Accelerate Oligonucleotide Therapeutics Ebook

S E PT E M B E R 2 0 2 2
Resources and Tools to

Accelerate Oligonucleotide
Therapeutics to Market
Designation of Oligo Profiling Oligo
Oligo Trends
Oligo Starting and Impurity Identification
and Insights
Materials Control and Quantitation
This custom ebook is sponsored by ThermoFisher Scientific,

and presented in partnership with BioPharm International.
Oligo Trends Designation of Oligo Oligo Profiling and Oligo Identification
and Insights Starting Materials Impurity Control and Quantitation
vijay0401/stock.adobe.com
Oligonucleotide Therapeutics:
Trends and Insights
Q&A with Chris Rosenau, Product Manager, Nucleic Acid Therapeutics, Thermo Fisher Scientific
C
hris Rosenau received his BS in
Bacteriology and Zoology from the
University of Wisconsin–Madison. In his
20+ year career, he has held a variety of
R&D positions in molecular biology, ran the Blood
Center of Wisconsin’s Solid Organ Histocompatibility
Transplant Laboratory, and managed cell-free DNA
diagnostic products. Since mid-2020, he has been a
Product Manager for Nucleic Acid Therapeutics at
Thermo Fisher Scientific, responsible for materials
used in development and manufacturing of
oligonucleotide therapeutics and mRNA vaccines,
such as phosphoramidites and modified nucleoside
triphosphates.
SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 2 SPONSORED CONTENT

What are oligonucleotide What are some trends and challenges

therapeutics? in ASO and RNAi therapeutics?
Oligonucleotide therapeutics are short The success of mRNA-based COVID-19
DNA or RNA molecules that modulate the vaccines has brought renewed interest
expression of target RNA for the treatment to other nucleic acid drugs, including
of diseases. These oligos are about 20 ASO and RNAi therapeutics. In the next
nucleotides long, can be single- or double- 5 years, the oligo therapeutics market
stranded, and are traditionally chemically is expected to grow by 25–30%; over
synthesized using phosphoramidites as 1,000 oligonucleotide molecules are
building blocks. This drug family includes currently in the pipeline. Major drugs such
antisense oligonucleotides (ASOs), small as ONPATTRO®, inclisiran, and others
interfering RNAs (siRNAs) involved in have shown that chemically synthesized
RNA interference (RNAi), small activating oligonucleotide drugs have the potential to
RNAs (saRNAs), aptamers, and guide RNAs be best-in-class molecules, helping millions
(gRNAs) involved in CRISPR gene editing of patients around the world with genetic,
technology. As of 2022, there are 14 such cardiometabolic, and other diseases.
drugs currently commercialized worldwide.
Twelve of these are approved in the United
States, the majority of which are ASOs
and siRNAs. They treat diseases including SPONSORED CONTENT
muscular dystrophy, polyneuropathy, and Phosphoramidites for
hypercholesterolemia, with many more oligonucleotide synthesis
therapeutics currently in development and
other diseases being targeted
What are the main differences As both ASOs and siRNAs are short
between ASO and RNAi oligonucleotides, their stability in the body,
technologies? Phosphoramidites for immunogenicity to the host, specificity to
They both act on target RNA, resulting
oligonucleotide in
synthesis their target sequences, and delivery to target
RNA cleavage and ultimately preventing organs have always been challenges.
protein translation. However, their structure
and mechanisms are different. ASOs are As the building blocks for oligonucleotide
single-stranded oligos that recruit RNase H1 therapeutics, modifications to
to cleave their RNA targets in the nucleus phosphoramidite structures have been
or the cytoplasm. On the other hand, RNAi developed to improve performance
involves double-stranded siRNAs that recruit and mitigate these issues. For example,
the RNA-induced silencing complex (RISC) to 2′-O-methyl (2′-OMe) and 2′-fluoro (2′-Fl)
cleave their RNA targets in the cytoplasm. modifications are commonly incorporated

into siRNA oligonucleotides to improve “Thermo Fisher has been

resistance to nucleases and thermal
stability, and reduce off-target effects. providing phosphoramidites
For ASOs, 2′-O-methoxyethyl (2′-MOE)– on a global scale with
modified amidites, among others, can be
incorporated to produce similar effects. leading quality systems and
For delivery, encapsulating oligonucleotide >40 years of nucleic acid
therapeutics in lipid nanoparticles (LNPs) or
conjugating them to N-acetylgalactosamine chemistry innovation.”
(GalNAc) moieties that have been
developed in recent years facilitate meeting their capacity needs. We are
improved durability and specificity of organ committed to growing with our customers
targeting for nucleic acid drugs. as their demand for oligo synthesis
increases. We have recently expanded our
What do customers care about when phosphoramidite capacity, significantly
requesting phosphoramidites for adding to our already industrial-scale
oligonucleotide synthesis? manufacturing suite to both serve our large
Customers look for quality assurance, customers and ensure that we have safety
documentation support, capacity, and batch- stock available for all customers.
to-batch consistency in their products.
Scale-up and future capacity are also
Customers commonly approach us with important when our customers are starting
quality requirements. Providing quality with a custom modification but also looking
products starts with solid manufacturing toward their large-scale manufacturing down
processes housed within a well-established the road. At Thermo Fisher, our technical
quality system, which we have developed Process Development team develops their
over decades in partnership with our small-scale methods with eventual industrial
customers, and the documentation that scale in mind, shortening lead times when
supports it. Thermo Fisher has been our customers require larger volumes. Scale-
providing phosphoramidites on a global up and tech transfer are built in from the
scale with leading quality systems and >40 beginning of our process.
years of nucleic acid chemistry innovation.
Our manufacturing site is certified for ISO Batch-to-batch consistency is very important
9001:2015 for quality assurance, traceability, for our customers’ processes as well. We
and documentation support. have the knowledge and experience to
minimize batch-to-batch variation, starting
Customers also want to make sure their from raw material qualification and incoming
phosphoramidite supplier is capable of raw material testing, ensuring quality is

maintained in manufacturing via in-process

testing, and eventually ensuring every product
meets final product release specifications. SPONSORED CONTENT
Oligonucleotide synthesis
What are your organization’s for therapeutics
capabilities?
A broad range of nucleic acid chemistry
options are available, including both off-
the-shelf and custom phosphoramidites. How does your organization work
Our phosphoramidite offerings include with a customer who is interested in
standard phosphoramidites, modified custom phosphoramidites?
phosphoramidites with 2′ sugar It starts with a discussion between the
modifications (e.g., 2′-OMe, 2′-Fl, and 2′- customer and our business development
MOE) and various protection modifications, team to get a clear understanding of the
dye labels, structural moieties, linkers, and customer’s needs, timeline, budget, quality
spacers. We have also partnered for >40 requirements, potential scale-up, and so
years to co-innovate with our customers, on. Our technical team then reviews details
including custom manufacturing (e.g., involving the technical specifications
GalNAc) and analytical services. of the products, specifics regarding the
processes we will use, whether there
Thermo Scientific™ DNA and RNA are any IP issues, etc. Our sourcing team
phosphoramidites are suitable for reviews the raw material needs for the
oligonucleotide manufacturers for the project and manages the supply chain.
development of therapeutic, diagnostic, and Our project management team oversees
research applications. Thermo Scientific™ the project and communicates with the
TheraPure™ phosphoramidites undergo customer in collaboration with a dedicated
additional quality control release testing internal technical team on inputs, design,
compared to our standard phosphoramidites, planning, milestones, etc. In summary, we
helping ensure that impurities and residual partner with our customers from day one
solvents are controlled to the stringent levels to project completion so that they are
required by our customers for their oligo informed of the project status and progress
therapeutics and diagnostic applications. along the way, delivering the final product
exactly as requested.
Outside of our phosphoramidite team,
Thermo Fisher provides reagents for How is your phosphoramidite team
oligonucleotide synthesis and analysis tools contributing in this area?
for synthesized oligonucleotides, as well as Thermo Fisher provides global scale,
the ability to synthesize multi-gram yields. service excellence, cost efficiency, co-

innovation, and high quality standards in • Co-innovation: We have partnered with

phosphoramidite offerings. our customers for >40 years on co-
innovation and custom development
• Global scale: We can provide raw projects, including supporting the
materials on a global scale with a manufacturing of customized GalNAc
secure supply chain headquartered in phosphoramidites.
the US. We have also expanded our • High quality standards: Our quality
manufacturing capacity in 2020 and attributes include purity profiles,
2021, with 5x additional capacity online impurity profiles, materials-of-origin
in 2022. information, and traceability. We also
• Service excellence: Our customer provide our customers access to
service team strives for a <48-hour quality management and scientific
response to orders and >98% delivery staff for continual regulatory and
to customers by the promised date. We technical support. We have also
have phosphoramidites with standard and initiated multiple sustainability
common modifications available in bulk, programs to further optimize our
allowing for ease of purchase processes and reduce waste.
For Research Use Only. Not for use in diagnostic procedures. © 2022 Thermo Fisher Scientific Inc. All rights reserved.
All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. Onpattro is a
registered trademark of Alnylam Pharmaceuticals, Inc. EXT3694 0822

TheraPure phosphoramidites:
the ongoing pursuit of excellence
Discover the value of Thermo Scientific™ TheraPure™ phosphoramadites
• Designed for oligonucleotide • Manufactured with expertise • Produced in our state-of-the-art
manufacturers requiring highly in protection chemistries, Milwaukee, Wisconsin facility,
defined impurity profiles and modifications, manufacturing with an ISO 9001:2015–certified
strictly controlled processes scale-up, and innovative quality management system
purification strategies
World-class manufacturing Customer-focused technologies

• Expanded manufacturing capacity with • Nucleic acid chemistry and custom
metric-ton capability chemistry expertise for co-innovation and
custom development projects
• Regionally diverse supply chain
• State-of-the-art manufacturing facility
• Phosphoramidites available in bulk packaging
• Sustainability program to optimize
processes and reduce waste
Dedicated customer service

• <48-hour* response to orders
• >98%* delivery to customer at promised date
• Standard synthesizer packaging and custom

packaging options available, ready for
prompt delivery
Continually improved quality Analytical capabilities

system with extensive testing • Analytical method development
• ISO 9001–compliant quality management • Method verification/validation
system refined over decades of partnership with
our customers • Test method transfer
• Impurity profiling, control of critical impurities, and • Stability studies

traceability for materials of origin
• High-quality TheraPure phosphoramidite offerings

* May not be applicable in all regions.
Ordering information
Product Cat. No. Product Cat. No.
TheraPure DNA phosphoramidites TheraPure locked nucleic acids

TheraPure dA β-Cyanoethyl Phosphoramidite 27-2030 TheraPure Locked Nucleic Acid A (Bz) Phosphoramidite 27-1340
TheraPure dC β-Cyanoethyl Phosphoramidite 27-2032 TheraPure Locked Nucleic Acid 5-Me-C (Bz)
27-1348
TheraPure dG β-Cyanoethyl Phosphoramidite 27-2034 Phosphoramidite
TheraPure DMF-dG Phosphoramidite 27-1737 TheraPure Locked Nucleic Acid G (DMF) Phosphoramidite 27-1347
TheraPure dT β-Cyanoethyl Phosphoramidite 27-2036 TheraPure Locked Nucleic Acid T Phosphoramidite 27-1346
TheraPure RNA phosphoramidites Standard DNA phosphoramidites
TheraPure Bz-rA Phosphoramidite 27-1903 dA β-Cyanoethyl Phosphoramidite 27-1730

TheraPure Ac-rC Phosphoramidite 27-1805 dC β-Cyanoethyl Phosphoramidite 27-1732
TheraPure iBu-rG Phosphoramidite 27-1906 dG β-Cyanoethyl Phosphoramidite 27-1734
TheraPure rU Phosphoramidite 27-1804 T β-Cyanoethyl Phosphoramidite 27-1736
TheraPure 2ʹ-OMe phosphoramidites Standard RNA phosphoramidites
TheraPure 2ʹ-OMe-Bz-A Phosphoramidite 27-2042 Bz-rA Phosphoramidite 27-1403

TheraPure 2ʹ-OMe-Ac-C Phosphoramidite 27-2043 Ac-rC Phosphoramidite 27-1405
TheraPure 2ʹ-OMe-iBu-G Phosphoramidite 27-2046 iBu-rG Phosphoramidite 27-1406
TheraPure 2ʹ-OMe-U Phosphoramidite 27-2044 rU Phosphoramidite 27-1404
TheraPure 2ʹ-MOE phosphoramidites Standard 2ʹ-OMe phosphoramidites
TheraPure 2ʹ-MOE-A Phosphoramidite 27-1019 2ʹ-OMe-PAC-A Phosphoramidite 27-1822

TheraPure 2ʹ-MOE-5mC Phosphoramidite 27-1020 2ʹ-OMe-iPrPAC-G Phosphoramidite 27-1826
TheraPure 2ʹ-MOE-G Phosphoramidite 27-1022 2ʹ-OMe-Bz-A β-Cyanoethyl Phosphoramidite 27-1842
TheraPure 2ʹ-MOE-T Phosphoramidite 27-1021 2ʹ-OMe-Ac-C β-Cyanoethyl Phosphoramidite 27-1823
2ʹ-OMe-iBu-G β-Cyanoethyl Phosphoramidite 27-1846
TheraPure 2ʹ-fluoro phosphoramidites
2ʹ-OMe-U β-Cyanoethyl Phosphoramidite 27-1825
TheraPure 2ʹ-Fluoro-Bz-A Phosphoramidite 27-1601
Thermo Scientific™ DyLight™ dye-labeled phosphoramidites
TheraPure 2ʹ-Fluoro-Acetyl-C Phosphoramidite 27-1604
TheraPure 2ʹ-Fluoro-iBu-G Phosphoramidite 27-1607 DyLight DY547 Phosphoramidite SY6332
TheraPure 2ʹ-Fluoro-U Phosphoramidite 27-1602 DyLight DY647 Phosphoramidite SY6334
DyLight DY677 Phosphoramidite SY6336
Fast deprotect DNA phosphoramidites
Linkers/spacers/5ʹ-modifier phosphoramidites
PAC-dA Phosphoramidite 27-1723
iPrPAC-dG Phosphoramidite 27-1726 5ʹ-Aminohexyl Linker Phosphoramidite 27-0035
iBu-dC Phosphoramidite 27-1725 TFA Amino Linker Phosphoramidite 27-1792
Ac-dC Phosphoramidite 29-1727 Phosphorylating Phosphoramidite 27-1794
TheraPure DMF-dG Phosphoramidite 27-1737
Reverse Abasic Phosphoramidite 27-1998
Structural phosphoramidites
dU Phosphoramidite 27-1738
dI Phosphoramidite 27-1744
N6-Me-dA Phosphoramidite 27-1746
5-Me-dC Phosphoramidite 27-1748
N4-Ethyl-dC Phosphoramidite 27-1743
Learn more at thermofisher.com/amidites

For Research Use Only. Not for use in diagnostic procedures. © 2022 Thermo Fisher Scientific Inc. All rights
reserved. All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified.
COL119002 0822
luchschenF/stock.adobe.com
Perspectives on the Designation of

Oligonucleotide Starting Materials
By William F. Kiesman,1 Andrew K. McPherson,2 Louis J. Diorazio,3 Leo Van den Bergh,4 Peter
D. Smith,5 John M. Northall,6 Alec Fettes,7 Tiejun Wang,8 Martin Mehlmann,9 Syed Raza,10
and Gary Held11
Reprinted from William F. Kiesman, Andrew K. McPherson, Louis J. Diorazio, Leo Van den Bergh, Peter D. Smith, John M. Northall, Alec
Fettes, Tiejun Wang, Martin Mehlmann, Syed Raza, and Gary Held. Perspectives on the Designation of Oligonucleotide Starting Materials.
Nucleic Acid Therapeutics. Apr 2021. 93-113. http://doi.org/10.1089/nat.2020.0909
ABSTRACT
The designation of starting materials (SMs) for pharmaceuticals has been a topic of great
interest and debate since the first ICH quality guidance was published. The increase in the
number and variety of commercialized oligonucleotides (antisense oligonucleotides—ASOs,
small interfering RNAs—siRNAs, etc.) in recent years has reignited dialogue on this topic
because of the unique complexity of the monomeric nucleotides and other contributory
materials used to manufacture oligonucleotides. The SM working group in the European
Pharma Oligonucleotide Consortium (EPOC) was formed to help establish simple, risk-based
criteria to guide the justification of oligonucleotide SMs. This article provides a description
of the common types of SMs, classes of SM impurities, and control strategies that will be
helpful to maintain manufacturing consistency.

