Role of the Polycomb Methyltransferase Ezh1 in Myelodysplastic Syndrome Induced By Ezh2 Insufficiency

Ezh1 and Ezh2, the catalytic components of polycomb-repressive complex 2 (PRC2), negatively control gene expression by catalyzing mono, di, and tri-methylation of histone H3 at lysine 27 (H3K27me1/me2/me3). Loss-of-function mutations of EZH2, but not those of EZH1, have been found in patients with hematologic malignancies such as myelodysplastic syndrome (MDS), myeloproliferative neoplasms (MPNs), and MDS/MPN overlap disorders.

We previously demonstrated that hematopoietic cell-specific Ezh2 knockout mice (Ezh2^Δ/Δ) developed hematologic malignancies including MDS and MDS/MPN. Although deletion of Ezh1, another enzymatic component of PRC2, (Ezh1^-/-) did not significantly affect global H3K27me3 levels or hematopoiesis, deletion of both Ezh1 and Ezh2 in mice (Ezh1^-/-Ezh2^Δ/Δ) caused rapid exhaustion of hematopoietic stem cells (HSCs). Given that only Ezh1 and Ezh2 are known as enzymatic components of PRC2, we concluded that residual PRC2 enzymatic activity is required for HSC maintenance and development of hematologic malignancies in the setting of EZH2 insufficiency frequently observed in MDS. However, the role of Ezh1 in Ezh2-insufficient hematologic malignancies is still not fully understood since hematopoiesis could not be maintained in Ezh1^-/-Ezh2^Δ/Δ mice. Here we analyzed the impact of Ezh1 heterozygosity on Ezh2-null hematopoiesis (Ezh1^+/-Ezh2^Δ/Δ), in which PRC2 activity is mediated by a single allele of Ezh1, for better understanding of Ezh2-deficient hematologic malignancies.

We first transplanted BM cells from Ezh1^+/-Ezh2^flox/flox CD45.2 mice with CD45.1 wild-type competitor cells into lethally irradiated CD45.1 recipient mice and deleted Ezh2 by intraperitoneal injection of tamoxifen. Ezh1^+/-Ezh2^Δ/Δ cells exhibited a lower repopulation capacity than Ezh2^Δ/Δ but established persistent repopulation for at least 6 months after the deletion of Ezh2 while double knockout cells (Ezh1^-/-Ezh2^Δ/Δ) were outcompeted by competitor cells immediately. We next transplanted BM cells from Ezh1^+/-Ezh2^flox/flox CD45.2 mice without CD45.1 wild-type competitor cells into lethally irradiated CD45.1 recipient mice and deleted Ezh2 by intraperitoneal injection of tamoxifen. Importantly, recipient mice reconstituted with Ezh1^+/-Ezh2^Δ/Δ cells exhibited MDS-like phenotypes including anemia and morphological myelodysplasia, which were more pronounced than those of Ezh2^Δ/Δ mice. Ezh1^+/-Ezh2^Δ/Δ mice also showed more advanced hematological abnormalities such as erythroid differentiation block, increased apoptosis of erythroid cells, and extramedullary hematopoiesis in the spleen than Ezh2^Δ/Δ mice did. These results suggest that Ezh1 heterozygosity promotes the development of myelodysplasia in the setting of Ezh2insufficiency.

Next we examined the molecular mechanism by which the loss of Ezh1 promotes myelodysplasia. Western blot and ChIP-sequence analyses revealed that global levels of H3K27me3 were not significantly changed but H3K27me3 levels at promoter regions of the PRC2 target genes were obviously reduced by Ezh1 heterozygosity in Ezh2^Δ/Δ HSPCs. As a consequence, PRC2 target genes were highly de-repressed in Ezh1^+/-Ezh2^Δ/Δ LSK HSPCs compared with Ezh2^Δ/Δ HSPCs. Among these, several genes appeared to be associated with MDS such as S100A9, encoding an inflammatory protein implicated in dyserythropoiesis in MDS. Furthermore, gene set enrichment analysis showed that the genes highly expressed in myeloid cells were positively enriched by Ezh1 heterozygosity in Ezh2^Δ/ΔHSPCs.

These findings indicate that dosage of Ezh1 is critical in the maintenance of Ezh2-insufficient hematopoiesis as well as the progression of MDS with Ezh2 insufficiency.