內質網胺肽酶II之蛋白質表現研究與探討

Abstract

在高等的脊椎動物中，大部分的抗原胜肽是由泛素-蛋白解體途徑所產生之胜肽衍生而來。26S蛋白解體系統將病毒蛋白質降解為2-25個胺基酸之胜肽後，有少部分的胜肽會被運輸至內質網中，經ERAP1 (endoplasmic reticulum aminopeptidase 1) 和ERAP2 (endoplasmic reticulum aminopeptidase 2) 處理為成熟之抗原胜肽，進而與第一類主要組織相容性複合體 (major histocompatibility complex class I) 分子結合而呈現於細胞表面，此即為免疫系統的監視機制 (immunosurveillance)。由於ERAP1與ERAP2具有不同的受質專一性，因此，本論文欲探討造成這個差異的可能原因。 為了進一步分析ERAP1與ERAP2之特性，首先必須表現及純化全長之ERAP1及ERAP2蛋白質。然而，先前的研究結果指出，若藉由大腸桿菌表現系統大量表現，容易形成降解片段，甚至累積於不溶體中的情形。因此本論文嘗試利用桿狀病毒表現系統，表現並純化全長之ERAP1及ERAP2。然而在此系統中，仍無法順利表現出具有胺肽酶活性之ERAP2。為了探討是否是由於ERAP2之5’ 及3’ 序列的影響，造成其基因表現不易，本論文成功建構了數個5’ 或3’ 端截短之ERAP2載體以及chimera融合基因載體，並將其表現於大腸桿菌表現系統、哺乳類動物細胞表現系統以及桿狀病毒表現系統。結果顯示，N端及C端截短之ERAP2蛋白質雖可經由大腸桿菌表現，但卻是以不溶體之形式存在，因此本論文亦建立並探討了純化不溶體之最適條件與方法。除此之外，雖然ERAP1可成功地表現於293T細胞中，但是卻無法順利將帶有Kozak序列之ERAP2、3’ 截短之ERAP2或ERAP1與ERAP2的chimera基因轉染至HeLa或是293T細胞中。同樣地，chimera之重組蛋白質亦無法正常表現於桿狀病毒表現系統中。綜合以上結果，推測ERAP2之5’ 極可能具有特殊之序列或形成複雜的RNA二級結構，而造成其表現上的不易。針對ERAP1與ERAP2之RNA二級結構進行比較分析後發現，相較於ERAP1，ERAP2之RNA結構的確較為複雜，因此除了不同表現系統中密碼子使用 (codon usage) 的問題外，RNA的二級結構亦可能是造成表現困難的原因之一。

In higher vertebrates, most of the antigenic peptides are derived from peptides generated during protein degradation by the ubiquitin-proteasome pathway. The 26S proteasome is able to degrade viral proteins to peptides of about 2~25 residues long. However, a small fraction of peptides escapes complete degradation and is transported into the endoplasmic reticulum, where they are processed by ERAP1 (endoplasmic reticulum aminopeptidase 1) and ERAP2 (endoplasmic reticulum aminopeptidase 2) into mature antigenic peptides. These mature antigens then bind to major histocompatibility complex (MHC) class I molecules and are presented to the cell surface. This process is known as immunosurveillance. Since ERAP1 and ERAP2 possess substrate specificity, our primary interest is to investigate further the probable reasons behind this difference. To further analyze the unique properties of ERAP1 and ERAP2, we first need to express and purify recombinant ERAP1 and ERAP2 proteins. Previous studies have shown that if E. coli expression system is used, it is easy to observe proteins as degraded fragments or even in the form of inclusion bodies. Therefore, we tried expressing full-length ERAP1 and ERAP2 using the baculovirus protein expression system. However, in this alternate expression system, we still failed to express ERAP2 with aminopeptidase activity. In order to determine whether the 5’-end and 3’-end sequence of ERAP2 affect its ability to express, we successfully generated several 5’- or 3’-truncated ERAP2 constructs as well as chimera constructs and expressed them in E. coli expression system, mammalian expression system and the baculovirus expression system. The results have shown that, although N-terminal and C-terminal truncated ERAP2 proteins can be expression in E. coli, they accumulated as inclusion bodies. Consequently, we have set up the optimal condtions for purifying inclusion bodies. Interestingly, we have found that ERAP1 can be expressed in 293T cells while Kozak-ERAP2, C-terminal truncated ERAP2 and chimera fusion proteins cannot. Chimera fusion proteins also cannot be successfully expressed using baculovirus expression system. From the above mentioned results we have come to suspect ERAP2’s unique 5’ sequence or complex RNA secondary structure may have an effect on its expression. It is therefore not surprising when we found the RNA structure of ERAP2 is much more complicated than that of ERAP1 after comparing and analyzing the RNA secondary structures of both proteins. Thus, it is safe to conclude that other than codon usage in different expression systems, the RNA secondary structure of ERAP2 is the most likely factor contributing to expression difficulties.