Imprinted X‐chromosome inactivation impacts primitive endoderm differentiation in mouse blastocysts

Epigenetic and transcriptome alterations are essential for lineage specification, represented by imprinted X‐chromosome inactivation (iXCI) in female mouse preimplantation embryos. However, how various factors affect transcriptome states and lineage commitment remains unclear. We found that in vitro culture duration strongly influences transcriptional variation compared to iXCI loss. Single‐cell analysis of the inner cell mass (ICM) for major transcription and epigenomic factors revealed that sex‐specific differences in expression are diminished by loss of iXCI in the primitive endoderm (PrE) but not in the epiblast. Females had a higher proportion of ICM compared to that in males, and PrE development was affected by iXCI states in female embryos. Our findings provide insight into sex differences and iXCI function in lineage specification.

Epigenetic and transcriptome alterations are essential for lineage specification, represented by imprinted X-chromosome inactivation (iXCI) in female mouse preimplantation embryos. However, how various factors affect transcriptome states and lineage commitment remains unclear. We found that in vitro culture duration strongly influences transcriptional variation compared to iXCI loss. Single-cell analysis of the inner cell mass (ICM) for major transcription and epigenomic factors revealed that sex-specific differences in expression are diminished by loss of iXCI in the primitive endoderm (PrE) but not in the epiblast. Females had a higher proportion of ICM compared to that in males, and PrE development was affected by iXCI states in female embryos. Our findings provide insight into sex differences and iXCI function in lineage specification.
Keywords: imprinted X-chromosome inactivation; lineage specification; sex difference Epigenetic dynamics and transcriptome alterations are observed after fertilization in mammals [1]. One representative epigenetic event is the establishment of X-chromosome inactivation (XCI) in female mouse embryos [2]. Generally, one of two X-chromosomes in female cells is inactivated to compensate for X-linked gene expression dosage [2]. In mice, prior to establishing XCI, the long noncoding RNA Xist is transcribed from the paternal X-chromosome at approximately the 4-cell stage. Parental origin-specific Xist expression in females leads to imprinted X-chromosome inactivation (iXCI) [3,4]. Establishment of iXCI occurs at approximately the late blastocyst stage, and hallmarks of heterochromatin such as H3K27me3 foci overlapped with Xist transcripts emerge [5]. Once iXCI is established, states are stably maintained during embryonic development in the primitive endoderm (PrE) and trophoblast lineages. In contrast, iXCI is reprogrammed in the epiblast [6,7]. Consequently, both female X-chromosomes are transiently present in an active state, and then, one of two X-chromosomes is randomly inactivated as differentiation proceeds.
A recent study using RNA-seq revealed that Xist-dependent iXCI already occurs without the H3K27me3 hallmark in early preimplantation phases in female mice [8]. Interestingly, the iXCI state is affected by the culture situation, and embryos produced by in vitro culture show an imbalanced sex ratio in birth rates [9].
During preimplantation development in mice, sex differences are observed in terms of the developmental speed by the blastocyst stage [10]. This phenotype is attributed to the presence of the Y-chromosome [10]. Moreover, the developmental competency of embryos derived from in vivo culture is superior to that obtained in vitro in mice after implantation [11]. These studies indicate that culture environment, sex differences, and iXCI impact the transcriptome and developmental propensity.
In the present study, we investigated the effects of transcriptome states by RNA-seq analysis of mouse blastocysts derived from different in vitro culture durations as well as following the loss of iXCI. Clustering analysis revealed that the largest clusters were formed based on culture conditions. We identified differentially expressed genes (DEGs) in sex-specific-, in vitro culture period-, and iXCI-dependent manner. Interestingly, upregulated genes in the female embryo derived from a short-term in vitro culture (in vivo generated embryos) compared to in the male embryo were associated with a pluripotency network. Moreover, a high-fidelity quantitative gene expression assay targeting representative transcription and epigenetic modifiers at single-cell resolution in the inner cell mass (ICM) revealed that DEGs between males and females depend on iXCI in the PrE lineage. Finally, we found that the proportion of PrE development differed between males and females, and loss of iXCI diminished the PrE differentiation potential. These findings provide insight into the relationship of iXCI with lineage specification and developmental differences based on sex.

