Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Oct 1.
Published in final edited form as: Nat Cell Biol. 2013 Sep 8;15(10):1141–1152. doi: 10.1038/ncb2839

SON connects the splicing-regulatory network with pluripotency in human embryonic stem cells

Xinyi Lu 1,2,#, Jonathan Göke 1,#, Friedrich Sachs 1,3, Pierre-Étienne Jacques 4, Hongqing Liang 1, Bo Feng 5, Guillaume Bourque 6, Paula A Bubulya 7, Huck-Hui Ng 1,2,3,8,9,11
PMCID: PMC4097007  NIHMSID: NIHMS612091  PMID: 24013217

Abstract

Human embryonic stem cells (hESCs) harbour the ability to undergo lineage-specific differentiation into clinically relevant cell types. Transcription factors and epigenetic modifiers are known to play important roles in the maintenance of pluripotency of hESCs. However, little is known about regulation of pluripotency through splicing. In this study, we identify the spliceosome-associated factor SON as a factor essential for the maintenance of hESCs. Depletion of SON in hESCs results in the loss of pluripotency and cell death. Using genome-wide RNA profiling, we identified transcripts that are regulated by SON. Importantly, we confirmed that SON regulates the proper splicing of transcripts encoding for pluripotency regulators such as OCT4, PRDM14, E4F1 and MED24. Furthermore, we show that SON is bound to these transcripts in vivo. In summary, we connect a splicing-regulatory network for accurate transcript production to the maintenance of pluripotency and self-renewal of hESCs.

Human embryonic stem cells (hESCs) have the unique ability to self-renew and differentiate into cells of the three embryonic lineages1,2. As such, hESCs possess clinical potential, whereby clinically relevant cell types can be derived and potentially used for cell-based therapy3,4. To fully harness the potential of hESCs, it is important to understand the molecular mechanisms that govern hESC identity.

We have previously performed a genome-wide RNA interference (RNAi) screen to identify candidate genes that are important for the maintenance of the hESC identity5. Transcription factors and transcription co-factors are over-represented among the top candidates, and this is consistent with their important functions in embryonic stem cells6–14 (ESCs). One of the top ten candidates is the spliceosome-associated factor SON, which was initially proposed to be a potential DNA-binding protein15,16. Recent studies suggest that SON is a protein localized in nuclear speckles17,18, and that it is involved in cell-cycle progression and pre-messenger RNA processing19–22. Furthermore, SON was shown to physically interact with the spliceosome and deficiency of SON compromises spliceosome function in HeLa cells19,21.

In this study, we examined the function of SON in hESCs and found that SON is essential for hESC survival and pluripotency. We show that SON is required for proper splicing of transcripts encoding the cell-cycle protein TUBG1 and the pluripotency regulators OCT4, PRDM14, MED24 and E4F1 in hESCs. Using an RNA immunoprecipitation assay, we further confirm that SON binds to these transcripts. Moreover, we show that SON depletion induces the expression of a previously un-annotated isoform of PRDM14 that lacks the reprogramming capacity of the known isoform. Taken together, our study expands our understanding of hESC pluripotency beyond the transcriptional level and demonstrates that specific regulation of RNA splicing serves as an important regulatory mechanism in hESCs.

RESULTS

SON regulates cell survival and maintenance of pluripotency in hESCs

Our previous genome-wide RNAi screen had identified a repertoire of regulators essential for the maintenance of hESC identity5. Although transcription factors and transcription co-factors are enriched among the top candidate genes, we were particularly interested in a candidate (SON) that is ranked ninth among the 21,000 analysed genes. SON is also ranked the highest among all splicing-related factors. SON is expressed in multiple cell types at varying levels (Supplementary Fig. S1a). Importantly, induction of pluripotency in fibroblasts (MRC5) using the 4 Yamanaka factors (OCT4, SOX2, KLF4 and C-MYC) results in increased expression of SON, comparable to that observed in hESCs (Supplementary Fig. S1b,c). In addition, hESC differentiation induced by PRDM14 or NFRKB depletion or removal of FGF is accompanied by downregulation of SON (Supplementary Fig. S1b,d), further highlighting the particular relevance of SON in the context of pluripotency. To confirm the finding of our previous RNAi screen result5, we designed four short hairpin RNAs (shRNAs) that efficiently deplete SON at both transcript and protein levels (Fig. 1a,b and Supplementary Fig. S2a,b). Remarkably, depletion of SON resulted in the loss of alkaline phosphatase staining and induced hESC differentiation into fibroblast-like cells (Fig. 1c). The differentiation phenotype of SON-knockdown cells was also supported by a downregulation of pluripotency-associated genes (Fig. 1d and Supplementary Fig. S2b) and hESC surface markers (Supplementary Fig. S2c,d). On the other hand, expression of differentiation markers was upregulated whereas the expression of housekeeping genes remained unchanged (Fig. 1e and Supplementary Fig. S2b–e). Interestingly, depletion of SON during reprogramming in the presence of 4 Yamanaka factors (OCT4, SOX2, KLF4 and C-MYC) leads to a reduction in the efficiency of induced pluripotency (Supplementary Fig. S2f). Furthermore, co-expression of an RNAi-resistant form of SON in hESCs could rescue the effect of SON knockdown, confirming that the observed hESC differentiation phenotype and concomitant loss of pluripotency marker expression are SON-depletion specific (Fig. 2a,b and Supplementary Fig. S2g). Together, these results show that expression of SON is enriched in hESCs and confirm that SON is important for the maintenance of the pluripotent state as well as induction of pluripotency.

Figure 1.

Figure 1

Open in a new tab

The effects of SON depletion in hESCs. (a) Expression level of SON transcript after SON shRNA treatment as determined by qPCR. Four shRNAs were designed to target SON independently. shRNA against the Luciferase gene was used as the control. Samples were analysed 4 days after transfection. Mean±s.e.m. from biological triplicate data is shown in all graphs. (b) Immunostaining of SON in hESCs. DAPI positively stains the nucleus. (c) Alkaline phosphatase staining of hESCs after SON depletion. (d) Expression of pluripotency markers as determined by qPCR after SON depletion. Biological triplicate data for qPCR are presented as mean±s.e.m. (e) Expression of differentiation markers as determined by qPCR after SON depletion. Biological triplicate data are presented as mean±s.e.m. Scale bars, 20 μm.

Figure 2.

Figure 2

Open in a new tab

Rescue of SON depletion phenotype in hESCs. (a) Expression of SON and pluripotency markers in hESCs expressing SON shRNA against SON UTR, SON shRNA+RNAi immune SON or control shRNA. Mean±s.e.m. of biological triplicate data is shown in all graphs. The P value was determined by two-tailed Student's t -test. (b) Immunostaining of pluripotency cell surface marker after transfecting hESCs carrying RNAi immune SON with shRNA targeting the UTR of the SON transcript. (c) Cell morphology of hESCs transfected by SON shRNA and treated with ROCK inhibitor (inh.). (d) Expression of SON and pluripotency markers after depletion of SON in the presence of ROCK inhibitor. Mean±s.e.m. from biological triplicate data is shown. (e) Expression of differentiation markers after depletion of SON in the presence of ROCK inhibitor. Mean±s.e.m. from biological triplicate data is shown. Scale bars, 20 μm.

