|Molecular Vision 2000;
Received 13 March 2000 | Accepted 13 July 2000 | Published 27 July 2000
Three dimensional structure of the seventh transmembrane helical domain of the G-protein receptor, rhodopsin
Philip L. Yeagle,1
Chad Danis,2 Gregory Choi,2 James L. Alderfer,2
Arlene D. Albert1
1Department of Molecular and Cell Biology, University of Connecticut, Storrs, CT; 2Department of Biophysics, Roswell Park Cancer Institute, Buffalo, NY
Correspondence to: Philip L. Yeagle, Department of Molecular and Cell Biology, University of Connecticut, Storrs, CT, 06269; Phone: (860) 486-4363; FAX: (860) 486-4331; email: email@example.com
Purpose: The three dimensional structure of a peptide comprising the sequence of the seventh transmembrane segment of the G-protein coupled receptor, rhodopsin, was determined in solution.
Methods: The sequence of the seventh transmembrane segment of rhodopsin, which contains the NPxxY sequence that is highly conserved among G-protein coupled receptors and lys296 that forms the Schiff base with the retinal, was synthesized by solid phase peptide synthesis. The three dimensional structure was determined in solution by high-resolution nuclear magnetic resonance (NMR).
Results: The structure revealed a helix-break-helix motif for this sequence. Two families of structures were observed which differed in the angle between the two helical segments. The sequence of this transmembrane segment overlapped significantly the sequence of a peptide from the carboxyl terminal of rhodopsin, the structure of which was solved previously. The redundant sequence formed a helix in both peptides. It was therefore possible to superimpose the redundant sequence of both peptides and construct a structure for rhodopsin encompassing residues 291-348.
Conclusions: This structure reveals locations of the lys296 and the acylation sites of rhodopsin that are consistent with the known biochemistry of this receptor. This segmentation approach to membrane protein structure provides important structural information in the absence of an X-ray crystal structure of rhodopsin. The approach is expected to be useful for other G-protein coupled receptors.
G-protein coupled receptors constitute a large family of receptors which regulate a wide variety of cellular functions. These receptors respond to an array of signals, which include peptide and glycoprotein hormones, neurotransmitters, chemokines, adenine nucleotides and light. Heterotrimeric GTP-binding proteins are then activated to transduce the extracellular signals to intracellular events. To understand this mechanism of signal transduction, a high-resolution structure of the protein is essential. Because G-protein coupled receptors are integral membrane proteins, crystallization for X-ray diffraction studies is difficult. Therefore, no complete three-dimensional structure for any G-protein coupled receptor has been reported.
Rhodopsin, the receptor for visual signal transduction and widely regarded as the archetype of G-protein receptors, is one of the most extensively investigated members of the family of G-protein coupled receptors. Circular dichroism studies  and low-resolution structural data from two dimensional crystals of rhodopsin [2,3] have established that rhodopsin is built around a bundle of seven transmembrane helices, but the structures of the cytoplasmic face that activates the G-protein and the intradiskal surface remain unknown.
Studies on the four-helix bundle, myohemerythrin, a soluble protein, have shown that small peptides from the sequence of the protein retain the secondary structure of the native protein in solution . Likewise, studies on the 7 helix bundle, bacteriorhodopsin, an integral membrane protein, have shown that the secondary structure of this protein, including loops and helices, is preserved in the structure of small peptides in solution [5,6].
These results suggest that an alternative path to structural information on proteins like rhodopsin is to determine the structures of segments of the protein. Several reports demonstrated that peptides containing sequences from the cytoplasmic loops and carboxyl terminus of rhodopsin exhibit biological activity in solution [7-11]. In this laboratory, we have shown that the cytoplasmic loops and the carboxyl terminal exhibit defined structure as peptides in solution [10,12-14]. Furthermore, the intradiskal loops and the amino terminus also are structured as peptides in solution . Therefore, the individual loops and carboxyl terminus and amino terminus of rhodopsin can be regarded as small segments each of whose secondary structure is stabilized by short range interactions and fully coded by the local primary sequence.
