Bacteriorhodopsins purple membrane structure

Olle Edholm, Oliver Berger, and Fritz Jahnig. Structure and fluctuations of bacteriorhodopsin in the purple membrane. J. Mol. Biol., 250 94 111, 1995. 

Indeed, hydrophilic N- or C-terminal ends and loop domains of these membrane proteins exposed to aqueous phases are able to undergo rapid or intermediate motional fluctuations, respectively, as shown in the 3D pictures of transmembrane (TM) moieties of bacteriorhodopsin (bR) as a typical membrane protein in the purple membrane (PM) of Halobacterium salinarum.176 178 Structural information about protein surfaces, including the interhelical loops and N- and C-terminal ends, is completely missing from X-ray data. It is also conceivable that such pictures should be further modified, when membrane proteins in biologically active states are not always present as oligomers such as dimer or trimer as in 2D or 3D crystals but as monomers in lipid bilayers. 

Many of the early genetic studies were done on H. halobium and its related strains. The popularity of these strains stemmed from the fact that several interesting spontaneous mutations could be readily detected. Among the most characterized mutations are those that affect the production of the protein part of the purple membrane, the heavily studied light-driven proton pump bacteriorhodopsin. Spontaneous mutations occur at a frequency of lO 4. Analysis of these mutations showed that in almost every case a foreign DNA sequence was introduced into the bacterioopsin (bop) structural gene or into sequences surrounding it. 

Belrhali H, Nollert P, Royant A, Menzel C, Rosenbusch JP, Landau EM, Pebay-Peyroula E (1999) Protein, lipid, and water organization in bacteriorhodopsin crystals A molecular view of the purple membrane at 1. 9 A. Structure Fold. Des. 7 909-917... 

An additional basic difference between rhodopsin and bacter-iorhodopsin is associated with the structure of the pigment in the membrane. In contrast to the well-established rotational and translational mobility of the rhodopsin molecule in the viscous disk membrane (31-35), X-ray diffraction methods have shown that bacteriorhodopsin in the purple membrane is organized as a rigid two-dimensional hexagonal lattice with a 63 A unit cell (36-38). ... 

On the basis of novel electron diffraction methods, Unwin and Henderson have shown that each protein molecule consists of seven ot-hexical rods, about 40 A long and 10 A apart, all perpendicular to the plane of the membrane (39,40). The lattice structure results in clustering of the BR molecules as trimers around one of the threefold axes. Analysis of CD spectra in terms of exciton interactions between the chromophores (41,42), as well as linear dichroism measurements (43-45), have yielded values of 20-24° for the tilt angle (out of the membrane plane) of the retinal transition moment, which closely coincides with the long axis of the molecule. The absence of any rotational mobility of bacteriorhodopsin in the purple membrane (on a time scale of 60 min) was also confirmed by linear dichroism measurements (45). 

Bacteriorhodopsin, is a retinal-containing protein in the purple membrane of a halophilic, (salt-loving) archaebacterium, Halobacterium halobium, which pumps protons out of the cell on activation by light.The three-dimensional structure of bacteriorhodopsin resembles that of rhodopsin in the eye. 

Photoreceptor Pigments. There have been several reviews on the structures, photochemistry, and functioning of the retinal-protein photoreceptor pigments involved in the processes of visionand in the purple membrane of Halobacteria (bacteriorhodopsin). ° ° In addition to the papers quoted earlier on the spectroscopy of these pigments, many other reports have appeareddealing with rhodopsin and intermediates in its photocycle, especially photochemistry, chromophore-protein conformation and binding, and reaction kinetics. Similar studies on bacteriorhodopsin have also been described." "-"" ... 

