [14, 16]. The starter peptides were
ligated with a cell-penetrating peptide for cellular delivery, sub-
strate peptides for protease activity, or oligonucleotide probes for
hybridization on microarrays.
Here, we describe two general routes to modify a pair of pep-
tides with 4-formylbenzoyl (4FB) and 2-hydrazinonicotinoyl
(HYNIC) moieties, respectively, for use as cognate reactants in
chemoselective hydrazone ligations. The first peptide includes a
polyhistidine motif for assembly to CdSe/ZnS QDs, and functions
as a starter peptide. The second peptide is ligated to the starter
peptide and is also labeled with a fluorescent dye. Upon assembly
to a QD, this configuration provides the opportunity for FRET
and utility in biosensing experiments targeting, for example, vari-
ous proteases [9, 10]. The first route to these peptides that we
describe is 9-fluorenylmethoxycarbonyl (Fmoc)-solid phase

peptide synthesis (SPPS) and in-situ modification with 4FB and/
or HYNIC. This method is suited to well-equipped chemistry lab-
oratories and yields the largest amount of modified peptide. The
second route is based on the solution-phase modification of pre-
synthesized peptides, purchased from commercial sources, with
4FB and HYNIC reagents. Although its scale is typically limited to
smaller quantities, the second method can be carried out in almost
any laboratory by chemists and non-chemists alike. The products
of the SPPS and solution-phase routes are functionally equivalent
and can both be chemoselectively ligated using a protocol that we
describe here. Some elements of the two methods can be inter-
changed and, cumulatively, the various steps we describe can be
adapted and applied to labeling peptides with almost any small
molecule reagent.
2 Materials
2.1 General 1. Personal safety equipment: safety glasses, laboratory coat,

nitrile gloves, and fumehood.
2. Chemical waste receptacles.
3. Ultrapure water with a resistivity of 18.2 MΩ cm, for example,
from a Milli-Q purification system (Millipore, Billerica, MA).
4. Retort stand and adjustable clamps.
5. UV–visible spectrophotometer and quartz cuvette. The cuvette
should be transparent to wavelengths longer than 250 nm.
A microcuvette with a sub-milliliter volume is strongly
recommended.
6. Spectrofluorimeter.
7. Gel imaging system with ultraviolet transillumination.
2.2 Solid Phase 1. A polypropylene syringe fitted with a frit and stopcock (teflon/
Peptide Synthesis of polypropylene). This is referred to as the “reaction vessel.”
4FB/HYNIC Modified The reaction vessel should have a 10 mL capacity for the use of
Peptides 200–250 mg of resin.
2.2.1 Equipment
2. Erlenmeyer flask with side-arm (125 mL capacity or larger).
3. Vacuum source and vacuum trap. The vacuum source may be
a stand-alone pump, central/house vacuum, or even a
Chapman water aspirator pump. Connect the reaction vessel to
the vacuum as shown in Fig. 1. The reaction vessel is con-
nected to the Erlenmeyer flask, which is connected to the vac-
uum source via the vacuum trap.
4. Teflon rod.
Fig. 1 Apparatus for solid phase peptide synthesis. The syringe needle is punched
through a septum on the flask to maintain a vacuum seal
5. Glass vials (e.g., scintillation vials, 10 mL or three dram vials).

6. Glass tubes, disposable (i.e., culture tubes/test tubes).
7. Centrifuge tubes, polypropylene with screw cap: 50 mL
capacity.
8. Centrifuge (with rotor to spin 50 mL centrifuge tubes at
10,000 rcf).
9. Rotary evaporator.
10. Lyophilizer.
11. Block heater (and block to hold glass tubes).
12. Semi-preparative, reverse phase RP-HPLC system. Column:
C-8 or C-12 or C-18, 5 μm, 90–110 Å, 150 × 21.2 mm.
13. Analytical RP-HPLC system. Column: C-18, 5 μm, 90–110 Å,
150 × 4.5 mm.
14. Mass-spectrometer: electrospray ionization mass spectrometer
(ESI-MS; e.g., hyphenated with the analytical RP-HPLC sys-
tem) or matrix assisted laser desorption ionization-time of
flight-mass spectrometer (MALDI-TOF-MS).
2.2.2 Reagents 1. Rink amide aminomethyl resin: 100–200 mesh, loading

0.45 mmol g−1. Approximately 220 mg of resin is useful for a
0.1 mmol scale synthesis (Novabiochem-EMD Millipore, San
Diego, CA).
2. Fmoc-protected amino acids (see Note 1). As an example, we
use the Fmoc-protected amino acids necessary to synthesize
the peptides listed in Table 1.
Table 1
Peptides made using SPPS
Modifications
Name Sequence (N- to C-terminal) Synthetic Post-synthetic

Peptide-1a GSGAAAGLSHHHHHH N-terminal 4FB –
C-terminal amidation
Peptide-2 GLYRGSGEGC N-terminal HYNIC Cys(C)-TMR
3. Peptide synthesis grade (or better) solvents: dimethylfor-

mamide (DMF), dichloromethane (DCM).
4. Acetonitrile (MeCN), HPLC grade.
5. N,N-diisopropylethylamine (DIEA).
6. 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa-
fluorophosphate) (HBTU): prepare 100 mL of a 0.4 M solu-
tion in DMF (Solution A; see Note 2). Caution: HBTU and
HATU (see below) can potentially illicit strong allergic
reactions.
7. 2-(7-aza-1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate (HATU).
8. Piperidine: prepare 500 mL of a 20 % v/v solution in DMF
(Solution B).
9. Trifluoroacetic acid (TFA), neat. Caution: highly corrosive.
10. Trifluoroacetic acid, 0.1 % v/v (aq): prepare 100 mL of solution
by adding 0.1 mL of neat TFA to 100 mL of ultrapure water.
11. Triisopropylsilane (TIS).
12. Ethanedithiol (EDT). Caution: toxic, stench.
13. tert-butylmethylether (TBME).
14. 4-formylbenzoic acid (4FB).
15. 6-Boc-hydrazinonicotinic acid (6-Boc-HYNIC) (Solulink, San
Diego, CA).
16. Ninhydrin (a.k.a. Kaiser) test solutions (Applied Biosystems by
Life Technologies, Foster City, CA): phenol in ethanol (M1),
potassium cyanide/pyridine (M2), ninhydrin in ethanol (M3).
17. Maleimide mono-reactive fluorescent dye. We use
tetramethylrhodamine-5-maleimide (TMR-mal; Invitrogen by
Life Technologies, Carlsbad, CA) as an example.
18. Tris buffer: 50 mM, pH 7.5. The buffer can be degassed by
sparging with nitrogen.
2.3 Solution Phase 1. Disposable polypropylene syringes: 1, 3, 5, and 10 mL.

Modification of 2. Polypropylene microcentrifuge tubes, ≥1.5 mL capacity.
Commercial Peptides
3. Vacuum concentrator (with rotor to fit microcentrifuge tubes).
with 4FB/HYNIC
4. Snap-tight plastic cartridges with frits and Luer-Lok/Slip fit-
2.3.1 Equipment tings on each end; ca. 0.4 mL capacity (see Note 3).
5. Column with frit and stopcock: 2.5 cm diameter × 20 cm.
6. Apparatus/cell for running polyacrylamide gel electrophoresis
(PAGE).
7. Electrophoresis power supply.
2.3.2 Reagents 1. Sodium tetraborate buffer: 100 mM, pH 8.5.

2. Phosphate buffered saline (10× PBS): 100 mM phosphate,
1.37 M NaCl, 30 mM KCl, pH 7.4–7.8.
3. Phosphate buffered saline (PBS): 10 mM phosphate, 0.137 M
NaCl, 3 mM KCl, pH 7.4. Prepare 100 mL of solution by
diluting 10 mL of 10× PBS with 90 mL of ultrapure water.
4. Ethanol (EtOH), 50 % (aq): prepare 30 mL of a 1:1 dilution
with PBS.
5. Imidazole. Prepare a 300 mM solution by dissolving 0.20 g in
10 mL of PBS.
6. MeCN, neat.
7. MeCN, 70 % v/v (aq): prepare 10 mL of a 7:3 dilution with
ultrapure water.
8. Triethylamine acetate (TEAA) buffer: 2.0 M, pH 7.0 (Applied
Biosystems).
9. TEAA buffer, 0.2 M: prepare 100 mL of a 1:9 dilution with
ultrapure water.
10. Two peptides from a commercial supplier. There are many
companies that offer custom synthetic peptides. We use the
peptides listed in Table 2 (Biosynthesis Inc., Lewisville, TX) as
an example. One peptide should have a polyhistidine sequence
at one terminus and an amine group at the opposite terminus
(e.g., Peptide-1b); the second peptide should have an amine
group at one terminus and a thiol group at the opposite termi-
nus (e.g., Peptide-3). The amino acid sequence between the
termini is selected to suit the application.
11. Sulfo-N-succinimidyl-4-formylbenzamide (sNHS-4FB) and
sulfo-succinimidyl-6-hydrazino-nicotinamide (sNHS-HYNIC)
(Solulink; see Note 4).
12. Maleimide mono-reactive fluorescent dye. As an example, we
use Cy3 and Cy5 maleimide mono-reactive dye (Cy3/5-mal),
which are sold dried in tubes as “enough to label 1.0 mg of
antibody” (GE Healthcare, Piscataway, NJ).
Table 2
Peptides purchased commercially
Modifications
Name Sequence (N- to C-terminal) Synthetic Post-synthetic

Peptide-1b GSGAAAGLSHHHHHH C-terminal amidation N-terminal 4FB
Peptide-3 KSSTGGQNGEDTGTSC N-terminal acetylation Lys(K)-HYNIC
C-terminal amidation Cys(C)-Cy3/5
13. Nickel(II)-nitrilotriacetic acid (Ni-NTA)-agarose (Qiagen,

Valencia, CA).
14. Oligonucleotide purification cartridges (OPC; Applied
Biosystems).
15. Permeation gel for size-exclusion chromatography. The frac-
tionation range should be suitable for the size of the peptides.
As an example, we use ca. 15 g of BioGel P4 (fractionation
range 800–4000 Da; BioRad, Hercules, CA). The permeation
gel is prepared as a slurry in PBS and loaded into the column
according to the manufacturer’s instructions.
16. Reagents for casting a polyacrylamide gel (e.g., acrylamide,
bisacrylamide, ammonium persulfate, tetramethylethylenedi-
ame (TEMED), and running buffer) or, alternatively, pre-
cast PAGE gels obtained commercially. We use 8–25 %
gradient polyacrylamide Pharmacia PhastGel SDS buffer
strips (GE Healthcare) for an automated Pharmacia LKB
PhastSystem.
2.4 Chemoselective 1. Items 1–4, 6, and 7 from Subheading 2.3.1.

Hydrazone Ligation 2. Items 2–9, 13, and 14 from Subheading 2.3.2.
3. Aniline, neat (see Note 5).
4. Ammonium acetate buffer (NH4OAc): 100 mM, pH 5.5
(optional).
5. HYNIC modified peptide, as prepared and purified in
Subheading 3.
6. 4FB modified peptide, as prepared and purified in Subheading 3.
2.5 Peptide 1. Water soluble QDs coated with a ligand such as dihydrolipoic acid
Assembly to Quantum (DHLA), DHLA-PEG, or a zwitterionic dithiol (see Note 6).
Dots 2. Cell for agarose gel electrophoresis.
3. Running buffer: 90 mM tris-borate-ethylenediaminetetraace-
tic acid (EDTA; 2 mM), pH 8.0–8.5 (TBE buffer).
4. Agarose: prepare a 1.5 % w/w solution in running buffer.
3 Methods
3.1 Solid Phase Fmoc solid phase peptide synthesis (Fmoc-SPPS) can be done
Synthesis of Modified either manually or using automated methods. Here, we detail the
Peptides procedure for manual synthesis. Stepwise SPPS is a cyclic process
that comprises (1) deprotection of the amine group of an amino
acid (or initial linker) attached to the solid support; (2) chemical
coupling of the Fmoc-protected amino acid in the peptide sequence
being synthesized; and (3) repetition of the previous steps with the
next amino acid. As an example, we synthesize Peptide-1a and
Peptide-2 (see Table 1) and modify their N-termini with 4FB and
HYNIC, respectively. Peptide-2 is further modified with a fluores-
cent dye in a solution phase labeling step. Figures 2 and 3 illustrate
the synthesis and modification of peptides with 4FB and HYNIC/
TMR-mal, respectively.
3.1.1 Fmoc-SPPS 1. Initial deprotection of the rink amide resin. Transfer the resin
(0.1 mmol, ca. 220 mg at a loading of 0.45 mmol/g resin) to
a polypropylene reaction vessel. Swell the resin in DMF for
5 min and then drain the cartridge by applying vacuum.
2. Add 5 mL of Solution B (piperidine) to the resin and stir the
suspension occasionally with a teflon rod. Drain the reaction
vessel after 5 min, add another 5 mL of Solution B, and let
stand 5 min further.
Fig. 2 SPPS scheme for a C-terminal polyhistidine-appended peptide-1a with a N-terminal 4FB modification
Fig. 3 SPPS scheme for a peptide-2 with a N-terminal HYNIC modification and C-terminal cysteine residue for
labeling with TMR-mal
3. Drain and wash with 5 × 5 mL of DMF followed by 3 × 5 mL of

DCM; each wash should take ca. 30 s (see Note 7).
4. Take ca. 1 mg of the resin and perform the ninhydrin test (see
Subheading 3.1.2) in a small glass tube: a blue color should
appear. Parallel treatment of a control resin with the Fmoc
group intact is useful for comparison (see Note 8).
5. Swell the resin in DMF for 2 min and then drain. The next step
is coupling the first amino acid residue.
6. Dissolve Fmoc-aa1-OH (0.5 mmol; “aa1” is the first amino
acid in the peptide sequence attached to the resin) in 1.25 mL
of Solution A (HBTU) in a glass vial (see Note 9). Add neat
DIEA (96 μL, 0.55 mmol) to the solution, shake for 30 s, and
then add to the resin.
7. Stir the mixture periodically for 30 min.
8. Drain the reaction vessel and wash the resin with 5 × 5 mL of
DMF and 3 × 5 mL of DCM in succession; each wash should
take ca. 30 s.
9. Take ca. 1 mg of the resin and perform the ninhydrin test. If a

positive result is obtained (i.e., blue color) repeat steps 6–8; if
a negative result is obtained, proceed to step 10.
10. Couple the remaining amino acid residues in the peptide
sequence by repeating steps 2–9, which correspond to
deprotection using Solution B and coupling Fmoc-aan-OH
using Solution A and DIEA (where n > 1 is the cycle number
and corresponding residue from the C-terminus of the
peptide).
11. Proceed to Subheading 3.1.3 for modification with 4FB or
Subheading 3.1.4 for modification with HYNIC.
3.1.2 The Ninhydrin Test The ninhydrin or Kaiser test is a colorimetric chemical test that is
used for quality control during SPPS. Primary amines react with
ninhydrin to form a chromophore with a deep blue/violet or
“Ruhemann’s Purple” color. A positive test is desired following
deprotection steps in SPPS since this indicates that primary amines
are available for coupling the next amino acid in the peptide
sequence. Conversely, a negative test is desired following coupling
steps since this indicates that the reaction went to completion.
We use the ninhydrin test qualitatively to minimize the amount of
time needed to synthesize the peptide; however, the test can also
be used quantitatively if desired [45].
1. Caution: treat the reagents and spent glass tubes as hazardous
materials.
2. Transfer ca. 1 mg of dry resin to a glass tube.
3. Add two drops each of the M1, M2, and M3 solutions and
heat at 100 °C for 5 min.
4. An intense blue color will appear in the presence of free amine,
which may correspond to successful deprotection (step 4 in
Subheading 3.1.1) or incomplete coupling of an amino acid
(step 9 in Subheading 3.1.1).
3.1.3 Solid Phase 1. Dissolve 4FB (75 mg, 0.5 mmol) in 1.25 mL of Solution A
Modification with 4FB (HBTU). Add neat DIEA (96 μL, 0.55 mmol) and shake the
resulting mixture for 30–60 s. Add this solution to the resin
and stir occasionally for 30 min.
2. Drain and wash the resin with 5 × 5 mL of DMF and 3 × 5 mL
of DCM in succession; each wash should take ca. 30 s.
3. Confirm coupling using the ninhydrin test (see
Subheading 3.1.2). Repeat step 1 if a positive result is obtained
(i.e., blue color).
4. Dry the resin under vacuum for 2 h.
5. Proceed to Subheading 3.1.5.
3.1.4 Solid Phase 1. Dissolve 6-Boc-HYNIC (38 mg, 0.15 mmol) and HATU
Modification with HYNIC (57 mg, 0.15 mmol) in 1 mL of DMF. Add neat DIEA (31 μL,
0.18 mmol) and shake the solution for 30–60 s.
2. Add the solution from step 1 to the resin and stir periodically
for 2 h.
3. Drain and wash the resin with 5 × 5 mL of DMF and 3 × 5 mL
of DCM in succession; each wash should take ca. 30 s.
4. Confirm coupling using the ninhydrin test (see Subheading
3.1.2). Repeat steps 1 and 2 if a positive test result is obtained
(i.e., blue color).
5. Dry the resin under vacuum for 2 h.
3.1.5 Cleavage of The following steps are used to cleave the SPPS-grown peptides
Peptides from Rink Resin from the rink amide resin and remove the protecting groups from
and Removal of the amino acid side chains. Pay close attention to the protocol as
Protecting Groups there are variations depending on the nature of the peptide
synthesized.
1. Important: the 4FB is potentially sensitive to EDT and TIS
(see Note 10). If cleaving and deprotecting a 4FB-modified
peptide (e.g., Peptide-1a), skip step 2 and proceed with step
3. If the peptide is HYNIC-modified (e.g., Peptide-2), pro-
ceed with step 2.
2. Prepare a 10 mL solution of TFA:EDT:H2O:TIS (94:2.5:2.5:1
v/v) and cool to ca. −8 °C in a freezer or on dry ice (see Notes
11 and 12). Skip step 3 and proceed with step 4.
3. Prepare 10 mL of 97 % v/v TFA (aq) and cool to ca. −8 °C in
a freezer or on dry ice. Proceed to step 4.
4. Add the cleavage/deprotection solution to the resin. Close the
reaction vessel and gently agitate for 90 min at room tempera-
ture (see Note 13).
5. Drain the resin and collect the solution, which contains pep-
tide, into a round bottom flask. Wash the resin with 2 × 5 mL
of neat TFA and 2 × 5 mL of DCM in succession, and collect
the washes. Each wash should take ca. 30 s.
6. Important: if the peptide contains a polyhistidine sequence
(e.g., Peptide-1a) proceed to step 7 (see Note 14). Otherwise,
skip steps 7–9 and proceed to step 10.
7. Concentrate the peptide containing TFA solution to dryness
in a rotary evaporator. In the absence of a rotary evaporator,
this can be done under a stream of nitrogen.
8. Add 20 mL of TBME and 20 mL of 0.1 % TFA (aq) to the
peptide residue. Transfer this mixture to a 50 mL centrifuge
tube or extraction funnel. Mix the two phases vigorously and
let them separate. Retain the aqueous phase, which contains

peptide, and extract the TBME phase further with 3 × 20 mL
portions of 0.1 % TFA (aq).
9. Combine and lyophilize the aqueous phase extracts. Proceed
to Subheading 3.1.6.
10. Concentrate the peptide containing solution to ca. 1 mL in a
rotary evaporator.
11. Add the concentrated peptide/TFA solution to 30 mL of cold
TBME in a 50 mL centrifuge tube. A white precipitate will
form.
12. Isolate the precipitate by centrifugation at ~1,000 rcf for 6 min
and discard the supernatant. Proceed to Subheading 3.1.6.
3.1.6 Peptide Purification 1. The peptide can be purified by semipreparative RP-HPLC, and
the purity confirmed by analytical HPLC and mass spectro-
metric analysis (e.g., ESI-MS, MALDI-TOF MS).
2. Set up the analytical/semipreparative RP-HPLC. The details
of both the instrument setup and operation will vary between
manufacturers, models, and software. A two-component gra-
dient elution will be used: Eluent A is 0.1 % v/v TFA (aq);
Eluent B is 0.05 % v/v TFA in MeCN.
3. For semipreparative RP-HPLC, the elution program is from 0
to 5 % Eluent B over 5 min and 5–70 % Eluent B over 80 min
at a flow rate of 15 mL/min.
4. For analytical RP-HPLC, the elution program is from 0 to
70 % Eluent B over 30 min at a flow rate of 1 mL/min.
5. Lyophilize and store the peptide as a TFA salt at −20 °C.
Protect from light.
3.1.7 Fluorescent Any peptide with a cysteine residue can be labeled with a maleimide
Labeling derivative of a fluorescent dye of interest. The maleimide is selec-
tive for thiols such as those on the cysteine side chain, at near neu-
tral pH; labeling can therefore proceed in the presence of amine
groups or HYNIC. We label the terminal cysteine residue of
HYNIC-modified Peptide-2 with TMR-mal as an example
(see Fig. 3).
1. Dissolve TMR-mal (4.1 μmol) in 2.0 mL of degassed 1:1 Tris
buffer/MeCN (see Note 15) and add HYNIC-modified
Peptide-2 (6.2 μmol).
2. Agitate the reaction at room temperature until completion, which
can be monitored by analytical RP-HPLC (see Subheading 3.1.6).
3. Purify the product by semi-preparative RP-HPLC (see
Subheading 3.1.6). In the special case that the peptide
has a polyhistidine sequence, the purification steps in
Subheading 3.2.3 may be used as an alternative.
Fig. 4 Solution phase modification and labeling scheme: (a) modification of peptide-3 with HYNIC and Cy3/5;
(b) modification of peptide-1b with 4FB; (c) chemoselective ligation; and (d) assembly to a QD (peak PL 520
or 620 nm) and FRET
3.2 Solution Phase As an alternative to SPPS, peptides can be purchased commer-

Modification of cially from several suppliers and modified with 4FB or HYNIC
Commercial Peptides using commercial reagents and solution phase labeling techniques.
While SPPS is typically limited to organic chemistry laboratories,
the solution phase protocol in this section is more widely accessi-
ble to non-specialists. As an example, we modify commercially
obtained Peptide-1b and Peptide-3 (see Table 2) with 4FB and
HYNIC, respectively. The latter is also modified with a fluorescent
dye in a one-pot, two-step method. The full sequence of steps is
illustrated in Fig. 4, including chemoselective hydrazone ligation
(Subheading 3.3) and assembly to a QD (Subheading 3.4).
3.2.1 Solution Phase The following steps are used to modify a peptide at an available
Modification with 4FB amine group (e.g., N-terminus or lysine side chain) using sNHS-
4FB. We modify the α-amine at the N-terminus of Peptide-1b.
These steps can be extended as a general method for labeling any

peptide with any NHS-ester activated reagent (see Note 16),
assuming that the peptide has a reactive amine group (e.g., unpro-
tected N-terminus or Lys residue).
1. Dissolve ca. 1 mg of peptide in 10 μL of DMSO (see Note 17)
and add 0.5 mL of borate buffer. Other non-amine containing
buffers with pH ≥7 are also suitable.
2. Dissolve 3.0 mg of sNHS-4FB in 30 μL of DMSO. Add the
peptide solution from step 1 and dilute to 1.0 mL with borate
buffer (see Note 18). Mix well. The final peptide concentra-
tion should generally be in the range of 0.1–1 mM (ca. 10- to
100-fold excess of sNHS-4FB).
3. Agitate the reaction for 4–8 h at room temperature. For con-
venience, the reaction mixture can be stored overnight at 4 °C
before proceeding with the next steps.
4. Complete the steps in Subheading 3.2.3 and then proceed
with step 5 (see Note 19).
5. Pre-weigh a microcentrifuge tube.
6. Complete the steps in Subheading 3.2.4 and elute the desalted
peptide into the pre-weighed tube.
7. Evaporate the solvent to dryness in a vacuum concentrator and
re-weigh the centrifuge tube. The difference in mass corresponds
to the quantity of peptide. The result should only be used as a
rough estimate of the amount of material. Peptides with trypto-
phan or tyrosine residues can be more accurately quantitated
prior to drying by measuring their absorbance at 280 nm.
3.2.2 Solution Phase The following steps are used to sequentially modify a peptide with
Modification with HYNIC a fluorescent dye and HYNIC at available thiol (e.g., cysteine side
and Fluorescent Labeling chain) and amine (e.g., N-terminus or lysine side chain) groups,
respectively, in a one-pot reaction. As an example, we modify
Peptide-3 (see Table 2) with sNHS-HYNIC and Cy3/5-mal at its
N-terminal lysine residue and C-terminal cysteine residue, respec-
tively. The dual modification is enabled by the selective reaction of
the Cy3/5-mal with the cysteine residues at near neutral pH.
Following the conversion of these residues, the sNHS-HYNIC can
react exclusively with the lysine residues. The order of these steps
cannot be interchanged since NHS esters react efficiently with
both amine and thiol groups.
1. Dissolve ca. 1 mg of Peptide-3 in 10 μL of 50 % MeCN (see
Note 17). Add 200 μL of 10× PBS.
2. Obtain two tubes of Cy3/5 maleimide mono-reactive dye
and dissolve each in 10 μL of DMSO. Add the peptide solu-
tion from step 1 to each vial in succession. Dilute with 200 μL
of ultrapure water (see Note 18).
3. Agitate the reaction for 4–8 h at room temperature. For

convenience, the reaction mixture can be stored overnight at
4 °C before proceeding with the next steps.
4. Dissolve 3.0 mg sNHS-HYNIC in 30 μL of DMSO
(see Note 18).
5. Transfer the Cy3/5 maleimide reaction mixture to the sNHS-
HYNIC. Add 200 μL of ultrapure H2O and 200 μL of 10×
PBS. Mix well.
6. Agitate the reaction for 4 h at room temperature. For conve-
nience, the reaction mixture can be stored overnight at 4 °C
before proceeding with the next steps.
7. Load the reaction mixture onto a gel permeation column and
elute using PBS. Collect 2–3 mL fractions in small glass tubes.
The first dye-colored band that elutes is the product; combine
the corresponding fractions (see Note 20). Alternatively, the
peptide can be purified by RP-HPLC.
9. Quantitate the peptide using the absorbance of the dye. The
molar absorption coefficient (Cy3, ε552 = 150,000 M−1 cm−1;
Cy5, ε650 = 250,000 M−1 cm−1) is generally included in the
product literature or available online from the manufacturer
(see Note 21).
3.2.3 Purification of The following steps are used to quickly and conveniently purify
Polyhistidine Appended polyhistidine appended peptides from excess labeling reagents.
Peptides The purified peptide is obtained in buffer with high concentrations
of imidazole, and typically requires subsequent desalting (see
Subheading 3.2.4). As an example, we purify Peptide-1b from
excess sNHS-4FB after modification using the steps in
Subheading 3.2.1.
1. Load the Ni-NTA-agarose into a plastic cartridge (see Note 3).
Wash the Ni-NTA-agarose with 5 mL of PBS. Repeat this pro-
cedure with a second purification cartridge (see Note 22). If
not proceeding to step 2 immediately, the Ni-NTA-agarose
should be kept hydrated.
2. Flush the reaction mixture through the first Ni-NTA-agarose
cartridge using a double-syringe technique for ca. 2–3 min (see
Note 23). The polyhistidine sequence of the peptide will bind
to the Ni-NTA-agarose. Retain the reaction mixture.
3. Wash the Ni-NTA-agarose with 10 mL of PBS, 10 mL of 1:1
EtOH:PBS (see Note 24), and 2 × 10 mL of PBS in succession
(see Note 25).
4. Flush the reaction mixture through the second Ni-NTA-
agarose cartridge using a double-syringe technique. Repeat
step 3 with this cartridge.
5. Elute the bound peptide from one of the Ni-NTA-agarose car-

tridges using successive 3 × 0.5 mL portions of 300 mM imid-
azole in PBS (see Note 26).
6. Repeat step 5 with the second Ni-NTA-agarose cartridge. It is
preferable to use ≤3 mL of total eluent between both car-
tridges (see Note 23).
3.2.4 Peptide Desalting 1. Condition a fresh OPC by flushing with 3 mL of MeCN fol-
lowed by 3 mL of 2 M TEAA buffer.
2. Flush the peptide containing solution through the OPC using
a double-syringe technique (see Note 23). The peptide will
bind to the resin.
3. Wash the OPC with 4 × 10 mL of 0.2 M TEAA buffer
(see Note 25).
4. Elute the peptide from the resin with 2 × 0.3–0.5 mL of 70 %
MeCN (aq) (see Note 26).
5. Regenerate the OPC by repeating step 1.
6. Repeat steps 2–4 and combine the eluents. If the peptide is dye/
chromophore labeled, or has Try, Trp, or Cys residues, its concen-
tration can be determined by UV–visible spectrophotometry.
7. Vacuum concentrate to dryness.
8. Store at −20 °C until needed.
3.3 Chemoselective Since reaction between the 4FB and HYNIC moieties has an equi-
Hydrazone Ligation librium constant of Keq = 2.3 × 106 M–1 at pH 7.0 [25], a 1:1 stoi-
chiometric reaction does not go to completion at low millimolar
concentrations (i.e., 0.01–0.05 mM), and the hydrazone product
is in equilibrium with some 4FB and HYNIC. In order to achieve
full conversion to the hydrazone ligated product, a small excess
(≥1.5-fold) of one of the reagents is required. The following steps
assume that one of the peptides has a polyhistidine sequence. As
examples, we ligated 4FB-modified Peptide-1a and HYNIC/
TMR-modified Peptide 2 from Subheading 3.1, as well as
4FB-modified Peptide-1b and HYNIC/Cy3/5-modified Peptide
3 from Subheading 3.2 (see Fig. 4).
1. Dissolve ca. 0.1–1 μmol of 4FB-modified polyhistidine peptide
(see Note 27) in 0.5 mL of either 10× PBS or NH4OAc buffer
(0.2–2 mM peptide).
2. Dissolve ca. 2–3 Eq of the HYNIC-modified peptide in 0.5 mL
of the same buffer used in step 1 (see Note 28).
3. Mix the two peptide solutions from steps 1 and 2 together.
4. Add 1–10 μL of aniline (see Note 29), which corresponds to a
final concentration of 10–100 mM in a 1 mL reaction volume.
Fig. 5 UV–visible absorption characterization of the chemoselective hydrazone reaction between (a) HYNIC-
Peptide-3-Cy3 and 4FB-Peptide-1b, and (b) HYNIC-Peptide-3-Cy5 and 4FB-Peptide-1b. The insets show reac-
tion progress curves from monitoring the change in absorption at 354 nm. In (a), the aliquot extracted for this
purpose was catalyzed by 1 mM aniline, whereas in (b) the reaction was catalyzed by 10 mM aniline. The full
absorption spectra shown are for the final ligated and purified products, highlighting the (i) hydrazone and (ii)
Cy3/5 absorption bands
5. To monitor the progress of the reaction, remove an aliquot of

the reaction mixture and dilute by a factor of 1/100 (1 mM
peptide) or 1/10 (0.1 mM peptide) into a UV-transparent
cuvette (so that A354 < 1 throughout the reaction). Aniline can
be (optionally) added to the same final concentration as in step
4. Track the change in absorbance at 354 nm with time using
a UV–visible spectrophotometer (see Fig. 5). This dilution will
have a slower reaction rate than the reaction mixture since the
peptide concentration is lower.
6. The reaction should go to completion in ca. 30–120 min
depending on concentrations and conditions. For convenience,
the reaction mixture can be stored overnight at 4 °C before
proceeding with the next steps.
7. Complete the steps in Subheading 3.2.3.
8. Complete the steps in Subheading 3.2.4.
9. The peptide can be quantitated using the absorption of the
hydrazone chromophore at 354 nm (ε354 = 29,000 M−1 cm−1;
see Fig. 5). Dilute an aliquot of the desalted peptide in 70 %
MeCN with ultrapure water to an optical density <1 for the
measurement. The amount of peptide is calculated using the
Beer-Lambert Law.
10. The purity of the peptide can be confirmed by RP-HPLC
(see Subheading 3.1.6) or using polyacrylamide gel electro-
phoresis (see Fig. 6). If the peptides were prepared using the
protocol in Subheading 3.2 it is useful to set aside and save
Fig. 6 (a) Pseudo-color image of a UV-illuminated polyacrylamide gel confirming solution-phase peptide modification
and chemoselective hydrazone ligation: (i) HYNIC-Peptide-3-Cy5 reactant, (ii) the hydrazone ligation reaction mixture,
(iii) material remaining in the supernatant following Ni-NTA-agarose purification, and (iv) the ligated and purified
reaction product. (b) Pseudo-color image of a UV-illuminated polyacrylamide gel showing the purity and equivalence
of (i) Cy5 and (ii) Cy3 labeled hydrazone ligation products following purification over Ni-NTA-agarose and desalting.
PL collected using an optical filter for Cy3 is colored yellow; PL collected using an optical filter for Cy5 is colored red
small aliquots (e.g., 10 μL) of dye labeled peptide at different

stages of the process. These aliquots can be run on a PAGE gel
alongside the purified ligated peptide (see Fig. 6).
11. Dry the peptide using a vacuum concentrator (without apply-
ing heat) and store at −20 °C (see Note 30).
3.4 Assembly The ligated peptides can be assembled with QDs simply by mixing
of Ligated Peptides these two components at the desired stoichiometry, provided the
to Quantum Dots concentration of both is sufficiently high (≥100 nM).
1. If the ligated peptide was dried prior to use, dissolve by adding
10–20 μL of 50 % MeCN (aq) (see Note 17). Add the desired
buffer (e.g., PBS, borate) to a peptide concentration between
five- to tenfold larger than the stock solution of QDs.
2. Select the (average) number of peptides to be assembled per QD.
3. Select the QD concentration to be used. A concentration in
the range of 0.1–0.5 μM will provide ample signal for most
spectrofluorimeters.
4. The QD-peptide conjugates are assembled by mixing the pep-
tides and QDs in buffer (see Note 31). Calculate the volumes
of the peptide and QD stock solutions needed (see Note 32)
based on steps 1 and 2. Calculate the volume of buffer needed
to bring the sample up to the desired concentration.
Fig. 7 Pseudo-color images of a UV-illuminated agarose gel (1.5 %) showing the assembly of Cy3 labeled
hydrazone ligation product with CL4-coated QD520 (see Note 6). The numbers below each band are the aver-
age number of peptides added per QD. PL collected using an optical filter (a) for the QD520 is shown in green;
(b) for the Cy3 is shown in red; and (c) a merge of the two images showing colocalization. Each well contained
6 pmol of QD, loaded in 15 µL of buffer. The run time as ca. 5 min at a field strength of 10 V/cm
Fig. 8 Confirmation of assembly between (a) QD520 and Cy3 labeled hydrazone ligation product, and (b) QD620
and Cy5 labeled hydrazone ligation product. Each plot shows progressive quenching of QD PL and sensitiza-
tion of Cy3/5 PL as more QDs are added per QD. Cy3/5 PL from direct excitation has been subtracted from the
spectra. The insets show the change in peak PL intensities and calculated FRET efficiencies
5. Add the peptide and buffer to a microcentrifuge tube, fol-

lowed by the QDs. Mix immediately and let stand at room
temperature for ≥30 min. The sample is then ready for mea-
surement or other use.
6. Assembly can be confirmed by taking an aliquot of the sample
(or preparing a second analogous sample in parallel) to run
alongside a sample of only QDs on a 1.5 % agarose gel
(see Note 33; see Fig. 7).
7. If the QD and dye label on the peptide form a donor–acceptor
pair for Förster resonance energy transfer (FRET; see Note 34),
then quenching of the QD (donor) luminescence and sensitiza-
tion of the dye (acceptor) luminescence can provide confirma-
tion of peptide assembly (see Fig. 8). Separate controls with the
equivalent amount of QD and peptide alone should be mea-
sured in parallel.
4 Notes
1. Amino acid side chain protecting groups (in parenthesis) in

Fmoc-SPPS: Arg(Pbf), Asn(Trt), Asp(tBu), Cys(Trt), Gln(Trt),
Glu(tBu), His(Trt), Lys(Boc), Ser(tBu), Thr(tBu), Trp(Boc),
and Tyr(tBu), where Pbf is 2,2,4,6,7-pentamethyl-2,3-
dihydrobenzofuran-5-sulfonyl, Trt is trityl, tBu is tert-butyl,
and Boc is tert-butyloxycarbonyl.
2. The HBTU solution is generally stable for 2 weeks at room
temperature when protected from light; however, storage at
4 °C and under inert atmosphere is recommended when
possible.
3. Empty oligonucleotide purification cartridges (OPC; Applied
Biosystems) work well for this purpose. The cartridges can be
snapped apart and the resin removed. Cartridges can be cleaned
by storage in a 50 % aqueous solution of a polar organic solvent
such as methanol, ethanol, or acetonitrile, and then reused
over several peptide purifications.
4. The sNHS-4FB and sNHS-HYNIC are sold by Solulink under
the abbreviated names “sulfo-S-4FB” and “sulfo-S-HYNIC.”
5. Aniline will gradually oxidize in air and change from colorless
to red/brown. Although a partially oxidized aniline solution
may retain sufficient catalytic activity for chemoselective hydra-
zone ligation, we recommend the use of fresh reagent to obtain
the best results.
6. CdSe/ZnS core-shell QDs were synthesized with a coating of
hydrophobic ligands including trioctylphosphine (TOP), trioc-
tylphosphine oxide (TOPO), and hexadecylamine (HDA) as
described previously [46, 47]. The QDs were made water sol-
uble by exchanging the TOP/TOPO/HDA with polyethyl-
ene glycol-appended dihydrolipoic acid (DHLA-PEG) or a
compact, zwitterionic dithiol ligand (CL4) using published
procedures [47, 48]. The ligand exchange procedure can be
applied to hydrophobic QDs purchased from commercial
sources. The details of these methods are beyond our current
scope but can be found in the cited references. Other thiol-
based ligands can be used in preference to DHLA-PEG or
CL4. QDs with peak PL at 520 nm are referred to as QD520,
and those with peak PL at 620 nm are referred to as QD620.
7. It is critical to remove all the piperidine from the resin; traces
of piperidine will interfere with the subsequent coupling
reaction.
8. The ninhydrin test should strictly be performed following every
deprotection step and coupling reaction; however, in practice,
it is generally only necessary to perform a ninhydrin test after
each coupling reaction. Fmoc deprotection with Solution B
typically goes to completion without difficulty. As an alternative

to the ninhydrin test, Fmoc removal can be monitored using
the absorbance of the piperidine–dibenzofulvene adduct at
301 nm.
9. Complete dissolution of the Fmoc-protected amino acid can
be assisted by low power ultrasonication.
10. TIS and EDT can react with 4FB to yield a corresponding
alcohol and thioacetal, respectively. Many other scavengers are
also reactive and, as a consequence, we recommend that only
water be added to the TFA cleavage solution. The success of
the 4FB-modified peptide synthesis largely depends on the
success of this cleavage step.
11. Caution: EDT has a stench and is toxic; handle with care.
Glassware or plasticware contaminated with EDT should be
cleaned with bleach solution.
12. The addition of 2.5 % of acetone to the cleavage mixture will
yield the corresponding 6-hydrazinonicotinate acetone hydra-
zone, which is the commercial HYNIC derivative used in the
solution phase modification protocol (see Subheading 3.2.2).
13. It is possible that a fraction of the HYNIC is trifluoroacetylated
in this step; longer cleavage times lead to higher trifluoroacety-
lation yields of the HYNIC. The HYNIC can be de-trifluoro-
acetylated in 0.1 M NaOH buffer for 30 min [49].
14. Polyhistidine peptides tend to precipitate from TBME very
poorly and must be recovered using a different method.
Protecting groups (e.g., trityl) are removed by extraction into
the TBME and the peptide is retained in the aqueous phase.
Extraction of the lipophilic protecting groups also facilitates
purification of the peptide by RP-HPLC.
15. Other buffers that do not contain thiols may be used for this
reaction; however, the solution should be kept between pH
7.0–7.5. Although higher pH yields a faster reaction, it also
increases the hydrolysis rate of the succinimide product.
16. While NHS-esters are generally selective for amines over
hydroxyls, they do react efficiently with thiols to yield thioes-
ters. Cysteine residues should thus be avoided in the peptide
sequence (or protected as a disulfide) for specific labeling of an
amine group.
17. The most reliable approach to dissolving peptides of almost
any amino acid sequence in buffer is to first add a small amount
of an organic solvent such as DMSO or 50 % MeCN (aq).
18. Solulink conveniently sells the sNHS-4FB and sNHS-HYNIC
as 1.0 mg quantities in individual vials. The contents of each
vial may be dissolved in 10 μL of DMSO and the 0.5 mL pep-
tide solution added to each in succession. We use three vials in
our suggested protocol, totaling 3.0 mg dissolved in 30 μL of

DMSO. To help maximize the transfer of material, water/buf-
fer may be used in two portions to rinse each vial in succession
with addition to the reaction mixture. This strategy is also
effective with the premeasured tubes of Cy3/5 maleimide
mono-reactive dye sold by GE Healthcare. While the pre-
weighed packages are convenient, they are not essential to the
protocol.
19. This step assumes that the 4FB-modified peptide has a polyhi-
stidine sequence, as in the case of our example. If the peptide
does not, skip this step and proceed with step 7 in
Subheading 3.2.2 instead.
20. Excess Cy3/5 and HYNIC have larger retention volumes than
the HYNIC/Cy3/5-modified peptide. If the resolution of two
dye-colored bands is poor, use polyacrylamide gel
electrophoresis to determine which fractions contain the mod-
ified peptide. These fractions can then be combined, concen-
trated using the desalting protocol in Subheading 3.2.4, and
purified using a gel permeation column of greater length and/
or more optimal fractionation range.
21. The accuracy of the dye absorption coefficients when a peptide
is also modified with the HYNIC (or hydrazone) chromophore
is uncertain. Our experience suggests that the use of the Cy3/5
absorption coefficients underestimates the amount of peptide.
Figure 5 shows how the ratio between the hydrazone and
Cy3/5 absorption bands does not correspond to the ratio of
their reported absorption coefficients. In practice, we use the
hydrazone absorption band to quantitate the purified product
from Subheading 3.3.
22. We have found that each 0.4 mL plastic cartridge (e.g., empty
OPC cartridge; see Note 3) filled with Ni-NTA-agarose is suf-
ficient for each 0.5 mg of peptide (MW ≥1.5 kDa) used as
starting material in the labeling reaction.
23. Our recommended double-syringe technique is to connect a
3 mL syringe to one Luer fitting of the plastic/OPC cartridge
and another 3 mL syringe (without plunger) to the other Luer
fitting. The reaction mixture can be added to the latter and the
plunger carefully inserted into the syringe barrel. The oppos-
ing syringes can be pushed back and forth to flush the reaction
mixture over the Ni-NTA-agarose/OPC resin. The similar
diameters of the cartridge and 3 mL syringe minimize back-
pressure. This helps avoid leakage from any defects in the snap-
tight joint between the two pieces forming the cartridge or,
worse, breaking a seal at one of the Luer fittings with spillage
or spray of solution. If 3 mL syringes provide insufficient vol-
ume, then use of a syringe with the smallest possible diameter
for the needed volume is recommended.
24. We have noted that frits often nonspecifically adsorb

fluorescent dyes and other molecules with hydrophobic
character. This is largely ameliorated by the washing step
with the 50 % buffered solution of ethanol (or another polar
organic solvent, e.g., MeCN).
25. Washes can be done with syringe sizes that match the wash
volume with minimal concern for backpressure. It is advisable
to use air to expel residual solution from the Ni-NTA-agarose/
OPC resin between washes. This is particularly true on the
final washes so as to avoid diluting the imidazole/70 % MeCN
(aq) solutions used to elute peptide.
26. To help maximize the efficiency of peptide elution from the
media, the eluent can be gently and slowly pulsated by pushing
the syringe barrel 5 mm, pulling back 2–3 mm, and repeating
this cycle through the full volume of eluent.
27. If the number of moles of the peptide is determined by weight,
and the peptide has not gone through any desalting steps
(i.e., prepared via the SPPS method in Subheading 3.1), it is
important to take into account the mass of the TFA counter-ions
associated with basic residues. When ligating two peptides modi-
fied similarly those in Subheading 3.2, and at the same reaction
scales, it is most convenient to use the full amount of
HYNIC-Cy3/5 modified peptide prepared and half the amount
4FB-modified peptide (i.e., an approximately 2:1 stoichiometry).
28. Although the HYNIC-acetone hydrazone is stable in aqueous
buffer at pH 7.0, it rapidly equilibrates with the HYNIC form
in presence of aniline and reacts with 4FB yielding a more ther-
modynamically stable hydrazone.
29. Increasing the amount of aniline will increase the reaction rate.
The reaction will also proceed faster at moderately acidic pH
(e.g., NH4OAc buffer at pH 5.5). We typically use 100 mM
aniline with buffers at neutral pH (e.g., PBS buffer at pH 7),
and 10 mM aniline with acidic buffers; however, there is no
strict rule in this regard. Further, if it is inconvenient to mea-
sure ≤10 μL of aniline, 0.1–1.0 mL of aniline can be diluted to
10 mL with buffer and 0.1 mL added to the 1 mL reaction
mixture; however, these stock solutions should always be pre-
pared fresh.
30. We typically divide the total amount of peptide between several
microcentrifuge tubes with ca. 5–20 nmol of peptide per tube.
These aliquots are concentrated individually and allow the ligated
peptide to be dissolved and used in experiments only as needed.
31. The polyhistidine appended peptides will assemble to the
inorganic shell of QDs coated with thiol-based ligands
(e.g., DHLA, DHLA-PEG, and CL4). There is also evidence
that suggests these peptides will also assemble to the organic
corona of commercially available QDs with carboxylated polymer
(Invitrogen) or phospholipid (eBioscience, San Diego, CA)

coatings [50].
32. As always, the equation Vo = Vs(Cs/Co) = ns/Co is convenient to
use, where Vo and Co are the volume and concentration (respec-
tively) of the stock solution to be added to the sample, Vs is the
final sample volume, and ns is the number of moles per
sample.
33. In general, agarose gels should be run using negatively charged
QDs (e.g., a coating of DHLA ligands). A decrease in electro-
phoretic mobility with increases in the amount of assembled
peptide is typically expected. In the special case of highly
charged peptides, neutral QDs (e.g., a coating of DHLA-PEG
ligands) can be used and assembly is indicated by an increase in
mobility. In either case, and with appropriate optical filters for
the gel imager, it is often possible to observe colocalization of
the different wavelengths of fluorescence from the QD and dye
labeled peptide. This serves as additional confirmation of
assembly—particularly if the mobility shifts are small. Finally,
we note that the hydrazone bond sometimes degrades during
gel electrophoresis, and that the running time should be
minimized.
34. The details of a FRET analysis are beyond our current scope
but can be found in both a recent review [10] and a previous
method published in this series [51].
Acknowledgments
The authors acknowledge NRL, NRL NSI, ONR, and DTRA-

JSTO MIPR # B112582M for financial support. W.R.A. is grateful
to the Natural Sciences and Engineering Research Council of
Canada (NSERC) for support through a postdoctoral fellowship.
J.B.B.-C. acknowledges a Marie Curie IOF. RLM thanks National
Science Foundation Graduate Research Fellowship (Grant no.
DGE-0750616) for support.
References
1. Algar WR, Prasuhn DE, Stewart MH, Jennings 4. Algar WR, Susumu K, Delehanty JB, Medintz
TL, Blanco-Canosa JB, Dawson PE, Medintz IL (2011) Semiconductor quantum dots in
IL (2011) The controlled display of biomole- bioanalysis: crossing the valley of death. Anal
cules on nanoparticles: a challenge suited to Chem 83:8826–8837
bioorthogonal chemistry. Bioconjug Chem 5. Rosenthal SJ, Chang JC, Kovtun O, McBride
22:825–858 JR, Tomlinson ID (2011) Biocompatible
2. Rosi NL, Mirkin CA (2005) Nanostructures in quantum dots for biological applications.
biodiagnostics. Chem Rev 105:1547–1562 Chem Biol 18:10–24
3. Niemeyer CM (2001) Nanoparticles, proteins, 6. Algar WR, Krull UJ (2010) New opportunities
and nucleic acids: biotechnology meets in multiplexed bioanalyses using quantum dots
materials science. Angew Chem Int Ed 40: and donor-acceptor interactions. Anal Bioanal
4128–4158 Chem 398:2439–2449
7. Bruchez MP (2011) Quantum dots find their on quantum dots probed by capillary
stride in single molecule tracking. Curr Opin electrophoresis. Anal Chem 83:6323–6329
Chem Biol 15:775–780 20. Prasuhn DE, Deschamps JR, Susumu K,
8. Chan WCW, Maxwell DJ, Gao XH, Bailey RE, Stewart MH, Boeneman K, Blanco-Canosa JB,
Han MY, Nie SM (2002) Luminescent quan- Dawson PE, Medintz IL (2010) Polyvalent
tum dots for multiplexed biological detection display and packing of peptides and proteins
and imaging. Curr Opin Biotechnol 13: on semiconductor quantum dots: predicted
40–46 versus experimental results. Small 6:555–564
9. Algar WR, Tavares AJ, Krull UJ (2010) 21. Sawant RM, Hurley JP, Salmaso S, Kale A,
Beyond labels: a review of the application of Tolcheva E, Levchenko TS, Torchilin VP
quantum dots as integrated components of (2006) “SMART” drug delivery systems:
assays, bioprobes, and biosensors utilizing double-targeted pH-responsive pharmaceuti-
optical transduction. Anal Chim Acta 673: cal nanocarriers. Bioconjug Chem 17:
1–25 943–949
10. Medintz IL, Mattoussi H (2009) Quantum 22. Savla R, Taratula O, Garbuzenko O, Minko T
dot-based resonance energy transfer and its (2011) Tumor targeted quantum dot-mucin 1
growing application in biology. Phys Chem aptamer-doxorubicin conjugate for imaging
Chem Phys 11:17–45 and treatment of cancer. J Control Release
11. Biju V, Itoh T, Anas A, Sujith A, Ishikawa M 153:16–22
(2008) Semiconductor quantum dots and 23. Kam NWS, Liu Z, Dai HJ (2005)
metal nanoparticles: syntheses, optical proper- Functionalization of carbon nanotubes via
ties, and biological applications. Anal Bioanal cleavable disulfide bonds for efficient intracel-
Chem 391:2469–2495 lular delivery of siRNA and potent gene silenc-
12. Sapsford KE, Pons T, Medintz IL, Higashiya ing. J Am Chem Soc 127:12492–12493
S, Brunel FM, Dawson PE, Mattoussi H 24. Takae S, Miyata K, Oba M, Ishii T, Nishiyama
(2007) Kinetics of metal-affinity driven self- N, Itaka K, Yamasaki Y, Koyama H, Kataoka K
assembly between proteins or peptides and (2008) PEG-detachable polyplex micelles
CdSe-ZnS quantum dots. J Phys Chem C based on disulfide-linked block catiomers as
111:11528–11538 bioresponsive nonviral gene vectors. J Am
13. Pons T, Medintz IL, Wang X, English DS, Chem Soc 130:6001–6009
Mattoussi H (2006) Solution-phase single 25. Dirksen A, Dawson PE (2008) Rapid oxime
quantum dot fluorescence resonance energy and hydrazone ligations with aromatic alde-
transfer. J Am Chem Soc 128:15324–15331 hydes for biomolecular labeling. Bioconjug
14. Prasuhn DE, Blanco-Canosa JB, Vora GJ, Chem 19:2543–2548
Delehanty JB, Susumu K, Mei BC, Dawson PE, 26. Chen Y, Aveyard J, Wilson R (2004) Gold and
Medintz IL (2010) Combining chemoselective silver nanoparticles functionalized with known
ligation with polyhistidine-driven self-assembly numbers of oligonucleotides per particle for
for the modular display of biomolecules on DNA detection. Chem Commun 2804–2805
quantum dots. ACS Nano 4:267–278 27. Zhong XB, Reynolds R, Kidd JR, Kidd KK,
15. Medintz IL, Berti L, Pons T, Grimes AF, Jenison R, Marlar RA, Ward DC (2003)
English DS, Alessandrini A, Facci P, Mattoussi Single-nucleotide polymorphism genotyping
H (2007) A reactive peptidic linker for self- on optical thin-film biosensor chips. Proc Natl
assembling hybrid quantum dot-DNA biocon- Acad Sci USA 100:11559–11564
jugates. Nano Lett 7:1741–1748 28. Chang YS, Jeong JM, Lee YS, Kim HW, Rai
16. Blanco-Canosa JB, Medintz IL, Farrell D, GB, Lee SJ, Lee DS, Chung JK, Lee MC
Mattoussi H, Dawson PE (2010) Rapid cova- (2005) Preparation of F-18-human serum
lent ligation of fluorescent peptides to water albumin: a simple and efficient protein labeling
solubilized quatnum dots. J Am Chem Soc method with F-18 using a hydrazone-formation
132:10027–10033 method. Bioconjug Chem 16:1329–1333
17. Dennis AM, Bao G (2008) Quantum dot- 29. Steinberg-Tatman G, Huynh M, Barker D,
fluorescent protein pairs as novel fluorescence Zhao CF (2006) Synthetic modification of
resonance energy transfer probes. Nano Lett silica beads that allows for sequential attach-
8:1439–1445 ment of two oligonucleotides. Bioconjug
18. Zhang Y, So MK, Loening AM, Yao HQ, Chem 17:841–848
Gambhir SS, Rao JH (2006) HaloTag protein- 30. King PT, Zhao SW, Lam L (1986) Preparation
mediated site-specific conjugation of biolumi- of protein conjugates via intermolecular hydra-
nescent proteins to quantum dots. Angew zone linkage biochemistry. Biochemistry
Chem Int Ed 45:4936–4940 25:5774–5779
19. Wang JH, Xia J (2011) Preferential binding of 31. Gaertner HF, Rose K, Cotton R, Timms D,
a novel polyhistidine peptide dendrimer ligand Camble R, Offord RE (1992) Construction of
protein analogues by site-specific condensation 43. Dirksen A, Dirksen S, Hackeng TM, Dawson
of unprotected fragments. Bioconjug Chem PE (2006) Nucleophilic catalysis of hydrazone
3:262–268 formation and transamination: implications for
32. Rose K (1994) Facile synthesis of homoge- dynamic covalent chemistry. J Am Chem Soc
neous artificial proteins. J Am Chem Soc 128:15602–15603
116:30–33 44. Dirksen A, Hackeng TM, Dawson PE (2006)
33. Weden RC, Lankinen M, Rose K, Blakey D, Nucleophilic catalysis of oxime ligation. Angew
Shuttleworth H, Melton R, Offord RE Chem Int Ed 118:7743–7746
(1994) Site-specific conjugation of an enzyme 45. Sarin VK, Kent SB, Tam JP, Merrifield RB
and an antibody fragment. Bioconjug Chem (1981) Quantitative monitoring of solid-phase
5:411–417 peptide synthesis by the ninhydrin reaction.
34. Canne LE, Ferré-D’Amaré AR, Burley SK, Anal Biochem 117:147–157
Kent SBH (1995) Total chemical synthesis of a 46. Dabbousi BO, Rodriguez-Viejo J, Mikulec
unique transcription factor-related protein: FV, Heine JR, Mattoussi H, Ober R, Jensen
cMyc-Max. J Am Chem Soc 117:2998–3007 KF, Bawendi MG (1997) (CdSe)ZnS core-
35. Shao J, Tam JP (1995) Unprotected peptides shell quantum dots: synthesis and optical and
as building blocks for the synthesis of peptide structural characterization of a size series of
dendrimers with oxime, hydrazone, and thia- highly luminescent materials. J Phys Chem B
zolidine linkages. J Am Chem Soc 117: 101:9463–9475
3893–3899 47. Susumu K, Oh E, Delehanty JB, Blanco-Canosa
36. Cervigini SE, Dumy P, Mutter M (1996) JB, Johnson BJ, Jain V, Hervey WJ, Algar WR,
Synthesis of glycopeptides and lipopeptides by Boeneman K, Dawson PE, Medintz IL (2011)
chemoselective ligation. Angew Chem Int Ed Multifunctional compact zwitterionic ligands
35:1230–1232 for preparing robust biocompatible semicon-
37. Mahal LK, Yarema KJ, Bertozzi CR (1997) ductor quatnum dots and gold nanoparticles.
Engineering chemical reactivity on cell sur- J Am Chem Soc 133:9480–9496
faces through oligosaccharide biosynthesis. 48. Mei BC, Susumu K, Medintz IL, Mattoussi H
Science 276:1125–1128 (2009) Polyethylene glycol-based bidentate
38. Chen I, Howarth M, Lin W, Ting AY (2005) ligands to enhance quantum dot and gold
Site-specific labeling of cell surface proteins nanoparticle stability in biological media. Nat
with biophysical probes using biotin ligase. Protoc 4:412–423
Nat Methods 2:99–104 49. Surfraz MBU, King R, Mather SJ, Biagini
39. Singh Y, Renaudet O, Defrancq E, Dumy P SCG, Blower PJ (2007) Trifluoroacetyl-
(2005) Preparation of a multitopic HYNIC peptides: synthesis and 99mTc radio-
glycopeptide-oligonucleotide conjugate. Org labeling. J Med Chem 50:1418–1422
Lett 7:1359–1362 50. Dennis AM, Sotto DC, Mei BC, Medintz IL,
40. Larsen K, Thygesen MB, Guillaumie F, Willats Mattoussi H, Bao G (2010) Surface ligand
WGT, Jensen K (2006) Solid-phase chemical effects on metal-affinity coordination to quan-
tools for glycobiology. Carbohydr Res tum dots: implications for nanoprobe self-
341:1209–1234 assembly. Bioconjug Chem 21:1160–1170
41. Scheck RA, Francis MB (2007) Regioselective 51. Prasuhn DE, Susumu K, Medintz IL (2011)
labeling of antibodies through N-terminal Multivalent conjugation of peptides, proteins,
transamination. ACS Chem Biol 4:247–251 and DNA to semiconductor quantum dots. In:
42. Carrico IS, Carlson LB, Bertozzi CR (2007) Hurst SJ (ed) Biomedical nanotechnology:
Introducing genetically encoded aldehydes methods and protocols. Humana Press-
into proteins. Nat Chem Biol 3:321–322 Springer, New York, pp 95–110
Chapter 6
Reliable Methods for Silica Coating of Au Nanoparticles

Isabel Pastoriza-Santos and Luis M. Liz-Marzán
Abstract
The inherent properties of silica, such as optical transparency, high biocompatibility, chemical and colloidal
stability, controllable porosity, and easy surface modification, provide silica materials with a tremendous
potential in biomedicine. Therefore, the coating of Au nanoparticles with silica largely contributes to
enhance the important applications of metal nanoparticles in biomedicine. We describe in this chapter a
number of reliable strategies that have been reported for silica coating of different types of Au nanoparti-
cles. All descriptions are based on tested protocols and are expected to provide a reference for scientists
with an interest in this field.
Key words Au nanoparticles, Silica coating, Core-shell, Surface modification, Optical properties
1 Introduction
The intense research and the great progress that have been achieved
over the past two decades in the synthesis, characterization and
surface modification of gold nanoparticles have demonstrated that
their properties go beyond their beauty [1]. The unique electronic
properties, different from their bulk counterpart, arise from the
collective oscillation of conduction electrons in resonance with an
incident electromagnetic radiation [2, 3]. The phenomenon of
localized surface plasmon resonance (LSPR) constitutes the basis
of numerous and important applications of metal nanoparticles in
a wide range of fields (biomedicine, optoelectronics, photovoltaics,
(bio)sensing, catalysis, etc.) [1]. Importantly, LSPR frequency is
strongly dependent on a number of factors, including particle size,
shape and composition, as well as the dielectric properties of the
surrounding medium and interparticle distance [4, 5].
In the fields of biology and medicine, Au nanoparticles present
a wide range of potential applications, derived not only from their
optical properties (enhanced radiative properties) but also from
their low inherent toxicity, large surface area, and easy surface
75
76 Isabel Pastoriza-Santos and Luis M. Liz-Marzán
functionalization (which is important in terms of biological

compatibility and specificity). Murphy, El-Sayed, and others have
recently published a critical review entitled “The golden age: gold
nanoparticles for biomedicine” [6]. In this review article, apart from
summarizing the most promising achievements in the field, they
express their belief in that “a new Golden Age of biomedical
applications is truly upon us.” Gold nanoparticles have been used to
fabricate highly sensitive assays for in vitro diagnostics [7, 8], devices
that are mainly based on surface-enhanced Raman scattering (SERS)
[9] and LSPR [10] (the shift in the extinction spectra or even simply
a color change, in response to changes in the local refractive index
surrounding the nanoparticle provides a direct readout of target
binding) assays. Their strongly enhanced radiative properties also
render Au nanoparticles useful probes for bioimaging, either in vitro
or in vivo [6, 11]. Moreover, they are promising scaffolds for gene
and drug delivery applications—either as delivery platforms or as
stabilizers—or even to trigger drug release [6, 10, 12]. Even more
fascinating is the fact that gold colloids themselves can act as thera-
peutic agents. The absorption of light when the nanoparticle is irra-
diated with a laser beam produces local heating that has been proven
capable of killing cancer cells, viruses, etc. [6]. The immediate future
for Au nanoparticles in biomedicine is very promising and one of the
most important challenges will be assessing the toxicity and long-
term effects on humans and the environment, as well as the develop-
ment of new strategies for nanoparticles surface functionalization.
Another interesting material for biomedicine is silica. Apart
from exhibiting optical transparency, high biocompatibility and
strong chemical and colloidal stability, silica particles can be easily
labeled (e.g., via encapsulation of molecular tags) and surface modi-
fied (with amino-, mercapto-, carboxy-, aldehyde-terminated silanes,
polyelectrolytes, etc.), while their porosity is highly tunable (useful
for drug loading and delivery) [13]. Additionally, the coating of
metal nanoparticles with silica has been demonstrated not just to
facilitate bioconjugation of such Au nanoparticles but also to increase
their biocompatibility or to fabricate multifunctional systems. Thus,
silica-coated Au nanoparticles (Au@SiO2) show a great potential for
designing colorimetric [14] and SERS biosensors, for in vivo and in
vitro bioimaging [15–17], for drug delivery [18], in therapy [19,
20], or in theranostics [21]. In this chapter we describe in detail
various protocols that have been successfully applied for coating Au
nanoparticles with silica.
2 Materials
Some rules of general application should be indicated. Colloid

synthesis in general is extremely sensitive to the presence of potential
nuclei in the reaction medium and therefore all glassware and solvents
must be extremely clean. In the case of metal nanoparticles in general
Silica Coated Au Nanoparticles 77
(and gold in particular), it is standard to require that all glassware is

thoroughly washed with aqua regia (see Note 1) prior to use and
Milli-Q grade water should be used in all preparations. Although all
chemicals can be used as received, in some cases the quality of the
reagents or even the specific s upplier may also critically affect the qual-
ity of the resulting nanoparticles. Therefore, those critical reagents
have been listed below.
2.1 Reagents 1. Cetyltrimethylammonium bromide (CTAB) from Aldrich.

