SETD5 modulates homeostasis of hematopoietic stem cells by mediating RNA Polymerase II pausing in cooperation with HCF-1

Abstract

SETD5 mutations were identified as the genetic causes of neurodevelopmental disorders. While the whole-body knockout of Setd5 in mice leads to embryonic lethality, the role of SETD5 in adult stem cell remains unexplored. Here, a critical role of Setd5 in hematopoietic stem cells (HSCs) is identified. Specific deletion of Setd5 in hematopoietic system significantly increased the number of immunophenotypic HSCs by promoting HSC proliferation. Setd5-deficient HSCs exhibited impaired long-term self-renewal capacity and multiple-lineage differentiation potentials under transplantation pressure. Transcriptome analysis of Setd5-deficient HSCs revealed a disruption of quiescence state of long-term HSCs, a cause of the exhaustion of functional HSCs. Mechanistically, SETD5 was shown to regulate HSC quiescence by mediating the release of promoter-proximal paused RNA polymerase II (Pol II) on E2F targets in cooperation with HCF-1 and PAF1 complex. Taken together, these findings reveal an essential role of SETD5 in regulating Pol II pausing-mediated maintenance of adult stem cells.

Fig. 1. Setd5 deficiency causes phenotypic HSPC expansion.

A FACS analysis of T, B, and myeloid cells frequency in PB cells; n = 5. B Relative frequency of immature cells (Lin^– and c-Kit⁺) in BM. Lin: lineage cocktail; n = 5. C The absolute cell number of HPC (LK: Lin^–c-Kit⁺Sca-1^–), CMP (Lin^–c-Kit⁺Sca1^–CD34⁺CD16/32^low), GMP (Lin^–c-Kit⁺Sca1^–CD34⁺CD16/32^high) and MEP (Lin^–c-Kit⁺Sca1^–CD34^–CD16/32^low) populations in Setd5^fl/fl and Setd5^CKO mice; n = 5. D–F FACS analysis of LSK⁺s (Lin^–Sca1⁺c-Kit⁺) and SLAM-HSCs (Lin^–Sca1⁺c-Kit⁺CD150⁺CD48^–) and absolute cell number in BM. SLAM-MPP: Lin^–Sca1⁺c-Kit⁺CD150^–CD48^–, HPC1: Lin^–Sca1⁺c-Kit⁺CD150^–CD48⁺, HPC2: Lin^–Sca1⁺c-Kit⁺CD150⁺CD48⁺; n = 5. G Representative FACS profiles of Ki67 staining of SLAM-HSCs. H, I The frequencies of G0, G1, S/G2/M phases in SLAM-HSCs and LSK⁺s are shown; n = 4. J, K Apoptosis analysis of HSPCs in Setd5^fl/fl and Setd5^CKO mice, LT-HSC: Lin^–Sca1⁺c-Kit⁺CD34^–Flt3^low, ST-HSC: Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3^low, MPP: Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3⁺; n = 5. Data are represented as mean ± SEM. * P < 0.05, ** P < 0.01, *** P < 0.001.

Fig. 2. Setd5 deletion by Vav-Cre impairs repopulation capacity of HSCs.

A–E Serial competitive repopulation assays with Setd5^fl/fl and Setd5^CKO LSK⁺ cells. Strategy for serial competitive repopulation assays (A), BMT: BM transplantation. Quantification of donor-derived (CD45.2) cells (B), donor-derived myeloid (M: CD11b⁺) cells (C), donor-derived T (CD3⁺) cells (D) and B (B220⁺) cells (E) in PB at indicated time points in primary (n = 9) and secondary (n = 10) recipients. F Proportion of each lineage in donor derived PB cells; n = 5. G Donor contribution of indicated cell populations in BM cells of primary and secondary recipient mice at indicated times; n = 5. H Poisson statistical analysis from the limiting dilution assay. LTR-HSCs: long-term repopulating (LTR)-HSC. Symbols represent the percentage of negative mice for each dose of cells. Solid lines indicate the best-fit linear model for each dosage. Dotted lines represent 95% confidence intervals. I Frequencies of functional HSCs were calculated according to Poisson statistics. Data are represented as Mean ± SEM; * P < 0.05, ** P < 0.01, *** P < 0.001.

Fig. 3. Setd5 regulates HSC pool in adult hematopoietic cells and plays an intrinsic role in HSCs.

A Schematic diagram of Mx1-Cre mediated Setd5 knockout in adult hematopoiesis. B, C The absolute number of LSK⁺s and SLAM-HSCs in Setd5^fl/fl and Setd5^IKO mice; n = 5. D, E BrdU incorporation of LSK⁺ and HSC populations and representative flow cytometry histogram of LT-HSC populations of BrdU incorporation 24 h after injection; n = 4. F Schematic diagram for transplantation assay with Setd5^fl/fl and Setd5^IKO BM cells. G–J Quantification of donor-derived (CD45.2) cells in the PB of recipient animals at indicated time points; n = 9. K Proportion of each lineage in donor derived PB cells; n = 5. L Donor contribution of indicated cell populations in BM cells of recipient mice 20 weeks after transplantation; n = 5. M Percentage of donor-derived cells in the PB of Setd5^fl/fl and Setd5^IKO secondary recipients at the indicated time points; n = 9. Mean ± SEM; * P < 0.05, ** P < 0.01, *** P < 0.001.

Fig. 4. Transcriptome profiling of Setd5-deficient HSCs reveals altered stem cell property and cell cycle signature.