Besides the differentiation phenotype, SON depletion induces a marked reduction of cell number. TUNEL (TdT-mediated dUTP nick end labelling) assay reveals an increased incidence of DNA fragmentation, suggesting that this reduction in cell number is due to cell death (Supplementary Fig. S3a). Notably, despite the cell death and differentiation phenotype of SON knockdown in hESCs, knockdown of SON in MRC5 fibroblast cells slows down the growth of the cells but does not lead to massive apoptosis (Supplementary Fig. S3b–d), suggesting that SON has a cell-type-specific role in hESCs. Moreover, in agreement with previous findings in HeLa cells19,21,22, SON depletion in hESCs leads to cell-cycle defects (Supplementary Fig. S3e–g). It has been suggested that cell-cycle arrest may initiate differentiation in hESCs (ref. 23). To exclude the possibility that the observed cell differentiation phenotype is a secondary effect of apoptosis induced by cell-cycle defects, we depleted the cell survival genes TUBA1C and CCNA2 as a control. TUBA1C is a major constituent of microtubules, which are crucial for cell-cycle progression and cell survival24. CCNA2 is a protein important to cell-cycle progression through G1/S and G2/M phase25. Depletion of TUBA1C or CCNA2 in hESCs resulted in increased cell death, but failed to induce fibroblast-like morphology (Supplementary Fig. S3h–i). Furthermore, expression of pluripotency markers remained unchanged following TUBA1C or CCNA2 depletion (Supplementary Fig. S3j–k), indicating that cell death and differentiation in hESCs could occur independently.

To further distinguish the cell death phenotype from the differentiation phenotype, we first analysed the kinetics following SON knockdown in hESCs. However, the kinetics of marker induction and cell-cycle changes is similar (Supplementary Fig. S4a–e). Therefore, to investigate the two processes (cell survival versus pluripotency), we inhibited cell apoptosis with the ROCK inhibitor26. Despite a reduction in apoptosis, hESCs underwent differentiation after depletion of SON with concomitant downregulation of pluripotency markers and upregulation of differentiation markers (Fig. 2c–e and Supplementary Fig. S4f–g). Moreover, fluorescence-activated cell sorting analysis following TUNEL assay revealed that only ~50% cells undergo apoptosis after SON depletion, whereas TUBA1C depletion led to massive cell death (~85%) in hESCs (Supplementary Fig. S4h). Notably, addition of ROCK inhibitor reduced the number of cells undergoing apoptosis after SON depletion from 50% to ~4%. Furthermore, staining of the pluripotency marker SOX2 together with the apoptotic marker γ-H2AX (ref. 27) reveals that apoptotic cells expressed the pluripotency gene at a similar level to living cells after SON depletion (Supplementary Fig. S4i). In addition, pluripotency marker TRA-1-60 expression is downregulated in cells at all cell-cycle stages examined after SON knockdown (Supplementary Fig. S4j). These observations indicate that cell death and differentiation occur concurrently but independently after SON depletion.

Genome-wide profiling identifies SON-regulated transcripts

SON is known to regulate splicing of transcripts encoding cell-cycle proteins in HeLa cells21,22. To further investigate the function of SON in hESCs, we performed paired-end mRNA sequencing (RNA-seq) after a 3-day depletion of SON (Fig. 3a and Supplementary Fig. S5a). First, we calculated differential gene expression between control knockdown and SON knockdown in hESCs. Genes that show increased expression after depletion of SON are significantly enriched in developmental gene ontology (GO) terms, confirming that cells start to undergo differentiation (Fig. 3b and Supplementary Fig. S5b). When we examined the RNA-seq data for TUBG1, a known target of SON, we found that SON depletion indeed results in severe splicing defects (Fig. 3c). Most notably, the data show increased levels of transcripts with intronic sequences, indicating that SON depletion induces intron inclusion in mature mRNA. Importantly, the data also confirm the specificity of SON-regulated splicing, as only introns 7–10 of TUBG1 were affected (Fig. 3c).

Figure 3.

Figure 3

Open in a new tab

RNA-seq analysis of transcripts regulated by SON. (a) Scheme of the RNA-seq analysis. Cells were transfected with constructs expressing SON shRNA (shRNA1) or control-shRNA-expressing constructs for 3 days. RNA was extracted from pooled samples. Next, poly(A)-tailed mRNA was enriched and fragmented for sequencing. Data analysis was carried out for three aspects: gene expression, intron expression and alternative splicing. (b) Gene ontology analysis of upregulated genes after 3-day SON depletion according to RNA-seq data. (c) Fold change of RNA-seq reads from control and SON knockdown at the TUBG1 locus (hg19). Shown are fold changes with P < 0.4. Introns that are known SON targets are marked by asterisks. (d) Intron inclusion is estimated using the ratio of spliced over unspliced reads that cover exon–intron boundaries. See Supplementary Information for alternative estimation of intron inclusion. (e) Gene ontology analysis of genes with intron inclusion after SON depletion. (f) Number of events with increased and decreased intron inclusion after depletion of SON. (g) Log ratio of spliced over unspliced reads at SON-targeted introns for control and SON knockdown. Significance was estimated using the Wilcoxon signed-rank test.

To identify splicing defects following SON depletion, we first identified all splice events using the set of mapped RNA-seq reads that span a splice junction (spliced reads, Fig. 3d, in total 466,085 junctions). We further restricted our analysis to high-confidence junctions that can be associated with known exons (Methods for details). The reduced set covers 173,874 introns and 99% of all spliced reads. We then calculated the number of spliced and unspliced reads that span exon–intron boundaries to estimate the rate of intron inclusion (Supplementary Information for results from alternative estimation of intron expression). Genes that show increased intron inclusion following SON depletion are involved in different biological processes, such as the cell cycle, as has been reported previously19–21 (Fig. 3e). To identify the most severely affected introns after SON depletion, we calculated a t-test statistic score and selected a test-score cutoff of −8 (Supplementary Fig. S5c); and we further required that at least 5 spliced reads support the splice site. Using these thresholds, we obtained a set of 1,994 introns from 1,127 genes whose splicing was impaired following SON depletion (Supplementary Table S1). From this set, we validated 30 intron inclusion events using PCR with reverse transcription (RT–PCR), confirming the robustness of our method to correctly identify splicing defects after depletion of SON (Supplementary Fig. S5d,e). Reassuringly, our analysis identifies previously known targets of SON (TUBG1, TUBGCP2, AURKB, FANCG and HDAC6, Supplementary Table S1). Consistent with what we observed for TUBG1, most of these introns show a significant increase of intronic signal when compared with exonic signal after the knockdown of SON, confirming intron inclusion as a major splicing defect (Fig. 3f,g and Supplementary Fig. S5f).

