區域官能基地圖揭露人類蛋白激酶之選擇性與機制

Abstract

蛋白激酶在訊息傳遞途徑上扮演重要角色，蛋白激酶若失去調控將引發一些複雜疾病，這使蛋白激酶成為許多疾病治療的熱門標的，如癌症、糖尿病、神經失調、心血管疾病等。由於518個人類蛋白激酶多具有演化上高度保留的ATP結合位，所以大多數的蛋白激酶抑制劑皆與多種蛋白激酶交互作用，這使得設計選擇性蛋白激酶抑制劑成為新興且極具挑戰的任務。 我們提出「激酶-抑制劑家族(KIF)」這個新概念，希望能有助於探討蛋白激酶選擇性與機制。我們共建構了775個KIFs，包含171,265個蛋白激酶-抑制劑交互作用、37,601個蛋白激酶抑制劑和518個人類蛋白激酶。對一個KIF中的蛋白激酶成員而言，我們用14個錨點之區域官能基地圖(SiMMap)來探討保留的蛋白激酶子結構環境與其抑制劑一致的化學官能基之間的交互作用。SiMMap中一致的錨點是子結構環境與化學官能基之間的交互作用在統計上具有顯著性，可被視為一個熱點，代表此保守的結合環境除了參與抑製劑結合外，還具生物功能。KIF的定義為：(1)蛋白激酶序列顯著相似(BLAST E-value ≤ 10-40)、(2)抑制劑之拓樸構型與官能基組成顯著相似、(3)蛋白激酶-抑制劑之結合介面(kinase-inhibitor binding interfaces)與交互作用力(interaction)相似。 透過大規模蛋白激酶分析的實驗結果顯示，在同一KIF中的蛋白激酶通常擁有相似的抑制劑。更進一步還發現SiMMap能區別不同的蛋白激酶構型，例如DFG-in, DFG-out, A-loop active, A-loop inactive，且與抑制劑選擇性有關，例如第一型抑制劑與第二型抑制劑。在我們收集的638個與疾病相關之殘基突變的實驗資料顯示，區域官能基地圖所辨識出的重要殘基(key residue)如果突變常常產生抗藥性。我們建構KIDFamMap資料庫來表達KIF以及激酶-抑制劑-疾病(KID)的關係，希望可以了解蛋白激酶選擇性與機制。此資料庫包含775個KIFs、962 個KIDs、171,265個蛋白激酶-抑制劑交互作用、37,601個蛋白激酶抑制劑、518個人類蛋白激酶、339個疾病以及638個與疾病相關之突變。我們相信KIF與KIDFamMap有助於提供生物見解與窺探蛋白激酶選擇性及機制。

Protein kinases play central roles in signaling pathways and are promising therapeutic targets for many diseases, such as cancer, diabetes, neuronal disorder and cardiovascular disease. Designing selective protein kinase inhibitors is an emergent and challenge task because 518 human protein kinases share an evolutionary conserved ATP-binding site and the majority of protein kinase inhibitors interact with multiple protein kinases. Here, we proposed a new concept “kinase-inhibitor family (KIF)” for protein kinase selectivity and mechanisms. We have constructed 775 KIFs including 171,265 protein kinase-inhibitor interactions, 37,601 protein kinase inhibitors and 518 human protein kinases. For the protein kinase members in a KIF, we used SiMMap with 14 anchors to discover the conserved structural subsites interacting with consensus moieties of their inhibitors. The consensus anchor, the subsite-moiety interactions with statistical significance, of a SiMMap can be regarded as a ”hot spot” that represents the conserved binding environments involved in inhibitor bindings and biological functions. Here, a KIF can be defined as follows: (1) the protein kinases in a KIF with the significant sequence similarity (BLAST E-value ≤ 10-40); (2) the inhibitors in a KIF with the significant topology and moiety composition similarity; (3) protein kinase-inhibitor interactions in a KIF with similar inhibitions and binding interfaces. Experimental results reveal that the protein kinases and inhibitors in the same KIF often possess the similar inhibitors based on large-scale protein kinase profiling. Furthermore, the SiMMAP anchors are able to reflect protein kinase conformations (e.g., DFG-in, DFG-out, A-loop active, and A-loop inactive), protein kinase functions (638 disease allelic variants are often the conserved interacting residues), and protein kinase selectivity (e.g., Type I, Type II, and Type III inhibitors). We build the KIDFamMap database to represent the KIF and kinase-inhibitor-disease (KID) relationships for protein kinase selectivity and mechanisms. This database includes 775 KIFs, 962 KIDs, 171,265 protein kinase-inhibitor interactions, 37,601 protein kinase inhibitors, 518 human protein kinases, 339 diseases, and 638 disease allelic variants. We believe that KIF and KIDFamMap are useful in providing biological insights and guiding the processes of discovering protein kinase selectivity and mechanisms.