Linker histone regulates the myeloid versus lymphoid bifurcation of multipotent hematopoietic stem and progenitors

Abstract

Myeloid-biased differentiation of multipotent hematopoietic stem and progenitor cells (HSPCs) occurs with aging or exhaustion. The molecular mechanism(s) responsible for this fate bias remain unclear. Here, we report that linker histone regulates HSPC fate choice at the lymphoid versus myeloid bifurcation. Linker histones package nucleosomes and compact chromatin. HSPCs expressing a doxycycline (dox)-inducible H1.0 transgene favor the lymphoid fate, display strengthened nucleosome organization, and reduced chromatin accessibility at subsets of genomic regions. The genomic regions showing reduced chromatin accessibility host many known marker genes of myeloid-biased HSCs. The transcription factor Hlf is located in one of the most differentially closed regions, whose chromatin accessibility and gene expression are reduced in H1.0^high HSPCs. Failure to reduce Hlf expression in multipotential HSPCs abrogates the H1.0-endowed lymphoid potential. Furthermore, HSPCs display aspartyl protease-dependent H1.0 decreases, especially in response to interferon alpha (IFNα). Aspartyl protease inhibitors preserve endogenous H1.0 levels and promote the lymphoid fate of wild type HSPCs. Thus, our work elucidates a molecular scenario of how myeloid bias arises and uncovers a point of intervention for correcting myeloid skewed hematopoiesis.

Keywords: inflammation; linker histone; myeloid bias.

Fig. 1.

H1.0 expression in HSPCs confers increased lymphopoiesis. (A). Diagram for transgene targeting strategy to introduce H1.0-GFP into the Hprt locus located on chromosome X. rtTA is expressed from the Rosa26 locus. Due to random X chromosome inactivation, heterozygous female cells could have the iH1.0-GFP transgene on either X chromosomes. The cells having iH1.0-GFP on the active X (Xa), but not those having it on the inactive X (Xi), will show dox-inducible transgene expression. (B). Scheme of the experimental setup for noncompetitive whole bone marrow (WBM) transplantation using X^wtX^iH1.0-GFP heterozygous females as donors. All recipients are fed with dox water. (C). Representative FACS dot plots for the recipients in B. Donor-derived CD45.1 + cells in peripheral blood further divided into GFP- and GFP+ compartments. Representative lineage markers CD11b+ and B220+ cells at 16 wk posttransplantation. (D). Quantification of lineage marker-positive cells within donor-derived CD45.1 + cells divided into GFP- and GFP+ compartments as shown in C. Lymphoid/myeloid ratio is determined as (%B220+ + %CD3+)/%CD11b+ in individual recipient mice. n = 6 each. (E). Scheme of the experimental setup for noncompetitive WBM transplantation using X^{iH1.0- GFP}Y male mice as donors. The recipients are treated with regular or dox water. In the dox water-treated recipients, donor (CD45.1 +) cells are further divided by their H1.0-GFP fluorescence intensity. (F). Representative FACS dot plots for the recipients shown in E, at 16 wk posttransplantation. – dox and + dox total CD45.1 +, total GFP+, and GFP^high cells in peripheral blood stained for lineage markers (CD11b and B220) are shown. GFP^high denotes the top 10% among total GFP+ cells. (G). Quantification of the lymphoid/myeloid ratios in the recipients shown in F. (H). Scheme of the experimental setup for noncompetitive WBM transplantation using X^iH1.0-GFPY or X^{iHMGN1-mCherry}Y male mice as donors. The recipients are treated with dox water. (I). Quantification of the number of MPP4 and CLP per leg (femur and tibia) in recipient mice 16 wk posttransplantation shown in the Left panel; the numbers of LT-HSC, ST-HSC, MPP2, and MPP3 shown in the Right panel. n = 5 per group. (J). Scheme of the experimental setup for noncompetitive WBM transplantation using X^iH1.0-GFPX^{iHMGN1-mCherry} females as donors. All recipients are fed with dox water. (K). Representative FACS dot plots showing the donor-derived (CD45.1 +) peripheral blood cells at 4 and 8 wk posttransplantation in recipients. (L). Quantification of the lymphoid/myeloid ratios within the respective GFP+ and mCherry+ cells within individual recipients. n = 4 for 4 to 12 wk; n = 3 for 16 wk. (M). Quantification of bone marrow HSPC subsets positive for H1.0-GFP or HMGN1-mCherry in recipient mice 12 to 16 wk posttransplantation. n = 4 each. (N). Quantification of lymphoid colony-forming units by WBM cells at 16 wk posttransplantation. n = 3 each. (O). Quantification of myeloid colony-forming units by WBM cells at 16 wk posttransplantation. n = 3 each. (P). Scheme of the experimental setup for secondary noncompetitive WBM transplantation using X^iH1.0-GFPX^{iHMGN1- mCherry} as donors. All recipients are fed with dox water. (Q). Quantification of bone marrow HSPCs in secondary recipients 16 wk posttransplantation (n = 4 to 7 each). (R). Quantification of lymphoid and myeloid colony-forming units by WBM cells at 16 wk postsecondary transplantation. Individual values as mean ± SD are shown. *P < 0.05, **P < 0.01, ***P < 0.001, by unpaired, 2-tailed Student’s t test except for D and L (paired, 2-tailed Student’s t test) and G (Ordinary One-Way ANOVA with post hoc Tukey). See also SI Appendix, Fig. S1.

Fig. 2.

