Figures and data
mESCs require Stag3 to exit pluripotency.
a) Relative mRNA expression of the three Stag paralogs by qRT-PCR in naïve (2i) mESCs. Data is from 15-20 independent biological replicates. The central line represents the median. Asterisks indicate a statistically significant difference as assessed using two-tailed t-test. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001; ns, not significant.
b) Whole-cell protein extracts (WCE) from mESCs, at several timepoints during EpiLC differentiation, in mouse embryonic fibroblast cells (MEF) and in Testis tissue were analysed by immunoblot (IB) for levels of STAG1, STAG2, STAG3, SMC3 and NANOG. alpha-TUBULIN (aTUB) serves as a loading control.
c) Left, relative mRNA expression of Stag3 by qRT-PCR in untreated (UT) mESCs or upon treatment with siLuc or siStag3. Data is from 20 independent biological replicates. Right, quantification of STAG3 protein levels in naive mESCs treated with siLuc and siStag3 assessed by ImageJ relative to the signal in UT cells. Data is from 5 independent biological replicates. Quantifications and statistical analysis as above.
d) Left, relative mRNA expression of Nanog by qRT-PCR in UT, siLuc and siStag3 treated mESCs. Data is from 12 independent biological replicates. Right, quantification of NANOG protein levels in naive mESCs treated as above assessed by FACS analysis of NANOG-GFP GeoMean. Data is from 3 independent biological replicates.
e) MA plot of RNA-seq obtained from 3 biological replicates of siStag3 and siLuc treated naive mESCs. Differential expression analysis was performed to plot data is as log2 fold change in KD conditions relative to siLuc controls. Labelled genes represent statistically significant (p<0.05) fold change based on students t-test.
f) Left, relative expression of Dppa3 mRNA by qRT-PCR in UT mESCs or upon treatment with siLuc or siStag3. Data is from 12 independent biological replicates. Right, quantification of DPPA3 protein levels in naive mESCs treated as above assessed by FACS analysis of DPPA3-GFP GeoMean relative to UT. Data is from 5 independent biological replicates.
g) Enrichment score (ES) plots from gene set enrichment analysis (GSEA) using curated naive or primed pluripotency gene sets (see Methods). Negative and positive normalised enrichment scores (NES) point to the gene set being over-represented in the top-most down-or upregulated genes in Stag3 KD mESCs, respectively. Vertical bars refer to individual genes in the gene set and their position reflects the contribution of each gene to the NES.
h) Area occupied by AP colonies relative to total colony area in mESCs treated with siLuc and siStag3 from 3 independent biological replicates where n > 50 colonies/condition were counted.
Loss of Stag3 prevents commitment to PGCLC.
a) GSEA of Stag3 KD RNA-seq data reveals a negative normalised enrichment score of -1.50 (adj p= 0.044) for the Spermatogenesis signature (see Methods).
b) Relative mRNA expression of PGC regulators in siLuc and siStag3 mESCs. Data is shown as fold-change from UT and is from 9 independent biological replicates. Quantifications and statistical analysis as before.
c) Representative brightfield (top row), epi-fluorescent (middle row) and FACS profiles of PRDM1-GFP mESCs at select timepoints during in vitro differentiation into PGCLCs. Scale bar is 200um.
d) Relative mRNA expression of select mESC (Nanog), EpiLC (Dnmt3b), PGCLC (Prdm1, Dppa3) and cohesin (Stag1, 2, 3 and Smc1b) genes from various timepoints during PGCLC differentiation in vitro. EB GFP+ cells were FACS sorted at the timepoints shown and collected for qRT-PCR. Gene expression is relative to levels in mESCs. Data is from 3 independent biological replicates.
e) (Top) Representative brightfield and epi-fluorescent images of DPPA3-GFP mESCs treated with siLuc or siStag3 at three timepoints of embryoid body (EB) differentiation. (Bottom) Representative FACS profile of DPPA3-GFP d6 EB treated with siLuc or siStag3. Scale bar is 200μm.
f) Left, quantification of DPPA3-GFP GeoMean or right, the % of DPPA3-GFP+ cells assessed by FACS (relative to UT mESCs) at different stages of PGCLC differentiation and upon siRNA treatment. Data is from 4 independent biological replicates. Quantifications and statistical analysis as before.
g) Relative mRNA expression of select genes in mESCs or d6 EB GFP+ cells in siLuc and siStag3 conditions. Data is shown as fold-change from UT and is from 3 independent biological replicates. Quantifications and statistical analysis as above.