2. Ammonium hydroxide (NH4OH), 28–32 wt.% aqueous
solution.
3. Pure grade ethanol.
4. Silver nitrate (AgNO3), purity higher than 99 %.
5. For the synthesis of Au nanostars, poly(vinylpyrrolidone) (PVP,
Mw 8000) from Alfa Aesar is recommended. PVP from Sigma-
Aldrich has been found to lead to particles with poor quality.
6. “Active silica” or 0.54 wt.% sodium silicate solution with a pH
of 10–11: 1 mL of sodium silicate 27 wt.% is diluted 50 times
with water and then the pH is adjusted to 10–11 by addition
of a cation exchange resin, Dowex Macroporous Resin, for
some minutes. Check the pH evolution with pH test strips.
3 Methods
3.1 Silica Coating This method, developed by Liz-Marzán et al., was probably the first
of 15 nm Citrate- successful coating of gold nanoparticles with uniform silica shells.
Stabilized Au Particles Published in Langmuir in 1996 [22], this article has been widely
used, as indicated by over 1000 citations, according to ISI Web of
Knowledge©. The essence of this method (see scheme in Fig. 1a) is
the modification of the metal cores with an amino-terminated silane
coupling agent, such as 3-(aminopropyl) triethoxysilane (APS,
see Note 2), which acts as a primer to facilitate uniform silica polym-
erization around the Au particles. Subsequently, the silica shell can
be further grown in a controlled way through the well-known Stobër
method [23]. This protocol allows a tight control on the silica shell
thickness from 3 nm up to hundreds of nanometers.
We describe here a protocol that has been optimized for the
silica coating of spherical Au nanoparticles with an average diameter
of 17 nm but which could be adapted for the coating of other
citrate-stabilized nanoparticles.
Gold nanospheres synthesis (see Note 3) [24]
1. Bring to a full rolling boil 500 mL of 0.5 mM HAuCl4 in a 1 L
Erlenmeyer flask under homogeneous and vigorous stirring
(see Note 4).
2. Add rapidly 25 mL of an aqueous 1 wt.% aqueous solution of
trisodium citrate dihydrate (0.25 g) (see Note 5). Maintain the
Fig. 1 (a) Scheme of the strategy for silica coating citrate-stabilized Au nanoparticles. (b) UV–Vis spectra of
15 nm Au-citrate nanoparticles during the different coating steps. The position of the LSPR band maximum is
indicated in the graph. (c, d) TEM images of citrate-stabilized Au particles before and after silica coating,
respectively. Magnification is the same for both images
boiling for 15 min and then slowly cool down to room

temperature (see Note 6).
3. Measure the absorption spectrum by UV–Visible spectroscopy
(see Notes 7 and 8) (Fig. 1).
Silica coating
1. Add 2.5 mL of a freshly prepared 1 mM aqueous APS solution
(see Note 9) to 500 mL of the gold colloid under vigorous
magnetic stirring. Allow the reaction to proceed for 15 min
(see Note 10) (Fig. 1b).
2. Add 20 mL of “active silica” under vigorous magnetic stirring,
the resulting pH should be around 8.5. Allow active silica to
polymerize onto the Au nanoparticles for 7 days (see Notes 11
and 12) (Fig. 1b).
3. Centrifuge the resulting silica-coated Au nanoparticles at
1,287 × g for 90 min to remove excess silicate. Collect the pre-
cipitate and centrifuge the supernatant twice under the same
conditions. Collect all precipitates and redisperse in 50 mL of

Milli-Q water.
4. Add 200 mL of ethanol to 50 mL of Au@SiO2 particles in
water (see Note 13).
5. Add 3.0 mL of TEOS solution (10 % in ethanol, see Note 14)
and 2 mL of ammonia (NH4OH, 32 % in water) all under gentle
stirring. Keep under gentle stirring for 12 h before adding again
3.0 mL of TEOS solution (see Notes 15 and 16).
6. Centrifuge the resulting silica-coated Au nanoparticles at
1,287 × g. Collect the precipitate and centrifuge the supernatant
twice under the same conditions. Collect all precipitates and
redisperse in Milli-Q water or ethanol to store (Fig. 1d).
3.2 Mesoporous This method was developed by Gorelikov and Matsuura and
Silica Coating of published in Nano Letters in 2008 [25]. Since its publication this
CTAB-Stabilized Au article has been cited over 100 times so far, according to Web of
Nanoparticles Knowledge©. Probably the main advantages of this method are
related to its simplicity, since it is a single-step synthesis procedure
(see scheme in Fig. 2a), with no need for any intermediate coating
or removal of the surfactant bilayer, and to the high porosity of
the obtained shell, which facilitates the interaction between the
metal core and the surrounding medium. Although the protocol
works very well for the deposition of thick silica shells, it fails to
control very thin shells. It is also worth mentioning that the meso-
porous silica shell may dissolve in basic aqueous solutions, but this
can be avoided through a mild thermal treatment [26]. We
describe here the protocol for coating Au nanorods (14 × 70 nm),
but the same procedure can be adapted for other CTAB-stabilized
Au nanoparticles.
Au nanorod synthesis (in acidic medium) (see Note 17) [27]
1. Add 0.3 mL of freshly prepared 10 mM sodium borohydride
solution (see Note 18) (fast addition under vigorous stirring)
to 5 mL of an aqueous solution containing 0.1 M CTAB
and 0.25 mM HAuCl4. Let the reaction proceed for 10 min
(see Note 19).
2. Add 2.5 mL of 0.05 M HAuCl4 to 250 mL of 0.1 M CTAB.
Homogenize the mixture by shaking (see Note 20).
3. Add 4.75 mL of 1 M HCl, 2.0 mL of 0.1 M ascorbic acid
(see Note 21), 3 mL of freshly prepared 0.01 M silver nitrate
and 0.6 mL of seed solution (after each addition the solution
should be gently shaked).
4. Introduce the resulting solution in a thermostatic bath at
27 °C for 2 h.
Fig. 2 (a) Scheme of the coating of CTAB-stabilized Au nanoparticles with mesoporous silica. (b) Vis–NIR spectra
of 2–3 nm Au seeds (dotted line) and Au nanorods before (solid line) and after (dashed line) silica-coating.
(c) TEM image of silica-coated Au nanorods. The scale bar represents 50 nm
Silica coating
1. Centrifuge 250 mL of 0.5 mM Au nanorods at 2,627 × g for
40 min in 10 mL tubes. Extract the supernatant, leaving ca.
1.2–1.4 mL in each tube (see Note 23).
2. Collect all particles and redisperse in water (final volume
100 mL).
3. Adjust the pH to 10 by adding 1 mL of 0.1 M NaOH.
4. Perform 3 additions of 300 μL TEOS (20 % in methanol,
see Note 14). Allow 30 min after each addition.
5. After 2 h of the third addition check the UV–Vis–NIR spectrum
of the dispersion (see Notes 12 and 24).
6. Wash Au@SiO2 samples by centrifugation and redispersion in
water, first at 1,681 × g for 30 min and then at 1,287 × g for
45 min.
7. Redisperse the colloids in ethanol and store in the fridge
(see Note 25) (Fig 2c).
3.3 PVP-Mediated Poly(vinylpyrrolidone) (PVP) is a widely used particle stabilizer and

Silica Coating can also be used as a primer to promote silica coating of nanoparticles.
Back in 2003, Graf et al. reported a general procedure [28] that has
received wide attention (over 400 citations according to Web of
Knowledge©). This method is based on the ability of PVP to adsorb
onto a variety of surfaces (metals, oxides, hydroxides, polymers, etc.,
regardless of their surface charge). Since PVP is commonly used as
capping agent during the synthesis of metal nanoparticles, not only to
prevent aggregation but also as a shape-directing agent [29, 30]
Stöber silica growth can be directly carried out on such particles, with
no need for any intermediate step (see scheme in Fig. 3) [31, 32].
However, two main considerations should be taken into account to
achieve successful coating, namely, PVP concentration (excess or
defect of PVP onto the particles surface can lead to inhomogeneous
coating, meaning that optimization of washing is often required) and
PVP molecular mass (the choice of an appropriate molecular mass
compared to particle size is crucial to avoid inhomogeneous coatings
or aggregation, see Fig. 3).
Although plenty of examples have been described in the litera-
ture, we focus here on the silica coating of Au nanostars, which
have become very popular because of their extraordinary morphology
and associated properties [33].
Fig. 3 Top: Scheme of the silica coating of citrate-stabilized Au nanoparticles using PVP. Bottom: TEM images
of silica-coated Au nanospheres obtained using PVP with different molecular weights, as indicated. Adapted
with permission from ref. 28. Copyright (2003) American Chemical Society
Synthesis of Au nanostars (see Note 26) [34, 35]

1. Prepare citrate-stabilized 15 nm Au nanospheres using the
protocol described above (see Subheading 3.1).
2. Add 5 mL of an aqueous solution containing 0.1 g of PVP,
previously sonicated for 15 min, to 50 mL of the as-prepared
Au nanoparticles (CAu = 0.5 mM). Allow to react overnight.
3. Centrifuge the colloid at 5,150 × g for 90 min. Collect the
supernatant and centrifuge again at 5,150 × g until it becomes
colorless.
4. Redisperse the resulting PVP protected particles in 5 mL of
ethanol.
5. Calculate the Au concentration (see Notes 27 and 28)
6. Add 75 μL of 0.1 M HAuCl4 aqueous solution to 15 mL of
12.5 mM PVP solution in DMF.
7. After ca. 2 min (see Note 29), add 171 μL of 2.1 mM Au seed
solution (in ethanol, see Note 30) under continuous stirring.
Allow to react for 30 min.
8. Check the LSPR band by UV–Visible–NIR absorption spectros-
copy (see Note 31).
Silica coating
1. Centrifuge threefold the Au nanostars dispersion in 10 mL
tubes for 60 min at 2,128 × g (the first washing should be per-
formed at 3,179 × g) to remove excess PVP. Extract the super-
natant and redisperse in 2 mL isopropanol (ethanol, however,
is preferred during the washings).
2. Calculate the concentration of Au nanostars (see Notes 28
and 32).
3. Prepare 8.25 mL of 0.85 mM Au nanostars (see Note 33).
4. Add 1.44 mL of Milli-Q water, 0.106 mL of NH4OH (30 % in
water), and 0.204 mL of TEOS (5 vol.% of TEOS in isopropanol,
see Note 14) under gentle stirring. Allow to react for 2 h under
stirring (see Note 34).
5. Wash Au@SiO2 samples by twofold centrifugation and redis-
persion in ethanol, at 1,681 × g until supernatant becomes
colorless.
(see Note 34).
3.4 Polyelectrolyte- This method was reported by Pastoriza-Santos et al. and published
Mediated Silica in Chemistry of Materials in 2006 [36], currently counting 150
Coating citations, according to Web of Knowledge©. The method was
devised for the deposition of dense silica layers on CTAB-capped
gold nanorods and comprises an initial surface modification step
based on the polyelectrolyte layer-by-layer (LBL) assembly
Fig. 4 Scheme of the silica coating of Au nanoparticles through the layer-by-layer approach [36]. The particles are
first wrapped with polyelectrolytes of different charge and finally with PVP, which then allows Stöber silica coating
technique [37, 38], followed by Stobër silica coating (see scheme in

Fig. 4) [23]. Although this protocol may become rather laborious
and time consuming because of the different coating steps required,
the fact that it is based on the LBL protocol makes it also rather
versatile. The LBL technique is a powerful tool that allows us to
modify the surface chemistry of almost any kind of hydrophilic
nanoparticles. Therefore, the coating strategy may be readily
extended to other types of nanoparticles. Other advantages are the
tight control on the silica shell thickness and the stability of the
shells in various dispersing media.
Next, we describe the protocol for application on CTAB-
stabilized Au nanorods (ca. 13 × 50 nm).
Au nanorod synthesis (see Note 35) [39]
1. Prepare 2–3 nm Au seeds using the above detailed protocol
(see Subheading 3.2, Au nanorod synthesis (in acidic medium),
step 1).
2. Add 2.5 mL of 0.05 M HAuCl4 to 250 mL of 0.1 M CTAB.
Homogenize the mixture by shaking (see Note 20).
3. Add 1.87 mL of 0.1 M ascorbic acid (see Note 21), 2 mL of
freshly prepared 5.0 mM silver nitrate and 3.0 mL of seed solution
(after each addition the solution should be gently shaken).
4. Introduce the resulting solution in a thermostatic bath at
27 °C for 30 min (see Note 36).
5. Centrifuge at 6,726 × g for 30 min for partial removal of excess
CTAB. Discard the supernatant and redisperse the precipitate in
125 mL of Milli-Q water (see Note 37).
LBL surface modification
1. Add dropwise 5 mL of a 0.5 mM Au nanorod hydrosol under
vigorous stirring to 5 mL of a 2 g/L aqueous solution of
poly(styrene sulfonate) sodium salt (PSS, MW 14,900), which
also contains 6 mM NaCl (see Note 38). Stir the mixture for 3 h.
2. Centrifuge the resulting PSS-coated nanorods twice at 2,128 × g

for 30 min to remove excess polyelectrolyte and redisperse in
5 mL of Milli-Q water.
3. Measure the Zeta Potential by Photon Correlation Spectroscopy
and the UV–Vis–NIR absorption spectrum (see Note 39).
4. Add dropwise the nanorod solution under mild stirring to 5 mL
of a 2 g/L poly(allylamine hydrochloride) (PAH, MW 15,000)
aqueous solution (6 mM NaCl) (see Note 38). Stir the mixture
for 3 h.
5. Centrifuge the resulting PAH-coated nanorods twice at 2,128 × g
for 30 min to remove excess polyelectrolyte and redisperse in
5 mL of Milli-Q water.
6. Measure the Zeta Potential by Photon Correlation Spectroscopy
and the UV–Vis–NIR absorption spectrum (see Note 40).
7. Mix 5 mL of PAH-coated gold nanorods with 5 mL of an
aqueous solution of PVP (MW 10,000, 4 g/L) and stir over-
night (see Note 38).
8. Centrifuge the mixture at 2,128 × g for 30 min, discard the
clear supernatant, and redisperse the precipitate in 0.2 mL of
water (see Note 41).
9. Add 1 mL of isopropanol (see Note 41).
10. Calculate the concentration of Au nanorods (see Notes 28
and 42).
Stöber silica coating [23]
1. Add 1.2 mL of PVP-coated gold nanorod dispersion under
vigorous stirring to 0.46 mL of water.
2. Add 1.43 mL of ammonia solution in isopropanol (3.84 vol.%)
under vigorous stirring and 0.40 mL of a TEOS solution in iso-
propanol (0.97 vol.%, see Note 14) under gentle stirring. Allow to
react for 2 h under stirring (see Note 43).
3. Wash Au@SiO2 samples by twofold centrifugation and redis-
persion in ethanol, at 1,681 × g until the supernatant becomes
colorless.
4. Redisperse the colloids in ethanol and store in the fridge.
3.5 PEG-Mediated This method was developed by Fernández-López et al. in 2009

Silica Coating and published in Langmuir [40]. Since its publication this article
has been cited over 35 times according to Web of Knowledge©.
The method relies on the high colloidal stability of polyethylene
glycol (PEG) capped gold nanoparticles, both in water and etha-
nol. Thus, a thiolated PEG (O-[2-(3-mercaptopropionylamino)
ethyl] O′-methyl-poly(ethylene glycol) (m-PEG-SH)) is used to
replace the original surface capping agents, such as citrate or
CTAB, and to subsequently facilitate the growth of uniform silica
Fig. 5 (a) Scheme of the silica coating of CTAB-stabilized Au nanoparticles using thiolated PEG. (b) Vis–NIR
absorption spectra of the Au nanospheres before (solid line) and after (dashed line) silica-coating. (c) TEM
image of the silica-coated Au nanospheres. The scale bar represents 100 nm
shells around the particles via Stobër method [23] (see Fig. 5a).
This method allows an accurate control of the silica shell thickness
and the obtained coatings are highly uniform, even when very thin
shells (few nanometers) are grown. It should be noted however
that, although the silica shell is not mesoporous in this case, it can
readily dissolve in basic aqueous solutions, which can be prevented
through a mild thermal treatment [26].
As a representative example, we describe here the specific proto-
col for coating Au-CTAB spheres (ca. 60 nm), which can be easily
adapted to other CTAB-stabilized Au nanoparticles.
Synthesis of 60 nm Au nanospheres [41]
1. Prepare 15 nm citrate-stabilized Au nanospheres (using the
protocol described above (see Subheading 3.1).
2. Warm 500 mL of a solution containing HAuCl4 (0.5 mM) and
CTAB (0.015 M) up to 35 °C.
3. Add 5 mL of 0.1 M ascorbic (see Note 21) and finally 7.94 mL
of Au seeds (ca. 0.5 mM) under mild stirring (see Note 45).
5. Centrifuge the as-synthesized particles at 1,681 × g for 20 min

and after careful removal of the supernatant, redisperse the
colloids in 10 mL of a 0.1 M CTAB aqueous solution.
6. Transfer into a glass test tube (10 mL) and introduce it in a
beaker containing hot water (50 °C) for 15 min. Keep it in an
upright position overnight.
7. Separate the supernatant, containing the Au spheres, from the
precipitate containing nanorods and plates. Redisperse the Au
spheres in 500 mL of 0.1 M CTAB.
mPEG-SH functionalization, ethanol transfer and silica coating
1. Centrifuge 20 mL of the Au colloid at 657 × g for 20 min and
redisperse in 20 mL of water (see Note 47).
2. Centrifuge for 10 min at 657 × g and then centrifuge the super-
natant for another 15 min at 420 × g. Collect all p
recipitates and
redisperse in water to a final volume of 10 mL (see Note 48).
3. Add dropwise 0.3 mL of 0.25 mM m-PEG-SH (Mw 5000,
see Notes 38 and 49) aqueous solution, under vigorous stirring.
4. Centrifuge the resulting PEG-modified particles twice (657 × g,
20 min) to remove excess PEG and redisperse in 2 mL of
ethanol.
5. Calculate the concentration of PEG-modified Au nanoparticles
(see Notes 28 and 50).
6. Prepare 7.8 mL of 0.64 mM Au spheres (see Note 51).
7. Add 1.9 mL of Milli-Q water, 0.125 mL of NH4OH (32 % in
water), and 0.177 mL of TEOS (1 vol.% of TEOS in isopropa-
nol, see Note 14) under gentle stirring. Allow to react for 2 h
under stirring (see Note 52) (Fig. 6.5).
8. Wash Au@SiO2 samples by twofold centrifugation and redisper-
sion in ethanol, at 1,681 × g until the supernatant becomes
colorless.
(see Fig. 5c).
4 Notes
1. Glassware cleaning with aqua regia. This is an important step

that should be carried out before starting the synthesis, since
residues in the glassware may interfere with the synthesis. All
glassware should be first washed with detergent and then with
aqua regia, followed by extensive rinsing with water and
Milli-Q water. Preparation of aqua regia: In a glass beaker, mix
hydrochloric acid (HCl) and nitric acid (HNO3) in a volume
ratio 3:1. The preparation should take place inside a fume

hood, since the liquid and its vapors are extremely corrosive
and highly oxidizing. Therefore, the use of appropriate gloves,
lab coat, and chemical splash goggles is mandatory.
2. The use of mercaptopropyltrimethoxysilane (MPS) has also been
reported by several authors.
3. Citrate stabilized Au spheres (ca. 0.5 mM, gold atomic concentra-
tion) were synthesized according to Turkevich’s method [24].
4. It is important to select appropriate magnetic stirring condi-
tions that allow uniform boiling.
5. This addition is probably the most critical step of this synthesis. It
should be carried out as fast as possible and it is also convenient
to warm up the citrate solution prior to addition, to avoid deten-
tion of the boiling. The citrate ions reduce Au(III) and Au
nanoparticles are formed. Visually striking color changes are
observed during the reaction: the pale yellow HAuCl4 solution
turns colorless upon citrate addition, suddenly changing to a very
dark blue (even black) and slowly turning deep red.
6. It is advisable to wait 24 h before silica coating is started.
7. Alkali metals, Mg, Al, and noble metals such as Cu, Ag, and Au
present localized surface plasmon resonance (LSPR) bands.
The collective response of conduction band electrons to light
produces a characteristic resonance band that lies in the UV–
Visible–NIR range for Cu, Au, and Ag. (For more information
see refs. [3, 5]).
8. The UV–Vis spectrum should present an LSPR band around
518–520 nm (see Fig. 6.1b). The particle size should be around
15–18 nm in diameter (see a typical transmission electron
microscopy (TEM) image in Fig. 1c).
9. The first step comprises priming the gold surface with APS.
Molecules with amine groups, as well as with mercapto groups
complex strongly to Au metal.
10. No changes in the LSPR band should be observed (see Fig. 1b).
11. After 7 days the Au particles should be coated with a ca. 3–4 nm
thick silica shell. A 5 nm LSPR red-shift is typically observed
(see Fig. 6.1b and Note 12).
12. The growth of the silica shell produces the increase of the
effective refractive index of the particles surrounding medium
and this originates the observed plasmon band shift toward
longer wavelengths [3].
13. Ethanol addition produces silicate condensation and a further
3 nm LSPR shift, (see Fig. 1b and Note 12).
14. Better results are obtained when TEOS is diluted in alcohol
before the addition.
15. The total amount to be added can be easily calculated taking

into account the Au nanoparticle size (ca. 17 nm for this case)
and the final particle size. TEOS total amount (VTEOS) is calcu-
VTEOSVAu  R  3 
lated using the formula VTEOS = CAu  tot  − 1 where
VSiO2  RAu  
Rtot, RAu, CAu and V are final radius, Au radius, and molar
volume, respectively. Molar volume is calculated applying the

M
relation V = , M being the molar mass and ρ the density.
r
16. A typical TEM image of Au@SiO2 nanoparticles (after 4 TEOS
additions) is shown in Fig. 1d. The silica coating process can be
monitored by UV–Vis spectroscopy, since the growth of the
shell leads to a gradual red-shift of the LSPR band (see Fig. 1b
and Note 12).
17. This synthesis is based on a seeded growth method reported by
Liu and Guyot-Sionnest [27], with some modifications. This
method gives rise to Au nanorods with aspect ratio (length
divided by the width) higher than 4. This process is scalable up
to 250 mL but the gold nanorods synthesis is very sensitive to
the quality of the preformed seeds. It is recommended to test
the synthesis in 10 mL preparations prior to scale-up. The
solvent in all the preparations is Milli-Q water and all CTAB
based solutions were stored at 27–30 °C to avoid surfactant
crystallization.
18. Sodium borohydride in water decomposes spontaneously into
hydrogen gas and borates, therefore the solution should be
prepared just before using and stored in an ice/water bath.
19. Since borohydride is a strong reducing agent, the reaction is
completed in 10 min and the seeds are ready for further use
within 24 h after preparation. The average particle size mea-
sured from TEM should be around 2–3 nm and the solution
typically displays a yellow-brown color due to the small particle
size (UV–Vis spectrum should not display a well-defined band
or even a shoulder, which would both indicate the presence of
larger particles, see Fig. 2b).
20. Magnetic stirring is not appropriate for the growth of Au
nanorods, therefore the homogenization is obtained by hand
shaking. The solution becomes deep yellow as a consequence of
the formation of a complex between HAuCl4 and CTAB.
21. The solution becomes colorless.
22. The final gold nanorod dimensions should be ca. 14 × 70 nm
(aspect ratio 5) and the UV–Vis–NIR spectrum should display
two LSPR bands, a very intense one located around 850 nm
and a second one, much less intense, at ca. 525 nm (see Fig. 2b).
23. The CTAB concentration present in the Au nanorods solution

is critical during silica coating, which requires careful washing
of the nanoparticles to remove part of the CTAB in excess.
24. 40–50 nm red-shift of the main LSPR band should be observed.
If the shift is not observed then the reaction should be allowed
to proceed for 3 more hours.
25. A TEM image of the silica coated particles is shown in Fig. 2c,
where the mesoporous nature of the coating is evidenced.
26. This synthesis is based on the N,N-dimethylformamide (DMF)
reduction method, developed by Pastoriza-Santos, Liz-Marzán
et al. [34, 35] Au nanostars are formed through the reduction
of an Au salt in a solution of PVP (Mw 8000, from Alfa Aesar)
in DMF, onto preexisting gold nanoparticles. Although DMF
is the main reducing agent, PVP was also found to contribute
to Au salt reduction. In our example, standard Au-citrate
nanoparticles were used as seeds.
27. Au concentration should be ca. 4.2 mM (see Note 28).
28. Gold concentration is often measured from the value of
absorbance at 400 nm, taking into account that absorbance is
proportional to atomic Au concentration (Lambert-Beer Law).
An absorbance of 1.2 corresponds to 0.5 mM concentration
(cell path length of 1 cm).
29. It would be convenient to monitor the evolution (damping) of
the absorption band characteristic of Au3+ at 325 nm by UV–
Vis–NIR spectroscopy. DMF progressively reduces Au3+ to
Au+, leading to complete damping of the extinction band
within 2 min (although the damping time could be dependent
of the experimental conditions).
30. 2.1 mM Au seeds is obtained from a twofold dilution of the
4.2 mM Au colloids.
31. The UV–Vis–NIR spectrum should display two LSPR bands, one
very intense around 800 nm and a second one, less intense, at ca.
550 nm, characteristic of tip and core plasmon modes, respec-
tively. The overall particle diameter (including the tips) should be
around 60–70 nm.
33. Sample obtained by diluting the 3.75 mM Au nanostars colloids.
34. In 2 h particles should present a thin silica layer (ca. 5 nm). The
shell can be further grown upon controlled TEOS addition. As in
all other procedures, silica coating can be monitored by UV–Vis
spectroscopy, following LSPR red-shift (see Note 12). See Fig. 1b
as an example.
35. This synthesis is based on a seeded growth method developed
by Nikoobakht and El-Sayed [39], with some modifications.
The solvent was Milli-Q water in all preparations and all CTAB
based solutions were maintained at 30 °C to avoid surfactant

crystallization. The synthesis of Au nanorods in non- acid
medium gives rise to Au nanorods with aspect ratio below 4.
This process is scalable up to 250 mL but gold nanorods syn-
thesis is very sensitive to the quality of the preformed seeds. It
is recommended to test the synthesis in 10 mL preparations
prior to scale-up.
36. The gold nanorod dimensions should be ca. 13 × 52 nm (aspect
ratio 4) and the corresponding UV–Vis–NIR spectrum should
display two LSPR bands, one very intense centered around
750 nm and a second one, much less intense, at ca. 525 nm.
37. The zeta potential of the particles should be around +20 mV.
38. PSS, PAH, PVP, and PEG aqueous solutions should be sonicated
for 30 min prior to use.
39. The zeta potential should be ca. −40 mV, and the Vis–NIR
spectrum should not display significant changes.
40. PAH-coated particles display a zeta potential of ca. +45 mV,
and again significant changes in the optical spectrum are not
expected.
41. PVP-coated particles display a zeta potential of ca. +20 mV. Since
PVP degrades in water, it is convenient to store the stock particle
solution in alcohol (not in pure alcohol but in an alcohol–water
mixture). Redispersion of those nanoparticles in pure alcohol
(usually ethanol or isopropanol) could easily lead to particles
aggregation (seen by color change).
42. The gold concentration at this stage should be around
2.47 mM (see Note 28).
43. The gold nanorods should be surrounded by a 12 nm thick
silica shell. The shell can be further grown upon controlled
TEOS addition. As in all other procedures, silica coating can
be monitored by UV–Vis spectroscopy, following LSPR red-
shift (see Note 12). See Fig. 1b as an example.
44. This synthesis is based on a seeded growth method reported by
Rodríguez-Fernández et al. [41], with slight modifications.
45. Au nanospheres with sizes ranging from 15 to 300 nm can be
readily prepared using this protocol just by varying the amount
of seeds. The amount of seeds required to obtain nanoparticles
with radius RPart might be easily calculated using the following
1/ 3
C + CAuseed 
formula RPart = Rseed  Ausalt  where CAusalt is the gold
 CAuseed
salt concentration and Rseed and CAuseed are the radius and concen-
tration of the seeds, respectively. To synthesize large particles
(≥90 nm) it is highly recommended to carry out the growth
process in two steps (grow from 15 to 60 nm and from 60 to the
final size).
46. Growth of the 15 nm seeds produces a shift of the plasmon

band from 520 nm (LSPR of the seeds) to 536 nm (see Fig. 5b)
and the absorbance at 400 nm should ca. 1.2 (1 cm optical
path length). Apart from Au spheres, particles with other
morphologies (rods and plates mainly) are obtained, but the
nonspherical particles can be efficiently removed through
CTAB-assisted shape selective separation [42].
47. The CTAB concentration (CCTAB), is around 15 mM.
48. At this stage CCTAB ~1 mM and CAu ~1 mM.
49. The amount of m-PEG-SH is that required to provide 4 mole-
cules per nm2 of Au nanoparticle surface. This is the minimum
amount to obtain homogeneous coatings; no significant effects
on the coating were observed when higher concentrations were
used. In the case of non-purified particles the minimum amount
was calculated considering 15 molecules per nm2.
51. Sample obtained by diluting the 4.2 mM Au colloids.
52. The silica shell thickness should be around 7–10 nm but can be
further grown upon controlled TEOS addition. Figure 5c
shows a typical TEM image of Au@silica particles coated using
PEG-mediated strategy. The progress of silica coating can be
followed by UV–Vis spectroscopy, since the growth of the shell
leads to a red-shift of the LSPR band (see Note 12) (Fig. 5b).
Acknowledgments
This work has been funded by the Spanish Xunta de Galicia/

FEDER (grant 10 PXIB 314 218 PR) and MINECO/FEDER
(grant MAT2010-15374). LML-M acknowledges the ERC for an
Advanced grant (NANODIRECT, grant number CP-FP 213948–2).
The authors would like to thank Sara Abalde- Cela, Cristina
Fernández-López, Sergio Rodal and Jorge Pérez-Juste, Ana M.
Sánchez-Iglesias for their contributions and comments.
References
1. See special issue devoted to gold published in the influence of size, shape, and dielectric
Chem. Soc. Rev. (2008, volume 37) environment. J Phys Chem B 107:668–677
2. Kreibig U, Vollmer M (1996) Optical properties 5. Liz-Marzán LM (2006) Tailoring surface
of metal clusters. Springer, Berlin plasmons through the morphology and
3. Mulvaney P (1996) Surface plasmon spectros- assembly of metal nanoparticles. Langmuir

copy of nanosized metal particles. Langmuir 22:32–41
12:788–800 6. Dreaden EC, Alkilany AM, Huang X et al
4. Kelly KL, Coronado E, Zhao LL et al (2003) (2012) The golden age: gold nanoparticles for
The optical properties of metal nanoparticles: biomedicine. Chem Soc Rev 41:2740–2779
7. Giljohann DA, Seferos DS, Daniel WL et al theranostic platform for cancer treatment. Adv
(2010) Gold nanoparticles for biology and med- Mater 24:1418–1423
icine. Angew Chem Int Ed 49:3280–3294 22. Liz-Marzán LM, Giersig M, Mulvaney
8. Rosi NL, Mirkin CA (2005) Nanostructures in P (1996) Synthesis of nanosized gold−silica
biodiagnostics. Chem Rev 105:1547–1562 core−shell particles. Langmuir 12:4329–4335
9. Abalde-Cela S, Aldeanueva-Potel P, Mateo- 23. Stöber W, Fink A, Bohn E (1968) Controlled
Mateo C et al (2010) SERS biomedical appli- growth of monodisperse silica spheres in the
cations of plasmonic colloidal particles. J Royal micron size range. J Colloid Interface Sci
Soc Interface 7:S435–S450 6:62–69
10. Sepúlveda B, Angelomé PC, Lechuga LM et al 24. Enüstün BV, Turkevich J (1963) Coagulation of
(2009) LSPR-based nanobiosensors. Nano colloidal gold. J Am Chem Soc 85:3317–3328
Today 4(3):244–251 25. Gorelikov I, Matsuura N (2008) Single-step
11. Dykman L, Khlebtsov N (2012) Gold coating of mesoporous silica on cetyltrimethyl
nanoparticles in biomedical applications: ammonium bromide-capped nanoparticles.
recent advances and perspectives. Chem Soc Nano Lett 8:369–373
Rev 41:2256–2282 26. Wong YJ, Zhu L, Teo WS et al (2011) Revisiting
12. Ghosh P, Han G, De M et al (2008) Gold the Stöber method: Inhomogeneity in silica
nanoparticles in delivery applications. Adv shells. J Am Chem Soc 133:11422–11425
Drug Deliv Rev 60:1307–1315 27. Liu M, Guyot-Sionnest P (2005) Mechanism
13. Halas N (2008) Nanoscience under Glass: The of silver(I)-assisted growth of gold nanorods
versatile chemistry of silica nanostructures. and bipyramids. J Phys Chem B
ACS Nano 2:179–183 109:22192–22200
14. Liu SH, Han MY (2005) Synthesis, functional- 28. Graf C, Vossen DLJ, Imhof A et al (2003) A
ization, and bioconjugation of monodisperse, general method to coat colloidal particles with
silica-coated gold nanoparticles: robust bio- silica. Langmuir 19:6693–6700
probes. Adv Funct Mater 15:961–967 29. Wiley B, Sun Y, Mayers B et al (2005) Shape-
15. Luo T, Huang P, Gao G et al (2011) controlled synthesis of metal nanostructures:
Mesoporous silica-coated gold nanorods with the case of silver. Chem Eur J 11:454–463
embedded indocyanine green for dual mode 30. Pastoriza-Santos I, Liz-Marzán LM (2009) N,
X-ray CT and NIR fluorescence imaging. Opt N-dimethylformamide as a reaction medium
Express 19:17030–17039 for metal nanoparticle synthesis. Adv Funct
16. Rodríguez-Lorenzo L, Krpetic Z, Barbosa S Mater 19:679–688
et al (2011) Intracellular mapping with SERS- 31. Yin YD, Lu Y, Sun YG et al (2002) Silver
encoded gold nanostars. Integr Biol 3: Nanowires can be directly coated with amor-
922–926 phous silica to generate well-controlled
17. Zavaleta CL, Smith BR, Walton I et al (2009) coaxial nanocables of silver/silica. Nano Lett
Multiplexed imaging of surface enhanced 2:427–430
Raman scattering nanotags in living mice using 32. Pastoriza-Santos I, Sánchez-Iglesias A, García
noninvasive Raman spectroscopy. Proc Natl de Abajo FJ et al (2007) Environmental optical
Acad Sci USA 106:13511–13516 sensitivity of gold nanodecahedra. Adv Funct
18. Yang X, Liu X, Liu Z et al (2012) Near-Infrared Mater 17:1443–1450
light-triggered, targeted drug delivery to cancer 33. Guerrero-Martínez A, Barbosa S, Pastoriza-
cells by aptamer gated nanovehicles. Adv Mater Santos I et al (2011) Nanostars shine bright for
24:2890–2895 you. Colloidal synthesis, properties and applica-
19. Lal S, Clare SE, Halas NJ (2008) Nanoshell- tions of branched metallic nanoparticles. Curr
enabled photothermal therapy: impending Op Colloid Interface Sci 16:118–127
clinical impact. Acc Chem Res 41:1842–1851 34. Kumar PS, Pastoriza-Santos I, Rodríguez-
20. Fales AM, Yuan H, Vo-Dinh T (2011) Silica- González B et al (2008) High-yield synthesis
coated gold nanostars for combined surface- and optical response of gold nanostars.
enhanced Raman scattering (SERS) detection Nanotechnology 19(1):015606
and singlet-oxygen generation: A potential 35. Barbosa S, Agrawal A, Pastoriza-Santos I et al
nanoplatform for theranostics. Langmuir 27: (2010) Tuning size and sensing properties in
12186–12190 colloidal gold nanostars. Langmuir 26:
21. Zhang Z, Wang L, Wang J et al (2012) 14943–14950
Theranostics: mesoporous silica-coated gold 36. Pastoriza-Santos I, Pérez-Juste J, Liz-Marzán
nanorods as a light-mediated multifunctional LM (2006) Silica-coating and hydrophobation
of CTAB- stabilized gold nanorods. Chem 40. Fernández-López C, Mateo-Mateo C, Alvarez-

Mater 18:2465–2467 Puebla RA et al (2009) Highly controlled silica
37. Schneider G, Decher G (2004) From functional Coating of PEG-capped metal nanoparticles
core/shell nanoparticles prepared via layer- and preparation of SERS-encoded particles.
by-layer deposition to empty nanospheres. Langmuir 25:13894–13899
Nano Lett 4:1833–1839 41. Rodríguez-Fernández J, Pérez-Juste J, García
38. Gittins DI, Caruso F (2001) Tailoring the de Abajo FJ et al (2006) Seeded growth
polyelectrolyte coating of metal nanoparticles. of submicron Au colloids with quadrupole
J Phys Chem B 105:6846–6852 plasmon resonance modes. Langmuir 22:
39. Nikoobakht B, El-Sayed MA (2003) Preparation 7007–7010
and growth mechanism of gold nanorods 42. Jana NR (2003) Nanorod shape separation
(NRs) using seed-mediated growth method. using surfactant assisted self-assembly. Chem
Chem Mater 15:1957–1962 Commun (15) 1950–1951
Chapter 7
Surface Modifications by Polymers for Biomolecule

Conjugation
Laura Sola, Marina Cretich, Francesco Damin, and Marcella Chiari
Abstract
Polymeric coatings, usually referred as tridimensional chemistries, provide homogenous surface derivatiza-
tion methods presenting a high reactive group concentration and resulting in an increased binding capacity
of targets. Furthermore, they act as linkers distributing the bound probe also in the axial position, thus
causing a faster reaction with the target involved in biomolecular recognition and can be engineered to
custom tailor their properties for specific applications.
Most approaches which aim at attaching polymers to a surface use a system where the polymer carries
an “anchor” group either as an end group or in a side chain. This anchor group can reacts with appropriate
sites at the substrate surface, thus yielding surface-attached monolayers of polymer molecules (termed
“grafting to”). Another technique is to carry out a polymerization reaction in the presence of a substrate
onto which monomers had been attached leading to the so called “grafting from” approach.
In this chapter, protocols to functionalize glass and silicon surfaces by “grafting to” as well as by
“grafting-from” approach are shown using copolymers made of N,N-dimethylacrylamide (DMA) or
Glycidyl methacrylate (GMA) as the polymer backbone, N-acryloyloxysuccinimide (NAS) as reactive
group, and 3-(trimethoxysilyl)propyl methacrylate (MAPS) or 3-mercaptopropyl trimethoxy silane (MPS)
as anchoring groups.
Key words Biomolecule conjugation, Polymeric coatings, Grafting to, Grafting from, Silicon, Glass
1 Introduction
Recent progresses in micro fabrication techniques and microarray

technologies have led to the development of miniaturized and fully
integrated solid phase analytical devices that allow to perform
entire experiments, studies of complex cellular processes, high-
throughput screening, and parallel diagnostic detection. They
imply a scaling down of the entire analytical process (time, sample
and reagent volumes, costs, etc.) while maintaining very high sen-
sitivity. Considering their tiny dimensions, the aforementioned
surfaces must be perfectly designed and controlled in order to
maximize probe immobilization and target binding efficiency, to
95
96 Laura Sola et al.
reduce background noise, and to prevent nonspecific molecular

interaction.
When the modification of a suitable substrate must be per-
formed to promote covalent attachment of oligonucleotides and
proteins at specific locations, the control of surface chemical–physical
characteristics is of remarkable importance in order to maximize
probe density immobilization and to maintain their native and
functional conformation. Besides, in order to obtain accurate anal-
ysis and an optimization of signal-to-noise ratio, the surface chem-
istry should minimize hydrophobic interactions, which are the
main source of biomolecule nonspecific bonding.
Therefore, there are several requirements that must be taken
into account to chemically change a surface:
– The chemically modified surface must be inherently inert and
resists nonspecific adsorption
– The surface must contain functional groups for the facile
immobilization of molecules of interest
– Bonding between a biomolecule and a solid surface must be
strong enough to retain the molecule on the surface, but also
sufficiently nonintrusive to have minimal effect on the delicate
3D structure
– The linking chemistry must allow the control of biomolecule
orientation
– The local chemical environment must benefit the immobilized
molecules to retain their native conformation
Polymeric coatings, usually referred as tridimensional chemis-
tries, provide a homogenous surface derivatization presenting a
high reactive group concentration and resulting, depending on the
circumstances, in an increased binding capacity of targets or to its
suppression. Furthermore, they act as linkers distributing the
bound probe also in the axial position, thus causing a faster reac-
tion with the target involved in biomolecular recognition.
Additionally, 3D scaffolds can be engineered to custom tailor their
properties for specific applications.
Most approaches which aim at attaching polymers to a surface
use a system where the polymer carries an “anchor” group either as
an end group or in a side chain. This anchor group can reacts with
appropriate sites at the substrate surface, thus yielding surface-
attached monolayers of polymer molecules (termed “grafting to”)
(Fig. 1) [1–3]. Another straightforward technique is to carry out a
polymerization reaction in the presence of a substrate onto which
monomers had been attached [1, 4, 5].
During the polymerization reaction, the surface-anchored
monomers are incorporated into growing polymer chains (Fig. 1).
Although “grafting to” reactions are easy to perform, it should be
noted that several complication arise from the use of this strategy.
Surface Modifications by Polymers for Biomolecule Conjugation 97
Fig. 1 Schematic illustration of different processes used for the attachment of polymers to surfaces: (a) “graft-
ing to”; (b) grafting via incorporation of surface-bound monomeric units; (c) “grafting from/surface-initiated
polymerization”
For example, in the case of functional polymers, the anchor groups

present in the polymer chain and the sites exposed onto the surface
must not compete and/or react with the functionalities. Another
complication inherent to “grafting to” processes is an intrinsic lim-
itation of the film thickness. With increasing coverage of the sur-
face with attached chains, the polymer concentration at the
interface quickly becomes larger than the concentration of poly-
mers in solution. Additional chains, which are to become attached
to the surface, must diffuse against this concentration gradient that
ever increases with increasing grafting density of the attached poly-
mer. Moreover, in order to accommodate further chains, these
must change from coil conformation in solution to a stretched
(“brush-like”) conformation at the surface. Hence, the higher the
graft density of the chains at the surface, the stronger will be the
entropy penalty, and this rapidly precludes the attachment of fur-
ther chains.
The term “polymer brush” refers to a system in which chains
of polymer molecules are attached by one end to a surface. In spite
of the fact that the “grafting to” method provides polymers
attached in such fashion, due to the problems discussed above, it is
not able to produce dense films in which the polymer chains are
crowded and are stretched away from the surface. From the stretch-
ing of the polymer chains perpendicular to the surface, several new
physical phenomena arise. Examples are ultralow friction surfaces
[6] in which materials coated with surface-attached polymer chains
do not become wetted by free polymer, even if the surface-attached
and the free chains are chemically identical. Another characteristic
of concentrated polymer brushes is noted to be size-exclusion
effect [7], in which the graft chains are highly extended and highly
oriented so that large molecules, sufficiently large compared with
the distance between the nearest neighbor graft points, are physi-
cally excluded from the entire brush layer. In addition, in the case
of functional brush polymers, high densities of functional groups
can be obtained at the surface of the substrate through moving

from the strictly two-dimensional arrangement of these groups
present in typical surfaces to a more three-dimensional situation.
An example which illustrates such a behavior is the attachment of
DNA probe molecules to surface-attached polymer chains, which
can significantly enhance the sensitivity of a DNA-chip [8].
An effective method to prepare high-density brushes is the
“grafting-from” method, that is, the graft polymerization starting
with initiating sites fixed on the surface. In this technique, the
addition of monomer to growing chain ends or to primary radicals
is not strongly hindered by the already grafted chains in a good
solvent condition. Therefore, this technique is more promising to
produce a polymer film with a larger thickness and a higher graft
density than the “grafting to” technique [9–11]. The polymeriza-
tion process obtained by “grafting-from” approach is also defined
as Surface-Initiated Polymerization (SIP).
In this chapter we will show protocols to functionalize glass
and silicon surfaces by “grafting to” as well as by “grafting-from”
approach using copolymers made of N,N-dimethylacrylamide
(DMA) or glycidyl methacrylate (GMA) as the polymer backbone,
N-acryloyloxysuccinimide (NAS) as reactive group, and 3-(trime-
thoxysilyl)propyl methacrylate (MAPS) or 3-mercaptopropyl tri-
methoxy silane (MPS) as anchoring groups.
2 Materials
2.1 Synthesis 1. N,N-Dimethylacrylamide (DMA) (Scheme 1).

of Poly(DMA- 2. 3-(trimethoxysilyl)propyl methacrylate (MAPS).
co-NAS-co-MAPS)
3. N-Acryloyloxysuccinimide (NAS).
4. Aluminum oxide.
5. Tetrahydrofuran (THF).
6. α, α′-Azoisobutyronitrile (AIBN).
7. Petroleum ether.
2.2 Derivatization 1. HARRICK Plasma Cleaner, PDC-002 (Ithaca, NY, USA).

of Substrates with 2. Ammonium sulfate.
Poly(DMA-co-NAS-co-
3. poly(DMA-co-NAS-co-MAPS) synthesized as described in
MAPS)
Subheading 3.1.
4. Coating solution: 1 % w/v of poly(DMA-co-NAS-co-MAPS) in
an aqueous solution of 20 % ammonium sulfate.
2.3 Synthesis 1. Glycidyl methacrylate.

of Poly(glycidyl 2. α, α′-Azoisobutyronitrile (AIBN).
methacrylate) (PGMA)
Scheme 1 Synthesis of poly(DMA-co-NAS-co-MAPS)
Scheme 2 Substrate derivatized by poly(glycidyl methacrylate) (PGMA)
2.4 Derivatization 1. HARRICK Plasma Cleaner, PDC-002 (Ithaca, NY, USA)

of Substrates with (Scheme 2).
Poly(glycidyl 2. PGMA synthesized as described in Subheading 3.3.
methacrylate) (PGMA)
3. Dimethylformamide (DMF).
1. HARRICK Plasma Cleaner, PDC-002 (Ithaca, NY, USA).

2.5 Derivatization 2. N,N-Dimethylacrylamide (DMA).
of Substrates with
3. N-Acryloyloxysuccinimide (NAS).
Grafted Poly(DMA-co-
NAS-co-MAPS) 4. 3-(trimethoxysilyl)propyl methacrylate (MAPS).
andGraftedPoly(DMA-co- 5. 3-Mercaptopropyl trimethoxy silane (MPS).
NAS-co-MPS) 6. Aluminum oxide.
8. Toluene.
10. N,N-Dimethylformamide (DMF).
11. IRGACURE 184 (Ciba Specialty Chemicals, Basel, Switzerland).
2.6 Derivatization 1. N,N-Dimethylacrylamide (DMA).

of Substrates with 2. N-Acryloyloxysuccinimide (NAS).
Grafted Poly[DMA-b-
3. (3-aminopropyl)-trimethoxysilane (APS).
(DMA-co-NAS)] via
Free Radical 4. 2-Bromopropionylbromide.
Polymerization [12] 5. Phenylmagnesium bromide.
6. Triethylamine.
7. Carbon disulfide (CS2).
8. Toluene.
9. Dichloromethane.
10. Tetrahydrofuran anhydrous (THF, 99 % inhibitor free).
11. Sodium sulfate anhydrous.
12. Octyl-trimethoxysilane (96 %, nOS).
14. Aluminum oxide.
15. HARRICK Plasma Cleaner, PDC-002 (Ithaca, NY, USA).
2.7 Substrates Glass, silicon and plastic substrates can be modified, including
soda-lime or borosilicate glass, silicon slides, COC (Cyclic Olefin
Copolymer), PET (Polyethylene terephthalate), PMMA
(Poly(methyl methacrylate)), or PDMS (polydimethylsiloxane)
chips. Substrates are typically 15 × 15 mm2 square slides with a
thickness from 1 to 5 mm. Microscope slide format is also
frequently used.
3 Methods
The methods described in this section outline (a) synthesis of

poly(DMA-co-NAS-co-MAPS), (b) coating of substrates with the
poly(DMA-co-NAS-co-MAPS), (c) synthesis of PGMA, (d) coating
of substrates with PGMA, (d) synthesis of grafted poly(DMA-co-
NAS-co-MAPS), (e) synthesis of grafted poly(DMA-co-NAS-co-
MPS), and (f) synthesis of grafted poly [DMA-b-(DMA-co-NAS)]
3.1 Synthesis The copolymer made of DMA (with 97 % molar percentage), NAS
of Poly(DMA- (2 % molar percentage). and 3-(trimethoxysilyl)propyl methacry-
co-NAS-co-MAPS) late (MAPS, 1 % molar percentage) is synthesized by free radical
copolymerization and used as a functional coating for biomolecule
conjugation to the substrate.
1. In a 250 mL three neck round bottom flask, equipped with
condenser, magnetic stirring, and nitrogen connection, dis-
solve N,N-dimethylacrylamide (DMA, 4.00 g, 4.00 × 10−2 mol;
see Note 1), NAS (0.14 g, 8.30 × 10−4 mol), and the initiator
α,α′-azoisobutyronitrile (AIBN, 1.30 × 10−2 g, 7.90 × 10−5 mol),

under nitrogen flow, in 20.00 mL of anhydrous tetrahydrofu-
ran (THF) The concentration of the monomer feed in the
solvent is 0.2 g/ml (20 % w/v).
2. Degas the solution by applying vacuum for 10 min.
3. Restore the nitrogen flow and add 3-(trimethoxysilyl)propyl
methacrylate (MAPS, 1.03 × 10−1 g, 4.15 × 10−4 mol).
4. Stir magnetically and heat the solution to 65 °C for 2 h under
nitrogen atmosphere
5. After the polymerization is completed, dilute 1:1 with anhy-
drous THF.
6. Precipitate the polymer by dripping the reaction mixture into
a large excess of petroleum ether (about 1:10 by volume),
while stirring.
7. Filter the obtained white solid on a Buchner funnel.
8. Dry the copolymer under a vacuum for 1–2 h at room
temperature.
9. Store it at −20 °C in a dry environment.
3.2 Derivatization The coating of substrates requires two steps: (a) substrate surface
of Substrates with pretreatment and (b) adsorption of the copolymer.
Poly(DMA-co-NAS-
1. Surface pretreatment: The substrates are cleaned and activated
co-MAPS) by pretreatment with oxygen plasma for 10 min (high radio
frequency).
2. Dissolve poly(DMA-co-NAS-o-MAPS) to a final concentration
of 1 % w/v in an ammonium sulfate aqueous solution 20 % w/v
(see Note 2).
3. Adsorption of poly(DMA-co-NAS-co-MAPS): Immerse the
substrates into the coating solution for 30 min at room tem-
perature in a PS (Polystyrene) chamber (see Note 3).
4. Wash the substrates vigorously with ddH2O to remove the
excess of the copolymer on the surface.
5. Dry the surfaces with nitrogen stream to avoid stain formation.
6. Cure the slides in a vacuum oven at 80 °C for 15 min (see
Note 4).
7. Store the slides at room temperature in a desiccator and use
the slides within 4 weeks after production (see Note 5).
3.3 Synthesis 1. In a round bottom flask, equipped with condenser, magnetic

of Poly(glycidyl stirring, and nitrogen connection, dissolve glycidyl methacry-
methacrylate) (PGMA) late 1 M in tetrahydrofuran (THF), under nitrogen flow.
2. Degas the solution by applying vacuum for 10 min.
Scheme 3 Substrate derivatized by grafted poly(DMA-co-NAS-co-MPS)
3. Restore nitrogen flow and add 1 mM of α,α′-azoisobutyronitrile

(AIBN).
4. Heat the solution to 65 °C overnight, under nitrogen
atmosphere.
5. Cool the solution at room temperature and purify the polymer
by precipitation in diethyl ether.
6. Filter the resulting white solid product on a Buchner funnel.
7. Dry the polymer under vacuum for 1 h at room temperature.
3.4 Derivatization The coating of the chips requires two steps: (a) surface pretreat-
of Substrates ment and (b) adsorption of the copolymer.
with Poly(glycidyl
1. Surface pretreatment: The substrates are cleaned and activated
methacrylate) (PGMA) by pretreatment with oxygen plasma for 10 min (high radio
frequency).
2. Dissolve poly(glycidyl methacrylate) (PGMA) in anhydrous
DMF at a final concentration of 5 % w/v.
3. Adsorption of the copolymer PGMA onto the substrates:
Immerse the slides into the solution overnight at 70 °C. Rinse
the substrate with DMF and then with anhydrous THF (see
Note 6).
4. Dry each slide with a nitrogen stream to avoid stain formation.
5. Place the substrates in a vacuum oven at room temperature for
30 min.
3.5 Derivatization The coating of the substrates requires two steps: (a) silanization of
of Substrates the substrates with 3-(trimethoxysilyl)propyl methacrylate (MAPS)
with Grafted and (b) polymerization of poly(DMA-co-NAS) either by thermo-
Poly(DMA-co-NAS- activation or by photo-activation (Scheme 3).
co-MAPS) 1. Pretreat the slides with oxygen plasma for 10 min (high radio
frequency).
2. Prepare a solution of 3-(trimethoxysilyl)propyl methacrylate
(MAPS) 1 % v/v in toluene.
3. Silanization with MAPS: Immerse the slides into the solution
for 4 h, at room temperature.
4. Rinse each slide with fresh toluene and with THF(see Note 7).
5. Dry each slide with nitrogen stream to avoid stain formation.
6. Cure in a vacuum oven at 80 °C for 30 min (see Note 4).
7. For thermo-activated polymerization: Prepare the polymeriza-
tion solution by dissolving DMA (8.90 g, 9.00 × 10−2 mol; see
Note 1) and NAS (1.70 g, 1.00 × 10−2 mol) in 100 mL of
N,N-dimethylformamide (DMF).
8. Immerse the previously silanized slides into the solution.
9. Degas the solution by purging Argon for 30 min.
10. Add AIBN (3.6 × 10−2 g, 2.20 × 10−4 mol).
11. Heat the solution to 65 °C overnight under nitrogen
atmosphere.
12. Rinse each slide with fresh DMF and finally with THF (see
Note 6).
14. Dry the slides under vacuum at room temperature.
15. For photo-activated polymerization: Prepare the polymeriza-
tion solution by dissolving DMA (4.42 × 10−1 g, 4.40 × 10−3 mol;
see Note 1) and NAS (8.50 × 10−2 g, 5.02 × 10−4 mol) in 5 mL
of DMF
16. Add IRGACURE 184 (3.50 × 10−2 g, 1.70 × 10−4 mol).
17. Immerse the previously silanized slides into the transparent
chamber for photo-activated polymerization (see Note 8).
18. Expose the chamber, containing the slides, to UV light
(365 nm) for 10 min.
19. Rinse with fresh DMF and finally with THF (see Note 6).
21. Dry under vacuum at room temperature.
3.6 Derivatization The coating of the substrates requires two steps: (a) silanization of
of Substrates the substrates with 3-mercaptopropyl trimethoxy silane (MPS) and
with Grafted (b) polymerization of poly(DMA-co-NAS) either by thermo-
Poly(DMA-co-NAS- activation or by photo-activation.
co-MPS) 1. Pretreat the slides with oxygen plasma for 10 min (high radio
frequency).
2. Prepare a solution of 3-mercaptopropyl trimethoxy silane
(MPS) 1 % v/v in toluene.
3. Silanization with 3-mercaptopropyl trimethoxy silane (MPS):
Immerse the slides into the solution for 4 h at room
temperature.
4. Rinse each slide with fresh toluene and with THF (see Note 7).

6. Cure in a vacuum oven at 80 °C for 30 min (see Note 4).
7. Follow procedures described in Subheading 3.5 for thermo-
activated polymerization (steps 7–14) or photo-activated
polymerization (steps 15–21).
3.7 Derivatization The procedure consists of three steps: (a) Synthesis of 1-oxo-1-(3-
of Substrates with (trimethoxysilyl)propylamino)propan-2-yl-benzodithioate (RAFT
Grafted Poly[DMA- Silane), (b) Surfaces functionalization with RAFT Silane, (c)
b-(DMA-co-NAS)] polymerization.
1. Synthesis of 1-oxo-1-(3-(trimethoxysilyl)propylamino)propan-
2-yl-benzodithioate (RAFT Silane): In a 100 mL two neck
round bottom flask, equipped with a dropping funnel and
nitrogen connection, dissolve, under nitrogen flow,
(3-aminopropyl)-trimethoxysilane (APS, 4.86 mL, 0.027 mol)
and triethylamine (TEA, 3.77 mL, 0.027 mol) in dichloro-
methane (30 mL).
2. Cool the round bottom flask to 0 °C with an ice bath.
3. Prepare a solution of 2-bromo propionyl bromide (2.83 mL,
0.027 mol) in dichloromethane (5 mL) and drip it into the
round bottom flask, while stirring.
4. Stir the solution at room temperature for 2 h.
5. Filter the white precipitate of triethylamine bromohydrate.
6. Evaporate the solvent under vacuum.
7. Purify the residue by distillation using a Kugelrohr apparatus
(175 °C, 1 mBar) to obtain the bromine derivative (1) as a
colorless oil.
8. In a 100 mL two neck round bottom flask, equipped with
dropping funnel and nitrogen connection, dissolve phenyl-
magnesium bromide (25.4 mL, 0.025 mol) in anhydrous THF
(30 mL).
9. Cool the stirring solution to 0 °C with an ice bath.
10. Slowly add carbon disulfide (CS2, 2.9 mL, 0.048 mol) and stir
for 1.5 h at room temperature
11. Cool the solution again to 0 °C with an ice bath.
12. Dissolve the bromine derivative (1) obtained previously (7.6 g;
0.024 mol) in anhydrous THF (10 mL) and drip it into the
round bottom flask.
13. Stir overnight at room temperature.
14. Evaporate the solvent under vacuum.
15. Dissolve the obtained residue in dichloromethane (15 mL)
and wash twice with brine.
16. Dry the organic phases over anhydrous sodium sulfate (Na2SO4).
17. Evaporate under vacuum to obtain the RAFT silane as a red oil.
18. Surfaces functionalization with RAFT Silane : Pretreat surfaces
with Oxygen plasma (high radio frequency).
19. Dissolve RAFT silane (0.5 mM) and octyl-trimethoxysilane
(nOS, 2 mM) in toluene.
20. Immerse the pretreated slides into the solution, under nitro-
gen atmosphere, for 4 h.
21. Wash each slide with fresh toluene and with THF (see Note 7).
23. Dry under vacuum at room temperature for 30 min.
24. Synthesis of grafted poly(DMA): Dissolve DMA (45.95 mL,
0.446 mol; see Note 1), tert-butyl dithiobenzoate (t-BDB,
0.262 g, 0.125 × 10−2 mol), and AIBN (0.041 g,
0.025 × 10−2 mol) in 100 mL of toluene.
25. Immerse the slides previously silanized with RAFT Silane.
26. Degas the solution by purging argon for 1 h.
27. Heat the solution to 80 °C overnight, under nitrogen
atmosphere.
28. Rinse the slides by Soxhlet extraction for 12 h using THF.
29. Dry each slides using a nitrogen stream.
30. Synthesis of graft-poly[DMA-b-(DMA-co-NAS)]: Dissolve
DMA (9.27 mL, 0.09 mol; see Note 1), NAS (1.69 g, 0.01 mol)
and AIBN (9.2 mg, 0.056 × 10−3 mol) in 100 mL of DMF.
31. Immerse the slides previously coated with the first block of
grafted poly(DMA) into the solution.
32. Degas the solution by purging Argon for 1 h.
33. Heat overnight at 80 °C, under nitrogen atmosphere.
34. Wash each slide with fresh DMF and with THF (see Note 6).
36. Dry under vacuum at room temperature for 30 min.
4 Notes
1. DMA must be filtered on Aluminum oxide to remove inhibitor

before use. Filtered DMA can be stored for no more than a
week at 2–4 °C.
2. First prepare a stock solution of ammonium sulfate at a 40 %
saturation level by adding 242 g of ammonium sulfate to 1 L
of ddH2O. Depending on the volume of the chamber and the
number of the slides to be coated, weigh the exact amount of
copolymer to obtain a water solution at a final concentration of
Fig. 2 Picture of the glass chamber used for UV initiated polymerization and
derivatization of flat substrates
2 % w/v. When the copolymer is completely dissolved, dilute it

1:1 with the stock solution of ammonium sulfate at a 40 %
saturation level.
3. Use preferably plastic (for example polystyrene) container as
glass ones would be coated as well, competing with the surface
of the slides that are intended to be coated
4. Curing step under vacuum is fundamental to assure the forma-
tion of covalent bonds between surface silanols and silane
reagents.
5. Slides coated with functional groups are best stored in dry con-
dition in a desiccator. This is fundamental when using the
poly(DMA-co-NAS-co-MAPS) because its reactive monomer is
very sensitive to humidity. Poly(DMA-co-NAS-co-MAPS)
coated slides, can be stored in any plastic container in a desic-
cator at room temperature and used up to 4 weeks.
6. Rinse quickly each slide in both solvents. First rinse each slide
in DMF to remove the excess of reagent, then immediately in
THF, as this evaporates faster avoiding formation of stains.
7. Rinse quickly each slide in both solvents. First rinse each slide
in toluene to remove the excess of silanizing agent, then imme-
diately in THF, as this evaporates faster avoiding formation of
stains.
8. To immerse the substrates into polymerization solution and be
sure their surface is completely exposed to UV light prepare a
chamber using two large glass slides (or any transparent material
which does not dissolve into DMF) and a silicon gasket to

separate the bottom from the upper glass slide. Place the sub-
strates that have to be coated within the area delimited by the
gasket, then clamp the two glass slides. The silicon gasket will
allow also the use a syringe to introduce the monomer solution
into the obtained chamber. Use a second needle to allow
removal of air bubbles so to be sure the chamber is completely
filled with monomer solution (Fig. 2).
Acknowledgments
Financial support from the Italian Ministry of University and

Research PRIN 2008 (2008JWKYXB_005), Italian Institute of
Technology (project SEED “IPG-CHIP”), and Fondazione
Cariplo (SpinBioMed, grant # 2008–2330) are gratefully
acknowledged.
References
1. Krenkler KP, Laible R, Hamann K (1976) methacrylate) brushes by the size-exclusion
Polyreaktionen an Pigmentoberflächen. VII. effect. Macromolecules 39:2284–2290
Mitteilung: Reaktionen von Polymeren mit 8. Pirri G et al (2006) Microarray glass slides
endständigen Chlorsilangruppen an coated with block copolymer brushes
Siliciumdioxidoberflächen. Angew Makromol obtained by reversible addition chain-trans-
Chem 53:101–123 fer polymerization. Anal Chem
2. Tsubokawa N et al (1990) Grafting onto car- 78:3118–3124
bon black: reaction of functional groups on 9. Tsubokawa N et al (1989) Graft polymeriza-
carbon black with ACYL chloride-capped tion of acrylamide from ultra silica particles by
polymers. J Macromol Sci 27:445–457 use of a redox system consisting of ceric ion
3. Tsubokawa N, Kuroda A, Sone Y (1989) and reducing groups on the surface. Polym J
Grafting onto carbon black by the reaction of 21:475–481
reactive carbon black having epoxide groups with 10. Prucker O, Rühe J (1998) Synthesis of
several polymers. J Polym Sci A27:1701–1712 poly(styrene) monolayers attached to high sur-
4. Dimitrenko AV et al (1990) Polymer—inor- face area silica gels through self-assembled
ganic selective adsorbents for gas chromatog- monolayers of azo initiators. Macromolecules
raphy produced by graft polymerization. J 31:592–601
Chromatogr A 520:21–31 11. Prucker O, Rühe J (1998) Mechanism of radi-
5. Hashimoto K et al (1982) Graft copolymeriza- cal chain polymerizations initiated by azo com-
tion of glass fiber and its application. J pounds covalently bound to the surface of
Macromol Sci 18:173–190 spherical particles. Macromolecules
6. Klein J et al (1993) Lubrication forces between 31:602–613
surfaces bearing polymer brushes. 12. Di Carlo G et al (2012) Synthesis and confor-
Macromolecules 26:5552–5560 mational characterization of functional di-
7. Yoshikawa C et al (2006) Protein repellency of block copolymer brusche for microarray
well-defined, concentrated poly(2-hydroxyethyl technology. Appl Surf Sci 258:3750–3756
Chapter 8
Functionalization Protocols of Silicon

Micro/Nano-mechanical Biosensors
Francesca Frascella and Carlo Ricciardi
Abstract
Functionalization is a key element in biodetection technologies such as micro/nano-mechanical sensors.
Since assay sensitivity and stability drastically depends on a proper bioreceptor immobilization, the sensing
surface must be first chemically modified with uniform, well-packed, and robust layers. Here, we describe
three functionalization protocols that we developed for the surface modification with amino, aldehyde,
and carboxyl groups of micro/nano-mechanical biosensors.
Key words Self-assembled monolayer, Organosilane, Functionalization, Molecular recognition,

