The Effect of TET Mutations on DNA Methyltransferase Cytotoxicity and ERV Induction in B and T Cell Malignancies

Abstract Cancers exhibit altered DNA methylation compared to normal cells, including methylation loss at regions normally silenced for genome stability and gain at promoter regions of tumor suppressor genes. 5-azacytidine (AZA) and 5-aza-2'-deoxycytidine are DNA methyltransferase inhibitors (DNMTi) that remove DNA methylation and are FDA approved for the treatment of myelodysplastic syndrome. We (Chiappinelli KB et al., Cell 2015; Li H et al., Oncotarget 2014) have shown that DNMTi upregulate immune signaling, including the interferon response, tumor antigens, and antigen presentation in solid tumors. DNMTi activate a canonical interferon signaling pathway through upregulated expression of dsRNA, specifically hypermethylated endogenous retroviruses (ERVs) that activate dsRNA sensors. The interferon response activated by ERV signaling recruits immune cells, promoting tumor clearance and sensitizing to immune therapy (Stone ML et al., PNAS 2017). We sought to determine whether DNMTis cause a similar ERV activation and cell killing in B and T cell malignancies, focusing first on mantle cell lymphoma (MCL). We further hypothesized that mutations in ten-eleven translocation methylcytosine dioxygenases (TET), enzymes that normally demethylate DNA, might lead to a hypermethylated state making cells more resistant to direct cytotoxicity from DNMTi. We also sought to test the effects of TET mutation on ERV induction and immunogenicity. For 5-Azacytidine (Aza) to inactivate DNMTs it must be incorporated into the DNA of actively dividing cells. Since Aza is only stable in the body for < 30 minutes and MCL cells have slow doubling times, only a fraction of cells divide during that 30 minute window. To improve response in these cancers, extended drug availability is necessary. We used three different treatment regimens to demonstrate that extending the availability of active drug shows a better response, significantly more so than increasing the drug dose or treating for a longer period of time. The treatments included: one dose a day for 3 days (3 Day Tx; blue), one dose a day for 5 days (5 Day Tx; red), and three doses a day for three days (3X3 Tx; green). The three doses a day were spaced three hours apart to ensure that most of the active drug was gone before the subsequent treatment was given. This was done for a dose response curve of 250, 500, 1000, 2000, and 4000 ηg/mL, compared to PBS control (0). Figure 1 shows the percentage of live cells at two time points, Day 6 and Day 9 after the start of treatment. In the TET wild type MCL cell lines (Maver, Mino, Jeko) the three treatment a day group (3X3) showed better killing when compared to one dose per day for three days or five days. The differences between the 3X3 treatment and the single dose per day was also more pronounced in the slowest growing cell lines. Mino, which showed the largest difference in killing, doubles between 50-72 hours, whereas Jeko (doubling time 26-33 hours) shows an appreciable but less drastic difference between the 3 dose versus 1 dose per day. This further supports the hypothesis that slower growing tumors benefit from sustained drug delivery rather than an increase in drug dose. TET mutated leukemia cell lines (Karpas 299, CCRF-CEM), were less AZA sensitive. This supports our hypothesis that TET mutation leads to a hypermethylated state more resistant to DNMTi. The degree of resistance correlates with the extent of the TET mutations. Karpas, the most resistant, has mutations in both TET1 and TET2, whereas CCRF only has a mutation in TET3. These results somewhat contradict recent reports that leukemias and lymphomas with TET2 mutations show higher response rates to AZA. However, this may be due to increased immunogenicity, rather than just direct cytotoxicity, which is supported by the increase in ERV expression shown in Figure 2 (ERVs profiled: Syncytin-1, ervFC2, and envW2). CCRF showed increased ERV expression in both the five day treatment and the 3X3 treatment. More extensive genomic ERV analyses spanning all 5 cell lines are currently underway and will be presented at the conference. In conclusion: (1) ERVs, a potential immunotherapy target, are expressed in mantle cell lymphoma and expression is increased by DNMTi treatment; (2) extended exposure to AZA appears to be important, and needs to be considered in clinical development; (3) TET mutations can direct personalized treatment where DNMTi can be used in conjunction with checkpoint blockade immunotherapy. Disclosures No relevant conflicts of interest to declare.