A deep learning-driven discovery of berberine derivatives as novel antibacterial against multidrug-resistant Helicobacter pylori through targeting outer membrane protein transport and assembling

Helicobacter pylori (H. pylori) is currently recognized as the primary carcinogenic pathogen associated with gastric tumorigenesis, and its high prevalence and resistance make it difficult to tackle. A graph neural network-based deep learning model, employing different training sets of 13,638 molecules for pre-training and fine-tuning through rigorous iterative learning processes, was aided in predicting and exploring novel molecules against H. pylori. A positively predicted novel berberine (BBR) derivative 8 with 3,13-disubstituted alkene exhibited a potency against all tested drug-susceptible and resistant H. pylori strains with minimum inhibitory concentrations (MICs) of 0.25–0.5 μg/mL. Strikingly, pharmacokinetic studies demonstrated an ideal gastric retention of 8, with the stomach concentration significantly higher than its MIC value at 24 h post dose. Oral administration of 8 and proton pump inhibitor omeprazole (OPZ) showed a 2.2-log reduction in the gastric bacterial burden compared with the control group, which is comparable to the triple-therapy, namely OPZ + amoxicillin (AMX) + clarithromycin (CLA), and partially restored the diversity of the intestinal flora as well as the abundance of probiotics. A combination of OPZ + AMX + CLA + 8 could further decrease the bacteria load (2.8-log reduction). More importantly, the mono-therapy of 8 showed comparable eradicative efficacies compared with both triple-therapy (OPZ + AMX + CLA) and the quadruple-therapy (OPZ + AMX + CLA + bismuth citrate) groups. SecA and BamD, playing a major role in outer membrane protein (OMP) transport and assembling, were identified and verified as the direct targets of 8 by employing the chemoproteomics technique. The treatment of 8 induced the death of H. pylori caused by OMP deficiency, and the subsequent reduced adhesion to gastric epithelial cells. In summary, by targeting the relatively conserved OMPs transport and assembling system, 8 has the potential to be developed as a novel anti-H. pylori candidate, especially for the eradication of drug-resistant strains. The deep learning model established in this study might provide a reliable prediction tool for future anti-H. pylori candidate discovery.

幽门螺杆菌（Helicobacter pylori, H. pylori）是目前公认的与胃肿瘤发生密切相关的主要致癌病原体，其高流行率与耐药性使得防控难题难以攻克。本研究构建了基于图神经网络的深度学习模型，采用13638个分子的不同训练集，通过严谨的迭代学习流程完成预训练与微调，以辅助预测并探索抗幽门螺杆菌的新型分子。经正向预测得到的新型小檗碱（berberine, BBR）衍生物8（3,13-二取代烯烃结构）对所有受试药敏株与耐药幽门螺杆菌菌株均展现出强效活性，最低抑菌浓度（minimum inhibitory concentration, MIC）范围为0.25~0.5 μg/mL。值得注意的是，药代动力学研究显示衍生物8具备理想的胃滞留特性：给药24小时后，胃内药物浓度显著高于其MIC值。与对照组相比，单独口服衍生物8联合质子泵抑制剂奥美拉唑（omeprazole, OPZ）可使胃内细菌负荷降低2.2 log，效果与由OPZ+阿莫西林（amoxicillin, AMX）+克拉霉素（clarithromycin, CLA）组成的三联疗法相当，同时还可部分恢复肠道菌群多样性与益生菌丰度。而OPZ+AMX+CLA+衍生物8的联合疗法可进一步降低细菌载量（降幅达2.8 log）。更重要的是，单独使用衍生物8的根除效果可与三联疗法（OPZ+AMX+CLA）及四联疗法（OPZ+AMX+CLA+枸橼酸铋）相媲美。通过化学蛋白质组学技术，本研究鉴定并验证了在外膜蛋白（outer membrane protein, OMP）转运与组装过程中发挥核心作用的SecA与BamD为衍生物8的直接作用靶点。衍生物8可通过诱导外膜蛋白缺陷引发幽门螺杆菌死亡，进而降低其对胃上皮细胞的黏附能力。综上，通过靶向相对保守的外膜蛋白转运与组装系统，衍生物8具备开发为新型抗幽门螺杆菌候选药物的潜力，尤其适用于耐药菌株的根除。本研究构建的深度学习模型可为未来抗幽门螺杆菌候选药物的发现提供可靠的预测工具。