Microfluidic-assisted engineering of multilayered microcapsules for 3D

stem cell culture

P. Carreras*a, b, M.Gallardob, c, A. Ortizb, c, M.L. Moralesb, c, T. Cedenab, c, I. Gonzaleza and J.
Martinez b, c, d
a
CSIC, Spanish National Research Council, Madrid, 28006, Spain0345; b Hospital 12 Octubre,
Hematology Department, Research institute i+12, Madrid, 28040, Spain; c CNIO, Spanish national
cancer research Centre, Hematological malignancies research unit, Madrid, 28029, Spain; d UCM,
Complutense University Madrid, Medical faculty, Madrid, 28040.

ABSTRACT

Regenerative medicine has increasingly made use of adult stem cells in the last years (1, 2). Micro-engineering a
biomimetic three dimensional structure provides a realistic approach for stem cell niche studies, and further translational
applications. A promising approach for engineering artificial stem cell niches is provided by high-throughput
microfluidic technologies. In this work, a droplet-based microfluidic-assisted encapsulation device for the generation of
multi-layered cellular structures on demand using alginate and Puramatrix is presented. This novel technology is based
on gravity-driven flows, passive mixing principle and a gelation system where the use of a double laminar oil flow where
only one contains the cross-linking agent allows both the uniform gelation of the inner core and the continuous
generation of a stream of cross-linked hydrogel beads. The soft consecutive coating of the inner core with a second and a
third layer without exposing the encapsulated cells to external forces that might reduce their viability represents a
promising technology towards 3D stem cell encapsulation. Furthermore, we demonstrate the suitability of the presented
technology for encapsulation of stem cells by using human Mesenchymal Stem cells (hMScs) and human Hematopoietic
stem cells (hHScs). Preliminary results demonstrate a niche model capable of mid-term culture of primitive hHScs in a
microfluidic environment. Therefore, the presented method could apply for the artificial reconstruction of the stem cell
niche components as an efficient approach to study stem cell behaviour in vitro under controlled conditions, opening a
wide field of potential applications within uTAS for 3D cell culture and tissue engineering applications.
Keywords: Microfluidics, Microencapsulation, Droplet technology, Tissue engineering, Stem cell niche

1. INTRODUCTION
Regenerative medicine has increasingly made use of adult stem cells in the last years due to their therapeutic potential1, 2.
It has been previously reported that three-dimensional (3D) cell culture structures promotes many biological relevant
functions not observed in 2D monolayer cell culture3. Therefore, the best approach to mimic the native stem cell
microenvironment would include microengineering 3D structures able to allocate stem cells.
Approaches for 3D-organized cell culture have included a myriad of scaffold-free systems4, 5 and scaffold supported
systems6,7. Stem cell maintenance studies in these structures have also been performed in suspension or in spinner flask,
resulting in large heterogeneity in spheroid size8, and hanging-drop techniques that provide some control over the
spheroid size but lacks of aggregated generation control9, 10.
Previous work from our group reported several methods used for the generation of multicellular aggregates to
accommodate stem cells11. It was concluded that most of the reported methods to date were lacking of stability on the
template12 or control on the geometry of the cell constructs13. Even when using scaffolds, cell location inside the droplet
could not be determined as most of the studies used single bead culture or co-culture spheroids14, 15. Hence, there is still a
need to develop a technology able to create highly complex, and well-controlled 3D dynamic environment for stem cell
culture in order to mimic in vivo systems.

