Functional interaction between bovine rhodopsin and G protein transducin.

To elucidate the mechanisms of specific coupling of bovine rhodopsin with the G protein transducin (G(t)), we have constructed the bovine rhodopsin mutants whose second or third cytoplasmic loop (loop 2 or 3) was replaced with the corresponding loop of the G(o)-coupled scallop rhodopsin and investigated the difference in the activation abilities for G(t), G(o), and G(i) among these mutants and wild type. We have also prepared the Galpha(i) mutants whose C-terminal 11 or 5 amino acids were replaced with those of Galpha(t), Galpha(o), and Galpha(q) to evaluate the role of the C-terminal tail of the alpha-subunit on the specific coupling of bovine rhodopsin with G(t). Replacement of loop 2 of bovine rhodopsin with that of the scallop rhodopsin caused about a 40% loss of G(t) and G(o) activation, whereas that of loop 3 enhanced the G(o) activation four times with a 60% decrease in the G(t) activation. These results indicated that loop 3 of bovine rhodopsin is one of the regions responsible for the specific coupling with G(t). Loop 3 of bovine rhodopsin discriminates the difference of the 6-amino acid sequence (region A) at a position adjacent to the C-terminal 5 amino acids of the G protein, resulting in the different activation efficiency between G(t) and G(o). In addition, the binding of region A to loop 3 of bovine rhodopsin is essential for activation of G(t) but not G(i), even though the sequence of the region A is almost identical between Galpha(t) and Galpha(i). These results suggest that the binding of loop 3 of bovine rhodopsin to region A in Galpha(t) is one of the mechanisms of specific G(t) activation by bovine rhodopsin.

Visual pigment is one of the G protein-coupled receptors that has diverged into a photoreceptive protein in the retinal visual cells (1). It is a 35-55-kDa membrane protein consisting of a single polypeptide opsin and a chromophore 11-cis-retinal. The opsin contains seven transmembrane ␣-helices, the structural motif typical of the G protein-coupled receptors. Light isomerizes the 11-cis-retinal into the all-trans form, which induces the conformational changes of the opsin to activate the retinal G protein. Several lines of evidence have now revealed that there are at least three subtypes of visual pigments, each of which couples with the G t , 1 G q , or G o subtype of the G protein, respectively (2). Among the G protein subtypes, the latter two have been widely distributed in a variety of cells in various animals (3), whereas the G t has been identified only in the photoreceptor cells (4) and some types of taste cells in vertebrates (5,6). 2 In addition, G t biochemically exhibits unique characteristics different from other types; that is, G t can be easily solubilized in an aqueous solution without detergent (7), and it exhibits a high affinity to GDP (8). The latter characteristic of G t is physiologically important because it is one of the key characteristics to account for the low noise of the signal transduction. Therefore, together with the inverse agonist nature of the chromophore in visual pigment (9), G t is thought to be the highly differentiated G protein to mediate the signal transduction cascade with low noise.
Although vertebrate rhodopsins couple with G t in the native photoreceptor cells, they exhibit the ability to activate G i and G o as evidenced by the in vitro studies (10). From this point of view, bovine rhodopsin could be classified as a receptor that couples with the G i family of the G protein. However, among most G i /G o -coupled receptors, only a few receptors exhibit efficient activities to G t (11)(12)(13)(14), and therefore, vertebrate rhodopsins could have acquired the specific sequence(s) or the structure in the molecules during the course of evolution. 3 The process of interaction between bovine rhodopsin and G t has been extensively investigated by various techniques, and it was suggested that the second, third, and extra cytoplasmic loops of bovine rhodopsin are involved in the G t activation (15)(16)(17)(18)(19)(20)(21)(22)(23)(24). However, almost all of the experiments were performed using only G t and its peptide fragments, and it is still unclear which of the interaction sites is specific to the G t subtype. That is, there is no evidence for the interaction site(s) to account for the activation ability specific for G t . Therefore, elucidation of the specific interaction site(s) between vertebrate rhodopsins and G t is important for furthering our understanding of the signal transduction system in vertebrates.
