Intramolecular Interactions That Induce Helical Rearrangement upon Rhodopsin Activation

Background: Identification of the intramolecular interactions in metarhodopsin IIa is essential for understanding the activation mechanism of rhodopsin. Results: Environmental changes around the chromophore, Ala-164, His-211, and Phe-261 were observed by probing with cysteine S-H vibrations. Conclusion: The interactions involving these residues are altered before the helical rearrangements. Significance: The intramolecular interactions by which rhodopsin adopts the transducin-activating conformation are shown. Rhodopsin undergoes rearrangements of its transmembrane helices after photon absorption to transfer a light signal to the G-protein transducin. To investigate the mechanism by which rhodopsin adopts the transducin-activating conformation, the local environmental changes in the transmembrane region were probed using the cysteine S-H group, whose stretching frequency is well isolated from the other protein vibrational modes. The S-H stretching modes of cysteine residues introduced into Helix III, which contains several key residues for the helical movements, and of native cysteine residues were measured by Fourier transform infrared spectroscopy. This method was applied to metarhodopsin IIa, a precursor of the transducin-activating state in which the intramolecular interactions are likely to produce a state ready for helical movements. No environmental change was observed near the ionic lock between Arg-135 in Helix III and Glu-247 in Helix VI that maintains the inactive conformation. Rather, the cysteine residues that showed environmental changes were located around the chromophore, Ala-164, His-211, and Phe-261. These findings imply that the hydrogen bond between Helix III and Helix V involving Glu-122 and His-211 and the hydrophobic packing between Helix III and Helix VI involving Gly-121, Leu-125, Phe-261, and Trp-265 are altered before the helical rearrangement leading toward the active conformation.

G-proteins. GPCRs have common structural motifs consisting of seven transmembrane helices (1)(2)(3). Ligands bind to the extracellular surface of a GPCR and activate it. The activation occurs via rearrangements of transmembrane helices (4 -6) that are likely to be common among GPCRs (7). Thus, the ligand-induced changes in the intramolecular interactions that induce helical rearrangement should be clarified to understand the mechanisms of activation of GPCRs.
Many GPCRs have been identified, but rhodopsin, which absorbs photons in rod photoreceptor cells, is biochemically, structurally, and physicochemically the best characterized. Rhodopsin consists of a protein moiety called opsin and an 11-cis-retinal chromophore that absorbs a photon. The chromophore is covalently bound to a Lys residue in Helix VII of opsin through a Schiff base linkage. Photoisomerization of the chromophore initiates conformational changes of opsin, leading to the formation of several intermediate states. Among these states, metarhodopsin II (Meta-II) is thought to activate the G-protein transducin (Gt).
Meta-II has an absorption spectrum in the near-UV region, unlike the dark state or the other intermediates. This is mainly because the Schiff base linkage between opsin and chromophore is deprotonated in Meta-II. Meta-II is activated via the significant rearrangement of the transmembrane helices (4 -6), as confirmed by the crystal structure of the complex of Meta-II and C-terminal peptide of Gt (8 -10). Meta-II forms a pH-and temperature-dependent equilibrium with its precursor metarhodopsin I (Meta-I) (11)(12)(13)(14). Recent extensive studies on metarhodopsins have demonstrated that Meta-II is the equilibrium mixture composed of several species (Scheme 1) (14 -18). Meta-I is in pH-independent thermal equilibrium with Meta-II a and Meta-II b . Although both Meta-II a and Meta-II b have a deprotonated Schiff base, Meta-II a displays only minor helical rearrangements. This mixture is in pH-dependent equilibrium with Meta-II b H ϩ , which binds to and activates Gt.
The intramolecular interactions in rhodopsin are altered essentially by the proton movement from the protonated Schiff base to the counterion (Glu-113). Although the helical arrangement of Meta-II a is comparable with those of the dark state or its precursors, the intramolecular interactions in Meta-II a should be altered, and these alterations should readily induce significant helical rearrangements. Thus, identification of the intramolecular interactions in Meta-II a is essential for understanding the mechanism by which photoactivated rhodopsin adopts the active conformation (18).
