Encapsulation boosts islet-cell signature in differentiating human induced pluripotent stem cells via integrin signalling

Abstract

Cell replacement therapies hold great therapeutic potential. Nevertheless, our knowledge of the mechanisms governing the developmental processes is limited, impeding the quality of differentiation protocols. Generating insulin-expressing cells in vitro is no exception, with the guided series of differentiation events producing heterogeneous cell populations that display mixed pancreatic islet phenotypes and immaturity. The achievement of terminal differentiation ultimately requires the in vivo transplantation of, usually, encapsulated cells. Here we show the impact of cell confinement on the pancreatic islet signature during the guided differentiation of alginate encapsulated human induced pluripotent stem cells (hiPSCs). Our results show that encapsulation improves differentiation by significantly reshaping the proteome landscape of the cells towards an islet-like signature. Pathway analysis is suggestive of integrins transducing the encapsulation effect into intracellular signalling cascades promoting differentiation. These analyses provide a molecular framework for understanding the confinement effects on hiPSCs differentiation while confirming its importance for this process.

Figure 1

Comparison of the hiPSC differentiation outcome according to the stage of encapsulation. (a) Scheme depicting the three cell populations analysed by immunofluorescence. (b) Proportion of the differentiated hiPSC-cells expressing insulin, glucagon or somatostatin in the three distinct populations analysed, quantified by Imaris software. (c) Proportion of bihormonal cells in the three distinct populations analyzed. (d) Proportion of the differentiated hiPSC-cells expressing PDX1 or NKX6.1 in the three distinct populations analysed. (e) Proportion of insulin + cells coexpressing PDX1 or NKX6.1. (f) High magnification confocal images of cells inside alginate capsules stained for insulin (green), glucagon (red), somatostatin (purple) and DAPI (blue) by whole mount immunofluorescence. (g) Whole mount immunofluorescence of encapsulated cells stained for insulin (green), NKX6.1 (red), PDX1 (purple) and DAPI (blue), gamma correction 0.4. Scale bars: 10 µm. Graphs data are shown as mean ± SEM.

Figure 2

Global proteome analysis of hiPSC differentiating either on Matrigel or encapsulated in alginate capsules. (a) Scheme illustrating the cell populations and differentiation stages considered for global proteomics. (b) Experimental design of the conditions compared in TMT 11-plex proteomics. (c) Hierarchical clustering of normalized TMT-ratios (n = 2, 2, 2, 2, 3). (d) Analysis workflow depicting the comparisons employed and the assessed corresponding effect. (e) The number of proteins showing a dynamic of regulations compatible with an islet-promoting pattern in response to each of the four effects considered. Arrows depict the generic prerequisite direction of regulation for group inclusion. (f) Pies charts depicting the proportion of proteins following islet-promoting and islet-antagonizing regulation patterns in each of the four effects considered. The graph bars represent the proportion of proteins reaching abundance levels indistinguishable from those detected in native human islets.

Figure 3

Pathway analysis and heatmaps of proteins following islet-promoting regulation patterns in response to the different effects assessed. (a) Tables depicting the top canonical pathways and predicted upstream regulators in response to Differentiation Cocktail Effect, (b) Heatmap representing the regulation of selected markers in response to Differentiation Cocktail Effect (c) Tables depicting the top canonical pathways and predicted upstream regulators in response to Confounding Effect, (d) Heatmap representing the regulation of selected markers in response to Confounding Effect, (e) Tables depicting the top canonical pathways and predicted upstream regulators in response to Early Encapsulation Effect, (f) Heatmap representing the regulation of selected markers in response to Early Encapsulation Effect, (g) Tables depicting the top canonical pathways and predicted upstream regulators in response to Late Encapsulation Effect, (h) Heatmap representing the regulation of selected markers in response to Late Encapsulation Effect (orange A – predicted activation, blue I – predicted inhibition, red # signals shared pathways or predicted upstream regulators between left and right effects).

Figure 4

Pathways analysis of proteins displaying islet-promoting regulation patterns in response to more than one effect. (a) IPA-generated tables of the Top 5 canonical pathways for proteome landscapes regulated in response to all three effects considered (purple, left Venn diagram) as well as to Differentiation Cocktail Effect and Early Encapsulation Effect solely (green, middle Venn diagram). Selected regulated pancreatic islet markers are shown for the proteome landscape responding to Differentiation Cocktail Effect and Late Encapsulation Effect solely (yellow, right Venn diagram). (b) IPA-generated tables of top canonical pathways and selected regulated pancreatic islet markers for proteome landscapes regulated by encapsulation regardless of differentiation stage of encapsulation (purple, left Venn diagram), only by encapsulation during the early stages (S0- > S5) of differentiation (green, middle Venn diagram) and only by encapsulation during the late stages (S5- > S7) of differentiation (yellow, right Venn diagram). (c) Selected IPA circular networks for the proteome regulated exclusively by either Early Encapsulation Effect (left) or Late Encapsulation Effect (right). The heatmap represents the direction of regulation towards islet-abundance values in S5^bead[S0-S5] compared to S5 (left) and S7^bead[S5-S7] compared to S7 population. Orange A – predicted activation, blue I – predicted inhibition, green – observed downregulation, red – observed upregulation. The circular networks and the top canonical pathways were generated through the use of IPA (QIAGEN Inc., https://www.qiagenbio-informatics.com/products/ingenuity-pathway-analysis).

Figure 5

Pathway analysis of the proteome landscape generated by the direct comparison between cells differentiating in alginate capsules and Matrigel differentiated-cells. (a) Analysis workflow and IPA-generated tables of the top canonical pathways predicted activated or inhibited at stage 5 of differentiation between S5^bead[S0-S5] as compared and S5-cells. (b–d) IPA-generated graphical representation of RhoGDI signalling, PI3K/AKT Pathway and Integrin Signalling, (e) Analysis workflow and IPA-generated tables of the top canonical pathways predicted activated or inhibited at stage 7 of differentiation between in S7^bead[S5-S7] as compared and S7-cells (blue - predicted inhibited, orange – predicted activated, green – observed downregulation, red – observed upregulation, arrow heads point to integrin involvement, magenta # points at shared top pathways between the comparisons). The pathways were generated through the use of IPA (QIAGEN Inc., https://www.qiagenbio-informatics.com/products/ingenuity-pathway-analysis).