幾丁質酵素和幾丁質結合蛋白之基因選殖、蛋白質結構和反應機制以及其應用之研究

Abstract

本論文旨於研究幾丁質水解之相關蛋白，其中將探討仙人掌桿菌幾丁質酵素(ChiNCTU2)之結構和其反應機構。我們也將探討如何利用由粘質沙雷氏菌(Serratia marcescens)選殖所得之幾丁質酵素Chi_NCTU，生產幾丁雙醣。在企圖有效促進幾丁質水解的催化反應，我們亦引進了粘質沙雷氏菌幾丁質結合蛋白(chitin binding protein, CBP)，此蛋白質除可增進幾丁質水解速率外，我們亦利用CBP建立了一套經濟又有效之蛋白質純化系統。 仙人掌桿菌幾丁質酵素ChiNCTU2，屬於醣類水解酵素18家族，基因內含一段27個胺基酸的訊息胜肽 (signal peptide) 與一段333個胺基酸的成熟蛋白。經過胺基酸多重比對後，發現E145及Y227可能為ChiNCTU2中參與催化反應的重要胺基酸。透過定點突變與結構解析(與同步輻射陳俊榮 教授合作)後，由定點突變研究顯示，E145G突變酵素之活性較之野生株酵素約喪失2800倍，顯示此殘基在催化作用中扮演重要的角色。由數種突變酵素與幾丁寡醣之共結晶所得之結構發現，在E145Q+(NAG)2中，-1位置之醣基構形為椅型 (chair form)，而在E145Q/Y227F+(NAG)2 和E145G/Y227F+(NAG)4中，-1位置之醣基構形則為船型 (boat form)； 在E145Q/Y227F+(NAG)2 和E145G/Y227F+(NAG)4兩個共結晶結構中我們亦發現D143、E145、E190和Y193等胺基酸殘基同時存在雙重構形；由野生株酵素與突變酵素與幾丁寡醣之共結晶結構比較發現，ChiNCTU2可能利用dynamic loop的移動促進受質結合於催化位置，此特性尚未見諸於其他幾丁質水解酵素，屬於ChiNCTU2特有之性質。根據共結晶所給的資訊，我們可推測出ChiNCTU2可能的水解反應機制細節。 另一幾丁質水解酵素乃由Serratia marcescens 染色體DNA選殖而得，該酵素可成功大量於大腸桿菌中表達，一般而言，其水解產物為單醣與雙醣之混合體，利用條件的控制(20mM NaOAc, pH 5.5)我們以可大量製備幾丁二醣。此外，由分子模擬的技術已得一多重突變點(K367D368→367G368P；T416A417Y418T419→416G417G418G419G)之酵素ChiA_NCTU_m1，其產物則為幾丁單醣，雖然我們尚無法利用ChiA及其突變酵素製備参醣或更長之醣鏈產物，但我們已可成功大量製備單醣與雙醣。 同時，為了增加ChiA_NCTU水解膠狀chitin產生N-乙醯幾丁二醣的速率，我們亦從Serratia marcescens 染色體DNA選殖出幾丁質結合蛋白(chitin binding protein, CBP)，CBP具有破壞chitin直鏈結構的作用，憑藉CBP21的幫助，幾丁質酵素可加快幾丁質水解速度約20％。而在先期試驗中得知CBP在pH 8.0條件下能與幾丁質結合，而在pH<7條件其結合力遽減而被洗脫。因此，可將其建構成純化蛋白質的工具或作為實行酵素固定化之用。我們應用分生技術，將CBP基因及在其下游續接連接肽（linker）、蛋白酶切位點（protease cut site）及限制酵素多重切點（MCS）等DNA序列，建構於pRSET A表達載體上，再於限制酵素MCS位置分別嵌入數種糖類水解酵素之基因，並利用大腸桿菌進行表現。利用自製幾丁質細粒管柱進行純化，已成功純化得純度90 % 以上且回收率達40 % 以上的CBP融合的標的蛋白質。 我們進一步以CBP作為攜帶蛋白，利用(EAAAK)5連接攜帶蛋白與標的蛋白，實驗證實，利用CBP一個步驟就可以純化到融合蛋白。在實驗過程中，我們發現重複連接胜肽EAAAK二至五次均具備一相當特別的性質，在pH 6-7此連接胜肽具有自動裂解的現象，使得本純化系統不需額外使用蛋白水解酶(protease)來分離融合蛋白。目前已有數個蛋白利用建立的純化系統成功的獲得標的蛋白。

