Cytoprotection of human progenitor and stem cells through encapsulation in alginate templated, dual crosslinked silk and silk - gelatin composite hydrogel microbeads

Abstract

Susceptibility of mammalian cells against harsh processing conditions limit their use in cell transplantation and tissue engineering applications. Besides modulation of the cell microenvironment, encapsulation of mammalian cells within hydrogel microbeads attract attention for cytoprotection through physical isolation of the encapsulated cells. The hydrogel formulations used for cell microencapsulation are largely dominated by ionically crosslinked alginate, which suffer from low structural stability under physiological culture conditions and poor cell-matrix interactions. Here we demonstrate the fabrication of alginate templated silk and silk/gelatin composite hydrogel microspheres with permanent or on-demand cleavable enzymatic crosslinks using simple and cost-effective centrifugation-based droplet processing. Composite microbeads display structural stability under ion exchange conditions with improved mechanical properties compared to ionically crosslinked alginate microspheres. Human mesenchymal stem and neural progenitor cells are successfully encapsulated in the composite beads and protected against environmental factors, including exposure to polycations, extracellular acidosis, apoptotic cytokines, UV irradiation, anoikis, immune recognition, and particularly mechanical stress. Microbeads preserve viability, growth, and differentiation of encapsulated stem and progenitor cells after extrusion in viscous polyethylene oxide (PEO) solution through a 27-gauge fine needle, suggesting potential applications in injection-based delivery and 3D bioprinting of mammalian cells with higher success rates.

Graphical Abstract

Alginate templated, dual crosslinked on-demand degradable silk and silk/gelatin composite hydrogel microbeads are fabricated by centrifugation-based droplet formation. Composite microbeads display size-selective permeability, and better mechanical properties and UV absorption than alginate-only controls. Human stem and progenitor cells are successfully encapsulated in composite microbeads with minimum impact on viability, and protected against antibody recognition, polycations, apoptotic cytokine, extracellular acidosis, UV-C radiation, and hydrodynamic forces during capillary flow.

1. Introduction

The last two decades of regenerative medicine studies have made extensive use of stem and progenitor cells for transplantation and tissue engineering because of their self-renewing capacity and multilineage differentiation ^[1]. Mesenchymal stem cells (MSCs) hold promise in ex vivo engineering of tissues and organs ^[2] as well as in cell therapy for immunomodulation to reduce inflammation and promote tissue repair in rheumatoid arthritis, spinal cord injuries, myocardial infarctions, Crohn’s Disease, or cartilage and meniscus repairs, among other needs ^[3]. Similarly, neural progenitor cells (NPCs) have been utilized for neural tissue engineering ^[4] and cell transplantation therapies for diseases affecting the central and peripheral nervous systems including spinal cord injuries ^[5], brain tumors ^[6] and Parkinson’s disease ^[7]. Use of mammalian cells involves large numbers of cells cultured and processed ex vivo and then delivered in vivo, subjecting cells to varying degrees of environmental stress ^[8], including polycations ^[9], UV irradiation ^[10], attachment deprivation (anoikis) ^[11], non-physiological pH ^[12], or inflammatory cytokines ^[13]. Moreover, mammalian cells lack a protective cell wall or exoskeleton and are surrounded only by a fragile lipid membrane, rendering them highly susceptible to mechanical stress ^[14] applied in bioreactors ^[15], or during centrifugation ^[16], delivery through capillaries ^[17], 3D bioprinting ^[18], injection-based transplantation ^[19], and blood circulation in vivo ^[20]. Extrusion in a viscous carrier gel or bioink solution could subject cells to extremes of mechanical stress that can reach up to hundreds of kPa, which impact cell fate ^[21] and can cause necrotic death by disrupting the cell membrane ^{[19, 22]}. Impaired cell viability also reduce production yield ^[23] or the rate of therapeutic success ^{[21b, 24]} and may cause acute inflammatory responses at the implantation site ^[25]. Thus, it is essential to provide protection against environmental stresses to preserve cell viability and function during ex vivo processing and delivery.

Encapsulation of mammalian cells within microbeads of synthetic or natural polymeric matrices has been investigated to engineer the cell microenvironment and to provide cytoprotection by acting as a barrier against physical and biochemical factors ^[26]. Hydrogel matrices have been of particular interest in cell microencapsulation studies due to their tunable mechanical or chemical properties and porosity that mimics microenvironments in native tissues and provides cytoprotection ^[26]. Microscale hydrogels provide more efficient transfer of nutrients and waste products, reduce risk of hypoxia through higher surface to volume ratios ^[27], and help with cost efficiency due to the lower amounts of materials needed for fabrication ^[28]. The polysaccharide alginate has been the most common biopolymer ^[29] used for the fabrication of monodisperse hydrogel microspheres using microfluidics ^[30] and electrostatic ^[31] or pressure-induced droplet generators ^[32] by taking advantage of its rapid ionic crosslinking in aqueous calcium solution ^[33]. Despite the simplicity and cytocompatibility of the crosslinking method, alginate gels rapidly lose structural stability under metabolic activity due to ongoing ion exchange mechanisms ^[34]. Moreover, alginate lacks cell recognition motifs, limiting cell-matrix interactions, survival, and growth of the encapsulated cells ^[35]. Other biopolymers most commonly used for cell encapsulation include thermosetting agarose ^[36] and gelatin ^[37], which also suffer from poor structural stability under physiological conditions due to their rapid sol-gel transition above 30°C ^[38]. Enzymatic crosslinking of hydrogel microspheres through horseradish peroxidase (HRP) / hydrogen peroxide (H₂O₂) reactions offers a cytocompatible alternative with enhanced structural stability and mechanical properties ^[39]. The approach has been demonstrated with several phenol-conjugated biopolymers including alginate ^[40], gelatin ^[37], dextran ^[28] and hyaluronic acid (HA) ^[41] for the fabrication of cell encapsulating hydrogel microbeads. Enzymatically crosslinked polysaccharides and synthetic polymers still suffer from limited cell-matrix interactions ^[42], while extracellular matrix-derived polymers undergo rapid enzymatic degradation ^[43]. Therefore, it is important to develop hydrogel formulations for the fabrication of hydrogel microbeads with structural stability under physiological conditions, robust mechanical properties, and high bioactivity for encapsulation and cytoprotection of mammalian cells.

