Tankyrase inhibition promotes a stable human naïve pluripotent state with improved functionality

Abstract

The derivation and maintenance of human pluripotent stem cells (hPSCs) in stable naïve pluripotent states has a wide impact in human developmental biology. However, hPSCs are unstable in classical naïve mouse embryonic stem cell (ESC) WNT and MEK/ERK signal inhibition (2i) culture. We show that a broad repertoire of conventional hESC and transgene-independent human induced pluripotent stem cell (hiPSC) lines could be reverted to stable human preimplantation inner cell mass (ICM)-like naïve states with only WNT, MEK/ERK, and tankyrase inhibition (LIF-3i). LIF-3i-reverted hPSCs retained normal karyotypes and genomic imprints, and attained defining mouse ESC-like functional features, including high clonal self-renewal, independence from MEK/ERK signaling, dependence on JAK/STAT3 and BMP4 signaling, and naïve-specific transcriptional and epigenetic configurations. Tankyrase inhibition promoted a stable acquisition of a human preimplantation ICM-like ground state via modulation of WNT signaling, and was most efficacious in efficiently reprogrammed conventional hiPSCs. Importantly, naïve reversion of a broad repertoire of conventional hiPSCs reduced lineage-primed gene expression and significantly improved their multilineage differentiation capacities. Stable naïve hPSCs with reduced genetic variability and improved functional pluripotency will have great utility in regenerative medicine and human disease modeling.

Keywords: Differentiation; Ground state; Human embryonic stem cell; Induced pluripotent stem cell; Naïve pluripotency.

Fig. 1.

Multilineage differentiation and genetic variability of conventional hiPSCs derived from unique donors via different reprogramming methods. (A) Hematopoietic differentiation [% day 14 human embryoid body (hEB) cells] of independent donor-derived hPSCs (see Table S2). hESC lines: H9, H7, H1, ES03 (gold). Fibro-iPSCs: IMR4, HUF3, HUF5, 7ta, WT2 (green). 7F-E and 4F-E hematopoietic-iPSCs: lines generated without stromal activation (sa) were mononuclear CB-7F-E (iCB 2.5, iCB8, iCB9; purple) and MP-CB-4F-E (E17C1, E20C2, E32C7; purple) and lines generated with sa were saMPCB-4FE (E5C3, E12C5, E17C6; red). Included are MP-PB-4FE lines from the same donor (ZPB) that were derived with sa (E29C1, E29C4, E29C6; red) and without sa (E29C10, E29C11, E29C12; purple). Four 7F-E CB-derived sa-MP-iPSC lines were all from the same donor (D003) generated with sa (6.2, 6.13, 19.11) (Fig. S2). *P<0.05, **P<0.005 (one-way ANOVA) for averaged groups relative to the hESC group. (B) Osteogenic differentiation (quantitative Alizarin Red dye), and (C) endodermal differentiations of individual lines in each hPSC class from the same unique donors as above. Error bars indicate s.e.m. (D) Number of genes differentially expressed between hiPSCs and hESCs (# hiPSC ≠ hESC) (ANOVA, P<0.05; fold change >1.5×). (E) Dendrogram of hPSCs from above clustered on lineage-primed gene targets of the PRC2 complex (Table S1). (F) PCA of ESC and core module genes (Table S1) showing variance in gene expression among donor cells and the hPSC lines shown in Fig. S2A. Red spheres indicate early passage sa-CB-derived MP-iPSCs (n=12, average passage=11.4); gold circles indicate hESCs (n=5, average passage=58.5); green circles indicate AdF-iPSCs (n=9, average passage=20.4); gray diamonds indicate donor fibroblasts; gray triangles indicate donor CB cells.

Fig. 2.

Tankyrase inhibition promotes stable reversion of hPSCs to an mESC-like pluripotency in classical WNT-MEK/ERK 2i conditions. (A) Clonal passaging of conventional hPSCs in LIF-3i permitted stable reversion to dome-shaped colony morphology, with bi-directional reversion of phenotype when re-cultured in bFGF. (B) qRT-PCR analysis of pluripotency-associated transcripts of (female) sa-MP-iPSC line 6.2 cultured before (primed) and after four single-cell passages in LIF-3i. Transcript levels of NANOG, STELLA, NR5A2 and XIST are shown. **P<0.001 (paired t-test); NS, not significant. (C) Representative SSEA4⁺ TRA-1-81⁺ expression of stable LIF-3i-reverted hPSCs. (D) Confocal immunofluorescence microscopy of LIF-3i-cultured hPSCs. Shown are representative nuclear and cytoplasmic protein co-expressions of NANOG, OCT4, TRA-1-81 and SSEA4 with naïve-specific factors (E-cadherin, NR5A2 and STELLA). (E) Representative teratoma of LIF-3i-reverted E5C3 (Table S3B). Mesoderm (Meso; e.g. cartilage), ectoderm (Ecto; e.g. pigmented retinal epithelium) and endoderm (Endo; e.g. glandular epithelium) are indicated. (F) Immunofluorescence (right) and quantitation per field (left) of X-chromosome activation status of primed versus LIF-3i-reverted line 6.2 with XIST RNA-FISH probe; normal human dermal fibroblasts (NHDF) provided controls. Error bars indicate s.e.m. Scale bars: 500 µm in A; 50 µm in D; 100 µm in E; 20 µm in F.

