PRosettaC: Rosetta based modeling of PROTAC mediated ternary complexes

Abstract

Proteolysis-targeting chimeras (PROTACs), which induce degradation by recruitment of an E3 ligase to a target protein, are gaining much interest as a new pharmacological modality. However, designing PROTACs is challenging. Formation of a ternary complex between the protein target, the PROTAC and the recruited E3 ligase is considered paramount for successful degradation. A structural model of this ternary complex could in principle inform rational PROTAC design. Unfortunately, only a handful of structures are available for such complexes, necessitating tools for their modeling. We developed a combined protocol that alternates between sampling of the protein-protein interaction space and the PROTAC molecule conformational space. Application of this protocol -PRosettaC -to a benchmark of known PROTAC ternary complexes results in near-native predictions, with often atomic accuracy prediction of the protein chains, as well as the PROTAC binding moieties. It allowed the modeling of a CRBN/BTK complex that recapitulated experimental results for a series of PROTACs. PRosettaC generated models may be used to design PROTACs for new targets, as well as improve PROTACs for existing targets, potentially cutting down time and synthesis efforts.

Current strategies for the design of PROTAC linkers: a critical review

“…Group epitope mapping is a well-established method and has proven reliable in determining solvent exposed positions in a range of VHL ligands [155]. Computational modelling of the TC as an alternative to experimental techniques to inform rational design is also likely to see significant expansion in the future, and an increasing number of groups are developing methods to do this [156,157]. The continued development of TC modelling may provide avenues for linker development in lieu of significant empirical experimental optimisation.…”

Improved Accuracy for Modeling PROTAC-Mediated Ternary Complex Formation and Targeted Protein Degradation via NewIn SilicoMethodologies

“…Occasionally, some true positives are grouped in a cluster separate from the largest double cluster -for example, for 6HR2, 16 crystal-like poses do not belong to the largest double cluster (15 belong to the 4 th -largest cluster, and one is a singleton) -and thus these desirable true positive results are discarded when only the largest double cluster is considered. While it is certainly possible to evaluate smaller double clusters as well, as has been done elsewhere, 42 unless additional experimental information is available for use to evaluate these clusters, it seems likely that more noise than signal will generally result from this more expansive approach. Moreover, as can be seen in (Table 1), whereas the accompanying darker bars show the populations within the largest L3xP10 double cluster.…”

“…In 2019 we reported 41 a suite of computational modeling tools to facilitate PROTAC design, thus enabling the cost-and time-saving advantages of an iterative design campaign guided by in silico modeling results. The goal of our original work was to develop a generally applicable PROTAC modeling protocol, i.e., one that could give accurate results across different protein targets, E3 ligases, and linkers; such an approach had been absent from the literature at that time, with all previous PROTAC modeling studies limited to a specific TPD system (although preprints following our model have recently appeared 42,43 ). To that end, we established "Method 4" as the most accurate approach for both VHL-and cereblon-based PROTACs.…”

“…We expect that not only will the specific results and techniques presented below prove useful for researchers using our proposed Methods, but moreover many of the findings discussed herein should be general and transferrable, so that the accuracy of independently developed PROTAC modeling procedures, such as those recently described by Li et al 44 or elsewhere in preprint manuscripts,. 42,43 can be further refined.…”

Atomic-Resolution Prediction of Degrader-mediated Ternary Complex Structures by Combining Molecular Simulations with Hydrogen Deuterium Exchange

“…14,[22][23][24][25] Previous work to computationally predict ternary structures mostly consists of protein-protein docking protocols, perhaps followed by refinement of the initial structures with molecular dynamics (MD) simulations to assess the stability of the predicted models. [24][25][26][27][28][29][30] However, these docking protocols fail to predict high-resolution structures (sub-2.0Å) with high fidelity. That is, while protein-protein docking protocols have shown some promise in generating structural models of ternary complexes with reasonable resolution (often characterized as sub-10Å RMSD to an x-ray structure), the best structures typically fall somewhere within a long list of possible poses (often in the hundreds or thousands), demonstrating the challenge associated with the selection of high-accuracy ternary structure models.…”