Zinc-induced Decrease of the Thermal Stability and Regeneration of Rhodopsin*

Zinc is present at high concentrations in the photoreceptor cells of the retina where it has been proposed to play a role in the visual phototransduction process. In order to obtain more information about this role, the study of the effect of zinc on several properties of the visual photoreceptor rhodopsin has been investigated. A specific effect of Zn2+ on the thermal stability of rhodopsin, obtained from bovine retinas and solubilized in dodecyl maltoside detergent, in the dark is reported. The thermal stability of rhodopsin in its ground state (dark state) is clearly reduced with increasing Zn2+ concentrations (0–50 μmZn2+). The thermal bleaching process is accelerated in the presence of Zn2+ with k rate constants, at 55 °C, of 0.028 ± 0.002 min−1 (0 μmZn2+) and 0.056 ± 0.003 min−1 (50 μm Zn2+), corresponding tot 1 2 values of 24.4 ± 1.6 min and 11.8 ± 0.1 min, respectively. Thermodynamic parameters derived from Arrhenius plots show a significant E a increase at 50 μm Zn2+ for the process, with ΔG ‡ decrease and increase in ΔH ‡ and ΔS ‡ possibly reflecting conformational rearrangements and reordering of water molecules. The stability of the metarhodopsin II intermediate is also decreased and changes in the metarhodopsin II decay pathway are also detected. The extent of rhodopsin regeneration in vitro is also reduced by zinc. These effects, specific for zinc, are also seen for rhodopsin in native disc membranes, and may be relevant to the suggested role of Zn2+ in normal and pathological retinal function.

Rhodopsin is the photoreceptor protein of the vertebrate retina (1)(2)(3) belonging to the G-protein-coupled receptor (GPCR) 1 superfamily (4 -6). It is the main protein component of the rod outer segments (ROS) of the retinal photoreceptor cells, and its easy isolation from bovine retinas has made of this receptor a widely used model for the GPCR superfamily. Rhodopsin is a key molecule in the biochemistry of vision and alterations in its sequence have been associated with retinal disease. In particular, a high number of mutations in the opsin gene have been associated with the autosomal form of the retinal degenerative disease retinitis pigmentosa (7,8). A number of factors have been proposed to be related to retinal function, and among them a possible role for Zn 2ϩ in the retina and its metabolism has been proposed (9). Zn 2ϩ is present at particularly high concentrations in the retina (10) being a component of the disc membranes in the rod outer segments of the photoreceptor cells (11). Zn 2ϩ has been histochemically localized to ROS (12) and it has been shown to copurify with ROS proteins as well (13). Despite this presence, the role of Zn 2ϩ in the visual cycle and specifically in its interaction with rhodopsin remains unclear.
Zn 2ϩ is required for the function of numerous proteins, serving both as a part of the active site in, for example, metalloenzymes, and acting to stabilize protein domains, such as the Zn 2ϩ fingerbinding motif in transcription factors (14,15). The structure of many Zn 2ϩ -binding sites is known from x-ray crystallography of Zn 2ϩ -binding proteins and, thus, the geometry of the interaction between Zn 2ϩ and different coordinating residues is well characterized (14,15). This, together with the small size of the zinc (II) ion, makes artificially generated Zn 2ϩ -binding sites a highly useful approach for probing structure-function relationships in proteins. In GPCRs, for example, construction of bis-His Zn 2ϩ -binding sites has led to important information about both the organization of the transmembrane helices and their movements during receptor activation to be obtained (16 -22). Therefore, engineered Zn 2ϩ binding sites created by site-directed mutagenesis is an interesting tool for probing intramolecular interactions in these types of receptors (23). Particularly in the case of rhodopsin, important structural information about helical orientation, connectivities, and the conformational change upon light activation, has been obtained from engineered Zn 2ϩ -binding sites in this visual photoreceptor (19).