Silk fibroin (SF) extracted from the cocoons of domesticated silkworm Bombyx mori is a biocompatible structural protein that displays extraordinary mechanical properties and slow degradation owing to intrinsic self-assembly of the repetitive hydrophobic domains into crystalline β-sheets ^[44], making it an excellent candidate for microencapsulation of cells for cytoprotection and preservation. Fabrication of SF-based microspheres has been demonstrated before through physical crosslinking by micro-emulsification in a diffusing organic phase ^[45], electrospraying into an alcohol bath ^[46], or co-flow microfluidics using an oil phase with methanol ^[47] or aqueous polyvinyl alcohol solution ^[48] as the continuous phase. These approaches, however, were limited to drug delivery applications and have not been utilized in cell encapsulation due to potential cytotoxicity of the crosslinking methods or the small bead diameter below the average size of mammalian cells. Ionically crosslinked silk/alginate composite microbeads were employed for cell encapsulation, but they were still sensitive to disintegration by ion exchange ^[49]. In addition to physical crosslinking through self-assembly, aqueous SF solutions also undergo enzymatic crosslinking via tyrosine side chains into highly elastic cytocompatible hydrogels ^[50], enabling covalent crosslinking with phenol-conjugated biopolymers such as hyaluronic acid ^[51] and gelatin ^[52] to obtain composite hydrogels with improved properties. Composites hydrogels of silk with phenol-enriched gelatin, for example, displayed improved gelation kinetics and bioactivity ^[52]. Moreover, chemical modification of SF through carbodiimide chemistry ^[53] was employed to introduce additional phenol groups on SF to further modulate the gelation kinetics and mechanical properties of enzymatically crosslinked hydrogels ^[52]. Here we hypothesized that unique properties of silk and silk/gelatin hydrogel formulations could be utilized in the fabrication of composite microbeads for the encapsulation and cytoprotection of mammalian cells.

In this study, we employed alginate templating of silk and silk/gelatin blends for the fabrication of dual crosslinked hydrogel microbeads. The rapid ionic crosslinking of alginate was exploited as a template for the enzymatic ^[54] or photo-crosslinking of silk hydrogels ^[55] before, but not for cell encapsulating microbeads to our knowledge. Centrifugation-based droplet formation, a method that was developed as a simple approach for the fabrication of ionically crosslinked alginate microbeads with low polydispersity without using droplet generators, microfluidic setups, and oil phases ^[56], was adapted for alginate templating of enzymatically crosslinked microbeads. For the on-demand release of encapsulated cells, alginate was chemically conjugated with phenol groups with reduction-sensitive linkers and incorporated into the enzymatically crosslinked polymer network, allowing for disintegration of microbeads on-demand by mild reduction. Alginate/silk and alginate/silk/gelatin composite microbeads did not disintegrate completely under ion exchange and displayed better mechanical properties compared to ionically crosslinked alginate-only controls. Human NPCs and MSCs were successfully encapsulated without a significant reduction in cell viability. Composite microbeads provided cytoprotection against anoikis, high molecular weight (MW) polyethyleneimine (PEI), acidic pH, tumor necrosis factor (TNF)-α and UV-C irradiation. Microencapsulated hMSCs and hNPCs survived extrusion through a 27-gauge fine needle with a diameter of 210 μm, and their proliferation and differentiation capacities were preserved. Our results suggest that the microbead formulations introduced here could be utilized for cytoprotection of mammalian cells for processing, preservation, and delivery.

3. Results

3.1. Fabrication of alginate-templated composite microbeads for cell encapsulation

Alginate templating was adapted to fabricate enzymatically crosslinked silk or silk/gelatin hydrogel microbeads via centrifugation-based droplet formation strategy (Figure 1A) using common lab consumables (Supplementary Figure S1) and different pre-hydrogel formulations (Table 1). Centrifugation of Alg/SF/G-TA solution with HRP at 2250 g into a CaCl₂/H₂O₂ crosslinking bath resulted in formation of spherical hydrogel microbeads, which disintegrated in sodium citrate solution (data not shown). Microbeads prepared using the same formulation with tyramine-conjugated SF (SF-TA) instead of SF, on the other hand, (Figure S2A-i) did not dissolve upon sodium citrate treatment but their opacity increased, and average diameter decreased from 159 ± 7 um to 146 ± 8 um (Figure S2A-ii). Live/dead confocal micrographs of hNPCs and hMSCs encapsulated in composite microbeads (Figure S2B) revealed that most of the cells survived the treatment. Both cell types gradually migrated out and were able to proliferate both on the outer surfaces of the microbeads and on the culture plate.

Figure 1.

Fabrication of alginate-templated, dual crosslinked silk and silk/gelatin composite hydrogel microbeads. Schematics illustrating (A) fabrication of hydrogel microbeads via centrifugation-based droplet formation and (B) 2-step chemical modification of alginate with disulfide-linked phenol (DSP) through carbodiimide coupling of cystamine and 4-hydroxyphenylacetic acid (4-HPAA). (C) Brightfield micrographs of Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA hydrogel microbeads (i) before or (ii) after treatment with sodium citrate or (iii) with a mixture of sodium citrate and glutathione (GSH). Scale bars: 200 μm.

Table 1.

Composition of hydrogel formulations used for microbead fabrication.

	Hydrogel composition (w/v)
Formulation	Alg	Alg-DSP	SF-TA	G-TA
Alg/SF /G-TA	2.5%	-	3.75%	1.25%
Alg/SF-TA/G-TA	2.5%	-	3.75%	1.25%
Alg/Alg-DSP/SF-TA	1.5%	2%	3.5%	-
Alg/Alg-DSP/SF-TA/G-TA	1.5%	2%	2.5%	1%

Quantification of alginate content by DMMB assay (Figure S3A) showed that the amount of alginate released from enzymatically crosslinked microbeads, which were fabricated using 40 μL of pre-hydrogel solution, was 0.31 mg and 0.2 mg before and after citrate treatment, respectively. The amount of alginate released from only ionic crosslinking (No H₂O₂ in crosslinking bath) control was ~0.99 mg (Figure S3B), indicating that sodium citrate treatment released only ~0.11 mg of alginate from the dual crosslinked beads, and the remaining ~0.88 mg was withheld within the microbeads.

3.2. Microencapsulation in on-demand degradable composite microbeads

3.2.1. Characterization of alginate with disulfide-linked phenol moieties (Alg-DSP)

To achieve disintegration of microbeads on-demand, Alg-DSP with reduction sensitive phenol side chains was synthesized through 2-step carbodiimide coupling of cystamine and 4-HPAA groups on alginate backbone (Figure 1B). The ¹H NMR analysis of the Alg-DSP revealed the appearance of new peaks between 6.8–7.5 ppm corresponding to the aromatic hydrogens on the phenol units, and between 3 – 3.9 ppm corresponding to aliphatic hydrogens on the cystamine and 4-HPAA (Figure S4). Alg-DSP solution displayed significantly higher UV absorption between 250 – 290 nm compared to unmodified alginate (Figure S5A), further confirming the conjugation of phenol moieties. The amount of phenol groups was quantified as 250 nmol per mg Alg-DSP from the OD₂₈₀ value using the calibration curve prepared with known concentrations of 4-HPAA (Figure S5B).

