HHEX expression drives AML development

In this issue of Blood, Takeda and colleagues¹  reveal the importance of hematopoietically expressed homeobox (HHEX) in promoting acute myeloid leukemia (AML) transformation in cooperation with mutant additional sex combs-like 1 (ASXL1) and identify MYB and ETV5 as critical transcriptional targets shared by HHEX and mutant ASXL1.

AML results in the accumulation of immature clonal myeloid cells, driven by acquired genetic mutations in hematopoietic stem and progenitor cells (HSPCs). The largest class of recurrent AML gene mutations affect epigenetic modifiers, including DNMT3A, IDH1/2, TET2, EZH2, MLL/KMT2A, and ASXL1, which are often found as early initiating events.²  Although mutations in any one of these epigenetic-modifying genes result in significant changes in the epigenetic landscape and gene expression in HSPCs, they are often insufficient to result in overt transformation to leukemia, suggesting that additional genetic and/or molecular changes are required for AML. For example, the cooccurrence of mutant DMNT3A and FLT3-ITD is associated with an unfavorable prognosis in patients and results in rapid and penetrant AML in mouse models.^3-5  In other cases, an initiating genetic mutation in the context of an altered cellular state coinciding with changes in the expression of critical genes is sufficient for leukemic cell transformation. Although frequently cooccurring gene mutations have been extensively studied in AML, much less is known about the relevant altered gene expression changes that cooperate with leukemia-initiating gene mutations.

ASXL1 is one of the most frequently mutated genes in myeloid malignancies, occurring in 5% to 11% of patients with AML. ASXL1 belongs to a family of chromatin-binding Polycomb proteins and is involved in controlling gene expression by interacting with epigenetic regulators, such as Polycomb repressive complex 2 (PRC2) and BAP1. Interaction of ASXL1 and BAP1 at promoters results in monoubiquitination of histone H2A at lysine 119 (H2AK119ub), catalyzed by PRC1 complexes, and subsequent repression of target gene expression. The majority of ASXL1 mutations are frameshift or nonsense mutations that disrupt the C-terminal plant homeodomain finger region resulting in C-terminally truncated ASXL1 mutant proteins (referred herein as ASXL1-MT). ASXL1-MT retains its ability to interact with several of its epigenetic binding partners and induces aberrant histone modifications in a dominant-negative or gain-of-function manner. Thus, these epigenetic changes induced by ASXL1-MT are thought to contribute to the development of myeloid neoplasms.

The contribution of ASXL1 mutations has been illuminated by analysis of ASXL1-MT knockin (ASXL1-MT-KI) mice, which develop mild anemia, impaired erythroid differentiation, myeloid skewing, and dysplasia.⁶  Although these hematologic changes are indicative of early-stage myelodysplastic syndromes (MDS), the ASXL1-MT-KI mice never develop overt leukemia, suggesting that additional elements are required for myeloid transformation of ASXL1-mutated cells. To identify cooperating molecular events with ASXL1-MT, the team performed a retrovirus-mediated insertional mutagenesis screen and identified HHEX as a candidate that may cooperate with ASXL1-MT to induce myeloid leukemia.

HHEX is a homeobox-containing transcriptional repressor that is overexpressed in various subtypes of AML, including those with ASXL1 mutations.⁷  In the current study, the team investigated the potential cooperation between ASXL1-MT and HHEX overexpression in myeloid leukemogenesis. Utilizing retroviral models, overexpression of HHEX expanded ASXL1-MT–expressing HSPCs by preventing apoptosis and blocking differentiation. However, overexpression of HHEX had nominal effects on normal HSPCs. Importantly, overexpression of HHEX in ASXL1-MT HSPCs accelerated the development of AML, whereas HHEX depletion attenuated the leukemic cell function of human ASXL1-MT–expressing AML cells. To identify the mechanistic basis for the cooperation between HHEX and ASXL1-MT, they performed RNA-sequencing and chromatin immunoprecipitation and found that HHEX and ASXL1-MT coregulated MYB and ETV5 expression, which they demonstrated to be essential for the HHEX-induced development of AML. Mechanistically, they found that expression of ASXL1-MT enhanced the binding of HHEX to the promoter loci of MYB and ETV5. Deletion of MYB or ETV5 was sufficient to induce apoptosis or differentiation of ASXL1-MT–expressing AML cells, respectively. In contrast, expression of MYB or ETV5 was able to restore the function of HHEX-depleted ASXL1-MT–expressing AML cells. Collectively, these findings revealed a novel HHEX-MYB/ETV5 axis that promotes myeloid transformation of ASXL1-mutated preleukemia cells (see figure).

Although the authors found that cells expressing ASXL1-MT showed binding of HHEX to MYB and ETV5 promoter loci, precisely how ASXL1-MT promotes accessibility of HHEX to the target loci requires additional investigation. It is possible that ASXL1-MT promotes accessibility of HHEX via reduction of H2AK119ub at these key loci. Alternatively, the ASXL1-MT/BAP1 complex may enhance promoter activity independent of histone modifications by recruiting transcriptional factors that enhance promoter activity, including MYB and ETV5. Future studies will need to determine the precise cooperation of target gene expression control by HHEX and ASXL1-MT.

One of the interesting and relevant findings of this work is that HHEX overexpression is not exclusive to ASXL1-MT AML, but is broadly elevated in AML as compared with normal hematopoietic cells. Based on this observation, the authors extended their analysis of HHEX and found that overexpression of HHEX in hematopoietic cells from RUNX1-ETO9a and FLT3-ITD mice similarly resulted in AML, suggesting that increased HHEX expression is critical for the progression of AML. The regulation of HHEX expression in AML remains unknown and will require future investigation. However, recent studies have described HHEX expression to be regulated by LMO2/ERG/FLI1 in T-cell leukemia and by RUNX1 in hematopoietic cells.^8,9  Given the paucity of effective treatments for AML, interfering with HHEX-mediated activity may provide new opportunities for therapeutic developments.