INTRODUCTION therapeutics in a series of technical white

A harmonized approach for the designation papers. These draw on its collective subject
and justification of starting materials (SMs) matter expertise, complementing that in
for new chemical entities (NCEs) has been the literature and will serve as a reference
outlined in recent regulatory guidance [1,2] for industry practice and to help establish
and proposals from industry groups [3]. development principles for oligonucleotides.
These risk-based approaches provide insights The consortium aims to be proactive and
into how SMs can impact drug substance inclusive, and anticipates initiating wider
quality and also mechanisms for control of discussion on oligonucleotide CMC practice
critical attributes of SMs that may impact and policy to expedite access to potentially
drug substance quality. life-changing medicines.
As chemically synthesized active Within EPOC, the Oligonucleotide SMs

ingredients, oligonucleotides have the Working Group was launched to examine
potential to share similar risk-based member company practices and propose
justifications as more traditional, small risk-based strategies for more uniform
molecule NCEs. This anticipation is oligonucleotide SM justification packages.
hindered, however, by the lack of
recognized standards and the small This article summarizes general approaches
numbers of approved oligonucleotide to the justification packages that include the
products. Such a situation could lead to following:
justification of SMs for oligonucleotide
products being subject to inconsistent • Determination of the criticality of SM
expectations by agencies in different impurities
regions or, indeed, by sponsor companies. • Illustration using deoxy
phosphoramidites with typical quality
The European Pharma Oligonucleotide attributes of SMs and analytical
Consortium (EPOC) [4] was created in methods used for controls
2018 to address this and similar situations. • Application of justification to more
EPOC is a collaboration between multiple complex phosphoramidites, for
pharma companies with the aim of sharing example, 2′-(methoxyethoxy)ribose
chemistry, manufacturing and control (MOE) and locked nucleic acids (LNAs)
(CMC) knowledge, and strategies to • Extension to convergent syntheses
enable harmonization of oligonucleotide of oligonucleotides from smaller
development and commercialization oligonucleotides such as dimers/
practices. The consortium will publish blockmers
science-based recommendations for • Approaches for components of
the development of oligonucleotide conjugates (linkers and ligands)

A broad range of SMs has been applied three main activities that are conceptually
to the manufacture of therapeutic similar to those employed in standard small-
oligonucleotides and it is not feasible molecule preparation. These are Synthesis,
to cover all of the options. Rather than Work-up (often referred to as Downstream
provide hard and fast rules, this report Processing for oligonucleotides), and Drug
illustrates principles to consider for simpler Substance Isolation (FIGURE 1).
SMs and elaborates this as the perceived
complexity increases. In this way, sponsor When examined in more detail, however,
companies can adapt and apply these the oligonucleotide process is different from
principles to justification and specification of small molecule manufacturing. The chemical
oligonucleotide SMs in the context of their synthesis of a therapeutic oligonucleotide is
own drug projects and corporate approaches most often carried out on a functionalized
to regulatory filing. solid support using an automated
synthesizer. The oligonucleotide chain is
Oligonucleotide therapeutics are becoming extended through iterative synthetic cycles
more prevalent in the global marketplace where each cycle results in the incorporation
and manufacturing scales for these
complex products are increasing as larger
FIGURE 1: Process overview of
volume indications become legitimate
oligonucleotide manufacturing
targets. A flexible, harmonized risk-based
operations.
SM justification approach shared by
regulators and developers will help ensure
sustainable patient access to affordable,
high-quality products.
DISCUSSION
The manufacturing process
Before discussing SMs, it is necessary
to understand the oligonucleotide
manufacturing process to provide a context
for justification of certain materials as SMs.
The range of operations that are conducted
may act as purging steps for SM-derived
impurities and must be considered in any
justification.
As an end-to-end manufacturing process,

the various operations can be grouped into

of one additional nucleotide unit (FIGURE 2). ~80 synthetic steps carried out sequentially
The cycle consists of four successive steps: without pause in the process or isolated
intermediates. The solid-supported synthesis
• Detritylation: removal of a is performed on a packed column with all
4,4′-dimethoxytrityl (DMT) protecting reagents, solvents, and building blocks
group at the site where chain delivered as solutions under computer-
elongation will occur programmed control. When combined with
• Coupling: reaction with an activated the advantages already mentioned (robust,
phosphoramidite-functionalized high yielding, and highly selective chemistry
building block to enable introduction with extensive column washing steps), this
of a single-nucleotide unit into the results in a highly controlled and predictable
growing oligonucleotide chain outcome for the synthesis phase.
• Sulfurization/Oxidation: introduction
of a sulfur or oxygen atom at the newly
created internucleotide phosphotriester
SPONSORED CONTENT
linkage, resulting in conversion from
P(III) to P(V) Phosphoramidites
dilution tables
• Capping: addition of a reactive
acylating reagent to effect capping
of any unreacted hydroxyl center
remaining as a result of incomplete Once the oligonucleotide sequence has
coupling or undesired deprotection been completed, deprotection steps are
side-reactions and reduce propagation required before the work-up/downstream
of such impurities processing steps (purification, desalting,
and concentration). The first operation is
Each step is highly selective and very high an amine wash to effect removal of the
yielding and the solid support is thoroughly phosphorous backbone protecting group
washed with solvent between each successive (typically 2-cyanoethyl, CE) and results
operation in the cycle. In combination with in a global backbone deprotection of the
the high solubility of excess reagents and phosphate/phosphorothioate esters, giving
associated byproducts, this ensures that there the triethylamine salt of the resin-bound
is no carry-over of reagents, building blocks, oligonucleotide.
and nontethered impurities between the
different synthetic steps.* Treatment of the resin-supported
oligonucleotide with aqueous ammonia
The solid-supported synthesis is carried out removes various amine protecting groups
as a single continuous operation. For a 20- on the nucleobases, as well as triggering
mer oligonucleotide, this means a total of cleavage of the resin linker, resulting in

FIGURE 2: Typical oligonucleotide synthesis process.

release of the oligonucleotide from the solid possible to carry out a sequence of ethanol-
support (often referred to as cleavage and based oligonucleotide precipitations and
deprotection). Subsequent filtration and subsequent reconstitutions from water
washing (to remove the solid support) result as a means to remove low-molecular
in a solution of crude, 5′-DMT protected weight impurities and concentrate the
oligonucleotide, ready for purification. oligonucleotide.
The crude oligonucleotide solution is purified After the desalting and concentration
by liquid chromatography, typically strong operations, the API can be provided directly
anion ion exchange (SAX). Impurities not as an aqueous concentrate [5] or isolated as
closely related to the active pharmaceutical a solid—typically by lyophilization.
ingredient (API; eg, by virtue of significant
difference in chain length—shortmers/
longmers) are readily separated during SPONSORED CONTENT
the chromatographic purification step; Thermo Scientific DNAPac
however, full-length (and close to full family of columns: superior
length) oligonucleotide impurities will not oligonucleotide analysis
be removed during this step. The eluate
is progressed forward to the desalting/
Thermo Scientific DNAPac
concentration step. Family of Columns
The repetitive nature of this overall
oligonucleotide synthesis process where
Superior Oligonucleotide Analysis
Depending on the precise nature of the individual cycles apply similar conditions
process, the 5′-DMT group can be removed to substrates and reagents of broadly
during the solid-supported synthesis before similar reactivity delivers a predictable
amine treatment, during chromatography, or as outcome and is routinely used as the
a standalone postchromatography operation. method of choice for Good Manufacturing
Practice (GMP) manufacture of therapeutic
In the desalting/concentration step the oligonucleotides. In this respect, the high
counter ion is exchanged (if needed) and the degree of automation, absence of in-
oligonucleotide solution is concentrated. process testing, and robust performance
This can be achieved by ultrafiltration/ are reminiscent of a well-understood, well-
diafiltration, for example, through use defined commercial manufacturing process,
of a tangential flow filtration apparatus even for preclinical manufactures.
equipped with membranes. During this step,
residual organic solvents, salts, and low- Although oligonucleotide therapeutics
molecular weight impurities are removed are explicitly excluded from the scope of
according to the pore diameter cutoff ICHQ6A (and by reference to ICHQ11),
size of the membrane. Alternatively, it is there are concepts in the quality guidelines

that are currently being generally applied to • Oligonucleotides with related

oligonucleotide manufacturing with the most sequences or size typically possess
relevant being ICH M7 (regarding mutagenic similar physical properties; so purging
impurities) [6], ICH Q3C (residual solvents) of impurities with the same or similar
[7] and ICH Q3D (elemental impurities) [8], number of nucleotides as the desired
and ICH Q7 (Good manufacturing practices) product (full length impurities) is
[9]. One notable diversion from the guidance challenging.
relates to ICH Q3A (Impurities in New • Effective SM specifications supported
Drug Substances [10]). While the general by detailed understanding of the
concepts of reporting limits, identification fate and control of impurities in
limits, and qualification limits still apply, phosphoramidites or other SMs are vital
higher thresholds have been accepted for aspects of the overall oligonucleotide
oligonucleotides than for small molecules, control strategy.
to date [11]. • Analysis of the risk of impurity
carryover across the unit operations
ICH Q11 provides a science-driven, is of greater priority than number of
risk-based framework addressing the chemical transformations to reduce the
propensity for SMs to influence the quality risk of contamination and support the
of the drug substance. This requires a control strategy throughout the product
thorough understanding of actual and lifecycle.
potential impurities, as well as their fate • The operation of many steps without
in downstream processing gained from interspersed analysis during solid-
knowledge of the synthetic route coupled supported synthesis does not support
with risk assessments. Although ICH Q11 application of traditional stepwise
explicitly states that oligonucleotides are impurity fate and purging approaches.
out of scope, the ethos for SM selection • Ideally, the SM can be sourced as a
outlined in ICH Q11 remains applicable, but commodity with controlled quality
should be considered with an appreciation
of oligonucleotide processing. Working Phosphoramidite building blocks (henceforth
in this way, a number of principles can described as amidites) are manufactured
be considered in the designation of using standard chemical manufacturing
oligonucleotides SMs: technology and are well controlled. Several
amidites are widely available from third-
• A defined and stable structure with party commercial suppliers with controlled
characteristic chemical and physical quality. To date, 2′-deoxyribose amidites
properties (deoxyamidites) and 2′-(2-methoxyethoxy)
• A significant structural fragment toward ribose amidites (MOE amidites) have
the structure of the drug substance been accepted as appropriate SMs for

oligonucleotides, demonstrating that how regulatory guidance can be applied to

amidites can be acceptable in accordance oligonucleotide SMs before introducing more
with the principles of ICH Q11 guidelines. complex examples.
In this situation, the application of
GMPcontrols for oligonucleotide drug Deoxyamidites are incorporated into the
substance manufacturing processes starts oligonucleotide as significant structural
from the amidite SMs with an appropriate fragments and therefore fulfill the most
control strategy to ensure quality of the basic requirement for SMs. The most
finished oligonucleotide. prevalent examples are the 2-cyanoethyl-
N,N-diisopropylaminophosphoramidites,
Deoxyamidites which were first applied in oligonucleotide
The deoxyamidites, for example, 1–4 (FIGURE 3), synthesis in 1984 [12]. They were rapidly
whose core structures are found naturally adopted as the standard approach due
in DNA, are the most widely used building to supporting highly efficient coupling
blocks in oligonucleotide APIs. As such, following appropriate activation. Supply
they provide a convenient introduction for of these materials has increased to a point
where they are commercially available in
large quantities (up to hundreds of kg batch
FIGURE 3: Deoxyamidites (atoms
size) from multiple vendors worldwide with
marked in blue are incorporated into
multiton annual capacity available in the
the oligonucleotide).
market. In addition, there are many more
vendors capable of supplying medium- to
small-scale amounts of material.
As might be expected for such

established materials, the structure and
physiochemical properties of deoxyamidites
have been rigorously characterized.
A variety of analytical techniques, for
example, 1H/13C/31P/2D-NMR (two-
dimensional nuclear magnetic resonance
spectroscopy), specific rotation, and high-
performance liquid chromatography with
ultraviolet and mass spectrometry detection
(HPLC-UV-MS) have been applied in EPOC
member companies and elsewhere, providing
a confidence in the robustness of their
quality. These studies have led to a detailed

understanding of potential and actual

impurities, as well as their fate and impact FIGURE 4: General method used to
on the quality of the target oligonucleotide. manufacture deoxyamidites.
Consequently, the material attributes of
the deoxyamidites that may impact the API
critical quality attributes (CQAs) [1] (eg,
reactive critical, impurities) can be defined,
and analytical methods with appropriate
acceptance criteria can be validated, leading
to specifications to purchase materials.
This comprehensive understanding further
supports acceptance of deoxyamidites
as SMs, and a more detailed discussion is
presented in the next section.
Deoxyamidites are free-flowing,

nonhygroscopic, amorphous solids. They are
derived from the corresponding nucleosides
that are obtained from non-animal sources
through fermentation and readily available
in ton quantities in stereochemically
pure form. The amidites are a mixture of nucleosides (12–15, often referred to as
the two diastereoisomers at the P atom PNS). The final step is the phosphitylation
and typically seen as double peaks both of PNS with 2-cyanoethyl-N,N,N′,N′-
in 31P NMR and high-performance liquid tetraisopropylphosphordiamidite (often
chromatography (HPLC). Commercially referred to as Phos reagent or P-reagent)
available deoxyamidites are synthesized in the presence of an activator. Preferred
by the following general synthesis scheme activators are small, weak, nonhygroscopic
(FIGURE 4). organic acids.† Although no systematic
study has been conducted, anecdotal
The 2′-deoxyribonucleosides 5–7 are evidence derived from more than 20 years
acylated on the exocyclic primary amines of of experience within the EPOC partners
the nucleobases, that is, adenine, cytosine, related to deoxyamidites synthesized using
and guanine, with desired protecting two common activators (1H-tetrazole, DCI)
groups (step 2). Since thymine does not indicates that activator choice does not
have an exocyclic amine, this step is not appear to influence the amounts of reactive
performed. Protection of the 5′-hydroxyl as impurities that result in oligonucleotide drug
the DMT ether affords the fully protected substance impurities. For similar reasons,

solvents used in the reaction are not

considered to impact SM CQAs. FIGURE 5: Examples of amidite
impurity types.
In current commercial scale processes,
although no new carbon stereocenter is
created from the nucleoside, the synthesis of
the amidite results in an R/S stereochemical
mixture at phosphorus. Crucially, the
absolute configuration of the phosphorus
atom of deoxyamidites does not impact
the distribution of oligonucleotide
diastereoisomers. This is determined by
other variables during the coupling reaction
in the oligonucleotide chain extension
cycle; therefore, control of phosphorus
stereochemistry in deoxyamidites is
unimportant [13,14].
Impurities in deoxyamidites can be assigned

to two broad groups based on an assessment
of their reactivity during oligonucleotide
coupling and the ability to purge any
resulting impurity during the manufacturing
process. The most important group results
in impurities in the crude drug substance these are commonly known as noncritical
that are not subsequently purged. These impurities. These might be such species
are known as reactive, critical (or critical) as related nucleosides and nucleotides or
impurities and generally contain both residual solvents that do not react with the
phosphoramidite functionality and an acid- oligonucleotide chain during coupling. Such
labile protecting group, such that they can inert components are known as nonreactive,
propagate chain elongation (FIGURE 5). It is noncritical impurities. There is a second
the individual and total amounts of these subset of noncritical impurities that do react
critical impurities that constitute the CQAs with the evolving oligonucleotide chain
of deoxyamidites [15]. during coupling, but do not affect product
quality. This might be due to the resulting
The second group comprises deoxyamidite impurity in the oligonucleotide being readily
impurities that have no impact on the final purged, for example, during chromatography,
drug substance purity and, unsurprisingly, or because the impurity motif is lost during

processing and results in the target product. containing 16 are not purged during
These are the reactive, noncritical impurities. downstream processing, thereby rendering
Examples of each class are presented it critical; 16 is controlled by purification
(TABLE 1). at the PNS and amidite stages and by
specification.
The reactive, critical impurity 16 will
incorporate into the oligonucleotide It is important to note that the
during the coupling reaction through a deoxyamidite examples provided in
5′-5′ internucleotide linkage. Detritylation TABLE 1 were used to illustrate the
during the subsequent deprotection step definitions of the classes of impurities,
will actually release a terminal 3′-hydroxyl but in actual practice, there are almost
at the expected 5′-terminus leading to a no traces of critical impurities in
subsequent 3′-3′ internucleotide linkage deoxyamidites. Specifically, the levels
before reverting to normal progress in of 16, when present at all, are very
later cycles. This introduces a reversed low (~0.04%) due to the high degree
nucleotide to a sequence. Due to its of selectivity between the 3′- and
similarity to the parent, oligonucleotides 5′-hydroxyls in the tritylation reaction.
TABLE 1: Origin and control of indicative critical and noncritical impurities in

deoxyamidites.
Impurity type Source Method(s) of control
16 Reaction of DMT-Cl with 3′-OH PNS purification; deoxyamidite

rather than 5′-OH during PNS purification; deoxyamidite
Reactive, critical formation specification
17 Impurity in DMT-Cl reagent with Vendor DMT-Cl specification;

Reactive, noncritical one chlorine on one aryl ring deoxyamidite specification
Control of water during PNS

18 Amidite hydrolysis during phosphitylation; deoxyamidite
Nonreactive, noncritical phosphitylation purification; deoxyamidite
specification
19 Bz protecting group was either PNS purification; deoxyamidite

Reactive, critical or reactive, never installed or was installed, but purification; deoxyamidite
noncritical later cleaved specification
DMT, 4,4′-dimethoxytrityl; PNS, protected nucleoside.