In the present study, using Fourier transform infrared (FTIR) spectroscopy (19), we detected the structural changes in the transmembrane region by monitoring changes in the S-H stretching vibration of the cysteine residues present in native rhodopsin or of cysteine residues systematically introduced into the transmembrane region. There are some advantages of using cysteine as a probe. First, depending on the environment, the cysteine S-H stretching frequency is in the 2580 -2525 cm Ϫ1 region, which is well separated from vibrations of the other groups present in protein (20 -23). Thus, changes of the environment of the cysteine can be probed without interference from other vibrational modes. Second, cysteine can be accommodated in the transmembrane region because of its relatively small and hydrophobic nature. Third, cysteine residues are introduced using a well established point mutation technique, which is simpler than the recently developed method using a non-natural amino acid, p-azido-L-phenylalanine (24).
We have developed this method for mapping the amino acid residues that undergo environmental changes (cysteine scanning) and applied it to bathorhodopsin, whose high resolution crystal structure is available (23). It was confirmed that cysteine residues introduced near the chromophore showed a clear shift in the S-H stretching mode in response to the local conformational change, but those introduced into the other part were completely silent. In the present study, Meta-II a (15,17,18) was characterized by cysteine scanning. We targeted the cysteine residues introduced into Helix III because it contains several functionally and structurally important residues, such as Glu-113, Glu-122, and Glu-134 (25)(26)(27)(28)(29), and the separation of Helix III from Helix VI is the key event for the activation of rhodopsin.

EXPERIMENTAL PROCEDURES
Preparation of Rhodopsin and Its Cysteine Mutants-Wildtype and mutant bovine opsin genes were expressed in HEK 293S cell lines as reported previously (30,31). cDNAs were fully sequenced before introduction into the expression vector, pUSR␣. To reconstitute the pigments, membrane fragments containing expressed opsins were incubated with 11-cis-retinal for more than 3 h at 4°C. The pigments were extracted with 1% (w/v) dodecyl-␤-D-maltoside in buffer P (50 mM HEPES, 140 mM NaCl, pH 6.5). The extracts were incubated with rho 1D4 antibody-agarose gel at room temperature overnight. After washing with buffer A (0.02% dodecyl-␤-D-maltoside in buffer P), the pigments were eluted with buffer A containing the C-terminal octadecapeptide of rhodopsin (DEASTTVSK-TETSQVAPA). The UV-visible spectra were recorded at this stage to estimate the absorption maximum in the visible region and optical purity index with a Shimadzu MPS-2000 recording spectrophotometer.
Solubilized rhodopsins were supplemented with a 100-fold molar excess of PC (Type XI-E, Sigma P2772) and dialyzed against buffer P for 3-4 days at 4°C. Rhodopsins in PC liposomes had absorbance in the range of 0.02-0.1. Liposomes were then collected by centrifugation and resuspended in 1 mM phosphate buffer (pH 5.7) supplemented with 5 mM NaCl. Sixty microliters of suspensions were placed on a BaF 2 window and dried under vacuum using an aspirator. The sample was sealed using another BaF 2 window and a spacer, after ϳ1 l of H 2 O was put beside the sample.
Spectroscopy-FTIR spectra were recorded using a Bio-Rad FTS40K spectrometer. An Oxford DN-1704 cryostat connected to an Oxford ITC-4 temperature controller was used for maintaining the sample temperature at 280.0 Ϯ 0.1 K. Irradiation light (Ͼ520 nm) was generated using a 1-kW tungsten halogen lamp (Rikagaku Seiki) and passed through a glass cutoff filter (O54, Toshiba). The Meta-II a minus rhodopsin difference spectra (Meta-II a /Rho spectra) were obtained by irradiation with Ͼ520-nm light at 280 K for 30 s. For each measurement, 256 interferograms at 2 cm Ϫ1 resolution were recorded. Meta-II a /Rho spectra were the averages of at least two measurements.

Effect of Mutation on the Absorption Maxima-Wild-type
rhodopsin has 10 cysteine residues ( Fig. 1), among which Cys-322 and Cys-323 are palmitoylated, and Cys-110 and Cys-187 form a disulfide bond. Thus, 6 of 10 native cysteine residues have free S-H groups and were replaced with serine (C140S, C167S, C185S, C222S, C264S, and C316S). We also prepared a double mutant, C110A/C187A, because a previous FTIR study focusing on S-H groups suggested the possible cleavage of the disulfide bond formed by these residues (21). In addition, 15 cysteine-introduced mutants at positions 117, 118, and 122-134 were prepared to probe the environmental changes along Helix III. We first examined the effects of these replacements of the native cysteine residues and introductions of cysteine residues.
