Principles of signaling pathway modulation for enhancing human naive pluripotency induction

Abstract

Isolating human MEK/ERK signaling-independent pluripotent stem cells (PSCs) with naive pluripotency characteristics while maintaining differentiation competence and (epi)genetic integrity remains challenging. Here, we engineer reporter systems that allow the screening for defined conditions that induce molecular and functional features of human naive pluripotency. Synergistic inhibition of WNT/β-CATENIN, protein kinase C (PKC), and SRC signaling consolidates the induction of teratoma-competent naive human PSCs, with the capacity to differentiate into trophoblast stem cells (TSCs) and extraembryonic naive endodermal (nEND) cells in vitro. Divergent signaling and transcriptional requirements for boosting naive pluripotency were found between mouse and human. P53 depletion in naive hPSCs increased their contribution to mouse-human cross-species chimeric embryos upon priming and differentiation. Finally, MEK/ERK inhibition can be substituted with the inhibition of NOTCH/RBPj, which induces alternative naive-like hPSCs with a diminished risk for deleterious global DNA hypomethylation. Our findings set a framework for defining the signaling foundations of human naive pluripotency.

Keywords: cross-species chimerisim; embryonic stem cells; extra-embryonic stem cells; iPSC; naive pluripotency; reprogramming.

Graphical abstract

Figure 1

Reporter systems for functional screening for enhanced human naive pluripotency conditions (A) Strategy for generating hESCs with TET-OFF regulated expression of METTL3. (B) Western blot analysis for hESCs with TET-OFF METTL3. (C) Representative images for METTL3 TET-OFF hESCs in different conditions. White arrow indicates viable PSCs at P+2. (D) Scheme depicting strategy for conducting a screen to identify additives to NHSM that allow maintaining pluripotency in TET-OFF METTL3 hESCs with DOX. (E) W3G4 cells were reverted in the presence of DOX for up to four passages in different conditions and stained for OCT4. (F) OCT4+ pluripotency maintenance in NHSM conditions supplemented with various TNKi and DOX. (G) Quantification of ΔPE-OCT4-GFP reporter by FACS in a variety of primed (red) and naive conditions (blue). MFI, mean fluorescence intensity. (H) ΔPE-OCT4-GFP naive pluripotency reporter activity in a variety of conditions with various concentrations of ACTIVIN A. (I) Representative phase-contrast images of hESCs expanded in different NHSM+TNKi-based conditions.

Figure 2

Defining enhanced human naive conditions compatible with blocking TGF/ACTIVIN signaling (A) Scheme depicting strategy for conducting a screen to identify a small molecule or cytokine additive to optimize NHSM conditions after the addition of TNKi that would allow maintaining pluripotency in TET-OFF METTL3 human stem cells (clone W3G4) and without supplementing exogenous ACTIVIN. (B) Immunostaining for OCT4+ pluripotency marker maintenance in optimized naive conditions without TGF/ACTIVIN (n = 3 per condition). (C) Summary of small molecules and their concentrations used in HENSM conditions. (D) Phase-contrast images showing naive domed-like morphology of hESCs expanded in the indicated HENSM platform. (E) Mass-spectrometry-based quantification of m⁶A on mRNA. (F) Teratoma staining and analysis. (G) Strategy for generating hESCs with TET-OFF-regulated expression of DNA methyltransferase enzyme, DNMT1. (H) Western blot analysis for DNMT1 expression in HENSM conditions supplemented with either DOX or BRAF inhibitor (SB590885). (I) DNMT1 TET-OFF ESC clone was maintained in the presence of DOX for up to four passages in different conditions and stained for OCT4. (J) Nanog+ percentage in mouse or hESCs was measured in the indicated naive and primed conditions in the absence of exogenous L-glutamine supplementation.