There is experimental evidence indicating the presence of naturally occurring Zn 2ϩ -binding sites in several GPCR proteins. These include the dopamine transporter (24), D 1 dopamine receptor (25) and all D 2 -like dopamine receptors: D 2L , D 3 , and D 4 (26). Molecules participating in the visual phototransduction process are already known to have Zn 2ϩ -binding sites, including rhodopsin (27) and cGMP phosphodiesterase (PDE6␣␤) (28).
Previous reports showed direct Zn 2ϩ binding to rhodopsin, both in ROS membranes and in detergent-solubilized-purified protein systems, by means of 65 Zn competition binding experiments, indicating that there may be coordination sites for Zn 2ϩ in native rhodopsin (27). This binding occurs in the dark and it increases upon illumination of the protein (27). Also, previous reports indicated that Zn 2ϩ enhanced azido-[␣-32 P]ATP binding to rat ROS, suggesting a structural effect of Zn 2ϩ on rhodopsin (29). Four Zn 2ϩ were included in the crystal structure of rhodopsin at 2.8-Å resolution (30,31). Recently this crystal structure was refined, and seven Zn 2ϩ were iden-* This work was supported by Grant PM98-0134 from Dirección General de Enseñ anza Superior e Investigación Científica, from the Spanish Ministry of Science and Technology, and in part by grants from Fundación Lucha contra la Ceguera and Dirección General de la ONCE. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  tified, three per monomer and one which was at a boundary region of the two monomers forming the unit cell of the crystal, as well as water molecules that would play a functional role in rhodopsin activation (32). Four of these ions are located in the intradiscal domain of the receptor. Although these binding sites occur under crystallization conditions, they also provide evidence for direct Zn 2ϩ binding to rhodopsin.
In the present study we have undertaken the study of the Zn 2ϩ effect on rhodopsin structure and stability, by taking advantage of the proposed endogenous Zn 2ϩ -binding sites in the receptor. We find a clear specific effect of Zn 2ϩ on the thermal stability of rhodopsin when compared with other metal ions tested. At the same time we can also see an effect for this metal ion on several properties of rhodopsin like pigment regeneration and metarhodopsin II (MetaII) stability. The results obtained indicate that Zn 2ϩ clearly reduces the thermal stability and regeneration of rhodopsin in detergent solubilized form and in ROS membranes. These results suggest that Zn 2ϩ binding to rhodopsin results in specific effects on the structure and stability of the receptor. This effect alone may not be accountable for the physiological role proposed for Zn 2ϩ in visual function. Other effects like those related to increased rhodopsin phosphorylation, or the proposed effect upon other proteins of the visual phototransduction cascade, or upon some component of the retinoid cycle, may also play a role in vivo.
Water used for buffer preparation was Milli-Q (Millipore) quality and was received with 10 -18 m⍀. Traces of zinc were detected using a polarographic instrument (Metrohm Ltd., Herisau, Switzerland). The zinc concentration was determined to be Ͻ5 ϫ 10 Ϫ2 M by the inverse voltammetric (anodic stripping) method, using the hanging mercury drop electrode.
ROS membranes were prepared under dim red light from frozen bovine retinas using a sucrose gradient method. The membranes were suspended in isolation buffer (buffer A): 70 mM potassium phosphate, 1 mM MgCl 2 , and 0.1 mM EDTA (pH 6.9) and pelleted by centrifugation. The membrane pellets were resuspended in 5 mM Tris-HCl (pH 7.5) and 0.5 mM MgCl 2 (hypotonic buffer). Two alternating washes with these buffers were carried out to decrease the amount of any further contaminating proteins. Finally, ROS membranes were split in several aliquots and stored in the dark at Ϫ20°C. Typically, 1 mg of ROS membranes was solubilized in 2 ml of buffer A containing 1% DM. Solubilization was done at 4°C in the dark for 1 h (33). The solution was then centrifuged for 30 min in a T865 Sorvall rotor at 30000 rpm, and the supernatant was used for the assays as solubilized rhodopsin (Rho). Rho was usually stored in the dark at 4°C or kept at Ϫ20°C for long term storage. Rho concentration was determined by using a molar extinction coefficient value ⑀ 500 of 40,600 M Ϫ1 cm Ϫ1 , and a molecular weight of 40,000 Da. The preparations showed UV/vis absorbance ratios (A 280 /A 500 ) in the 1.8 -2.0 range.
