| AI-Guided Discovery of Allosteric Src Inhibitors for Chronic Borrelia Infection |
| Paper ID : 1122-IPCA5 (R2) |
| Authors |
|
Fatemeh Shams1, Elina Khanehzar1, Zakkyeh Telmadarraiy2, Faezeh Faghihi3, Amirsajad Jafari *4 11. Student Research Committee, Shiraz University of Medical Sciences, Shiraz, Iran 2. Medicinal and Natural Products Chemistry Research Center, Shiraz University of Medical Sciences, Shiraz, Iran 21. Department of Medical Entomology and Vector Control, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran 2. Rahyan Novin Danesh (RND) University, Sari, Mazandaran, Iran 31. Cellular and Molecular Research Center, Iran University of Medical Sciences, Tehran, Iran 2. Immunogenetics Research Center, Mazandaran University of Medical Sciences, Sari, Iran 41. Department of Basic Sciences, School of Veterinary Medicine, Shiraz University, Shiraz, Iran
2. Medicinal and Natural Products Chemistry Research Center, Shiraz University of Medical Sciences, Shiraz, Iran |
| Abstract |
| Iran faces a significant public health threat from tick-borne Borrelia pathogens, including Lyme disease (Borrelia burgdorferi) and relapsing fever (B. theileri, B. persica). The primary Lyme disease vector, Ixodes ricinus, is concentrated in the humid Caspian region (Mazandaran, Guilan, Kurdistan), with 2–10% of ticks carrying pathogenic genospecies. Other ticks, such as Rhipicephalus annulatus (linked to B. theileri) and Ornithodoros tholozani (transmitting B. persica), further contribute to transmission, though the vector competence of R. sanguineus and Hyalomma spp. remains uncertain. Rural populations in forested areas are at high risk of exposure. Chronic Lyme disease, driven by persistent inflammation and immune dysregulation, lacks effective treatments, prompting this study to employ AI in designing novel allosteric Src kinase inhibitors—a potential therapeutic breakthrough. Ongoing surveillance of tick populations and pathogens is essential to mitigate zoonotic risks in Iran and globally. To address these challenges, we employed ChatGPT-4o, a commercial LLM, to design novel allosteric inhibitors targeting the inactive conformation of Src (PDB: 1FMK). The model generated ten structurally diverse compounds inspired by known Src inhibitors (PP2, SU6656, and Dasatinib), optimized for Lipinski compliance. The SwissADME database was used to evaluate pharmacokinetic properties and toxicity profiles. Molecular docking simulations via PyRx assessed binding affinities, with PP2 serving as the reference inhibitor. ChatGPT-4o successfully designed 10 novel compounds, nine of which exhibited higher binding affinity than PP2. Four candidates—IsoPP2-Triazole, Ethoxy-SU6656, Quinazoline-iPr, and N-Methyl-PP2 Ether—met lead-likeness criteria with favorable pharmacological profiles. The top-performing molecule, IsoPP2-Triazole, demonstrated exceptional binding to Src’s allosteric cleft (−9.2 kcal/mol vs. −7.4 kcal/mol for PP2), positioning it as a prime candidate for further development. This study presents an AI-driven framework for designing allosteric Src inhibitors to combat chronic inflammation from tick-borne diseases like Lyme disease. The identification of Borrelia-carrying ticks in Iran’s northern and western regions underscores the need for targeted therapies. By selectively inhibiting Src-mediated inflammatory pathways, the discovered compounds, particularly IsoPP2-Triazole, offer a promising strategy to alleviate persistent inflammation. This AI-assisted approach not only accelerates drug discovery but also opens new avenues for treating the complex inflammatory sequelae of tick-borne infections in Iran and globally. |
| Keywords |
| Lyme, LLM, Inflammation, In silico, Drug Design |
| Status: Abstract Accepted |
