Proc. Nat. Acad. Sci. USA Vol. 72, No. 1, pp. 381-385, January 1975

Light-Stimulated Phosphorylation of in the Retina: The Presence of a Kinase That Is Specific for Photobleached Rhodopsin (protein phosphorylation/cAMP/cGMP) MALCOLM WELLER, NOELLE VIRMAUX, AND P. MANDEL Centre de Neurochimie, 11 Rue Humann, 67085 Strasbourg Cedex, France Communicated by Seymour S. Kety, September 30, 1974

ABSTRACT A protein kinase has been extiacted from Preparation of Tris Extracts from ROS (unless otherwise bovine rod outer segments by a mild procedure. The en- stated all extractions were performed at 40 in dim red light). zyme acts specifically on photobleached, not unbleached, rhodopsin and will not catalyze the phosphorylation of ROS were homogenized by hand at a concentration of about histones, phosvitin, or casein. We propose the name 0.5 mg of protein per ml in 0.01 M Tris*HCl (pH 7.0) and " kinase" for the enzyme, which is not affected by centrifuged at 100,000 X g for 60 min. cyclic nucleotides but which is inhibited by theophylline. Preparations of purified rod outer segments, however, Preparation of "Purified Rhodopsin" by Differential Extrac- appear to contain only low concentration of opsin phos- tion of ROS with Sodium Dodecyl Sulfate. The pellet remaining phatase activity. after extraction with Tris - HCl was extracted twice with 0.1%o It has recently been shown in several laboratories that when sodium dodecyl sulfate in 10 mM Tris*HCl (pH 7.0). The rod outer segments (ROS) prepared from frog (1, 2) or ox insoluble material was washed three times with 0.066 M Na (3-5) retinas are incubated with ATP and Mg2+ rhodopsin is phosphate buffer, pH 7.0. This procedure removes not only phosphorylated and the reaction is markedly stimulated by nonrhodopsin but also bleached rhodopsin. On the light. basis of polyacrylamide gel electrophoresis opsin appears to be There seem to be three ways in which light could act. First, the only protein in the preparation (14). by directly stimulating the activity of the kinase, second, by Protein Determination was by the method of Lowry et al. altering the conformation of the rhodopsin so that it becomes a (15) using bovine plasma albumin as standard. substrate for the kinase, and third, by altering the concentra- tion of some cofactor which changes the activity of the kinase. Radioactive A TP 7y-labeled [32P]ATP was prepared (16) and In support of the third possibility, it has been observed that purified (17) by previously described methods. the rate of production of 3': 5'-cAMP and cGMP in ROS is Measurement of Protein Phosphorylation. All incubations lowered on exposure to light (6-10) and, since rates of protein were carried out in the presence of 50 mM Tris * HCl pH 7.4, 1 phosphorylation are, in so many cases, controlled by cyclic mM [a2PIATP and 1 mM MgCl2 (unless otherwise stated) at nucleotides, it was at first thought that the effect of light on 370 in a volume of 0.5 ml with 100-200 ltg of protein. Reactions cyclic nucleotide concentration could be correlated with the were terminated by the addition of 2 ml of ice-cold 10% effect on protein phosphorylation (3, 4). A decrease in cyclic trichloroacetic acid (or 20% to precipitate histones) and 0.1 nucleotide concentration could increase the net rate of protein mg (0.1 ml) of bovine serum albumin added as protein carrier. phosphorylation if the cyclic nucleotide inhibited a protein The protein precipitate was washed twice with 10% trichloro- kinase or, alternatively, stimulated a protein phosphatase. acetic acid, M orthophosphoric acid, then resuspended in 0.5 Cyclic-AMP-inhibited protein kinases have been discovered ml of 0.1 N NaOH and incubated for 10 min at 37°, after which in the slime mold (11) and cAMP-stimulated protein phos- 2 ml of 10 or 20% trichloroacetic acid was added and the pro- phatases in the toad bladder membrane (12). tein was spun down. The protein precipitates were washed once In this paper, we