2020-10-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/652271摘要：本計劃旨在開發一個新的技術平台，結合生物資訊和組合式胜&#32957;庫 (combinatorial peptide library) 合成，從前人未曾探索過的天然物世界中，找尋新的抗生素。 此技術平台奠基於一個簡單的概念：生物分子中，保守的結構模體 (conserved structural motif) 與重要功能息息相關。一個與已知的抗生素結構類似、有著相同模體的分子，多半也是個抗生素。因此，若想要發現新的抗生素，應該優先探索已知抗生素之保守模體附近的化學空間。「非核糖體&#32957; (nonribosomal peptides (NRP)) 」是天然物種類中為數最多、最多樣化的家族，在目前已知的諸多抗生素中它也是最大宗，故本計劃將以非核糖體&#32957;為核心。無奈相較於其他生物分子，如DNA和蛋白質，我們所知的NRP數目太少，以致於無法從中歸納出保守模體。這得歸咎於傳統天然物的發現，必需在實驗室中培養微生物，然而絕大部分的微生物在實驗室中不容易培養。因此，靠這套傳統方法所發現的小分子，僅僅佔天然物多樣性的極小一部分 (<1％)，許多科學家於是致力於開發不需仰賴培養微生物的天然物發現方法。若干生物資訊演算法 (bioinformatic algorithms)，能僅就NRP的合成酵素之蛋白質序列，來預測該NRP的結構。我將利用這些演算法分析微生物基因體資料庫，預測傳統天然物發現方法所未能探索的NRP。接下來，這些靠預測所得到的NRP，將與已知的非核糖體&#32957;資料庫結合，創造一個新的虛擬資料庫，再從中找尋保守模體。本計劃中，每一個組合式胜&#32957;庫的設計，都會對應一個保守的NRP模體：模體保持不變，其他胺基酸則隨機化，藉此有效地探索模體附近的化學空間。從這些組合式胜&#32957;庫中透過活性篩選所得到的新抗生素，在增量合成後，仔細研究它的最小抑菌濃度、作用機制、以及結構活性關係，並且積極申請專利，往藥物開發之路邁進。 由於本計劃所提出的技術平台不需仰賴微生物培養，它理當能夠探索過去傳統天然物發現方法未能深入的世界。<br> Abstract: Here I propose to develop a new antibiotic discovery pipeline that draws inspiration from natural products; it will combine bioinformatic analysis and combinatorial peptide synthesis to tap into previously unexplored natural product chemical spaces. The new discovery approach is founded on the notion that functional importance in biomolecules correlates highly with conserved structural features (motifs). The chemical space close to the conserved motif of a known antibiotic is highly likely to harbor more antibiotics, and molecules in this space should be prioritized in a search for new antibiotics. Nonribosomal peptides (NRPs) are the largest, most diverse natural product family. They have contributed to known antibiotics more than any other families and will the focus of this proposal. Unfortunately, compared to other biomolecules, such as DNA and proteins, there are too few known NRPs for conserved motifs to be identified. This is because the traditional natural product discovery approach based on examining microbial culture broths has accessed only a fraction (<1%) of the biosynthetic diversity encoded in microorganisms. The disparity is attributed to our inability to culture the vast majority of microbes under laboratory conditions, and scientists are seeking various new approaches that are independent of culture to circumvent this bottleneck. One possible starting point is the use of bioinformatic algorithms to predict the structure of a NRP based on the primary sequence of its biosynthetic genes alone. I will apply these algorithms to the vast collection of microbial genomes to predict NRPs the traditional approached failed to discover. Predicted NRPs will then be combined with the database of known natural products to generate a virtual library and searched in silico to identify conserved NRP motifs. Each of my combinatorial peptide library will be designed to carry an invariant conserved NRP motif flanked by variable residues to effectively survey the chemical space nearby. Finally, such a NRP inspired combinatorial peptide library will be synthesized and screened for new antibiotics; hits will be resynthesized in bulk for validation and characterization, including their minimum inhibitory concentrations, mechanism of actions, and structure-activity relationship studies. Because culture is not a prerequisite, my new approach will be able to explore the immense natural product world that has thus far remained unexplored by the traditional culture-based discovery approach.非核糖體&#32957天然物抗生素依鈣抗生素抗真菌藥達托霉素(救必辛)保守模體生物資訊演算法非核糖體&#32957結構預測組合式胜&#32957庫生物活性篩選nonribosomal peptidesnatural productsantibioticscalcium-dependent antibioticantifungaldaptomycinconserved motifbioinformatic algorithmsNRP structure predictioncombinatorial peptide librarybioactivity screening