Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jul 21;12(7):e0180922.
doi: 10.1371/journal.pone.0180922. eCollection 2017.

Direct Targets of pSTAT5 Signalling in Erythropoiesis

Free PMC article

Direct Targets of pSTAT5 Signalling in Erythropoiesis

Kevin R Gillinder et al. PLoS One. .
Free PMC article


Erythropoietin (EPO) acts through the dimeric erythropoietin receptor to stimulate proliferation, survival, differentiation and enucleation of erythroid progenitor cells. We undertook two complimentary approaches to find EPO-dependent pSTAT5 target genes in murine erythroid cells: RNA-seq of newly transcribed (4sU-labelled) RNA, and ChIP-seq for pSTAT5 30 minutes after EPO stimulation. We found 302 pSTAT5-occupied sites: ~15% of these reside in promoters while the rest reside within intronic enhancers or intergenic regions, some >100kb from the nearest TSS. The majority of pSTAT5 peaks contain a central palindromic GAS element, TTCYXRGAA. There was significant enrichment for GATA motifs and CACCC-box motifs within the neighbourhood of pSTAT5-bound peaks, and GATA1 and/or KLF1 co-occupancy at many sites. Using 4sU-RNA-seq we determined the EPO-induced transcriptome and validated differentially expressed genes using dynamic CAGE data and qRT-PCR. We identified known direct pSTAT5 target genes such as Bcl2l1, Pim1 and Cish, and many new targets likely to be involved in driving erythroid cell differentiation including those involved in mRNA splicing (Rbm25), epigenetic regulation (Suv420h2), and EpoR turnover (Clint1/EpsinR). Some of these new EpoR-JAK2-pSTAT5 target genes could be used as biomarkers for monitoring disease activity in polycythaemia vera, and for monitoring responses to JAK inhibitors.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Fig 1
Fig 1. pSTAT5 binds promoters and enhancers of key erythroid genes in concert with GATA1 and KLF1.
(A) 302 peaks were annotated with respect to the nearest TSS as defined by RefSeq. Only ~23% of peaks reside within gene promoters (1 kb upstream or 100 nt downstream of a TSS). The majority are intronic or distal (within 50 kb of a gene body) consistent with enhancer occupancy. (B) De novo motif discovery on all 302 peaks using MEME (see Methods) identified highly significant enrichment of a (i) palindromic GAS element (TTCYMRGAA), a (ii) GATA binding site (WGATAR), and a (iii) KLF binding site (CCMCRCCCN). No other significantly enriched motifs were found. (C) Venn diagram and of erythroid co-occupancy between pSTAT5, KLF1 and GATA1 at key erythroid enhancers and promoters. KLF1 ChIP-seq was generated in K1-ER cells and GATA1 ChIP-seq was generated in G1-ER4 cells after induction with tamoxifen. Co-occupancy defined as ChIP summits within 500 bp. (D) Density heat-map of ChIP signal centred on pSTAT5 peak summits from GATA1 and KLF1 as above (C) and additional STAT5 ChIP in other murine tissues. The Y-axis represents individual peak regions, and the X-axis represents 10 kb surrounding the summit. Read intensities were normalized to number of reads in dataset and hierarchically clustered according to intensity within 500 bp of peak centre.
Fig 2
Fig 2. EPO-induced changes in erythroid gene expression.
Statistically significant DEGs are plotted as fold change (EPO +/-) versus mean FPKM between conditions. Yellow points represent genes where pSTAT5 binds within 100 kb of TSS (presumably enhancers) and red points are DEGs with promoter bound pSTAT5. Grey points represent genes with no significant pSTAT5 binding site within 100 kb.
Fig 3
Fig 3. Immediate direct transcriptional targets of EPOR-pSTAT5 signalling.
(A) ChIP-seq and 4sU-RNA-seq in erythroid cells following 30 mins of EPO induction across the Cish gene. Read density per million mapped reads (y-axis) for pSTAT5 (ochre), KLF1 (orange) and GATA1 (brown) ChIP. Coloured bars above each track represent peak summits as called by MACS2. Read density profiles (y-axis) for 4sU-RNA-seq from J2E cells (blue) and J2E cells post EPO-induction (red) are displayed for both forward (+ values) and reverse (- values) strands. A schematic of gene structure and alternative transcripts from RefSeq is shown in black with scale bar. The green bar represents the region selected for CAGE tag counts shown in (C). (B) qRT-PCR for Cish primary pre-spliced transcripts and processed mRNA over a 24-hour time course following