show that neither cAMP nor cGMP have in 1 ml of ethanol-ether (1/1) and dissolved in 2 ml of 0.1 N any effect on the phosphorylation of ROS protein. The protein NaOH at 1000 and the Cerenkov radiation was measured (18). kinase in ROS is specific to photobleached, not unbleached rhodopsin, and the increase in substrate concentration which Measurement of Protein Dephosphorylation. Samples of ROS results from exposure of ROS to light is alone sufficient to were incubated with 1 mM MgCl2, 1 mM [a2P]ATP and 50 account for the increase in protein phosphorylation. mM Tris HCl (pH 7.4) at a protein concentration of about 0.4 mg/ml for 30 min at 37°. The suspension was then centri- MATERIALS AND METHODS fuged at 100,000 X g for 10 min and the pellet was washed Preparation of Rod Outer Segments from Bovine Retinas was twice by resuspension and centrifugation from 50 mM sodium as previously described (13). The segments routinely showed a phosphate buffer, pH 7.0, before final resuspension at a con- ratio of absorbance at 280 to 500 nm of about 2.2. Incubation centration of 1-2 mg of protein per ml in the same buffer. with 1 1-cis-retinaldehyde only increased the absorption at 500 Aliquots of the [82P]ROS were then incubated with 50 mM nm by 5-10%, indicating that the material was only 5-10% Tris-HCl (pH 7.4) at a protein concentration of about 0.4 bleached. mg/ml with the additions stated in the text. Aliquots (0.5 ml) were then taken at different times and mixed with 2 ml of ice- cold 10% trichloroacetic acid, the denatured protein being Abbreviation: ROS, rod outer segments. washed and its radioactivity measured as described above. 381 Downloaded by guest on September 28, 2021 382 Biochemistry: WeHer et al. Proc. Nat. Acad. Sci. USA 72 (1975)

TABLE 1. The effect of phosphodiesterase inhibitors and cyclic nucleotides on the o a0% bleach intrinsic protein kinase activity of ROS 0~ Va- -0 0 I 0 rC12 0 Phosphodiesterase Cyclic % of control L I L inhibitor nucleotide L 4 /%50/ bleach (no additions) 0 White light E SQ 20009 - 61 i 2.4 (6) m It c0 SQ20009 1 mM cAMP (I mM) 58.5 ±4(6) o A~~~~24%blech mI SQ 20009 cGMP (0. 1 mM) 55.5 i 3.8 (4) t A _ ----4 _ I ._.__. C Theophylline - 61. 5 ± 7.1 (6) X A Dark ° 10 30 50 70 90 110 10 mM cAMP 6 12 18 24 30 c Relative percentage oF Theophylline (1 mM) 61.8 ± 6 (4) Time (min) unbleached rhodopsin Theophylline cGMP (0.1 mM) 60, 70 FIG. 1. The effect of bleaching on the posphorylation of Papaverine - 89 ± 8.2 (7) rhodopsin. (a) Samples of purified rhodopsin were suspended at Papaverine 2 mM cAMP (1 mM) 79 i 12 (4) about 1 mg of protein per ml in 50 mM sodium phosphate buffer Papaverine cGMP (0. 1 mM) 85, 91 (pH 7.0) and bleached to various extents by exposure to white cAMP (I mM) 94.8 ±i 8 (6) light. Aliquots were taken to determine the extent of bleaching by - cGMP (0. I mM) 93 i 4.6 (3) measuring the decrease in absorbance at 500 nm and the remain- - Dibutyryl cAMP 77.2 i 11.6 (10) ing material was incubated under the phosphorylation conditions Dibutyryl cGMP 102. 5 i 4. 1 (4) described in the text. Broken lines show the time course of phos- Dim red light phorylation of intact ROS, A in dim red light; A or in white light. 20009 Similar results were obtained with three separate experiments. SQ 95 ±t 5 (4) (b) The maximum amount of phosphate which could be in- SQ 20009 1 mM cAMP (I mM) 91.2 i 4.7 (6) corporated into rhodopsin which had been bleached to various SQ 20009 cGMP (0. I mM) 102 extents was plotted against the amount of unbleached rhodopsin Theophylline 86. 7 ± 13.1 (6) in the sample, shown as a percentage of the amount in the Theophylline 10 mM cAMP (1 mM) 99. 3 i 1. 