3.2.2. Microbead fabrication

Unlike unmodified alginate that formed spherical microbeads at 4% w/v (Figure S6A), the minimum concentration of Alg-DSP required for microbead fabrication by centrifugation was 6% w/v (Figure S6B). Alg-DSP microbeads fabricated by dual crosslinking swelled after sodium citrate treatment but did not disintegrate but dissolved away when incubated in a mixed solution of sodium citrate and glutathione (GSH), confirming successful enzymatic crosslinking through reduction sensitive bridges (Figure S6B). Alg-DSP was then utilized in the fabrication of composite hydrogels by blending with SF-TA and G-TA. Both Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA pre-hydrogel solutions formed spherical microbeads upon centrifugation at 2,250 g into dual crosslinking bath (Figure 1C-i), swelled upon sodium citrate treatment (Figure 1C-ii), and completely disintegrated after incubation in a mixture of sodium citrate and GSH (Figure 1C-iii). This observation confirmed that sodium citrate treatment cleaved the ionic bridges between Alg-DSP and unmodified alginate chains, while covalent crosslinks between Alg-DSP and SF-TA or G-TA were cleaved by glutathione treatment through the reduction of disulfide bridges in the side chains of Alg-DSP into thiol moieties (Figure S7). Spherical beads were obtained at all centrifugation forces between 2,000 and 3,000 g, and the increasing centrifugation speed gradually decreased the average diameter of the microbeads from 180 ± 9 μm to 142 ± 7 μm (Figure S8).

3.2.3. In vitro cell response

Live/dead confocal micrographs of the cell-laden microbeads fabricated by centrifugation at 2250 g showed many calcein-positive viable cells and only a few dead cells (Figure 2A), indicating cytocompatibility of the hydrogel formulation and centrifugation conditions. Survival rates of hNPCs and hMSCs released from the composite microbeads prepared by centrifugation at 2000 g were quantified on the live/dead micrographs (Figure 2B-i) as 85% and 80%, respectively (Figure 2B-ii), which dropped to ~75% and ~50%, respectively, when the centrifugation force was increased to 3000 g during microbead fabrication. Confocal micrographs of the cell-laden Alg/Alg-DSP/SF-TA/G-TA composite microbeads over 3 weeks revealed that many hNPCs migrated out of both the untreated and citrate treated beads by day 14 (Figure 2C). Similarly, hMSCs (Figure S9A) were also able to migrate out of the citrate treated microbeads and proliferate both on the culture plate and the microbead surfaces, binding them together at later time points. Despite initial swelling after citrate treatment, cell-laden microbeads contracted significantly over 3 weeks, with the average diameter decreasing from ~196 um to 143 um (Figure S9B).

Figure 2.

In vitro cell response to composite microbeads. (A) Live/dead confocal micrographs of hNPCs and hMSCs encapsulated in dual crosslinked Alg/Alg-DSP/SF-TA hydrogel microbeads fabricated by centrifugation at 2,250g. Scale bars: 250 μm. (B) (i) Live/dead confocal micrographs and (ii) survival rates of the cells released from the microbeads immediately after encapsulation showing the impact of centrifugation forces at 2,250 or 3,000 g during microbead fabrication. (C) Live/dead confocal micrographs of hNPCs microencapsulated in Alg/Alg-DSP/SF-TA/G-TA microbeads and cultured over 3 weeks. Green: calcein (live), red: EthD-1 (dead), scale bars: 200 μm.

3.3. Characterization of on-demand degradable hydrogel microbeads

3.3.1. Permselectivity

Digital quantification of diffusivity analysis performed on confocal micrographs (Figure 3A) of microbeads incubated with FITC-labeled dextran with various MWs showed that the diffusivity ratios of dual crosslinked Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA microbeads decreased gradually from ~0.8 to ~0.4 with increasing dextran MW (Figure 3B), suggesting a barrier effect against large molecules with a MW larger than 10 kDa. After citrate treatment, however, the diffusivity ratios increased significantly, reaching ~0.7 for 70 kDa dextran.

Figure 3.

Characterization of hydrogel microbeads. (A) Confocal micrographs (Scale bars: 200 μm) and (B) quantitative diffusivity ratios before and after sodium citrate treatment of Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA hydrogels incubated in 0.1 mg/mL FITC-Dextran (4–70 kDa) solutions for 24 h at 37°C (n=20). (C) UV absorbance spectra of composite hydrogel microbeads compared to 4% w/v alginate-only control beads at 250–400 nm. (D) Elastic moduli of alginate-only, Alg/Alg-DSP/SF-TA and untreated and citrate treated Alg/Alg-DSP/SF-TA/G-TA hydrogel microbeads. (n=6, *p < 0.05, **p < 0.01 and ***p < 0.001)

3.3.2. UV absorption

UV absorption of microbeads between 250–400 nm was quantified using UV spectroscopy. Both Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA microbeads displayed significantly higher UV absorption, more than 2-fold in the UV-C region below 280 nm, compared to the ionically crosslinked alginate-only control (Figure 3C).

3.3.3. Stiffness

Elastic moduli of dual crosslinked composite microbeads and ionically crosslinked alginate-only controls were measured by AFM nanoindentation analysis. The average elastic moduli of Alg/Alg-DSP/SF-TA and Alg/Alg-DSP/SF-TA/G-TA microbeads were quantified as 49 ± 15 kPa and 53 ± 14 kPa, respectively, while that of ionically crosslinked alginate-only control was 6 ± 2 kPa (Figure 3D). Sodium citrate treatment of Alg/Alg-DSP/SF-TA/G-TA microbeads lowered elastic modulus more than 5-folds down to 10 ± 3 kPa, suggesting on-demand modulation of the mechanical properties of dual crosslinked microbeads by cleavage of ionic crosslinks.

3.4. Cytoprotection of microencapsulated cells

3.4.1. Barrier effect against large macromolecules

Control and microencapsulated cells were incubated with antibodies specific to surface receptors, or with polycation PEI with an average MW of 70 kDa or cytokine TNF-a to evaluate barrier properties of microbeads against large macromolecules. Flow cytometry analysis (Figure S10) showed that hMSCs were positive for surface markers CD90 and CD44, which was further confirmed by confocal micrographs of immunostained sample (Figure 4A-i), where the surfaces of individual hMSCs were positive for the FITC signal from secondary antibody. Microencapsulated cells, on the other hand, were negative for immunostaining as no FITC signal was collected from the surface of the cells encapsulated in dual crosslinked Alg/Alg-DSP/SF-TA microbeads (Figure 4A-ii), indicating lack of antibody diffusion into the microgels.

Figure 4.

