Molecular modelling studies and identification of novel phytochemical inhibitor of DLL3

Prostate cancer has been recently considered the most diagnosed cancer in male. DLL3 is overexpressed in CRPC-NE but not in localised prostate cancer or BPH. There are no effective treatments for neuroendocrine differentiated prostate cancer due to a lack of understanding of DLL3 structure and function. The structure of DLL3 is not yet determined using any experimental techniques. Hence, the structure-based drug discovery approach against prostate cancer has not shown great success. In present study, molecular modelling techniques were employed to generate three-dimensional structure of DLL3 and performed its thorough structural analysis. Further, all-atom molecular dynamics simulation was performed to obtain energetically favourable conformation. Further, we used a virtual screening using a library of >13800 phytochemicals from the IMPPAT database and other literature to select the best possible phytochemical inhibitor for DLL3 and identified the top five compounds. Relative binding affinity was calculated using the MM-PBSA approach. ADMET properties of the screened compounds reveal the toxic effect of Gnemonol C. We believe these studied physicochemical properties, functional domain identification, and binding site identification would be very useful to gain more structural and functional insights of DLL3; also, it can be used to infer their pharmacodynamics properties of DLL3 which was recently reported as an important prostate cancer target. The current study also proposes that Ergosterol Peroxide, Dioslupecin A, Mulberrofuran K, and Caracurine V have strong affinities and could serve as plausible inhibitors against DLL3. We believe this study would further help develop better drug candidates against neuroendocrine prostate cancer.
Communicated by Ramaswamy H. Sarma

前列腺癌现已被认定为男性中确诊率最高的癌症。Delta样配体3（DLL3）在去势抵抗性神经内分泌前列腺癌（CRPC-NE）中呈过表达状态，而在局限性前列腺癌或良性前列腺增生（BPH）中则无过表达现象。由于对DLL3的结构与功能缺乏深入认知，目前尚无针对神经内分泌分化型前列腺癌的有效治疗手段。目前尚未有任何实验技术成功解析出DLL3的三维结构，因此基于结构的前列腺癌药物开发策略尚未取得显著突破。

本研究采用分子建模技术构建了DLL3的三维结构，并对其开展了全面的结构分析。随后通过全原子分子动力学模拟获取了能量最优的稳定构象。进一步，我们基于印度药用植物、植物化学与治疗学数据库（IMPPAT）及其他文献来源的超过13800种植物化学物文库进行虚拟筛选，以筛选出针对DLL3的最优植物化学抑制剂，并鉴定出排名前五的候选化合物。采用分子力学-泊松玻尔兹曼表面积（MM-PBSA）方法计算了各化合物的相对结合亲和力。对筛选所得化合物的ADMET（吸收、分布、代谢、排泄、毒性）性质分析显示，格尼莫醇C（Gnemonol C）存在毒性风险。

我们认为，本研究中对DLL3的理化性质分析、功能结构域鉴定以及结合位点解析，将有助于进一步阐明其结构与功能机制；同时可用于推断其药效学特性——而DLL3作为近期被报道的重要前列腺癌靶点，其相关研究具有重要价值。本研究同时提出，麦角甾醇过氧化物（Ergosterol Peroxide）、薯蓣鲁配辛A（Dioslupecin A）、桑呋喃K（Mulberrofuran K）以及卡拉库里碱V（Caracurine V）对DLL3具有较强结合亲和力，可作为潜在的DLL3抑制剂。我们相信本研究将助力开发更有效的神经内分泌前列腺癌治疗药物。

本文由Ramaswamy H. Sarma提交。