In the present study, we have tried to obtain a clue to the specific interaction site(s) between bovine rhodopsin and G t . For this purpose, we have constructed bovine rhodopsin mutants whose second or third cytoplasmic loops were replaced with the corresponding loop of the G o -coupled scallop rhodopsin (see Fig. 1A) and investigated the difference in the activation efficiencies for G t , G o , and G i among these mutants and wild type. Furthermore, to elucidate the mechanism of the specific interaction between the cytoplasmic loops of bovine rhodopsin and the C-terminal region of the G protein ␣-subunit, we have constructed the G␣ i mutants whose C-terminal 11 or 5 amino acids were replaced with those of G␣ t , G␣ o , and G␣ q (see Fig.  1B) and investigated the difference in the activation efficiencies for these mutants between the wild type and the loopreplaced mutants of rhodopsin. These results showed that the third cytoplasmic loop of bovine rhodopsin discriminates between G t and G o by recognizing the region including 6 amino acids (see Fig. 1B, region A) adjacent to the region including the C-terminal 5 amino acids (see Fig. 1B, region B), whereas that of the SCOP2 does not discriminate region A. There was little difference in the activation efficiency of G i between the wild type rhodopsin and the loop 3-replaced mutant, which suggested that the interaction between loop 3 of bovine rhodopsin and region A is not important for catalyzing the GDP-GTP exchange reaction on G i . From these findings, the molecular mechanism of the coupling specificity of bovine rhodopsin with G t and the functional diversity among the G proteins are discussed.

Materials-[ 35 S]GTP␥S (37 TBq/mmol) was purchased from
PerkinElmer Life Sciences. The cDNA encoding rat G␣ i1 was isolated from the cDNA library of rat brain (kindly provided by Dr. H. Itoh, Tokyo Institute Technology). The hybridoma that produces the antibovine rhodopsin monoclonal antibody was generously supplied by Dr. R. Molday (University of British Columbia). Oligomers of DNA for the PCR mutagenesis were purchased from Amersham Biosciences. The DNA encoding bovine opsin, which has 30 unique restriction sites, was synthesized to carry out the cassette mutagenesis.
Preparation of Loop-replaced Mutant-The third loop-replaced mutant cDNA fragment was constructed by using standard PCR mutagenesis techniques with the SCOP2 cDNA as the template. The PCR fragments were constructed into the bovine opsin cDNA restriction sites, XhoI and XbaI for the loop 2 replacement and BsrGI and NdeI for the loop 3 replacement, and the constructed DNAs were ligated into the expression vector SR␣ using the restriction sites HindIII and EcoRI. Expression and purification of the rhodopsins were carried out according to our previous methods (25). Briefly, the visual pigments were extracted in 1% dodecylmaltoside in a buffer composed of 50 mM HEPES and 140 mM NaCl (pH 7.0). They were absorbed into the rhodopsin antibody 1D4 and then eluted by the buffer supplemented with the epitope for the antibody, a C terminus nonapeptide (TSQVAPA), and 0.02% dodecylmaltoside. The visible spectra were recorded with a Shimadzu MPS-2000 recording spectrophotometer at 4°C (26).
Preparation of G Proteins-G t and G o were purified from bovine retinas and brains according to Refs. 27 and 28, respectively. The rat G␣ i1 subunit was expressed in Escherichia coli strain BL21 using the G␣ i1 cDNA constructed into the pQE6 plasmid vector and was prepared as already described (29). For expression of the mutant G␣ i1 s having the C-terminal 11 amino acids of G␣ q , G␣ o , and G␣ t (see Fig. 1), PCR methods were applied to replace the C terminus sequence of the G␣ i1 cDNA by those of the rat G␣ q , G␣ o , and G␣ t cDNAs. The wild type and mutant G␣ i1 s were unmyristoylated. 4 The purified wild type and mutant G␣ i1 s were mixed with an equal amount of transducin ␤␥ before use.
G t concentration was determined by the method of Bradford (30). G o and G i concentration were estimated from the saturated intrinsic GTP␥S binding to them in the presence of 10 M GTP␥S at 20°C.
GTP␥S Binding Assay-Activation of the G protein by visual pigments was measured by monitoring the amount of GTP␥S bound to the G protein according to our previous report (31,32). All of the procedures were carried out at 15°C (G t assay) and 0°C (G o and G i assays). The rhodopsin solutions were irradiated with orange light for 30 s or kept dark and then immediately mixed with the G protein solution. The buffer composition of the mixture is the same in the G t , G i , and

. Comparison of primary sequence of bovine and scallop G o -coupled rhodopsins at loops 2 and 3 (A)
and G proteins at the C terminus of G␣ (B). The amino acids in loops 2 and 3 are denoted by the bold characters. In A, the numbering is shown as the bovine rhodopsin system. In B, the numbers above the sequences refer to the amino acid positions on the context of G␣ t , and the C-terminal 11 amino acids of G␣s followed by the ␣5 domain are divided into two parts, regions A and B (details in the text).
