Hijacking the ubiquitin and autophagy system for the development of chemical degraders

Targeted protein degradation (TPD) marks an emerging novel technology within the chemical biology and drug discovery communities. The classical drug discovery process has focused on target driven inhibition, where a small molecule is designed to bind to the protein of interest (POI) and then interfere with its function. For enzymes, inhibitors are often designed to target the active site (orthosteric), inhibiting the protein´s enzymatic activity by cofactor or substrate competitive binding. However, approximately 75% of human proteins lack typical active sites and are considered “undruggable” by conventional inhibitors, including a large number of proteins linked to diseases, such as e.g. the transcription factors p53 and MYC. TPD is an emerging approach capable to target a much larger fraction of the human proteome. Here, target proteins are not inhibited at a functional binding site but they are degraded, leading to a complete elimination of all protein functions, including enzymatic activities and scaffolding functions, which is impossible to achieve by conventional inhibitors. To target a protein for degradation, so-called PROteolysis TArgeting Chimeras (PROTACs) can be used, which comprise bi-functional small-molecules containing two binding moieties, connected via a chemical linker. This arrangement induces proximity between both targeted proteins. In PROTACs, one of the targeted proteins marks an E3 ligase, while the other ligand recruits a POI as a non-natural substrate (neosubstrate) to the E3 ligase. Since E3 ligases act as substrate recruitment sites of the ubiquitin system, the chemically induced recruitment to the E3 can trigger ubiquitination and subsequent degradation of the POI. To undergo degradation, a complex and highly controlled mechanism known as the ubiquitin proteasomal system (UPS), consisting of a cascade of enzymatic reactions for ubiquitin transfer, is required. Due to the complexity of the mechanism, many developed proximity-inducing compounds do not lead to final protein degradation because they fail to mediate all required steps in the UPS cascade. Therefore, this thesis focused on the investigation why and where PROTAC mediated degradation can fail and how to rationally optimize small molecule degraders. To address this question, bespoke cellular assays for each of the different key steps within the PROTAC-mediated degradation cycle have been developed. Starting with the establishment of assay systems for binary complexes, a model system focusing on the baculoviral IAP repeat-containing (BIRC) proteins was established using a variety of biochemical measurements and the development of a cell-based selectivity screening platform. This system was crucial for the validation of the live-cell assays using NanoBRET-based technologies which were subsequently developed and yielded good correlation between biophysical in vitro assays and live-cell assays. Therefore, live-cell assays were chosen as the most promising assay system for the evaluation of novel degrader and E3 ligase ligands (e.g. for GID4), since this assay system additionally enabled information about the fraction of compounds entering living cells. Moreover, live cells allowed the use of full-length protein together with all binding partners which was previously shown to be crucial for robust small molecule-protein interaction (e.g. the CRBN-DDA1-DDB1 complex). Using NanoBRET experiments evaluating a variety of PROTACs, cell penetration was identified as an important parameter which was successfully assessed by measuring live-cell target engagement in comparison with permeabilized cells. Using this assay system, a relative ratio of cell penetration was determined which correlated well with data measured on well-established but work-intensive systems (e.g. Caco2 or PAMPA assay). The use of BRET-based assay technology also enabled measurements in relevroperties, demonstrated by significantly weaker live cell target engagement compared to permeabilized cells. During this work, it became obvious that a database for tracer molecules – ligands of POIs labelled with fluorescent dyes – which are necessary key reagents for cellular target engagement assays (e.g. for TR-FRET or NanoBRET), would have significantly sped up the assay development process. Therefore, an online database, called “tracerDB” was developed. Within this crowdsourced project, we established a repository of experimentally validated tracers that can be identified in a web-based searchable format. In addition, tracers developed by the community can be submitted which resulted in worldwide submissions from numerous institutions. Generally, this database project received an overall great acceptance and support from the scientific community. Due to the weak binding of some PROTACs, for one of the POIs (WDR5) a tracer-free alternative technique called HiBiT-based cellular thermal shift assay (HiBiT-CETSA) was established for binary complex assessment. HiBiT-CETSA (also called BiTSA), a Split-NLuc system, measures the thermal stability of HiBiT-tagged target proteins which in theory increases when bound by a compound. The increase in stability of the protein-compound complex compared to the unbound state is a measure for the interaction affinity of the compound with its target protein. In the case of WDR5, HiBiT-CETSA was found to be an excellent tool to rank PROTACs according to their affinity. However, due to the high melting temperature, it was not possible to unambiguously distinguish between measurements in intact and already thermally lysed cells.