The presence of an extra chlorine atom in oligonucleotide impurity than for the final
reactive, noncritical impurity 17 does not cycle where it may just lead to parent
materially affect the rate of subsequent oligonucleotide. In such cases, criticality will
detritylation, which removes this impurity need to be assessed for the process chosen
motif from the chain. Oligonucleotide and the appropriate limit set for 19.
chains that have incorporated 17 continue
to elongate into full-length parent Deoxyamidites typically contain trace
oligonucleotide. Therefore, while quantities of water and solvents (eg,
impurity 17 is reactive during the acetonitrile, ethyl acetate, and toluene).
synthesis, it is noncritical with respect to These do not react with the growing
oligonucleotide CQAs. support-bound oligonucleotide and are
washed away or purged, as such they do
The nonreactive, noncritical impurity 18 not affect drug substance purity and are
(H-phosphonate) is unreactive under normal therefore considered noncritical. A steady
amidite coupling conditions and therefore evolution of deoxyamidite manufacturing
passes through the synthesis column and means that there are almost no traces of
away from the oligonucleotide during amidite critical impurities observed (TABLE 2).
delivery and subsequent washing steps.
While noncritical, some control over 18 is Of 72 lots of materials manufactured by 4
advantageous for yield consistency. vendors on multi-kg scales, 13 contained a
single critical impurity in the range 0.04%–
Criticality for other reactive impurities will 0.11% with a further 2 batches containing
not always be so clearly defined. Reactive only 19 in the range 0.2%–0.3%. Only two
impurity 19 might be present in 1 and will batches contained >1 critical impurity,
couple as normal, but will introduce branching but still totaling <0.2%. Similar summary
impurities throughout the rest of the statistics on assay, purity by 31P NMR, water
synthesis. During subsequent coupling cycles, content, and residual solvents (TABLE 3) show
one or both growing branches may fail to deoxyamidites to be of consistently high
extend and be capped, generating a complex quality from multiple vendors.
mixture of branched species of various lengths
with as many as two DMT groups present on Specifications for deoxyamidites focus on
the 5′-termini. Depending on which coupling control of critical impurities and overall
cycle was compromised, the resulting mixture purity (TABLE 4). The need to control
of branched oligonucleotide impurities may be critical impurities will be evident from
easier or harder for the chosen downstream the foregoing discussion, but control of
purification method to purge. Incorporation overall purity can also be important. For
in the first coupling cycle would result in a example, reversed-phase purification
much larger (and generally easier to separate) ruggedly purges failure sequences, whereas

TABLE 2: Deoxyamidite Impurity Lot History Summary

Residual
Lots with critical Assay range P purity range
31
Water
Vendor No. of lots solvents
impurities (% a/a) (% a/a) (% w/w)
(% w/w)
1 31 8 94.0–99.6 96.8–100.0 0.11–0.35 0.23–2.90
2 22 5 94.4–99.7 98.8–100 0.07–0.67 0.84–2.42
3 10 1 97.0–99.5 99.3–99.9 0.10–0.28 0.22–2.36
4 9 1 94.9–100.5 97.9–100 0.11–0.23 ND–1.1
Total 72 15
ND, not detected.
TABLE 3: Additional summary statistics for 72 lots of deoxyamidites.

Characterization Mean Median SD
Assay (% w/w) 97.5 98.1 1.62
Purity by 31P NMR (% a/a) 99.2 99.4 0.83
Water (% w/w) 0.19 0.15 0.108
Residual solvents (% w/w) 1.07 0.95 0.568
NMR, nuclear magnetic resonance spectroscopy.
with some oligonucleotide sequences, to aid process robustness (and to avoid

strong anion exchange chromatography paying for expensive noncritical impurities).
is sensitive to quality of input materials. When considering critical impurities, it is
If the chosen process for oligonucleotide also necessary to consider the multiplicity
manufacture cannot completely purge of deoxyamidite incorporation due to the
coupling failures, then coupling efficiency impurity family approach typically applied to
becomes a critical process parameter and, oligonucleotide impurities [11]. The presence
by extension, the overall purity of the of a single, critical reactive amidite impurity
deoxyamidites becomes critical. Even if at 0.05% w/w in each amidite during the
the oligonucleotide process can purge all synthesis of a 20-mer oligonucleotide could
failed sequences, it is still advantageous lead to a maximum of 1.0% (20 × 0.05) of
to control the overall deoxyamidite purity the corresponding oligonucleotide impurity.

TABLE 4: Additional summary statistics for 72 lots of deoxyamidites.
Test Method Acceptance criterion
Appearance Visual inspection White to yellow powder
Identification LC-UV-MS MoIM of the sample and the reference standard agree to
within an amu limit. Retention times of both main peaks
of the sample and reference standard agree to within a
limit
Assay LCa,b NLT 90.0%

Critical impurity
NMT 0.20%
Any unspecified critical impurityc
NMT 0.15%
Total critical impurities
Impurity profile LCa,b NMT 0.50%
Purity 31
P NMR ≥95.0% a/a
Water content KF NMT 1.0% w/w
Residual solvents GC NMT 4.0% w/w
LC methods may be reported as either area percent or weight percent at a specific wavelength that can vary depending upon the
a
method.
Both UV and MS detection methods have been applied.
b
Any impurity that contains both an amidite moiety and DMT protecting group may be critical and should be investigated; all other
c
impurities are noncritical.
GC, gas chromatography; KF, Karl Fischer; LC-UV-MS, liquid chromatography with ultraviolet and mass spectrometry detection; MoIM,
monoisotopic mass; NLT, not less than; NMT, not more than.
This is due to amplification depending upon More complex amidites

how often the individual amidite is used As the oligonucleotide space has matured,
in a specific sequence since the motif will a broader range of amidites has been
be incorporated at low level during each accommodated within sequences (FIGURE 6).
coupling cycle. This may not be a problem if The more popular of these have been those
amidite impurities are qualified appropriately, derived from RNA nucleotides 20 such as
but could lead to surprises if not anticipated 2′-F [16] and 2′-OMe [17], more elaborate
and controlled in the deoxyamidite versions such as the LNAs 21 [18], and other
specification. ring systems such as the morpholinos 22 that

FIGURE 6: Examples of more complex FIGURE 7: MOE amidites (atoms

amidites. marked in blue are incorporated
into the oligonucleotide). MOE,
2′-(methoxyethoxy)ribose.
lead to the phosphomorpholino

oligonucleotides (PMOs) [19], as well
as many others. It should come as no
surprise that the same approaches and
impurity classifications identified for the
deoxyamidites can be applied to these more These materials are coupled in the
complex amidites. oligonucleotide synthesis in the same manner
as deoxyamidites. They are also stable
One of the most popular ribose-derived solids and their raw materials (non-animal
monomers used in oligonucleotide APIs are sourced nucleosides 27) are available in ton
the 2′-O-(2-methoxyethyl)ribonucleoside quantities. The synthesis of commercially
amidites (MOE amidites, 23–26 R = OMOE) available MOE amidites follows a general
(FIGURE 7). MOE amidites are SMs for a approach dependent on whether they bear
number of marketed products (Kynamro, purine or pyrimidine bases (FIGURE 8).
Tegsedi, Spinraza, and Waylivra) and will be
used to extend the deoxyamidite argument Introduction of the MOE group onto the
into a more complex case. 2′-hydroxyl of nucleoside 27 is achieved

in one of two ways. If 24 possesses

FIGURE 8: General approaches to
a purine nucleobase (ie, adenine and
MOE amidites.
guanine), direct alkylation of the nucleoside
at the 2′-hydroxyl is achieved with an
activated form of 2-methoxyethanol [20].
Thus, analogous to the deoxyamidites,
all the ribose stereocenters of, for
example, 23 and 25, are derived directly
from the corresponding sugar and
stereochemical integrity is maintained
during conversion to the amidites.
Stereoisomers of 23 or 25 should, therefore,
not be treated as CQAs.
When the nucleobase in 27 is a

pyrimidine (eg, 5-Me cytosine and
5-Me uracil), it is first converted to a
bicyclic oxazolidine 30, inverting the
stereochemistry at the 2′-position following
an SN2 mechanism [21]; 30 can only be
formed as a single stereochemical isomer
following neighboring group displacement
of the activated 2′-OH by the nucleobase.
Ring opening of 30 with 2-methoxyethanol
or a nucleophilic variant also results in
inversion of C2′, that is, the overall effect
can only be for double inversion at C2′, thus
retaining the natural configuration.
For both ring types, nucleosides 28 and

31 are converted into the corresponding
MOE amidites in the same way as for
deoxyamidites. The exocyclic primary
amines of the nucleobases are acylated
(adenine, methylcytosine, and guanine),
the 5′-hydroxyl is protected as the DMT presence of an activator. For pyrimidines,
ether, and finally, the 3′-hydroxyl is nucleoside 31 (X = O) can also be converted
phosphitylated with Phos reagent in the into 31 (X = NH) if required.

In the case of pyrimidine MOE amidites, specificity for 34 in the release methods,
there is a theoretical risk that 30 could for example, 24 and 26. As a consequence
be subject to an alternative reaction of this observation and, since the integrity
involving neighboring group attack from the of all other stereocenters remains intact
3′-hydroxyl leading to epoxide 32 (FIGURE 9). during conversion of 27 to 24 or 26, the
In this situation, epoxide opening of 32 by stereochemical integrity of pyrimidine MOEs
methoxyethanol would give 33 as an is not treated as a CQA.
impurity in 31, which, on completion
of the synthesis, would be expected to In the same way as for deoxyamidites,
provide 34 as a critical, diastereomeric MOE amidite impurities can be treated
MOE amidite impurity. As with all such as noncritical or critical and the same
investigations, it is always important to conditions apply. Species that do not
ensure that the potential for impurity react (eg, solvents, phosphonates,
formation through reasonable reaction phosphonoamidates, and phosphoramidates),
pathways has been considered. In this either do not contain an activatable amidite
case, the isomer pathway leading to 34 has group or are analogous to the reactive,
never been observed, despite extensive noncritical deoxyamidite impurities
investigations (McPherson A, 2020, described earlier and are readily removed
unpublished data), although it might during purification. Introduction of the
be viewed as prudent to demonstrate alkyl side chain at the 2′-O position does
introduce some 2′-O impurities 35 that can
FIGURE 9: Theoretical C-2′ inversion be incorporated into the oligonucleotide
of MOE amidite precursors. during synthesis and should be considered
critical impurities (FIGURE 10). These
alkylation impurities are monitored in the
MOE amidites by HPLC-UV-MS and are
readily controlled in the SM specifications.
As with deoxyamidites, MOE amidites

carry the potential for regioisomeric
impurities 36 (sometimes called
inverted amidites) if tritylation occurs
on the 3′-hydroxyl and phosphitylation
occurs on the 5′-hydroxyl (FIGURE 11).
Furthermore, an alternative 3′-alkylation
of purine MOE nucleosides could occur
followed by 2′-phosphitylation to form a
different set of regioisomer impurities 37.

Any oligonucleotide impurity derived

FIGURE 10: 2′-O impurities in MOE
from 36 and 37 would be not be removed
amidites.
during downstream processing due to the
similarity with the parent; 36 and 37 are
therefore controlled by purification in the
MOE amidite SM synthesis and by SM
specifications.
As might be expected from the additional

manipulations in their synthesis, MOE
amidites typically have more critical
impurities than their deoxyamidite analogs.
Of 104 lots of materials manufactured
on multi-kg scales by 4 vendors, critical
impurities ranged from below the limit of
FIGURE 11: Regioisomeric MOE detection (<0.04% a/a) to a high of 0.8% a/a
amidite impurities. (TABLE 5). Similar summaries on assay, purity
by 31P NMR, water content, and residual
solvents show that MOE amidites are
manufactured to a consistently high quality
from multiple vendors.
Generally, MOE amidite specifications

track water content and solvents similar to
TABLE 5: MOE amidite impurity range summary.

Purity by Total critical Residual
Water
Amidite No. of lots Assay (% w/w) 31
P NMR impurities solvents
(% w/w)
(% a/a) (% a/a) (% w/w)
MOE
24 91–101 95–100 0.04–0.8 0.1–0.5 0.3–1.7
Me
U
MOE
28 95–101 97–100 0.04–0.25 0.1–0.3 0.2–2
Me
C
MOE A 25 97–101 97–100 0.04–0.4 0.1–0.7 0.3–1.7
MOE G 27 92–98 96–99 0.04–0.7 0.2–0.9 0.3–3.2
MOE amidites in Table 6 have the base protection schemes depicted in Fig. 5.
MOE, 2′-(methoxyethoxy)ribose.

TABLE 6: Example impurity specifications for MOE amidites.

Critical Impurity NMT 0.2–0.4% a/a
Any unspecified critical impurityb NMT 0.15% a/a
Impurity profile HPLCa Total critical impurities NMT 0.5–0.8% a/a
Purity 31
P NMR NLT 95.0% a/a
a
Both UV and MS detection methods have been applied.
b
ny impurity containing both a amidite moiety and a DMT protecting group may be critical and should be investigated; all other
A
impurities are noncritical.
HPLC, high-performance liquid chromatography.
those for deoxyamidites, but mainly focus the deoxyamidites and MOE amidites. In
on control of critical impurities and overall addition, multiple synthetic routes might be
purity. Typical values for impurity limits used employed to deliver materials.
in clinical and commercial products by EPOC
partners are outlined (TABLE 6) (Note: limits The LNA derivatives are typical examples of
should be defined and be fit for each new this additional complexity and, in the case
oligonucleotide sequence and tailored to of the more challenging constrained ethyl
the controls for the specific manufacturing (cEt) amidites (cEts, 21 R = Me) (FIGURE 6), a
process—one size does not fit all). The number of chemical routes to these materials
specifications for individual critical impurities have been published [22–24]. A common
and the totals are somewhat higher than characteristic is that all follow lengthy
for the corresponding deoxyamidites. This synthetic sequences, although the final
reflects that higher levels of detectable stages are similar.
impurities have been observed and used
successfully in clinical and commercial Salinas et al.’s approach [23] (FIGURE 12)
manufacturing by EPOC partners. is a linear synthesis and starts from the
appropriate RNA nucleoside 26 in a similar
All the points raised for deoxyamidites and manner to the MOE amidites. Although only
MOE amidites can be extended further for demonstrated for cEt MeU 42, extension to
even more complex amidites. These are other amidites should be possible.
often proprietary in nature, which brings
the additional complication that supply The Seth et al. [22] and Blade et al. [24]
chains may not be so well established as syntheses offer a more complete approach