Cytoprotection of cells by encapsulation in Alg/Alg-DSP/SF-TA/G-TA composite hydrogel microbeads against environmental factors. (A) Confocal micrographs of control and microencapsulated hMSCs after immunostaining on CD44 and CD90 surface markers (Green: FITC-labeled secondary antibody, scale bars: 100 μm.). (B) (i) Live/dead fluorescent micrographs (Green: calcein (live), red: EthD-1 (dead), scale bars: 200 μm.) and (ii) survival rates of control and microencapsulated hNPCs after exposure to 0.1 mg/mL PEI for 10 min, 40 ng/mL TNF-α for 48 h, attachment deprivation on PDMS coated wells for 48 h, pH 5.0 buffer for 30 min, or UV-C irradiation at 254 nm for 20 min. (n = 3, *p < 0.05, **p < 0.01 and ***p < 0.001)

Diffusion of PEI and TNF-a into the hydrogel microbeads were evaluated through the changes in the viabilities of microencapsulated hNPCs and hMSCs on live/dead fluorescence micrographs (Figure 4B-i, Figure S11A) used for quantification of survival rates (Figure 4B-ii, Figure S11B). None of the control cells survived incubation in 0.1 mg/mL PEI solution for 10 min, revealing its high cytotoxicity on cells in suspension. hNPCs and hMSCs encapsulated in dual crosslinked Alg/Alg-DSP/SF-TA microbeads, on the other hand, had viabilities of 62% and 74%, respectively, after release from microbeads following incubation with PEI. Cell response to incubation with TNF-α for 48 h varied with cell type: only ~18% of hNPCs were viable after treatment, while ~74% of the hMSCs survived incubation, likely because of the regulatory and secretory roles of MSCs rather than apoptosis in response to inflammatory cytokines ^[57]. Microencapsulation in dual crosslinked Alg/Alg-DSP/SF-TA/G-TA microbeads significantly increased survival rate of hNPCs to ~72%, and the viability of microencapsulated hMSCs increased slightly to ~79%.

3.4.2. Protection against anoikis and extracellular acidosis

Apoptotic cell death via attachment deprivation (anoikis) was induced though 48 h of culture on PDMS-coated culture plates. Control hNPCs and hMSCs formed clusters with many dead cells in the core and some viable cells in the perimeter (Figure 4B-i, Figure S11A), resulting in survival rates of ~33% and ~40%, respectively (Figure 4B-ii, Figure S11B). Microencapsulated cells released after attachment deprivation were well-dispersed and ~80% of them were viable, indicating survival signals provided by Alg/Alg-DSP/SF-TA/G-TA microbeads. The impact of exposure to acidic pH was analyzed by incubation in MES buffered saline at pH 5.0 for 30 min, which reduced the viabilities of hNPCs and hMSCs down to 8% and 33%, respectively. Microencapsulation in Alg/Alg-DSP/SF-TA microbeads significantly improved survival rates after exposure to extracellular acidosis as the ~70% and 82% hNPCs and hMSCs remained viable after treatment.

3.4.3. Shielding against UV-C radiation

Considering the high UV absorbance of SF and phenol-conjugated polymers, UV shielding of the composite microbeads was assessed by exposing cell-laden beads to UV-C irradiation at 254 nm for 20 min. Cell viability 2 h after UV treatment was recorded as 61% and 50% for control hNPCs and hMSCS, respectively, while more than 75% of both cell types microencapsulated in dual crosslinked Alg/Alg-DSP/SF-TA microbeads survived the treatment (Figure 4B-ii, Figure S11B). Live/dead fluorescence micrographs of hNPCs 48 h after UV treatment (Figure S12A) revealed that almost all cells in the control group died, suggesting apoptosis at later time points, while many cells in the microencapsulated group were viable and able to spread (Figure S12B), confirming the UV shielding effect of the composite microbeads.

3.4.4. Preservation of cell viability and growth against hydrodynamic forces

Control and microencapsulated cells were subjected to hydrodynamic forces during extrusion in viscous PEO solution through 27-gauge needles with an inner diameter of 210 μm. Many dead cells were observed on live/dead micrographs 2h after extrusion (Figure 5A-i), and survival rates of hNPCs and hMSCs were ~32% and ~44%, respectively (Figure S8A-ii), revealing severe cell damage caused by mechanical stress during capillary flow in PEO solution. In the case of microencapsulated cells, on the other hand, no significant difference was recorded in cell viability before and after extrusion as ~82% and 80% of hNPCs and hMSCs were viable, respectively. Live/dead micrographs of hNPCs over 3 weeks of culture after extrusion (Figure S13A) showed that both control cells and microencapsulated cells after release were able to spread and proliferate, reaching confluency by day 21. Microencapsulated cells displayed higher metabolic activity than extrusion control at days 1 and 7 but lower than the untreated control at all time points tested (Figure S13B). hMSCs in the extrusion control group, on the other hand, were not able to reach confluency over 21 days after extrusion (Figure 5B-i) and their metabolic activity was significantly lower than the untreated control and extrusion microencapsulated groups from day 1 to 21 (Figure 5B-ii). Unlike extrusion control, hMSCs extruded in microbeads were able to reach confluency after release by day 21, and their metabolic activity lined up with that of untreated control by day 14.

Figure 5.

Influence of encapsulation in Alg/Alg-DSP/SF-TA/G-TA composite microbeads on cell survival and growth after extrusion in 6% w/v PEO solution through 27G needle at 3 mL/min. (A) (i) Live/dead fluorescent micrographs (green: calcein (live), red: EthD-1 (dead), scale bars: 200 μm) and (ii) survival rates of control and microencapsulated cells 2 h after extrusion. (B) (i) Live/dead fluorescent micrographs (MC: microencapsulated, scale bars: 200 μm) and (ii) metabolic activities of control and microencapsulated hMSCs over 3 weeks of culture after extrusion compared to untreated control (n=4, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001).

3.4.5. Preservation of cell differentiation after extrusion-based delivery

Osteogenic and adipogenic differentiation of hMSCs and neural differentiation of hNPCs were evaluated after extrusion by qPCR analysis of lineage specific marker genes (Figure 6A) and histological staining (Figure 6B) or immunostaining (Figure 6C). hMSCs in the untreated control group displayed upregulation of bone marker genes alkaline phosphatase (ALP), osteopontin (OP) and osteocalcin (OC) by ~12-, 7- and 1.4-folds, respectively, after induction of osteogenesis for 21 days (Figure 6A-i). Adipose markers adiponectin, leptin and peroxisome proliferator activated receptor gamma (PPARg) were upregulated by ~72000-, 231- and 32-folds, respectively (Figure 6A-ii), after incubation in adipogenic medium for 14 days. Cells cultured in osteogenic or adipogenic media were also positive for calcium deposition or oil droplet formation determined by alizarin red or oil red staining (Figure 6B), respectively, indicating their multipotent differentiation capacity. Interestingly, expression of osteogenic and adipogenic markers by the cells in the extrusion control group was significantly lower compared to the untreated control, and no calcium deposition or oil droplets were detected after histological staining. Cells in the extrusion after microencapsulation group, on the other hand, displayed slightly higher expression of ALP and OP (Figure 6A-i) but comparable levels of expression for adipogenic genes with the untreated control (Figure 6A-ii). Like untreated control, cells released from microbeads following extrusion were positive for both alizarin red or oil red staining after induction of osteogenesis or adipogenesis, respectively (Figure 6B).

Figure 6.

