Yi-han Chien1, You-Ting Chen1, Judith Röske2, Kaixuan Zhang2, Graham P. Marsh3, Sebenzile Myeni4, Hannah J. Maple3, Alex Moloney3, Marjolein Kikkert4, Rolf Hilgenfeld2,5, Mark Brönstrup1,6, Katharina Rox1,6
1Department of Chemical Biology (CBIO), Helmholtz Centre for Infection Research (HZI), 2Institute of Molecular Medicine, University of Lübeck, 3Bio-Techne (Tocris), 4Leiden University Medical Centre, 5German Center for Infection Research (DZIF), 6German Center for Infection Research (DZIF)
Objectives: Proteolysis-targeting chimeras (PROTAC) are novel chemical concepts taking advantage of the proteasome ubiquitination system to degrade the protein of interest (POI). Compared to oncological PROTACs, which have already entered clinical trials [1], the development of PROTACs against infectious diseases has only recently accelerated [2]. When designing PROTACs, potent inhibitors of the targeted protein are commonly deployed as the POI binding ligand. With the viral protein as the main target for antiviral drugs, covalent inhibitors (both reversible and irreversible) are increasingly explored due to less toxicity concerns. Thus, covalent inhibitors speak for using them as the POI binding ligand in anti-viral PROTAC design [3]. Currently, PROTACs are mostly equipped with a reversible non-covalent POI binding ligand. To date, the consequences of a more robust or irreversible binding to the target protein with respect to degradation efficiency remain obscure. Methods: In this study we extended the current mechanistic model of reversible non-covalent PROTAC to study the differences between antiviral PROTACs, deploying reversible non-covalent, reversible covalent and irreversible binding ligands in inducing degradation as well as the downstream phenotypic effect, mechanistically. The extended model is verified using one of our in-house PROTAC compound PROTAC-1. Simulations and comparisons were conducted using hypothetical PROTACs equipped with different published inhibitors as POI binding ligands, covering different binding mechanisms, as real-case scenarios. To inform the in vivo antiviral efficacy, the mechanistic PD model is linked to a viral dynamics model, constructed based on viral load data published for hACE2-K18 mice, which is a mouse model used to mimic human SARS-CoV-2 infection [4]. A compartmental PK profile was constructed for PROTAC-1 based on the in-house PK data from mice, and was linked to the PD model as well as the viral dynamics model, subsequently. The in-house in vivo efficacy data of PROTAC-1 was used to verify the simulation result. The modeling work was conducted using package nlmixr2 in R [5]. Results: The degradation effect, as well as the downstream phenotypic effect were compared between the different binding mechanisms using the hypothetical PROTACs. It was found that a more sustained binding to the target protein does not significantly improve the maximum degradation. Furthermore, based on the simulation, redesigning an irreversible inhibitor into a PROTAC would lead to the least improvement in EC50 fold change, with the majority of the phenotypic effect being contributed by the inhibition effect of the PROTAC. It suggests that the improved degradation effect by forming an irreversible covalent bond could be phenotypically overshadowed by the as well improved inhibition effect of a PROTAC. The superiority of a PROTAC over its parent inhibitor is thus clearer if it binds reversibly. The viral dynamics model was constructed based on viral load data reported for hACE2-K18 mice [4]. The model was constructed in a way that some of the vital parameters are estimated in units that would enable comparisons to human’s [6] and can be used for further translational purpose. Despite of the valid in vitro POI degradation induced by PROTAC-1, the constructed model does not predict a significant anti-viral effect in vivo following the applied dosing regimen. This prediction comply with our in-house in vivo efficacy observations of the compound. Using the model, an intensified dosing regimen, characterized by more frequent dosing and higher dose amounts, is recommended in order to observe a significant difference on the viral load from the vehicle control at the planned sampling time point. Conclusions: The extended modelling framework compares different target protein binding mechanisms in PROTAC design mechanistically. It offers feedback on compound design and can be used to have an initial assessment with regard to degradation efficiency using limited affinity data of the PROTAC compound, potentially minimizing unnecessary experimentation and accelerate the compound screening process. By linking the PK/PD model to a viral dynamics model, the in vivo efficacy of an anti-viral PROTAC can be predicted. This offers valuable feedback on the compound selection process and on guiding the dosage regimen. Furthermore, the viral dynamic model was constructed for a mouse model that was deigned to mimic human disease, and the efficacy results obtained in this mouse model provide valuable insights for translating therapeutic effects to humans. Finally, to our knowledge, this study is the first one to validate simulation results with in vivo experimental models in the field of antiviral PROTACs.
[1] Békés M, Langley DR, Crews CM. PROTAC targeted protein degraders: the past is prologue. Nature Reviews Drug Discovery. 2022;21(3):181-200. [2] Liang J, Wu Y, Lan K, Dong C, Wu S, Li S, et al. Antiviral PROTACs: Opportunity borne with challenge. Cell Insight. 2023;2(3):100092. [3] Boike L, Henning NJ, Nomura DK. Advances in covalent drug discovery. Nature Reviews Drug Discovery. 2022;21(12):881-98. [4] Dong W, Mead H, Tian L, Park J-G, Garcia JI, Jaramillo S, et al. The K18-Human ACE2 Transgenic Mouse Model Recapitulates Non-severe and Severe COVID-19 in Response to an Infectious Dose of the SARS-CoV-2 Virus. Journal of Virology. 2022;96(1):e00964-21. [5] Fidler M, Wilkins JJ, Hooijmaijers R, Post TM, Schoemaker R, Trame MN, et al. Nonlinear Mixed-Effects Model Development and Simulation Using nlmixr and Related R Open-Source Packages. CPT: Pharmacometrics & Systems Pharmacology. 2019;8(9):621-33. [6] Gonçalves A, Bertrand J, Ke R, Comets E, de Lamballerie X, Malvy D, et al. Timing of Antiviral Treatment Initiation is Critical to Reduce SARS-CoV-2 Viral Load. CPT: Pharmacometrics & Systems Pharmacology. 2020;9(9):509-14.
Reference: PAGE 33 (2025) Abstr 11372 [www.page-meeting.org/?abstract=11372]
Poster: Drug/Disease Modelling - Infection