We now extend these studies to the seventh transmembrane segment of rhodopsin. This helix is of particular interest because it includes lys296 that forms the Schiff base with the 11-cis retinal. It is only when the retinal chromophore is bound that the receptor is poised to respond to light. The seventh transmembrane helix also includes the NPxxY sequence. Although the role of this sequence is unclear, it is highly conserved within the family of G-protein receptors, suggesting structural significance. We find that this transmembrane segment is predominantly helical in structure, with a break in the helix at the NPxxY sequence. Furthermore, this segment was sufficiently long to exhibit sufficient overlap with the previously determined structure of the carboxyl terminal. By assembling the structures of these two segments, a structure for the portion of rhodopsin containing residues 291-348 was obtained.
The peptide PAFFAKTSAVYNPVIYIMMNKQFRN (rhoviih), representing the sequence of the seventh transmembrane helix of rhodopsin, was synthesized through solid phase synthesis in the Biotechnology Center at the University of Connecticut.
Figure 1 shows this sequence in the primary structure of rhodopsin . The purity of this peptide was analyzed by HPLC. All solvents were analytical grade from Fisher Scientific (Suwanee, GA). Peptides were analyzed on a Voyager system from Amersham Pharmacia Biotech (Piscataway, NJ), equipped with a C-18 reverse phase analytical column. Two solvent systems were used: 0.1% TFA in distilled water (A) and 0.1%TFA in acetylnitrile (B). A solvent gradient was used from A to B. A single peak dominated the HPLC with a retention of 38 min. About 95% of the material eluted in a single peak.
All NMR spectra were recorded on a Bruker AMX-600 spectrometer at 30 °C in DMSO because this peptide was not stable in aqueous solution, or in trifluoroethanol, in agreement with the results on the solubility of the seventh transmembrane helix of the tachykinin receptor . The peptide was used at a concentration of 3 mM. Standard pulse sequences and phase cycling were employed to record double quantum filtered (DQF) COSY and NOESY (400 ms mixing time) . All spectra were accumulated in a phase sensitive manner using time-proportional phase incrementation for quadrature detection in F1. Chemical shifts were referenced to the residual protonated DMSO signal as a reference, defining it as 2.49 ppm with respect to TMS. The behavior of the 1H NMR chemical shifts was examined as a function of peptide concentration and over a concentration range of one order of magnitude, there was no change in the 1H chemical shifts. These data are consistent with the presence of monomers. Sequence-specific assignments were obtained using standard approaches.
Attempts were made to solubilize rhoviih in other organic solvents and in currently available deuterated detergents. Introducing rhoviih into chloroform:methanol mixtures resulted in a cloudy suspension. This peptide was taken up in three different detergents: octylglucoside, sodium dodecyl sulfate and dodecylphosphocholine. In the deuterated forms of these detergents, NMR spectra were accumulated. In no case were useful NMR data obtained because the peptide was not adequately solubilized by the detergent micelles. This is likely due to the presence of the sequence, lys296-thr297-ser298, in the middle of the peptide. These residues are incompatible with the interior of a micelle and probably keep the peptide from being incorporated into detergent micelles.
The sequence-specific assignment of the 1H NMR spectrum for the peptide was carried out using standard methods employing FELIX (Molecular Simulations, Inc., San Diego, CA). Assigned NOE cross peaks were segmented using a statistical segmentation function and characterized as strong medium and weak corresponding to upper bounds distance range constraints of 0.27, 0.37, and 0.50 nm, respectively. Lower bounds between non-bonded atoms were set to the sum of their van der Waals radii (approximately 0.18 nm). Pseudoatom corrections were added to interproton distance restraints where necessary . Distance geometry calculations were carried out using the program DIANA  within the SYBYL 6.4 package (Tripos Software Inc., St. Louis, MO). First generation DIANA structures, 150 in total, were calculated. Energy refinement calculations (restrained minimizations/dynamics) were carried out on the best distance geometry structures using the SYBYL program implementing the Kollman all-atom force field. Statistics on structures were obtained from X-plor. These calculations were performed on a Silicon Graphics R10000 computer (Silicon Graphics, Mountain View, CA). Imaging and superposition of the resulting structures and construction of the transmembrane helices of rhodopsin was performed on a PowerMac with MacImdad (Molecular Applications Group, Palo Alto, CA). A total of 443 constraints were used. The final structures showed 12 violations of the constraints greater than 0.03 nm.
