Regulation of Transducin GTPase Activity by Human Retinal RGS*

The intrinsic GTPase activity of transducin controls inactivation of the effector enzyme, cGMP phosphodiesterase (PDE), during turnoff of the visual signal. The inhibitory γ-subunit of PDE (Pγ), an unidentified membrane factor and a retinal specific member of the RGS family of proteins have been shown to accelerate GTP hydrolysis by transducin. We have expressed a human homologue of murine retinal specific RGS (hRGSr) in Escherichia coli and investigated its role in the regulation of transducin GTPase activity. As other RGS proteins, hRGSr interacted preferentially with a transitional conformation of the transducin α-subunit, Gtα GDPAlF 4 − , while its binding to GtαGTPγS or GtαGDP was weak. hRGSr and Pγ did not compete for the interaction with Gtα GDPAlF 4 − . Affinity of the Pγ-Gtα GDPAlF 4 −interaction was modestly enhanced by addition of hRGSr, as measured by a fluorescence assay of Gtα GDPAlF 4 −binding to Pγ labeled with 3-(bromoacetyl)-7-diethylaminocoumarin (PγBC). Binding of hRGSr to Gtα GDPAlF 4 −complexed with PγBC resulted in a maximal ∼40% reduction of BC fluorescence allowing estimation of the hRGSr affinity for Gtα GDPAlF 4 −(K d 35 nm). In a single turnover assay, hRGSr accelerated GTPase activity of transducin reconstituted with the urea-stripped rod outer segment (ROS) membranes by more than 10-fold to a rate of 0.23 s−1. Addition of Pγ to the reconstituted system reduced the GTPase level accelerated by hRGSr (k cat 0.085 s−1). The GTPase activity of transducin and the PDE inactivation rates in native ROS membranes in the presence of hRGSr were elevated 3-fold or more regardless of the membrane concentrations. In ROS suspensions containing 30 μm rhodopsin these rates exceeded 0.7 s−1. Our data suggest that effects of hRGSr on transducin’s GTPase activity are attenuated by Pγ but independent of a putative membrane GTPase activating protein factor. The rate of transducin GTPase activity in the presence of hRGSr is sufficient to correlate it with in vivo turnoff kinetics of the visual cascade.

In vertebrate photoreceptor cells, the signal is transduced from light-activated rhodopsin to the effector enzyme, cGMPphosphodiesterase (PDE), 1 via the heterotrimeric G-protein, transducin (G t␣␤␥ ). The GTP-bound ␣-subunit of transducin (G t␣ GTP) relieves the inhibition imposed by two inhibitory PDE ␥-subunits (P␥) on the enzyme catalytic ␣␤ subunits (P␣␤). Activation of PDE leads to a closure of cGMP-gated channels in the photoreceptor plasma membranes (1)(2)(3). The inactivation of PDE is a critical component of the turnoff mechanism in the visual transduction cascade. This inactivation is controlled by the intrinsic GTPase activity of transducin which hydrolyzes GTP to GDP. The GDP-bound G t␣ (G t␣ GDP) has a substantially reduced affinity for P␥ and releases P␥ to reinhibit P␣␤ (1, 4 -6). The rate of GTP hydrolysis by transducin measured in vitro (7,8) is too slow to account for the fast photoresponse turnoff in vivo (9,10). The P␥ subunit (11,12) and a distinct membrane-associated protein factor (13,14) have been shown to enhance transducin GTPase activity in the activated membrane-bound transducin-PDE complex to a level comparable with the rate of transducin inactivation in vivo. A recent study has shown that a retinal specific member of the RGS family, RGSr, serves as a GTPase-activating protein (GAP) for transducin, providing an additional dimension to an already complex picture of the regulation of transducin GTPase activity (15). Functional relationships between RGSr, the ␥-subunit of PDE, and a putative membrane GAP factor are currently not understood.
Here, we study the interaction between transducin and a human retinal specific RGS (hRGSr), and regulation of transducin GTPase activity by hRGSr. We examine the effects of P␥ and photoreceptor membrane concentration on modulation of the GTPase activity by hRGSr.
