Ferreira G C, Pedersen P L
Department of Biological Chemistry, Johns Hopkins University, School of Medicine, Baltimore, Maryland 21205.
J Biol Chem. 1992 Mar 15;267(8):5460-6.
In order to better understand why higher eukaryotic membrane proteins, in contrast to soluble proteins, are not readily expressed in Escherichia coli, the gene encoding the liver mitochondrial phosphate transporter (H+/Pi symporter) (Ferreira, G. C., Pratt, R. D., and Pedersen, P. L. (1989) J. Biol. Chem. 264, 15628-15633), was subcloned into a plasmid (pFOG402) containing the alkaline phosphatase promoter and leader sequence. Although this system is highly efficient in overexpressing soluble mitochondrial proteins in E. coli, e.g. alpha and beta subunits of the liver ATP synthase, it fails to express the H+/Pi transporter. Expression is not obtained by truncation of the transporter gene from either the 3' or 5' end, by fusing the mature transporter gene to genes encoding either the alpha or beta ATP synthase subunits, or by using different expression plasmids. Significantly, the H+/Pi transporter is overexpressed in E. coli provided its cDNA is first truncated at the 3' end (carboxyl-terminal end) and fused to a cDNA fragment derived from the ATP synthase alpha subunit gene. In fact, progressive deletions from the 3' end of the transporter cDNA produce a ladder of increasingly overexpressed fusion proteins which account from the largest to the smallest for approximately 2.5-14% of the total bacterial cell protein. The minimal truncation necessary from the 3' end is 192 base pairs corresponding to 64 COOH-terminal amino acids. This corresponds to 20% of the transporter and involves removal of one of the six predicted membrane-spanning segments. In a variety of additional experiments designed to define the molecular basis for E. coli's inability to express the complete liver H+/Pi transporter, problems related to cell toxicity and transcription were ruled out. However, in vitro transcription-translation assays revealed that the complete transporter is readily expressed when eukaryotic, but not prokaryotic, ribosomes are present. Significantly, the fused transporter gene (i.e. Pi transporter cDNA truncated at the 3' end + ATP synthase alpha subunit cDNA) is expressed when prokaryotic ribosomes are present. These results support the view that the difficulty in expressing higher eukaryotic membrane proteins in bacteria may be related in some cases to a problem at the level of translation.
为了更好地理解为何与可溶性蛋白质不同,高等真核生物膜蛋白不易在大肠杆菌中表达,编码肝线粒体磷酸盐转运体(H⁺/Pi同向转运体)的基因(费雷拉,G.C.,普拉特,R.D.,和佩德森,P.L.(1989年)《生物化学杂志》264,15628 - 15633)被亚克隆到一个含有碱性磷酸酶启动子和前导序列的质粒(pFOG402)中。尽管该系统在大肠杆菌中高效过表达可溶性线粒体蛋白方面非常有效,例如肝ATP合酶的α和β亚基,但它无法表达H⁺/Pi转运体。无论是从转运体基因的3'端还是5'端进行截短,将成熟转运体基因与编码ATP合酶α或β亚基的基因融合,或者使用不同的表达质粒,都无法获得表达。值得注意的是,如果肝H⁺/Pi转运体的cDNA首先在3'端(羧基末端)被截短并与来自ATP合酶α亚基基因的cDNA片段融合,那么它在大肠杆菌中会过表达。事实上,从转运体cDNA的3'端进行逐步缺失会产生一系列过表达程度越来越高的融合蛋白,这些融合蛋白占细菌总细胞蛋白的比例从最大到最小约为2.5% - 14%。从3'端所需的最小截短是192个碱基对,对应64个羧基末端氨基酸。这相当于转运体的20%,并且涉及去除六个预测的跨膜区段中的一个。在旨在确定大肠杆菌无法表达完整肝H⁺/Pi转运体的分子基础的各种额外实验中,排除了与细胞毒性和转录相关的问题。然而,体外转录 - 翻译分析表明,当存在真核而非原核核糖体时,完整的转运体很容易表达。值得注意的是,当存在原核核糖体时,融合的转运体基因(即3'端截短的Pi转运体cDNA + ATP合酶α亚基cDNA)会表达。这些结果支持了这样一种观点,即在某些情况下,在细菌中表达高等真核生物膜蛋白的困难可能与翻译水平的问题有关。