J. Mol. Biol. (1989) 210, 829-847 Tertiary Structure of Bacteriorhodopsin Positions and Orientations of Helices A and B in the Structural Map Determined by Neutron Diffraction Jean-Luc Popott Institut de Biologie Physico-Chimique and Collége de France, 13 rue Pierre et Marie Curie F-75005 Paris, France Donald M, Engelman Yale University, Department of Molecular Biophysics de Biochemistry New Haven, CT 06511, U.S.A Ogan Gurelt and Giuseppe Zaccai Institut Laue-Langevin and CNRS U.A. 1333, 156X, F-38042 Grenoble, France (Received 4 April 1989, and in revised form 10 August 1989) Positions and rotations of two helices in the tertiary structure of bacteriorhodopsin have been studied by neutron diffraction using reconstituted, hybrid purple membrane samples. Purple membrane was biosynthetically *H-labeled at non-exchangeable hydrogen positions ‘of leucine and tryptophan residues. Two chymotryptic fragments were purified, ‘encompassing either the first two or the last five of the seven putative transmembrane segments identified in the amino acid sequence of bacteriorhodopsin. The 7H-labeled fragments, diluted to variable extents with the identical, unlabeled fragment, were mixed with their unlabeled counterpart; bacteriorhodopsin was then renatured and reconstituted. ‘The crystalline purple membrane samples thus obtained contained _ hybrid bacteriorhodopsin molecales in which certain transmembrane segments had been selectively 2HLlabeled to various degrees. Neutron diffraction powder patterns were recorded and analyzed both by calculating difference Fourier maps and by model building. The two analyses yielded consistent results. The first and second transmembrane segments in the sequence ‘correspond to helices 1 and 7 of the three-dimensional structure, respectively. Rotational orientations of these two helioes were identified using best fits to the observed diffraction intensities. The data also put restrictions on the position of the third transmembrane segment. These observations are discussed in the context of folding models for bacteriorhodopsin, the environment of the retinal Schiff base, and site-directed ‘mutagenesis experiments. {{ Author to whom all correspondence should be sent. iP ka 4 Prosont address: Columbia University, College of Physicians & Surgeons, New York, NY 10082, U.S.A. {§ Abbroviations used: BR, bacteriorhodopein; C-1 chymotryptie fragment of BR, residues 72 to 248; C2, It is now more than a decade since the low- resolution structure of bacteriorhodopsin (BR&) revealed that it has seven transmembrane helices shymottyptic fragment of BR, residues 1 to 71; (Unwin & Henderson, 1975; Henderson & Unwin, D,-PITC, “H-labeled phenylisothioeyanate; em. 1975), and almost as long since the amino acid electron microscopy; num. nuclear magnetic resonance; sequence was determined (Ovchinnikov et al, 1979; PM, purple membrane; s0., standard deviation Khorana ef al., 1979). Seven putative trans 820 (0022-2836/80240820-19 408.0010 . © 10960 Academic Pres Limited830 IL, Popot ot al, STAAAGDGASPEP AE, shRPEVA (b) Figure 1, (a) Transmembrane folding model of BR (from Engelman of al., 1982). Residues that carried 7H in the present ‘periments are indicated by filled aquares. Chymotrypsin cleaves bacterioopsin after PheT! (\). Fragment C-2 covers residues I to 71, fragment C-I covers residues 72 to 248. (b) Distribution of “A atoma inthe 7 helices, projected on. ‘plane normal to the axis of the helix (shown by +). membrane segments have been identified within fairly narrow limits of uncertainty by hydrophobi- city analysis of the sequence and a number of modi- ation or immunolabeling studies (ngelman etal 1982, 1986; and references therein). ‘The aim of obtaining a chemical model for BR by combining ‘the structural and sequence data has been approached by a variety of means, yet no reliable positioning of the polypeptide chain in the electron density map has emerged so far. ‘Neutron diffraction coupled with *H labeling has been used to identify features of the molecule in planar projection, most notably the position of the retinal chromophore (Jubb et al, 1984; Seiff et al., 1985, 1986a; Heyn ef al., 1988). As the retinal is linked by a Schiff