should be noted that we selected the experimental conditions under which no intrinsic uptakes of GTP␥S in G t , G i , and G o were observed to precisely compare the activation efficiencies of rhodopsin and its chimeras to one of the G proteins of different characteristics. We also compared the efficiencies of activation of the various G proteins by rhodopsin and chimeras after confirming that the activation efficiencies of G proteins were linearly correlated with the concentrations of pigments and that these pigments concentrations exhibited the linear kinetics up to 90 s in each G protein activation assay (see Fig. 2). Fig. 1A shows the alignment of the amino acid sequences of the second and the third cytoplasmic loops of the rhodopsins. According to the alignment, we prepared mutants of bovine rhodopsin in which the second or the third cytoplasmic loop was replaced with the corresponding loop of SCOP2, the G ocoupled scallop rhodopsin. All of the mutants and wild type rhodopsin exhibited almost the same absorption maxima at 499 nm, and upon irradiation at 4°C, they were converted to meta II intermediates whose maxima were located at 380 nm (data not shown). These results enabled us to use the same irradiation conditions for all of the mutant and wild type rhodopsins for the GTP␥S binding assay.

RESULTS
The ability of the mutant rhodopsins to activate the G protein subtypes was assayed by monitoring the light-dependent GTP␥S binding to the G proteins (Fig. 2). We selected different experimental conditions under which no intrinsic exchanges of GTP␥S in G t , G i , and G o were observed to precisely compare the activation efficiencies of rhodopsin and its chimeras to one of the G proteins of different characteristics. Fig. 2 provides three kinds of information about our G protein activation assays. First, no intrinsic GTP␥S binding to G protein was detected under our experimental conditions. The experimental results of most effective rhodopsin to each G protein are shown in the figure to demonstrate that no intrinsic GTP␥S binding was observed in any rhodopsins. Second, a linear relationship between pigment concentration and G protein activation in each G protein assay is shown, where the G protein activation was measured 45 s after the activation of rhodopsin or the loopreplaced mutant. Finally, these pigment concentrations exhibited the linear kinetics up to 90 s in each G protein activation assay. We also confirmed that other rhodopsins exhibited a linear effect on G protein activation when the same amounts of these rhodopsins were added to the reaction mixture.
It should be noted that the G o assay was carried out at a low temperature (0°C) in the presence of high GDP and low GTP␥S concentrations to diminish a large intrinsic uptake of GTP␥S by G o . As a result, we added a relatively large amount of rhodopsin to effectively activate G o , although the initial rate of G o activation was still smaller than those of others. In the following experiments, we calculated the initial rates of G protein activation from the amounts of GTP␥S binding to G proteins for 45 s and compared the G protein activation abilities of wild type and mutant rhodopsin. The rhodopsin concentrations used in Figs. 3-6 (10 nM for G t , 100 nM for G o , and 10 nM for G i ) are in the range of the concentration that gave a linear effect on G protein activation.
Effects of Loop 2 Replacement on the Activation Efficiencies for G t , G o , and G i - Fig. 3 shows the effects of the loop 2 replacement of bovine rhodopsin with that of SCOP2 on the activation efficiencies for G t , G o , and G i . The replacement caused about a 40% loss of activity to G t , suggesting that the loop 2 sequence of SCOP2 affects the activation ability of bovine rhodopsin. A similar loss of activation ability was observed when G o was subjected to the experiments, although the loss of the G i activation ability was not very prominent. These results showed that the replacement effect was not specifically observed for the G t activation, suggesting that the loss of the activation ability is not due to the deficiency of the specific coupling with G t but due to the inadequate sequence of the SCOP2 for the formation of the active state of bovine rhodopsin. In a previous study, we have demonstrated that the Nterminal 7 amino acids of loop 2 are important for the formation of active state of bovine rhodopsin (32). Among the 7 amino acids in the introduced loop 2 of SCOP2, the cysteine at position 134 has a characteristic different from that of the glutamic acid in the ERY motif of bovine rhodopsin. In fact, the replacement of Cys 134 with glutamic acid in the loop 2-replaced mutant rescued its G t activation efficiency to 80% of the wild type (data not shown).