All of these mutants were reconstituted into pigments with 11-cis-retinal. In addition, difference FTIR spectra between these mutants and their bathorhodopsin are comparable with that of wild type (23). These findings indicate that the overall structure of rhodopsin is little perturbed by these mutations. Absorption maxima ( max ) of mutants and optical purity indexes (ratio of absorbance at 280 nm to absorbance at max ) of the samples are listed in Table 1. Whereas the absorption maxima of most mutants were similar to that of the wild type (499 nm), mutations near the chromophore caused relatively large SCHEME 1. Meta-II equilibrium mixture composition.
blue shifts of the absorption maxima (A117C, T118C, and W126C). However, the difference FTIR spectra for bathorhodopsin of these mutants agreed with those of wild-type except for the chromophore bands, indicating that the perturbation of these mutations is limited to the vicinity of the chromophore (23).
In some samples, the optical purity index was Ͼ3, indicating that a significant amount of opsin and/or other proteins was present in the sample. However, because the difference FTIR spectra before and after visible light irradiation were measured in the following experiments, the proteins that are not photoactive (e.g. opsin) should not contribute to the FTIR spectra.
Formation of Metarhodopsin II a in Hydrated Phosphatidylcholine Liposomes-Meta-II a is the intermediate state in which the chromophore Schiff base is deprotonated but the helical arrangement is similar to that of Meta-I (15,17,18,32). In native rod outer segment (ROS) membrane, small amount of Meta-II a is in equilibrium with Meta-I and Meta-II b (Scheme 1). It has been reported that the amount of Meta-II a in the equilibrium is increased in 1,2-dioleoyl-sn-glycero-3-phosphocholine membranes (18), and Meta-II a in the equilibrium with significant amounts of Meta-I and Meta-II b has been partially characterized by FTIR spectroscopy (18).
In the course of characterization of Meta-II by FTIR, we found that a difference FTIR spectrum whose characteristics agree with those of the Meta-II a /Rho spectrum is obtained by irradiation of rhodopsin in a hydrated film sample at 280 K, which was prepared by drying rhodopsin in PC liposomes suspended in the phosphate buffer at pH 5.7 followed by hydration by ϳ1 l of H 2 O (Fig. 2). Although Meta-II a is favored at higher FIGURE 1. Secondary structural model of bovine rhodopsin. Among the 10 native cysteine residues, 6 cysteine residues, which have free S-H groups (red), were replaced by serine residues (C140S, C167S, C185S, C222S, C264S, and C316S). Cys-110 and Cys-187, which form a disulfide bond (blue), were replaced by alanine residues (C110A/C187A). Alternatively, a cysteine residue was introduced into Helix III (A117C, T118C, E122C, I123C, A124C, L125C, W126C, S127C, L128C, V129C, V130C, L131C, A132C, I133C, and E134C) (green). pH, this product is trapped at acidic pH, and is likely to be in the protonated form (Meta-II a H ϩ ) (see below). This putative Meta-II a H ϩ /Rho spectrum is shown in Fig. 2 in comparison with Meta-I/Rho and Meta-II b H ϩ /Rho spectra. The Meta-I/Rho spectrum has an amide-I band at 1635 cm Ϫ1 , whereas the Meta-II b H ϩ /Rho spectrum has one at 1644 cm Ϫ1 . Because the Meta-II a H ϩ /Rho spectrum has 1645 and 1635 cm Ϫ1 bands, it appears to be a mixture of Meta-I/Rho and Meta-II b H ϩ /Rho spectra. In fact, synthetic spectra composed of 80% Meta-I/Rho and 20% Meta-II b H ϩ /Rho spectra showed comparable intensities of the 1645 and 1635 cm Ϫ1 bands (Fig. 3). However, the 949 cm Ϫ1 band typical of the Meta-I/Rho spectrum (33,34) disappeared in the Meta-II a H ϩ /Rho spectrum, indicating that the Meta-II a H ϩ /Rho spectrum cannot be generated by a linear combination of Meta-I/Rho and Meta-II b H ϩ /Rho spectra. These findings confirm that the Meta-II a H ϩ /Rho spectrum obtained in the present condition is the difference FTIR spectrum between isolated Meta-II a H ϩ and rhodopsin.
It should be noted that the irradiation of native rhodopsin in the hydrated film of ROS in this condition generated a Meta-II b H ϩ /Rho spectrum, unlike the spectrum obtained in PC liposomes (Fig. 3). When the hydration level was reduced, a Meta-I-like photoproduct was produced in ROS membrane. Thus, the stabilization of Meta-II a in our experimental condition is likely to be caused by PC.
