Inhibition of Rhodopsin Kinase by Recoverin

Recoverin is a 23-kDa Ca2+-binding protein found predominantly in vertebrate photoreceptor cells. Recent electrophysiological and biochemical studies suggest that recoverin may regulate the photoresponse by inhibiting rhodopsin phosphorylation. We find in both cell homogenates and reconstituted systems that the inhibition of rhodopsin phosphorylation by recoverin occurs over a significantly higher free Ca2+ range than previously reported. Half-maximal inhibition occurs at 1.5-3 μM free Ca2+ and is cooperative with a Hill coefficient of ∼2. Measurements of transducin activation demonstrate that this inhibition prolongs the lifetime of catalytically active rhodopsin. Ca2+-recoverin directly inhibits rhodopsin kinase activity, and Ca2+-dependent binding of recoverin to rod outer segment membranes is not required for its action. Extrapolation of the in vitro data to in vivo conditions based on simple mass action calculations places the Ca2+-recoverin regulation within the physiological free Ca2+ range in intact rod outer segment. The data are consistent with a model in which the fall in free Ca2+ that accompanies rod excitation exerts negative feedback by relieving inhibition of rhodopsin phosphorylation. Recoverin is a 23-kDa Ca2+-binding protein found predominantly in vertebrate photoreceptor cells. Recent electrophysiological and biochemical studies suggest that recoverin may regulate the photoresponse by inhibiting rhodopsin phosphorylation. We find in both cell homogenates and reconstituted systems that the inhibition of rhodopsin phosphorylation by recoverin occurs over a significantly higher free Ca2+ range than previously reported. Half-maximal inhibition occurs at 1.5-3 μM free Ca2+ and is cooperative with a Hill coefficient of ∼2. Measurements of transducin activation demonstrate that this inhibition prolongs the lifetime of catalytically active rhodopsin. Ca2+-recoverin directly inhibits rhodopsin kinase activity, and Ca2+-dependent binding of recoverin to rod outer segment membranes is not required for its action. Extrapolation of the in vitro data to in vivo conditions based on simple mass action calculations places the Ca2+-recoverin regulation within the physiological free Ca2+ range in intact rod outer segment. The data are consistent with a model in which the fall in free Ca2+ that accompanies rod excitation exerts negative feedback by relieving inhibition of rhodopsin phosphorylation. A drop in free Ca2+ following illumination appears to be a key regulatory step during recovery of the photoresponse and during light adaptation (see Refs. 1McNaughton P.A. Physiol. Rev. 1990; 70: 847-883Crossref PubMed Scopus (131) Google Scholar and 2Miller J.L. Picones A.P. Korenbrot J.I. Curr. Opin. Neurobiol. 1994; 4: 488-495Crossref PubMed Scopus (44) Google Scholar for review). It has been shown to regulate a variety of enzymes involved in the phototransduction cascade (3Koch K.W. Stryer L. Nature. 1988; 334: 64-66Crossref PubMed Scopus (471) Google Scholar, 4Hsu Y.-T. Molday R.S. Nature. 1993; 361: 76-79Crossref PubMed Scopus (310) Google Scholar, 5Lagnado L. Baylor D.A. Nature. 1994; 367: 273-277Crossref PubMed Scopus (108) Google Scholar, 6Kawamura S. Murakami M. Nature. 1991; 349: 420-423Crossref PubMed Scopus (185) Google Scholar) that include guanylate cyclase, the cGMP-gated channel, rhodopsin, and cGMP phosphodiesterase. Following Kawamura(7Kawamura S. Nature. 1993; 362: 855-857Crossref PubMed Scopus (323) Google Scholar), several studies have suggested that the Ca2+ effect on cGMP phosphodiesterase might be mediated by a Ca2+-binding protein, recoverin, through its inhibition of rhodopsin phosphorylation (8Klenchin V.A. Calvert P.D. Bownds M.D. Biophys. J. 1994; 66 (abstr.): A48Google Scholar, 9Chen C.-K. Hurley J.B. Invest. Ophthalmol. 