FIGURE 12: Salinas cEt route. FIGURE 13: Blade cEt route.
to delivering a range of cEt amidites. In

both cases, a linear sequence provides a
common intermediate such as 46 (FIGURE 13),
which supports a divergent approach to to that of Seth’s from 39 onward. A similar
the required amidites; 46 contains the divergent approach is applied toward the
key skeletal elements present in the final LNAs (21, R = H) [25,26].
amidite, although not in the final, structural
presentation. Given acceptable molecular The nature of the route (divergent vs.
properties, 46 also represents a convenient linear) is important since critical reactive
storage point if flexibility is required, for impurities would generally be expected to
example, to support a broad development possess comparable kinetics to the required
portfolio. The Salinas synthesis is identical phosphoramidites during coupling. The linear

approach reflects a more traditional situation which further mitigate the potential impact
for SMs where provision of each cEt amidite of this apparent complexity:
can be viewed as an independent activity
with impurity identification and purging only • The stereochemistry at C-3′ is inverted
relevant to that specific SM. from that originally in di(acetone)
glucose as a result of oxidation and
In the case of the divergent approaches, the later reduction. This is a common
situation is somewhat more complex. Any reaction sequence on protected glucose
inherent impurity not purged before isolation to invert C-3′ and, hence, obtain less
of 46 could, in principle, lead to analogous common sugars [27]. Steric crowding
impurities in all cEt amidites generated ensures that delivery of hydride during
from that batch, and in turn result in higher reduction of 49 (FIGURE 14) occurs from
levels of the corresponding oligonucleotide the convex face of the [3.3.0] ring
impurity. Impurities generated downstream system, resulting in the desired 3′-(S)
from 46 will be discrete to the individual configuration in 50.
cEt amidites, although the chemistry across
the divergent stages is quite similar and FIGURE 14: Stereochemical control in
therefore one might see common issues to cEt amidites.
various extents.
In practice, the linear approach is not

pursued at scale and cEt amidites are
produced by one or other of the divergent
syntheses. As with deoxyamidites and
MOE amidites, the precursor di(acetone)
glucose is chiral, naturally derived, and
well characterized with unambiguous
stereochemistry. A key difference in the case
of the cEt amidites is that reaction occurs at
four of the five original stereocenters with
overall inversion at each, potentially resulting
in a more complex chemistry to follow
for the SM. Confidence in stereochemical
integrity can be provided by approaches
such as X-ray structural elucidation or
NMR correlation studies following key
transformations. Some general mechanistic
observations can also be applied, however,

• The base is added through a • Control of stereochemistry for the

Vorbrüggen reaction [28,29], whereby pendant 6′-Me on the 2′-4′ bridging
an equilibrating mixture of activated group
C-1′ acetate isomers 51 is internally • 6′-Me-deletion impurity (M-14, 55).
displaced by participation of the
C-2′ acetate, giving acetoxy-bridged For all approaches, there is potential for
intermediate 52 (FIGURE 14). This is low levels of the 6′-(R) diastereomer 54 to
a well-documented process and is be formed (FIGURE 15). Dependent on
followed by opening of 52 at the the synthesis employed, this is either a
activated C-1′ by the nucleobase consequence of incomplete stereochemical
leading to 47. This again occurs from inversion during an oxidation/reduction
the convex face of a [3.3.0] bicyclic sequence (39→40) or through activation
system resulting in the desired 1′-(R) of the secondary alcohol rather than the
configuration. primary during an epoxide formation
• Hydrolysis of the C-2′ acetate (45→46); 54 can be readily identified at the
in 47 induces nucleophilic attack of point of formation, but reacts in subsequent
the oxyanion on to the C-6′ mesylate steps in a similar manner to the parent 6′-(S)
to provide the cEt bicyclo [2.2.1] isomer and the resulting impurities can be
framework 53 (FIGURE 14). This tracked through the synthesis.
mandates the relationship between
C-2′ and C-4′ since both the 2′-OH
and the mesylate-bearing C-4′ side FIGURE 15: Critical cEt amidite
chain must be on the same face to
impurities.
successfully react. Since C-2′ retains
the natural configuration found in
glucose, this provides additional
confidence for the stereochemistry at
both C-2′ and C-4′. Also, in this step,
SN2 displacement of the C-6′ mesylate
sets the stereochemistry at C-6′
by inversion of the natural glucose
stereochemistry.
Over and above the previously described Impurity 55 is a feature of both the
concerns for deoxyamidites and MOE Seth and Salinas approaches arising
amidites, the major novel challenges for from incomplete reaction during methyl
cEt amidites arise from the following: addition to an aldehyde (39→40).
Such an impurity is to be expected and

the requirement for a demonstrable

understanding of its fate should be FIGURE 16: Critical cEt amidite
anticipated. The Blade approach avoids impurities.
this methylation step completely and
a paper assessment had ruled out this
impurity motif. Surprisingly, 55 was
observed as a byproduct, proposed
to result from an unanticipated bond
cleavage mechanism. This observation
further emphasizes the need for vigilance
during SM synthesis as with the previous
situation relating to the anticipated, but Alternative synthetic approaches to
unobserved 2′-diastereomer for MOE oligonucleotides
pyrimidine amidites 34. The importance In comparison with small molecule synthesis
of a detailed understanding of generation where convergent approaches are viewed
and fate of impurities rather than as desirable, oligonucleotide synthesis
taking an assumed position cannot be has largely remained as a linear (and
overemphasized. Once such an impurity lengthy) exercise. In an effort to introduce
is observed and identified, actions can convergency, alternative approaches
be taken to proactively purge if deemed have been considered such as the use of
appropriate to reach desired quality dinucleotide amidites [12,30–33] or the
levels (eg, through reactive chemistry or combination of shorter oligonucleotides as
recrystallization of an intermediate). The demonstrated using the templated ligation
impact of purging on these impurities is approach exemplified by Crameri et al.
presented ( TABLE 7). (FIGURE 16) [34].
TABLE 7: Purging of impurities 54 Such approaches are usually described as

And 55 during constrained ethyl blockmers as exemplified by the deoxy GT
manufacture. blockmer amidite 56 (FIGURE 17). Similar
criteria to that described previously should
Stage 54 (% a/a) 55 (% a/a) be applied when designating SMs for
Initial formation 0.24–0.30 0.7–1.0 these alternative approaches. A number
of obvious additional challenges can be
26 0.05–0.12 0.5
identified due to the existence of the
cEt amidite ND 0.1–0.2 internucleotide phosphorus linker. The
first is associated with the coupling cycle
cEt, constrained ethyl.
required to deliver 56. This results in a series

of impurities more typically associated with the blockmer introduces a second, variable,
finished oligonucleotides, for example, chiral element leading to all discrete species
coupling failures, overcoupling, P = O (for being present as 4 diastereomers. In the case
X = S). Thus the set of reactive, critical of a phosphate linkage, this complexity does
impurities may be more significant than for not extend through to the API where the
the corresponding monomer deoxyamidites. phosphate is achiral.
The presence of the second phosphorous This additional stereochemistry has a

functional group compounds this issue. dual impact since signal/noise is reduced
The amidite group has already been and the number of potentially observable
identified previously as bringing a mixture components increased, both by a factor
of diastereomers. Aside from the special of approximately two over standard
case of a stereo-specific phosphorothioate, monomer amidites. The consequence is
that blockmers present greater technical
FIGURE 17: Representative blockmer challenges in identifying/quantifying
deoxy amidite. impurities, thus increasing complexity during
the development of an appropriate control
strategy for these molecules. A blockmer
approach will also require a larger number of
potential SMs across a portfolio of projects
rather than the limited number of monomer
amidites generally employed to manufacture
oligonucleotides.
No marketed oligonucleotides currently

apply such convergent approaches, but as
the pressure to supply ever larger quantities
of oligonucleotides to meet growing patient
demands continues to build, scalable
alternatives to the current solid-supported
manufacturing process may entertain these
types of routes. The Q&A for ICH Q11
advises that convergent syntheses are
acceptable and in answer to question 3, “ICH
Q11 general principles apply to the selection
of starting materials for linear or convergent
syntheses [1]. The ICH Q11 general
principles should be applied independently

to each branch of a convergent synthesis, agents that are small molecules attached
unless the point of convergence of the to the oligonucleotide (FIGURE 18) such
branches occurs upstream of an appropriate as cholesterol [35], tocopherol [36],
starting material.” In addition, although anisamide [37], folic acid [38], peptides [39],
a blockmer contains multiple nucleotide anandamide [40], N-acetyl-d-galactosamine
subunits, they may be considered SMs as (GalNAc) [41], and poly(ethyleneglycol)
highlighted in answer to question 2, “ICH ethers (PEGs) [42]. These are generally
Q7 states that an ‘API starting material’ is attached to the oligonucleotide through a
a raw material, intermediate, or an API that cleavable tether often referred to as a linker
is used in the production of an API. When a (highlighted in red, FIGURES 18 and 19). There
chemical, including one that is also an API, is further activity, whereby large molecules
is proposed to be a SM, all ICH Q11 general such as antibodies [43] and aptamers [44]
principles still need to be considered.” More are applied, although this is out of scope for
optimistically, it should be noted in the this discussion.
conceptually similar field of peptides that,
at least one EPOC member company has Most of these modifications can occur at
reported successful justification of 2-mer either the 3′ or 5′ end of the oligonucleotide,
peptides as SMs for the construction of although other modifications are also possible
larger peptides. In that case, several agencies such as an internucleotide phosphonate [45].
accepted a specification for dipeptides
that comprised comprehensive identity N-acetyl-d-galactosamine (GalNAc)
(composition and sequence), purity and conjugation has become increasingly popular
impurity profile, chiral purity, water (Karl for the targeted delivery of chemically
Fischer [KF]) or limit of detection (LoD), and modified oligonucleotides to hepatocytes
residual solvents (A. Charaf, 2019, personal through binding to ASGR (asialoglycoprotein
communication). receptor) [46,47]. Although the chemical
modification can take several forms,
Conjugating agents they retain a common feature, in that,
Due to the cost of oligonucleotides, several GalNAc moieties (typically 3—the
efforts to reduce patient dosage have triantennary structure) are connected to
become more important. Besides modified an oligonucleotide through a linker for
nucleotide chemistries, the use of optimal binding. Beyond this, a variety
conjugates has become more common to of differences in the nature of the spacer
improve pharmacokinetics and distribution (eg, alkyl, ethylene glycol) and the point of
and facilitate cellular uptake mainly for attachment (eg, tris and lysine-lysine) to the
antisense oligonucleotide (ASO) and oligonucleotide have been applied. These are
small interfering RNA (siRNAs). The most schematically summarized (FIGURE 19) for the
common modifications are conjugate most widely used approaches [48,49].

FIGURE 18: Selected small molecule conjugates.
Attachment of GalNAc at the discussed, although the principles set out

5′-oligonucleotide terminus serves as a are equally applicable.
useful example of the treatment of conjugate
fragments. The conjugation can be done For 5′-GalNAc, completion of the
after the solid-phase synthesis or starting oligonucleotide fragment synthesis is
with the oligonucleotide construct loaded on typically followed by the solid-phase
the solid support (FIGURE 20) [50]. coupling of a spacer amidite (such as 57)
and, finally, by the solution-phase coupling
As an example of the approach, of the fully assembled GalNAc cluster,
this publication will focus on post- for example, 58. Compared to naked
oligonucleotide synthesis conjugation, oligonucleotides, GalNAc-conjugated
that is, “5′-GalNAc,” specifically bis-lysine oligonucleotides therefore contain two
cluster 58. “GalNAc on oligonucleotide” additional significant structural elements of
and “3′-GalNAc” [41] will not be the API. In a similar way to the (deoxy)ribose

FIGURE 19: Selection of GalNAc attachment motifs. GalNAc, N-acetylgalactosamine.
amidites, this fulfills the most fundamental is typically introduced in the final synthetic
criterion for an SM. step offering fewer transformations than
would be considered acceptable in the realm
Based on ICH Q11, “enough of the drug of traditional small molecules. However, the
substance manufacturing process should be extensive downstream processing, including
described in the application. …” In the case of chromatographic purification, desalting by
5′-GalNAc conjugates, the 5′-GalNAc cluster ultrafiltration/diafiltration, and isolation,

FIGURE 20: 5′-GalNAc conjugation options.
may compensate for the reduced number of triethyleneglycol spacer to enable additional
chemical transformations. separation of GalNAc from oligonucleotide
(FIGURE 21).
The linker used in 5′-GalNAc conjugates
is typically a protected 6-aminohexyl The convergent synthesis commences with
phosphoramidite (6-AH) such as 57. This benzyl glycinate 59, which is converted
is of a similar size to conventional small to triethyleneglycol spacer 60 in a
molecule SMs and can therefore be treated diazotization-promoted substitution [51].
as such; it will not be discussed in any Glycosidation with N-acetyl-d-galactosamine
further detail. GalNAc cluster 58 is a typical tetraacetate 61 mediated by TMSOTf
example of its type and features a bis-lysine provides β-anomer 62. The bis-lysine
moiety to allow for sufficient branching and a coupling partner 65 is readily prepared from

Boc-protected l-lysine benzyl ester 63, which Crude 65 is used without further

is coupled to bis-Boc-protected l-lysine purification and coupled with 62 using
under standard conditions (T3P, DIPEA) to T3P to give the protected GalNAc cluster,
give 64 followed by amine deprotection which is taken to the global deprotection
under acidic conditions to give 65. step (aq. NaOH and MeOH) without further
FIGURE 21: Synthesis of GalNAc Cluster 58.

purification. GalNAc cluster 58 is purified (DS) CQA can be excluded based on their
using preparative HPLC. specification level in 58 and hence, no
additional fate-of-impurity data are required.
58 is an isolated, well-characterized, and Impurities 68 and 69 are controlled by
stable amorphous material manufactured the process and their levels are below
using a reliable and robust process. Its the reporting limit in 58. It is advisable to
precursors are naturally occurring amino acid ensure that analytical methods can assess
or sugar building blocks, which are readily the configuration of stereocenters, which
available in ton scale in enantiomerically pure are synthetically derived and/or prone
form. To date, the synthesis was carried out to epimerization (eg, anomeric center of
on multi-kg scale with appropriate analytical galactosamine or α-position of lysine amino
controls using standard techniques (HPLC). acids in 70). The nonreactive, noncritical
impurity 71 has been shown to be depleted
It is important to point out preparative HPLC in the downstream purification step. In the
purification of 58 should not be considered context of ICH Q11, “impact on DS quality”
a specific “unit operation” in the sense laid is defined as level above the identification
out in ICH Q11 Q&A. The choice of this threshold. Since the amount present in
technique is motivated rather by the lack GalNAc is significantly lower than the ID
of crystallinity of 58 than by failure of other threshold in the API, an impact can be
purification techniques to provide material of excluded.
sufficient quality.
The specification of 58 is based on
Even at a relatively early stage of current knowledge and will be revised as
development, a number of potential and additional batch history data and/or process
actual impurities in 58 have been identified development data become available before
(FIGURE 22). Potential process-related the manufacture of the commercial drug
impurities will be continuously evaluated substance batches.
during further development. Any new
unknown impurities detected above the Given the breadth of different GalNAc
reporting limit of the analytical method clusters used across the industry and
in future batches will be characterized, the above-mentioned lack of regulatory
and their fate will be investigated in the guidance for oligonucleotides, a general
subsequent processing steps, if needed. recommendation regarding their
acceptability as SM is difficult. Sponsors
The only impurities present in 58 at a level are encouraged to evaluate their GalNAc
>0.10% area are critical impurities 66 (0.27% cluster using a science-based approach
area‡) and 67 (0.15% area§). An impact of founded in the principles set forth in ICH
these two impurities on a drug substance Q11 and provide data demonstrating safety

FIGURE 22: Impurities in GalNAc cluster 58 (impurity motif highlighted).
to patients.** The holistic approach outlined presented illustrates scientific concepts that
in ICH Q11 and associated Q&A provide EPOC member companies feel are relevant
a good framework for this assessment. to support GalNAc and related structures.
Understanding of how the impurity profile
of the SM affects the drug substance quality Solid support
is necessary and should be supported by One important element in most
a sound specification for release testing oligonucleotide syntheses is the synthesis
of GalNAc. The application of GalNAc support (eg, NittoPhaseHL® and †† Primer
conjugates in a commercial setting is a Support 5G®‡‡) and their derivatives
very immature area; therefore, the advice functionalized with an appropriate linker

(eg, UnyLinker®§§ and succinate). Typically,

the first manufacturing step begins after
FIGURE 23: Relationship of
NittoPhase UnyLinker® and
deprotection of the commercial support (eg,
oligonucleotide.
NittoPhase UnyLinker HL350®) followed by
coupling of the first nucleotide. Since the
synthesis resin functioned as a stationary
phase (or has even been described as a
3′-protecting group) for the automated
synthesis, it contributed no material to the
oligonucleotide at the end of the process and
would be described as a noncontributory raw
material (FIGURE 23).
As with other noncontributory raw materials

and reagents, the resin still exerts an effect
on the synthesis and robust performance
during oligonucleotide synthesis under GMP
requires monitoring of appropriate attributes
of the solid support. Sample preparation is
key to examine impurities derived from solid
supports and varies with the support [54], a
few typical attributes are listed (TABLE 8).
Although amidite coupling is generally very

high yielding, there can be situations where
the first coupling can be more challenging
(eg, due to added steric hindrance/
lower reactivity at the secondary alcohol
of UnyLinker). Since oligonucleotide
manufacturing is quite expensive, knowledge
of the yield for the first coupling may be
advantageous, but is not readily ascertained
during manufacture because of the
inability to sample the packed column. One
approach to circumvent this difficulty is to
apply preloaded supports where the first
nucleotide or even 3′-GalNAc is already
coupled to the resin. Such resins are available

from most suppliers and loadings can be performance characteristics of the support
assayed, therefore increasing confidence for may involve derivatization and/or wet
the subsequent manufacture. chemical techniques to examine loading
capacity or impurities derived from the
In these cases, the input support element that is bound to the loaded
contains structural elements of the final support (TABLE 9). There are complexities
oligonucleotide and the noncontributory involved in the validation and routine use
status cannot be applied. As a result, of these types of tests; so their necessity
sponsors may seek to justify the preloaded should be informed by a well-developed
support as an SM. It should be anticipated risk assessment that examines the
that loaded solid supports would require capability of the assays and their ability to
more extensive characterization than control for critical properties of the loaded
unmodified supports. For instance, solid support.
TABLE 8: Standard quality attributes generally monitored for solid supports.

