pubmed.ncbi.nlm.nih.gov

A cohesin-OCT4 complex mediates Sox enhancers to prime an early embryonic lineage - PubMed

  • ️Thu Jan 01 2015

A cohesin-OCT4 complex mediates Sox enhancers to prime an early embryonic lineage

Nesrine Abboud et al. Nat Commun. 2015.

Abstract

Short- and long-scales intra- and inter-chromosomal interactions are linked to gene transcription, but the molecular events underlying these structures and how they affect cell fate decision during embryonic development are poorly understood. One of the first embryonic cell fate decisions (that is, mesendoderm determination) is driven by the POU factor OCT4, acting in concert with the high-mobility group genes Sox-2 and Sox-17. Here we report a chromatin-remodelling mechanism and enhancer function that mediate cell fate switching. OCT4 alters the higher-order chromatin structure at both Sox-2 and Sox-17 loci. OCT4 titrates out cohesin and switches the Sox-17 enhancer from a locked (within an inter-chromosomal Sox-2 enhancer/CCCTC-binding factor CTCF/cohesin loop) to an active (within an intra-chromosomal Sox-17 promoter/enhancer/cohesin loop) state. SALL4 concomitantly mobilizes the polycomb complexes at the Soxs loci. Thus, OCT4/SALL4-driven cohesin- and polycombs-mediated changes in higher-order chromatin structure mediate instruction of early cell fate in embryonic cells.

PubMed Disclaimer

Figures

Figure 1
Figure 1. Embryonic pattern of expression of SOX-17/OCT4/SALL4

Whole-mount staining of E7.5 mouse embryos with (a) anti-SOX-17- and alexa546-conjugated secondary antibody, (b) anti-OCT4- and alexa488-conjugated secondary antibody, (c) merged image (a,b). (d) Anti-OCT4 whole-mount immunostaining of E7.5 embryos derived from a Sox17cre/+ breeder crossed with Rosa26tDTomato female. (e) the embryo proper was enzymatically dissociated and the cells were plated and fixed on coverslip. The cells were immunostained by an anti-OCT4 (upper panel) or anti-SALL4 antibody. Scale bars, 10 µm. Note that OCT4 is cytosolic in most of cells indicating a downregulation of the transcription factor. (f) Whole-mount staining of embryo using an anti-OCT4 and anti-SALL4 antibodies; a stack of images was acquired in confocal microscopy and reconstituted as a 3D image using AMIRA software. The merged image is shown in the centre. DE, definitive endoderm, VE visceral endoderm. (g) Anti-OCT4 immunostaining of E7.5 embryos derived from a MesP1cre/+ breeder crossed with Rosa26tDTomato female. Whole-mount staining. (h) The posterior region of the embryo was enzymatically dissociated and cells were plated and immediately fixed on coverslip to be stained by an anti-OCT4 antibody and a secondary anti-alexa488 secondary antibody. Arrows indicate OCT4 + cells in the MesP1 cell lineage. OCT4 is either nuclear (h1) or cytosolic (h2, h3 and h4). Scale bars, 50 µm.

Figure 2
Figure 2. OCT4 induces SALL4+/SOX-17+ mesendoderm by targeting SALL4 promoter

(a) OCT4 increase in expression gives rise to a mesendodermal cell population. OCT4-iA cDNA was nucleofected in HUESCs to increase by twofold the protein level of expression (inset: left panel: western blot anti-OCT4. Right panel: graphs representative of three gels, bands were quantified using Quantity one (Biorad) and normalized using actin as a housekeeping protein). Real-time PCR shows the fold increase or decrease in gene expression normalized to undifferentiated HUESCs ((mean ± s.e.m., from four experiments, *Student’s t-test P≤0.01). (b) OCT4-binding sites on SALL4 promoter from ChIP-on-chip assay. Inset: ChIP-PCR anti-OCT4 from mock or OCT4OE cells. The result is represented as a fold change in occupancy in mesendodermal cells versus the undifferentiated HUESCs after normalization to the input sample. (c) OCT4 occupancy in SALL4 promoter (pr) region was amplified by real-time quantitative PCR for mock or OCT4OE cells (mean ± s.e.m., from three experiments, **Student’s t-test, P≤ 0.01). (d) Oct4OE cells express both SALL4 and SOX-17. Mock (pcDNA nucleofected or Oct4OE cells were stained with anti-SALL4 and anti-SOX-17 antibodies and visualized at × 40 magnification by confocal microscopy. Scale bar, 10 µm. (e) Western blot anti-SALL4 of mock and Oct4OE cells. The histone H4 was used to normalize the blot.