SON regulates splicing of specific short introns with weak splice sites

To better understand the mechanism behind SON-regulated splicing, we investigated whether SON-regulated introns exhibit specific properties. For this analysis we selected introns from the same genes, but which did not show significant inclusion, as a control to correct for gene to gene variations. Interestingly, we find that SON-regulated introns are significantly shorter than unaffected introns from the same genes (Fig. 4a, Wilcoxon test P = 3.85×10−131; Supplementary Fig. S6a,b). Furthermore, we found that genes with introns regulated by SON had a significantly weaker splicing strength when compared with unaffected introns (Fig. 4b,c, Wilcoxon test P = 1.62 × 10−16 (5′); P = 3.51×10−11 (3′); Supplementary Fig. S6c,d; ref. 28). These observations are consistent, as splicing strength tends to positively correlate with intron length29. As splicing strength is a property of the splice junction sequence, we calculated the nucleotide frequency for SON-targeted introns. Strikingly, SON-targeted introns show a very distinct sequence from the genome-wide average (Fig. 4d,e). Even when we considered only SON-targeted introns and non-targeted introns from the same genes, we still observed a difference in sequence composition, most notably a higher GC content in SON-targeted introns (Fig. 4f and Supplementary Fig. S6e–g). Introns with high levels of GC content have been associated with intron inclusion events30. In summary, these results highlight the large-scale splicing deficiency at preferentially short introns with weak splice sites following SON knockdown in hESCs, and indicate that sequence-specific properties might be involved in SON-mediated regulation of splicing.

Figure 4.

Figure 4

Open in a new tab

SON regulates splicing of small introns with weak splice sites. (a) Size distribution (box plot) of SON-targeted introns, and non-target introns from the same genes. Significance was estimated using the Wilcoxon signed-rank test. (b) Splice strength for the 5′ splice sites, estimated using MaxEntScan; only introns from genes that had at least one affected intron are compared to account for local properties. (c) Splice strength for the 3′ splice sites, estimated using MaxEntScan; only introns from genes that had at least one affected intron are compared to account for local properties. (d) Nucleotide composition of 5′ and 3′ splice sites of all introns. (e) Nucleotide composition of 5′ and 3′ splice sites of introns affected by SON depletion. (f) GC content for all introns (black), and introns from genes that have at least one intron affected after depletion of SON (red: affected introns, green: unaffected introns from the same genes).

As the activity of SON could be due to a general targeting effect of SON on all weak spliced introns, we determined the expression of other introns with weak splicing sites after SON depletion in hESCs to exclude this possibility. We found that splicing of weak introns from CEBPZ, SPAG5 and SMARCE1 is not affected by SON depletion (Supplementary Fig. S6h). In addition, we examined the effect of depletion of another weak splicing regulator (TIA1) to determine whether it functions the same as SON in hESCs. Knockdown of TIA1 does not cause morphological changes in hESCs (Supplementary Fig. S6i), and the expression of pluripotency markers as well as differentiation markers remains unchanged (Supplementary Fig. S6j). There are also no changes in splicing for introns regulated by SON after TIA1 knockdown (Supplementary Fig. S6k). These data suggest that SON exerts a specific function in regulating splicing of transcripts.

SON regulates pluripotency by promoting splicing of pluripotency genes

As SON expression is required for maintaining pluripotency in hESCs, we investigated whether splicing of pluripotency-associated genes is affected by SON depletion. By overlapping SON-regulated transcripts with genes essential for hESC identity maintenance5, we identified 41 candidates where the observed splicing defects might directly affect pluripotency (Fig. 5a, P = 0.0093). Using an independent set of hESC-specific genes, we confirmed the significant enrichment of pluripotency-associated genes in the set of SON-targeted genes (Supplementary Fig. S7a, P value = 8.8×10−5). Moreover, by overlapping SON-targeted genes with genes essential for hESC cell survival5, we found that SON regulates splicing of 45 genes essential for hESC survival (Fig. 5b, P = 0.00016). These introns regulated by SON from pluripotency genes and cell survival genes are also shorter than non-targeted introns (Supplementary Fig. S7b–d). Notably, OCT4, PRDM14, MED24 and E4F1 are among the pluripotency genes targeted by SON (Fig. 5a,c). Whereas all four genes are important regulators of the hESC identity, both OCT4 and PRDM14 are also key transcription factors involved in induced pluripotency5,31.

Figure 5.

Figure 5

Open in a new tab

SON targets specific introns of transcripts that encode pluripotency regulators. (a) Venn diagram showing the overlap of hESC identity maintenance genes (green) with genes targeted by SON. Positive control genes are highlighted in blue. Genes essential for hESC identity maintenance were defined as genes with average pluripotency F score < −2 (Supplementary Table S1 from genome-wide RNAi screening in hESCs (ref. 5)). Significance was estimated using Fisher's exact test. (b) Venn diagram showing the overlap of self-renewal genes (red) with genes regulated by SON. Genes essential for hESC survival were defined as genes with average cell survival N score < −2 (Supplementary Table S2 from genome-wide RNAi screening in hESCs (ref. 5). The P value was calculated using Fisher's test. (c) Plot of quantile-normalized H3K27 acetylation at promoters in H1 hESCs and 20 non-pluripotent cell lines47. SON-targeted genes are highlighted in red. Examples of SON-targeted genes with a hESC-specific epigenetic profile are highlighted. (d) Schematic illustration of the structural features of genes regulated by SON, showing: control introns (line, grey label), introns regulated by SON (line, red label), and control exons (filled rectangle, green label). Primer locations are labelled as arrows. (e) qPCR analysis of intron inclusion in transcripts of pluripotency genes (PRDM14, OCT4, E4F1 and MED24), positive control gene TUBG1 and negative control gene TUBA1B after 3-day depletion of SON with SON shRNA1. Mean±s.e.m. from biological triplicate data is shown. The P value was determined by Student's t-test. (f) RNA-CLIP analysis on interaction of SON with intronic sequences of pluripotency-associated transcripts. Biological triplicate data are presented as mean±s.e.m.

To confirm that SON depletion affects splicing of these key factors, we validated these predictions using quantitative PCR (qPCR). Indeed, SON-targeted introns from OCT4, PRDM14, MED24 and E4F1 transcripts were upregulated (Fig. 5d,e), whereas non-targeted introns from the same genes were not affected following knockdown of SON (Fig. 5d,e). A time-course analysis confirms that intron inclusion for these transcripts increased with extended time of SON depletion (Supplementary Fig. S7e). This may explain why significant cell differentiation occurs only at a certain time point after a sufficient amount of unspliced transcripts accumulated at day 3.5. By comparing the expression of all the introns and exons in the OCT4 transcript, we found that only specific introns are regulated by SON (Supplementary Fig. S7f). Notably, another important pluripotency regulator, NANOG, is not affected by the depletion of SON (Supplementary Fig. S7g). This suggests that SON might be governing hESC pluripotency in a specific manner.

To address whether SON is directly bound to these transcripts in hESCs, we performed RNA crosslinking and immunoprecipitation (RNA-CLIP) assays. SON was first crosslinked to RNA by ultraviolet irradiation. The SON–RNA complexes were then purified using SON-specific antibodies. Real-time PCR analysis showed enrichment of unspliced transcripts of OCT4, PRDM14, MED24, E4F1 and TUBG1 for SON RNA-CLIP, but not for control antibody RNA-CLIP (Fig. 5f). The intronic sequence for the negative control transcript TUBA1B was not enriched by SON RNA-CLIP (Fig. 5f). Hence, the RNA-CLIP results indicate that SON physically interacts with specific introns from OCT4, PRDM14, MED24, E4F1 and TUBG1. In summary, these data indicate that SON regulates hESC pluripotency and cell survival by promoting specific splicing of introns from transcripts encoding pluripotency regulators and cell-cycle proteins.

