Fig 2: Cell–cell communication analyses of human–mouse chimeric embryos reveal that laminin and collagen signaling-related pathways are dominant in PODXL overexpressing cells. A) The total number of interactions and interaction strength were displayed as bar plots showing the chimeric human RUES1-EPSCs and the chimeric mouse embryos. B,C) In the cell–cell communication networks, circle plots of the number of interactions or interaction strength between -Dox and +Dox conditions are shown. Red (or blue) colored edges represented increased (or decreased) signaling in the +Dox compared to the -Dox conditions in the chimeric human RUES1-EPSCs or the chimeric mouse embryos, respectively. D,E) Scatter plots showing significant changes in “outgoing” or “incoming” signals between -Dox and + Dox chimera human RUES1-EPSCs or between -Dox and +Dox chimera mouse cells, respectively. F) Scatter plot showing the differential signals that change from -Dox to +Dox in the Pre-EPI cell population and Early TE-like cell population in the chimeric human cells. Positive values indicated the increase in the +Dox chimera human RUES1-EPSCs while negative values indicated the increase in the -Dox chimeric human cells. G) Heatmap plot showing the “incoming” signals in different cell populations between -Dox and +Dox chimeric mouse cells. H) Scatter plot showing the signal changes from -Dox to +Dox and associated with Polar TE/ExE populations in the chimeric mouse cells. Positive values indicated the increase in the +Dox chimera mouse cells while negative values indicated the increase in the -Dox chimeric mouse cells. I) Significant signaling pathways ranked based on differences in the overall information flow within the inferred networks between -Dox and +Dox in the chimeric human RUES1-EPSCs (right) and the information flow as converted into a 1.0 scale (left). The top signaling pathways colored red were enriched in the -Dox condition and those colored blue were enriched in the +Dox condition. The right panel showed collagen and laminin which were upregulated in +Dox cells shown by information flow (please refer to method details). J) Bubble plot showing the significant context-specific signals upregulated or downregulated from the Early TE-like cell population to other cell clusters in the chimeric human RUES1-EPSCs.

Fig 3: Inducing PODXL overexpression extends pluripotency potential in vitro and teratoma formation in vivo. A) Schematic view of the setup of an inducible Tet-on-PODXL vector in hEPSCs. B) qRT-PCR assays accessing the cleavage-stage specific gene expression levels in RUES1 derived-hEPSCs after inducing PODXL overexpression (+Dox) or cholesterol supplementation for 3 days. Target gene expression in each sample was compared to that in RUES1 hESCs. The error bars indicate the averages ± SDs (n = 3). One-way ANOVA tests with Dunnett's multiple comparison test were performed (***p < 0.001, **p < 0.01, *p < 0.05). C) Western blotting shows the upregulation of ZSCAN4 after inducing PODXL overexpression in RUES1-EPSCs after treatment with doxycycline (+Dox) for 3 day in hEPSCs. +: Dox, 2 × 10-6 m; ++: Dox, 4 × 10-6 m. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA (***p < 0.001, **p < 0.01, *p < 0.05) with Dunnett's multiple comparisons test performed relative to -Dox EPSCs of target proteins. D) Representative images of hEPSCs and hESCs (RUES1 and RUES2) for P3 cells after TSC induction. CT27 and bTS11 are TSC-positive controls.[ 22 ] CT27: TSCs derived from a trimester placenta, bTS11: TSCs derived from a blastocyst. E) qRT-PCR analyses for trophoblast marker genes in primed RUES2 hESCs, RUES2-EPSCs -Dox, and RUES2-EPSCs +Dox cultured in hTSC medium were measured. Relative gene expression levels were compared to primed RUES2 hESCs cultured in TSC medium. Error bars indicate averages ±SD (n = 3). P-values were determined with an unpaired Student's t-test (***p < 0.01, **p < 0.01, *p < 0.05) relative to the corresponding -Dox control group. F) Teratomas generated from engrafted cells (RUES1-hESCs, RUES1-EPSC -Dox, RUES1-EPSC +Dox) by subcutaneous injection. H&E staining and immunohistochemistry analyses for four germ layers in teratomas are shown as indicated.

