WNT signalling is a major regulator of mammalian development and tumorigenesis through control of cellular functions such as proliferation and differentiation. Many canonical WNT pathway mutations occur in tumours at or upstream of b-catenin; thus, if signalling is blocked at or below b-catenin, an inhibitor should be active against multiple tumours driven by WNT-activating mutations. With this in mind, we employed a cell-based screen to identify small molecules that block WNT signalling at or downstream of b-catenin. With no prior knowledge of the molecular target we used hypothesis-driven medicinal chemistry optimization to progress a series of 3,4,5-trisubstituted pyridines to a potent orally bioavailable small-molecule WNT-pathway inhibitor with <i>in vivo</i> activity. Knowledge of the potency and structure-activity relationships gained during the series optimisation allowed us to use a chemo-proteomics strategy to find the molecular target. We identified the mediator-associated kinases CDK8/19, that were confirmed as targets by biophysical approaches including SPR and X-ray crystallography. We used phospho-proteomics to identify robust target engagement biomarkers, microarray gene expression profiling to demonstrate regulation of specific transcription factors and super-enhancer activity, and CRISPR gene knockout to show that CDK8 inhibition was the predominant driver of tumour cell responses. We subsequently discovered and optimised an additional chemically-distinct series that identified a 3-methyl-1H-pyrazolo[3,4-b]pyridine. Having two potent and selective exemplars from structurally differentiated chemical series in hand, we were well positioned to investigate whether dual CDK8/19 modulation had a sufficient therapeutic window that would justify the clinical development of these compounds. We established that both series had antiproliferative and antitumor activity in WNT-dependent models, but tolerability studies found severe multi-organ toxicity and no therapeutic window. In light of the these observations the clinical development of these series was halted. In this presentation I will give an overview of the target ID, validation and our learnings. Our data showed CDK8/19 inhibitors elicit multiple effects beyond WNT-signalling, suggest caution in progressing them to the clinic and that although targeting pleiotropic mechanisms may have the potential to overcome or prevent resistance, they come with an increased potential for toxic liabilities.<br/>

Lessons learnt from the discovery of CDK8/19 protein kinase inhibitors: From phenotypic screen to selective chemical probes

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.