*[email protected]; phone 0034915618806; fax 0034914117651; www.itefi.csic.es

Microfluidic techniques offers the opportunity to encapsulate cells into templated matrices in a controlled manner. Some
studies have used laminar co-flows for the generation of multi-layered beads16, 17, however these methods were not
suitable for the generation of double layered alginate-alginate beads while requiring a complicated set-up.
Our group already reported a technology to produce a double layered hydrogel bead able to accommodate adult stem
cells where controlled generation of the structure geometry, composition and cell distribution was achieved11. This
technology has been improved by including an additional third layer in order to be able to better mimic the stem cell
niche environment and recapitulate more complex biological systems.
In the presented droplet-based microdevice, hydrogel droplets are produced by hydrodynamic focusing techniques and
gelled by the circulation of the droplet between two laminar flows, one containing the cross-linking agent. Generated
droplets are consecutively coated by two hydrogel layers by synchronizing droplet production rates.
The presented technology permits the rapid extraction of the generated cellular constructs for immediate washing and
culture, limiting the time exposure of the cells to the cross-linking agent. Additionally, it does not requires the use of any
external force to handle and mix the droplets, resulting in a specially mild and suitable method for the encapsulation of
adult stem cells. This three-layered hydrogel bead generator represents a promising technique able to provide optimal
requirements for 3D multicellular stem cell culture and to recapitulate in vitro the in vivo stem cell niche environment
opening a wide range of possibilities within microTAS and tissue engineering.

2. EXPERIMENTAL

2.1 Materials/Chemicals
Mineral oil (Sigma Aldrich) was used as the main oil carrier. Sudan red dye (Sigma Aldrich) was used to color the oil
containing the cross-linking agent and visualize the interface between the two main carrier oil flows. Alginic acid sodium
salt powder (Sigma Aldrich) was used to prepare the alginate solutions. The aqueous phase was composed of deionized
water unless otherwise stated. Puramatrix (Becton Dickinson) solution was prepared at the corresponding concentrations
by sonication of the gel for 30 minutes at 30C. Cresol red dye (Sigma Aldrich) was used as Ph. indicator for the inner
core droplet. Acetic acid (Sigma Aldrich) was used as the cross-linking agent. Nano-calcium carbonate powder was
purchased from Skyspring Nanomaterials and used following the vendor protocol.

2.2 Cell assays

Phosphate Buffered Saline PBS was acquired from Cultek. Calcein-AM (0.1ug/ul) and Dapi 10mg (200ug/ml) were
purchased from Sigma-Aldrich (St. Louis, MO). HLA-DR V450 was purchased from Becton Dickinson, US. Phosphate
Buffered Saline PBS was acquired from Cultek, Spain
Bone Marrow CD34+ cells were purchased from Lonza. Human Mesenchymal stem cells were purchased from DSMZ.
The cell line was cultured in DMEM (Biowest, France) supplemented with 10% fetal bovine serum. Cells were cultured
and used from passages 2–6 in a 37°C humidified incubator with 5% CO2. Trypsin-EDTA (0.25%), phenol red was
acquired from Thermo Fisher Scientific (Waltham, MA).

2.3 Microfabrication
The microfluidic devices were designed using Solidworks CAD (Dassault Systèmes) software. A computer numerical
control (CNC) program was then created using Mastercam (CNC Software Inc.), according to the microfluidic circuit
design and dimensions.
The outline of the device was extracted from a PMMA plastic sheet (MacMaster-Carr) using a Laser engraver (Universal
Laser Systems, VLS 6.30). The plastic cut out was then taken to a CNC machine (HAAS Super Minimill), where the
microfluidic circuit was created based on the Mastercam program. Milling conditions – milling direction, spindle speed,
feed rate – were optimized for surface quality. Finally, fluidic ports were created using a drill press (Ellis).
The PMMA device was sealed with a self-adhesive pouch (Opko diagnosis). Main channel dimensions are displayed in
Figure 1 The device was connected to open barrels of 10 ml disposable plastic syringes (BD biosciences) via silicone
tubing (ID 1.1mm O.D. 2.16mm) (Fisher scientific). The driving force used for the experiments was gravity, so liquid
pressure on the microdevice was controlled by the heights of the open syringes. Pieces of polytetrafluoroethylene (PTFE)
tubing (OD 2 mm) (Cole Parmer) were used to interface the reservoirs to the chip.

