- 學位論文
探討多抗黴素雙羥化酶催化連續羥基化及非天然腈基化之結構機轉
Structural study of polyoxin dihydroxylase from Streptomyces cacaoi
摘要
細菌對抗生素產生抗藥性是現今公共衛生領域面臨的重大挑戰之一，因此，開發新的治療策略為一迫切的需求。改造現有的核苷類抗生素以產生新的抗生素，是對抗細菌耐藥性的方法之一。因此，詳細瞭解核苷類抗生素的生物合成機制有其必要性。多抗黴素(Polyoxin)是一種真菌核苷類似物，主要由核苷骨架、多肟酸（POIA）和氨甲醯多聚草氨酸 (CPOAA) 三部分組成。多抗黴素雙羥化酶 (PolL) 是參與氨甲醯多聚草氨酸 (CPOAA) 生合成途徑的重要蛋白，它催化 α-氨基-δ-氨基甲酰基羥基戊酸 (ACV) 的連續羥基化反應，首先在四號碳骨架上催化生成第一個羥基，接著在三號碳催化上產生第二個羥基，完成氨甲醯多聚草氨酸 （CPOAA）的生合成。電腦分析 (In silico analysis) 結果顯示，PolL之立體結構應為典型的桶狀折疊 (cupin fold)，催化中心中包含具高度保留性、與二價鐵離子(Fe(II)) 結合的蛋白序列 HX(D/E)XnH。功能研究進一步顯示 PolL 需要氧氣 (O2) 和2-酮戊二酸 (ɑ-Ketoglutarate) 作為輔受質，以及二價鐵離子 Fe(II) 作為輔因子。基於這些特徵， 推測PolL應屬於非血基質鐵 (non-heme iron) /α-酮戊二酸 (α-ketoglutarate) 依賴型雙氧化酶蛋白家族 (2OGXs)。此外，此蛋白家族成員催化中心結構均具有由八股反平行的β-褶板摺疊之雙鏈 β-螺旋 (DSBH) 核心，內含由兩個組胺酸 (Histidine)與一個天門冬胺酸 (Aspartic acid) 或麩胺酸 (Glutamic acid)形成與二價鐵離子結合的模體。在催化過程中，2OGXs家族蛋白會利用具強氧化力的四價鐵離子Fe(IV)-oxo 中間物來催化各式反應。此外，目前已發現，若在受質上加上疊氮基 (azido group) 此輔助性官能基，則PolL 會由原先催化的連續羥基化反應轉變成催化非天然之腈基化反應 (nitrile installation)，顯示出2OGXs家族蛋白催化反應的多樣性。因此，詳細的了解2OGXs家族蛋白的催化機制有助於將其用於潛在的生物催化劑。 為了探索PolL如何達成立體特異性且連續之羥基化反應，本研究的首要目標是解析PolL的立體結構，以探討其催化反應的詳細機制。來自可可鏈黴菌的PolL基因已被建構於表達質體，且重組 PolL 蛋白可以成功地於大腸桿菌中表達並純化至均質。然而，純化的 PolL 在室溫且高蛋白濃度條件下並不穩定，並且在結晶篩選過程中極容易形成嚴重沉澱，且僅有球晶 (spherulites) 生成。雖然經過微調後得到晶體，但X光繞射實驗結果顯示是鹽類晶體。為了得到蛋白晶體，我們進行了緩衝溶液篩選 (buffer screen) 以尋找更合適的緩衝溶液條件，提升蛋白穩定性以促進蛋白結晶，篩選結果顯示 PolL 在磷酸緩衝溶液中展現更好的穩定性。然而，儘管 PolL 在掃晶實驗過程中沈澱情形大幅下降，但僅生成了鹽類晶體，而若將磷酸緩衝溶液濃度調降，則此時蛋白的穩定度便顯著下降，在掃晶過程當中形成較嚴重的沈澱情形。 18-冠醚-6 (18-crown-6-ether) 是一種環狀醚，會通過穩定帶正電荷氨基酸的側鏈構形來調節蛋白質表面亂度以助蛋白結晶，因此也被添加到 PolL 中。儘管添加了18-冠醚-6 的蛋白樣品確實表現出更高的穩定性，蛋白樣本能濃縮至更高的濃度 (10 mg/mL)，在掃晶實驗中沈澱的情況也較以往輕微。然而，於此同時，樣本變得非常粘稠，且生成的晶體經 X 光繞射分析顯示仍為鹽類。後又在新的結晶條件中發現球晶 (spherulites) 生成，由硬度判斷應是蛋白堆疊物，未來將可進行微調，以得到蛋白晶體。在原本的重組蛋白質體外，我們也嘗試不同的基因架構以幫助形成蛋白晶體。根據序列比對及二級結構預測的結果，我們建構一個N端截短的 PolL 蛋白以促進結晶，雖然此蛋白可以成功被大腸桿菌表達，但在純化過程中出現溶解度不佳的問題。於是，我們又再設計了另一個能夠降低的表面亂度的 PolL 突變質體 (PolL_SER)，以增加我們獲得晶體的機會。目前 PolL_SER 蛋白已可成功被大腸桿菌表達，且可以成功被純化，日後將可持續搜尋PolL_SER 的結晶條件。
並列摘要
Antibiotic resistance is one of the most serious threats to public health. Therefore, there is an urgent need to develop new therapeutic strategies for combating this menace. Repurposing of nucleoside antibiotics is one of the efficient approaches against bacterial drug resistance, and elucidating the biosynthetic pathway of nucleoside antibiotics may contribute to the design of new chemoenzymatic synthetic procedures. Polyoxin, an antifungal nucleoside analog, is made up of three components, a nucleoside skeleton, polyoximic acid (POIA), and carbamoylpolyoxamic acid (CPOAA). Polyoxin dihydroxylase (PolL), an essential component of the carbamoylpolyoxamic acid (CPOAA) biosynthesis pathway, catalyzes sequential hydroxylation, first on C-4 and then C-3, of α-amino-δ-carbamoylhydroxyvaleric acid (ACV) to form