SON-regulated transcripts are subjected to nonsense-mediated decay

SON depletion in hESCs leads to the accumulation of a number of transcripts containing unspliced introns; these unspliced transcripts, which may contain a premature stop codon, could be regulated by the nonsense-mediated mRNA decay (NMD) pathway32–34. Interestingly, the expression of cytoplasmic proteins essential for the NMD pathway, such as UPF1, is affected by SON depletion as well (Fig. 6a–c). This might explain why intron retention is observed in poly(A) mRNA after SON depletion. The degradation of transcripts through the NMD pathway eventually leads to the reduction of corresponding protein expression (Fig. 6c). Depletion of SON together with UPF1 and UPF2 leads to increased accumulation of unspliced transcripts (Fig. 6d), confirming that unspliced transcripts are regulated by the NMD pathway.

Figure 6.

Figure 6

Open in a new tab

Unspliced transcripts induced by SON depletion are potentially targeted by the NMD pathway. (a) The fold change of RNA-seq reads from control and SON knockdown at the UPF1 locus (hg19). Shown are fold changes with P < 0.4. (b) Expression of UPF1 and UPF2 after transfection of shRNAs according to qPCR. Mean±s.e.m. from biological triplicate data is shown. (c) Expression of proteins regulated by SON depletion determined by western blotting. (d) Expression of introns regulated by SON after depletion of SON together with UPF1 and UPF2 depletion according to qPCR. Mean±s.e.m. from biological triplicate data is shown. Uncropped images of blots are shown in Supplementary Fig. S9.

SON regulates the alternative splicing in hESCs

As SON has an important splicing regulatory function, we studied whether alternative splicing can be observed following SON depletion in hESCs besides intron inclusion. To detect alternative splicing events, we calculated the fold change of the read count at every exon–exon junction (Fig. 7a). As we were interested in a change of junction usage relative to all other junctions of the same gene, we calculated differential splicing for every gene separately, thereby correcting for the overall change of expression. In total, we identified 940 splicing events in 694 genes that show a change after depletion of SON (Supplementary Table S2). Strikingly, most of these alternative splicing events are skipped exons after knockdown of SON (Fig. 7b) and most encode minor isoforms (Fig. 7c). Interestingly, many alternative splicing events overlap SON-regulated introns (Fig. 7d, Fisher's test P = 5.3×10−123), suggesting that exon skipping is indeed a result of loss of the splicing regulatory function of SON.

Figure 7.

Figure 7

Open in a new tab

SON regulates alternative splicing in hESCs. (a) Scheme to illustrate our approach to identify splicing differences between control and SON knockdown. (b) Histogram showing the number of skipped exons for SON-knockdown-induced alternative splicing events and the genome-wide distribution. Significance was estimated using Fisher's exact test. (c) Fraction of splice junction usage induced by SON depletion and the genome-wide distribution. (d) Percentage of exon-skipping events induced by SON depletion that overlap intron inclusion events induced by SON depletion. Significance was estimated using Fisher's exact test.

SON knockdown induces the alternative splicing of PRDM14

As for intron inclusion events, we find that SON-knockdown-induced alternative splicing events significantly overlap with hESC-specific genes (Fig. 8a, Fisher's test P = 9.66×10−8) as well as genes regulating pluripotency and survival of hESCs (Supplementary Fig. S8a,b). One of the top candidates of SON-regulated alternative splicing events was again from the gene coding for the pluripotency regulator PRDM14. The RNA-seq data show that the expression of exon 2 but not other exons from PRDM14 was specifically downregulated after SON depletion (Fig. 8b). At the same time, there was a significant increase in transcripts that connect exons 1 and 3, corresponding to a previously un-annotated shorter isoform of PRDM14 (Fig. 8b). Therefore, we validated the isoform expression of PRDM14 after SON depletion using RT–PCR. The results strikingly show the isoform switch following SON depletion (Fig. 8c). We identified a shorter isoform of PRDM14 present in hESCs and this isoform is upregulated after SON depletion whereas the full-length known isoform of PRDM14 is downregulated (Fig. 8c). By sequencing the RT–PCR product, we confirmed that exon 2 of PRDM14 was skipped in the PRDM14 short isoform (Fig. 8c). By co-expressing either PRDM14 isoform with a reporter carrying PRDM14 motifs, we found that the PRDM14 long isoform but not the short isoform is functional as an activator of gene expression (Fig. 8d). Moreover, only the long PRDM14 isoform could enhance reprogramming efficiency, whereas the short isoform failed, suggesting that the long isoform is the functional isoform in the maintenance and gain of pluripotency (Fig. 8e). Previously it was shown that PRDM14 is essential for silencing the expression of developmental genes35. Thus, this isoform switch of pluripotency factor PRDM14 may potentially serve as a main downstream event driving the subsequent differentiation of hESCs. Together, these data indicate that SON links splicing with the core regulatory network that maintains pluripotency by regulating correct splicing of transcripts encoding pluripotency factors.

Figure 8.

Figure 8

Open in a new tab

SON knockdown induces an isoform switch of PRDM14. (a) Overlap of SON-regulated transcripts with hESC-specific transcripts. Significance was estimated using Fisher's exact test. (b) Top: the fold change of RNA-seq reads from control and SON knockdown at the PRDM14 locus (hg19). Shown are fold changes with P < 0.4. Bottom: fold change of normalized splice junction RNA-seq reads from control and SON knockdown; splicing events are highlighted in red (increased in SON knockdown) and green (decreased in SON knockdown). (c) RT–PCR analysis of PRDM14 isoform expression after SON depletion. Primers were placed at the beginning and end of the transcript. The exon-skipping event was confirmed by sequencing. (d) Luciferase assay analysis of the gene activation potential of the PRDM14 isoform. The reporter structure is shown at the top of the figure. Biological triplicate data are presented as mean±s.e.m. (e) Reprogramming in the presence of PRDM14 isoforms. OSKC: the 4 Yamanaka reprogramming factors: OCT4 (O), SOX2 (S), KLF4 (K) and c-MYC (C). Biological triplicate data are presented as mean±s.e.m. (f) Schematic of SON-regulated splicing in hESCs. In the presence of SON, introns are properly spliced out and correct isoforms of transcripts are expressed to maintain hESC pluripotency and self-renewal. In the absence of SON, exon skipping and intron retention occur, subsequently leading to differentiation and cell death of hESCs.

DISCUSSION

We propose a model (Fig. 8f) whereby SON interacts with other splicing factors in the spliceosome19,21 to promote the splicing of introns and inclusion of alternative exons from genes essential for hESC identity. SON may provide a control switch for appropriate processing of a specific set of transcripts that alter cell fate. Hence, splicing regulation by SON is important for normal expression of these pluripotency genes and cell survival genes in the maintenance of hESC identity.

Transcript splicing is a key regulatory step for the proper expression of intron-containing genes36,37. Although it is known that profound changes in alternative splicing occur during ESC differentiation38,39, the precise mechanism(s) that regulate this process is not well studied. In mouse ESCs, alternative splicing has been found to play a major role in cell fate determination during differentiation40. In hESCs, a deficit in the splicing factor FOX2 leads to cell death41. An ESC-specific splicing switch in FOXP1 transcripts has been reported to play a role in the maintenance of pluripotency42. More recently, MBNL proteins were found to inhibit ESC-specific alternative splicing43.