Transmembrane helix 7 of rhodopsin
A 25mer (rhoviih) corresponding to rhodopsin residues 291-315 was synthesized. This amino acid sequence was predicted to form the seventh transmembrane helix of bovine rhodopsin based on hydropathy plots of the primary sequence of rhodopsin . The length of the peptides was chosen to be sufficiently long to exhibit a well-defined structure, but short enough to maintain solubility. Previous observations in this laboratory have indicated that peptides with approximately 17-43 residues from rhodopsin show considerable secondary structure in solution as detected by circular dichroism and NMR. The solution structure of this peptide (rhoviih) was determined by standard homonuclear two-dimensional NMR protocols, as described in Methods. The chemical shifts and coupling constants are reported in Table 1. The sequence used can be seen in Figure 1. From the NOESY map, 443 constraints were obtained: 208 intra-residue, 148 sequential and 87 long range. Five dihedral angle constraints were obtained from coupling constants. Figure 2 shows the connectivities. Figure 3 shows the number of constraints per residue. Figure 4 shows the chemical shift index  as a function of residue number. This analysis shows a strong indication of helix for residues 18-25, and a moderate suggestion of helix for residues 1-9 (internal numbering scheme for the peptide).
This peptide exhibits defined structure as a free peptide in solution. Two families of structures were found consistent with the constraints obtained from the NOESY maps. The structure families for this peptide appear in Figure 5A and Figure 5B, as an overlay of the six best structures obtained from DIANA, after energy minimization. The structures in Figure 5B are favored in the DIANA calculation if the dihedral angle constraints are included (PDB 1FDF). A Ramachandran plot shows most of the residues in allowed regions, with two in disallowed regions. The structure is predominantly an a-helix motif, with a break in the helix next to the proline of the NPxxY sequence in both families, consistent with previous suggestions . These two helical segments are consistent with the chemical shift index in Figure 4, which suggests two helical segments separated by a break. The angle between the two helical segments is significantly different in the two families. The angles the two helical segments make with each other are 100° for one family (Figure 5A) and 155° for the other (Figure 5B). In the study by Konvicka et al.,  which showed a broken helix, the corresponding angles between the two helical segments were 90° and 135°. In the structure of the sixth transmembrane segment of the sarcoplasmic reticulum Ca2+ ATPase, two helical segments were observed, with an angle between them of 140° . Another structural study on a transmembrane helix of integral membrane proteins reported a similar bend in the helix axis at a proline .
This structure was determined using DMSO as a solvent because of difficulties solubilizing the peptide in other media (see Methods). Studies on three loops  and one transmembrane helix (Katragadda et al., unpublished data) of bacteriorhodopsin as peptides in DMSO have shown good fidelity between the structures of the peptides seen in solution and the corresponding region in the crystal structure of the protein.
Connection of helix 7 with the carboxyl terminus of rhodopsin
The structure of the carboxyl terminus of rhodopsin (rhodopsin residues 306-348) was reported previously . The sequence of that structure and the sequence of rhoviih (rhodopsin residues 291-315) reported here have 10 residues in common. These two structures together offer an opportunity for structure determination not heretofore available. The carboxyl terminal region of rhoviih reported here consists of an a-helix. The amino terminal region of the peptide comprising residues 306-348 is also an a-helix . These observations reveal that the sequence 306-315 of rhodopsin is an a-helix regardless of the peptide in which it is located.
Therefore, this region of the two structures can be superimposed as a means of joining the two structures into one. Of these overlapping residues, the terminal five of each peptide were not used for the superposition because of disordering that occurs in the solutions to the structures at the ends of each of the peptides. The remaining five residues in common were superimposed with a least squares fit and the results appear in Figure 6. The rmsd of the superposition of the 5 alpha carbon atoms was 0.13 nm. As can be seen from the figure, the structure of this 5-residue overlap region is very similar in the two peptide structures. Thus, it is possible to superimpose the structures of both molecules, using the residues in common to both, to reveal the structure for the entire represented region (residues 291-348) of rhodopsin (Figure 7).
