Abstract Recent large sequencing and cancer dependency studies have accelerated the identification of candidate targets for precision medicine. However, the current drug development paradigm starting with target identification and validation can be slow and has thus far yielded a limited variety of successful targets. We sought to return to an empirical approach to drug discovery and performed a high throughput screen to identify small molecules that were both potent and selective. In a primary screen of 2000 compounds in two cell-lines: A549 and H1734, three compounds only affected H1734 viability. One of which validated in a dose-response experiment with great potency and specificity, we called this small molecule ‘Compound 1B’. In an effort to identify the target of Compound 1B, we profiled 766 genomically-characterized cancer cell lines and found that approximately 4% were sensitive to our compound. Sensitivity was not restricted to a particular tissue of origin. Interestingly, expression of Phosphodiesterase 3A (PDE3A) correlated with cytotoxicity. We further showed that Compound 1B specifically inhibited the enzymatic activity of PDE3A and PDE3B in a panel of 11 different phosphodiesterase family members. However, only a subset of other PDE3 inhibitors shared the same cytotoxic phenotype of Compound 1B. In a rescue screen of 1600 bioactive compounds, we identified the non-lethal PDE3 inhibitors as compounds that were able to rescue cell death induced by Compound 1B. Biochemical assays showed that both Compound 1B, cytotoxic and non-cytotoxic PDE3 inhibitors compete for binding to PDE3A. Knockdown of PDE3A did not affect cell viability and inhibited response of sensitive cell lines to Compound 1B. Thus we have identified a potent and selective small molecule that likely acts through PDE3A to induce cancer cell-line cytotoxicity. Our data suggest a hyper- or neomorphic function of PDE3A induced upon binding of Compound 1B. By cross-referencing integrative datasets with compound-sensitivity data, we show that reversal of the current drug-development paradigm can elucidate novel cancer targets, which are not yet identifiable by analysis of large next-generation sequencing datasets. Citation Format: Luc M. de Waal, Tim A. Lewis, Lara Gechijian, Aviad Tsherniak, Willmen Youngsaye, Matthew Rees, Oliver Mikse, Mark Hickey, Patrick Faloon, Nicola Tolliday, Angela Koehler, Monica Schenone, Kwok Wong, Alykhan Shamji, Benito Munoz, Stuart L. Schreiber, Heidi Greulich, Matthew L. Meyerson. An integrated genomic characterization of the target of a small molecule identifies a novel cancer dependency. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4596. doi:10.1158/1538-7445.AM2014-4596
Read full abstract