Soft Poly(ethylene) Glycol Hydrogels Support Epiblast-like Morphogenesis in 3D Cultured Human Pluripotent Stem Cells

Introduction: Human embryonic models derived from human pluripotent stem cells (hPSCs) have enjoyed a recent explosion in popularity, owing to their ability to capture early developmental milestones which are impossible to study in vivo. Established methods for producing these models, however, rely on reconstituted basement membrane (rBM) matrices to provide a 3D microenvironment conducive to cell self-assembly. Though these methods have been successfully applied to understanding the cellular machinery involved in morphogenesis, rBM matrices suffer from a poorly defined composition, inability to tune their biochemical or biophysical properties, such as cell-matrix adhesion or matrix stiffness, and poor control over cell fates in long-term culture. As such, they are inadequate for understanding how biophysical cues from the extracellular matrix (ECM) influence cell morphogenesis, something which remains poorly understood. To this end, we designed a “blank-slate”, modular, non-adhesive hydrogel platform for elucidating the role of matrix physical cues in early development. Hydrogels were assembled through the crosslinking of PEG thiol and acrylate macromers, and could support primed hPSC self-assembly into lumenal, epiblast-like structures without additional biochemical cues.

Materials and

Methods:

Methods: PEG hydrogels were formed via Michael addition crosslinking between 8-arm PEG acrylate and 4-arm PEG thiol. PEG hydrogels of different stiffness were formed by tuning polymer concentration between 1.5-3% w/v. To investigate the role of matrix biophysical cues on morphogenesis, cells were encapsulated in soft (~250 Pa), intermediate, (~1000 Pa), and stiff (~3000 Pa) PEG hydrogels. Aggregates were then characterized for epiblast morphological and molecular hallmarks, proliferation, and germ-layer differentiation capacity.

Cell Culture: Upon reaching confluency in 2D, hPSCs were dissociated and mixed with 8-arm PEG Acrylate (Jenkem, Plano, TX) and 4-arm Thiol (Jenkem) macromers. Following encapsulation, cells were cultured in Essential-8 (E8) medium (Fisher Scientific, Waltham, MA) supplemented with 10 µM ROCK inhibitor Y-27632 (STEMCELL Technologies, Cambridge, MA) for up to 6 days. Media was refreshed daily.

Aggregate Morphological Characterization: Frequency of lumen formation, lumen volume, and aggregate volume were derived from phalloidin staining. E-cadherin immunofluorescence was used to determine cellular shape and alignment.

Self-Renewal Capacity, Proliferation, and Apical-basal Polarization: Self-renewal capacity was assessed through the immunofluorescent analysis of pluripotency markers OCT4, SOX2, and NANOG. Similarly, proliferation and apical-basal polarity were determined through Ki67 and tight junction marker ZO-1, respectively.

Trilineage Differentiation: After 4 days in self-renewing conditions, cells were exposed to differentiation media. Ectoderm and endoderm lineages were derived using Stemdiff™ ectoderm and endoderm media (STEMCELL Technologies). Mesoderm was derived by treating aggregates with 6 µM CHIR99021 (STEMCELL Technologies) for 48 hours. Immunofluorescence of Pax6 (Ectoderm), Brachyury (Mesoderm), and SOX17 (Endoderm) was used to quantify differentiation efficiency.

Results, Conclusions, and Discussions:

Results: Hydrogel stiffness was found to influence aggregate morphology. Intermediate stiffness hydrogels were the most conducive to lumenogenesis, as cells cultured in these matrices had a higher rate of lumen formation and expansion than those cultured in soft hydrogels. Aggregates in intermediate hydrogels also contained a more centrally placed lumen, which was flanked by a single layer of cells, while aggregates grown in soft hydrogels often contained multiple layers (Figure 1A, 1D). Cells cultured in stiff hydrogels, conversely, failed to develop lumens after 6 days of culture. E-cadherin staining revealed cells in all matrices to be initially disordered by day 3, after which they became radially aligned towards the center of the aggregate, like the human epiblast, by day 6 (Figure 1A). While cells in the soft and stiff hydrogels largely stained positive for proliferation marker Ki67 after 6 days, cells in the intermediate stiffness hydrogels experienced a sharp decrease in expression by that time (Figure 1B). Aggregates also bore molecular hallmarks of the human peri-implantation epiblast. Immunofluorescence revealed that aggregates maintained their stemness throughout 6 days in culture (Figure 1C). The apical enrichment of ZO-1, indicating apical-basal polarization, was observed when cells were cultured in soft and intermediate stiffness hydrogels, though not in stiff hydrogels (Figure 1D). Cells were also found to maintain their trilineage differentiation capacity, with >85% of cells staining positive for their respective lineage marker (Figure 1E).

Discussion and

Conclusion: The observed differences in aggregate morphology and polarization when cells are grown under different matrix stiffnesses suggest that biophysical matrix cues play a role in hPSC morphogenesis. Additionally, cells cultured in intermediate stiffness hydrogels produce aggregates with a strong morphological and molecular resemblance to the peri-implantation human epiblast, indicating their value as potential models of this stage of human development. As Michael-addition chemistry allows the facile incorporation of adhesion ligands, this platform may also be utilized in elucidating the role biochemical matrix cues play in shaping hPSC morphogenesis.

Acknowledgements (Optional): This work was supported by the NIH [R01HD101130] and the NSF [CMMI-2130192 and CBET-1943798], and by Bioinspired Seed Funds from Bioinspired Syracuse: Institute for Material and Living System, Syracuse University.