35 (abstr.): 1485Google Scholar, 10Gorodovikova E.N. Senin I.I. Philippov P.P. FEBS Lett. 1994; 353: 171-172Crossref PubMed Scopus (54) Google Scholar) (recoverin is called S-modulin in the frog; for simplicity, we refer to it as frog recoverin). Consistent with this inhibition of rhodopsin phosphorylation, physiological experiments have shown that recoverin and its homologues can slow photoresponse recovery(11Gray-Keller M.P. Polans A.S. Palczewski K. Detwiler P.B. Neuron. 1993; 10: 523-531Abstract Full Text PDF PubMed Scopus (138) Google Scholar). Recent experiments on transgenic mice lacking recoverin also seem to support this hypothesis(12Dodd R.L. Makino C.L. Chen J. Simon M.I. Baylor D.A. Invest. Ophthalmol. 36 (abstr.): S641Google Scholar). Recoverin is expressed only in the retina and, with the exception of a subset of bipolar cells, is specific to photoreceptor cells(13Milam A.H. Dacey D.M. Dizhoor A.M. Visual Neurosci. 1993; 10: 1-12Crossref PubMed Scopus (235) Google Scholar). It is a 23-kDa protein that contains several EF-hand motifs characteristic of Ca2+-binding proteins (14Ray S. Zozulya S. Niemi G.A. Flaherty K.M. Brolley D. Dizhoor A.M. McKay D.B. Hurley J. Stryer L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 5705-5709Crossref PubMed Scopus (97) Google Scholar, 15Flaherty K.M. Zozulya S. Stryer L. McKay D.B. Cell. 1993; 75: 709-716Abstract Full Text PDF PubMed Scopus (225) Google Scholar) and is heterogenously fatty acid acylated at its N terminus(16Dizhoor A.M. Ericsson L.H. Johnson R.S. Kumar S. Olshevskaya E. Zozulya S. Neubert T.A. Stryer L. Hurley J.B. Walsh K.A. J. Biol. Chem. 1992; 267: 16033-16036Abstract Full Text PDF PubMed Google Scholar). This modification allows recoverin to bind membranes upon binding Ca2+, suggesting that recoverin's biological activity may be related to its Ca2+-dependent membrane binding(17Zozulya S. Stryer L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 11569-11573Crossref PubMed Scopus (294) Google Scholar, 18Dizhoor A.M. Chen C.-K. Olshevskaya E. Sinelnikova V.V. Phillippov P. Hurley J.B. Science. 1993; 259: 829-832Crossref PubMed Scopus (169) Google Scholar). This paper addresses several questions that are important in establishing the physiological relevance of the Ca2+-recoverin system in ROS.1( 1The abbreviations used are: ROSrod outer segment(s)RKrhodopsin kinaseGTPγSguanosine 5′-O-(3-thiotriphosphate)BAPTA1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid. )1) What is the free Ca2+ concentration range over which recoverin inhibition of rhodopsin phosphorylation is relaxed? 2) What is the mechanism of recoverin inhibition of rhodopsin phosphorylation? 3) Do the in vitro biochemical data support the proposed physiological role of recoverin in living photoreceptors? The data show that Ca2+-recoverin acts directly on RK to decrease its catalytic activity and that no components other than rhodopsin, kinase, and recoverin are required. Extrapolation of the recoverin inhibition of RK observed in vitro at micromolar free Ca2+ to conditions in the intact ROS suggests a role for the Ca2+-recoverin system in normal ROS function. rod outer segment(s) rhodopsin kinase guanosine 5′-O-(3-thiotriphosphate) 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid. γ-32PATP and γ-35SGTP were purchased from DuPont NEN; Percoll and Heparin HiTrap columns were from Pharmacia Biotech Inc., potassium isethionate was from Kodak; Chelex 100 resin was from Bio-Rad; 1 M CaCl2 solution was from BDH; BAPTA and Fluo-3 were from Molecular Probes; and bis-Tris propane was from Calbiochem. Other chemicals were obtained from Sigma. The standard buffer used in all experiments contained 105 mM potassium isethionate, 5 mM sodium isethionate, 10 mM HEPES, 2 mM MgCl2, pH 7.8. It was purified from calcium ion contamination on a Chelex 100 column prior to addition of MgCl2 so that Ca2+ concentration was below 1 μM. Intact frog ROS were purified as described(19Biernbaum M.S. Bownds M.D. J. Gen. Physiol. 