Heterologous expression of correctly folded eukaryotic membrane proteins is currently the primary limitation to their structure/function study. The lack of suitable expression techniques most severely limits structural determination by crystallography, which has been possible only for those eukaryotic membrane proteins that are available in high abundance from natural sources. Escherichia coli has been the most extensively used heterologous expression vehicle for prokaryotic membrane proteins, but a general finding is that eukaryotic proteins improperly fold in the E. coli membrane. This application is to develop a novel technology for correctly folded eukaryotic membrane protein production in E. coli. We will exploit the recent discovery that lipids in E. coli cells assist membrane protein folding by specific interactions with folding intermediates in a manner analogous to protein molecular chaperones. Our approach is to genetically engineer strains with lipid types and compositions approximating that of the eukaryotic organism from which the target protein to be expressed derives. The """"""""foreign"""""""" membrane proteins will therefore be expressed and assembled into membranes that more closely represent their native lipid environment. First, we will construct strains mimicking the lipids of the eukaryotic alga Chlamydomonas reinhardtii plasma membrane, and express microbial rhodopsin apoproteins encoded in the genome of that organism. Rhodopsin apoproteins form a membrane-embedded retinal-binding pocket comprised of residues from each of their seven transmembrane helices, and retinal-complexation produces a pigment with visible absorption and characteristic photochemical activity. The unique value of the microbial rhodopsins to this study is their ability to be assayed rapidly and quantitatively for proper folding by laser flash kinetic spectroscopic measurement of their photochemical reaction cycles in a large background of highly scattering cell material, such as whole cell suspensions. Second, we will construct strains mimicking the lipids of mammalian plasma and endoplasmic reticulum membranes. Target proteins will be human rod and cone visual pigments, cytochrome P450 isoforms and 12-helix human adenylyl cyclase isoforms, each having efficient assays for their proper folding. If exploiting the lipid chaperone function is successful, a breakthrough in the much-needed expression technology for eukaryotic membrane proteins will be achieved. Submitted in response to PA-03-100 High Impact/High Risk Research and PA-02-060 Structural Biology of Membrane Proteins.