Apart from cell survival and cell fate determination, SON-regulated splicing is also essential for the maintenance of pluripotency in hESCs. Our study has demonstrated that depletion of the splicing factor SON led to both cell death and differentiation of hESCs. The differentiation phenotype observed after knockdown of SON is conversely unique to hESCs. Hence, we propose that SON regulates hESC identity by influencing the splicing of transcripts encoding pluripotency factors such as OCT4, PRDM14, MED24 and E4F1. Notably, OCT4 is a member of core transcriptional circuitry in hESCs, and mis-regulation of OCT4 expression causes cell differentiation31,44. PRDM14 was found to play a central role in the maintenance and gain of pluripotency5,35. Mediator complexes were also found to play an important role in maintaining ESC identity45, and MED24 depletion in hESCs similarly causes cell differentiation5. E4f1, a transcription factor that was also previously identified in a hESC RNAi screen5, has been shown to be essential to mouse ESC maintenance46. Hence, the inappropriate splicing of OCT4, MED24 and E4F1 transcripts and isoform switch of PRDM14 may contribute to the differentiation phenotype induced by SON deficiency in hESCs. Together, our study links the splicing-regulatory network with pluripotency through SON, which targets genes that encode for key hESC regulators, demonstrating how accurate splicing is required to maintain hESC identity.

METHODS

hESC culture and transfection

The hESC line H1 (WA-01, passage 30) was cultured feeder-free in conditioned medium on Matrigel (BD; ref. 48). Conditioned medium contains 20% Knock-out serum replacement (Gibco), 1 mM l-glutamine (Gibco), 1% non-essential amino acids (Gibco) 0.1 mM 2-mercaptoethanol (Gibco) and 4 ng ml−1 human basic fibroblast growth factor (bFGF; Invitrogen), and medium was supplemented with additional bFGF (8 ng ml−1) before use. The hESCs were passaged for expansion with 1 mg ml−1 collagenase IV (Gibco) every 5–7 days. For transfection, the hESCs were passaged with 0.25% trypsin (Invitrogen) for dissociation and transfected with 1.5 μg shRNA and 4.5 μl Mirus LT1 transfection reagents the following day. Cells were selected under 1 μg ml−1 puromycin after transfection and collected 4 days after transfection for RNA. For RNA-seq, cells were collected for mRNA 3 days after transfection of SON shRNA1. For rescue experiments, CAG–SON was transfected into hESCs to establish hESCs overexpressing SON. shRNA (1 μg) targeting the UTR region of SON or control shRNA was transfected into hESCs. Cells were collected 4 days after transfection for RNA or cell imaging. For inhibition of apoptosis, ROCK inhibitor was added 3 days after depletion of SON and incubated for 1 day before collection. For depletion of SON, UPF1 and UPF2 at the same time, 1.2 μg SON shRNA was transfected together with 1.2 μg of UPF1 and UPF2 shRNA. Sequences of shRNAs: 5′-GATGAAATGGGTAAGTACA-3′ (control shRNA), 5′-GCTGAGCGCTCTATGATGT-3′ (SON shRNA1), 5′-GATACAGAACTACGATATA-3′ (SON shRNA2), 5′-GGTCTTTCGTGGTCAGTAA-3′ (SON shRNA3), 5′-AATGTCAGTGGAGTATCA-3′ (SON shRNA4), 5′-GGACTAGCGAGGAGGAGTT-3′ (SON shRNA UTR), 5′-GGCTGCCCTTGAGAAGGATTA-3′ (TUBA1C), 5′-CCATTGGTCCCTCTTGATT-3′ (CCNA2), 5′-GATGCAGTTCCGCTCCATT-3′ (UPF1; ref. 49), 5′-GAAGTTGGTACGGGCATC-3′ (UPF2; ref. 50), 5′-GCTCTAATTCTGCAACTCTTT-3′ (TIA1 shRNA1; ref. 51), 5′-GCTCTAATTCTGCAACTCTTT-3′ (TIA1 shRNA2; ref. 51). The cell line used has not been authenticated recently but it has been tested for mycoplasma contamination.

Virus production and reprogramming

SON shRNAs, PRDM14 or the PRDM14 isoform was cloned into pMX vector for retrovirus-mediated knockdown. Retroviruses were packaged using the Pantropic Retroviral Expression System (Clontech) and concentrated with centrifugal filter devices (Millipore). MRC-5 human lung primary fibroblast cells obtained from ATCC were cultured in 15% FBS/DMEM. Confluent MRC-5 cells from a 6-well plate were split into 24 wells one day before and then transduced with retroviruses containing 4 Yamanaka factors (OCT4, SOX2, KLF4 and C-MYC) with or without custom viruses in the presence of 4 μg ml−1 Polybrene (Sigma). After 24 h, the cells were changed to fresh 15% FBS/DMEM medium. MRC-5 cells were passaged on CF-1 feeder cells and maintained in conditioned hESC medium for 3–4 weeks before collection for immunostaining.

Reverse transcription and qPCR

Total RNA was extracted from cells using the TRIzol reagent (Invitrogen) and was then treated with DNase (Promega) for 1 h. The concentration of RNA was determined by NanoDrop 2000 (ThermoScientific). DNase-treated total RNA (500 ng) was reverse transcribed with oligo(dT) and SuperScript II (Invitrogen). qPCR was performed with 2 μl 5× diluted cDNA per 10μl reaction and KAPA SYBR Green (KapaBiosystems). Quantitative real-time PCR was performed on the PRISM 7900HT system (Applied Biosystem). Gene expression levels were then normalized to that of β-actin. For semi-quantitative RT-PCR, 2 μg total RNA was reversed transcribed. For downstream PCR, 1/20 of the reverse transcription product was used. Band intensity was determined by ImageJ and the percentage of unspliced transcript was calculated. Primers used for detection of intron inclusion are listed in Supplementary Table S3.

Alkaline phosphatase staining

The alkaline phosphatase detection kit (Chemicon) was used according to the manufacturer's instructions to determine alkaline phosphatase activity. H1 hESCs were cultured in a 6-well plate, and then fixed with 4% formaldehyde at 25 °C for 30 min. Cells were then permeabilized in PBS containing 0.1% Tween-20 for 10 min. Staining solution was later added to cells and incubated for 30 min. Cells were washed with PBS 3 times and visualized by microscopy. Images of ×200 magnification were captured by the Leica Application Suite using a Leica DM IRD microscope. Experiments were repeated 3 times.

TUNEL assay and fluorescence-activated sorting analysis

TUNEL assays were performed using the in situ Cell Death Detection Kit (Roche). H1 hESCs were cultured in a 6-well plate, fixed with 4% formaldehyde at 4 °C overnight and incubated with the TUNEL labelling mixture for 1 h at 37 °C. Plates were washed with PBS and imaged at ×200 magnification. Experiments were repeated 3 times. To quantify the TUNEL assay, fluorescence-activated sorting analysis was used to determine the composition of the cell population after staining. H1 hESCs were cultured in a 10 cm plate and treated with SON shRNA and control shRNA. After 4-day treatment, cells were detached from the plate by trypsin treatment and dissociated into single cells. TUNEL assay was performed with these cells, and cells were analysed at 530 nm. For cell-cycle analysis, SON was depleted from 2 to 3.5 days. Cells were then collected and fixed in ethanol overnight. Propidium iodide was used to stain the cells for cell-cycle analysis by fluorescence-activated sorting analysis.