1985; 85: 83-105Crossref PubMed Scopus (51) Google Scholar), except that CaCl2 concentration in Percoll was decreased to 0.1 mM. ROS were resuspended in the standard buffer and homogenized with a motorized tissue grinder to minimize any diffusional limitations caused by disk stacks(20Dumke C.L. Arshavsky V.Y. Calvert P.D. Bownds M.D. Pugh Jr., E.D. J. Gen. Physiol. 1994; 103: 1071-1098Crossref PubMed Scopus (50) Google Scholar). Bovine ROS were purified under infrared illumination as described(21McDowell J.H. Hargrave P.A. Methods in Neurosciences: Photoreceptor Cells. Academic Press, San Diego, CA1993: 123-130Google Scholar). The same method of bovine ROS isolation was used for RK purification, but all sucrose solutions were prepared in 10 mM Tris, 5 mM MgCl2, pH 7.5, and the procedure was performed in room light. Bovine and frog recoverin were purified (22Polans A.S. Crabb J. Palczewski K. Hargrave P.A. Methods in Neurosciences: Photoreceptor Cells. Academic Press, San Diego, CA1993: 248-260Google Scholar) and stored at −70°C. The concentration of recoverin was determined by absorbance at 280 nm using a molar extinction coefficient of 36,400.2( 2A. Polans, personal communication. )RK was extracted as described (23Palczewski K. Buczylko J. Van Hooser P. Carr S.A. Huddleston M.J. Crabb J.W. J. Biol. Chem. 1992; 267: 18991-18998Abstract Full Text PDF PubMed Google Scholar) and purified based on published procedures(24Kelleher D.J. Johnson G.L. J. Biol. Chem. 1990; 265: 2632-2639Abstract Full Text PDF PubMed Google Scholar, 25Buczylko J. Gutmann C. Palczewski K. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 2568-2572Crossref PubMed Scopus (66) Google Scholar). Briefly, extracted RK was dialyzed against 10 mM Tris, 0.4% Tween 80, pH 8.0, and loaded on a 1 × 3.5-cm DEAE-cellulose column at 0.3 ml/min. The column was washed with 300 ml of buffer A (20 mM Tris, 0.2% Tween 80, pH 8.0) and then with 100 ml of 35 mM NaCl in buffer A, and RK was eluted with a 35-135 mM NaCl gradient (0.2 ml/min; total volume, 45 ml). Fractions containing RK were loaded on a 1-ml Heparin HiTrap column equilibrated with buffer B (10 mM bis-Tris propane, 0.064% Tween 80, 2 mM MgCl2, pH 7.8). The column was washed with 125 mM KCl in buffer B, and RK was eluted with 250 mM KCl at 0.06 ml/min. A set of 4× stock solutions with different CaCl2 concentrations and fixed BAPTA concentration in standard buffer was prepared (1× = 5 mM BAPTA). Free Ca2+ in 4-fold diluted solutions was measured. For the free Ca2+ range of 10 nM to 5 μM, Fluo-3 dye was used. The Kd of Fluo-3 for Ca2+, 450 nM, was derived from its fluorescence in solutions of 10-70 nM free Ca2+ buffered with BAPTA according to the estimates of the program BAD(26Brooks S.P.J. Storey K.B. Anal. Biochem. 1992; 201: 119-126Crossref PubMed Scopus (324) Google Scholar). Preparation of accurate BAPTA solutions required gravimetric determination of the water content (∼10%) of the commercially obtained BAPTA. For free Ca2+ that was ≥10 μM, a Ca2+-selective electrode (Microelectrodes, Inc., Londonderry, NH) was used following recommendations of the manufacturer. A control determination showed that frog ROS suspensions containing up to 20 μM rhodopsin do not change free Ca2+. Samples of intact frog ROS with different rhodopsin concentrations (19Biernbaum M.S. Bownds M.D. J. Gen. Physiol. 1985; 85: 83-105Crossref PubMed Scopus (51) Google Scholar) were mixed with SDS-polyacrylamide gel electrophoresis loading buffer, heated for 30 min at 60°C, and run on a 15% polyacrylamide gel in parallel with purified recoverin standards. The intensity of recoverin bands was determined by densitometric analysis with the aid of a Foto/analyst system and Collage software (Fotodyne, New Berlin, WI) using laplacian edge detection and local background subtraction. Calculations assumed molecular masses of 23 and 39 kDa for recoverin and rhodopsin, respectively. Frog RK activity was measured in ROS suspensions of 20 μM rhodopsin unless otherwise indicated. Bovine