Characteristics of Meta-II a H ϩ -The small amide-I band in the Meta-II a H ϩ /Rho spectrum indicates that Meta-II a H ϩ undergoes small conformational change. The positive 1712 cm Ϫ1 band shows the protonation of Glu-113 in Meta-II a H ϩ (35). Asp-83 and Glu-122 show 1748/1769 cm Ϫ1 and 1745/ 1728 cm Ϫ1 bands, respectively, in the Meta-II b H ϩ /Rho spectrum (36). Similar bands for Asp-83 and Glu-122 are also observed in the Meta-II a H ϩ /Rho spectrum. However, the weak intensity of the positive 1748 cm Ϫ1 band would result from the reduced 1745/1728 cm Ϫ1 bands of Glu-122. These characteristics of the Meta-II a H ϩ /Rho spectrum are consistent with those of Meta-II a (18).
To assess the protonation states of Glu residues in Meta-II a H ϩ , the Meta-II a H ϩ /Rho spectrum of the wild type was compared with those of E122C and E134C (Fig. 2, d and e). The Meta-II b H ϩ /Rho (c) spectra. Meta-I/Rho and Meta-II b H ϩ /Rho spectra were recorded using bovine ROS by irradiation at 240 and 290 K, respectively. Meta-II a H ϩ /Rho spectra of E122C (d) and E134C (e) are superimposed on that of wild type (cyan lines). The double difference spectra (wild-type spectra minus mutant spectra) are shown below (blue lines). double difference spectra between the Meta-II a H ϩ /Rho spectra of wild-type and E122C showed that the CϭO stretching mode of Glu-122 shifted from 1728 to 1745 cm Ϫ1 upon formation of Meta-II a H ϩ . These frequencies are comparable with those of Meta-II b H ϩ /Rho, implying that the hydrogen bond involving Glu-122 in Meta-II a H ϩ is not altered in Meta-II b H ϩ . For E134C, the double difference spectrum showed only the positive band at 1740 cm Ϫ1 , indicating the protonation of Glu-134. Thus, the protonation site in Meta-II a H ϩ is Glu-134, like that in Meta-II b H ϩ . However, the frequency is significantly higher than that in Meta-II b H ϩ (1713 cm Ϫ1 ) (16). Thus, the hydrogen bond involving Glu-134 is weak in Meta-II a H ϩ and is strengthened in Meta-II b H ϩ . Glu-134 in Meta-II b H ϩ would be strongly hydrogen-bonded with a water molecule penetrating into the transmembrane region, as suggested by the crystal structures of Meta-II (8) and opsin (37,38).
Meta-II a H ϩ /Rho spectra (single difference spectra) of wild type and the mutants in the 1800 to 800 cm Ϫ1 region are shown in Fig. 4. They all exhibited no marker band of Meta-I at 949 cm Ϫ1 and weak intensity of the 1644 cm Ϫ1 band specific for Meta-II b H ϩ , implying that the introduction of cysteine does not largely affect the formation of Meta-II a H ϩ , although the different intensity of the 1644 cm Ϫ1 band suggests that the equilibrium constant between Meta-II a H ϩ and Meta-II b H ϩ varied among the mutants. It should be noted that the conformational changes of A117C, T118C, and W126C were also comparable with those of wild type, whereas they showed significantly blueshifted absorption spectra.

Mapping of Structural Changes upon Formation of Meta-II a H ϩ Probed by S-H Stretching
Modes-Meta-II a H ϩ /Rho spectra in the S-H stretching region are shown in Fig. 5. Meta-II a H ϩ /Rho spectra of wild type (cyan lines) are superimposed on those of mutants (red lines), and the double difference spectra between wild-type and mutant spectra are shown below (blue lines). Among the spectra of the seven cysteine-replaced mutants (Fig. 5, a-g), the Meta-II a H ϩ /Rho spectra of C167S and C185S were different from that of wild type, indicating environmental changes at positions 167 and 185. The double difference spectra between wild type and C167S and between wild-type and C185S exhibited bilobic bands at 2556/2569 cm Ϫ1 and 2546/2561 cm Ϫ1 , respectively (Fig. 5, b and c), implying that the hydrogen bonds involving Cys-167 and Cys-185 are strengthened upon formation of Meta-II a H ϩ . Although the conformation around Cys-316 changes upon formation of the Gt-activating state (Meta-II b H ϩ ) (39), the S-H stretching vibrations of Cys-316 were not perturbed in Meta-II a H ϩ .