Preparation of ROS Membranes, G t␣␤␥ , G t␣ GTP␥S, G t␣ GDP, G t␤␥ , and P␥BC-Bovine ROS membranes were prepared as described previously (16). Urea-washed ROS membranes (uROS) were prepared according to protocol in Ref. 17. Hypotonically washed ROS membranes were prepared as described in Ref. 14. Transducin, G t␣␤␥ , was extracted from ROS membranes using GTP as described in Ref. 18. The G t␣ GTP␥S was extracted from ROS membranes using GTP␥S and purified by chromatography on Blue-Sepharose CL-6B by the procedure described in Ref. 19. G t␣ GDP was prepared and purified according to protocols in Ref. 20. P␥BC was obtained and purified as described in Ref. 6. The purified proteins were stored in 40% glycerol at Ϫ20°C or without glycerol at Ϫ80°C.
Cloning and Expression hRGSr-A BLAST search at NCBI (Bethesda, MD) to compare the mouse mRGSr cDNA sequence against DNA sequence data bases revealed a human homologue, A28-RGS14p (the GenBank accession number U70426), which is 85% identical to mRGSr. DNA prepared from the amplified human retinal cDNA gt10 library (kindly provided by J. Nathans, Johns Hopkins University) was used as a template for the polymerase chain reaction amplification with primers that were synthesized based on the A28-RGS14p sequence. The polymerase chain reaction was performed in 30 l of reaction mixture containing 100 ng of the template DNA and 0.5 M of the following primers: ATACTCTAGACATGTGCCGCACCCTGGC (5Ј) and ATGC-CTCGAGACTCAGGTGTGTGAGG (3Ј). The polymerase chain reaction product (620 bp) was digested with XbaI and XhoI (the restriction sites are underlined) and subcloned into the pGEX-KG vector (21) for GST-hRGSr fusion protein expression. The DNA sequence was verified by automated DNA sequencing at the University of Iowa DNA Core Facility using the 3Ј-pGEX sequencing primer (Pharmacia) and the 5Јprimer shown above. The subcloned sequence was different from A28-RGS14p at two nucleotide positions of the open reading frame (C 125 3 T and A 160 3 G), leading to substitutions of amino acid residues Ser 42 3 Phe and Asn 54 3 Asp. Typically, expression host E. coli DH5␣ cells were grown on 2 ϫ TY medium and induced at OD 600 ϭ 0.5 by addition of isopropyl-1-thio-␤-D-galactopyranoside (0.4 mM final concentration). After a 4-h induction at 37°C, cells were harvested and sonicated in 20 mM Tris-HCl (pH 8.0) buffer containing 100 mM NaCl, 2 mM MgCl 2 , 6 mM ␤-mercaptoethanol, and 5% glycerol (buffer A). The supernatant (100,000 ϫ g, 1 h) was loaded on a glutathione-agarose column. GST-hRGSr was eluted from the column using 50 mM Tris-HCl buffer (pH 8.0) containing 10 mM glutathione. GST-hRGSr was then passed through a PD-10 column (Pharmacia) to separate glutathione and digested with thrombin (0.25 NIH units/mg) for 90 min at room temperature. hRGSr was reapplied on a glutathione-agarose column to remove GST.
Binding of Transducin to GST-hRGSr-agarose-G t␣ GDP or G t␣ GTP␥S (6 M final concentration) were mixed with glutathioneagarose retaining ϳ10 g of hRGSr in 40 l of 20 mM HEPES buffer (pH 7.6), 100 mM NaCl, and 2 mM MgCl 2 (buffer B). Where indicated, the buffer contained 30 M AlCl 3 , 10 mM sodium fluoride, and 10 -30 M P␥. After incubation for 20 min at room temperature, the agarose beads were spun down, washed with 1 ml of buffer B, and the bound proteins were eluted with a sample buffer for SDS-polyacrylamide gel electrophoresis.