RK activity was measured using urea-treated ROS membranes as a substrate (final concentration, 10 μM rhodopsin). ROS, a CaCl2/BAPTA stock solution, and recoverin if necessary were mixed in a final volume of 15 μl in the dark, the suspension was illuminated (bleached), and 5 μl of γ-32PATP (0.4-0.8 mM) was added (control experiments have shown that the order in which reactants are mixed is not important; the same extent of RK inhibition is obtained if the reaction is initiated by bleaching of rhodopsin in the presence of ATP). After 1-2 min, the reaction was stopped by the addition of μl of acid 100 mM 100 mM pH was measured using of the following For of rhodopsin, the rhodopsin bands from SDS-polyacrylamide gel electrophoresis were in For of were through washed with × 1 ml of 100 mM sodium pH 7.5, and in 2 ml of acid. The were by The of binding following a light was measured by binding under the same conditions as used for rhodopsin phosphorylation with the exception that contained 100 μM γ-35SGTP and The reaction was to for 5 min and with 100 mM 100 mM pH 7.8. used the in The data from experiments shown in and were to a (see and then by the of rhodopsin in order to data from experiments that showed on a coefficient = for data set was from the which were then by the at 0.1 of The data from experiments shown in were to the is the rhodopsin phosphorylation and is concentration of recoverin and then by the of data from different experiments were and were performed data shown are the of at of recoverin in frog ROS and inhibition of rhodopsin phosphorylation as a of recoverin A, determination of the in intact frog recoverin of recoverin and ROS are shown data from 5 The data are consistent with a of a gel used for recoverin to and of recoverin and ROS and of rhodopsin). The bands in the ROS were as recoverin based on and the that run at the same as purified bovine recoverin the ROS is by The in of ROS is caused by gel with to rhodopsin. B, of the inhibition of rhodopsin phosphorylation on recoverin concentration in 10 mM rhodopsin frog ROS at μM and 2 mM free Ca2+. The is according to the with = and = The of rhodopsin phosphorylation on free Ca2+ for frog ROS in the and presence of recoverin is shown in A of the free Ca2+ is This is by that inhibition of rhodopsin phosphorylation occurs at micromolar and free Ca2+ The data can be to the 2) free Ca2+ concentration and is the of rhodopsin by rhodopsin phosphorylation in this a of Ca2+ with a free Ca2+ concentration and a Hill coefficient of and Ca2+ with and 2) that in a Ca2+ 3) that the The that a of Ca2+ with recoverin concentration, other do not change The Ca2+ inhibition in the range of mM free Ca2+ and is to be of physiological The shown below (see that it not recoverin and is of inhibition of purified RK by Ca2+ was previously K. J.H. Hargrave P.A. 1988; PubMed Scopus Google Scholar). We that the inhibition of rhodopsin phosphorylation at micromolar free Ca2+ is and refer to it as the the in 1 are to V.A. Calvert P.D. Bownds M.D. Biophys. J. 1994; 66 (abstr.): A48Google Scholar, 10Gorodovikova E.N. Senin I.I. Philippov P.P. FEBS Lett. 1994; 353: 171-172Crossref PubMed Scopus (54) Google Scholar, S. P. Biochem. Biophys. 1994; PubMed Scopus Google Scholar). The important is that we find a higher free Ca2+ for the recoverin effect 2 a free Ca2+ that is order of higher than that by other has to this We a for the but that it is to several in calcium with A that is is that the of for Ca2+ and its pH in in free Ca2+ if in the concentrations of Ca2+ are Refs. D.J. G.L. J. Physiol. PubMed Google Scholar and D. A. J. Physiol. 1990; Scopus Google Scholar). and of calcium we have directly measured free Ca2+ in solutions used for the experiments (see and that We have determined the of recoverin in intact ROS and found that its molar to rhodopsin is The inhibition of rhodopsin phosphorylation in frog ROS as a of the total concentration of recoverin at free Ca2+ concentrations can then be described by a simple with at 0.3 μM recoverin that the effect of recoverin on cGMP phosphodiesterase activity S. Murakami M. Nature. 