TARGET DISCOVERY FROM PHENOTYPIC SCREENS AND KNOWN COMPOUNDS OF UNKNOWN TARGET

Peter Juhasz

Biogen, Cambridge, MA, USA

Abstract

Phenotypic screen based drug discovery is an attractive alternative to target-based efforts as it can yield druggable pathways and targets that would not be characterized, in the canonical sense, as such. A key component of phenotypic screen based discovery programs is identification of pathways or the molecules that are targeted by the screen hits and mediate the phenomenology explored by the screen.

Available tools to deconvolve targets of small molecules come in a wide variety that enables: a) unbiased profiling of protein binding (chemoproteomics, protein arrays, Y3H) or functional effects (siRNA/CRISPR screens, various omics and profiling techniques) and b) validation of hypotheses emerging from these profiling efforts. Successful targetID typically requires a unique mix of these approaches that rarely repeats from one program to another – leaving lot of room to scientific intuition and creativity. To illustrate the concepts we report a study on regulators of non-canonical NF-κB signaling wherein ~150k compounds were screened in search if pathway inhibitors while counterscreening against inhibitors of the canonical pathway. During target deconvolution we have discovered a novel role for cyclin dependent kinase 12 (CDK12) in transcriptionally regulating the non-canonical NF-κB pathway. A novel compound was found which inhibits lymphotoxin β receptor (LTβR)- and FN14-dependent p52 nuclear translocation, but not TNFα receptor (TNFR)-mediated, canonical NF-κB p65 nuclear translocation. Chemoproteomics identified cyclin dependent kinase 12 (CDK12) as the target of this compound. siRNA screen and pharmacological modulation have confirmed this hypothesis. The experiments have shown that our molecule affects global transcriptional changes and prevent LTβR and FN14-dependent MAP3K14 (NIK) mRNA induction and subsequent protein accumulation through reduced phosphorylation of RNA Pol II CTD.