Non-Coding HOX Fusions in Pediatric Non-Down Syndrome Acute Megakaryoblastic Leukemia

The homeobox (HOX) genes are a highly conserved family of transcription factors involved in embryonic patterning as well as adult hematopoiesis. Dysregulation of HOX genes, in particular upregulation of HOXA cluster genes, is a frequent event in Acute Myelogenous Leukemia (AML). Recently, we performed a detailed genomic analysis on pediatric non-Down Syndrome Acute Megakaryoblastic Leukemia (non-DS-AMKL) and identified novel fusions involving a HOX cluster gene in 14.9% of the cases. While most fusions were predicted to lead to an in-frame functional protein, several fusions included a non-coding HOX antisense gene (PLEK-HOXA11-AS, C8orf76-HOXA11-AS, HOXA10-AS-CD164) that were predicted to result in a loss of function of these regulatory transcripts. The functional consequence of these events, however, remain unknown. HOXA11-AS (human) and Hoxa11os (mouse) have been previously shown to have mutually exclusive expression with the Hoxa11 transcript throughout development. We therefore hypothesized that loss of function of non-coding HOX antisense genes as a result of these structural variations would cause upregulation of nearby coding HOXA genes that in turn promote leukemogenesis.

To test this hypothesis, using CRISPR-Cas9 technology, we genome edited the human AMKL cell line CMK to carry the PLEK-HOXA11-AS translocation. qRT-PCR of HOXA11-AS and HOXA9-11 transcripts in this cell line recapitulated the pattern seen in patient specimens. Specifically, HOXA11-AS expression was significantly diminished while HOXA10 and HOXA11 transcripts were upregulated 1.8-2.5-fold when compared to parental CMK cells (p=0.0385 and p=0.006 respectively). To further investigate the loss of HOXA11-ASin vivo a CRISPR-Cas9 Hoxa11os knockout mouse model was established. qRT-PCR on bone marrow confirmed the loss of Hoxa11os transcripts in heterozygous (Hoxa11os1^+/-) and homozygous (Hoxa11os^-/-) mice of both genders (p=<0.0001-0.0012). Consistent with Hoxa11os knockdown, Hoxa11 transcripts were upregulated in male (1.8-fold p=0.0023 Hoxa11os^+/-, and 2-fold p=0.0052 Hoxa11os^-/-)and female (1.3-fold p=0.0074 Hoxa11os^+/- and 2.2-fold p=0.0226 Hoxa11os^-/-) bone marrow compared to wild type gender matched littermates. Interestingly, flow cytometry analysis of progenitor subsets revealed gender specific findings. We found a significant increase in the frequency of the lineage negative, Sca-1 and c-Kit positive (LSK) population in males (0.13% of total bone marrow Hoxa11os^+/+, 0.19% p=0.0214 Hoxa11os^+/-, and 0.25% p=0.0001 Hoxa11os^-/-) compared to wild type male littermates but not in female mice at 8 weeks of age. In contrast an increase in the megakaryocyte-erythroid (MEP) population was seen only in the female setting (0.07% Hoxa11os^+/+, 0.15% p=0.0055 Hoxa11os^+/-, and 0.165% p<0.0001 Hoxa11os^-/-). Limiting dilution colony forming assay confirmed the higher LSK frequency with a 2-fold increase in the number of colonies for male knockout marrow compared to wild type marrow in contrast to the female setting where no significant differences were seen. As hormonal signals have been shown to regulate expression of HOX genes and differences in clonogenicity of male and female stem cells has been previously demonstrated, we reasoned this phenomenon could be secondary to extrinsic stimuli in vivo. The relatively uniform Hoxa11 levels in male and female knockout mice, however, suggested that cell intrinsic factors may also play a role. We therefore overexpressed HOXA11 into male and female wild type bone marrow ex vivo for colony forming assays to determine if elevated levels of the HOXA11 protein led to functional differences. This assay demonstrated a clear enhancement of self-renewal in male but not female bone marrow in contrast to HOXA9 overexpression which serially replated in both genders.

Combined these data demonstrate that loss of function alterations in Hoxa11os transcripts lead to upregulation of Hoxa11 and gender specific hematopoietic progenitor cell perturbations. Ongoing efforts include competitive transplant studies as well as RNA and ChIP sequencing to identify gender specific downstream targets of Hoxa11 in the hematopoietic compartment in order to understand the selective expansion of progenitor subsets and male specific self-renewal capacity of this protein. These data will contribute to our understanding on how HOXA11-AS translocations promote oncogenesis.