2.4 Microscopy
Fluorescence images were acquired using inverted laser scanning confocal microscopy (LSM510 META, Axiovert RT
200M, Zeiss, Germany) equipped with an environmental chamber maintaining the cells under physiological condition,
using the laser lines 405 nm and 488 nm.
For all other experiments, a Motic Stereoscope (Motic) and Moticam 2.0 software (Motic) were used. Droplet sizes were
analyzed on screen using Image J software.

2.5 Research methodology

The presented device schematic and operational process is depicted in Figure 1.

Figure 1: a) Schematic of the three layered hydrogel bead generator and snapshot of a recorded video comprising (from top to
bottom), the hydrodynamic focusing generation of the inner core, the passive mixing process for the formation of the first layer
and the passive mixing process for the outer layer b) CAD design of the manufactured platform.

First, alginate droplets containing calcium carbonate nanopowder are generated by hydrodynamic focusing using gravity
to drive pressure. The droplets then travel through a meandering channel containing two laminar oil flows. By the
addition of acetic acid to the second laminar oil flow continuous phase, calcium ions are released after droplet formation
by the diffusion of the acetic acid across the oil interface. The platform is designed so that the second oil flow containing
the cross-linking agent has a double function for both cross-linking and adjusting the droplet production rates of the
generated droplet without affecting the gelation process.
The so generated droplets are then passed by a second inlet were a stream of second sample layer alginate droplets are
generated in a synchronous way. Hence, when the inner core droplet reaches the area of the second droplet production,
the two droplets (one gelated and another ungelated) passively mix along the meander, generating a double layered
hydrogel bead. Then, the generated double layered droplet continues travelling along the mender, leading to the gelation
of the outer layer. When the double layered gelated bead reaches the third inlet layer where a third droplet is generated, a
three layered structure is generated again using the same mixing principle.
The passive nature of the mixing principle underlying this construction process enhances the suitability of these resulting
structures for accommodating stem cells.
Preliminary experiments demonstrated that both the gelation process and the three layered bead generation technique are
suitable for the encapsulation of living cells by using human Mesenchymal stem cells (hMSCs) and human
Hematopoietic stem cells (hHSCs). Double and triple cell laden structures were generated, washed and cultured in order
to asses encapsulated cell viability.
This mild three layered bead generation technique allowed us to encapsulate the cells in the middle layer with high
viability using alginate (sample 1)- alginate (sample 2)- alginate (sample 3) scaffolds and alginate (sample 1)- alginate
(sample 2)- Puramatrix (sample 3) scaffolds.

3. RESULTS AND DISCUSSION

Premixed aqueous samples containing alginate (1,6 wt %) and Calcium carbonate nanopowder (1 wt %) were prepared
as sample 1, 2 and 3. When using alginate as Sample 3, Puramatrix (0,3% v/v) was prepared following the vendor
instructions. The second oil was prepared by adding glacial acetic acid to a pre-dyed Sudan red mineral oil (0.5% v/v).

First characterization studies were run by varying the height of the oil 2 and using alginate as a scaffold in all layers.
Adjustments of the samples flow rates were performed so that the inner core produced was synchronized with the second
and third droplet produced. Measurements of the three droplets constituent layers production rate were performed as well
as its mixing efficiency by areas (droplet 1 with droplet 2 and droplet 2 with droplet 3). Also the total mixing efficiency
of the device resulting from the measurements of the number of droplet 1 droplet 2 and droplet 3 merged were measured
respect droplet 2 production rate and droplet 3 production rate. These preliminary experiments showed the capacity of
the geometry for the suitable synchronization of the flow rates so that the desired three layered bead can be produced.

Figure 2 a) Preliminary characterization of the three layered device. Operational conditions for a inner layer, middle and outer
layer 1,6% wt alginate, b) CFC imaging of a three layered bead containing fluorescent latex beads on the middle layer.
A set of experiments using fluorescent latex beads to be added to the middle layer sample was also performed. The
objective of this trial was both to assess the stability of the layers and the approximate shape of the collected beads.
Figure 2b shows pictures taken in a confocal microscope of three layered beads containing latex fluorescent beads on
sample 2 (middle layer).