Fig 4: PODXL is necessary for hPSC self-renewal and viability. A) Bright-field images of shRNA-treated HUES6 hESCs. Scale bar, 200 µm. B) Alamar blue (AB) assays and ALP activity assays (a pluripotent marker) were performed in shRNA-treated HUES6 and H9 cells. ALP levels were normalized to the relative cell numbers measured in the AB assays. The error bars indicate the averages ± SDs (n = 8 in HUES6; n = 6 in H9). P-values were determined with one-way ANOVA (****p < 0.0001) with Dunnett's multiple comparison test performed relative to shRFP hESCs. C) Quantification of c-MYC and TERT in shRNA-treated HUES6 measured by Western blotting. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA (**p < 0.01, *p < 0.05) with Dunnett's multiple comparison test performed relative to shRFP hESCs. D) Measurements of apoptosis/necrosis, as demonstrated by fluorescence-assisted cell sorting (FACS) analyses using annexin V-PI in HUES6. E) Quantification of annexin V- and PI-positive cells. The error bars indicate the averages ± SEMs (n = 4). P-values were determined with one-way ANOVA (**p < 0.01, *p < 0.05) with Dunnett's multiple comparison test performed relative to shRFP hESCs. F) Flow chart of iPSC generation. (Created with BioRender.com) CRL-2097 fibroblasts were stably infected with lentivirus expressing shRNAs and lentivirus encoding four reprogramming factors at the same time. 4 days after infection, cells were split into matrigel-coated plates in a feeder-free condition. G) Reprogrammed colonies were stained to assess ALP, and ALP-positive colonies were counted. After 20 days of virus induction, ALP staining was performed and ALP-positive colonies were counted with Image J software. The error bars indicate the averages ± SDs (n = 3). P-values were determined with one-way ANOVA (***p < 0.001, **p < 0.01) with Dunnett's multiple comparison test performed relative to shRFP hESCs. H) Bright-field images of shRNA-treated H9 and HUES6-derived EPSCs. KSR: 5% KnockOut serum replacement. I) AB assays and ALP activities were assessed in shRNA-treated HUES6- or H9-derived hEPSCs. The ALP activity levels were normalized to the relative cell numbers (AB assays). The error bars indicate the averages ± SDs (n = 8 in H9-EPSCs; n = 12 in HUES6-EPSCs). P-values were determined with one-way ANOVA (****p < 0.0001, ***p < 0.001, **p < 0.01) with Dunnett's multiple comparison test performed relative to shRFP hESCs. J) c-MYC and TERT were downregulated after PODXL knockdown in HUES6-derived EPSCs as observed by Western blotting. The error bars indicate the averages ± SEMs (n = 5). P-values were determined with one-way ANOVA (****p < 0.0001, ***p < 0.001) with Dunnett's multiple comparison test performed relative to shRFP hESCs.

Fig 5: PODXL regulates de novo cholesterol biosynthesis and exogenous addition of cholesterol rescues PODXL knockdown-dependent loss of pluripotency. A) Gene expression analyses of PODXL-overexpressing HUES6 hESCs using cDNA microarrays. B) Schematic of the cholesterol biosynthesis pathway. Inhibitors used in this study are illustrated targeting each specified enzyme. C,D) mRNA levels of cholesterol genes were analyzed by qRT-PCR in shPODXL HUES6 hESCs. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05) with Dunnett's multiple comparison test performed relative to shRFP hESCs. For PODXL-overexpressing hESCs, the error bars indicate the averages ± SEMs (n = 3), and p-values were determined with an unpaired Student's t-test (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05). E) Quantification of Western blots showing HMGCR, SREBP1, SREBP2, c-MYC, and TERT levels in PODXL-overexpressing HUES6 cells. The error bars indicate the averages ± SEMs (PODXL, TERT, c-MYC, HMGCR, and GAPDH, n = 5; SREBP1 and SREBP2, n = 3). P-values were determined with an unpaired Student's t-test (***p < 0.001, **p < 0.01, *p < 0.05). F) Quantification of Western blots showing HMGCR, SREBP1, SREBP2, c-MYC, and TERT levels in shPODXL-transduced HUES6 cells. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05) with Dunnett's multiple comparison test performed relative to shRFP hESCs. G) Cellular cholesterol content was determined with an Amplex Red assay kit. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA (**p < 0.01) with Dunnett's multiple comparison test performed relative to shRFP hESCs. For PODXL overexpression, p-values were determined with an unpaired Student's t-test (**p < 0.01) relative to GFP control cells. H) Bright-field images of HUES6 hESCs treated separately with simvastatin, AY9944, and MBCD for 3 days. Scale bar, 200 µm. I) IC50 analyses of the three inhibitors in hESCs, hBMSCs, hNSCs, and fibroblasts assessed by cell growth (AB assays) under serum-free or serum-supplemented conditions. J) Bright-field images of HUES6 hESCs with PODXL knockdown supplemented with cholesterol for 7 days. Scale bar, 200 µm. K) Alamar blue and ALP activity assays were performed in HUES6 hESCs with PODXL knocked down supplemented with cholesterol. The error bars indicate the averages ± SDs (n = 3). P-values were determined with one-way ANOVA with Dunnett's multiple comparison test (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05) performed relative to shRFP hESCs in solvent condition (0). L) Quantification of Annexin V-/PI-positive cells in HUES6 hESCs with PODXL knocked down supplemented with cholesterol. The error bars indicate the averages ± SEMs (n = 3). P-values were determined with one-way ANOVA with Dunnett's multiple comparison test (ns, not significant) performed relative to shRFP hESCs in solvent condition (0% w/v). M) Quantification of Western blots was performed in HUES6 hESCs with PODXL knocked down supplemented with cholesterol for 7 days. The error bars indicate the averages ± SEMs (n = 4). P-values were determined with one-way ANOVA (ns, not significant) with Dunnett's multiple comparison test performed relative to shRFP hESCs in solvent condition (0% w/v) for each of the target proteins.