- Email: [email protected]
Three-dimensional structure of bacteriorhodopsin
311 The molecular dynamics simulations demonstrated that the neurotransmittermolecules fluctuated rapidly between different conformations as they approach the receptor. In simulations with the neurotransmitter at the putative binding site in the central core of the receptors, the positively charged amino group in the neurotransmitter became oriented towards the negatively charged aspartie residues in helices 2 and 3, and was able to introduce conformation changes in the side chain of these residues. The simulations clearly demonstrated that neurotransmitter-receptor interactions should be regarded as dynamic processes, in order to truly understand the molecular mechanisms.
References Weiner, S. L, Kollman,P. A., Case, D. A., Singh, U. C., Ghio, C., Alagona,G., Profeta Jr., S. and Weiner, P. L, 1984, A new force field for molecularmechanicalsimulationof nucleic acids and proteins. J. Am. Chem. Soc. 106, 765. Weiner, S. J., Kollman,P. A., Nguyen, D. T. and Case, D. A., 1986, An all atom force field for simulationsof proteins and nucleic acids. J. Comput. Chem. 7, 230. Wang, H., Lipfert, L., Malbon, C. C. and Bahouth, S., 1989, Site-directedanti-peptideantibodiesdefine the topography of the I~-adrenergic receptor. J. Biol. Chem. 264, 14424. Dahl, S. G., Edvardsen, O. and Sylte, I., 1991, Moleculardynamicsof dopamineat the D2 receptor. Proe. Natl. Acad. Sci. U.S.A., in press.
Three-dimensional structure of bacteriorhodopsin
Ceska, T.A.*, Henderson, R., Baldwin, J.M., Zemlin, F.**, Beckmann, E.** and Downing, K.*** MRC Lab. of Mol. Biol., Hills Road, Cambridge CB2 2QH, U.K., *Current address: EMBL, Meyerhofstrasse 1, 6900 Heidelberg, F.I~ G., **Fritz-Haber-blstitut der Max-Planck-Gesellschafi, Faradayweg 4-6, D IO00 Berlin 33, F.R~G., ***Cell and Molecular Biology Division, Lawrence Berkeley Lab., Berkeley, CA 94720, U.S.A. Key words: Bacteriorhodopsin, Electron crystallography, Structure
The structure of bacteriorhodopsin, the protein component of purple membrane from Halobacterium halobium has been analysed by electron diffraction and electron microscopy (Henderson et al, 1990). The procedures that have been developed to enable an atomic model of the molecule to be constructed will be described. So far, a resolution of about 3.5A has been achieved, and 2.5A should be achievable. The structure of bacteriorhodopsin was originally determined to a resolution of 7A by unstained electron microscopy (Henderson and Unwin, 1975), and shown to consist of 7 transmembrane rods of density presumed to be alpha helices. Several attempts were made to obtain high resolution phases, for which the high resolution diffraction amplitudes can be easily collected. Using the high resolution diffraction data from two different crystal forms, Rossmann and Henderson (1982) described extending the phases from 6A to 3.3A in projection. Ceska and Henderson (1990) studied the use of heavy atom derivatives of purple membrane in obtaining phase information, in an analogous manner to the way phase information is obtained for protein X-ray crystallography. These approaches to determine the high resolution structure of bacteriodlodopsin achieved only limited success. The most successful approach to obtain the phases of the high resolution diffraction amplitudes has been to analyse images of purple membrane which contain high resolution information. Obtaining these images is singularly difficult. Several problems had to be overcome to obtain an atomic resolution model of the bacteriorhodopsin structure. The special problem of electron microscopy of unstained biological molecules is that the electron exposure that is allowed is limited by the radiation damage from the electron beam. This produces extremely low contrast images in which individual molecules cannot be discerned. The quantum noise from the small number of electrons that form the image exceeds the signal from the contrast between the atoms of the structures by a factor of 10 or 20. This means that very large areas of the specimen must be averaged to obtain a statistically significant structure.