Immunostaining

Human ESCs were fixed with 4% formaldehyde in PBS. After permeabilization in 1% Triton X-100/PBS for 30 min, cells were blocked in 1% BSA for 30 min. Immunostaining was performed using the following primary antibodies: SSEA-4 (clone 813–70, sc-21704, Santa Cruz), GATA6 (sc-7244, Santa Cruz), MSX1 (sc-17726, Santa Cruz), Tra-1-81 (clone TRA-1–80, sc-21706, Santa Cruz), TRA-1-60 (clone TRA-1-60, sc-21705, Santa Cruz), OCT4 (clone C30A3, #2840, Cell Signaling), Gamma-H2AX (clone JBW301, 05-636, Millipore), AURORA A (clone 35C1, ab13824, Abcam), α-tubulin (clone B-5-1-2, T5168, Sigma). Secondary antibodies used were Alexa Fluor 546 anti-mouse IgG (A11003, Invitrogen), Alexa Fluor 546 anti-rabbit IgG (A11010, Invitrogen), Alexa Fluor 546 anti-goat IgG (A11056, Invitrogen), Alexa Fluor 488 anti-rabbit IgG (A21441, Invitrogen), Alexa Fluor 488 anti-goat IgG (A21467, Invitrogen), Alex Fluor 488 (A21200, Invitrogen) anti-mouse (A21200, Invitrogen), and Alexa Fluor 647 anti-mouse IgG (A21463, Invitrogen). Hoechst (Invitrogen) was used for staining the nuclei. Immunostaining experiments were repeated 3 times.

Western blot

Protein concentration was measured with a Bradford assay kit (Bio-Rad). Cell lysate (50 μg) was resolved on a 10% SDS–polyacrylamide gel and transferred to a polyvinylidinedifluoride membrane (Millipore). After blocking in 5% skimmed milk for 1 h, the appropriate primary antibodies were added: anti-SON (rabbit polyclonal antibody generated against the amino-terminal 250 amino acids of SON), anti-Oct4 (1:5,000, ab19857, Abcam), anti-NANOG (1:800, AF1997, R&D), anti-UPF1 (1:1,000, sc-48802, Santa Cruz), anti-DNMT1 (1:1,000, sc-20701, Santa Cruz), anti-PRDM14 (generated against the N-terminus of PRDM14)5, anti-SF2 (1:1,000, sc-28724, Santa Cruz), anti-GAPDH (1:1,000, sc-25778, Santa Cruz) or anti-β-actin (1:1,000, clone C4, sc-47778, Santa Cruz) primary antibodies for 1 h. After washing with PBS/0.1% Tween-20, horse-radish peroxidase (HRP)-conjugated anti-rabbit IgG (1:5,000, NA934V, Amersham), HRP-conjugated anti-goat IgG (1:5,000, sc-2020, Santa Cruz) or HRP-conjugated anti-mouse IgG (1:5,000, NA931V, Amersham), secondary antibodies were added for 1 h. After washing with PBST, signals were then detected using the West Dura Extender Duration Substrate (ThermoScientific). Western blot experiments were repeated twice.

RNA preparation and sequencing

Trizol (Invitrogen) was used for extraction of total RNA from hESCs treated with control shRNA or SONshRNA for 3 days. RNA (4 μg) extracted from biological duplicate samples was poly(A) enriched by incubating with oligo(dT)-coated Sepharose beads for two rounds (Ambion). Of the resultant enriched mRNA, 500 ng was fragmented, size selected and reverse transcribed according to the IlluminaTruSeq RNA Sample Preparation Guide. Multiplexed samples were sequenced with the IlluminaHighSeq system. Each library was sequenced for 2 × 80 million of 75-nucleotide reads.

Bioinformatics analysis of RNA-sequencing data

The reads of SON-shRNA-treated and control-shRNA-treated samples were mapped to the human genome assembly hg19 using TopHat52 version 1.4.1 with parameters -M –butterfly-search – microexon-search -n 2 -m 1 -N 3 -i 30 –coverage search, we also used the –GTF option with transcriptome annotations from Ensembl (GRCh37.69). Differential expression was estimated using cufflinks version 2.02 (ref. 53). Genes with increased expression in Fig. 3b were selected on the basis of the expression log-fold change between control and knockdown (cutoff 0.5). Intron inclusion was estimated by two different scores. As a first score we calculated the ratio of spliced over unspliced reads. Reads were counted that either mapped to the two nucleotides 5′ and 3′ of the splice site (unspliced reads) or that mapped only to the exonic nucleotide and to a second exon for the rest of the read (spliced read). For every sample we then divided the number of spliced reads for every junction by the number of unspliced reads that cover the 3′ or 5′ splice site. A pseudo-count of 1 was added both for spliced and unspliced reads. As the intron inclusion score we calculate the t-test statistic for the ratio of spliced over unspliced reads between control and knockdown. To obtain a set of high-confidence intron inclusion events, we required a test score below –8, a minimum number of 5 spliced reads that support the junction, and a minimum number of at least 3 unspliced reads for the 3′ and 5′ splice sites (Supplementary Table S1). For unaffected introns we required a test score between–5 and 1.5 and a minimum number of 5 reads that support the splice junction. To control for gene to gene variation we compared introns from the set of genes that contain both affected and unaffected introns. As a second score, we estimated intron expression. We counted reads that map into introns divided by the length of the introns and normalized this to the number of reads mapping into the adjacent exons (divided by the length of the exons). MaxEntScan was used to estimate splicing strength for all splice sites detected in our RNA-seq data using the Maximum Entropy model29. To estimate alternative splicing, we calculated the read count for all junctions used in this study (see main text). We associated every junction with the gene that was annotated for the adjacent exons. The log-fold change and significance were estimated using DESeq2 (ref. 54) for every gene separately to get estimates independent of the change in expression. Box plots were generated in R using standard settings. All bioinformatics analyses were carried out using R version 3.0 and Bioconductor version 2.12.

RNA crosslinking and immunoprecipitation

RNA immunoprecipitation experiments were performed according to a published protocol with modifications55. Briefly, hESCs were ultraviolet-irradiated at 254 nm, 600 mJ cm−2 (Stratagene Stratalinker) and cells were lysed in Triton buffer with short disruptive sonication. DNA was removed from nuclear lysate by 15 min DNase treatment and RNA was fragmented by 5 min RNase A treatment at 37 °C. Nuclear lysates were precleared with Protein G Dynabeads (Invitrogen) for 2 h at 4 °C, and then incubated along with antibodies overnight. RNA was immunoprecipitated with Protein G Dynabeads. Proteins were removed by Proteinase K (Fermentas) treatment after RNA elution. RNA was extracted using Trizol (Invitrogen) and RT–qPCR was performed as described above.

Luciferase assay

For the Luciferase assay, 75 ng pGL4 reporter with minimal promoter or plus 5xPRDM14-binding motifs was transfected into HEK293T cells (ATCC) together with 75 ng PRDM14 or PRDM14 isoform and 3 ng SV40–Renilla expression plasmid. Transfected cells were maintained in culture for 3 days. Cells were collected and the Luciferase assay was performed according to the Dual-Luciferase Reporter Assay System Protocol (Promega). The activity of the Luciferase gene is normalized against the activity of Renilla. The effect of PRDM14 or the PRDM14 isoform on reporter activity was normalized against minimal promoter or promoter with 5xPRDM14-binding motifs. The cell line used has not been authenticated recently but has been tested for mycoplasma contamination.

