Title. Bad leaving groups make better strategy: KL-50, DNA methyltransferases, and drug-resistant glioblastoma.

Abstract. Enzymes that repair DNA damage are expressed in healthy tissue but display reduced activity in many cancers. O⁶-Methylguanine-DNA methyltransferase (MGMT) is one such enzyme that converts O⁶-alkylguanosine to guanosine by nucleophilic displacement. MGMT is silenced in approximately half of glioblastomas and 70% of lower-grade gliomas. Patients with MGMT-silenced glioblastoma typically receive surgery followed by radiation therapy and the DNA alkylation agent temozolomide. Temozolomide generates O⁶-methylguanosine under physiological conditions and the drug’s mechanism of action relies on an intact DNA-mismatch repair pathway. The treatment provides just a ~7 mo. survival benefit since acquired resistance via silencing of the mismatch repair pathway is common. More than 95% of these patients die within 5 years.

Here I will describe the design, synthesis, and mechanism of action of the first MGMT-dependent, mismatch repair-independent DNA alkylation agent, a novel fluoroethyl imidazotetrazine (KL-50). KL-50 produces DNA interstrand cross-links specifically in MGMT-silenced cells. These interstrand cross-links form by initial production of O⁶-(2-fluoroethyl)guanosine, followed by a slow cyclization to an N¹,O⁶-ethanoguanine intermediate, and a similarly slow ring opening by the adjacent thymidine base. The selectivity of KL-50 derives from the faster relative rate of reversal of O⁶-(2-fluoroethyl)guanosine in healthy cells. KL-50 is efficacious and well-tolerated in animal models of temozolomide-resistant glioblastoma. The structural similarity to temozolomide also provides optimism that it may be translatable to the clinic. Integrating the relative rates of chemical DNA modification and biochemical DNA repair into design of chemotherapies may provide opportunities for the treatment of other cancers harboring specific, tumor-associated DNA repair defects.