Mouse embryo collection
Xist KO males harboring a repeat A deletion (B6.Cg-Xist<tm5Sado>) were a kind gift from Sado (Kindai University). Xist KO males (>9 weeks) were mated with 8to 10-week-old B6D2F1 females purchased from CLEA and Sankyo Labo Service (Tokyo, Japan) in vivo or in vitro according to standard methods [3]. All animals were maintained and used in accordance with the Guidelines for the Care and Use of Laboratory Animals of the Japanese Association for Laboratory Animal Science and National Research Institute for Child Health and Development (NRICHD) and Tokai University of Japan. All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee of the NRICHD (Permit Number: 05-006) and Tokai University (Permit Number: 192028).
Both in vivo and in vitro fertilized embryos were cultured in KSOM (EMD Millipore, Burlington, MA, USA) until the late blastocyst stage in an atmosphere of 37°C 5% CO 2 . For in vivo samples, at embryonic day 2.5 (E2.5) after fertilization, embryos were flushed from the uterus using M2 medium and cultured for 2.5 days in KSOM medium until E5.0 and used for analysis. For WT embryos collection, the embryos from the littermate are used. In contrast, TYH medium [3] was used as the fertilized medium and the embryos were cultured as described above. At E5, the embryos derived from in vitro were used for analysis.

Single-cell isolation from ICM
Blastocysts at E5 were subjected to immune surgery to isolate the ICM using a conventional method [12]. In total, seven male, seven Xist KO female, and eight WT female embryos were used for single-cell collection. The isolated ICM was placed in 0.05% trypsin (Thermo Fisher Scientific, Waltham, MA, USA) containing 0.1% polyvinyl alcohol (PVA; Sigma, St. Louis, MO, USA) in PBS (À) and incubated for 5 min at room temperature. The trypsin-treated ICM was separated into single cells using a micromanipulator and washed with PBS (À) containing 0.1% PVA for use in the single-cell qPCR assay.

Transcriptome analysis
Total RNA was extracted using an RNeasy micro kit with DNase treatment (Qiagen, Hilden, Germany) following the manufacturer's instructions. Total RNA obtained from each sample was subjected to sequencing library construction using an Ovation Solo RNA-Seq System (NuGEN Technologies, San Carlos, CA, USA) according to the manufacturer's protocol. The Ovation SoLo RNA-Seq System integrates AnyDeplete technology for rRNA depletion and a molecular barcode system. Library quality was assessed with an Agilent 2100 Bioanalyzer High Sensitivity DNA Kit (Agilent Technologies, Santa Clara, CA, USA). The pooled libraries of the samples were sequenced using the Illumina Hiseq 1500 system (Illumina, San Diego, CA, USA) in 101-base-pair paired-end reads.
Sequencing adaptors, low-quality reads, and bases were trimmed with the Trimmomatic-0.32 tool. The sequence reads were aligned to the Mouse reference genome (mm10) using TOPHAT 2.1.1 (BOWTIE2-2.2.3, https://ccb.jhu.edu/sof tware/tophat/index.shtml) with the option '-g 1 --no-mixed --no-discordant'. Sequence libraries of the Ovation SoLo RNA-Seq System included 8 bp molecular barcodes to detect unique RNA molecules for accurate gene expression analysis. NUDUP tool ver.2.2 (NuGEN Technologies, San Carlos, CA, USA) was used to analyze redundancy of the molecular barcodes to identify unique molecules. Files of gene model annotations and known transcripts were downloaded from the Illumina's iGenomes website (http://sup port.illumina.com/sequencing/sequencing_software/igenome. html), which were necessary for whole transcriptome alignment with TOPHAT. The aligned reads were subjected to downstream analyses using STRANDNGS 3.2 software (Agilent Technologies). The read counts allocated for each gene and transcript (RefSeq Genes 2013.04.01) were quantified using a reads per kilobase per million mapped reads (RPKM) method [13].
Genes showing a very low normalized signal value (RPKM < 1) in at least 1 sample in any group were filtered. A karyotype plot of differentially expressed genes (DEGs) was created in R 3.4 package karyoplote R version 1.4.2 [14].
Embryo sex was determined based on the expression of Eif2s3y mapped to the Y-chromosome. Xist KO females were confirmed by the absence of both Xist and Eif2s3y expression. The original data are available in GEO (GSE133845).
For pathway analysis, we used DAVID bioinformatics (https://david.ncifcrf.gov/). To identify the significantly enriched pathway based on KEGG in DAVID, Fisher's exact test (P < 0.05) is used to identify significant pathways.

