Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Nov;18(11):1127-1138.
doi: 10.1038/ncb3424. Epub 2016 Oct 17.

Regulation of Transcriptional Elongation in Pluripotency and Cell Differentiation by the PHD-finger Protein Phf5a

Affiliations
Free PMC article

Regulation of Transcriptional Elongation in Pluripotency and Cell Differentiation by the PHD-finger Protein Phf5a

Alexandros Strikoudis et al. Nat Cell Biol. .
Free PMC article

Abstract

Pluripotent embryonic stem cells (ESCs) self-renew or differentiate into all tissues of the developing embryo and cell-specification factors are necessary to balance gene expression. Here we delineate the function of the PHD-finger protein 5a (Phf5a) in ESC self-renewal and ascribe its role in regulating pluripotency, cellular reprogramming and myoblast specification. We demonstrate that Phf5a is essential for maintaining pluripotency, since depleted ESCs exhibit hallmarks of differentiation. Mechanistically, we attribute Phf5a function to the stabilization of the Paf1 transcriptional complex and control of RNA polymerase II elongation on pluripotency loci. Apart from an ESC-specific factor, we demonstrate that Phf5a controls differentiation of adult myoblasts. Our findings suggest a potent mode of regulation by Phf5a in stem cells, which directs their transcriptional programme, ultimately regulating maintenance of pluripotency and cellular reprogramming.