Preservation of cell differentiation against hydrodynamic stress during extrusion through 27G needle by microencapsulation in Alg/Alg-DSP/SF-TA/G-TA microbeads (A) qPCR analysis of the expression of (i) bone marker genes alkaline phosphatase (ALP), osteopontin (OP) and osteocalcin (OC) or (ii) adipose marker genes adiponectin, leptin, and peroxisome proliferator activated receptor gamma (PPARg) by hMSCs, and (iii) neural marker genes Nestin, β-III tubulin (TUBB3) and microtubule associated protein 2 (MAP2) by hNPCs upon incubation in induction media for 21 or 14 days following 48 h of culture in expansion media after extrusion (n = 3, *p < 0.05, **p < 0.01 and ***p < 0.001). (B) Brightfield micrographs of hMSCs after histological staining with alizarin red for calcium deposition 21 days after osteogenic induction, or with oil red for lipid droplets 14 days after adipogenic induction. Scale bars: 200 μm. (C) Immunostaining of hNPCs for β-III tubulin 14 days after extrusion. Green: β-III tubulin, blue: DAPI (nuclei), scale bars: 100 μm. MC: microencapsulated.

Untreated hNPCs incubated in neurogenic medium for 14 days displayed ~5-, 9- and 49-fold upregulation of neural marker genes neuroepithelial stem cell protein (Nestin), tubulin beta 3 class III (TUBB3), and microtubule associated protein 2 (MAP2), respectively (Figure 6A-iii). While no significant difference was recorded in the expression of Nestin and MAP2, cells in the extrusion control group had significantly levels of TUBB3, which was also visualized by immunostaining of beta-III tubulin (Figure 6C). Expression of neural markers by the cells released from microbeads after extrusion was comparable with the untreated control with no statistically significant difference (Figure 6A-iii), and the cells were positive for beta-III tubulin on day 14 (Figure 6C). These observations suggested preservation of differentiation capacities of both hMSCs and hNPCs after exposure to hydrodynamic forces during capillary flow.

3. Discussion

Centrifugation-based droplet formation strategy was adapted using alginate as a template for fabrication of enzymatically crosslinked silk and silk/gelatin composite hydrogel microbeads as an alternative to alginate-only microgels for cell encapsulation and cytoprotection. Considering that a maximum pre-hydrogel volume of 20 μL can be centrifuged per needle and a total capacity of sixty tubes for the rotor, it is possible to convert more than 1 mL of solution into microbeads and encapsulate ten million cells in a minute after a single run, which is significantly higher than microfluidics approaches where the flow rate of the aqueous phase is generally below 150 μL/min ^[58] and it would take around 10 min to achieve the same yields at the highest possible flow rates. Ionic crosslinks of alginate-only hydrogels are susceptible to the ionic strength of physiological buffers such as PBS ^[59] as well as the ion exchange mechanisms due to the metabolic activity of encapsulated cells ^[60]. These factors result in poor mechanical properties and the deterioration of structural stability required for cytoprotection. Here we proposed incorporation of silk fibroin into the hydrogel formulation and enzymatic crosslinking of the polymer network to improve the mechanical properties and structural stability of the microgels though crystallinity of silk ^[61] and stability of covalent bonds ^[40]. In addition to the susceptibility of ionic crosslinks to environmental factors, Ca²⁺-crosslinked alginate hydrogels also suffer from high hydrophilicity that impacts the absorption profiles of proteins ^[62] along with the lack of cell recognition sequences ^[63] that impair cell-matrix interactions and survival in alginate microbeads ^{[56a, 56b]}.

Although the repetitive hydrophobic segments of silk balance the hydrophilicity of alginate ^[64] and enhance its cytocompatibility by modulating the adsorption profile of proteins from the culture medium ^[65], the lack of cell binding sequences in silk from B. mori ^[66] could still limit cell-matrix interactions within the composite hydrogel microbeads. Thus, hydrogel formulation was also supplemented with gelatin, which significantly improves the bioactivity of alginate ^[67] and silk ^[68] hydrogels through the intrinsic arginine-glycine-aspartic acid (RGD) cell binding sequences ^[69]. Indeed, both human stem and progenitor cells encapsulated in Alg/SF-TA/G-TA and Alg/Alg-DSP/SF-TA/G-TA hydrogels migrated out of the gels, attached on microbead surfaces and proliferated, suggesting support for the bioactivity of the hydrogel matrices. Cleavage of the ionic crosslinks by sodium citrate treatment allowed for a higher number of cells to migrate out and proliferate on the culture plates compared to their untreated counterparts at earlier time points. This finding could be attributed to the increase in cell mobility and proliferation at lower crosslinking densities / higher porosity of hydrogel matrices ^[70]. Encapsulation in composite microbeads also offered protection against anoikis, which is defined as the apoptotic death of adherent mammalian cells when appropriate cell–matrix interactions are missing ^[71]. This can impact cell viability in bioinert matrices ^[72] or at injection sites after cell transplantation therapy ^[73]. Although some control hNPCs and hMSC remained viable under attachment deprivation by forming cell clusters, likely through a similar mechanism observed during organoid formation ^[74], encapsulation in composite microbeads significantly increased viability, indicating structural support in suspension culture and survival signals by the hydrogel matrix. Bioactivity of silk/alginate hydrogels could also be improved by modification with RGD peptides ^[61], which could be in situ conjugated into hydrogels during the enzymatic crosslinking of silk ^[75]. This strategy, however, results in longer gelation times and lower crosslinking densities and mechanical properties due to the occupation of some tyrosine side chains ^[52], which is not desirable in cell encapsulation for cytoprotection. Enzymatic crosslinking of silk with tyramine-conjugated gelatin, on the other hand, not only improved the viability and proliferation of encapsulated cells but also accelerated gelation and enhanced hydrogel stiffness through increased crosslinking density ^[52].

Shrinking and opacification of Alg/SF-TA/G-TA microbeads upon sodium citrate treatment confirm enzymatic crosslinking and suggest self-assembly of SF-TA chains at elevated ionic strength ^[76] of sodium citrate solution upon increased chain mobility within the matrix ^[77] after the cleavage of ionic crosslinks. Although the composite microbeads maintain structural stability and the encapsulated cells could migrate out over time after sodium citrate treatment, on-demand release of the cells is not possible, limiting transient encapsulation for cytoprotection purposes. Alg-DSP construct we developed combines structural stability of enzymatic crosslinks under ion exchange and on-demand degradability by reduction, providing a powerful alternative to other enzymatically crosslinked phenol-conjugated alginate constructs ^[78] that cannot be dissolved on-demand for the release of encapsulated cells. Unlike the organic synthesis pathway previously employed for enzymatically crosslinked, reduction sensitive polyglutamic acid hydrogels ^[79], here we used a much simpler, all aqueous strategy through 2-step carbodiimide coupling for the synthesis of Alg-DSP. Unmodified alginate was included in the composite hydrogel formulation as a thickening agent due to lowered viscosity Alg-DSP solution likely due to the conversion of some hydrophilic carboxylic acid groups into hydrophobic phenol moieties, failing to provide appropriate flow mechanics to form individual droplets in quasi-static dripping regime during centrifugation through the nozzle ^[56a]. Swelling behavior of Alg-DSP/Alg/SF-TA/G-TA microbeads contrary to shrinking of Alg/SF-TA/G-TA upon cleavage of ionic bridges by sodium citrate treatment suggested limited mobility and self-assembly of SF-TA chains when Alg-DSP was enzymatically crosslinked to the polymer network. Similar observations were made before with enzymatic crosslinking of silk with phenol-substituted hyaluronic acid (HA) chains, which inhibited β-sheet formation and hydrogel stiffening ^{[51, 80]} while simply blending silk with unmodified HA induced self-assembly and stiffening ^[81]. The gradual decrease in the diameter of cell-laden microbeads could be attributed to the cleavage of reduction-sensitive crosslinks by basal glutathione release from metabolically active cells ^[82], increasing SF chain mobility and enabling self-assembly. Increasing the centrifugation force above 2250 g during droplet fabrication allows for the fabrication of microbeads with a diameter smaller than 150 μm. However, this process negatively impacts the viability of the encapsulated cells likely due to higher hydrodynamic (shear and extensional) forces within the needle at higher flow rates of the viscous pre-hydrogel solution ^[18] during droplet formation.