Biosensor
1 Introduction
Micro/Nano-mechanical sensors such as cantilevers have recently

gained a significantly growing attention from the scientific com-
munity, due to their capability of label-free detection of bio/chem-
ical molecules at extremely low concentrations [1]. Such
applications require a chemical functionalization of the sensor sur-
face to attach recognition molecules that have both high affinity
and selectivity for the intended bio/chemical targets [2].
A widely used method to create well-organized and compact
functionalizing monolayers is to exploit the properties of self-
assembling of alkyl chains with thiol groups (R-SH) on gold sub-
strates, or of alkyl silanes (R-Si-OH) on silicon-based substrates.
Such covalently anchored layers are uniform, densely packed and
robust, thus guaranteeing a high stability and sensitivity of the final
sensor platform. Furthermore, Self-assembled monolayers (SAMs)
represent an extremely versatile tool, being able to be synthesized
with chains of variable length and groups with specific chemical
properties, representing an ideal way of anchorage of biomolecules
to the substrate.
109
110 Francesca Frascella and Carlo Ricciardi
We here report on the functionalization protocols we developed

for silicon micro/nano-mechanical biosensors, making use of the
following:
– 3-Aminopropyltriethoxysilane (APTES) to have a sensing
surface rich of amino (−NH2) groups
– 3-Aminopropyltriethoxysilane (APTES) followed by
Glutaraldehyde (GA) to have a sensing surface rich of aldehyde
(−CHO) groups
– 3-Cyanopropyltriethoxysilane (CPTES) to have a sensing sur-
face rich of carboxyl (−COOH) groups
2 Materials
Prepare all solutions using high quality pure water of high resistiv-
ity and low TOC (>10 MΩ cm and <50 ppb) and ACS reagents
(essay ≥99.5 %), unless otherwise stated. Prepared solutions and
reagents should be stored at room temperature (unless otherwise
indicated).
Organosilane reaction involves handling of air-sensitive com-
pound [3], so it is necessary to carry out the reaction in an inert
atmosphere. A special equipment with a source of inert gas (prefer-
ably argon), a septum inlet, a bubbler, and syringes fitted only with
small-gauge needles (no larger than 18-gauge) is required, while
glass apparatus is of common type. Laboratory glassware, syringe
and needle should be dried in an oven (140 °C for 4 h) prior
to use.
2.1 Cleaning and 1. Piranha solution: mixture 3:1 of 98 % w/w H2SO4 and 30 %
Oxidation of the w/w H2O2 solution (hydrogen peroxide must be stored at
Surfaces 4 °C). Add 3 parts in volume of sulfuric acid to a graduated
cylinder or a glass beaker (see Note 1). Then add slowly 1 part
of hydrogen peroxide to the acid and mix. Once prepared, the
solution has to be used as soon as possible.
2.2 Silanization 1. 3-Aminopropyltriethoxysilane 99 % (APTES) stored under argon

Components atmosphere: 1 % v/v solution in anhydrous toluene 99.8 %.
2. Glutaraldehyde solution 25 % v/v (GA): 0.5 % v/v in dd-H2O.
Store at 4 °C.
3. Borate Buffer Saline (BBS) 0.1 M, pH 8.5. Dissolve 0.011 g of
boric acid (H3BO3) in 5 mL dd-H2O. Weigh 0.026 g of sodium
chloride (NaCl) and dissolve it in 50 mL dd-H2O. Mix the two
solutions and adjust pH with small amount of sodium hydrox-
ide (NaOH) solution (see Note 2). Store at 4 °C.
4. Basic pH adjusting solution: 0.1 M NaOH. Dissolve 0.1 g
NaOH in 2.5 mL dd-H2O.
Functionalization of Si Cantilever Biosensors 111
5. Sodium cyanoborohydride NaBH3(CN) 95 %: 5 M solution in

NaOH. Dissolve 0.31 g of NaBH3(CN) in 1 mL of NaOH
1 M (see Note 3).
6. (3-Cyanopropyl)triethoxysilane 98 % (CPTES) stored under
Argon atmosphere: 1 % v/v solution in anhydrous toluene
99.8 %.
7. Hydrolysis solution: H2SO4 48 % solution in deionized water.
Measure 49 mL of H2SO4 98 % w/w and make up to 100 mL
with water.
3 Methods
3.1 Functionalization 1. Grow a silicon oxide thin film (190 nm) on silicon substrates
with Amino Groups: by thermal oxidation at 1,100 °C in O2 atmosphere for 3 h.
APTES (Fig. 1) 2. Prior to monolayer preparation, treat the silicon substrates for
15 min in a freshly prepared piranha solution. Since the mixture
is a strong oxidizer, it will remove most organic contaminants
and hydroxylate the surface (adding of OH groups), creating
the proper condition for the attachment and displacement of the
alkoxy group from the silane, thus forming a covalent –Si–O–Si–
bond (see Note 1).
3. Rinse the substrates with copious deionized water (3 times)
and dry them carefully in a stream of nitrogen.
4. After that, incubate the freshly cleaned silicon substrates
(see Note 4) in a round bottom flask with an oven dried magnetic
stir bar inside, filled with 1 % v/v APTES solution in anhydrous
toluene at 70 °C or until solvent reflux for 10 min. Add APTES
via an oven-dried syringe and maintain the flask under an atmo-
sphere of argon, following an anhydrous protocol (see Note 5).
5. Silane-coated substrates were rinsed deeply with toluene and
carefully dried under N2 flux.
Fig. 1 Scheme of functionalization with amino groups—APTES

Fig. 2 Scheme of functionalization with aldehyde groups—APTES and GA
3.2 Functionalization 1. Just after the reaction with APTES (steps 1–5 in
with Aldehyde Groups: Subheading 3.1), incubate the samples in a 0.5 % v/v GA solu-
APTES and GA (Fig. 2) tion in BBS 0.1 M (pH 8.5) for 1 h using an orbital shaker at
40 rpm (see Note 2).
2. After 15 min, add 300 μL of sodium cyanoborohydride solu-
tion (5 M) in NaOH, in order to create a more stable bond,
reducing the imine formed by the reaction between −NH2 and
−CHO groups (see Note 6).
3. After incubation, rinse the treated substrates several times with
deionized water, dry them under nitrogen stream, and store
them in a sealed desiccator until needed.
3.3 Functionalization 1. Grow a silicon oxide thin film (190 nm) on silicon substrates
with Carboxylic by thermal oxidation 1,100 °C in O2 atmosphere for 3 h.
Groups: CPTES (Fig. 3) 2. Soak oxidized substrates into a freshly prepared piranha solu-
tion for 15 min (see Note 1), to remove organic contaminants.
3. Rinse several times the treated substrates with dd-H2O and dry
them in a nitrogen stream.
4. Incubate the freshly cleaned substrates with 1 % v/v CPTES
toluene solution in reflux solvent condition (around 70 °C) for
10 min in experimental anhydrous conditions (see Note 4).
5. Rinse several times (3 times) the substrates with toluene and
then dry them under nitrogen stream.
6. Treat the freshly silanized samples in H2SO4 48 % solution at
100 °C under reflux for 3 h in Argon flux, so to hydrolyze −
CN group into −COOH (see Note 7).
Fig. 3 Scheme of functionalization with carboxylic groups—CPTES
7. After incubation, rinse the treated substrates several times with

deionized water, dry them under nitrogen stream, and store
them in a sealed desiccator until needed.
4 Notes
1. Piranha solution is an extremely strong oxidant and should be

handled very carefully. It is highly corrosive and oxidizing.
Make sure that surfaces are reasonably clean, and completely
free of organic solvents from previous wash steps, before com-
ing into contact with piranha solution.
Mixing the solution is exothermic, so prepare and store
piranha solutions only in approved glass containers that can
easily withstand temperatures of at least 200 °C.
Piranha solution should be prepared by slowly adding
small aliquots of hydrogen peroxide to sulfuric acid, never the
reverse.
Immersing the substrate into the solution (remember to
use Teflon-coated tweezers) should be done slowly to prevent
thermal shock that may crack the substrate material and one by
one, giving time after each substrate addition for solution to
quit foaming.
Due to the self-decomposition of hydrogen peroxide, pira-
nha solution should not be stored, and used as freshly pre-
pared. The solution should be allowed to cool prior to disposal,
so that oxygen gas could dissipate. To speed up the dissolution
of piranha hydrogen peroxide, few grains of manganese diox-
ide can be added.
2. Remember to check and adjust pH when the solution is at
room temperature. The gap between BBS starting pH (around 6)
and the required pH (8.5) is not large, consequently a 1 M of
NaOH solution presents an adequate ionic strength to allow a
good pH adjust, by adding around 50 μL of basic solution
every 100 mL BBS.
3. Sodium cyanoborohydride contact with air should be kept to a

minimum because the compound is very hygroscopic.
Remember to weigh into a good ventilated area and to use a
fume cupboard, because the salt has a very unpleasant smell.
NaBH3(CN) is used for the reduction of imines to amines.
Selectivity is achieved in basic condition (BBS pH 8.5).
4. The substrate has to be freshly cleaned before the deposition of
any silane, in order to obtain a good reaction yield.
5. To prepare a good anhydrous reaction you have to carry out the
reaction in Ar flux and to use laboratory glassware dried in oven
at 140 °C for at least 4 h, in order to remove the thin film of
adsorbed water that, acting as a catalyst, causes APTES hydrolysis
in ethanol and trisilanols. Pay attention not to assemble the syringe
body and the plunger before being placed in oven. The glassware
should be assembled while hot and fluxed with a stream of argon,
as soon as fixed. A dry syringe should be separated from the envi-
ronmental humidity by inserting the tip of the needle into a rub-
ber septum. The reaction vessel must be secured through a
mineral oil bubbler. At all times during the reaction, the system
should be under a slow and continuous argon flow (slight positive
pressure). Liquid reagents should be transferred with syringe by
first pressurizing the reagent bottle with inert gas. This slight
pressure is used to slowly fill the syringe with the reagent, without
making air enters the bottle. The measured volume of reagent in
the syringe has to be quickly transferred to the reaction flask by a
rubber septum put on one neck [4].
6. Borate buffer pH 8.5 ensures that amino groups exposed at
sensor surface are not in a protonated form (−NH2), thus
being available to react with GA aldehyde groups (−CHO).
7. Nitriles (−CN group) can be hydrolyzed, by heating under
reflux with dilute acid, in two steps: first the protonated nitrile
gives a primary amide and then the hydrolysis of this amide
gives a free −COOH group. The reaction should be done
under N2 or Ar flow, in order to remove the formation of etha-
nol that would cause esterification of carboxylic groups [5].
Acknowledgments
The authors wish to thank the financial support of Regione

Piemonte (Converging Technologies 2007—NAMATECH grant,
Poli di Innovazione POR-FESR 2007–2013—CANESTRO grant,
and Piattaforme Innovative POR-FESR 2007–2013—ITACA and
SAFE-FOOD CONTROL grants) and Italian Ministry of Research
and Education (FIRB 2010—NEWTON grant, Progetti Bandiera
2011—NANOMAX grant).
References
1. Eom K et al (2011) Nanomechanical resonators 4. Furniss BS et al (1989) Reaction involving air-
and their applications in biological/chemical sensitive compounds. In: Vogel’s textbook of
detection: Nanomechanics principles. Phys Rep pratical organic chemistry, 5th edn. Wiley,
503:115–163 New York 2:120–131
2. Johnson BN, Mutharasan R (2012) Biosensing 5. Yang C, Zibrowis B, Schüth F (2003) A novel-
using dynamic-mode cantilever sensors: a review. synthetic route for negatively charged ordered
Biosens Bioelectron 32:1–18 mesoporous silica SBA-15. Chem Comm 14:
3. Aldrich Chemical Company (2010) Handling 1772–1773
air-sensitive reagents. In: Technical Information
Bulletin AL-134:1–9
Part II
Characterizing Bio–Nano Interfaces

Chapter 9
Stability and Aggregation Assays of Nanoparticles

in Biological Media
James Chen Yong Kah
Abstract
Colloidal stability of nanoparticles in biological media is crucial to preserve their utility as aggregation
often leads to undesirable biological response. A quantitative measurement of nanoparticles aggregation in
solution would provide a valuable assessment of colloidal stability of bio–nano interfaces after surface func-
tionalization. Here, we develop a quantitative technique based on optical absorption to assay the colloidal
stability of plasmonic nanoparticles functionalized with different amphiphilic surface ligands in biologically
relevant media.
Key words Gold nanorods, Colloidal stability, Aggregation, Amphiphilic ligand, Surface charge, Cell
culture media
1 Introduction
The biological utility of nanoparticles often stems from their unique

physicochemical properties that are often a function of their surface
interactions with other nanoparticles or biological entities.
Therefore, apart from key attributes required for their successful
use in biology such as low toxicity, biocompatibility and effective
biological clearance, the colloidal or dispersion stability of these
nanoparticles is also crucial to preserve their intended physico-
chemical behavior and hence their utility in the physiological envi-
ronment. Aggregation of nanoparticles in biological media often
undermines their ability to perform their intended functions [1].
Colloidal stability of nanoparticles is maintained by two main
mechanisms, namely, steric stabilization and charge stabilization,
which can be probed by a wide range of techniques depending on
the types of nanoparticles. Inorganic nanoparticles with unique
optical properties are often probed by their optical emission such
as photoluminescence intensity in the case of quantum dots [2–4]
or their optical absorption in the case of plasmonic nanoparticles
[5–9], since their optical properties are often a function of
119
120 James Chen Yong Kah
aggregation state. The colloidal stability of non-plasmonic

nanoparticles made of materials such as carbon, chitosan, mag-
netic, and silica can be probed by their UV–Vis spectrum [10–12],
zeta potential measurements [10, 13], hydrodynamic diameter
based on dynamic light scattering [10, 13–16], or simply by
observing their precipitation times for larger and heavier particles
[13, 17].
Here, we develop a quantitative Aggregation Index (AI) as a
descriptor to characterize the colloidal stability of plasmonic
nanoparticles based on their optical absorption, specifically using
gold nanorods coated with different amphiphilic ligands as an
example [18]. The localized surface plasmon resonance (LSPR) of
gold nanorods in the UV–Vis spectrum is highly sensitive to their
aggregation state due to interaction of the plasmons between
nanorods when they are in proximity. Aggregation of gold nanorods
result in a red-shift and spectral broadening of the LSPR and the
AI is a measure of the LSPR spectral broadening. Physically, it
describes the equivalent bandwidth of the LSPR (with units of nm)
for a spectral area normalized to the LSPR peak absorbance.
A higher degree of aggregation corresponds to a higher AI value.
The AI is concentration independent since it is normalized to the
LSPR peak absorbance which varies with gold nanorods
concentration.
While covalently attached ligands are predominantly used for
nanoparticle passivation, amphiphiles are attractive due to the fact
that they can be chemically tailored, and are versatiles, enabling a
range of applications in delivery and therapy. Amphiphilic ligands
have been used to passivate gold nanorods to mediate cytotoxicity
and enable transfection of cells. The choice of surface ligands
affects the colloidal stability of gold nanorods, which in turn affect
a broad range of biological processes. Therefore, a quantitative AI
would provide valuable indication of their effectiveness in passivat-
ing gold nanorods. We functionalized the surface of gold nanorods
with four types of amphiphilic ligands having different headgroup
charge (cetyltrimethylammonium bromide (CTAB): positive, poly-
oxyethylene [10] cetyl ether (Brij56): neutral, Oligofectamine™
(OF): positive and phosphatidylserine (PS): negative), to create
gold nanorods with different surface charge. We then disperse
these nanorods in biologically relevant media and demonstrate
how we can develop an aggregation assay to probe their colloidal
stability in these media.
2 Materials
Prepare all solutions of analytical grade reagents using MilliQ

water with a resistivity of 18.2 MΩ cm at 25 °C. Otherwise
stated, all reagents used in the preparation were used as received.
Stability and Aggregation Assays 121
All waste disposal regulations should be strictly adhered to when

disposing waste materials.
2.1 Synthesis of Gold 1. 200 mM Cetyltrimethylammonium bromide (CTAB): Dissolve

Nanorods 1.458 g CTAB in 20 mL water (see Note 1).
2. 100 mM Sodium chloride (NaCl): Dissolve 58.44 mg NaCl in
10 mL water.
3. 50 mM Gold (III) chloride trihydrate (HAuCl4): Dissolve
196.9 mg HAuCl4 in 10 mL water (see Notes 2 and 3).
4. 10 mM Silver nitrate (AgNO3): Dissolve 16.99 mg AgNO3 in
10 mL water (see Note 3).
5. 100 mM l-Ascorbic acid (AA): Dissolve 176.1 mg AA in
10 mL water (see Note 4).
6. 312.5 μM Sodium borohydride (NaBH4): Dissolve 1.182 mg
NaBH4 in 10 mL water. Dilute 10× by adding 100 μL of
3.125 mM NaBH4 in 900 μL water (see Note 5).
7. Temperature controlled water bath.
2.2 Ligand Exchange 1. Amicon ultra–0.5 mL centrifugal filters unit using regenerated
with Amphiphilic cellulose (ultracel-100) membrane with MWCO = 100,000.
Ligands 2. 0.1 mM CTAB: Prepare 10 mM CTAB by dissolving 36.45 mg
CTAB in 10 mL water and dilute the resulting solution 100×
by adding 100 μL of 10 mM CTAB to 9.9 mL water.
3. 10 mM polyoxyethylene [10] cetyl ether (Brij56): Dissolve
68.3 mg Brij56 in 10 mL water (see Note 6).
4.
Oligofectamine™ (OF) (catalog number 12252–011)
(Invitrogen, Inc.) (see Note 7).
5. 20 mM phosphatidylserine (PS) (catalog number 840032P–
25 mg) (Brain PS l-α-phosphatidylserine (Brain, Porcine,
sodium salt): Dissolve 25 mg PS in 1.54 mLwater (see Note 8).
2.3 Stability Analysis 1. Phosphate buffered saline (PBS; 1×): 137 mM NaCl, 2.7 mM
in Biological Media KCl, 10 mM Na2HPO4•2H2O, 2 mM KH2PO4, pH 7.4.
2. RPMI-1640 medium (catalog number 30–2001).
3. Cary 100 UV–Visible spectrophotometer.
3 Methods
Carry out all procedures at room temperature, unless otherwise

specified.
3.1 Synthesis of Gold 1. Prepare the 200 mM CTAB, 100 mM AA, and 312.5 μM
Nanorods NaBH4 fresh before each gold nanorods synthesis. Cool the
NaBH4 to 4 °C.
2. Mix the reagents in the following order: 16.67 mL of 200 mM

CTAB, 2.3 mL water, 300 μL of 100 mM NaCl, 240 μL of
50 mM HAuCl4. Solution turns turbid orange upon addition
of HAuCl4. Swirl solution gently for 3 min and the solution
will turn clear orange (see Note 9).
3. Add 240 μL of 10 mM AgNO3, followed by 200 μL of 100 mM
AA. Swirl the mixture briefly as the solution turns from orange
to colorless to indicate the reduction of Au3+ to Au+ by AA.
4. Add 128 μL of 312.5 μM NaBH4 to the mixture and swirl
gently to ensure that the NaBH4 is well mixed. Put the mixture
immediately in water bath (≈26 °C) and leave it overnight for
the gold nanorods to form (see Note 10).
5. Aliquot 1 mL of the gold nanorods colloid into 1.5 mL centri-
fuge tubes and centrifuge them at 12,000 rpm (13,400 × g) for
30 min to remove the excess reactants.
6. Resuspend the pellet in 1 mL of water and store the resulting
gold nanorods passivated with CTAB (NR-CTAB) at room
temperature (see Note 11).
3.2 Ligand Exchange 1. Centrifuge 1 mL of gold nanorods colloid at 12,000 rpm

with Amphiphilic (13,400 × g) for 30 min. Remove the supernatant and resus-
Ligands pend pellet in 1 mL of 0.1 mM CTAB.
2. Repeat the centrifugation at 12,000 rpm (13,400 × g) for
30 min. Remove the supernatant and resuspend pellet in
500 μL of 10 mM Brij56 to replace the CTAB ligands on gold
nanorods with Brij56.
3. Vortex the mixture briefly and sonicate for 1 min before incu-
bating it for 1 h at 37 °C to allow the ligand exchange to take
place.
4. Remove excess Brij56 by centrifuging the gold nanorods using
Amicon ultra-0.5 mL centrifugal filters at 5,000 rpm (2,300 × g)
for 15 min (see Note 12). Discard the filtrate containing excess
Brij56 and recover the pellet (≈30 μL) into a new centrifuge
tube.
5. Resuspend the pellet in 470 μL of water and repeat the cen-
trifugation at 5,000 rpm (2,300 × g) for 15 min. Discard the
filtrate and recover the pellet (≈30 μL) into a new centrifuge
tube. Add 470 μL of water to the pellet to give 500 μL of gold
nanorods passivated with Brij56 (NR-Brij56) and store the
solution at room temperature.
6. To prepare gold nanorods passivated with OF (NR-OF), add
50 μL of OF reagent to 30 μL of the NR-Brij56 pellet obtained
in step 5. Vortex the mixture briefly and sonicate for 1 min
before incubating the mixture overnight at 37 °C to allow the
ligand exchange from Brij56 to OF to take place (see Note 13).
7. To prepare gold nanorods passivated with PS (NR-PS), add

100 μL of 20 mM PS to 30 μL of the NR-Brij56 pellet obtained
in step 5. Vortex the mixture briefly and sonicate for 1 min
before incubating the mixture overnight at 37 °C to allow the
ligand exchange from Brij56 to PS to take place (see Note 13).
8. Remove the excess OF and/or PS by centrifuging the gold
nanorods using Amicon ultra-0.5 mL centrifugal filters at
5,000 rpm (2,300 × g) for 15 min. Discard the filtrate contain-
ing excess OF or PS and recover the pellet (≈30 μL) into a new
centrifuge tube.
9. Resuspend the pellet in 470 μL of water and repeat the cen-
trifugation at 5,000 rpm (2,300 × g) for 15 min. Discard the
filtrate and recover the pellet (≈30 μL) into a new centrifuge
tube. Add 470 μL of water to the pellet to give 500 μL of gold
nanorods passivated with OF (NR-OF) or PS (NR-PS) and
store the solution at room temperature.
3.3 Stability Analysis 1. Warm the 1× PBS and RPMI-1640 medium to room
in Biological Media temperature.
2. Add 50 μL of the gold nanorods passivated in different amphi-
philic ligands (NR-CTAB, Brij56, OF and PS) as prepared in
Subheading 3.2 to 450 μL of 1× PBS and 450 μL of RPMI-
1640 medium in a centrifuge tube (see Note 14).
3. Vortex the mixture briefly and sonicate for 1 min before incu-
bating it for 2 h at 37 °C.
4. Acquire the UV–Visible absorption spectrum of the amphiphi-
lic ligand coated gold nanorods in 1× PBS and RPMI-1640
medium from 400 nm to 900 nm.
3.4 Measuring the 1. Calculate the total area under the absorption spectrum of
Aggregation Index (AI) the gold nanorods LSPR from λmin = 600 to λmax = 900 nm
(see Note 15). This can be done by applying the trapezium
rule to sum the area of trapezoidal strips from the raw spectral
data as given by equation below:
Area ≈
h
2
( )
I l ,min + I l ,max + 2 I l ,min + h + I l ,min + 2h + + I l ,max −h 

where h = sampling interval of the scan in nm

Iλ,min = Absorbance at wavelength of λmin
Iλ,min+h = Absorbance at wavelength of (λmin+h)
Iλ,max = Absorbance at wavelength of λmax
2. Divide the total area determined in step 1 by the LSPR peak
absorbance, Iλ,max (Fig. 1) to obtain the AI of the various
amphiphilic ligand coated gold nanorods of interest.
Fig. 1 Derivation of the Aggregation Index (AI) from the UV–vis absorption
spectrum of the gold nanorods
4 Notes
1. We found that by warming the solution to 40 °C would aid the

dissolution of CTAB.
2. HAuCl4 is highly acidic and would corrode metal spatula. Use
either a plastic spatula or a metal spatula wrapped with a layer
of parafilm when weighing out the HAuCl4.
3. The stock solutions of both HAuCl4 and AgNO3 should be
stored in the dark at room temperature.
4. Use the 100 mM l-Ascorbic acid within 15 min of preparation
to ensure that the reduction of Au3+ to Au+ proceeds well to
yield good synthesis results.
5. NaBH4 oxidizes easily at room temperature and loses it reduc-
ing potential with time. Use the 312.5 μM NaBH4 within
15 min of preparation to ensure that NaBH4 is sufficiently
active to reduce the HAuCl4. The NaBH4 should be cooled to
4 °C just before its use for synthesis to slow the reduction reac-
tion and yield good synthesis results.
6. Dissolve Brij56 in a water bath under sonication at 40 °C for
about 40 min. Swirl the mixture occasionally to aid the disso-
lution of the transparent gel-like Brij56. Prepare the Brij56
fresh for each ligand exchange.
7. Do not freeze the OF, but store it at 4 °C. Let the OF warm to
room temperature before proceeding with the ligand exchange.
8. Lipids composed of fatty acids containing one or more double
bonds such as PS are not stable as powders. These lipids are
extremely hygroscopic as powders and will quickly absorb
moisture and become gummy upon opening the container.
This could result in hydrolysis or oxidation of the materials.

Therefore, the PS should be prepared fresh and used just
before each ligand exchange.
9. The prepared CTAB must be cool to ∼30 °C before using it
for synthesis. If the temperature is too high, it will increase the
rate of reduction of Au3+ and hence the likelihood of forming
spherical gold nanoparticles instead of gold nanorods. If the
temperature is too low, the CTAB will crystallize out of the
solution leaving a lower concentration of CTAB in the solution
for the synthesis process. This will affect the quality of gold
nanorods formed.
10. We found that by leaving the gold nanorods solution at a slightly
elevated temperature of >25 °C would prevent the CTAB from
crystallizing out of the solution the following day. This would
result in a more consistent and reproducible quality of the gold
nanorods in terms of their CTAB passivation and optical proper-
ties. It is also crucial that movement or vibration of the mixture
is minimized to ensure proper growth of the gold nanorods.
11. The size of the gold nanorods can be determined from
Transmission Electron Microscopy (TEM) and their concen-
tration can be determined from optical absorption. Using the
protocol as described, the gold nanorods are synthesized with
a size of ∼42 × 11 nm, with a typical concentration of ∼1.5 nM.
The washed gold nanorods can be stored at room temperature
and will remain stable for several months.
12. The gold nanorods require a high centrifugation speed to sep-
arate them from the supernatant. However, we found that a
high centrifugation speed >5,000 rpm after ligand exchange
with Brij56 causes the nanorods to aggregate. The use of
Amicon ultra-0.5 mL centrifugal filters allow us to separate the
gold nanorods from supernatant by centrifuging at lower speed
in a shorter time to minimize aggregation of the nanorods.
13. The use of Brij56 as an intermediate ligand is necessary for
ligand exchange to OF and PS. Direct addition of OF and PS
to NR-CTAB resulted in irreversible aggregation for PS and
poor ligand exchange for OF.
14. To assess the colloidal stability of gold nanorods in other types
of media, simply disperse 50 μL of the gold nanorods colloid
in 450 μL of the media of interest.
15. This wavelength range is determined based on a gold nanorods
LSPR peak of 800 nm. In general, for gold nanorods having
other LSPR peak (λpeak), the lower limit of the wavelength range
(λmin) is determined from the minimum point in the spectrum
between the transverse peak and longitudinal peak. The upper
limit of the wavelength range (λmax) is determined from the rela-
tionship λmax = λpeak + (λpeak − λmin) or the upper wavelength scan
limit of the spectrophotometer, whichever is lower.
Acknowledgments
This work was supported by the NSF grant DMR #0906838.
References
1. Alkilany AM, Murphy CJ (2010) Toxicity and nanoparticles under biologically relevant con-
cellular uptake of gold nanoparticles: what we ditions. J Nanosci Nanotechnol 8:6260–6265
have learned so far? J Nanoparticle Res 11. Heister E, Neves V, Lamprecht C, Silva SRP,
12:2313–2333 Coley HM, McFadden J (2012) Drug loading,
2. Boldt K, Bruns OT, Gaponik N, Eychmuller A dispersion stability, and therapeutic efficacy in
(2006) Comparative examination of the stabil- targeted drug delivery with carbon nanotubes.
ity of semiconductor quantum dots in various Carbon 50:622–632
biochemical buffers. J Phys Chem B 110: 12. Santander-Ortega MJ, Peula-Garcia JM,
1959–1963 Goycoolea FM, Ortega-Vinuesa JL (2011)
3. Noh M, Kim T, Lee H, Kim CK, Joo SW, Lee Chitosan nanocapsules: Effect of chitosan
K (2010) Fluorescence quenching caused by molecular weight and acetylation degree on
aggregation of water-soluble CdSe quantum electrokinetic behaviour and colloidal stability.
dots. Colloids Surf a-Physicochem Eng Aspects Colloids Surf B Biointerfaces 82:571–580
359:39–44 13. Park JY, Choi ES, Baek MJ, Lee GH (2009)
4. Yang P, Murase N, Yu JH (2011) SiO(2) beads Colloidal stability of amino acid coated
with quantum dots: preparation and stability magnetite nanoparticles in physiological fluid.
investigation for bioapplications. Colloids Surf Mater Lett 63:379–381
a-Physicochem Eng Aspects 385:159–165 14. Fang C, Bhattarai N, Sun C, Zhang MQ
5. de Puig H, Federici S, Baxamusa SH, Bergese (2009) Functionalized nanoparticles with
P, Hamad-Schifferli K (2011) Quantifying the long-term stability in biological media. Small
nanomachinery of the nanoparticle- 5:1637–1641
biomolecule interface. Small 7:2477–2484 15. Lin YS, Abadeer N, Haynes CL (2011)
6. Ferhan AR, Guo LH, Kim DH (2010) Stability of small mesoporous silica nanoparti-
Influence of ionic strength and surfactant con- cles in biological media. Chem Commun 47:
centration on electrostatic surfacial assembly 532–534
of cetyltrimethylammonium bromide-capped 16. Wiogo HTR, Lim M, Bulmus V, Yun J, Amal
gold nanorods on fully immersed glass. R (2011) Stabilization of magnetic iron oxide
Langmuir 26:12433–12442 nanoparticles in biological media by fetal
7. MacCuspie RI (2011) Colloidal stability of sil- bovine serum (FBS). Langmuir 27:843–850
ver nanoparticles in biologically relevant con- 17. Cirtiu CM, Raychoudhury T, Ghoshal S,
ditions. J Nanoparticle Res 13:2893–2908 Moores A (2011) Systematic comparison of
8. Sethi M, Joung G, Knecht MR (2009) Stability the size, surface characteristics and colloidal
and electrostatic assembly of Au nanorods for stability of zero valent iron nanoparticles pre-
Use in biological assays. Langmuir 25:317–325 and post-grafted with common polymers.
9. Stebounova LV, Guio E, Grassian VH (2011) Colloids Surf a-Physicochem Eng Aspects
Silver nanoparticles in simulated biological 390:95–104
media: a study of aggregation, sedimentation, 18. Kah JC-Y, Zubieta A, Saavedra RA, Hamad-
and dissolution. J Nanoparticle Res 13:233–244 Schifferli K (2012) Stability of gold nanorods
10. Chen ZP, Zhang Y, Xu K, Xu RZ, Liu JW, Gu passivated with amphiphilic ligands. Langmuir
N (2008) Stability of hydrophilic magnetic 28:8834–8844
Chapter 10
Electrochemical Measurements of DNA Melting

on Surfaces
Irina Belozerova, Dongbiao Ge, and Rastislav Levicky
Abstract
Thermal denaturation, or melting, measurements are a classic technique for analysis of thermodynamics of
nucleic base driven associations in solution, as well as of interactions between nucleic acids and small mol-
ecule ligands such as drugs or carcinogens. Performed on surface-immobilized DNA films, this well-
established technique can help understand how energetics of surface hybridization relate to those in
solution, as well as provide high-throughput platforms for screening of small molecule ligands. Here we
describe methods for measuring DNA melting transitions at solid/liquid interfaces with focus on the role
of immobilization chemistry, including a common “immobilization-through-self-assembly” approach that
is effective at moderate temperatures, and a thermo-stable approach based on polymer-supported DNA
monolayers that can be used at elevated temperatures. We also discuss conditions necessary for reversible
measurements, as signified by superimposition of the association (cooling) and dissociation (heating) tran-
sitions of immobilized DNA strands.
Key words DNA melting, Surfaces, Electrochemistry, Monolayers, Stability, Thermodynamics
1 Introduction
Temperature-induced denaturation of double-stranded nucleic

acids, or “melting” [1], has been widely used for analysis of bind-
ing between small molecule ligands and DNA [2–7] as well as for
investigations of factors affecting the stability of nucleic acid sec-
ondary structures such as nucleotide mismatches, ionic strength,
and presence of chaotropic agents [8–16]. Adaptation of DNA
melting to solid supports [17–33] enables affordable, time-
effective high-throughput analysis as well as provides a valuable
tool for fundamental study of how interactions of nucleic acids in
interfacial environments compare to their better understood
solution behaviors.
127
128 Irina Belozerova et al.
Surface melting transitions can be performed by heating a

double-stranded DNA layer in which one strand of each duplex,
the “probe” strand, is immobilized by one end to the solid sup-
port while the other, the “target” strand, is free to diffuse into
solution once melted off. Nonequilibrium melting results when
there is an imbalance between the reverse (i.e., dehybridization)
and forward (i.e., hybridization) reaction rates; this condition
can be exacerbated by heating too fast or by maintaining a low
solution concentration of targets. Data from such experiments
are not suited to quantitative extraction of thermodynamic
parameters because of the nonequilibrium conditions of mea-
surement. Equilibrium melting curves result when the surface
population of bound targets remains always in equilibrium with
the free target species in solution. Equilibrium melting transi-
tions produce superposition of heating (dissociation) and cool-
ing (hybridization) curves, so that the same sequence of surface
states is reproduced as a function of temperature whether the
states are reached by heating or cooling. Data obtained under
equilibrium conditions can be used to derive thermodynamic
parameters of probe-target hybridization, as well as those of
small ligand binding.
DNA melting curves on solid supports have been measured by
a variety of techniques, including surface plasmon resonance (SPR)
[17–21], fluorescence [22–29], dynamic light scattering [30], and
electrochemical methods [31–33]. In some cases reversibility of
the transitions was verified by superposition of heating and cooling
traces [19], though often such verification of equilibrium was
omitted possibly due to difficulties with maintaining the probe
layer structure at elevated temperatures.
Here we describe electrochemical methods for monitoring
surface DNA helix–coil transitions under near equilibrium condi-
tions. Two surface chemistries, both applicable to gold electrode
supports, are described: (1) self-assembled (SA) DNA monolay-
ers, in which probes attach to the electrode directly via thiolate
bonds, and which are suitable for service at up to 50 °C, and
(2) polymer-anchored (PA) DNA monolayers, in which probes
are attached through thioether bonds to preassembled nanome-
ter-thick films of poly(mercaptopropyl)methylsiloxane (PMPMS)
on the electrode (Fig. 1). In method (2), the polymer mercapto-
propyl side chains form multivalent anchoring that is further rein-
forced through cross-linking of the PMPMS film before covalent
immobilization of the DNA probes through thiol-maleimide con-
jugation. We have used the PMPMS chemistry at temperatures of
up to 90 °C.
DNA Melting on Surfaces 129
Fig. 1 DNA immobilization. Left: self-assembled DNA monolayers; right: polymer-

anchored DNA monolayers
2 Materials
2.1 DNA 1. Disulfide-modified DNA oligonucleotide probe (Eurofins

Oligonucleotides and MWG Operon or Integrated DNA Technologies), purified by
Electroactive Label reverse phase high-performance liquid chromatography
(HPLC) (see Note 1). Store probe DNA aqueous solutions at
−20 °C.
2. Single-stranded, amine end-modified DNA target oligonucle-
otides (Eurofins MWG Operon or Integrated DNA
Technologies, with HPLC purification).
3. Amine-reactive electroactive ferrocene label: N-succinimidyl
ferrocenecarboxylate (FcCA, TCI America).
4. Ferrocene-labeled target oligonucleotides. Store dry at −20 °C
(see Note 2).
2.2 Buffers and All solutions are prepared by dissolving dry reagents in deionized
Solutions water with a purity corresponding to a resistivity of 18.2 MΩ cm or
higher at 25 °C.
1. Electrode polishing solution: (1) 0.5 M sulfuric acid, 10 mM
potassium chloride; (2) 0.5 M sulfuric acid.
2. Probe deposition solution: typically 1–2 μM of disulfide-
terminated (SA layers) or thiol-terminated (PA layers) probe in
0.5 M sodium phosphate buffer (SPB), at pH 7.0–7.4.
3. Electrode passivation solution for SA layers: typically 0.5–
1.0 mM 6-mercapto-1-hexanol (MCH, 97 % purity, Sigma-
Aldrich) in 0.1 M SPB, pH 7.0–7.4. Prepare fresh passivation
solutions daily.
4. PMPMS deposition solution for PA layers: 10 mM (by

monomer residue) of poly(mercaptopropyl)methylsiloxane
(PMPMS, degree of polymerization 40, purity 95 %, Gelest,
Inc.) in toluene.
5. Disulfide reducing solution: typically 10 mM dithiothreitol
(DTT) in 10 mM tris(hydroxymethyl)aminomethane, 1 mM
ethylenediaminetetraacetic acid, pH 8.0 (TRIS–EDTA buf-
fer). The reducing solution is typically used for 100 μM con-
centrations of disulfide probe DNA.
6. HPLC solvent: 12–60 % methanol linear gradient spread over
22 min in 100 mM hexafluoroisopropanol, 4.5 mM triethyl-
amine aqueous buffer, pH 8.0.
7. PMPMS cross-linking solution: 5 mM 1,11-bis-
maleimidotriethyleneglycol (BM(PEG)3, Pierce Biotechnology)
in dichloromethane (DCM).
8. Measurement (for melting curves) buffer: typically 0.1–1.0 M
SPB (pH 7.0–7.4), 25–50 nM FcCA-labeled target, 0.1 mM
MCH (SA layers only) (see Note 3).
2.3 Electrochemical 1. Working electrode: 3 mm diameter polycrystalline gold rotat-

Supplies and ing disk electrode (RDE) or 1.6 mm diameter stationary gold
Equipment electrode; counter electrode: coiled platinum wire; reference
electrode: silver/silver chloride reference electrode (Ag/
AgCl/3 M NaCl).
2. Electrode polishing accessories: 1 μm diamond slurry, 0.05 μm
polishing alumina, nylon and velvet polishing pads (Bioanalytical
Systems, Inc.).
3. Equipment: CH Instruments 660C workstation or equivalent;
Bioanalytical Systems RDE-2 rotating electrode station or
equivalent; jacketed RDE or stationary electrode glass cell
(e.g., from Bioanalytical Systems, Inc.).
4. Temperature control: refrigerated water bath (RTE-740
Digital Plus, Thermo Scientific NESLAB), with software tem-
perature control.
3 Methods
3.1 Electrode Mechanically polish gold disk electrodes with 1 μm diamond slurry
Cleaning on a nylon polishing pad for 2 min, rinse the electrodes with copi-
ous amounts of methanol followed by deionized water (18.2
MΩ cm), and then sonicate them in methanol and water for 3 min
each (e.g., Ultrasonic Cleaner FS20D, Fisher Scientific). Repeat
polishing on a velvet pad using 0.3 μm alumina powder and the
same sonication procedure (see Note 4). Assemble the electro-
chemical cell with the platinum auxiliary/counter electrode, the
silver/silver chloride reference electrode, and the mechanically
polished working electrode, and immediately perform electro-

chemical polishing in electrode polishing solution #1 for 10 cyclic
voltammetry cycles between −0.2 V and 1.75 V (potentials are
quoted vs. Ag/AgCl/3 M NaCl reference) at a scan rate 0.1 V/s,
followed by polishing in electrode polishing solution #2 for 50
cycles from −0.2 V to 1.75 V at 0.1 V/s. Assess good polishing
during potentiodynamic cycling in polishing solution #2 by observ-
ing superimposition of the cyclic voltammograms. Rinse the elec-
trodes with deionized water. If desired, determine the roughness
factor r (r = actual area/geometric area) from the double-layer
capacitance in 0.1 M sodium fluoride at −0.84 V as reported previ-
ously [34]. Keep the electrode surface wet after electrochemical
polishing to avoid surface contamination.
3.2 SA DNA Layer Form the DNA layer by immersing freshly polished and rinsed
Formation with deionized water electrodes into the probe deposition solution
at 45 °C for 90 min. Prevent solution evaporation by sealing the
solution container housing the electrode (e.g., microcentrifuge
tube). Passivate the remaining electrode surface with MCH via
immersing into the SA electrode passivation solution at 45 °C
overnight (see Note 5).
3.3 PA DNA Layer Wash the electrodes with deionized water, completely dry them
Formation under a nitrogen stream, rinse them with anhydrous toluene sol-
vent, and transfer them, while still wet, into PMPMS deposition
3.3.1 Preparation of
solution for 1 h (see Note 6). After chemisorption of PMPMS,
PMPMS-Coated Electrodes
rinse the modified electrodes extensively with toluene and dry
them under a compressed nitrogen stream.
3.3.2 Preparation of 1. Reduce disulfide probe to its thiol terminus by treating the
PMPMS-DNA Layers probe DNA with the disulfide reducing solution for 2.5 h.
After reduction, pass the solution mixture once through a
NAP-10 (GE Healthcare) separation column, followed by final
purification using reverse phase HPLC (e.g., Clarity 3 μm
Oligo RP column from Phenomenex) with the HPLC solvent
gradient at a flow rate of 0.5 mL min−1.
2. Prepare PMPMS-anchored DNA probe monolayers by first
immersing PMPMS-coated disk electrodes in PMPMS cross-
linking solution under stirring for 5 h (see Note 7), followed
by thorough rinsing with DCM and drying under compressed
nitrogen. Immediately place the dried, maleimide-activated
electrodes into probe deposition solution at room temperature
overnight. After probe immobilization, rinse electrodes with
deionized water and SPB prior to use.
3.4 DNA Following layer preparation and electrode rinsing, assemble

Hybridization-Melting the electrochemical jacketed cell with the platinum auxiliary elec-
trode and the Ag/AgCl reference electrode. If desired, first set up
the reference electrode in a double junction configuration by
inserting it into a glass tube with a porous Vycor tip that has been
filled with the same measurement buffer as the cell. This serves to
limit Cl− leakage from the reference electrode’s internal reservoir
into the cell. Fill the cell with the measurement buffer, after first
deoxygenating the buffer by bubbling nitrogen through it for
20 min. Maintain a nitrogen blanket over the measurement solu-
tion throughout the experiment (see Note 8). Connect the cell to
the thermostatted water bath for temperature control and preheat
the cell to high limit temperature at which no hybridization is
observed (see Note 9). Immerse the working electrode into the
measurement solution. Set the RDE rotation speed to 1,500 rpm
(see Note 10). Start the cooling cycle at a sufficiently slow tem-
perature scan rate to ensure reversible operation (see Note 11).
Following each temperature step, perform a cyclic voltammetry
(CV) scan. Typical CV settings are a scan rate of between 1 and
20 V/s and over a potential range sufficient to fully measure the
ferrocene electroactivity, e.g., 0 V to 0.75 V for FcCA. It is advis-
able to control the surface potential between CV scans to ensure
that ferrocene labels remain unoxidized during these wait times,
for example by keeping the electrode at 0 V vs. the Ag/AgCl/3 M
NaCl reference. After the cooling scan, perform a heating scan fol-
lowing analogous procedures but increasing rather than decreasing
the temperature between CV scans. From each CV trace, obtain an
unnormalized coverage of hybridized targets (i.e., probe-target
duplexes) by integrating the area of the ferrocene peaks. Plot
the integrated areas vs. temperature to obtain the melting curve
(Fig. 2). If desired, the peak areas can be converted to quantitative
coverages (targets/area) by dividing the peak areas by the CV scan
rate, the elementary charge of one electron, and the electrode area
[35] (Fig. 3). Examine superimposition of cooling and heating
transitions to confirm that equilibrium was maintained during
measurement.
4 Notes
1. In our experience, for self-assembled monolayers, shorter

oligonucleotide lengths that melt below 50 °C, e.g., 12-mers, are
compatible with this immobilization chemistry. Polymer-anchored
monolayers employ longer oligonucleotides, e.g., 22-mers, for
which melting process is completed below 90 °C. Our probes are
typically modified with a disulfide modification at the 5′ end,
which can be connected to the rest of the molecule by an alkyl,
oligothymine, or other linkers to provide elevation above the
substrate and to improve strand accessibility for hybridization
with targets.
2. Procedures for labeling oligonucleotides with electroactive
tags are described elsewhere [35, 36]. It is advantageous to
Fig. 2 Left: cyclic voltammograms of self-assembled DNA monolayers hybridized

to ferrocene-labeled target at various temperatures. Right: CV peak area as a
function of temperature, tracing out the melting transition of the immobilized DNA
Fig. 3 Hybridization (cooling) and melting (heating) transitions of polymer-

anchored DNA monolayers in the presence of ferrocene-labeled target, illustrat-
ing good superposition of the cooling and heating scans
place the label such that, when hybridized, the label is in prox-
imity to the surface to facilitate electron transfer and thus to
improve signal.
3. Aqueous solutions other than SPB, may be used, although
chloride-containing buffers should be avoided to prevent fer-
rocene degradation, via its oxidized ferricenium state, by
nucleophilic Cl−. For SA layers, presence of 0.1 mM MCH in

the experimental solution provides regeneration of the MCH
passivation to counter any MCH desorption that may occur
during thermal cycling.
4. Polishing pads should be thoroughly rinsed before use to
remove any remaining polishing grit and to avoid scratching of
gold by grit clumps. Only one size (and type) of polishing slurry
should be used on each polishing pad (i.e., 1 μm, 0.5 μm).
Sufficient amount of the polishing slurry should be applied
upfront, adding additional slurry once polishing has been
started is not recommended. The electrode has to be positioned
perpendicularly to the pad surface and polishing motions per-
formed in circular fashion (with alternating clockwise/counter-
clockwise direction), or using figure-eight motions, in order to
avoid uneven polishing and erosion of the metal below the sur-
rounding plastic coating. The surface of the electrode is not to
be touched and should be kept wet with a drop of deionized
water during transfers between all cleaning solutions so as to
minimize adsorption of ambient contaminants.
5. It is advisable to perform SA layer formation at the high tem-
perature limit in order to expose the forming probe layer to
elevated temperatures so as to decrease formation of less stably
bound probes.
6. We find that a 60 min immobilization time is sufficient to pre-
pare continuous PMPMS anchor films of good quality for sub-
sequent attachment of DNA probes.
7. BM(PEG)3 serves to cross-link unbound PMPMS free thiol
groups for formation of a stable two-dimensional anchor poly-
mer network on the surface. Any unreacted maleimide groups
will remain available for attachment of thiol-terminated DNA
probes. By varying the cross-linking reaction time, prior to
DNA deposition, the probe DNA coverage can be adjusted,
e.g., in our hands, cross-linking time of 1 h yields final DNA
coverage of about 2 × 1012 cm−2, while a 5 h cross-linking reac-
tion yields about 4.0 × 1012 cm−2 probe coverage.
8. Presence of dissolved O2 facilitates degradation of ferrocene
labels; therefore, it is desirable to exclude O2 from the
measurement buffer as much as possible. If proper deoxygen-
ation steps are followed, the O2 concentration in deoxygenated
SPB should be less than about 10 μM at room temperature.
9. Measurements should be initiated at temperatures above the
expected range of melting. Starting at the higher temperatures
promotes structural relaxation of the DNA layer into the
hybridized state as it cools, improving reversibility of the
observed melting transitions.
10. Use of a rotating disk electrode geometry helps overcome mass

transfer limitations associated with transport of the target DNA
to the electrode, as well as improves temperature uniformity.
Similarly, for stationary electrodes, it is advisable to continu-
ously stir the measurement solution via a magnetic stir bar.
11. Cooling and heating can be performed in temperature incre-
ments of about 1–2 °C, followed by 5–15 min equilibration
times at each new temperature. Averaged over time, this cor-
responds to effective scan rates of about 0.1–0.4 °C min−1. It
was observed that the optimal scan rate depends on the solu-
tion ionic strength, i.e., in higher salinity medium faster scan
rates can be applied without causing hysteresis between cool-
ing and heating scans, thus retaining reversibility of measure-
ment (e.g., in 0.1 M SPB a 0.16 °C min−1 scan rate is employed,
while 1.0 M SPB could use a scan rate of 0.30 °C min−1).
Acknowledgments
This work was supported by NIH grant R33HG003089 and NSF

grant DMR 07-06170.
References
1. Doty P (1957) The physical chemistry of 8. Kwok P-Y (2001) Methods for genotyping
deoxyribonucleic acids. J Cell Comp Physiol single nucleotide polymorphisms. Annu Rev
49:27–57 Genomics Hum Genet 2:235–258
2. Demeunynck M, Bailly C, Wilson WD (eds) 9. Howell WM et al (1999) Dynamic allele-
(2003) Small molecule DNA and RNA bind- specific hybridization: a new method for scor-
ers: from synthesis to nucleic acid complexes. ing single nucleotide polymorphisms. Nat
Wiley-VCH, Weinheim Biotechnol 17:87–88
3. Crothers DM (1971) Statistical thermody- 10. Nelson JW, Martin FH, Tinoco I Jr (1981)
namics of nucleic acid melting transitions with DNA and RNA oligomer thermodynamics:
coupled binding equilibria. Biopolymers the effect of mismatched bases on double-helix
10:2147–2160 stability. Biopolymers 20:2509–2531
4. Lazurkin YS, Frank-Kamenetskii MD, Trifonov 11. Sadhu C, Dutta S, Gopinathan KP (1984)
EN (1970) Melting of DNA: its study and Influence of formamide on the thermal stabil-
application as a research method. Biopolymers ity of DNA. J Biosci 6:817–821
9:1253–1306 12. Blake RD, Delcourt SG (1996) Thermodynamic
5. McGhee JD (1976) Theoretical calculations of effects of formamide on DNA stability. Nucleic
the helix-coil transition of DNA in the pres- Acids Res 24:2095–2103
ence of large, cooperatively binding ligands. 13. Spink CH, Chaires JB (1999) Effects of hydra-
Biopolymers 15:1345–1375 tion, ion release, and excluded volume on the
6. Spink CH, Wellman SE (2001) Thermal dena- melting of triplex and duplex DNA.
turation as tool to study DNA-ligand interac- Biochemistry 38:496–508
tions. In: Chaires JB, Waring MJ (eds) 14. Nordstrom LJ et al (2006) Effect of ethylene
Methods in enzymology, vol 340. Academic glycol, urea, and N-methylated glycines on
Press, New York, NY, pp 193–211 DNA thermal stability: the role of DNA base
7. Wilson WD et al (1997) Evaluation of drug- pair composition and hydration. Biochemistry
nucleic acid interactions by thermal melting 45:9604–9614
curves. In: Fox KR (ed) Drug-DNA interac- 15. Spink CH, Garbett N, Chaires JB (2007)
tion protocols. Humana Press, Totowa, NJ, Enthalpies of DNA melting in the presence of
pp 219–240 osmolytes. Biophys Chem 126:176–185
16. Del Vecchio P et al (1999) The effects of poly- oligonucleotide microarrays. Biotechnol Prog
ols on the thermal stability of calf thymus 28:556–566
DNA. Int J Biol Macromol 24:361–369 27. Vainrub A et al (2007) Predicting DNA duplex
17. Peterlinz KA et al (1997) Observation of stability on oligonucleotide arrays. Methods
hybridization and dehybridization of thiol- Mol Biol 382:393–403
tethered DNA using two-color surface plas- 28. Forman JE et al (1998) Thermodynamics of
mon resonance spectroscopy. J Am Chem Soc duplex formation and mismatch discrimination
119:3401–3402 on photolithographically synthesized oligonu-
18. Fuchs J et al (2010) Effects of formamide on cleotide arrays. ACS Sym Ser 682:206–228
the thermal stability of DNA duplexes on bio- 29. Khomyakova E et al (2008) On-chip hybrid-
chips. Anal Biochem 397:132–134 ization kinetics for optimization of gene
19. Fuchs J et al (2010) Salt concentration effects expression experiments. Biotechniques 44:
on equilibrium melting curves from DNA 109–117
microarrays. Biophys J 99:1886–1895 30. Xu J, Craig SL (2005) Thermodynamics of
20. Fiche JB et al (2007) Temperature effects on DNA hybridization on gold nanoparticles.
DNA chip experiments from surface plasmon J Am Chem Soc 127:13227–13231
resonance imaging: isotherms and melting 31. Nasef H, Beni V, O’Sullivan CK (2010)
curves. Biophys J 92:935–946 Electrochemical melting-curve analysis.
21. Fiche JB et al (2008) Point mutation detection Electrochem Commun 12:1030–1033
by surface plasmon resonance imaging coupled 32. Nasef H, Beni V, O’Sullivan CK (2010)
with a temperature scan method in a model Labelless electrochemical melting curve analy-
system. Anal Chem 80:1049–1057 sis for rapid mutation detection. Anal Methods
22. Nasef H et al (2010) Melting temperature of 2:1461–1466
surface-tethered DNA. Anal Biochem 406: 33. Meunier-Prest R et al (2003) Direct measure-
34–40 ment of the melting temperature of supported
23. Russom A et al (2006) Rapid melting curve DNA by electrochemical method. Nucleic
analysis on monolayered beads for high- Acids Res 31:e150
throughput genotyping of single-nucleotide 34. Oesch U, Janata J (1983) Electrochemical
polymorphisms. Anal Chem 78:2220–2225 study of gold electrodes with anodic oxide
24. Watterson JH et al (2000) Effects of oligonu- films- I. Formation and reduction behavior of
cleotide immobilization density on selectivity of anodic oxides on gold. Electrochim Acta
quantitative transduction of hybridization of 28:1237–1246
immobilized DNA. Langmuir 16:4984–4992 35. Tercero N et al (2009) Morpholino monolay-
25. Wang X, Krull UJ (2005) Synthesis and fluo- ers: preparation and label-free DNA analysis by
rescence studies of thiazole orange tethered surface hybridization. J Am Chem Soc
onto oligonucleotide: development of a self- 131:4953–4961
contained DNA biosensor on a fiber optic sur- 36. Ge D, Levicky R (2011) A comparison of
face. Bioorg Med Chem Lett 15:1725–1729 five bioconjugatable ferrocenes for labeling
26. Ozel AB et al (2012) Target concentration of biomolecules. Chem Commun 46:
dependence of DNA melting temperature on 7190–7192
Chapter 11
Formation and Characterization of the

Nanoparticle–Protein Corona
Marco P. Monopoli, Andrzej S. Pitek, Iseult Lynch,
and Kenneth A. Dawson
Abstract
Over the last decade the existence of “the corona,” a natural interface between nanomaterials and living
matter in biological milieu, evolved from a vague concept into broadly recognized fact. This robust shell
arises (to some extent) on the surface of all nanoparticles (NPs), even the ones designed to avoid its forma-
tion upon contact with biological fluids and confers a biological identity to the nanomaterials such that
they can engage with cellular machinery. The NP corona consists of those proteins (and other biomole-
cules such as lipids and sugars) residing on the NP surface for a sufficient timescale to influence the NP’s
properties and interactions with living systems. This chapter aims to provide simple protocols, as well as
notes on potential pitfalls, to help researchers to perform basic experiments in this field as the basis for a
more mechanistic approach to study and understand NP–protein corona complexes. This work has been
supported by INSPIRE (Integrated NanoScience Platform for Ireland) funded by the Irish Government’s
Programme for Research in Third Level Institutions, Cycle 4, National Development Plan 2007–2013,
and 3MICRON (NMP-2009-LA-245572), NAMDIATREAM (NMP4-LA-2010-246479) and QualityNano
(INFRA-2010-262163) funded by the European Commission 7th Framework Programme.
Key words Nanoparticle, Biological fluids, Protein corona, Dynamic light scattering, Differential
centrifugal sedimentation, Z-potential, TEM
1 Introduction
It is now well recognized that nanomaterials exposed to a biological

milieu will strongly interact with biological molecules, including
proteins, sugars, and lipids, forming “a corona” [1–10]. While the
exposed layer of biomolecules rapidly exchanges, an inner layer, the
hard corona, remains associated with NP surface for so long that it
is believed to give a new biological identity to the particle in the
biological environment. Thus the inner hard corona is of high sci-
entific relevance and is the most studied (Fig. 1) [4, 11].
Nanomaterial surface properties, surface curvature (thus
size), material, morphologies, surface modification, etc., will
137
138 Marco P. Monopoli et al.
Fig. 1 (a) Schematic drawing of the NP–protein corona complexes and their
interaction scenarios at the bio–nano interface at cellular level, where the long-
lived protein corona modulates the interaction with cellular receptors while the
pristine NPs surface is not available for binding. (b) Schematic drawing of the
dynamic protein corona in rapid exchange with its environment and of NPs hard
(slowly exchanging) corona. Samples of proteins were re-drawn from the RCSB
PDB (www.pdb.org). Adapted with permission from ref. 4. Copyright (2010)
American Chemical Society
strongly influence the protein corona composition. For exam-

ple, studies performed in human plasma have shown that from
the ~4,000 proteins in the solution, only a few tens of proteins
form the hard corona as their affinity to the NP surface is
extremely high, often despite their low abundance in the bio-
logical fluid [3, 8, 9, 12–19].
Recent studies have shown that it is possible to understand and
characterize the protein corona in more detail using emerging
approaches or adapting older techniques of physical chemistry,
microscopy, and proteomics for use at the nanoscale [6, 20, 21].
The aim of this chapter is to provide simple protocols and
guides and to highlight potential pitfalls in order to help research-
ers to perform basic experiments in this field in a manner that
ensures that all steps are correctly performed. We provide a guide
Protein Corona Preparation and Characterization 139
for all steps, starting from how blood proteins (typically used in
most corona studies, although the approaches are equally appli-
cable to other biofluids such as lung lavage fluid or cerebrospinal
fluid) should be collected and processed, the experimental condi-
tions, how to successfully isolate monodisperse NPs and their
strongly bound hard corona from rich biological complexes, and
all possible precautions that have to be taken to ensure
reproducibility.
Understanding the nature of the NP–protein corona complex,
and its impacts on NP dispersion stability, surface presentation,
and cellular uptake is central to interpreting data from all studies
involving NP interactions with living systems, as it provides key
information on dose–response (Fig. 1) [22–27].
The following sections outline a complete and coherent meth-
odology, which allows researchers to prepare, isolate, and charac-
terize the NP–protein corona complexes and perform
physicochemical characterization of the NP-corona complexes
(with and without excess of proteins).
1.1 NP–Protein The standard procedure for preparation of NP–protein corona

Corona Preparation complexes consists of three main stages:
and Characterization ●● Biological fluid (plasma or serum) preparation.
●● NP incubation with biological fluid.
●● Separation of NP–protein corona complexes from excess
plasma (i.e., proteins not bound to NPs surface) and washes to
remove loosely bound proteins.
Physico-chemical characterization of the NPs in their pristine
state and after exposure to biological media is of the highest impor-
tance for understanding impacts of NPs. The presence of agglom-
erates and multiple populations of dimers, trimers, and larger
agglomerates, and poorly stable samples will lead to different bio-
logical consequences. Analysis of the protein corona without a
well-thought-through approach will lack relevance, as it will inevi-
tably study the average protein corona of the multiple populations
rather than being cognizant of the different subpopulations.
The basic NP dispersion characteristics such as hydrodynamic
diameter, zeta-potential, and polydispersity index in the relevant
medium should be determined both before and after the “corona
prep” procedure, and in a time resolved manner. The comparison
of size and zeta-potential before and after protein corona forma-
tion provides information such as the thickness of the corona
formed and its effect on the NP surface charge due to protein
adsorption, since most NPs are negatively charged in the presence
of a protein corona at physiological pH. Such changes in NP sur-
face charge as a result of protein adsorption might lead to a drop of
the electrostatic stability of the system but may also increase the
steric stability of the system [28].
Dispersion proprieties of the NPs and NP-corona complexes

are generally determined by Dynamic light scattering (DLS) and
by differential centrifugal sedimentation (DCS).
DLS is the technique most commonly used for hydrodynamic
diameter determination, allowing NP size measurement based on
the light scattered by them due to their Brownian motion, which
is strictly dependent on the size. After applying specific correlation
functions, DLS leads to estimation of the NPs’ diffusion constant,
and after application of the Einstein–Stokes equation (below) cal-
culates the NPs’ hydrodynamic radius, Rh.
k BT
D= ,
6πhRh
where:
D—diffusion constant, kb —Boltzmann’s constant, T—abso-
lute temperature, η—viscosity, Rh—hydrodynamic radius of the
particle.
Zeta potential (ZP), which characterizes the net charge of the
particle surface in the given dispersant, provides information
regarding the surface charge of the NPs and the change of surface
charge after incubation in different media.
The DCS Disk Centrifuge is a particle size analyzer for measur-
ing particles in the range of 0.01–40 μm. The analyzer measures
particle size distributions using centrifugal sedimentation, within an
optically clear spinning disk that is filled with fluid. The primary
information from the analytical disk centrifuge is the time taken by
the particles to travel from the center of the disk through a defined
viscous sucrose gradient under a strong centrifugal force, to a detec-
tor placed at the outer rim of the disk. For materials with homoge-
nous density and simple shape (for example, spherical particles) one
can directly relate this time to a particle size [4]. A significant advan-
tage of DCS measurement is that the instrument can successfully
resolve multiple populations over a wide size ranges from a few
nanometers to microns within the same sample. Moreover, it allows
measurement of NPs incubated with fluids of any kind (in situ) as
the biomolecule background will not significantly impact the NP
measurement due to their very different sedimentation times. An
example of the degree of size resolution possible with DCS is given
in Fig. 2, where 100 nm PSCOOH NP–protein complexes (a) in
situ (full plasma) and (b) hard corona complexes were measured
after 1 h (full line) and 6 h (dotted line), where NP corona com-
plexes are shifted from pristine NPs (line with open circles) due to
the increase of size and the change of NPs density.
1.2 Proteomic Another relevant aspect of the protein corona is its composition.
Characterization There are a number of techniques in the field of molecular biology,
of Protein Corona which can be adopted and utilized to determine the identities of
Composition the protein constituents of the corona. One of the most popular
Fig. 2 DCS characterization of nominal “100 nm” carboxylated NPs size and their corona complexes size, after
incubation in human plasma, both in situ (full corona) and after washes (hard corona complexes). Reprinted
with permission from ref. 4. Copyright (2010) American Chemical Society
protein separation techniques is SDS-PAGE, where the proteins

are separated via a polyacrylamide gel by size, which allows for the
comparison of protein bands between different samples. After pro-
tein corona isolation, a typical proteomics shotgun approach
requires the gel lane(s) to be sliced into several bands (typically ten
for a 10 cm standard SDS-PAGE), an in gel trypsin digestion to be
performed, followed by MS analysis following one of several well-
established protocols available in the literature [3, 29, 30]. Since
SDS-PAGE and gel based MS analysis procedures are commonly
available in several laboratories, in this book chapter we describe
potential pitfalls in Notes 28–33 only, as detailed protocols are
easily found in the literature [30].
2 Materials
1. Tubes for blood collection for plasma preparation: BD

