Vulnerabilities of MMR-deficient glioblastoma

Glioblastoma (GBM) is the most common and lethal brain cancer, with inherent or adaptive resistance to all existing treatments. One important mechanism by which GBM develops resistance to temozolomide, the frontline chemotherapy used in its treatment, stems from mutations in genes such as MSH6 and MSH2 critical in DNA mismatch repair (MMR)—a mechanism found in colon and other cancers as well. These MMR-deficient GBMs become hypermutated and particularly aggressive as well as resistant to many chemotherapies, and there is a pressing need to identify therapies that are effective against them. While a recent report describes two cases of pediatric MMR-deficient, hypermutated GBM that responded to immunotherapeutic checkpoint inhibitors, our own and others’ clinical experience has uniformly indicated a lack of responses in adult patients with MMR-deficient, hypermutated GBM treated with these agents. This proposal tests novel therapeutic approaches to MMR-deficient, hypermutated GBMs and uses unique tools to do so. Our preliminary studies expand on prior reports suggesting possible activity of calcium channel inhibition against MMR-deficient cancers to show for the first time that combining inhibitors of different calcium channels—carboxyamidotriazole (CAI), mibefradil, and verapamil—has preferential and synergistic activity against MMR-deficient GBM lines versus parental lines. In addition, our preliminary results further suggest TGF-β as a potential target in this setting, and that the chemotherapy drug irinotecan and the anti-cholesterol statins can be repurposed as TGF- β inhibitors with preferential activity against MMR-deficient GBM. These therapeutic strategies are being tested against matched sets of GBM lines that each include a parental line and an MMR-deficient line derived from it, as well as sets of GBM lines that spontaneously developed MMR deficiency and control GBM lines. The MMR- deficient lines each have MSH6 or MSH2 insufficiency derived either from mutations secondary to in vivo temozolomide treatment or from stable expression of shRNA. This proposal will use these matched parental/MMR-deficient GBM lines to test the effects of calcium channel blockade combinations and TGF-β inhibition in vitro and in vivo. In addition, given that MMR deficiency and hypermutation and both therapeutic approaches are likely to impact the anti-GBM immune response, we will also develop an immunocompetent mouse model of MMR-deficient GBM to test these effects. Both targeted and unbiased studies of mechanism will be performed, including assessing connections between the two therapeutic strategies and phosphoproteomic and RNA-seq analyses. The proposed studies will yield new biologic and therapeutic insights that could rapidly impact the treatment of MMR-deficient, hypermutated GBM and other cancers.