base to lysine 216, which lies in helix 7, these studies place constraints on the poss- ‘ble location of that part of the polypeptide chain Further modification or labeling studies have been directed to other parts of the molecule (see Discussion). To this day, however, attempts to posi tion any of the transmembrane sequence segments have been plagued by major uncertainties. BR has been labeled biosynthetically by inoor- poration of *H-labeled amino acids. Analysis of the neutron diffraction pattems followed two strategies difference Fourier methods (Engelman & Zaceai, 1980) and model building (Trewhella et al., 1983) Bach approach involved complications arising from intrinsie- methodological limitations and from thePath of the Polypeptide Chain in Bacteriorhodopsin 831 Table 1 Flow-chart of experiments (0) Purifeation of BR fragments ‘Native PM Per *Flabeed | ‘PM *HLlabeled fase i rer Unlabeled ©: Per-*HL-labeled | CA and 0-2 7HL-abeled | fon Lew and Trp (2) Reconstitution of hybrid samples (ample number) ©. labeled ‘Unlabeled C1 Iabeled con Lou and Trp, C1 on Leu and Trp unlabeled C-1 Unlabeled C-2 267, 657 655, 860, 661 Por?HL-labeled C2 a2 i (€3 labeled on Leu and ep 654 = - ©2 Iabeled on Lau and Trp + ‘unlabeled O-2 658, 659 = = (8) Recording of neutron diffraction patterns (4) Analysis Selection of models Difference Fourier maps (Comparison of exparimental and predicted maps fact that the label was spread throughout. the molecule. A further difficulty with the earlier model- building approach was the large number of models ‘that needed to be considered (greater than 10°). The analysis required a stepwise screening that risked eliminating the correct folding model at an early Tn the present work, we have reduced the combi: natorial problem and restricted the label distribu tion by reassembling BR molecules from two fragments, one of them *H-labeled and the other not. A flow-chart of the experiments is shown in ‘Table 1. We have shown previously that two-dimen- sional crystals of BR with the native structure and ‘geometry can be reformed following renaturation of BR from these fragments (Popot ef al., 1986, 1987).. Fragment C-2 contains the first two of the trans- membrane helices and fragment C-1 the last five (Fig. 1). 7H can be introduced selectively into one of the two fragments. As an initial effort in this diree. tion, we per-*H-labeled the two-helix fragment (C-2) and localized it to one end of BR structural map (Popot ef al., 1986; Trewhella et al., 1986). It was not possible, however, to determine a unique posi- tion for each of the two helices nor to define their rotational orientation. ‘The improvements included in the present work concern sample preparation and data analysis. By labeling only the non-exchangeable positions of leucine residues and the a and f carbon atoms of ‘tryptophan residues, the label was confined to the transmembrane region. Moreover, helices that. were not labeled to the same extent could be distinguished one from another. Hybrid BR samples were constructed with either the C-1 or the C-2 fragment *H-labeled to various extents. Because of the weaker labeling, data analysis could be carried ‘out by both difference Fourier methods and model uilding. The model-building methodology deve- oped for the present work followed the strategy used by Jubb et al. (1984) for localizing *H-labeled retinal; model data are not caleulated entirely ab initio, but by combining structure factors derived from’ intensities observed with native purple membrane (PM) and structure factors calculated for the label. This approach has two advantages over that used by Trewhella ef al, (1983, 1986); it mini izes errors due to approximations in the model and it reduces considerably the computational cost of extensive model searches. ‘The present work establishes the positions of each of the first two transmembrane sequence segments in the structural map and provides some indication of the probable position of the third segment. It defines a preferred rotational orientation for these helices about their axes. An overall folding model in which transmembrane helices pack without inter leaving, as originally suggested by Engelman et al