Figures

Figure 1
Figure 1. Phf5a is required for maintenance of ESC self-renewal
(a) Western blot analysis of Phf5a and Nanog, Oct4 proteins during ESC differentiation (see Supplementary Figure 7). (b) Histogram FACS plots representing loss of GFP fluorescence in Nanog-GFP transcriptional reporter ESCs following knockdown with shControl or shPhf5a respectively. (c) Alkaline phosphatase (AP) staining of ESCs following knockdown with shControl or shPhf5a (2 different hairpins), respectively. Scale bars, 100 µm. (d) Heatmap of Affymetrix microarrays for differentially expressed genes of Nanog-GFP ESCs following knockdown with shControl or shPhf5a respectively. Red: upregulated genes, blue: downregulated genes. Q-value<0.05 Fold change (log2)>1.5. (e and f) Bar graphs showing expression levels by qRT-PCR of Phf5a, pluripotency markers (e) and differentiation markers (f), respectively, following shPhf5a knockdown in ESCs. n=6 biologically independent replicates (see Supplementary Table 5). Phf5a, Nanog, Pou5f1, Sox2, Zfp42 and Nr0b1: **p=0.0001, respectively. Gata6, Gata4, Nkx2-5, Sox1, and Meox1, **p=0.0001, respectively, Brachyury: n.s: non-significant, p=0.5632, two-sided Student’s t-test, values represent the mean ± s.d.. (g) GO-Circle plot displaying gene-annotation enrichment analysis. Blue and red indicate downregulated or upregulated gene-associated GO Terms, respectively, relative to the z-score of the analysis. (h) GO-Chord plot displaying relationships between several representative downregulated and upregulated GO Terms and associated genes. Distinct categories linked to pluripotent or differentiated cells cluster separately. (i) Western blot analysis of pluripotency factors following CRISPR-Cas9 mediated Phf5a depletion in ESCs. (see Supplementary Figure 7).
Figure 2
Figure 2. Phf5a regulates ESC pluripotency and cellular reprogramming
(a and b) Comparison of mass (a) and size (b) of teratomas generated in SCID mice following injection of doxycycline-induced ESCs engineered to express shControl or shPhf5a cassettes from the Col1a1 locus. n=4 biologically independent replicates (see Supplementary Table 5). **p=0.001 two-sided Student’s t-test, values represent the mean ± s.d. (c) Comparison of Phf5a transcript levels between differentiated fibroblasts and pluripotent stem cells by qRT-PCR. n=4 biologically independent replicates (see Supplementary Table 5). ESCs: **p=0.0035, iPSCs: **p=0.0013, two-sided Student’s t-test, values represent the mean ± s.d. (d) Western blot analysis of Phf5a protein in differentiated fibroblasts or pluripotent stem cells. (see Supplementary Figure 7). (e and f) Alkaline phosphatase (AP) staining (e) and comparison of AP-positive ESC-like colony number (f), respectively, of reprogrammable OKSM MEFs on day14 post-initial doxycycline induction following shPhf5a knockdown. n=4 biologically independent replicates (see Supplementary Table 5). **p=0.001, respectively, two-sided Student’s t-test, values represent the mean ± s.d.
Figure 3
Figure 3. Phf5a physically associates with the Paf1 complex
(a) Ingenuity systems-generated pathway of Phf5a interacting proteins following purification and mass spectrometry in ESCs. Solid or dashed lines illustrate established direct or indirect interactions, respectively. (b) Validation of Phf5a interactions with the Paf1 complex in ESCs using Flag-Phf5a purification. Tagged Phf5a, Wdr61 (positive control) and GFP (negative control) were transiently expressed in engineered Tet-inducible ESC lines following addition of doxycycline. Bait proteins are tagged (marked with a star) and migrate slower than endogenous proteins (see Supplementary Figure 7). (c) Endogenous protein immunoprecipitations for Phf5a and Paf1C subunits in ESCs (see Supplementary Figure 7). (d) Paf1-complex subunits Paf1, Cdc73 and Wdr61 were cloned in HA-tag expressing vectors and subjected into in-vitro transcription and translation. Phf5a protein was expressed and purified from bacteria. In vitro binding of HA-tagged subunits and Phf5a was interrogated by a pull-down assay using HA-immunoprecipitation and western blot analysis (see Supplementary Figure 7). (e) Phf5a interacting proteins from ESCs were subjected to glycerol gradient sedimentation followed by fractionation and western blot analysis resulting in overlapping distributions of Phf5a and Paf1-complex subunits. A control analysis for GFP is shown in the lower panel. (see Supplementary Figure 7).
Figure 4
Figure 4. Phf5a controls interactions among Paf1C subunits and its silencing abrogates Paf1C recruitment on pluripotency genes in ESCs
(a) Western blot analysis of Paf1C subunit immunoprecipitations in 293T cells following knockdown with shControl or shPhf5a, respectively, showing loss of interactions between different Paf1C members upon Phf5a depletion. Two different shRNA hairpins are shown. Left Panel: Blot for Leo1; Right Panel: blot for Cdc73; IP: immunoprecipitation. IB: immunoblot. (see Supplementary Figure 7). (b) Venn diagram showing number of genes bound by individual Paf1C subunits in ESCs using ChIP-sequencing with antibodies against endogenous Leo1, Cdc73 and Paf1 proteins. (c) Heatmap representations of normalized read density for Leo1 binding in ESCs following shControl or shPhf5a silencing, respectively. (d) Venn diagrams showing the numbers of genes bound by Leo1, Cdc73 and Paf1 in ESCs in the presence or absence of Phf5a, respectively. (e) Binding profiles for genomic distribution of Leo1 peaks (upstream, promoter, coding region, 5'UTR, 3'UTR, downstream and intergenic) in ESCs, showing preferential (32%) binding within gene bodies. (f) Gene set enrichment analysis (GSEA) enrichment plots showing significant enrichment of the top Paf1 targets for genes linked to embryonic stem cell signatures. (g) Snapshots of Leo1 and Cdc73 binding on representative pluripotency gene targets (Sall4, Klf4, Zfp42, Esrrb, Sox2, Pou5f1 and Nanog) in the presence (blue) of absence (red) of Phf5a, respectively.
Figure 5
Figure 5. Phf5a controls transcriptional elongation and RNA-PolII pause-release of pluripotency genes in ESCs