Fig. 3.

Acquisition of mESC-like signaling pathways in LIF-3i-reverted hPSCs. (A) Stabilization of nuclear P-STAT3 and cytoplasmic activated (activ) β-catenin in LIF-3-reverted hPSCs. Nuclear (Nu), cytoplasmic (Cy) and total (Tot) fractions are shown for H9 hESC, E5C3 sa-MP-iPSC and C1.2 fibro-iPSC lines. TBP and actin are protein loading controls. +, LIF-3i; −, primed culture. (B) Inhibition/proliferation assays of sa-MP-iPSC line E5C3. Shown are mESC signaling pathways [e.g. LIF/JAK/STAT, LIF-receptor/gp130, CREB, BMP4 (dorsomorphin) and PI3K]. Cumulative proliferations of SSEA4⁺ TRA-1-81⁺ E5C3 cells were measured after 12 days of culture in the presence of indicated inhibitors, normalized to LIF-3i-alone conditions (at 100%). (C) Western blots of key WNT components AXIN1 and activated β-catenin in both nuclear and cytoplasmic fractions in primed (−) versus LIF-3i reverted (+) hPSC cultures. (D,E) Confocal microscopy of WNT proteins in indicated primed (bFGF) versus LIF-3i-reverted hPSC lines. Activated β-catenin is shown to be distinctly sequestered outside of the nuclear compartment, whereas uniform expression of OCT4 was localized strictly within nuclei. Scale bars: 20 µm.

Fig. 4.

Transcriptional and epigenetic profiling of LIF-3i-reverted hPSCs. (A) Genome-wide cross-species hierarchical clustering. Shown is a dendrogram of expression microarrays of mESC [serum/LIF; naïve (LIF-2i)], primed mEpiSC, and isogenic hPSC samples from this study before (hPSC primed) and after 5 passages in LIF-3i (hPSC naïve). Human PSC lines (n=12) included: three hESC lines H9, H7 and ES03 (gold); six sa-MP-iPSC lines E5C3, E5C1, E17C6, LZ6+2, LZ6+10 and 6.2 (red); and three fibro-iPSC lines 7ta, C1.2 and C2 (green). (B) (Top) CpG methylation. Box plot shows beta values of genome-wide autosomal differentially methylated region (DMR) CpG probes from Infinium methylation arrays [16,282 of 473,864 autosomal probes significantly (P<0.05) differentially methylated (SD>0.15); see supplementary Materials and Methods for further details] in the same isogenic primed (−) versus LIF-3i-reverted (+) hPSC samples used for the microarrays above. Gold, hESCs (n=3); red, sa-MP-iPSCs (n=6); green, fibro-iPSCs (n=3). ***P<0.001 (paired two-way t-test). (Bottom) Global 5MC and 5hMC levels from dot blot immunoassays (relative to primed) for representative LIF-3i-reverted hPSCs. Genomic DNA samples were collected before (−) and after (+) LIF-3i reversion from H9 (gold), E5C3 (red) and C1.2 (green). (C) Activities of proximal enhancer (PE) and distal enhancer (DE) elements of the human OCT4 promoter in primed (bFGF) versus LIF-3i-reverted E5C3. Shown are relative firefly luciferase activities following normalization with Renilla luciferase and negative control basal activities ±s.d. (n=3). *P<0.05 (paired t-test). (D) Stable BAC reporter transgenic OCT4 PE/DE mutant lines. (Top) Cytometry plots of representative LIF-3i-reverted C2 hiPSC subclones (n=3) stably transfected with full-length OCT4-GFP-2A-PURO PE/DE sequences (control), mutant ΔPE-OCT4-GFP-PURO constructs, or non-transfected (no construct) controls. (Bottom) Percentage GFP⁺ cells among naïve cultures of individual hiPSC subclones (n=3) expressing control or mutant ΔPE sequences. (E) Pluripotency circuits in LIF-3i-reverted hPSCs. (Top) Mean beta values of core module-specific CpG DMRs in primed (−) versus LIF-3i-reverted (+) hPSC; (bottom) corresponding log₂ mean subtracted normalized expression of core module genes (Table S1) of the same independent hPSC samples (identical to those used above for expression microarrays). Gold, hESCs (n=3); red, sa-MP-iPSCs (n=6); green, fibro-iPSCs (n=3). (F) Pluripotency gene-specific promoter CpG methylation. Heatmap-dendrogram clustering and box plots of mean beta values of ESC module gene-specific CpG DMRs [P<0.001 (paired two-way t-test)] of LIF-3i (+) versus primed (−) hPSCs. Samples are the same 12 hPSC lines in each category, as described above. Percentages represent reduction of median beta value following LIF-3i reversions.

