Four other membrane transporters of Methanococcus jannaschii were as well studied: a putative phosphate permease transporter (designated Phopp), a sodium-decarboxylate transporter (designated Dassp), a sodium transporter (designated Nssp) and a potassium uptake protein (designated Trkp). These four transporters were cloned using the same method as for Aatp, and we tried to heterologously overexpress them in Saccharomyces cerevisiae, with the final aim to purify them and perform crystallisation trials. As for Aatp, the hyperthermophilic nature of these transporters was expected to facilitate the purification in a heat treatment step.
7.1.1. Expression constructs
As with the AAT gene, each transporter genes PHOP, DASS, NSS and TRK were inserted into the multi-copy E.coli / yeast shuttle vector pITy-QC resulting in expression vectors pPHOP, pDASS, pNAA and pTRK (see the section material and methods, chapter 13.1.1.2). The constructs were generated so as to encode a C-terminally His6 tagged variant of each gene for facilitated purification by metal-affinity chromatography. Yeast expression strain BJ5464 was transformed with all of these constructs.
7.1.2. Selection of suitable culture conditions by expression level
The expression of the 4 putative membrane transporters was tested unter the same conditions as those described for Aatp (see chapter 6.2): growth in selective medium to an OD600 between 4 to 5, and induction with galactose for 8h.
As with Aatp, total membranes were isolated from each induced yeast strain and stripped under high salt conditions in order to remove extrinsic proteins, and to enrich the sample in the desired transport protein. On a Western-blot, the total (stripped) membranes were analysed for expression of the transporters and total stripped membranes of a strain expressing Aatp were used as the reference in the densitometric analysis (Fig. 52). There is clear evidence that Phopp, Dassp, Trkp and Nssp were extremely poorly expressed in comparison to Aatp.
Fig. 52: Comparison of the expression level of Aatp with that of Phopp, Dassp, Trkp and Nssp in total stripped yeast membranes. Total membranes were isolated from induced expression strains and 5 µg were loaded on a Western-blot gel. The anti-His6 antibody used to probe the Western-blotlabeled specifically the proteins of interest. The negative control presents membranes from untransformed yeast cells.
Aatp Phopp Dassp Trkp negative Aatp Nssp control
non-specific bands
In the expression system and under the conditions we used, none of the 4 putative membrane transporters of Methanococcus jannashii was overexpressed, or even expressed at a level sufficient to obtain enough material for further investigations. It seems that our Saccharomyces cerevisiae expression system was only efficient for Aatp.
In consequence, we did not perform further investigations with these four transporters.
Discussion
Membrane proteins are vital elements in the communication of the cell with its environment.
Among them, the G-protein coupled receptors (GPCRs) have raised a great deal of interest lately since they comprise around one third of the pharmaceutically promising drug targets. Their structural analysis is of outstanding medical and pharmaceutical interest in order to understand the mechanism of drug action. However, we have limited knowledge of membrane protein structure and function in general and of GPCRs in particular. So far, there is only one 3D-structure of a GPCR resolved, which is that of the bovine rhodopsin (Palczewski et al., 2000).
The membrane proteins from extremophiles such as hyperthermophilic archaea are also of scientific interest. It was found that many of them could be phylogenetically related to membrane proteins of eukaryotic organisms. As these membrane proteins naturally function in organisms that are extremely resistant to high temperatures, high pressure and extreme pH, they are expectedly more robust than their mesophilic counterparts. This robustness can be very useful to overcome the problems usually encountered during solubilisation and purification of integral membrane proteins.
Hyperthermophilic proteins usually contain a higher proportion of amino-acid residues capable of forming intrinsic forces of stabilisation like salt bridges, hydrogen bonds or hydrophobic interactions, this leading to an expected higher stability of the transmembrane domain interactions under high temperatures (Shiraki et al., 2001; Jaenicke and Bohm, 1998; Vetriani et al., 1998;
Jaenicke et al., 1996). For example, a comparison between the glutamate dehydrogenase of the hyperthermophile Thermococcus litoralis and the one from Pyrococcus furiosus showed that the less stable T.litoralis enzyme has a decreased number of ion pair interactions, modified patterns of hydrogen bonding, and substitutions that decrease its packing efficiency (Britton et al., 1999).
Hyperthermophilic membrane proteins are also thought to present a better helical packing, making them more resistant to the decrease of lateral pressure (Engelman et al., 2003; Schneider et al., 2002).
So, thermostability is possibly caused by numerous subtle sequence differences modifying the intrinsic stabilisation forces in the proteins. If these forces are disturbed during the solubilisation and/or the purification procedure, loss of hyperthermostability would result (Arnott et al., 2000).
Until now, several soluble proteins from hyperthermophilic organisms have been overexpressed, purified and crystallised. However, only 7 structures of hyperthermophilic membrane proteins (López-Marqués et al., 2005; Ferreira et al., 2004; Van den Berg et al., 2004; Kamiya and Shen, 2003; Jiang et al., 2002b; Jiang et al., 2002a; Nogi et al., 2000; Soulimane et al., 2000b;
Soulimane et al., 2000a)), and only one structure of a hyperthermophilic amino-acid transporter, the glutamate transporter homologue from Pyrococcus horikoshii (Yernool et al., 2004) have as yet been resolved.
Though there is plenty of experience with the expression, purification and crystallisation of soluble proteins, we are indeed far from having the same possibilities for membrane proteins. Only few generally applicable methods and protocols have been developed that can be referred to, so that the field of membrane protein study appears rather vague, and every new study is actually pioneering work. The challenge to produce high amounts of correctly folded, biologically active, and pure protein for biophysical, biochemical and structural analyses becomes even more demanding with integral membrane proteins that are naturally low abundant and whose overexpression is confined to the biomembrane as the natural environment.
This study’s goal was to acquire high resolution 3D structural data of integral membrane proteins.
For the crystallisation experiments, we had to provide chemically and structurally homogenous, and highly purified membrane protein preparations in high quantities (mg).