STAG3 is localised to the cytoplasm in mESCs.
a) Immunoblot analysis of STAG3 levels in whole cell extract (WCE), cytoplasmic (C.plsm), nucleoplasmic (N.plsm) and chromatin (Chrom.) fractions in mESC (ES) and 48hr EpiLCs (Epi). Inset, over-exposed IB of the chromatin fraction to show the STAG3 bandshift (compare red and blue arrows). alpha-TUBULIN (aTUB) and H3 serve as fractionation and loading controls.
b) Immunoblot analysis of STAG3 levels in fractionated lysates from FACS-sorted G1 or G2 mESC populations. aTUB and H3 serve as fractionation and loading controls as above.
c) Representative confocal images of endogenous LAMIN B (demarcating the nuclear membrane) and STAG3 assessed by co-IF in mESCs and counterstained with DAPI. Shown are four independent cells with different STAG3 signal profiles in the cytoplasm. Scale bars, 3µm.
d)-e) Representative confocal images of d) aTUB or e) gamma-(g-)TUBULIN (gTUB) with STAG3 and LAMIN B and counterstained with DAPI by co-IF in mESCs. Arrows indicate notable regions of overlap with STAG3 in the cytoplasm. NB. the STAG3 signal at the aTUB ‘bridge’ shown in the last cell from d). Scale bars, 3µm.
f) ImageJ quantification of the MFI of endogenous (eSTAG3) in wildtype (WT) mESC in either the nucleus (light grey) or at gTUB foci (orange). Quantifications and statistical analysis are as shown previously. Data are from n > 100 independent cells/condition in 3 biological replicates. AU, arbitrary units.
g) Immunoblot analysis of STAG3, gTUB and aTUB levels in WCE of UT mESCs or after treatment with siLuc or siStag3. Shown also is the % reduction in gTUB levels relative to the signal in UT cells, as assessed by quantification of the bands using ImageJ.
h) ImageJ quantification of the MFI of left, eSTAG3; middle, gTUB and right, gTUB foci size in WT mESC treated with siLuc (grey) or siStag3 (blue). Quantifications and statistical analysis were done as above. Data are from n > 100 independent cells/condition in 2 biological replicates. AU, arbitrary units.
Characterisation of the STAG3 protein interaction network in mESCs.
a) Scatter plot displaying the log2 protein group intensity across three biological replicates for the 337 proteins identified in STAG3-v5 IP-MS data produced from v5-TRAP in WT and Stag3-v5 mESCs. Coloured dots represent STAG3-v5 specific and enriched interactors (148 proteins in total). Blue dots/labels represent proteins which were uniquely detected in the v5 IPs, representing STAG3-v5-specific interactors. Red dots/labels represent enriched STAG3-v5 interactors with an abundance increase of at least 2-fold and p<0.05 in STAG3-v5 compared to WT. NB. For clarity we have not labelled all of the ribosome proteins here, please refer to Figure S4 and Table S2.
b) -Log10 transformed adjusted p-value (FDR) for enrichment of centrosome interactome data from O’Neill et al. and Carden et al., with the STAG3-V5 interactome. ‘PanCent’ comprises centrosome-interacting proteins common to at least two of the four human cancer cell lines tested in Carden et al.
c) Most enriched GO terms and their corresponding FDR values, arranged by category, for the 147 STAG3-v5 interactors (STAG3 was not included as an interactor). For full GO term enrichment, see Table S2.
d) Simplified STAG3-v5 interaction network of protein–protein interactions identified in mESCs using STRING. Node colours describe the major enriched categories with squares denoting RBPs. Proteins are grouped according to Cellular Compartment enriched terms. All proteins present in both the 147 STAG3-v5 PPI and each enriched GO term are indicated here.
STAG3 mediates post-transcriptional regulation of Dppa3 in mESCs and destabilises TNRC6C.
a) Relative mRNA expression of Stag3 (top) and Dppa3 (bottom) in siLuc and siStag3 mESCs treated with cyclohexamide (CHX) for 4 or 8 hours to inhibit translation from three independent replicates. Significance refers to the average expression at each timepoint using t-test as before.
b) Scatter plot from the TMT analysis displaying the statistical significance (-log10 p-value) versus the protein abundance difference from siLuc and siStag3 mESCs. Data was derived from three biological replicates. Vertical dashed lines represent changes of 1.5-fold (green dots). Horizontal dashed line represents a p-value of 0.05. Blue dots, STAG3-V5 interactors from Fig 4a.
c) Immunoblot analysis of STAG3 and TNRC6C levels in cytoplasmic (C.plsm), nucleoplasmic (N.plsm) and chromatin (Chrom.) fractions of mESCs treated with siLuc and siStag3. aTUB and H3 serve as fractionation and loading controls.
d) Representative confocal images of coIF of endogenous STAG3, LAMIN B and TNRC6C in WT mESCs treated with siLuc and siStag3. Arrows indicate notable regions of overlap.
e) (Above) Representative confocal images of coIF of STAG3-v5, LAMIN B and TNRC6C under siLuc and siStag3 conditions. (Below) ImageJ quantification of the MFI of left, STAG3-v5 and right, TNRC6C in siRNA-treated mESCs. Quantifications and statistical analysis were done as above. Data are from n > 100 independent cells/condition in 2 biological replicates. AU, arbitrary units.
f) STRING analysis of the STAG3-V5 interactors enriched within the P-body interactome data from Hubstenberger et al. -log10 transformed adjusted p-values (FDR) are in the text. Node colours describe the major enriched categories of the enriched proteins.
Antibodies used in this study