Fig. 5.

Comparative transcriptional/epigenetic meta-analyses of LIF-3i-reverted hPSCs. (A) Mean genome-wide gene expression versus CpG methylation DMRs crossplot of LIF-3i-reverted hPSCs versus their isogenic primed hPSC counterparts [same 12 samples as in Fig. 4, i.e. hESCs (n=3), MP-iPSCs (n=6), fibro-iPSCs (n=3)]. Shown are DMRs (CpG beta values) of LIF-3-reverted versus isogenic primed hPSC counterparts (y-axis, P≤0.05) versus their corresponding differential gene expressions [red, x-axis, log₂ fold changes (FC); P≤0.05, FC ±≥1.5]. (B) Curated GSEA reactome pathways over-represented (FDR<0.01; P<0.01) in LIF-3i-reverted sa-MP-iPSCs versus their isogenic primed sa-MP-iPSC counterparts (n=6; same samples as Fig. 4A). (C,D) Comparative platform-normalized PCA of whole-genome expression of primed (blue) or naïve (red) PSC lines from different laboratories. (C) Human morula/blastocyst (red circles) or (D) mESC-serum/mESC-LIF-2i (black/gray circles) samples were used as benchmarks. Human morula/blastocyst PCA in C is clustered on a module of the most differentially expressed genes in E4-E5 human pluripotent epiblast (Petropoulos et al., 2016) (see Table S1). Z, this study and includes the n=12 independent hPSC lines in Fig. 4A; H, Hanna et al., 2010; G, Gafni et al., 2013; M, Takashima et al., 2014; S, Theunissen et al., 2014; V, Vassena et al., 2011. (E) Comparison of differentially expressed (P≤0.05, FC±≥1.5) naïve-specific and lineage-primed transcripts in naïve hPSCs derived in this work or other labs. FC: normalized ratios of naïve/primed expression microarray signal intensities. Ratios are of LIF-3i-reverted versus primed hPSC samples (n=12 hPSCs, as above) versus samples of those published as indicated.

Fig. 6.

Multilineage differentiation of isogenic primed versus LIF-3i-reverted hiPSC lines. (A) Differential expression of lineage-primed genes in the Polycomb (PRC2) circuit (ANOVA, P<0.001; Table S1) in seven hiPSC lines before (−) and after (+) LIF-3i reversion. Shown are heatmaps and associated log2 mean subtracted expression of PRC2 module genes of the LIF-3i-reverted versus isogenic-primed hiPSCs used in the differentiation studies below. Red, 4F-E sa-MP-iPSCs (n=4): circle, E5C3; square, E5C1; and triangle, E17C6 (or LZ6+10 for neural differentiations). Green, fibro-iPSCs (n=3): circle, C1.2; square, C2; and triangle, 7ta. *P<0.05 (paired t-tests). (B) Definitive endoderm differentiations (FOXA2⁺, CXCR4⁺ SOX17⁺) of isogenic LIF-3i-reverted versus primed hPSCs. Neural differentiations. (C,D) Kinetics of SOX1⁺ nestin⁺ and PAX6⁺ nestin⁺ neural progenitors in the same primed versus LIF-3i-reverted isogenic sa-MP-PSC (n=3) and fibro-hiPSC (n=3) lines described above. *P<0.05, **P<0.01 (paired t-tests). (E) Confocal microscopy of CDr3⁺ dye-binding neural progenitor rosettes (Yun et al., 2012). Neural rosettes were evaluated following passage of day 7 neural-induced LIF-3i-reverted (+) versus isogenic primed (−) E5C1 hiPSCs. Scale bars: 100 µm. (F) Isogenic vascular-endothelial hEB differentiations. Flow cytometry kinetics of CD31⁺ CD146⁺ (left) and KDR⁺ CD73⁺ (right) VP populations of the same isogenic sa-MP-iPSC lines as above (n=3). Error bars indicate s.e.m.

Fig. 7.

Summary schemas. (A) Relationships between primed pluripotency, WNT signaling, and amenability to stable naïve LIF-3i reversion. (B) Proposed molecular mechanisms underlying naïve-promoting effects of tankyrase inhibition in the presence of LIF-2i.