2 (3) material before exposure to light (relative percentage of un- Theophylline cGMP (0.1 mM) bleached rhodopsin). Results are shown as means standard Papaverine - 106.25 ±- 4.8 (4) deviations. Papaverine 2 mM cAMP (1 mM) 114.3 t 27 (4) Papaverine cGMP (0.1 -mM) RESULTS - cAMP (I mM) 104 ± 9 (5) 1 In agreement with the results of other workers (1-5), we found cGMP (0. mM) 97.7 ± 7.9 (4) Dibutyryl cAMP 98.9 ± 11.6 (10) that isolated ROS-on incubation with a [32P]ATP show light- Dibutyryl cGMP 99.3 ± stimulated incorporation of 32p (Fig. 1). The effect of light is 15.5 (4) much more pronounced than that observed by Frank et al. (4) who, after Protein phosphorylation was carried out as described in the 2-min incubation, only obtained a stimulation of text for 2 min either in white or dim red light. Results are shown as about 1.3 times while we obtained a stimulation of 10-20 means ± standard deviations with the number of observations times, in agreement with other workers (1, 3, 5). shown in parentheses. The incorporated radioactivity is bound to protein and not lipid, since it could neither be extracted into chloroform- was slightly inhibitory in the light (Table 1). The effect of methanol (2/1) nor into chloroform-methanol-concentrated phosphodiesterase inhibitors could be due to an increased HCl (300/200/1). Polyacrylamide gel electrophoresis in the accumulation of cyclic nucleotides, which cause an inhibition presence of sodium dodecyl sulfate showed, in agreement with of phosphorylation. Cyclic nucleotides, however, do not affect other workers (1-5), that opsin was the if not major, the only, phosphorylation, but it may be that it is difficult protein to be phosphorylated. It has been found that for them to serine reach the kinase inside the ROS membranes. To solve this and to a lesser extent threonine are the major, if not only, site of problem we examined the effect of cyclic nucleotides on the phosphorylation (3, 5). soluble kinase (see below). Other workers have reported, without giving any details, that the phosphorylation of ROS is unaffected by cAMP Extraction of Protein Kinase Activity from ROS. It has (3, 4). The effect of cGMP, which is also present in high con- previously been reported that rhodopsin kinase activity may centrations in ROS (8), was not examined. To examine the be extracted from ROS by sonication in Tris EDTA (5). We effects of cyclic nucleotides on the phosphorylation of ROS it treated ROS by a milder procedure, described in the Materials was necessary to add a phosphodiesterase inhibitor to prevent and Methods section. About 8-10% of the total ROS protein the hydrolysis of the cyclic nucleotide. is solubilized and the insoluble material remaining after ex- We found that theophylline and SQ 20009 (which is struc- traction showed a loss of intrinsic protein kinase activity, turally very similar to theophylline) are both quite strong particularly when measured in the light. The activity was inhibitors of ROS phosphorylation in the light but have less restored by addition of the Tris extract (Table 2) which did effect in the dark, while papaverine is only slightly inhibitory not, however, show any self-phosphorylation. in the light and has no effect in the dark (Table 1). Neither It was also found that the Tris extract catalyzed the phos- cAMP nor cGMP have any effect in the presence or absence of phorylation of purified photobleached, but not unbleached, phosphodiesterase inhibitors in the light or in the dark, and rhodopsin (Fig. 1). Exposure of the Tris extract to light did similarly dibutyryl cGMP has no effect in the light or dark, not cause it to phosphorylate unbleached rhodopsin, nor did it while dibutyryl cA;M.P, though having no effect in the dark, affect the rate with which it could catalyze the phosphoryla- Downloaded by guest on September 28, 2021 Proc. Nat. Acad. Sci. USA 72 (1975) An Opsin-Specific Protein Kinase in Retinal Rods 383

tion of bleached rhodopsin. The Tris extract thus contains a -E c- 8 /