Appearance Visual examination Mixture of powder and aggregates
Sample spectrum conforms to standard

Identification FTIR by ATR or NIR
spectrum
DMT loading Spectrophotometry Measured in μmol/g
Impurities HPLC NMT 0.10% a/a
ATR, Attenuated total reflection spectroscopy; FTIR, Fourier transform infrared spectroscopy; NIR, near-infrared spectroscopy.
TABLE 9: Modified solid support control and characterization.

Physical characteristics Loading capacity Quality attributes (cleavage may be required)
Particle size (microscopy) Use test Identity of loaded support (IR) or loaded entity
(HPLC)
Swell volume (eg, in MeCN) Use test Related substances (HPLC)
Water (KF)
Residual solvents (GC)
Bulk density Testing for stoichiometry Loss on drying
IR, infrared.

CONCLUSIONS Impurity classes in these materials are well

In this white paper, we provide guidance characterized and well understood
on the application of risk-based strategies, The criticality of the various impurity
founded on the collective experience of classes toward impact on API quality are
member companies of the European Pharma well understood (reactive, critical; reactive,
Oligonucleotide Consortium. The purpose noncritical; and nonreactive, noncritical)
is to enable a more uniform approach to the A high level of control is achievable (enabling
justification of various general classes of very stringent purity-focused specifications)
oligonucleotide SMs.
For the related ribonucleoside amidite
As we describe, oligonucleotides are building blocks, such as the MOE
explicitly out of scope with respect to amidites 23–26 and cEt amidites such
ICH Q11. EPOC member companies as 42, similar arguments can be applied.
have, however, sought to apply the Although these materials are of greater
ethos and principles of ICH Q11 in their synthetic complexity and can contain larger
development activities, but in conjunction numbers of impurities than deoxyamidites,
with awareness of the practices and the control strategies employed and
knowledge of oligonucleotide processing. acceptance criteria in specifications are
We recognize not only that step count in its comparable and driven by the same risk-
traditional sense is not a helpful construct based principles. Analogously, with the
for SM justification during oligonucleotide appropriate understanding and control of
processing but also that chromatographic impurities, amidite dimers or blockmers
purification and other downstream such as 56 would follow the ICH Q7 and
operations should serve to provide some Q11 SM principles as part of convergent
mitigation. We describe how a detailed oligonucleotide syntheses.
understanding of the synthetic steps in
both SM and oligonucleotide processing To exemplify oligonucleotide conjugates, a
can support a regulatory SM proposal 5′-GalNAc case study of a linker-modified
addressing the importance and fate of oligonucleotide was used to illustrate two
impurities and providing confidence for further SM types to consider—the linker
patient safety. amidite 57 and the GalNAc cluster 58.
The GalNAc clusters were identified as a
We introduced our position using larger challenge; consequently, a general
deoxyamidites 1–4 as the simplest and most recommendation of their acceptability as SM is
common of the SMs used in oligonucleotide less straightforward; however, we recommend
API manufacture. These clearly satisfy the that sponsors apply the same risk- and science-
criteria for SMs, in line with the guiding based approaches defined in ICH Q11 to
principles of ICH Q11: assess their own particular situation.

This article also gives consideration to REFERENCES:

the solid supports themselves. For the 1. International Conference on Harmonisation
standard resins and the linkers used for of Technical Requirements for Registration of
attachment of the initial nucleotide unit Pharmaceuticals for Human Use, RFPFHU. (2018).
to the resin, these are clearly defined as ICH Q11 development and manufacture of drug
noncontributory raw materials, and the substances—questions and answers (chemical
guidance around key attributes associated entities and biotechnological/biological
with use of these noncontributory raw entities). 2020. https://database.ich.org/sites/
materials is reiterated. In specific cases default/files/Q11_Q%26As_Q%26As.pdf.
where a sponsor may choose to employ Accessed October 3, 2020.
a functionalized solid support where a 2. International Conference on Harmonisation
significant contributory element of the of Technical Requirements for Registration of
API is already present on the support, it Pharmaceuticals for Human Use, RFPFHU.
is possible that further justification as an (2012). ICH Q11 development and manufacture
SM may be required, on a case-by-case of drug substances (chemical entities and
basis, with a focus on the characterization biotechnological/biological entities). https://
and performance characteristics of the database.ich.org/sites/default/files/Q11%20
functionalized support. Guideline.pdf. Accessed October 3, 2020.
3. Faul MM, MD Argentine, M Egan, MC Eriksson,
In general, through the cross-industry Z Ge, F Hicks, WF Kiesman, I Mergelsberg,
examples provided in this article, we JD Orr, et al. (2015). Part 3: designation and
outline a clear and uniform approach to the justification of API starting materials: proposed
designation of SM status for the amidite framework for alignment from an industry
building blocks utilized in oligonucleotide API perspective. Org Process Res Dev 19:915–924.
manufacture, firmly rooted in the principles 4. Tivesten, A, N Akhtar, J Stolee, Y Fillon, R
of ICH Q11. We have also provided insight Orr, ML Hill, L Van den Bergh, T Huybrechts,
and guidance on the designation of more J Muslehiddinoglu, et al. (2018). European
complicated SMs such as those required for Pharma Oligonucleotide Consortium: a move to
oligonucleotide conjugates, as well as setting consolidate oligonucleotide knowledge and share
out a clear position relating to the solid experience within the community. Ther Innov
supports required for synthesis. We believe Regul Sci 52:687–688.
that publication of these arguments will be 5. Muslehiddinoglu J, R Simler, ML Hill, C Mueller,
of great value in enabling both sponsors JHA Amery, L Dixon, A Watson, K Storch, C
and regulatory bodies to achieve greater Gazziola, et al. (2020). Technical considerations
clarity and harmony in justifying the status for use of oligonucleotide solution API. Nucleic
of SMs for use in cGMP oligonucleotide Acid Ther 30:189–197.
manufacture. 6. International Conference on Harmonisation
of Technical Requirements for Registration of

Pharmaceuticals for Human Use, RFPFHU. (2017). N Sharpe, B Andrews, et al. (2017). Impurities

ICH M7(R1) assessment and control of DNA in oligonucleotide drug substances and drug
reactive (mutagenic) impurities in pharmaceuticals products. Nucleic Acid Ther 27:309–322.
to limit potential carcinogenic risk. https:// 12. Sinha ND, J Biernat, J McManus and H Koster.
database.ich.org/sites/default/files/M7_R1_ (1984). Polymer support oligonucleotide
Guideline.pdf. Accessed October 3, 2020. synthesis XVIII: use of beta-cyanoethyl-N,N-
7. International Conference on Harmonisation dialkylamino-/N-morpholino phosphoramidite
of Technical Requirements for Registration of of deoxynucleosides for the synthesis of DNA
Pharmaceuticals for Human Use, RFPFHU. fragments simplifying deprotection and isolation
(2016). ICH Q3C(R6) impurities: guideline of the final product. Nucleic Acids Res 12:4539–
for residual solvents. https://database.ich. 4557.
org/sites/default/files/Q3C-R6_Guideline_ 13. Cheruvallath ZS, H Sasmor, DL Cole and VT
ErrorCorrection_2019_0410_0.pdf. Ravikumar. (2000). Influence of diastereomeric
Accessed October 3, 2020. ratios of deoxyribonucleoside phosphoramidites
8. International Conference on Harmonisation on the synthesis of phosphorothioate
of Technical Requirements for Registration oligonucleotides. Nucleosides Nucleotides Nucleic
of Pharmaceuticals for Human Use, Acids 19:533–543.
RFPFHU. (2019). ICH Q3D(R1) guideline 14. Cheruvallath ZS, TK Wyrzykiewicz,
for elemental impurities. https://database. DL Cole and VT Ravikumar.
ich.org/sites/default/files/Q3D-R1EWG_ (1999). Diastereomeric process control in
Document_Step4_Guideline_2019_0322.pdf. the synthesis of oligodeoxyribonucleotide
Accessed October 3, 2020. phosphorothioates. Nucleosides and
9. International Conference on Harmonisation Nucleotides 18:1191–1194.
of Technical Requirements for Registration of 15. Capaldi CD and AN Scozarri.
Pharmaceuticals for Human Use, RFPFHU. (2000). (2008). Manufacturing and analytical
ICH Q7 good manufacturing guide for active processes for 2′-O-(2-methoxyethyl)-modified
pharmaceutical ingredients. https://database. oligonucleotides. In: Antisense Drug Technology:
ich.org/sites/default/files/Q7%20Guideline.pdf. Principles, Strategies and Applications, 2nd
Accessed October 3, 2020. ed. Crooke ST, ed. CRC Press, Boca Raton, FL,
10. International Conference on Harmonisation pp 401–434.
of Technical Requirements for Registration of 16. Kawasaki AM, MD Casper, SM Freier, EA Lesnik,
Pharmaceuticals for Human Use, RFPFHU. MC Zounes, LL Cummins, C Gonzalez and PD
(2006). Q3A(R2) impurities in new drug Cook. (1993). Uniformly modified 2′-deoxy-2′-
substances. https://database.ich.org/sites/ fluoro phosphorothioate oligonucleotides as
default/files/Q3A%28R2%29%20Guideline.pdf. nuclease-resistant antisense compounds with
Accessed October 3, 2020. high affinity and specificity for RNA targets. J Med
11. Capaldi D, A Teasdale, S Henry, N Akhtar, C Chem 36:831–841.
den Besten, S Gao-Sheridan, M Kretschmer, 17. Majlessi M, NC Nelson and MM Becker.

(1998). Advantages of 2′-O-methyl nucleosides. J Org Chem 80:5337–5343.

oligoribonucleotide probes for detecting RNA 25. Youssefyeh RD, JPH Verheyden and JG Moffatt.
targets. Nucleic Acids Res 26:2224–2229. (1979). 4′-Substituted nucleosides. 4. Synthesis
18. Braasch DA and DR Corey. (2001). Locked nucleic of some 4′-hydroxymethyl nucleosides. J Org
acid (LNA): fine-tuning the recognition of DNA Chem 44:1301–1309.
and RNA. Chem Biol 8:1–7. 26. Koshkin AA, J Fensholdt, HM Pfundheller and
19. Summerton J and D Weller. (1997). Morpholino C Lomholt. (2001). A simplified and efficient
antisense oligomers: design, preparation, and route to 2′-O, 4′-C-methylene-linked bicyclic
properties. Antisense Nucleic Acid Drug Dev 7:187– ribonucleosides (locked nucleic acid). J Org
195. Chem 66:8504–8512.
20. Taj SA, P Gurumurthy, R Suresh, S Narayanan, 27. Collins PM. (1965). The kinetics of the acid
S Suman Meenakshi and YS Sanghvi. catalysed hydrolysis of some isopropylidene
(2003). Process development for the synthesis furanoses. Tetrahedron 21:1809–1815.
of 5′-O-(4,4′-dimethoxytrityl)-N2-isobutyryl- 28. Vorbruggen H, K Krolikiewicz and B Bennua.
2′-O-(2-methoxyethyl)-guanosine—a potential (1981). Nucleoside synthesis with trimethylsilyl
precursor for the second generation antisense triflate and perchlorate as catalysts. Chem
oligonucleotides: an improved process for the Ber 114:1234–1255.
preparation of 2′-O-alkyl-2,6-diaminopurine 29. Wittenburg E. (1964). A new Synthesis of
riboside. Nucleosides Nucleotides Nucleic Nucleosides. Zeitschrift für Chemie 4:303–304.
Acids 22:1327–1330. 30. Eleuteri A, ZS Cheruvallath, DC Capaldi,
21. Ross BS, Q Song and M Han. (2005). Kilo-scale AH Krotz, DL Cole and VT Ravikumar.
synthesis process for 2′-O-(2-methoxyethyl)- (1999). Oligodeoxyribonucleotide
pyrimidine derivatives. Nucleosides Nucleotides phosphorothioates: substantial reduction of
Nucleic Acids 24:815–818. (N-1)-mer content through the use of blockmer
22. Seth PP, G Vasquez, CA Allerson, A Berdeja, H phosphoramidite synthons. Nucleosides
Gaus, GA Kinberger, TP Prakash, MT Migawa, Nucleotides 18:1211–1213.
B Bhat, et al. (2010). Synthesis and biophysical 31. Krotz AH, P Klopchin, DL Cole and VT Ravikumar.
evaluation of 2′,4′-constrained 2′O-methoxyethyl (1997). Phosphorothioate oligonucleotides:
and 2′,4′-constrained 2′O-ethyl nucleic acid largely reduced (n-1)-mer and phosphodiester
analogues. J Org Chem 75:1569–1581. content through the use of dimeric
23. Salinas JC, MT Migawa, BL Merner and S phosphoramidite synthons. Bioorg Med Chem
Hanessian. (2014). Alternative syntheses of Lett 7:73–78.
(S)-cEt-BNA: a key constrained nucleoside 32. Cusack NJ, CB Reese and JHV Boom.
component of bioactive antisense gapmer (1973). Block synthesis of oligonucleotides
sequences. J Org Chem 79:11651–11660. by the phosphotriester approach. Tetrahedron
24. Blade H, D Bradley, L Diorazio, T Evans, BR Lett 7:2209–2212.
Hayter and GP Howell. (2015). Modular synthesis 33. Miura K, K Sawadaishi, H Inoue and E Ohtsuka.
of constrained ethyl (cEt) purine and pyrimidine (1987). Blockwise mechanical synthesis of

oligonucleotides by the phosphoramidite RNAi-mediated gene silencing. J Am Chem

method. Chem Pharm Bull (Tokyo) 35:833–836. Soc 136:16958–16961.
34. Crameri A, ML Hill, SL Lovelock, M Schober, DG 42. Bonora GM, E Ivanova, V Zarytova, B
Tew, and PJ Thomas. (2016). Novel Processes Burcovich and FM Veronese. (1997). Synthesis
for the Production of Oligonucleotides. WO and characterization of high-molecular
2018/011067 A2. mass polyethylene glycol-conjugated
35. Holasová Š, M Mojžíšek, M Bunček, D oligonucleotides. Bioconjug Chem 8:793–797.
Vokurková, H Radilová, M Šafářová, M Červinka 43. Arnold AE, E Malek-Adamian, PU Le, A Meng,
and R Haluza. (2005). Cholesterol conjugated S Martinez-Montero, K Petrecca, MJ Damha
oligonucleotide and LNA: a comparison of cellular and MS Shoichet. (2018). Antibody-antisense
and nuclear uptake by Hep2 cells enhanced by oligonucleotide conjugate downregulates a key
Streptolysin-O. Mol Cell Biochem 276:61–69. gene in glioblastoma stem cells. Mol Ther Nucleic
36. Nishina T, J Numata, K Nishina, K Yoshida-Tanaka, Acids 11:518–527.
K Nitta, W Piao, R Iwata, S Ito, H Kuwahara, et 44. Sun H, X Zhu, PY Lu, RR Rosato, W Tan and Y
al. (2015). Chimeric antisense oligonucleotide Zu. (2014). Oligonucleotide aptamers: new tools
conjugated to alpha-tocopherol. Mol Ther Nucleic for targeted cancer therapy. Mol Ther Nucleic
Acids 4:e220. Acids 3:e182.
37. Nakagawa O, X Ming, L Huang and RL Juliano. 45. Lera M and CJ Hayes. (2001). An olefin cross-
(2010). Targeted intracellular delivery of antisense metathesis approach to vinylphosphonate-linked
oligonucleotides via conjugation with small- nucleic acids. Org Lett 3:2765–2768.
molecule ligands. J Am Chem Soc 132:8848–8849. 46. Lee YC, RR Townsend, MR Hardy, J Lonngren,
38. Habus I, J Xie, RP Iyer, WQ Zhou, LX Shen and J Arnarp, M Haraldsson and H Lonn.
S Agrawal. (1998). A mild and efficient solid- (1983). Binding of synthetic oligosaccharides to
support synthesis of novel oligonucleotide the hepatic Gal/GalNAc lectin. Dependence on fine
conjugates. Bioconjug Chem 9:283–291. structural features. J Biol Chem 258:199–202.
39. Venkatesan N and BH Kim. (2006). Peptide 47. Rensen PC, LA Sliedregt, M Ferns, E Kieviet,
conjugates of oligonucleotides: synthesis and SM van Rossenberg, SH van Leeuwen, TJ van
applications. Chem Rev 106:3712–3761. Berkel and EA Biessen. (2001). Determination
40. Willibald J, J Harder, K Sparrer, KK Conzelmann of the upper size limit for uptake and processing
and T Carell. (2012). Click-modified anandamide of ligands by the asialoglycoprotein receptor
siRNA enables delivery and gene silencing on hepatocytes in vitro and in vivo. J Biol
in neuronal and immune cells. J Am Chem Chem 276:37577–37584.
Soc 134:12330–12333. 48. Prakash TP, J Yu, MT Migawa, GA Kinberger,
41. Nair JK, JL Willoughby, A Chan, K Charisse, WB Wan, ME Ostergaard, RL Carty, G
MR Alam, Q Wang, M Hoekstra, P Kandasamy, Vasquez, A Low, et al. (2016). Comprehensive
AV Kel’in, et al. (2014). Multivalent structure-activity relationship of triantennary
N-acetylgalactosamine-conjugated siRNA N-acetylgalactosamine conjugated antisense
localizes in hepatocytes and elicits robust oligonucleotides for targeted delivery to

hepatocytes. J Med Chem 59:2718–2733. AUTHOR AFFILIATIONS:

49. Huang Y. (2017). Preclinical and clinical 1
Antisense Oligonucleotide Development
advances of GalNAc-decorated nucleic acid and Manufacturing, Biogen, Inc., Cambridge,
therapeutics. Mol Ther Nucleic Acids 6:116–132. Massachusetts, USA.
50. Cedillo I, D Chreng, E Engle, L Chen, AK 2
Process Organic Chemistry, Ionis Pharmaceuticals,
McPherson and AA Rodriguez. (2017). Synthesis Inc., Carlsbad, California, USA.
of 5′-GalNAc-conjugated oligonucleotides: 3
Chemical Development, Pharmaceutical Technology
a comparison of solid and solution-phase & Development, Operations, AstraZeneca,
conjugation strategies. Molecules 22:1356. Macclesfield, United Kingdom.
51. Lill J and R Trussardi. (2016). Processes for 4
API Small Molecule Development, Janssen, Beerse,
the preparation of GalNAc acid derivatives. Belgium.
WO/2017/021385. 5
Early Chemical Development, Pharmaceutical
52. Janas MM, CE Harbison, VK Perry, B Carito, JE Sciences, R&D, AstraZeneca, Macclesfield, United
Sutherland, AK Vaishnaw, ND Keirstead and G Kingdom.
Warner. (2018). The nonclinical safety profile 6
Chemical Development, Product Development
of GalNAc-conjugated RNAi therapeutics in and Supply, GlaxoSmithKline, Stevenage, United
subacute studies. Toxicol Pathol 46:735–745. Kingdom.
53. Husser C, A Brink, M Zell, MB Muller, E Koller 7
Pharmaceutical Division, Small Molecule Technical
and S Schadt. (2017). Identification of GalNAc- Development, Department of Process Chemistry
conjugated antisense oligonucleotide metabolites and Catalysis, F. Hoffmann-LaRoche Ltd., Basel,
using an untargeted and generic approach based Switzerland.
on high resolution mass spectrometry. Anal 8
Global Regulatory Affairs, CMC & Devices, Sanofi,
Chem 89:6821–6826. Bridgewater, New Jersey, USA.
54. Brooks JL, P Olsen, L Chen and AA 9
External Technical Oversight Analytics, F. Hoffmann-
Rodriguez. (2017). UnyLinker dimer La Roche Ltd., Basel, Switzerland.
impurity characterization and process 10
Amidite Manufacturing and Process Development,
improvement. Tetrahedron Lett 58:1050–1052. Thermo Fisher Scientific, Milwaukee, Wisconsin,
USA.
ACKNOWLEDGMENTS 11
Amidite Quality Control and Analytical
The authors thank Ayman Charaf (Sanofi), Rudolf Development, Thermo Fisher Scientific, Milwaukee,
Goeller (Sanofi), Tobias Metzenthin (Sanofi), and Wisconsin, USA.
Mahir Ozdemir (Janssen) for helpful comments and
discussion during the development of this article and AUTHOR DISCLOSURE STATEMENT
the European Pharma Oligonucleotide Consortium W.F.K. is employed by Biogen; A.K.M. is employed
(EPOC) Steering Committee for providing editorial by Ionis Pharmaceuticals; L.J.D. and P.D.S. are
insights. employed by AstraZeneca; L.V.B. is employed by
Janssen; J.M.N. is employed by GlaxoSmithKline;

A.F. and M.M. are employed by F. Hoffmann-La ‡ Following oligonucleotide impurity conventions, this
Roche Ltd.; T.W. is employed by Sanofi; and S.R. level accounts for three different impurities, which
and G.H. are employed by Thermo Fisher Scientific. have a single diethylene glycol unit in one of the three
No competing financial interests exist. spacer units.
§ This level accounts for three different impurities where
FUNDING INFORMATION any one of the three acetyl groups was cleaved.
No funding was received for this article. ** It

is important to understand that the GalNAc cluster
is a targeting moiety, which is cleaved from the active
* EPOC article in preparation. compound during metabolism and does not modulate
† Discussions with deoxyamidite vendors the target. Toxicological coverage of the GalNAc
indicate that typical activators applied include cluster (including related impurities) is achieved
4,5-dicyanoimidazole (DCI), 1H-tetrazole, 5-(ethylthio)- through data generated with the API in animals and
1H-tetrazole (ETT), 5-(benzylthio)-1H-tetrazole (BTT), humans [52,53].
5-(3,5-bis(trifluoromethylphenyl)-1H-tetrazole), †† NittoPhase is a trademark of Nitto Denko Corporation.
5-(4-nitrophenyl)-1H-tetrazole, benzimidazolium triflate, ‡‡ Primer 5G is a trademark of General Electric
and N-methylimidazolium triflate. Activators must companies.
possess an aqueous pKa value comparable to acetic acid §§ UnyLinker is a trademark of Ionis Pharmaceuticals, Inc.
[pKa(aq) = 4.5] and produce a conjugate base that is both
a good nucleophile and leaving group. The choice of
activator used to phosphitylate the nucleoside precursor
does not appear to impact amidite CQAs.

vijay0401/stock.adobe.com
Therapeutic Oligo Quality:

Profiling and Controlling for Raw
Material Impurities
By Indra Pal*, Grant Fernstrum, Chandrashekar Gudise and Gary Held
Minimizing Introduction
and controlling As a growing number of therapeutic oligonucleotide
compounds continue to be introduced into the clinical
upstream single pipeline, and advancing into larger, late phase clinical
critical impurities trials, an increasingly stringent demand is placed upon the
can help reduce phosphoramidite supply chain.
stringency of
With global raw material suppliers scaling up production to meet
oligo purification market demands, heightened concern surrounds the increased
and increase potential for generating novel, as well as previously identified,
overall yields. impurities. The impurities found within the material supply chain
can directly impact the quality of phosphoramidite synthesis,
and thus potentially affect the quality of a therapeutic oligo.

Therefore, an increased focus has been The impurities found within

placed upon controlling incoming raw
materials, understanding the impact to the material supply chain can
phosphoramidite chemistry, including directly impact the quality of
impurity profile, and subsequent effects
phosphoramidite synthesis,
to oligo purity. Thermo Fisher Scientific
Milwaukee continues to undertake a and thus potentially affect the
comprehensive approach to supply quality of a therapeutic oligo.
chain management through partnerships
with raw material suppliers, as well
as customers, to define raw material
specifications, including control of impurity Discussion
levels to satisfy the dynamic quality AMTr-Chloride in DMTr-Chloride generates
requirements. non-critical impurity B. HMTr-Chloride
generates the critical impurities C, D, E and
Recently, the Process Development F. Impurities A, B, C, D and F have been
team performed a deep investigation synthesized and further characterized by 31P
into the quality of an integral raw NMR, HPLC and LC-MS.
material, 4-4’-Dimethoxytrityl Chloride
(I, DMTr-Cl). The team capably identified All the critical impurities will produce
the role and potential deleterious impact deletion sequences during oligo synthesis.
of two potential impurities, 4-Acetoxy- The number and percentage of total
4’-methoxytrityl Chloride (ii, AMTr- impurities in an oligonucleotide due to
Cl) and 4-Hydroxy-4’-methoxytrityl these critical impurities in phosphoramidites
Chloride (iii, HMTr-Cl), in the synthesized will depend on the sequence length. For
phosphoramidite (illustrated in example, a 0.1% critical impurity in the
Synthetic Schemes). phosphoramidite may generate up to 2.0% of
impurities in a 20mer oligo.
Conclusions
SPONSORED CONTENT Thermo Fisher Scientific collaborates
Classification and with our suppliers and therapeutic oligo
characterization manufacturing and developmental
of impurities in
phosphoramidites used partners to offer phosphoramidites that
in making therapeutic reflect the high standards for which our
oligonucleotides TheraPure Phosphoramidtes have been
known since 2002.

Therefore, an increased focus has been placed upon controlling incoming raw mat
Oligo Trends Designation
chemistry, including of Oligo
impurity profile,Oligo
and Profiling
subsequentand effects
OligotoIdentification
oligo purity. Therm
take a comprehensive approach to supply chain management through partnership
to define raw material specifications, including control of impurity levels to satisfy th
Recently, the Process Development team performed a deep investigation into the q
Dimethoxytrityl Chloride (I, DMTr-Cl). The team capably identified the role and pot
Synthetic Schemes: 5’-DMT-N4-Bz-dC- Phosphoramidite (A); 5’-AMT-N4-Bz-dC- Phosphoramidite (B);
4-Acetoxy-4’-methoxytrityl Chloride (ii, AMTr-Cl) and 4-Hydroxy-4’-methoxytrityl Ch
5’-HMT-N4-Bz-dC-di Phosphoramidite (C); HMT Phosphoramidite (D); HMT-di Phosphoramidite (E); and
N4-Bz-dC-3’ramidite (illustrated(F).in Synthetic Schemes, below).
,5’-di Phosphoramidite
Fig. 1: 31P
HMT Phos
Fig. 2: 31P
HMT-N4-
Fig. 4: HPLC com

5’-HMT-N4-Bz-dC
Conclusions: Therm
ufacturing and develo
which our TheraPure
Minimizing and contro

reduce stringency of
Discussion: AMTr-Chloride in DMTr-Chloride generates non-critical impurity B. HMTr-Chloride
generates the critical impurities C, D, E and F. Impurities A, B, C, D and F have been synthe- Our continuous comm
evels to satisfy
stigation into thethe dynamic
quality quality
of anmaterial requirements.
integralsuppliers,
raw material, 4-4’-
hrough partnerships with raw as well as customers,
igation
d
ty the role
levels into
to and the
satisfy
Oligo Trends quality
potential of anofquality
deleterious
the dynamic
Designation integral raw
Oligo impact
Oligo ofmaterial,
requirements. two
Profiling 4-4’-
and potential
Oligo impurities,
Identification
the role and
vestigation intopotential
4’-methoxytrityl the qualitydeleterious
Chloride (iii,
of an integralimpact
HMTr-Cl), in the
raw of two 4-4’-
potential
synthesized
material, impurities
phospho-
methoxytrityl
ied the role andChloride (iii, HMTr-Cl),
potential deleterious impactin of
thetwo synthesized phospho-
potential impurities,
y-4’-methoxytrityl Chloride
FIGURE 1: P NMR (iii,comparison
Chemical shift
31 HMTr-Cl), in the
of 5’-DMT-N synthesized phospho-
-Bz-dC-Phosphoramidite (A) and
4
HMT Phosphoramidite (D).
Fig. 1: 31P NMR Chemical shift comparison of 5’-DMT-N4-Bz-dC-Phosphoramidite (A) and

HMT Phosphoramidite (D)
FIGURE 2: 31P NMR Chemical shift comparison of 5’-AMT-N4-Bz-dC-Phosphoramidite (B) and
5’-HMT-N4-Bz-dC-di
Fig. 1: 31P Phosphoramidite (C).
NMR Chemical shift comparison of 5’-DMT-N4-Bz-dC-Phosphoramidite (A) and
Fig.HMT
1: Phosphoramidite
31 P NMR Chemical(D)shift comparison of 5’-DMT-N -Bz-dC-Phosphoramidite (A) and
4
HMT Phosphoramidite (D)
Fig. 2: 31P NMR Chemical shift comparison of 5’-AMT-N4-Bz-dC-Phosphoramidite (B) and 5’-
HMT-N4-Bz-dC-di Phosphoramidite (C)
Fig. 2: 31P NMR Chemical shift comparison of 5’-AMT-N4-Bz-dC-Phosphoramidite (B) and 5’-
HMT-N4-Bz-dC-di Phosphoramidite (C)
FIGURE 3:Fig.
31 2: 31P NMR Chemical shift comparison
P NMR Chemical shift of N4-Bz-dC-3’ of ,5’-AMT-N 4-Bz-dC-Phosphoramidite (B) and
5’ Phosphoramidite (F). 5’-
HMT-N -Bz-dC-di Phosphoramidite (C)
4
F B A
C
F B A
C
SEPTEMBER 2022 | BIOPHARM INTERNATIONAL D 52 SPONSORED CONTENT
D
FIGURE 4: HPLC comparisons of 5’-DMT-N4-Bz-dC-Phosphoramidite (A), 5’-AMT-N4-Bz-dC-

Phosphoramidite (B), 5’-HMT-N4-Bz-dC-di Phosphoramidite (C), HMT Phosphoramidite (D) and
N4-Bz-dC-3’,5’-di Phosphoramidite (F)
F B A
C
Fig. 4: HPLC comparisons of 5’-DMT-N4-Bz-dC-Phosphoramidite (A), 5’-AMT-N4-Bz-dC-Phosphoramidite (B),

5’-HMT-N4-Bz-dC-di Phosphoramidite (C), HMT Phosphoramidite (D) and N4-Bz-dC-3’,5’-di Phosphoramidite (F)
Minimizing and controlling upstream single

Conclusions: Thermo
critical impurities asFisher Scientific
demonstrated collaborates with our suppliers and therapeutic oligo man
in DMT-
ufacturing
Cl, canand
helpdevelopmental partners
reduce stringency of oligoto offer phosphoramidites SPONSORED
that reflect the high standards f
CONTENT
whichpurification
our TheraPure Phosphoramidtes
and increase overall yields.have been known since 2002.
Phosphoramidite quote and
information request
Our continuous
Minimizing commitment
and controlling to thesingle
upstream pursuit
critical impurities as demonstrated in DMT-Cl, can he
of stringency
reduce deeper control, analytical
of oligo refinement
purification andand
increase overall yields.
quantitation of the phosphoramidite supply Acknowledgement: We thank Amal Audi,
chain will maintain
Our continuous ThermotoFisher
commitment Scientific
the pursuit Syedcontrol,
of deeper Raza, Rick Matusik,refinement
analytical Roger Knight, Amy
and quantitatio
as an industry leader and a sustainable Pasley, and Michael Singer for their support.
of the phosphoramidite supply chain will maintain Thermo Fisher Scientific as an industry leader
and apartner for thepartner
sustainable continued
for growth and safety
the continued growth and safety of oligotherapeutic medicines.
of oligotherapeutic medicines.
Acknowledgement: We thank Amal Audi, Syed Raza, Rick Matusik, Roger Knight, Amy Pasley,
and Michael Singer for their support.
© 2017 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific and its
subsidiaries unless otherwise specified.