Sample Preparation-Zinc binding to Rho was carried out in the dark. Rho was diluted with 20 mM HEPES, containing 145 mM NaCl, 2 mM MgCl 2 , 0.5 mM CaCl 2 , and 0.012% DM in the spectrophotometric cuvette to a final DM concentration of 0.02% DM. These preparations were monitored by UV/vis spectroscopy. A stock solution of 1 M ZnCl 2 was prepared in the same dilution buffer. Small aliquots of the stock zinc solution were added to obtain the final desired zinc concentration in the sample. The sample was then incubated for 1 h at 15°C or 20°C in the dark, and it was monitored by recording UV/vis spectra at 5-min intervals over cycle mode to ensure that there was no spectral change with time. The procedure for the control experiments was the same, but zinc addition was omitted. Sample preparation for ROS rhodopsin in membranes (unsolubilized samples) was carried out essentially as described for the detergent-solubilized samples. Other metal ions (Ca 2ϩ , Co 2ϩ , Cd 2ϩ , and Cu 2ϩ ) were assayed as chloride salts to rule out unspecific charge effects. In these cases, sample preparation was carried out in the same way as with zinc. Choline was used as a control model compound to completely rule out the possible unspecific electrostatic nature of the observed effects.
UV/Vis Spectra of Rhodopsin-All measurements were made on a Cary 1E spectrophotometer (Varian, Australia), equipped with waterjacketed cuvette holders connected to a circulating water bath. Temperature was controlled by a Peltier accessory connected to the spectrophotometer. All spectra were recorded, in the 250 -650-nm range, with a bandwidth of 2 mm, a response time of 0.1 s, and scan speed of 250 nm/min. The protein sample binding reaction was carried out in the spectrophotometric cuvette during 1 h at 20°C and was monitored by UV/vis spectroscopy as described above. These reaction samples were used for photobleaching, thermal stability, and regeneration experiments. The experiments were performed in triplicate. Fourth derivative spectra were obtained, from five absorption spectra coadded to improve the signal-to-noise ratio, by using the Savitzsky-Golay algorithm and 200 points with the Grams/32 software (Galactic Industries).
Photobleaching and Acidification of Rho-Samples were bleached with a 150 watt power source equipped with an optic fiber guide and using a 495 nm cut-off filter. Dark-adapted Rho samples with 50 and 100 M zinc or without zinc were illuminated for 10 s to ensure complete photoconversion to 380-nm absorbing species. Acidification was carried out, immediately after photobleaching, by addition of 2 N H 2 SO 4 to a final pH of 1.9 (1% of the sample volume), and after 2 min to allow for stabilization an absorption spectrum was recorded.
Photobleaching Efficiency and Metarhodopsin III (MetaIII) Formation-In the case of the photobleaching efficiency experiments, darkadapted Rho samples with 100 M zinc or without zinc were illuminated with subsaturating light (with a 25 watt light source and a distance between the sample and the light source of 5 cm) at 5-s intervals. Immediately at the end of each bleaching interval, changes in the absorption spectra were recorded until complete photobleaching was achieved (complete disappearance of the 500-nm chromophoric band).
The formation of MetaIII was measured, after Rho illumination for 60 s (saturating illumination), in the presence of different zinc concentrations or without zinc (control) at 15°C. Complete absorption spectra in the range studied were recorded starting immediately after illumination and at 2.5-min intervals thereafter. MetaIII formation was monitored at A 480. All absorbance values were normalized to the same initial A 500 nm before photoconversion.
