Synthesis and SAR Studies of novel thiazolidine analogs for melanoma

3963

Introduction: Currently there is no effective pharmacological treatment for advanced melanoma. Exploring novel agents that interfere with cell signaling pathways could reduce drug resistance and provide a better therapeutic index.

Methods: We first screened a selected library consisting of 91 compounds. We tested their effects on two melanoma cell lines (SKMEL-188 and WM-164), and on fibroblasts served as control (model of stromal cell that ideally should not be affected by anti-neoplastic treatments). We also purchased dacarbazine (DTIC), Taxol, and synthesized Sorafenib. We use these drugs for assay control and as standards to assess our compounds. Cell viability after compound treatment for 48 hours was measured using the sulforhodamine B assay. To understand the mechanism of actions, we performed lactate dehydrogenase (LDH) release assay, caspase activation assay and measured DNA content with flow cytometry. We also assessed phosphorylation of MAPK and AKT with western blot. Metabolic alternations were studied with ¹H and ³¹P NMR.

Results: The initial screening identified three potent lead structures, but only thiazolidine structure displayed both high potency (IC₅₀ is 0.57 uM) and selectivity. DTIC is not active due to the lack of bioactivation in vitro. Taxol has a lower IC₅₀, but it only inhibited cell growth and did not kill cancer cells even at high concentration. In contrast, thiazolidine analogs effectively killed cancer cells. Sorafenib has been used extensively in clinical trials for melanoma. It is about 10 times less potent than our lead compound (IC₅₀ is 5 uM), but it has a higher selectivity against cancer cells (25, compared with 10 for our lead compound). This is consistent with the fact that Sorafenib is well tolerated in patients in clinical trials for melanoma. Extensive synthesis and SAR studies of new thiazolidine analogs improved the selectivity while largely preserved their potency. Western blot analysis indicated that thiazolidines inhibit phosphorylation of MAPK by ~60%. pAkt levels were well below the detection limit for all three cell lines. These indicated that neither MAPK nor AKT pathway are the major target for thiazolidines in melanoma. Thiazolidine analogs induced LDH releases, did not activate caspases, and caused DNA fragmentation pattern characteristic of necrosis. Hence they mainly induce necrotic rather than apoptotic cell death. ¹H and ³¹P NMR studies of extracted metabolites indicated metabolic alternations in phospholipids upon incubation with the compounds, indicating that thiazolidines may interact with receptors in the cell membrane. Conclusions: these novel thiazolidine compounds are highly active against melanoma in vitro. Further optimization for future in vivo testing and mechanistic studies are in progress.