nobeastsofierce/stock.adobe.com
Identification and Quantitation of

Oligonucleotides, Impurities, and
Degradation Products
By Haichuan Liu1, Kevin Guo2, Jennifer Sutton1, Keeley Murphy1, Min Du2
The ddMS2 approach Introduction

enables confident Oligonucleotides are synthesized, polymeric sequences of
nucleotides (RNA, DNA, and their analogs) that are increasingly
identification, being developed as direct therapeutic agents against a wide
mapping, and range of disease conditions. They have attracted increasing
relative quantitation attention from the biopharmaceutical industry due to the
of oligonucleotides successes of applying this new modality for the treatment of
rare diseases as well as their potential in treating common
and their impurities diseases and even coronavirus disease-2019 (COVID-19).1-3
in a single Therapeutic oligonucleotides produced by chemical synthesis
experiment. carry various types of product-related impurities, including
deletion sequences (‘shortmers’), addition sequences
(‘longmers’), and the modified full-length species.4-7

The n–x shortmers, the most common impurities, however, it does not provide
impurities present in synthetic base-by-base sequence information and
oligonucleotides, are formed due to failed localization of modifications. Additionally,
base coupling at the 5’ end followed by it is challenging to apply the MS1-
incomplete capping, which may also result based method for the identification
in n–1 impurities with different single of impurities with modifications and
deletions.4 The longmers are mostly the n+1 degradation products. By comparison,
or n+2 species, while the modified impurities an HRAM based ddMS2 method allows
correspond to the full-length product confident identification and mapping of
with modifications on its nucleobases or unmodified and modified oligonucleotides,
phosphorothioate linkages.4 Degradation of as demonstrated in our recent application
synthetic oligonucleotides may introduce note.8 The introduction of powerful
additional species in the products. Although Oligonucleotide Analysis tools in Thermo
chemically modified oligonucleotides are Scientific™ BioPharma Finder™ 4.0 software
normally quite stable, the rate of degradation enables fast processing and annotation
may be affected by their sequences and the of ddMS2 data, as well as a comparative
presence of different stressors.4-6 analysis of multiple raw files simultaneously.
In this application note, the capability of

SPONSORED CONTENT
ddMS2 was extended to the characterization
of impurities and degradation products of
Oligonucleotide analysis:
robust, accurate a synthetic oligonucleotide using a Thermo
characterization of your Scientific™ Orbitrap Exploris™ 240 mass
oligonucleotide therapeutic
spectrometer coupled with a Thermo
Scientific™ Vanquish™ Horizon UHPLC
system, a Thermo Scientific™ DNAPac™ RP
Advanced analytical tools are indispensable column, and BioPharma Finder software.
for the characterization of various The optimization of ddMS2 was performed
oligonucleotide impurities and degradation using a strategy described previously.8 The
products, some of which are present at optimal ddMS2 method was then applied
a very low level. One popular method to identification, mapping, and relative
for oligonucleotide analysis is ion-pair quantitation of oligonucleotide impurities
reversed-phase liquid chromatography in samples purified with different methods
coupled with mass spectrometry (IP- or treated under different stress conditions.
RP LC-MS).4-6 The MS1-based LC-MS The power of HRAM and ddMS2 for
method offers intact mass confirmation structural characterization of high MW
for oligonucleotides and their common impurities will also be highlighted.

Experimental Methanol (P/N A456-1)

Equipment • 1,1,1,3,3,3-Hexafluoro-2-propanol (HFIP),
• Thermo Scientific Orbitrap Exploris 240 >99.0%, TCI Chemicals (P/N H0424)
mass spectrometer (P/N BRE725535) • N,N-Diisopropylethylamine (DIPEA),
• Thermo Scientific Vanquish Horizon >99.0%,
UHPLC system (P/N 5400.0105) TCI Chemicals (P/N D1599)
• Solvent A: 2% (~190 mM) HFIP and
0.1% (~5.7 mM) DIPEA in water (pH 7.8)
• Solvent B: 0.075% (~7.1 mM) HFIP and
0.0375% (~2.1 mM) DIPEA in methanol
Oligonucleotide samples
A lyophilized DNA 21mer (CAG TCG
ATT GTA CTG TAC TTA) was purchased
from Integrated DNA Technologies (IDT)
(Coralville, IA, USA) in the following two
forms by choosing different purification
Software methods available from IDT.
• Thermo Scientific BioPharma Finder 4.0
software (P/N OPTON-30988) A. 21mer purified with standard
desalting, and
Columns B. 21mer purified using HPLC
• Thermo Scientific DNAPac RP column
(4 μm, 2.1 × 50 mm, P/N 088924) No further purification was performed in
house prior to LC-MS analysis. The 21mer
Vials and closures sample described in this application note,
• Thermo Scientific™ 11 mm Autosampler unless otherwise mentioned as HPLC
Snap-It Caps (P/N C4011-50B) purified, is referred to as Sample A.
• Thermo Scientific™ 11 mm Plastic
Crimp/Snap Top Autosampler Vials (P/N A stock solution of 1 mg/mL was prepared
C4011-13) for two 21mer samples of different purity.
• Eppendorf™ DNA LoBind This solution was diluted to 100 µg/mL
Microcentrifuge Tubes (P/N 022431005) before LC-MS analysis.
Solvents Forced degradation

• Thermo Scientific™ UHPLC-MS Water In a heat stress study, the 21mer (1 mg/mL)
(P/N W81) was heated at 80 °C for 1, 2, 4, 6, and 24
• Thermo Scientific™ UHPLC-MS hours. The samples were let to cool to room

temperature and then diluted to 100 µg/mL Mass spectrometry

for LC-MS analysis. The Orbitrap Exploris 240 mass
spectrometer was operated with Thermo
The oxidative stress was performed by Scientific™ Xcalibur™ 4.4 software and
incubating the 21mer (1 mg/mL) with controlled by Orbitrap Exploris Series 2.0
5% H2O2 at room temperature for instrument control software. Instrument
1, 2, 4, 6, and 24 hours. The samples calibration was performed using Thermo
were diluted to 100 µg/mL before Scientific™ Pierce™ FlexMix™ calibration
LC-MS analysis. solution. Data acquisition was performed
in negative ion mode. The ddMS2 methods
Ion-pair reversed-phase liquid were based on templates provided with
chromatography the Orbitrap Exploris instrument control
Oligonucleotide separations were performed software (ICSW). TABLES 1 and 2 list the global
with a DNAPac RP column using a Vanquish and scan parameters of the ddMS2 methods,
Horizon UHPLC system. The autosampler respectively.
was held at 5 °C while the column was
maintained at 60 °C with the column oven
TABLE 1: Ion source properties and
Thermostatting Mode set to Still Air. The
global method settings of ddMS2
solvents were prepared in the original methods
UHPLC solvent bottles to minimize salt
Ion source properties Value
contamination. The LC gradient used in this
study is shown in FIGURE 1. Ion source type H-ESI
Spray voltage Static
Negative ion (V) 2,500
FIGURE 1: LC gradient for Sheath gas (Arb) 40
oligonucleotide separation at a flow Aux gas (Arb) 20
rate of 0.4 mL/min Sweep gas (Arb) 1
Ion transfer tube temp. (°C) 320
Time Vaporizer temp. (°C) 300
(min) %B 100 1
Method settings Value
0 10 80 0.8
Application mode Peptide
Flow (mL/min)
0.5 10
Percent B
60 0.6
4 22 Method duration (min) 15
40 0.4
10.5 35 Expected LC peak width (s) 6
11.7 80 20 0.2
Advanced peak
13.5 80 Checked
0
0 2 4 6 8 10 12 14
0 determination
13.6 10 Time (min)
15 10
Default charge state 1
Internal mass calibration Off

TABLE 2: Full Scan only and Full Scan/ TABLE 2: Full Scan only and Full Scan/
ddMS2 method settings ddMS2 method settings (continued)
Full scan only Full Scan/ddMS2
Full scan Value Full scan Value

Orbitrap resolution 60,000
Scan range (m/z) 550–2,000
RF lens (%) 70
Orbitrap resolution 120,000 AGC target Custom
Scan range (m/z) 550–2,000 Normalized AGC target (%) 100
RF lens (%) 70 Max. injection time mode Custom
Max. injection time (ms) 100
AGC target Custom
Microscans 1
Normalized AGC target (%) 100
Data type Profile
Max. injection time mode Custom Polarity Negative
Max. injection time (ms) 100 Intensity Value
Microscans 1 Intensity threshold 5.0e3
Data type Profile Charge state Value
Polarity Negative Include charge state(s) 1–20

Dynamic exclusion Value
Dynamic exclusion mode Auto
Oligonucleotide analysis in BioPharma

Finder software
The details about ddMS2 method
optimization and data analysis using
BioPharma Finder software were described
in an application note published recently.8
Briefly, the identification and mapping of
ddMS2 Value
21mer and its impurities were performed Isolation window (m/z) 2
in the Oligonucleotide Analysis workflow Isolation offset Off
within BioPharma Finder software. The Collison energy mode Stepped
Collision energy type Normalized
main identification parameters include: Use HCD collision energies (%) 18,20,22i
MS/MS = Use All MS/MS, Mass Accuracy Orbitrap resolution 30,000
= 5 ppm, and minimum confidence = 0.80. Scan range mode Auto
AGC target Standard
The task of “Find All Ions in the Run” under
Max. injection time mode Custom
Component Detection was chosen for Max. injection time (ms) 200
most of the data analysis. The summed Microscans 1
Data type Profile
peak area of all charge states of 21mer was
Data dependent mode Cycle time
obtained by choosing the task of “Find All Time between master scans (s) 1
Masses in the Run”. Stepped NCE was varied from 10-12-14 to 20-22-24 with an
i
incremental value of 1. A stepped NCE of 18-20-22 was used here

for impurity identification, mapping, and relative quantitation in one
experiment, as described in detail below.

Accurate masses of high MW impurities were coupling at the 5’ end. In BioPharma Finder
obtained by performing the deconvolution software, the impurities are assigned as
of MS1 data using the Intact Mass Analysis “x-y” where x and y represent the position
workflow within BioPharma Finder numbers of the first and last bases,
software. The Xtract (Isotopically Resolved) respectively. The “x-y” species may carry -OH
deconvolution algorithm was used together or -PO4H at 5’ end. For example, “A2-A21
with the Sliding Windows algorithm for [A2+Dephosphorylation]” assigned in the
generating source spectra in the retention BioPharma Finder result stands for the n-1
time (RT) range above the full-length product impurity of 21mer with -OH at 5’ end, while
(7.5–9.5 min). The output mass and charge “A2-A21 [Nonspecific]” corresponds to that
state ranges were set 5,000 to 20,000 and 3 with 5’-PO4H. Unless otherwise mentioned,
to 30, respectively. The sequence matching the “n-x” impurities in this application are
mass tolerance was set to 5 ppm. referred to as those with 5’-OH.
Results and discussion

HRAM mass spectrometry of 21mer
SPONSORED CONTENT
The Vanquish Horizon UHPLC system
gRNA analysis: where’s the coupled with a DNAPac RP column provides
best place to start?
robust and reproducible separation of
oligonucleotides and impurities. FIGURE 2
displays a representative mass spectrum of
BioPharma Finder software provides 21mer. A resolution setting of 120,000 at m/z
relative quantitation results of impurities 200 provided baseline separation of the 21mer
in the modification results table under isotopes, thereby allowing accurate mass
the Modification Summary section. The measurement of this intact oligonucleotide
components used for % abundance calculation (~1 ppm). The inset in FIGURE 2 highlights the
can be modified in the components table. low salt adduct (< 2%) under the experimental
The % abundance of the impurities checked conditions employed in this study. To minimize
in the modification result table are visualized the level of salt adducts, it is recommended to
in a modification plot. Some of the plots in carefully choose the ion-paring reagents and
this application note were generated in a LC solvents.
spreadsheet using the export function from
the modification results table. Optimization of the stepped NCE for 21mer
and impurities
Nomenclature of impurities The optimal NCE varies depending on the
In this application note, the conventional oligonucleotide sequence, size, modification,
“n-x” nomenclature is applied to deletion and charge state, and hence needs to be
sequences produced from incomplete base carefully determined. As described previously,8

FIGURE 2: Mass spectrum of 21mer acquired at a resolution setting of 120,000 (at

m/z 200). The inset shows the isotopic envelope of the 4- and its sodium adduct
present at a very low level (< 2%).
4-
9-
10-
8- +Na
4-
m/z
7-
6- 5-
m/z
the ability to compare multiple raw files in The average structural resolution (ASR) is
BioPharma Finder software led to a quick a score that evaluates the completeness of
survey and determination of the optimal fragment coverage of an oligonucleotide. An
stepped NCE value or range for an oligo. The ASR of 1.0 indicates a full fragment coverage
same strategy for optimizing stepped NCE was of an oligonucleotide sequence. The larger
applied to 21mer in this study. the ASR is than 1.0, the less complete the
fragment coverage map is. FIGURE 4 shows
FIGURE 3 displays fragment cover maps and ASR scores measured for eleven ddMS2 raw
annotated MS spectra of the 9- of 21mer
2
files of 21mer (9-) and a selected impurity
acquired using two different stepped NCEs. (n-7, 3-) acquired at stepped NCEs ranging
While a stepped NCE of 10-12-14 provided from 10-12-14 to 20-22-24. While a wide
insufficient fragmentation and hence range of stepped NCEs (11-13-15 to 20-22-
incomplete coverage of 21mer (FIGURE 3A 24) provided complete or nearly complete
AND 3B), a stepped NCE of 14-16-18 gave fragment coverage for the 9- of 21mer
100% coverage for this oligo. (ASR = 1.0-1.2; FIGURE 4A), higher stepped

FIGURE 3: Fragment coverage maps and annotated MS2 spectra of the 9- of

21mer acquired at stepped NCEs of 10-12-14 (a-b) and 14-16-18 (c-d)
(a) (b)
(c) (d)
m/z
NCEs (NCE ≥ 18-20-22) worked better for confirmation of oligonucleotide sequence

the 3- of the n-7 species (ASR = 1.3-1.5; and localization of modifications (if any),
FIGURE 4B). A stepped NCE of 18-20-22 was but it can also confidently identify very low
determined to give good coverage for 21mer abundant impurities that are not discernible
and its impurities; hence it was used for in a chromatogram. FIGURES 5A and 5B display
studies described in the following sections. total ion chromatograms (TICs) of two
However, if large impurities (e.g., n-1) are of 21mer samples (desalting and HPLC) with
interest, it is recommended to use stepped shaded colors highlighting the impurities
NCEs optimal for the full-length product, identified in each sample. The inset of
which is between 12-14-16 and 16-18-20 FIGURE 5A shows selected ion chromatogram
for the 21mer studied here (FIGURE 4A). (SIC) (top) and zoomed TIC of n-16 present
at a very low level (~0.015%, NL: 6.9E+4).
Identification, mapping, and relative However, this impurity was confidently
quantitation of impurities of 21mer purified identified by its high-quality MS2 spectrum
using a standard desalting vs. an HPLC method (FIGURE 5C) and a complete fragment
The ddMS2 method not only allows quick coverage map (FIGURE 5D).