Fluorescence Assays-Fluorescence assays were performed on a F-2000 Fluorescence Spectrophotometer (Hitachi) in 1 ml of buffer B essentially as described in Ref. 6. Where indicated, the buffer contained 30 M AlCl 3 and 10 mM sodium fluoride. Typically, hRGSr was added to P␥BC prior to addition of transducin. The assays were carried out at equilibrium which was reached in less than 3 s after mixing of the components except when G t␣ GDP and AlF 4 Ϫ were used. In the latter experiments the equilibrium due to G t␣ GDP activation by AlF 4 Ϫ was reached in less than 15 s. Fluorescence of P␥BC was monitored with excitation at 445 nm and emission at 495 nm. Concentration of P␥BC was determined using ⑀445 ϭ 53,000.
Analytical Methods-Single turnover GTPase activity measurements were carried out essentially as described in Ref. 22. The reaction was initiated by mixing bleached ROS membranes with 200 nM [␥-32 P]GTP (ϳ5 ϫ 10 4 dpm/pmol) in a total volume of 20 l. The reaction was quenched by addition of 100 l of 7% perchloric acid. Nucleotides were then precipitated using charcoal, and 32 P i formation was measured by liquid scintillation counting. Concentrations of transducin in different preparations of ROS membranes were determined using the [ 35 S]GTP␥S binding assay. ROS membranes were incubated with 2 M GTP␥S (10 5 dpm/pmol) for 10 min at room temperature in 20 l of buffer B, and the mixture was applied onto GF/B filters (Millipore). The filters were washed with 3 ml of 40 mM Tris-HCl buffer (pH 8.0) containing 100 mM NaCl and 4 mM MgCl 2 and counted in a liquid scintillation counter. To determine if hRGSr or P␥ may affect the nucleotide binding to G t␣␤␥ , 200 nM [ 35 S]GTP␥S was added to uROS membranes (5 M rhodopsin) reconstituted with 0.4 M G t␣␤␥ and 1 M hRGSr (and/or 1 M P␥) in 20 l of buffer B. The mixture was immediately (Ͻ2 s) diluted into 3 ml of cold 40 mM Tris-HCl buffer (pH 8.0) containing 100 mM NaCl, 4 mM MgCl 2 , and 2 M GTP␥S. The GTP␥S binding was then measured as described above. The PDE activity was measured using the proton-evolution assay as described in Ref. 23. Protein concentrations were determined by the method of Bradford (24) using IgG as a standard or using calculated extinction coefficients at 280 nm. SDS-polyacrylamide gel electrophoresis was performed by the method of Laemmli (25) in 12% acrylamide gels. Rhodopsin concentrations were measured using the difference in absorbance at 500 nm between "dark" and bleached ROS preparations. The K d and IC 50 values were calculated as described in Ref. 26. The results are expressed as the mean Ϯ S.E. of triplicate measurements.

RESULTS
Expression and Purification of hRGSr-The sequence encoding for the human homologue of mouse retinal RGSr protein was polymerase chain reaction amplified from the retinal cDNA library, subcloned into the pGEX-KG vector, and expressed as a GST fusion protein as described under "Experimental Procedures." Eluate of GST-hRGSr from a column with glutathione-agarose is shown in Fig. 1 (lane 2). GST-hRGSr migrated on SDS gels with the expected mobility of a 52-kDa protein. Appearance of an additional doublet at ϳ28 -30 kDa, which most likely contains GST polypeptide, is typical for the GST fusion expression system (27). hRGSr, purified after cleavage of the fusion protein GST-hRGSr with thrombin, migrated as a 25-kDa protein (Fig. 1, lane 3). The yield of hRGSr was ϳ6 mg/liter of culture.
