Soft Polyethylene Glycol Hydrogels Support Human PSC Pluripotency and Morphogenesis

Lumenogenesis within the epiblast represents a critical step in early human development, priming the embryo for future specification and patterning events. However, little is known about the specific mechanisms that drive this process due to the inability to study the early embryo in vivo. While human pluripotent stem cell (hPSC)-based models recapitulate many aspects of the human epiblast, most approaches for generating these 3D structures rely on ill-defined, reconstituted basement membrane matrices. Here, we designed synthetic, nonadhesive polyethylene glycol (PEG) hydrogel matrices to better understand the role of matrix mechanical cues in iPSC morphogenesis, specifically elastic modulus. First, we identified a narrow range of hydrogel moduli that were conducive to the hPSC viability, pluripotency, and differentiation. We then used this platform to investigate the effects of the hydrogel modulus on lumenogenesis, finding that matrices of intermediate stiffness yielded the most epiblast-like aggregates. Conversely, stiffer matrices impeded lumen formation and apico-basal polarization, while the softest matrices yielded polarized but aberrant structures. Our approach offers a simple, modular platform for modeling the human epiblast and investigating the role of matrix cues in its morphogenesis.


Hydrogel Fabrication:
4-arm and 8-arm PEG acrylate containing hydrogels with dithiol crosslinker DTBA were fabricated at a acrylate to thiol molar ratio of 1:1 at 5% w/v total polymer concentration.The procedure for fabricating the hydrogel is the same as described in the main manuscript.

S1. Preliminary screening of PEG macromers and thiol crosslinkers
We screened 4-arm and 8-arm PEG acrylate for their ability to form a hydrogel and support hiPSCs viability post encapsulation.Our results indicated that the PEG hydrogels formed using multi-arm PEG acrylate and dithiol crosslinker (DTBA) showed low cell viability 3 days post encapsulation in the hydrogels.
Comparatively hydrogels formed using 8-arm PEG acrylate supported higher cell viability than hydrogels formed using the 4-arm PEG acrylate (Table S2).Moreover, we could not form hydrogels of less than 5% w/v polymer concentration using the 4-arm PEG acrylate or 8-arm PEG acrylate when using dithiol crosslinker.Thus, we next screened hydrogels composed of 8-arm PEG acrylate and 4-arm PEG thiol.
Thus, hydrogels formed with 8-arm PEG acrylate and 4-arm PEG thiol were used for further studies.
Table S2: Preliminary screening of PEG hydrogel for their ability to support hiPSCs viability post encapsulation.
Note: a. DTBA is a dithiol crosslinkers with two terminal thiols while 4-arm PEG thiol has 4 terminal thiols.Single hiPSCs were encapsulated in 5% w/v PEG hydrogels at a density of 2.5x10 6 cells/ml and cultured for 3 days.E8 media was supplemented with ROCKi was used.Viability staining was performed, and fluorescent images were captured on a Leica Thunder imager.Scale bars are 100 µm.Cells viability was low in all conditions.While the most viable cells were found in 8-arm gels, no significant difference was found between arms and time points.Ordinary one-way ANOVA with Tukey post-hoc correction was used to analyze each data (ns = not significant; p>0.05).
Preliminary screening of 5% w/v 8-arm PEG-Ac crosslinked with 4-arm PEG thiol hydrogels showed that although the encapsulated single hiPSCs can form aggregates, the viability remained low 3 days post encapsulation (Figure S2C&D).This may be due to a very high stiffness of these gels compared to gels formed using lower polymer concentration (1.5 -3% w/v; Figure 1 and S2A&B).For all experiments, n = 3 gels.A) Young's modulus values were estimated from G' data obtained through rheometery.B) A frequency sweep revealed these gels to be mainly elastic.C) Single hiPSCs were encapsulated at a density of 2.5x10 6 cells/ml and cultured for 3 days in E8 media supplemented with ROCKi, after which viability staining was performed.Unlike in 5% DTBA gels, aggregates formed after 3 days of culture.Scale bars are 200 µm.D) Aggregate viability on day 3 was quantified in ImageJ and presented as the ratio of live to dead cell area.

S3. Effect of ROCK inhibition on hiPSCs cultured in PEG hydrogels
Figure S3: Effects of ROCKi removal on aggregate viability.A) Cells were encapsulated in intermediate stiffness gels at a density of 2.5x10 6 cells/ml and cultured under self-renewing conditions with ROCKi.After allowing cells to aggregate for 3 days, ROCKi was removed, and the cells were grown until day 6.B) Aggregate viability on day 6 was calculated from fluorescent viability staining images and presented as the ratio of live to dead cell area.N = 2 for all gel samples.Viability data was analyzed using an unpaired, two-tailed t-test with Welch's correction (ns = not significant).