The structure obtained from the NOESY data for the seventh transmembrane helix of rhodopsin shows a helix-break-helix motif, with a break in the helix occurring next to the proline of the conserved NPxxY sequence. This is a very similar structure to the structure reported previously for a shorter version (13mer) of the seventh transmembrane helix of the tachykinin receptor, also containing the NPxxY sequence . In the latter structure, the data also show a helix-break-helix motif, with a break next to the proline. In a recent modeling study of the NPxxY sequence, two conformations around the proline were predicted . One result showed a much more acute angle between the two helical segments than the other. In our studies, the available constraints were compatible with two conformations. Both of these conformations were similar to those reported in the modeling studies. Unfortunately, insufficient constraints are available from the NOESY maps to quantitatively evaluate the relative weight one should give to the two structures, although the inclusion of the dihedral angle constraints appears to favor the more open conformation (Figure 5B) as noted above. For this reason and others noted previously , the statistics provided are on that family of structures. As was suggested earlier , it is possible that helix 7 changes from one of these conformations to another as the receptor changes state. However, it is also possible that in rhodopsin, either the conformation in Figure 5A or the conformation in Figure 5B may be favored .
Direct structure determination by nuclear magnetic resonance (NMR) has been successful for soluble proteins and for peptides. However, integral membrane proteins are large and are not soluble in aqueous solution. Thus, they are not amenable to this technique. Recent work on the largely helical protein, myohemerythrin (a soluble protein made up of a four-helix bundle) has suggested it may be feasible to elucidate some protein structures by piecing together structural segments. It was shown that when the protein structure was divided into small structural segments, the structures of peptides which represented most of these segments exhibited the same secondary structure (as determined by NMR solution methods) as found in the structure of the intact protein . Studies on the integral membrane protein, bacteriorhodopsin, showed that both helices and loops of this protein were preserved in the structures of peptides in solution [6,25]. These studies, along with others [26-28], clearly suggest that secondary structures (helices and loops) are stabilized by short-range interactions. That suggestion is supported by the observation of elements of secondary structure in the solution structure of peptides comprising the cytoplasmic loops and the carboxyl terminal of rhodopsin [11,13], and the PTH receptor .
One other question remains that concerns the solvent system used in this study: do peptides adopt the same structure in DMSO as the same sequence adopts in the native protein? This question has been investigated in two ways. First, the structure of three loops of bacteriorhodopsin were determined in DMSO and compared with the native protein. Excellent agreement was obtained from the structures in solution and the crystal structures of the protein . Second, the top of the seventh transmembrane helix of rhodopsin was determined in water as part of the carboxyl terminal  and in DMSO in this report. The same structure was observed in each case. Therefore, the use of DMSO shows no artifacts.
The data presented here suggest that these concepts may be exploited to obtain structural elements of a membrane protein. The peptide in this study was designed to overlap significantly with the sequence of the peptide we previously studied from the carboxyl terminal of rhodopsin. We observed that the overlap region of these two peptides is helical in both structures. This observation is a clear example of the principles elucidated above that a secondary structure like a-helix is stabilized by short-range interactions. The portion of the rhodopsin sequence that is common between these two peptides is helical whether it is in a peptide comprising the carboxyl terminal of the protein or in a peptide comprising the seventh transmembrane helix of rhodopsin.
Therefore, it is possible to superimpose the two structures utilizing the sequence in common, as described above. The result is a unified structure for the seventh transmembrane helix and the carboxyl terminus of rhodopsin. In principle, one might consider exploring other regions of the structure of a membrane protein like rhodopsin in an analogous manner.