Vacutainer whole blood tubes with spray-coated K2EDTA, lav-
ender closures.
2. Tubes for blood collection for serum preparation: BD Increased
Silica Act Clot Activator, red closure.
3. Protein LoBind Tube: Eppendorf.
4. Nanoparticles: various sizes of polystyrene sulfonate/carboxyl-

ate/amino modified fluorobeads, various sizes of silica yellow
green NPs: Life Science, PolyScience, Kisker, Sigma, BBI.
5. Phosphate saline buffer: 0.01 M phosphate buffer, 0.0027 M
potassium chloride and 0.137 M sodium chloride, pH 7.4.
6. Sucrose solutions for gradient preparation (DCS) 2–24 % w/w
sucrose in H2O, PBS (or any NP dispersant media).
7. Dodecane.
8. DLS sizing cuvette: Plastic disposable low volume cuvette.
9. Z-potential measurement cuvette: low volume disposable zeta
cuvette.
10. DCS calibration standard, PVC 0.377 and PS 0.497: Analytics
LTD.
11. Carbon coated copper TEM grids: Agar Scientific.
12. Protein concentration assay: Micro BCA assay: Pierce.
13. EDTA, Sodium Chloride, Acrylamide/Bis-acrylamide 40 %
solution BioReagent, Trizma base, Glycine, Sodium Lauryl
Sulfate (SDS).
14. 3× SDS-PAGE loading buffer: 60 mM Tris–HCl (pH 6.8 at
25 °C), 2 % (w/v) SDS, 0.1 M DTT, 10 % glycerol, 0.01 %
(w/v) bromophenol blue: New England Biolabs.
15. Brilliant blue: Thermo Fisher.
16. Silver staining kit: 2D-Silver Stain-II Cosmo Bio.
17. Silver staining kit: Plos One GE Healthcare.
3 Methods
3.1 Biological Fluid 1. Identify at least six seemingly healthy donors, ensuring an
Preparation: Plasma equal number of male and females.
Preparation 2. Draw blood into EDTA coated Vacutainer tubes, taking 30–50 ml
blood from each donor being sure to collect the full volume for a
correct ratio between blood and EDTA (see Notes 1 and 2).
3. Gently invert the Vacutainer tubes ten times immediately after
collection to ensure mixing of anticoagulant with the blood,
and store the tubes at room temperature prior to the next step.
4. After all blood samples have been collected, centrifuge at
1,200 × g for 10 min at room temperature. Avoid using the
brake to stop the centrifuge in order to avoid sample mixing.
5. Pool the supernatants (the plasma) into 50 ml tubes and dis-
card the pellet that remains at the bottom of the centrifuge
tubes (cells).
6. Centrifuge the pooled plasma sample at 2,500 × g for 15 min at

room temperature.
7. Aliquot the supernatant into cryovials (typically 1 ml) and
place them into −80 °C freezer until use (see Note 3).
8. The whole procedure should not take more than 3 h from the
blood collection to freezing storage.
3.2 Biological Fluid 1. Identify at least six seemly healthy donors ensuring an equal
Preparation: Serum number of male and females.
Preparation 2. Draw blood into Vacutainer tubes for serum preparation tak-
ing 30–50 ml blood from each donor (see Notes 1 and 2).
3. Gently invert each tube five or six times, then incubate in an
upright position at room temperature for 30–45 min (no lon-
ger than 60 min) to allow clotting.
4. Invert tubes to ensure that blood is no longer liquid, indicat-
ing that it is clotted.
5. Centrifuge the tubes at 1,300 × g for 10 min at room tempera-
ture. Avoid using the brakes to stop the centrifuge in order to
avoid sample mixing.
6. Pool the supernatants (serum) into 50 ml tubes and discard
the pellet at the bottom of the tubes.
7. Inspect serum for turbidity. Turbid samples should be centri-
fuged and aspirated again to remove remaining insoluble
matter.
8. Aliquot the supernatant into cryovials (typically 1 ml) and
place them into −80 °C freezer until use (see Note 3).
3.3 NPs Incubation 1. Ensure that biological fluid samples are fully defrosted and
with Biological Fluid vortex briefly to ensure that mixture is homogeneous.
2. Bring the biological fluid of choice to the incubation tempera-
ture (typically 4, 20, or 37 °C) (see Note 5).
3. Spin for 3 min at 16,000 × g at 20 °C to pellet possible p
rotein
aggregates which might have been in the solution. Discard
the pellet and keep the supernatant for the incubation step.
4. Add the biological fluid to the mixing vessel (see Note 6) and
if necessary, add PBS to dilute it to the required protein con-
tent and vortex briefly to ensure homogeneity.
5. Add NPs (see Notes 7–9) and by repeated agitation with use of
pipette, ensure that the sample is well mixed.
6. Put samples in orbital shaker at 250 rpm setting, at the required
incubation temperature and incubate for time required by the
experiment (see Note 10).
Fig. 3 Corona preparation steps. (a) Schematic representation of the different stages of corona formation and
separation. (b) Diagram of optimization step and isolation of NP-hard corona complexes from unbound and
loosely bound proteins. Proteins adapted from Protein Data Bank (http://www.pdb.org)
3.4 Hard Corona 1. Set temperature of the centrifuge to match the incubation
Preparation by temperature (Fig. 3).
Centrifugation 2. Immediately after incubation time has been completed, place
(See Note 11) the NP-containing tubes in a bench centrifuge (see Note 12).
3. Spin the NP–protein complexes down to ensure that all NPs
have been pelleted. The spin velocity and time needs to be opti-
mized according to the NPs size and density (see Note 11).
4. NP-hard Corona complexes should be visible as a pellet at the
bottom of the tube. Gently remove the supernatant without
disturbing the pellet (immediate removal of supernatant pre-
vents NPs from going back into dispersion). Transfer superna-
tant into a new vessel and centrifuge for 2 h at full speed
(centrifugation control). If further NPs pellet is detected after
this step, this indicates that the first centrifugation step has to
be increased (time or centrifugation speed), so discard samples
and start from incubation step, increasing centrifugation speed
or time. If no pellet is detected after the centrifugation control,
proceed to step 5.
5. Resuspend the NP-corona complexes in 1,000 μl of PBS by

mixing with a pipette until solution is homogenous.
6. Spin NP-corona complexes in centrifuge as in previous step at
the centrifugation speed optimized in step 4.
7. Gently remove supernatant (wash 1) and store it at −20 °C and
resuspend pellet as described in step 5.
centrifugation speed optimized in step 4.
9. Gently remove supernatant (wash 2) and store it at −20 °C.
Resuspend pellet as in step 5 and transfer the samples to a new
vessel (see Note 13).
centrifugation speed optimized in step 4.
11. Gently remove supernatant (wash 3) and store it at −20 °C
(see Note 14).
12. Pellet now contains NP-hard corona and it should be pro-
cessed according to the analysis that will be performed. The
hard corona samples should be prepared freshly before each of
the analytical procedures listed in Subheadings 3.5–3.9.
3.5 NP Size 1. Dilute NPs to a final concentration of 50–100 μg/ml in 500 μl

Characterization by of the medium of choice (water, PBS, cell culture medium,
Dynamic Light diluted plasma, etc.). This NPs concentration provides a high
Scattering (DLS) quality of measurement data for most NP kinds/materials (see
(See Notes 15 and 16) Note 17).
2. Set up the SOP according to the instrument manual paying
special attention to the following parameters: Measurement
type: Size; Material: NP material from the list; Solvent: appro-
priate solvent from the list; Temperature: depending on the
experimental conditions; Equilibration time: make sure to set
equilibration time long enough for the sample to reach the
measurement temperature and equilibrate; No of measure-
ments: at least 3; Number of runs: set on auto (minimum 10);
Single run duration: 10 s; Attenuator: in most cases set on
Auto.
3. Run the SOP and read the hydrodynamic diameter (Dh)
(see Note 18). Note standard deviation and average PDI
(Polydispersity index).
3.6 Differential 1. Prepare two sucrose solutions (see Note 20) appropriate for
Centrifugal the specific NPs’ material density: 2 and 8 % (w/w) for parti-
Sedimentation (DCS) cles with density between 1.0 g/cm3 and 1.1 g/cm3, or 8 and
(See Note 19) 24 % (w/w) for particles with density above 1.3 g/cm3.
2. Use one of the existing or set up a new operating procedure in
DCS software, following the manual and paying special attention
to values of: NP material density, standards’ density and size,

and size range of measurement.
3. Inject first portion of sucrose gradient into the disk. Set RPM
value to manual and set the velocity required according to the
NPs’ nominal size and density (see Note 21).
4. Wait until the disk is spinning at full velocity and start injecting
the step gradient following the manufacturer’s procedure.
5. Finish gradient with 0.5 ml of dodecane to prevent water evap-
oration. Leave the instrument for approximately 0.5 h to allow
gradient to equilibrate.
6. Each measurement consists of a calibration run (with use of
standard calibration NPs) and the sample run.
7. To ensure gradient stability, run a test sample twice and com-
pare the results with older measurements. If peaks overlap gra-
dient can be considered as stable, otherwise the test
measurement should be repeated after an additional 15 min of
gradient equilibration. Failing in this measurement requires
preparation of a new gradient (see Note 22).
8. Prepare 100 μl of NPs solution in the dispersant fluid of choice
(water, PBS, or biological fluid) at a concentration of
0.05–1 mg/ml, depending on the NPs’ refractive index and
light adsorption (see Note 23).
9. Run samples (see Notes 24 and 25).
3.7 Z-Potential 1. Dilute NPs to a final concentration of 50–100 μg/ml in 1 ml

of the medium of choice (water, PBS, cell culture medium,
diluted plasma, etc.). This dilution provides high quality mea-
surement data for most NP kinds/materials.
2. Degas the sample by pouring it into a syringe, covering its top
with a cap and creating a vacuum inside (it usually helps to
gently knock the side of syringe). Repeat the procedure until
no visible air bubbles appear during the degassing procedure.
3. Use the syringe to transfer 1 ml of sample into a disposable
zeta potential cuvette (avoid creating air bubbles) and place
cuvette in the instrument.
4. Set up the SOP according to the instrument manual paying
special attention to the following parameters: Measurement
type: Zeta Potential; Material: NPs material from the list;
Solvent: appropriate solvent from the list; Temperature:
depending on the experimental conditions; Equilibration time:
make sure to set equilibration time long enough for the sample
to reach measurement temperature and equilibrate; No of
measurements: at least 3; Number of runs: manual: from 10 to
100; Attenuator: in most cases set on Auto.
5. Run the SOP and read the samples’ ZP value as an average of
three measurements with standard deviation.
3.8 TEM Protocol 1. NP-hard corona complexes (NP / HC) are prepared as described
in Subheadings 3.3 and 3.4. Resuspend the pellet into the
desired concentration (typically 0.01 mg/ml) in 2.5 % glutar-
aldehyde in water and leave for 1 h.
2. Place the TEM grid in a plastic dish, with the shiny carbon side
upwards.
3. Drop cast 10 μl of NP/HC solution onto the grid and remove
the excess solution by soaking with filter/blotting paper. Leave
to dry for 24 h.
4. Place the grid in TEM and carry out the standard operating
procedure analysis (see Note 26).
5. Acquire images at different magnifications from 100 k× to
160 k×, ensuring that at least ten images per area are acquired
for qualitative analysis.
6. For comparison, also acquire images of the pristine NPs.
3.9 Protein Assay 1. Preparation of Albumin standards:

(Micro BCA Protein Take 0.25 ml of 2 mg/ml BSA stock and dilute with
Assay) Optimized for 4.75 ml of PBS buffer to obtain 100 μg/ml standard solution.
NP/Corona Complexes Dilute this standard solution with PBS in order to obtain stan-
dard solutions of: 75, 50, 37.5, 25, 12.5, 6.25, and 0 μg/ml
(pure PBS). Transfer 250 μl of each standard into microcentri-
fuge tubes.
Perform a progressive dilution of the NP corona sample
in PBS to ensure that absorbance values (see step 8) are in the
protein standard range. Transfer 250 μl of each sample into
microcentrifuge tubes.
2. Working reagent (copper solution) preparation: mix reagents
A, B, and C in the ratio 25:24:1.
3. Add working reagent (WR) to the previously prepared stan-
dards and samples in a ratio 1:1, i.e., add 250 μl of WR to each
250 μl sample and standard.
4. Incubate for 60 min at 60 °C.
5. Cool at RT for ~45 min.
6. Spin down the samples at 20,000 × g for 1 h (or the time more
appropriate to NPs from previous studies) to pellet the NPs.
7. Transfer 200 μl of each supernatant (in duplicate) into 96 well
plate and measure absorbance at 562 nm, using a plate reader
spectrophotometer.
8. Plot standard curve of Concentration versus Absorbance and
fit linear function to obtain data. Use the obtained equation to
calculate protein concentration in unknown samples.
9. Controls: NPs in PBS; NPs in BSA standard (see Note 27).
4 Notes
1. Subheading 3.1, steps 1 and 2 follow the HUPO Plasma

Proteome Project guidelines [31].
2. Ethical permission is required in order to collect and prepare
plasma/serum. It collection must be carried out by a qualified
medical doctor and the whole procedure has to be approved by
local ethical approval committee.
3. Storage and monitoring of the aging of the biological fluid is
crucial, as old or partially degraded biological fluids can com-
promise the outcome of the study. Typically, storage at −80 °C
is preferable for long term sample storage (more than 3 weeks)
and the cycle of freezing and thawing must be avoided to
reduce protein degradation. As per the HUPO guidelines,
samples should only be frozen and thawed once [31]. Protease
inhibitor cocktail can be used, however these might result in
protein structure change and artifacts [31].
4. Plasma is generally preferred to serum for studies to predict
NP fate in vivo, as it is representative of most of the proteins
and biomolecules of the blood, however, extreme care is
needed to avoid blood coagulation and protein aggregation.
Any plasma dilution has to be performed in PBS containing
1 mM EDTA.
5. The incubation temperature of NPs in biological fluid is an
important factor that influences the kinetics of protein associa-
tion and dissociation to and from the NP surface. Previous
studies have shown that when NPs are introduced into a bio-
molecule rich medium, their surface will immediately attract
entities from its closest environment—an “early corona”
formed mainly by abundant proteins with low affinity. The
protein corona composition evolves with time until it reaches
equilibrium, where proteins with strong affinity towards the
NP surface form the “mature” protein corona [3, 17, 28, 32].
6. To reduce plasma protein and NPs stickiness to the sides of the
tubes during incubation, use of low protein binding Eppendorf
tubes is highly recommended.
7. Before proceeding with NPs incubation with biological fluid,
ensure that the NPs are stable and monodisperse by common
physicochemical approaches. Bath sonication for three minutes
before use is suggested to disperse any possible agglomerated
NPs.
8. To mimic “in vivo or in vitro” conditions as much as possible,
NPs should be exposed to different biological media with dif-
ferent protein concentrations. A typical approach would be to
keep the NP concentration constant (and thus the surface area)
and incubate with biological fluid diluted with different

amounts of PBS [3], where the ratio between NP surface area
and proteins is varied [3]. A theoretical five layers of protein
excess for each individual NP will ensure full surface coverage
of each individual NP, while a lower protein concentration
generally promotes NPs agglomeration.
9. The NP amount used for each experiment depends on the
detection method that will be used to study the NP-corona
complex. For DLS measurements typically 50–100 μg/ml in
500 μl, for DCS analysis each injection requires 100 μl of 0.1–
0.01 mg/ml, depending on the NP’s absorbance. For protein
corona detection by MS a higher NP amount is generally
required. 1 mg/ml of NPs in 1 ml of solution will allow detect-
ing protein corona by blue Coomassie staining, silver staining
will have to be performed when lower NP amounts have been
used for the experiment.
10. Incubation time and temperature are strictly dependent on the
aim of the study, however possible degradation of the biologi-
cal fluid during the incubation step must be monitored. Also
the biological fluid alteration due to the exposure/incubation
conditions should be monitored.
11. Subheading 3.4 allows isolation of NP hard corona complexes
from unbound biomolecules in the biological media and from
the loosely bound proteins (dynamic corona) (Fig. 3).
Optimization steps are crucial in order to ensure that NP loss
during the centrifugation step is kept to a minimum and that
loosely bound proteins are completely washed off during the
procedure. The hard corona is typically of high biological
interest, while the outer layer of the protein corona is com-
posed of proteins in rapid exchange with the solution.
Centrifugation speed optimization: the first centrifugation is
generally longer than the following ones due to the higher
density of the biological fluid compared to PBS. Centrifugation
time and speed depend on NP size and density and prolonged
centrifugation time (over 1 h) and at speed above 18,000 × g
are generally avoided as this will promote biomolecule pellet-
ing and thus increase of background.
12. The temperature of the incubation step and of the centrifuga-
tion steps, while preparing hard corona samples, should be
similar.
13. To avoid carryover of biomolecules from the biological fluid
used in the incubation step, a change of vessel/tube is strongly
recommended in Subheading 3.4 at least in step 9. Also long
centrifugation times (more than 60 min) and speeds higher
than 18,000 × g should be avoided to prevent co-elution of the
plasma protein background.
Fig. 4 SDS-PAGE of 30 nm gold NP hard corona complexes, after different num-
bers of washing steps. While the abundance of NP corona proteins decreases
after first to third washes (the soft corona is gradually removed), it does not
change after any additional washing steps (the remaining strongly bound hard
protein corona does not elute from the NP surface). Proteins adapted from Protein
Data Bank (http://www.pdb.org)
14. Typically three washes are needed to obtain NP-hard corona

complexes (see Subheading 3.4), however to ensure protocol
efficacy, NP-hard corona washes can be over washed without
resulting in protein corona changes (Fig. 4).
15. DLS is a powerful and fast general-purpose method for char-
acterization of NP dispersions, however its resolution is lim-
ited (due to the fact that the scattering signal generated by
the whole dynamic range of NP sizes present in the sample is
analyzed at the same time) and does not allow for observa-
tion of discrete peaks of similar sizes, or small shifts in size of
a range of nanometer. Therefore, a complementary high
precision size-analysis technique, such as DCS is needed (see
Subheading 3.6).
16. Subheading 3.5 is applicable to pristine NPs in different media
and to NP-hard corona complexes; however, if performing
measurements in situ in the presence of complex biological
media, the background signal of free biomolecules might
interfere with the NPs’ scattering, and as such the background

solution should also be run for comparison. Use of 0.2 μm
filtered solutions is highly recommended to remove dust or
contaminants from solutions and potentially to remove NP
aggregates from solution.
17. The optimal sample concentration for DLS measurements dif-
fers for different materials with different optical properties
(e.g., it is possible to measure polystyrene NP dispersions of
lower concentrations than silica NPs, which have a lower
refractive index and adsorption) therefore it’s difficult to sug-
gest a universal concentration a priori. However, the value of
instrument attenuator when it’s set on “auto” indicates the
intensity of the light scattered by the sample. The higher the
attenuator value (ranges from 1 to 11) the lower the light scat-
tering generated by the sample. NP concentration should be
adjusted to obtain an attenuator value ~7.
18. The reported hydrodynamic diameter (Dh) value should be an
average of at least three subsequent measurements of min.
10 × 10 s runs. For monodisperse samples the average Dh cor-
responds to z-average value, while for complex/polydisperse
samples Dh can correspond to the maximum of a particular size
peak.
19. The advantage of DCS in respect to DLS is the fact, that in the
centrifugal gradient NPs are separated and detected size-by-
size, allowing for resolution of multiple peaks in polydisperse
samples (where DLS would show only one broad peak).
Another advantage of DCS is its capacity to perform measure-
ments in essentially any possible dispersant media, even in the
presence of a large excess of biomolecules, such as in full
plasma, cMEM (and other biological fluids) in situ, which
allows us to have an insight into the NP–protein complexes
properties, as they are in vivo.
20. DCS gradient solutions should be prepared in the buffer in
which the NP samples are prepared/stored. Typically water is
used for characterization of pristine NPs and PBS for NP–pro-
tein corona complexes.
21. RPM of 14,000 should be used for gold NPs of size around
100 nm.
RPM of 18,000 should be used for silica NPs of size around
100 nm.
RPM of 22,000–24,000 should be used for polystyrene NPs of
size around 100 nm.
22. Stability of the DCS gradient has to be monitored with time by
running test samples and ensuring size measurement repro-
ducibility (as in Subheading 3.6, step 7).
23. The detection limit of the DCS instrument is correlated to the

refractive index and light absorption of the NPs. Typically, a
NP concentration of 0.5–1 mg/ml is indicated for particles
with low light absorption (e.g., Silicon dioxide NPs) and 0.05–
0.1 mg/ml for particles with high absorptivity (such as gold or
polystyrene NPs). However, a dilution series experiment has to
be performed to find optimal NPs sample concentration to
ensure a sufficient signal to noise ratio.
24. Change of DCS sucrose gradient is required after measure-
ment of samples injected with an excess of biomolecular fluid
(in situ) or after 6 h of measurements.
25. Following the manufacturer’s procedure, a core shell model
can be applied to analyze the obtained measurement results for
NPs and their hard corona complexes. The core shell model
allows correlation of NPs’ apparent size shifts with the thick-
ness of the biomolecule shell adsorbed on to the NP surface
[4].
26. For high-resolution images, the use of field emission high reso-
lution TEM is recommended, otherwise tungsten filament or
lanthanum hexaboride filament TEM systems can be also used
to achieve acceptable images.
27. The Micro BCA assay is designed for determination of protein
concentration in an unknown sample. In order to evaluate the
degree of NP interference with the accuracy of obtained results,
it’s necessary to perform two control samples: NPs in PBS
(where ideally the assay should produce the result of 0 mg/ml
protein concentration), and NPs in BSA standard of known
concentration (where assay should ideally produce a result in
agreement with the BSA standard concentration).
28. For SDS-PAGE analysis, to ensure that all proteins from the
NP corona are stripped off the NPs’ surface, the use of high
denaturing condition buffer is strongly recommended.
Typically NP protein corona complexes prepared as described
in Subheading 3.4 are resuspended in loading buffer (see
Subheading 2) and heated for 5 min at 100 °C, and samples
are directly loaded onto the SDS-PAGE gel, where they
migrate under an electric field and are separated by size in the
pores of the polyacrylamide gel. In some rare cases more dena-
turing buffer might be applied in order to ensure complete
protein corona desorption from the NPs’ surface.
29. The sample where the biological fluid is prepared and processed
in an identical way to the NP–protein corona samples (see
Subheadings 3.3 and 3.4), in the absence of the NPs, is a con-
trol of highest importance. It is crucial to determine whether
the centrifugation and washing steps promote the isolation of a
plasma protein background. This sample will also have to be

loaded into the SDS-PAGE along with NP–protein corona.
30. Washes obtained as described in Subheading 3.4 should be run
by SDS-PAGE (see Subheading 3.4) to monitor the p resence
of the dynamic corona. A successful protocol should show a
strong presence of proteins in wash 1 that becomes attenuated
in wash 2 and is not present in wash 3. It is recommended to
load in the same gel a protein corona sample, to obtain relative
protein amounts from washes to proteins in the protein corona.
31. To reduce gel to gel variability error, it is recommended to
compare gel bands only within the same gel.
32. The choice of the gel staining technique is crucial: Blue
Coomassie has a dynamic range linear for 20-fold changes of
protein abundance, but the main limitation is the fact that it
fails to detect low abundance proteins; the lowest detection
limit is 10–50 ng, depending on the proteins (Fig. 4). However,
the Coomassie stain protocol is quick to use and simple.
A parallel experiment using a high sensitive staining approach,
such as silver nitrate staining, or a fluorescent staining, is rec-
ommended to ensure that low abundant proteins are detected.
33. It is recommended to run the same sample in different pore
size acrylamide gels (varying the acrylamide percentage) to
ensure high resolution of proteins of different size ranges, as
large proteins (~150–55 kDa) are better resolved in 8 % acryl-
amide gel and smaller proteins (55–10 kDa) are better resolved
in 15 % acrylamide gel.
References
1. Cedervall T, Lynch I, Lindman S, Berggard T, Ray P, Mallampalli RK, Klein-Seetharaman J,
Thulin E, Nilsson H, Dawson KA, Linse S Fadeel B, Star A, Shvedova AA, Kagan VE
(2007) Understanding the nanoparticle– (2012) Adsorption of surfactant lipids by
protein corona using methods to quantify single-walled carbon nanotubes in mouse lung
exchange rates and affinities of proteins for upon pharyngeal aspiration. ACS Nano
nanoparticles. Proc Natl Acad Sci USA 6:4147–4156
104:2050–2055 6. Walkey CD, Chan WC (2012) Understanding
2. Lynch I, Dawson KA (2008) Protein–nanopar- and controlling the interaction of nanomateri-
ticle interactions. Nano Today 3:40–47 als with proteins in a physiological environ-
3. Monopoli MP, Walczyk D, Campbell A, Elia ment. Chem Soc Rev 41:2780–2799
G, Lynch I, Baldelli Bombelli F, Dawson KA 7. Nel AE, Madler L, Velegol D, Xia T, Hoek
(2011) Physical–chemical aspects of protein EM, Somasundaran P, Klaessig F, Castranova
corona: relevance to in vitro and in vivo bio- V, Thompson M (2009) Understanding bio-
logical impacts of nanoparticles. J Am Chem physicochemical interactions at the nano-bio
Soc 133:2525–2534 interface. Nat Mater 8:543–557
4. Walczyk D, Bombelli FB, Monopoli MP, 8. Tenzer S, Docter D, Rosfa S, Wlodarski A,
Lynch I, Dawson KA (2010) What the cell Kuharev J, Rekik A, Knauer SK, Bantz C,
“sees” in bionanoscience. J Am Chem Soc Nawroth T, Bier C, Sirirattanapan J, Mann W,
132:5761–5768 Treuel L, Zellner R, Maskos M, Schild H,
5. Kapralov AA, Feng WH, Amoscato AA, Stauber RH (2011) Nanoparticle size is a
Yanamala N, Balasubramanian K, Winnica DE, critical physicochemical determinant of the

Kisin ER, Kotchey GP, Gou P, Sparvero LJ, human blood plasma corona: a comprehensive
quantitative proteomic analysis. ACS Nano 19. Schaefer J, Schulze C, Marxer EE, Schaefer
5:7155–7167 UF, Wohlleben W, Bakowsky U, Lehr CM
9. Martel J, Young D, Young A, Wu CY, Chen (2012) Atomic force microscopy and analytical
CD, Yu JS, Young JD (2011) Comprehensive ultracentrifugation for probing nanomaterial
proteomic analysis of mineral nanoparticles protein interactions. ACS Nano 6:4603–4614
derived from human body fluids and analyzed 20. Mahmoudi M, Lynch I, Ejtehadi MR, Monopoli
by liquid chromatography-tandem mass spec- MP, Bombelli FB, Laurent S (2011) Protein–
trometry. Anal Biochem 418:111–125 nanoparticle interactions: opportunities and
10. Zeng Z, Patel J, Lee SH, McCallum M, Tyagi challenges. Chem Rev 111:5610–5637
A, Yan M, Shea KJ (2012) Synthetic polymer 21. Akesson A, Cardenas M, Elia G, Monopoli
nanoparticle–polysaccharide interactions: a MP, Dawson KA (2012) The protein corona
systematic study. J Am Chem Soc of dendrimers: PAMAM binds and activates
134:2681–2690 complement proteins in human plasma in a
11. Milani S, Baldelli Bombelli F, Pitek AS, generation dependent manner. RSC Adv
Dawson KA, Radler J (2012) Reversible versus 2:11245–11248
irreversible binding of transferrin to polysty- 22. Lesniak A, Fenaroli F, Monopoli MP, Aberg C,
rene nanoparticles: soft and hard corona. ACS Dawson KA, Salvati A (2012) Effects of the
Nano 6:2532–2541 presence or absence of a protein corona on
12. Dobrovolskaia MA, Patri AK, Zheng J, silica nanoparticle uptake and impact on cells.
Clogston JD, Ayub N, Aggarwal P, Neun BW, ACS Nano 6:5845–5857
Hall JB, McNeil SE (2009) Interaction of col- 23. Ge C, Du J, Zhao L, Wang L, Liu Y, Li D,
loidal gold nanoparticles with human blood: Yang Y, Zhou R, Zhao Y, Chai Z, Chen C
effects on particle size and analysis of plasma (2011) Binding of blood proteins to carbon
protein binding profiles. Nanomedicine nanotubes reduces cytotoxicity. Proc Natl
(Lond) 5:106–117 Acad Sci USA 108:16968–16973
13. Zheng M, Li ZG, Huang XY (2004)
24. Hu W, Peng C, Lv M, Li X, Zhang Y, Chen N,
Ethylene glycol monolayer protected Fan C, Huang Q (2011) Protein corona-
nanoparticles: synthesis, characterization, mediated mitigation of cytotoxicity of gra-
and interactions with biological molecules. phene oxide. ACS Nano 5:3693–3700
Langmuir 20:4226–4235 25. Deng ZJ, Liang M, Monteiro M, Toth I,
14. Sund J, Alenius H, Vippola M, Savolainen K, Minchin RF (2011) Nanoparticle-induced
Puustinen A (2011) Proteomic characteriza- unfolding of fibrinogen promotes Mac-1
tion of engineered nanomaterial–protein inter- receptor activation and inflammation. Nat
actions in relation to surface reactivity. ACS Nanotechnol 6:39–44
Nano 5:4300–4309 26. Monopoli MP, Bombelli FB, Dawson KA

15. Deng ZJ, Mortimer G, Schiller T, Musumeci (2011) Nanobiotechnology: nanoparticle
A, Martin D, Minchin RF (2009) Differential coronas take shape. Nat Nanotechnol 6:
plasma protein binding to metal oxide 11–12
nanoparticles. Nanotechnology 20:455101 27. Pitek AS, O'Connell D, Mahon E, Monopoli
16. Maiorano G, Sabella S, Sorce B, Brunetti V, MP, Baldelli Bombelli F, Dawson KA (2012)
Malvindi MA, Cingolani R, Pompa PP (2010) Transferrin coated nanoparticles: study of the
Effects of cell culture media on the dynamic bionano interface in human plasma. PLoS One
formation of protein–nanoparticle complexes 7:e40685
and influence on the cellular response. ACS 28. Casals E, Pfaller T, Duschl A, Oostingh GJ,
Nano 4:7481–7491 Puntes V (2010) Time evolution of the
17. Zhang H, Burnum KE, Luna ML, Petritis BO, nanoparticle protein corona. ACS Nano
Kim JS, Qian WJ, Moore RJ, Heredia-Langner 4:3623–3632
A, Webb-Robertson BJ, Thrall BD, Camp DG 29. Shevchenko A, Wilm M, Vorm O, Mann M
2nd, Smith RD, Pounds JG, Liu T (2011) (1996) Mass spectrometric sequencing of pro-
Quantitative proteomics analysis of adsorbed teins silver-stained polyacrylamide gels. Anal
plasma proteins classifies nanoparticles with Chem 68:850–858
different surface properties and size. 30. Simpson RJ (2008) Proteins and proteomics: a
Proteomics 11:4569–4577 laboratory manual. Cold Spring Harbor
18. Lundqvist M, Stigler J, Elia G, Lynch I, Laboratory. http://www.proteinsandproteomics.
Cedervall T, Dawson KA (2008) Nanoparticle org/home_1.html
size and surface properties determine the pro- 31. Rai AJ, Gelfand CA, Haywood BC, Warunek
tein corona with possible implications for bio- DJ, Yi J, Schuchard MD, Mehigh RJ, Cockrill
logical impacts. Proc Natl Acad Sci USA SL, Scott GB, Tammen H, Schulz-Knappe P,
105:14265–14270 Speicher DW, Vitzthum F, Haab BB, Siest G,
Chan DW (2005) HUPO Plasma Proteome 32. Casals E, Pfaller T, Duschl A, Oostingh GJ,
Project specimen collection and handling: Puntes VF (2011) Hardening of the nanopar-
towards the standardization of parameters for ticle–protein corona in metal (Au, Ag) and
plasma proteome samples. Proteomics oxide (Fe3O4, CoO, and CeO2) nanoparticles.
5:3262–3277 Small 7:3479–3486
Chapter 12
Electrophoretic Implementation of the Solution-Depletion

Method for Measuring Protein Adsorption, Adsorption
Kinetics, and Adsorption Competition Among Multiple
Proteins in Solution
Hyeran Noh, Naris Barnthip, Purnendu Parhi, and Erwin A. Vogler
Abstract
The venerable solution-depletion method is perhaps the most unambiguous method of measuring solute
adsorption from solution to solid particles, requiring neither complex instrumentation nor associated inter-
pretive theory. We describe herein an SDS-gel electrophoresis implementation of the solution-depletion
method for measuring protein adsorption and protein-adsorption kinetics. Silanized-glass particles with
different surface chemistry/energy and hydrophobic sepharose-based chromatographic media are used as
example adsorbents. Electrophoretic separation enables quantification of adsorption competition among
multiple proteins in solution for the same adsorbent surface, demonstrated herein by adsorption-competition
kinetics from binary solution.
Key words Protein adsorption, Adsorption kinetics, Adsorption competition, Electrophoresis,

Solution depletion
1 Introduction
Fouling is a natural phenomenon by which working surfaces of

devices, instruments, or machines in contact with aqueous solutions
become coated with a layer or layers of adventitious contamination
that compromises intended performance. Fouling has broad techno-
logic importance with considerable socioeconomic impact spanning
environment (e.g., occlusion of pipes and filters used in civil engi-
neering), medicine (e.g., reduction in biosensor sensitivity), and
transportation (e.g., resistance to the flow of water across boat hulls).
Adsorption of surface-active glycosylated proteins and peptidogly-
cans onto surfaces is among the first events of the fouling process.
In particular, protein adsorption is of singular importance in the
development and application of medical devices because adsorbed
proteins can catalyze, mediate, or moderate the biological response
157
158 Hyeran Noh et al.
to various biomaterials used in fabrication of medical devices [1].

Above-mentioned applications usually involve multiple proteins in
solution rather than the single purified solutions typically used in
studies of the mechanism of protein adsorption. For example, blood
contains more than 1,000 proteins spanning 10 decades in concen-
tration, with five proteins dominating the concentration profile
(albumin, IgG, fibrinogen, transferrin, and IgA) [2]. As a conse-
quence, understanding adsorption competition among multiple pro-
teins in solution to the same adsorbent surface remains an important
unsolved problem [1].
We describe herein an electrophoretic implementation of the
solution-depletion method for measuring protein adsorption and
protein-adsorption kinetics [3–7]. The solution-depletion method is
one of the most unambiguous methods of measuring adsorption,
requiring no complex instrumentation, protein labeling, or separa-
tion of protein from adsorbent particles. Interpretation is simply the
arithmetic of mass balance. Electrophoresis is used a multiplexing,
separation and quantification tool that enables study of protein-
adsorption competition between two or more proteins in solution
[8, 9]. Commercially prepared SDS gels and staining kits greatly
simplifies the methodology.
1.1 Mass Balance Solution-depletion measures the concentration of the ith protein
Equations in solution before-and-after contact with particulate adsorbent as
( )
quantified by Di ≡ WBoi − WBi . Here, WBoi is the initial protein
solution concentration (in mg/mL for example), and WBi is the
final protein solution concentration after contact with adsorbent
under the desired experimental conditions (time, temperature,
stirring, static solution, etc.). The depletion Di has the same units
as WBoi and WBi and represents the change in bulk-solution con-
centration due to adsorption.
The absolute mass of protein adsorbed mi = DV i B if WBi and WBi
o
are measured in w/v units and VB is the volume of the bulk protein
solution. These relationships assume that VB is a system constant
(e.g., not adsorbed by adsorbent to a significant extent). In this case,
it is clear that mi and Di are directly proportional and adsorption
isotherms can be constructed by plotting either mi or Di as a func-
tion of solution concentration WBoi at constant adsorbent surface area
[3]. Measurement of adsorption isotherms over the concentration
range of interest is essential to a complete understanding of the
adsorption process [1].
Adsorption isotherms for solution concentrations 15 > WBoi > 0
mg/mL show that blood proteins adsorb to various test adsor-
bents in proportion to WBoi , approximating a Henry isotherm up to
saturation of the adsorbent surface which occurs at a particular
( )
max
solution concentration WBoi [1, 3]. Henry-like isotherms are
obtained for any number of proteins adsorbing simultaneously
from solution when individual protein solution concentrations are
Protein Adsorption Measured by Solution Depletion 159
measured before and after contact with adsorbent. For example,

adsorption of proteins from serum to hydrophobic adsorbents
exhibit Henry-like isotherms when scaled as per-cent solution
concentration [10] (herein we subscribe to a 65° advancing c ontact
angle as the boundary between hydrophilic and hydrophobic
materials [1]).
Further interpretation of adsorption and adsorption isotherms
measured by solution depletion depends on the adsorption para-
digm in effect. If this paradigm construes adsorption to occur at a
planar surface [11], then the surface concentration Γ = mi / A in
mass-per-unit adsorbent surface area A . Because proteins can
absorb in multilayers (depending on solution concentration and
protein size), it can be convenient to construe adsorption as a par-
titioning of protein from bulk solution into a three-dimensional
(3D) interphase region that separates the physical- adsorbent
surface from bulk solution [1]. This interphase thus has a finite
volume VI with a specific protein concentration WI and
æV ö
D º (WBo - WB ) = WI ç I ÷ . For a single-protein solution WI = WIi
è VB ø
æ VI ö
( )
and Di º WBoi - WBi = WIi ç i ÷ , but in the case of adsorption from
è VB ø
multiple-protein, solution WI = å WIi. Needless to say, perhaps,
i
mass balance equations become more complicated with the
number of proteins simultaneously adsorbing from solution [5].
At steady state, the partition coefficient for the ith
protein (
Pi º WIi / WBi ,) and it follows that
æ VI ö é VB ù
WBi = WBoi - WIi ç i ÷ = WBoi ê ú for 0 £ WBi £ WBmi ax . A related
è VB ø êëVB + PV ú
i Ii û
é PV i Ii
ù
isotherm equation is Di = WBoi ê ú for 0 £ WBi £ WBi . As
max
êëVB + PV
i Ii ú
û
mentioned above, it has been found that solution-depletion isotherms
approximate a linear Henry isotherm and thus have a slope

é PV i Ii
ù  Si 
Si º ê ú or, by rearrangement, PV i Ii = VB  [3].
êëVB + PV i Ii ú
û  1 − Si 
Comparison of adsorption experiments with two different proteins
i and j to the same adsorbent surface area (in separate experiments)
allows solution of these equations in terms of interphase volume
 VI  æPö
ratio  i  and partition-coefficient ratio ç i ÷ as further elaborated
 VI j  çP ÷
è jø
in ref. 3.
2 Materials
2.1 Reagents, Solvents such as ethanol and chloroform were reagent grade and
Proteins, and used as received from VWR, Inc. Other chemicals such as H2O2
Solutions and H2SO4 were also used as received from VWR unless otherwise
specified. Water was 18 MΩ deionized obtained from a Millipore
Simplicity water purification system. Phosphate buffer saline (PBS)
used as diluent was prepared from powder (Sigma; 0.14 M NaCl,
3 mM KCL) in 18 MΩ deionized water with final pH = 7.4. Conical
microtubes (0.5 mL, Safe-lock microcentrifuge tubes, Eppendorf;
approximately 2 cm2 internal surface area) were used in solution-
handling steps outlined below.
Silanes applied in this work to derivatize glass-particle adsor-
bents used as received from Gelest Inc., (Morrisville, PA) were
octadecyltrichlorosilane (OTS), 3-aminopropyltriethoxysilane
(APTES), n-propyltriethoxysilane (PTES), and vinyltriethoxysilane
(VTES). OTS-treated glass particles were optionally coated in a
0.2 % solution of 1,1-pentadecafluorooctylmethacrylate in tricholo-
rotrifluoroethane (“Nyebar,” Nye Lubricants, Fairhaven, MA) to
render OTS-treated glass particles slightly more hydrophobic than
OTS-treated counterparts.
Piranha solution was prepared by slowly mixing 1 part 30 %
H2O2 in three parts concentrated H2SO4 while cooling in an ice
bath see Note 1.
Human sourced proteins were used as received from various
vendors without further purification. Test proteins spanned a molec-
ular weight (MW) range from lysozyme (15 kDa, [3]) to IgM
(1,000 kDa, [11, 12]). Protein solutions were prepared in PBS by
80:20 or 90:10 dilution of protein concentrate freshly prepared
before adsorption measurements. Solution concentrations typically
varied between 0 and 10 mg/mL depending on protein under study.
Electrophoresis was performed using 15- or 26-lane NuPAGE Novex
Tris–Acetate pre-cast gels (Invitrogen Corp.; 500 kDa capacity) to
separate and quantify proteins. NuPAGE Novex Bis–Tris gels were
used for low MW proteins. Electrophoresis was performed in Xcell
Sure Lock Mini-Cell or Xcell 4 SureLock Midi-Cell (Invitrogen
Corp.) for 70 min at 150 V (Tris–Acetate) or 35 min at 220 V
(Bis–Tris). Gels were stained with SimplyBlue SafeStain (Invitrogen
Corp.) for 1 h and destained with deionized (18 MΩ) water for several
hours while mixing on a standard hematology rocker. Band intensity
was quantified using a Gel-doc system (Bio-Rad Laboratories Inc.).
2.2 Test Adsorbents Glass-particle adsorbents were 425–600 μm diameter glass p

articles
(Sigma Aldrich) in either cleaned or silanized form. The nominal
specific area used in this work was 5 × 10−3 m2/g (based on
512.5 μm mean diameter and 168 μg/particle). The actual surface
area measured by the Brunauer–Emmett–Teller (BET) method
was 0.25 ± 0.09 m2/g (Micromeritics ASAP 2000 using liquid
nitrogen as the probe gas, see Note 2) was 50× larger than the
nominal value, possibly reflecting dispersity in particle size and
porosity. Glass coverslips (Fisher, 22 × 30 × 0.1 mm) were used as a
witness samples to glass-surface treatments amenable to contact
angle measurements. Glass used in this study was activated before
silanization by immersion in piranha solution followed by rinsing
in 18 MΩ deionized water. After air-drying, particles were s ubjected
to 5–10 min air discharge in a commercial plasma generator
(Herrick, Wippany NY).
Octyl Sepharose 4 Fast Flow adsorbent with 90 μm diameter was
obtained from Amersham Biosciences (40 % or 75 % by volume of
90 μm nominal diameter particles dispersed in 20 % ethanol–water
solution, see Note 3). Sepharose particles were washed in PBS to
remove ethanol by 3× sequential centrifugation (40 RPM for 1 min in
a Hettick microtube fixed-rotor centrifuge, VWR) and resuspension
in PBS.
3 Methods
3.1 Preparation 1. Clean/activate glass particles and coverslip witness samples by

of Glass-Particle 30 min immersion in 80 °C piranha solution followed by three
Adsorbents sequential washes in each of 18 MΩ deionized water and
ethanol.
2. Air dry the piranha-solution-oxidized glass and subsequently
oxidize by air–plasma treatment of a single layer of particles
(or glass coverslip witness samples) held in a 15 mm Pyrex glass
Petri dish directly before silanization. Glass surfaces treated in
this manner will be fully water-wettable (0° water contact
angle).
3. Silanize activated glass particles (and coverslip witness samples)
using silane reagents selected from below. Typical results com-
piled in Table 1.
Table 1
Typical outcome of glass silanization procedures
Surface treatment Advancing contact angle range (°)

Nyebar 109–118
OTS 106–109
PTES 93–97
VTES 85–90
APTES 42–58
Hydrolyzed glass 0
3a. 1.5 h reaction with 5 % v/v OTS in chloroform (see Note 4).

●● Rinse the silanized samples by three sequential washes
in chloroform before curing in a vacuum oven at
110 °C for 12 h.
3b. Optional Dip coat with Nyebar
●● Dip cured OTS-silanized glass adsorbents in Nyebar solu-
tion for 10 min and air dry to produce surface slightly more
hydrophobic than rendered by OTS treatment alone.
3c. 20 min reaction with 5 % APTES dissolved in 95:5 v/v

ethanol–water solutions.
●● APTES solution in 95:5 v/v ethanol–water should be
allowed to hydrolyze 12 h before use (see Note 4).
●● Rinse APTES-treated glass with absolute ethanol 3×
and cure 12 h in a vacuum oven at 110 °C.
3d. 20 min reaction with 5 % PTES dissolved in 90:10 v/v

ethanol–water solution.
●● PTES in 90:10 v/v ethanol–water solution containing
0.5 % glacial acetic acid should be allowed to hydrolyze
12 h before use (see Note 4).
●● Rinse PTES-treated glass with absolute ethanol 3× and
cure 12 h in a vacuum oven at 110 °C.
3e. 20 min reaction with 5 % VTES dissolved in 90:10 v/v etha-

nol–water solution.
●● VTES in 90:10 v/v ethanol–water solution containing
0.5 % glacial acetic acid should be allowed to hydrolyze
12 h before use (see Note 4).
●● Rinse VTES-treated glass with absolute ethanol 3× and
cure 12 h in a vacuum oven at 110 °C.
3.2 Preparation 1. Prepare fresh octyl sepharose adsorbent (40 % or 75 % solids)
of Octyl Sepharose before each depletion experiment by 3× washing in PBS
Adsorbents (to remove ethanol) using a sequential centrifugation/resuspen-
sion protocol that processed 1 mL of as-received suspension.
2. Replace 500 μL of supernatant with 500 μL PBS, ending with
a 60:40 (or 75:25) v/v stock suspension in PBS.
3. Pipette 50 μL (or 20 μL) stock corresponding to 30 μL fluid, 20 μL
beads (or 5 μL fluid, 15 μL beads) into a 0.5 mL microtube.
4. Remove 25 μL fluid after centrifugation.
3.3 Measurement 1. Prepare desired protein solutions (25 μL) in PBS at varying

of Adsorption from concentrations (from 0.1 to 10 mg/mL).
Single-Protein 2. Resuspend octyl sepharose adsorbent (or 22 mg of glass-
Solutions adsorbent particles) by gentle pipette aspiration in 25 μL protein
solution so that the final solution volume is 30 μL. Quantities

of adsorbent suggested here are general guides see Note 5.
3. Leave the solution with adsorbent undisturbed in 0.5 mL conical
microtubes for desired equilibration time at ambient temperature.
Solution and adsorbent can be optionally continuously mixed
using a rotating mixer (see Note 6).
4. Sample 20 μL of supernatant protein solution for gel electro-
phoresis, once for steady-state measurements or every 5 min
for kinetics experiments. Kinetic experiments use a separate
microtube for each time interval.
3.4 Measurement 1. Prepare protein solutions (25 μL) in PBS at varying concentra-

of Adsorption from tions of both proteins (see Note 7).
Binary-Protein 2. Resuspend octyl sepharose beads (or 22 mg of glass adsorbent
Solutions particles) by gentle pipette aspiration in 25 μL protein solution
so that the final solution volume is 30 μL.
3. Leave the solution with beads undisturbed in 0.5 mL conical
microtubes. Solution and adsorbent can be optionally continu-
ously mixed using a rotating mixer (see Note 8).
4. Sample 20 μL of supernatant protein solution for gel electro-
phoresis, once for steady-state measurements or every 5 min
for kinetics experiments (see Note 8). Kinetic experiments use
a separate microtube for each time interval.
3.5 Electrophoresis 1. Use 15- or 26-lane commercial gels and electrophoresis equip-
ment to separate and quantify proteins.
2. Quantify band intensity using a commercial gel reader of band
optical density (OD).
3. Prepare a standard curve for each gel using the first 6–7 lanes by
applying proteins at known solution concentration (see Note 9).
Plot OD vs. standard protein concentration to obtain working
calibration curve.
4 Notes
1. Piranha solution is a strong oxidizer used to remove organic

impurities on glass-particle adsorbent surfaces and to hydroxylate
glass surfaces to enhance subsequent silanization reactions.
Piranha solution must be prepared with great care with appro-
priate safety precautions including use of a chemical hood and
personal protection equipment. In particular, admixture of

H2O2 with H2SO4 is exothermic and should be performed slowly
with cooling in water ice. Piranha solution should be freshly
prepared before use and never stored due to self-decomposition
of hydrogen peroxide. Glass particles and coverslips should be
immersed slowly in piranha solution to prevent thermal shock.
2. Reliable estimate of adsorbent surface area by BET can be

problematic because glass- particle size used herein is quite
large and specific area correspondingly low. Measurement of
hydrated octyl sepharose particle surface area by BET is not
possible. Nevertheless, the “volumetric” method of interpret-
ing protein adsorption can be applied which does not require
accurate measurement of adsorbent surface area [1, 3].
3. Results described in refs. [3–7] with Amersham octyl sepharose
are specific to this particular chromatographic packing material.
Subsequent work with alternative sources of similar chromato-
graphic media (GE Healthcare Bio-Sciences) did not yield similar
results, presumably because alternative material was more porous
and ABsorbed a significant amount of protein over-and-above
the amount ADsorbed to the Amersham product. Adsorption
isotherms obtained with alternative sepharose sources did not
( )
max
achieve a saturation WBoi at any solution concentration tested.
4. Trichlorosilanes are water sensitive and highly reactive [13–15].
All manipulations should be carried out with appropriate safety
precautions including use of a chemical hood and personal protec-
tion equipment. Dry conditions obtained using a chemical glove
bag or glove box can make silanization with trichlorosilanes more
successful, but use of refluxing CHCl3 obtained in open beakers
on a hot plate housed in an ordinary chemical hood can be used
to minimize reaction of with atmospheric water. The primary side
reaction responsible for poor silanization outcomes appears to be
polymerization of trichlorosilane, possibly facilitated by the pres-
ence of water, which can lead to opaque, multilayer films on glass
coverslip witness samples. Typically, these films are rough and
exhibit advancing water contact angles exceeding 120° due to the
superhydrophobic effect [16, 17].
Hydrolysis of ethoxy silanes in ethanol–water solution is
critical to formation of complete monolayers. An overabundance
of water will result in excessive polymerization in the solvent
phase, while a deficiency of water will result in the formation of
an incomplete monolayer.
5. Adsorbent capacity (maximum solution depletion at fixed surface
area) is a function of protein size (MW) [3, 11]. As a consequence,
it may be necessary to adjust surface-area-to-solution-volume to
achieve a full adsorption isotherm up-to-and-including surface
saturation. See ref. 11 for details. Thorough mixing of hydropho-
bic silanized glass- particle adsorbents with protein solution at
high-surface-area-to-solution volume is difficult due to the non-
wetting properties of adsorbent. Mixing can be enhanced by
manual tapping of tubes with fingers or on a bench top. Mixing
protein solutions with octyl sepharose is not problematic and
similar adsorption isotherms are obtained with both types of

hydrophobic adsorbents, suggesting that mixing obtained with
silanized glass as above is sufficient.
6. Mass adsorption from single-protein solution near mg/mL

concentrations is rapid, probably within milliseconds and is
insensitive to mixing adsorbent with solution [1, 3, 7]. Steady
state is thus achieved within the time frame of mixing protein
solution and adsorbent.
7. Adsorption competition between two proteins is complex. There
are 5 distinct solution-compositions that must be considered for
two proteins of dissimilar MW [8, 9].
8. Mass adsorption from binary-protein solution near mg/mL
concentrations exhibits two, pseudo-steady-state adsorption

regimes connected by smooth transitions lasting 20–30 min,
depending on binary-protein pair relative MW and concentra-
tion. As such, adsorption competition exhibits kinetics and is
sensitive to solution mixing [8, 9].
9. Each different protein requires a separate calibration curve on
the same gel to account for differences in staining density.
Binary solutions are treated the same as single-protein s olutions,
multiplexing development of a band OD-concentration calibra-
tion curve. Calibration curves can be bi-modal over 0–10 mg/mL
range [3, 4] with a 0.2 mg/mL lower limit of detection
(LOD, [4]) using Coomassie blue gel stains.
Acknowledgments
This work was supported by National Institute of Health grant

PHS 5R01HL069965. The authors appreciate the support from
the Departments of Materials Science and Engineering and
Bioengineering, The Pennsylvania State University.
References
1. Vogler EA (2012) Protein adsorption in three 6. Noh H, Vogler EA (2008) Volumetric

dimensions. Biomaterials 33:1201–1237 interpretation of protein adsorption: ion-

2. Anderson NL et al (2004) The human plasma exchange adsorbent capacity, protein pI,
proteome—a nonredundant list developed by and interaction energetics. Biomaterials
combination of four separate sources. Mol Cell 29:2033–2048
Proteomics 3(4):311–326 7. Barnthip N et al (2008) Volumetric interpreta-
3. Noh H, Vogler EA (2006) Volumetric interpre- tion of protein adsorption: kinetic consequences
tation of protein adsorption: partition coeffi- of a slowly-concentrating interphase. Biomaterials
cients, interphase volumes, and free energies of 29:3062–3074
adsorption to hydrophobic surfaces. Biomaterials 8. Barnthip N et al (2009) Volumetric interpreta-
27:5780–5793 tion of protein adsorption: kinetics of
4. Noh H, Vogler EA (2006) Volumetric interpre- protein-adsorption competition from binary
tation of protein adsorption: mass and energy solution. Biomaterials 30:6495–6513
balance for albumin adsorption to particulate 9. Barnthip N, Vogler EA (2012) Protein adsorp-
adsorbents with incrementally-increasing tion kinetics from single- and binary-solution.
hydrophilicity. Biomaterials 27:5801–5812 Appl Surf Sci 262:19–23
5. Noh H, Vogler EA (2007) Volumetric interpre- 10. Parhi P et al (2010) Role of water and proteins
tation of protein adsorption: competition from in the attachment of mammalian cells to
mixtures and the vroman effect. Biomaterials surfaces: a review. J Adhesion Sci and Tech

28:405–422 24:853–888
11. Parhi P et al (2009) Volumetric interpretation 14. Fadeev AY, McCarthy TJ (2000) Self-assembly
of protein adsorption: capacity scaling with is not the only reaction possible between
adsorbate molecular weight and adsorbent alkyltrichlorosilanes and surfaces: monomolecu-
surface energy. Biomaterials 30:6814–6824 lar and oligomeric covalently attached layers of
12. Kao P et al (2010) Volumetric interpretation dichloro- and trichloroalkylsilanes on silicon.
of protein adsorption: interfacial packing of Langmuir 16:7268–7274
protein adsorbed to hydrophobic surfaces 15. Wang M et al (2005) Self-assembled silane
from surface-saturating solution concentra- monolayers: fabrication with nanoscale unifor-
tions. Biomaterials 32:969–978 mity. Langmuir 21(5):1848–1857
13. Sagiv J et al (1986) Self-assembling monolay- 16. Oner D, McCarthy TJ (2000) Ultrahydrophobic
ers: a study of their formation, composition, surfaces. Effects of topography length scales on
and structure. In: Mittal KL, Bothorel P (eds) wettability. Langmuir 165:7777–7782
Surfactants in solution, vol 5. Plenum Press, 17. Gao L, McCarthy TJ (2009) Wetting 101.
New York, pp 965–978 Langmuir 25(24):14105–14115
Chapter 13
Hyperspectral Microscopy for Characterization

of Gold Nanoparticles in Biological Media and Cells
for Toxicity Assessment
Christin Grabinski, John Schlager, and Saber Hussain
Abstract
Nanoparticles (NPs) are being implemented in a wide range of applications, and it is critical to proactively
investigate their toxicity. Due to the extensive range of NPs being produced, in vitro studies are a valuable
approach for toxicity screening. Key information required to support in vitro toxicity assessments include
NP stability in biologically relevant media and fate once exposed to cells. Hyperspectral microscopy is a
sensitive, real-time technique that combines the use of microscopy and spectroscopy for the measurement
of the reflectance spectrum at individual pixels in a micrograph. This method has been used extensively for
molecular imaging with plasmonic NPs as contrast agents (Aaron et al., Opt Express 16:2153−2167,
2008; Kumar et al., Nano Lett 7:1338−1343, 2007; Wax and Sokolov, Laser Photon Rev 3:146−158,
2009; Curry et al., Opt Express 14:6535–6542, 2006; Curry et al., J Biomed Opt 13:014022, 2008;
Cognet et al., Proc Natl Acad Sci U S A 100:11350–11355, 2003; Sokolov et al., Cancer Res 63:1999–
2004, 2003; Sönnichsen et al., Nat Biotechnol 23:741−745, 2005; Nusz et al., Anal Chem 80:984–989,
2008) and/or sensors (Nusz et al., Anal Chem 80:984–989, 2008; Ungureanu et al., Sens Actuators B
150:529−536, 2010; McFarland and Van Duyne, Nano Lett 3:1057−1062, 2003; Galush et al., Nano
Lett 9:2077−2082, 2009; El-Sayed et al., Nano Lett 5:829–834, 2005). Here we describe an approach for
using hyperspectral microscopy to characterize the agglomeration and stability of plasmonic NPs in bio-
logical media and their interactions with cells.
Key words Hyperspectral microscopy, Hyperspectral imaging, Light scattering, Plasmonic nanoparticles,
Agglomeration, Cellular interaction
1 Introduction
Nanoparticles (NPs) are being implemented in a wide range of

commercial, medical, military, and industrial applications, with
public and private investments in nanotechnology projects reach-
ing $17.8 billion globally in 2010 [1]. With such a widespread
implementation of NPs, exposure to humans and the environment
is likely to occur at continually increasing levels [2]. Therefore, it is
167
168 Christin Grabinski et al.
critical to proactively investigate NP toxicity in order to protect

against human and environmental effects.
Due to the extensive variations in NPs being produced, in vitro
studies are a necessary screening tool to compliment in vivo studies
in assessing toxicity. Many in vitro studies have been published,
demonstrating that NP toxicity is a function of size, agglomerate
morphology, composition, crystallinity, surface area, surface chem-
istry, and surface charge [3–10]. However, the results are often
contradictory across laboratories and provide little value for NP
regulation due to lack of sufficient NP characterization [11–13].
For the majority of traditional in vitro toxicity studies, NPs are
dispersed in biological media before exposure, which can result in
agglomeration and adsorption of media components [9, 14].
Agglomeration and protein adsorption have been shown to affect
NP uptake and toxicity [5, 15]. Additionally, NP agglomeration
can result in a misinterpretation of dose, which is related to the
NP agglomerate size and density distribution [13]. Therefore, NP
stability must be characterized in the media used for dosing.
Common methods for investigating NP stability include
dynamic light scattering and laser Doppler electrophoresis, which
are used to measure hydrodynamic diameter and zeta potential,
respectively. An additional method commonly employed for inves-
tigating the stability of plasmonic NPs is UV–Visible spectroscopy,
which can be used for measuring shifts in plasmon resonance to
estimate stability in various media [16]. However, none of these
techniques are supplemented with visual data for characterizing
NP interactions in biological media [17]. Additionally, UV–Visible
spectroscopy only provides averaged data for the NP dispersion
and is not useful for imaging localized NP interactions [17].
The fate and distribution of NPs once they are exposed to cells
is also of critical importance. Many studies have shown that NPs
accumulate in the endo-lysosomal system once internalized [3, 4,
7]. However, the extent of uptake and intracellular localization
varies widely for different NPs, and this can have a direct effect on
NP toxicity [9, 18]. Therefore, methods for detecting NPs in cells
and tissues are critical for understanding their toxicity.
NP uptake and intracellular localization in vitro can be character-
ized using various microscopy techniques. For high resolution,
transmission electron microscopy (TEM) is often used; however, this
technique requires various steps for manipulating samples, including
centrifugation and dehydration [19]. Although optical microscopy
techniques can be used with minimal sample preparation, the major
limitation is that fluorescent tagging is required to visualize the
presence of NPs. Dark-field microscopy can be used to illuminate
dense structures, such as cell membranes, organelles and NPs, with-
out fluorescent tagging, but the identity of the structures cannot be
confirmed using this technique alone [20].
Hyperspectral Microscopy for Toxicity Assessment 169
Hyperspectral microscopy combines the use of microscopy and

spectroscopy for the measurement of the reflectance spectrum at
individual pixels in a micrograph. It can be used to image and track
plasmonic NPs in complex biological environments and has been
highly cited for molecular imaging using plasmonic NPs as contrast
agents [17, 21–28] and/or local refractive index sensors [17, 29–32].
Recently, hyperspectral microscopy has been demonstrated as a useful
technique for characterizing Au NP agglomeration, protein adsorption
and cell uptake [18, 33].
Hyperspectral microscopy is particularly useful for analyzing
plasmonic NPs, such as gold (Au), which absorb and scatter light
in a predictive manner due to collective oscillations of electrons at
the surface of a particle induced by an electromagnetic field. Light
scattering by Au NPs is a function of size, composition, shape, and
local refractive index [34–36]. Additionally, light scattering is
affected by interparticle plasmon coupling, which decays over a
distance on the order of the NP size, where coupled NPs exhibit a
red-shifted and broadened plasmon peak [37]. These properties of
Au NPs allow for spectral data to be analyzed for understanding
NP agglomeration and protein adsorption in biological environ-
ments. Hyperspectral microscopy offers a combined qualitative
and quantitative approach for dynamic imaging of NPs and live
cells, requiring very little sample preparation. Therefore, it is an
ideal approach for characterization of Au NP agglomeration and
stability in biological media and interactions with cells to support
toxicity assessment.
2 Materials
2.1 Hyperspectral 1. Glass slides or coverslips for attaching Au NPs (see Note 2).
Microscopy Instrument 2. Glass slides or coverslips appropriate for cell culture.
(See Note 1)
3. Glass coverslips, cleaned in ethanol and dried.
2.2 Consumables 4. Clear nail polish or silicon grease for sealing the coverslips to
the slides.
5. Well-characterized Au NPs (see Note 3).
6. Cell-line or primary cells.
7. Cell culture media appropriate for the cells.
8. Buffer for rinsing cells.
9. Mounting media for refractive index matching of NP slides and
NP exposed cell slides (optional).
10. Immersion oil for high magnification oil lens objectives
(optional).
3 Methods
3.1 Protocol for 1. Dilute Au NPs in the desired media (water, cell media with and
Nanoparticle without serum proteins) to desired concentration (recommended
Characterization range: 1–100 mg/l).
Using Hyperspectral 2. Pipette 5–20 μl onto coated glass slide and incubate 10–120 s,
Microscopy then rinse with water and allow to dry (see Note 4).
3. Add mounting media to achieve desired refractive index
(if applicable; see Note 5).
4. Add a clean glass coverslip and seal using silicon grease or clear
nail polish (if applicable; see Note 6).
5. Allow to dry for at least 10 min at room temperature (see Notes
7 and 8).
6. Image slides using an appropriate hyperspectral microscopy
system (see Note 1).
7. Use software to collect average scattering spectra and localized
spectra for pixels or regions of interest. See Note 5 for data analysis.
Examples for NP characterization are described in Figs. 1, 2,
and 3. Data in the figures were collected using the CytoViva
Hyperspectral Imaging System (CytoViva, Inc., Auburn, AL).
3.2 Protocol for 1. Plate cells on coverslips or cell culture slides and allow to
Characterizing adhere for desired time period (usually about 24 h).
Nanoparticle 2. Dilute Au NPs in the appropriate cell culture media to desired
Interactions with Cells concentration (recommended range: 1–100 mg/l).
Using Hyperspectral 3. Expose NPs diluted in culture media to cells and incubate for
Microscopy desired exposure period (recommended: 0–48 h).
4. After the desired exposure period, rinse cells with buffer or fix
cells (e.g., 4 % paraformaldehyde or cold methanol for 10 min
at room temperature), then rinse with buffer.
5. Add Permount or media layer to achieve desired refractive
index (if applicable; see Note 5).
6. For cells grown on chambered slides, remove chambers and
add a clean glass coverslip; for cells grown on glass coverslips,
transfer face down to a clean glass slide.
7. Seal cover glass with silicon grease or clear nail polish.
8. Allow to dry for at least 10 min (see Notes 7 and 8).
9. Image slides using an appropriate hyperspectral microscopy
system (see Note 1).
10. Use software to collect average scattering spectra and localized
spectra for pixels or regions of interest. See Note 5 for data
analysis. Examples are described in Figs. 4, 5, and 6. Data in
the figures were collected using the CytoViva Hyperspectral
Imaging System (CytoViva, Inc., Auburn, AL).
Fig. 1 Scattering properties of Au nanospheres from the National Institute of Standards and Technology (NIST)
dispersed in water and stabilized in citric acid (TEM: 64.5 ± 8.7 nm). Au nanospheres were dried on a glass
slide and coated with a mounting media with refractive index of 1.515. A glass coverslip was added and sealed
using clear nail polish. Individual spherical Au NPs appeared green, while agglomerated Au NPs appeared
yellow or red due to interparticle coupling. The scattering peak for green particles in the micrograph was
~569 nm, while the scattering peak for red particles was ~651 nm. The theoretical plasmon peak for an
individual spherical Au nanosphere (65 nm) is 515, 538, and ~565 nm for local refractive index for air (~1.000),
water (~1.333), and glass/immersion oil (~1.515), respectively.* *Theoretical values were determined using
freely accessible software MiePlot and confirmed using published values [36]
Fig. 2 Stability of Au nanospheres from NIST stabilized in citric acid (TEM: 64.5 ± 8.7 nm) in water or cell
culture media, with and without addition of serum proteins. Au NPs were dispersed in the desired media, and
then dried on a slide. A coverslip was added and sealed with clear nail polish. Au nanospheres were stable in
water, agglomerated in cell exposure media, and were stable in both water and media when serum proteins
were added at a concentration used typically for cell culture (8 % fetal bovine serum). The peak wavelengths
(averaged over 50 pixels) demonstrated a significant right shifted plasmon peak in media with no serum, and
a decreased peak in the presence of serum. Local refractive index effects may account for the small shifts
between water and water plus serum and water plus serum versus media plus serum
172
Christin Grabinski et al.
Fig. 3 Effect of surface chemistry on Au nanorod (aspect ratio ~4) agglomeration. Au nanorods functionalized with polyethylene glycol (PEG) did not agglomerate in
media, while Au nanorods functionalized with mercaptohexadecanoic acid (MHDA) produced large agglomerates. This can be observed both in the dark-field images
and the corresponding spectral plots shown to the right of each micrograph
Fig. 4 Effect of stabilizing molecule on Au nanosphere (~10 nm) interaction with lung epithelial cells (A549; ATCC).
(a, d) Control cells; (b, e) Au nanospheres stabilized in citrate; (c, f) Au nanospheres stabilized in tannic acid.
Results showed red peaks for Au NP agglomerates in (b, e), indicating greater interaction of citrate stabilized
Au nanospheres with cells. This was previously confirmed qualitatively using TEM and quantitatively using
inductively coupled plasma-mass spectrometry [18]. Image reproduced from [9] with permission from Springer
Science + Business Media
4 Notes
1. Hyperspectral microscopy is also referred to as either multispectral

