Structural Membrane Proteomics
Project Group of Department of Molecular Membrane Biology (MMB)
The project aims to isolate, identify and to crystallise as many proteins and membrane protein complexes from membranes of the hyperthermophilic bacterium Aquifex aeolicus. The isolated membrane proteins and membrane protein complexes are also functionally characterised.
Proteins from hyperthermophilic organisms are considered to be more stable and more rigid than their mesophilic counterparts. Therefore, the probability of obtaining stable, homogeneous and crystallisable membrane protein complexes should be higher with complexes from thermophilic organisms than with complexes from mesophilic ones. Thus, we have chosen the hyperthermophilic eubacterium Aquifex aeolicus as our target organism. The relatively small size of its genome (less than 1/3 of that of Escherichia coli) and the consequently lower number of protein complexes implies that it is comparatively easy to purify individual membrane protein complexes. Most genes in the Aquifex genome are found in separate operons, but genes belonging to one pathway or one protein complex are often dispersed throughout the genome. Several homologues of one gene often exist, leading to the necessity to identify which homologue of one protein subunit is present in the respective complex.
We isolate and characterise as many membrane proteins and membrane protein complexes from native membranes of Aquifex aeolicus as possible by conventional biochemical techniques. Because the complete genome sequence is known, the purified proteins are identified by mass spectrometry, subjected to crystallisation trials, identification of function and mechanistic investigations. For crystallszed proteins annotated as putative proteins, we use both bioinformatics tools and biochemical experiments to indentify their function.
We have purified and characterised the NADH:quinone oxidoreductase, a possible respiratory supercomplex consisting of the cytochrome bc1 complex and of the cytochrome c oxidase, the F1Fo-ATP synthase, the sulfide-quinone oxidoreductase, and several complexes of hypothetical proteins. All protein complexes that we have purified to homogeneity in sufficient quantities, have yielded diffracting crystals (Fig. 2). Crystals of two membrane protein complexes diffract X-rays to resolutions better than 2 Å. One of them is sulfide-quinone oxidoreductase, its structure has been solved at 2 Å resolution.
Examples of purified membrane protein complexes
Aquifex respiratory complex I is highly stable and active, the specific activity for electron transfer from NADH to Q10 of the isolated complex is 29 U/mg at 80ºC with a half-life of about 10 h. Single particle electron microscopy revealed many details in its cytoplasmic arm, a pronounced invariant angle (90º) between the cytoplasmic arm and the membrane arm indicates a good preservation of the enzyme and a homogeneous preparation, a promising property for crystallisation attempts. Recently, the structure of the enzyme hydrophilic domain has been determined using X-ray crystallography.
We have identified all subunits of the F1Fo-ATP synthase. Interestingly, two versions of the b subunit, b1 and b2, with only low sequence homology to each other, were found. Electron microscopic single particle analysis displays the structural details with particular emphasis on the peripheral stalk, and the central stalk appears to be tilted and/or kinked shown in several orientations.
Sulfide-quinone oxidoreductase (SQR) is a monotopic membrane enzyme that belongs to the superfamily of flavoprotein disulfide reductases. It catalyses sulfide oxidation to polysulfur through direct flavin-quinone electron transfer. The enzyme could be purified to a highly stable, homogeneous and active form from A. aeolicus. We could solve its 3-D X-ray structure at 2.0 Å by de novo phase determination. The protein is a homotrimer. The structure reveals how the SQR interacts with the membrane and how it selectively binds sulfide and quinone analogues. Most interestingly, it shows that FAD is bound to the protein through an unusual putative persulfide bridge and that the product of the reaction is covalently bound to the active site and cocrystallises with the enzyme. Through the structure, we obtained a detailed insight into A. aeolicus SQR and we formulated a new hypothesis on the mechanism by which this interesting enzyme may catalyse sulfur-polymerisation.
One of the best diffracting crystals (at 1.8 Å) was obtained with the hypothetical protein Aq_1862, an outer membrane porin protein. The protein is one of the most abundant proteins in Aquifex membranes. We have analyzed the function of this porin after incorporation into black lipid membranes (BLM). A sequence comparison demonstrated homology to two classes of outer membrane porins, the phosphate-selective porin O and P family and the putative ammonium transporter FmdC. Secondary and tertiary structure predictions (using PRED-TMBB and 3D-PSSM) predicted an 18-stranded b-barrel as well as similarity to the fold of the outer membrane porin PhoE of E. coli.
AQ_1760 is another “hypothetical protein”. It is not predicted to be a membrane protein, but it is always present in the membrane fraction, neither in the cytosol, nor in the culture medium. A sequence comparison demonstrated that it contains the Linocin_M18 bacteriocin domain. Proteins sharing this domain show various functions including antimicrobial effects and a proteolytic activity. All form homo-oligomeric complexes.
In a collaboration with the Institute of Oceanology of the Chinese Academy of Sciences, Qingdao, we isolate and try to crystallize protein complexes from algae and sea animals. In a collaboration with Prof. V. Mueller (Frankfurt University, Frankfurt am Main), we are also crystallizing A1Ao-ATP synthases from archaea.