Several observations arise from examination of the structure of the region 291-348 of rhodopsin. Cys 322 and cys 323 are the sites of palmitoylation on rhodopsin. In the structure presented here, these cysteines are found at the base of the fourth cytoplasmic loop. The location of the cysteines likely delineates the top of the hydrophobic core of the bilayer. The portion of the helix that appears "above" the location of the cysteines in Figure 6 is highly charged and likely not buried in the bilayer. The part of the helix at and below the level of the cysteines is hydrophobic, until the region of the lysine is reached. The helix appears, from other work , to extend further towards the bottom of the figure than is represented in the peptide whose structure is reported here. Therefore, the other side of the hydrophobic core of the bilayer is probably not reached by this peptide.
It is interesting to note the orientation of cys 322 and cys 323 relative to lys 296. Lys 296 forms a Schiff base linkage with the chromophore, 11-cis retinal. It is therefore expected that this lysine residue will be oriented toward the interior of the protein. If the lysine in the structure presented here is oriented in this manner, the observed structure dictates that the cysteines would be on the outside of the helical bundle. This placement permits the acylated fatty acids to be in the lipid bilayer, as expected. The relative placement of the cysteines and the lysine in the structure reported here are therefore consistent with the known structural and biochemical properties of rhodopsin.
Lys296 is near the break in the helix induced by the proline of the conserved NPxxY sequence. Conserved prolines are found in the adjacent helices in the bundle, helices 4, 5, and 6. If all of these prolines induce analogous breaks in their respective helices, then a pocket would be formed in the interior of the bundle of transmembrane helices of rhodopsin, suitable for binding the retinal.
This work was supported by National Institutes of Health Grant EY03328 and in part by CA16056. We thank A. Chopra for his assistance in the structure solution and Dr. R. Wollman for his assistance in data acquisition.
1. Albert AD, Litman BJ. Independent structural domains in the membrane protein bovine rhodopsin. Biochemistry 1978; 17:3893-900.
2. Schertler GF, Villa C, Henderson R. Projection structure of rhodopsin. Nature 1993; 362:770-2.
3. Unger VM, Hargrave PA, Baldwin JM, Schertler GF. Arrangement of rhodopsin transmembrane alpha-helices. Nature 1997; 389:203-6.
4. Dyson HJ, Merutka G, Waltho JP, Lerner RA, Wright PE. Folding of peptide fragments comprising the complete sequence of proteins. Models for initiation of protein folding. I. Myohemerythrin. J Mol Biol 1992; 226:795-817.
5. Kahn TW, Sturtevant JM, Engelman DM. Thermodynamic measurements of the contributions of helix-connecting loops and of retinal to the stability of bacteriorhodopsin. Biochemistry 1992; 31:8829-39.
6. Katragaddaa M, Alderfer JL, Yeagle PL. Solution structure of the loops of bacteriorhodopsin closely resembles the crystal structure. Biochim Biophys Acta 2000; 1466:1-6.
7. Konig B, Arendt A, McDowell JH, Kahlert M, Hargrave PA, Hofmann KP. Three cytoplasmic loops of rhodopsin interact with transducin. Proc Natl Acad Sci U S A 1989; 86:6878-82.
8. Takemoto DJ, Takemoto LJ, Hansen J, Morrison D. Regulation of retinal transducin by C-terminal peptides of rhodopsin. Biochem J 1985; 232:669-72.
9. Marin EP, Krishna AG, Zvyaga TA, Isele J, Siebert F, Sakmar TP. The amino terminus of the fourth cytoplasmic loop of rhodopsin modulates rhodopsin-transducin interaction. J Biol Chem 2000; 275:1930-6.
10. Yeagle PL, Alderfer JL, Albert AD. Structure of the carboxyl-terminal domain of bovine rhodopsin. Nat Struct Biol 1995; 2:832-4.
11. Yeagle PL, Alderfer JL, Salloum AC, Ali L, Albert AD. The first and second cytoplasmic loops of the G-protein receptor, rhodopsin, independently form beta-turns. Biochemistry 1997; 36:3864-9.
12. Yeagle PL, Alderfer JL, Albert AD. Structure of the third cytoplasmic loop of bovine rhodopsin. Biochemistry 1995; 34:14621-5.