Figure 3

HENSM endows hPSCs with naive-like transcriptional features (A) Unbiased hierarchical clustering was performed on RNA-seq measurement obtained from different hESCs and iPSCs expanded in HENSM, HENSM-ACT naive, and primed conditions. The data were also clustered with previous independently generated RNA-seq on hPSCs expanded in 5iLA conditions (naive 5iLA^∗; Theunissen et al., 2016), reset conditions (naive reset^∗; Takashima et al., 2014) (composed of NANOG-KLF2 transgenes and 2iLGo), or primed conditions (primed^∗). (B) Principal-component analysis (PCA) of samples represented in (A). (C) RT-PCR analysis for naive pluripotency markers. Primed expression levels were set as 1. Student’s t test, ^∗p < 0.01. (D) Gene set enrichment analysis (GSEA) based on comparison with gene sets produced from differentially expressed genes (DEGs) in different developmental conditions. (Yan et al., 2013) and (Tyser et al., 2020) positive and negative enrichment scores enriched in HENSM, 5iLA, and primed DEGs are shown. (E) Uniquely accessible regions were recognized in 2/8-cell and ICM of human early development (Wu et al., 2018). The percentages of upregulated regions (fold change [FC] > 2) in naive (blue) or primed (red) are indicated. Analysis was done for three distinct naive systems: HENSM, t2iL/Gö, and 5iLA. Accessible regions in naive tend to have a higher overlap with ICM regions than with 2- or 8-cell regions.

Figure 4

Cross-species chimerism and epigenetic hallmarks in HENSM conditions (A) Schematic and FACS results following using WIBR2 (female 46XX) 29-8 hESC line that carries GFP and tdTomato on each of the X chromosomes in the MECP2 locus, in the indicated conditions. (B) Average methylation as calculated from primed samples and naive samples that were maintained in various HENSM conditions, along with previously published reset-naive samples and human ICM samples. (C) Global methylation histogram measured on the same samples as in (B). Dark blue, percentage of highly methylated CpGs (>0.9 methylation level); light blue, percentage of lowly methylated CpGs (<0.1 methylation level). (D) Western blot analysis for DNA methylation regulators: DNMT1 and UHRF1 enzymes. (E) Representative images of whole-mount in toto imaged mouse embryos after microinjection of the indicated naive hiPSCs are shown in comparison to non-injected WT embryos. White squares in tiles outline zoomed-in regions in subsequent panels. GFP staining was used to trace hiPSC-derived progeny and CellTracker and Hoechst as counter staining. (F) Representative images of IHC for TUJ1, SOX17, SOX2, and GFP of injected (upper panels) and non-injected E15.5 mouse embryos (lower panels) for different indicated anatomical regions are shown. GFP served as human cell tracer. GFP and TUJ1, SOX17, and SOX2 overlap as shown also in merged zoomed-in regions of tissues, depicted in red squares in the tiles. White arrowheads in insets depict co-localization of GFP and lineage markers. Tile scale bars, 200 and 100 μm. Zoomed-in scale bar, 50 μm.

Figure 5

Signaling principles for inducing human naive pluripotency (A) FACS results following using WIBR2 (female 46XX) 29-8 hESC line that carries GFP and tdTomato on each of the X chromosomes in the MECP2 locus. Indicated components from HENSM conditions were individually depleted or added during induction phase, and cells were subjected to FACS analysis after 10 days. (B) Quantification of ΔPE-OCT4-GFP naive pluripotency reporter in variety of induction conditions after three passages. (C) Quantification of ΔPE-OCT4-GFP knock-in naive pluripotency reporter in βCAT^+/+ and isogenic βCAT^−/− cells in variety of conditions after three-passage transfer from HENSM starting conditions. (D) Measuring WNT activity via luciferase reporter assay. Values were normalized to HENSM no-TNKi conditions (defined as 1; n = 3 per condition). Student’s t test, ^∗p < 0.01. (E) Morphological changes in βCAT^+/+ and isogenic βCAT^−/− cells after two passages from induction of naive pluripotency in the presence or absence of TNKi. (F) Mouse and human βCAT^−/− ESCs were rendered transgenic with a validated tamoxifen-inducible βCAT⁻ERT transgene. Acquisition or loss of naive domed-like morphology was assayed after two passages of the indicated conditions. (G) FACS analysis for naive ΔPE-OCT4-GFP pluripotency maker expression in human WIBR3-ΔPE-βCAT KO; βCAT ERT-Tg line before and after tamoxifen addition for 48 h. (H) Mean normalized gene expression of selected WNT ligands in naive and primed hPSCs. (I) RT-PCR expression of naive markers in LIS41 hESCs expanded in HENSM conditions with IWP2 instead of XAV939 (relative to primed cells set as 1). (J) X chromosome reactivation status in 29-8 reporter cell line expanded in HENSM containing IWP2 as WNTi instead of TNKi (XAV939).