Further characterization was performed using Puramatrix on layer 3 of the structure. Experimental procedure and
parameter analysis was performed as described before. In order to be able to visualize the layers with the on-site
experimental setup, cresol red dye was added to the inner core, while keeping non-coloured alginate for the middle layer
and Puramatrix (optically transparent) for the outer layer.

Figure 3 shows the results of this characterization. These results showed that higher droplet production rates favors the
three layered bead production. Optimal conditions could be determined for which 100% of the produced inner core
droplets were coated by a second and a third layer of hydrogel sample.

Figure 3 a) Characterization for the three layered device operational conditions using low droplet flow rates ( droplet production
rates below 38 droplets/min) and b) high flow rates (droplet production rates below 50 droplets/min) c) Comparison between the
total number of three layered beads for droplet production rates in a) ( left) and b) (right).

Shapes of the final structures varied between a concentric shape or a non-concentric structures due to the slight
differences that occurred on the scaffold gelation points at the time of the passive mixing. The following results on
viability and cell encapsulation capacity, suggested that they were not affected by the shape of the resulting final three
layered structure, while still being a three layered spheroid.

3.1 Cell encapsulation and culture

The suitability of the presented method for cell encapsulation was demonstrated by accommodating living human
Mesenchymal stem cells (hMSCs) into the middle layer.
Prior to resuspension in PBS and mixture in premixed alginate emulsification, human Mesenchymal stem cells were
routinely cultured and trypsinized. We used two cell density ranges, 3 and 15× 106 cells/mL. The generated cell laden
structures were collected and washed to remove residual acid. Calcein-AM (0.1 μM) which fluoresces green upon the
reaction of intracellular esterase and stains live cells; and DAPI 0,1 M which binds to the DNA of dead membrane
compromised cells were employed to visualize the distribution of living and dead cells right after encapsulation.

Beads containing 15x10e6 cells /ml hMSCs on the middle layer were produced, collected and observed under confocal
microscopy in order to assess its stability (Figure 4).
Results of the life dead assay demonstrated that the encapsulated cells remained with high viability, therefore
demonstrating the device suitability for stem cell encapsulation.

Figure 4: a) : Life dead assay on a three layered bead using 15x10e6 hMScs/ml in alginate 1,6% wt. for the middle core, and empty
alginate 1,6% wt. for the middle and the outer layer. From top left to bottom right, four different configurations of shape for the
three layered beads b) Z-stack confocal imaging of a life dead assay on a three layered bead using 15x10e6 hMScs/ml in alginate
1,6% wt. for the middle core, and empty alginate 1,6% wt. for the inner and the outer layer.

An additional experiment was performed encapsulating 3x10e6 cells/ ml hMSC in both the inner and the outer layer to
complete the viability assay in the three layers (Figure 5).The material on the outer layer was set to be Puramatrix 0,3 %
v/v in order to assess the viability of the middle layer encapsulated cells in alginate with Puramatrix as an outer coating
layer. The calcein-DAPI assay was performed following the same protocol as before, demonstrating high cell viability
after encapsulation.

Figure 5: Life dead assay using 3x10e6 hMScs/ml in alginate 1,6% wt for the inner core, empty alginate 1,6% wt for the middle
layer and 3x10e6 hMScs/ml in 0,3% v/v Puramatrix on the outer layer. a) bright field microscopy image b) Life dead assay.