13. Yeagle PL, Alderfer JL, Albert AD. Structure determination of the fourth cytoplasmic loop and carboxyl terminal domain of bovine rhodopsin. Mol Vis 1996; 2:12 <http://www.molvis.org/molvis/v2/a12/>.
14. Dorey M, Hargrave PA, McDowell JH, Arendt A, Vogt T, Bhawasar N, Albert AD, Yeagle PL. Effects of phosphorylation on the structure of the G-protein receptor rhodopsin. Biochim Biophys Acta 1999; 1416:217-24.
15. Yeagle PL, Salloum A, Chopra A, Bhawsar N, Ali L, Kuzmanovski G, Alderfer JL, Albert AD. Structures of the intradiskal loops and amino terminus of the G-protein receptor, rhodopsin. J Pept Res 2000; 55:455-65.
16. Hargrave PA, McDowell JH, Curtis DR, Wang JK, Juszczak E, Fong SL, Rao JK, Argos P. The structure of bovine rhodopsin. Biophys Struct Mech 1983; 9:235-44.
17. Berlose JP, Convert O, Brunissen A, Chassaing G, Lavielle S. Three-dimensional structure of the highly conserved seventh transmembrane domain of G-protein-coupled receptors. Eur J Biochem 1994; 225:827-43.
18. Kumar A, Ernst RR, Wuthrich K. A two-dimensional nuclear Overhauser enhancement (2D NOE) experiment for the elucidation of complete proton-proton cross-relaxation networks in biological macromolecules. Biochem Biophys Res Commun 1980; 95:1-6.
19. Wuthrich K, Billeter M, Braun W. Pseudo-structures for the 20 common amino acids for use in studies of protein conformations by measurements of intramolecular proton-proton distance constraints with nuclear magnetic resonance. J Mol Biol 1983; 169:949-61.
20. Guntert P, Braun W, Wuthrich K. Efficient computation of three-dimensional protein structures in solution from nuclear magnetic resonance data using the program DIANA and the supporting programs CALIBA, HABAS and GLOMSA. J Mol Biol 1991; 217:517-30.
21. Wishart DS, Sykes BD, Richards FM. The chemical shift index: a fast and simple method for the assignment of protein secondary structure through NMR spectroscopy. Biochemistry 1992; 31:1647-51.
22. Findlay JB, Pappin DJ. The opsin family of proteins. Biochem J 1986; 238:625-42.
23. Konvicka K, Guarnieri F, Ballesteros JA, Weinstein H. A proposed structure for the transmembrane segment 7 of G protein-coupled receptors incorporating an asn-Pro/Asp-Pro motif. Biophys J 1998; 75:601-11.
24. Soulie S, Neumann JM, Berthomieu C, Moller JV, Maire M, Forge V. NMR conformational study of the sixth transmembrane segment of sarcoplasmic reticulum Ca2+-ATPase. Biochemistry 1999; 38:5813-21.
25. Hunt JF, Earnest TN, Bousche O, Kalghatgi K, Reilly K, Horvath C, Rothschild KJ, Engelman DM. A biophysical study of integral membrane protein folding. Biochemistry 1997; 36:15156-76.
26. Yang AS, Hitz B, Honig B. Free energy determinants of secondary structure formation: III. beta-turns and their role in protein folding. J Mol Biol 1996; 259:873-82.
27. Yao J, Dyson HJ, Wright PE. Three-dimensional structure of a type VI turn in a linear peptide in water solution. Evidence for stacking of aromatic rings as a major stabilizing factor. J Mol Biol 1994; 243:754-66.
28. Fezoui Y, Connolly PJ, Osterhout JJ. Solution structure of alpha t alpha, a helical hairpin peptide of de novo design. Protein Sci 1997; 6:1869-77.
29. Mierke DF, Royo M, Pelligrini M, Sun H, Chorev M. Third cytoplasmic loop of the PTH/PTHrP receptor. J Am Chem Soc 1996; 118:8998-9004.
30. Chopra A, Yeagle PL, Alderfer JA, Albert AD. Solution structure of the sixth transmembrane helix of the G-protein-coupled receptor, rhodopsin(1). Biochim Biophys Acta 2000; 1463:1-5.