protein kinase which is itself unaffected by light but which 40 a. catalyzes tbe phosphorylation of bleached rhodopsin. CoL ;i 4 If ROS were illuminated for 2 min before Tris treatment the o 0 quantity of protein that could be extracted was reduced from about 60 to 20 to 30 mg/g of total rhodopsin protein. More interestingly, the extract prepared in this way contained much less protein kinase activity (Fig. 2). `0 oE0 20 40 60 80 This could be explained if the kinase binds rather strongly C pg of tris extract protein to its substrate, bleached rhodopsin. Since there is much more FIG. 2. Opsin kinase activity of Tris extracts from ROS. of this substrate in light-exposed ROS, it is easier to extract ROS were extracted with 10 mM Tris-HCl as described in the the enzyme from material kept in the dark. text either in the dark (0), or after the ROS were exposed for 2 Because of the higher concentration of protein kinase activ- min to white light (0). The protein kinase activity of the extract ity all the Tris extracts referred to below as soluble enzyme towards purified bleached rhodopsin was then determined as described in the 100 of and a reaction were prepared from material kept in the dark. text, using ,ug rhodopsin time of 4 min. Similar results were obtained with. three separate The Effect of Cyclic Nucleotides on Protein Kinase Activity of experiments. Tris Extract Prepared from ROS. Working with a mixture of soluble enzyme and purified rhodopsin, it was possible to phosphorylation of rhodopsin catalyzed the Tris examine the direct effect of cyclic nucleotides on the protein by extract indicates that the kinase for this reaction kinase. As was the case with intact ROS, it was found that protein responsible is not the same as that which the theophylline and SQ 20009 were strongly inhibitory to the probably catalyzes phospho- of histones. In addition it was found that the phosphorylation of the purified rhodopsin, while cAMP or rylation Tris if cGMP alone, or in the presence of phosphodiesterase inhibi- extract showed slight any activity towards histones, casein, tors, have no effect (Table 3). It thus appears that the effect of or phosvitin, even in the presence of cyclic nucleotide (Table theophylline (or the derivative SQ 20009) on the phosphoryla- 4). Similarly we found that a histone kinase purified from beef tion of intact ROS was not due to an increase in cyclic nucle- kidney (21) could not catalyze the phosphorylation of otide concentration caused by the inhibition of phosphodies- bleached or unbleached rhodopsin either in the presence or A terase activity, but was due to a direct inhibition of protein absence of cyclic AMP. similar result has recently been kinase activity. The fact that more inhibition is observed in reported for a beef muscle histone kinase (20). the light than in the dark can be explained by the fact that The time course of protein phosphorylation of intact ROS is bleached rhodopsin is the substrate for the enzyme. In the essentially the same as the time course of the phosphorylation dark the concentration of bleached rhodopsin is very low so of purified bleached rhodopsin catalyzed by a Tris extract that the protein kinase is present in excess; partial inhibition when the two are recombined in the same proportion as in the of its activity would not cause a great reduction in protein intact material, (Fig. la). There is thus no evidence that opsin phosphorylation. In the light, however, there is much more bleached rhodopsin substrate so that the enzyme is no longer TABLE 3. The effect of phosphodiesterase inhibitors and cyclic present in excess, hence inhibitors would cause more reduction nucleotides on the opsin kinase activity of Tris extracts prepared of protein phosphorylation. from ROS Specificity of the Protein Kinase Activity of Tris Extracts % of Preparedfrom ROS. There have been reports of the occurrence Cyclic activity Phosphodiesterase nucleotide in control in whole retina and even in ROS an enzyme isolated of that is inhibitor (0. 