A Heat map showing differential expression of 443 genes in Setd5^fl/fl and Setd5^CKO SLAM-HSCs, |Log₂foldchange | > 1, p value < 0.05; n = 3. B GSEA analyses for genes affected in the SLAM-HSCs of Setd5^fl/fl and Setd5^CKO mice. NES, normalized enrichment score. C Relative expression levels of cell cycle and multi-potency genes in HSC cells; n = 3. mRNA levels were normalized to the expression of 18 s. D. Experimental design for single cell sequencing with Smart-seq2, LSK⁺ cells were indexed sorted into 96-well plates containing lysis buffer. E LSK⁺ cells were projected onto Nestorowa et al. data [30]. F UMAP visualizations of single cell transcriptomes of identified clusters using Seurat. G–I Diffusion maps of all cells were colored according to the expression levels of selected genes. J, K Histograms showing the compositions of surface markers defined four populations and transcriptome-defined five clusters by ten cell types annotated with Nestorowa et al. data. C1-C5 means cluster 1-5. Ten annotated cells including E-SALM (EPCR⁺CD48^–CD150⁺), LT-HSC (Lin^–Sca1⁺c^-Kit⁺CD34^–Flt3^low), ST-HSC (Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3^lowCD48^–CD150^–), MPP1 (Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3^lowCD48^–CD150⁺), MPP2 (Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3^lowCD48⁺CD150⁺), MPP3 (Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3^lowCD48⁺CD150^–), LMPP (Lin^–Sca1⁺c-Kit⁺CD34⁺Flt3⁺), CMP, GMP and MEP. L GSEA of LT-HSC, proliferation and quiescence signatures comparing between Setd5^CKO and Setd5^fl/fl in HSC enriched cluster 1 and cluster 2. M, N Proportion of SLAM-HSCs, HPC1s and five clusters identified with Seurat in each of the cell cycle stages between two groups.

Fig. 5. SETD5 regulates the expression of E2F targets associated with HCF-1.

A Venn diagrams showing the overlap between Hela/MEL MS-identified proteins and reported SETD5 interactors. B Anti-FLAG immunoprecipitation of lysates from Hela and MEL cells transfected with SETD5-FLAG (OE) and empty FLAG vector (EV). Input: cell extract before immunoprecipitation, FLAG-IP: immunoprecipitation elute with 2×loading buffer. C Reciprocal immunoprecipitation with antibodies against HCF-1 in MEL cells transfected with SETD5-FLAG and empty FLAG vector. D Pie chart showing the distribution of ChIP-seq peaks for SETD5 with respect to promoter (define as a 3.0-kb proximal region centered on the gene start sites, including TSS), 5’UTR (5’-untranslated region), 3’UTR (3’-untranslated region), exon, intron and intergenic regions. E MEME motif analysis of SETD5 peaks, TF: transcription factor. F–G Density plots of SETD5 (blue) and E2F1/HCF-1 (green) normalized ChIP-seq signals, TES: Transcription End Site. H IGV tracks of target genes of SETD5-FLAG, and input DNA in MEL cells. The coverage data are represented as normalized reads of bins per million (BPM). I ChIP-qPCR of the binding of SETD5-FLAG to the promoters of indicated genes, Prc1 is a negative control which was not found in FLAG-SETD5’s ChIP-seq data. J Relative expression levels of target genes in control and Setd5^CKO LSK⁺ cells measured by real-time PCR; n = 3. mRNA levels were normalized to the expression of 18 s. Mean ± SEM; * P < 0.05, ** P < 0.01, *** P < 0.001.

Fig. 6. SETD5 regulates the promoter-proximal paused Pol II release on E2F targets.

A Density plots of SETD5 (blue) and PAF1 (green) normalized ChIP-seq signals. B, C Occupancy of total Pol II on all genes and DEGs (Fold change>1.5 and p adj value<0.05), ChIP normalization was implemented by integrating Spike-in Chromatin. D, E Density plot of all genes and DEGs with reduced average Pol II pausing index in Setd5^CKO. F IGV tracks comparing occupancy of Pol II within Cdk1 locus. G, H ChIP-qPCR of Pol II and Pol II Ser2P at the promoters of indicated genes in c-Kit⁺ cells from control or Setd5^CKO mice; n = 2. I ChIP-qPCR analysis of PAF1 at the promoters of indicated genes in c-Kit⁺ cells; n = 2. J The H3K9 dimethylation levels at Cdc20 and Cenpe promoters; Gapdh as an internal control; n = 2. Mean ± SEM; *P < 0.05, **P < 0.01, *** P < 0.001.

Fig. 7. Setd5CKO deficiencies could be partially rescued by super elongation complex-related inhibitors.

A c-Kit⁺ BM cells from Setd5^fl/fl and Setd5^CKO mice were treated with BAY-1143572 1 μM, JQ1 500 μM, and BMS-387032 300 nM for 24 h before cell cycle analysis. B The frequencies of G0, G1, S/G2/M LSK⁺s in drug-treated c-Kit⁺ cells for 24 h from Setd5^fl/fl and Setd5^CKO mice were shown; n = 4. C A model for the role of SETD5 in modulating hematopoietic stem cell homeostasis. In normal HSCs, SETD5 occupies at E2F-responsive promoters in association with HCF-1, maintains a paused Pol II state with PAF1 complex, leading to transcriptional silencing of E2F targets and a quiescent state of HSCs. Once Setd5 is depleted, PAF1 occupancy is weakened and triggers Pol II elongation to active transcription of E2F target genes and promote G1 to S phase transition.