Binding of GST-hRGSr to G t␣ GDPAlF 4 Ϫ , G t␣ GTP␥S, and G t␣ GDP-A number of RGS proteins (RGS1, RGS4, and mouse RGSr) have been shown to interact preferentially with a transitional G ␣ GDPAlF 4 Ϫ conformation of G ␣ subunits (15,28,29). We examined the ability of GST-hRGSr bound to glutathioneagarose to co-precipitate G t␣ in different conformations. Fig. 2A shows that GST-hRGSr precipitated stoichiometric amounts of G t␣ GDPAlF 4 Ϫ , while amounts of G t␣ GTP␥S and G t␣ GDP that co-precipitated with GST-hRGSr were significantly lower. In control experiments, G t␣ GDPAlF 4 Ϫ , G t␣ GTP␥S, and G t␣ GDP did not co-precipitate with glutathione-agarose that contained no bound GST-hRGSr (not shown). The results suggest that the affinity of hRGSr for the G t␣ conformations decreases in the following order: G t␣ GDPAlF 4 Ϫ Ͼ Ͼ G t␣ GTP␥S Ͼ G t␣ GDP.

Effects of P␥ on the Interaction between hRGSr and G t␣ GDPAlF 4
Ϫ -To determine if P␥ can compete with hRGSr for the interaction with G t␣ GDPAlF 4 Ϫ , we initially tested effects of P␥ on G t␣ GDPAlF 4 Ϫ binding to GST-hRGSr. Even at high concentrations (up to 30 M) P␥ did not affect binding of G t␣ GDPAlF 4 Ϫ to GST-hRGSr immobilized on glutathione-agarose (Fig. 2B). We next investigated effects of hRGSr on the interaction between P␥ and G t␣ GDPAlF 4 Ϫ or G t␣ GTP␥S using a fluorescence assay. Addition of G t␣ GDPAlF 4 Ϫ to a fluorescently labeled P␥, P␥BC, produced an approximately 7.5-fold maximal increase in the BC fluorescence (Fig. 3A), while G t␣ GTP␥S enhanced the fluorescence of P␥BC by more than 6-fold (not shown). The K d values for the G t␣ GDPAlF 4 Ϫ and G t␣ GTP␥S binding to P␥BC were 2.8 Ϯ 0.1 and 2.1 Ϯ 0.1 nM, respectively. The affinity of G t␣ GDPAlF 4 Ϫ binding to P␥BC was somewhat higher in the presence of 100 nM hRGSr (K d 1.2 Ϯ 0.1 nM), suggesting that hRGSr and P␥ bind to G t␣ GDPAlF 4 Ϫ noncompetitively (Fig. 3A). Addition of hRGSr had no effect on the fluorescence of P␥BC alone (not shown), but resulted in a dose-dependent decrease in the fluorescence enhancement of P␥BC caused by the latter binding to G t␣ GDPAlF 4 Ϫ (Fig. 3B). The fluorescence was decreased maximally by ϳ40% with an IC 50 of ϳ35 nM. Since hRGSr and P␥ interact with G t␣ GDPAlF 4 Ϫ noncompetitively, this IC 50 value may serve as an estimate for the affinity of hRGSr interaction with G t␣ GDPAlF 4 Ϫ . In control experiments, hRGSr did not affect the fluorescence of the G t␣ GTP␥S⅐P␥BC complex (not shown).
Effects of hRGSr on Transducin's GTPase Activity-Single turnover measurements of GTPase activity were carried out as described under "Experimental Procedures." Under these conditions ([GTP] Ͻ [G t ]) the GTPase reaction can be analyzed using an exponential function: % GTP hydrolyzed ϭ 100(1-e -kt ), where k is a rate constant for GTP hydrolysis. First, transducin GTPase activity was measured in the reconstituted system with uROS membranes. According to previous reports, uROS membranes lack the GAP activity of a membrane factor, and the P␥ subunit did not affect the GTPase activity of transducin when uROS membranes were used (14). The calculated rate of GTP hydrolysis by transducin (0.4 M) reconstituted with uROS containing 5 M rhodopsin was 0.022 Ϯ 0.001 s Ϫ1 (Fig.  4A). Addition of 1 M hRGSr resulted in acceleration of the GTPase activity by more than 10-fold (k ϭ 0.23 Ϯ 0.01 s Ϫ1 ). The basal GTPase activity of transducin was not significantly altered in the presence of 1 M P␥ (k ϭ 0.019 Ϯ 0.003 s Ϫ1 ). However, P␥ substantially reduced the accelerated GTPase activity of the G t␣ ⅐hRGSr complex (k ϭ 0.085 Ϯ 0.005 s Ϫ1 ). In control experiments, hRGSr, P␥, or the two proteins combined had no notable effect on the binding of [ 35 S]GTP␥S to transducin under similar conditions (not shown). The reaction was complete in less than 2 s.