Single-cell qPCR analysis
The isolated single cells were moved to lysis buffer using a mouth pipette under a microscope. A Single Cell-to-CT TM qRT-PCR Kit (Thermo Fisher Scientific) was used for the assay with slight modifications as described previously [15]. Briefly, half volumes of all reagents were used in the Taq-Man probe assay. Cdx2 with < 30 C t values were considered as trophectoderm cells and removed from the analysis. C t values > 35 were considered as no detection. Following a previous report [16] with slight modifications, we obtained single-cell expression data using qPCR. To calculate the expression levels per cell based on the C t values, a C t value of 30 was used as background baseline and subtracted from each C t value. In this analysis, C t values of no detection samples were calculated as a C t value of 35. The resulting values were used as the normalized value for gene expression analysis.
For embryo sexing, cells in which Eif2s3y was detected were used as male cells, whereas other cells were used as females in WT embryos. Because we used sperm with Xist deletion to generate Xist KO embryos, cells with no Eif2s3y expression were regarded as Xist KO females. The normalized values (Log2) were used for clustering analysis with EZR (http://www.jichi.ac.jp/saitama-sct/SaitamaHP.files/sta tmedEN.html). To divide the cells into the epiblast or PrE lineage, the cells were classified based on the expression states of Nanog, Sox2, Gata4, Gata6, and Sox17, which were identified as each lineage marker. Based on the clustering results, the cells were subjected to statistical analysis (one-way analysis of variance, ANOVA, with post hoc test for multiple comparisons among embryo types). All TaqMan probes used in this study are shown in Table S4.

Immunofluorescence combined with DNA FISH
The zona pellucida was removed with acid Tyrode's solution (Sigma), and embryos were fixed with 2% paraformaldehyde in PBS containing 0.1% PVA (Sigma) for 20 min at room temperature and then permeabilized with 0.25% Triton X in PBS-PVA for 15 min at room temperature. After blocking with 1.5% bovine serum albumin (Sigma), the samples were incubated with primary antibodies for NANOG (1 : 200; ReproCell, Beltsville, MD, USA) and GATA4 (1 : 200; Abcam, Cambridge, UK). Alexa Fluor 488-or 546-conjugated anti-mouse or anti-rabbit IgG secondary antibodies (Thermo Fisher Scientific) were diluted with PBS-PVA and incubated with the samples for 1 h at room temperature. DNA was stained with DAPI (Vector Laboratories, Burlingame, CA, USA). Before capturing the images, the samples were attached to slide glass and covered with a cover slip. The images were observed using a LSM510 laser scanning confocal microscope (Carl Zeiss, Oberkochen, Germany), and the quantification assay was carried out using IMAGEJ software (NIH, Bethesda, MD, USA).
DNA FISH experiments were performed as described previously [17]. Briefly, after washing with PBS, the samples were incubated with PBS (À) containing RNase A for 1 h at 37°C. The samples were treated with pepsin (Sigma) for 8-10 min, washed with PBS (À), and incubated with sexing probe (Chromosome Science Labo., Hokkaido, Japan) at 85°C for 10 min and then incubated with 37°C overnight. After incubation, the samples were washed as described previously [17] and images were captured.

Results
Effect of transcriptome states on culture conditions, sex, and loss of Xist Mammalian embryos can develop to the late blastocyst stage in vitro, and blastocyst transfer into the uterus is widely used as a reproductive technique [18]. Thus, we considered that the late blastocyst is suitable for examining the effects of culture conditions, sex differences, and iXCI on transcriptome states in mice. We conducted strand-specific RNA-seq using individual late blastocysts derived from in vivo or in vitro fertilization mated with wild-type (WT) and Xist knockout (KO) male mice [19] and cultured in vitro for different durations, referred to as in vivo and in vitro samples,   respectively (Fig. 1A). The cell numbers in blastocysts from both origins were not significantly different (Fig. S1a), indicating that the developmental stage is similar in both in vivo and in vitro embryos. The correlation coefficient among the samples was more than 0.93, exhibiting a high reproducibility (Fig. S1b). Using the genes with > 1 RPKM in at least one group, we performed hierarchical clustering analysis (Table S1). Clusters were clearly formed based on culture duration, genotype, and sex (Fig. 1B). Interestingly, the largest cluster was formed based on the in vitro culture duration, indicating that transcriptome states are greatly affected by the developmental environment.