Thermal Stability Assay-Rho thermal stability in the dark was followed by monitoring the rate of A 500 loss, and the appearance of A 380 as a function of time, at constant temperature. Rho, in the presence of 5, 15, 30, 50, 100, and 200 M zinc, or in the absence of zinc (control), was thermally bleached in the dark at 55°C. The process was carried out in a similar way as previously described (34) over 3 h. Complete absorption spectra were recorded at 5 min intervals. Other experiments, using the same conditions but with 50 M calcium and other metal ions, were performed to rule out a possible unspecific charge effect. Thermodynamic parameters for the dark thermal bleaching process were determined as a function of temperature. This method regards thermal bleaching as an irreversible process (34). The thermodynamic parameters E a , ⌬G ‡ , ⌬H ‡ , and ⌬S ‡ were calculated from the decay rate constants versus 1/T (Arrhenius plot) using the values for the slopes of these plots, which were derived from least-squares analysis. For this purpose, Rho samples in the dark and in the presence of 50 M zinc (or without zinc in the control sample) were thermally bleached in the dark at different temperatures in the range of 53 to 58°C.
Regeneration of Rhodopsin-The regeneration of Rho was carried at 15°C in the dark. 11-cis-Retinal in an ethanolic stock solution was added to a 2.5-fold molar excess over dark Rho in the presence of 50, 100, and 200 M zinc or without zinc (control). Then, the sample was illuminated for 60 s (light-saturated), and absorption spectra were immediately recorded at 2.5-min cycles over 4 h. Rho regeneration rate was monitored at 500 nm. Samples were eventually acidified to confirm the presence of Schiff base-linked species. Regeneration experiments were also carried out with ROS rhodopsin samples essentially in the same way as for the detergent-solubilized samples.

UV Spectroscopic Characterization of the Zn 2ϩ -Rhodopsin
Complex-The presence of Zn 2ϩ , at a concentration of 200 M, did not affect the spectral properties of rhodopsin in the dark or in the light (Fig. 1). The characteristic 500-nm visible chromophoric band remained unchanged in the Zn 2ϩ containing sample (compare Fig. 1, A and B). The presence of Zn 2ϩ did not interfere with the ability of rhodopsin to form the 380-nm band, corresponding to the MetaII intermediate, upon illumination under light-saturating conditions (Fig. 1B) indicating the same photobleaching behavior than the non-zinc containing sample (Fig. 1A). In both cases the presence of a protonated Schiff base linkage was detected by acidification of the illuminated samples, with the corresponding shift of the band to 440 nm, confirming that the 380-nm band can be attributed to the MetaII species, and indicating that the protonation status of the Schiff base linkage was not affected by Zn 2ϩ (Fig. 1B). These results are in agreement with studies using wild-type recombinant rhodopsin in which even in the presence of 5 mM Zn 2ϩ similar results were obtained (19), suggesting that the rhodopsin-Zn 2ϩ complex is functional, although some differences in the transducin activation ability were detected in the presence of zinc (19). The photobleaching efficiency is not altered by zinc when rhodopsin is illuminated with subsaturating light intensities ( Fig. 2A). However, MetaIII formation is decreased in the presence of zinc when compared with the non-zinc-containing sample (Fig. 2B). Furthermore, when Rho samples are acidified at different times after photobleaching a 2-fold decrease under the studied conditions in the MetaII stability can be detected (the determined t1 ⁄2 values are 30 and 15 min in the absence and in the presence of Zn 2ϩ , respectively).
Fourth Derivative of the UV Absorption Spectra-Fourth derivative spectroscopy allows separation of the different electronic components corresponding to Phe, Tyr, and Trp residues in proteins. This allows a detailed analysis of the structural environment of these aromatic amino acids to be performed (35). The fourth derivative represents the best compromise between the best resolution and the best signal-to-noise ratio (35). This technique has been successfully applied to the study of conformational changes in native and recombinant proteins (36). No significant differences were detected in the fourth derivative UV spectra of rhodopsin in the dark upon Zn 2ϩ addition to the sample (Fig. 3, A and B and Table I). The wavelengths of the fourth derivative peaks at 285.6 and 292.2 nm and troughs at 289.0 and 295.5 nm remained unchanged upon zinc addition to the sample (compare a traces in Fig. 3, A  and B; see Table I). However, some differences were detected after illumination of the control rhodopsin sample in the absence of zinc (compare a and c traces in Fig. 3A; see Table I). These differences were increased when Zn 2ϩ was present in the sample (compare Fig. 3, A and B; see Table I) with larger blue-shifts in the 275-300-nm region and increased intensity changes in some of the bands. In contrast, peaks in the 240 -275 nm underwent only minor changes when compared with the spectra of the non-zinc-containing samples.