FIGURE 4: Comparative analysis of FIGURE 5: Figure 5. TICs of 21mer

raw files acquired using 11 different purified using (a) standard desalting vs.
stepped NCE settings for (a) 21mer (b) HPLC method with the shaded
(9-) and (b) n-7 (3-). representation of oligonucleotide and
Shaded in green, yellow, and red are ASRs of 1.0-1.2, impurities identified.
1.3-1.5, and >1.5, respectively. The inset in (a) displays the zoom-in of TIC (bottom) and SIC
(top) of n-16, a very low abundant (~0.015%) impurity in
21mer. (c) MS2 spectrum and (d) fragment coverage map of
(a) 21mer, 9- n-16 confirming the identity of this low-level impurity. Note:
Not all impurities identified are shown in Figures 5a and 5b.
(a)
(b) n-7, 3-
(b)
(c)
FIGURE 6 compares the MS area of 21mer and

% abundances of its n-x impurities in two
samples purified with the desalting or HPLC
m/z
method. While the total peak area of the full- (d)
length product (21mer) remained nearly the

same in two samples (FIGURE 6A), a drastic
decrease in % abundance of n-x impurities
(from 2.55% to 1.18%), particularly the
shorter ones, was measured in the HPLC-
purified sample as compared to the desalting

FIGURE 6: (a) Total MS area of all detected charge states (4- to 11-) of 21mer in
two samples purified with the desalting or HPLC method.
(b) Modification plots and (c) % abundance table of n-x impurities of 21mer in the
desalting vs. HPLC samples.
This table was created using the spreadsheet export functionality in BioPharma Finder software.
(a) 21mer 4.54E+08 (b) n-x

4.33E+08
0.8 2.55%
4.00E+08
3.00E+08 0.6
% Abundance
1.18%
MS Area
2.00E+08 0.4
1.00E+08 0.2
0.00E+00 0
Desalting HPLC Desalting HPLC
(c) n-x
Impurity Desalting HPLC Impurity Desalting HPLC

n-1 0.58 0.19 n-11 0.05 0
n-2 0.15 0.30 n-12 0.04 0
n-3 0.14 0.24 n-13 0.21 0
n-4 0.11 0.09 n-14 0.08 0
n-5 0.59 0.15 n-15 0.04 0
n-6 0.10 0.02 n-16 0.01 0
n-7 0.10 0.14 n-17 0.04 0
n-8 0.06 0.05 n-18 0.04 0
n-9 0.14 0 n-19 0.01 0
n-10 0.06 0 Total 2.55 1.18
sample (FIGURE 6B and 6C), as anticipated. A Forced degradations of 21mer

similar trend was observed for n-x impurities Degradation of DNA can occur at high rates
with phosphate at 5’ end (data not shown). in vivo or under physiological conditions.6
These results showcase the ease of using To understand degradation pathways of
BioPharma Finder software for relative synthetic oligonucleotides, the effects of
quantitation of oligonucleotide impurities acid, base, oxidative, thermal, and photolytic
and quick assessment of sample purity. stresses on these molecules have been

extensively studied.4-6 In this work, thermal stress to 24 hours led to nearly complete
and oxidative stresses on 21mer were degradation of 21mer and larger impurities.
investigated.
The ddMS2 results of heat stress samples
FIGURE 7 illustrates that heating 21mer revealed different behaviors for different
for an extended period (up to 24 hours) impurities (FIGURE 8). While the % abundance
resulted in significant degradation of this of n-1 gradually decreased, the ratio of n-5
oligonucleotide. A drastic decrease in the increased at 1 h and then steadily declined
abundance of 21mer was observed between (FIGURE 8A). By contrast, a significant increase
2 to 4 hours of heat stress. At 6 hours, in % abundance of small impurities (e.g.,
abundant degradation products in various n-18) was seen at 24 h (FIGURE 8A). However,
lengths were detected. Extending the heat the main changes upon heat degradation
FIGURE 7: (a) TICs and (b) SIC of the 21mer samples heated at 80 °C for various
time points. (c) Total MS area of all charge states of 21mer in the control and heat
stress samples.
(b)
(a)
100 NL: 1.13E9
100 0h
0h
0
50 100 NL: 8.63E8 1h
0 0
100
NL: 6.80E8
Relative intensity
100 2h
1h
0
50 100 NL: 4.20E8
4h
0 0
100
100 NL: 1.37E8 6h
2h
Relative intensity
0
50 100 NL: 5.79E4
24 h
0 0
100 1 2 3 4 5 6 7 8 9 10
4h RT (min)
50 (c)
0
100
6h
50
MS Area
0
100
24 h
50
0
1 2 3 4 5 6 7 8
RT (min) 0h 1h 2h 4h 6h 24 h

FIGURE 8: % Abundances of (a) selected n-x (5’-OH) impurities and total %

abundances of (b) n-x (5’-PO4H) impurities, (c) base loss products, and (d) products
associated with deprimidination and depurination in the control (0 h) and heat
stress (1–24 h) samples
1 25
n-18 n-13
(a) (c) Base Loss
n-5 n-1
0.8 20
% Abundance
% Abundance
0.6 15
0.4 10
0.2 5
0 0
0h 1h 2h 4h 6h 24 h 0h 1h 2h 4h 6h 24 h
50 25
(b) n-x (5'-PO4H) (d) Depyrimidination
Depurination
40 20
% Abundance
% Abundance
30 15
20 10
10 5
0 0
0h 1h 2h 4h 6h 24 h 0h 1h 2h 4h 6h 24 h
of 21mer were significant increases in % abundance of oxidized 21mer increased

abundances of n-x impurities with 5’-PO4H dramatically in the 24 h course of oxidative
(FIGURE 8B), base loss products (FIGURE stress, as anticipated. Taken together, the
8C), and the impurities associated with results of forced degradation showcased the
depyrimidination or depurination (FIGURE 8D). power of using ddMS2 for identification and
relative quantitation of impurities in a single
In contrast to heat stress, oxidative stress experiment as well as for gaining insight into
did not result in a significant change in the degradation pathways of oligonucleotides.
TICs of 21mer (FIGURE 9A). The level of n-x
impurities increased slightly upon H2O2 High MW impurities of 21mer
treatment but then remained stable over In the 21mer sample purified with the
time (FIGURE 9B). Similarly, the MS area of desalting method, a series of high MW
21mer (FIGURE 9C) did not show a significant impurities were detected at RTs later than
change (FIGURE 9C). However, the relative that of the full-length product. The main

FIGURE 9: (a) TICs of the 21mer control (no H2O2 treatment) and 21mer samples
treated by 5% H2O2 for 1 h, 2 h, 4 h, 6 h, and 24 h;
(b) % Abundances of n-x impurities and MS area of (c) non-oxidized and (d)
oxidized 9- in the control and oxidative stress samples
(a) (c) 9-
10-
No H2O2
1h
Relative Intensity
2h
MS Area
4h
6h
24 h
0h 1h 2h 4h 6h 24 h
RT (min)
8 (b) (d) 9-, oxidized
n-x
n-x (5'-PO4H)
6
% Abundance
MS Area
0
0h 1h 2h 4h 6h 24 h 0h 1h 2h 4h 6h 24 h
high MW species was eluted out at ~9 workflow within BioPharma Finder software
min (FIGURE 10A). Interestingly, these MW revealed a series of peaks whose MWs match
species were completely absent in the perfectly with the total masses of a 21mer
HPLC sample (FIGURE 10B), showing the plus its n-x impurities, i.e., “M + (n-x)” (FIGURE
effectiveness of HPLC for the removal of 10D). The mass errors measured for the eight
these species. The chromatographic peak at largest impurities were all less than 1 ppm
RT ~ 9 min corresponds to two major charge (TABLE 3). It should be mentioned that similar
state envelopes, one of which is displayed high MW species have been observed for
in FIGURE 10C. The HRAM capability of the oligonucleotides with A at 3’ end.9 In that
Orbitrap Exploris 240 mass spectrometer study, a branched structure, where the 3’-OH
allowed accurate mass measurement of these of n-x is linked to the exocyclic amino group
high MW impurities. The deconvolution of the 3’-A of 21mer, was proposed for these
of MS1 data using the Sliding Windows high MW impurities. However, no MS2 data
algorithm in the Intact Mass Analysis of high MW impurities was acquired.

FIGURE 10: Zoom-ins of TICs of 21mer purified with (a) desalting or (b) HPLC
method showing high MW species eluting at ~9 min; (c) mass spectrum of a
high MW species with inset showing the isotopically resolved envelope of its
17-charge state; (d) deconvolution spectrum of high MW species assigned as
“21mer + (n-x)” with x = 1-19
17-
8 (a) Desalting 100 100 17-
Relative Abundance
21mer (c)
80
Relative Abundance
80 60
6
19- 40
8.9 60 15- 20
4 0
40 738.0 738.5 739.0 739.5
13- m/z
20 10-
2 7-
Relative intensity
0
0 600 800 1000 1200 1400 1600 1800
m/z
8 21mer (b) HPLC x=1
100
(d)
21mer + (n-x)
6
Relative Abundance
80 2
(x=1-19)
60
4
40
2 4
20
18 12 10 6
16 14 8
0 0
7.0 7.5 8.0 8.5 9.0 9.5 7000 8000 9000 10,000 11,000 12,000 13,000
RT (min) Mass
In this study, ddMS2 was employed to

TABLE 3: Mass errors for selected high
facilitate structural elucidation of these high
MW species. M = 21mer
MW impurities. A stepped NCE of 11-13-
Species Exp. MW Theo. MW Error (ppm) 15 was used to acquire ddMS2 data of high
M + n-1 12261.033 12261.044 0.9 MW impurities. FIGURE 11 displays fragment
M + n-2 11956.994 11956.998 0.4
coverage maps of the “M + (n-1)” impurity
obtained by searching its ddMS2 data
M + n-3 11667.947 11667.952 0.4
against a 41 nt sequence with 3’- (FIGURE
M + n-4 11354.898 11354.894 0.3
11A) or 5’-end (FIGURE 11B) of n-1 placed to
M + n-5 11050.853 11050.848 0.4 the 3’ end of 21mer. Interestingly, mapping
M + n-6 10721.802 10721.796 0.6 the former sequence did not identify any
M + n-7 10417.755 10417.750 0.5
MS2 fragments (e.g., w1–w20) covering the
3’ end of the 41 nt sequence (FIGURE 11A).
M + n-8 10128.712 10128.703 0.9
By contrast, a series of w and x fragments

FIGURE 11: Fragment coverage maps of (a) “M + 3’ (n-1)” and (b) “M + 5’ (n-1)”, in
which the 3’ and 5’ ends of n-1 are placed to the 3’ end of M (21mer), respectively,
to the sequences for oligonucleotide mapping in BioPharma Finder software
(a) M + 3’ (n-1)
(b) M + 5’ (n-1)
were identified using the 41 nt sequence the linkage occurring on the 5’ side of n-x
where the 5’ end of n-1 was placed to the instead of its 3’ end, as proposed in the
3’ end of 21mer (FIGURE 11B). Similar ddMS2 previous study9. These results show that
results were also observed for other “M ddMS2 can provide additional insight into
+ (n-x)” impurities (data not shown). The the structure of high MW impurities that
two chains (21mer and n-x) likely form may not be obtained by MS1 alone. Further
a similar branched structure described structural elucidation of high MW impurities
before9. However, our ddMS2 data indicate will be the subject of future study.

Conclusion 3. Rossi, J.J., et al. Oligonucleotides and the

COVID-19 pandemic: a perspective. Nucleic Acid
In summary, the ddMS2 approach described
Therapeutics, 2020, 30, 129.
in this application note enables confident
4. Sutton, J.M., et al. Current state of oligonucleotide
identification, mapping, and relative characterization using liquid chromatography-
quantitation of oligonucleotides and their mass spectrometry: insight into critical issues.
impurities in a single experiment. The high Journal of the American Society for Mass
confident identification can be obtained Spectrometry. 2020, 31, 1775.
for very low abundant impurities that are 5. El Zahar, N.M., et al. Chromatographic approaches
for the characterization and quality control of
not discernible at the chromatographic
therapeutic oligonucleotide impurities. Biomedical
level. The three case studies presented
Chromatography, 2018, 32, e4088.
here demonstrate the ability of ddMS2 6. Pourshahian, S. Therapeutic Oligonucleotides,
for quick assessment of oligonucleotide impurities, degradants, and their characterization
purity, comparison of different purification by mass spectrometry. Mass Spectrometry Reviews,
methods, and in-depth understanding of 2019. doi: 10.1002/mas.21615.
forced degradations of oligonucleotides. 7. Capaldi, D., et al. Impurities in oligonucleotide
drug substances and drug products. Nucleic Acid
Additionally, the capabilities of HRAM and
Therapeutics, 2017, 27, 309.
ddMS2 offered by the Orbitrap Exploris 240
8. Liu, H.C., et al. Oligonucleotide mapping using
mass spectrometer provide insight into the BioPharma Finder software. 2020, Thermo
structure of high MW impurities, which Scientific Application Note 73789. http://assets.
cannot be obtained using the MS1 method thermofisher.com/TFS-Assets/CMD/Application-
alone. The new Oligonucleotide Analysis Notes/an-73789-lc-ms-identification-mapping-
workflow introduced in BioPharma Finder oligonucleotides-an73789-en.pdf
9. Kurata, C., et al. Characterization of high
software provides a streamlined process
molecular weight impurities in synthetic
from sequence creation and oligonucleotide
phosphorothioate oligonucleotides. Bioorganic &
mapping to relative quantitation of impurities Medicinal Chemistry Letters, 2006, 16, 607.
and result reviewing. An array of tools for
comparative analysis offered by BioPharma Acknowledgments
Finder software allow easy optimization We would like to acknowledge all the colleagues
of ddMS2 and comparisons of data from who have contributed to the work and reviewed this
different studies, including impurity analysis application note.
and time-course degradation.
References Haichuan Liu, Jennifer Sutton, Keeley Murphy

1. Wang, F., et al. RNA therapeutics on the rise. Thermo Fisher Scientific, San Jose, CA
Nature Reviews Drug Discovery, 2020, 19, 441.
2. Bajan, S., et al. RNA-based therapeutics: from Kevin Guo, Min Du
antisense oligonucleotide to miRNAs. Cells, 2020, Thermo Fisher Scientific, Boston, MA
9, 137.

Comprehensive
oligonucleotide analysis
Oligonucleotide analysis presents multiple analytical and data
processing challenges. Thermo Scientific™ BioPharma Finder™
software helps overcome these hurdles to provide confidence
with comprehensive capabilities.
Bf
Learn more at thermofisher.com/BioPharmaFinder

For Research Use Only. Not for use in diagnostic procedures. © 2022 Thermo Fisher Scientific Inc. All rights reserved. All
trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. FL000759-EN 0522M

Accelerate Oligonucleotide Therapeutics Ebook

Uploaded by

Copyright:

Available Formats

Accelerate Oligonucleotide Therapeutics Ebook

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

Accelerate Oligonucleotide Therapeutics Ebook

Uploaded by

Copyright:

Available Formats

S E PT E M B E R 2 0 2 2

Resources and Tools to

This custom ebook is sponsored by ThermoFisher Scientific,

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 2 SPONSORED CONTENT

What are oligonucleotide What are some trends and challenges

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 3 SPONSORED CONTENT

into siRNA oligonucleotides to improve “Thermo Fisher has been

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 4 SPONSORED CONTENT

maintained in manufacturing via in-process

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 5 SPONSORED CONTENT

innovation, and high quality standards in • Co-innovation: We have partnered with

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 6 SPONSORED CONTENT

World-class manufacturing Customer-focused technologies

Dedicated customer service

• >98%* delivery to customer at promised date

• Standard synthesizer packaging and custom

Continually improved quality Analytical capabilities

• Impurity profiling, control of critical impurities, and • Stability studies

• High-quality TheraPure phosphoramidite offerings

TheraPure DNA phosphoramidites TheraPure locked nucleic acids

TheraPure RNA phosphoramidites Standard DNA phosphoramidites

TheraPure Bz-rA Phosphoramidite 27-1903 dA β-Cyanoethyl Phosphoramidite 27-1730

TheraPure 2ʹ-OMe phosphoramidites Standard RNA phosphoramidites

TheraPure 2ʹ-OMe-Bz-A Phosphoramidite 27-2042 Bz-rA Phosphoramidite 27-1403

TheraPure 2ʹ-MOE phosphoramidites Standard 2ʹ-OMe phosphoramidites

TheraPure 2ʹ-MOE-A Phosphoramidite 27-1019 2ʹ-OMe-PAC-A Phosphoramidite 27-1822

Learn more at thermofisher.com/amidites

Perspectives on the Designation of

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 9 SPONSORED CONTENT

INTRODUCTION therapeutics in a series of technical white

As chemically synthesized active Within EPOC, the Oligonucleotide SMs

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 10 SPONSORED CONTENT

As an end-to-end manufacturing process,

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 11 SPONSORED CONTENT

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 12 SPONSORED CONTENT

FIGURE 2: Typical oligonucleotide synthesis process.

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 13 SPONSORED CONTENT

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 14 SPONSORED CONTENT

that are currently being generally applied to • Oligonucleotides with related

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 15 SPONSORED CONTENT

oligonucleotides, demonstrating that how regulatory guidance can be applied to

As might be expected for such

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 16 SPONSORED CONTENT

understanding of potential and actual

Deoxyamidites are free-flowing,

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 17 SPONSORED CONTENT

solvents used in the reaction are not

Impurities in deoxyamidites can be assigned

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 18 SPONSORED CONTENT

TABLE 1: Origin and control of indicative critical and noncritical impurities in

Impurity type Source Method(s) of control

16 Reaction of DMT-Cl with 3′-OH PNS purification; deoxyamidite

17 Impurity in DMT-Cl reagent with Vendor DMT-Cl specification;

Control of water during PNS

19 Bz protecting group was either PNS purification; deoxyamidite

DMT, 4,4′-dimethoxytrityl; PNS, protected nucleoside.

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 19 SPONSORED CONTENT

SEPTEMBER 2022 | BIOPHARM INTERNATIONAL 20 SPONSORED CONTENT

TABLE 2: Deoxyamidite Impurity Lot History Summary

into siRNA oligonucleotides to improve “Thermo Fisher has been