Figure 3
Figure 3. SALL4 mediates the cardiogenic action of OCT4

(a) Anti-SALL4 immunofluorescence in mock (scrambled Sh-nucleofected cells) or SALL4Sh-nucleofected cells showed that SALL4 was downregulated in HUESC by nucleofection of a ShRNA. Scale bar, 50 µm. Right panel: western blot analysis of SALL4 in mock or SALL4Sh-transfected cells. (b,c) OCT4OE + scrambled Sh or OCT4OE + SALL4Sh-transfected cells were cultured on matrigel-coated dishes in the presence of FGF2- and MEF-conditioned medium for 4 days after nucleofection, to avoid interference of MEF chromatin in ChIP. (b) ChIP was performed using an anti-OCT4 antibody from mock (GFP nucleofected), OCT4OE + scrambled Sh cells or OCT4OE + SALL4Sh cells. PCR was performed using primers specific of SOX-17 or SOX-2 enhancers A and B or matching a 3′-SOX-17 R2 region (mean ± s.e.m., n = 3; **Student’s t-test, P≤0.01.). (c) Anti-SALL4 ChIP-PCR analysis of SOX-17 and SOX-2 promoters. SOX17R2 and SOX2B enhancers were monitored to check the ChIP specificity. (d) Gene expression was monitored by real-time PCR in mock, OCT4OE cells or OCT4OE + SALL4Sh-transfected cells cultured on matrigel-coated dishes in the presence of FGF2- and MEF-conditioned medium for 4 days after nucleofection. Results are expressed as fold changes in OCT4OE cells versus undifferentiated HUESCs (mean ± s.e.m., n = 3, **Student’s t-test P≤ 0.01. (e) After 4 days post nucleofection, cells were dispersed with dispase and allowed to spontaneously aggregate to form EBs. Six days later, EBs were dissociated with trypsin, then plated for 14 h on fibronectin-coated labtech plates and proteins were visualized by immunofluorescence. The immunostained cells whose nuclei were labelled with DAPI were scanned with a BioRad confocal microscope. Background was set using undifferentiated HUESC stained with the same antibodies. Scale bar, 20 µm. Cells positive for the transcription factors were scored using DAPI as a reference to segment cells using Image J (right graph). The right graph shows statistical data (mean ± s.e.m., from three experiments). **Student’s t-test, P≤0.01. DAPI, 4,6-diamidino-2-phenylindole

Figure 4
Figure 4. SALL4 recruits the polycomb complexes PRC1 and PRC2 to the Sox-2 locus and the PRC1 complex to the Sox-17 locus

Sequential ChlP-PCR anti-SUZ12 or anti-BMI1 and then anti-SALL4 were performed from chromatin of control ZHBTc6 (+ tetracyclin) ESCs (CTRL) or OCT4-overexpressing (ZHBTc6 – tetracyclin) mesendodermal cells (Oct4OE). Both Sox-2C (a) and Sox-17 (b) enhancers and promoters were interrogated. Sox2B and Sox17R2 regions were used as negative controls. (c) ChIP anti-SUZ12 (left panel) and anti-BMI (right panel) were performed from chromatin of control ZHBTc6 (+ tetracyclin) ESCs (CTRL) or OCT4-overexpressing (ZHBTc6 – tetracyclin) mesendodermal cells (Oct4OE). Both Sox-2C and Sox-17 enhancers were interrogated. Data are expressed as mean ± s.e.m., from three experiments, *Student’s t-test, P≤0.01.

Figure 5
Figure 5. RAD21- and CTCF-mediated OCT4-induced changes in the 3D configuration of Sox-17 locus

(a) NIBPL was downregulated by a siRNA. Anti-NIPBL immunostaining of ZHBTc6 cells before differentiation within embryoid bodies of OCT4OE induced mesendodermal cells transfected with a scrambled siRNA or the NIPBL siRNA. Inset: OCT4 and NANOG expressions were found not affected in NIPBL-downregulated cells. Scale bar, 50 µm. (b) Anti-OCT4 ChIP was performed in control ESCs (CTRL) and OCT4-overexpressing mesendodermal cells (OCT4OE) in the presence or absence of NIPBL. Both the Sox-2C and the Sox-17 enhancers were interrogated. (c) Anti-CTCF and anti-RAD21 ChIP-PCR were performed from chromatin of control ZHBTc6 (+ tetracyclin) ESCs (CTRL) or OCT4-overexpressing (ZHBTc6-tetracyclin) mesendodermal cells (Oct4OE). (d) Sequential ChIP-PCR anti-OCT4 and then anti-RAD21 were performed from chromatin of control ZHBTc6 ESCs (CTRL) or OCT4-overexpressing mesendodermal cells (OCT4OE). Both Sox-2C (blue bars; c,d) and Sox-17 (green bars; c,d) enhancers were interrogated in real-time PCR. (e) Anti-OCT4 immunoprecipitation of whole cell extract from control (Ctrl) or OCT4-overexpressing mesendodermal cells (Oct4OE). The blot was incubated with both anti-OCT4 and anti-RAD21 antibodies. (f) Anti-OCT4 ChIP-PCR was performed from chromatin of control ZHBTc6 ESCs (Ctrl) or OCT4-overexpressing mesendodermal cells (Oct4OE). Sox-17 promoter was interrogated. Data are representative of three to four experiments (mean ± s.e.m). *Student’s t-test, P≤0.01.