S7. Directed trilineage differentiation of WTC-11 hiPSCs
The trilineage differentiation of WTC-11 hiPSC aggregates was performed using the same procedure as described for RUES-GLR cells.Briefly, cells were first encapsulated at 2.5x10 6 cells/ml in intermediate stiffness gels and grown for 4 days under self-renewing conditions.Media was then replaced with

Figure S1 :
Figure S1:Representative images indicating hiPSCs viability post encapsulation in PEG hydrogels made using DTBA dithiol crosslinker and 4-arm or 8-arm PEG acrylate.Cell viability of hiPSCs encapsulated in PEG hydrogels composed of A) 4-arm PEG Ac-DTBA B) 8-arm PEG Ac-DTBA.Single hiPSCs were encapsulated in 5% w/v PEG hydrogels at a density of 2.5x10 6 cells/ml and cultured for 3 days.E8 media was supplemented with ROCKi was used.Viability staining was performed, and fluorescent images were captured on a Leica Thunder imager.Scale bars are 100 µm.Cells viability was low in all conditions.While the most viable cells were found in 8-arm gels, no significant difference was found between arms and time points.Ordinary one-way ANOVA with Tukey post-hoc correction was used to analyze each data (ns = not significant; p>0.05).

Figure S2 :
Figure S2: Rheological analysis and viability screening of 5% 8-arm PEG-Ac / 4-arm PEG-SH hydrogels.For all experiments, n = 3 gels.A) Young's modulus values were estimated from G' data obtained through rheometery.B) A frequency sweep revealed these gels to be mainly elastic.C) Single hiPSCs were encapsulated at a density of 2.5x10 6 cells/ml and cultured for 3 days in E8 media supplemented with ROCKi, after which viability staining was performed.Unlike in 5% DTBA gels, aggregates formed after 3 days of culture.Scale bars are 200 µm.D) Aggregate viability on day 3 was quantified in ImageJ and presented as the ratio of live to dead cell area.

Figure S4 :
Figure S3: Effects of ROCKi removal on aggregate viability.A) Cells were encapsulated in intermediate stiffness gels at a density of 2.5x10 6 cells/ml and cultured under self-renewing conditions with ROCKi.After allowing cells to aggregate for 3 days, ROCKi was removed, and the cells were grown until day 6.B) Aggregate viability on day 6 was calculated from fluorescent viability staining images and presented as the ratio of live to dead cell area.N = 2 for all gel samples.Viability data was analyzed using an unpaired, two-tailed t-test with Welch's correction (ns = not significant).C) Cells treated with ROCKi until day 6 and D) cells grown until day 6 after ROCKi removal on day 3. Aggregates with dead cells were clearly visible with the removal of ROCKi.Scale bars are 200 µm.

Figure S6 :
Figure S6: Reporter cells RUES2-GLR show lumen formation and maintain pluripotency after encapsulation in PEG hydrogels of intermediate stiffness.A) Representative bright field images showing aggregation and lumen formation in cells cultured for 4 days in PEG hydrogels of intermediate stiffness (2% gels).B) Representative fluorescent images showing expression of SOX-2 expression (green) in cells cultured in PEG hydrogels for 4 days.White arrows indicate the presence of lumen in the cell aggregates.
Stemdiff TM Trilineage Endoderm and Ectoderm media, as the manufacturer's protocol, or CHIR99021 for 48 hours.All germ layer markers were analyzed via immunofluorescence.Similar to RUES-GLR the WTC-11 cells also showed differentiation into the three germ layers as indicated by high percentage expression of respective markers.

Figure S7 :
Figure S7: Directed differentiation of WTC-11 hiPSCs encapsulated in intermediate stiffness PEG hydrogels.A) Representative fluorescent images of aggregates at the end of each differentiation procedure.B) Expression of germ layer markers Pax6, T-brachyury, and SOX17 were derived from immunofluorescence imaging.Expression was represented as the average percentage of lineage marker expressing cells, averaged across multiple samples (n≥3 gels).At least three images were taken per gel.Ordinary one-way ANOVA with Tukey post-hoc correction was used to analyze for differences in germ layer activation between lineages (ns = not significant)

Table S1 :
Antibodies used in immunofluorescent analysis of pluripotency, polarization, and trilineage differentiation capacity.