The Role of Cellular and Protein Context in Targeted Protein Degradation

Targeted protein degradation (TPD) is an ever-advancing tool for disrupting protein function in basic research and therapeutics. TPD harnesses cellular E3 ligases and protein degradation machinery to ubiquitylate and target proteins of interest (POI). Small molecule degraders induce proximity between endogenous E3 ligases and POIs for targeted POI ubiquitylation and degradation. Prototypic degraders, including molecular glues and proteolysis-targeting chimeras (PROTACs), rely on the formation of a productive ternary complex in cells between the POI, degrader and E3 to achieve POI ubiquitylation and degradation by the proteasome. It is becoming clearer how degrader design and the nature of the ternary complex influences TPD. However, inside the cell environment, characteristics of the E3s and POIs themselves could equally impact degradation efficiency. This thesis investigates the impact of POI localisation and its context on TPD. To explore how POI subcellular distribution affects TPD, dTAG and HaloTag POIs were first localised to the nucleus, cytoplasm, outer mitochondrial membrane, endoplasmic reticulum, Golgi, peroxisome, or lysosome. CUL4CRBN and CUL2VHL E3 ligases were then re-directed to degrade these POIs by using PROTACs targeting dTAG (dTAG-13 & novel degrader dTAG-VHL) or HaloTag (HaloPROTAC-E). POI degradation efficiency by dTAG-13 and dTAG-VHL varied with differential subcellular distribution, with no degradation observed at the peroxisome, lysosome or Golgi for both. There was a difference in the degradation efficiency of the same dTAG POI when using dTAG-13 (CUL4CRBN-targeting) or dTAG-VHL (CUL2VHL-targeting), suggesting the choice of E3 ligase impacts TPD. HaloPROTAC-E (CUL2VHL-targeting) efficiently degraded HaloTag POIs at all subcellular locations except the Golgi lumen. POI localised to the Golgi with the HaloTag repositioned to face the cytoplasm was subsequently shown to be efficiently degraded, demonstrating that POI and E3 accessibility are important cellular factors to consider in TPD. Both VHL and CRBN displayed mainly cytoplasmic localisation. Differential dTAG- and HaloTag-based degradation at the same POI location supports that the structure of the POI-E3 complex induced by the PROTACs impacts TPD efficiency. Often proteins exist and function as part of multi-protein or other macromolecular complexes, allowing a single protein to control different signalling networks. Consequently, bringing a POI and E3 into proximity by a degrader in a specific cellular environment may also recruit POI-associated proteins and lead to their codegradation. In the context of the casein kinase 1α isoform (CK1α), targeted degradation by lenalidomide led to co-degradation of FAM83F in complex but no other FAM83 protein, suggesting that FAM83F-CK1α complexes can be selectively targeted by lenalidomide. CK1α has several roles in the cell to control processes involved with cell death, cell cycle, proliferation, differentiation and motility. In vertebrates, CK1α is regulated through its interaction with the eight members (A-H) of the FAM83 family to elicit these cellular functions. So far, it has been shown that FAM83D-CK1α localises to the mitotic spindle, FAM83F-CK1α at the plasma membrane and FAM83G-CK1α complexes primarily in the cytoplasm which both activate canonical β-catenin/WNT signalling. Loss of FAM83G-CK1α interaction by FAM83G point mutations, that have been implicated in palmoplantar keratoderma skin disorder, results in a reduced WNT signalling response. This thesis has characterised a novel FAM83G mutation and shown loss of CK1α interaction and reduced WNT activity in patient-derived cells. Since lenalidomide could selectively target FAM83F-CK1α complexes, it was hypothesised that IMiD derivatives that could form different arrangements of the ternary complex may allow for TPD of other unique FAM83-CK1α complexes. Two 16 classes of lenalidomide-derived selective degraders of CK1α developed by the Woo (Harvard) and Rankovic (St. Jude) labs were characterised in this thesis to assess codegradation of different FAM83-CK1α complexes. Several FAM83 proteins were codegraded by these compounds with efficiency comparable to CK1α degradation. FAM83 degradation was mediated through recruitment of CK1α to CUL4CRBN E3 in a dose- and time-dependent manner. FAM83F and FAM83G were most consistently codegraded with CK1α by all compounds tested. Moreover, when tested in PPK patientderived fibroblasts, FAM83GR265P protein was not degraded while CK1α was, providing strong evidence that the loss of FAM83G-CK1α interaction in this context affected the ability of IMiDs to co-degrade FAM83-CK1α complexes. Overall, this thesis demonstrates that the cellular and protein context play crucial roles on TPD. Unique features of the POI context can dictate the degradation efficiency and outcome. From a clinical perspective, it is very important to understand basic POI function within a specific cellular environment to be aware of possible indirect effects of and on TPD technologies. From a research perspective, differential TPD effects may lead to degraders that can target specific protein complexes. In the context of FAM83-CK1α biology, degraders that selectively degrade one FAM83 member together with CK1α, like that of lenalidomide, but no other FAM83-CK1α complex, could be employed to target a selective function of one FAM83-CK1α complex without affecting functions of other complexes. This may address the challenge of targeting isoform-specific CK1 cellular roles in both basic research and therapeutics.