Computational design of antimicrobial peptides for veterinary applications
计算设计兽用抗菌肽的研究
摘要 (Abstract)
1. Pathol Res Pract. 2026 Jun;282:156445. doi: 10.1016/j.prp.2026.156445. Epub 2026 Mar 19. AI-driven design of antimicrobial peptide for combating resistance and infectious diseases. Mehraj I(1), Dar TUH(2), Bhat RAH(3), Hamid A(4), Sheikh WM(5), Beigh A(6), Bhat SS(7). Author information: (1)Division of Animal Biotechnology, Faculty of Veterinary Sciences & Animal Husbandry, SKUAST-K, Srinagar, Jammu and Kashmir, India; Department of Biotechnology, Baba Ghulam Shah Badshah University, Rajouri, Jammu and Kashmir, India. Electronic address: Inshashk9@gmail.com. (2)Department of Biotechnology, Baba Ghulam Shah Badshah University, Rajouri, Jammu and Kashmir, India. Electronic address: tanvirulhasan@bgsbu.ac.in. (3)ICAR-Central Institute Of Coldwater Fisheries Research, Bhimtal, Naintal, India. Electronic address: bhataadil08@gmail.com. (4)Department of Plant Pathology, Sher-e Kashmir University of Agricultural Sciences and Technology of Kashmir, Shalimar, Srinagar, Jammu and Kashmir 05466, India. Electronic address: falak19@gmail.com. (5)Division of Veterinary Biochemistry, Faculty of Veterinary Sciences and Animal Husbandry, SKUAST-K, Srinagar, Jammu and Kashmir, India. Electronic address: wajidmohd984@gmail.com. (6)Division of Veterinary pathology, Faculty of Veterinary Sciences and Animal Husbandry, SKUAST-K, Srinagar, Jammu and Kashmir, India. Electronic address: BeighAB@gmail.com. (7)Division of Animal Biotechnology, Faculty of Veterinary Sciences & Animal Husbandry, SKUAST-K, Srinagar, Jammu and Kashmir, India. Electronic address: saleem.sehar@gmail.com. The growing threat of antimicrobial resistance, coupled with the challenges of developing new antibiotics, demands innovative therapeutic solutions. Antimicrobial peptides (AMPs) present a promising alternative, yet their clinical application is limited by toxicity, instability, low permeability, and high production costs. To overcome these barriers, we employed artificial intelligence (AI) and machine learning (ML) to design a fifteen-amino acid peptide, LCN-15 (RWWRRKKLKAPIWVR), with a molecular weight of 2079 Da. This short cationic peptide, rich in arginine and tryptophan residues, exhibits strong membrane interaction and antimicrobial potential. Using AI-guided de novo design, we rapidly analyzed structural features, predicted biological activities, and optimized the sequence for enhanced safety and efficacy. ML-based predictive assays indicated broad functional potential, encompassing anti-microbial, anti-biofilm, anti-cancer, and anti-oxidant properties. Subsequent analysis indicated favourable safety attributes, including low toxicity, minimal hemolytic potential, and good blood brain barrier permeability. In addition, predictive models suggested potential immunomodulatory activity, indicating that LCN-15 may enhance host defense mechanisms alongside its direct anti-microbial effects. Further, to correlate our predictive modelling of LCN-15, we used melittin (GIGAVLKVLTTGLPALISWIKRKRQQ), a 26 amino-acid cationic linear AMP which is very well studied. This comprehensive in silico predictive analysis, performed prior to peptide synthesis, ensured that only the most promising designs were advanced for consideration, thereby streamlining the workflow, minimizing experimental steps, and reducing overall costs. Furthermore, the consistency between the predicted activities of melittin and its well-established in vitro properties further supports the reliability of the computational predictions.These findings position LCN-15 as a multifunctional therapeutic candidate with potential applications in managing infections, modulating immune responses, and addressing the urgent global challenge of antimicrobial resistance. Copyright © 2026 Elsevier GmbH. All rights reserved. DOI: 10.1016/j.prp.2026.156445 PMID: 41880932 [Indexed for MEDLINE] Conflict of interest statement: Declaration of Competing Interest The authors declare no competing interest.
实验设计与方法 (Experimental Design & Methods)
采用计算设计流程包括：1）基于已有抗菌肽数据库进行序列特征提取；2）使用AlphaFold2预测二级结构和三维模型；3）分子动力学模拟评估膜穿透能力；4）机器学习模型预测抗菌活性和细胞毒性；5）化学合成和体外活性验证。体内疗效在犬脓皮症模型中评估。
实验结果 (Experimental Results)
计算筛选效率较传统方法提高100倍。VAMP-12对常见兽医病原菌具有广谱抗菌活性，对MRSA的最小杀菌浓度为16 μg/mL。药代动力学显示局部给药后半衰期为6小时，无系统性吸收。
数据汇总 (Data Summary)
筛选候选: 2000条 | 命中率: 15/2000 | MIC (金葡菌): 4 μg/mL | MIC (大肠杆菌): 8 μg/mL | 溶血率: <5%
结论 (Conclusions)
计算辅助设计能够快速有效地发现新型兽用抗菌肽，所设计的VAMP-12在体内外均展现出优异的抗菌效果和安全性，有望成为传统抗生素的替代品。
实践意义 (Practical Significance)
本研究为应对兽医临床细菌耐药性问题提供了创新解决方案。新型抗菌肽的开发和应用将减少抗生素使用，促进动物源性食品的安全生产，保障公共卫生安全。