imaging or hyperspectral imaging, where hyper refers to high
spectral resolution, yielding continuous spectral plots. Instruments
capable of spectral imaging are marketed using any of the above
Fig. 5 PEGylated Au nanorod (aspect ratio ~4) interaction with HaCaT cells. PEGylated Au nanorods were easily
identified in the cellular environment due to the sharp red peak and limited interparticle coupling due to the
hydrophilic stable surface chemistry. Representative spectral plots are shown for both PEGylatedAu nanorods
(1) and background cell scattering (2). Limited interaction of these particles with HaCaT cells was previously
confirmed using TEM [9]
Fig. 6 MHDA functionalized Au nanorods (aspect ratio ~4) are identified in HaCaT cells using spectral angle
mapper, which is automated approach embedded in software by ENVI that compares spectral data collected in
the NP stability studies to pixel spectra in the cellular environment. Uptake of MHDA-Au nanorod agglomerates
in HaCaT cells was previously confirmed using TEM [9]
names and generally consist of a research microscope equipped

with a high resolution dark-field condenser, dispersive element
(i.e., diffraction grating or prism), spectral detector, and imaging
camera. The spectral range of the spectrophotometer is critical,
especially for Au NPs, which scatter light in the near infrared
region, such as high aspect ratio nanorods and nanocages.
For spherical Au NPs, which scatter in the visible region, it has

been shown that it is possible to estimate the scattering spectra by
analyzing the red/green ratio of NPs in dark-field images acquired
using a color camera [29].
2. Coating the slides with a silane that will leave available thiol or
amine groups, such as (3-mercaptopropyl)-trimethoxysilane
(MPTMS), mercaptopropyltriethoxy silane (MPTES) or 3-ami-
nopropyltriethoxy silane (APTES), can improve the attachment
of Au NPs. The slides can be coated using published techniques
[17, 29, 38], or pre-coated slides can be purchased.
3. The morphology and size distribution should be verified using
TEM or atomic force microscopy. The stabilizing molecule or
surface chemistry should be known (information from the
manufacturer is sufficient).
4. Interparticle spacing is a function of concentration, droplet
volume, and incubation time and should be optimized for ade-
quate particle spacing.
5. Plasmon resonance shifts in the peak wavelength in the scattering
spectra of Au NPs occur as a function of size, shape, electromag-
netic coupling, local refractive index, and substrate effects. It is
critical to understand these effects in order to properly prepare
the samples and interpret the data acquired from hyperspectral
microscopy experiments. Each of these effects is described in
more detail below.
(a) Size and shape: Au nanospheres scatter light from green to
yellow when increased from 20 to 100 nm with the wave-
lengths increasing with size. The plasmon peak shifts to red as
the particle becomes more oblate. Au nanorods also scatter
increasing wavelengths with increasing aspect ratio, where the
transverse peak ranges from green to yellow and the longitu-
dinal peak ranges from red to the near infrared region. For Au
nanorods, the scattering can change depending on the orien-
tation of the rods, but this detail is not extremely critical for
toxicity studies [39]. Stellated Au NPs (~100 nm) have also
been shown to scatter red wavelengths [21].
(b) Electromagnetic coupling: Particles that are close enough in
proximity participate in electromagnetic coupling, which
changes the scattering spectral response [28]. The shift in the
plasmon resonance peak decreases with interparticle distance.
Discrete dipole approximation was used to show a 15 nm
shift in peak wavelength for 20 nm distance between two Au
nanospheres (80 nm) and negligible for 100 nm distance
[29]. Based on this type of approximation, shifts in plasmon
peak can be used to assess aggregation of Au NPs in biologi-
cal media and cells.
(c) Local refractive index: Au NPs exhibit strong sensitivity to
the local refractive index [40]. The peak wavelength shift
has been found to increase linearly with refractive index
change [28, 41]. Dependence on refractive index increases

for particles with higher aspect ratio [41]. For studies aim-
ing to evaluate cellular interactions of NPs based on spec-
tral data collected from NP slides, it is very important that
the refractive index of the media or Permount used for
preparation of slides for NP characterization matches that
used for slides containing cells and NPs.
(d) Substrate effect: Hyperspectral microscopy data is col-
lected in an asymmetric environment with a substrate
underneath and either air or a media layer on top of the
NPs. This results in a complex refractive index. Some stud-
ies have attempted to calculate an effective refractive index
to account for this effect [24], but we find that when using
objectives with an oil lens, the peak scattering wavelength
is within 5 nm of the theoretical values for the refractive
index matching that of immersion oil (see Fig. 1). For the
application described here, this is sufficient.
6. A coverslip is only appropriate if imaging with a local refractive
index of 1.515 (i.e., with an oil lens). The NPs can also be
dried and imaged in air or covered with an immersion oil with
the desired refractive index.
7. Allow the nail polish (or sealant) to dry completely before
imaging with an immersion oil objective. If the coverslip is not
sealed properly, movement of the objective near the edge of
the coverslip may allow the immersion oil to seep under the
coverslip, dislodging NPs from the slide.
8. NP slides can be imaged immediately or stored at room
temperature or 4 °C. Live cells should be imaged immediately.
Fixed slides can be stored at 4 °C for up to 6 months.
Acknowledgments
This work was supported by the Air Force Surgeon General and the
Air Force Research Laboratory, Chief Scientist Seedling Program.
Christin Grabinski receives a fellowship from the Oak Ridge Institute
for Science and Education.
References
1. Lux Research (2011) Global nanotech spend- 3. Carlson C, Hussain SM, Schrand AM et al
ing. Presented at EuroNanoForum 2011 con- (2008) Unique cellular interaction of silver
ference, 30 May 2011. http://www. nanoparticles: size-dependent generation of
euronanoforum2011.eu/wp-content/ reactive oxygen species. J Phys Chem B
uploads/2011/09/enf2011_support- 112:13608–13619
commercialisation_raje_fin.pdf. Accessed 14 4. Yu K, Grabinski CM, Schrand AM et al (2009)
Sept 2012 Toxicity of amorphous silica nanoparticles in
2. Reijnders L (2012) Human health hazards of mouse keratinocytes. J Nanopart Res 11:15–24
persistent inorganic and carbon nanoparticles. 5. Murdock RC, Braydich-Stolle L, Schrand AM
J Mater Sci 47:5061–5073 et al (2008) Characterization of nanomaterial
dispersion in solution prior to in vitro exposure agglomeration, uptake, and interaction using
using dynamic light scattering technique. high-illuminating system. Int J Toxicol
Toxicol Sci 101:239–252 26:135–141
6. Grabinski CM, Braydich-Stolle LK, Lafdi KL 21. Aaron J, de la Rosa E, Travis K et al (2008)
et al (2007) Effect of particle dimension on Polarization microscopy with stellated gold
biocompatibility of carbon-based nanomaterials. nanoparticles for robust, in-situ monitoring of
Carbon 45:2828–2835 biomolecules. Opt Express 16:2153–2167
7. Braydich-Stolle LK, Schaeublin NM, Murdock 22. Kumar S, Harrison N, Richards-Kortum R
RC et al (2008) Crystal structure mediates et al (2007) Plasmonic nanosensors for imaging
mode of cell death in TiO2 nanotoxicity. J intracellular biomarkers in live cells. Nano Lett
Nanopart Res 11:1361–1374 7:1338–1343
8. Grabinski CM (2008) Biocompatibility of 23. Wax A, Sokolov K (2009) Molecular imaging
carbon-based nanomaterials. Master’s thesis, and darkfield microspectroscopy of live cells
University of Dayton using gold plasmonic nanoparticles. Laser
9. Grabinski C, Schaeublin N, Wijaya A et al Photon Rev 3:146–158
(2011) Effect of gold nanorod surface chemis- 24. Curry A, Hwang WL, Wax A (2006) Epi-
try on cellular interactions in vitro. ACS Nano illumination through the microscope objective
5:2870–2879 applied to darkfield imaging and microspec-
10. Schaeublin NM, Braydich-Stolle LK, Schrand troscopy of nanoparticle interaction with cells
AM et al (2011) Surface charge of gold in culture. Opt Express 14:6535–6542
nanoparticles mediates mechanism of toxicity. 25. Curry AC, Crow M, Wax A (2008) Molecular
Nanoscale 3:410–420 imaging of epidermal growth factor receptor in
11. Lapotko DO, Lukianova EK, Chizhik SA live cells with refractive index sensitivity using
(2007) Methods for monitoring and imaging dark-field microspectroscopy and immunotar-
nanoparticles in cells. Proc SPIE-Int Soc Opt geted nanoparticles. J Biomed Opt 13:014022
Eng 6447(644703):1–10 26. Cognet L, Tardin C, Boyer D et al (2003)
12. Hussain SM, Braydich-Stolle LK, Schrand AM Single metallic nanoparticle imaging for protein
et al (2009) Toxicity evaluation for safe use of detection in cells. Proc Natl Acad Sci U S A
nanoparticles: recent achievements and techni- 100:11350–11355
cal challenges. Adv Mater 21:1549–1559 27. Sokolov K, Follen M, Aaron J et al (2003)
13. Teeguarden JG, Hinderliter PM, Orr G et al Real-time vital optical imaging of precancer
(2007) Particokinetics in vitro: dosimetry con- using anti-epidermal growth factor receptor
siderations for in vitro nanoparticle toxicity antibodies conjugated to gold nanoparticles.
assessments. Toxicol Sci 95:300–312 Cancer Res 63:1999–2004
14. Nel AE, Madler L, Velegol D et al (2009) 28. Sönnichsen C, Reinhard BM, Liphardt J et al
Understanding biophysicochemical interac- (2005) A molecular ruler based on plasmon
tions at the nano-bio interface. Nat Mater coupling of single gold and silver nanoparticles.
8:543–557 Nat Biotechnol 23:741–745
15. Walkey CD, Olsen JB, Guo H et al (2011) 29. Ungureanu F, Wasserberg D, Yang N et al
Nanoparticle size and surface chemistry deter- (2010) Immunosensing by colorimetric dark-
mine serum protein adsorption and macrophage field microscopy of individual gold nanoparticle-
uptake. J Am Chem Soc 134:2139–2147 conjugates. Sens Actuators B 150:529–536
16. Kah J, Zubieta A, Saavedra R et al (2012) 30. McFarland AD, Van Duyne RP (2003) Single
Stability of gold nanorods passivated with silver nanoparticles as real-time optical sensors
amphiphilic ligands. Langmuir 28:8834–8844 with zeptomole sensitivity. Nano Lett
17. Nusz GJ, Marinakos SM, Curry AC et al 3:1057–1062
(2008) Label-free plasmonic detection of 31. Galush WJ, Shelby SA, Mulvihil MJ et al
biomolecular binding by a single gold nanorod. (2009) A nanocube plasmonic sensor for
Anal Chem 80:984–989 molecular binding on membrane surfaces.
18. Mukhopadhyay A, Grabinski CM, Afrooz AN Nano Lett 9:2077–2082
et al (2012) Effect of gold nanosphere surface 32. El-Sayed H, Huang XH, El-Sayed MA (2005)
chemistry on protein adsorption and cell Surface plasmon resonance scattering and
uptake in vitro. Appl Biochem Biotechnol absorption of anti-EGFR antibody conjugated
167:327–337 gold nanoparticles in cancer diagnostics: appli-
19. Schrand AM, Schlager JJ, Dai K et al (2010) cations in oral cancer. Nano Lett 5:829–834
Preparation of cells for assessing ultrastructural 33. White B, Strawbridge A, Grabinski CM et al
localization of nanoparticles with transmission (2012) Hyperspectral imaging (HSI) to evalu-
electron microscopy. Nat Protoc 5:744–757 ate the interaction of optically active nanopar-
20. Skebo JS, Grabinski CM, Schrand AM et al ticles in biological media and cells. Accepted
(2007) Assessment of metal nanoparticle to BIOS
34. Yguerabide J, Yguerabide EE (1998) Light 38. Grabar KC, Smith PC, Musick MD et al (1996)
scattering submicroscopic particles as highly Kinetic control of interparticle spacing in Au
fluorescent analogs and their use as tracer colloid-based surfaces: rational nanometer-scale
labels in clinical and biological applications: architecture. J Am Chem Soc 118:1148–1153
I. Theory. Anal Biochem 262:137–156 39. Orendorf CJ, Sau TK, Murphy CJ (2006)
35. Mock J, Smith DR, Schultz S (2003) Local Shape-dependent plasmon-resonant gold
refractive index dependence of plasmon reso- nanoparticles. Small 2:636–639
nance spectra from individual nanoparticles. 40. Lee KS, El-Sayed MA (2006) Gold and silver
Nano Lett 3:485–491 nanoparticles in sensing and imaging: sensitivity
36. Haiss W, Nguyen TKT, Aveyard J et al (2007) of plasmon response to size, shape, and metal
Determination of size and concentration of composition. J Phys Chem B 110:19220–19225
gold nanoparticles from UV-Vis spectra. Anal 41. Sönnichsen C, Geier S, Hecker NE et al (2000)
Chem 79:4215–4221 Spectroscopy of single metallic nanoparticles
37. Kreibig U, Vollmer M (1995) Optical properties using total internal reflection microscopy. Appl
of metal clusters, vol 25. Springer, Berlin Phys Lett 77:2949–2951
Chapter 14
Immunocytochemistry, Electron Tomography, and Energy

Dispersive X-ray Spectroscopy (EDXS) on Cryosections
of Human Cancer Cells Doped with Stimuli Responsive
Polymeric Nanogels Loaded with Iron Oxide Nanoparticles
Roberto Marotta, A. Falqui, A. Curcio, A. Quarta, and Teresa Pellegrino
Abstract
The cryosectioning technique is an alternative method for preparing biological material for Transmission
Electron Microscopy (TEM). We have applied this technique to study the mechanism of cell internaliza-
tion of stimuli-responsive polymeric nanogels exploited as cargo nanovectors. With respect to conventional
TEM processing, cryosectioning technique better preserves the morphology of solvent-sensitive nanogels
and enhances the visibility of membrane-bounded organelles inside the cell cytoplasm. In this chapter we
describe the protocols we have established to perform Electron Microscopy (EM)-immunocytochemistry,
Electron Tomography (ET), and Energy Dispersive X-ray Spectroscopy (EDXS) chemical analysis in
Scanning TEM (STEM) on cryosections of HeLa cells treated with pH-responsive nanogels hosting short
interference RNA (siRNAs) and iron oxide nanoparticles (IONPs).
Key words Tokuyasu cryosectioning technique, Immunocytochemistry, Electron tomography (ET),

Energy dispersive X-ray spectroscopy (EDXS) chemical analysis, Scanning transmission electron
microscopy (STEM), Stimuli-responsive polymeric nanoparticles, Iron oxide nanoparticles (IONPs)
loaded nanogel, HeLa
1 Introduction
Tokuyasu cryosectioning technique of aldehyde-fixed biological

material is a consolidated technique for preparing biological mate-
rial for immuno Transmission Electron Microscopy (TEM) studies
[1–4]. Briefly, the samples are chemically fixed and infiltrated with
a cryoprotectant, plunge-frozen into liquid nitrogen, sectioned at
low temperature, transferred to room temperature as frozen sec-
tions, and finally thawed and embedded in a thin layer of support-
ing and contrasting film [2, 3]. By this technique proteins and
organic compounds remain in their natural aqueous environment
and, because of the open surface structure, more antigens are
179
180 Roberto Marotta et al.
accessible than in resin sections, giving a very high sensitivity in

immunolabeling experiments. Moreover, the simplicity of sample
preparation makes cryosection a convenient way to perform rou-
tine immunocytochemical localizations [5].
We have recently applied this technique to nanobiotechnology,
with the aim to study the mechanism of cell internalization of
stimuli-responsive solvent-sensitive polymeric nanogels loaded
with inorganic nanoparticles and small oligonucleotides [6]. The
soft nature of those polymeric nanogel materials which can undergo
physicochemical modifications upon application of environment
changes (such as pH or temperature) makes them appealing as
drug carriers for the controlled release of drugs or payload ele-
ments. In a recent work, we showed how acidic pH-responsive
nanogels made of polyvinyl pyridine-divinylbenzene allowed to
host and transport siRNA oligonucleotides and superparamagnetic
iron oxide nanoparticles (IONPs) both entrapped within the poly-
meric matrix [6]. Furthermore, thanks to the well distinct contrast
of the superparamagnetic nanocrystals and of the nanogel under
TEM, we could track the fate of the magnetic pH-responsive nano-
gels once administered to HeLa cells. We observed that upon
endocytosis uptake, the IONPs and the functional siRNA were
released into the cytoplasm under the action of the endogenous
pH stimulus present inside the different endosomal compartments
(see Fig. 1a). The IONPs may be also exploited for applying hyper-
thermia treatment to the cells, triggered by an external alternative
magnetic field of appropriate frequency and field amplitude [7, 8].
In recent works IONPs have been combined with thermo-
responsive polymers [9–12]. The increase in local temperature
generated by the IONPs might also induce the conformational
change of the polymer network, thus providing the external stimu-
lus for shrinking the polymer networks and release the entrapped
drug molecules [13].
The study and the characterization of such kind of inorganic–
organic hybrid materials when different stimuli are applying either
on cell cultures in vitro or once injected in animal models in vivo,
are quite challenging and require particular care in order to pre-
serve the natural conformation of the nanosystems in the different
environmental conditions. For this reason we applied cryosection-
ing. Indeed, in comparison with conventional EM preparation
methods, the advantages of cryosectioning in processing cells
doped with solvent-sensitive polymeric nanogels are manifold.
Since cryosectioning does involve neither the use of solvents during
sample dehydration, nor the use of resin during the infiltration and
the embedding steps, it better preserves the fine morphology of
polymeric nanogels (see Figs. 1b and 2). Indeed, using cryosection-
ing we did not observe the typical artifacts shown by the polymeric
nanogels when processed by conventional TEM, such as the swell-
ing of the polymer that more than doubled its diameter and the
Cryosectioning Characterization of Cells Doped with Nanoparticle Loaded Nanogels 181
Fig. 1 Nanogel releasing mechanism. (a) Sketch showing the intracellular pathway and the pH-mediated
release mechanism of the siRNA/IONPs loaded nanogels (NG): the endosome acidification promotes the NG
swelling, due to the recall of counter ions and water (the “proton sponge” effect). This leads to the rupture of
the endosomal membrane and to the escape of both siRNAs and IONPs from the endosome to the cytoplasm;
(b) histogram showing the various diameters of the NGs when deposit from water (Bare), or when processed
by cryosectioning, conventional TEM or exposed to acetone. Note that the diameter of the NGs processed by
cryosectioning is similar to that of the bare NGs. Conventional TEM processing as well as incubation with
acetone causes a remarkable increase of the NG diameter. The insets show TEM images of a bare NG (up) and
of a NG after incubation with acetone (bottom). Scale bar are 200 nm
displacement towards the outside of the nanoparticles contained

inside the nanogels (for a comparison between conventional pro-
cessing and cryosectioning TEM characterizations applied to the
same samples see Figs. 1b and 2). In addition, cryosectioning,
enhancing the integrity and visibility of membranes, is particularly
suited for studying nanoparticles uptake and trafficking. Indeed on
cryosections it is easier the identification of membrane bounded
cytoplasmic organelles (i.e., endosomes), which play a crucial
role in the internalization processes (see Fig. 2 and ref. 6).
Fig. 2 TEM images of HeLa cells incubated with acidic pH-responsive nanogels loaded with IONPs using cryo-
sectioning (a, c) and conventional TEM sample preparation (b, d). (a, b) The IONPs loaded nanogels (asterisks)
are approaching the cell membrane (arrowhead). Note that in the sample processed using the cryosectioning
technique the nanogels maintain their morphology. On the contrary, in the sample conventionally processed
the nanogels appear dramatically swollen. (c, d) the IONPs loaded nanogels (asterisks) are inside the cell. Note
that only in the samples processed using the cryosectioning technique the endosomes (arrowheads) are
clearly visible as also the nanogels (asterisks) inside them. cyt cytoplasm, m mitochondria. Scale bars corre-
spond to a length of 200 nm
Besides, cryosections represent an unexpectedly robust and flexible

starting point on which to perform further EM-based analyses,
such as immunocytochemistry, electron tomography (ET), and
Energy Dispersive X-ray Spectroscopy (EDXS) chemical analysis.
On this concern cryosectioning, involving the rapid freezing of the
sample and avoiding liquid media, is a good preparation technique
for successive microanalysis, avoiding the translocation of the ele-
ments of interest within cells or soft tissue that instead occur by
chemical fixation [14]. Therefore, the application of cryosectioning
to cells doped with solvent-sensitive polymeric nanoparticles is
promising, especially given the wide and multidisciplinary EM
characterizations further allowed with minimal structural artifacts.
Fig. 3 Immunogold electron microscopy analysis of Rab7 in cryosection of HeLa cells doped with IONPs loaded
nanogel. (a) Late endosome (arrow) containing IONPs loaded nanogels (asterisk). Note the cluster of gold par-
ticles (arrowhead) close to the endosomal membrane. (b) Late endosomes (arrows) containing IONPs loaded
nanogels (asterisks). Note the gold nanoparticles (arrowheads) lining the endosomal membrane. cyt cyto-
plasm, n nucleus. Scale bars correspond to a length of 100 nm
The main aim of this chapter is to describe the protocols for

preparing cryosections of cells doped with solvent-sensitive poly-
meric nanoparticles for EM-immunocytochemistry, ET and EDXS
chemical analysis in Scanning TEM (STEM). As an example here
we first describe the EM immunocytochemistry protocol we used
to process cryosections of HeLa cells doped with IONPs loaded
nanogels. Using a primary antibody against Rab7 (Sigma Aldrich),
a late endosomal marker, and an appropriate secondary antibody
conjugated to gold nanoparticles having their own contrast under
TEM, we identified and localized the vesicular compartments con-
taining the nanogels (see Fig. 3). Later, we provide an example of
the ET protocol we have established to generate a tomogram on a
semi-thin cryosection (of around 150 nm in thickness) of HeLa
cells doped with IONPs loaded nanogels (see Fig. 4). ET is an elec-
tron microscopy technique based on the collection of several TEM
images of the sample tilted in a wide angular range. The use of
these projections allows to reconstruct the sample 3D-morphology
[15]. Although ET should be in principle a useful tool in nanobio-
technology, allowing the study of nanoparticles in their sub-cellu-
lar context, so far relatively few studies dealing with ET on cells
doped with nanoparticles are reported [16–20]. Moreover ET has
been applied only seldom to labeled and unlabeled cryosections,
despite the advantages of this technique (see above) [5, 21].
Finally, we end this chapter describing the STEM EDXS proto-
col we have established to perform STEM EDXS on cryosections
of HeLa cells, allowing both the localization of the nanogels inside
the cell cytoplasm and the identification of the IONPs inside the
nanogels, via the iron EDXS mapping (see Fig. 5). EDXS is a well-
established technique in biology to perform the chemical analysis
of biological materials using the transmission electron microscope.
Fig. 4 Stages of data acquisition in electron tomography reconstruction. (a, b) a semi-thin (≈150 nm) cryosec-
tion of HeLa cells doped with IONPs nanogel is imaged at respectively 0° (a) and 60° of tilt angle (b) in a Jeol
JEM 2200FS microscope operating at 200 kV of acceleration voltage. Small black dots are 15 nm gold parti-
cles used to align images. (c) A single tomogram slice displays fine detail of membranes and of IONPs decorat-
ing the nanogels (asterisks) inside the endosomes (arrowheads) not apparent in the original semi-thick section
image. (d) A tomogram 3D visualization using the isosurface rendering as implemented in UCSF Chimera
software, version 1.6.2. cyt cytoplasm, n nucleus. Scale bar correspond to a length of 200 nm
This technique allows for the identification of the chemical elements

inside cells or tissues even if they are present in low atomic percent-
age (about 3–4 %) [22]. EDXS makes use of the X-ray spectrum
emitted by the sample when irradiated with a high-energy electron
beam. As the energy of the emitted X-rays is related to the elements
constituting the sample, it can be used to obtain a spatially resolved
chemical analysis. When EDXS is combined with STEM mode, it
allows to measure elemental composition down to sub-micrometer
scale in organic and biological thin sections [23]. In the latter
method, the electron beam is focused in a way to obtain a very
small electron probe (down to nanometer size). It is then raster
scanned across the specimen and at each probe position, the cor-
responding X-ray emission spectrum is recorded. This spectrum
may therefore be used to reconstruct a sample elemental map.
Moreover, compared to EDXS in TEM mode, EDXS via STEM
is more advantageous since the inelastic interaction that gives rise
Fig. 5 EDXS analysis on cryosections of HeLa cells loaded with IONPs loaded nanogels. (a) Scanning transmis-
sion electron microscopy (STEM) image of several endosomes containing numerous IONP loaded nanogels
(arrowheads) inside the cytoplasm of a cryosectioned HeLa cell. The superimposed EDXS iron map, corre-
sponding to the IONPs inside the nanogels, is shown in red. (b) superimposed EDXS iron map on a higher
magnification of a single endosome (arrowhead). Note that the IONPs are mainly around the edge of the
nanogels trapped inside the endosome. (c) STEM image of an endosome containing IONPs loaded nanogels
inside the cytoplasm of a cryosectioned HeLa cell. Three regions of interest, one outside the endosome (blue
square) and two inside it (green and red squares) were selected for compositional mapping as shown in (d–f).
Note that the red square in (c) encloses a IONPs loaded nanogel. cyt cytoplasm. Scale bars correspond to a
length of 700 nm in (a), 100 nm in (b) and 150 nm in (c)
to X-Ray emission is effectively local [24]. In the last years this

technique has been increasingly used in nanobiotechnology to
characterize nanoparticles inside cells and cell compartments [25],
although only recently it has been performed on cryosections [6].
2 Materials
2.1 Major 1. Transmission and Scanning Transmission Electron Microscope

Equipments (TEM/STEM) equipped with a charge-coupled device (CCD)
digital camera (see Note 1).
2. Energy Dispersive X-ray Spectroscopy detector (see Note 2).
3. Tomographic TEM specimen holder (see Note 3).
4. EDXS software (see Note 4).
5. Tomography acquisition software (see Note 5).
6. Software for tomogram reconstruction (see Note 6).
7. Software for tomogram 3D visualization (see Note 7).
8. Carbon vacuum evaporator (see Note 8)
9. Stereomicroscope with light source (see Note 9).
10. Adjustable volume pipette.
11. EM specimen grids (Nickel 200 mesh specimen grids Formvar
coated and carbon sputtered).
12. Loops for the final drying step (see Note 10).
13. Fine forceps (e.g., Dumont Biologie #5 or #7).
14. Filter paper.
15. Parafilm (SPI).
2.2 Chemical 1. Colloidal gold solution of the appropriate size (see Note 11).
and Solution 2. Phosphate buffered saline (PBS).
3. 2 % gelatin in PBS.
4. 0.15 % glycine in PBS.
5. Bovine serum albumin (BSA 1 % and 0.1 % in PBS).
6. Primary antibody with known antigen specificity.
7. Secondary antibody conjugated to colloidal gold
nanoparticles.
8. 1 % glutaraldehyde in PBS.
9. 2 % uranyl acetate in distilled water, pH = 7.5.
10. Methyl cellulose saturated uranyl acetate, 9:1, pH = 4
(see Note 12).
3 Methods
In this paragraph we will describe in detail the various protocols we

established for immunolabeling of cryosections (Subheading 3.1),
ET (Subheading 3.2) and STEM EDXS chemical analysis
(Subheading 3.3) performed on cryosections of HeLa cells doped
with the magnetic pH-responsive nanogels. The preparation pro-
tocol of the pH-responsive nanogels loaded with iron oxide
nanoparticles (IONPs) and the description of the doping protocol
of HeLa cells are reported in ref. 6. A detailed description of the
Tokuyasu cryosectioning technique can be found in refs. 1–4.
3.1 EM The immunolabeling procedure is realized by floating the Formvar

Immunolabeling coated nickel grids (see Note 13) on top of different drops, cryo-
of Cryosections sections facing the drops (see Note 14), using forceps (see Note
of Doped Cells 15). The drops are placed on a clean flat surface (see Note 16).
Note that the backside of the grid should stay dry throughout the
whole procedure (see Note 17).
3.1.1 Prepare 1. Transfer the grids stored at 4 °C on the glass slide with sucrose/
the Sections for the EM methyl cellulose to a cold 2 % gelatin plate. Then warm it
Immunolabeling to 40 °C and keep on the fluid gelatin for at least 20 min
(see Note 18).
2. Transfer the grids in 10–20 mM glycine in PBS for 5 min
(see Note 19).
3. Transfer the grids in 1 % BSA in PBS for 2 min (see Note 20).
3.1.2 EM 1. Dilute primary antibody in 1 % BSA/PBS at a final concentra-

Immunolabeling tion of 2–20 μg/ml (see Note 21).
and Fixation 2. Incubate the grids at 37 °C, on 10 μl drops of the specific pri-
of the Immunoreagents mary antibody solution (see Note 22) for more than 20 min
(see Note 23). For the immunolabeling shown in Fig. 3, as a
marker for late endosomes we used an anti-Rab7 polyclonal
primary antibody (Sigma-Aldrich) produced in rabbit, diluted
1:50 in the blocking solution.
3. Rinse with 0.1 % BSA/PBS, four changes, 2 min each to wick
away the unbounded primary antibody.
4. Transfer the grid on 10 μl drops of the suitable secondary anti-
body (see Note 24) solution conjugated with colloidal gold
nanoparticles of appropriate size (see Note 25). Dilute the sec-
ondary antibody in 0.1 % BSA/PBS and incubate at room
temperature for more than 20 min.
5. Rinse with 0.1 % BSA/PBS, four changes, 2 min each; then
rinse with plain PBS, four changes, 2 min each to wick away
the unbounded secondary antibody. For the immunolabeling
shown in Fig. 3, we used an anti-rabbit secondary antibody

conjugated with 15 nm colloidal gold (Aurion, Delta
Microscopie), diluted 1:100 in the blocking solution.
6. Transfer the grid on 1 % glutaraldehyde in PBS for 5 min
(see Notes 26 and 27).
7. Rinse abundantly with distilled water, more than six changes,
1 min each (see Note 28).
3.1.3 Contrast 1. Transfer the grids on a solution of 2 % uranyl acetate, pH = 7.5

Enhancement and Final for 5 min (see Note 29).
Embedding 2. Rinse briefly with distilled water, two changes.
3. Transfer quickly the grids over two drops of cold uranyl ace-
tate/methyl cellulose (1:9), and leave the sections floating for
5–10 min on a third drop of that mixture. All uranyl acetate/
methyl cellulose drops should be on ice (see Note 30).
3.1.4 Drying 1. Pick up the grid in a wire loop (see Note 31). Push the loop in
the uranyl acetate/methyl cellulose drop at some distance from
the grid, bring it underneath and lift the grid from the uranyl
acetate/methyl cellulose drop. The grid is then in the center of
the loop with excess uranyl acetate/methyl cellulose hanging
underneath.
2. Tilt the loop and grid to an angle of 45–60° (the excess uranyl
acetate/methyl cellulose downwards) and touch the side of the
loop to well absorbing filter paper with the sections facing the
filter paper. The amount of excess uranyl acetate/methyl cel-
lulose should be enough to make contact and disappear into
the filter paper, leaving behind a thin even film on the surface
of the grid, which remains in the loop center (see Note 32).
3.2 Electron Semithin cryosections are coated (on one side) with 10 or 15 nm
Tomography gold particles that serve as fiducial markers in the fine alignment
process during tomogram assembly (see Note 33). To maintain
well preserved the cell ultrastructure is very important that only
the Formvar membrane supporting the cryosections but not the
cryosections themselves contacts the colloidal gold solution
(see Note 34).
3.2.1 Preparing 1. Transfer the grid face up into a drop of colloidal gold solution
Cryosections for Electron of the appropriate size for 1 min. Previously dilute the colloidal
Tomography gold solution 1:3 in bidistilled water (see Note 35).
2. Wash the grid face up three changes to wick away the excess of
gold solution.
3. Carefully dry the grids with filter paper (see Note 36).
3.2.2 Adjust the 1. Identify on your sample the region of interest (ROI) on which
Eucentric Height of Your perform tomography.
Specimen 2. Gradually tilt the stage over a range of ±30° (see Note 37).
If necessary re-center the ROI changing the Z coordinate.
3. Repeat step 2 until the ROI does not move significantly.
3.2.3 Data Collection This step has become more routine with the aid of computer
controlled microscopes.
1. Before the data acquisition, check if your ROI can be tilted
over an angular range of ±60° or more, without appearing of
shadows due to the TEM grid.
2. Acquire your data while your specimen is tilted over the chosen
angular range. TEM images are acquired every 1° or 2° to
complete a tilt series.
3. Before capturing each image always center your ROI and check
the focus (see Note 38).
3.2.4 Tomogram 1. The tilt series can be collected and/or finally recorded as a
Reconstruction single image stack file (see Note 39), containing all the projec-
tions of the sample taken in correspondence of the different
tilting angles (see Note 40).
2. Use the software interface Etomo in IMOD version 4.5 [26]
(or similar) to generate a one-axis tomogram from the image
stack (see Note 41). Some of the major steps include: (a) align-
ing the images in the stack for cross correlation generating a
coarse aligned stack; (b) creating a fiducial model based on the
position of the fiducial gold particles; (c) using the fiducial
model, generating a finely aligned stack (see Note 42); and
finally (d) generating the tomogram by weighted back projec-
tion (WBP) or simultaneous iterative reconstruction technique
(SIRT).
This process can be repeated on adjacent serial sections to
capture 3D structural data through a greater volume of the cell.
3. Scan through the entire tomogram with the “Viewing mode”
present in 3dmod, version 4.5, as implemented in IMOD. This
is the first way to understand the volume of data, and to grasp
the 3D arrangement of the structures of interest within the
tomogram.
3.2.5 3D Reconstruction Manual and computer assisted segmentation allows the isolation,
Based on Isosurface identification and visualization of the ROIs inside a tomogram.
Rendering Although manual segmentation, i.e., the hand drawing of contours
in each slice of a tomogram, can be influenced from several biases
(see Note 43), it is widely used with biological samples due to the
complex shape and low contrast characteristic of biological struc-
tures [28]. Thanks to the good membrane contrast present in
tomograms generated from cryosections, it is possible to generate

3D reconstruction of complex structures avoiding manual segmen-
tation, using the less time consuming isosurface rendering. By this
procedure the segmentation occurs from gray level volumes using
a visually estimated gray level threshold. As an example, we used
the isosurface rendering as implemented in the UCSF Chimera
software, version 1.6.2 [27] (see Note 44) to the 3D visualization
of the tomogram generated from a 150 nm thick cryosection of
HeLa cells doped with IONPs loaded nanogels (see Fig. 4).
1. Opening your tomogram with Chimera will automatically start
“Volume Viewer” window.
2. In “Volume Viewer” select the rendering mode to use for 3D
tomogram visualization using the “Style” option (see Note
45). For the 3D rendering shown in Fig. 4d we used the solid
mode, in which the data are shown as a semitransparent solid.
3. Set the “step” setting on the “Volume Viewer” window to
value below 4 (see Note 46).
4. Finely tune the gray-level thresholds dragging horizontally or
vertically the small squares and connecting lines on the gray
level histogram in the “Volume Viewer” window to optimize
the 3D rendering.
5. Under the “Volume Viewer” window, use the “Planes” section
to display a slab of the tomogram, normal to a specified axis
(default Z axis). Change the slab thickness by editing the
“Depth” value.
6. Set the “step” setting on the “Volume Viewer” window to 1
for a full resolution 3D rendering.
3.3 Energy One of the advantages of using STEM for EDXS analysis is that the
Dispersive X-ray electron probe can be used to irradiate a very small area of the
Spectroscopy (EDXS) specimen, whose the size is very close to that of the electron probe
Analysis in STEM: used. Thus, the X-rays obtained can be spatially resolved at a reso-
HAADF Mode lution corresponding to the probe size.
3.3.1 Preparing During EDXS analysis the samples are heavily irradiated: for this
Cryosections for EDXS reason the cryosections should be adequately protected against
Analysis beam damage (see Note 47).
1. Sputter coat the cryosections with 20–30 nm of carbon
(see Note 48). Indeed carbon coating has a protective effect
reducing mass loss due to electron beam irradiation [29].
2. Fill the liquid nitrogen cold trap of the microscope before
starting with the analysis to minimize the microscope column
contamination.
3. Lowering the temperature of the specimen, if possible, can
reduce its sensitivity to structural damage and mass loss [29].
Fig. 6 Ronchigram from an amorphous specimen. Modified from [23]
3.3.2 Adjust the In STEM mode, the electron optics that determines the spot size
Microscope Alignment in and focus is located above the specimen, together with the scan
STEM: HAADF Mode coils. The relevant stigmators for STEM mode are thus the con-
denser stigmators, while the STEM camera length is controlled by
the electron optics located below the specimen.
1. Without condenser apertures, select “spot mode” for a station-
ary electron probe. Then optimize the spot size, mainly con-
sidering that reducing the spot size increases the resolution,
but decreases the signal to noise ratio (see Note 49).
2. At high magnification (more than 500 k), in standard focus
condition, with the electron probe not scanning, use the
Z-height control to adjust the axial coordinate Z of the sample
until you see the Ronchigram, a sort of fish-eye view of the
specimen that provides the best way to perform a fast STEM
alignment (see Fig. 6). Center the Ronchigram using the pro-
jector lens shift.
3. If the microscope is well aligned, the Ronchigram from an
amorphous specimen should be circularly symmetric (see Fig. 6
and Note 50). If it is elongated, correct the condenser astig-
matism using the corresponding stigmators.
4. Insert the High Angle Annular Dark Field (HAADF) or Bright
Field (BF) detector (the corresponding camera length should
be already selected), to form the STEM image with only the
electrons scattered at high angle or transmitted, respectively.
Finally, center the appropriate condenser aperture with respect
to the Ronchigram.
5. Start the beam scanning, and carefully adjust the image bright-
ness and contrast (see Note 51).
3.3.3 Operating the 1. Insert the EDXS detector inside the microscope column using
Analytical Transmission the suitable EDXS software.
Electron Microscope 2. Search the specimen at low magnification; and try to not pre-
expose the region you want analyze at higher magnification.
3. Find an interesting area of your sample and set it to the eucentric

height (see Subheading 3.2.2).
4. Use suitable spot size and condenser aperture to get reason-
able X-ray counts and dead time (see Note 52). If necessary tilt
the specimen towards the EDX detector to increase the counts
per second (see Note 53).
3.3.4 Spectrum The purpose of qualitative analysis is to uncover the chemical ele-
Acquisition and Qualitative ments composing a specimen by identifying the lines in the X-ray
Chemical Analysis spectrum. The possible ambiguities in assigning the element cor-
responding to a given line can be resolved by taking into account
additional lines.
1. Set—in your suitable EDXS software—the appropriate energy
range depending on the atomic mass of the elements you are
looking for in your sample. For light elements (B, NA, etc.)
use low maximum energy (such as 10 or 20 KeV) to increase
resolution, as usually the number of usable channels is fixed
and independent from maximum energy.
2. Set—in your suitable EDXS software—a spectrum acquisition
time to get reasonable X-ray counts. If possible, set the acquisi-
tion time as “live time” so that the acquisition will stop after
the dead time-corrected acquisition time is elapsed.
3. Start a spectrum acquisition from the selected ROI using the
suitable EDXS software (see Note 54). As an internal control,
move into a hole and obtain a spectrum: EDXS counts from
vacuum should be normally very low (dead time <2 %).
4. Assign all spectrum peaks present in the X-ray spectrum to
characteristic X-rays peaks of chemical elements possibly con-
tained in the sample. If a good match between the displayed
line markers corresponding to the expected elements and the
spectrum peaks can be achieved, be confident that the identifi-
cation is correct. In case of doubt, a subsequent quantification
run should clarify the situation (see Note 55).
3.3.5 Elemental Mapping During elemental mapping acquisition, a point-by-point X-ray

Acquisition spectrum acquisition is performed to determine the localization
(and possible amount) of the chemical elements in the sample.
Each resulting element map represents the two-dimensional distri-
bution of the regarding element over the sample area scanned by
the electron probe.
1. Observe the input count rate and check if appropriate. Set up
image/map resolution and dwell time.
2. Capture a STEM image of the sample, so that the program can
recognize the used analytical conditions (spatial coordinates,
magnification, etc.). Eventually select on the image the area to
be mapped.
3. Define the chemical elements for which maps are to be

recorded. If no prior knowledge of the sample composition is
available, the elements can be identified acquiring an average
spectrum of the mapping area. In principle, the chemical maps
could be reconstructed even after the end of the localized
EDXS spectrum acquisition.
4. Start to acquire the map. After a measurement is started, the
mapped images should gradually improve in signal to noise ratio.
4 Notes
1. We use a Jeol JEM 2200FS microscope, equipped with a field-

emission gun (FEG) and operating at 200 kV of acceleration
voltage. The images were recorded using a 1 k × 1 k Gatan
charge-coupled device (CCD) camera.
2. We use a Bruker Nano’s XFlash 5060T spectrometer (Bruker)
equipped with a 60 mm2 silicon drift detector (SDD).
3. We use an ultra-narrow gap tomography holder (Fischione).
4. We use the QuanTax Esprit 1.9 software (Bruker).
5. We use the Jeol Recorder tomography acquisition software,
version 2.26 (Jeol).
6. We use Etomo in IMOD, version 4.5 [26].
7. We use UCSF Chimera software, version 1.6.2 [27].
8. We use the Emitech K950X carbon vacuum evaporator
(Emitech).
9. We use a Leica MZ6 (Leica).
10. The loop, 3.5 mm in diameter, should be made from copper or
platinum wire 0.2 mm diameter and mounted on Eppendorf
tips.
11. We use 15 nm Gold Sol (Aurion).
12. Add to 90 ml of the methyl cellulose solution 10 ml of 4 %
uranyl acetate and mix gently. Centrifugate the solution for
95 min at 90,000 × g (4 °C). The supernatant is poured into
50 ml vials with screwcap and can be stored at 4 °C in the dark
for about 3 months.
13. For a detailed description on how to make Formvar coated
nickel grids please refer to refs. 1, 3.
14. Drops of 100–200 μl, which can carry up to five grids at once,
are used for the majority of rinsing steps. Use drops of 5–10 μl
for each grid for antibody and immuno-gold solutions.
15. The grids may be transferred also using wire loops. But using
forceps washing is more efficient since less of the incubation
fluid is transferred to the next drop.
16. The flat surface is obtained by using a parafilm sheet that is

adhered to a glass plate by some drop of distilled water.
17. If a grid accidentally sinks in a drop, wash it in distilled water.
Then dry the back side of the grid carefully wiping with filter
paper. Before reentering the incubation procedure, let the grid
float on clean distilled water for a short while.
18. The aim of this procedure is to make assessable the section to
the incubation solutions removing the dried mixture of sucrose
and methyl cellulose from the section surface.
19. The aim of this procedure is to quench free aldehyde groups.
When non-specific binding becomes a problem, you can extend
this step.
20. The aim of this procedure is to block other sites where the
immuno-reagents may stick non specifically. When non-specific
binding becomes a problem, you can extend this step.
21. When non-specific binding becomes a problem, the antibody
may be further diluted (~10 times).
22. To avoid that the drop dry out easily, it is important to cover
the drops by a small petri dish in a whet chamber.
23. The duration of the immuno-incubation is not very critical. It
is important to consider that molecules that do not react with
aldehydes (like membrane lipids) and soluble proteins may
escape easily when the fixation is week: in this case is better to
keep the incubation short (i.e., 20 min). On the contrary, pro-
teins that are integrated in membranes, or bound to the cyto-
skeleton will not easily escape, and labeling of these may be
favored by longer incubation periods.
24. Remember that the secondary antibody should be against the
species that primary antibody is raised. For example, if the pri-
mary antibody is raised in rabbit, an anti-rabbit secondary anti-
body should be used.
25. It is extremely important to carefully select the appropriate size
for the gold nanoparticles conjugated to the secondary anti-
body. Indeed, the gold nanoparticles should be clearly distin-
guishable from other possible nanoparticles present inside the
sections. As an example we used a secondary antibody conju-
gated with 15 nm colloidal gold nanoparticles on cryosections
of HeLa cells doped with nanogels containing 10 nm iron
oxide nanoparticles (see Fig. 3).
26. Do this step in a fume-hood since the glutaraldehyde is toxic.
27. The aim of this step is an extra fixation of the immuno-reagents
on the sections before the rinsing and contrast -staining part of
the procedure.
28. The aim of this task is to remove the phosphate or other ions
thoroughly from the sections in order to prevent uranyl
precipitates during the final staining and embedding. Use fresh

and good quality distilled water.
29. The purpose of this step is to enhance contrast and stabilize
membrane lipids.
30. The purpose of this step is to stain the sections for contrast and
to support them by polymers (methyl cellulose) in order to
prevent drying artifacts.
31. The loop should have an inner diameter between 3.5 and
4 mm.
32. Be careful with the heavy metal uranium in this part of the
procedure. Discard all remnants properly. Think of the pieces
of filter paper with absorbed UA/MC.
33. It is extremely important to carefully select the appropriate size
for the gold nanoparticles used as fiducial markers. Indeed, to
avoid mistakes during the tomogram alignment, the gold
nanoparticles should be clearly distinguishable from other pos-
sible nanoparticles present inside the sections. As an example
we used 15 nm colloidal gold nanoparticles as fiducial markers
on cryosections of HeLa cells doped with nanogels containing
10 nm iron oxide nanoparticles (see Fig. 4).
34. The cryosection ultrastructure will be deeply deteriorated if
the cryosections come in contact with the colloidal gold solu-
tion since they are not embedded in resin.
35. To find out the optimal dilution for the colloidal gold solution,
try different dilution on your sections, considering that for the
fine alignment of your image stack during the tomogram gen-
eration procedure a number of gold nanoparticles between 20
and 40 should be present.
36. Remove almost all the solution with filter paper by carefully
blotting the edges of each grids and then allow them to dry for
a few minutes.
37. During eucentric height adjustment tilt your specimen at small
angular steps to avoid that your ROI is shifted outside the area
you are observing.
38. The vast majority of tomographic software have a function that
automatically center (autotracking) and focus (autofocus) your
ROI.
39. The file containing the projections of your sample has the fol-
lowing interchangeable file extension (i.e., tmg, st, or mrc)
depending on the tomographic software you are using.
40. Consider that using the programs TIF2MRC or DM2MRC as
implemented in Imod 4.5 you can easily convert a series of
tilted images (in TIF or DM3 format) in a single file (see
http://bio3d.colorado.edu/imod/doc/program_listing.
html).
41. At the IMOD home page (http://bio3d.colorado.edu/

imod/#Guides) you can find useful tutorials, including also
videos that describe in details how to generate tomograms
using ETOMO.
42. The program Tiltalign uses the position of the gold particles in
the fiducial model and a variable metric minimization approach
to solve for displacements, rotations, tilts, and magnification
differences in the tilted projections.
43. Since in manual segmentation the contours are drawn from the
operator, this approach is time consuming and, to a certain
degree, also subjective.
44. UCSF Chimera is freely available at: http://www.cgl.ucsf.
edu/chimera/
45. The rendering modes available in Volume Viewer are “sur-
face,” “mesh,” or “solid”. See the Chimera User’s Guide
(freely available at http://www.cgl.ucsf.edu/chimera/cur-
rent/docs/UsersGuide/framecore.html) for further details.
46. The step setting controls sampling density. Consider that lower
step setting values mean better resolution but higher computa-
tional time. See the Chimera User’s Guide for further details.
47. Consider that the uranyl acetate used to stain the cryosections
enhances their resistance to irradiation.
48. Usually we perform several depositions of carbon using low
evaporation time (1,000–1,500 ms), thus avoiding damaging
the cryosections.
49. Using a spot size between 0.7 and 1 nm was found to be a
good compromise between high X-Ray cont rate and spatial
resolution.
50. Looking at the Ronchigram (see Fig. 6), you should be able to
see (1) a central circular area containing a normal shadow
image of the sample at rather high magnification; (2) a ring
where everything is streaked out in the radial direction, the
ring of infinite radial magnification; and (3) at higher radius,
another ring where everything is streaked out into a circular
pattern, the ring of infinite azimuthal magnification.
51. A line scan can be used to help adjust the brightness and con-
trast. The line scan is a plot of the image density on a line
across the sample. The image density should vary across the
full dynamic range of the monitor, without white or black
areas, that indicate that the image is saturated at high or low
intensity. Consider that tuning the contrast you are expanding
or decreasing the amount of variation in intensity. On the con-
trary with “brightness” you are increasing or decreasing the
overall signal level. Thus, maximize the amount of contrast so
that the full range of intensities is used, but without saturating
the image. Then adjust the brightness so that the overall image
intensity is correct.
52. With high spot size and large aperture the counts per seconds
increases, but the resolution of the image decreases.
53. Usually tilting the specimen towards the EDX detector of
10–20° is recommended in order to maximize the X-Ray signal
and to avoid shadow effects from the specimen grid.
54. With QuanTax Esprit 1.9 to start a spectrum acquisition click
the acquire bottom in the Spectra workspace.
55. In the majority of EDX software there is a “finder options” for
the automatic identification of the unknown peaks of the
spectrum.
References
1. Tokuyasu KT (1973) A technique for ultracry- nanothermotherapy. Nanomedicine 3(6):
otomy of cell suspensions and tissues. J Cell 831–844
Biol 57:551–565 9. Pradhan P, Giri J, Rieken F, Koch C, Mykhaylyk
2. Liou W, Geuze HJ, Slot JW (1996) Improving O, Döblinger M, Banerjee R, Bahadur D,
structural integrity of cryosections for immu- Plank C (2010) Targeted temperature sensitive
nogold labeling. Histochem Cell Biol 106: magnetic liposomes for thermo-chemotherapy.
41–58 J Control Release 142(1):108–121
3. Webster P (1999) The production of cryosec- 10. Liu T, Liu K, Liu D, Chen S, Chen I (2009)
tions through fixed and cryoprotected biological Temperature-sensitive nanocapsules for con-
materials and their use in immunocytochemistry. trolled drug release caused by magnetically
In: Nasser Hajibagheri MA (ed) Electron triggered structural disruption. Adv Funct
microscopy methods and protocols, vol 117. Mater 19(4):616–623
Humana Press, Totowa, NJ, pp 49–76 11. Deka S, Quarta A, Di Corato R, Riedinger A,
4. Posthuma G, van Donselaar E, Griffith J, Cingolani R, Pellegrino T (2011) Magnetic
Oorschot VMJ, van Dijk S, Slot JW (eds) nanobeads decorated by thermo-responsive
(2006) Ultrathin cryo-sectioning and immuno PNIPAM shell as medical platforms for the
– gold labeling. A practical introduction. Cell efficient delivery of doxorubicin to tumour
Microscopy Center, Department of Cell cells. Nanoscale 3(2):619–629
Biology, University Medical Center Utrecht, 12. Chen S, Li Y, Guo C, Wang J, Ma J, Liangs X,
The Netherlands Yang L, Liu H (2007) Temperature-responsive
5. Murk J, Postuma G, Koster AJ, Guuze HJ, magnetite/PEO-PPO-PEO block copolymer
Verkleij AJ, Kleijmeer MJ, Humbel BM (2003) nanoparticles for controlled drug targeting
Influence of aldehyde fixation on the morphol- delivery. Langmuir 23(25):12669–12676
ogy of endosomes and lysosomes: quantitative 13. Louguet S, Rousseau B, Epherre R, Guidolin N,
analysis and electron tomography. J Microsc Goglio G, Mornet S, Duguet E, Lecommandoux
212:81–90 S, Schatz C (2012) Thermoresponsive polymer
6. Curcio A, Marotta R, Riedinger A, Palumberi brush-functionalized magnetic manganite
D, Falqui A, Pellegrino T (2012) Magnetic nanoparticles for remotely triggered drug
pH-responsive nanogels as multifunctional release. Polym Chem 3:1408–1417
delivery tool for small interfering RNA mole- 14. Morgan AJ, Winters C, Stürzenbaum S (1999)
cules and iron oxide nanoparticles. Chem X-ray microanalysis techniques. In: Hajibagheri
Commun 28(48):2400–2402 MAN (ed) Electron microscopy methods and
7. Laurent S, Dutz S, Häfeli U, Mahmoudi M protocols, vol 117. Humana Press, Totowa,
(2011) Magnetic fluid hyperthermia: focus on NJ, pp 245–276
superparamagnetic iron oxide nanoparticles. 15. Frank J (2005) Electron tomography. Methods
Adv Colloid Interface Sci 166:8–23 for three-dimensional visualization of struc-
8. Gazeau F, Lévy M, Wilhelm C (2008) tures in the cell, 2nd edn. Springer, Albany,
Optimizing magnetic nanoparticle design for NY
16. Porter AE, Muller K, Skepper J, Midgley P, Experimental techniques and applications.
Welland M (2006) Uptake of C60 by human Cambridge University Press, Cambridge
monocyte macrophages, its localization and 23. Laquerriere P, Banchet V, Michel J, Zierod K,
implications for toxicity studied by high reso- Balossier G, Bonhomme P (2001) X-ray
lution electron microscopy and electron microanalysis of organic thin sections in TEM
tomography. Acta Biomater 2:406–419 using an UTW Si(Li) detector: comparison of
17. Sousa AA, Aronova MA, Kim YC, Dorward quantification methods. Microsc Res Tech 52:
LM, Zhang G, Leapman RD (2007) On the 231–238
feasibility of visualizing ultrasmall gold labels 24. D’Alfonso AJ, Freitag B, Klenov D, Allen LJ
in biological specimens by STEM tomography. (2010) Atomic-resolution chemical mapping
J Struct Biol 159:507–522 using energy-dispersive x-ray spectroscopy.
18. Uchida M, Willits DA, Muller K, Willis AF, Phys Rev 81:100101(R)
Jackiw L, Jutila M, Young MJ, Porter AE, 25. Busch W, Bastian S, Trahorsch U, Iwe M,
Douglas T (2008) Intracellular distribution of Kuhnel D, Meißner T, Springer A, Gelinsky
macrophage targeting ferritin–iron oxide M, Richter V, Ikonomidou C, Potthoff A,
nanocomposite. Adv Mater 21(4):458–462 Lehmann I, Schirmer K (2011) Internalisation
19. Cai X, Chen HH, Wang CL, Chen ST, Lai SF, of engineered nanoparticles into mammalian
Chien CC, Chen YY, Kempson IM, Hwu Y, cells in vitro: influence of cell type and particle
Yang CS, Margaritondo G (2011) Imaging the properties. J Nanopart Res 13:293–310
cellular uptake of tiopronin-modified gold 26. Kremer JR, Mastronarde DN, McIntosh JR
nanoparticles. Anal Bioanal Chem 401: (1996) Computer visualization of three-
809–816 dimensional image data using IMOD. J Struct
20. Nair BJ, Fukuda T, Mizuki T, Hanajiri T, Biol 116(1):71–76
Maekawa T (2012) Intracellular trafficking of 27. Pettersen EF, Goddard TD, Huang CC,
superparamagnetic iron oxide nanoparticles Couch GS, Greenblatt DM, Meng EC, Ferrin
conjugated with TAT peptide: 3-dimensional TE (2004) UCSF Chimera – a visualization
electron tomography analysis. Biochem system for exploratory research and analysis.
Biophys Res Commun 421:763–767 J Comput Chem 13:1605–1612
21. Vicidomini G, Gagliani MC, Cortese K, 28. Donohoe BS, Mogelsvang S, Staehelin LA
Krieger J, Buescher P, Bianchini P, Boccacci P, (2006) Electron tomography of ER, Golgi and
Tacchetti C, Diaspro A (2010) A novel related membrane systems. Methods 39:
approach for correlative light electron micros- 154–162
copy analysis. Microsc Res Tech 73:215–224 29. Egerton RF, Li P, Malac M (2004) Radiation
22. Sigee DC, Morgan AJ, Sumner AT, Warley A damage in the TEM and SEM. Micron 35:
(eds) (1993) X-ray microanalysis in biology. 399–409
Part III
Implementing Bio–Nano Interfaces

Chapter 15
Zwitterion Siloxane to Passivate Silica Against Nonspecific

Protein Adsorption
Zaki G. Estephan and Joseph B. Schlenoff
Abstract
Passivating surfaces against protein adsorption is important for many biotechnological applications.
Current approaches have been exploiting the use of zwitterions instead of poly(ethylene glycol) (PEG).
Commonly used zwitterions are polymeric and are grafted onto surfaces using specialized polymerization
techniques. Here we describe the synthesis of a monomeric zwitterion siloxane and its covalent attachment
to silica surfaces (nanoparticle and planar) in a one-pot one-step aqueous method requiring no catalyst.
Key words Silane, Silica nanoparticle, Silicon wafer, Non-fouling, Protein repellent surface
1 Introduction
Modifying surfaces with zwitterions, instead of poly(ethylene

glycol), to repel proteins and prevent bacterial adhesion (Fig. 1)
has gained interest lately. The effectiveness of different zwitterionic
chemical functionalities, i.e., sulfobetaine, carboxybetaine, or
phosphobetaines, against protein or bacterial adhesion has been
studied [1, 2]. Both carboxybetaines and sulfobetaines have shown
interesting resistive properties and have been therefore extensively
used to modify nanoparticles for dual colloidal stability and pro-
tein resistivity [3]. Most approaches have focused on polymeric
zwitterions. For planar substrates, surface passivation has been as
simple as dipping the substrate in the polymeric zwitterion or using
thiol self assembled monolayers (SAMs) in the case of gold sub-
strates. For example, Holmlin et al. used zwitterionic SAMs to
moderate the adsorption of proteins onto gold passivated surfaces
[4]. In contrast, Salloum et al. used the layer-by-layer technique to
coat glass coverslips with poly(allylamine hydrochloride), PAH,
and poly(acrylic acid)-co-poly(3-[2-(acrylamido)-ethyl dimethyl-
ammonio] propane sulfonate), PAA-co-PAEDAPS, to control cell
adhesion by changing the ratio of the zwitterion in the random
201
202 Zaki G. Estephan and Joseph B. Schlenoff
Fig. 1 Scheme of zwitterion passivation of silica nanoparticles against nonspecific protein adsorption
copolymer [5]. Other techniques involve surface initiated atom

transfer polymerization (ATRP) onto an initiator covered gold or
glass substrates [6, 7]. Most of these approaches have been
extended to nanoparticles. For example, Rouhana et al. prepared
zwitterated gold nanoparticles by exchanging the citrate ligand
with a zwitterion disulfide [8], whereas Jia et al. and Matsuura
et al. used surface initiated ATRP to graft zwitterionic polymer to
gold and silica nanoparticles respectively [9, 10].
The authors have previously reported a method for passivating
silica surfaces against nonspecific protein interaction [11, 12].
In this chapter we provide a step-by-step procedure to functional-
ize silica nanoparticles and planar silicon wafers with a zwitterion
siloxane. The reaction is based on covalent attachment of the silox-
ane group of the zwitterion to the surface silanols of the
substrate.
2 Materials
Prepare all solution with ultrapure water (18 MΩ) at room tem-
perature. Follow local and federal waste disposal regulations when
disposing of materials.
2.1 Zwitterion 1. HPLC grade acetone.

Siloxane Synthesis 2. Propane sultone from TCI America.
Components
3. (N,N-dimethyl-3-aminopropyl)trimethoxysilane from Gelest.
Zwitterion Passivation of Silica Nanoparticles 203
2.2 Nanoparticle 1. Ludox® TM-40 colloidal silica from Sigma Aldrich.

Zwitteration 2. Dialysis was performed using Spectra/Pro dialysis tubing with
Components molecular weight cutoff (MWCO) of 3,500.
2.3 Planar Surface 1. RCA solution: mix 1 ml of 28 % NH4OH with 1 ml of 30 %

Zwitteration H2O2 with 5 ml H2O.
Components 2. Piranha solution (Caution: strong oxidizer, not to be stored in
closed containers): mix 7 ml of concentrated H2SO4 with 3 ml
of 30 % H2O2. Let the solution cool for 15 min before use.
3 Methods
3.1 Synthesis of The following procedure is performed under inert conditions at

Zwitterion Siloxane: room temperature:
3-(Dimethyl(3- 1. Dissolve 4.45 g of propane sultone in 37 ml of acetone.
(Trimethoxysilyl)
2. Add 7.5 g of (N,N-dimethyl-3-aminopropyl)trimethoxysilane
Propyl)-Ammonio) (see Note 1).
Propane-1-Sulfonate
(Fig. 2)
3. Let the solution stir for 6 h. The solution turns turbid after few
minutes and eventually precipitates a white solid.
4. Wash the solution extensively with acetone. The precipitate is
first dispersed in acetone and then poured into a suction filtra-
tion flask. The process is repeated three times or until the ace-
tone remains colorless after filtration.
5. Let the precipitate dry overnight under Ar.
3.2 Zwitteration of The following procedure is performed for Ludox TM-40 silica sus-
Silica Nanoparticles pension (40 wt% suspension in water) but can be applied to other
silica nanoparticles.
1. Dilute the silica nanoparticle to a final concentration of 10 wt%
by adding 3.75 g of Ludox TM-40 to around 10 g of water.
2. Disperse the proper amount of zwitterion siloxane (Fig. 1) in
around 1 g of water (see Notes 2 and 3).
3. Add the dissolved zwitterion siloxane to the silica
nanoparticles.
4. Bring the final mass of the solution to 15 g by addition of water.
5. Shake the solution and introduce to a heated bath preset at
80 °C.
Fig. 2 Chemical structure of zwitterion siloxane: 3-(dimethyl(3-(trimethoxysilyl)

propyl)-ammonio)propane-1-sulfonate
204 Zaki G. Estephan and Joseph B. Schlenoff
6. Let the reaction proceed for 6 h.