The basal GTPase activity of transducin in the suspension of untreated native ROS membranes containing 5 M rhodopsin and 0.41 M transducin was higher (k ϭ 0.052 Ϯ 0.002 s Ϫ1 ) than in the reconstituted system with uROS (Fig. 4B). However, the elevation of transducin GTPase activity by hRGSr was only ϳ3.3-fold (k ϭ 0.17 Ϯ 0.01 s Ϫ1 ). Perhaps, the presence of PDE containing the P␥ subunit in native ROS preparations prevented a larger enhancement of the GTPase activity by hRGSr as seen using uROS. To test this idea we have prepared hypotonically washed dROS membranes depleted of PDE. The GTPase activity in suspensions of dROS containing 5 M rhodopsin and 0.38 M transducin (k ϭ 0.031 Ϯ 0.001 s Ϫ1 ) was accelerated by ϳ7.5-fold (k ϭ 0.23 Ϯ 0.02 s Ϫ1 ) with addition of hRGSr (Fig. 4C). The P␥ subunit increased transducin's GTPase activity by more than 2-fold to a rate of 0.075 Ϯ 0.003 s Ϫ1 . Effects of hRGSr and P␥ on transducin were not additive. Moreover, the GTPase activity of transducin was lower in the presence of both hRGSr and P␥ (k ϭ 0.18 Ϯ 0.01 s Ϫ1 ) than in the presence of hRGSr alone. Interestingly, hRGSr enhanced the GTPase rates in suspensions of untreated ROS to levels similar to those in suspensions of dROS membranes reconstituted with P␥.
The GTPase activity of transducin has been shown to elevate with increasing concentrations of ROS membranes presumably due to the action of a putative membrane GAP factor (11)(12)(13)(14). At low concentrations of ROS membranes the G t␣ GTP⅐P␥ complex dissociates from P␣␤ and is mainly soluble (30,31). Increasing the membrane concentration shifts the equilibrium toward the membrane bound complex (G t␣ GTP) 2 P␣␤␥ 2 (32)(33)(34) allowing the interaction of G t␣ GTP with a putative GAP factor. The calculated rate for the GTP hydrolysis in the suspension of native ROS membranes containing 30 M rhodopsin was ϳ3fold higher (k ϭ 0.16 Ϯ 0.02 s Ϫ1 ) than that seen using 5 M rhodopsin (Fig. 4D). hRGSr was at least as effective at higher concentrations of ROS membranes as it was in diluted ROS suspensions. We estimate that the rate of GTP hydrolysis in Ϫ , and G t␣ GTP␥S to GST-hRGSr immobilized on glutathione agarose was carried out as described under "Experimental Procedures." Lanes: 1, G t␣ GDP; 2 and 3, G t␣ GDP and G t␣ GDPAlF 4 Ϫ bound to GST-hRGSr, respectively; 4, G t␣ GTP␥S; 5, G t␣ GTP␥S bound to GST-hRGSr. B, effects of P␥ on binding of G t␣ GDPAlF 4 Ϫ to GST-hRGSr. Acceleration of PDE Inactivation by hRGSr-We measured PDE inactivation in suspensions of bleached native ROS membranes using the proton evolution assay (23). The activation and inactivation of PDE was monitored after addition of GTP under single turnover conditions using a pH microelectrode. The PDE activity was maximal in less than 1 s after addition of GTP. Therefore, we were unable to resolve the activation phase. As shown earlier, the rate of PDE inactivation can be well approximated by fitting the inactivation phase with a single exponential decay function (14). The PDE activity in the assay is proportional to the concentration of active G t␣ GTP. The change of pH due to hydrolysis of cGMP under single GTP turnover conditions can