Tissue specific requirement for the inv(16) oncogene CBFB::MYH11 implies unique leukemia survival mechanisms in the bone marrow microenvironment

Abstract Inversion of chromosome 16 [inv(16)] generates the oncogenic fusion gene Cbfb::MYH11 (CM) which isknown to initiate leukemogenesis in Acute Myeloid Leukemia (AML). Considering the fusion protein is invariably found in patients at relapse, it is assumed that CM is also required after leukemic transformation. Inhibitors of CM have been proposed as a potential treatment for inv(16) AML, which makes it critical to understand the fusion protein's role in leukemia maintenance. To test the role of Cbfb::MYH11in leukemia maintenance, we used leukemia cells from a knockin mouse model expressing the fusion gene under the control of the inducible Mx-1 promoter (Cbfb+/56M, Mx1-Cre+). These leukemia cells were transduced with a doxycycline (dox) inducible shRNA against the MYH11 domain of the fusion gene (shMYH11), that also constitutively expresses GFP. shMYH11 transduced leukemia cells were transplanted into sub-lethally irradiated mice and monitored for leukemic engraftment, and then given dox in their drinking water to induce knockdown (KD) of CM. We found that decreased CM expression nearly eliminated the leukemia cells in the peripheral blood (PB) and spleen (SP) starting at day 7 of dox treatment. In the bone marrow (BM), we observed a decrease in the percentage of leukemia cells, although the percentage of leukemia cells remained above 20% through day 28 of dox treatment, despite these cells having significantly lower CM RNA and protein. In addition, at day 7 of KD, shMYH11+ cells harvested from SP, but not the BM, showed a significant increase in apoptosis. These results indicate that leukemia cells in the SP and PB require CM expression, but that there are mechanisms unique to the BM that promote leukemia survival.To investigate mechanisms contributing to the survival of CM KD cells in the BM, we performed bulk RNA sequencing. KEGG enrichment analysis revealed that genes associated with pro-survival and stress adaptation pathways are upregulated in CM KD BM cells. Specifically, the autophagy associated genes ATF6, ATG14, ATG2A, BECN1 and mTOR were significantly upregulated. Western blot analysis revealed higher expression of the autophagy marker LC3B in CM KD BM, but not SP, cells as compared to control leukemia cells. To test if there is a distinct subpopulation of leukemia cells capable of survival in the BM after loss of CM, we carried out single-cell RNA sequencing. Unsupervised clustering analysis revealed the presence of distinct clusters, some of which expanded with CM KD, implying that there are specific sub-populations able to survive withoutthe fusion gene and that CM-independence is not a general feature of all leukemia cells in the BM. Interestingly, we did not observe common changes in gene expression or deregulated pathways in response to CM KD associated with the CM-independent clusters. This implies that CM-independent cells are not inducing pro-survival, anti-apoptosis genes in response to KD of the fusion protein. In fact, we found that in control leukemia cells in the BM, the CM-independent clusters shared a transcriptional profile associated with cytokines and chemokine activity. Collectively, our results imply that the CM-independent leukemia cells constitutively express a pro-survival program associated with signaling from the BM microenvironment.To test if CM-independent leukemia cells can re-establish disease, we maintained a cohort of mice on dox. We found that ~50% of dox treated mice regained GFP+ cells in their PB while maintaining KD of the fusion gene. To identify molecular mechanisms allowing CM KD cells to re-emerge from BM, we performed a whole transcriptome sequencing on relapse and control leukemia cells. Enrichment analysis indicated that genes associated with interferon regulator factor (IRF) signaling were significantly deregulated. Specifically, IRF1, IRF7, and IRF9 are decreased in the relapse samples compared to control. Notably, loss of these genes is associated with relapse in AML patients. This implies that suppression of IRF signaling might be a common mechanism by which leukemia cells acquire the ability to survive outside of the BM.Taken together, our study supports the development of inhibitors targeting the CM fusion protein and that such drugs would likely be effective at managing leukemic burden. However, a better understanding of the survival mechanisms in the BM is critical to cure inv(16) AML in all patients.