This thesis discussed the function, structure, catalysis and potential applications of proteins that involved in chitin degradation. Two bacterial chitinases were cloned and overexpressed for these subjects. Chitin-binding protein (CBP), recognized as an enhancer for the catalysis of chitinase, was further introduced in the study. An unexpected outcome was derived from the study of CBP that, consequently, allowed us to develop a simple and cost-effective system for protein purification. The results for all subjects were summarized as follows: A chitinase gene from Bacillus cereus NCTU2 (ChiNCTU2) was previously cloned and identified as a member of family 18 glycoside hydrolases. ChiNCTU2 is a hydrolytic enzyme, which cleaves β-1,4-glycosidic bonds of the natural biopolymer chitin to generate di-N-acetyl-chitobiose. The matured protein containing 333-aa was over-expressed in E. coli. Amino acid multi-alignment revealed that E145 and Y227 of ChiNCTU2 are conserved. Mutagenic study showed that the catalytic activity of E145G mutant was reduced by 2900 fold, indicating the essential role of E145 for ChiNCTU2 catalysis. However, the function of Y227 is different from that of conserved Y390 (ChiA) and Y214 (ChiB) in Serratia marcescens. Besides of the mutagenic examination, we present the crystal structure of ChiNCTU2. The structure composes of only a catalytic domain without the commonly observed chitin-binding domain in other chitinases. The structures of the cocrystallized mutants, E145Q+(NAG)2, E145Q/Y227F+(NAG)2 and E145G/Y227F(NAG)4, have been refined at high resolution and the interactions with the substrate have been characterized. The obtained results clearly show that the enzyme bends and rotates the substrate in the vicinity of the scissile bond. Furthermore, the enzyme imposes a critical “chair” to “boat” conformational change on the sugar residue bound to the -1 subsite. The presence of substrate induces a significant conformational change (about 5.7Å) on a dynamic loop (from residues I106 to V112). The distance is equal to the size of a mono-saccharide imply the possibility that the saccharide can slide along the cleft. The residues, D143, E145, E190 and Y193, were also found in double conformations. Through the kinetic analysese of various mutants, the extensive inspection of sugar-complex protein structure, and a sequence comparison with homologous chitinases, the mechanistic action of ChiNCTU2 can be extensively elucidated. Another chitinase (chiA) from Serratia Marcescens was cloned by PCR. The recombinant enzyme was characterized and tested for the preparation of chitobiose and other oligosaccharides. In general, the recombinant chtinase A exhibited an exo-type catalytic activity toward colloidal chitin and released both N-acetylglucosamine and N,N-diacetyl chitobiose as products. Although massive efforts were spent for developing mutants with anticipated power to produce uniform oligosaccharide was unsuccessful, we discovered when the enzymatic reaction was performed in sodium acetate buffer at pH 5.5 N,N-diacetyl chitobiose were produced as the predominant product; under such conditions, an enzymatic process is established for the production of the disaccharide on a 100-g scale. In order to enhance the catalytic hydrolysis of chitin by chitinase, CBP from Serratia marcescens was cloned and employed to furnish the goal. In the presence of CBP, the catalytic activity chitinase was enhanced by 15-20%. In addition to the application of CBP as an activity enhancer of chitinase, we attempted to use CBP as the protein carrier for protein purification by β-chitin affinity column. We reported herein an unexpected discovery that the presence of the repeated EAAAK peptide linker in a CBP-fusion protein possessed a pH-dependent auto-cleavage feature. An expression vector derived from pREST was constructed to compose the gene of the CBP and the nucleotide sequence of the (EAAAK)5 peptide linker following restriction sites for target gene insertion. Fusion proteins were expressed with E. coli and purified with a chitin column. In the range pH 6-7, the target protein becomes automatically released from the fusion protein without proteolytic treatment. Although the mechanism of this auto-cleavage property of an (EAAAK) 5 linker remains unclear, this feature has been successfully employed for many cases of protein purification without the tag of a fusion protein. Proteins with high purity (> 90% homogeneity for all cases) were easily obtained. The recovery yield was more than 40 %.