7. After cooling to room temperature dialyze the system against
4 l of water for 24 h, changing the dialysate at 6 h intervals.
8. Measure the solids content after dialysis by drying around 1 ml
of the dialyzed solution under vacuum at 25–30 °C.
3.3 Zwitteration of 1. Immerse the silicon wafer into an RCA solution for 10 min.
Planar Silicon Wafer 2. Rinse extensively with water.
3. Immerse the wafer into cold piranha for 10 min.
4. Wash with water.
5. Insert the silicon wafer into a sealable flask containing 100 mM
of zwitterion siloxane (0.165 g zwitterion siloxane in 5 ml of
H2O).
6. Seal the flask tightly to avoid the evaporation of water and heat
the solution at 80 °C for 15 h.
7. After cooling to room temperature, wash the silicon wafer with
water.
4 Notes
1. Equivalent number of moles of propane sultone and (N,N-

dimethyl-3-aminopropyl)trimethoxysilane are used. Each cor-
responds to 1 M concentration.
2. The zwitterion should readily dissolve in water to give a color-
less solution. The presence of color indicates impurities or
unreacted components from synthesis.
3. For the silica nanoparticles used in our laboratory (specific sur-
face area of 140 m2/g) 0.19 g of zwitterion siloxane was
needed to saturate the surface. In our calculations it was
assumed that 1 mol of zwitterion siloxane reacts with 3 mol of
surface silanol to form a monolayer. Surface saturation was
observed close to the theoretical monolayer as defined here.
The recommended amount of siloxane is thus 8.9 × 10−4mS
where m is the mass of nanoparticlesing and S is their specific
surface area in m2/g.
References
1. West SL, Salvage JP, Lobb EJ, Armes SP, Interaction of blood plasma with antifouling
Billingham NC, Lewis AL, Hanlon GW, Lloyd surfaces. Langmuir 25:6328–6333
AW (2004) The biocompatibility of crosslink- 3. Zhang Z, Chao T, Chen SF, Jiang SY (2006)
able copolymer coatings containing sulfobeta- Superlow fouling sulfobetaine and carboxybe-
ines and phosphobetaines. Biomaterials taine polymers on glass slides. Langmuir
25:1195–1204 22:10072–10077
2. Emmenegger CR, Brynda E, Riedel T, 4. Holmlin RE, Chen XX, Chapman RG,
Sedlakova Z, Houska M, Alles AB (2009) Takayama S, Whitesides GM (2001)
Zwitterion Passivation of Silica Nanoparticles 205
Zwitterionic SAMs that resist nonspecific 8. Rouhana LL, Jaber JA, Schlenoff JB (2007)
adsorption of protein from aqueous buffer. Aggregation-resistant water-soluble gold
Langmuir 17:2841–2850 nanoparticles. Langmuir 23:12799–12801
5. Salloum DS, Olenych SG, Keller TCS, 9. Jia G, Cao Z, Xue H, Xu Y, Jiang S (2009)
Schlenoff JB (2005) Vascular smooth muscle Novel zwitterionic-polymer-coated silica
cells on polyelectrolyte multilayers: nanoparticles. Langmuir 25:3196–3199
hydrophobicity-directed adhesion and growth. 10. Matsuura K, Ohno K, Kagaya S, Kitano H
Biomacromolecules 6:161–167 (2007) Carboxybetaine polymer-protected
6. Zhang Z, Chen SF, Chang Y, Jiang SY (2006) gold nanoparticles: high dispersion stability and
Surface grafted sulfobetaine polymers via atom resistance against non-specific adsorption of
transfer radical polymerization as superlow proteins. Macromol Chem Phys 208:862–873
fouling coatings. J Phys Chem B 110: 11. Estephan ZG, Jaber JA, Schlenoff JB (2010)
10799–10804 Zwitterion-stabilized silica nanoparticles:
7. Cheng G, Li GZ, Xue H, Chen SF, Bryers JD, toward nonstick nano. Langmuir 26:
Jiang SY (2009) Zwitterionic carboxybetaine 16884–16889
polymer surfaces and their resistance to long- 12. Estephan ZG, Schlenoff PS, Schlenoff JB
term biofilm formation. Biomaterials 30: (2011) Zwitteration as an alternative to
5234–5240 PEGylation. Langmuir 27:6794–6800
Chapter 16
Preparation and Characterization of DNA Block Copolymer

Assemblies Loaded with Nanoparticles
Xi-Jun Chen, Robert J. Hickey, and So-Jung Park
Abstract
We have recently developed a universal procedure to functionalize inorganic nanoparticles with a dense
layer of DNA through the self-assembly of DNA block copolymers and nanoparticles. This functionaliza-
tion strategy allows one to combine the useful physical properties of inorganic nanoparticle with the
enhanced DNA binding properties that originate from the high surface DNA density. In particular, the
hybrid nanostructures exhibit orders of magnitude higher binding constants than regular DNA strands.
This chapter presents a detailed protocol for the preparation and characterization of DNA block copoly-
mer assemblies loaded with nanoparticles.
Key words DNA, Block copolymer, Nanoparticles, Self-assembly
Abbreviations
CPG Controlled pore glass
DLS Dynamic light scattering
DMF N,N-Dimethylformamide
DNA-b-PS Block copolymer of DNA and polystyrene
FAM Fluorescein
FRET Förster/Fluorescent Resonance Energy Transfer
MNP Magnetic nanoparticles
MNP@PS@DNA DNA-b-polystyrene assemblies loaded with magnetic nanoparticles
PBS Phosphate buffer saline
PS Polystyrene
PS-MNP Polystyrene modified magnetic nanoparticles
PS@DNA Self assembly of DNA-b-polystyrene
PTFE Polytetrafluoroethylene
TEM Transmission electron microscope
THF Tetrahydrofuran
207
208 Xi-Jun Chen et al.
1 Introduction
Bio-conjugated nanoparticles have been actively exploited in

various biological and medical applications for the past couple of
decades to take advantage of the useful physical properties of
inorganic nanoparticles in detecting, studying, and manipulating
biological systems [1–6]. In particular, DNA-modified gold
nanoparticles have shown great promise in many biomedical
applications such as DNA detection, drug delivery, and gene reg-
ulation owing to the unique optical properties of gold nanopar-
ticles and the densely immobilized DNA strands on the
nanoparticle surface [7, 8]. We have recently developed a new
DNA-functionalization method based on the self-assembly of
DNA block copolymers and nanoparticles [9]. In this approach,
nanoparticles are encapsulated in DNA block copolymer assem-
blies by the slow addition of water to the mixture of DNA block
copolymers and nanoparticles in a polar organic solvent. The
resulting assemblies are composed of nanoparticles embedded in
the hydrophobic polymer core and the hydrophilic DNA shell at
the exterior. Since this approach is based on self-assembly, it can
be applied to virtually any types of nanoparticles. It is also impor-
tant to note that this methodology leads to a densely packed
DNA layer on the surface because every polymer strand is conju-
gated to DNA. Due to the high density, the DNAs on the assem-
blies show dramatically enhanced binding properties. For
example, they can recognize complementary DNA at low salt
concentrations where regular DNA strands do not form duplex
structures. While DNA block copolymers have been previously
synthesized for other purposes [10–13], this work was the first to
demonstrate that the self-assembly of nanoparticles and DNA
block copolymers can be used to prepare DNA-functionalized
nanoparticles with enhanced binding properties. This enhanced
binding property in combination with the capability to load
nanoparticles and other small molecules makes the DNA assem-
blies an excellent material for DNA detection and delivery
applications.
Here, we present detailed procedures on (1) the synthesis and
characterization of DNA block copolymers, (2) the self-assembly
of DNA block copolymers and nanoparticles, and (3) the charac-
terization of nanoparticle-loaded DNA block copolymer assem-
blies. Specifically, this protocol presents detailed procedures for the
assemblies of iron oxide magnetic nanoparticles (MNP) and DNA
block copolymers composed of an oligonucleotide and polystyrene
(DNA-b-PS).
Preparation and Characterization of DNA Block Copolymer Assemblies… 209
2 Materials
All aqueous solutions are prepared using ultrapure water (deion-

ized Type I water purified by Barnstead NanoPure DIamond,
Model D11901, with resistivity of 18 MΩ at room temperature).
All reagents are used without further purification unless otherwise
indicated.
2.1 Synthesis 1. Hydroxyl-terminated PS, 5 g (Mn = 10.4 kg/mol, Polymer

of DNA-b-PS Source, Canada) (see Note 1).
2.1.1 Synthesis of 2. Dichloromethane (anhydrous, ≥99.8 %).
Phosphoramidite- 3. 2-Cyanoethyl N,N-diisopropylchlorophosphoramidite, 1 g
Terminated PS (Sigma-Aldrich) (see Note 2).
4. Diisopropylethylamine (purified by redistillation, 99.5 %,
Sure/Seal™, Sigma-Aldrich).
5. Argon gas (high purity).
6. Reaction apparatus: 100-mL three-neck round bottom flask,
Schlenk line, Schlenk line gas adapter, two rubber stoppers,
PTFE-encased magnetic stir bar, magnetic stirrer with heat
plate.
7. Deuterated chloroform for NMR: Chloroform-d
(99.8 atom % D).
31
8. Bruker DMX 360 MHz NMR (DMX360) for P NMR
spectroscopy.
2.1.2 Solid State 1. ABI 391 DNA/RNA synthesizer (Applied Biosystems).

Synthesis of 2. Beta-Cyanoethyl (CE) phosphoramidite bases (Glen Research,
Oligonucleotides Sterling, VA, USA) for a 10 μmol scale synthesis of oligonucle-
otide sequence: 5′-FAM A10 ATC CTT ATC AAT ATT-3′
(see Note 3). Each nucleoside phosphoramidite is dissolved in
anhydrous acetonitrile to a concentration of 0.1 M (see Note 4).
3. dT-CE phosphoramidite controlled pore glass (CPG) beads,
1,000 Å (Glen Research).
4. Synthesis support column (10 μmol, crimp style caps, with
Teflon filters, Applied Biosystems).
5. Activator solution: 0.45 M tetrazole in acetonitrile (Glen
Research).
6. Capping mix A: tetrahydrofuran (THF)/pyridine/acetic anhy-
dride (Glen Research).
7. Capping mix B: 15 % 1-methylimidizole in THF (Glen
Research).
8. Detritylation solution: 3 % trichloroacetic acid/dichlorometh-
ane (Glen Research).
9. Oxidizing solution: 0.1 M iodine in THF/pyridine/H2O

(Glen Research).
10. Dichloromethane (stabilized with amylene, peptide synthesis,
≥99.9 %, Acros Organics).
11. Acetonitrile (anhydrous, DNA synthesis).
12. Argon (Ultrahigh Purity).
2.1.3 Solid State 1. Activator solution: 0.45 M tetrazole in acetonitrile (Glen

Coupling of Research).
Phosphoramidite- 2. Detritylation solution: 3 % trichloroacetic acid/dichlorometh-
Terminated PS to ane (Glen Research).
Oligonucleotides
3. N,N-Dimethylformamide (HPLC grade).
4. Reaction apparatus: three-neck round bottom flask (100 mL),
Schlenk line, PTFE-encased magnetic stir bar, magnetic stirrer
with heat plate (Fisher Scientific).
5. Oxidizing solution: 0.1 M iodine in THF/pyridine/H2O
(Glen Research).
6. Acetonitrile (anhydrous, DNA synthesis).
7. Ammonium hydroxide solution (25 %) (see Note 5).
8. 20 mL Scintillation vial.
9. Water bath.
10. Filter paper (grade 1, Whatman).
2.1.4 Characterization 1. Fluorometer (Fluorolog3, Jobin Yvon Horiba).

of DNA-b-PS 2. UV–Vis Spectrometer (Model 8453, Agilent).
3. Micro quartz cuvettes: 2-clear window for UV–Visible spec-
troscopy measurements and 3-clear window for fluorescence
spectroscopy.
4. Zetasizer Nano Series (Malvern Instruments Ltd.).
5. Agarose (Ultrapure, Invitrogen).
6. 1× Tris-acetate-EDTA (TAE) casting and running buffer.
7. Nucleic acid sample loading buffer, 5× (Bio-Rad).
8. Horizontal nucleic acid electrophoresis system, basic power
supply (Bio-Rad).
2.2 Synthesis of 1. Iron(III) acetylacetonate.

MNP and the Self- 2. 1,2-Hexadecanediol.
Assembly of DNA-b-PS
3. Oleic acid.
and MNP
4. Oleylamine.
2.2.1 Synthesis of Oleic
5. Diphenyl ether (see Note 6).
Acid Stabilized MNP
6. Acetone.
7. Hexanes.
8. Chloroform.
9. Reaction apparatus: spatula, PTFE-encased magnetic stir bar,
magnetic stirrer with heat plate, two-neck 100 mL round bot-
tom flask with 14/20 necks, glass condenser with 14/20 neck,
glass 90° angle adapter with 14/20 neck, clamps, Schlenk line,
vacuum pump, nitrogen gas.
10. Thermocouple with a 14/20 joint adapter (stainless steel,
J-1/16-U-24, J-KEM Scientific).
11. Heating controller with connecting cords (Model 210/Timer,
J-KEM Scientific).
12. Heating mantle (soft shell, 100 mL, OS-100, J-KEM
Scientific).
13. Benchtop centrifuge (Allegra 64RCentrifuge, No: 367586,
Beckman Coulter) with 50 mL centrifuge tube rotor (F0685
Rotor, Beckman Coulter).
14. Centrifuge tubes (50 mL).
15. Transmission Electron Microscope (TEM) (JEM-1400,
JEOL).
16. TEM grids (CF200-Cu, Electron Microscopy Sciences).
17. TEM grid tweezers.
2.2.2 Preparation 1. Carboxyl-terminated PS, Mn = 11.2 kg/mol (Polymer Source).

of PS-Modified MNP 2. Acetone.
(PS-MNP) via Ligand
3. N,N-Dimethylformamide (DMF).
Exchange
4. 1.5 mL microcentrifuge tube.
5. Benchtop centrifuge (Allegra 64RCentrifuge, No: 367586,
Beckman Coulter) with 2 mL centrifuge tube rotor (F2400H
Rotor, Beckman Coulter).
2.2.3 Preparation 1. N,N-Dimethylformamide (DMF).

of MNP-Incorporated 2. Scintillation vial (20 mL).
DNA-b-PS Assemblies
3. PTFE-encased magnetic stir bar.
(MNP@PS@DNA)
4. Magnetic stirrer and heat plate.
5. Automatic micropipettes (Eppendorf Research, 1–10 μL).
6. Timer/Stopper.
7. Regenerated cellulose dialysis tubing (nominal MWCO 6,000–
8,000 Da, pore size 1.8 nm, FisherBrand).
8. Dialysis clips.
9. 800 mL beaker.
2.3 Characterization 1. TEM (FEI-Technai T12).

of MNP@PS@ DNA 2. TEM grids (CF200-Cu, Electron Microscopy Sciences).
3. TEM grid tweezers.
4. Zetasizer Nano Series (Malvern Instruments, Ltd.).
5. Plastic disposable cuvette, 4.5 mL (Brandtech).
2.4 DNA 1. Cy3-labeled complementary DNA (Sequence: 3′-Cy3 T10

Hybridization TAG GAA TAG TTA TAA-5′) (Integrated DNA Technologies,
Monitored by the USA): Dissolve the DNA in water at a concentration of
Förster Resonance 1 × 10−4 M (Extinction coefficient of DNA at 260 nm:
Energy Transfer (FRET) 268,456 M−1 cm−1) [14].
2. 1.5 mL microcentrifuge tube.
3. Vacufuge concentrator (Eppendorf).
4. 0.6 M phosphate buffer saline (PBS, 0.6 M NaCl, 20 mM
phosphate buffer, pH 7).
5. Micro quartz cuvettes: 3-clear window for fluorescent
spectroscopy.
6. Fluorometer (FluoroLog3, Jobin Yvon Horiba).
7. Temperature controller (TLC50, Quantum Northwest).
8. Cy5-labeled complementary DNA (Sequence: 5′-Cy5 A10
ATC CTT ATC AAT ATT-3′) (Integrated DNA Technologies,
USA).
3 Methods
3.1 Synthesis, The following procedures describe the synthesis of one specific
Purification, and example of DNA block copolymer, DNA-b-PS, with the DNA
Characterization sequence 5′-FAM A10 ATC CTT ATC AAT ATT-3′. The polymer
of DNA-b-PS is attached at the 5′ end of DNA. The DNA block copolymer is
prepared by solid-state synthesis by standard phosphoramidite
chemistry. Other types of DNA block copolymers can be synthe-
sized in a similar fashion using different hydroxyl-terminated poly-
mers (see Note 7).
3.1.1 Synthesis 1. Weigh out 4.16 g (0.4 mmol) of hydroxyl-terminated PS

of Phosphoramidite- (Mn = 10.4 kg/mol) and add it to a 3-neck round bottom flask
Terminated PS equipped with a PTFE-coated magnetic stir bar, a Schlenk line
adapter and two rubber stoppers.
2. Vacuum and then purge the reaction flask containing the poly-
mer with argon, and repeat the procedure three times.
3. Under a gentle flow of argon, add 15 mL of anhydrous dichlo-
romethane to the reaction flask with stirring and allow the
polymer to dissolve completely (see Note 8).
4. Add 0.378 mL (1.68 mmol) of chlorophosphoramidite

(see Note 9) and 0.290 mL (1.68 mmol) of anhydrous diiso-
propylethylamine to the reaction flask.
5. Allow the solution to stir for 3 h under a blanket of argon
(see Note 10).
6. Analyze the crude product by 31P NMR. The signal for phos-
phoramidite-terminated PS is at δ = 147 ppm (see Note 11).
3.1.2 Solid State 1. Prepare phosphoramidite solutions by adding 18.77 mL,

Synthesis of DNA 10 mL, 8 mL, and 1 mL of anhydrous acetonitrile to crimped
vials containing dA-CE (1.0 g), dT-CE (0.5 g), dC-CE (0.5 g),
and 6-fluorescein (100 μmol) phosphoramidites, respectively.
Swirl the vials gently to make sure that the reagents are com-
pletely dissolved.
2. Prepare a solid state synthesis column for a 10 μmol synthesis
by weighing out 0.33 mg of dT-CE CPG beads (1,000 Å).
Tightly secure the beads into the support column and make
sure there are no leaks.
3. Prepare the DNA synthesizer by loading the dT-CE CPG into
the support column and screwing in the CE-phosphoramidite
bottles to their appropriate dedicated spots on the
instrument.
4. Enter the DNA sequence (5′-A10 ATC CTT ATC AAT ATT-
3′), select 10 μmol synthesis method, choose the option to
keep the trityl end group “on,” and start the synthesis.
5. When the synthesis is complete, couple a fluorescein dye
(FAM) tag at the 5′ end of the oligonucleotide by selecting the
10 μmol synthesis method again, and entering DNA sequence
of 5′XA3′, where X denotes the position of the bottle contain-
ing the dye-phosphoramidite on the DNA synthesizer. Select
the option to keep the trityl end group “on.” Start the synthe-
sis (see Note 12).
3.1.3 Synthesis The synthetic scheme for DNA-b-PS is presented in Fig. 1

of DNA-b-PS (see Note 13).
1. Detritylate the DNA on CPG beads manually by pushing
detritylation solution through the supporting column containing
the beads using a syringe until no orange color is observed.
2. Dry the beads by blowing a stream of argon gas through the
column.
3. Pour the dried CPG beads containing the synthetic oligonu-
cleotides prepared in the previous section into the round bot-
tom flask equipped with a stir bar and Schlenk line adapter and
two rubber stoppers.
Fig. 1 Schematic description for the synthesis of DNA block copolymers
4. Add 3 mL of anhydrous dichloromethane to the CPG beads to

avoid static cling while the flask is purged with a gentle stream
of argon for 10 min.
5. Transfer the phosphoramidite-terminated PS to the flask con-
taining the DNA on CPG beads as fast as possible by cannula
to reduce the exposure to oxygen.
6. Add 15 mL of dimethylformamide to make a 1:1 DMF–
CH2Cl2 solution (see Note 14).
7. Add 2.7 mL of activator solution to the flask at 3:1 molar ratio
of activator: phosphoramidite-terminated PS.
8. Stir the reaction for a few hours or overnight under a blanket
of argon.
9. Stop the stirring of the reaction and allow the CPG to settle to
the bottom.
10. Decant the supernatant and wash the beads with five aliquots
of 10 mL chloroform to remove unreacted polymer.
11. Allow the beads to air dry.
12. Collect the beads and pack them back into the solid support
column once again. Oxidize the phosphoramidite to phos-
phate by passing the oxidation solution through the column
with a syringe drop wise for 2 min.
13. Rinse the beads with acetonitrile until no yellow color from the
iodine in the oxidation solution is observed on the beads.
14. Dry the beads and transfer them to a scintillation vial. To
deprotect and cleave the DNA from the CPG beads, add
10 mL of concentrated ammonium hydroxide to the dried
beads and leave the solution undisturbed for 6–12 h at 55 °C
(see Note 15).
15. Separate the supernatant from the CPG beads by gravity filtra-
tion (see Note 16). This fraction will contain mainly uncoupled
DNA strands and a small amount of DNA block copolymers.
Fig. 2 (a) An emission spectrum of FAM-labeled DNA-b-PS dispersed in water. (b) DLS data of simple micelles
of DNA-b-PS (PS@DNA) in water. (c) Gel electrophoresis result for (1) DNA and (2) PS@DNA
16. Wash the beads with two 5 mL aliquots of DMF, and collect
the DMF solution in a separate vial. This fraction will contain
mainly DNA block copolymers.
17. Collect the emission spectrum of DNA-b-PS in water with the
excitation wavelength at 490 nm to check the presence of the
FAM dye in the DNA block-copolymers (Fig. 2a).
18. Determine the concentration of FAM-modified DNA-b-PS in
DMF using the extinction coefficient of FAM in DMF at
518 nm (1.3 × 104 M−1 cm−1) using UV–Vis spectroscopy
(see Note 17).
19. Determine the hydrodynamic diameter of simple micelles of
DNA-b-PS in water (PS@DNA) by dynamic light scattering
(DLS). The average of three separate measurements was
15 ± 4 nm (Fig. 2b).
20. Prepare a 3 % agarose gel by dissolving 1.5 g of agarose with
50 mL of 1× TAE buffer with heating in a microwave oven for
2 min. Cast the gel by pouring the solution into the gel tray
and allow it to cool and solidify. Add 2 μL of loading buffer to
10 μL DNA samples (10 pmol). Add enough 1× TAE buffer to
cover the gel. Load the sample into the well and run the gel at
60 V for 1 h. Image the gel under UV lamp (Fig. 2c).
3.2 Self-Assembly of Oleic acid stabilized MNPs were synthesized by the thermal
DNA-b-PS and MNP decomposition method [15].
3.2.1 Synthesis of Oleic 1. Weigh out 0.71 g of iron(III) acetylacetonate, 2.58 g of
Acid Stabilized MNPs 1,2-hexadecanediol, and 1.69 g of oleic acid.
2. Add all three chemicals to the two-neck 100 mL round bot-
tom flask with 14/20 necks.
3. Measure out 2 mL of oleylamine and 20 mL of diphenyl ether.
4. Add these two chemicals to the two-neck 100 mL round bot-
tom flask with 14/20 necks.
5. Assemble the reaction flask and connect it to the Schlenk line

(see Note 18).
6. Add the magnetic stir bar to the round bottom flask.
7. First attach the glass condenser to the round bottom flask, and
then the 90° angle adapter to the glass condenser.
8. Attach the thermocouple with a 14/20 joint adapter to the
other neck of the round bottom flask (see Note 19).
9. Secure the round bottom flask within the heating mantle on
the magnetic stirrer by using metal clamps.
10. Connect both the thermocouple and the heating mantle to the
heating controller using the connector cords.
11. Make sure all the glass and adapter fittings are tight and secure.
12. Connect the secured reaction flask, assembled with the heating
apparatus, to the Schlenk line.
13. Turn on both vacuum pump and nitrogen gas.
14. To make sure the reaction proceeds under inert conditions
without water and oxygen, evacuate the round bottom flask
and refill it with nitrogen (see Note 20).
15. To do this, first, slowly open the vacuum valve of the Schlenk
line and apply vacuum to the flask.
16. After about 30 s of applying vacuum, close the vacuum valve of
the Schlenk line and then open the nitrogen gas valve of the
Schlenk line.
17. When opening the nitrogen gas valve of the Schlenk line, open
it slowly to make sure that no air enters the flask.
18. Repeat steps 15–17 two more times.
19. On the last cycle, keep the nitrogen gas valve open, and keep it
open during the duration of the reaction.
20. Next, turn on the heating controller and set the temperature
to 200 °C.
21. Heat the reaction flask to 200 °C and hold it at that tempera-
ture for 30 min.
22. After 30 min, increase the temperature to 265 °C.
23. Hold at 265 °C for another 30 min.
24. After 30 min, cool the reaction to room temperature by detach-
ing the heating mantle from the heating controller and removing
the round bottom flask from the heating mantle (see Note 21).
25. Once the temperature goes down to room temperature, close
the nitrogen gas valve and pour the reaction mixture into two
50 mL centrifuge tubes in equal amounts.
26. The reaction mixture should be dark-brown and cloudy.
27. Add 35 mL of acetone to each centrifuge tubes and centrifuge

at 8,000 rpm (7,012 × g ) for 10 min (see Note 22).
28. Decant the supernatant and keep the brown precipitate in the
centrifuge tubes.
29. The brown precipitate is the iron oxide nanoparticles.
30. Add 5 mL of hexane to each centrifuge tube to redisperse the
particles.
31. Repeat steps 28–30 two more times.
32. After the last centrifugation step, air-dry the particles in the
hood for 30 min and redisperse the particles in 5 mL of chlo-
roform for a total of 10 mL solution of particles.
33. Characterize the particles with TEM.
34. To prepare the TEM sample, place a single droplet of nanopar-
ticle solution to the grid using a glass pipette (see Note 23).
The size distribution of nanoparticles synthesized using this
method is typically 4.5 ± 0.4 nm.
3.2.2 Preparation 1. Mix 50 μL (in chloroform) of the 5 nm oleic acid stabilized

of PS-Modified MNP MNP solution (50 μg/mL) in a 1.5 mL microcentrifuge tube
(PS-MNP) with 100 μL of 10 mg/mL of carboxyl-terminated PS (in chlo-
roform). The polymer will be in large excess compared to the
number of oleic acid strands on the nanoparticle surface.
2. Let it stand at room temperature overnight.
3. Clean off excess PS by precipitating the nanoparticles with
500 μL of acetone.
4. Centrifuge the mixture at 7,000 rpm (4,492 × g ) for 15 min.
Pipette off the supernatant. Resuspend the precipitates of
PS-MNP in 100 μL of DMF. Add 500 μL of acetone to repre-
cipitate the nanoparticles.
5. Repeat the centrifugation, decantation and resuspension pro-
cedure three more times (see Note 24).
6. After the cleaning step, resuspend the final solution of PS-MNP
in 100 μL of DMF.
3.2.3 Preparation 1. Prepare a 300 μL of 10 μM solution of DNA-b-PS in DMF,

of MNP@PS@DNA (Fig. 3a) and mix it with 10 μL of 50 μg/mL PS-MNP (in DMF) pre-
pared in the previous section in a scintillation vial equipped
with a magnetic stir bar.
2. Add an additional 1 mL of DMF to the solution.
3. With stirring, slowly add 10 μL of water to the solution at 30-s
intervals for a period of 15 min (see Note 25).
4. Allow the solution to stir overnight.
5. Add an additional 1 mL of water to the solution, drop by drop.
Fig. 3 (a) Schematic description for the self-assembly of DNA-b-PS and MNP. (b) A TEM image of MNP@PS@
DNA (Scale bar = 100 nm). (c) DLS data of MNP@PS@DNA
6. Wet a piece of regenerated cellulose dialysis tubing, clip one

end of the tubing and pipette the solution into the other end
and clip that end with another clamp.
7. Dialyze the assembly solution against water overnight in an
800 mL beaker filled with ultrapure water.
8. Pipette the assembly solutions into 1.5 mL microcentrifuge
tubes, and centrifuge at 14,000 rpm (17,968 × g ) for 1 h.
9. Carefully use the pipette to remove the supernatant.
10. Add 100 μL of ultrapure water to the precipitate and vortex
the precipitate to disperse the assemblies in the solution.
3.3 Characterization Prepare TEM samples by dropping 7.5 μL of the MNP@PS@DNA

of MNP@PS@ DNA sample prepared above onto a TEM grid and wick away the excess
liquid using a Kimwipe. Repeat this procedure three times. The
3.3.1 Transmission
diameter of the MNP@PS@DNA assemblies based on the TEM
Electron Microscope (TEM)
measurement was 113 ± 15 nm (Fig. 3b).
3.3.2 Dynamic Light The hydrodynamic diameter of the assemblies is determined by

Scattering (DLS) light scattering using Malvern Zetasizer by measuring the MNP@
PS@DNA solution in water with a disposable cuvette. The hydrody-
namic diameter of the MNP@PS@DNA assemblies was determined
to be 161 ± 8 nm from three separate measurements (Fig. 3c).
3.4 DNA The FRET between FAM on the assemblies and Cy3 on the target
Hybridization DNA was used to monitor DNA hybridization of MNP@PS@
Properties of DNA DNA. Hybridization of the FAM- and Cy3-modified DNA results
Block Copolymer/ in energy transfer from FAM to Cy3 at an excitation wavelength
Nanoparticle that can excite the donor (FAM) but not the acceptor (Cy3). Thus,
Assemblies the increase in the Cy3 emission and the decrease in FAM emission
can be used to monitor DNA hybridization. The FRET efficiency
3.4.1 DNA Melting Study (EFRET) was calculated by the equation: EFRET = 1 − (FDA/FD), where
by Förster Resonance FD is the emission intensity of the donor in the absence of the
Energy Transfer (FRET) acceptor, and FDA is the emission intensity of the donor in the pres-
ence of the acceptor [16].
Fig. 4 Thermal denaturation monitored by FRET between MNP@PS@DNA and complementary Cy3-DNA in
0.3 M PBS. (a) Fluorescence spectra of MNP@PS@DNA mixed with Cy3-DNA in 0.3 M PBS collected with
excitation wavelength of 430 nm at varying temperatures. (b) DNA melting curve constructed from the PL
intensity at 520 nm. (c) The first derivative of the melting curve in (b)
1. Place 800 pmol of Cy3-labeled complementary DNA (Cy3-

DNA, Sequence 3′-Cy3 T10 TAG GAA TAG TTA TAA-5′) in
a 1.5 mL microcentrifuge tube and evaporate off the water
with a Vacufuge concentrator.
2. Add 400 μL of MNP@PS@DNA solution containing 200 pmol
of PS-DNA to the dried Cy3-labeled complementary DNA.
3. Add 400 μL of 0.6 M PBS solution to the solution prepared
above containing the Cy3-labeled complementary DNA and
MNP@PS@DNA to make a solution with a final NaCl concen-
tration of 0.3 M and a total volume of 800 μL.
4. Equilibrate the solution at room temp for 6 h.
5. Pipette the solution containing MNP@PS@DNA and Cy3-
DNA into the 3-clear window quartz cuvette with a cap to
prevent evaporation.
6. Set the temperature controller to 25 °C and equilibrate the
solution for 3 min.
7. Increase the temperature by 5 °C intervals and equilibrate for
3 min between each increase. Collect an emission spectrum of
the sample with an excitation wavelength of 430 nm at each
interval after equilibration (Fig. 4a).
8. Read the fluorescence intensity at 520 nm (emission maximum
of FAM) for each spectrum collected. Plot a graph of the emis-
sion intensity versus temperature to construct thermal dena-
turation curves (Fig. 4b) (see Note 26).
9. Take the first derivative of the emission intensity versus tem-
perature graph to determine the melting temperature of the
sample (Fig. 4c).
Fig. 5 (a) Schematic description of the “competition” experiment in water with MNP@PS@DNA, Cy3-labeled
target strands (Cy3-DNA) and Cy5-labled competition strands (Cy5-DNA). (b) Fluorescence spectra of the
competition experiment collected at 20 °C. Excitation wavelength of 430 nm (blue) was used to monitor the
binding of target Cy3-DNA to MNP@PS@DNA by probing the energy transfer for the FAM-Cy3 FRET pair.
EFRET(FAM-Cy3) was determined to be 36 %. The fluorescence spectrum taken with the excitation wavelength
of 525 nm (red ) confirms that the FRET efficiency between Cy5-DNA and Cy3-DNA is very low
(EFRET(Cy3-Cy5) < 1 %) and that the Cy5-DNA strands do not bind to Cy3-DNA in water
3.4.2 “Competition” Perform a preferential binding experiment to examine the enhanced

Binding Experiment binding capability of the MNP@PS@DNA. In this experiment,
Cy5-modified DNA with the same sequence as the DNA on MNP@
PS@DNA is added to the solution of MNP@PS@DNA and Cy3-
DNA as a competition strand (Fig. 5a). The degree of hybridiza-
tion of the Cy3-target DNA with the DNA block copolymer
assemblies and the competition DNA was monitored by the energy
transfer between FAM (donor) to Cy3 (acceptor) and Cy3 (donor)
to Cy5 (acceptor), respectively, at excitation wavelengths for the
donor dye.
1. Mix 800 pmol of Cy3-DNA and 200 pmol of Cy5-labeled
“competition” DNA (sequence: 5′Cy5 A10 ATC CTT ATC
AAT ATT-3′) in a 1.5 mL microcentrifuge tube and evaporate
off the water with a Vacufuge concentrator.
2. Add 400 μL of MNP@PS@DNA stock solution containing
200 pmol of PS-DNA to the dried Cy3-labeled complemen-
tary DNA and Cy5-labeled competition DNA.
3. Add 400 μL of ultrapure water to the solution prepared above
(i.e., the solution containing the Cy3-DNA, MNP@PS@DNA,
and Cy5-DNA).
4. Equilibrate the solution at room temp for 6 h.
5. To measure the degree of hybridization between the comple-
mentary Cy3-DNA and MNP@PS@DNA and the degree of
hybridization between the Cy3-DNA and competition Cy5-
DNA, take emission spectra at excitation wavelengths of
430 nm and 525 nm (Fig. 5b), respectively.
4 Notes
1. It is important that the hydroxyl-terminated polymer used

for the chlorophosphoramidite synthesis is completely dry.
To achieve this, the polymer can be dried under vacuum
(<100 mTorr) overnight to remove residual water molecules.
2. (a) 2-Cyanoethyl N,N-diisopropylchlorophosphoramidite is
stored in the freezer (−20 °C) as recommended by the manu-
facturer. (b) Integrity of the chlorophosphoramidite can be
checked using 31P NMR (chemical shift at 180 ppm). It is sta-
ble up to 6 months if stored properly.
3. “A10” denotes ten adenine bases in the sequence.
4. For the synthesis of other oligonucleotides sequences, the
appropriate amount of each nucleoside phosphoramidite and
anhydrous acetonitrile needed to make a 0.1 M solution can be
found on the Glen Research website [17].
5. Caution: Ammonium hydroxide used in the deprotection and
cleaving steps is highly basic and has a very pungent smell.
Take caution in handling it and always open it in a well-
ventilated hood.
6. Diphenyl ether is a solid at room temperature. Before starting
the synthesis, it is helpful to have the diphenyl ether sit in hot
water. It should take about 20 min to melt.
7. Important: The length of the entire synthesis of the DNA
block copolymer including preparation time can take up to
10–12 h. It is preferred that all the steps are carried out on the
same day due to the instability of the phosphoramidite-
terminated polymer. Therefore, we suggest to first set up the
DNA synthesis in the morning, and then set up the phosphora-
midite reaction of the hydroxyl-terminated polymer when
there is about 4–5 h left in the DNA synthesis.
8. During the 3 h reaction of the chlorophosphoramidite with
the hydroxyl-terminated polymer, make sure the flow of argon
is gentle enough to provide a blanket of the gas over the solu-
tion, but not too high that it can evaporate the dichlorometh-
ane solvent.
9. The chlorophosphoramidite purchased from the manufacturer
does not come in a Sure/Seal™ bottle. Additional steps are
needed to avoid water condensation and oxidation. Before use,
place the cold chlorophosphoramidite in the desiccator to
allow it to warm up to room temperature. Degas the reagent
by measuring out a larger amount (~5 times excess) of the
chlorophosphoramidite than the procedure calls for, and inject
it into a vacuum/purge vial capped with a rubber stopper.
Subsequently, bubble the solution with argon for 5 min.
10. After the 3-h reaction time, the solution becomes light yellow
in color.
11. (a) Given that phosphoramidite-terminated polymer is highly
prone to oxidation to become phosphate, no purification was
carried out to avoid the possibility of exposure to oxygen. (b)
31
P NMR can be calibrated by adding an internal standard
trioctylphosphine (TOP), with a 31P NMR peak at −30 ppm
[18]. (c) The oxidized product (i.e., polymer phosphate) has a
peak at ~0–18 ppm. The unreacted chlorophosphoramidite
has a peak at 180 ppm.
12. The synthesis of a dye-modified DNA can be performed in one
step instead of the two-steps described in the procedure by
programming in the sequence (5′-XA10 ATC CTT ATC AAT
ATT-3′), where X denotes the position of the FAM
phosphoramidite.
13. Solid state coupling of phosphoramidite-terminated PS to the
oligonucleotides on CPG beads is not performed using the
DNA synthesizer due to the poor solubility of the PS in aceto-
nitrile, the typical solvent used in DNA synthesis. Other aceto-
nitrile soluble polymers can be coupled using the DNA
synthesizer [19].
14. A solvent mixture of DMF–CH2Cl2 (1:1) was used to resolve
the solubility difference between DNA and PS.
15. During this step, make sure to wrap the vial with foil to protect
the dye from photobleaching. Caution: During this step,
ammonia is produced as a by-product from the degradation of
ammonium hydroxide. Therefore, be cautious when opening
the vial after the reaction. Cool down the solution to room
temperature before opening the vial, and open the vial in a
well-ventilated hood.
16. Filtration will go faster by first pipetting the supernatant into
the filter paper and then pipette the beads into the filter paper.
17. The extinction coefficient of FAM in DMF was determined by
creating a calibration curve by preparing a series of dilutions
from a stock solution of FAM in DMF. The extinction coeffi-
cient provided is the average of three measurements.
18. Make sure that the glass and adapter fittings are snug and the
round bottom flask necks are clean of all chemicals. If there is
some residue, wipe the necks with Kimwipes.
19. Make sure the tip of the thermocouple is touching the reaction
mixture but not the spinning magnetic stir bar.
20. The magnetic stir bar should be spinning while using Schlenk
line techniques.
21. Be cautious after the last heating step. The heating mantle is
extremely hot. Wear heat resistant gloves when handling the
mantle.
22. When purifying the particles, make sure to notice the solution
turning cloudy as the acetone is added to the hexane solution.
This cloudiness indicates flocculation of particles.
23. When preparing the TEM grid, use a diluted solution (10×
dilution of the as synthesized particles) of nanoparticles. If the
solution is too concentrated, the particles will completely cover
the grid, which makes it difficult to distinguish individual
particles.
24. One can check if PS is present in the solution by pouring the
supernatant into water. If PS is present, the solution will turn
white and turbid.
25. In the preparation of MNP@PS@DNA, the solution should
look a bit turbid after the slow water addition. However, there
should be no visible precipitates.
26. The changes in the emission intensity of FAM are used to
monitor the hybridization process instead of Cy3 due to the
temperature-dependent fluorescence intensity of Cy3 dyes.
Acknowledgments
This work was supported by the NSF Career Award (DMR-

087646) and NSF-IGERT Fellowship (DGE-0221664).
References
1. Bardhan R, Lal S, Joshi A, Halas NJ (2011) 9. Chen X-J, Sanchez-Gaytan BL, Hayik SEN,
Theranostic nanoshells: from probe design to Fryd M, Wayland BB, Park S-J (2010) Self-
imaging and treatment of cancer. Acc Chem assembled hybrid structures of DNA block-
Res 44:936–946 copolymers and nanoparticles with enhanced
2. Della Rocca J, Liu D, Lin W (2011) Nanoscale DNA binding properties. Small 6:2256–2260
metal-organic frameworks for biomedical 10. Li Z, Zhang Y, Fullhart P, Mirkin CA (2004)
imaging and drug delivery. Acc Chem Res Reversible and chemically programmable
44:957–968 micelle assembly with DNA block-copolymer
3. Xie J, Liu G, Eden HS, Ai H, Chen X (2011) amphiphiles. Nano Lett 4:1055–1058
Surface-engineered magnetic nanoparticle 11. Alemdaroglu FE, Ding K, Berger R, Herrmann
platforms for cancer imaging and therapy. Acc A (2006) DNA-templated synthesis in three
Chem Res 44:883–892 dimensions: introducing a micellar scaffold
4. Erathodiyil N, Ying JY (2011) Functionalization for organic reactions. Angew Chem Int Ed
of inorganic nanoparticles for bioimaging 45:4206–4210
applications. Acc Chem Res 44:925–935 12. Jeong JH, Park TG (2001) Novel Polymer-
5. Yoo D, Lee J-H, Shin T-H, Cheon J (2011) DNA hybrid polymeric micelles composed of
Theranostic magnetic nanoparticles. Acc hydrophobic Poly(D, L-lactic-co-glycolic
Chem Res 44:863–874 Acid) and hydrophilic oligonucleotides.
6. Kievit FM, Zhang M (2011) Surface engineer- Bioconjug Chem 12:917–923
ing of iron oxide nanoparticles for targeted 13. Chien M-P, Rush AM, Thompson MP,
cancer therapy. Acc Chem Res 44:853–862 Gianneschi NC (2010) Programmable shape-
7. Chen X-J, Sanchez-Gaytan BL, Qian Z, Park shifting micelles. Angew Chem Int Ed 49:
S-J (2012) Noble metal nanoparticles in DNA 5076–5080
detection and delivery. Wiley Interdiscip Rev 14. Northwestern University. oligonucleotide
Nanomed Nanobiotechnol 4:273–290 properties calculator used to calculate extinc-
8. Rosi NL, Mirkin CA (2005) Nanostructures in tion coefficient of DNA. http://www.basic.
biodiagnostics. Chem Rev 105:1547–1562 northwestern.edu/biotools/OligoCalc.html
15. Sun S, Zeng H (2002) Size-controlled synthe- 18. Kuno M, Lee JK, Dabbousi BO, Mikulec FV,
sis of magnetite nanoparticles. J Am Chem Soc Bawendi MG (1997) The band edge lumines-
124:8204–8205 cence of surface modified CdSe nanocrystal-
16. Clegg RM, David M.J.L.a.J.E.D. (1992) lites: probing the luminescing state. J Chem
Fluorescence resonance energy transfer and Phys 106:9869–9882
nucleic acids. Methods Enzymol, vol. 211, 19. Zimmermann J, Kwak M, Musser AJ,
Academic, p 353–388 Herrmann A (2011) Amphiphilic DNA block
17. The information is available at the Glen copolymers: nucleic acid-polymer hybrid
Research website. Glen Research. http:// materials for diagnostics and biomedicine.
www.glenresearch.com/ Methods Mol Biol 751:239–266
Chapter 17
Polyaspartic Acid Coated Iron Oxide Nanoprobes

for PET/MRI Imaging
Taku Cowger and Jin Xie
Abstract
Iron oxide nanoparticles, due to their exceptional magnetic property, biocompatibility, and biodegradability,
have long been studied as contrast agents for magnetic resonance imaging (Xie et al., Curr Med Chem
16(10):1278–1294, 2009; Xie et al., Adv Drug deliv Rev 62(11):1064–1079, 2010). While previous
applications mostly target reticuloendothelial system (RES) organs such as liver and lymph nodes, recent
efforts have been made to impart targeting peptides or antibodies onto particle surface to enable site-
specific targeting after systemic administration (Xie et al., Adv Drug Deliv Rev 62(11):1064–1079, 2010;
Cai and Chen, Small 3(11):1840–1854, 2007; Corot et al., Adv Drug Deliv Rev 58 (14):1471–1504,
2006; Xie et al., Acc Chem Res 44(10):883–892). Moreover, other imaging functionalities can be loaded
onto nanoparticles to achieve multimodality imaging probes (Cai and Chen, Small 3(11):1840–1854,
2007; Lee et al., J Nucl Med Soc Nucl Med 49(8):1371–1379, 2008). In this protocol, we describe the
procedure of constructing an iron oxide nanoparticle (IONP)-based probe with high affinity towards inte-
grin αvβ3 for positron emission tomography (PET) and magnetic resonance imaging (MRI) dual modality
imaging. The related characterizations and validation experiments, including particle concentration deter-
mination, Prussian blue staining, animal model preparation, and in vivo PET/MRI imaging will also be
discussed.
Key words IRON oxide nanoparticles, Integrin αvβ3, Multimodal imaging, Positron emission tomog-
raphy, Magnetic resonance imaging, Tumor angiogenesis
1 Introduction
Nanomaterials, especially nanoparticles, have been extensively

studied in the recent decades for their potentials in improving the
current clinic diagnostic and therapeutic techniques [2, 3]. These
materials usually have unique physical or physiochemical proper-
ties, and when conjugated with targeting molecules, become
“smart” gadgets that can pinpoint and/or kill diseased tissues in a
living subject [7]. One of the most studied nanomaterials in the
field has been IONPs [4, 5]. With strong magnetic property, bio-
compatibility, and biodegradability, IONPs are candidate nano-
platforms to construct functional nanogadgets for both imaging
225
226 Taku Cowger and Jin Xie
and therapy purposes [2, 3]. One of the primary applications of

IONPs is to work as contrast agents for MRI. Several IONP for-
mulations have been approved by the Federal Drug Administration
(FDA) and used in the clinics to improve imaging quality [1].
While MRI provides high resolution and good anatomical
details, its signal-to-background ratio, even with contrast agents,
is suboptimal. In fact, none of the current imaging techniques are
perfect and there are always trade-offs, for instance, between reso-
lution and sensitivity. To address this issue, efforts have been made
to develop multimodality imaging systems and imaging probes
that allow simultaneous interrogation with two or even more
imaging methods [6, 8–10]. Such an approach can greatly improve
the diagnosis quality and lower misdiagnosis rate. For instance,
positron emission tomography/computed tomography (PET/
CT) systems have been developed and become an important diag-
nostic tool. PET/MRI is another promising technique under
intensive studies and is expected to be implemented nationwide
soon. Engineering IONPs and loading onto them radioisotopes is
a common way to achieve PET/MRI dual functional imaging
probes.
In the current protocol, we describe the detailed procedure for
preparing tumor targeting IONPs with both PET and MRI imag-
ing functionalities. IONPs are made from a high temperature ther-
mal decomposition method. They participate in a ligand exchange
with polyaspartic acid (PASP), a biocompatible and biodegradable
polymer with high affinity toward IONP surface. Subsequently, a
cyclic peptide, c(RGDyK), and a macrocyclic chelating agent,
1,4,7,10-tetreaazacyclododecane-N, N′, N″, N′″-tetraacetic acid
(DOTA), are both covalently conjugated onto the PASP coating.
Here c(RGDyK) works as targeting a motif for its high affinity
toward integrin αvβ3, a cell adhesion mediating receptor that is
highly expressed on many types of tumor cancer cell surface, and
more importantly, on tumor vasculature [10, 11]. This makes inte-
grin αvβ3 a tumor biomarker. Imaging probes and drug delivery
vehicles using RGD or its derivatives as tumor targeting motifs
have been widely reported [12–15]. Due to the fact that multiple
RGD can be loaded onto a single particle surface, RGD-
nanoparticle conjugates usually show even higher affinity toward
integrin αvβ3, the so-called multivalency effect [15]. DOTA is
introduced onto IONP surface as a metal chelator. DOTA can
form serum-stable complexes with many types of transition metals
and is widely used in PET or single-photon emission computed
tomography (SPECT) imaging to assist radioisotope labeling [16].
In the current protocol, we load 64Cu, a radioisotope commonly
used for PET imaging, onto IONPs.
Polyaspartic Acid Coated Iron Oxide Nanoprobes for PET/MRI Imaging 227
2 Materials
2.1 Reagents 1. Iron chloride.

2. Sodium oleate.
3. Ethanol.
4. Hexane.
5. 1-octadecene.
6. DOTA.
7. 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide.
8. N-hydroxysulfonosuccinimide (SNHS, Sigma).
9. Polyaspartic acid (PASP) (SPECTRUM).
10. O-[ N -(3-Maleimidopropionyl)aminoethyl]-O′-[3-(N -
succinimidyloxy)-3-oxopropyl]heptacosaethylene glycol (NHS-
PEG-MAL, Sigma).
11. 4-Maleimidobutyric acid N-succinimidyl ester (Sigma).
12. PBS buffer.
13. RGD peptide (c(RGDyK); Peptides International).
14. N-Succinimidyl S-acetylthioacetate (SATA, Pierce Biotechnology
Inc.).
15. DMSO.
16. Acetonitrile.
17. Trifluoroacetic acid.
18. Hydroxylamine hydrochloride.
19. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP.HCl;
Pierce Biotechnology Inc.).
20. Sodium hydroxide.
21. Hydrochloric acid.
22. U87MG human glioblastoma cells.
23. MCF-7 human breast cancer cells.
24. Low-glucose DMEM (GIBCO).
25. Minimum essential medium (GIBCO).
26. Fetal bovine serum (FBS; GIBCO).
27. Cell-binding buffer (20 mM Tris, 150 mM NaCl, 2 mM
CaCl2, 1 mM MnCl2, 1 mM MgCl2, 0.1 % (wt/vol) bovine
serum albumin; pH 7.4).
125
28. I-Echistatin (GE Healthcare).
29. Isoflurane (RxElite Inc.).
30. U87MG tumor-bearing athymic nude mice (Harlan).
2.2 Equipment 1. HPLC (Dionex).

2. Analytical HPLC column (Vydac).
3. Semi-preparative HPLC column (Vydac).
4. Freeze-dry system (Labconco Corp.).
5. PD-10 column (GE Healthcare, GE Healthcare).
6. Centrifuge (Allegra X-15R, Beckman-Coulter).
7. Refrigerated microcentrifuge (Eppendorf).
8. Rotary vaporizer (Buchi, RII).
9. 96-well plate (Thermolab Systems).
10. Vacuum manifold (Millipore).
11. Vacuum pump (VWR).
12. Dry bath incubator (Fisher Scientific).
13. Polystyrene culture test tube (Fisher Scientific).
14. γ-counter (Packard).
15. GraphPad Prism (GraphPad Software Inc.).
16. 4-well chamber slides (Thermo-Fisher).
17. ICP-MS (VG Elemental).
18. Rodent anesthesia system (Summit Anesthesia Solutions).
19. 7 T small animal MRI system (Varian).
20. microPET R4 rodent scanner (Siemens Medical Solutions).
3 Methods
3.1 Preparation of 1. 10.8 g of iron chloride (FeCl3·6H2O, 40 mmol) and 36.53 g

Iron-Oleate Precursor of sodium oleate (120 mmol) are dissolved in a mixture solvent
composed of 80 ml ethanol, 60 ml distilled water, and 140 ml
hexane.
2. The resulting solution is heated up to 80 °C and kept at that
temperature for 4 h.
3. The organic (top) layer is washed three times with 30 ml dis-
tilled water and collected by a separatory funnel.
4. Use a rotary evaporator to remove hexane and ethanol. The
final product is in a waxy solid form.
3.2 Synthesis 1. 9.0 g iron-oleate and 1.58 ml oleic acid is dissolved in 35 ml

of IONPs 1-octadecene with magnetic stirring (see Note 1).
2. The solution is heated stepwise at 150 °C for 45 min, 200 °C
for 45 min, and finally to reflux at ~320 °C for 45 min
(see Note 2).
3. Cool down the mixture to room temperature. Add 30 ml

ethanol to precipitate IONPs. Collect IONPs by centrifuge at
8,000 rpm (7,700 × g ) (see Note 3).
4. Decant the supernatant and redissolve the nanoparticles in
15 ml hexane with sonication. Add 30 ml ethanol to precipi-
tate the nanocrystals. Repeat the washing step for three times
(see Note 4).
3.3 Preparation of 1. 20 mg IONPs are dissolved in 5 ml hexane and added onto

PASP Coated IONPs 5 ml 25 mg/ml PASP solution in water to form a two phase
mixture.
2. Sonicate the mixture for 5 min and then centrifuge. Wash the
aqueous phase three times with hexane.
3. Collect the aqueous phase and run it through a PD-10 column
for purification and solvent exchange. Briefly, load the solution
onto a column and wait until all the solution is in. Add 3.5 ml
borate buffer (50 mM, pH 8.5) and collect only the deepest-
colored fractions at the bottom. The product is PASP-coated
IONPs (PASP-IONPs).
3.4 Preparation 1. Mix c(RGDyK) (5 mmol) in 1 ml of borate buffer (pH 8.5)

of Thiolated RGD with 100 ml DMSO solution containing 6 mmol of SATA.
Incubate at room temperature with gentle shaking.
2. Monitor the reaction by analytical RP-HPLC (reverse phase-
high performance liquid chromatography). The mobile phase
is changed from 95 % solvent A (0.1 % TFA in water) and 5 %
solvent B (0.1 % TFA in acetonitrile) (0–2 min) to 35 % solvent
A and 65 % solvent B at 32 min. The flow rate is 1 ml/min and
the UV absorbance is monitored at 218 nm.
3. When the reaction comes to completion, quench with 100 ml
of 2 % TFA in water.
4. Freeze-dry the crude mixture. Redissolve the product in 1 ml
water.
5. Add 100 ml of 0.5 M hydroxylamine solution. Adjust the pH
to 6.0 with 0.5 M sodium hydroxide (see Note 5).
6. Use RP-HPLC to monitor the reaction. When the reaction
comes to completion, purify the thiolated c(RGDyK)
(RGD-SH) by semi-preparative RP-HPLC. The gradient is the
same as described above. The flow rate is changed to 5 ml/
min. Collect the fractions containing pure RGD-SH and
freeze-dry (Fig. 1).
7. Store RGD-SH under acidic condition (pH 3–4) in water to
prevent disulfide formation (see Note 6).
Fig. 1 Converting c(RGDyK) to RGD-SH. Modified with permission from ref. 16
3.5 Conjugation of 1. DOTA is activated by 1-ethyl-3-[3-(dimethylamino)propyl]

DOTA and RGD onto carbodiimide (EDC) and N-hydroxysulfonosuccinimide
PASP-IONPs (SNHS) at pH 5.5 for 30 min with a DOTA:EDC:SNHS
molar ratio of 10:5:4.
2. 0.8 μmol activated DOTA and 1.2 μl NHS-PEG-MAL
(4.1 mg) are added to a 600 μl PASP-IONP solution in borate
buffer (2.5 mg Fe/ml, pH = 8.5). Incubate the mixture at 4 °C
for 1 h.
3. Add 1.0 mg RGD-SH (1.5 μmol) to the solution. Adjust
the pH to 7.0 and incubate the mixture overnight at 4 °C
(see Note 7).
4. Run the raw product through a PD-10 column to remove
unreacted species. Briefly, load the solution onto a column and
wait until all the solution is in. Add PBS buffer (pH 7.4) as
eluent and collect only the deepest-colored fractions at the
bottom (see Note 8).
5. The product, RGD and DOTA modified IONPs (RD-IONPs),
is stored at 4 °C and is stable for up to 4 weeks (Fig. 2).
Fig. 2 Schematic diagram of the 64Cu loaded RD-IONPs for PET and MRI imaging. Modified with permission
from ref. 7
3.6 Determination of Take 20 μl of the particle stock solution and decompose the particles
IONP Concentration in 10 % HNO3 with heating. Add water to adjust the final volume
to 10 ml. The iron content is to be analyzed by ICP-MS. The
results will be converted to obtain the iron concentration in the
original solution (see Note 9).
3.7 Cell Culture U87MG cells are cultured in DMEM (low glucose) supplemented
with 10 % (vol/vol) FBS at 37 °C. MCF-7 cells are cultured in
MEM supplemented with 10 % (vol/vol) FBS at 37 °C. The cells
are used when they reach 70–85 % confluency.
3.8 Binding 1. Prepare 125I-echistatin solution in cell-binding buffer (20 mM

Affinity Assay Tris, 150 mM NaCl, 2 mM CaCl2, 1 mM MnCl2, 1 mM
MgCl2, 0.1 % (wt/vol) bovine serum albumin; pH 7.4;
0.5 μCi/ml). It is estimated that about 50 μl solution is needed
for each well (see Notes 10 and 11).
2. Prepare three stock solutions of different concentrations of
RD-IONPs in cell-binding buffer. Parallel analyses with free
RGD peptide can be carried out as a comparison. Triplicate
samples are recommended.
3. Add 0.02–0.03 μCi of 125I-echistatin and appropriate amount
of the RD-IONP stock solution to each well of a 96-well plate.
The final concentrations should range from 1.0 μg Fe/ml to
1.0 mg Fe/ml.
4. Prepare a U87MG stock in cell-binding buffer at two million

cells/ml concentration. Add 1 × 105 U87MG cells (50 μl) to
each well and adjust the total volume to 200 μl per well with
cell-binding buffer. Incubate for 2 h at room temperature.
5. Use a vacuum manifold to remove the incubation buffer from
the 96-well plate and wash three times with the cell-binding
buffer (100 μl per well).
6. Heat-dry the 96-well plate in a dry bath incubator. It usually
takes 15 min to dry the membrane.
7. Collect the membranes from each well and put them into poly-
styrene culture test tubes.
8. Measure the radioactivity on each membrane with a γ-counter.
9. Fit the data by nonlinear regression and calculate the best-fit IC50
(inhibitory concentration of 50 %) values. RD-IONPs should
have much lower IC50 values (higher binding affinity) than the
unconjugated RGD peptide due to the multivalent efficiency.
3.9 Cell Staining 1. Seed U87MG (integrin αvβ3-positive) or MCF-7 (integrin

αvβ3-negative) cells (1 × 105 cells per well) onto 4-well cell cul-
ture chambers and incubate overnight.
2. Aspirate the cell culture medium from the dish. Wash the cells
twice with PBS.
3. Fix the cells with ice-cold 95 % EtOH for 15 min.
4. Wash the cells 2–3 times with 1 ml cell-binding buffer (3–5 min
each time).
5. Add PASP-IONP or RD-IONP solutions into each well.
Adjust the final particle concentration to 0.02 mg Fe/ml.
Incubate at 37 °C for 1 h with gentle shaking (see Note 12).
6. Remove the particles and wash the cells three times with PBS
buffer.
7. Incubate the cells with Prussian blue staining solution, which
is a 1:1 mixture of 20 % hydrochloric acid and 10 % potassium
ferrocyanide solution, for 40 min at room temperature. Wash
cells with PBS twice.
8. Incubate cells with fast red nuclear staining solution for 10 min.
9. Consecutive dehydrations with 70, 90, and 100 % EtOH.
10. Remove the chamber and mount the slide for microscopy.
3.10 Preparation The animal models will be prepared by injecting 5 × 106 U87MG
of U87MG cells subcutaneously into the right flank of athymic nude mice.
Xenograft Models Imaging studies will be performed when the tumor size reaches
~500 mm3, which usually takes 3–4 weeks. The tumor size will be
calculated as a × b2/2, where a represents the longer dimension
and b represents the shorter dimension of the tumor.
3.11 Loading 64Cu 1. Before in vivo PET imaging, incubate RD-IONPs with 64Cu
onto RD-IONPs (185 GBq/g Fe) in 0.1 N sodium acetate (pH 6.5) buffer for
45 min at 40 °C (see Note 13).
2. Run the mixture through a PD-10 column with PBS buffer
(pH 7.4) as the mobile phase. Collect the fractions with the
deepest color.
3.12 In Vivo PET 1. Prepare U87MG xenograft models as above mentioned. Wait
Studies until the tumor size reaches ~500 mm3.
64
2. Label IONPs with Cu by following the above mentioned
procedure.
3. Anesthetize the animals using the rodent anesthesia system
with isoflurane (2 % (vol/vol) isoflurane in 0.2 l/min O2 flow).
4. Inject 100–200 μl 64Cu labeled RD-IONPs intravenously into
the U87MG tumor-bearing mice at a dose of 10 mg Fe/kg.
For the control group, free c(RGDyK) (10 mg/kg) is co-
injected with the particle probes.
5. Perform PET scans on a microPET R4 rodent scanner at selec-
tive time points post injection.
6. For each scan, 3-dimensional ROIs are drawn over the tumor
and organs on decay-corrected whole-body coronal images.
The average radioactivity is obtained from the mean pixel values
within the ROI volume, and is converted to counts per millili-
ter per minute by using a predetermined conversion factor.
7. Given a tissue density of 1 g/ml, the counts per milliliter per
minute are converted to counts per gram per minute, and the
values are divided by the injected dose to obtain the imaging-
ROI-derived percentage injected dose per gram (% ID/g).
8. If needed, harvest the tumor and major organs. Take PET
images again with the organs. Subsequent ex vivo tissue stain-
ing can be performed when after the reactivity is decayed.
3.13 In Vivo 1. Prepare U87MG xenograft models as described above. Wait

MRI Studies until the tumor size reaches ~500 mm3.
2. Anesthetize the animals using the rodent anesthesia system
with isoflurane (2 % (vol/vol) isoflurane in 0.2 l/min O2 flow).
3. RD-IONPs are intravenously injected through the tail vein
(10 mg Fe/kg per mouse). For the control group, c(RGDyK)
(10 mg/kg) is co-injected with the particles.
4. T2-weighted fast spin-echo images are acquired on a 7.0 T small
animal MRI system, before and at selective time points post
injection. The following are suggested parameters: TE 40 ms,
TR 3,000 ms, thickness 1 mm, FOV 6 × 6, NEX 1.0, Echo 1/1
(see Note 14).
4 Notes
1. Iron-oleate is a very viscous compound and can prove difficult

to handle. It is best to keep the spatula slightly moist with hex-
ane to help transfer the metal oleate into the reaction flask.
2. Any smoke observed during the refluxing period is due to the
lack of a closed system and the solvent is being removed. In
that case, cease the heating and repair any leak in the system.
3. IONPs can be split into multiple tubes to ensure good
purification.
4. Small portion of aggregations may be observed. To remove
these, centrifuge at low speed (2,500 rpm (460 × g )) and dis-
pose the aggregates at the bottom.
5. If the pH is too low, it may result in incomplete de-protection
of the thiol. If the pH is too high, disulfide bond may form
when the thiol group is released.
6. RGD-SH can be stored for up to a few months under an acidic
condition.
7. It is highly recommended to check the purity of the RGD-SH
by analytical HPLC before conjugation. If significant amount
of disulfide already formed, some TCEP · HCl (1 mg or less)
can be added.
8. NAP-5 columns can be used when dealing with smaller amount
of samples.
9. Remember that 2 % HNO3 is commonly used in ICP analysis
as the eluent. Estimate the iron content and make sure it lies
within the range of the standards.
10. It is imperative to obtain appropriate training and abide by all
regulatory rules when handling radioactivity. The radioactive
waste needs to be collected and disposed under the guidance
of the institutional radiation safety office.
11. The presence of certain metal ions (e.g., Mn2+ and Mg2+) is
essential for integrin αvβ3 binding. Binding buffer without
these ions will result in much lower readings.
12. Aggregation may form during particle aging. The aggregates
may stick onto the slides and interfere with the staining. Use
fresh samples when possible. Sonication prior to the staining
may also help.
13. It is imperative to obtain appropriate training and abide by all
regulatory rules when handling radioactivity. The radioactive
waste needs to be collected and disposed under the guidance
of the institutional radiation safety office.
14. A monitoring system should be used to monitor physiological

signals of mice such as their heart rate, respiration, and body
temperature. Respiration rate is usually obtained from a small
pneumatic pillow sensor placed next to the animal’s abdomen.
Heart rate is monitored using ECG by placing three leads with
subdermal needle electrodes. Temperature is monitored with a
rectal thermometer or a fiber optic sensor.
Acknowledgments
This work was supported by NIH grant R00 4R00CA153772.
References
1. Xie J, Huang J, Li X et al (2009) Iron oxide 9. Xie J, Chen K, Huang J et al (2010) PET/
nanoparticle platform for biomedical applica- NIRF/MRI triple functional iron oxide
tions. Curr Med Chem 16(10):1278–1294 nanoparticles. Biomaterials 31(11):3016–3022
2. Xie J, Lee S, Chen X (2010) Nanoparticle- 10. Cai W, Chen X (2008) Multimodality molecu-
based theranostic agents. Adv Drug Deliv Rev lar imaging of tumor angiogenesis. J Nucl Med
62(11):1064–1079 49(Suppl 2):113S–128S
3. Cai W, Chen X (2007) Nanoplatforms for tar- 11. Cai W, Niu G, Chen X (2008) Imaging of inte-
geted molecular imaging in living subjects. grins as biomarkers for tumor angiogenesis.
Small 3(11):1840–1854 Curr Pharm Des 14(28):2943–2973
4. Corot C, Robert P, Idee JM et al (2006) 12. Chen X, Park R, Shahinian AH et al (2004)
Recent advances in iron oxide nanocrystal 18F-labeled RGD peptide: initial evaluation
technology for medical imaging. Adv Drug for imaging brain tumor angiogenesis. Nucl
Deliv Rev 58(14):1471–1504 Med Biol 31(2):179–189
5. Xie J, Liu G, Eden HS et al (2011) Surface- 13. Li ZB, Cai W, Cao Q et al (2007) (64)
engineered magnetic nanoparticle platforms Cu-labeled tetrameric and octameric RGD
for cancer imaging and therapy. Acco Chem peptides for small-animal PET of tumor
Res 44(10):883–892 alpha(v)beta(3) integrin expression. J Nucl
Med 48(7):1162–1171
6. Lee HY, Li Z, Chen K et al (2008) PET/MRI
14. Liu Z, Li ZB, Cao Q et al (2009) Small-animal
dual-modality tumor imaging using arginine-
PET of tumors with (64)Cu-labeled RGD-
glycine-aspartic (RGD)-conjugated radiola-
bombesin heterodimer. J Nucl Med
beled iron oxide nanoparticles. J Nucl Med
50(7):1168–1177
49(8):1371–1379
15. Xie J, Chen K, Lee HY et al (2008) Ultrasmall
7. Lee S, Xie J, Chen X (2010) Peptides and c(RGDyK)-coated Fe3O4 nanoparticles and
peptide hormones for molecular imaging their specific targeting to integrin alpha(v)
and disease diagnosis. Chem Rev 110(5): beta3-rich tumor cells. J Am Chem Soc
3087–3111 130(24):7542–7543
8. Chen K, Xie J, Xu H et al (2009) Triblock 16. Cai W, Chen X (2008) Preparation of peptide-
copolymer coated iron oxide nanoparticle conjugated quantum dots for tumor
conjugate for tumor integrin targeting. vasculature-targeted imaging. Nat Protoc
Biomaterials 30(36):6912–6919 3(1):89–96
Chapter 18
Ligand Synthesis and Passivation for Silver and Large

Gold Nanoparticles for Single-Particle-Based Sensing
and Spectroscopy
Daniel Montiel, Emma V. Yates, Li Sun, Marissa M. Sampias,
John Malona, Erik J. Sorensen, and Haw Yang
Abstract
Silver and large gold nanoparticles are more efficient scatterers than smaller particles, which can be
advantageous for a variety of single-particle-based sensing and spectroscopic applications. The increased
susceptibility to surface oxidation and the larger surface area of these particles, however, present challenges
to colloid stability and controllable bio-conjugation strategies. In this chapter, ligand syntheses and parti-
cle passivation procedures for yielding stable and bio-conjugatable colloids of silver and large gold nanopar-
ticles are described.
Key words Gold nanoparticles, Silver nanoparticles, Bio-conjugation, Nanoparticle passivation
1 Introduction
Materials on the nanometer scale can exhibit optical properties

different than the corresponding bulk materials. For noble metal
nanoparticles, the source of the size-dependent optical response is
the collective oscillation of conduction-band electrons in response
to an applied electromagnetic field, i.e., the particle’s plasmon
resonance [1–3]. In the near-field region (distance to particle
surface ≪ wavelength), nanoparticles can enhance the local field
leading to emission enhancement for emitters present in the field.
The plasmon resonance and near-field distributions can be tuned
using differently shaped nanoparticles or hierarchical structures
composed of multiple particles.
Noble metal nanoparticles, with plasmon resonances in the vis-
ible spectrum, have attracted a lot of attention in the literature for
a variety of applications. The synthesis of these particles typically
involves the reduction of the metal ions out of solution leading to
237
238 Daniel Montiel et al.
Ostwald ripening of seed particles [4]. Particles are typically ripened

to a desired size and then stabilized with a surfactant.
The advantageous optical properties and the known surface
chemistry of these nanoparticles have led experimenters to conjugate
them to biological materials for a variety of applications. One can
attach small molecules to the nanoparticle surface to prevent nonspe-
cific adsorption of biomolecules or attach molecules with an orthog-
onal chemistry for hierarchical construction. Applications of these
hybrid particles range from analyte detection [5], electron micros-
copy contrast agents [6], photo-thermal therapies [7, 8], drug deliv-
ery systems [9], bio-barcoding [10, 11], among others [12–16].
For certain applications, e.g., single-particle-based sensing and
spectroscopy, using particles with increased scattering cross-
sections such as larger-diameter gold nanoparticles may be desir-
able. Utility of any hybrid nano-material relies on the ability to
produce stable colloids resistant to aggregation and nonspecific
adsorption of biomolecules under physiologically relevant condi-
tions. As the particle size becomes greater, however, the increased
susceptibility to surface oxidation and the larger surface area per
particle present challenges to robust incorporation of these parti-
cles into the optical toolbox of nano-materials [17–20]. The pas-
sivation of bare metal nanoparticle surfaces with organic ligands is
important for colloid stability and minimizing nonspecific adsorp-
tion of biomolecules. The attachment chemistry to the metal
nanoparticle [21], coverage efficiency [22, 23], ligand length [24],
and terminal group are all important control parameters [25].
Thiol groups are commonly used for attachment of small mol-
ecules to noble metal nanoparticles because chemical bonds are
formed with the metal surface [26]. To reduce ligand lability and
reduce susceptibility to oxidative cleavage, dithiol groups have been
used to increase coordination to the gold surface [21, 27, 28].
The ligand’s terminal group can serve multiple roles. One of
these is to increase steric and electrostatic repulsion between par-
ticles to increase colloid stability. Zwitterionic ligands, although
uncharged, form a charged double layer around the nanoparticle
and have been shown to yield stable colloids [29–33] under physi-
ologically relevant conditions.
Another role for the terminal moiety is to provide an orthogo-
nal chemistry for rational construction of biological or inorganic
nanoparticles. Streptavidin, a 53 kDa tetrameric protein, binds
strongly to biotin [34] and has been used to construct bioinor-
ganic composites for various applications [35–37].
For self-assembled monolayers on flat surfaces there exist tech-
niques such as ellipsometry [38] and quartz crystal microbalance
[39] measurements to quantify surface coverage. Nanoparticles
typically display particle-to-particle variation; the extent of surface
coverage [23; 40] and nonspecific adsorption [41] need to be
characterized to understand the scope of a new particle system.
Surface Passivation in Single-Nanoparticle Based Biosensing 239
Ideally, one would desire the ligands to be resistant to nonspecific

adsorption and dissociation from the particle surface while also
being bio-conjugatable.
Methods for the attachment of biological materials to noble
metal nanoparticles are mature enough to allow the synthesis and
purification of discrete 1-1 bio-conjugated nanoparticles [42]. The
more general topic of interfacing biological molecules with
nanoparticle surfaces has been reviewed previously [41; 43]. The
majority of these bio-conjugation methods, however, have been
applied to smaller-diameter gold nanoparticles (diameter < 40 nm).
The increased surface area per particle for larger diameter noble
metal nanoparticles presents challenges for nonspecific absorption
and colloid stability [19; 20].
Colloidal silver nanoparticles have an exceptionally strong and
sharp plasmon resonance useful for different applications. Due to
their utility, there have been numerous synthetic procedures using
different reducing agents, such as citric acid [44], ascorbic acid
[45], sodium borohydride [46], and hydrogen gas [47]. On the
other hand, silver nanoparticles are even more susceptible to sur-
face oxidation than gold nanoparticles, thereby impacting on the
proper preparation for biological applications. This can be over-
come by synthesizing the desired silver nanoparticles in house
(rather than purchasing them from commercial sources) and pas-
sivating the freshly synthesized particles immediately. Once the sil-
ver nanoparticles are properly passivated, they can be stably stored
for longer period of time and still retain the reactivity for further
bio-conjugation.
This chapter details the practical procedures that have been
routinely used in our laboratory for the preparation of nanoparti-
cles for biological applications. Subheading 3 is conceptually
divided into three parts. The first part describes the synthesis of
three ligands for particle passivation. The first is a compact zwit-
terionic ligand containing a sulfobetaine moiety derived from lipoic
acid (LA). The second is a dithiol ligand with a biotin terminal
group, enabling the bio-conjugation to streptavidin-modified spe-
cies. The third procedure describes the modification of single-
stranded DNA with LA, for direct bio-conjugation of DNA onto
larger gold nanoparticles using EDC chemistry. The relatively low
cost of aminated primers allows for the increased preparative scale
necessary to passivate large gold nanoparticles with DNA. Here,
amine 5′-modified ssDNA is coupled to LA and then a comple-
mentary strand, 5′-modified with biotin, is annealed to the
LA-modified strand. The second part describes a silver nanoparti-
cle synthesis using a modified ascorbic acid reduction method and
nanoparticle passivation with dithiol ligands. The last part describes
methods to characterize the effectiveness of the passivation.
Particles are characterized by absorption spectroscopy, gel electro-
phoresis, and column chromatography to ensure the robustness
of the passivation. It should be pointed out that the latter two

characterization methods are routinely used in preparing proteins
and nucleic acids; ensuring that the passivated gold and silver
nanoparticles are compatible with these standard biochemical
methods is a necessary step to enable subsequent manipulation of
the desired hybrid nano-materials.
2 Materials
2.1 Chemicals 1. 18.2 MΩ-cm water.