1 mM) stimulated by cAMP and that catalyzes the phosphorylation (-additions) of histones (5, 19, 20, 22). The lack of effect of cAMP on the White light SQ 20009 1 mM 21, 20 TABLE 2. Restoration of intrinsic protein kinase activity to SQ 20009 1 mM cAMP 18, 17 Tris-extracted ROS cAMP 100, 103 SQ 20009 1 mM cGMP 17, 18.5 Intrinsic protein kinase activity* Theophylline 10 mM - 51.5 4 6.3 (6) Theophylline 10 mM cAMP 50.7 4 4.1 (6) Material Dim red light White light - Dibutyryl cGMP 100,97 Dibutyiyl cAMP 98 ± 2 (4) Dark-extracted ROS 75 ±4 20 5 Papaverine 2 mM - 88.25 4± 6.4 (4) Dark extract 1 3 2 5 2 mM ± Dark-extracted ROS plus Papaverine cAMP 75.5 7.5 (4) dark extract 73 5 74 4 Dim red light SQ 20009 1 mM - 45,41 ROS were extracted with 10 mM Tris *HCl and protein SQ 20009 1 mM cAMP 60,45 phosphorylation was carried out for 2 min as described in the text. - cAMP -98, 94 Where Tris extract was added it was at a concentration of 30,ug of protein per ml. Results are shown as mean ±- standard devi- Incubations were carried out for 2 min as described in the text ation and are taken from four experiments. with 200,g/ml of rhodopsin and 32 ,ug/ml of Tris extract protein. * % of activity of untreated ROS measured in dim red light Results are shown as means 4 standard deviations with the and white light. number of observations in parentheses. Downloaded by guest on September 28, 2021 'qQA UV= Biochemistry: Weller et al. Proc. Nat. Acad. Sci. USA 72 (1976) TABLE 4. Protein kinase activity of Tris extracts from ROS towards various protein substrates

Protein kinase activity, nmol/min per mg Tris extract protein 8 mM MgC12 1 mM MgC12 + 10pM cAMP -cAMP + 10,McAMP -cAMP Histone 0.08 A 0.08 (8) 0.05 4 0.06 (6) 0.8 i 0.8 (6) 0.5 0.6 (4) Phosvitin 0.0440.06(6) 0.09 4 0.1 Casein 0.1 4 0.1 Bleached rhodopsin 16 4 (6) 17 4 5 (6)

Protein kinase activity was determined as described in the text with 50-100 ,sg of Tris extract protein per ml. Results are shown as means A standard deviations with the number of observations in parentheses.

is phosphorylated in the intact ROS by enzymes apart from Frank et al. (4), who showed that there was a roughly linear the "opsin kinase" found in the Tris extract. relationship between the rate of phosphorylation of ROS and the extent of bleaching, but is in contrast to the results of Effect ofRhodopsin Photostimulation on Protein Phosphoryla- Bownds et al. (1), who initially reported that ROS prepared tion. If "opsin kinase" acts only on bleached rhodopsin, from frog retinas showed maximal phosphorylation after only exposure of ROS to light will increase the amount of substrate 1% of the rhodopsin is bleached. In a recent publication, available to the enzyme and hence the net amount of phos- however, they reported that further bleaching did cause an phate which can be incorporated into the ROS protein will increase in phosphorylation, though small bleaches were much also be increased. To demonstrate this point purified rhodopsin more effective (2, 22). Our experiments give no indication of was exposed to light for various times, producing various such an effect. The method which we used for the preparation degrees of bleaching, and the time course of phosphorylation of rhodopsin extracts bleached material (14). Examination of was determined. It was found that increased rhodopsin bleach- Fig. lb shows the line to cross the abscissa at 101%, suggesting ing increased the amount of phosphate which could be in- that the preparation of rhodopsin is only 1% bleached before corporated (Fig. 1). Both the initial rate of phosphorylation exposure to light. Intacts ROS show more phosphorylation in and the total amount of phosphate which could be incorpo- the dark than does a mixture of rhodopsin and Tris extract rated were increased. No alteration in the time course of phos- (Fig. la). phorylation was observed if the bleached rhodopsin was left Several workers have observed a maximal incorporation of in the dark for times up to 10 min before the start of phos- about 1 mole of phosphate/mole of rhodopsin (1, 3, 5). We, phorylation. This result indicates that the increased phos- however, like