Design, Synthesis, and Biological Evaluation of Potent and Selective PROTAC Degraders of Oncogenic KRASG12D.

KRASG12D, the most frequent KRAS oncogenic mutation, is a promising target for cancer therapy. Herein, we report the design, synthesis, and biological evaluation of a series of KRASG12D PROTACs by connecting the analogues of MRTX1133 and the VHL ligand. Structural modifications of the linker moiety and KRAS inhibitor part suggested a critical role of membrane permeability in the degradation activity of the KRASG12D PROTACs. Mechanism studies with the representative compound 8o demonstrated that the potent, rapid, and selective degradation of KRASG12D induced by 8o was via a VHL- and proteasome-dependent manner. This compound selectively and potently suppressed the growth of multiple KRASG12D mutant cancer cells, displayed favorable pharmacokinetic and pharmacodynamic properties in mice, and showed significant antitumor efficacy in the AsPC-1 xenograft mouse model. Further optimization of 8o appears to be promising for the development of a new chemotherapy for KRASG12D-driven cancers as the complementary therapeutic strategy to KRAS inhibition.

Learning protein fitness landscapes with deep mutational scanning data from multiple sources.

One of the key points of machine learning-assisted directed evolution (MLDE) is the accurate learning of the fitness landscape, a conceptual mapping from sequence variants to the desired function. Here, we describe a multi-protein training scheme that leverages the existing deep mutational scanning data from diverse proteins to aid in understanding the fitness landscape of a new protein. Proof-of-concept trials are designed to validate this training scheme in three aspects: random and positional extrapolation for single-variant effects, zero-shot fitness predictions for new proteins, and extrapolation for higher-order variant effects from single-variant effects. Moreover, our study identified previously overlooked strong baselines, and their unexpectedly good performance brings our attention to the pitfalls of MLDE. Overall, these results may improve our understanding of the association between different protein fitness profiles and shed light on developing better machine learning-assisted approaches to the directed evolution of proteins. A record of this paper's transparent peer review process is included in the supplemental information.

Design, Synthesis, and Bioevaluation of Pyrido[2,3-d]pyrimidin-7-ones as Potent SOS1 Inhibitors.

The use of small molecular modulators to target the guanine nucleotide exchange factor SOS1 has been demonstrated to be a promising strategy for the treatment of various KRAS-driven cancers. In the present study, we designed and synthesized a series of new SOS1 inhibitors with the pyrido[2,3-d]pyrimidin-7-one scaffold. One representative compound 8u showed comparable activities to the reported SOS1 inhibitor BI-3406 in both the biochemical assay and the 3-D cell growth inhibition assay. Compound 8u obtained good cellular activities against a panel of KRAS G12-mutated cancer cell lines and inhibited downstream ERK and AKT activation in MIA PaCa-2 and AsPC-1 cells. In addition, it displayed synergistic antiproliferative effects when used in combination with KRAS G12C or G12D inhibitors. Further modifications of the new compounds may give us a promising SOS1 inhibitor with favorable druglike properties for use in the treatment of KRAS-mutated patients.

Computing the relative binding affinity of ligands based on a pairwise binding comparison network.

Structure-based lead optimization is an open challenge in drug discovery, which is still largely driven by hypotheses and depends on the experience of medicinal chemists. Here we propose a pairwise binding comparison network (PBCNet) based on a physics-informed graph attention mechanism, specifically tailored for ranking the relative binding affinity among congeneric ligands. Benchmarking on two held-out sets (provided by Schrödinger and Merck) containing over 460 ligands and 16 targets, PBCNet demonstrated substantial advantages in terms of both prediction accuracy and computational efficiency. Equipped with a fine-tuning operation, the performance of PBCNet reaches that of Schrödinger's FEP+, which is much more computationally intensive and requires substantial expert intervention. A further simulation-based experiment showed that active learning-optimized PBCNet may accelerate lead optimization campaigns by 473%. Finally, for the convenience of users, a web service for PBCNet is established to facilitate complex relative binding affinity prediction through an easy-to-operate graphical interface.