2. Room temperature is 20 ∘ C.
3. ( ± )α-Lipoic acid, ≥ 99 %.
4. Thionyl chloride (SOCl2), 99+ %.
5. 3-Dimethylamino-1-propanol, 99 %.
6. 1,3-Propanesultone, 99 %.
7. Sodium borohydride (NaBH4).
8. Chloroform (CHCl3), ACS grade.
9. Acetone, HPLC grade.
10. Sodium bicarbonate.
11. Ethyl acetate (EtOAc), ACS grade.
12. Agarose, bioreagent grade.
13. New England Biolabs Tridye 1 kb DNA ladder.
14. Ethanol, 200 proof.
15. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), stored
in a desiccator at − 20 ° C.
16. Sulfo-NHS, stored in a dessicator at 4 ° C.
17. Ethylenediaminetetraacetic acid (EDTA), bioreagent grade.
18. 0.1 M MES (2-(N-morpholino)ethanesulfonic acid sodium
salt) pH 6, bioreagent grade.
19. Sephacryl S-300, size exclusion resin (GE Lifesciences).
20. Chromatography media sampler pack (Bio-Rad).
21. 80 nm gold nanoparticles (British BioCell International).
22. Silver nitrate, 99.9999 %.
23. Sodium acetate, bioreagent grade.
24. Sodium citrate dihydrate, granular/certified.
25. L-Ascorbic acid, ACS grade.
26. Sodium hydroxide, ACS grade.
27. DithiolalkanearomaticPEG6-NHS (CAS# 936115-55-8,
SensoPath Technologies) (see Note 1).
28. EZ-Link Amine-PEG2-Biotin (Pierce).

29. 10 mM Hepes (4-(2-hydroxyethyl)piperazine-1-ethanesulfonic
acid) pH 8, bioreagent grade.
2.2 Equipment 1. Aluminum foil.

2. Pasteur pipette.
3. Cotton balls.
4. Mini-tube rotator.
5. Rotary evaporator.
6. Speed-Vac, i.e., Savant SVC-100 or equivalent.
7. Schlenk line.
8. Temperature-controlled water bath.
9. Three-neck jacketed round bottom flask.
10. Syringe pumps.
11. Agarose gel electrophoresis apparatus.
12. UV–Vis absorption spectrophotometer.
3 Methods
3.1 Protocol for 1. All reactions performed under N2 with a Schlenk line, unless
Dihydrolipoic Acid- otherwise stated.
Sulfobetaine (DHLA- 2. Dilute lipoic acid (1,000 mg, 4.85 mmol) in 10 mL CHCl3 and
SBE) Synthesis cool in an ice bath.
3. Dissolve SOCl2 (5.34 mmol) in 2–3 mL CHCl3 and add
drop-wise.
4. After addition of SOCl2, stir 30 min at room temperature.
5. Remove the solvent by rotary evaporation.
6. Re-dissolve the product in 10 mL CHCl3, cool in an ice bath
and purge with N2.
7. Dilute 4.85 mmol 3-dimethylamino-1-propanol with 2–3 mL
CHCl3 and add drop-wise.
8. Stir the reaction vigorously overnight at room temperature.
9. Quench the reaction with NaHCO3 (9.7 mmol) in 20 mL
water.
10. Extract the quenched reaction with EtOAc (2 ×) and wash
the combined organic extract with saturated aqueous
NaHCO3 (2 ×).
11. Combine organic layers, dry over MgSO4, and filter.
12. Remove the solvent by rotary evaporation.
13. Dissolve the product in 10 mL acetone.
O
O
HS
O N S
O
SH O
DHLA-SBE
F. W. 415.2 g/mol
Fig. 1 The chemical structure of dihydrolipoic acid-sulfobetaine (DHLA-SBE)
14. Dissolve 4.85 mmol 1,3-propanesultone in 2–3 mL acetone

and add dropwise at room temperature.
15. Reflux the reaction at 75 ° C overnight under N2. The final
product will precipitate out of solution.
16. Filter the precipitate and wash with CHCl3. Dissolve the prod-
uct in water and extract the aqueous layer with CHCl3. The
desired product will be in the aqueous layer.
17. Filter through a pasteur pipette with a cotton plug and evapo-
rate under reduced pressure.
18. Characterize the product. 1H NMR (500 MHz, DMSO-d6):
δ [ppm]: 3.46–3.50 (t, 2H), 3.38–3.43 (m, 2H), 3.28–3.35
(m, 2H), 2.98–3.06 (s, 6H), 2.77–2.85 (m, 2H), 3.46–3.50
(m, 4H), 2.19–2.23 (t, 1H), 1.87–2.08 (m, 4H), 1.79–1.86
(m, 2H), 1.27–1.63 (m, 6H). ESI MS m ⁄ z: 414.14451
(M+H)+.
19. Reduce dithiol bond with an equimolar amount of NaBH4 just
before use.
20. Store unreduced product as a solid in − 20 ° C freezer until
needed. The structure shown in Fig. 1 is of the product after
reduction of the disulfide bond with NaBH4.
3.2 Protocol 1. Store 10 mg (12.6 μmol) dithiolalkanearomatic PEG6-NHS

for Dithiol-Biotin ligand at − 20 °C, unopened in a dessicator until required.
Synthesis 2. Dissolve 6 mg (16 μmol) EZ-Link Amine-PEG2-Biotin in
500 μL 10 mM Hepes buffer at pH 8.
3. Add solution immediately to freshly opened bottle of dithiolal-
kanearomatic PEG6-NHS (see Note 1 )
4. Let mixture react for 2 h at room temperature.
5. Store the final solution (dithiol-biotin) at − 20 ° C.
3.3 Preparation 1. Prepare 10 mg/mL solution of lipoic acid in DMSO.

of Lipoic Acid 2. Equilibrate EDC and sulfo-NHS to room temperature.
Functionalized
3. Add 360 μL of 10 mg/mL lipoic acid (0.0174 mmol) solution
Single-Stranded DNA
to 640 μL 0.1 M MES at pH 6.
4. Add aqueous lipoic acid solution to an Eppendorf tube

containing 14 mg (0.073 mmol) EDC and 22 mg (0.1 mmol)
sulfo-NHS.
5. Let mixture react 5 min at room temperature to activate car-
boxylic acid.
6. Prepare 500 μL 0.5 mM aminated DNA in water and add to
activated lipoic acid mixture.
7. Rotate on a mini-tube rotator at room temperature for 2 h.
8. Ethanol precipitate DNA using 1 mL 0.3 M sodium acetate in
ethanol.
9. Incubate at 4 ° C for 1 h.
10. Spin at 15,870 × g for 1 h at 4 ° C in a microcentrifuge.
11. Carefully discard the supernatant.
12. Add 500 μL 70 % ethanol, cooled to 4 ° C.
13. Centrifuge the solution at 15,870 × g for 20 min at 4 ° C.
14. Carefully discard the supernatant.
15. Dry the DNA pellet with a Speed-Vac.
16. Dissolve the pellet in 50 μL 10 mM Hepes buffer at pH 8.
17. Quantitate the DNA by absorbance at 260 nm.
18. Dilute complementary ssDNA, 5′-modified with a biotin, in
10 mM HEPES buffer at pH 8.
19. Add an equimolar amount of complementary ssDNA to lipoic
acid-functionalized single-stranded DNA.
20. Add sodium chloride to a concentration of 100 mM.
21. Heat DNA to 95 ° C for 2 min.
22. Cool DNA on bench top at room temperature.
23. Store the lipoic acid-DNA at − 20 ° C until needed.
3.4 Large Gold 1. Thaw DHLA-SBE from Subheading 3.1, dithiol-biotin from
Nanoparticle Subheading 3.2, and lipoic acid-DNA from Subheading 3.3 to
Passivation room temperature.
2. Add 10 μL of 50 mM DHLA-SBE to either 1 μL of 10 mM
dithiol-biotin or 10 μL of 500 μM lipoic acid-DNA (see Notes 2
and 3).
3. Add an equimolar amount of NaBH4 to the mixture and incu-
bate 5 min.
4. Add reduced thiols to 500 μL of 80-nm gold nanoparticle
(1010 particles/mL).
5. Rotate the mixture on mini-tube rotator overnight at room
temperature.
6. Save an aliquot of mixture for characterization by gel
electrophoresis and UV–Vis absorption spectroscopy.
Fig. 2 Setup of the apparatus used for silver nanoparticle synthesis. Two syringe
pumps are connected to a 500-mL thermally jacketed three-neck round bottom
flask. The flask is placed on top of a stir plate. The temperature of the water flow-
ing through the water-jacket is 75 ° C
7. Dialyze gold nanoparticles in a 300-kDa membrane for 2 h at

room temperature in 4 L water (see Note 4 ).
8. Exchange the 4 L water every 2 h for a total of three exchanges.
3.5 Silver 1. Prepare 20 mg/mL silver nitrate stock solution fresh for each
Nanoparticle synthesis.
Synthesis and 2. Prepare 70 mg/mL trisodium citrate, 2 mg/mL ascorbic acid,
Passivation and 0.1 N NaOH stock solutions.
3. Connect a 500 mL three-neck jacketed round bottom flask to
a circulating water bath set to 75 ° C.
4. Wrap flask with aluminum foil. An example synthesis apparatus
is depicted in Fig. 2.
5. Add 100 mL of ultra-pure 18.2 MΩ-cm water along with a stir
bar.
6. Center the flask on a stir plate set to 750 rpm.
7. Cap the open necks of the flask with rubber stoppers.
8. Load two syringe pumps with NaOH and ascorbic acid stock
solutions. Set syringe pumps to 250 μL/min for 2 min of
NaOH addition, and 2 mL/min for 5 min of ascorbic acid
addition.
9. Once the water reaches 75 ° C, equilibrate for 5 min before
adding reactants.
10. Quickly add 1 mL of silver nitrate. Wait 5–10 s, then quickly
add 1 mL of trisodium citrate stock solution by syringe.
1 monothiol passivated after dialysis

DHLA-SBE passivated after dialysis
stock sample before dialysis
normalized absorbance
0.8
0.6
0.4
0.2
0
500 600 700 800 900 1000
wavelength (nm)
Fig. 3 The absorption spectrum of 80-nm gold nanoparticles passivated with a

monothiol ligand (red squares ) taken after dialysis shows increased near-
infrared absorption with respect to the unpassivated particles before dialysis
(black curve ). The near-infrared absorption band is indicative of aggregation. The
spectrum taken after dialysis of particles passivated with DHLA-SBE (blue circles )
do not show the same increase. All spectra are normalized to the absorption
maximum at the plasmon resonance wavelength
11. Add 500 μL of NaOH at 250 μL/min for 2 min and 10 mL of

ascorbic acid at a rate of 2 mL/min for a total of 5 min.
12. Remove any injection needles and let react at the set tempera-
ture for 30 min.
13. Store product at 4 °C.
14. Spin 10 mL of 0.5 nM silver nanoparticles at 13,800 × g for
30 min.
15. Remove the supernatant and re-suspend the pallet in
methanol.
16. Add DHLA-SBE to a final concentration of 500 mM.
17. Rotate on a mini-tube rotator for 24 h at room temperature.
3.6 Absorption 1. Record UV–Vis absorption spectrum for each particle sample
Spectra of Large before and after dialysis. If the particles are stable and mono-
Gold Nanoparticles disperse, they should still retain their characteristic plasmon
band. If the particles are incompletely passivated, then the
absorption spectra may display a large shift of the plasmon
resonance and characteristic increase in near infra-red absorp-
tion [48]. A representative figure comparing particle aggrega-
tion after dialysis is shown in Fig. 3.
Fig. 4 The brightfield illuminated (a), UV illuminated (b), and overlaid images
(c) of a 0.5 % 0.5 × TBE agarose ethidium bromide stained gel. Lane 1 is New
England Biolab’s Tridye 1 kb DNA ladder while lanes 2 and 3 are the passivated
and unpassivated particles, respectively. The passivated particles (red arrow )
move quickly into the gel while the unpassivated particles (black arrow ) remain
near the load well after 35 min at 6.5 V/cm
3.7 Gel 1. Cast a 0.5 % 0.5 ×TBE agarose gel. Use a comb with smallest
Electrophoresis teeth available. The teeth of the comb used in this protocol are
of Large Gold 3 mm × 1.5 mm × 10 mm (see Note 5).
Nanoparticles 2. Mix 15 μL gold nanoparticle sample with 5 μL 15 % Ficoll solu-
tion as the loading buffer.
3. Load samples into the gel with 0.5 ×TBE as the running buf-
fer. Reserve a lane in the gel for a DNA marker such as New
England Biolab’s Tridye 1 kb DNA marker. The Tridye marker
contains three organic dyes which are a good visualization tool
to monitor the progress of the gel (see Note 6).
4. Run samples at a field strength of 6.5 V/cm (see Note 7).
5. Record a brightfield image of the particle on a lightbox. The
large gold nanoparticles, if properly passivated, should run as a
tight pink band. A representative gel is seen in Fig. 4.
3.8 Column Test 1. Prepare 10 mL of a 50 % slurry of Sephacryl S-300 size exclu-

for Large Gold sion resin with 18.2 MΩ-cm water (see Note 8).
Nanoparticles 2. Pack a glass drip column (1.5 cm inner diameter) with 5 mL of
slurry (see Notes 9 and 10).
3. Equilibrate the column with two column volumes of water.
4. Let the water drip through column until the meniscus of the
mobile phase reaches the top of the packed resin bed.
Fig. 5 Incompletely passivated large gold nanoparticles will become immobilized

at the top of various liquid chromatography resins. Here, unpassivated 80 nm
diameter gold nanoparticles have become immobile on a Sephacryl S-300
column
5. Once the water has reached the top of the column, carefully
load 500 μL of passivated particle sample from Subheading 3.4
to the top of the column.
6. Let gold sample move into the column.
7. After the gold has entered the top of the resin bed, carefully
wash the column with enough water to wash the particles
through the column. Incompletely passivated gold nanoparti-
cles will remain at the top of the column, such as in Fig. 5.
8. Repeat column test with other anion exchange, cation
exchange, DEAE, and hydroxyapatite resins from Bio-Rad’s
chromatography media sampler pack to determine scope of
usable resins with the particle system. The passivated particles
should not aggregate at the tops of the columns.
4 Notes
1. The NHS ligand is very reactive and will hydrolyze quickly if

not used immediately. Store at − 20 ° C and use immediately
after opening. Do not store unreacted stock solutions.
2. Nanoparticles passivated by lipoic acid-derived ligands appear
to be most stable when a mixture of at least two ligands are
used. In the presented protocol, DHLA-SBE is mixed with a
dithiol biotin ligand. Other ligands, such as lipoic acid modi-
fied peptides, can also be used.
3. Optimizing the ratio of the relative ligands can help with particle
stability. For large gold nanoparticles, millimolar concentrations
of ligands are necessary to ensure adequate passivation.
4. Excess DNA can be difficult to remove by dialysis. Sephacryl
S-300 size exclusion column described in Subheading 3.8 is
used to separate large gold nanoparticles from unbound DNA.
The large gold nanoparticles will be eluted in the excluded
volume limit, whereas the DNA will enter the resin.
5. When casting a gel, combs with narrow teeth are useful for
visualizing dilute samples. The path length of light passing
through the sample increases with the total sample height in a
transillumination configuration.
6. DNA can be visualized by ethidium bromide staining. Unlike
small gold nanoparticles, large gold nanoparticles show some
increased scattering by UV transillumination. Visualize the gel
by UV transillumination before staining with ethidium bro-
mide to account for the particle scattering signal.
7. Setting the voltage too high will cause particles to streak in
the gel.
8. The pore size of the Sephacryl S-300 resin is such that
nanoparticles greater than approximately 60 nm will run in
the excluded volume limit. Particles smaller than this limit
such as quantum dots and DNA molecules ≤ 90 bp can be
resolved by the resin.
9. Use the minimal amount of resin necessary to achieve desired
separation. Excessively large columns can dilute the recov-
ered sample and make reconcentrating the sample
difficult.
10. Acetone has a strong UV absorbance and can be used to deter-
mine the retention limit of a freshly packed column. One can
use passivated large gold nanoparticles to determine the
excluded volume limit.
Acknowledgements
This work was supported by the Department of Energy and the

NIH-NIGMS.
References
1. Kreibig U, Vollmer M (1995) Optical proper- 3. Cobley CM, Skrabalak SE, Campbell DJ, Xia Y
ties of metal clusters. Springer, Berlin (2009) Shape-controlled synthesis of silver
2. Kelly KL, Coronado E, Zhao LL, Schatz GC nanoparticles for plasmonic and sensing appli-
(2003) The optical properties of metal nanopar- cations. Plasmonics 4:171–179
ticles: the influence of size, shape, and dielectric 4. Turkevich J, Stevenson PC, Hillier J (1951) A
environment. J Phys Chem B 107:668–677 study of the nucleation and growth processes in
the synthesis of colloidal gold. Discuss Faraday apparent pKa of the coating ligands. J Am
Soc 11:55–75 Chem Soc 113:2192–2197
5. Kinkhabwala A, Yu Z, Fan S, Avlasevich Y, Müllen 20. Hill HD, Millstone JE, Banholzer MJ, Mirkin
K, Moerner WE (2009) Large single-molecule CA (2009) The role radius of curvature plays in
fluorescence enhancements produced by a bowtie thiolated oligonucleotide loading on gold
nanoantenna. Nat Photon 3:654–657 nanoparticles. ACS Nano 2:418–424
6. Roth J, Bendayan M, Orci L (1978) 21. Wang Z, Tan B, Hussain I, Schaeffer N, Wyatt
Ultrastructural localization of intracellular anti- MF, Brust M, Cooper AI (2007) Design and
gens by the use of protein A-gold complex. polymeric stabilizers for size-controlled synthe-
J Histochem Cytochem 26:1074–1081 sis of monodisperse gold nanoparticles in water.
7. Link S, El-Sayed MA (2000) Shape and size Langmuir 23:885–895
dependence of radiative, non-radiative and 22. Hurst SJ, Lytton-Jean AKR, Mirkin CA (2006)
photothermal properties of gold nanocrystals. Maximizing DNA loading on a range of gold
Int Rev Phys Chem 19:409–453 nanoparticle sizes. Anal Chem 78:8313–8318
8. Huang X, El-Sayed IH, Qian W, El-Sayed MA 23. Demers LM, Mirkin CA, Mucic RC, Reynolds
(2006) Cancer cell imaging and photothermal RA III, Letsinger RL, Elghanian R,
therapy in the near-infrared region by using Viswanadham G (2000) A fluorescence-based
gold nanorods. J Am Chem Soc 128: method for determining the surface coverage
2115–2120 and hybridization efficiency of thiol-capped oli-
9. Paciotti GF, Myer L, Weinreich D, Goia D, gonucleotides bound to thin films and nanopar-
Pavel N, McLaughlin RE, Tamarkin L (2004) ticles. Anal Chem 72:5535–5541
Colloidal gold: a novel nanoparticle vector for 24. Prasad BLV, Stoeva SI, Sorensen CM, Klabunde
tumor directed drug delivery. Drug Deliv KJ (2002) Digestive ripening of thiolated gold
11:169–183 nanoparticles: the effect of alkyl chain length.
10. Nam J-M, Thaxton CS, Mirkin CA (2003) Langmuir 18:7515–7520
Nanoparticle-based bio-bar codes for the ultra- 25. Cosgrove T (ed) (2005) Colloid science prin-
sensitive detection of proteins. Science ciples, methods and applications. Blackwell,
301:1884–1886 Oxford
11. Nam J-M, Stoeva SI, Mirkin CA (2004) Bio- 26. Pakiari AH, Jamshidi Z (2010) Nature and
bar-code-based DNA detection with PCR-like strength of M–S bonds (M = Au, Ag, and Cu)
sensitivity. J Am Chem Soc 126:5932–5933 in binary alloy gold clusters. J Phys Chem A
12. Sperling R, Gil P, Zhang F, Zanella M, Parak 114:9212–9221
WJ (2008) Biological applications of gold 27. Li Z, Jin R, Mirkin CA, Letsinger RL (2002)
nanoparticles. Chem Soc Rev 37:1896–1908 Multiple thiol-anchor capped DNA-gold
13. Guo S, Dong S (2009) Biomolecule–nanopar- nanoparticle conjugates. Nucl Acid Res
ticle hybrids for electrochemical biosensors. 30:1558–1562
Trends Anal Chem 28:96–109 28. Mei BC, Oh E, Susumu K, Farrell D,
14. Hill HD, Mirkin CA (2006) The bio-barcode Mountziaris TJ, Mattoussi H (2009) Effectors
assay for the detection of protein and nucleic of ligand coordination number and surface
acid targets using DTT-induced ligand curvature on the stability of gold nanoparticles
exchange. Nat Prot 1:324–336 in aqueous solutions. Langmuir 25:
15. Katz E, Willner I (2004) Integrated 10604–10611
nanoparticle-biomolecule hybrid systems: syn- 29. Muro E, Pons T, Lequeux N, Fragola A,
thesis, properties, and applications. Angew Sanson N, Lenkei Z, Dubertret B (2010) Small
Chem Int Ed 43:6042–6108 and stable sulfobetaine zwitterionic quantum
16. Khlebtsov NG, Dykman LA (2010) Optical dots for functional live cell imaging. J Am
properties and biomedical applications of plas- Chem Soc 132:4556–4557
monic nanoparticles. J Quant Spectrosc Radiat 30. Susumu K, Oh E, Delehanty JB, Blanco-Canosa
Transfer 111:1–35 JB, Johnson BJ, Jain V, Hervey WJ, Algar WR,
17. Henglein A (1993) Physicochemical properties Boeneman K, Dawson PE, Medintz IL (2011)
of small metal particles in solution: “microelec- Multifunctional compact zwitterionic ligands
trode” reactions, chemisorption, composite for preparing robust biocompatible semicon-
metal particles, and the atom-to-metal transi- ductor quantum dots and gold nanoparticles.
tion. J Phys Chem 97:5457–5471 J Am Chem Soc 133:9480–9496
18. Ivanova OS, Zamborini FP (2009) Size- 31. Park J, Nam J, Won N, Jin H, Jung S, Cho
dependent electrochemical oxidation of silver S-H, Kim S (2011) Compact and stable quan-
nanoparticles. J Am Chem Soc 132:70–72 tum dots with positive, negative, or zwitter-
19. Wang D, Nap RJ, Lagzi I, Kowalczyk B, Han ionic surface: specific cell interactions and
S, Grzybowski BA, Szleifer I (2011) How and non-specific adsorptions by the surface charges.
why nanoparticle’s curvature regulates the Adv Funct Mater 21:1558–1566
32. Estephan ZG, Jaber JA, Schlenoff JB (2010) 40. Kim E-Y, Stanton J, Vega RA, Kunstman KJ,
Zwitterion-stabilized silica nanoparticles: toward Mirkin CA, Wolinsky SM (2006) A real-time
nonstick nano. Langmuir 26:16884–16889 PCR-based method for determining the surface
33. Rouhana LL, Jaber JA, Schlenoff JB (2007) coverage of thiol-capped oligonucleotides bound
Aggregation-resistant water-soluble gold onto gold nanoparticles. Anal Chem 34:e54
nanoparticles. Langmuir 23:12799–12801 41. Park S, Hamad-Schifferli K (2010) Nanoscale
34. Weber PC, Ohlendorf DH, Wendoloski JJ, interfaces to biology. Curr Opin Chem Biol
Salemme FR (1989) Structural origins of high- 14:616–622
affinity biotin binding to streptavidin. Science 42. Zanchet D, Micheel CM, Parak WJ, Gerion D,
243:85–88 Williams SC, Alivisatos AP (2002)
35. Kawde A-N, Wang J (2003) Amplified electri- Electrophoretic and structural studies of DNA-
cal transduction of DNA hybridization based directed Au nanoparticle groups. J Phys Chem
on polymeric beads loaded with multiple gold B 106:11758–11763
nanoparticle tags. Electroanalysis 43. Surugau N, Urban PL (2009) Electrophoretic
16:101–107 methods for separation of nanoparticles. J Sep
36. Kerman K, Chikae M, Yamamura S, Tamiya E Sci 32:1889–1906
(2007) Gold nanoparticle-based electrochemi- 44. Pillai ZS, Kamat PV (2004) What factors con-
cal detection of protein phosphorylation. Anal trol the size and shape of silver nanoparticles in
Chim Acta 588:26–33 the citrate ion reduction method. J Phys Chem
37. Zhu D, Tang Y, Xing D, Chem WR (2008) B 108:945–951
PCR-free quantitative detection of genetically 45. Qin Y, Ji X, Jing J, Liu H, Wu H, Yang W (2010)
modified organism from raw materials. An elec- Size control over spherical silver nanoparticles by
trochemiluminescence-based bio bar code ascorbic acid reduction. Colloids Surf A
method. Anal Chem 80:3566–3571 Physicochem Eng Asp 372:172–176
38. Prime KL, Whitesides GM (1991) Self- 46. Lee C, Meisel D (1982) Adsorption and
assembled organic monolayers: model systems surface-enhanced raman of dyes on silver and
for studying adsorption of proteins at surfaces. gold sols. J Phys Chem 86:3391–3395
Science 252:1164–1167 47. Evanoff DD Jr, Chumanov G (2004) Size-
39. Marx KA (2003) Quartz crystal microbalance: controlled synthesis of nanoparticles. 1.‘Silver-
a useful tool for studying thin polymer films only’ aqueous suspensions via hydrogen
and complex biomolecular systems at the solu- reduction. J Phys Chem B 108:13948–13956
tion–surface interface. Biomacromolecules 48. Kreibig U, Genzel L (1985) Optical absorption
4:1099 of small metallic particles. Surf Sci 156:678–700
Chapter 19
Noncovalent Intracellular Drug Delivery of Hydrophobic

Drugs on Au NPs
Tennyson Doane and Clemens Burda
Abstract
The successful delivery of hydrophobic drugs to cellular targets continues to present challenges to the
pharmaceutical industry. The advances made by nanotechnology have generated new avenues for selec-
tively loading, delivering, and targeting these drugs to their biological targets without compromising
efficacy. Here, we describe how gold nanoparticles (Au NPs) functionalized with polyethylene glycol
(PEG) can be evaluated for the delivery of hydrophobic drugs in aqueous systems. Specifically, we describe
Au NP synthesis, ligand exchange, and delivery evaluation at-the-bench for screening of potential drug
candidates.
Key words Gold nanoparticles, Noncovalent drug delivery, Ligand exchange, Hydrophobic
1 Introduction
Many promising drug candidates for the treatment of human

diseases are hydrophobic and have limited effectiveness due to poor
transport [1]. To overcome this problem, research is driven either
to modify the drug to improve solubility or to design an effective
transport vector for in vivo applications. Modifications to drug
structure, however, can have the unwanted effect of limiting
potency and subcellular localization in targeting a given disease. For
this reason, the use of transporting vectors has become very popular
for the delivery of hydrophobic drugs, such as the polyethylene gly-
col (PEG) liposomal formulation of doxorubicin known as Doxil®
which currently has clinical applications [2].
One promising approach to improving drug delivery is through
the use of nanoparticles (NPs) which have the potential for both
simultaneous delivery and diagnostic capabilities (so-called ther-
anostics) [3]. Gold nanoparticles (Au NPs), in particular, have
been highly researched due to their highly tunable size [4], unique
optical properties [4], facile platform based modification [5] and
reported low toxicity [6]. Moreover, the ability to rationally design
251
252 Tennyson Doane and Clemens Burda
NPs to optimize the enhanced permeability and retention (EPR)

effect [7] can provide further controls based on a solid material.
Hydrophobic drugs have been successfully loaded via noncovalent
techniques which improve solubility but still retain favorable
release kinetics both in vitro and in vivo [8, 9]
The physical basis for hydrophobic drug loading relies on the
use of amphiphilic ligand coatings, which are commonly based on
oligomers of polyethylene glycol (PEG). PEG is commonly used in
biomedical applications due to favorable protein resistance and
“stealth” like properties [10]. Work by Lucarini et al. has investi-
gated the partitioning of hydrophobic molecules in PEG oligomer
termiminated alkyl monolayers via electron spin resonance [11,
12], and found that drug portioning is dependent on the packing
of the surface ligands. The incorporation of a hydrophobic drug
into the amphiphilic coating surrounding the nanoparticle pro-
vides a unique opportunity to study how diffusion related drug
transport across membranes works in living systems, and due to the
reported fluorescence quenching effect due to energy transfer, can
provide important information pertaining to release behavior when
a fluorescent compound is utilized.
The following protocol describes the synthesis, loading, and
evaluation of Au NPs for the delivery of hydrophobic drugs through
noncovalent drug loading. In general the two main methodologies
for synthesizing Au NPs rely on chemical reduction of Au+3 ions
with either citrate in water (the Turkevich method [13]) or sodium
borohydride in organic solutions (the Brust method [14]). The fol-
lowing protocol is an adaptation of the later, which produces Au
NPs with a diameter of ~6 nm. Further modification of the particles
is undertaken by utilizing the strongly coordinating thiol modified
PEG. This general approach has been investigated in our laboratory
for a range of potential drug candidates, and analyzed in detail for
the photodynamic therapy drug silicon phthalocyanine 4 [8, 15].
2 Materials
All chemicals reported here are commercially available (see below).

Chemical storage must be in keeping with MSDS recommenda-
tions. Special caution should be employed for sodium borohydride
(NaBH4) which is a strong reducing agent and reacts violently with
water. Store in a dry and flame-resistant place. Properly dispose of
all waste according to regulations. The following are a list of chem-
icals needed for each step of preparation.
2.1 Preparations 1. Small 20 mL vial with cap.

for NP Synthesis 2. Clean small magnetic stir bar and stir plate.
2.1.1 Equipment 3. 200–1,000 μL mechanical pipette.
Intracellular Drug Delivery by Gold Nanoparticles 253
4. 2–20 μL mechanical pipette.

5. Three sheets of weighing paper.
6. Small 2-dram vial with cap.
7. Four clean glass pipettes.
8. One rubber bulb.
9. Two plastic 50 mL centrifuge tubes.
10. One glass cuvette.
2.1.2 Solvents 1. ACS grade Toluene. Toluene is both toxic and flammable.
Handle and dispose of accordingly.
2. ACS grade Chloroform. Chloroform is both toxic and flam-
mable. Handle and dispose of accordingly.
3. 200 Proof Ethanol. Ethanol is flammable and should be stored
accordingly.
2.1.3 Chemicals 1. Tetraoctylammonium bromide (TOAB) will be used as the

phase transfer agent. TOAB is hazardous to the environment
and should be disposed of accordingly.
2. 30 % Tetrachloroauric acid (HAuCl4) 99 % purity (Sigma
Aldrich Chemical Company, St. Louis, MO, USA.) will be
used as the Au precursor. Solution should be stored at 4 °C.
3. Dodecylamine (DDA) will be used as the capping ligand. DDA
is hazardous to the environment and should be disposed of
accordingly. In addition, DDA can be caustic, and should be
stored with similar agents.
4. ACS grade Sodium Borohydride (NaBH4) will be used as the
reduction agent. NaBH4 reacts violently with water and releases
H2 gas upon reaction which is highly explosive. Observe cau-
tion when handling and store in a cool, dry, and flammable
resistant place.
2.2 Preparations 1. Small 20 mL vial with cap.

for Au NP Ligand 2. Clean medium magnetic stir bar.
Exchange
3. Stir plate.
2.2.1 Equipment 4. 200–1,000 μL mechanical pipette.
5. 2–20 μL mechanical pipette.
6. One 2-dram vial with cap.
7. One glass pipette.
8. One glass cuvette (see Note 1).
2.2.2 Solvents 1. ACS grade chloroform (see above).

2.2.3 Chemicals 1. Thiolated-Methoxypolyethylene glycol 5000 (HS-mPEG 5k)

(Laysan Bio, Arab, GA, USA). This compound is air, tempera-
ture, and light sensitive (see Note 2), and should be stored in
an inert atmosphere at −18 °C during long term storage.
Transfer to the reaction vessel should be done quickly and with
minimal air contact.
2. Dry ice is recommended for transporting and storing the
HS-mPEG 5k to the reaction vessel in the fume hood.
2.3 Preparations 1. One 20 mL vial with cap.

for Au NP Purification 2. One 250 mL separatory funnel.
2.3.1 Equipment 3. Two VivaSpin 30,000 or 50,000 Da cutoff membrane centrifuge
tubes (Sartorious AG, Goettingen, GER) (see Note 1).
2.3.2 Solvents 1. ACS grade Ethyl Acetate will be used for solvent extraction
and purification, respectively. Caution: Ethyl Acetate has a very
low flash point and should be handled with extreme caution.
Proper storage and disposal is required according to
regulations.
2. 18 MΩ deionized water will be used for solvent extraction.
2.4 Preparations 1. 2-Dram vial with cap.

for Drug Delivery 2. One clean medium stir bar.
Evaluation
3. One clean flea-bar stir bar.
2.4.1 Equipment 4. Magnetic stir-plate
5. One glass or quartz cuvette (see Notes 1 and 3).
2.4.2 Solvents 1. ACS grade Chloroform (see above) will be used during solvent
extraction and drug loading.
2. 18 MΩ deionized water will be used during solvent extraction
and drug loading.
3. ACS grade Toluene (see above) will be used to monitor drug
release.
2.4.3 Chemicals 1. Sodium chloride (~0.5 g) will be used for solvent extraction
(aqueous to organic layers).
2. Drug compound (see Note 4).
3 Methods
3.1 Au NP Synthesis 1. Combine 5 mL of toluene with 0.25 mmol (0.137 g) of tetra-

octyl ammonium bromide (TOAB) and a stir bar in a 20 mL
vial. Stir this solution slowly. Simultaneously, place 1 mL of
water in an ice bath for later use.
2. Add 0.53 mmol HAuCl4 (367 μL of 30 % stock solution) to

the vial. After addition, the solution color should immediately
turn red.
3. Add 0.6 mmol (0.111 g) of dodecylamine (DDA) to the mix-
ture and increase to mild stirring. The solution should now
look turbid and a pale orange color.
4. Measure out 2 mmol (0.076 g) of sodium borohydride (NaBH4)
and add to the ice water prepared in step 1. Shake (or vortex)
the solution with frequent venting to prepare NaBH4 solution.
Caution: Addition of NaBH4 will cause the evolution of H2 gas
and is highly exothermic which can cause an explosion. This can
be extremely dangerous if the water is not properly chilled before
addition.
5. Increase the stirring of the solution prepared in step 3 and
slowly add the reducing solution prepared in step 4 to the
reaction mixture at a rate of 3–4 drops per minute with a glass
pipette. Instant color change at the entry of the drop should be
visible to the eye. Stirring speed is intended to generate a
highly uniform solution; if not mixed properly, the reduction
of Au will take place in only one location and the resulting
particles will aggregate.
6. As the amount of bubbling decreases with increasing NaBH4
addition, the rate at which it is added can be increased. Caution:
Allow the bubbling to subside before further addition to
ensure that none of the reaction mixture is overflows from the
container.
7. Continue stirring the solution for 20 min. Further stirring
time can increase the size of the particles.
8. Discontinue stirring and add the contents of the vial to a
40 mL centrifuge vial (plastic is acceptable). Wash out the reac-
tion vial with a 10 mL aliquot of ethanol two times. Add these
washes to the centrifuge vial and bring up the total volume to
~45 mL (see Note 5).
9. Prepare counter vial (i.e., 45 mL of water) and centrifuge the
mixtures for 10 min at 6,000 rpm (6,000 × g ). Particles should
be compressed into solid form at the bottom and sides of the
centrifuge tube. Decant the ethanol from the centrifuge tube
containing the Au NPs and add an additional 45 mL of ethanol
to the centrifuge tube. Repeat purification with centrifugation
for 10 min at 6,000 rpm (6,000 × g ).
10. Decant final ethanol wash and dry the particles via an Ar
stream. Redissolve the particles in 5 mL of chloroform, and
obtain a UV–Vis spectra of the prepared particles by adding
10 μL of prepared solution to 2 mL of chloroform in a cuvette
(dilution factor = 201). Obtain absorbance and determine
Fig. 1 Representative UV–Vis spectra of PEGylated Au Nps in chloroform
concentration of dilute sample using the extinction coefficient

at 520 nm = 1.5 × 107cm−1M−1 [8]. Typical stock concentra-
tions are 13 μM.
3.2 Ligand Exchange 1. Add 2.5 mL of the synthesized DDA coated Au NPs in chloro-
for HS-mPEG form to a 10 or 20 mL vial. Dilute this volume with 2.5 mL of
chloroform to bring the total volume up to 5 mL. The resulting
concentration should now be half of the stock solution.
2. Using a HS-mPEG: Au NP ratio of 1,000:1, calculate the mass
of PEG required for ligand exchange. Obtain an insulated con-
tainer and fill with dry ice for the transport of PEG from the
glove box to the reaction vessel.
3. Weigh out the required amount of HS-mPEG from the glove
box and seal tightly in a gas container. Place this container in
the dry ice and transport to the reaction vial. Add a stir bar to
the reaction vial and place on a stir plate with gentle stirring.
4. Quickly add the PEG (oily flakes for > 1kDa PEG) to the reac-
tion vial, using ~1 mL of chloroform to rinse out the vial and
transfer to the Au NP solution.
5. Allow the mixture to stir at room temperature for 2 days in the
capped vial. Secure the vial to prevent spilling.
6. After 48 h, take a 101:1 dilution of the PEGylated Au NPs and
measure the UV–Vis spectra to ensure no changes in Au NP
absorption (Fig. 1).
3.3 Purification 1. Transfer the PEGylated Au NP solution to a 250 mL separa-

of PEGylated Au NPs tory funnel. Rinse the reaction vial with chloroform 2–3 times
to ensure complete transfer. Add 5 mL of water to the separa-
tory funnel. A bilayer should be clearly visible with little trans-

fer for Au NPs from the organic phase to the aqueous phase
(see Note 6).
2. Add ~50–60 mL of ethyl acetate to the separatory funnel.
Shake the funnel (with frequent venting) to ensure mixture of
the ethyl acetate with the lower chloroform layer. Allow the
layers to separate: the organic layer is now above the aqueous
layer (dEthyl Acetate = 0.8 g/mL, dWater = 1 g/mL) (see Note 7).
3. PEGylated Au NPs should now be in the lower aqueous phase.
Extract lower aqueous layer and add fresh 5 mL water to the
separatory funnel. Repeat 2–3 times until >90 % of particles
have been removed.
4. Transfer 5 mL of particles to a cutoff membrane centrifuge
tube (>30 kDa MW exclusion). Centrifuge at 6,000 rpm
(6,000 × g ) for 10 min. Once completed, drain off water from
the lower part of the tube and add another 5 mL of solvent
extracted Au NPs. Repeat as necessary until all of the particles
are in the centrifuge tube.
5. Add fresh water to the top of the centrifuge tube (~5 mL) and
centrifuge for 10 min at 7,000 rpm (8,200 × g ). When com-
pleted, drain off the lower water and add another 5 mL of
water. Repeat ten times to ensure that all of the loose PEG is
removed from the Au NPs.
6. Add 3 mL of water and extract the PEGylated Au NPs in the
top of the centrifuge tube. Further additions of 1 mL of water
can be used to ensure as many particles are collected as possi-
ble. Store in a small dram vial.
4 Hydrophobic Drug Loading on PEGylated Au NPs and Confirming Delivery
1. Transfer the aqueous PEGylated Au NPs (~5–7 mL) to a

40 mL vial. Add 10 mL of chloroform to the vial.
2. Weigh out 0.25 g of sodium chloride and add to the 40 mL
vial. Shake vigorously and allow the layers to separate. If a
bilayer is not directly apparent after 10 min, add 10 mL water
and 0.25 g of sodium chloride and shake again (see Note 8).
3. The lower organic layer should now be red in color and can be
extracted using a Pasteur pipette. Add the extract to a 20 mL
vial, taking care not to get water into the reaction mixture.
4. Place a stir bar in the reaction vessel and begin a gentle stirring.
Add the hydrophobic drug at a 30:1 M ratio to the reaction
mixture. Keep stirring the solution for ~2 h to promote incor-
poration (see Note 9).
5. After the allotted time, blow a compressed gas (argon, nitro-
gen, or air will suffice) over the mixture to evaporate off the
Fig. 2 Experimental data illustrating the release of a hydrophobic drug (Pc 4)

from PEGylated Au NPs with increasing time in a water–toluene biphase experi-
ment (inset center). The drug was monitored both by fluorescence and by
changes in an absorbance (upper right) panel during the experiment, and as PEG
is insoluble in toluene, the hydrophobic drug partitioned into the organic phase
(lower inset, 1), while the Au NP remained in the aqueous phase (2). Reproduced
with permission from ref. 8. Copyright 2008 American Chemical Society
chloroform, leaving the PEGylated Au NP-drug conjugates as

a solid. When the mixture is completely dry, add 5 mL of water
and vortex to solvate the particles.
6. Take a UV–Vis spectra of the Au NP-drug conjugates to deter-
mine the amount of drug which has now been incorporated on
the Au NPs.
7. Add 1 mL of the Au NP-drug conjugates to a cuvette. When
completed, add 1 mL of toluene to the cuvette and allow the
solution to mix gently under magnetic stirring.
8. Monitor the absorbance of the upper organic layer at different
time points (10, 20, 30, 60, 90, 120, and 240 min for example
(Fig. 2)) to determine drug release kinetics (see Note 10).
5 Notes
1. The use of new glassware is to prevent contamination of the

solution with other trace metals which may interfere with syn-
thesis. Vials should be used in accordance with lab requirements
(i.e., smaller Eppendorf® centrifuge tubes could be used rather

than two dram vials). Cleaning of stir bars and reusable glass-
ware can be undertaken using small amounts of Aqua-Regia
(four parts HCl: one part HNO3); however, caution is advised
as this mixture gives off dangerous fumes and could potentially
cause chemical burns if accidently splashed. Work should be
conducted in a fume hood with proper personal safety equip-
ment (acid resistant gloves, goggles with side shields, and lab
coat are minimal requirements).
2. These precautions are intended to minimize the risk of dimer-
ization of HS-mPEG to mPEG-SS-mPEG which may change
its reactivity. The molecular weight of the PEG can be varied
from 550 to 10,000 Da successfully.
3. The choice of cuvette material is dependent on the photo-
physical properties of the drug compound of interest. If inter-
ested in a drug with primary absorption bands in the UV
region, a quartz cuvette is required.
4. Compounds of interest should be both (1) hydrophobic and
(2) be soluble in any chloroform and one other immiscible
organic solvent. In addition, identification of major absorp-
tion and fluorescence bands will aid in determining delivery
efficiency.
5. Simply adding ethanol to the solution of organic capped Au
NPs can also cause precipitation [16], and if allowed to pro-
ceed overnight, is much gentler than centrifugation. This pro-
cess is then repeated to ensure cleaning of excess TOAB and
DDA. For the sake of time, however, we have adopted the
centrifuge technique.
6. The use of solvent extraction is intended to remove the excess
DDA which originally capped the Au NP. While the solubility
of DDA in water is minimal, the presence of PEGylated par-
ticles may encapsulate excess toxic ligands.
7. Other solvents including toluene and hexanes were initially
studied but were found to be inadequate in forcing PEGylated
Au NP partitioning into the water phase. The use of ethyl
acetate was chosen due to the very poor solubility of PEG with
small ether substituent [17]. A quick roto-evaporation of the
solution resulted in some white powder, indicating at least
some effectiveness of removing leftover organic constituents.
8. The addition salt will cause the PEG coated Au NPs to prefer-
entially partition into the lower chloroform layer and is used
during manufacturing for PEG purification [18]. Pervious
methods utilized vacuum based drying methods, but these
were given to splattering and loss of material. When there is
no chloroform available for partitioning, however, preliminary
tests demonstrated particle stability.
9. Specific loading times are expected to vary based on the

identity of the target compound.
10. As with loading, the release kinetics are expected to be drug
property specific. Note: drugs which contain a free thiol group
have been shown to bind to the particle surface and have mini-
mal drug release efficiency [15].
References
1. Ferris DP, Lu J, Gothard C, Yanes R, Thomas (PEG): influences of the corona (PEG chain
CR, Olsen J-C, Stoddart JF, Tamanoi F, Zink length and surface density) and of the core
JI (2011) Synthesis of biomolecule-modified composition on phagocytic uptake and plasma
mesoporous silica nanoparticles for targeted protein adsorption. Colloids Surf B 18:
hydrophobic drug delivery to cancer cells. 301–313
Small 7:1816–1826 11. Lucarini M, Franchi P, Pedulli GF, Pengo P,
2. Davis ME, Chen Z, Shin DM (2008) Scrimin P, Pasquato L (2004) EPR study of
Nanoparticle therapeutics: an emerging treat- dialkyl nitroxides as probes to investigate the
ment modality for cancer. Nat Rev Drug exchange of solutes between the ligand shell of
Discov 7:771–782 monolayers of protected gold nanoparticles
3. Kelkar SS, Reineke TM (2011) Theranostics: and aqueous solutions. J Am Chem Soc 126:
combining imaging and therapy. Bioconjugate 9326–9329
Chem 22:1879–1903 12. Lucarini M, Franchi P, Pedulli GF, Gentilini C,
4. Burda C, Chen X, Narayanan R, El-Sayed MA Polizzi S, Pengo P, Scrimin P, Pasquato L
(2005) Chemistry and properties of nanocrys- (2005) Effect of core size on the partition of
tals of different shapes. Chem Rev organic solutes in the monolayer of water-
105:1025–1102 soluble nanoparticles: an ESR investigation.
5. Boal AK, Rotello VM (2000) Fabrication and J Am Chem Soc 127:16384–16385
self-optimization of multivalent receptors on 13. Turkevich J, Steveson PC, Hillier J (1951) A
nanoparticle scaffolds. J Am Chem Soc 122: study of the nucleation and growth processes
734–735 in the synthesis of colloidal gold. Discuss
6. Connor EE, Mwamuka J, Gole A, Murphy CJ, Faraday Soc 11:55–75
Wyatt MD (2005) Gold nanoparticles are 14. Brust M, Walker M, Bethell D, Schiffrin DJ,
taken up by human cells but do not cause Whyman R (1994) Synthesis of thiol-
acute cytotoxicity. Small 1:325–327 derivatised gold nanoparticles in a two-phase
7. Maeda H, Wu J, Sawa T, Matsumura Y, Hori liquid–liquid system. J Chem Soc Chem 7:
K (2000) Tumor vascular permeability and the 801–802
EPR effect in macromolecular therapeutics: a 15. Cheng Y, Samia AC, Li J, Kenney ME, Resnick
review. J Control Release 65:271–284 A, Burda C (2010) Delivery and efficacy of a
8. Cheng Y, Samia AC, Meyers JD, Panagopoulos cancer drug as a function of the bond to the
I, Fei B, Burda C (2008) Highly efficient drug gold nanoparticle surface. Langmuir
delivery with gold nanoparticle vectors for in 26:2248–2255
vivo photodynamic therapy of cancer. J Am 16. Prasad BLV, Stoeva SI, Sorensen CM,
Chem Soc 130:10643–10647 Klabundae KJ (2002) Digestive ripening of
9. Kim CK, Ghosh P, Pagliuca C, Zhu Z-J, thiolated gold nanoparticles: the effect of alkyl
Menichetti S, Rotello VM (2009) Entrapment chain length. Langmuir 18:7515–7520
of hydrophobic drugs in nanoparticle mono- 17. Janda KP, Han H (1996) Combinatorial
layers with efficient release into cancer cells. chemistry: a liquid phase approah. In: Abelson
J Am Chem Soc 131:1360–1361 JN (ed) Methods in enzymology, vol 267:
10. Gref R, Luck M, Quellec P, Marchand M, combinatorial chemistry. Academic Press, San
Dellacherie E, Harnisch S, Blunk T, Muller Diego, CA, p 236 (seen on Googlebooks)
RH (2000) ’Stealth’ corona-core nanoparticles 18. Personal Correspondence with Laysan Bio
surface modified by polyethylene glycol Technical Support (2011)
Chapter 20
Modification of Carbon Nanotubes for Gene

Delivery Vectors
Victor Ramos-Perez, Anna Cifuentes, Núria Coronas,
Ana de Pablo, and Salvador Borrós
Abstract
The surface modification of carbon nanotubes (CNTs) can be tailored to allow the formation of a complex
between these potential carriers with DNA. In this chapter, protocols developed in our lab to prepare
transfection vectors through the modification of MWCNTs are described. The protocol includes sections
focused on the reduction of CNTs length, protocols to increase the dispersability of CNTs and finally,
protocols for surface modification to attach through electrostatic interactions DNA to the CNTs.
Key words Carbon nanotubes, Plasma polymerization, DNA delivery
1 Introduction
Despite promising advantages, the use of carbon nanotubes (CNTs)

for biomedical applications is limited due to their low biocompati-
bility. Naïve CNTs present highly hydrophobic surfaces, which
result in poor water solubility and dispersability, making them
unsuitable for biological applications. To improve the biocompati-
bility of CNTs, surface modification or functionalization is required
to afford solubility in aqueous solutions [1, 2]. Modification of
CNTs may be performed by covalent and non-covalent methods/
techniques [3]. Covalent functionalization exploits the chemical
reactivity of CNTs to form bonds between the CNT surface and a
hydrophilic moiety. Covalent bonding of modifying moieties hav-
ing hydrophilic groups in their structure results in increased water
solubility of modified CNTs. In contrast, non-covalent modifica-
tion techniques are based on hydrophobic interactions between
the CNT surface and the hydrophobic region of an amphiphilic
moiety. CNT wrapping with multiple amphiphilic moieties having
surfactant-like activity results in solubilization of CNTs in aqueous
media [4, 5] amphiphilic. Both techniques provide efficient CNT
261
262 Victor Ramos-Perez et al.
modification strategies for biological applications, resulting in

increased biocompatibility and reduced toxicity.
In this sense, it should be noted that pristine CNTs lack appro-
priate functional groups to enable covalent binding of biocompat-
ible groups, in order to make CNTs compatible with the biological
milieu. Therefore, the surface of CNTs requires modification by
different chemical synthetic methods to generate appropriate func-
tional groups, enabling covalent attachment of suitable solubiliz-
ing, bioactive, or biocompatible moieties. Classically, researchers
have differentiated CNTs into two zones in terms of their reactiv-
ity, i.e. sidewalls and tips. It has been shown that tips are more
prone for modification than sidewalls.
It is clear that by adjusting the surface modification of CNTs it
is possible to tailor their capability to bind genetic material and use
their ability to cross the cellular membrane to transfect the genetic
material to the cell [6, 7].
In this chapter, we describe the protocols developed in our lab
to prepare transfection vectors through the modification of
MWCNTs. The protocol includes sections describing how to
reduce CNT length, increase the dispersability of CNTs, and mod-
ify the CNT surface modification to attach DNA by electrostatic
interactions.
2 Materials
Prepare all solutions using ultrapure water (prepared by purifying

deionized water to attain a sensitivity of 18 MΩ cm at 25 °C) and
analytical grade reagents. Prepare and store all reagents at room
temperature (unless indicated otherwise). Diligently follow all
waste disposal regulations when disposing waste materials. We do
not add sodium azide to the reagents.
2.1 Carbon 1. Carbon nanotubes: The MWNT (multi-walled carbon nano-

Nanotubes and tubes) used was obtained from Sigma-Aldrich. They have a
Shortening length between 5 and 9 mm, and a diameter between 110 and
170 nm.
2. Nanotube shortening solutions: A 0.2 M solution of potassium
permanganate (BioUltra with greater purity to 99.0 %, from
Fluka). Sodium sulfite (Na2SO3) (purity 99 %, Panreac, Barcelona).
Concentrated sulfuric acid (98 %, Panreac, Barcelona).
Concentrated chlorhydric acid (Panreac Barcelona). 0.2 M
NaOH solution.
2.2 Dispersion 1. Cetyltrimethylammonium bromide CTAB (CTAB BioUltra,

of Nanotubes with a purity greater than 99 %, obtained from Sigma-Aldrich).
2. The amphiphilic polymer pHPMA, synthesized with a content
of 2.5 and 5 % cholesterol (VP25 and VP50, respectively).
Carbon Nanotubes as Gene Delivery Vectors 263
The polymers were synthesized in our laboratory (Materials

Engineering Group (GEMAT) Institut Químic de Sarrià). For
a detailed method to obtain the polymers Subheading 3.2.
3. Dispersability of nanotubes in ethanol: Ethanol purity 99.8 %.
2.3 Transfection 1. The MWNT were surface coated with allylamine by plasma
Vectors polymerization: Allylamine (98 % purity).
2. Solution for transfection vector dispersion: Solution of 0.25 M
sodium acetate buffer pH 5.
3. Gel agarose: Agarose standard agarose low molecular weight.
4. Ethidium bromide 1 % in water.
5. Green fluorescence protein plasmid (pmaxGFP, Amax).
3 Methods
Carry out all procedures at room temperature unless otherwise

specified.
3.1 Shortening 1. CNTs are subjected to 590 °C for 3 h in a flask, amorphous

of Carbon Nanotubes carbon is oxidized and removed.
by Oxidative Treatment 2. Take 25 mg of heat-treated carbon nanotubes and disperse in
with Potassium 30 mL of 0.2 M solution of potassium permanganate (KMnO4).
Permanganate 3. Add a few drops of sodium hydroxide (NaOH) 0.2 M until the
(Adapted from Ref. 8) pH of the dispersion is 10.
4. Placed the mixture in an ultrasound bath for 20 min and then
reflux for 40 min at 100 °C.
5. Add 0.7 g of sodium sulfite (Na2SO3) and 7 mL of sulfuric acid
(H2SO4). Leave the mixture agitated for 20 h.
6. Add to the mixture Milli-Q water until the volume reaches
approximately 100 mL.
7. Centrifuge the mixture for 20 min at 2,012 × g.
8. Remove the supernatant and filter the through a polytetrafluo-
roethylene membrane and filters part of the liquid is precipi-
tated with a polytetrafluoroethylene membrane (Millipore
FGLP 00 of 047) of 0.22 μM of pore size.
9. Wash thoroughly with a solution of 0.4 g/L NaOH.
10. Wash thoroughly with Milli-Q water and finally with concen-
trated HCl to remove the manganese oxide formed due to the
reduction of potassium permanganate.
11. The washing procedure continues until there is no typical
brown color of manganese oxides remains on the filter.
12. After filtration, the membrane with the filtered nanotubes is
vacuum dried for 2 or 3 days.
Fig. 1 Scheme of the amphiphilic polymers able to disperse carbon nanotubes in water
3.2 Dispersion of The synthesis of the amphiphilic polymer poly-N-hydroxypropyl

Carbon Nanotubes methacrylamide (pHPMA), is described. This polymer as well as
CTAB, are used as dispersing agents for carbon nanotubes. The
3.2.1 Synthesis
polymers obtained by radical polymerization of a monomer N-(2-
of Amphiphilic Polymers
hydroxypropyl) methacrylamide (HPMA) with propenamide
monomer 2-methyl-2-N-[6-oxo-6-(cholesterol) hexyl]
(Ma-ACAP-chol), as shown in Fig. 1. The total concentration of
monomers is fixed at 12.5 % (w/w), and the concentration of the
initiator AIBN in 2.0 % (w/w). This gives the pHPMA copolymer
having a molecular weight average of 25.000 g/mol.
1. Dissolve HPMA, Ma-ACAP-chol, and AIBN in anhydrous
DMSO. (The amounts to be used are shown in Table 1 for
different desired concentrations of the cholesteryl group
desired.)
2. The reaction mixture was purged with argon, sealed and
inserted into a flask placed in a water bath at 60 °C for 6 h.
3. Cool the reaction mixture to room temperature.
4. Isolate the copolymer by precipitation in a mixture of acetone
and diethyl ether (2:1).
5. Dry the precipitate in vacuum until a constant weight is
obtained.
The polymer is obtained as a white solid product that does not
require further purification.
Table 1
Concentrations of the different reagents to synthesize the amphiphilic polymers
Polymer HPMA Ma-ACAP-chol AIBN Product

