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和應付緊急疫情。再者，過去的「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)上也已經遇上瓶頸。在這樣的情況下，「舊藥新用」因為成本較低，而且可以迅速進入第二期臨床實驗觀察其效果，因此成為愈來愈重要的議題。在本計劃中，我們提出一個新概念「分子間藥理作用介面家族(molecular pharma-interface family)」，以及新方法「作用位置分割地圖(site-moiety map, SiMMap)」，希望達成三項目的：為若干種上市藥物(包括中草藥)找出新用途、開發多標的(multi-target)藥物，並且建立一個針對國人常見癌症的磷酸化酵素-藥物-疾病關聯性網絡資料庫。

現在全球以分子立體結構為基礎的藥物設計工具通常難以找出蛋白質-配體結合介面上控制其生物功能的關鍵性位置(例如藥效基團位置)，針對這個問題，本研究團隊在過去已發展出一套高準確率的分子嵌合工具GEMDOCK，迄今累積了近十年的一系列研究，是國內少數有能力自行研發藥物設計技術的團隊；我們也積極與國內外實驗室合作，已先後為幾個重要病原體找出抑制劑(≤10 μM)，如登革病毒、流感病毒、胃幽門螺旋菌等等，且經過實驗證實有抑制效果。GEMDOCK 亦曾於2007 年榮獲國家新創獎。

目前我們已提出一套藥效基團分析技術(即SiMMap)，可以將蛋白質-配體結合介面依其立體形狀及藥效基團之性質做分群，形成「分子間藥理作用介面家族」，屬於同一家族的蛋白質也會與同一小分子化合物結合。換言之，這一群蛋白質共有一塊由若干物化性質相似、空間位向相近的胺基酸片段組成的空間區域。SiMMap 能定位出這些藥效作用區域，進一步為上市藥找出未知的標的蛋白質(target protein)，或用於開發新藥、及評估藥物副作用。

在本計劃中，本團隊將利用過去在系統生物學上累積的研究優勢(包括已發表的「局部結構搜尋技術」等)，與GEMDOCK 和SiMMap 結合，針對舊藥新用、開發新藥及建立國人常見癌症的磷酸化酵素-藥物-疾病關聯性網絡資料庫三項重要議題進行研究。我們是全球第一個提出「分子間藥理作用介面家族」概念的團隊，而我們的計算結果將與生物實驗室合作，加以驗證。當我們將這套觀念運用在分析國內社會常見癌症如肺癌、肝癌、腸癌、胰臟癌時，我們的初步成果顯示， 上市藥物imatinib 除了能抑制KIT 和ABL1 磷酸化酵素而達到治療胃腸道基質瘤(gastrointestinal stromal tumor)和慢性骨髓細胞白血病(chronic myeloid leukemia)的效果之外，MAPK8(腸癌、胰臟癌)、MAPK10(腸癌、胰臟癌)、KDR(血管增生，與肝癌及非小細胞肺癌有關)也具有可與KIT 及ABL1 歸屬於同一藥理作用介面家族的配體結合位置，換言之，imatinib 也可能抑制MAPK8、MAPK10 及KDR；這一點在比對前人發表的實驗論文後得到了證實。換句話說，這是研究舊藥新用的一個良好起點。我們相信當這套技術與研究成果成熟，將對疾病治療、降低藥物副作用及藥物成本，以及增進國民健康有所貢獻。

The current costly and time-consuming framework of drug development is often 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, to discover a new drug from old drugs is a promising approach since old drugs are often approved by regulatory agencies for human and the new uses can be fast evaluated in phase II clinical trial. Here, we will propose a novel concept "molecular pharma-interface family" and Site-Moiety Map (SiMMap) for finding new uses (or side effects) for old drugs (or Chinese herbs), discovering multi-target drugs, and developing a kinase-drug-cancer (disease) network database.

As protein structures increase rapidly, structure-based drug design and virtual screening approaches are becoming important and helpful in drug discovery. However, the major weakness of virtual screenings is likely due to incomplete understandings of ligand binding mechanisms and the subsequently imprecise scoring algorithms. These approaches generally cannot identify the key binding environments (e.g., pharmacophore spots) on protein-ligand interfaces that are essential to trigger or block the biological responses of the target protein. To address these issues, we have developed a molecular docking toolkit, namely GEMDOCK, which is one of wide-used tools in the world. We have cooperated with over 10 web laboratories for identifying lead compounds (such as dengue virus envelop protein, influenza virus neuraminidase, and anti-cancer inhibitors) and protein mechanisms (such as cAMP Receptor-Like Protein CLP and milk beta-lactoglobulin). Due to the achievement of GEMDOCK, we got the 2007 National Innovation Award.

Recently, we have reported a structure-based pharmacophore (SiMMap) to show that members of molecular pharma-interface family often share a homologous fold and conserved structural features for interaction with chemically similar ligands throughout evolution. For example, protein kinases and isozymes that catalyze the same chemical reactions with different molecular structures or sequences are similar in binding environments; that is, these proteins are often share a set of space-related segments with similar physico-chemical properties and geometric shapes. The SiMMap can establish core binding environments, which comprise both conserved binding sub-sites among proteins of molecular pharmainterface family and compound moieties, to discover the inhibitors of these proteins.

In this project, we will apply structural systems biology and Space-Related Pharmamotif (derived form our published method, 3D-BLAST) to enhance GEMDOCK and SiMMap for developing kinasedrug-disease network databases, discovering new kinase inhibitors for cancers, and identifying new uses for old drugs (potent lead compounds) and Chinese herbs. Protein kinases have a fundamental role in signal transduction pathways and play important roles in many cancers, such as non-small cell lung cancer (NSCLC), colorectal cancer, gastrointestinal stromal tumor (GIST), and leukemia. To our best knowledge, our approaches and databases are the first to acquire pharmacophore models (site-moiety map) and networks for understating biological mechanisms and identifying multi-target inhibitors based on molecular pharma-interface family. These computational results will be evaluated by bioassays. Our preliminary results show that the drug imatinib can inhibit kinases KIT and ABL1 for GIST and chronic myeloid leukemia; kinases MAPK10 and MAPK8 for colorectal and pancreatic cancers; and kinase insert domain receptor (KDR) for some solid tumors (e.g., hepatocellular carcinoma and NSCLC). We believe that the kinase-drug-disease network and molecular pharma-interface family are useful for new uses for old drugs and multi-target drugs. In addition, our results will make a great impact on public health.