The Hypomethylating Agent 5-Aza-4’-Thio-2’-Deoxycytidine Targets Early B Progenitor Cells and Leads to B Cell Precursor Acute Lymphoblastic Leukemia in Rag1 Knockout Mice

5-Aza-4’-thio-2’-deoxycytidine (Aza TdCyd or ATC) is a cytidine analog used as a promising DNA methyl transferase inhibitor (DNMTi) in preclinical studies against solid tumors. Previously, we demonstrated that ATC functions as a potent mutagen capable of introducing C>G transversions in vivo in wild-type C57BL/6 mice and in vitro in human leukemia cell lines, specifically within a 5’-NCG-3’ context. ATC-treated mice primarily developed acute lymphoblastic leukemia of both B cell and T cell origin.
To investigate effects of ATC on hematopoietic differentiation in vitro, we treated wild type (WT) bone marrow (BM) or bone marrow from mice that are predisposed to leukemic transformation. Although ATC treatment had no discernible effect on myeloid, erythroid, or B cell differentiation of WT BM using methylcellulose culture, there was a clear effect of ATC treatment of BM from mice that are predisposed to myeloid, T and B cell leukemia due to expression of a NUP98::HOXD13 (NHD13) transgene. Extended methylcellulose culture of NHD13 BM cells revealed the presence of dense colonies resembling CFU-Blast (CFU-Bl) colonies. Treatment of NHD13 BM with ATC resulted in loss of the CFU-Bl colonies and appearance of mast cell (CD16/32+Kit+) colonies identical to those seen in WT BM. This phenotypic result was accompanied by loss of Cdkn2a methylation in the ATC treated BM colonies.
Given the in vitro effect of ATC on myeloid differentiation, we were curious as to why ATC treatment did not result in malignant transformation of myeloid cells in vivo. We hypothesized that ATC treatment did not produce myeloid leukemia in vivo simply because the mice died of a B or T lymphoid leukemia prior to any putative myeloid transformation. To determine whether ATC could be oncogenic in non-lymphoid cells, we examined the mutagenic effects of ATC treatment in Rag1 knockout (Rag1 KO) mice. Given that Rag1 KO mice lack mature T and B cells, we reasoned that the Rag1 KO mice would be “protected” from the development of B or T lineage leukemia. Surprisingly, in vivo treatment of Rag1 KO mice with ATC invariably led to the development of CD19+ or CD19+ B220+ B-lineage ALL, which demonstrate a gene expression signature similar to a previously described murine precursor B-cell acute lymphoblastic leukemia. We identified early B cell precursor (BCP) populations with CD19+ or CD19+ B220 expression as potential targets for ATC-induced mutations and leukemia following in vivo selection of malignant clones. Whole exome sequencing (WES) revealed a dramatic increase in C>G transversions in ATC-treated samples, which occurred almost exclusively within a 5’-NCG-3’ context. Examination of the methylation status of known C>G mutated CpGs identified from WES data using bisulfite sequencing showed that the mutated CpGs identified in ATC-induced BCP ALL tumor samples were heavily methylated. Additionally, genes frequently mutated in BCP ALL, such as Bcor, Trp53, Jak1, Jak3, and Ptpn11, were identified as recurrent targets of ATC-induced C>G transversions, connecting ATC treatment to mutation to malignant transformation. In summary, these findings corroborate the mutagenic effect of ATC and identify a B cell precursor population in Rag1-KO mice that is capable of leukemic transformation upon treatment with ATC.