Human Naive Pluripotent Stem Cells Model X Chromosome Dampening and X Inactivation

Abstract

Naive human embryonic stem cells (hESCs) can be derived from primed hESCs or directly from blastocysts, but their X chromosome state has remained unresolved. Here, we show that the inactive X chromosome (X_i) of primed hESCs was reactivated in naive culture conditions. Like cells of the blastocyst, the resulting naive cells contained two active X chromosomes with XIST expression and chromosome-wide transcriptional dampening and initiated XIST-mediated X inactivation upon differentiation. Both establishment of and exit from the naive state (differentiation) happened via an XIST-negative X_aX_a intermediate. Together, these findings identify a cell culture system for functionally exploring the two X chromosome dosage compensation processes in early human development: X dampening and X inactivation. However, remaining differences between naive hESCs and embryonic cells related to mono-allelic XIST expression and non-random X inactivation highlight the need for further culture improvement. As the naive state resets X_i abnormalities seen in primed hESCs, it may provide cells better suited for downstream applications.

Keywords: X chromosome; X chromosome dampening; X chromosome inactivation; XIST; embryonic stem cells; human development; human stem cells; lncRNA; naive pluripotency; pluripotent stem cells.

Figure 1. The X_i of primed hESCs reactivates in the 5iLAF culture condition

(A) Representative RNA FISH images for primed UCLA1 at passage 16 (P16), detecting XIST and nascent transcription foci of UTX, (escapes XCI) and HUWE1 (subject to XCI). Panels show HUWE1 and UTX expression without XIST (left), XIST only (middle), and all three channels together (right) in DAPI-stained nuclei (blue). A nucleus with the most prevalent pattern is highlighted with a dotted box and enlarged, and its pattern depicted by the cartoon on the right. (B) As in (A), but for the X-linked genes ATRX (top) or THOC2 (bottom), showing only the merged panel with XIST and UTX. (C) Quantification of the RNA FISH patterns for XIST and the X-linked genes HUWE1, ATRX, and THOC2 in primed UCLA1 as shown in (A) and (B). Only cells with bi-allelic UTX expression were considered. (D) Representative RNA FISH images of the most prominent X-chromosome state in naïve UCLA1 at P19 in 5iLAF media, detecting HUWE1, XIST and UTX, similar to (A). (E) As in (D), but detecting the X-linked genes ATRX (top) or THOC2 (bottom). (F) Quantification of the RNA FISH patterns for XIST with HUWE1, ATRX, and THOC2, respectively, in naive UCLA1 as shown in (D) and (E) in cells with bi-allelic UTX expression. (G) Representative RNA FISH images similar to (D), but for the bi-allelic XIST (top) and XIST-negative (bottom) X_aX_a patterns. See also Figure S1.

Figure 2. X_i-reactivation leads to an XIST negative X_aX_a intermediate before induction of XIST expression from the X_a