Up-regulated in
To determine the gene expression states under each condition, we performed pairwise comparison and screened DEGs showing more than twofold changes ( Fig. 1C and Table S2). We identified numerous DEGs in the comparison of in vivo and in vitro conditions in both male and female blastocysts, supporting the clustering results (total > 929 genes; Fig. 1C). In contrast, comparison of males and females under each condition revealed a relatively small number of DEGs (<493 genes; Fig. 1C). However, among the up-regulated genes in females compared to males, 26% (in vivo) and 40% (in vitro) were enriched in X-chromosomes (Fig. 1C). We also confirmed that X-linked genes were not dominantly expressed in this stage (Fig. S1b).
As expected, a large population of X-linked genes was up-regulated in Xist KO females (in vivo: 37% and in vitro: 26%; Fig. 1C), confirming the effect of Xist-dependent iXCI [8]. Moreover, Xist deletion resulted in a large number DEGs irrespective of the in vitro culture duration (in vivo: 982 genes and in vitro: 1172 genes; Fig. 1C).
Next, to evaluate the features of the DEGs identified in each pairwise comparison, we constructed a Venn diagram and performed KEGG pathway analysis. Several known pathways were detected in the DEGs by the comparison of in vivo with in vitro in males and females (Fig. S2a). For instance, 'Protein digestion and absorption' and 'Metabolic pathways' were commonly down-and up-regulated pathways, respectively (Fig. S2a). Interestingly, when we examined the DEGs in males and females in each culture condition, the pathway linked to 'signaling pathway regulating pluripotency of stem cell' was identified as significant (Fig. S2b). The pathway is based on the DEGs specifically up-regulated in vivo in females compared to in vivo in males (Fig. S2b). The finding that 26% (in vivo) and 40% (in vitro) of up-regulated genes in females compared to males were enriched in the Xchromosome (Fig. 1C) suggests that X-linked gene overdosage affects pluripotency networks.
In terms of the effect of Xist deletion, up-regulated DEGs in Xist KO females compared to in WT females under both culture conditions were commonly linked to 'mRNA surveillance pathway' and 'Metabolic pathway' (Fig. S2c). Although a significant pathway was not found in the up-regulated DEGs specifically in in vivo Xist KO females, 'p53 signaling pathway' was identified as a significant pathway in the up-regulated DEGs, specifically in in vitro Xist KO female embryos (Fig. S1c).
Interestingly, pluripotency-related pathway, mitogen-activated protein kinase (MAPK) signaling, and transforming growth factor-b signaling were detected as down-regulated DEGs in Xist KO females, specifically in vivo (Fig. S2c). These results indicate that although loss of iXCI causes developmental disorder in each developmental condition [19,17], the pathways influenced differ between the conditions. These analyses were conducted with all genes, including autosomes; however, Xist deletion caused mainly X-linked gene overdosage (Fig. 1B). Therefore, we focused on X-linked gene expression states. Of the up-regulated DEGs in Xist KO females, the number of X-linked genes was high under each condition, and 73% of common DEGs were mapped to X-chromosomes (Fig. 1D). Furthermore, we found that X-linked up-regulated DEGs in the Xist KO female were randomly distributed over the entire X-chromosome under both culture patterns (Fig. 1E). Interestingly, among commonly up-regulated genes in Xist KO females, large families of Rhox, Xlr, and Magea were identified (Fig. 1E), suggesting that these regions might be subjected to common epigenetic regulation.
Throughout RNA-seq analysis in embryos under various conditions, we found many DEGs. In particular, many X-linked genes were up-regulated in females compared to males under both in vivo and in vitro conditions, indicating that these conditions cause transcriptome differences between males and females. Moreover, although down-regulation of many autosomal genes was shown in Xist KO females irrespective of culture duration, Xist-dependent up-regulated genes on the X-chromosome were not mainly influenced by in vitro culture, indicating that the substantial effects of Xist RNA on cis-acting elements were stable (Fig. 1E).

