RiDYMO® Platform Advances PLK1 Inhibitors Development

BEIJING, Nov. 19, 2024 -- With the assistance of the RiDYMO® platform, researchers can delve deeply into various potential targets in drug development, such as protein-protein interactions, kinases, GPCRs, ion channels, and transcription factors.

Polo-like kinase 1 (PLK1) is a highly conserved serine/threonine protein kinase that plays an essential role during cellular mitosis. Overexpression of PLK1 is linked to various malignancies, including colorectal, pancreatic, and breast cancers, and is associated with poor prognosis. Consequently, PLK1 has become a significant target in cancer drug development, with many candidates currently undergoing research and trials. Notably, Cardiff Oncology's onvansertib (PCM-075) has made considerable progress, showing promising efficacy in treating patients with metastatic colorectal cancer (mCRC).

The RiDYMO® platform, developed by DP Technology, is a state-of-the-art computational drug design platform that integrates multiple artificial intelligence techniques, physical algorithms, and experimental validation to support hit discovery and optimization. Utilizing the co-crystal complex of PLK1 and PCM-075 (2YAC), the RiDYMO platform was employed to synthesize and test over 100 compounds. This led to the identification of preclinical candidate compounds with potential clinical value and significantly expedited the drug development process.

An in-depth analysis of the interactions between PCM-075 and the PLK1 protein was conducted. By preserving key interactions and applying molecular generation, evaluation, and screening techniques, a novel scaffold lead compound, DP101, was identified with an activity of 56 nM. During the optimization process based on the DP101 structure, modifications were made to several scaffold sites, including hydrogen bonding networks (R3) and hydrophobic group encapsulation zones (R1 & R2). Experimental data showed a high consistency between computational predictions and experimental outcomes, demonstrating the platform's effectiveness in optimizing kinase inhibitors.

Leveraging the RiDYMO® platform's activity prediction ranking and physicochemical property forecasting, multiple rounds of testing and optimization were conducted, ultimately resulting in several compounds with enhanced in vitro activity and pharmacokinetic properties. Notably, DP226 exhibited an activity of 0.19 nM, which was nearly 300 times better than DP101 and outperformed PCM-075. Furthermore, DP226 had an oral bioavailability of 76.8%, significantly higher than PCM-075's 24%. At low doses, DP226 displayed superior anti-tumor effects compared to PCM-075, and when combined with bevacizumab, it led to tumor regression, positioning it as a potential PCC molecule for further research.

Dr. Dongdong Li, Director of Medicinal Chemistry at DP Technology and Project Leader, highlighted the comprehensive exploration of various potential targets in drug development facilitated by the RiDYMO® platform. The platform's capabilities in activity predictions, rankings, and experimental validations significantly enhance the efficiency of activity optimization, effectively shortening optimization cycles and accelerating overall project development timelines. The integrated capabilities of the RiDYMO® platform show promise for providing high-efficiency support in early-stage drug discovery, and collaborations with experienced pharmaceutical companies are anticipated to advance the project to the next milestone.

The RiDYMO® Hit Discovery and Optimization Platform, developed by DP Technology, utilizes AI for Science to explore a broader chemical space encompassing small molecules, macrocycles, and cyclic peptides. By integrating advanced artificial intelligence, physical algorithms, and high-throughput experimentation, the platform excels in designing oral macrocyclic compounds and rapidly delivers innovative drug candidates.

One of the core algorithms, Reinforced Dynamics (RiD), offers a significant advantage in sampling efficiency for molecular dynamics simulations. By leveraging the high-dimensional representation capabilities of neural networks, RiD efficiently captures dynamic conformational changes in complex biomolecular systems. The core RiD algorithm has been published in Nature Computational Science, showcasing its capability to provide a comprehensive free energy surface within a short simulation time compared to traditional methods.

RiDYMO® has been successfully employed in various drug discovery projects, including the development of c-Myc small molecules, GPX4 small molecules, and β-catenin cyclic peptides, demonstrating its strong potential for innovative drug discovery.