(A) Quantification of the RNA FISH patterns for XIST and HUWE1 as UCLA1 progressed from the primed state at P12 to the naïve pluripotent state up to P7 in 5iLAF media. All counts were in naïve cells with two UTX foci. (B) Allelic expression proportions of X-linked genes in the primed UCLA1 population, the naïve UCLA1 clone4 (at early passage, XIST-negative state), and the naïve XIST-positive UCLA1 clones 9 and 12, based on reads covering indicated SNPs, ordered along the X-chromosome (depicted with red lines in the X-chromosome image below), in replicate RNA-seq data (rep1 and rep2, closely-spaced bars). SNPs located in the same gene were placed next to each other and marked with an asterisk (*). ND = Not Determined due to insufficient read coverage. (C) Analysis of allelic expression proportions of X-linked genes in single XIST-negative and XIST-positive naïve UCLA1 cells. Single cells of naïve UCLA1 at early passage, when the population was still largely XIST-negative, and at later passage, when most cells were XIST-positive, were grouped into XIST-positive and XIST-negative cells based on their XIST-expression levels. The two graphs on the right show the allelic expression proportion of X-linked genes for each single cell with sufficient SNP coverage in these two groups, with each circle representing the proportion of one SNP in one single cell. Highlighted with yellow is the 20–80% range region where we considered X-linked gene expression from both X-chromosomes to take place. For comparison, the graph on the left plots the allelic ratio of the same SNPs used in the right in primed UCLA1 as described in (B). (D) Heatmap of unsupervised hierarchical clustering of RRBS-based methylation levels of CpGs within X-linked CGIs. The number of CpGs considered is given. Mb (megabase) indicates the relative position on the X-chromosome. (E) Heatmap as in (C), but for all X-linked CpGs covered by RRBS. (F) Methylation averages of X-linked and autosomal CpGs outside (top) and within CGI’s (bottom) in primed UCLA1 and the naïve clones 4 (XIST-negative), 9 and 12 (XIST-positive). (G) Empirical cumulative distribution functions of X-linked gene expression for naïve UCLA1 clone4 at early (XIST-negative) and late (XIST-positive) passage, and the XIST-positive clones 9 and 12, based on replicate RNA-seq data sets where available (rep1, rep2). The inset shows autosomal expression data for the same samples in the same format. Asterisk (*) indicates statistically significant difference between the distributions of X-linked gene expression of both replicates of early passage (XIST-negative) clone4 compared to all other samples (p<0.0035 by Wilcoxon rank sum test with continuity correction). The X-linked and all autosomal distributions for all other samples were not significantly different from each other. (H) Same as in (G), but for early (XIST-negative) and late (XIST-positive) passage naïve UCLA4 hESCs. X-linked but not autosomal gene expressions of both replicates of early passage (XIST-negative) samples were significantly different from replicate 2 (but not replicate 1) of late passage (XIST-positive) cells (p<0.05). See also Figures S2 and S3.

Figure 3. The XIST expressing X_aX_a state is found in naïve hESC lines regardless of source

(A) Immunofluorescence detection of H3K27me3 (red) combined with RNA FISH for XIST (green) in normal female fibroblasts (top), the naïve UCLA1 clone12 (middle), and blastocyst-derived naïve UCLA20n (bottom) cells. (B) As in (A), except for RNA polymerase II (RNA PolII) and XIST. Yellow arrowheads demonstrate regions in nuclei devoid of RNA PolII signal under the XIST signal. (C) Quantification of RNA FISH patterns of transcription foci of HUWE1, ATRX, and THOC2, respectively, and of XIST in primed, 5iLA- and 4iLA-cultured naïve WIBR3 hESCs. (D) Representative RNA FISH images detecting XIST, HUWE1, and UTX in the blastocyst-derived naïve hESC lines UCLA19n and UCLA20n. RNA FISH patterns were stable throughout passaging and in the graph are quantified at P3 for both lines. See also Figure S4.

Figure 4. Naïve hESCs initiate XCI upon differentiation

(A) Representative RNA FISH images for XIST, HUWE1, and UTX in naïve UCLA1 at P42 in 5iLAF, in the re-primed condition at P7, and after seven days of differentiation (Diff) from the re-primed state. A nucleus with the most prevalent pattern is highlighted with a dotted box and enlarged, and also depicted by the cartoon on the right. (B) Quantification of XIST-expressing differentiated cells with different HUWE1 transcription patterns (bi-allelic with equal transcription foci on both X-chromosomes, bi-allelic with a smaller transcription focus on the XIST-coated chromosome, and mono-allelic) in cells with bi-allelic UTX expression. (C) Representative RNA FISH images as in (A), but for the naïve hESC line UCLA20n at P8 (naïve), P4 (re-primed), and after seven days of differentiation of re-primed state.