Zn 2ϩ -induced Decrease in the Thermal Stability of Rhodopsin-
The thermal stability of Rho in the presence of different concentrations of Zn 2ϩ was determined at 55°C in the dark. The kinetics of the thermal bleaching process of rhodopsin in the dark was spectroscopically monitored by measuring the decay of the 500-nm chromophoric band at constant temperature (55°C) as a function of time (Fig. 4A). At this temperature, a progressive faster decay of this band was observed as a function of Zn 2ϩ concentration with regard to the control sample without Zn 2ϩ . This reflects a decrease in the thermal stability of rhodopsin in its ground state (dark state) with increasing Zn 2ϩ concentrations. Thus, the thermal process is accelerated in the presence of Zn 2ϩ with k rate constants of 0.028 Ϯ 0.002 min Ϫ1 (control, 0 M Zn 2ϩ ) and 0.056 Ϯ 0.003 min Ϫ1 at 50 M Zn 2ϩ (Fig. 4A, inset). It corresponds to thermal bleaching processes with t1 ⁄2 values of 24.4 Ϯ 1.6 min and 11.8 Ϯ 0.1 min, respectively. Thermodynamic parameters derived from Arrhenius plots obtained from experiments performed at different temperatures (Fig. 4B) are shown in Table II. Activation energy (E a ), and the ⌬H ‡ component were significantly increased (by about 58%) for the thermal bleaching process of rhodopsin in the presence of Zn 2ϩ . A ⌬G ‡ decrease of 17% was also observed in the presence of Zn 2ϩ . In the case of the entropic component, this showed an increase of about 63% in the zinc-containing sample.
The previous experiments were done with Rho samples. We also performed similar experiments with rhodopsin in its native membrane environment (ROS membranes). Although rhodopsin in membranes is more stable than Rho (compare Fig. 5 FIG. 1. UV/vis spectroscopic characterization of the Zn 2؉ -rhodopsin complex. A, rhodopsin in the absence of Zn 2ϩ ; B, rhodopsin in the presence of 200 M Zn 2ϩ . UV/vis absorption spectra of rhodopsin were measured in buffer A with 0.02% DM at 20°C. Absorption spectra were measured before (dark spectra) and immediately after illumination for 10 s with a fiber-optic light equipped with a Ͼ495-nm long-pass filter (light spectra). Acid spectra were recorded after the addition of 2 N H 2 SO 4 (to a final pH 1.9) to the illuminated samples. Comparison of the spectral properties of Zn 2ϩ -and non-Zn 2ϩ -containing samples indicates that conversion of rhodopsin to MetaII and the acidification behavior (protonation of the Schiff base) were not affected by the formation of the Zn 2ϩ -rhodopsin complex.  (traces b and c). Spectrum labeled as light (trace c) was obtained 9 h after illumination. Fourth derivative spectra were obtained from five coadded absorption spectra by using the Savitzky-Golay algorithm as described under "Experimental Procedures." and Fig. 4A), a decrease in stability is also observed upon addition of 100 M zinc when compared with the sample with no zinc added (the t1 ⁄2 derived for the process is 121.6 min in the absence of zinc and 53.0 min in the presence of 100 M zinc) (Fig. 5).
Specific Effect of Zn 2ϩ on the Thermal Stability of Rhodopsin-In order to assess the presumed specificity of zinc on the thermal stability of rhodopsin several other metal divalent cations were tested. To this aim the decrease of the visible band at 500 nm with time was recorded as a measure of the thermal stability of the protein as described under "Experimental Procedures." The results indicate that in the presence of Cd 2ϩ , Co 2ϩ , and Ca 2ϩ the thermal stability of rhodopsin was similar to that of rhodopsin alone (Fig. 6). Among the different metal ions tested, only Cu 2ϩ showed a significant effect with a similar decay to that observed in the case of Zn 2ϩ . This result is in agreement with the reported ability of Cu 2ϩ to compete with Zn 2ϩ for rhodopsin binding, both in membranes and in purified form in detergent (27). We also tested the model compound choline to rule out a mere possible electrostatic effect. Choline increased the thermal stability of rhodopsin and this effect could be related to interactions of the molecule with rhodopsin micelles.
