Synthetic Lethality Targeting the Aid/RAD51 Axis for Treatment of AICDA-Expressing Lymphoma

Activation Induced Cytidine Deaminase (AICDA or AID) is a DNA directed cytidine deaminase that induces somatic hypermutation and immunoglobulin class switching in stimulated B-lymphocytes. AICDA is a DNA damaging enzyme, producing DNA base pair mismatches that can subsequently be converted into mutations, DNA single strand breaks (SSB), or DNA double strand breaks (DSB). Although AICDA is normally expressed at high levels only in activated B-cells, a wide range of cancer types show constitutive overexpression of AICDA. In malignant cells, as in B-cells, AICDA promotes hypermutation and genomic instability. We previously demonstrated that cells expressing AICDA are critically dependent on the DNA repair factor RAD51 for their survival. We have developed novel small molecule inhibitors of RAD51 that selectively kill AICDA overexpressing cells. One of these, CYT01A, depletes the nucleus of RAD51, thus reducing DNA double strand break repair activity. We have shown CYT01A to be a potent (EC50~150nM) RAD51 inhibitor that reduces the efficiency of RAD51 focus formation, leading to sensitization to excess DNA damage while preserving baseline recombination/repair activity. Here we present the in vitro and in vivo pharmacologic characterization of CYT01A. We first analyzed potential off-target liabilities by carrying out kinome inhibition and safety panel (Panlabs) screens. CYT01A showed negligible inhibitory activity against 429 kinases, receptors, or other pharmacologically important non-RAD51 targets. These data are consistent with high specificity for AICDA and predict in vivo tolerability. We next measured metabolic stability of CYT01A in mouse liver microsome incubation assays, demonstrating CYT01A stability in both the absence and presence of NADPH. Measurement of Caco-2 cell permeability, as a predictor of oral bioavailability, revealed low to moderate efflux of CYT01A. Consistent with these permeability data, CYT01A showed modest oral availability, but relatively long clearance times. Daily oral dosing at 80 mg/kg for 7-14 days resulted in significant reduction in AICDA expressing lymphomas, in a human-to-mouse xenograft model. This demonstrates efficacious iin vivo exposure despite relatively low oral availability. By contrast, AICDA null lymphomas showed no significant tumor response versus vehicle controls, consistent with the synthetic lethal relationship between AICDA and RAD51. Finally, we demonstrated that repeat daily administration of CYT01A up to the maximum dose tested (80mg/kg) was well tolerated and resulted in no observable changes in complete blood counts (CBC), bone marrow function or output, or body weights. Taken together these data represent a definitive proof-of-principle demonstration that in vivo targeting of RAD51 is an effective strategy to treat AICDA expressing cancers. These data form the basis for continued preclinical development of an AICDA/RAD51 synthetic lethal therapeutic paradigm that may be applicable to a wide range of both solid tumors and hematologic malignancies.