Figure 6

KLF17 is a key promoter of human naive pluripotency (A) Motif enrichment in HENSM-specific and primed-specific accessible chromatin regions (n = 20,642 and b = 36,927, respectively). Motif families are indicated at the right. Shades represent enrichment fold change. (B) Primed WT and KO hESCs for KLF4 and/or KLF17 were reverted in HENSM and HENSM-ACT conditions, imaged, and assayed for OCT4-GFP by FACS analysis (values indicated per frame). (C) WT and KO mESCs for Klf17 were expanded in 2iL, phase imaged, and assayed for OCT4-GFP by FACS analysis (values indicated per frame). (D) Genomic annotation of binding sites of KLF17, KLF4, TFAP2C, NANOG, OCT4, and SOX2 measured in HENSM naive or primed conditions, showing the preference of KLF17 to bind promoters, compared to all TFs that prefer to bind enhancers (distal intergenic + introns annotations). (E) KLF17 binding pattern (average Z score + STD) in naive-specific regions (n = 20,642, blue), primed-specific regions (n = 36,927, red), or all promoters (n = 43,463, gray). (F) Overlap between target genes of the indicated TFs, showing highest overlap between Sox2 and Oct4 target genes (p ~0) and a relative lower overlap with KLF17 and KLF4 targets (p < 10⁻²¹). Scaled p value is presented (Method details). (G) ChIP-seq profile of KLF17, OCT4, and SOX2 in selected regions and different conditions in hPSCs. IGV range common to all tracks is indicated.

Figure 7

NOTCH/RBPj inhibition generates alternative human naive cells without using MEK/ERKi (A) FACS analysis showing status of X activation in female 29-8 cells following decreasing concentrations of ERKi in HENSM conditions. (B) Schematic showing screen strategy for finding small molecule supplements that could allow maintaining GFP+/tdTomato+ 29-8 cells in HENSM conditions in which ERKi is completely omitted (0HENSM) or partially depleted (tHENSM). (C) FACS analysis following supplementing 0HENSM or tHENSM with DBZ (NOTCHi). (D) Summary of small molecules and their concentrations used in the optimized tHENSM and 0HENSM conditions. (E) ΔPE-OCT4-GFP naive pluripotency reporter in HENSM, tHENSM, and 0HENSM conditions. (F) Phase images of WIBR1 hESCs in different naive pluripotency conditions. (G) RT-PCR analysis for naive pluripotency markers in different naive and primed conditions. Values were normalized to ACTIN and GAPDH. Primed expression levels were set as 1. (H) PCA of gene expression profiles of cells grown in primed condition (red) or in HENSM, HENSM-ACT, tHENSM, or 0HENSM conditions (blue shades). (I) Global methylation histogram calculated from primed samples and naive samples that were maintained in various conditions. Dark blue, percentage of highly methylated CpGs (>0.9 methylation level); light blue, percentage of lowly methylated CpGs (<0.1 methylation level); yellow dots, sample methylation average. (J) Western blot analysis for phosphorylated ERK levels in the indicated growth conditions. (K) RT-PCR analysis for changes in naive pluripotency marker expression in the indicated naive conditions with or without RBPj small molecule inhibitor RIN1. Student’s t test, ^∗p < 0.05; ^∗∗p < 0.01; NS, not significant. (L) RT-PCR analysis for changes in naive pluripotency marker expression in RBPj^+/+ and RBPj^−/− hESCs upon removal of DBZ from 0HENSM conditions. Values were normalized to ACTIN and GAPDH. RBPj^+/+ 0HENSM no-DBZ sample expression levels were set as 1. Student’s t test, ^∗p < 0.01; NS, not significant.