1991; 349: 420-423Crossref PubMed Scopus (185) Google Scholar, S. Nature. 1993; 362: 855-857Crossref PubMed Scopus (323) Google Scholar) is not a of cGMP phosphodiesterase by Ca2+-recoverin and that the recoverin action prolongs the lifetime of catalytically active rhodopsin, we the recoverin effect be on the of transducin The the lifetime of catalytically active rhodopsin, the the of transducin it is to that the of transducin is not the total of to transducin following a be a of the of rhodopsin that addition of recoverin to ROS in 10 μM free Ca2+ of binding that is consistent with inhibition of rhodopsin phosphorylation. The for the mechanism of recoverin action is inhibition of The Ca2+ of rhodopsin phosphorylation obtained with purified bovine recoverin, and urea-treated ROS is shown in and the protein are shown in It is that all of Ca2+ inhibition of rhodopsin phosphorylation calcium the Ca2+ range for the recoverin and a Hill coefficient to 2) can be reconstituted with purified proteins of any 5 data from experiments several of recoverin inhibition of recoverin might of RK to rhodopsin. For a inhibition the extent of the recoverin effect be at substrate to the that the recoverin effect not on the of rhodopsin Ca2+-dependent membrane binding of recoverin might a role in recoverin function. find only recoverin is to the extent of the recoverin effect in the presence of 10 and 100 μM rhodopsin was the Kd of recoverin for ROS membranes is μM S. Stryer L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 11569-11573Crossref PubMed Scopus (294) Google Scholar), this of membrane concentration to a of the of a was found to Ca2+ in the presence of concentrations of proteins and is the that Ca2+-recoverin for so that the μM used in the experiments was not in the presence of this the recoverin effect with that it is not the the same is obtained with 100 μM and 1 mM of rhodopsin phosphorylation by recoverin can be observed in vitro 1 μM free Ca2+ and the presence of micromolar concentrations of recoverin of the recoverin effect that it a inhibition of RK by Ca2+-recoverin also Refs. 9Chen C.-K. Hurley J.B. Invest. Ophthalmol. 35 (abstr.): 1485Google Scholar and 10Gorodovikova E.N. Senin I.I. Philippov P.P. FEBS Lett. 1994; 353: 171-172Crossref PubMed Scopus (54) Google Scholar). The that the recoverin effect is of the concentration of RK A and suggests that recoverin inhibits the catalytic activity of of suggest that membrane binding of recoverin not its to it has been shown that recoverin that not bind to membranes the same activity as the protein S. P. Biochem. Biophys. 1994; PubMed Scopus Google D. A. and M. D. in the data in with the of recoverin for S. Stryer L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 11569-11573Crossref PubMed Scopus (294) Google Scholar), that membrane binding is not required for recoverin inhibition of data of the of recoverin for Ca2+. that RK not rhodopsin, the extent of the recoverin effect is determined by the of RK that is to This is by 3) is recoverin and and are for the a Hill coefficient to 2 was found and for we binding with binding for Ca2+. analysis of this under conditions of buffered Ca2+ that the total recoverin concentration is higher than the recoverin effect is of RK concentration and 2) at free Ca2+, the recoverin concentration required for the of the recoverin effect is to only to be in order to find a for the A of μM, with a of μM is found to a for the data of μM for is in with of Ca2+ binding to recoverin that with of and and to recoverin that show cooperative Ca2+ binding with a Hill coefficient of and at μM J.B. M. Stryer L. J. Biol. Chem. 1995; Full Text Full Text PDF PubMed Scopus Google Scholar). This analysis also we that previously free Ca2+ for the recoverin effect in vitro are in order to a of μM free Ca2+ for the recoverin effect recoverin is S. Nature. 