Regeneration of Rhodopsin in the Presence of Zn 2ϩ -The effect of zinc on pigment regeneration with 11-cis-retinal after rhodopsin photobleaching is shown in Fig. 7A. A clear decrease The wavelength for the maximum and minimum peak values ( max and min ) were obtained from UV fourth derivative spectra (see Fig. 3). Tyr and Trp residues of rhodopsin (control, in the absence of Zn 2ϩ ) in the dark and after bleaching have different wavelength location reflecting a blue-shift (⌬). In the presence of Zn 2ϩ (200 M Zn 2ϩ ), these peaks become further blue-shifted reflecting changes in Tyr and Trp residues of rhodopsin to a more polar environment.

TABLE II
Thermodynamic parameters for the thermal bleaching of rhodopsin These parameters were determined from experiments in the absence and in the presence of 50 M Zn 2ϩ . The E a parameter was determined using values from the slopes of the Arrhenius plots (Fig. 4B), which were derived from least-squares analysis. ⌬H ‡ ϭ E a Ϫ RT; In k ϭ Ϫ⌬G ‡/ RT; and ⌬G ‡/RT ϭ ⌬H ‡/RT ϩ ⌬S ‡ /R. All values were derived for T ϭ 55°C: R ϭ 8.3 J⅐ mol Ϫ1⅐ K Ϫ1 . These parameters were significantly different for the Zn 2ϩ -rhodopsin complex (*, p Ͻ 0.05).  is observed for maximal rhodopsin regeneration with increasing zinc concentration. This decrease is about 30% at 200 M added zinc. Higher zinc concentrations tested indicated further reduction of the regeneration percentage (to very low regeneration level), but the corresponding values could not be reliably obtained due to a high aggregation propensity of the samples with time. Regeneration experiments carried out with rhodopsin in ROS membranes (non-solubilized samples, Fig. 7B) indicated a similar behavior to that found for Rho samples (compare Fig. 7, A and B). In the case of ROS membranes the maximal regeneration observed was slightly lower than that observed for the Rho samples, but the effect of zinc addition was comparable.

DISCUSSION
Previous studies have reported specific Zn 2ϩ binding to purified rhodopsin and disc membranes in both dark and light conditions (27,29). Furthermore, Zn 2ϩ was shown to enhance rhodopsin phosphorylation (29). Structurally bound zinc has been also found in the recent crystal structure of rhodopsin (30 -32). Despite these reports, a detailed study of the effect of Zn 2ϩ on rhodopsin properties has not been previously reported. We find a clear effect of zinc on several rhodopsin properties. Our spectroscopic results indicate that zinc does not alter the retinal binding pocket of rhodopsin and the formation of the MetaII intermediate. Also, the status of the protonated Schiff base is not affected by the presence of zinc ions.
Under physiological conditions, inactive dark state rhodopsin changes almost entirely to the active MetaII conformation upon photon absorption. In vitro there is an equilibrium between the MetaI and MetaII conformations that is modulated by several experimental conditions (37)(38)(39)(40). It has been recently shown that this equilibrium is also influenced by salts, which shift the equilibrium to the MetaII side (41)(42)(43). In our case and under our experimental conditions, zinc does not seem to alter the MetaI-MetaII equilibrium.
We find that the MetaII intermediate of the Zn 2ϩ -rhodopsin complex formed upon illumination is destabilized, and MetaIII formation is reduced. These two facts could be reconciled if we assume as a likely interpretation that the faster MetaII decay is also accompanied by direct formation of opsin and free alltrans-retinal (without going through a MetaIII intermediate).
