StemBond hydrogels control the mechanical microenvironment for pluripotent stem cells

Studies of mechanical signalling are typically performed by comparing cells cultured on soft and stiff hydrogel-based substrates. However, it is challenging to independently and robustly control both substrate stiffness and extracellular matrix tethering to substrates, making matrix tethering a potentially confounding variable in mechanical signalling investigations. Moreover, unstable matrix tethering can lead to poor cell attachment and weak engagement of cell adhesions. To address this, we developed StemBond hydrogels, a hydrogel in which matrix tethering is robust and can be varied independently of stiffness. We validate StemBond hydrogels by showing that they provide an optimal system for culturing mouse and human pluripotent stem cells. We further show how soft StemBond hydrogels modulate stem cell function, partly through stiffness-sensitive ERK signalling. Our findings underline how substrate mechanics impact mechanosensitive signalling pathways regulating self-renewal and differentiation, indicating that optimising the complete mechanical microenvironment will offer greater control over stem cell fate specification.

d -Cell attachment on fibronectin-coated tissue culture plastic (TCP), stiff hydrogels and soft mid AHA hydrogels. 175'000 cells were plated for 24hrs in culture in 2i or serum+LIF medium before being detached and counted (n = 2 independent samples).    Figure 3): mESC lineage commitment is not impaired on soft and stiff substrates a -(Left) Flow cytometry gating strategy: singlets were selected in gate "Singlets", then debris were gated out in gate "G1", and finally population of high and low Rex1 expressing cells were determined on the histogram. The counts and % of parent population are given for the example shown. (Right) Flow cytometry profiles of reporter line Rex1GFP::d2 cultured on TCP (grey), stiff (red) or soft (black) high AHA StemBond hydrogels for 0hr (left), 24hrs (centre) and 48hrs (right) in N2B27. The green filled histogram is a negative control. For each condition, histograms of two replicate samples were averaged and smoothed.

Supplementary Figure 4 (in relation to
b -(Top) Gene expression of neuroectoderm markers Pax3, Pax6 and Sox1, and naïve marker Nanog after differentiation in N2B27 on soft and stiff (high AHA) StemBond hydrogels, and on TCP. Expression are normalised to TCP (day 3). Note that the neural markers were very lowly expressed in 2i, and not expressed in mesoendoderm differentiation conditions (not shown). (Bottom) Gene expression of mesoendoderm markers Eomes, Foxa2, T/Brachyury, and naïve marker Nanog after differentiation in Activin A + CHIRON on soft and stiff (high AHA) StemBond hydrogels, and on TCP. Expression is normalised to TCP (day 3). Note that mesoderm markers were very lowly expressed in 2i, and not expressed in neurectoderm differentiation (not shown). Bars show mean +/-standard error (n = 2 independent samples).
(Right) Quantification of the number of AP positive colonies after 3 passages as % of replated cells (n=8). In all panels error bars mean +/-show standard deviation and P-values indicate the outcome of an ANOVA test.
g -Quantification of pERK activity on TCP, soft and stiff StemBond hydrogels in HNES1 cells cultured in complete media PDLGX or in LGX media lacking PD03. (Top) Western Blot showing total ERK protein (red) and phospho-ERK1/2 (green). (Bottom) Mean +/-standard deviation of phospho-ERK normalised to total ERK levels. n = 2. Cells were seeded in PDLGX on laminin for 2 days before PD03 was removed. Cells were lysed 3 days after PD03 removal.
h -Expression (mean +/-standard deviation) of naïve pluripotency markers Klf4, Klf17 and Tfcp2l1 in HNES1 cells seeded on TCP and StemBond hydrogels. Cells were seeded in PDLGX on laminin for 2.5 days before PD03 was removed. Cells were lysed 2 and 4 days later and RNA extracted. Gene expression was measured by RT-qPCR and normalised to Gapdh. n = 3. P values are from one-way ANOVA. n.s.: non-significant.

): Reprogramming of EpiSCs is complete after 8 days on StemBond hydrogels
Expression (mean +/-std) of naïve pluripotency genes in iPSCs after 1 passage on TCP in 2i+LIF+bsd, following reprogramming on TCP, stiff or soft substrates. Expression for each replicate was normalised relative to Gapdh then to the highest value across all samples. n = 2 independent experiments. a -Gene expression (mean +/-std) of naïve pluripotency genes Esrrb, Nanog and Tfcpl21 for cells cultured for 24hr in serum+LIF on fibronectin-coated TCP soft and stiff hydrogels and low AHA (light blue), mid AHA (cyan) and high AHA (dark blue) concentrations. Expressions were normalized to TCP (N=3). P values indicate significant differences (n-way ANOVA test) due to different stiffness. AHA levels did not lead to significant differences (P > 0.05).
b -Network visualization of enriched biological processes due to substrate stiffness. In red are the processes enriched in upregulated genes, in blue the processes enriched in downregulated genes. Node size is proportional to the number of genes involved, and colour-charts inside each node indicate the adhesiveness for which the pathways were significantly enriched (padj < 0.1). Edge width indicates the similarity coefficient between nodes.
c -Intersection of significantly modulated genes (p < 0.05, abs( log2fc )> 0, FPKM > 1) between soft and stiff substrates in all media conditions. In bold are the numbers of systematically modulated genes (i.e. in at least 3 conditions). Examples of some genes are given on the side with font size proportional to the average up/downregulation.    Figure 6f): ERK activity depends on substrate stiffness, but not on other environmental variables a -Immunofluorescence for phospho-ERK (red) and DAPI (grey) on stiff (left) and soft (right) substrates. Cells were seeded on the gels for 24hrs in serum+LIF. Scale bar 20µm.
b -Pie charts giving the percentage of genes whose expression was higher on soft (grey) or on stiff (black) at 2hrs (left) and 12hrs (right) after removing PD03. Percentages are relative to all the ERKregulated genes at 2hrs, and to the ERK-regulated genes with|log2 (t=12hrs/t=0hr)| > 1 at 12hrs.
c -Average log2 fold change of expression between soft and stiff for the ERK-activated, ERKsuppressed and all expressed genes. P values obtained by a two-sided permutation test over random samples of all expressed genes. n.s.: non-significant.
d -Average fold change of expression on soft vs stiff hydrogels for ERK−targets in serum+LIF, 24hrs. ERK targets identified from PD03 removal of 2i+LIF at 2hrs or 12hrs (see Figure 6f). Fold changes for all genes (target genes and non−target genes) are shown for comparison. (Left) The expression of those targets is taken from data set of Figure 6a-b (cells seeded on different substrates). (Right) The expression of those targets is taken from data set of Figure 6c. Statistics show significant difference to the distribution of fold changes of all genes (P values computed using a one-sided permutation test), n.s.: non-significant. Statistics are computed over n = 117, 34, 176, 130 genes for each of the categories.