The barrier effect of dual crosslinked microbeads against molecules larger than 10 kDa was exploited for protection of encapsulated cells against cytotoxic factors such as high MW PEI, which is used as a structural supplement in cell encapsulating hydrogels ^[83] but could cause necrosis by damaging the integrity of negatively charged cell membrane ^[84]; and TNF-α, a ~17 kDa inflammatory cytokine that is abundant in inflammation site in sepsis, osteoporosis, multiple sclerosis, rheumatoid arthritis, and inflammatory bowel diseases ^[85]. Besides inflammatory cytokines, protection by encapsulation in composite microgels against extracellular acidosis, which disrupts cell function through disorders in organelles ^[86] and inhibition of protein synthesis ^[87] in the tumor or inflammatory microenvironment ^[88], holds promise in adaptive immune cell transfer therapy ^[89]. Preservation of viability at pH 5.0 could be attributed to the polyanionic nature of the alginate within the microbeads that is rich in carboxylic acids, likely providing buffering at low pH ^[90]. Due to the sensitivity of ionic crosslinks in alginate hydrogels to low pH through the protonation of carboxylic acid groups ^[91], dual crosslinking of composite microbeads was essential for the preservation of structural stability at acidic pH. Unlike hydrogel microbeads of unmodified alginate, high UV absorption of the composite microbeads between 250 and 290 nm by the phenol side chains ^[92] on Alg-DSP, SF-TA and G-TA provides a barrier against UV-C and UV-B radiation that can be utilized in chemical-free sterilization ^[93] or photo-crosslinking of hydrogels ^[94]. For example, 2-hydroxy-4-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959), a cytocompatible photoinitiator that is widely used in photocrosslinking of cell encapsulating hydrogels, could not be utilized at its absorption peak of 260 nm due to DNA damage and apoptotic cell death from UV-C radiation ^[95]. Instead, 365 nm is used to initiate photocrosslinking, which takes significantly longer than 260 nm due to lower molar extinction coefficient in the UV-A region ^[96]. Our findings suggest that the suspensions of cells microencapsulated in dual crosslinked composite microbeads can safely be sterilized rapidly or included in hydrogels crosslinked under UV-C radiation.

Extremes of extensional and shear forces applied during capillary flow can cause transient or permanent damage to the plasma membrane and result in formation of pores, distortion of cell shape and size, and leakage of cytosolic contents ^[97]. The mechanical stress-associated reduction in cell viability has been a major issue in injection based delivery and extrusion 3D bioprinting, where cell suspensions in a viscous solution flow through microscale fine needles or nozzles ^{[21a, 24]}, reducing cell viability by up to 90% ^{[22a, 98]}. The use of microencapsulated cells in a spinner vessel ^[27], during injection-based therapeutic delivery ^[99], or 3D bioprinting ^[100] were previously demonstrated, but the extent of mechanical stress or cytoprotection were not discussed in detail. Here we showed that composite microbeads had significantly higher stiffness than alginate-only controls and could be utilized in cytoprotection against hydrodynamic stress during capillary flow. Stiffness of ionically crosslinked alginate microbeads could be improved by increasing the polymer concentration ^[101], which would increase solution viscosity ^[102] and could compromise the flow mechanics required for droplet formation ^[56a] as well as the viability of encapsulated cells because of higher hydrodynamic forces within the needle ^[103]. Stiffer alginate hydrogels could also be obtained via higher crosslinking density by increasing the CaCl₂ concentration or duration of incubation in the crosslinking bath ^{[33, 104]}, which could impair cell metabolism and proliferation ^[105] and even cause apoptosis ^[106]. Dual crosslinked composite microbeads, on the other hand, enable superior mechanical properties without a significant increase in solution viscosity or longer crosslinking times. Considering the influence of matrix stiffness over stem cell fate ^[107], sodium citrate treatment of composite microbeads also offers on-demand reduction of stiffness, which could be utilized in future studies for the modulation of cell behavior in the microbeads.

Extrusion of cells in viscous PEO solution not only impaired cell viability significantly through necrotic death, but also impaired growth and differentiation, which could be attributed to alterations in gene expression due to chromosomal remodeling ^[108]. hMSCs failing to undergo osteogenesis or adipogenesis after extrusion could also be explained with commitment into a fibroblast-like phenotype after mechanical stress during extrusion in the presence of death signals from dead cell debri ^[109]. Although hNPCs were able proliferate and express elevated levels of Nestin and MAP2 after extrusion, expression of TUBB3 was significantly impaired. TUBB3 encodes β-III tubulin that forms cytoskeletal microtubules in neurons within central and peripheral nervous systems ^[110], and it was shown to be non-essential for neurogenesis but required for axon growth ^[111]. Considering that relative levels of tubulin isoforms may change in response to microtubule dynamics ^[112], hydrodynamic stress could be speculated to induce cytoskeletal remodeling in hNPCs. Composite microbeads successfully preserved viability, growth and multipotency of both cell types after extrusion, which emphasizes the potential applications of the approach in cell therapy and 3D bioprinting where survival, self-renewal and stemness are essential in success rate.

4. Conclusions

Alginate-templated, dual crosslinked silk and silk/gelatin composite hydrogel microbeads were fabricated and used for encapsulation of human stem and progenitor cells by employing the simple and cost-effective centrifugation-based droplet formation approach. In addition to introducing the synthesis of enzymatically crosslinkable and on-demand degradable alginate in the system, the present study also demonstrates dual crosslinked composite microbead formulations for improved structural stability and mechanical properties compared to ionically crosslinked alginate-only predecessors. Bone marrow-derived hMSCs and NPCs were successfully microencapsulated and released from the gels, maintaining ≥80% survival. Hydrogel microbeads protected encapsulated cells against cytotoxic molecules or factors, acidity, anoikis, UV irradiation and particularly hydrodynamic stress during extrusion-based delivery. Preservation of cell viability, growth and differentiation capabilities after extrusion supports this new microencapsulation approach as an effective strategy for cell transplantation applications through injection-based delivery or for use in 3D bioprinting.