312 Another problem concerns the recording of images of highly tilted specimens, which are needed for the 3-dimensional structure determination. The large height change from one side of a tilted object to the other, if it is in the form of an extended 2-dimensional crystal, means that special steps are necessary to allow the retrieval of the structural information. The three-dimensionalmap of bacteriorhodopsin has been obtained, at near-atomic resolution, by collecting and analysing electron diffraction patterns and electron micrographs from crystals of bacteriorhodopsin preserved at very low temperatures. The map shows a resolution of 3.5A in a direction parallel to the plane of the membrane, but poorer resolution perpendicularly. It shows many features well resolved from the main density of the seven alpha-helices, which we interpret as the bulky sidechains of tyrosine, phenylalanine and tryptophan as well as a very dense feature which is the beta-ionnne ring of the retinal chromophore. Using these bulky sideehains as starting points and taking account of bulges of density for the smaller sidechains such as leucine, an atomic model for the residues between 8 and 225 has been built. There are 21 amino acids from all 7 helices surrounding the retinal, and 26 amino acids contributed by 5 helices that form the proton channel. Ten of the amino acids in the middle of the proton channel are also part of the retinal-binding site. The model provides a useful basis for considering the mechanism of proton pumping and in the interpretation of other experimental data. In particular, the model suggests that the pK changes in the Schiff base must act as the means by which light energy is converted to proton pumping through the channel. Asp-96 is on the pathway from the cytoplasm to the Schiff base and asp85 on the pathway from the Schiff base to the extracellular surface. The structure of bacteriorhodopsin will most likely be relevant to understanding the structure of other 7-helix membrane proteins by providing a framework for the membrane components of these proteins. These 7-helix membrane proteins include those mediating their action through a G-protein cascade (such as visual rhodopsin), as well as other membrane receptors for hormones and neuropeptides which mediate their action through other second messengers. References Ceska, T.A. and Henderson,R., 1990, Analysisof high-resolutionelectron diffractionpatterns from purple membranelabelled with heavy atoms. J. Mol. Biol. 213, 539-560. Henderson,R. and Unwin,P.N.T., 1975,Three dimensionalmodelof purplemembraneobtainedby electronmicroscopy.Nature257, 28-32. Henderson, R. and Sehertler, G.F.X., 1990,The structure of bacteriorhodopsinand its relevance to the visual opsins and other seven-helix G-protein coupled receptors. Phil. Trans. R. Soe. Lond. B326, 379-389. Henderson,R., Baldwin,J.M., Ces~, T.A.,Beckanann,E., Zemlin,F. and Downing,K.H., 1990,Model for the structureof bacteriorhodopsin based on high-resolutionelectron co,o-microscopy. L Mol. Biol. 213, 899-929. Rossmann,M.G. and Henderson,R,, 1982,Phasingelectrondiffractionamplitudeswith the molecularreplacementmethod. Acta Co,st. A38, 13-20.
References Weiner, S. L, Kollman,P. A., Case, D. A., Singh, U. C., Ghio, C., Alagona,G., Profeta Jr., S. and Weiner, P. L, 1984, A new force field for molecularmechanicalsimulationof nucleic acids and proteins. J. Am. Chem. Soc. 106, 765. Weiner, S. J., Kollman,P. A., Nguyen, D. T. and Case, D. A., 1986, An all atom force field for simulationsof proteins and nucleic acids. J. Comput. Chem. 7, 230. Wang, H., Lipfert, L., Malbon, C. C. and Bahouth, S., 1989, Site-directedanti-peptideantibodiesdefine the topography of the I~-adrenergic receptor. J. Biol. Chem. 264, 14424. Dahl, S. G., Edvardsen, O. and Sylte, I., 1991, Moleculardynamicsof dopamineat the D2 receptor. Proe. Natl. Acad. Sci. U.S.A., in press.