It has been speculated that cleavage of the disulfide bond between Cys-110 and Cys-187 might occur upon formation of Meta-II based on the finding that the Meta-II/Rho spectrum shows only positive bands in the S-H stretching region (21). However, our results clearly showed that the Meta-II a H ϩ /Rho spectra of C110A/C187A were identical with that of wild type (Fig. 5g), indicating that the disulfide bond is not cleaved.
The Meta-II a H ϩ /Rho spectra of the cysteine-introduced mutants are shown in Fig. 5, h-v. Although the native cysteine residues were not replaced in these mutants, the vibrational band of the introduced cysteine was assessed by the double difference spectra between cysteine-introduced mutants and wild type, in which the contribution of native cysteine residues should be canceled.
Ala-117, Thr-118, and Glu-122 are located in the immediate vicinity of the chromophore, and the introduction of the cysteine residue in these positions causes the blue-shifted absorption spectra ( Table 1). The double difference spectrum for Meta-II a H ϩ of A117C had a broad and weak positive band at 2573 cm Ϫ1 (Fig. 5h). A small positive band at 2553 cm Ϫ1 and negative bands at 2565 and 2532 cm Ϫ1 were observed for T118C (Fig. 5i). The double difference spectrum for E122C shows a negative band at 2574 cm Ϫ1 and two positive bands at 2580 and 2563 cm Ϫ1 (Fig. 5j). The two positive bands are likely to be caused by the heterogeneity of the S-H group in the Meta-II a H ϩ . The possible heterogeneity at position 122 is also indicated by the CϭO stretching band of Glu-122 having a spectral shoulder (arrow in Fig. 2d).
I123C had a positive band at 2580 cm Ϫ1 in the double difference spectrum for the Meta-II a H ϩ /Rho spectrum (Fig. 5k). The band at 2580 cm Ϫ1 suggests that this S-H group does not form a hydrogen bond in Meta-II a H ϩ (20). Because the side chain of Ile-123 is located on the opposite side of Helix III from the chromophore and surrounded by Helices II, III, and IV, small movements of these helices would take place upon Meta-II a H ϩ formation to disrupt the hydrogen bond of Cys-123.
L125C and L128C displayed bilobic bands, indicating the frequency shift of S-H stretching modes. However, A124C, W126C, and S127C displayed only weak positive bands. Because the absence of the complementary band implies an intensity change with little frequency change of the S-H stretching band, these results suggest that only minor changes occurred in these positions. No environmental change of the S-H group was observed for V129C, V130C, L131C, A132C, I133C, or E134C, although Glu-134 is involved in the conserved ERY sequence. Because large helical movement is suppressed in this condition, the environment around Cys-134 would not be altered.
The environmental changes of amino acid residues probed in this study are shown in Fig. 6. The positions of cysteine residues that showed frequency shifts in Meta-II a H ϩ /Rho spectra are indicated in red, those with only intensity change are shown in pink, and those with no change are shown in white. The amino acid residues within 4 Å of amino acid residues where introduced cysteine showed frequency shifts were listed using PyMOL software. The results demonstrated that Ala-164, His-211, and Phe-261 were located within 4 Å of two or more positions. It is an advantage of cysteine scanning over the conventional point mutation method that the key amino acid residue is identified without mutation of itself, because mutation of a key residue may substantially alter its nature.

DISCUSSION
The structural changes of rhodopsin in the transmembrane region were detected here by use of the cysteine S-H group as an internal probe. All mutants prepared in this study formed pigments with 11-cis-retinal and were converted to Meta-II a H ϩ , which exhibited chromophore vibrational bands as well as amide-I bands very similar to those of the wild type. These results indicate that the cysteine residue acts as a highly sensitive probe for conformational changes of the protein, although the introduction of a cysteine residue into the transmembrane region hardly perturbs the native structure. In the present study, a photoproduct which has a deprotonated chromophore but undergoes small conformational change was trapped in a hydrated film sample of PC liposome at pH 5.7. Although the characteristics of the FTIR spectrum of this photoproduct regarding the amide-I, Glu-83, and Glu-122 bands agreed with those of Meta-II a reported previously (18), Glu-134 in this photoproduct is protonated, whereas it is not protonated in Meta-II a (Fig. 2). Thus, this photoproduct was identified as Meta-II a H ϩ .
