Institut für Parasitologie, Biomedizinisches Forschungszentrum Seltersberg (BFS), Justus-Liebig-Universität Giessen, Giessen, Germany.
Grupo de Helmintologia e Malacologia Médica, Instituto René Rachou, Fundação Oswaldo Cruz-FIOCRUZ, Belo Horizonte, Brazil.
Front Cell Infect Microbiol. 2022 Jul 5;12:913301. doi: 10.3389/fcimb.2022.913301. eCollection 2022.
Schistosomiasis is a parasitic neglected disease with praziquantel (PZQ) utilized as the main drug for treatment, despite its low effectiveness against early stages of the worm. To aid in the search for new drugs to tackle schistosomiasis, computer-aided drug design has been proved a helpful tool to enhance the search and initial identification of schistosomicidal compounds, allowing fast and cost-efficient progress in drug discovery. The combination of high-throughput data followed by phenotypic screening assays allows the assessment of a vast library of compounds with the potential to inhibit a single or even several biological targets in a more time- and cost-saving manner. Here, we describe the molecular docking for screening of predicted homology models of five protein kinases (JNK, p38, ERK1, ERK2, and FES) of against approximately 85,000 molecules from the Managed Chemical Compounds Collection (MCCC) of the University of Nottingham (UK). We selected 169 molecules predicted to bind to SmERK1, SmERK2, SmFES, SmJNK, and/or Smp38 for screening assays using schistosomula and adult worms. In total, 89 (52.6%) molecules were considered active in at least one of the assays. This approach shows a much higher efficiency when compared to using only traditional high-throughput screening assays, where initial positive hits are retrieved from testing thousands of molecules. Additionally, when we focused on compound promiscuity over selectivity, we were able to efficiently detect active compounds that are predicted to target all kinases at the same time. This approach reinforces the concept of polypharmacology aiming for "one drug-multiple targets". Moreover, at least 17 active compounds presented satisfactory drug-like properties score when compared to PZQ, which allows for optimization before further screening assays. In conclusion, our data support the use of computer-aided drug design methodologies in conjunction with high-throughput screening approach.
血吸虫病是一种被忽视的寄生虫病，目前使用吡喹酮（PZQ）作为主要治疗药物，尽管它对早期蠕虫的疗效较低。为了寻找治疗血吸虫病的新药，计算机辅助药物设计已被证明是一种有用的工具，可以帮助寻找和初步鉴定杀血吸虫化合物，从而在药物发现方面实现快速和具有成本效益的进展。高通量数据与表型筛选测定相结合，允许以更节省时间和成本的方式评估具有抑制单一甚至多种生物靶标的巨大化合物库。在这里，我们描述了针对五个蛋白激酶（JNK、p38、ERK1、ERK2 和 FES）的预测同源模型进行分子对接筛选，这些激酶是从英国诺丁汉大学的 Managed Chemical Compounds Collection (MCCC) 中大约 85,000 种化合物中筛选出来的。我们选择了 169 种预测与 SmERK1、SmERK2、SmFES、SmJNK 和/或 Smp38 结合的分子进行筛选实验，使用的是毛蚴和成虫。总共，有 89 种（52.6%）分子在至少一种测定中被认为是活性的。与仅使用传统高通量筛选测定相比，这种方法的效率要高得多，在传统高通量筛选测定中，最初的阳性命中是从测试数千种分子中检索到的。此外，当我们关注化合物的混杂性而不是选择性时，我们能够有效地检测到同时预测针对所有激酶的活性化合物。这种方法强化了多药理学的概念，旨在实现“一种药物-多个靶点”。此外，与 PZQ 相比，至少有 17 种活性化合物的药物样性质评分令人满意，这允许在进一步的筛选测定之前进行优化。总之，我们的数据支持将计算机辅助药物设计方法与高通量筛选方法结合使用。