From the operational point of view, it was concluded that as long as both oil flow rates are initially equilibrated to co-
flow equally along the main meander, heights of sample 1, 2 and 3 can be fine-tuned to achieve a situation of close to
full synchronization to obtain a three layered bead generation.
Longer crosslinking times, i.e. extended exposure of the encapsulated cells to the acidic environment on the meander or
at collection, or higher concentrations of acetic acid on the second oil would result in a lower cell viability (data not
shown). However the presented technique able to minimize the acid exposure together with the low mechanical stresses
that cells experiment to be embedded in this three layered construct, constructs presented high cell viability after
encapsulation. Furthermore, the cells were clearly compartmentalized in a middle layer and remained; the shape of the
three layered droplets remained unchanged after gelation and no leakage of cells after encapsulation was observed.
Considering that cell viability could be increased by adjustment of the hydrogel’s mechanical properties, this method
could represent a technology not only for living cell encapsulation but also for long term culturing of cells in highly
controlled microgel matrices in the near future.

In order to test this technology as a tool for stem cell maintenance and culture, human Hematopoietic Stem cells hHSCs
were encapsulated in the inner core of the three layered device. For this experiment, 5x10e6 hMScs/ml were in alginate
1,6% wt. empty alginate was used for the middle core, and Puramatrix 0,3% was used as the outer layer. Cells were
enclosed in the inner core of the structure as it was considered the most critical condition due to the double cover of
hydrogel. The generated structures were produced as described before.

Collected beads were exposed for 20 minutes to DMEM containing mineral oil and introduced in the incubator in order
to accelerate the outer layer gelation. All the samples were afterwards washed with PBS twice before adding culture
media.Generated structures were cultured and imaged at 24/48/72 hours and then over a week using DMEM media
10%FBS.No media change was performed during the 7 days of cell culture.
FITC CD34+ which signals hHSCs green and DAPI 0,1 M which binds to the DNA of dead membrane compromised
cells were employed to visualize the distribution of hHSCS living and dead cells right after 24,72 and 1 week of cell
culture.

Beads containing hHSCs in the inner layer were produced, collected and observed under confocal microscopy in order to
assess its viability (Figure 6).

Figure 6: a) : FITC CD34+-DAPI assay on a three layered bead using 5x10e6 hHScs/ml in alginate 1,6% wt. for the middle core,
and empty alginate 1,6% wt. for the middle and PM for the outer layer. From left to right, three samples of beads after 24,72 and 1
week of cell culture-

Results of the FITC CD34+-DAPI assay demonstrated that the encapsulated cells could be cultured after being generated
in the presented device at least for a week and morevoer remained with high viability without the use of any additive on
the media or supporting cell, therefore demonstrating the suitability of the presented technology for human stem cell
encapsulation.
 4. CONCLUSIONS

In conclusion, we present a droplet-based microfluidic platform for the generation of three layered hydrogel beads
suitable for adult stem cell encapsulation. This novel technology is based on a passive mixing principle and a gelation
system where use of a double laminar oil flow where only one contains the cross-linking agent allows both the uniform
gelation of the inner core and the continuous generation of a stream of cross-linked hydrogel beads. The soft consecutive
coating of the inner core with a second and a third layer without exposing the encapsulated cells to external forces that
might reduce their viability represents a promising technology towards 3D stem cell encapsulation.

Furthermore, we demonstrate the suitability of the presented technology for encapsulation of stem cells by using human
Mesenchymal Stem cells (hMScs) in the middle layer. Moreover, hHSCS were also encapsulated in the inner core as a
text proof of its application as a method for the artificial reconstruction of the hematopoietic stem cell niche components.
The presented device represents an efficient attempt to engineer stem cell niches ex vivo in a multilayered three
dimensional matrix, and a suitable approach to study stem cell behavior in vitro under controlled conditions. Therefore,
the presented device is a valuable and unique tool towards understanding stem cell behavior in 3-dimensional
environments.

ACKNOWLEDGEMENTS

This work was supported by the European Commission H2020 Marie Curie Research Grants Scheme MSCA-IF-GF
(705163).

