利用定點突變研究蝦肝胰臟核酸水解酶活性中心之重要胺基酸殘基

Abstract

蝦肝胰臟核酸水解酶（shrimp hepatopancreatic nuclease, 簡稱 shNuclease），之前被認為是一種 DNase I，它的最適 pH 值為 7.5，作用時需要二價金屬離子（鎂離子、錳離子及鈣離子）的活化，產物是帶有5’-磷酸根的寡核苷酸，在和牛胰臟 DNase I 相同的作用條件下，iodoacetate 烷化作用會對其酵素活性造成抑制，這些特性皆與牛胰臟 DNase I 非常相似。但其分子量較牛胰臟 DNase I 大，約為 45 kD；沒有醣化的現象。由胺基酸組成來看，shNuclease 具有高比例的 Glu、Gly及Cys。這些特徵又與熟知的牛胰臟 DNase Ⅰ 大不相同。後來的研究發現它在鎂離子及鈣離子的存在下，具有水解 RNA 的活性，因此更名為 nuclease。ShNuclease 是由21個胺基酸的信號胜肽（signal peptide）及381個胺基酸的成熟蛋白質組成。它含有十一個半胱胺酸，其中十個形成五對分子內雙硫鍵，另一個半胱胺酸和一外來的含硫化合物形成鍵結。自然界中，shNuclease 並沒有序列相似的蛋白質，但它的第205到255個胺基酸和 Serratia marcescens 等數種微生物的核酸水解酶具有保留性，推測此區域應為酵素活性區，且 His211 可能為其活性中心。 關於 shNuclease 的研究，一開始時我們構築在 pET15b 質體上，並試圖在 E.coli BL21 (DE3) pLysE 進行表現。嘗試以固定濃度的 IPTG（isopropyl-β-D- thiogalactopyranoside）誘導不同長度時間，或以不同濃度的 IPTG 誘導相同長度時間，卻一直無法得到具有活性的重組蛋白。因此推測可能是 shNuclease 在 E.coli 中摺疊（folding）不正確所致。於是我們更換成哺乳細胞（mammalian cell）表現系統，同時為了使 shNuclease 能夠被偵測到，重新將 shNuclease 構築到具有 His tag 的 pcDNA3.1 myc-His 質體上，並在前端加上豬脾臟去氧核糖核酸水解酶的信號胜肽，再於人類 293T 細胞進行表現。轉染含有 shNuclease 的質體後，收集細胞培養液，並製備細胞萃取物。經由西方墨點分析法以及活性染色分析法，得知具有活性的 shNuclease 重組蛋白在細胞培養液及細胞萃取物中都有表現，至此確立了可得到具有活性之 shNuclease 重組蛋白的表現系統。 另一方面，之前 Friedhoff 等人證實 Arg87、His89及 Glu127在 Serratia nuclease 的催化反應上相當重要，並且提出反應機制：His89扮演著共軛鹼（general base）的角色；Glu127 可能是必要輔因子（essential cofactor）Mg2+的 ligand；Arg87 可能藉由和受質結合，以及使蛋白質骨架位於適當位置以利於親核基攻擊（nucleophile attack）而間接參與催化反應，對應到 shNuclease 則是 Lys209、His211 及 Glu249。且之前本實驗室經由以 14C-labelled iodoacetate 對 shNuclease 做定點修飾，初步證實 His211 可能在催化活性上扮演相當重要的角色，但卻缺乏直接的證據證明。因此我們針對 His211 進行定點突變，將其置換為 Gln（H211Q）。亦將 Lys209 置換為官能基相似的 Arg（K209R）及不帶電荷的 Leu（K209L）；Glu249 置換為類似的 Asp（E249D）及不具電荷的 Gln（E249Q）。並分別表現野生型 shNuclease 及各突變株，比較其比活性的差異。將各株突變株定出和野生型的相對比活性後，發現 K209R 比活性不但沒有下降，反而上升到接近野生型的兩倍。其他四株突變株比活性都大幅下降，特別是 H211Q 及 E249Q 更是如此，顯示這些殘基在催化上的確扮演重要的角色。由於實驗結果和 Friedhoff 等人相當接近，推測 shNuclease 的催化機制可能和 Serratia nuclease 相當類似。

Shrimp hepatopancreatic nuclease（shNuclease）, previously designated as DNase I, has an optimal pH of 7.5. The catalytic activity requires divalent metal ions, such as Ca2+, Mn2+, and Mg2+. Its products are oligonucleotides with 5’-phosphate. The enzyme activity can be inhibited by iodoacetate is as bovine DNase I. These characteristics are very similar to those of bovine DNase I. However, in contrast to bovine DNase I, shNuclease is not a glycoprotein, rich in Glu, Gly and Cys and its molecular weight is 45 kD, larger than that of bovine DNase I. In the presence of Mg2+ and Ca2+, it possessed a low level of hydrolytic activity towards RNA. Therefore shNuclease, which was named incorrectly, is now referred to as shrimp hepatopancreatic nuclease. The cDNA contains an open reading frame encoding a putative 21-residue signal peptide and a 381-residue mature protein. The enzyme has 11 Cys residues, forming five intramolecular disulfides and the eleventh Cys residue form a disulfide with an unknown thiol compound. A sequence alignment search revealed no homologous proteins. However, residues 205–255 shared a conserved active-site motif within a distinct group of nucleases, including Serratia marcescens. This region was proposed as active site of the enzyme, and His211 showed to be the center of the active site. To further investigate the active site of shNuclease, we constructed shNuclease into pET15b expression vector and expressed it in E.coli BL21 (DE3) pLysE. However, after IPTG induction with different concentrations or different times, the resulted recombinant protein was inactive. We speculated that the folding of shNuclease in E.coli was not correct. To obtain active recombinant proteins, we changed the expression system from E.coli to mammalian cells. We created a new construct with the addition of porcine spleen DNase II signal sequence at the N-terminus and hexa-histidine tag at the C-terminus by insert the cDNA of shNuclease into pcDNA3.1 myc-His expression vector in order to expressed it in human 293T cell. Active recombinant shNuclease was found both in media and in the cytoplasm, as indicated by western blot and zymogram analysis. Thus, we set up a workable expression system to obtain active shNuclease. Friedhoff et al proposed that Arg87, His89, and Glu127 of Serratia marcescen are very important in catalysis. His89 is likely to function as the general base, Glu127 could be ligand of the essential cofactor Mg2+ and Arg87could be indirectly involved in catalysis by binding to the substrate and positioning the protein backbone for nucleophile attack. Arg87, His89, and Glu127 of of Serratia marcesce are equivalent to Lys209, His211, and Glu249 of shNuclease. His211 was also shown to be critical for catalysis, based on the site-specific modification with 14C-labelled iodoacetate. We therefore constructed H211Q variant to prove its importance. Because Lys209 and Glu249 also seemed to be involved in catalysis, based on sequence alignment with Serratia nuclease, four other mutants were constructed：K209R and K209L, as well as E249D and E249Q. The expressed proteins showed that the specific activities of K209R increased almost twice as that of wild type shNuclease, while those of other variants decreased dramatically, especially H211Q and E249Q, indicating the critical roles of these residues in catalysis. Our results were very closed to that of Friedhoff et al, therefore we propose that the catalytic mechanism of shNuclease may be similar to Serratia nuclease.