Chowdhury Rayhan, Saima Samia Akter, Amin Md Al, Habib Md Kawsar, Mohiuddin Ramisa Binti, Wasaf Hasan Ali Mohamod, Khanam Roksana, Mahmud Shahin
Department of Biotechnology and Genetic Engineering, Mawlana Bhashani Science and Technology University, Santosh, Tangail-1902, Bangladesh.
Department of Pharmacy, Mawlana Bhashani Science and Technology University, Santosh, Tangail-1902, Bangladesh.
J Genet Eng Biotechnol. 2025 Sep;23(3):100514. doi: 10.1016/j.jgeb.2025.100514. Epub 2025 Jun 3.
Antibiotic resistance poses a significant global challenge as bacteria evolve in response to antibiotic use, leading to prolonged hospitalizations, increased healthcare costs, and higher mortality rates. Cephalosporins, a class of beta-lactam antibiotics, are commonly employed to manage infections; however, their misuse and overuse have contributed to resistance development. In response, in silico methods have emerged as cost-effective and efficient tools for drug discovery. This research aims to predict new compounds using ligand-based pharmacophore models while optimizing existing drugs. We employed a de novo approach to synthesize models of cephalosporin structural motifs, integrating the β-lactam core with potential antibiotic candidates. A shared features pharmacophore (SFP) model was constructed using cephalosporins from PubChem, including cephalothin, ceftriaxone, and cefotaxime. The model comprises hydrogen bond acceptors, hydrogen bond donors, aromatic rings, hydrophobic regions, and negatively ionizable sites. Its robustness was evidenced by a goodness-of-hit (GH) score of 0.739. The generated pharmacophore model, with a score of 0.9268, was utilized to screen a drug library, initially assessing 19 compounds. After the drug-likeness screening, seven promising compounds were identified. These candidates were then fused with the cephalosporin core using genetic algorithms and fragment-based design, resulting in 30 novel synthetic models. Most of these models demonstrated a cephalosporin core, over 70 % average similarity, a TPSA (NO) ≤ 99.85 Å, a drug-likeness (QED) ≥ 0.6, and a Synthetic Accessibility Score (SAScore) ≤ 4.3. Molecular docking and MD simulation evaluations highlighted two candidates-Molecule 23 and Molecule 5, demonstrating superior binding affinities to Penicillin-binding protein 1a (PDB ID: 2V2F) compared to controls. To ensure feasible synthesis, molecular architecture comparison and computational retrosynthesis were performed, confirming the likelihood of successful laboratory synthesis. These findings advance the fight against antimicrobial resistance by establishing a method for designing new, highly effective antibiotic drugs.
抗生素耐药性是一项重大的全球挑战，因为细菌会随着抗生素的使用而进化，导致住院时间延长、医疗成本增加和死亡率上升。头孢菌素是一类β-内酰胺抗生素，常用于治疗感染；然而，它们的滥用和过度使用导致了耐药性的产生。作为回应，计算机辅助方法已成为药物研发中经济高效的工具。本研究旨在使用基于配体的药效团模型预测新化合物，同时优化现有药物。我们采用从头开始的方法合成头孢菌素结构基序模型，将β-内酰胺核心与潜在的抗生素候选物整合在一起。使用来自PubChem的头孢菌素构建了一个共享特征药效团（SFP）模型，包括头孢噻吩、头孢曲松和头孢噻肟。该模型包括氢键受体、氢键供体、芳香环、疏水区域和可电离的负电荷位点。其稳健性通过0.739的命中优值（GH）得分得到证明。生成的药效团模型得分为0.9268，用于筛选药物库，最初评估了19种化合物。经过类药性筛选，确定了7种有前景的化合物。然后使用遗传算法和基于片段的设计将这些候选物与头孢菌素核心融合，产生了30个新的合成模型。这些模型中的大多数都展示了头孢菌素核心，平均相似度超过70%，拓扑极性表面积（TPSA）（NO）≤99.85 Å，类药性（QED）≥0.6，以及合成可及性得分（SAScore）≤4.3。分子对接和分子动力学模拟评估突出了两个候选物——分子23和分子5，与对照相比，它们对青霉素结合蛋白1a（PDB ID：2V2F）表现出更高的结合亲和力。为确保合成可行，进行了分子结构比较和计算逆合成，证实了实验室成功合成的可能性。这些发现通过建立一种设计新型高效抗生素药物的方法，推进了对抗菌素耐药性的斗争。