Effects of Loop 3 Replacement on the Activation Efficiencies for G t , G o , and G i -
The replacement of loop 3 with the corresponding loop of SCOP2 resulted in a 60% loss of the activation ability for G t (Fig. 4A); the extent of the loss is more prominent than that observed in the loop 2-replaced mutant (Fig. 3A). In contrast, the mutant activated G o four times more efficiently than the wild type (Fig. 4B), suggesting that loop 3 of SCOP2 is more suitable for the G o activation than that of bovine rhodopsin, or loop 3 of bovine rhodopsin could not take a proper conformation to efficiently activate G o .
Because bovine and scallop rhodopsins are colocalized with G t and G o , respectively (2), in the native photoreceptor cells, it is reasonable that these rhodopsins exhibit some specificity to G t and G o . The present results showed that loop 3 is one of the regions specifically interacting with the G proteins, which is in contrast to the role of loop 2 that could be related to the conformational change of the rhodopsin molecule to form the state activating the G proteins (32). However, we could not directly compare the activation efficiency of G t by the wild type or the loop 3-replaced mutant with that of G o , because, in addition to the different character of each G protein, the experimental conditions to assay G protein activations were different among the G proteins. In other words, we could not estimate whether both the loops of the wild type and mutant or either of the loops exhibit different affinities to these G proteins.
One of the ways to overcome these difficulties is the construction of G protein mutants from one of the G protein subtypes, whereas the mutants have the sequences responsible for the specific coupling with the receptor subtypes. Several mutational analyses of the G protein ␣-subunit suggested that the C terminus is one of the determinants for the characteristics of the G protein subtype in receptor coupling, which is well conserved among each subtype (33)(34)(35). Therefore, we have prepared the G␣ i mutants in which the C-terminal sequences were replaced with the corresponding sequences of G␣ t , G␣ o , and G␣ q , respectively. Under the same experimental conditions as wild type G i , no G i mutant exhibited a significant uncatalyzed rate of exchange of GTP␥S (data not shown).

Activation Efficiencies of G i Mutants by Wild Type and Loopreplaced Mutants of Bovine
Rhodopsin-To examine the effect of the C terminus replacements, we first prepared the G␣ i mutants whose C-terminal 5 amino acids were replaced with those of G␣ o and G␣ q . Hereafter, the mutants are referred to as G i/o 5 and G i/q 5, respectively. It should be noted that the Cterminal 5 amino acids of G␣ i are identical to those of G␣ t (Fig.  1B).
The wild type rhodopsin and the loop 3-replaced mutant exhibited a similar ability to activate G i /wild and G i /o5, whereas they exhibited little ability for G i /q5 (Fig. 5). The latter FIG. 3. Initial rates of G protein activations by wild type and loop 2-replaced mutant rhodopsin. Activation of G t , G o , and G i by wild type and mutant rhodopsin, which possesses Loop2 of G o -coupled rhodopsin (SCOP2), was measured by a GTP␥S binding assay. 10, 100, and 10 nM rhodopsins were used in G t , G o , and G i assay, respectively. In each G protein activation assay, the initial rate (fmol/min/mol rhodopsin) was calculated from the GTP␥S bound G protein for 45 s. It should be noted that the rate (G o activation/1 min/mol rhodopsin added) is smaller than those of others, because high GDP and low GTP␥S concentrations as well as a larger amount of rhodopsin were used in G o assay to diminish the large intrinsic uptake of GTP␥S by G o . The data are the means Ϯ S.D. of at least three separate experiments.
is consistent with the fact that the C terminus sequence of G q is different from those of the G i group and is one of the regions for the impairment of the G i /G o -coupled receptor (33). On the other hand, we observed differences in the activation ability between the wild type and the mutant when the G␣ i mutants having the C-terminal 11 amino acids of G␣ t and G␣ o (G i /t11 and G i /o11, respectively) were subjected to the experiments (Fig. 6).
As expected from the sequence similarity of the C-terminal 11 amino acids between G␣ i and G␣ t , the wild type rhodopsin activated the G i /t11 with an efficiency similar to that of G i /wild. However, it showed only about a 20% efficiency to activate G i /o11 (Fig. 6). The mutant having the loop 3 of SCOP2 activated G i /o11 as efficiently as G i /wild and G i /t11, and the activation efficiencies are similar to those of G i /wild and G i /t11 observed in the wild type (Fig. 6). These results indicated that loop 3 of bovine rhodopsin could not take the proper conformation to interact with the C-terminal 11 amino acids of G o , resulting in a less efficient activation of G i /o11. The interesting observation is that the lower activation efficiency of G i /o11 shown by the wild type relative to the mutant is comparable with the difference in the G o activation efficiency between these pigments (Fig. 4B). Therefore, the difference in the G o activation efficiency between the wild type and the mutant could be mainly due to the reduced activation ability of the wild type but not to the efficient activation ability of the mutant. Hereafter, the region of the 6 amino acids from the N terminus and that of the 5 amino acids from the C terminus of the 11 amino acids are referred to as regions A and B, respectively (Fig. 1B).