5. Experimental Section

Synthesis of tyramine-substituted silk fibroin (SF-TA) and gelatin (G-TA):

For the extraction of silk fibroin (SF) from Bombyx mori, 5 g of cut cocoons were degummed by boiling in 2 L of 0.02 M sodium carbonate (Sigma-Aldrich, St. Louis, MO) solution for 60 min and rinsing three times in deionized (DI) water. Dried fibers were solubilized in 9.3 M lithium bromide (Sigma-Aldrich, St. Louis, MO) solution and dialyzed using 3.5 kDa MWCO tubing against DI water for 3 days. The resulting solution was centrifuged 2 times at 9,000 rpm at 4°C for 20 min to remove insoluble particles and the protein concentration was determined by weighing a known volume of solution before and after drying overnight at 60°C. Tyramine-substituted SF (SF-TA) and gelatin (G-TA) were synthesized as previously described ^[52]. Briefly, 2% (w/v) SF or gelatin type A (from porcine skin) in 0.05 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH 6.0) were reacted with tyramine hydrochloride (Sigma-Aldrich, St. Louis, MO) (500 mg per 1 g protein) in the presence of 1-ethyl-3-(−3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Thermo Fisher Scientific, Rockford, IL) (184 mg per 1 g protein) and N-hydroxysuccinimide (NHS) (57 mg per 1 g protein) for 18 h at room temperature (RT, 24°C) under gentle stirring. Solutions were dialyzed against DI water using 3.5 kDa MWCO tubing with 6 changes over 3 days. SF-TA solution was concentrated to ~10% w/v by water evaporation in a fume hood and stored at 4°C for up to a month. G-TA solution was lyophilized and stored at −20°C until further use.

Synthesis and characterization of alginate conjugated with disulfide-linked phenol moieties:

Alginate with disulfide-linked phenol side chains (Alg-DSP) was synthesized through a two-step carbodiimide coupling pathway. A 1% (w/v) sodium alginate (G/M ratio: 1.05) (Sigma-Aldrich, St. Louis, MO) in 0.05 M MES buffer (pH 6.0) was reacted with cystamine hydrochloride (Sigma-Aldrich, St. Louis, MO) (500 mg per 1 g alginate) in the presence of EDC (370 mg per 1 g alginate) and NHS (120 mg per 1 g alginate) for 18 h at RT under gentle stirring. The resulting solution was dialyzed using 3.5 kDa MWCO tubing against DI water for 3 days with 6 water changes and concentrated overnight to ~2% w/v by water evaporation in a fume hood. The alginate-cystamine solution was then reacted with 4-hydroxyphenyl acetic acid (4-HPAA, 160 mg per 1 g alginate-cystamine) in 0.05 M MES buffer (pH 6.0) in the presence of EDC and NHS (140 mg and 85 mg per 100 mg 4-HPAA, respectively) for 18 h at RT and the solution was dialyzed against DI water for 3 days with 6 water changes. The solution was diluted to 1 mg/mL with dH₂O and sterile filtered through 0.22 um filter units, lyophilized and stored at −20°C. Lyophilized Alg-DSP powder was dissolved in deuterated water (D₂O) at a concentration of 20 mg/mL and analyzed using a 500 MHz Bruker Avance III NMR spectrometer. UV absorption spectra of 1 mg/mL unmodified alginate or Alg-DSP solutions in dH₂O were measured using a SpectraMax M2 multi-mode microplate reader (Molecular Devices, Sunnyvale, CA). Mol% of DSP chains conjugated on alginate was quantified using a calibration curve plotted by OD₂₈₀ values of known concentrations of 4-HPAA.

Microbead fabrication:

Alginate templated SF-TA/G-TA, Alg-DSP/SF-TA and Alg-DSP/SF-TA/G-TA solutions (Table 1) were prepared in HD buffer (40 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) with 5% w/v dextrose, pH 7.4) and supplemented with 100 U/mL HRP. A 2 mL Eppendorf tube was filled with 1.4 mL of crosslinking solution (40 mM HEPES, 4.5 g/mL dextrose, 150 mM NaCl, 5 mM KCl, 50 mM CaCl₂ and 0.5 mM H₂O₂) and a 34G needle (51 um inner diameter) (Figure S1-i) was placed on top using a polydimethylsiloxane (PDMS) plug (Figure S1-ii). 10–20 μL of pre-hydrogel solution was added into the needle (Figure S1-iii) and centrifuged at 2000 g - 3000 g for 1 min using an Eppendorf 5804 R centrifuge with a T-60–11 horizontal rotor (Figure S1-iv). The crosslinking solution was discarded immediately after centrifugation. For the cleavage of ionic bonds, hydrogel microbeads were incubated in 35 mM sodium citrate solution in HD buffer for 5 min. For the complete dissolution of microbeads through the cleavage of both ionic and disulfide bridges, beads were incubated in 35 mM sodium citrate with 10 mM glutathione (GSH) in HD buffer at 37°C for 20 min under gentle shaking. Morphology of the microbeads before and after citrate treatment was analyzed on the phase contrast micrographs collected using a BZ-X700 Fluorescence Microscope (Keyence Corp., Itasca, IL). Diameters of the microbeads were quantified using the image analysis software ImageJ (1.48v, NIH, USA). Alginate content of the microbeads was quantified before and after sodium citrate treatment using dimethylmethylene blue (DMMB) assay as described elsewhere ^[113].

Permselectivity:

Hydrogel microbeads were incubated in 0.1 mg/mL solutions of fluorescein isothiocyanate (FITC)-dextran constructs with MWs of 4, 10, 40 and 70 kDa in DMEM colorless media for 24 hours at 37°C with 5% CO₂. Microbeads were visualized using a TCS SP8 microscope from Leica Microsystems (Wetzlar, Germany) and the fluorescence intensities of the microbeads and the background were measured using ImageJ. Diffusivity of microbeads was estimated from the ratio of the fluorescence intensity within the hydrogel microbeads to the background fluorescence intensity using the equation (1):

$$\mathit{\text{Diffusivity}}\mathit{\text{Ratio}}=\frac{\mathit{\text{Integrated}}\mathit{\text{Density}}\mathit{\text{of}}\mathit{\text{Microbead}}}{\left(\mathit{\text{Mean}}\mathit{\text{Gray}}\mathit{\text{Value}}\mathit{\text{of}}\mathit{\text{Background}}\right)*\left(\mathit{\text{Area}}\mathit{\text{of}}\mathit{\text{Microbead}}\right)}$$ (1)

Mechanical testing:

Force measurements on hydrogel microbeads were performed on an Asylum Research MFP-3D-Bio Atomic Force Microscope (AFM) (Asylum Research, Santa Barbara, CA) integrated with an inverted Nikon Eclipse Ti optical microscope (Micro Video Instruments, Avon, MA). Force curves were collected by using the AFM force - volume mode at a mapping distance of 2 μm between points using a spherical AFM probe with a radius R = 6.5 μm (measured optically via a microscope with a 40x objective, NA 0.6), and RMS roughness (Rq) of 16.93±1.13nm (measured using tapping mode AFM), and a calibrated spring constant k = 0.12 N/m (measured by using the thermal calibration built in to the MFP3D software of the AFM). The elastic modulus for the beads was determined by fitting the force vs. indentation curves with the Hertz model for a spherical indenter as described previously ^[114]. Before measurements were performed on each new sample, the cantilever was calibrated first in air and then in the sample medium. To limit energy dissipation due to viscoelastic effects the indentation frequency was 0.4 Hz for all measurements.