Franks et al. (4), found a maximal incorporation phorylation is due to an increased concentration of substrate of about 0.5 mole/mole of bleached rhodopsin. It is amount of only for the kinase reaction. To make this point clear the tempting to explain this result by suggesting that the rhodop- phosphate which could be maximally incorporated into rho- sin used in our experiments was already partially phosphoryl- dopsin was plotted against the percentage of unbleached rho- ated. dopsin in the sample. A completely linear relationship was In contrast to these observations Bownds et al., working observed (Fig. lb), showing that the amount of phosphate with ROS from frog retinas, have reported that at bleaches of can be into rhodopsin is which maximally incorporated less than 5%, 10-15 moles of phosphate may be incorporated directly proportional to the number of bleached rhodopsin With amounts of with that of per mole of bleached rhodopsin. higher molecules present. This result is in agreement bleaching, only 1 mole of phosphate can be incorporated, and they suggest that at low levels of bleaching phosphate may be _100 bound to unbleached rhodopsin molecules (2). Our experi- *' 181 ~~a 0 no of such an effect. L ments with cattle retinas give indication 0 CL N 80 0-I _ Dephosphorylation of Rhodopsin. If rhodopsin phosphoryla- 0 0 tion has any physiological importance, enzymes must be E3 16 0 E present which can not only catalyze the phosphorylation of rhodopsin but also the dephosphorylation, otherwise, in vivo, not occur. X( 40 the turnover of rhodopsin phosphate could x It may be seen from Fig. 3a that ROS contain only slight E 14 14 21 o 0o 10 20 0L 7 protein phosphatase activity, which is increased in the pres- Time U Time (min) (min) ence of Mg2 + and inhibited by EDTA. cAMP had no effect on FIG. 3. Dephosphorylation of ROS proteins. (a) 32P-labeled the dephosphorylation and the same rate of reaction was ob- bleached ROS were incubated in dim red light at 370 at a protein served whether the experiment was carried out in the light or concentration of 300 ,ug/ml with 50 mM Tris. HOl (@), with in the dark. Other workers similarly have shown only very low 2-mM MgCl2 (0), or with 1 mM EDTA (A). (b) 3P-labeled ROS rates of dephosphorylation (4, 5). were added at a concentration of 0.4 mg of protein per ml to 50 Regeneration of phosphorylated bleached ROS did not the or mM Tris.HCl pH 7.4, 2 mM MgC12 in presence (/v) alter their rate of dephosphorylation. absence (0) of a homogenate of whole retina prepared at a ml of reaction mixture. Similar The low rate of dephosphorylation of phosphorylated rho- concentration of 0.5 retinas per the results were obtained with three separate experiments. dopsin is a matter for some surprise. Perhaps protein Downloaded by guest on September 28, 2021 Proc. Nat. Acad. Sci. USA 72 (1975) An Opsin-Specific Protein Kinase in Retinal Rods 385

phosphatase activity is soluble and lost during the preparation 1. Bownds, D., Dawes, J., Miller, J. & Stahlman, M. (1972) of the ROS. If phosphorylated ROS were incubated at 370 Nature New Biol. 237, 125-127. 2. Bownds, D., Brodie, A., Robinson, W. E., Palmer, D., with a crude retinal homogenate, they were certainly dephos- Miller, J. & Shedlovsky, A. (1974) Exp. Eye Res. 18, 253- phorylated more rapidly (Fig. 3b) but the reaction was still 269. not very fast. 3. Kuhn, H. & Dreyer, W. J. (1972) FEBS Lett. 20, 1-6. 4. Frank, R. N., Cavanagh, H. D. & Kenyon, K. R. (1973) DISCUSSION J. Biol. Chem. 248, 596-09. 