Topography of the rhodopsin molecule. Identification of the domain phosphorylated.

Studies on the light-stimulated phosphorylation of rod outer segments by [gamma-32P]ATP showed that although nearly 1 mol of [32P]phosphate was incorporated/mol of total opsin, only a small fraction of the molecules were phosphorylated, and these contained at least 2-3 mol of phosphate/mol. Rod outer segments containing the phosphorylated opsin were incubated with 11-cis-retinal to generate phosphorylated rhodopsin and then digested with papain to produce a cleaved complex comprising three fragments, heavy (H), medium (M) and light (L). It was shown that L-fragment of apparent mol.wt. 6000 contained all the phosphorylation sites. This suggests that one specific domain of rhodopsin is susceptible to multiple phosphorylation.     

Differential scanning calorimetry of bovine rhodopsin in rod-outer-segment disk membranes.

Rhodopsin-containing retinal rod disk membranes from cattle have been examined by differential scanning calorimetry. Under conditions of 67 mM phosphate pH 7.0, unbleached rod outer segment disk membranes gave a single major endotherm with a temperature of denaturation (Tm) of 71.9 +/- 0.4 degrees C and a thermal unfolding calorimetric enthalpy change (delta Hcal) of 700 +/- 17 kJ/mol rhodopsin. Bleached rod outer segment disk membranes (membranes that had lost their absorbance at 498 nm after exposure to orange light) gave a single major endotherm with a Tm of 55.9 +/- 0.3 degrees C and a delta Hcal of 520 +/- 17 kJ/mol opsin. Neither bleached nor unbleached rod outer segment disk membranes gave endotherms upon thermal rescans. When thermal stability is examined over the pH range of 4-9, the major endotherms of both bleached and unbleached rod outer segment disk membranes were found to show maximum stability at pH 6.1. The observed delta Hcal values for bleached and unbleached rod outer segment disk membranes exhibit membrane concentration dependences which plateau at protein concentrations beyond 1.5 mg/mL. For partially bleached samples of rod outer segment disk membranes, the calorimetric enthalpy change for opsin appears to be somewhat dependent on the degree of bleaching, indicating intramembrane nearest neighbor interactions which affect the unfolding of opsin. Delta Hcal and Tm are particularly useful for assessing stability and testing for completeness of regeneration of rhodopsin from opsin. Other factors such as sample preparation and the presence of low concentrations of ethanol also affect the delta Hcal values while the Tm values remain fairly constant. This shows that the delta Hcal is a sensitive parameter for monitoring environmental changes of rhodopsin and opsin.     

Light activation of one rhodopsin molecule causes the phosphorylation of hundreds of others. A reaction observed in electropermeabilized frog rod outer segments exposed to dim illumination

A rhodopsin phosphorylation reaction that occurs with high-gain is observed if measurements are made in electropermeabilized frog rod outer segments (ROS) stimulated by a dim flash of light in the operating range of the photoreceptor. Flashes of light exciting 1000 or fewer of the 3 x 10(9) rhodopsins present/ROS results in the incorporation of 1400 phosphates from ATP into the rhodopsin pool for each excited rhodopsin (Rho*). This amplification decreases with increasing light intensity, falling most sharply after each disk has absorbed one photon. The high-gain reaction is lost if the ROS are broken into vesicles by shearing, leaving a low-gain rhodopsin phosphorylation characterized in previous studies using brighter illumination. The high-gain but not the low-gain phosphorylation appears to be regulated by G-protein and by calcium levels in the range over which intracellular calcium changes when rod photoreceptors are illuminated. Kinetic measurements made on the phosphorylation observed at higher light intensities shows that it initially occurs rapidly enough for a role in terminating the photoresponse. The high-gain phosphorylation observed at lower light intensities may play a global role in regulating light-adaptation of the rod photoreceptor, and its existence suggests that a search for a similar high-gain modification in systems using the homologous beta-adrenergic or muscarinic acetylcholine receptors might be rewarding.     

A link between rhodopsin and disc membrane cyclic nucleotide phosphodiesterase. Action spectrum and sensitivity to illumination.

Frog (Rana pipiens) rod outer segment disc membranes contain guanosine 3',5'-cyclic monophosphate phosphodiesterase (EC 3.1.4.1.c) which, in the presence of ATP, is stimulated 5- to 20-fold by illumination. The effectiveness of monochromatic light of different wavelengths in activating phosphodiesterase was examined. The action spectrum has a maximum of 500 nm, and the entire spectrum from 350 to 800 nm closely matches the absorption spectrum of rhodopsin, which is apparently the pigment which mediates the effects of light on phosphodiesterase activity. trans-Retinal alone does not mimic light. Half-maximal activation of the phosphodiesterase occurs with a light exposure which bleaches 1/2000 of the rhodopsins. Half-maximal activation can also be achieved by mixing 1 part of illuminated disc membranes in which the rhodopsin is bleached with 99 parts of unilluminated membranes. Regeneration of bleached rhodopsin by addition of 11-cis-retinal is illuminated disc membranes reverses the ability of these membranes to activate phosphodiesterase in unilluminated membranes. If the rhodopsin regenerated by 11-cis-retinal is illuminated again, it regains the ability to activate phosphodiesterase. These studies show that the levels of cyclic nucleotides in vetebrate rod outer segments are regulated by minute amounts of light and clearly indicate that rhodopsin is the photopigment whose state of illumination is closely linked to the enzymatic activity of disc membrane phosphodiesterase.     

On a soluble system for studying light activation of rod outer segment cyclic GMP phosphodiesterase

Bovine rod outer segment (ROS) cyclic GMP phosphodiesterase (PDE) could be activated about 6-fold by light, an effect that could be simulated by isolated bleached rhodopsin. About 90% of PDE activity in ROS could be extracted with 10 mM Tris-HCl, pH 7.5, but light is ineffective in activating the soluble enzyme. However, bleached rhodopsin could activate it in the presence of a very low concentration of ATP, strongly suggesting the mediation of rhodopsin in the light activation of the enzyme in ROS. Direct evidence is presented to suggest that the phosphorylation of opsin (bleached rhodopsin) is unrelated to the activation of PDE by bleached rhodopsin and ATP. The reconstitution of the light activation of PDE in a soluble system presented here opens up a new direction to future investigations on the mechanism of light regulation of cyclic GMP levels in retina and its implication in the photoreceptor function.     

Kinetic study of transfer of 11-cis-retinal between rod outer segment membranes using regeneration of rhodopsin

The present study demonstrates some important facts on the regeneration of rhodopsin in rod outer segment membranes. 11-cis-Retinal added to a rod outer segment membrane suspension did not react directly with opsin but was rapidly solubilized into membranes and then recombined with opsin in the membrane. It was also revealed that the regeneration of rhodopsin was perturbed by the formation of retinylidene Schiff base with phosphatidylethanolamine in rod outer segment membranes, which decreased with increasing temperature. The activation energy of rhodopsin regeneration in rod outer segment membranes was 18.7 kcal/mol, being smaller than the value of 22 kcal/mol in 1% digitonin solution. 11-cis-Retinal could be found to transfer relatively fast (tau-1/k(1) R 10(3) s) between rod outer segment membranes by using the regeneration of rhodopsin. It was demonstrated that the kinetic measurement for the transport of membrane-soluble molecules such as retinal between membranes could be perform ed with ease and precisely by the method described in this paper.     

A radioimmunoassay specific for opsin

A radioimmunoassay is developed for bovine opsin using a rabbit antiserum against bovine rod outer segment membranes. The assay is specific for opsin. Rhodopsin, bacteriorhodopsin and hemoglobin do not show cross-reaction. It can be carried out rapidly, has a sensitivity of 0.01 pmol bovine opsin and gives accurate results, even in the presence of a large excess of rhodopsin. Under the conditions described, the assay can be used to measure bovine opsin and rhodopsin in each other's presence by running a sample before and after illumination, with a sensitivity 2000-times higher than with spectrophotometric methods. The opsin content of rather crude preparations such as bovine retina homogenates can be accurately determined. Rabbit and mouse opsin can also be assayed with a reasonable degree of accuracy using the same rabbit antiserum.     

A functional rhodopsin-green fluorescent protein fusion protein localizes correctly in transgenic Xenopus laevis retinal rods and is expressed in a time-dependent pattern

To study rhodopsin biosynthesis and transport in vivo, we engineered a fusion protein (rho-GFP) of bovine rhodopsin (rho) and green fluorescent protein (GFP). rho-GFP expressed in COS-1 cells bound 11-cis retinal, generating a pigment with spectral properties of rhodopsin (A(max) at 500 nm) and GFP (A(max) at 488 nm). rho-GFP activated transducin at 50% of the wild-type activity, whereas phosphorylation of rho-GFP by rhodopsin kinase was 10% of wild-type levels. We expressed rho-GFP in the rod photoreceptors of Xenopus laevis using the X. laevis principal opsin promoter. Like rhodopsin, rho-GFP localized to rod outer segments, indicating that rho-GFP was recognized by membrane transport mechanisms. In contrast, a rho-GFP variant lacking the C-terminal outer segment localization signal distributed to both outer and inner segment membranes. Confocal microscopy of transgenic retinas revealed that transgene expression levels varied between cells, an effect that is probably analogous to position-effect variegation. Furthermore, rho-GFP concentrations varied along the length of individual rods, indicating that expression levels varied within single cells on a daily or hourly basis. These results have implications for transgenic models of retinal degeneration and mechanisms of position-effect variegation and demonstrate the utility of rho-GFP as a probe for rhodopsin transport and temporal regulation of promoter function.     

Light-induced dephosphorylation of a 33K protein in rod outer segments of rat retina

Phosphorylated proteins may play an important role in regulating the metabolism or function of rod photoreceptors. In mammalian retinas, a photoreceptor protein of 33 000 (33K) molecular weight is phosphorylated in a cyclic nucleotide dependent manner in vitro. Since light initiates the activation of a photoreceptor-specific phosphodiesterase and a rapid reduction in guanosine cyclic 3',5'-phosphate concentration, phosphorylation of the 33K protein may be modulated by light in situ. In order to test this possibility, dark-adapted rat retinas were incubated for 30 min in the dark in phosphate-free Kreb's buffer containing [32P]orthophosphate. Following incubation, rod outer segments were detached by shaking, and the 32P-labeled rod outer segment proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, detected by autoradiography, and quantitated by densitometric scanning. The incorporation of radioactivity (32P) into the 33K protein was higher than into any other rod outer segment protein, and the amount of 32P-labeled 33K protein in the detached rod outer segments remained unchanged during 10 additional min of darkness. The addition of isobutylmethylxanthine to the incubation medium enhanced the incorporation of 32P into 33K protein to about 400% of the original level. Exposure of freshly detached rod outer segments to room light for 90 s decreased the amount of labeled 33K protein to 45% of its original level. The dephosphorylation of labeled 33K protein continued, reaching 12% of the original dark value 10 min after the previously illuminated sample was returned to darkness. Light initiated the phosphorylation of rhodopsin, and rhodopsin phosphorylation continued during the postillumination period of darkness.     

11-cis-retinal reduces constitutive opsin phosphorylation and improves quantum catch in retinoid-deficient mouse rod photoreceptors

Rpe65(-/-) mice produce minimal amounts of 11-cis-retinal, the ligand necessary for the formation of photosensitive visual pigments. Therefore, the apoprotein opsin in these animals has not been exposed to its normal ligand. The Rpe65(-/-) mice contain less than 0.1% of wild type levels of rhodopsin. Mass spectrometric analysis of opsin from Rpe65(-/-) mice revealed unusually high levels of phosphorylation in dark-adapted mice but no other structural alterations. Single flash and flicker electroretinograms (ERGs) from 1-month-old animals showed trace rod function but no cone response. B-wave kinetics of the single-flash ERG are comparable with those of dark-adapted wild type mice containing a full compliment of rhodopsin. Application (intraperitoneal injection) of 11-cis-retinal to Rpe65(-/-) mice increased the rod ERG signal, increased levels of rhodopsin, and decreased opsin phosphorylation. Therefore, exogenous 11-cis-retinal improves photoreceptor function by regenerating rhodopsin and removes constitutive opsin phosphorylation. Our results indicate that opsin, which has not been exposed to 11-cis-retinal, does not generate the activity generally associated with the bleached apoprotein.     

A comparison of the efficiency of G protein activation by ligand-free and light-activated forms of rhodopsin

Activation of the photoreceptor G protein transducin (Gt) by opsin, the ligand-free form of rhodopsin, was measured using rod outer segment membranes with densities of opsin and Gt similar to those found in rod cells. When GTPgammaS was used as the activating nucleotide, opsin catalyzed transducin activation with an exponential time course with a rate constant k(act) on the order of 2 x 10(-3)s(-1). Comparison under these conditions to activation by flash-generated metarhodopsin II (MII) revealed that opsin- and R*-catalyzed activation showed similar kinetics when MII was present at a surface density approximately 10(-6) lower than that of opsin. Thus, in contrast to some previous reports, we find that the catalytic potency of opsin is only approximately 10(-6) that of MII. In the presence of residual retinaldehyde-derived species present in membranes treated with hydroxylamine after bleaching, the apparent k(act) observed was much higher than that for opsin, suggesting a possible explanation for previous reports of more efficient activation by opsin. These results are important for considering the possible role of opsin in the diverse phenomena in which it has been suggested to play a key role, such as bleaching desensitization and retinal degeneration induced by continuous light or vitamin A deprivation.     

Calorimetric studies of bovine rod outer segment disk membranes support a monomeric unit for both rhodopsin and opsin

The photoreceptor rhodopsin is a G-protein coupled receptor that has recently been proposed to exist as a dimer or higher order oligomer, in contrast to the previously described monomer, in retinal rod outer segment disk membranes. Rhodopsin exhibits considerably greater thermal stability than opsin (the bleached form of the receptor), which is reflected in an ∼15°C difference in the thermal denaturation temperatures (T_m) of rhodopsin and opsin as measured by differential scanning calorimetry. Here we use differential scanning calorimetry to investigate the effect of partial bleaching of disk membranes on the T_m of rhodopsin and of opsin in native disk membranes, as well as in cross-linked disk membranes in which rhodopsin dimers are known to be present. The T_ms of rhodopsin and opsin are expected to be perturbed if mixed oligomers are present. The T_m remained constant for rhodopsin and opsin in native disks regardless of the level of bleaching. In contrast, the T_m of cross-linked rhodopsin in disk membranes was dependent on the extent of bleaching. The energy of activation for denaturation of rhodopsin and cross-linked rhodopsin was calculated. Cross-linking rhodopsin significantly decreased the energy of activation. We conclude that in native disk membranes, rhodopsin behaves predominantly as a monomer.     

Transient phosphorylation by ATP of a 160 000 dalton protein in rod outer segments of Bufo marinus

Radioactive phosphate was incorporated from [γ-³²P]ATP into a 160 000 dalton protein from preparations of highly purified toad retinal rod outer segment membranes. Maximal incorporation occurred at 1μM ATP, and turnover in the presence of nonradioactive substrate was rapid, showing that the 160 kdalton protein catalyzes ATP hydrolysis. The 160 kdalton intermediate was sensitive to hydroxylamine, suggesting an acyl linkage between the protein and phosphate. Ionic requirements for phosphorylation showed the ATPase is different from other membrane-bound ionic pumps. The phosphorylated intermediate was almost completely suppressed by 20 μM vanadate, and partial suppression occurred at lower concentrations. About one 160 kdalton protein was labelled per 30 000 molecules of rhodopsin. Although [γ-³²P]GTP labeled the protein, the ATPase was far more specific for adenine than guanine nucleotides. The specificity for ATP and sensitivity to vanadate of the intermediate suggest a relation to an ATP-dependent structural change which occurs in stacks of outer segment discs (Thacher, S.M.; (1980) Fed. Proc. 39, 2066).     

Regulation of rhodopsin dephosphorylation by arrestin

We have characterized the opsin phosphatase activities in extracts of rod outer segments and determined their relationship to known protein phosphatases. The opsin phosphatase activity in the extracts was not due to protein phosphatases 1, 2B, or 2C because it was neither stimulated by Mg2+ or Ca2+/calmodulin nor inhibited by protein phosphatase inhibitors-1 or -2. Opsin phosphatase activity in rod outer segment extracts was potently inhibited by okadaic acid (IC50 approximately 10 nM), a preferential inhibitor of protein phosphatase 2A. Moreover, during chromatography on DEAE-Sepharose, the opsin phosphatase activity co-eluted with three peaks of protein phosphatase 2A activity, termed protein phosphatases 2A0, 2A1, and 2A2. The opsin phosphatase activity of each peak was stimulated by polylysine, a known activator of protein phosphatase 2A. Finally, treatment of rod outer segment extracts with 80% ethanol at room temperature converted the activity from a high molecular weight form characteristic of the protein phosphatase 2A0, 2A1, and 2A2 species to a low molecular weight form characteristic of the protein phosphatase 2A catalytic subunit. We conclude that protein phosphatase 2A is likely to be the physiologically relevant rhodopsin phosphatase. The 48-kDa rod outer segment protein arrestin (S-antigen) was found to inhibit the dephosphorylation of freshly photolyzed rhodopsin by protein phosphatase 2A but did not inhibit the dephosphorylation of unbleached rhodopsin. Arrestin has no effect on the dephosphorylation of phorphorylase a, indicating that the effect was substrate-directed. It appears that dephosphorylation of the photoreceptor protein phosphorhodopsin occurs only after decay of the photoactivated protein and that this may be regulated in vivo by arrestin. The binding of arrestin to photolyzed phosphorylated rhodopsin, i.e. the binding of a regulatory protein to a protein phosphatase substrate to form a complex resistant to dephosphorylation represents a novel mechanism for the regulation of protein phosphatase 2A.     

Mechanism and specificity of rhodopsin phosphorylation.

Partial separation of protein kinase activity from rhodopsin in isolated bovine retinal photoreceptor outer segments was accomplished by mild ultrasonic treatment followed by ultracentrifugation. Residual kinase activity in the rhodopsin-rich sediment was destroyed by chemical denaturation which did not affect the spectral properties of the rhodopsin. The retinal outer segment kinase was found to be specific for rhodopsin, since in these preparations it alone of several bovine protein kinases was capable of phosphorylating rhodopsin in the light. The phosphorylation reaction apparently requires a specific conformation of the rhodopsin molecule since it is abolished by heat denaturation of rhodopsin, and it is greatly reduced or abolished by treatment of the visual pigment protein with potassium alum after the rhodopsin has been "bleached" by light. When kinase and rhodopsin or opsin fractions were prepared from dark-adapted and bleached outer segments and the resultant fractions were mixed in various combinations of bleached and unbleached preparations, the observed pattern of light-activated phosphorylation was consistent only with the interpretation that a conformational change in the rhodopsin molecule in the light exposes a site on the visual pigment protein to the kinase and ATP. These results rule out the possibility of a direct or indirect (rhodopsin-mediated) light activation of the kinase. Finally, phosphorylation of retinal outer segment protein in monochromatic lights of various wavelengths followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis indicates that both rhodopsin and the higher molecular weight visual pigment protein reported by several laboratories have the same action spectrum for phosphorylation. This result is consistent with the suggestion that the higher molecular weight species is a rhodopsin dimer.     

Influence of light and calcium on guanosine 5'-triphosphate in isolated frog rod outer segments.

Frog rod outer segments contain approximately 0.25 mol of GTP and 0.25 mol of ATP per mol of rhodopsin 3 min after their isolation from the retina. UTP and CTP are present at 10-fold and 100-fold lower levels, respectively. Concentrations of GTP and ATP decline in parallel over the next 4 min to reach relatively stable levels of 0.1 mol per mol of rhodopsin. Illumination reduces the concentration of endogenous GTP but not ATP. This light-induced decrease in GTP can be as large as 70% and has a half-time of 7 s. GTP is reduced to steady intermediate levels during extended illumination of intermediate intensity, but partially returns to its dark-adapted level after brief illumination. The magnitude of the decrease increases as a linear function of the logarithm of continuous light intensity at levels which bleach between 5 X 10(2) and 5 X 10(6) rhodopsin molecules/outer segment per second. This exceeds the range of intensities over which illumination causes decreases in the cyclic GMP content and permeability of isolated outer segments (Woodruff and Bownds. 1979. J. Gen. Physiol. 73:629-653). Thus, over 4 log units of light intensity, a sensitivity control mechanism functions to make extended illumination less effective in stimulating a GTP decrease. GTP levels in dark-adapted outer segments are sensitive to changes in calcium concentration in the suspending medium. If the external calcium concentration is reduced to 10(-8) M, GTP concentration is lowered to the same level caused by saturating illumination, and the GTP remaining is no longer light-sensitive. Lowering calcium concentration to intermediate levels between 10(-6) and 10(-8) M reduces GTP to stable intermediate levels, and the GTP remaining can be reduced by light. Restoration of millimolar calcium drives synthesis of GTP, but not of ATP, and GTP lability towards illumination is again observed. These calcium-induced changes in GTP are diminished by the addition of the divalent cation ionophore A23187. Lowering or raising magnesium levels does not influence the GTP concentration. These data raise the possibility that light activates either a calcium transport mechanism driven by the hydrolysis of GTP, or some other calcium-sensitive GTPase activity of unknown function. Known light-dependent reactions involving cyclic nucleotide transformations and rhodopsin phosphorylation appear to account for only a small fraction of the light-induced GTP decrease.     

Light-stable rhodopsin. I. A rhodopsin analog reconstituted with a nonisomerizable 11-cis retinal derivative.

With the aim of preparing a light-stable rhodopsin-like pigment, an analog, II, of 11-cis retinal was synthesized in which isomerization of the C11-C12 cis-double bond is blocked by a cyclohexene ring built around the C10 to C13-methyl. The analog II formed a rhodopsin-like pigment (rhodopsin-II) with opsin expressed in COS-1 cells and with opsin from rod outer segments. The rate of rhodopsin-II formation from II and opsin was approximately 10 times slower than that of rhodopsin from 11-cis retinal and opsin. After solubilization in dodecyl maltoside and immunoaffinity purification, rhodopsin-II displayed an absorbance ratio (A280nm/A512nm) of 1.6, virtually identical with that of rhodopsin. Acid denaturation of rhodopsin-II formed a chromophore with lambda max, 452 nm, characteristic of protonated retinyl Schiff base. The ground state properties of rhodopsin-II were similar to those of rhodopsin in extinction coefficient (41,200 M-1 cm-1) and opsin-shift (2600 cm-1). Rhodopsin-II was stable to hydroxylamine in the dark, while light-dependent bleaching by hydroxylamine was slowed by approximately 2 orders of magnitude relative to rhodopsin. Illumination of rhodopsin-II for 10 s caused approximately 3 nm blue-shift and 3% loss of visible absorbance. Prolonged illumination caused a maximal blue-shift up to approximately 20 nm and approximately 40% loss of visible absorbance. An apparent photochemical steady state was reached after 12 min of illumination. Subsequent acid denaturation indicated that the retinyl Schiff base linkage was intact. A red-shift (approximately 12 nm) in lambda max and a 45% recovery of visible absorbance was observed after returning the 12-min illuminated pigment to darkness. Rhodopsin-II showed marginal light-dependent transducin activation and phosphorylation by rhodopsin kinase.     

Recoverin and rhodopsin kinase activity in detergent-resistant membrane rafts from rod outer segments

Cholesterol-rich membranes or detergent-resistant membranes (DRMs) have recently been isolated from bovine rod outer segments and were shown to contain several signaling proteins such as, for example, transducin and its effector, cGMP-phosphodiesterase PDE6. Here we report the presence of rhodopsin kinase and recoverin in DRMs that were isolated in either light or dark conditions at high and low Ca2+ concentrations. Inhibition of rhodopsin kinase activity by recoverin was more effective in DRMs than in the initial rod outer segment membranes. Furthermore, the Ca2+ sensitivity of rhodopsin kinase inhibition in DRMs was shifted to lower free Ca2+ concentration in comparison with the initial rod outer segment membranes (IC50=0.76 microm in DRMs and 1.91 microm in rod outer segments). We relate this effect to the high cholesterol content of DRMs because manipulating the cholesterol content of rod outer segment membranes by methyl-beta-cyclodextrin yielded a similar shift of the Ca2+-dependent dose-response curve of rhodopsin kinase inhibition. Furthermore, a high cholesterol content in the membranes also increased the ratio of the membrane-bound form of recoverin to its cytoplasmic free form. These data suggest that the Ca2+-dependent feedback loop that involves recoverin is spatially heterogeneous in the rod cell.     

Characterization of transducin from bovine retinal rod outer segments. Mechanism and effects of cholera toxin-catalyzed ADP-ribosylation

Transducin, a guanine nucleotide-binding protein consisting of two subunits (T alpha and T beta gamma), mediates the signal coupling between rhodopsin and a membrane-bound cyclic GMP phosphodiesterase in retinal rod outer segments. The T alpha subunit is an activator of the phosphodiesterase, and the function of the T beta gamma subunit is to physically link T alpha with photolyzed rhodopsin. In this study, the mechanism of cholera toxin-catalyzed ADP-ribosylation of T alpha has been examined in a reconstituted system consisting of purified transducin and stripped rod outer segment membranes. Limited proteolysis of the labeled T alpha with trypsin indicated that the inserted ADP-ribose is located exclusively on a single proteolytic fragment with an apparent molecular weight of 23,000. Maximal incorporation of ADP-ribose was achieved when guanosine 5'-(beta, gamma-imido)triphosphate (Gpp(NH)p) and T beta gamma were present at concentrations equal to that of T alpha and when rhodopsin was continuously irradiated with visible light in the 400-500 nm region. The stimulating effect of illumination was related to the direct interaction of the retinal chromophore with opsin. These findings strongly suggest that a transient protein complex consisting of T alpha X Gpp(NH)p, T beta gamma, and a photointermediate of rhodopsin is the required substrate for cholera toxin. Single turnover kinetic measurements demonstrated that the ADP-ribosylation of T alpha coincided with the appearance of a population of transducin molecules having a very slow rate of GTP hydrolysis. The hydrolysis rate of the bound GTP for this population was 1.1 X 10(-3)/s, which was 22-fold slower than the rate for the unmodified transducin.