分子間藥理作用介面家族應用在磷酸化酵素-藥物-疾病網絡與機制之研究

Abstract

目前的藥物開發流程因其成本昂貴且需時極長，已逐漸不能滿足市場對新藥物的需求。再者，過去「單一藥物針對單一標的，治療單一疾病」的思維，在治療病理現象複雜的疾病(如癌症)上也遭遇諸多瓶頸。針對上述議題，本計劃的主要目標在於利用所提出之新概念「分子間藥理作用介面家族(molecular pharma-interface family)」以及新方法「區域官能基地圖(site-moiety map, SiMMap)」，希望達到三項目標: （一）為上市藥物與中草藥尋找新用途（舊藥新用）。（二）發展多標靶藥物(multi-target drugs)。（三）建立國人常見癌症之磷酸化酵素-藥物-疾病關聯性網絡資料庫，希望能藉此達到加速藥物發展及增進治療複雜疾病之效果。在本期（第一年）中我們已將相關成果發表於Biochemical Pharmacology（影響係數4.9）與Nucleic Acids Research（影響係數7.8），並即將再發表一篇期刊論文於PLoS ONE（影響係數4.4）。 本年度為五年期計畫之第二年，我們完成了以下事項： (一) 針對147個具有結晶結構之磷酸化酵素、1,216個上市藥物、89,425個天然物及330,993個化合物建立「分子間藥理作用介面家族」。(二)利用此介面家族，我們發現數十個潛力新型抑制劑及多標靶藥物(包含現有上市藥物之舊藥新用及天然物)。(三)建立磷酸化酵素－藥物－疾病關聯性網絡資料庫，包含147種具有結晶結構的磷酸化酵素、1,216個藥物與超過3萬個磷酸化酵素抑制劑、79種OMIM表型(Phenotype)。此外，我們挑選了64個潛力抑制劑及50種磷酸化酵素進行大規模磷酸化酵素－化合物之活性分析(3,200 spots)，以驗證及精進我們模型。我們也應用分子間藥理作用介面家族於瞭解藥物作用機制、尋找新型抗生素及潛在標靶蛋白質。我們與交通大學趙瑞益教授合作，建立化合物PT-262對ROCK1的可能作用機制 (已於2011年發表於期刊Biochemical Pharmacology)；與清華大學王雯靜教授合作，已找出三個μM等級之新型多標靶(multi-target)抗生素，可有效抑制幽門螺旋桿菌中的莽草酸磷酸化酵素(Shikimate Kinase)和莽草酸去氫化酵素(shikimate dehydrogenase) (審核於期刊PLoS ONE)。與義守大學袁行修醫師合作，替天然物-大金星蕨萃取物protoapigenone找出潛在標靶蛋白質-PI3K四種催化次單元（catalytic subunits）(準備投稿中)。此外，我們利用分子間藥理作用介面家族的概念建立以結構為基礎之蛋白質-胜肽抗原作用介面家族，能大規模搜尋病原體基因組數據庫找出潛在胜肽抗原(已於2011年發表於期刊Nucleic Acids Research)。 上述的研究成果已達到計畫今年之預期目標，初步的成果證實本計畫所提出的分子間藥理作用介面家族能有效地尋找新型抑制劑及了解藥物作用機制。下年度的工作構想，主要分為兩大部分：(一)藉由大規模磷酸化酵素－化合物活性分析實驗資料，修正分子間藥理作用介面作用家族之計算模型，並探討普遍性 (general inhibitors)和專一性 (specific inhibitors)抑制劑之機制，進而設計具高專一性之藥物。(二) 擴增磷酸化酵素－藥物－疾病關聯性網絡資料庫，預期囊括全部518種已知之人類磷酸化酵素、與約一百萬化合物、與磷酸化酵素相關之3,300種OMIM (Online Mendelian Inheritance in Man)基因遺傳疾病表型。我們相信當這套技術與成果，將對疾病治療、降低藥物副作用及藥物成本有所貢獻。

The current framework of drug development is cost high and time-consuming and is unable to afford the requirements for new market drugs and emerging diseases. In addition, the "one drug for one target for one disease" drug-discovery strategy cannot fully solve complex diseases, such as cancer and diabetes. Therefore, developing a new concept to discover new uses for old drugs and new multi-target inhibitors would be useful for drug development. The new uses for old drugs can be fast evaluated by clinical trials and the multi-target inhibitors can an effective approach for complex diseases through inhibiting multiple targets. In this project, we propose a novel concept "molecular pharma-interface family" and Site-Moiety Map (SiMMap) to achieve the three goals: (1) finding new uses for old drugs and natural products, (2) discovering multi-target drugs, and (3) developing a kinase-drug-cancer (disease) network database. We believe that the new concept and SiMMap should accelerate drug development process and be useful for the treatment of complex diseases. In the first year, we have achieved some preliminary results and published three papers in Biochemical Pharmacology, Nucleic Acids Research, and PLoS ONE (under revision on Impact factor: 4.4). We have achieved the following preliminary results. First, we have established molecular pharma-interface families by using 147 protein kinases with crystal structures, 1,216 marked drugs, 89,425 natural products, and 330,993 chemical compounds. Second, we have identified dozens of potential inhibitors and multi-target drugs from marked drugs and natural products by using the families. Third, we have created a preliminary kinase-drug-disease network database, which composes 147 protein kinases, 1,216 marked drugs, ~30,000 kinase inhibitors, and 79 OMIM (Online Mendelian Inheritance in Man) phenotypes. Finally, we will perform large-scale kinase profiling on 64 selected potential inhibitors and 50 kinases to verify our model. We have applied the molecular pharma-interface families to understand the mechanisms of drug action, identify multi-target antibiotics, and discover potential targets. We identified the binding mechanism of a ROCK1 inhibitor (PT-262) by cooperating with Professor Chao Jui-I, NCTU (published on Biochemical Pharmacology, 2011). Furthermore, we identified three novel multi-target antibiotics in μM level against shikimate kinase and shikimate dehydrogenase of Helicobacter pylori by collaborating with Professor Wang Wen-Ching, NTHU (revised in PLoS ONE). With the cooperation of Professor Yuan Shyng-Shiou, ISU, four catalytic subunits of PI3K were predicted as the potential target of protoapigenone. In addition, we applied molecular pharma-interface family to construct the structure-based "Peptide antigen family", which can identify the potential peptide antigens from the complete pathogen genome database (published on Nucleic Acids Research, 2011). In summary, we have achieved the anticipated results of the first year, and the results reveal that the proposed molecular pharma-interface families are useful to discover inhibitors and reveal binding mechanisms. In the following year, we will verify and refine our computational models by analyzing results of large-scale kinase profiling assays. In addition, we will focus on the design of specific inhibitors, which can reduce side effects. Second, we will extend the current kinase-drug-disease network database by collecting 518 protein kinases, ~1,000,000 chemical compounds, and 3,300 OMIM disease phenotypes related to kinases. We believe that our kinase-drug-disease network database and the molecular pharma-interface families are useful for treatment of diseases, reducing side effects, and enhancing human health.