5. Kuhn, H., Cook, J. H. & Dreyer, W. J. (1973) Biochemistry The results described in this paper show that ROS contain a 12, 2495-2502. protein kinase which is readily solubilized and which specifi- 6. Bitensky, M. W., Gorman, R. E. & Miller, W. H. (1971) cally catalyzes the phosphorylation of photobleached but not Proc. Nat. Acad. Sci. USA 68, 561-562. unbleached rhodopsin. Most protein kinases so far reported 7. Miki, M., Keirns, J. J., Marcus, F. R., Freeman, J. & fall into two broad those that the Bitensky, M. W. (1973) Proc. Nat. Acad. Sci. USA 70, groups: catalyze phos- 3820-3824. phorylation of phosvitin, and those that catalyze the phos- 8. Goridis, C., Virmaux, N., Urban, P. F. & Mandel, P. (1973) phorylation of histones, though, of course, both groups of FEBS Lett. 30, 163-166. enzymes catalyze the phosphorylation of a number of other 9. Bensinger, R. E., Fletcher, R. T. & Chader, G. J. (1974) proteins as well. The protein kinase present in the Tris Science 183, 86-87. 10. Goridis, C. & Virmaux, N. (1974) Nature 248, 57-58. extract prepared from ROS belongs to neither group, since it 11. Kuehn, G. D. (1971) J. Biol. Chem. 246, 6366-6369. can catalyze the phosphorylation of neither histones nor 12. De Lorenzo, R. J. & Greengard, P. (1973) Proc. Nat. Acad. phosvitin. It appears that it is an enzyme that specifically Sci. USA 70, 1831-1835. catalyzes the phosphorylation of bleached rhodopsin, and we 13. Virmaux, N., Urban, P. F. & Waenheldt, T. V. (1971) propose the name "opsin kinase." The fact that the protein FEBS Lett. 12, 325-328. 14. Virmaux, N., Waenheldt, T. V. & Urban, P. F. (1972) kinase acts only on bleached rhodopsin explains the observa- C.R.H. Acad. Sci. Ser. D. 275, 2041-2044. tion that exposure to light increases the phosphorylation of 15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, ROS protein. R. J. (1951) J. Biol. Chem. 193, 265-275. The fact that ROS proteins are phosphorylated through the 16. Glynn, I. M. & Chappell, J. B. (1964) Biochem. J. 90, 147- action of a specific protein kinase is one of considerable inter- 149. 17. Rodnight, R., Hems, D. A. & Lavin, B. A. (1966) Biochem. est. The synaptic membrane proteins of brain are phos- J. 101, 502-515. phorylated through the action of a tightly bound protein 18. Haviland, R. T. & Bieber, L. L. (1970) Anal. Biochem. 33, kinase which differs in properties in several respects from the 323-334. enzyme that catalyzes the phosphorylation of phosvitin or 19. Pannbacker, R. G. & Schock, D. D. (1973) J. Gen. Physiol. histones (23-25). Like opsin kinase it is inhibited by theophyl- 61, 257-258. line but not by papaverine (23, 25, 26). It is, however, stimu- 20. Frank, R. N. & Bensinger, R. E. (1974) Exp. Eye Res. 18, lated by cAMP (26, 27). 271-280. The time course of phosphorylation is much too 21. Gilman, A. G. (1970) Proc. Nat. Acad. Sci. USA 67, 305-312. rhodopsin 22. Bownds, D., Dawes, J. & Miller, J. (1973) in Biochemistry slow to be involved in the direct response to light but it may be and Physiology of Visual Pigments, ed. Langer, H. (Springer, postulated that the phosphorylation is concerned in the adap- New York), pp. 267-273. tation of the photoreceptors to light or dark conditions (5, 22). 23. Weller, M. & Rodnight, R. (1971) Biochem. J. 124, 393-406. It has been suggested before (25, 27-29) that the phosphoryla- 24. Weller, M. (1972) Ph.D. Dissertation, London University. tion of certain membrane proteins may control passive perme- 25. Rodnight, R. & Weller, M. (1972) in Effect of Drugs on ability to certain ions and it may be that the phosphorylation Cellular Control Mechanisms, eds. Rabin, R. D. & Friedman, B. R. (McMillan Press Ltd., London), pp. 175-192. of rhodopsin may similarly control passive permeability in 26. Weller, M. & Rodnight, R. (1973) Biochem. J. 132, 483-492. ROS, so mediating the responsiveness to a light impulse. 27. Weller, M. & Rodnight, R. (1970) Nature 225, 187-188. We thank G. Nullans for excellent technical assistance and 28. Kebabian, J. W. & Greengard, P. (1971) Science 174, 1346- Dr. S. M. Hess (Squibb Institute of Medical Research) for the 1348. gift of the compound SQ 20009. M.W. gratefully acknowledges 29. De Lorenzo, R. J., Walton, K. G., Curran, P. F. & Green- receipt of a Science Research Council Fellowship. gard, P. (1973) Proc. Nat. Acad. Sci. USA 70, 880-884. Downloaded by guest on September 28, 2021