Statistical analysis

Statistical analysis was performed using R. Data are presented as mean±s.e.m. from biological triplicate data unless noted otherwise. Two-tailed Student's t-test was appropriately used to estimate the difference between two groups unless noted otherwise. The data are expected to be normally distributed and the variance is expected to be similar between the groups that are being statistically compared.

Supplementary Material

a
b
c
supp

ACKNOWLEDGEMENTS

We are grateful to the Agency for Science, Technology and Research (A*STAR) for their generous funding, and also for NIH R15 GM084407 to P.A.B. We acknowledge T. C. Peow, L-P. Yaw, T. Tng and G. K. Jee for their technical assistance. We acknowledge the Genome Technology Biology (GTB) group from the Genome Institute of Singapore (GIS) for their help and expertise in performing the RNA sequencing. We thank T-L. Huber for critical comments on the manuscript. J. Göke is supported by a fellowship within the Postdoc-Programme of the German Academic Exchange Service, DAAD.

Footnotes

Accession numbers. The primary ArrayExpress accession number for the RNA-seq data generated in this study is E-MTAB-1687 (primary accession), which is available at the ArrayExpress. Public data for H3K27ac (referenced accession) were downloaded from the UCSC ENCODE portal (http://encodeproject.org/ENCODE/).

AUTHOR CONTRIBUTIONS X.L.: conception, design, collection of data, data analysis and manuscript writing; J.G.: design, collection of data, data analysis and manuscript writing; F.S., H.L. and B.F.: collection and data analysis; P-É.J. and G.B.: collection and data analysis; P.A.B: design, and provision of study material; H-H.N.: conception, design, supervision, data interpretation and manuscript writing.

COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests.