Three-dimensional structure of bacteriorhodopsin
Ceska, T.A.*, Henderson, R., Baldwin, J.M., Zemlin, F.**, Beckmann, E.** and Downing, K.*** MRC Lab. of Mol. Biol., Hills Road, Cambridge CB2 2QH, U.K., *Current address: EMBL, Meyerhofstrasse 1, 6900 Heidelberg, F.I~ G., **Fritz-Haber-blstitut der Max-Planck-Gesellschafi, Faradayweg 4-6, D IO00 Berlin 33, F.R~G., ***Cell and Molecular Biology Division, Lawrence Berkeley Lab., Berkeley, CA 94720, U.S.A. Key words: Bacteriorhodopsin, Electron crystallography, Structure
The structure of bacteriorhodopsin, the protein component of purple membrane from Halobacterium halobium has been analysed by electron diffraction and electron microscopy (Henderson et al, 1990). The procedures that have been developed to enable an atomic model of the molecule to be constructed will be described. So far, a resolution of about 3.5A has been achieved, and 2.5A should be achievable. The structure of bacteriorhodopsin was originally determined to a resolution of 7A by unstained electron microscopy (Henderson and Unwin, 1975), and shown to consist of 7 transmembrane rods of density presumed to be alpha helices. Several attempts were made to obtain high resolution phases, for which the high resolution diffraction amplitudes can be easily collected. Using the high resolution diffraction data from two different crystal forms, Rossmann and Henderson (1982) described extending the phases from 6A to 3.3A in projection. Ceska and Henderson (1990) studied the use of heavy atom derivatives of purple membrane in obtaining phase information, in an analogous manner to the way phase information is obtained for protein X-ray crystallography. These approaches to determine the high resolution structure of bacteriodlodopsin achieved only limited success. The most successful approach to obtain the phases of the high resolution diffraction amplitudes has been to analyse images of purple membrane which contain high resolution information. Obtaining these images is singularly difficult. Several problems had to be overcome to obtain an atomic resolution model of the bacteriorhodopsin structure. The special problem of electron microscopy of unstained biological molecules is that the electron exposure that is allowed is limited by the radiation damage from the electron beam. This produces extremely low contrast images in which individual molecules cannot be discerned. The quantum noise from the small number of electrons that form the image exceeds the signal from the contrast between the atoms of the structures by a factor of 10 or 20. This means that very large areas of the specimen must be averaged to obtain a statistically significant structure.
312 Another problem concerns the recording of images of highly tilted specimens, which are needed for the 3-dimensional structure determination. The large height change from one side of a tilted object to the other, if it is in the form of an extended 2-dimensional crystal, means that special steps are necessary to allow the retrieval of the structural information. The three-dimensionalmap of bacteriorhodopsin has been obtained, at near-atomic resolution, by collecting and analysing electron diffraction patterns and electron micrographs from crystals of bacteriorhodopsin preserved at very low temperatures. The map shows a resolution of 3.5A in a direction parallel to the plane of the membrane, but poorer resolution perpendicularly. It shows many features well resolved from the main density of the seven alpha-helices, which we interpret as the bulky sidechains of tyrosine, phenylalanine and tryptophan as well as a very dense feature which is the beta-ionnne ring of the retinal chromophore. Using these bulky sideehains as starting points and taking account of bulges of density for the smaller sidechains such as leucine, an atomic model for the residues between 8 and 225 has been built. There are 21 amino acids from all 7 helices surrounding the retinal, and 26 amino acids contributed by 5 helices that form the proton channel. Ten of the amino acids in the middle of the proton channel are also part of the retinal-binding site. The model provides a useful basis for considering the mechanism of proton pumping and in the interpretation of other experimental data. In particular, the model suggests that the pK changes in the Schiff base must act as the means by which light energy is converted to proton pumping through the channel. Asp-96 is on the pathway from the cytoplasm to the Schiff base and asp85 on the pathway from the Schiff base to the extracellular surface. The structure of bacteriorhodopsin will most likely be relevant to understanding the structure of other 7-helix membrane proteins by providing a framework for the membrane components of these proteins. These 7-helix membrane proteins include those mediating their action through a G-protein cascade (such as visual rhodopsin), as well as other membrane receptors for hormones and neuropeptides which mediate their action through other second messengers. References Ceska, T.A. and Henderson,R., 1990, Analysisof high-resolutionelectron diffractionpatterns from purple membranelabelled with heavy atoms. J. Mol. Biol. 213, 539-560. Henderson,R. and Unwin,P.N.T., 1975,Three dimensionalmodelof purplemembraneobtainedby electronmicroscopy.Nature257, 28-32. Henderson, R. and Sehertler, G.F.X., 1990,The structure of bacteriorhodopsinand its relevance to the visual opsins and other seven-helix G-protein coupled receptors. Phil. Trans. R. Soe. Lond. B326, 379-389. Henderson,R., Baldwin,J.M., Ces~, T.A.,Beckanann,E., Zemlin,F. and Downing,K.H., 1990,Model for the structureof bacteriorhodopsin based on high-resolutionelectron co,o-microscopy. L Mol. Biol. 213, 899-929. Rossmann,M.G. and Henderson,R,, 1982,Phasingelectrondiffractionamplitudeswith the molecularreplacementmethod. Acta Co,st. A38, 13-20.