be described using an exponential function: pH ϭ ⌬pH max (1-e -kt ) or [cGMP]hydrolyzed ϭ ⌬[cGMP] max (1-e -kt ). The PDE activity represents a derivative of this function, and it decays with the inactivation constant k from the equation above. In diluted suspensions of native ROS (5 M rhodopsin), PDE was inactivated with a rate of 0.096 s Ϫ1 . In the presence of hRGSr, PDE inactivation was enhanced to a rate of 0.31 s Ϫ1 (Fig. 5A). The increase in k inact (Ͼ3-fold) correlated well with the decrease in maximal amounts of cGMP hydrolyzed in the presence of hRGSr (Fig. 5A). Furthermore, the acceleration of PDE inactivation caused by hRGSr was proportional to the elevation of transducin GTPase activity. We next tested effects of ROS membrane concentration on the modulation of PDE inactivation by hRGSr. The PDE inactiva-tion rate (0.26 s Ϫ1 ) was significantly higher in suspensions of ROS membranes containing 30 M rhodopsin than in diluted ROS suspensions (Fig. 5B). However, this enhanced rate was not accompanied by an equivalent decrease in the maximal amount of hydrolyzed cGMP because the initial PDE activity after addition of GTP was higher. It has been shown previously that PDE activation by transducin is more efficient at higher concentrations of photoreceptor membranes (32)(33)(34). Addition of 1 M hRGSr to ROS membranes containing 30 M rhodopsin resulted in ϳ3-fold increase in the PDE inactivation rate. The PDE inactivation rate was as high as 0.75 s Ϫ1 and correlated well with the GTPase rate (Ͼ0.7 s Ϫ1 ) measured under the same conditions. To determine if hGRSr can serve as an antagonist for PDE and block the enzyme activation, we have measured effects of hRGSr on GTP␥S-induced PDE activity in suspensions of ROS membranes containing 5 M rhodopsin. The rates of the GTP␥S-induced cGMP hydrolysis were not significantly affected by hRGSr. In the presence of relatively high concentrations of hRGSr (5 M) PDE activity was suppressed by only ϳ15% (not shown).

DISCUSSION
Recent findings have established that members of a new family of RGS proteins serve as GAPs for heterotrimeric Gproteins and attenuate G-protein-mediated signal transduction (28,35,36). Evidence suggests that RGS proteins accelerate the rate of GTP hydrolysis by G ␣ subunits but do not affect GDP/GTP exchange induced by activated G-protein-coupled receptors (35). Precise mechanisms of RGS GAP activity are not yet clear. It has been demonstrated that at least some RGS proteins (RGS1, RGS4, GAIP, and mRGSr) interact preferentially with the AlF 4 Ϫ bound conformation of G ␣ subunits and thus may accelerate GTP hydrolysis through stabilization of the transitional state of G-proteins (15,28,29). In several signaling systems the GTPase activity of G-proteins is enhanced by their effector enzymes (11,37). Regulation of the visual G-protein appears to be even more complicated. The ␥ subunit of PDE in concert with an unidentified membrane factor accelerate GTPase activity of transducin (11)(12)(13)(14). The mouse retinal specific RGS (mRGSr) protein has been recently identified and preliminarily characterized. mRGSr was shown to enhance GTPase activity under steady-state conditions (15).