Although there is no direct evidence showing the interconversion between Meta-II a H ϩ and Meta-II b H ϩ , Meta-II a H ϩ and Meta-II b H ϩ would be in pH-independent equilibrium, like Meta-II a and Meta-II b . Meta-II a H ϩ would be enriched by suppression of the helical rearrangements in the hydrated film con- taining less water than in solution. PC is also likely to bias the equilibrium toward Meta-II a H ϩ . However, because Meta-II a H ϩ was trapped in the artificial condition, we could not exclude the possibility that the structure of Meta-II a that is transiently formed in the physiological condition may be different from that of Meta-II a H ϩ observed here.
It is reported that the deprotonated intermediate is produced by irradiation of rhodopsin crystal (40). The crystal structure of this intermediate (Protein Data Bank entry 2I37) demonstrated that the Helix V is elongated, but the overall helical arrangement is close to the dark state, unlike the G-protein-interacting conformation (Protein Data Bank entries 3DQB, 3PQR, and 2X72). Because the small conformational change would be derived from the crystal packing, 2I37 would represent the structure of Meta-II a .
We probed the environmental changes for Meta-II a H ϩ at 23 positions (8 native cysteine residues plus 15 introduced cysteine residues) by analyzing the S-H vibrational mode. Mapping of the cysteine residues showing S-H vibrational changes demonstrated that they are localized proximal to Ala-164, His-211, and Phe-261 (Fig. 6).
His-211, which stabilizes the active conformation (41), is hydrogen-bonded with Glu-122. The amino acid residue corresponding to Glu-122 of rhodopsin is Gln in cone pigments, and this is one of the reasons why the decay of Meta-II of cone pigments is significantly faster than that of rhodopsin (42,43). The environmental changes around His-211 strongly suggest that the hydrogen bond between Glu-122 and His-211 is perturbed in Meta-II a H ϩ . Environmental changes were also observed around Phe-261, which forms the hydrophobic interface between helices III and VI together with Gly-121, Leu-125, and Trp-265. These residues are essential for the correct folding of the pigment (44 -46). Notably, mutants of these residues, such as F261V and G121L/F261V, show significant constitutive activity (47), suggesting that the disruption of this packing may result in the active conformation. It should be noted that these interpretations are based on the notion that Meta-II a H ϩ is a direct precursor of Meta-II b H ϩ .
On the other hand, the S-H group of Cys-222 in Helix V forms a hydrogen bond with the backbone carbonyl oxygen of Ala-132 in Helix III in the dark state (Protein Data Bank entry 1U19), whereas this hydrogen bond is disrupted in G-proteininteracting conformations (Protein Data Bank entries 3DQB, 3PQR, and 2X72). The lack of change in the S-H vibration of Cys-222 suggests that the arrangements of Helices III and V are not altered in Meta-II a H ϩ . In addition, the lack of environmental change at position 134, which is involved in the conserved ERY sequence and proximate to the Arg-135/Glu-247 ionic lock, also supports the notion that the conformational change is not propagated to the cytoplasmic side in Meta-II a H ϩ . . Meta-II a H ؉ /Rho spectra in the S-H stretching region. Meta-II a H ϩ /Rho spectra for mutants (red lines) are superimposed on that of wild type (cyan lines). The double difference spectra are shown below (blue lines). Scale bar, 1 ϫ 10 Ϫ4 . Left, Meta-II a H ϩ /Rho spectra for the cysteine-substituted mutants (a-g). Double difference spectra were calculated by subtracting mutant spectra from wild-type spectra. Middle and right, Meta-II a H ϩ /Rho spectra for cysteineintroduced mutants (h-v). Double difference spectra were calculated by subtracting wild-type spectra from mutant spectra. The typical frequency of the S-H stretching mode is shown by the thick bar at the bottom of each panel.
In conclusion, the present results showed that the Glu-122/ His-211 hydrogen bond, which stabilizes the active conformation, and the Gly-121/Leu-125/Phe-261/Trp-265 hydrophobic packing between Helices III and VI, which maintains the correct folding and suppresses the constitutive activity, are likely to be perturbed prior to the helical rearrangement. Protonation of Glu-134 facilitates the deprotonation of the chromophore Schiff base but does not necessarily induce the great helical movements. The changes in the intramolecular interaction near the chromophore would induce the significant rearrangement of the transmembrane helices, resulting in the active conformation.