Proteolysis-Targeting Chimeras as Tools in Drug Development (II

  • 3.2. Off-target protein degradation

    Multiple studies have shown that PROTACs are not entirely selective and can degrade proteins other than the primary target. Degradation of a protein that is not directly bound by the PROTAC can arise as a consequence of “bystander degradation” where the degradation of a protein that is not directly bound by the PROTAC becomes ubiquitinated and hence degraded as part of the same complex with the PROTAC POI. In other instances, this is an off-target effect due to the neo-morphic interactions with so-called neosubstrates, which become ubiquitinated by the E3 ligase and subsequently degraded. Off-target effect can also arise from the mere binary engagement of the target binding ligand of the PROTAC to other proteins in the same way as the POI. A well-characterized example of off-target activities with potential safety liabilities for PROTAC comes from the immunomodulatory drugs (IMiDs) thalidomide, pomalidomide and lenalidomide, which function as ligands for the CRBN E3 ligase. It is important to assess this safety risk using in vitro assays and/or incorporating relevant endpoints in a preclinical investigative toxicology study.  

    PROTAC molecules can induce the degradation of neo-substrates that cannot be predicted based on the small molecule E3 ligase ligands alone. Therefore, ideally monitoring the degradation of the global proteome following PROTAC treatment in cell models or tissues from in vivo pre-clinical studies should be performed. The current method of choice to study the dynamics of protein degradation in cellular systems is a multiplexed isobaric tandem mass tag (TMT) labelling approach for simultaneous identification and quantification of thousands of proteins. A comprehensive proteomics strategy towards investigating PROTAC safety will consider two elements: selectivity and mechanism. Selectivity entails characterizing the binding and degradation selectivity of the PROTAC of interest using affinity based chemical and global proteomics, respectively. Chemical proteomics profiling identifies the binding partners of the PROTAC of interest, as well as the specific E3 ligase and POI ligands (along with their respective proteomics Kd’s). This information enables assessing safety risks derived from the degradation profile linked to specific binary binding events. Global proteomics techniques have been broadly applied in the PROTAC field as a way to measure degradation selectivity as well as identifying previously unknown degradation off targets across a proteome of 8,000-10,000 proteins as a result of PROTAC treatment. Further, global profiling allows to monitor downstream protein dynamics and pathway effects by varying cell priming conditions and/or PROTAC treatment conditions. This parameter is particularly relevant when degrading transcription factors and epigenetic targets and delivers mechanistic insights that link degradation selectivity to in vivo observations, including adverse safety events. In addition to understanding proteome wide effects of PROTAC treatment, mechanistic parameters of interest include absolute quantitation and relative stoichiometry of E3 ligase and protein of interest levels, as well as the turnover rates of the target of interest in disease and safety relevant input material, both of which will inform mechanistic models for understanding and predicting ternary complex formation.

    Following identification of off-target proteins that are degraded by PROTAC, it is necessary to investigate the safety risks of these undesired effects through understanding the function of the degraded proteins and the physiological consequences of reducing their levels in different tissues.

    3.3. Disruption of cellular proteostasis

    Since PROTAC utilize an E3 ligase to initiate proteasomal degradation of a POI, it is possible that PROTAC-POI complexes compete with natural substrates for binding to the E3 ligase for ubiquitination and degradation, resulting in accumulation of those substrates and potentially perturbing specific cellular pathways (Figure 4). Furthermore, there is also a risk that PROTAC increase the cellular concentration of ubiquitinated proteins resulting in saturation of the proteasome, which could have consequences on cellular homeostasis. The proteasome controls the cellular content of proteins that regulate many aspects of cellular biology including cell cycle, cell growth, immune homeostasis and metabolic activity and therefore, alterations in proteasome activity and protein accumulation could have deleterious effects.

    While the concern about blocking natural E3 substrate and proteasome is theoretically relevant, experimentally it has not been observed at low PROTAC concentrations. One reason is that as PROTAC acts catalytically and so crucially sub-stoichiometrically, PROTAC will only engage to a sub-population of E3 ligase molecules, particularly in the case of abundant E3 ligases (as is the case with CRBN and VHL). This means that a large population of E3 ligase molecules remain unbound by the PROTAC, and can carry on their natural activity against their native substrate molecules, with little to no effective inhibition being observed. Thus, further in vivo studies to better understand the pharmacokinetic/pharmacodynamic relationship vs the POI and key off-targets are therefore required.

    3.4. Implications of the “hook” effect

    At high concentrations, PROTAC saturates binding to the target and to the E3 ligase resulting in the formation of binary complexes instead of the productive ternary complex (E3-PROTAC-Target). This prevents target ubiquitination and degradation, a situation that has been observed with many PROTACs and has been termed the “hook” effect. In theory, the formation of binary complexes could have two adverse consequences. The first concern is an increase in off-target degradation activity via binding of the E3-PROTAC binary complex to lower affinity off-targets (Figure 5). The second concern is the potential for a pharmacological response driven by the PROTAC-POI binary complex interaction that is different from target degradation. This scenario would be realized, for example, if the POI ligand (and resulting PROTAC) had agonist activity. The hook effect is an intrinsic property of any PROTAC and the underlying mechanisms in terms of ternary complex formation is generally understood. However, while ternary complex formation is an essential step for PROTAC mediated protein degradation, this is not sufficient to induce degradation. Consequentially, the onset of the hook effect observed from binding experiments may be very different from what is observed in cell-based degradation assays.

    Conclusion

    PROTAC has progressed from an academic tool to degrade protein into a potential therapy in about 20 years. Many pharmaceutical and biotech companies have invested in PROTAC and molecules with this mechanism of action have now entered clinical trials. Routes to assess and better understand PROTAC safety risks have been identified, which is an essential step towards delivering efficacious and safe PROTAC to patients.

    References

    [1] Sakamoto KM, Kim KB, Kumagai A, et al. PROTACs Chimeric molecule that target proteins to the Skp1-Cullin-F box complex. Proc Natl Acad Sci USA 2001, 98:8554–8559.

    [2] Liu Jing, Ma Jia, Liu Yi et al. PROTACs: A novel strategy for cancer therapy. [J] Semin. Cancer Biol., 2020, undefined: undefined.

    [3] Moreau Kevin, Coen Muireann, Zhang Andrew X et al. Proteolysis-targeting chimeras in drug development: A safety perspective. [J] Br. J. Pharmacol., 2020, 177: 1709-1718.

    [4] Burslem George M, Crews Craig M, Proteolysis-Targeting Chimeras as Therapeutics and Tools for Biological Discovery. [J] Cell, 2020, 181: 102-114.

    [5] Naito Mikihiko, Ohoka Nobumichi, Shibata Norihito et al. Targeted Protein Degradation by Chimeric Small Molecules, PROTACs and SNIPERs. [J].Front Chem, 2019, 7: 849.