Sex-specific differences in expression states of transcription and epigenetic factors depend on iXCI in the primitive endoderm but not in the epiblast
The blastocyst is the first phase of differentiation and is composed of three parts: epiblast, PrE, and trophectoderm [20]. In terms of iXCI, the state is preserved in the PrE and trophectoderm lineages but not in the epiblast [2]. Therefore, we examined how sex differences and iXCI loss impact transcription and epigenomic factors in the ICM at single-cell resolution. To investigate these intrinsic features, we used in vivo-derived samples for high-fidelity gene expression analysis by quantitative PCR using a TaqMan probe [3]. We confirmed that our system was highly reproducible (Fig. S3). Using this system, we collected single cells from the in vivo-derived ICM by immune surgery and evaluated gene expression for 72 genes related to transcription or epigenetic factors as well as XCI-related genes in male, female, and Xist KO female embryos ( Fig. 2A). To avoid contamination and potential damage to the trophectoderm during immune surgery, we excluded the trophectoderm cells and screened cells with low expression levels of Cdx2, a trophectoderm maker [21], for further analysis. After removing the cells with high Cdx2 expression, we selected 74 cells for further analysis (Table S3).
We divided the cells into two lineages, epiblast (EPI) and PrE, which are known makers for EPI (Nanog/ Sox2) or PrE (Gata4/Gata6/Sox17) [16], to categorize the cell lineage. Hierarchical clustering analysis clearly revealed two populations based on the expression levels of these markers (Fig. 2B). Each embryo group (WT female, male, and Xist KO female) was randomly distributed in both the EPI and PrE groups (Fig. 2B).
Based on the clustering results, we compared the expression states in individual genes based on their origins. In EPI cells, six up-regulated genes in females were identified as significantly expressed genes (SEGs) in the comparison of WT females to males (Fig. 2C). In contrast, the comparison between WT and Xist KO females showed that all genes except for Eomes genes were not significantly expressed, indicating that the major transcription and epigenomic factors were not affected by two continuously active X-chromosomes (Fig. 2D). We also identified 12 SEGs in Xist KO females compared to in males, although all of them were up-regulated in Xist KO females (Fig. 2E). These results indicate that sex differences affect the expression levels of some types of transcription and epigenetic factors in EPI, and loss of iXCI did not affect the expression states of the analyzed genes in female embryos.
In PrE, all identified SEGs between WT females and males were up-regulated in females, as observed for EPI (Fig. 2C,F). Moreover, the identified SEGs were up-regulated in Xist KO females compared to in WT females, indicating that loss of iXCI leads to overexpression of the analyzed genes and X-linked genes (Fig. 2G), consistent with the results of a recent report [8]. Interestingly, comparison of Xist KO females and males revealed that no SEGs compared between WT females with males were significantly expressed (Fig. 2F,H), and only Kdm5c and Mycn were up-regulated in Xist KO females compared to males (Fig. 2H).
These results indicate that female-specific genes up-regulated in PrE were affected by iXCI; however, the effect was decreased by Xist deletion, suggesting the dependency of PrE development.

Loss of iXCI impacts lineage specification of PrE
Next, we asked how the gene expression states affect lineage specification in the late blastocyst. We conducted immunofluorescence analysis of NANOG and GATA4 following DNA-fluorescence in situ hybridization (FISH) using in vivo-derived WT female, Xist KO female, and male late blastocysts. As previously reported [10], we observed that the number of male cells was significantly larger than that of female cells (Fig. 3A). Interestingly, fewer Xist KO cells were detected compared to WT female and male cells (Fig. 3A), indicating that loss of iXCI cannot compensate for sex differences in terms of cell numbers.
Immunofluorescence analysis revealed a significantly decreased proportion of GATA4-positive cells in males compared to in WT females (6.8% in males and 9.8% in WT females; Fig. 3B,C). Moreover, a significant decrease in the GATA4-positive cell population was detected in Xist KO females compared to in WT females (9.8% in WT females and 5.6% in Xist KO females; Fig. 3B,C). Interestingly, the percentage of GATA4-positive cells in Xist KO females was comparable to that in males (Fig. 3B,C), supporting the quantitative PCR results.  females were higher than those in males (Fig. S4a).
Given that the proportions of GATA4-and NANOGpositive cells of female embryos were higher than those of males and the CDX2-positive cell proportion is higher in males than in females (Fig. 3C,D and Fig. S4b), male embryos developed the trophectoderm lineage well compared with females. Taken together, these results indicate that sex differences occur in both EPI and PrE lineage development. Particularly, iXCI impacts on PrE lineage specification in females.

