Histone variants can replace replication-coupled histones in the nucleosome. Among core histones, the H2A family exhibits highest sequence divergence resulting in the largest number of variants. Dysregulated expression of macrohistone (macroH2A) variants has been reported in different cancers. Their role in normal hematopoiesis and hematopoietic cancers remained so far poorly understood.

We found high expression of macrohistone H2AFY (macroH2A1) in hematopoietic stem and progenitor cells (HSPCs) and primary AML as compared to differentiated hematopoietic cells. To assess for the functional dependency of human AML cells on macroH2A1, we performed in vitro CRISPR-Cas9 negative selection competition assays. Genetic inactivation of H2AFY in 4/9 AML cell lines (Molm-13, MV4;11, THP-1, OCI-AML3) resulted in outcompetition of H2AFY-deleted cells against non-deleted competitors. To confirm these findings in vivo, we performed xenotransplantation experiments. Genetic inactivation of H2AFY in human AML cells resulted in delayed leukemia development in recipient mice (median survival for NT-control: 35.5 days vs. H2AFY-KO: 44.0 days; p=0.002).

To validate the functional impact of macroH2A1 on leukemic stem cells (LSCs), we generated a novel conditional knockout mouse model for H2AFY [B6(Cg)-H2afytm1Flh/J (H2afy flox/flox)]. Here, isolated lineage-depleted H2afy+/+, Mx1-Cre+ and H2afyfl/fl, Mx1-Cre+ bone marrow cells (Lin-BMCs) were retrovirally infected with MLL-AF9-GFP (MA9) and transplanted into sub-lethally irradiated recipient mice. Following establishment of leukemia, we transplanted 1x105 leukemic cells and inactivated H2afy in the leukemic cells by pIpC injection on day 7, 9 and 11 post-transplantation. Genetic deletion of H2afy resulted in delayed disease progression (median survival +/+: 83.5 days vs. -/-: 162.0 days, p=0.0013) and reduction of disease penetrance to 61%. Organ infiltration by leukemic cells led to disease development in a limited number of H2afy-/- recipient mice. Isolated H2afy-/- leukemic cells showed a decreased in proliferation and an increased in differentiation. Importantly, 37% of H2afy-/- recipient mice that developed AML showed only partial excision of H2afy, indicating selection of partially excised clones. Long-term survivors without leukemia development showed no molecular evidence for counterselection. Irrespective of the time of disease development and survival, immunophenotypic analysis of (Lin-cKit+Sca-1-FcγR+CD34+GFP+) LGMPs revealed decreased numbers of LSCs following H2afy-deletion (+/+: 2631/106 BMC; -/-: 727/106 BMC; p=0.0002). To assess for the functional consequences of H2afy-deletion on the LSC pool, we performed limiting dilution assays. When transplanting 1x105, 1x104, 1x103 and 1x102 H2afy+/+ or -/- leukemic cells into sub-lethally irradiated recipients, LGMP frequency in H2afy-deficient cells was significantly reduced (1/1 vs. 1/29928; p=1.25e-19). The reduction of LGMP frequency correlated with reduced plating efficiency and increased differentiation. Consistently, primary human AML cells showed impaired colony forming capacity following H2afy deletion (mean colony number NT-control: 60.0±17.1 vs. H2afy-KD: 30.1±8.9; p=0.0023).

To assess for the impact of H2afy deletion on normal HSPCs, we performed competitive repopulation assays. Immunophenotypic abundance of HSPC in H2afy+/+ vs. -/- animals revealed comparable numbers in all relevant subpopulations. Serial competitive transplantation of H2afy+/+ and -/- BMC into primary and secondary recipient animals showed no competitive disadvantage or lack of self-renewal capacity of H2afy-/- cells. Likewise, colony formation of healthy human donor CD34+ cells was not impaired following depletion of H2AFY by RNAi.

Transcriptome analysis of LGMP populations indicated increased of myeloid cell development and loss of LSC signatures in H2afy-deficient LGMPs. Likewise, ATAC-seq of H2afy+/+ and -/- LGMP population revealed profound changes in the chromatin structure and a strong correlation between chromatin accessibility and mRNA expression.

Together, our results provide first evidence for a critical function of H2afy (macroH2A1) in models of AML while being dispensable for normal HSPC function. Experiments to assess for the mechanistic consequences of H2afy deletion on chromatin are currently under way.

No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution