Transducin Activation State Controls Its Light-dependent Translocation in Rod Photoreceptors*

Light-dependent redistribution of transducin between the rod outer segments (OS) and other photoreceptor compartments including the inner segments (IS) and synaptic terminals (ST) is recognized as a critical contributing factor to light and dark adaptation. The mechanisms of light-induced transducin translocation to the IS/ST and its return to the OS during dark adaptation are not well understood. We have probed these mechanisms by examining light-dependent localizations of the transducin-α subunit (Gtα)in mice lacking the photoreceptor GAP-protein RGS9, or expressing the GTPase-deficient mutant GtαQ200L. An illumination threshold for the Gtα movement out of the OS is lower in the RGS9 knockout mice, indicating that the fast inactivation of transducin in the wild-type mice limits its translocation to the IS/ST. Transgenic GtαQ200L mice have significantly diminished levels of proteins involved in cGMP metabolism in rods, most notably the PDE6 catalytic subunits, and severely reduced sensitivity to light. Similarly to the native Gtα, the GtαQ200L mutant is localized to the IS/ST compartment in light-adapted transgenic mice. However, the return of GtαQ200L to the OS during dark adaptation is markedly slower than normal. Thus, the light-dependent translocations of transducin are controlled by the GTP-hydrolysis on Gtα, and apparently, do not require Gtα interaction with RGS9 and PDE6.

Heterotrimeric GTP-binding proteins (G proteins) propagate a variety of hormonal and sensory signals from specific cell surface receptors to intracellular effectors (1)(2)(3). The visual transduction cascade in vertebrate photoreceptors has served for many years as a paradigm for G protein signaling. In rod photoreceptor cells, illuminated rhodopsin stimulates GTP-GDP exchange on the retinal G protein, transducin (Gt), 2 resulting in dissociation of Gt␣GTP from Gt␤␥ and rhodopsin. Gt␣ in the active GTP-bound conformation stimulates the effector enzyme, cGMP phosphodiesterase (PDE6), by displacing the inhibitory ␥-subunits (P␥) from the PDE6 catalytic core (PDE6␣␤). cGMP hydrolysis by active PDE6 results in closure of cGMP gated channels in the plasma membrane (4,5). The turn-off phase of the visual signal is determined by reactions controlling the lifetimes of photoexcited rhodopsin (R*) and activated transducin. The catalytic function of R* is blocked by the rhodopsin-kinase mediated phosphorylation and the binding of arrestin to phosphorylated R* (6 -8). The lifetime of Gt␣GTP is controlled by intrinsic GTPase activity. Hydrolysis of GTP switches the Gt␣ molecule to the inactive GDP-bound conformation and allows reinhibition of PDE␣␤ by P␥. RGS9-1, a photoreceptor-specific member of the RGS (regulators of G protein signaling) family, in the complex with G␤5L acts as a GTPase-activating protein for transducin and thus is a major regulator of the turn-off kinetics of the visual signal (9 -11). The RGS9-1/G␤5L complex is anchored to disc membranes through the interaction with R9AP (RGS-9-1-anchor protein) that enhances the complex GAP activity (12)(13)(14). The recovery in the visual transduction cascade and light adaptation of photoreceptors are coupled to the changes in the intracellular Ca 2ϩ concentration. A decrease in intracellular Ca 2ϩ concentration resulting from the closure of cGMP-gated channels leads to activation of guanylate cyclase by specific Ca 2ϩ -binding guanylate cyclase-activating proteins (GCAPs) and restoration of cGMP levels (15,16). Enhanced guanylate cyclase activity in lightadapted photoreceptors adjusts the sensitivity of photoresponses and expands the receptive range of light intensities.
The light-induced movement of transducin from the rod outer segments (OS) into the inner segments (IS) and synaptic terminals (ST) of photoreceptor cells, and the movement of arrestin in the opposite direction were known for a long time (17)(18)(19). This phenomenon has attracted renewed attention with the application of quantitative approaches involving tangential microsectioning of flat mounted retinas (20). A reduction in the transducin content in the rod outer segment because of its translocation was correlated with a decrease in the amplification of the photoresponse, suggesting that the protein redistribution is a novel mechanism contributing to light adaptation (20). A recent study concluded that the distribution of arrestin is controlled by its interactions with rhodopsin in the OS and microtubules in the IS, and the movement of arrestin occurs by simple diffusion (21). The mechanisms of transducin translocation to the IS/ST in the light and back to the OS during dark adaptation remain largely unknown. Diffusion and active transport represent two potential general modes for the transducin translocation (20). In a diffusion model, the lightinduced dissociation of Gt␣GTP and Gt␤␥ from the disc membranes simply allows them to diffuse into the IS/ST. In the IS/ST, Gt␣ is converted into the GDP-bound state, forms a heterotrimer with Gt␤␥, and returns by diffusion to the OS during dark adaptation. Gt accumulates in the OS in the dark due to its affinity for the disc membranes and, perhaps, binding to the ground-state rhodopsin (22). Based on the model, the kinetics and/or the illumination threshold for Gt␣ movement to the IS should depend on the rate of transducin inactivation. GTPase-deficient mutants of Gt␣ can be predicted to undergo a rapid translocation to the IS and have slower return kinetics during dark adaptation. We have examined the mechanism of transducin movement using the RGS9 knock-out (11) and GTPase-deficient Gt␣Q200L mouse models. In agreement with the diffusion hypothesis, the illumination threshold for the Gt␣ movement to the IS/ST is lower in the RGS9 Ϫ/Ϫ mice, and the return of Gt␣Q200L to the OS during dark adaptation of the transgenic mice is markedly slower than normal. The Gt␣Q200L mice show dramatic down-regulation of the rod PDE6 catalytic subunits. Therefore, the observation of light-dependent movement of transducin in these mice suggests that the translocation mechanism is independent of the transducin/PDE6 interaction.

EXPERIMENTAL PROCEDURES
Generation of EE-tagged Gt␣ and Gt␣Q200L Transgenic Mice-All experimental procedures involving the use of mice were carried out in accordance with the NIH guidelines and the protocols approved by the University of Iowa, University of Utah, and Virginia Commonwealth University Animal Care and Use Committees. A pBRH Gt␣ transgenic construct containing the mouse Gt␣ genomic sequence of ϳ5.5 kb (23) flanked by the 4.4-kb mouse opsin promoter fragment (24) and polyadenylation signal was kindly provided by Dr. J. Lem (New England Medical Center). The Glu-Glu (EE) monoclonal antibody epitope was introduced into Gt␣ to quantitatively assess the level of expression of the transgene (see Fig. 1). Two substitutions required to convert Gt␣ sequence 162 GYVPTE 167 into the epitope sequence EYMPTE were created using the QuikChange mutagenesis kit (Stratagene) according to the manufacturer's protocol. Subsequently, a similar procedure was used to generate the transgene for expression of the EE-tagged Gt␣Q200L (see Fig. 1). DNA fragments of 9-kb were released from the pBRH-Gt␣EE and pBRH-Gt␣EEQ200L plasmids by restriction with NotI and gel purification. The Gt␣EE and Gt␣EEQ200L DNA fragments were microinjected into mouse embryos and implanted into pseudopregnant females in the transgenic core facilities at University of Utah. Transgenic mice were identified by PCR of mouse tail DNA with a pair of primers chosen to amplify a 300-bp fragment surrounding the junction between the Gt␣ and polyadenylation signal sequences. Fifteen potential Gt␣EE transgenic founders were generated; nine of them transmitted the transgenes after mating with C57BL/6 mice, and five transgenic lines expressed the Gt␣EE protein as assessed by Western blotting with anti-EE antibodies. Two Gt␣EEQ200L founders were obtained, and both transmitted the transgenes. One Gt␣EE line and one Gt␣EEQ200L line with high levels of transgene expression across the retina were selected for characterization and breeding with the rod Gt␣ knock-out mice (25) to move the transgenes into the hemizygous (Gt␣ ϩ/Ϫ ) background.
Immunoblot Analysis and Quantification of Transgene Expression-Total mouse retinal homogenates were obtained by solubilization of 1-2 retinas in 200 l of 10% SDS-Na using brief sonication and heating. Protein concentrations were determined using the Bio-Rad DC protein assay and bovine serum albumin dissolved in 10% SDS-Na as a standard. Typically, a total protein content of a homogenate obtained from a single mouse retina was ϳ400 g. Samples of retinal homogenates were subjected to SDS-PAGE in 10% gels, electrotransferred onto nitrocellulose membranes, and probed with antibodies against rod T␣ (K-20) The antibody-antigen complexes were detected using anti-rabbit, anti-mouse, or anti-sheep antibodies conjugated to horseradish peroxidase (Sigma) and ECL reagent (Amersham Biosciences). The EE-tagged transducinlike chimeric Gt␣ * (Gt␣ * EE) and the Q200L mutant (Gt␣ * Q200L) were constructed by PCR-directed mutagenesis similarly as described previously (26). Escherichia coli-expressed and purified Gt␣ * EE was used as a standard. In control experiments, the immunoblot signals of Gt␣ * EE were not notably affected by the presence of up to 40 g of retinal extracts from Gt␣ Ϫ/Ϫ mice (not shown). Nitrocellulose membranes were exposed to film, and integrated densities of scanned individual bands were measured with Scion Image software (version 4.0.3.2).
Immunofluorescence-For dark adaptation, mice were kept in the dark for an indicated period of time (up to 30 days). All dark procedures were performed under infrared illumination using night vision goggles. For light adaptation, the pupils were dilated by applying a drop of 1% tropicamide, followed by a drop of 2.5% phenylephrine hydrochloride 30 min prior to light exposure. During light exposure, mice that were free to move were kept in a Styrofoam box (20 ϫ 15 ϫ 15 cm) covered with semi-transparent polyethylene film to diffuse light. Light from white fluorescent bulbs in the ceiling was used for strong light exposure (ϳ500 lux, 45 min). Low level light exposures (2-8 lux, 25 min) were achieved with light from a light table reflected by a white ceiling. These conditions provided a uniform illumination from all directions as measured with a LX-1010B digital lux meter. The mice were euthanized with CO 2 . Mouse eyeballs were enucleated, poked through the cornea with a 21-gauge needle, and fixed in 4% formaldehyde in phosphate-buffered saline for 2 h at 22°C. After fixation, the eyeballs were cut in half, the cornea and lens were removed, and the eyecups were submersed in a 30% sucrose solution in phosphate-buffered saline for 5 h at 4°C. The eyecups were then embedded in tissue freezing medium (TBS) and frozen on dry ice. Radial sectioning (10 m) of the retina was performed using a cryomicrotome Microm HM 505E. Retinal cryosections were air-dried and kept at Ϫ80°C until use. Before staining, sections were warmed up to 25°C and incubated in 0.1% Triton/phosphate-buffered saline for 30 min followed by incubation with 2% normal goat serum/5% bovine serum albumin in phosphate-buffered saline for 30 min. Then sections were incubated with rabbit anti-rod Gt␣ antibody K-20  Electroretinography-The mice were dark-adapted overnight and prepared for recording the next morning under infrared illumination after anesthesia with a mixture of ketamine (150 mg/kg intraperitoneal) and xylazine (10 mg/kg intraperitoneal). The body temperature was maintained at 35-37°C using a homeothermic blanket. Pupils were dilated using equal parts of topical phenylephrine (2.5%) and tropicamide (1%). Drops of 0.9% saline were applied onto the cornea to prevent its dehydration and allow electrical contact with the recording electrode (a gold wire loop). A platinum subdermal needle was inserted under the scalp and between the two eyes to serve as the reference electrode.  Amplification (at 1-1000-Hz bandpass, without notch filtering), stimuli presentation, and data acquisition were performed on the UTAS-E 3000 visual diagnostic system from LKC Technologies Inc. Scotopic ERG responses to single flash presentations (10-sec duration) at increasing intensities covering a range from Ϫ3.6 to 1.4 log cds/m 2 were recorded. To ensure the dark-adapted state of the animals, interstimulus intervals have been programmed to increase from 10 s at lowest stimulus intensity to 2 min at highest stimulus intensity. Two Gt␣EEQ200L mice were analyzed by ERG with similar results.
Trypsin Protection Test-One Gt␣EE retina and one Gt␣EEQ200L retina from the dark-adapted mice (30-day dark adaptation) were homogenized in the dark by sonication in 400 l of 20 mM Hepes buffer (pH ϭ 8.0), containing 100 mM NaCl and 4 mM MgSO 4 . Aliquots of retinal homogenates (15 l) were incubated with trypsin (10 g/aliquot) for 30 min at 22°C, followed by the addition of 50 g of soy bean trypsin inhibitor. Additional aliquots of retinal homogenates were photobleached, followed by the addition of 100 M GTP␥S or GTP. After 5-min incubation at 22°C, the aliquots were treated with trypsin as above and analyzed by Western blotting using anti-EE antibodies Single Turnover GTPase Assay-Single turnover GTPase activity measurements were carried out in suspensions of uROS membranes (5 M rhodopsin) reconstituted with Gt␣* or Gt␣*Q200L (0.5 M) and Gt␤␥ (0.5 M) as previously described (27,28). The GTPase reaction was initiated by the addition of 50 nM [␥-32 P]GTP (ϳ4 ϫ 10 5 dpm/ pmol). The GTPase rate constants were calculated by fitting the exper-imental data to an exponential function: %GTP hydrolyzed ϭ 100(1 Ϫ e Ϫkt ), where k is a rate constant for GTP hydrolysis.

RESULTS
Transducin Translocation in RGS9 Knock-out Mice-A general translocation of Gt␣ in the rods of the RGS9 knock-out mice was examined following exposure of the dark-adapted mice to moderate illumination conditions (500 lux, 45 min). This light exposure caused redistribution of Gt␣ from the OS to the IS/ST, similarly to that observed in control C57BL/6 mice (Fig. 2). To assess the return of Gt to the OS, the RGS9 Ϫ/Ϫ and control mice were exposed to light (500 lux, 45 min) and then dark-adapted for 4, 10, or 24 h. The immunostaining of retinal cross-sections from the dark-adapted mice demonstrated that redistribution of Gt␣ during dark adaptation is similar in RGS9 Ϫ/Ϫ and control mice, and is largely complete within 10-h in the dark (Fig. 2). Quantification of protein distribution by immunofluorescence intensity has its limitations because of the potential antibody epitope masking and other related artifacts (20). Nonetheless, the estimates of Gt␣ content in the OS of the light-and dark-adapted wild type and RGS9 Ϫ/Ϫ mice by immunofluorescence are similar to those obtained previously by a more quantitative procedure of tangential retina microsectioning combined with immunoblotting ( Fig. 2B) (29).
The conditions of minimal illumination that cause movement of Gt␣ in control dark-adapted mice with dilated pupils were determined by varying the level of illuminance and the time of light exposure. The distribution of Gt␣ in mouse retinas from C57BL/6 mice after the light exposure equivalent to 2 photopic lux over 25 min was found to be identical to the Gt␣ distribution in the dark-adapted animals (Fig. 3). Small, but detectable movement of Gt␣ to the IS was observed in the control retinas using illumination of 4 and 8 lux for 25 min (Fig. 3). In contrast to control mice, the lowest level of light exposure (2 lux, 25 min) induced a considerable translocation of Gt␣ to the rod IS/ST in the RGS9 Ϫ/Ϫ mice (Fig. 4).
Transgenic Mouse Lines Expressing EE-tagged Gt␣ and the Q200L Mutant-To monitor the localization and translocation of the GTPasedeficient Gt␣Q200L mutant in the presence of native transducin the Glu-Glu (EE) antibody epitope was introduced into the Gt␣ helical domain. The EE-tag at the selected position had no effect on the signaling activity of different G protein ␣ subunits including G␣ q , G␣ s , and G␣ i (30,31). We first confirmed that the EE-epitope does not affect the key biochemical properties of Gt␣, whereas the Q200L mutation  impairs its GTPase activity in vitro. The EE-tagged transducin-like chimeric Gt␣* (26) and Gt␣*Q200L were constructed and expressed in E. coli. Gt␣*EE and Gt␣* displayed similar rates of GTP␥S binding in the presence of Gt␤␥ and R*, GTPase activities in single-turnover assays, and effector interactions in the binding assay with fluorescently labeled P␥ (not shown). The impact of the Q200L mutation on Gt␣ was assessed by measuring single turnover GTP hydrolysis in the reconstituted system with Gt␣*Q200L, Gt␤␥, and urea-treated ROS as a source of R*. The k cat for GTP hydrolysis by Gt␣* was 0.019 s Ϫ1 (Fig. 5). The Q200L mutation reduced the rate of GTP hydrolysis by Gt␣* under these conditions by at least 10-fold (Fig. 5). Even larger decreases in the rate constants of GTP-hydrolysis had been reported for the G␣ s and G␣ i Q/L mutants (32,33).
Two transgenic mouse lines expressing the highest levels Gt␣EEQ200L or Gt␣EE were selected for the initial characterization.
Examination of the levels of major phototransduction proteins revealed no significant differences between the Gt␣EE mice and control nontransgenic animals (not shown). In contrast, the rod PDE6 catalytic subunits were dramatically down-regulated in the Gt␣EEQ200L mice (Fig. 6A). The levels of GC-1, GC-2, GCAP-1, and GCAP2 were also significantly decreased (Fig. 6A). The reduction in the PDE6 content caused by the GTPase-deficient mutant is in agreement with a similar observation made in transgenic mice expressing the cone Gt␣Q204L mutant in mouse rods (34). The morphology of Gt␣EEQ200L animals at 1 month of age was found to be largely normal (Fig. 6B). The analysis of scotopic ERG responses of dark-adapted Gt␣EEQ200L mice revealed that the response threshold is elevated by three orders of magnitude in comparison to wild type controls, indicating a severe reduction of light sensitivity in transgenic rods (Fig. 6C).
After breeding the transgenic mice with the Gt␣ knock-out mice (25), the levels of total Gt␣ (native Gt␣ plus mutant), Gt␣EE, and Gt␣EEQ200L were assessed in mice with the hemizygous Gt␣ ϩ/Ϫ background using the Western blot analysis with anti-EE and anti-rod Gt␣ antibodies. The total levels of transducin were similar in the Gt␣EE and control mice and slightly lower in the Gt␣EEQ200L mice (ϳ85%). Relative to the native Gt␣ in the wild type mice, the hemizygous Gt␣EE mice were estimated to express ϳ50% of Gt␣EE, whereas the hemizygous Gt␣EEQ200L mice expressed ϳ40% Gt␣EEQ200L (Fig. 7).
Localization and Translocation of Transducin in the Gt␣EEQ200L Mice-The immunofluorescence staining of retinal sections from the light-(500 lux, 45 min) and dark-adapted hemizygous Gt␣EE transgenic mice using anti-rod Gt␣ and anti-EE antibodies demonstrated a wild type-like light-induced translocation of Gt␣EE to the IS/ST and its return to the OS during dark adaptation (Fig. 8). Similarly, Gt␣EEQ200L was primarily localized to the IS/ST compartments in the light-exposed Gt␣EEQ200L mice (Fig. 8). However, the return of Gt␣EEQ200L to the OS during dark adaptation of the transgenic mice

Light-dependent Movement of Transducin in Rods
DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 was markedly slower. Only a partial relocalization of Gt␣EEQ200L to the OS was observed after a 5-day period of dark adaptation. The mutant Gt␣ apparently continued to accumulate in the OS during extended dark adaptation for 30 days, but its return was still incomplete with a fraction of Gt␣EEQ200L present in the IS/ST (Fig. 8). The distri-bution of Gt␤␥ in the light-(500 lux, 45 min) and dark-adapted Gt␣EE and Gt␣EEQ200L mice was probed using anti-Gt␤ antibodies (Fig. 9). This analysis confirmed light-dependent translocation of transducin in rods expressing the GTPase-deficient mutant. A fraction of Gt␤␥ also remained in the IS/ST after a 30-day dark adaptation of the transgenic animals (Fig. 9).
Mouse retinal homogenates were obtained from the Gt␣EEQ200L and Gt␣EE mice dark-adapted for 30 days to determine the nucleotide occupancy of the Q200L mutant. The dark-kept or light-exposed retinal samples in the absence or presence of GTP or GTP␥S were tested for Gt␣ sensitivity to trypsin using immunoblotting with anti-EE antibodies. The tryptic cleavage site at Arg-204 is protected in the activated GTP/GTP␥S-bound Gt␣ leading to the formation of ϳ30-kDa proteolytic product. The trypsin protection test demonstrated that Gt␣EE is

GDP-bound (not protected) in the dark and GTP␥S-bound (protected)
in the light in the presence of GTP␥S (Fig. 10). However, a small fraction of Gt␣EEQ200L was GTP-bound in the dark. Unlike Gt␣EE, Gt␣EEQ200L was protected in the light not only in the presence of GTP␥S, but GTP as well (Fig. 10), which is a clear indication of the reduced GTPase activity (32). A partial protection of Gt␣EEQ200L because of endogenous GTP was detected in the light-exposed sample in the absence of added GTP (Fig. 10). The presence of the GTP-bound Gt␣EEQ200L after extended dark adaptation indicates that the impairment of GTPase activity because of the Q/L substitution in native Gt␣ is much more severe than that in the chimeric Gt␣* in vitro.
Localization of P␥ in Gt␣EEQ200L, RGS9 Ϫ/Ϫ , and Wild Type Mice-Intriguingly, the down-regulation of the rod PDE6 catalytic subunits in the Gt␣EEQ200L mutant mice did not alter the level of the inhibitory P␥ subunit (Fig. 6A). Perhaps, this P␥ was stabilized by the interaction with the GTP-bound Gt␣ mutant. Free P␥ is a small, soluble, and, presumably, readily diffusible protein. With the markedly reduced levels of PDE6␣␤ in the OS, free P␥ would be expected to distribute evenly between the OS and the IS/ST. However, the immunofluorescence staining of retinal sections from the light-adapted Gt␣EEQ200L mice showed localization of P␥ mainly to the IS/ST, which is consistent with the complex between P␥ and the QL mutant (Fig. 11). P␥ partially moved to the OS after 30 days of dark adaptation (Fig. 11). Evidently, the pool of GTP-bound QL mutant in the IS/ST becomes sufficiently small during prolonged dark adaptation to cause the release of P␥. The subsequent partial movement of P␥ to the OS might be because of simple equilibration of free P␥ between the IS and OS and/or the interaction with the small amount of PDE6␣␤ in the OS of Q200L mice. This observation and earlier studies showing the release of a soluble complex P␥-Gt␣GTP␥S from the membrane-bound PDE6 in vitro (35) pointed toward a possibility that P␥ might co-translocate with the light-activated Gt␣. Yet the immunofluorescence analysis of the retinal sections of the wild type, RGS9 Ϫ/Ϫ (Fig. 11), and Gt␣EE mice (not shown) showed that P␥ is localized in the dark in the OS and does not appreciably translocate with Gt␣ in response to light.

DISCUSSION
Light-dependent redistributions of transducin and arrestin in photoreceptor cells represent a novel mechanism of light/dark adaptation. The movements of transducin and arrestin are unaffected in mice lacking arrestin and transducin, respectively, demonstrating that the translocations of the two proteins are independent processes (36,37). Two general mechanisms of light-dependent redistribution of Gt␣ are plausible: diffusion and active transport with the help of motor proteins. Supporting the diffusion mechanism, the movement of arrestin, and perhaps transducin, is reported to be energy-independent (21). The diffusion mechanism is also favored for the recently discovered lightdependent translocation of recoverin (38). The rate of diffusion of soluble proteins in photoreceptors is fast (21), whereas the time course of Gt␣ translocation to the IS/ST upon exposure to light is relatively slow (20). If the activated Gt␣GTP moves by simple diffusion, then the rate of its translocation to the IS/ST must be limited by the fast inactivation of Gt␣ by the RGS9 GAP complex. Consequently, the movement of an individual Gt␣ molecule during continuous illumination would be composed of multiple steps of the Gt activation, release, and diffusion of Gt␣GTP followed by the GTP hydrolysis and rebinding of Gt to a disc membrane more proximate to the IS. The RGS9 knock-out mouse is a well characterized model with a slower GTP hydrolysis on Gt␣ and slowed recovery of photoresponses (11). Equivalent levels of illumination are expected to produce a larger steady-state concentration of Gt␣GTP in the RGS9 Ϫ/Ϫ mice than in control mice and lead to a greater translocation of Gt␣ according to the diffusion mechanism. The analysis of Gt␣ redistribution in the RGS9 Ϫ/Ϫ mice demonstrated that the Gt␣ translocation does not require RGS9, and the illumination threshold for this translocation is indeed lower. The redistribution of Gt from the IS/ST to the OS during dark adaptation was unaffected in the RGS9 Ϫ/Ϫ mice. This observation is in accord with the cellular localization of RGS9-1 exclusively in the OS. Absent of the RGS9 GAP complex, Gt␣GTP inactivates in the IS/ST and forms a heterotrimer with Gt␤␥ with the same rate in wild type and RGS9 Ϫ/Ϫ mice. Supporting the notion that Gt␣ and Gt␤␥ translocate to the OS in the dark as a heterotrimer, the time courses of Gt␣ and Gt␤␥ return to the outer segment are similar (20), whereas in the dark-or light-adapted Gt␣ knock-out mice, Gt␤␥ is distributed throughout the photoreceptor cells (37).