(a) Western blot analysis of total PolII and Ser-2 phosphorylated RNA-PolII in ESC following shControl or shPhf5a knockdown or ESCs differentiated in the absence of LIF, respectively. (see Supplementary Figure 7). (b and c) Scatter plot representing pausing indices of downregulated (b) or upregulated genes (c), respectively, 72h following shControl or shPhf5a knockdown using GRO-seq analysis. Read density of 500bp downstream of promoters (5'density) was normalized to read density in the rest of the gene bodies (3'density). Pausing Index= 5’density/3’density. Gray: All genes; Blue: Downregulated genes; Red: Upregulated genes. (d) Box plot showing pausing index ratios after GRO-seq analysis for downregulated (blue) or upregulated (red) genes, respectively, following shPhf5a knockdown. Only downregulated genes exhibit significant promoter-proximal pausing after shPhf5a depletion. n=3 biologically independent replicates, Wilcoxon signed rank test non-parametric. (e) Box plot showing pausing index ratios after GRO-seq analysis for specific GO Terms. Blue: Downregulated and Red: Upregulated GO Term categories, respectively. Only downregulated GO Terms exhibit significant promoter-proximal pausing after shPhf5a depletion. n=3 biologically independent replicates, Wilcoxon signed rank test non-parametric. (f) Comparison of GRO-seq read density profiles of genes 72h following shControl or shPhf5a knockdown, respectively, in ESCs. RPKM: Reads Per Kilobase per Million total reads. (g) Box plot representing comparison of log2 pausing index for downregulated genes, 72h following shControl, shPhf5a, shPaf1 knockdown, or flavopiridol-treated ESCs, respectively, using GRO-seq analysis. Flavopiridol treatment is used as a positive control of pause-release block. n=3 biologically independent replicates, Wilcoxon signed rank test non-parametric. (h) Scatter plot representing RNA-PolII pausing index for Paf1C targets and pluripotency genes based on normalized Ser5 (on TSSs)/Ser2 (on gene bodies) read density ratio of RNA-PolII ChIP-Seq in ESCs following shControl of shPhf5a silencing, respectively. In box plots (d, e and g) the central mark is the median, and the edges of the box are the first and third quartiles. Whiskers extend to the most extreme non-outlier data points.
Figure 6
Figure 6. Phf5a regulates the deposition of histone marks characteristic of transcriptional elongation in pluripotency gene loci
(a) Box plots representing log2 fold change of normalized read density for H3K4me3, H3K79me2 and H3K36me3 ChIP-seq in ESCs following shControl or shPhf5a silencing. Plots represent comparisons of all expressed transcripts in ESCs with direct Paf1 targets around transcription start sites (TSSs) (H3K4me3) or gene bodies (H3K79me2 and H3K36me3). n=3 biologically independent replicates, Wilcoxon signed rank test non-parametric. (b) Normalized read density profiles around TSSs (H3K4me3) or gene bodies (H3K79me2 and H3K36me3) on Paf1C targets and pluripotency genes in ESCs in the presence (blue) or absence (red) of Phf5a. (c) Box plots showing log2 fold change H3K79me2 occupancy on gene bodies of target genes. H3K79me2 occupancy is increased in upregulated genes, however, H3K79me2 occupancy is decreased in downregulated genes compared to all expressed genes. n=3 biologically independent replicates, Wilcoxon signed rank test non-parametric. (d) Snapshots of representative H3K4me3 H3K79me2 and H3K36me3 density tracks on pluripotency genes or control loci under conditions of shControl (blue) or shPhf5a silencing (red). In box plots (a and c) the central mark is the median, and the edges of the box are the first and third quartiles. Whiskers extend to the most extreme non-outlier data points.
Figure 7
Figure 7. Phf5a loss leads to Paf1C destabilization and inhibits myogenic differentiation
(a) Western blot analysis of myoblast self-renewal and myotube differentiation markers (myosin heavy chain and Pax7, respectively) following shControl or shPhf5a knockdown (see Supplementary Figure 7). (b) Schematic of Tet-inducible Rosa26rtTACol1a1TREshRNA animals for the derivation of primary myoblasts. Addition of doxycycline drives expression of shPhf5a from the Col1a1 locus. LSL: LoxP-stop-LoxP cassette. (c and d) Myosin heavy chain (MHC) immunofluorescence on primary myotubes purified from Rosa26rtTACol1a1TREshRNA animals (c) and CRISPR-Cas9-mediated Phf5a silencing on C2C12 cells (d), respectively, depicting suppression of myoblast differentiation. Scale bars, 100 µm. (e) Western blot analysis of Phf5a on primary myotubes purified from Rosa26rtTACol1a1TREshRNA animals. Addition of Doxycyclin induces shRNA hairpin expression and the silencing of Phf5a (see Supplementary Figure 7). (f) Venn diagram of Leo1-bound genes in myoblasts and myotubes following ChIP-sequencing. (g) Genome browser tracks showing peaks of Leo1 ChIP-sequencing for representative genes in myoblasts and myotubes. Histone-1 cluster genes, Myog, Myo1c, Myom3 and Ubc are shown as examples. (h) Venn diagram of Leo1 bound genes in myotube differentiation in the presence or absence of Phf5a following ChIP-sequencing. (i) Genome browser tracks showing peaks of Leo1 ChIP-sequncing for representative genes in shControl and shPhf5a conditions, respectively. Myog, and several olfactory, taste and smell receptors and G-protein coupled receptors, ion channels and neurotransmitter receptors are shown as examples. (j) Western blot analysis of Paf1C subunit composition using immunoprecipitations in C2C12 cells differentiated for 72h following knockdown with shControl or shPhf5a, respectively. A significant loss among Paf1C subunit interactions is observed upon Phf5a silencing (see Supplementary Figure 7).

Similar articles

See all similar articles

Cited by 15 articles

See all "Cited by" articles

References

    1. Evans MJ, Kaufman MH. Establishment in culture of pluripotential cells from mouse embryos. Nature. 1981;292:154–156. - PubMed
    1. Martin GR. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proceedings of the National Academy of Sciences of the United States of America. 1981;78:7634–7638. - PMC - PubMed
    1. Orkin SH, Hochedlinger K. Chromatin connections to pluripotency and cellular reprogramming. Cell. 2011;145:835–850. - PMC - PubMed
    1. Buckley SM, et al. Regulation of pluripotency and cellular reprogramming by the ubiquitin-proteasome system. Cell Stem Cell. 2012;11:783–798. - PMC - PubMed
    1. Trappe R, et al. Identification and characterization of a novel murine multigene family containing a PHD-finger-like motif. Biochem Biophys Res Commun. 2002;293:816–826. - PubMed
Feedback