Targeting receptor tyrosine kinases with novel methods in computer-aided drug discovery for the treatment of fibrotic renal disease

用计算机辅助药物发现的新方法靶向受体酪氨酸激酶来治疗纤维化肾病

• 批准号: 10197115
• 负责人: Benjamin Patrick Brown
• 金额: $ 5.1万
• 依托单位: VANDERBILT UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-07-01 至 2022-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10197115
• 关键词: Address; Age; Algorithms; Binding; Binding Proteins; Binding Sites; Biochemical; Cell model; Chemicals; Chemistry; Chronic Kidney Failure; Clinical; Collaborations; Collagen; Collagen Type IV; Computer Assisted; Computer software; Computing Methodologies; DDR2 gene; Data; Deposition; Descriptor; Development; Disease; Disease Progression; Docking; End stage renal failure; Evaluation; Family; Fibrosis; Fluorescence Resonance Energy Transfer; In Vitro; Individual; Injury to Kidney; Intervention; Joints; Kidney; Kidney Diseases; Knowledge; Laboratories; Lead; Learning Module; Libraries; Ligands; Link; Machine Learning; Mediating; Methodology; Methods; Modeling; Molecular Conformation; Myofibroblast; Patients; Pharmacology; Phase; Phosphotransferases; Population; Prevalence; Proteins; Quantitative Structure-Activity Relationship; Receptor Protein-Tyrosine Kinases; Risk Factors; Risk Management; Sampling; Scientist; Severities; Site-Directed Mutagenesis; Structural Models; Structure; Symptoms; Testing; Therapeutic; Therapeutic Agents; United States; Work; aged; base; clinically relevant; design; discoidin domain receptor 1; discoidin domain receptor 2; discoidin receptor; drug candidate; drug discovery; flexibility; global health; high throughput analysis; high throughput screening; hospital readmission; improved; in silico; in vivo Model; inhibitor/antagonist; innovation; kidney fibrosis; kinase inhibitor; learning algorithm; machine learning algorithm; mesangial cell; molecular dynamics; mortality; mouse model; multi-task learning; multitask; neural network; new technology; novel; novel lead compound; novel therapeutic intervention; novel therapeutics; protein structure; screening; small molecule; structural biology; targeted treatment; therapeutic target; tool; virtual; virtual model

PROJECT SUMMARY Chronic Kidney Disease (CKD) is a major disease multiplier in patients aged 65+. CKD is characterized by progressive renal fibrosis mediated through supraphysiologic type IV collagen deposition by renal myofibroblasts. As the US population continues to age, it becomes increasingly critical to identify new therapeutic strategies for CKD. Mouse models of kidney injury suggest reducing the activity of the receptor tyrosine kinase discoidin domain receptor 1 (DDR1) is protective against fibrotic renal disease. Inhibition of DDR1 kinase reduces mesangial cell deposition of type IV collagen. To develop targeted therapeutics for CKD, the laboratory of Jens Meiler (sponsor of this application) partners with the laboratories of Ambra Pozzi (co- sponsor of this application) and Craig Lindsley to create a comprehensive DDR1 kinase inhibitor discovery pipeline. The Meiler laboratory utilizes a combination of ligand-based quantitative structure-activity relationship (QSAR) modeling for virtual high-throughput screening (vHTS) and subsequent protein-ligand docking to identify lead compounds for synthesis/derivatization (Lindsley) and biochemical/functional evaluation (Pozzi). Selective targeting of individual kinases remains a significant challenge, and current methods in vHTS fail to account for protein binding pocket features contributing to binding selectivity. The central objectives of this proposal are to identify novel DDR1-selective inhibitors for the treatment of CKD and to develop new technologies to address current limitations in vHTS. In Specific Aim I, I will generate and use QSAR models to perform vHTS for potential DDR1 inhibitors. I will subsequently define a structural model of DDR1 kinase inhibitor selectivity using molecular dynamics (MD)-generated conformational ensembles of DDR kinases in conjunction with ROSETTA flexible docking. I will also perform in silico and in vitro site-directed mutagenesis to further characterize the determinants of DDR1 kinase inhibitor selectivity. In Specific Aim II, I will develop a multitasking machine algorithm within the Meiler lab BIOLOGY AND CHEMISTRY LIBRARY (BCL) which will leverage protein structural information in addition to conventional ligand-based descriptors to improve vHTS for selective DDR1 kinase inhibitors. The methods developed will address long-standing shortcomings in the field of computer-aided drug discovery (CADD) – namely, that protein structure-based methods are computationally prohibitive for vHTS while ligand-based methods do not include direct information on binding mode. As the methods developed in Aim II become available, they will be integrated in the discovery cycle described in Aim I to ultimately define a structural model of DDR1 kinase selectivity and identify novel therapeutic agents for the treatment of CKD through the use of new and established methods. Furthermore, novel computational methods established in these studies will be broadly applicable to other challenging targets in drug discovery.

项目概要 慢性肾脏病 (CKD) 是 65 岁以上患者的主要疾病倍增因素。 CKD 的特点是 肾超生理性 IV 型胶原沉积介导的进行性肾纤维化 肌成纤维细胞。随着美国人口持续老龄化，寻找新的人口变得越来越重要 CKD 的治疗策略。小鼠肾损伤模型表明受体活性降低 酪氨酸激酶盘状蛋白结构域受体 1 (DDR1) 可预防纤维化肾病。抑制 DDR1 激酶减少 IV 型胶原蛋白的系膜细胞沉积。开发针对 CKD 的靶向疗法， Jens Meil​​er 实验室（本申请的赞助商）与 Ambra Pozzi 实验室（共同 本申请的赞助者）和 Craig Lindsley 共同创建全面的 DDR1 激酶抑制剂发现 管道。 Meiler 实验室利用基于配体的定量结构-活性关系的组合 用于虚拟高通量筛选 (vHTS) 的 (QSAR) 建模以及随后的蛋白质配体对接 识别用于合成/衍生化 (Lindsley) 和生化/功能评估 (Pozzi) 的先导化合物。 选择性靶向单个激酶仍然是一个重大挑战，目前的 vHTS 方法无法 解释有助于结合选择性的蛋白质结合口袋特征。本次活动的中心目标 提议是确定用于治疗 CKD 的新型 DDR1 选择性抑制剂并开发新的 解决 vHTS 当前限制的技术。在具体目标 I 中，我将生成并使用 QSAR 模型来 对潜在的 DDR1 抑制剂进行 vHTS。我随后将定义DDR1激酶的结构模型 使用分子动力学 (MD) 生成的 DDR 激酶构象整体进行抑制剂选择性 与 ROSETTA 灵活对接配合。我还将进行计算机模拟和体外定点突变 进一步表征 DDR1 激酶抑制剂选择性的决定因素。在具体目标 II 中，我将制定一个 Meiler 实验室生物和化学图书馆 (BCL) 中的多任务机器算法将 除了传统的基于配体的描述符之外，还利用蛋白质结构信息来改进 vHTS 选择性 DDR1 激酶抑制剂。所开发的方法将解决该领域长期存在的缺陷 计算机辅助药物发现 (CADD) 的发展——即基于蛋白质结构的方法通过计算 禁止 vHTS，而基于配体的方法不包括有关结合模式的直接信息。作为 在目标 II 中开发的方法变得可用，它们将被集成到目标 I 中描述的发现周期中 最终定义 DDR1 激酶选择性的结构模型并确定新的治疗药物 通过使用新的和既定的方法治疗 CKD。此外，新颖的计算 这些研究中建立的方法将广泛适用于药物发现中的其他具有挑战性的目标。