Gt␣ moves against the concentration gradient when the translocation to the IS is nearly complete (20). This would indicate a presence of Gt␣ binding sites in the inner segments or an involvement of active transport of some sorts. A potential binding partner for Gt␣GDP, a G protein modulator LGN, has been identified in the IS of photoreceptor cells (39,40). Alternatively, Gt␤␥ itself might be a binding "sink" for Gt␣. The time course of Gt␣␤␥ return to the OS during dark adaptation is much slower than the light-induced translocation of Gt␣ and Gt␤␥ (20). A slower transport of Gt␣␤␥ to the OS may be because of the greater size and lipophilic nature of the heterotrimer and may lead to its accumulation in the IS/ST. Interesting insights into the mechanism of Gt translocation to the IS/ST were revealed from the GTPase-deficient Gt␣EEQ200L mutant mouse model. In the light-adapted transgenic mice, Gt␣EEQ200L accumulates in the IS/ST similarly to Gt␣ in control mice. Yet in the light-exposed retinal samples, this mutant is GTPbound. This observation suggests that Gt␣ may concentrate in the IS independently of the complex formation with Gt␤␥ and indicates a possible binding partner for Gt␣GTP. Although the GTP hydrolysis may not be essential for the accumulation of Gt␣ in the IS/ST, the Gt␣EEQ200L model demonstrates that it is necessary for the relocalization of Gt to the OS during dark adaptation. The return of Gt␣EEQ200L to the OS during dark adaptation of the transgenic mice

Light-dependent Movement of Transducin in Rods
DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 was markedly slower. This deficient relocalization of Gt␣EEQ200L was because of slow GTP hydrolysis by the mutant rather than because of potentially compromised interaction with Gt␤␥. The complex of Gt␣ with Gt␤␥ is required for Gt activation by R*. Because Gt␣EEQ200L readily becomes activated in the light-exposed retinal preparations from the dark-adapted transgenic mice, the mutant Gt␣ is competent of interaction with Gt␤␥. A fraction of Gt␣Q200L remains in the GTPbound state after prolonged dark adaptation, providing an explanation for the residual presence of the mutant Gt␣ and Gt␤␥ in the IS/ST of transgenic rods.
The analysis of the transgenic Gt␣EEQ200L mice revealed a dramatic down-regulation of the PDE6 catalytic subunits. This finding is in agreement with the previously reported reduction in PDE6␣␤ levels by expression of the GTPase-deficient cone Gt␣ in mouse rods (34). Moreover, Gt␣EEQ200L mice are highly insensitive to light, essentially lacking the rod ERG responses. Therefore, the light-dependent translocation of the native and mutant Gt␣ in Gt␣EEQ200L mice suggests that the movement of transducin does not require its interaction with PDE6 and the downstream signaling events. Interestingly, the P␥ levels are unchanged in Gt␣EEQ200L mice. In the absence of an equivalent concentration of PDE6␣␤, P␥ is concentrated in the IS/ST of light-adapted Q200L rods, apparently in the complex with the mutant Gt␣. P␥ partially diffuses to the OS during dark adaptation of Q200L mice. However, in the WT and RGS9 Ϫ/Ϫ mice with normal levels of PDE6␣␤, P␥ constantly resides in the OS and does not translocate with activated Gt␣ to the IS/ST.
The analysis of the RGS9 Ϫ/Ϫ and Gt␣Q200L mice suggests that transducin translocation is controlled by the nucleotide-bound state of Gt␣. Reduced Gt␣ GTPase activity in the OS facilitates the light-induced translocation of Gt␣GTP to the IS/ST. In a Gt␣ mutant with severely impaired GTPase activity, such as Gt␣Q200L, the GTP hydrolysis becomes a rate-limiting step in the relocalization of transducin from the IS/ST to the OS during dark adaptation.