MDR1 Mediated Drug Resistance to a Histone Methyltansferase Inhibitor (KMT)

Aberrant histone methylation is emerging as a major contributor to oncogenesis, and inhibitors of histone lysine methyltransferases (KMTs) are starting to enter early clinical studies. Currently, inhibitors for two KMTs are in phase I/II clinical trials, an inhibitor of the histone 3 lysine 79 (H3K79) methyltransferase DOT1L for MLL-rearrange leukemias (NCT02141828), and inhibitors for the H3K27 methyltransferase EZH2 for lymphomas or solid tumors with activating EZH2 mutations (NCT01897571 and NCT02082977). The next few months and years are likely to see more KMT inhibitors entering clinical development, such as compounds directed against the H3K36 methyltransferase MMSET/NSD2/WHSC1, which is commonly translocated in multiple myeloma and carries activating mutations in a subset of acute lymphoblastic leukemia. Typically, KMT inhibitors are structurally similar and are derived from the methyl donor S-adenosyl-methionine (SAM). Very little is known about possible mechanisms of resistance to these compounds.

We exposed a panel of leukemia cell lines to varying concentrations of the DOT1L inhibitor EPZ4777 and assessed the decrease of H3K79 di-methylation by Western Blotting. Our analysis revealed a dose-dependent decrease of H3K79 dimethylation in most cell lines. Interestingly, we identified a pair of cell lines, KG-1 and KG-1a, which were near completely resistant to EPZ4777 mediated DOT1L inhibition, with no effect of 10 μM EPZ4777 (4 day exposure). Even after exposure to EPZ4777 for 15 days, KG-1 showed only a slight decrease in H3K79me2 marking and no decrease in marking was seen for KG-1a up to a concentration of 3 μM. We searched the encode database for expression levels of DOT1L, as well as higher expression of proteins previously described as mediating drug resistance in AML cell lines. We found no significantly higher DOT1L expression level, suggesting that overexpression of the target is not the predominant mechanisms of resistance in KG-1/KG-1a. We confirmed similar baseline H3K79 methylation levels in KG1/KG1a and the other cell lines in our panel by Western Blotting. However, we noted a higher expression of multidrug resistance protein 1 (MDR1/p-glycoprotein/ABCB1) in the KG-1 and KG-1a cell lines compared to the other cell lines in our panel. To explore whether DOT1L inhibition by EPZ4777 could be restored in KG1 and KG1a cells when MDR1 was inhibited, cells were exposed to 0.3 μM EPZ4777 and varying concentrations of Cyclosporine A (CSA), a known inhibitor of MDR1. Increasing concentration of CSA with a fixed dose of 0.3 μM EPZ4777 led to a dose dependent decrease in H3K79me2 in KG-1/KG-1a, suggesting that inhibition of MDR1 increased the intracellularly effective dose of EPZ4777. CSA at a concentration of 0.3 μM and 0.3 μM EPZ4777 near completely abolished H3K79 methylation. In summary, our results suggest that KMT inhibitors can be subject to classic drug resistance mechanisms such as the MDR1/p-glycoprotein/ABCB1 efflux pump. These data provide insight for the use and design of small-molecule inhibitors for KMTs.