Figure 5. XCI in differentiating naïve cells is non-random

(A) Representative electropherograms from Sanger sequencing of SNP-containing regions in the HUWE1 and XIST cDNA obtained from the primed UCLA1 population (at P19 for HUWE1 and P4 for XIST analysis; top), and from an individual cell differentiated from the re-primed state after transition through the naïve state (bottom). Pie charts summarize the Sanger sequencing results for additional XIST-positive, single differentiated cells (43 individual cells for HUWE1, 47 for XIST), considering three expression categories: expression of only reference (Ref) allele, only alternate (Alt) allele, or both alleles. (B) Allelic expression proportions of XIST based on RNA-seq reads (number of reads indicated on top) that covered the indicated SNP in single cells differentiated from naïve UCLA1 after transition through the re-primed state. The X-axis refer to single cell coordinates in a 96-well plate. For comparison, the allelic XIST proportion in primed UCLA1 was estimated from the Sanger sequencing electropherograms in (A, top). (C) As in (B), but for the X-linked genes RBM3 (two different SNPs), IDS, MPP1, TAF9B. Only XIST-expressing single cells were analyzed. Allelic expression proportions in primed UCLA1 were calculated from replicate RNA-seq data of the primed cell population (Pri1 and Pri2). (D) Representative immunofluorescence images detecting MeCP2 and CD44 in hiPSC clones 16 and 17 harboring an X_a with wild type and mutant MECP2, respectively. Shown are primed hiPSCs (P17; top), re-primed cells (P3) after transition through the naïve 5iLAF for 3 passages (middle), and the day 7 differentiation product of the re-primed state (bottom). (E) Schematic summary of X-chromosome dynamics as primed X_aX_i cells with or without XIST expression from the X_i transition to naïve pluripotency and differentiate. See also Figure S5.

Figure 6. Naïve pluripotency enables XCI in primed hESCs with an aberrant X_aX_a state

(A) Representative RNA FISH images detecting XIST, HUWE1, and UTX in primed UCLA9 (P15) and upon seven days of differentiation. A nucleus with the most prevalent pattern is highlighted with a dotted box and enlarged, and also depicted by the cartoon. (B) Quantification of RNA FISH patterns of XIST in combination with HUWE1, ATRX, and THOC2, respectively, in the primed state and its differentiated daughter cells. Only cells with biallelic UTX expression were quantified. (C) Quantification of RNA FISH patterns of XIST and HUWE1 in UCLA9 at indicated passages post 5iLAF conversion. Only cells with biallelic UTX expression were quantified. (D) Representative RNA FISH images detecting XIST, HUWE1, and UTX in UCLA9 in the naïve state (P22), the re-primed state obtained by culturing naïve cells in the primed culture condition for 3 passages, and in day 7 differentiated cells derived from re-primed cells. In each row, a nucleus with the most prevalent pattern is highlighted with a dotted box and enlarged, and depicted by the cartoon on the right. (E) As in (B), but for re-primed state. (F) As in (B), but for differentiated UCLA9 described in (E). (G) Schematic summary of the X-chromosome state of UCLA9 in indicated states. See also Figure S6.

Figure 7. Naïve pluripotency erases X_i-erosion of primed hESCs

(A) Quantification of RNA FISH patterns of XIST and HUWE1, or XIST and ATRX in primed UCLA4 and upon seven days of differentiation from the primed state. Only cells with bi-allelic UTX expression were quantified. (B) Quantification of RNA FISH patterns of XIST and HUWE1 in UCLA4 at indicated passages post 5iLAF conversion. Only cells with biallelic UTX expression were quantified. (C) Representative RNA FISH images detecting XIST, HUWE1, and UTX in UCLA4 in the naïve state (P7), in the re-primed state obtained from the naïve state, and in day 7 differentiated cells derived from re-primed cells. A nucleus with the most prevalent pattern is highlighted with a dotted box and enlarged, and also depicted by the cartoon. (D) Quantification of RNA FISH patterns in re-primed UCLA4 and their differentiated progeny in cells with bi-allelic UTX expression as in (A). Only XIST-expressing cells were considered in the differentiated state. (E) Representative RNA FISH image detecting XIST RNA and nascent transcription foci of ATRX in differentiated cells originating from re-primed cells derived from naïve UCLA4. (F) Schematic summary of the X-chromosome states of UCLA4 in indicated states. See also Figure S7.