EPO stimulation. (C) CAGE tag counts expressed as relative log expression (RLE) over the indicated pSTAT5-occupied promoter (green bar in panel A) from 0 to 24 hours post EPO stimulation. ChIP-seq and 4sU-RNA-seq across the Bcl2l1 (D) and Suv420h2 (G) genes. Overall design and colour coding of tracks is the same as for (A). Purple bars indicate pSTAT5-occupied intronic enhancers. Dynamic EPO-induced CAGE tags over this region are shown in S4 Fig. qRT-PCR for Bcl2l1 (E) and Suv420h2 (H) primary pre-spliced transcripts and processed mRNA over a 24-hour time course following EPO stimulation as in (B). CAGE tag counts for Bcl2l1 (E) and Suv420h2 (I) over promoters (dark green) and alternative promoter (light green) from 0 to 24 hours post EPO stimulation as in (C).
Fig 4
Fig 4. Delayed pSTAT5-dependent and independent EPO-induced gene expression.
ChIP-seq and 4sU-RNA-seq in erythroid cells following 30 mins of EPO induction across the Gypc (A) and Thbs1 (D) genes. The overall design and colour coding of tracks is the same as for Fig 3. qRT-PCR and CAGE tag counts as in Fig 3B and 3C for Gypc and Fig 3E and 3F for Thbs1 respectively. Bidirectional CAGE tags over the intronic enhancer (purple bar in panel A) are shown in S4C Fig. Gypc provides an example of delayed transcriptional activation by pSTAT5, whereas Thbs1 appears to be activated independently of pSTAT5.
Fig 5
Fig 5. Erythroid genes and biological pathways directly regulated by pSTAT5.
(A) Clint1, also known as EpsinR, is a direct target gene of pSTAT5 (see Table 1). Clint1 directly interacts with Clathrin to facilitate clathrin-mediated endocytosis (CME) of the EpoR and associated proteins. The EpoR can either be recycled from early endosomes to the cell surface for re-use, or degraded via late endosomes, multi-vesicular bodies and eventually lysosomes. (B) The Bcl2l1 gene encodes Bcl-x which is essential for terminal erythroid differentiation. pSTAT5 binds with KLF1 and GATA1 to upregulate its expression in response to EPO. pSTAT5 also binds the Rmb25 promoter with GATA1 to drive expression. The encoded protein, RBM25, is an RNA-binding protein which can bind into the second exon of the Bcl2l1 RNA via a CGGGCA element to induce preferential splicing to generate Bcl-x(S) isoform in preference to the Bcl-x(L) isoform. The former isoform is pro-apoptotic in some contexts but may play an independent role in erythroid maturation and enucleation. (C) The Suv420h2 gene is a direct target of pSTAT5 via an intronic enhancer. SUV4-20h2 is a histone methyltransferase which recognises H4K20me1 and adds two additional methyl groups. H4K20 is first methylated by Setd8/PR-Set7, a methyltransferase that is essential for erythropoiesis. H4K20me3 inhibits access to H4K16 by the MSL acetylase machinery. Acetylation at H4K16 is associated with active gene transcription whereas H4K20me2/3 is associated with transcriptional pausing. So, upregulation of Suv420h2 by EPO-pSTAT5 is likely to promote global pausing of erythroid gene transcription and reduced production of RNA, a feature of terminal erythroid cell maturation. H4K20me3 has also been associated with cell cycle arrest and chromatin compaction, two hallmarks of erythroid cell maturation.

Similar articles

See all similar articles

Cited by 11 articles

See all "Cited by" articles


    1. McMullin MF. Congenital erythrocytosis. Int J Lab Hematol. 2016;38 Suppl 1: 59–65. doi: 10.1111/ijlh.12506 - DOI - PubMed
    1. Kralovics R, Passamonti F, Buser AS, Teo S-S, Tiedt R, Passweg JR, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med. Massachusetts Medical Society; 2005;352: 1779–1790. doi: 10.1056/NEJMoa051113 - DOI - PubMed
    1. Baxter EJ, Scott LM, Campbell PJ, East C, Fourouclas N, Swanton S, et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005;365: 1054–1061. doi: 10.1016/S0140-6736(05)71142-9 - DOI - PubMed
    1. Wu H, Liu X, Jaenisch R, Lodish HF. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell. 1995;83: 59–67. - PubMed
    1. Kieran MW, Perkins AC, Orkin SH, Zon LI. Thrombopoietin rescues in vitro erythroid colony formation from mouse embryos lacking the erythropoietin receptor. Proc Natl Acad Sci USA. 1996;93: 9126–9131. - PMC - PubMed

Grant support

The work was supported by core grant funding to A.C.P. from Mater Research. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.