The mutant activated G i /o11 as efficiently as G i /wild, suggesting that the interaction of the loop 3 of SCOP2 with region A is not essential for the activation mechanism of G i . These results also suggested the specific role of the interaction be- FIG. 4. Initial rates of G protein activations by wild type and loop 3 replaced mutant rhodopsin. Activation of G t , G o , and G i by wild type and mutant rhodopsin having loop 3 of G o -coupled rhodopsin (SCOP2) was measured by GTP␥S binding assay. 10, 100, and 10 nM rhodopsins were used in G t , G o , and G i assay, respectively. In each G protein activation assay, the initial rate (fmol/min/mol rhodopsin) was calculated from the GTP␥S bound G protein for 45 s. The data are the means Ϯ S.D. of at least three separate experiments. Note that the mutant rhodopsin, which possesses loop 3 of SCOP2, showed around four times greater G o activation than that of wild type rhodopsin.
FIG. 5. Initial rates of activation of G i mutants having the C-terminal 5 amino acids of different G␣s. Two kinds of G␣ i mutants having the C-terminal 5 amino acids of G␣ o (G i /o5) and G␣ q (G i /q5) and wild type G␣ i (G i /wild) were subjected to GTP␥S binding assay by using wild type and loop 3-replaced mutant rhodopsins. Note that the absolute amounts of light-induced GTP␥S-bound for 1 min could be compared because all G protein assays were carried out under identical conditions. The data are the means Ϯ S.D. of three separate experiments. Wild type and mutant G proteins bound ϳ8000 fmol GTP␥S in the presence of 60 nM of the light-stimulated mutant rhodopsin having loop 3 of G o -rhodopsin upon incubation for 10 min, and the initial rates of wild type and mutant G i s, which were catalyzed by unirradiated rhodopsins, were less than 15 mol/min/mol rhodopsin (data not shown). tween the loop 3 of bovine rhodopsin and region A of G t in the activation mechanism of G t (see "Discussion").

DISCUSSION
As already described, bovine rhodopsin can be classified as one of the receptors that couple with the G i family of the G protein, but it has a specific ability to activate G t . The present studies using the loop-replaced rhodopsin mutants and the C terminus-replaced G␣ i mutants strongly suggest the specific role of region A (Fig. 1B) in the activation mechanism of G t by bovine rhodopsin. That is, bovine rhodopsin can discriminate the difference in the amino acid sequence of region A between G t and G o , suggesting that region A is one of the recognition sites of G t by bovine rhodopsin. Our spectroscopic experiments by using the C-terminal peptide (regions A ϩ B) supported our conclusion that is based on the analyses of G protein C terminus mutants; the peptide of the C-terminal 11 amino acids of G␣ t stabilizes the meta II intermediate of bovine rhodopsin in the rod outer segments more than two times greater than that of G␣ o (dissociation constants are about 2 and 5 mM, respectively; data not shown). Because the sequence of G t is identical with that of G o up to the 9 amino acids adjacent to the Cterminal 11 amino acids, it is likely that the 6 amino acids in region A are enough for the difference in the stabilization of meta II between G t and G o . Therefore, there could be present a specific interaction between loop 3 of bovine rhodopsin and region A of G t for the efficient activation of G t .
On the other hand, the wild type and loop-replaced mutant showed the activation ability for G i similar to each other (Figs. 3C, 5, and 6). This is in contrast to the fact that they showed the activation ability for G t considerably different from each other (Fig. 4). Because the sequence of region A is almost identical between G t and G i , these facts could originate from the different role of the interaction between loop 3 and region A. The difference in the activation efficiency of G t between the wild type and the loop 3-replaced mutant could be explained either by the efficient activation of G t by the loop 3 of bovine rhodopsin or less activation of G t by loop 3 of SCOP2. In the former case, the binding of the loop 3 of bovine rhodopsin with region A induces a proper conformational change(s) in the G t molecule to efficiently activate G t , whereas the proper conformational change(s) could not occur upon binding of the loop 3 with region A of G i . In the latter case, there is a specific activation mechanism of G t different from G i and G o , and the mechanism could not work upon the interaction of region A of G t with loop 3 of SCOP2. In either case, the experimental results could be explained by the specific interaction between the loop 3 of bovine rhodopsin and region A of G t .