Supplementary Information is available in the online version of the paper

References

  • 1.Thomson JA, et al. Embryonic stem cell lines derived from human blastocysts. Science. 1998;282:1145–1147. doi: 10.1126/science.282.5391.1145. [DOI] [PubMed] [Google Scholar]
  • 2.Reubinoff BE, Pera MF, Fong CY, Trounson A, Bongso A. Embryonic stem cell lines from human blastocysts: somatic differentiation in vitro. Nat. Biotechnol. 2000;18:399–404. doi: 10.1038/74447. [DOI] [PubMed] [Google Scholar]
  • 3.Wobus AM, Boheler KR. Embryonic stem cells: prospects for developmental biology and cell therapy. Physiol. Rev. 2005;85:635–678. doi: 10.1152/physrev.00054.2003. [DOI] [PubMed] [Google Scholar]
  • 4.Kimbrel EA, Lu SJ. Potential clinical applications for human pluripotent stem cell-derived blood components. Stem Cells Int. 2011;2011:273076. doi: 10.4061/2011/273076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chia NY, et al. A genome-wide RNAi screen reveals determinants of human embryonic stem cell identity. Nature. 2010;468:316–320. doi: 10.1038/nature09531. [DOI] [PubMed] [Google Scholar]
  • 6.Chen X, et al. Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell. 2008;133:1106–1117. doi: 10.1016/j.cell.2008.04.043. [DOI] [PubMed] [Google Scholar]
  • 7.Göke J, Chan YS, Yan J, Vingron M, Ng HH. Genome-wide kinasechromatin interactions reveal the regulatory network of ERK signaling in human embryonic stem cells. Mol. Cell. 2013;50:844–855. doi: 10.1016/j.molcel.2013.04.030. [DOI] [PubMed] [Google Scholar]
  • 8.Göke J, et al. Combinatorial binding in human and mouse embryonic stem cells identifies conserved enhancers active in early embryonic development. PLoS Comput. Biol. 2011;7:e1002304. doi: 10.1371/journal.pcbi.1002304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Jiang J, et al. A core Klf circuitry regulates self-renewal of embryonic stem cells. Nat. Cell Biol. 2008;10:353–360. doi: 10.1038/ncb1698. [DOI] [PubMed] [Google Scholar]
  • 10.Loh YH, et al. Genomic approaches to deconstruct pluripotency. Annu. Rev. Genom. Hum. Genet. 2011;12:165–185. doi: 10.1146/annurev-genom-082410-101506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Macarthur BD, Ma'ayan A, Lemischka IR. Systems biology of stem cell fate and cellular reprogramming. Nat. Rev. Mol. Cell Biol. 2009;10:672–681. doi: 10.1038/nrm2766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Orkin SH, Hochedlinger K. Chromatin connections to pluripotency and cellular reprogramming. Cell. 2011;145:835–850. doi: 10.1016/j.cell.2011.05.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Young RA. Control of the embryonic stem cell state. Cell. 2011;144:940–954. doi: 10.1016/j.cell.2011.01.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Yuan P, et al. Eset partners with Oct4 to restrict extraembryonic trophoblast lineage potential in embryonic stem cells. Genes Dev. 2009;23:2507–2520. doi: 10.1101/gad.1831909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Khan IM, et al. The SON gene encodes a conserved DNA binding protein mapping to human chromosome 21. Ann. Hum. Genet. 1994;58:25–34. doi: 10.1111/j.1469-1809.1994.tb00723.x. [DOI] [PubMed] [Google Scholar]
  • 16.Sun CT, et al. Transcription repression of human hepatitis B virus genes by negative regulatory element-binding protein/SON. J. Biol. Chem. 2001;276:24059–24067. doi: 10.1074/jbc.M101330200. [DOI] [PubMed] [Google Scholar]
  • 17.Wynn SL, et al. Organization and conservation of the GART/SON/DONSON locus in mouse and human genomes. Genomics. 2000;68:57–62. doi: 10.1006/geno.2000.6254. [DOI] [PubMed] [Google Scholar]
  • 18.Saitoh N, et al. Proteomic analysis of interchromatin granule clusters. Mol. Biol. Cell. 2004;15:3876–3890. doi: 10.1091/mbc.E04-03-0253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Huen MS, et al. SON is a spliceosome-associated factor required for mitotic progression. Cell Cycle. 2010;9:2679–2685. doi: 10.4161/cc.9.13.12151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Sharma A, Takata H, Shibahara K, Bubulya A, Bubulya PA. Son is essential for nuclear speckle organization and cell cycle progression. Mol. Biol. Cell. 2010;21:650–663. doi: 10.1091/mbc.E09-02-0126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Ahn EY, et al. SON controls cell-cycle progression by coordinated regulation of RNA splicing. Mol. Cell. 2011;42:185–198. doi: 10.1016/j.molcel.2011.03.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Sharma A, et al. Son maintains accurate splicing for a subset of human pre-mRNAs. J. Cell Sci. 2011;124:4286–4298. doi: 10.1242/jcs.092239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Ruiz S, et al. A high proliferation rate is required for cell reprogramming and maintenance of human embryonic stem cell identity. Curr. Biol. 2011;21:45–52. doi: 10.1016/j.cub.2010.11.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Sorger PK, Dobles M, Tournebize R, Hyman AA. Coupling cell division and cell death to microtubule dynamics. Curr. Opin. Cell Biol. 1997;9:807–814. doi: 10.1016/s0955-0674(97)80081-6. [DOI] [PubMed] [Google Scholar]
  • 25.Pagano M, Pepperkok R, Verde F, Ansorge W, Draetta G. Cyclin A is required at two points in the human cell cycle. EMBO J. 1992;11:961–971. doi: 10.1002/j.1460-2075.1992.tb05135.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Watanabe K, et al. A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat. Biotechnol. 2007;25:681–686. doi: 10.1038/nbt1310. [DOI] [PubMed] [Google Scholar]
  • 27.Rogakou EP, Nieves-Neira W, Boon C, Pommier Y, Bonner WM. Initiation of DNA fragmentation during apoptosis induces phosphorylation of H2AX histone at serine 139. J. Biol. Chem. 2000;275:9390–9395. doi: 10.1074/jbc.275.13.9390. [DOI] [PubMed] [Google Scholar]
  • 28.Yeo G, Burge CB. Maximum entropy modeling of short sequence motifs with applications to RNA splicing signals. J. Comput. Biol. 2004;11:377–394. doi: 10.1089/1066527041410418. [DOI] [PubMed] [Google Scholar]
  • 29.Dewey CN, Rogozin IB, Koonin EV. Compensatory relationship between splice sites and exonic splicing signals depending on the length of vertebrate introns. BMC Genom. 2006;7:311–319. doi: 10.1186/1471-2164-7-311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Amit M, et al. Differential GC content between exons and introns establishes distinct strategies of splice-site recognition. Cell Rep. 2012;1:543–556. doi: 10.1016/j.celrep.2012.03.013. [DOI] [PubMed] [Google Scholar]
  • 31.Hay DC, Sutherland L, Clark J, Burdon T. Oct-4 knockdown induces similar patterns of endoderm and trophoblast differentiation markers in human and mouse embryonic stem cells. Stem Cells. 2004;22:225–235. doi: 10.1634/stemcells.22-2-225. [DOI] [PubMed] [Google Scholar]
  • 32.Chang YF, Imam JS, Wilkinson MF. The nonsense-mediated decay RNA surveillance pathway. Annu. Rev. Biochem. 2007;76:51–74. doi: 10.1146/annurev.biochem.76.050106.093909. [DOI] [PubMed] [Google Scholar]
  • 33.Ishigaki Y, Li X, Serin G, Maquat LE. Evidence for a pioneer round of mRNA translation: mRNAs subject to nonsense-mediated decay in mammalian cells are bound by CBP80 and CBP20. Cell. 2001;106:607–617. doi: 10.1016/s0092-8674(01)00475-5. [DOI] [PubMed] [Google Scholar]
  • 34.Maquat LE. Nonsense-mediated mRNA decay: splicing, translation and mRNP dynamics. Nat. Rev. Mol. Cell Biol. 2004;5:89–99. doi: 10.1038/nrm1310. [DOI] [PubMed] [Google Scholar]
  • 35.Chan YS, et al. A PRC2-dependent repressive role of PRDM14 in human embryonic stem cells and induced pluripotent stem cell reprogramming. Stem Cells. 2012;31:682–692. doi: 10.1002/stem.1307. [DOI] [PubMed] [Google Scholar]
  • 36.Fedor MJ. Alternative splicing minireview series: combinatorial control facilitates splicing regulation of gene expression and enhances genome diversity. J. Biol. Chem. 2008;283:1209–1210. doi: 10.1074/jbc.R700046200. [DOI] [PubMed] [Google Scholar]
  • 37.Faustino NA, Cooper TA. Pre-mRNA splicing and human disease. Genes Dev. 2003;17:419–437. doi: 10.1101/gad.1048803. [DOI] [PubMed] [Google Scholar]
  • 38.Yeo GW, et al. Alternative splicing events identified in human embryonic stem cells and neural progenitors. PLoS Comput. Biol. 2007;3:1951–1967. doi: 10.1371/journal.pcbi.0030196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Salomonis N, et al. Alternative splicing in the differentiation of human embryonic stem cells into cardiac precursors. PLoS Comput. Biol. 2009;5:e1000553. doi: 10.1371/journal.pcbi.1000553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Salomonis N, et al. Alternative splicing regulates mouse embryonic stem cell pluripotency and differentiation. Proc. Natl Acad. Sci. USA. 2010;107:10514–10519. doi: 10.1073/pnas.0912260107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Yeo GW, et al. An RNA code for the FOX2 splicing regulator revealed by mapping RNA-protein interactions in stem cells. Nat. Struct. Mol. Biol. 2009;16:130–137. doi: 10.1038/nsmb.1545. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Gabut M, et al. An alternative splicing switch regulates embryonic stem cell pluripotency and reprogramming. Cell. 2011;147:132–146. doi: 10.1016/j.cell.2011.08.023. [DOI] [PubMed] [Google Scholar]
  • 43.Han H, et al. MBNL proteins repress ES-cell-specific alternative splicing and reprogramming. Nature. 2013;498:241–245. doi: 10.1038/nature12270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Boyer LA, et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell. 2005;122:947–956. doi: 10.1016/j.cell.2005.08.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Kagey MH, et al. Mediator and cohesin connect gene expression and chromatin architecture. Nature. 2010;467:430–435. doi: 10.1038/nature09380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Le Cam L, Lacroix M, Ciemerych MA, Sardet C, Sicinski P. The E4F protein is required for mitotic progression during embryonic cell cycles. Mol. Cell Biol. 2004;24:6467–6475. doi: 10.1128/MCB.24.14.6467-6475.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.ENCODE Project Consortium. Myers RM, et al. A user's guide to the encyclopedia of DNA elements (ENCODE) PLoS Biol. 2011;9:e1001046. doi: 10.1371/journal.pbio.1001046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Xu C, et al. Feeder-free growth of undifferentiated human embryonic stem cells. Nat. Biotechnol. 2001;19:971–974. doi: 10.1038/nbt1001-971. [DOI] [PubMed] [Google Scholar]
  • 49.Paillusson A, Hirschi N, Vallan C, Azzalin CM, Muhlemann O. A GFP-based reporter system to monitor nonsense-mediated mRNA decay. Nucleic Acids Res. 2005;33:e54. doi: 10.1093/nar/gni052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Wittmann J, Hol EM, Jack HM. hUPF2 silencing identifies physiologic substrates of mammalian nonsense-mediated mRNA decay. Mol. Cell. Biol. 2006;26:1272–1287. doi: 10.1128/MCB.26.4.1272-1287.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Izquierdo JM, Alcalde J, Carrascoso I, Reyes R, Ludena MD. Knockdown of T-cell intracellular antigens triggers cell proliferation, invasion and tumour growth. Biochem. J. 2011;435:337–344. doi: 10.1042/BJ20101030. [DOI] [PubMed] [Google Scholar]
  • 52.Trapnell C, Pachter L, Salzberg SL. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics. 2009;25:1105–1111. doi: 10.1093/bioinformatics/btp120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Trapnell C, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat. Biotechnol. 2010;28:511–515. doi: 10.1038/nbt.1621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11:R106. doi: 10.1186/gb-2010-11-10-r106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Zhao J, et al. Genome-wide identification of polycomb-associated RNAs by RIP-seq. Mol. Cell. 2010;40:939–953. doi: 10.1016/j.molcel.2010.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

a
NIHMS612091-supplement-a.xlsx (326KB, xlsx)
b
NIHMS612091-supplement-b.xlsx (112.6KB, xlsx)
c
NIHMS612091-supplement-c.xlsx (13KB, xlsx)
supp
NIHMS612091-supplement-supp.pdf (2.6MB, pdf)

RESOURCES