Next-Gen Degraders & Molecular Glues

Targeted degraders have huge potential to broaden the druggable space but they come with unique safety considerations, in particular those molecules engaging the CRBN E3 ligase. This talk will showcase our in vitro screening cascade, which includes high-throughput luminescence-based assays to monitor specific proteins, fast proteomics to monitor wider proteome changes, assessment of toxicity in primary human hematopoietic stem and progenitor cells, as well as in vivo assessment via a CRBN knock-in transgenic model. It will end with a general discussion around novel E3 ligases and their safety implications.

Molecular properties influence oral bioavailability, but current practices often oversimplify this relationship. Parameters like clogP and TPSA alone are inadequate for guiding bRo5 drug design. While experimental measures such as ChromlogD and EPSA offer improvements, they fall short for lead optimization. A comprehensive set of lipophilicity and polarity descriptors across membrane regions, along with chameleonicity metrics, is crucial for the successful design of orally bioavailable macrocycles and PROTACs.

The talk will demonstrate how mechanistic PK/PD modelling can rationally guide the optimisation of targeted protein degraders (TPDs) by means of quantitative prediction of in vivo degradation profiles from in vitro degradation data, and its powerful impact on improving the experimental design and output of in vivo protein-degradation studies. Case examples will be shown i) applying PK/PD modelling for compound selection and progression, ii) estimating of target occupancy, iii) generating a translational strategy tailored for TPDs, and iv) investigating both expected and unexpected changes related to the unique mode of action of targeted protein degraders after repeated dosing.

Transcription factors are an exciting drug target class; however, they have proven difficult to drug with conventional methods. Recently, both small-molecule degraders and small-molecule activators of transcription factors have been described. Here, we describe our recent developments in small molecule-induced proximity of oncogenic drivers.

The degradation of age-related protein aggregates within cells is an exciting therapeutic approach. For the most common dementia, Alzheimer’s disease, the removal of tau aggregates is predicted to reverse cognitive decline. At TRIMTECH Therapeutics, we harness the innate properties of the ubiquitous E3 ligase TRIM21 to selectively degrade disease-causing aggregates, whilst leaving the healthy versions of these proteins untouched.

Dysregulation of E1A binding protein (EP300) has been implicated in oncogenesis, inflammation, fibrosis, and neurodegeneration via its activity as transcription factor co-activator. However, combined targeting of EP300 and its highly homologous paralog CBP results in dose-limiting toxicity that limits efficacy. Here, we describe our design of heterobifunctional degraders selective for EP300 over CBP and the optimisation of potency, selectivity, and PK/ADME resulting in improved in vitro and in vivo performance.

The field of induced-proximity pharmacology is a burgeoning area of drug discovery with a myriad of opportunities for the medicinal chemist. Design challenges for such proximity-inducing drugs are different from traditional rule-of-5 compliant small molecule drugs as they encompass several layers of selectivity, aspects of catalysis, and the need to achieve drug-like characteristics in an extreme physicochemical property space. In my lecture I am going to discuss emerging design principles, share case studies, and outline the remaining needs for further innovation.

Reprogramming gene expression holds substantial therapeutic potential, but current small-molecule modalities fall short. We develop a strategy based on chemically induced proximity that hijacks oncogenic kinases and acyl transferases to activate cell death genes ordinarily repressed by transcription factor BCL6 in lymphoma. The resulting molecules potently kill cancer cells while leaving healthy tissue unaffected, and they demonstrate the privileged role of molecular glues for targeting 'undruggable' transcription factors.

We discovered covalent ligands can function through a molecular glue mechanism to stabilize protein complexes in live cells. Target binding resulted in neo-protein-protein interactions that alter localisation of binding partners to modulate cell biology. I will describe a proof-of-concept example of how this covalent molecular glue mechanism can target specific cancer cells.

In this talk, I will present a novel and efficient computational approach to identify ligandable surfaces on human E3 ligases, aiming to expand the repertoire of E3 ligases. The approach has been prospectively validated on FBW7, where an allosteric pocket was identified and targeted with small molecules. Some of these molecules act as allosteric modulators, enhancing the degradation of c-Myc and c-Jun, revealing a new mechanism in targeted protein degradation.

I will be discussing the requirements of a successful molecular glue design platform which blends the speed and scalability of machine learning with the accuracy of physics-based methods. This can be used for optimal protein-protein matching and ternary complex structure prediction.