Several kinds of evidence demonstrated the involvement of the ␣5 helix adjacent to the C terminus in G protein activation (36 -41). In addition to the sequence of region A, the sequence of the ␣5 is also highly homologous between G t and G i (40). These facts suggest that the conformational changes of the ␣5 helix upon binding of the loop 3 of bovine rhodopsin to region A could be similar between G t and G i . Thus it is likely that the binding of loop 3 of bovine rhodopsin to region A (and region B) could facilitate an additional conformational change(s) in the region other than the ␣5 helix in the G t molecule to induce the GDP-GTP exchange reaction on G t , whereas an additional conformational change, if any, may not be essential for the exchange reaction on G i . The difference in structure in the region other than the ␣5 helix between G t and G i , which affects the GDP-GTP exchange reactions of these proteins (42), was recently reported. Thus identification of the region that is affected upon binding of loop 3 of bovine rhodopsin to the region A will be our future research.
It remains to be determined which region in loop 3 of bovine rhodopsin is responsible for the binding to region A, although our experiments clearly showed a specific interaction between the loop 3 and region A. 5 To specify the binding region in loop 3, we divided loop 3 of bovine rhodopsin into three segments and prepared six chimerical mutants between bovine rhodopsin and SCOP2, as has already been done in the loop 2 mutants of bovine rhodopsin (32). However, we were unable to specify the region, probably because the number of amino acids in the loop 3 of bovine rhodopsin is different from those in the loop 3 of SCOP2 (Fig. 1A). On the other hand, we found that the mutant of bovine rhodopsin, where amino acids at positions 237 to 249 were deleted (⌬237-249; Ref. 15), showed no ability for both G t and G o . Because the deleted mutant still contains the four hydrophobic residues (Val 250 , Thr 251 , Val 254 , and Ile 255 ) responsible for the selective interaction with the G i group of the G protein by binding to the C-terminal 5 amino acids (34), it is reasonable that the deleted region 237-249 in the loop 3 is important for the binding to region A. The interesting observation is that the deleted mutant showed a partial ability to activate G i (30% of that shown by the wild type), whereas it showed a loss of G t and G o activation as described above. Because the deleted region forms a unique loop structure with the extrusion to the cytoplasmic surface in the crystal structure of bovine rhodopsin (43), its binding to region A might cause the proper conformation of the region adjacent to the ␣5 helix in the G␣ t molecule, but the binding to region A may not be necessary for the conformational change of the corresponding region in the G i molecule.
In summary, the present study shows the presence of a specific interaction between the loop 3 of bovine rhodopsin with region A in G␣ t , which is one of the candidates to account for 5 During the course of the preparation of this manuscript, Khorana and co-workers (47,48) published the cross-linking data showing that the cysteine introduced at position 240 in the third cytoplasmic loop of the light-activated rhodopsin could contact with residues 342-345 in the C-terminal region, residues 310 -313 in the ␣4-␤6 loop, and residues 19 -28 at the N-terminal helix.
FIG. 6. Initial rates of activation of G protein mutants by rhodopsin mutants. Three kinds of G protein mutants were prepared by substitution of the C-terminal 11 amino acids of G␣ i (G i /wild) by those of G␣ t (G i /t11), G␣ o (G i /o11), and G␣ q (G i /q11). Activation of these G proteins by wild type and rhodopsin mutants, which possessed loop 3 of SCOP2, was measured. The data are the means Ϯ S.D. of three separate experiments. Wild type and mutant G proteins bound ϳ9000 fmol GTP␥S in the presence of 60 nM light-stimulated mutant rhodopsin having loop 3 of G o -rhodopsin upon incubation for 10 min, and the initial rates of wild type and mutant G i s, which were catalyzed by unirradiated rhodopsins, were less than 14 mol/min/mol rhodopsin (data not shown). Note that the different absolute amounts of G i/wild between Figs. 4 and 5 were due to different preparations of G proteins. the coupling specificity of bovine rhodopsin with G t . As far as we know, this is the first experimental evidence that accounts for the specificity of bovine rhodopsin to G t . Further experiments via a strategy of deletional and site-directed mutagenesis will shed light on the mechanism of the G t specificity of vertebrate rhodopsin and its relation to the diversity of the receptor-G protein coupling.