1993; 362: 855-857Crossref PubMed Scopus (323) Google Scholar, 10Gorodovikova E.N. Senin I.I. Philippov P.P. FEBS Lett. 1994; 353: 171-172Crossref PubMed Scopus (54) Google Scholar, S. P. Biochem. Biophys. 1994; PubMed Scopus Google Scholar), to set at μM, from the determined We have The data seem to recoverin as a at micromolar free Ca2+ the free Ca2+ concentration in ROS is nM and to on J.I. PubMed Scopus Google Scholar, L. L. P.A. J. Physiol. 1992; PubMed Scopus Google Scholar, Biophys. J. 1994; Full Text PDF PubMed Scopus Google Scholar, M.P. Detwiler P.B. Neuron. 1994; Full Text PDF PubMed Scopus (235) Google Scholar). The is by of the data to in vivo which the recoverin effect the physiological free Ca2+ that at μM total recoverin the free Ca2+ = is than the of μM used in the The of a is to the Kd of recoverin for Ca2+ of recoverin is the of RK to Ca2+-recoverin is a of the concentration of in vivo conditions of μM recoverin on mM rhodopsin concentration P.A. of New Inc., New Scholar) and the of that we the free Ca2+ for the recoverin effect is this to be also to the in vivo RK its only the of the inhibition and not the free Ca2+ range over which it The inhibition at μM recoverin, for from to for recoverin to RK of 1 and respectively. no bands in intensity to recoverin are found in the of kDa the rhodopsin and of the in the molecular of the we the that the free Ca2+ range of recoverin action in vivo is the of Ca2+-recoverin to bind to Zozulya and Stryer S. Stryer L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 11569-11573Crossref PubMed Scopus (294) Google Scholar) at membrane concentrations the Ca2+ of recoverin binding to membranes have a than recoverin's for Ca2+. The same for Ca2+ binding to of Ca2+-recoverin to membrane the of the binding Ca2+ and recoverin the of Ca2+-recoverin = μM and a Ca2+-recoverin Kd for membranes of μM rhodopsin a Ca2+ of recoverin under in vivo conditions of mM rhodopsin and μM recoverin has a of μM free Ca2+ under the membrane concentrations found in ROS, recoverin is to be with Ca2+ at μM, to the of RK not to Ca2+-recoverin in vivo as a of free Ca2+, not the recoverin for Ca2+ as we but its of μM. this in a of the range of the recoverin effect free Ca2+ a of nM free Ca2+, which is within the physiological This to important of recoverin is not required for its inhibition of RK and this under in vivo conditions this modification might the free Ca2+ range of the recoverin by J.B. M. Stryer L. J. Biol. Chem. 1995; Full Text Full Text PDF PubMed Scopus Google Scholar), important of recoverin is that it of Ca2+ binding by recoverin, it a of Ca2+ The of recoverin for Ca2+ has important physiological that the of Ca2+ and Ca2+-binding proteins is 2 × P. Scholar), the micromolar Kd of the Ca2+-recoverin that its is and the lifetime of the is on the order of of This to in free Ca2+ concentration we show the for Ca2+ not recoverin's to free Ca2+ concentrations significantly than that the of recoverin to membrane and concentrations of recoverin are The in this with S. Nature. 1993; 362: 855-857Crossref PubMed Scopus (323) Google Scholar, 9Chen C.-K. Hurley J.B. Invest. Ophthalmol. 35 (abstr.): 1485Google Scholar, 10Gorodovikova E.N. Senin I.I. Philippov P.P. FEBS Lett. 1994; 353: 171-172Crossref PubMed Scopus (54) Google Scholar, M.P. Polans A.S. Palczewski K. Detwiler P.B. Neuron. 1993; 10: 523-531Abstract Full Text PDF PubMed Scopus (138) Google Scholar, R.L. Makino C.L. Chen J. Simon M.I. Baylor D.A. Invest. Ophthalmol. 36 (abstr.): S641Google Scholar), a for the relevance of recoverin in The is not to be but to a solution for the in vitro data and the physiological free Ca2+ it suggests the that the of the Ca2+ feedback on the of rhodopsin phosphorylation might in be by in the of recoverin and the of the We Polans and Palczewski for with and for We also Arshavsky for on the of the