利用同源蛋白質-化合物複合物揭露現有藥物之新用途

Abstract

舊藥新用是對於已知功能的藥物給予新的治療用途，由於不需要進行藥物前期開發階段，如毒理試驗等，因此可減少藥物研發成本，快速進入臨床。目前開發舊藥新用的方法主要是根據相似蛋白質具有相似功能或相似化合物具有相似藥理活性的概念，但仍有相似蛋白質對相同化合物或相似化合物對相同蛋白質間的親合力不同的情況，原因是因為彼此間交互作用不相同，因此本研究引入分子交互作用的概念。為了達到舊藥新用的目的，我們以蛋白質與藥物間的分子交互作用為基礎，提出同源蛋白質-化合物複合物（homologous protein-compound complexes）的概念，一群同源蛋白質-化合物複合物中具有相同化合物結晶的蛋白質我們定義為相似藥理蛋白質 （SimpharmaPro），群內被同樣蛋白質結晶的化合物我們定義為相似藥理化合物 （SimpharmaCom），而相似藥理蛋白質與相似藥理化合物間的交互作用定義為相似藥理交互作用 （SimpharmaIntact）。我們將此概念應用於預測已知藥物潛在的標靶蛋白。為了驗證本研究的概念，我們對含有FDA核准藥物的結晶結構來建立134群同源蛋白質-化合物複合物，其中包含248種FDA核准藥物的新用途。而在134群中共有145筆我們認為具潛在交互作用的蛋白質-化合物在公開資料庫中有實驗值記錄，並且有126對蛋白質-化合物實驗紀錄≦10μM，準確率達86%。此外，我們挑選了三種藥物，並以生物實驗驗證其新用途。首先以dobutamine為例，dobutamine原為強心劑，我們找到其潛在的肺癌相關標靶蛋白質2腎上腺素受體，且經實驗驗證dobutamine在肺癌細胞株實驗IC50為56.7μM；另一實例則以podofilox為例，podofilox原為皮膚用藥，我們找到其潛在的標靶蛋白質DNA拓樸異構酶2-，DNA拓樸異構酶2-為與肺癌相關蛋白質，且podofilox在肺癌細胞株實驗IC50為2.2nM。以上的例子初步闡明同源蛋白質-化合物複合物的概念並說明本研究在「舊藥新用」上的可能性。

Drug repositioning, identifying new indications for an old drug, is able to provide additional therapeutic applications. The cost and time for drug developments are known can be reduced because the properties of old drugs, such as toxicology. It also accelerates new therapeutic strategies into clinical trials. The current approaches of drug repositioning are based on the concepts of similar proteins with similar function or similar compounds with similar pharmacology activity. However, the interactions may be different between similar proteins or similar compounds. Therefore, we introduce the concept of molecular interaction into this research. To identify the potential targets of repositioned drugs, we proposed the concept of “homologous protein-compound complexes” bases on interactions between proteins and compounds. In homologous protein-compound complexes, the proteins bind the same group of compounds are defined as “SimpharmaPro”, compounds bind the same group of proteins are defined as “SimpharmaCom”, and the interactions between SimpharmaPro and SimpharmaCom are defined as “SimpharmaIntact”. To verify the concept of this research, we collected the FDA-approved drugs with crystal structures and target proteins to construct 134 groups homologous protein-compound complexes, which contains 248 FDA-approved drugs for new uses. There are 145 potential interactions identified and annotated in public database, and 126 among them are with experimental data showing IC50 ≦ 10μM. The accuracy is 86%. To verify our concept, we select 3 potential interactions to confirm by cell toxicity experiments. The first example is dobutamine, a 1 adrenergic stimulator, originally use in the treatments of cardiogenic shock and heart failure. In our research, we predicted that dobutamine potentially targets to 2 adrenergic receptor, which is associated with lung cancer. Furthermore, the IC50 of dobutamine in a lung cancer cell line is 56.7μM. The second example is podofilox, originally use is to treating warts and keratoses. We predicted the podofilox potentially targets to DNA topoisomerase 2-, which is associated with lung cancer. Furthermore, the IC50 of podofilox in a lung cancer cell line is 2.2nM. We believe that the “homologous protein-compound complexes” proposed in this thesis can have the potential for understanding molecular binding mechanisms and giving new clues for drug repositioning for drug development.