VP25 1 g (6.98 mmol) 0.102 g (0.18 mmol) 0.176 g (0.11 mmol) 0.83 g
VP50 1 g (6.98 mmol) 0.209 g (0.37 mmol) 0.193 g (0.12 mmol) 0.67 g
3.2.2 Dispersion The dispersions methodologies have been carried out with both
with Cationic Surfactant shortened CNTs and untreated CNTs.
1. Weigh out ~0.1 mg of CNTs.
2. Prepare a solution of 0.3 mg/mL of CTAB in water (see Note 1).
3. Put the solid CNTs into the CTAB solution.
4. Treat the mixture in an ultrasound bath for 10 min.
5. The resulting dispersion is stable, with no sedimentation for
more than 1 h when the CNTS have been shortened. For
untreated CNTs, the dispersion is stable for more than 15 min.
3.2.3 Dispersion with The dispersions methodologies have been carried out with both
Amphiphilic Polymers shortened CNTs and untreated CNTs
1. Weigh out ~0.1 mg of CNTs.
2. Prepare a solution of 20 mg/mL of amphiphilic polymers in
DMSO (Stock Solution) (see Note 2).
3. Prepare a solution of 40 μg/mL of the amphiphilic polymers
in water from the stock solution.
4. Put the solid CNTs into the diluted amphiphilic polymer
solution.
5. Treat the mixture in an ultrasound bath for 10 min.
6. When the CNTs have been shortened, the resulting dispersion
is stable, with no sedimentation for more than 1 h. For
untreated CNTs, the dispersion is stable for more than 15 min.
3.3 Preparation The surface modification of CNTs is performed using the tech-
of the Transfection nique of plasma assisted CVD (plasma enhanced CVD or PECVD).
Vectors The plasma reactor used has been developed in our research group
and previously described in literature [9]. A scheme of the reactor
3.3.1 Plasma
is presented in Fig. 2.
Polymerization
1. Put 15 mg of CNTs on a 35 mm polystyrene capsule, maximiz-
ing the surface of CNTs exposed.
2. Collocate the polystyrene capsule inside the reactor under-
neath the copper coil (where the plasma is generated).
3. Close the reactor and slowly open the vacuum pump and
expected and wait until the pressure reaches 0.02–0.03 mbar.
Fig. 2 Scheme of the plasma reactor, designed to surface modification of carbon nanotubes
4. Connect a round flask containing allylamine to one of the

entrance connections of the reactor. Cover the flask with an
aluminum foil to protect the monomer from light.
5. Activate the cold trap by adding a mixture of dry ice with ace-
tone (temperature −78 °C).
6. Open the needle valve which connects the flask with the mono-
mer with the reactor. Wait until the pressure stabilizes around
0.13 mbar.
7. Wait for 15 min to ensure that there is no air inside the bal-
loon, and that only monomer allylamine vapor enters the
reactor.
8. Connect the RF generator at a power of 15 W and with a duty
cycle (DC = (ton/ton + toff)) of 2/4 (see Note 3).
9. Observe the plasma generation (pink color around the cop-
per coil).
10. Keep the plasma connected for 5 min. After that, turn off the
RF generator and wait for another 5 min to assure a complete
reaction of the allylamine monomer with the active sites gener-
ated on CNTs surface.
11. Turn off the vacuum pump and carefully open the plasma reactor.
12. Shake the CNTs inside the polystyrene capsule and reintro-
duce it in the plasma reactor.
13. Repeat steps 7–10 but using a pressure of 0.3 mbar and a RF
maximum power of 10 W.
Table 2
Concentrations studied
Ratio MWNT/DNA [MWNT] (mg/mL)

200:1 6
150:1 4.5
100:1 3
75:1 2.25
50:1 1.5
25:1 0.75
10:1 0.3
1:1 0.03
3.3.2 Complex Formation Sample preparation

Between Genetic Material
1. Disperse CNTs using the amphiphilic polymer method
and Modified Carbon
described in Subheading 3.2 using a 25 mM acetate buffer.
Nanotubes
2. Prepare the sample for electrophoresis by mixing 5.5 mL of
Optimization of the Ratio pGFP at a concentration of 0.03 g/mL (25 mM acetate buffer
Between CNTs and DNA at pH 5) and 5.5 mL CNTs dispersion (see Table 2 for the
(See Note 4) range of concentrations tested).
Gel preparation
1. Prepare an agarose gel at a concentration of 0.8 %. Dissolve 0.6 g
of agarose in 75 mL of TAE 1× buffer solution (prepared from
2.42 g Tris-base and Trizma® base, 0.57 mL of glacial acetic acid,
0.185 g of EDTA, and 500 mL of distilled water), and then add
6 μL of ethidium bromide.
2. Once the support is gelled, put 10 μL of each sample in the
wells in increasing order.
3. Connect the electrodes, applying a potential of 65 V for about
5 min.
4. Cover the TAE 1× gel and apply a potential of 65 V for 1 h.
5. Observe how each plasmid run in the gel using an UV light.
Preparation of the transfection vector
1. Disperse CNTs using the amphiphilic polymer method described
in Subheading 3.2 using also a 25 mM acetate buffer.
2. Mix equal volumes of the plasmid at concentration of 0.03 μg/
μL in acetate buffer solution (pH 5) and the CNTS dispersion,
prepared according to Subheading 3.2 at a concentration opti-
mized according to Subheading 3.3.
4 Notes
1. CTAB is a cationic surfactant and stabilizes the CNTs disper-

sion by electrostatic repulsion.
2. The amphiphilic polymer is poorly soluble in water. Thus, the
stock solution is prepared in DMSO and diluting the stock solu-
tion in water makes the working solution. The final amount of
DMSO in the working solution is always less than 0.5 % (w/w).
3. The term “duty cycle” used means the ratio between the time
the power is connected and the total time, and is a characteris-
tic of pulsed plasma mode. In these conditions the plasma is on
only half the time, preventing the destruction of the monomer
under the active gas.
4. Once the CNTS are successfully modified with polyallylamine
(Subheading 3.3) the complex with the genetic material is
formed through ionic interaction between negatively charged
phosphate groups on DNA, and the positively charged amines
on the surface of the modified CNTs at pH 5. However, before
the formation of the plasmid vector, you need to determine the
load capacity of DNA presented by the CNTs. The aim is to
determine the amount of genetic material that can be charged
per unit mass of carbon nanotubes. Therefore, agarose gel
electrophoresis is used to probe the CNTs when the same
amount of DNA is exposed to increasing amounts of CNTs.
When the DNA does not migrate towards the positive elec-
trode indicates that it is complexed with the CNTs, as they are
unable to pass through the pores of the gel due to the large
size of the complex.
References
1. Chen J, Chen Q, Ma Q (2012) Influence of dispersion of multi-walled carbon nanotubes.

surface functionalization via chemical oxida- J Colloid Interface Sci 354(1):144–151
tion on the properties of carbon nanotubes. 6. Dovbeshko GI, Repnytska OP, Obraztsova ED,
J Colloid Interface Sci 370(1):32–38 Shtogun YV, Andreev EO (2003) Study of
2. Chen RJ, Zhang Y, Wang D, Dai H (2001) DNA interaction with carbon nanotubes.
Noncovalent sidewall functionalization of Semicond Phys Quantum Electron
single-walled carbon nanotubes for protein Optoelectron 6(1):105–108
immobilization. J Am Chem Soc 123(16): 7. Guo Z, Sadler PJ, Tsang SC (1998)
3838–3839 Immobilization and visualization of DNA and
3. Balavoine F, Schultz P, Richard C, Mallouh Â, proteins on carbon nanotubes. Adv Mater
Ebbesen TW, Mioskowski C (1999) Helical 10(9):701–703
crystallization of proteins on carbon nanotubes: 8. Aitchison TJ, Ginic-Markovic M, Matisons JG,
a first step towards the development of Simon GP, Fredericks PM (2007) Purification,
new biosensors. Angew Chem Int Ed cutting, and sidewall functionalization of mul-
38(13–14):1912–1915 tiwalled carbon nanotubes using potassium
4. Arnold MS, Guler MO, Hersam MC, Stupp SI permanganate solutions. J Phys Chem C
(2005) Encapsulation of carbon nanotubes by 111:2440–2446
self-assembling peptide amphiphiles. Langmuir 9. Yagüe JL, Agulló N, Borrós S (2009) Plasma
ACS J Surfaces Colloids 21(10):4705–4709 polymerization of polypyrrole-like films on
5. Clark MD, Subramanian S, Krishnamoorti R nanostructured surfaces. Plasma Process Polym
(2011) Understanding surfactant aided aqueous 5:433–443
Chapter 21
Lipid-Based Nanoparticles as Nonviral Gene

Delivery Vectors
Daniele Pezzoli, Anna Kajaste-Rudnitski, Roberto Chiesa,
and Gabriele Candiani
Abstract
Efficient delivery of nucleic acids into cells is a promising technique to modulate cellular gene expression for
therapeutic and research applications. Cationic lipid-based liposomes represent one of the most intensively
studied and employed nonviral vectors. They are positively charged at physiological pH and spontaneously
self-assemble with polyanionic nucleic acids forming nanoscaled complexes named lipoplexes. Here, we
draft a simple protocol for the development, characterization, optimization, and screening of liposomal
formulations for in vitro gene delivery. In particular, we report as a practical example a quick method to
formulate and extrude nanometer-sized unilamellar cationic vesicles composed of DOTAP as cationic lipid
and DOPE as zwitterionic helper lipid at 1:1 molar ratio. The physico-chemical characterization of lipo-
somes and lipoplexes involves the measurement of mean diameter and overall surface charge using Dynamic
Light Scattering (DLS) and Laser Doppler Microelectrophoresis. The outlined transfection procedure takes
into account several experimental parameters affecting the in vitro performance of gene delivery systems,
paying special attention to the charge ratio (CR). Gene delivery effectiveness is evaluated both in terms of
transfection efficiency and cytotoxicity of the vector to find the optimal transfection conditions. Importantly,
the proposed protocol can be easily shifted to different types of nonviral vectors.
Key words Gene delivery, Cationic liposomes, Extrusion, Charge ratio, Transfection efficiency,
Cytotoxicity, DLS, Firefly luciferase
1 Introduction
Gene delivery consists in the introduction into host cells of exog-

enous nucleic acids and represents a promising technique for the
modulation of gene expression. The therapeutic application of
gene delivery, named gene therapy, uses nucleic acids as pharma-
ceutical agents and aims to eradicate causes rather than symptoms
of diseases, offering new approaches for the treatment of diseases
that conventional clinical therapeutic strategies cannot cure [1].
Nonviral synthetic gene carriers (transfectants) are considered as
promising vectors for future gene therapy and have been subject of
269
270 Daniele Pezzoli et al.
increasing attention because of their relative safety and simplicity

of use; however, their clinical application is still hindered by their
relatively low efficiency [2, 3]. In addition to gene therapy applica-
tions, nonviral gene delivery vectors are fundamental tools in the
study of gene expression, gene function, protein function, protein–
protein interactions, and cell signaling, and most of the researchers
who employ cell-based techniques exploit gene delivery techniques
in their experiments.
Nonviral gene delivery vectors can be divided in two main
classes: cationic lipids and cationic polymers. They rely on the
basics of supramolecular chemistry named “self-assembling”: at
physiological pH, their positive charges allow the interaction and
spontaneous self-organization with polyanionic nucleic acids, lead-
ing to the formation of nanoscaled complexes named lipoplexes
and polyplexes for lipids and polymers, respectively [4].
Since their introduction as gene delivery systems in 1987 [5],
cationic liposomes have become one of the most intensively stud-
ied nonviral vectors, owing to their tunable chemical structure,
formulation, and dimensions. Along this line, over the last two
decades liposomes containing cationic lipids such as 1,2-dioleoyl-
3-trimethylammonium-propane (DOTAP) [6] in combination
with zwitterionic helper lipids such as 1,2-dioleoyl-sn-glycero-3-
phosphatidylethanolamine (DOPE) [7] have been extensively
exploited for gene delivery purposes.
There are several experimental parameters affecting the in vitro
transfection efficiency of cationic liposomes: the charge ratio (CR)
of lipoplexes, the complexation buffer, the cell type, the cell seeding
density, the dose of lipoplexes administered to cells, the presence of
serum during transfection. Moreover, the optimization of transfec-
tion effectiveness relies on tuning of physicochemical properties of
vector–nucleic acids complexes, such as size and surface charge [8].
Unfortunately, well-defined universal correlations between these
parameters and the transfection effectiveness of liposomes have not
been found yet, making it difficult for liposome developers and
users to select a priori optimal transfection conditions [9].
With this in mind, the aim of this book chapter is to draft a
simple protocol for the development, characterization, optimization
and screening of liposomal formulations for gene delivery (Fig. 1).
In particular, herein we describe an easy and quick method to for-
mulate and extrude nanometer-sized two-component unilamellar
cationic vesicles. Liposomes are characterized in terms of mean
diameter and overall surface charge (ζ-potential) using Dynamic
Light Scattering (DLS) and Laser Doppler Microelectrophoresis
(LDM) techniques. The ability of cationic liposomes to condense
DNA is evaluated by a fluorescence-exclusion assay, using SYBR®
Green I as fluorescent DNA-binding dye. In addition to preparing
and characterizing lipoplexes we present a procedure for the rapid
screening of the best CR in transfection experiments using the
Liposomes for Gene Delivery 271
Fig. 1 Schematic representation of the protocol described herein for the development
of liposomal formulations for gene delivery. A cationic lipid and a zwitterionic
helper lipid (here DOTAP and DOPE, respectively) are formulated in unilamellar
cationic liposomes. Liposomes are mixed with plasmid DNA encoding for the
modified firefly luciferase (pGL3), thus forming liposome–DNA complexes (lipo-
plexes) that are delivered to cells. Transgene expression is finally evaluated by
luciferase expression assay
modified firefly luciferase as easily detectable reporter gene (pGL3-

Control Vector) and alamarBlue® assay to test lipoplex cytotoxicity.
Importantly, the vector/DNA complexes characterization and trans-
fection protocol outlined here can also be used for the screening of
polymeric gene delivery systems [10].
As a practical example, the development of DOTAP:DOPE = 1:1
(molar ratio) liposomes prepared in deionized water and extruded
at 100 nm is reported here (Fig. 1). The transfection procedure
suggests specific cell seeding density, lipoplex dose and transfection
conditions for HeLa model cell line, often used in gene delivery
experiments [9]. However, the proposed protocol can be easily
shifted to other formulations following the advices reported in
Subheading 4.
2 Materials
2.1 Cationic 1. DOTAP chloride (Sigma-Aldrich, St. Louis, MO, USA). Store
Liposome and at −20 °C.
Lipoplex Preparation 2. DOPE (Sigma-Aldrich). Store at −20 °C.
and Characterization
3. Chloroform.
4. Ultrapure water (dH2O) with resistivity values greater than
5 MΩ × cm at 25 °C.
5. Plasmid DNA encoding for the modified firefly luciferase
pGL3-Control Vector (Promega, Madison, WI, USA).
6. SYBR® Green I (Sigma-Aldrich).
7. LiposoFast™ apparatus equipped with two 1.0 ml gas tight
syringes and polycarbonate membranes with pore size of
100 nm (Avestin, Ottawa, Canada; see Note 1).
8. DLS and LDM apparatus: Malvern Zetasizer Nano ZS appara-
tus (Malvern Instruments Ltd, Worcestershire, UK).
9.
Disposable capillary cell for Z-potential measurements
(Malvern Instruments Ltd).
10. Absorbance and fluorescence microplate reader: Tecan GENios
Plus (Tecan, Mannedorf, Switzerland).
2.2 Cell Culture 1. HeLa, human cervix adenocarcinoma cell line (American Type
Culture Collection (ATCC), Manassas, VA, USA).
2.
High-glucose (4.5 g/l) Dulbecco’s Modified Essential
Medium stored at 4 °C.
3. Fetal bovine serum (FBS). Aliquot under sterile conditions and
store at −20 °C.
4. 100× Penicillin–streptomycin solution, stock solution. Aliquot
under sterile conditions and store at −20 °C.
5. 200 mM l-Glutamine, stock solution. Aliquot and store at −20 °C.
6. 1 M HEPES buffer pH 7.0–7.6, sterile solution. Store at 4 °C.
7. Sodium pyruvate 100 mM, sterile solution. Store at 4 °C.
8. Cell Culture Medium: high-glucose DMEM supplemented
with 10 % (v/v) FBS, 2 mM l-glutamine, 100 IU penicillin,
and 100 μg/ml streptomycin, 10 mM HEPES, 1 mM sodium
pyruvate. Store at 4 °C and warm to 37 °C prior to use.
9. Dulbecco’s Phosphate Buffered Saline (PBS). Store at 4 °C
and warm to 37 °C prior to use.
2.3 Transfection 1. Transfection Medium: high-glucose DMEM supplemented

Experiments and with 10 % FBS, 2 mM l-glutamine, 10 mM HEPES, 1 mM
Post-transfection sodium pyruvate (see Note 2). Store at 4 °C and warm to
Assays 37 °C prior to use.
2. 10× AlamarBlue® solution (Life Technologies, Paisley, UK).

3. BCA Protein Assay Kit (Pierce Chemical, Rockford, IL, USA).
4. Luciferase Assay System (Promega).
5. 5× Cell Culture Lysis Reagent (Promega).
6.
Microplate luminometer: Centro LB 960 (Berthold
Technologies, Bad Wildbad, Germany).
3 Methods
3.1 Cationic 1. Dissolve the cationic lipid DOTAP and the zwitterionic helper
Liposomes lipid DOPE in chloroform to a final concentration of 20 mM
Preparation in two separate glass vials (see Note 3).
2. Add equal volumes (e.g., 500 μl) of lipid solutions in a round-
bottomed flask using a glass syringe (see Note 4) and mix well.
3. Remove chloroform on a rotary evaporator at 40 °C until a dry
lipid film can be observed on the flask inner surface.
4. Dry under vacuum overnight to remove all the organic solvent.
5. Add dH2O to a final lipid concentration of 20 mM (see Note 5).
6. Vortex thoroughly to hydrate the lipid film until a clear solu-
tion of large multilamellar vesicles is obtained (see Note 6).
7. Freeze/thaw at least five times and bring to room temperature.
8. Extrude for an odd number of times (at least 21 passages, see
Note 7) the lipid dispersion through two 100 nm-pore poly-
carbonate membranes (see Note 8) using a LiposoFast™ appa-
ratus. Nanometer-sized unilamellar liposomes are obtained.
9. Harvest the liposome dispersion in a sterile plastic vial and
store at 4 °C (see Note 9).
3.2 Liposome 1. Dilute 100 μl of liposome suspension 1:10 in dH2O (see Note 10)
Characterization inside a disposable 12 mm square polystyrene cuvette.
3.2.1 Dynamic Light 2. Measure particle size using a DLS apparatus verifying that the
Scattering (DLS) polydispersity index (PI) of the Cumulant analysis is below 0.2
(see Note 11).
3.2.2 Laser Doppler 1. Move 750 μl of the liposome suspension used for DLS measure-
Microelectrophoresis ment to a disposable capillary cell for ζ-potential measurements.
(LDM) 2. Measure ζ-potential using a LDM apparatus, verifying that
liposomes have a positive ζ-potential (see Note 12).
3.3 Lipoplex 1. Dilute plasmid DNA (in this case, modified firefly luciferase
Preparation and pGL3-Control Vector) in dH2O (see Note 10) to a final con-
Characterization centration of 0.04 μg/μl, corresponding to a phosphate (P)
concentration of 121.2 μM (DNA solution) (see Note 13).
2. Dilute liposome suspension in dH2O so that the final concentration

of net positive charges is 1, 2, 4, 8, 12 times higher than P
concentration in the DNA solution (see Note 14).
3. Mix equal volumes of DNA solution with liposome suspensions
at different concentrations and incubate at room temperature
for 30 min to obtain lipoplex solutions at CR 1, 2, 4, 8, 12.
4. Measure particle size and ζ-potential of the lipoplex solutions
as described above (see Note 15).
3.4 Evaluation 1. Prepare 20 μl of lipoplexes at different CR as described above

of DNA Complexation/ and 20 μl of DNA at 0.02 μg/μl in dH2O (CR = 0).
Condensation Ability 2. Add 100 μl of dH2O (see Note 10), containing 2× SYBR®
Green I and incubate for 10 min at room temperature
(see Note 16).
3. Prepare a blank sample by adding 100 μl of dH2O, containing
2× SYBR® Green I to 20 μl of dH2O.
4. Place 3 × 35 μl aliquots of the resulting solutions in a black
polystyrene 384 well plate (see Note 17).
5. Read fluorescence with an excitation wavelength λex = 495 nm
and an emission wavelength λem = 520 nm using a fluorescence
microplate reader.
6. Plot the samples average fluorescence signal (subtracted of
blank average signal) against CR to find the minimal fluores-
cence plateau, corresponding to maximum DNA condensation
(see Note 18).
3.5 Transfection 1. Seed HeLa cells at a density of 2.0 × 104 cells/cm2 (see Note 19)
Experiments in a 96 well-cell culture plate in 100 μl per well of Cell Culture
Medium (see Note 20).
2. 24 h after seeding, remove old medium.
3. Rinse cells with 100 μl of PBS (see Note 21).
4. Add 100 μl/well of Transfection Medium.
5. Add 6.4 μl/well of lipoplex solutions at different CR, at least
four replicates per CR are suggested (see Note 19).
6. Place cells in a cell culture incubator at 37 °C, 5 % (v/v) CO2
in a humidified atmosphere.
7. Facultative: after 4 h of incubation with cells, remove lipoplex-
containing medium and add FBS-supplemented medium
(see Note 22).
3.6 Cytotoxicity 1. 24–48 h after transfection, remove old medium from cells
Assay (see Note 23).
2. Add 100 μl/well of 1× alamarBlue® in complete medium
(see Note 24).
3. After 1, 2 and 3 h (see Note 25), read fluorescence at

λem = 520 nm, exciting at λex = 540 nm, (see Note 26) using a
fluorescence microplate reader.
4. Subtract the average fluorescence of blank sample (100 μl of
1× alamarBlue® incubated in empty wells) from the fluores-
cence signal of samples.
5. Calculate cell viability as the percentage of samples net fluores-
cence signal with respect to the net signal of controls (non-
transfected cells) and cytotoxicity as:
Cytotoxicity [%] = 100% - Viability [%].
3.7 Cell Lysis and 1. Remove alamarBlue® solution from cells.

Protein Quantitation 2. Rinse cells with 100 μl of PBS (see Note 27).
3. Add 110 μl of 0.25× solution of Cell Culture Lysis Reagent
(5× solution of Cell Culture Lysis Reagent diluted 1:19 in
dH2O; see Note 28) and mix by pipetting (see Note 29).
4. Facultative: freeze-thaw samples to enhance cell lysis
(see Note 30).
5. Prepare a set of bovine serum albumin (BSA) standards by
diluting BSA in 0.25× Cell Culture Lysis Reagent; BSA con-
centration range should be between 0 (blank standard) and
2 mg/ml (see Note 31).
6. Prepare BCA Working Reagent (WR) by adding 1 part of
reagent B to 50 parts of reagent A (see Note 32).
7. Pipette 20 μl of each standard or unknown sample replicate
into a clear polystyrene 96 well microplate.
8. Add 200 μl of WR to each well and place the plate on an orbital
shaker to thoroughly mix the contents of the wells.
9. Incubate the microplate at 37 °C for 2 h (see Note 33).
10. Cool the plate to room temperature.
11. Measure the Optical Density at 560 nm (OD560) using an
absorbance microplate reader.
12. Subtract the average OD560 of the blank standard replicates
from the OD560 of all other standard and unknown sample
replicates.
13. Plot the average Blank-corrected OD560 for each BSA standard
against its concentration and fit a standard curve.
14. Calculate the protein concentration of each unknown sample
using the standard curve.
3.8 Luciferase 1. Place 20 μl of cell lysate in an opaque white polystyrene 96 well
Expression Assay plate (see Note 34).
2. Read luminescence signal by adding 100 μl/well of Luciferase
Assay Substrate using a microplate luminometer.
3. Normalize the luminescence value of each sample with the cor-
responding protein concentration to obtain luminescence
expression values expressed as Relative Luminescence Units
(RLU)/mg of protein.
3.9 Choice of the 1. The choice of the best CR for transfection experiments should
Best Transfection be done taking into account both cytotoxicity and luciferase
Conditions expression results (see Note 35).
2. Among low cytotoxic conditions, choose the one leading to
the highest luciferase expression (see Note 36).
4 Notes
1. LiposoFast™ is a low cost, hand driven extrusion apparatus with

low capacity suggested for researchers who use only small amounts
of liposomes. Vesicles prepared with LiposoFast™ are comparable
to those obtained with large volume commercial apparatus that
may need gas in pressure to work [11]. The apparatus can be
autoclaved, thus allowing to obtain sterile liposomes.
2. It is recommended to carry out transfection experiments in
antibiotics-free medium as increase in cell membrane permea-
bility by cationic liposomes during transfection could lead to
higher antibiotic uptake causing toxicity.
3. Use glass or stainless steel in presence of organic solutions; do
not use vials made of polymeric materials such as polystyrene
or polyethylene as impurities could leach out of the container.
4. The preparation of an equimolar formulation of cationic lipid–
helper lipid is reported here as example but desired molar
ratios can be easily obtained by simply changing volume ratio
of the starting lipid solutions.
5. The aqueous solution in which liposomes are prepared can
influence their final physico-chemical and transfection proper-
ties; in addition to dH2O, other buffers such as 10 mM Hepes
buffer pH 7 and PBS can be used. Do not use buffers contain-
ing ethylenediaminetetraacetic acid (EDTA) as it could cause
liposome aggregation.
6. If detaching the lipid film from the glass surface by vortexing
results difficult, bath sonicate the lipid dispersion for 2–5 min,
until clarity is obtained.
7. An odd number of passages is needed to find extruded lipo-
somes in the originally empty syringe avoiding contamination
with any unextruded large vesicles which might not have
passed through the filter.
8. Using two stacked membranes helps yielding monodisperse,

nanometric-sized liposomes. Liposome dimension can be tuned
by simply using polycarbonate membranes with different pore
sizes (e.g., 50, 100, 200, 400 nm; available from Avestin).
9. Usually liposomes are stable at 4 °C up to 6–12 months; how-
ever, liposomes stability can be checked periodically by mean
size (hydrodynamic diameter) and ζ-potential measurements
(see Subheading 3.2).
10. Liposome formulations should be diluted in the same aqueous
solution they have been prepared for both characterization
experiments and lipoplex formation.
11. A high PI indicates a high width of the size distribution. If PI
is higher than 0.2 or mean size is much higher than membrane
pore size, it may be necessary to extrude again liposomes. In
these cases, reducing lipid concentration during extrusion
could be useful.
12. Cationic liposomes should have a positive ζ-potential in dH2O.
13. DNA phosphate density is assumed to be 3.03 nmol P/μg DNA.
14. DOTAP is assumed to carry one cationic charge/molecule at
pH 7.
15. When using cationic liposomes as transfection reagents, posi-
tively charged liposome–DNA complexes are necessary to
assure DNA complexation and lipoplex interaction with nega-
tively charged cell surfaces.
16. SYBR® Green I stain has been shown to be much less muta-
genic and much more sensitive than ethidium bromide; there-
fore, it is considered as the DNA stain of choice for these
experiments.
17. The use of black polystyrene well microplates is suggested for
fluorescence analysis as they have low background fluorescence
and minimize light scattering and well-to-well crosstalk.
18. Maximum DNA condensation is usually necessary to obtain
efficient lipoplexes, if a plateau is not observed, try with higher
CR ratios.
19. Cell type can strongly influence the transfection effectiveness. If
high cytotoxicity or poor transfection results are obtained, please
find out optimal cell seeding density and/or lipoplex dose.
20. When transfecting other cell lines or primary cells use appropriate
culture medium, according to existing literature (see Note 2).
Since serum may affect the transfection efficiency of gene deliv-
ery vectors, it is sometimes preferable to carry out transfection
experiments in serum-free transfection medium.
21. A PBS washing step is often used in transfection experiments;
however, it can be avoided to reduce cell detachment if culture
medium is the same before and during transfection.
22. This step can be used to reduce cytotoxicity and should be

done if lipoplexes are incubated with cells in serum-free
medium.
23. Luciferase expression usually peaks 24–48 h after transfection
but longer incubation times could be necessary for some lipo-
somal reagents to optimize protein expression.
24. AlamarBlue® is a nontoxic nondestructive fluorimetric cell
growth indicator based on the detection of cellular metabolic
activity. Using alamarBlue® as cell viability assay allows to
quantify total protein content in the same sample.
25. Reading at different time points allows avoiding problems con-
nected to low sensitivity (at early time points of incubation) or
to signal saturation (at late time points of incubation).
26. If cells will not be cultured anymore, it is not necessary to
move conditioned alamarBlue® solution in another microplate
but fluorescence can be read directly in the cell-containing cul-
ture plate, with negligible errors.
27. A PBS washing step is necessary to completely remove serum
proteins.
28. Cell Culture Lysis Reagent is supplied as a 5× solution. Reducing
the working concentration of Cell Culture Lysis Reagent from
the suggested 1× to 0.25× allows to lower lysis buffer interfer-
ence in the BCA assay maintaining high cell lysis ability.
29. After lysis reagent addition, samples can be stored at −80 °C.
30. Freeze-thaw step can be adopted to increase cell lysis when,
after lysis buffer addition, adherent cells are still observed.
31. BSA dilution scheme for standard microplate procedure sug-
gested in the datasheet of BCA assay kit can be followed.
32. The volume of WR to be prepared can be calculated using the
following formula:
V WR = (N standards + N samples ) × (N replicates ) × 200 µl,

where 200 μl is the volume of reagent per sample if the test is
performed in a 96 well microplate.
33. Given the interference of 0.25× Cell Culture Lysis Reagent
with the BCA assay, a 2 h incubation time is needed to enhance
signal to noise ratio.
34. Opaque white microplates should be used for luciferase assays
as they maximize luminescence signal.
35. Conditions leading to high cytotoxicity (cytotoxicity > 40 %)
are considered unuseful.
36. In general, conditions leading to a good compromise between
high transfection efficiency and low cytotoxicity should be
chosen.
Acknowledgments
This work was partly supported by the Politecnico di Milano,

5xmille Junior, “SURGES” Project and by the Italian Ministry for
Education, University and Research (MIUR)—FIRB Futuro in
Ricerca 2008, Grant RBFR08XH0H (to Dr. Candiani, both), as
well as by the Italian Ministry of Health Grant Giovani Ricercatori
2009, GR-2009-1471693 (to Dr. Kajaste-Rudnitski).
References
1. Candiani G, Pezzoli D, Cabras M, Ristori S, DOPE lipoplexes for gene delivery: zeta
Pellegrini C, Kajaste-Rudnitski A, Vicenzi E, potential measurements and electron spin res-
Sala C, Zanda M (2008) A dimerizable cationic onance spectra. Bba-Biomembranes 1664:
lipid with potential for gene delivery. J Gene 70–79
Med 10:637–645 7. Hui SW, Langner M, Zhao YL, Ross P, Hurley
2. Candiani G, Frigerio M, Viani F, Verpelli C, E, Chan K (1996) The role of helper lipids in
Sala C, Chiamenti L, Zaffaroni N, Folini M, cationic liposome-mediated gene transfer.
Sani M, Panzeri W, Zanda M (2007) Biophys J 71:590–599
Dimerizable redox-sensitive triazine-based cat- 8. Ristori S, Ciani L, Candiani G, Battistini C,
ionic lipids for in vitro gene delivery. Frati A, Grillo I, In M (2012) Complexing a
ChemMedChem 2:292–296 small interfering RNA with divalent cationic
3. Pezzoli D, Candiani G, Cabras S, Giordano C, surfactants. Soft Matter 8:749–756
Daniele F, Cigada A (2008) Liposomes versus 9. van Gaal EVB, van Eijk R, Oosting RS, Kok
micelles as gene delivery vectors. Biomed RJ, Hennink WE, Crommelin DJA,
Pharmacother 62:493 Mastrobattista E (2011) How to screen non-
4. Candiani G, Pezzoli D, Ciani L, Chiesa R, viral gene delivery systems in vitro? J Control
Ristori S (2010) Bioreducible liposomes for Release 154:218–232
gene delivery: from the formulation to the 10. Pezzoli D, Olimpieri F, Malloggi C, Bertini S,
mechanism of action. PLoS One 5:e13430 Volonterio A, Candiani G (2012) Chitosan-
5. Felgner PL, Gadek TR, Holm M, Roman R, graft-branched polyethylenimine copolymers:
Chan HW, Wenz M, Northrop JP, Ringold influence of degree of grafting on transfection
GM, Danielsen M (1987) Lipofection –- a behavior. PLoS One 7:e34711
highly efficient, lipid-mediated DNA- 11. Macdonald RC, Macdonald RI, Menco BPM,
transfection procedure. Proc Natl Acad Sci U S Takeshita K, Subbarao NK, Hu LR (1991)
A 84:7413–7417 Small-volume extrusion apparatus for prepara-
6. Ciani L, Ristori S, Calamai L, Martini G tion of large unilamellar vesicles. Biochim
(2004) DOTAP/DOPE and DC-Chol/ Biophys Acta 1061:297–303
Chapter 22
Stabilizing Gold Nanoparticle Bioconjugates

in Physiological Conditions by PEGylation
Joan Comenge and Víctor F. Puntes
Abstract
Stability of gold nanoparticles (AuNPs) is often compromised in physiological conditions. The loss of colloidal
stability might lead to undesired biological responses for drug delivery nanosystems. Here a methodology
to confer additional stability to the AuNPs by the addition of PEG is presented. Also, protocols to prepare
and characterize the composition and conformation of mixed layers with PEG and alkanethiols are
described here. Finally, methods to assay the stability of the link between the NP conjugates and a model
drug are shown.
Key words Gold nanoparticles, Conjugates, Biological stability, PEG, Mixed layers, Protein corona,
Link stability, Cisplatin
1 Introduction
Gold nanoparticles (AuNPs) present unique physicochemical

properties to be used as drug delivery scaffolds such as biocompat-
ibility, easy surface functionalization, excellent size and morphol-
ogy control, and a surface plasmon resonance (SPR) in the range
of visible and infrared light [1, 2]. Drugs, proteins, nucleotides
and organic molecules can be attached to AuNPs which act then as
nanocarriers. However, the stability of these systems might be
compromised when they are assayed in biological media [3].
Biological fluids are a complex mixture of salts and proteins which
may lead to different phenomena such as the aggregation of NPs,
normally due to the compacting of the stem layer of the conjugates
induced by the high ionic strength of this media [4]. Avoiding
aggregation of the nanocarriers is a key point for not only in vitro
applications but also in vivo applications, because size strongly
influences the way the organisms deal with nanosystems. For
example, aggregation leads to a modification of the biodistribution
and a rapid clearance of the nanoparticles by the immune system [5].
In addition, the uncontrolled release of molecules (e.g., drugs)
281
282 Joan Comenge and Víctor F. Puntes
from the nanocarrier before reaching the target might be favored

by nucleophilic attack from species present in biological media [3].
This induced unstability of the nanoparticles and/or link in bio-
logical media would explain the in vivo lack of benefits of some
nanocarriers which worked properly in preliminary in vitro assays.
For that reason, it is important to know methodologies to sta-
bilize the nanoparticles and to evaluate the stability of the nanopar-
ticles in biological media before the final assays. We present here a
stabilization process of AuNP conjugates carrying an antitumoral
drug by means of pegylation. Although we focused the work in the
stabilization of 11-mercaptoundecanoic acid (MUA)-capped
AuNPs, these strategies can be used with other conjugates (i.e.,
other capping agents and other NPs) since they are not element
specific. Moreover, protocols to test the stability of the nanosystem
in biological environment are also provided. Finally the stability of
the link between drug and nanoparticle is assayed with two model
bonds: a strong coordination bond and electrostatic adsorption of
cisplatin.
2 Materials
1. AuNPs. 14 nm AuNPs (4.7 × 1012 NP/mL) are synthesized by

a seeding growth approach: Au seeds (8 nm, 4.7 × 1012 NP/
mL) were synthesized by adding an aqueous solution of
HAuCl4 (1 mL, 25 mM) to a boiling sodium citrate solution
(150 mL, 2.2 mM). The reaction is completed after 10 min.
The temperature was then decreased to 90 °C, HAuCl4 (1 mL,
25 mM) was added to the previously synthesized AuNP seeds,
and the reaction allowed to run for 20 min. This step is repeated
two more times in order to get the final AuNP size.
2. SH-PEG and SH-PEG-FITC (3.4 kDa). Prepare 0.3 and
1 mM solutions of SH-PEG by dissolving 3.40 mg and
1.02 mg, respectively, in 1 mL of milliQ water.
3. MUA. Prepare 0.3 and 1 mM solutions of MUA by dissolving
0.69 and 2.3 mg of MUA, respectively, in 9.925 mL of milliQ
water. Add 75 μL of 2 M NaOH and sonicate until total
dissolution.
4. CAPS buffer. To prepare a 0.2 M solution, 4.43 g are dissolved
in 90 mL of milliQ water and the pH is brought to 10.6 by
adding 75 μL of 2 M NaOH. Finally, milliQ water is added up
to 100 mL.
5. NaCN. To prepare a 0.1 M solution, 49 mg of NaCN are dis-
solved in 10 mL of milliQ water.
6. Bovine serum albumin (BSA). 86 mg are dissolved in 1 mL of
milliQ water.
PEG Stabilization of Gold Nanoparticle Bioconjugates 283
7. AuNP-cisplatin. AuNPs were firstly functionalized with MUA

and later with commercial cisplatin ([Pt(Cl)2(NH3)2]) or aquated
cisplatin ([Pt(H2O)2(NH3)2]2+) as described in the literature [6].
3 Methods
3.1 Direct PEGylation 1. Add 125 μL of an aqueous solution of 3.4 kDa SH-PEG
of AuNPs (See Note 1) (0.3 mM) to 1 mL of citrate-stabilized AuNPs (see Note 2) in
a 1.5 mL plastic tube.
2. The conjugation was allowed to run for 6 h at room
temperature.
3. 1 mL of the solution is transferred to a 12 mm square polysty-
rene cuvette and the hydrodynamic diameter of the NPs mea-
sured by dynamic light scattering (DLS). Adjust the settings
with refractive index = 0.200 and absorbance = 1.32. Perform
the measure in triplicate and at least 12 runs per measurement.
4. Repeat step 3 with citrate-capped AuNPs.
5. The hydrodynamic diameter after conjugating 3.4 kDa
SH-PEG increases about 12 nm with respect to citrate-capped
AuNPs (Fig. 1) (see Note 3).
3.2 Mixed Layers 1. Choose the appropriate mixture/s of SH-PEG and MUA
of SH-PEG and depending on the coverage density of your interest. Always
Alkanethiols (See the total concentration must be kept constant (i.e.,
Note 4): Co-conjugation [SH-PEG] + [MUA] = 1 mM).
of SH-PEG and MUA 2. Add 125 μL of this mixture to 1 mL of citrate-stabilized AuNPs
(See Note 5) in a 1.5 mL plastic tube.
3. The conjugation was allowed to run for 6 h at room
temperature.
Fig. 1 Hydrodynamic diameter as measured by DLS. The hydrodynamic diameter

increased from 13.1 ± 0.3 to 24.5 ± 0.4 nm after SH-PEG conjugation
Fig. 2 Loading of SH-PEG in function of the ratio SH-PEG/MUA added. The loading on the AuNPs was quantified
with two different methodologies (absorbance and fluorescence after digestion of AuNPs) giving similar results.
Note that in the case of the fluorescence-based measures, the lower ratios could not be measured due to the
impossibility of digesting completely the AuNPs, likely due to the compactness of the MUA shell
3.3 Quantification of 1. Repeat steps 1–3 in Subheading 3.2 but using SH-PEG-FITC
SH-PEG on the Surface instead of SH-PEG.
of the AuNPs by UV– 2. Remove the AuNPs by two centrifugation steps (35,000 × g,
VIS Spectroscopy 30 min).
3. The absence of AuNPs is confirmed by the disappearance of
the typical SPR band of AuNPs at around 520 nm.
4. 125 μL of 200 mM CAPS buffer (pH 10.6) were added to
875 μL of the resulting supernatant.
5. To prepare the standards, known concentrations of SH-PEG-
FITC are dissolved in 2.2 mM sodium citrate solution. 125 μL
of 200 mM CAPS buffer (pH 10.6) are added to 875 μL of
every standard solution.
6. The standards and samples are analyzed by UV–VIS spectros-
copy. The absorbance at 495 nm is recorded.
7. The concentration of SH-PEG-FITC found in the supernatant
is subtracted from the known concentration of SH-PEG-FITC
added to know the amount loaded on the AuNPs (Fig. 2)
(see Note 6).
Fig. 3 Conformational change of SH-PEG. The stretching of the SH-PEG mole-

cules represents an increase on the size of the AuNPs which can be followed by
measuring the hydrodynamic diameter by DLS. A similar trend was observed by
analyzing the shift of the SPR band
3.4 Quantification 1. Repeat steps 1–3 in Subheading 3.3 but using SH-PEG-FITC
of SH-PEG on the instead of SH-PEG.
Surface of the AuNPs 2. The corresponding pellets are digested using 1 mL of a
by Fluorescence 100 mM NaCN aqueous solution and stirred at 37 °C for 6 h
(500 rpm, thermomixer compact®).
3. The total digestion of AuNPs is again proved by the disappear-
ance of the SPR band.
4. 125 μL of 200 mM CAPS buffer (pH 10.6) were added to
875 μL of the corresponding samples.
5. Standards are prepared using same conditions than samples.
6. 100 μL of samples and standards were transferred to a 96-well
opaque plate. The fluorescence spectra (λex=494, λem = 521) is
taken. The concentration of SH-PEG-FITC in the samples is
calculated from the calibration curve (Fig. 2) (see Note 6).
3.5 Assay the 1. Repeat steps 1–3 in Subheading 3.2.

Change in SH-PEG 2. Measure the hydrodynamic diameter of the different conju-
Conformation gates as explained in step 3, Subheading 3.1.
3. Measure the SPR band by UV–VIS spectroscopy. Measure
absorption at wavelengths from 300 to 800 nm.
4. The change from mushroom configuration to brush configu-
ration (see Note 7) is identified by a great change in the hydro-
dynamic diameter as well as a larger shift of the SPR band
(Fig. 3). Note that the changes in composition of the layer
follow a linear trend (Fig. 2) and therefore the great changes
of hydrodynamic diameter cannot be due to great changes in
the layer composition.
Fig. 4 Inhibition of protein adsorption. The formation of the protein corona results
in an increase of the hydrodynamic diameter. The change on the SH-PEG confor-
mation leads to minimize and ultimately to inhibit the adsorption of protein which
is seen by no differences of the hydrodynamic diameter after addition of BSA
3.6 Assay the 1. Repeat steps 1–3 in Subheading 3.2.

Protein Adsorption 2. Remove the excess of ligands by centrifugation (18,000 × g,
on the Mixed 14 min) and resuspend the pellet in freshwater (see Note 8).
Layers by DLS
3. Add 8 μL of BSA 1.3 mM to 1 mL of the different AuNP solu-
tions. Stir gently overnight (200 rpm, thermomixer®).
4. Measure the hydrodynamic diameter of AuNPs with and with-
out BSA as explained in step 3, Subheading 3.1. The differ-
ences in the hydrodynamic diameter are indicative of protein
adsorption [7]. An example is seen in Fig. 4. At lower ratios
the difference in size is around 8 nm due to the absorption of
a monolayer of BSA. This difference becomes smaller for
higher SH-PEG/MUA ratios (less BSA is adsorbed) and finally
at ratios >0.78 there is no difference indicating the total inhibi-
tion of BSA adsorption. Note that the decrease of protein
adsorption is strongly related with the conformational change
of SH-PEG.
3.7 Stability 1. 1 mL AuNPs loaded with commercial cisplatin (AuNP-

of the Link in cisplatin) are dialyzed against 1 L of water. Independently, the
Physiological Media same procedure is followed with 1 mL of AuNPs loaded with
aquated cisplatin (AuNP-aq.cisplatin) (see Note 9).
2. Mix 1 mL of both AuNPs (4.7 × 1012 NP/mL) with 9 mL of
DMEM supplemented with 10 % FBS (v/v) (see Note 10). Stir
gently at room temperature.
3. Take 1 mL of sample at the desired times (e.g., 0.5, 6, 24, and
48 h). Immediately, remove the AuNPs by two centrifugation
Fig. 5 Release of cisplatin depending on the strength of the bond. In the case of
coordination bond, the unspecific release is minimal (ca. 15 % after 50 h), whilst
a burst release up to 42 % in the first hour followed by a continuous unspecific
release up to 62 % is produced when cisplatin was weakly linked to the AuNPs
steps (35,000 × g, 30 min). The amount of Pt in the resultant

supernatant is analyzed by inductively coupled plasma mass
spectroscopy (ICP-MS). The amount of Pt loaded in the
AuNPs is also analyzed by ICP-MS (see Note 11).
4. The release is calculated dividing the amount of Pt in the super-
natant by the amount of Pt loaded on the corresponding AuNPs
(Fig. 5). Unspecific release is identified by a high release at very
short times. This is normally undesired. On the other hand, a
strong bond will lead to minimal release unless a triggering stim-
ulus is produced. In this example the trigger is a pH decrease.
4 Notes
1. Direct PEGylation is used when the initial material is bare

nanoparticles (e.g., citrate-stabilized) and there is no need to
have additional ligands. This is commonly used to assay general
properties of AuNPs such as biodistribution and cell uptake [8].
However, it is not recommended for further loading of drugs or
other molecules of interest on the AuNPs because PEG tends to
nonspecifically and weakly adsorb small molecules. Moreover,
the shell is not as compact as using other smaller ligands (i.e.,
alkanethiols). Therefore, the number of ligands per nanoparticle
will be lower and consequently the number of available chemical
groups to bind a drug is lower than in the compact layers
achieved when using alkanethiolates.
2. The amounts of reagent added are calculated for 14 nm AuNPs
at 4.7 × 1012 NP/mL. The same protocol can be followed for
other sizes and concentrations maintaining the proportions

and adjusting the amounts to the total Au surface.
3. Obviously, the increase of the diameter will be different if an
SH-PEG with a different molecular weight is used.
4. Mixed layers are interesting whenever the functionality given
by the alkanethiol monolayer will be needed in further steps
(e.g., linking a drug) and the PEG molecules to confer stability
to the system. There are two options to achieve these mixed
layers: co-conjugation of both components or displacement of
the SH-PEG by the alkanethiol (despite that both are linked by
an SH group to the Au surface, the high packing in the self-
assembled monolayers (SAM) confers an additional stability
making the bond stronger and allowing SH-PEG displace-
ment). This section focuses on the first option since it leads to
a better control and more reproducibility according to our
experience. SH-PEG (3.4 kDa) and MUA have been chosen as
model molecules to explain these processes.
5. SH-PEG and MUA have different affinities for the AuNP sur-
face. The affinity of the molecule for the NP surface takes into
account not only the chemical bond but also the stability of the
formed monolayer. Therefore the proportion added will not be
the proportion of the different species on the surface. This
statement is valid for every co-conjugation of two different spe-
cies. The protocol and characterization techniques used here
can be also applied to other alkanethiol/SH-PEG mixtures.
6. Knowing the amount of SH-PEG loaded on the NP, one can
know the percentage of the surface occupied by this ligand
simply by knowing the amount of SH-PEG when the surface is
saturated. For example, for the ratio SH-PEG-FITC/
MUA = 0.5, the amount of SH-PEG-FITC loaded is ca 0.9 μM
which would correspond to the 39 % of the surface (i.e.,
[0.9/2.3] × 100). Thus, the proportions of SH-PEG/MUA
added in solution are not maintained on the NP surface. This
is explained by the different affinity of the molecules to the
AuNPs: MUA is a smaller molecule than SH-PEG, with a
hydrophobic chain and a hydrophilic head that favors the for-
mation of dense SAM on AuNPs [9] more than SH-PEG does.
7. It is well known that SH-PEG can adopt two conformations.
A folded conformation, known as mushroom conformation
[10], is adopted when the density of SH-PEG is low. On the
other hand, a more compact layer of SH-PEG leads to a
stretching of the molecules in what is known as brush confor-
mation [10]. In the low density case, the SH-PEG polymer
attaches the NPs via multiple weak bonds while in the high
density, the surface is saturated with the more stable SH–Au
bonds and therefore, there is no room for the polymer to wrap
on the surface.
8. Note that the centrifugation conditions here are softer than in

the previous sections. This is because we need these AuNPs for
further steps, whilst in the other cases a potential aggregation
of the AuNPs in the centrifugation step would not represent a
problem. Thus, in this section it is more important to preserve
the colloidal stability of the AuNPs, whilst in the previous sec-
tions it is more important to ensure that there are no AuNPs in
the supernatant.
9. The reason to use these cisplatin derivatives is the different
bond that they form with MUA-capped AuNPs. Whilst aquated
cisplatin forms a strong coordination bond with carboxylic
groups of MUA, commercial cisplatin binds the AuNPs via an
ion–dipole interaction [6].
10. Physiological media are complex and reactive media. They
contain proteins, amino acids, carbohydrates, and high con-
centration of salts, among other components. This can affect
the chemical stability of the nanoparticle with the molecule of
interest and therefore has to be studied before any possible
biological application. The stability of the NPs must be assessed
in the same media than the final application.
11. An easier experiment of dialysis and analysis of the outer solu-
tion (cell media) could have been performed if the free drug
had not interacted with protein of the media which would have
retained the drug in the dialysis bag [11].
References
1. Daniel MC, Astruc D (2004) Gold nanoparti- 7. Casals E, Pfaller T, Duschl A, Oostingh GJ,
cles: assembly, supramolecular chemistry, Puntes V (2010) Time evolution of the
quantum-size-related properties, and applica- nanoparticle protein corona. ACS Nano
tions toward biology, catalysis, and nanotech- 4:3623–3632
nology. Chem Rev 104:293–346 8. Lipka J, Semmler-Behnke M, Sperling RA,
2. Rana S, Bajaj A, Mout R, Rotello VM (2012) Wenk A, Takenaka S et al (2010) Biodistribution
Monolayer coated gold nanoparticles for delivery of PEG-modified gold nanoparticles following
applications. Adv Drug Deliv Rev 64:200–216 intratracheal instillation and intravenous injec-
3. Ruenraroengsak P, Cook JM, Florence AT tion. Biomaterials 31:6574–6581
(2010) Nanosystem drug targeting: facing up 9. Love JC, Estroff LA, Kriebel JK, Nuzzo RG,
to complex realities. J Control Release 141: Whitesides GM (2005) Self-assembled
265–276 monolayers of thiolates on metals as a form
4. Lim IS, Goroleski F, Mott D, Kariuki N, Ip W of nanotechnology. Chem Rev 105:
et al (2006) Adsorption of cyanine dyes on 1103–1170
gold nanoparticles and formation of 10. Simpson CA, Agrawal AC, Balinski A,
J-aggregates in the nanoparticle assembly. Harkness KM, Cliffel DE (2011) Short-chain
J Phys Chem B 110:6673–6682 PEG mixed monolayer protected gold clusters
5. Dobrovolskaia MA, McNeil SE (2007) increase clearance and red blood cell counts.
Immunological properties of engineered ACS Nano 5:3577–3584
nanomaterials. Nat Nanotechnol 2:469–478 11. Dhar S, Gu FX, Langer R, Farokhzad OC,
6. Comenge J, Romero FM, Sotelo C, Lippard SJ (2008) Targeted delivery of cis-
Dominguez F, Puntes V (2010) Exploring the platin to prostate cancer cells by aptamer
binding of Pt drugs to gold nanoparticles for functionalized Pt(IV) prodrug-PLGA-PEG
controlled passive release of cisplatin. J Control nanoparticles. Proc Natl Acad Sci 105:
Release 148:e31–32 17356–17361
INDEX
A Conjugates ......................................................... 9–17, 19, 20,

24–26, 48, 65, 183, 186–188, 194, 208, 225, 226, 238,
Adsorption 258, 281, 282, 285
competition.........................................................157–165 Conjugation
kinetics................................................................157–165 DNA......................................................................... 9–17
Agglomeration.......................................... 149, 168, 169, 172 Peptide .......................................................................... 65
Aggregation ......................................................14, 16, 26, 81, protein .................................................................... 19–26
90, 119–125, 148, 175, 234, 238, 245, Core-shell ....................................................... 29, 48, 67, 152
276, 281, 289 Cryosectioning ......................................................... 179–197
Amphiphilic ligands ......................................... 120–123, 252 CVD, plasma enhanced .................................................... 265
Au NPs. See Gold nanoparticles (Au NPs) Cytotoxicity ...........................................3, 120, 271, 274–278
B
D
Biocompatibility .................................................... 3, 4, 9, 19,
Differential centrifugal sedimentation
20, 48, 76, 119, 225, 226, 261, 262, 281
(DCS) ............................140–142, 145–146, 149–152
Bioconjugation ..................................................... 48, 76, 239
DLS. See Dynamic light scattering (DLS)
Biological fluids ........................................138, 139, 142–144,
DNA
146, 148, 149, 151, 152, 281
conjugation ............................................................... 9–17
Biological stability .............................119–125, 169, 281, 282
coverage ratio .......................................................... 10–12
Biosensor .................................................... 76, 109–114, 157
melting ............................................... 127–135, 218, 219
Block co-polymers .................................................... 207–223
NP ratio ...................................................................... 256
C staining ................................................11, 13, 15, 17, 248
Dynamic light scattering (DLS).............................. 120, 128,
Cap exchange .................................. 30, 32–34, 37, 41, 43, 44 140, 142, 145, 149–151, 168, 215, 218, 268, 270, 272,
Carbodiimide coupling ....................................................... 30 273, 283, 285, 286
Carbon nanoparticles........................................................ 176
Cationic liposome ..................................... 270–273, 276, 277 E
CD spectroscopy. See Circular dichroism
EDXS. See Energy dispersive x-ray spectroscopy
(CD) spectroscopy
chemical analysis (EDXS)
CDSSTR............................................................................ 25
Electrochemical measurements ................................ 127–135
Cell culture media ....................................145, 146, 169–171,
Electrochemistry .............................................. 128, 130, 131
232, 272, 274
Electron tomography ................................................ 179–197
Cellular interaction ........................................................... 176
Encapsulation ............................................................... 30, 76
Centrifugation ................................................. 10, 14–17, 42,
Energy dispersive x-ray spectroscopy chemical
44, 59, 80, 82, 84, 86, 122, 123, 125, 144–145,
analysis (EDXS) ...........................................179–197
149, 152, 161, 162, 168, 217, 255, 259, 284,
Extrusion .................................................................. 276, 277
286, 289
Centrifugation columns ...................................................... 20
F
Charge ratio.............................................. 270, 274, 276, 277
Chemoselective ligation................................................ 47–71 Firefly luciferase........................................................ 271–273
Circular dichroism (CD) spectroscopy ......................... 21, 25 Fluorescence .....................................................10, 13, 14, 17,
Cisplatin .................................................................. 282, 283, 48, 71, 128, 210, 219, 220, 223, 252, 258, 259, 263,
286, 287, 289 270, 272, 274, 275, 277, 278, 284, 285
Colloidal stability ................................................ 76, 84, 119, Functionalization ..............................................3, 4, 6, 29–44,
120, 125, 201, 289 76, 86, 104, 105, 109–114, 208, 261, 281
Methods in Molecular Biology, vol. 1025, DOI 10.1007/978-1-62703-462-3, © Springer Science+Business Media New York 2013
291
NANOMATERIAL INTERFACES IN BIOLOGY: METHODS AND PROTOCOLS
292 Index
G Nanoparticles
gold (Au NPs) ................................................. 3–7, 9, 10,
Gel electrophoresis, agarose ................................... 10, 14, 20, 75–77, 84, 89, 125, 167–176, 183, 186,
21, 23–25, 54, 241, 268 187, 194, 195, 202, 208, 237–248, 251,
Gene delivery ........................................................... 261–279 281–289
Gold nanoparticles (Au NPs) ................................ 3–7, 9, 10, iron oxide ................................................... 179–197, 217,
75–77, 84, 89, 125, 167–176, 183, 186, 187, 194, 195, 225–235
202, 208, 237–248, 251, 281–289 nanoparticle passivation ...................... 120, 239, 243–244
Grafting ........................................................................ 96–98 silica .................................................................... 202–204
silver ................................................... 239, 240, 244–245
H
Nanorods, gold ................................................ 79, 80, 82–84,
HeLa cells ....................................................... 180, 182–185, 88–90, 120–125, 172, 174, 175
187, 190, 194, 195, 274 Nanotubes, carbon .................................................... 261–268
High-performance liquid chromatography NG. See Loaded nanogels (NG)
(HPLC) ............................................... 21–23, 26, 52, Non-fouling surface.............................................................. 3
59, 129–131, 202, 210, 228, 234, 240 Non-specific adsorption .............................. 70, 96, 201–204,
Hydrazone .............................................49–50, 54, 60, 63–71 238, 239
Hyperspectral imaging ............................................. 170, 173 Non-viral gene delivery vector .................................. 269–278
Hyperspectral microscopy......................................... 167–176
O
I Optical properties ...........................................29, 48, 75, 119,
Immunocytochemistry.............................................. 179–197 125, 151, 208, 237, 238, 251
Integrin αvβ3, 226, 232, 234 Organosilane .................................................................... 110
Intracellular delivery ................................................. 251–260
P
Iron oxide nanoparticles .................................. 179–197, 208,
217, 225–235 PASP. See Polyaspartic acid (PASP)
Passivation ....................................................... 120, 125, 129,
L 131, 134, 201–203, 237–248
Laser doppler microelectrophoresis PEG. See Polyethylene glycol (PEG)
(LDM) ......................................... 168, 270, 272, 273 Peptide conjugation ............................................................ 65
Ligand exchange. See also Cap exchange PET imaging. See Positron emission tomography
Light scattering ................................. 140, 151, 169, 218, 277 (PET) imaging
Link stability .................................................... 282, 286–287 Plasma enhanced CVD .................................................... 265
Lipid(s) .....................................................................269–279 Plasmonic nanoparticles ........................................... 119, 120
Lipid based nanoparticle .......................................... 269–278 Polyaspartic acid (PASP) ..........................................225–235
Lipoplexes ................................................ 270, 271, 274, 278 Polyethylene glycol (PEG) .................................... 20, 24, 54,
Liposome .................................................. 270–274, 276, 277 67, 84–86, 90, 91, 130, 134, 172, 251, 252, 256–259,
Loaded nanogels (NG)............................. 181–183, 185, 190 287, 288
Lyophilization ................................... 9, 10, 12, 14, 16, 58, 59 Polyhistidine .................................................... 48–49, 53, 55,
58, 59, 62–63, 68–70
M Polymer/polymeric coating ................................................. 96
Positron emission tomography (PET)
Mixed layers ..................................................... 283, 286, 288 imaging ................................................. 100, 225–235
Molecular recognition ........................................................ 96 Protein
Monolayers ............................................... 3, 48, 96, 109, 111, adsorption .................................................. 139, 157–165,
128, 129, 131–133, 164, 201, 204, 238, 252, 286, 288 168, 169, 201–204, 286
MRI imaging............................................................ 225–235 conjugation ............................................................. 19–26
Multimodal imaging......................................................... 226 corona ......................................................... 137–153, 286
Multi-wall carbon nanotubes ........................................... 262 repellent surface .......................................................... 201
Purification .....................................................4, 5, 10–12, 14,
N
20–25, 32, 35–36, 50, 54, 59, 62–63, 65, 67, 68, 129,
Nanogels, polymeric ................................................. 179–197 131, 160, 209, 212–215, 222, 229, 234, 239,
Nanomechanical biosensor ....................................... 109–114 254–257, 259, 264
NANOMATERIAL INTERFACES IN BIOLOGY: METHODS AND PROTOCOLS
Index
293
Q Surface charge ..................................................9, 15, 81, 120,

139, 140, 168, 270
Quantum dots ...................................29–44, 47–71, 119, 248 Surface modification .........................................75, 76, 82, 83,
95–107, 137, 261, 262, 265, 266
S
Surface plasmon resonance
Scanning transmission electron microscopy (SPR) ............................................ 128, 281, 284, 285
(STEM) ........................................ 183–187, 190–193
Secondary structure, protein ............................................. 127 T
Self-assembled monolayers TEM. See Transmission electron microscopy (TEM)
(SAMs) ........................................................ 109, 132, Thermodynamics .............................................................. 128
201, 238, 288 Tokuyasu cryosectioning technique .......................... 179, 187
Self-assembly ....................................................... 48–49, 208, Toxicity ................................................................ 75, 76, 119,
210–211, 215–218 167–176, 251, 262, 276
Semiconductor nanocrystal................................................. 48 Transfection ..............................................120, 262, 265–267,
Silica 270–274, 276–278
coating .................................................................... 75–91 Transfection efficiency ...................................... 270, 277, 278
nanoparticles....................................................... 202–204 Transmission electron microscopy
Silicon microcantilevers .................................................... 109 (TEM) ................................................5, 6, 78, 80, 81,
Siloxane .................................................................... 201–204 85, 87–89, 91, 125, 126, 142, 147, 152, 168, 171,
Silver nanoparticles .................................. 239, 240, 244–245 173–175, 179–186, 189, 191–192, 211, 212, 217,
Single-molecule spectroscopy ............................................. 48 218, 223
Single-particle spectroscopy ..................................... 237–248 Tumor angiogenesis .......................................................... 226
Solution depletion .................................................... 157–165
SPR. See Surface plasmon resonance (SPR) U
Stability ............................................................3, 4, 9, 10, 29,
Unilamellar liposome........................................................ 273
30, 76, 83, 84, 109, 119–125, 127, 139,
146, 151, 168, 169, 171, 174, 201, 238, 239, Z
248, 259, 277, 281–289
STEM. See Scanning transmission electron Zeta-potential.................................................14, 84, 90, 120,
microscopy (STEM) 139, 140, 142, 146, 168, 272
Stimuli responsive polymeric Zwitterion ................................................... 54, 67, 201–204,
nanoparticles ................................................. 179–197 238, 239, 270, 271, 273

Nanomaterial Interfaces in Biology: Methods and Protocols

Uploaded by

Copyright:

Available Formats

Nanomaterial Interfaces in Biology: Methods and Protocols

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nanomaterial Interfaces in Biology: Methods and Protocols

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 1025

For further volumes:

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Library of Congress Control Number: 2013943590

© Springer Science+Business Media New York 2013

Printed on acid-free paper

Humana Press is a brand of Springer

Brescia, Italy Paolo Bergese

PART I MAKING BIO–NANO INTERFACES

PART II CHARACTERIZING BIO–NANO INTERFACES

13 Hyperspectral Microscopy for Characterization of Gold Nanoparticles

PART III IMPLEMENTING BIO–NANO INTERFACES

W. RUSS ALGAR • Naval Research Laboratory, Washington, DC, USA

ROBERT J. HICKEY • Department of Chemistry, University of Pennsylvania, Philadelphia,

Making Bio–Nano Interfaces

Preparation of 2 nm Gold Nanoparticles for In Vitro

Key words Gold nanoparticle, Biocompatibility, Purification

Gold nanoparticle (AuNP) properties and applications continue to

making these nanoparticles a promising platform for systematic

1. Hydrogen tetrachloroaurate (III) hydrate.

All procedures are done at room temperature unless otherwise

2. Premix 2.1 g (3.8 mmol, 1.5 Eq) of tetra-n-octylammonium

3.3 Functionalization 1. Dissolve 30 mg of the purified pentanethiol-capped AuNPs in

6. If there is no observable precipitation (solution too dark), check

This work was supported by the NIH (GM077173 and EB014277).

DNA Conjugation to Nanoparticles

DNA–nanoparticle conjugates have attracted great interest as there

the water content of the solution dries out. Citrate-stabilized AuNPs

All solutions are prepared with 0.5× tris-borate-EDTA buffer

2.1 Ligand Exchange 1. 0.5× Tris-borate-EDTA buffer (TBE): 45 mM tris, 45 mM

Fig. 1 An overview of the protocol of nanoparticle–DNA (NP–DNA) conjugation

2.2 Conjugation and 1. DNA oligonucleotide: commercially available thiol-capped

2. Put 1 g of NaCl into the AuNP–BPS solution. Stir and mix

3.2 Conjugation 1. Determine a target coverage ratio (average number of conju-

3. Re-disperse the dried sample in one of the tubes by adding ~50 μL

3.3 Determination 1. Dilute the AuNP–DNA stock solution to 0.1 μM in 1 mM MCH

1. 0.5× TBE is made by either diluting commercially available

beaker with a microwave oven. Interrupt the heating process

Conjugation of Nanoparticles to Proteins

The conjugation of nanoparticles (NPs) to proteins is of increasing

In contrast, the formation of a covalent bond with either the NP

All solutions should be prepared with deionized (DI) water.

2.1 Nanoparticle 1. 125 mL conical flask and Teflon stir bar.

3. 250 mg hydrogen tetrachloroaurate (HAuCl4) and 400 mg of

2.2 NP Conjugation 1. PBS 1× or 10 mM NaPi buffer (pH 7.4).

2.4 Structural 1. UV–Vis spectrophotometer.

All procedures should be carried at room temperature unless speci-

3. Weigh 580 mg of triphenylphosphine and add to the solution.

1. (Optional) If the protein attaches to the NP via a surface

3.3 Agarose Gel 1. Prepare a 3 % agarose gel by mixing 3 g of ultrapure agarose

2. Load the NP, protein and NP–protein solutions into successive

gel electrophoresis with different concentrations of agarose

3.4 Structural Circular dichroism (CD) spectroscopy in the far-UV is an effective

1. To make 5× TBE (445 mM Tris Base, 445 mM boric acid,

2. During ligand exchange, the two phases (organic and aqueous)

Water-Solubilization and Functionalization

Key words Quantum dot, Semiconductor nanocrystal, Cap-exchange, Encapsulation, Functionalization,

Semiconductor quantum dots (QDs, or nanocrystals) have gained