H1.0-overexpressing HSPCs display strengthened nucleosome organization. (A). Quantification of the inferred abundance of H1 subtypes as percent of total H1 by LC–MS in WT and iH1.0- GFP+ LSK and GMP cells. n = 3 each. (B). Quantification of the inferred H1.0/nucleosome ratio in WT and iH1.0-GFP+ LSK and GMP cells. (C). ATAC-seq fragment lengths and density in WT and iH1.0 + LSK cells. Three types of chromatin with regard to their nucleosome repeat signals are detected: Type A contains open chromatin and no nucleosome repeat signal in both WT and iH1.0 LSK cells; Type B has weak nucleosome repeat signal with an average of ~4 bp increase in nucleosome repeat length in iH1.0 + LSK cells; Type C has no nucleosome repeat signal in WT LSK but exhibits strong nucleosome repeats of 190 bp in iH1.0 + LSK cells. (D). Linear regression model depicting GC content vs. % of ATAC seq reads in each of the 25 ChromHMM chromatin states in iH1.0 + relative to WT LSK cells. Reads coming from chromatin states with low GC content appear underrepresented in iH1.0 + relative to WT LSK cells. See also Dataset S3 for details. Individual values as mean ± SD are shown. *P < 0.05, **P < 0.01, ***P < 0.001, by unpaired, 2- tailed Student’s t test. See also SI Appendix, Fig. S2.

Fig. 3.

H1.0 expression imparts lymphoid fate potential by reducing chromatin accessibility and gene expression of Hlf. (A). Chromatin accessibility measured by ATAC-seq around the Hlf genomic region during WT HSPC differentiation from dataset GSE162551, aligned with those obtained from WT and iH1.0 LSK cells, further aligned with the ChromHMM chromatin states. n = 2 for WT and n = 3 for iH1.0 + LSK cells; one representative shown for each genotype. (B). ATAC qPCR quantification for the region highlighted in (A) from WT and iH1.0 + LSK cells, normalized to that of mRPL30. n = 2 for WT and n = 3 for iH1.0 + LSK cells. (C). Hlf and H1.0 mRNA levels in freshly sorted WT and iH1.0 + LSK cells from mice on dox water for 3 wk. n = 3 donors per group. (D). Hlf and H1.0 mRNA levels in cultured WT and iH1.0 LSK cells. n = 3 donors per group. (E). Quantification of lymphoid and myeloid CFU numbers using WT LSK cells transduced with indicated viral constructs. Each dot is a triplicate sample; results shown are representative of three independent experiments. Individual values as mean ± SD are shown. *P < 0.05, **P < 0.01, ***P < 0.001, by unpaired, 2- tailed Student’s t test. See also SI Appendix, Fig. S3.

Fig. 4.

H1.0 level is amenable to physiologic and pharmacologic regulation. (A). Scheme of the experimental setup for analyzing bone marrow HSPCs from X^iH1.0-GFPX^{iHMGN1-mCherry} female mice, which are unmanipulated except for 1 wk of dox water treatment. (B). Representative FACS plots showing the distribution of LSK subsets positive for H1.0-GFP or HMGN1- mCherry in mice diagramed in A. n = 3 each. (C). Left: representative FACS histogram showing GFP fluorescence intensity in indicated LSK subsets shown in B. Right: quantification for GFP mean fluorescence intensity (MFI) for the GFP+ cells in indicated populations. n = 3 each. (D). Scheme of the experimental setup for analyzing bone marrow HSPCs from X^iH1.0-GFPX^{iH1.0 GFP} female mice, which are unmanipulated except for 3 wk of dox water treatment. (E). Left: representative FACS histogram plot showing the GFP fluorescence intensity in indicated HSPC populations. Right: quantification for GFP mean fluorescence intensity (MFI) for the GFP+ cells in indicated populations. n = 5 each. (F). Representative FACS histogram showing GFP fluorescence intensity in BaF3 cells expressing viral constructs: control pan-cellular GFP (EV-GFP) and dox inducible H1.0-GFP. Cells were cultured in the presence or absence of pepstatin. See also SI Appendix, Fig. S4C. (G). Representative Western blot using H1.0 and GFP antibodies in BaF3 cells expressing viral EV-GFP or iH1.0-GFP, in the presence or absence of pepstatin. Actin was probed as a loading control. (H). Representative FACS histogram plot showing GFP fluorescence intensity in freshly sorted iH1.0-GFP LSK cells cultured in the presence or absence of pepstatin. Right: quantification of GFP mean fluorescence intensity (MFI) in the GFP+ iH1.0 LSK cells. n = 6 each. (I). Quantification of GFP MFI in GFP+ iH1.0 LSK cells cultured in the presence or absence of IFNα and pepstatin. n = 6 each. (J). Western blot of endogenous H1.0 and H1.0-GFP in WT and iH1.0 + LSK cells cultured in the presence or absence of IFNα/IFNγ ± pepstatin. Actin was probed as a loading control. Results are representative of three independent experiments. (K). Cathepsin D activity as measured by a fluorescent substrate. Results are normalized to relative fluorescence units (RFU) in WT LSK cell lysates. n = 5 each. (L). Quantification of lymphoid and myeloid CFUs in LSK cells treated with pepstatin. Each dot is an independent experiment. n = 4 each. (M). Western blot of endogenous H1.0 protein in human CD34+ cells cultured in the presence of pepstatin or atazanavir. Actin was probes as a loading control. (N). Quantification of lymphoid and myeloid CFUs is LSK cells treated with indicated protease inhibitors. Each dot is a triplicate sample; results shown are representative of three independent experiments. n = 3. Individual values as mean ± SD are shown. *P < 0.05, **P < 0.01, ***P < 0.001, by paired, 2- tailed Student’s t test, except for 4 N which is unpaired. See also SI Appendix, Fig. S4.