Discussion
Transcriptome analysis revealed that culture conditions greatly impact gene expression states and Xist deletion. Given that the developmental competency of embryos derived from in vitro fertilized and culture is inferior to in vivo embryos [11], the identified pathways may be responsible for the poor development in in vitroderived embryos. For example, the Wnt, MAPK, and vascular endothelial growth factor pathways were identified as commonly up-regulated in in vitro-derived male and female embryos (Fig. S1a). Because these pathways have been widely studied and small molecules are used for their repression [22], transient treatment during in vitro culture may have beneficial effects on development.
Because we observed that up-regulated genes in WT females compared to in males were enriched on the Xchromosome, sex differences in transcriptome would be caused by X-chromosome reactivation in the female epiblast [2]. Interestingly, we found that pathways linked to pluripotency were induced by up-regulated genes in female embryos in vivo compared to males (Fig. S1b). Recent studies using mouse embryonic stem cells indicated that having two active X-chromosomes blocks differentiation via Dusp9-mediated MAPK pathway suppression [23,24]. Consistent with the results of stem cell studies, Dusp9 was up-regulated in female embryos compared to male embryos (Table S1). Collectively, overexpression of X-linked genes was associated with pluripotent states and responsible for the high population of NANOG-positive cells in female compared to male embryos and in Xist KO females (Fig. 3D).
In terms of the effect of Xist deletion on the transcriptome, up-regulation of the Rhox and Magea families in Xist KO female is consistent with the results described in previous studies [8]. Importantly, overexpression of Rhox5 in mouse embryonic stem cells impairs the differentiation potential [25], indicating that dysregulation of Rhox-family genes is one factor responsible for the developmental disorder in Xist deletion. In contrast, although the results for the Xlr family have not been reported in Xist dependency in preimplantation embryos, given that the Xlr, Magea, and Rhox families are under the control of H3K9me2 [26,27] and Xist RNA overlaps with H3K9me2 [28], the results showing the Xist dependency of these families are reasonable.
Single-cell analysis showed that the number of SEGs in the comparison between males and Xist KO females was greater than that between in males and WT females (Fig. 2C,E). Although the exact reason for the difference remains unclear, and given that paternal Xchromosome reactivation occurs transiently in WT females [6,7] and RNA-seq analysis identified many DEGs between WT and Xist KO female embryos (Fig. 1C), the results suggest that EPI cells in Xist KO females are not exactly the same as those in WT females. Moreover, among of the up-regulated genes in Xist KO females compared with males, Sox2 and Esrrb were identified as SEGs [29,30]. As both genes are known as pluripotency-related genes, a defect to exit to pluripotency observed in female embryonic stem cells [23] may also occur in Xist KO female epiblasts.
In contrast, analysis of the PrE lineage revealed that that most transcriptional and epigenomic factors analyzed in this study were up-regulated in Xist KO female cells compared to WT female cells. These results are consistent with the genome-wide scale of transcriptional states in Xist KO trophoblasts [31]. Although the mechanisms underlying the up-regulation of X-linked genes and autosomal genes by loss of iXCI remain unknown, many transcription and epigenetic factors are mapped to X-chromosome such as Atrx and Mecp2. Indeed, Mecp2 is highly up-regulated in PrE and Xist KO female embryos and plays an important role in gene silencing [32]. Therefore, multiple derepression of gene silencers mapped to the X-chromosome may cause genome-wide up-regulation of autosomal genes.
Finally, in terms of ICM development, we found that PrE development in Xist KO female embryos was similar to that in males, supported by the single-cell qPCR results. Poor development of PrE in Xist KO female embryos was previously observed when SOX17positive cells were reduced [8]. In contrast to PrE development, we found no differences in EPI development in Xist KO females compared to WT females. Although the inactive X-chromosome is reactivated in EPI [2], Xist-dependent X-linked gene silencing occurs before X-chromosome reactivation [8,17]. Thus, these results indicate that having both X-chromosomes continuously active is dispensable for EPI specification in blastocysts. In conclusion, sex differences exist during early embryonic development in mice, and ICM development differs between males and females. In female embryos, iXCI impacts PrE development, revealing a novel role for iXCI in lineage specification.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.  Fig. S2. Venn diagram and KEGG pathway analysis. Fig. S3. Effect of preamplification on qPCR assay. Fig. S4. Expression states of lineage markers. Table S1. RNA-seq data with more than 1 RPKM. Table S2. Differentially expressed gene lists. Table S3. Single qPCR data. Table S4. TaqMan probe information.