Predictive Structure-Based Models of Malaria Resistance

描述（由申请人提供）：本项目的目标是开发未来恶性疟原虫（Pf）和间日疟原虫（Pv）二氢叶酸还原酶（DHFR）蛋白进化的预测模型，这将促进对野外可能的未来突变的假设生成，从而在这些突变之前发现针对耐药菌株的新型抗疟疾治疗方法。通过这项SC3研究，我们将对DHFR蛋白进化进行全面的基于结构的分析，以生成可能的氨基酸替代的位点特异性预测模型，并确定在响应选择压力时可能发生补偿性氨基酸替代的位置。研究将开始使用从公共领域数据库获得的DHFR蛋白序列生成一个全面的系统发育树。预测DHFR进化过程中关键支系的祖先序列。然后将为每个祖先序列生成3D同源模型，这些序列将依次添加到现有的基于结构的序列比对中，该序列比对是由来自22个物种的野生型（wt） DHFR实验确定的x射线晶体结构的叠加产生的。位点特异性氨基酸替换的预测模型将使用来自计算智能和机器学习领域的工具和技术，包括hmm和ann。这些模型将通过使用前70%的系统发育来测试和验证，以便预测剩余的30%。利用从这些预测模型中获得的见解，将对恶性疟原虫和间日疟原虫的DHFR突变序列进行分析，以研究导致耐药性的种内分化。假设的DHFR序列和同源性模型将代表疟原虫DHFR进化的下一步。将与现有的抗疟疾药物以及已知的DHFR抑制剂进行硅对接实验。从这些研究中预测的蛋白质-配体相互作用的检查将为耐药性的获得提供额外的见解。通过这项工作，实现了一种研究和建模DHFR蛋白进化的创新技术。将生成恶性疟原虫和间日疟原虫DHFR进化的潜在下一步预测模型，作为该方法概念的证明。这项技术具有深远的益处，包括对恶性疟原虫和间日疟原虫的种内分化和耐药性起源的假设，以及生成未来DHFR蛋白进化预测模型的能力，为在野外耐药性发展之前获得药物发现的“领先优势”提供了独特的机会。

DESCRIPTION (provided by applicant): The goal of this proposed effort is to develop predictive models of future Plasmodium falciparum (Pf) and Plasmodium vivax (Pv) dihydrofolate reductase (DHFR) protein evolution that will facilitate hypothesis generation for likely future mutations in the wild, leading to discovery of novel anti-malarial therapeutics againt drug resistant strains in advance of these mutations. Through this SC3 research, we will perform a comprehensive structure-based analysis of DHFR protein evolution in order to generate site-specific predictive models of likely amino acid replacements and identify locations where compensating amino acid replacements may be occurring in response to selection pressures. Research will commence with generation of a comprehensive phylogenetic tree using DHFR protein sequences obtained from public domain databases. Ancestral sequences will be predicted for key clades in DHFR evolution. 3D homology models will then be generated for each of these ancestral sequences that will be added sequentially to an already existing structure-based sequence alignment generated from a superposition of experimentally determined x-ray crystal structures of wild-type (wt) DHFR from 22 species. Predictive models of site-specific amino acid replacements will be generated using tools and techniques taken from the field of computational intelligence and machine learning that include HMMs and ANNs. These models will be tested and validated by using the first 70% of the phylogeny in order to predict the remaining 30%. Using the insights gained from these predictive models, an analysis will be performed with mutant DHFR sequences of P. falciparum and P. vivax to study the intraspecies differentiation that gave rise to drug resistance. Hypothetical DHFR sequences and homology models representing next steps in Plasmodium DHFR evolution will be generated. In silico docking experiments will be performed with existing anti-malarial drugs as well as known inhibitors of DHFR. Examination of the predicted protein-ligand interactions from these studies will provide additional insights into the acquisition of drug resistance. Through this proposed effort, an innovative technology for studying and modeling DHFR protein evolution is realized. Predictive models of potential next steps in P. falciparum and P. vivax DHFR evolution will be generated as a proof of concept of this approach. This technology has far-reaching benefits including the generation of hypotheses for intraspecies differentiation and origins of drug resistance in P. falciparum and P. vivax as well as the ability to generate predictive models of future DHFR protein evolution providing the unique opportunity of getting a "head start" on drug discovery before drug resistance develops in the wild. PUBLIC HEALTH RELEVANCE: Approximately forty-one percent of the world's population lives in areas where malaria is transmitted and each year and it is estimated that 350-500 million cases of malaria occur worldwide. Two of the most prevalent malaria strains, Plasmodium falciparum (Pf) and Plasmodium vivax (Pv) have developed clinical resistance to antifolate compounds that target the enzyme dihydrofolate reductase (DHFR) such as pyrimethamine and cycloguanil. Consequently, there remains a serious and immediate need for the development of novel antimalarial therapeutics that target drug-resistant strains. Using the proposed approach, we will develop predictive models of future Pf-DHFR and Pv-DHFR protein evolution that will facilitate hypothesis generation for likely future mutations in the wild. Afterthe completion of this SC3 research, we plan to integrate these predictive models into a comprehensive computational intelligence- based drug discovery platform thus providing the unique opportunity of getting a "head start" on drug discovery resulting in timely development of novel anti-malarial therapeutics to meet future needs. This approach will be tested on DHFR for novel antimalarial drug discovery; however, the methods developed can be applied broadly in early stage drug discovery and development.