We have investigated the regulation of transducin GTPase activity by human RGSr and examined effects of P␥ and photoreceptor membrane concentration on the functional activity of hRGSr. Our results support the data that retinal RGS binds tightly to the transitional conformation of transducin, G t␣ GDPAlF 4 Ϫ (15), and extend this observation by showing that it also weakly interacts with G t␣ GTP␥S and G t␣ GDP. Using a fluorescence assay of interaction between P␥BC and G t␣ GDPAlF 4 Ϫ , we demonstrated that P␥ and hRGSr interact with G t␣ GDPAlF 4 Ϫ noncompetitively. Furthermore, the affinity of the P␥/G t␣ GDPAlF 4 Ϫ interaction was modestly enhanced in the presence of hRGSr. This increase in affinity may reflect a stabilization of the G t␣ GDPAlF 4 Ϫ conformation which interacts with P␥ with high affinity. Binding of hRGSr to G t␣ GDPAlF 4 Ϫ affected the G t␣ conformation resulting in a decrease of the maximal fluorescence enhancement caused by G t␣ GDPAlF 4 Ϫ binding to P␥BC. We used this effect to estimate the affinity of the hRGSr/G t␣ GDPAlF 4 Ϫ interaction (ϳ35 nM). Interestingly, G t␣ GTP␥S and G t␣ GDPAlF 4 Ϫ had similar high affinities for P␥ and significantly different affinities for hRGSr, supporting the conclusion that G t␣ GDPAlF 4 Ϫ has distinct interfaces for interaction with P␥ and hRGSr. Comparison of the crystal structures of G t␣ GTP␥S and G t␣ GDPAlF 4 Ϫ shows limited conformational differences in the switch I and II regions of transducin (38). This suggests that hRGSr interacts with the switch I and/or switch II regions of G t␣ GDPAlF 4 Ϫ . Indeed, a crystal structure of RGS4 bound to G i␣1 AlF 4 Ϫ , that has been published during the review of this paper, shows that the RGS4-binding site is formed by the three switch regions of G i␣1 (39). The switch III, ␣3/␤5, and ␣4/␤6 regions of transducin have been earlier implicated in G t␣ interaction with P␥ (5, 40 -43). The P␥ and hRGSr-binding sites on G t␣ may possibly overlap, particularly, at the switch III region. Because the interactions between the switch III region and RGS protein are not extensive (39), such potential overlap would not cause a significant effect of hRGSr on apparent affinity of P␥ binding to G t␣ .
Binding of RGS proteins to the switch regions of G-proteins supports the idea that RGS proteins may serve as antagonists for some effectors (39). RGS4 has been shown to block activation of phospholipase C␤ by G q␣ GTP␥S (44). In our experiments, initial rates of cGMP hydrolysis in suspensions of ROS membranes after addition of GTP were unaffected in the presence of hRGSr (Fig. 5). Likewise, hRGSr has little effect on PDE activity in ROS membranes stimulated by GTP␥S. It appears that the main mode of RGS action in the transducin/ PDE signaling system is to accelerate G-protein and effector inactivation rather than to block effector activation.
Measurements of GTPase activity under single turnover conditions presented in this study demonstrate that hRGSr can dramatically increase the k cat of GTP hydrolysis by transducin. Effects of hRGSr on transducin's GTPase activity were attenuated but not eliminated by P␥. Even in the presence of P␥, hRGSr accelerated the GTPase activity by ϳ3-fold. Similar inhibitory effects of P␥ on stimulation of transducin GTPase activity by mouse RGSr have just been reported (45). However, our data indicate that P␥ attenuates effects of RGSr allosterically rather than competitively as suggested by Wieland et al. (45). In agreement with earlier observations, the rates of GTP hydrolysis were significantly higher in more concentrated suspensions of photoreceptor membranes (11)(12)(13)(14). hRGSr was equally potent as a GAP for transducin regardless of the membrane concentration, indicating that: (a) a putative membrane GAP factor for transducin represents a distinct non-RGS-like protein, and (b) the membrane factor does not compete with hRGS for binding to transducin. A newly discovered second retina-specific RGS protein (RET-RGS1) is a membrane-bound protein and therefore is a potential candidate for a putative GAP factor (46). Our study indicates that RET-RGS1 and an unidentified GAP factor are likely to be different proteins.
The PDE inactivation rates were increased by hRGSr proportionally to the acceleration of the GTPase rates. The evidence that transducin's GTPase activity represents a major mechanism for inactivation of PDE in the turnoff of visual signals (11,13,47) has been disputed (48,49). Our data support the conclusion that transducin's GTPase activity controls PDE inactivation. The rates of GTP hydrolysis and PDE inactivation (Ͼ0.7 s Ϫ1 ) observed at relatively high concentrations of photo-