A complex multi-chromophoric system comprises the purple membrane patches from Halobacterium salinarium. These patches are composed of about 3000 bacter-iorhodopsin proteins. The hyperpolarizability of solubilized monomeric bacterio-rhodopsin was measured by HRS and found to be 2100 x 10 esu at 1064 nm. This high value is due to the presence of a chromophore in the protein, the proto-nated Schiff base of retinal. A purple membrane patch can be treated as a two-dimensional crystal of bacteriorhodopsin proteins, and its structure is known in considerable detail. The analysis of the purple membrane tensor was performed by adding the hyperpolarizabilities of the individual proteins in the purple membrane. From (depolarized) HRS measurements on purple membrane suspensions, the structure of the purple membrane patches, and an average membrane size measured by atomic force microscopy, a fi value of 2200 x 10 esu was calculated for bacteriorhodopsin [22]. The organization of the dipolar protonated Schiff base chro-mophores in the membranes was found to be predominantly octopolar. 

Neutron diffraction of samples with added deuterated water [130], as well as hydrogen exchange studies [118], suggest that no aqueous channels exist between the protein molecules in purple membranes, or between the helices, i.e., the lipids completely fill the spaces. As expected from the crystalline structure and the relatively low lipid/protein ratio, the packing of protein in purple membrane is quite rigid, and the mobility of the protein in the lattice appears low, as determined by flash dichroism [131,132]. On the other hand, bacteriorhodopsin incorporated into liposomes at high lipid/protein ratios exists as a monomeric molecule [133], and shows rapid rotation by this criterion (relaxation time 15 /us) above the phase... 

The purple membrane lattice can be dissociated with mild detergents to yield bacteriorhodopsin monomers [45,167-169]. Dissociation will take place [167,169] without loss of the chromophore in both Triton X-100 and octylglucoside at pH 5, and virtually all of the lipids can be removed after this treatment by gel filtration in deoxycholate [170]. The significance of the lattice structure in the purple membrane is not very clear, but it appears from whole cell studies that crystalline bacteriorhodopsin is more effective in photophosphorylation than the monomeric pigment [171]. Remarkably, sodium dodecyl sulfate-denatured bacteriorhodopsin, with extensive loss of secondary structure, could be renatured to yield a product similar to native bacteriorhodopsin, which will spontaneously recrystallize [172]. 

Belrhali, H., Nollert, P., Royant, A., Menzel, C., Rosenbusch. J. P., Landau, E. M., Pebay-Peyroula, E. Protein, lipid and water organisation in Bacteriorhodopsin A molecular view of the purple membrane at 1.9 resolution. Structure with folding and design, 1999, 7, 909-917. 

Fig. 23. (A) The halophilic bacterium H. halobium with patches containing the "purple membrane" (B) Structure of the protein bacteriorhodopsin (left) and the structural formula for the chromophore retinal (right) (C) Covalent binding of retinal with iysine-216 forming a positively-charged Schiff base (D) Illumination of the bacteriorhodopsin retinal and transformation from a trans- to a cis-configuration and releases a proton from the Schiff base to the cell exterior relaxation to ttie trans-form, with uptake of a proton from the cytoplasmic interior. The combination of deprotonation and reprotonation on opposite sides of the membrane constitutes a proton pump. See text for other details. Figures partly adapted from Becker and Deamer (1991) The World of the Cell (2nd ed) Benjamin/Cummings PubI Co. p 215.

Notwithstanding this complexity, the need for three-dimensional, structural information at the atomic level of resolution is central and indispensable to biomembrane science. X-ray, and to a lesser extent neutron-diffraction, as the most important sources for such information have, therefore, been widely used in this field (for reviews, see Refs. 1-4). The success of this approach, however, has generally been less spectacular than for instance in the cases of protein or nucleic acid structure. The reasons for this lie in the very nature of biological membranes with few, notable exceptions (such as the purple membrane of halobacterium halobium, which can be viewed essentially as a two-dimensional crystal of bacteriorhodopsin with only little lipid. Refs. 5, 6,25) biological membranes are characterized by highly complex and variable molecular compositions, and by the structural dynamics, fluidity , which is in many cases essential for enzymatic, or other, functions of membranes. As a reflection of this most natural membranes do not crystallize, and a full, three-dimensional atomic structure analysis seems out of reach. 

---