It was our colleague’s clinical experiences evaluating CDK4/6 inhibition as a therapy for patients with well-differentiated and dedifferentiated liposarcoma that grabbed our interest and led to our current focus in SAGA. As is typical with new drugs, some patients do well and others do not. Being incorporated in their efforts to understand this es allowed us access to patient samples to determine whether CDK4/6 inhibitor induced senescence in cultures was of any clinical value. In seven patient samples collected pre- and post-treatment we were able to establish an association between MDM2 down-regulation and patient outcomes thus raising the possibility that SAGA was a clinically relevant mechanism of action for this class of drugs.Delineating the molecular mechanism of MDM2 regulation in quiescent cells induced by CDK4/6 inhibition led us to identify CDH18 and PDLIM7 as key players in this process. PDLIM7 binding to MDM2 inhibits its turnover, but PDLIM7 can be sequestered in CDH18 containing complexes away from MDM2. As expected CDH18 negative cells and cells engineered with a deletion in CDH18 failed to undergo senescence when treated with CDK4 inhibitors but would still exit the cell cycle into quiescence. Thus, we developed a CDH18 detection assay for use on clinical materials from patients who had enrolled in the CDK4/6 inhibitor trials and demonstrated that outcome was correlated to CDH18 status. This not only enforces the suggestion that SAGA and MDM2 down-regulation is an important clinical consideration for this drug, it opens the possibility that CDH18 might be a suitable biomarker to predict, pre-treatment, whether patients will respond well to this drug. We are currently expanding this into other diseases, such as breast cancer, and looking to carry out experiments to understand how CDH18 and PDLIM7 interactions are controlled, where CDH18 is localized, and how this might be manipulated for patient selection and combinatorial therapy.
But what about patients who are CDH18 negative or do not respond favorably to CDK4/6 inhibitors alone? We showed that ATRX dependent suppression of HRAS transcription was necessary for CDK4/6 inhibitor induced senescence, and thus we are developing cell line and xenograft models in liposarcoma and other cancer types to examine whether specifically manipulating HRAS activity might be an effective combinatorial therapy when applied sequentially to patients. We are also interested in examining new modalities that might be used in a sequential combinatorial fashion as we mine the data sets we obtain on SAGA and think about the other mechanisms of action advanced to explain how these drugs work in humans.
We will continue to weave our clinical colleagues into our basic science efforts to understand SAGA to bring new ideas and approaches in both directions and insure that we live up to the expectation of “More Science, Less Fear.”