CPOAA. PolL is a member of non-heme Fe(II)-dependent ɑ-KG dioxygenase (2OGXs) family. The 2OGXs structures feature an eight-stranded distorted double-stranded β-helix (DSBH) core fold with a 2-His-1-carboxylate motif for coordinating Fe(II). Mechanistically, 2OGXs utilize an Fe(IV)-oxo intermediate to catalyze a wide range of reactions. Due to the catalytic diversity of 2OGXs, repurposing of 2OGXs by protein or substrate engineering have been attempted to redirect reaction outcome. For example, when an azido group is introduced into substrate, PolL can catalyze non-native nitrile group installation rather than canonical hydroxylation. A detailed mechanistic understanding of 2OGXs should facilitate the use of 2OGXs as potential biocatalysts. In order to explore how the regio- and stereospecific hydroxylation of PolL is achieved, the goal of my thesis research is to determine the structures of PolL. Expression plasmids were constructed based on the PolL DNA sequence from Streptomyces cacaoi subsp. asoensis. Recombinant PolL can be successfully expressed in E. coli and purified to homogeneity. However, purified PolL appears less stable at room temperature and tend to precipitate during the crystallization screen. Although crystals formed through crystallization reagent optimization, the X-ray diffraction patterns indicated they were salts. To facilitate crystallization, we conducted buffer screen to search for a more appropriate buffer condition in which PolL exhibits better stability. Although PolL appeared more stable in a 100 mM phosphate-containing storage buffer during crystallization screen, only salt crystals were identified. However, PolL became less stable and still only salt crystals were produced when phosphate concentration was reduced. 18-crown-6-ether, a cyclic ether was used to modulate protein surface entropy through stabilization of the side chains of positively-charged amino acids, was also added to PolL. Although the crown ether-supplemented sample stock did not precipitate and PolL did exhibit higher stability, the solution became quite viscous and the X-ray diffraction patterns revealed that the majority of identified crystals were salts. Nevertheless, soft spherulites which likely resulted from ordered protein packing were observed in crystallization screen. Further optimization of the condition will be performed to obtain real protein crystals. Furthermore, we also constructed a N-terminal truncated PolL to facilitate crystallization but encountered solubility problem during sample preparation. As an alternative approach, we have designed PolL mutants with reduced surface entropy to improve our chances of obtaining crystals. The crystallization of PolL_SER is currently in progress.