Figure 6
Figure 6. OCT4 favours an enhancer/promoter loop within the Sox-17 locus and breaks a Sox-2 enhancer/promoter loop

3C experiments were performed on digested and religated chromatin from undifferentiated ZHBTc6 cells or OCT4-induced mesendodermal cells. Primers spanning within the Sox-2 and Sox-17 loci and numbered as in Table 2 were used in PCR. Control primers such as in exons were used as controls (cartoons, a-c). The graphs show the frequency of interactions within Sox loci on chromosomes 1 (a) or 3 (b) in both control (ZHBTc6 non-induced cells) and Oct4OE (ZHBTc6 induced) mesendodermal cells. Interactions frequencies were normalized using the GAPDH locus whose 3D configuration was assumed not to change, as GAPDH is expressed to the same extent in both pluripotent ESCs and OCT4OE (ZHBTc6 induced) mesendodermal cells. The relative interaction frequency value is thus set to one for the interaction measured between the control GAPDH fragments. Data are representative of four experiments (mean ± s.e.m.).

Figure 7
Figure 7. OCT4 breaks Sox-2/Sox-17 enhancer inter-chromosomal interaction

(a) 3C experiments were performed on digested and religated chromatin from undifferentiated ZHBTc6 cells or OCT4-induced mesendodermal cells. Primers spanning within the Sox-2 and Sox-17 loci and numbered as in Table 2 were used in PCR. Control primers, such as in exons, were used as controls (cartoons, a–c). Interaction frequencies were normalized using the GAPDH locus. The graphs show the frequency of interactions between chromosomes (Chr) 1 and 3. The gels show the specificity of the amplicons of the peaks of interaction; the sequence of amplicons was checked by sequencing. The relative interaction frequency value is set to one for the interaction measured between the control GAPDH fragments. Data are representative of four experiments (mean ± s.e.m.). (b) Illustration of FISH experiments using a Sox2alexa488 and Sox17alexa568 probes showing the interaction between chromosomes 1 and 3 for both alleles (the nucleus is representative of 116 imaged nuclei of pluripotent stem cells; upper panels). Alleles of chromosomes 1 and 3 were separated in mesendodermal OCT4-OE cells (the nucleus is representative of 110 nuclei; lower panels). The red, green channels and the merged image are shown. Scale bar, 2 µm. The probe hybridized region is magnified in the inset of the merged image. The right graph in the inset indicates the averaged percentage of cells featuring a Sox-2/Sox-17 interaction from three experiments (*Student’s t-test, P≤0.01)

Figure 8
Figure 8. NIPBL-mediated loading of the cohesin complex is required for the cardiogenic effect of OCT4

NIBPL was downregulated by a siRNA. Cardiogenesis was monitored by the gene expression of cardiac-enriched genes (a) and (b) by immunostaining of cardiac transcription factors Nkx2.5 and Isl1 and a sarcomeric protein (actinin) in day 12 EBs. Scale bar indicates 50 µm. (c) Ectodermal and endodermal genes were monitored. Data are representative of three experiments (mean ± s.e.m., *Student’s t-test, P≤0.01).

Figure 9
Figure 9. Mechanistic model of OCT4/SALL4 changes in higher-order chromatin structure

Cartoon illustrating the inter- and intra-chromosomal interactions within the Sox-2 and Sox-17 loci, as well as the location of bivalent epigenetic marks and polycomb complexes in pluripotent (inner cell mass (ICM)-like cells) and mesendodermal cells.

Similar articles

Cited by

References

    1. Stent GS. Developmental cell lineage. Int. J. Dev. Biol. 1998;42:237–241. - PubMed
    1. Benyajati C, Worcel A. Isolation, characterization, and structure of the folded interphase genome of Drosophila melanogaster. Cell. 1976;9:393–407. - PubMed
    1. McNairn AJ, Gerton JL. The chromosome glue gets a little stickier. Trends Genet. 2008;24:382–389. - PubMed
    1. Parelho V, et al. Cohesins functionally associate with CTCF on mammalian chromosome arms. Cell. 2008;132:422–433. - PubMed
    1. Wendt KS, et al. Cohesin mediates transcriptional insulation by CCCTC-binding factor. Nature. 2008;451:796–801. - PubMed

Publication types

MeSH terms

Substances