Cell viability:

Bone marrow-derived human MSCs and NPCs (ATCC, Manassas, VA) were used as mammalian cell models for microencapsulation and on-demand release studies. DMEM/Nutrient Mixture F12 medium supplemented with 10% FBS, 1% Glutamax, 1% non-essential amino acids, and 1% P/S; or with 5% FBS, 1X N2 supplement, bovine pituitary extract and 1% P/S were used for hMSCs and hNPCs, respectively. Cells suspended in 20 uL of FBS were mixed with the pre-hydrogel solutions with a final volume of 200 uL and a cell density of 10⁷ cells/mL, and 10 uL aliquots were centrifuged at 2250 or 3000g through the 34G needle into the crosslinking bath. Cells in Alg/Alg-DSP/SF-TA microbeads were released by sodium citrate and GSH treatment 2 h after encapsulation. Cell-laden Alg/Alg-DSP/SF-TA/G-TA microbeads cultured in expansion media for 21 days. Samples were stained with Live/Dead assay kit (Invitrogen, Carlsbad, CA) following the protocol provided by the manufacturer for suspended cells. The cells released from Alg/Alg-DSP/SF-TA microbeads were seeded in 96 well plates and imaged under fluorescence microscope. Five random images were taken (n=3) for each sample and analyzed using the live/dead staining macro of ImageJ to estimate the % viability. The cell-laden microbeads were imaged under confocal microscope after live/dead staining at days 0, 1, 7, 14 and 21.

Cytoprotection by microencapsulation:

Cell-laden Alg/Alg-DSP/SF-TA/G-TA hydrogel microbeads were incubated in 0.1 mg/mL PEI solution in 1X PBS for 10 min or in pH 5.0 MES buffered saline solution (10 mM MES, 137 mM NaCl, 3 mM KCl) for 30 min to evaluate cytoprotection against polycations and acidic environment. To investigate the protection against UV irradiation, microbeads suspended in 1X PBS were exposed to UV-C light at 254 nm using an 8 Watt UVP UVLMS-38 UV lamp (Analytik Jena AG, Thuringia, Germany) for 20 min. For the evaluation of protection against mechanical stress, cell-laden hydrogel microbeads were suspended in 6% w/v PEO (POLYOX WSR N750, Mwt 300,000 Da, ColorCon, Harleysville, PA) solution in growth media without FBS and extruded at 3 mL/min through 1.27 cm long Precisionglide 27G hypodermic needles with an inner diameter of 210 μm (BD, Franklin Lakes, NJ) using a Harvard PHD 2000 syringe pump (Harvard Apparatus, Holliston, MA). Cells encapsulated in Alg/Alg-DSP/SF-TA/G-TA microbeads were incubated in PDMS-coated well plates or suspended in 40 ng/mL tumor necrosis factor-α (TNF-α) solution in FBS-free growth media and incubated for 48 hours at 37°C with 5% CO₂ to assess protection against attachment deprivation or apoptotic factors, respectively. Immediately after exposure to environmental stress, cells were released from the hydrogel microbeads by sodium citrate-glutathione (GSH) treatment, seeded in well plates at a density of ~3×10⁵ cells/cm² and stained with Live/Dead viability assay kit following the protocol provided by the manufacturer for suspended cells. Cells were imaged using fluorescence microscope and survival rates were quantified from 5 random micrographs (n=3) using ImageJ as described above.

Proliferation and metabolic activity:

Cells released from Alg/Alg-DSP/SF-TA/G-TA microbeads after extrusion through 27G needle were seeded on culture plates at a density of 8000 cells/cm² for in vitro analysis over 21 days of culture. Cell proliferation was monitored through quantification of dsDNA using Quant-iT PicoGreen assay (n=3) and metabolic activity was measured using alamarBlue assay (n=4) and according to the instructions by the manufacturer (Invitrogen, Carlsbad, CA).

Quantitative Polymerase Chain Reaction (qPCR):

Total RNA extraction was performed using RNeasy Plus Micro Kit (Qiagen, Germantown, MD) according to manufacturer’s instructions. First strand cDNAs were reverse transcribed using RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Rockford, IL) from each sample according to manufacturer’s instructions using a thermal cycler (iCycler, BIO-RAD, USA). qPCR was performed using GoTaq® qPCR Mastermix (Promega, USA) according to manufacturer’s instructions using forward and reverse primers (Table S1) specific for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as the housekeeping gene; alkaline phosphatase (ALP), osteopontin and osteocalcin as osteogenic markers; adiponectin, leptin and peroxisome proliferator activated receptor gamma (PPARg) as adipogenic markers; and neuroepithelial stem cell protein (Nestin), microtubule associated protein 2 (MAP2) and tubulin beta 3 class III (TUBB3) as neural markers. Relative expression of the genes compared to expansion media control was calculated using the ΔΔCt method (n=3) as described elsewhere ^[115].

Histology and immunostaining:

Released cells were seeded at a density of ~4×10⁴ cells/cm² and after 48 h of culture, expansion media was replaced with neural differentiation medium (Neurobasal medium supplemented with 2% v/v B-27, 1X GlutaMAX and 1% P/S) for hNPCs and either with Gibco StemPro osteogenesis or adipogesis differentiation media (Thermo Fisher Scientific, Rockford, IL) for hMSCs. Induction medium was replaced every 3 days with fresh medium. After culture in neural an adipogenic differentiation media for 14 days and in osteogenic differentiation medium for 21 days, samples were fixed in 4% paraformaldehyde solution for 15 min and rinsed with 0.01 M PBS. hMSCs were stained with Alizarin Red S for calcium deposition or Oil Red O for formation of lipid droplets and imaged under a brightfield microscope. For the immunostaining of hNPCs, fixed samples were permeabilized in 1X PBS with 0.1% v/v Triton X-100 for 30 min and incubated in blocking solution (1% w/v BSA and 0.1% v/v Tween 20 in 1X PBS) at 37 °C for 1 h. Samples were incubated in primary antibody solution (Mouse anti-β-tubulin III, ab78078 in 0.1% w/v BSA solution) overnight at 4°C and in secondary antibody (Goat anti-mouse IgG Alexa Fluor 64, Thermo Fisher Scientific, Rockford, IL) for 1 h at RT. Cell nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) for 5 min at RT and the specimens were imaged under confocal microscope.

Statistical analysis:

All data are expressed as mean ± standard deviation for n ≥ 3. GraphPad Prism (GraphPad Software, La Jolla, CA) was used to perform One- or Two-way analysis of variance (ANOVA) with Tukey’s post hoc multiple comparison test to determine statistical significance (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001).