REFERENCES

[1] Singh, R., 2016, Stem Cells. Applications in Regenerative Medicine and Disease Therapeutics. Int J Cell Biol.
2016; 2016: 6940283
[2] Daley, G.Q. and Scadden, D.T., 2008. Prospects for stem cell-based therapy. Cell, 2008, 132, 544-548.
[3] Edmonson, R., Jenkins Broglie, J., Adcock F.A., and Yang, L. Three-Dimensional Cell Culture Systems and
Their Applications in Drug Discovery and Cell-Based Biosensors. Assay Drug Dev Technol. 2014 May 1;
12(4): 207–218.
[4] Lin R.Z. and Chang H.Y. Recent advances in three-dimensional multicellular spheroid culture for biomedical
research. Biotechnol J. 2008 ;3(9-10):1172-84
[5] Baraniak P.R. and McDevitt T.C. Scaffold-free culture of mesenchymal stem cell spheroids in suspension
preserves multilineage potential. Cell Tissue Res. 2012 Mar; 347(3): 701–711
[6] Chen A.A., Tsang V.L., Albrecht D.R., Bhatia SN. 3-D Fabrication Technology for Tissue Engineering.
BioMEMS and Biomedical Nanotechnology, Vol III: Therapeutic Micro/Nanotechnology.Springer.2007
[7] Faulkner-Jones A., Greenhough S., King J.A., Gardner J., Courtney A., Shu W., Development of a valve-based
cell printer for the formation of human embryonic stem cell spheroid aggregates Biofabrication., 2013
5(1):015013.
[8] Lee E.J., Park S.J., Kang S.K., Kim G.H., Kang H.J, Lee S.W., Jeon H.B., Kim H.S. Spherical bullet formation
via E-cadherin promotes therapeutic potency of mesenchymal stem cells derived from human umbilical cord
blood for myocardial infarction. Mol Ther. 2012; 20(7):1424-33.
[9] Kurosawa H. Methods for inducing embryoid body formation: in vitro differentiation system of embryonic stem
cells. J Biosci Bioeng, 2007,103: 389
[10] Banerjee M. and Bhonde R.R. Application of hanging drop technique for stem cell differentiation and
cytotoxicity studies. Cytotechnology. 2006 May; 51(1): 1–5
[11] Carreras, P. Cortrim, R., Gallardo,M., Ortiz, A., J. Martinez, S.K. Sia. Microengineering double layer hydrogel
structures for the recapitulation of the hematopoietic stem cell niche. Science bulletin, in
press,10.1016/j.scib.2018.09.003 2018
[12] Xu, S.Q., Nie, Z.H., Seo,M., Lewis, P., Kumacheva, E., Stone, H.A., Garstecki, P., Weibel, D.B., Gitlin I. and
Whitesides, G.M. Generation of monodisperse particles by using microfluidics: Control over size, shape, and
composition. Angew. Chem., Int. Ed., 2005, 44, 724–728.
[13] Markx, G. H., Carney, L., Littlefair, M., Sebastian, A., Buckle, A. M., Recreating the hematon: microfabrication
of artificial haematopoietic stem cell microniches in vitro using dielectrophoresis. Biomed Microdevices,
2009;11(1):143-50.
[14] Tumarkin E, Tzadu L, Csaszar E, et al., High-throughput combinatorial cell co-culture using microfluidics.
Integr Biol, 2011; 3: 653–6622011.
[15] Sakai S., Ito S., Inagaki H., Hirose K., Matsuyama T., Taya M, Kawakami K. akai et al. Cell-enclosing gelatin-
based microcapsule production for tissue engineering using a microfluidic flow-focusing system.
Biomicrofluidics, 2011 30;5(1):13402.
[16] Rondeau, E., Cooper-White J. Formation of multilayered biopolymer microcapsules and microparticles in a
multiphase microfluidic flow Biomicrofluidics, 2012, 6, 024125.
[17] Li, W., Zhand, L., Ge, X., Xu B., Zhang W., Qu L., Choi C., Xu j:, Zhang A., Lee H., Weitz D. Microfluidic
fabrication of microparticles for biomedical applications. Chem. Soc. Rev., 2018,47, 5646-5683