Detailed analysis of the fourth derivative results indicates small but significant differences as a consequence of illuminating the rhodopsin sample. Changes in the amplitude of the fourth derivative peaks together with blue-shifts of their max-ima can be interpreted as reflecting exposure of Trp and Tyr residues to a more hydrophilic microenvironment upon the rhodopsin-MetaII conformational transition. This is consistent with previous studies on UV photobleaching difference absorption spectra of ROS rhodopsin (44) and rhodopsin tryptophan mutants (45). The spectroscopic differences observed upon rhodopsin photobleaching are increased, mainly in the Tyr-Trp region, when Zn 2ϩ is present in the illuminated sample. The observed effect suggests differential Zn 2ϩ binding to dark and illuminated rhodopsin. Zinc ions found in the crystal structure of rhodopsin are mainly located at the intradiscal side of the protein (32). In particular, one of these zinc ions is located close to Tyr-96 in chain B of the receptor (32).
Thermodynamic parameters measure macroscopic changes representing the sum of all reaction events of the Zn 2ϩ -rhodopsin complex formation. The thermal bleaching process is accompanied by a decrease in ⌬G ‡ and an increase in ⌬H ‡ and ⌬S ‡ for the Zn 2ϩ -containing samples when compared with the values in the absence of Zn 2ϩ . Changes in enthalpy can be interpreted as reflecting conformational changes, and entropy effects can be interpreted as a reordering of water molecules as proposed previously (23,24). Similar interpretation has been provided for the binding of zinc to the D2-like dopamine receptor. In this latter case, the fact that the reaction was entropy driven suggested that zinc was essentially chelated by dopamine receptors, possibly resulting in the reordering of hydrogen bonds (23). Thermodynamic changes could reflect a structural rearrangement of native rhodopsin, during the formation of the Zn 2ϩ -rhodopsin complex, without affecting the retinal binding pocket and the protonated Schiff base linkage in the dark state of rhodopsin. The observed effect, reflecting reordering of water molecules, may be significant in the case of rhodopsin where a functional role for water molecules has been recently proposed (32). In this latter study more than ten water molecules are found structurally associated with each one of the two monomers, and two of these water molecules are in the intradiscal domain of the receptor, where four of the seven zinc atoms are located (32).
The preceding section of the discussion has dealt with results obtained with Rho (rhodopsin solubilized in DM detergent). However, it is known that rhodopsin is more stable in native ROS membranes than in detergent-solubilized form. Thus, it could be possible that the effect observed with Rho would be different when rhodopsin was assayed in disc membranes. Our results show that, although the rhodopsin samples are more stable in membranes, the differential effect found in the case of the zinc-containing sample is maintained, i.e. rhodopsin in membranes with zinc has lower thermal stability than control samples without zinc. Furthermore, the same behavior is detected in the case of the pigment regeneration experiments. In that case the reduction in rhodopsin regeneration induced by zinc is also observed when rhodopsin is found in its native membrane. These findings argue for the physiological relevance of the results obtained.
A structural effect of Zn 2ϩ on rhodopsin was also detected in a previous study, using engineered Zn 2ϩ -binding sites. In this case Zn 2ϩ binding to wild-type recombinant rhodopsin resulted in a somewhat reduced transducin activation (19). We find that zinc specifically alters several rhodopsin properties: a definite reduction in thermal stability of ground state dark rhodopsin, reduced stability of the active MetaII conformation, altered pathway of photointermediates after MetaII formation, and reduced pigment regeneration with 11-cis-retinal. These effects are accompanied by conformational changes of the protein that are shown to be specific of zinc and not of other divalent metal ions studied. The effect of zinc on retinal function seems to strongly depend on the local concentration and the ratio of free to bound cation (10). The proposed effect of Zn 2ϩ on visual function may be also related to effects upon other proteins of the visual phototransduction cascade (28) or upon some component of the retinoid cycle. Furthermore, the proposed increase in rhodopsin phosphoylation caused by Zn 2ϩ (29) could be also related to other processes, like receptor internalization, necessary for its recycling, thus changing the normal turnover of the receptor in the membrane.