Eur. J. Biochem. 209,1035-1040 (1992) r E B S 1992
c)
Characterization of fly rhodopsin kinase Yair N. DOZA', Baruch MJNKE', Michael CHOREV3 and Zvi SELINGER' Departments of Biologcal Chemistry Physiology' and Pharmaceutical Chemistry3, The Hebrew Univcrsity of Jerusalem and the Minerva center for studies of visual transduction, Jerusalem, Israel (Received July 20, 1992) - EJB 92 1033
Rhodopsin kinase activity of Musca dornestica was Characterized in a reconstitution assay, using urea-treated eye membranes as substrate and a purified fraction of eye cytosol as the enzymc. Analysis of kinase activity in fly eye, brain and abdomen extracts by reconstitution assays revealed that fly rhodopsin kinase is an eye-specific enzyme. It preferentially phosphorylates the light-activated form of rhodopsin (metarhodopsin) and has little activity with other protein substrates. Rhodopsin kinase binds to metarhodopsin and is released from rhodopsin-containing membranes. Metarhodopsin is a poor substrate for kinases from tissues other than the eye, making it a unique substrate for rhodopsin kinase. Rhodopsin kinase is inhibited by heparin, but not by the protein inhibitor of cAMP-dependent protein kinase. Its K,,, for ATP is 9 pM. Since fly rhodopsin is coupled to phospholipase C, studies of the interaction of rhodopsin with rhodopsin kinase can be useful in analysis of the reactions that lead to termination of thc inositol-phospholipid-signalingpathway.
'The photoreceptor protein, rhodopsin, is a major protein component in the retina of the compound eye of the fly. Like rhodopsin of vertebrates [I, 21, fly rhodopsin undergoes lightdependent phosphorylation [3]. Phosphorylation of vertebrate rhodopsin is the Grst event in quenching signal transduction initiated by light [4 - 61, and phosphorylation is important for desensitization of the P-adrenergic receptor [7 - 111 (for reviews see [12, 131). The family of kinases which phosphorylates receptors coupled to guanine-nucleotide-binding regulatory proteins (G-proteins) has been recently extended [14 - 161, and three members of this family have been cloned [I 7 191. A hallmark of these kinases is their specificity toward the active form of the receptor and their rather narrow range of substrate specificity. It has also been shown that these kinases translocate between the cytosol in unstimulated cells and the membranes of receptor-stimulated cells 120 - 231. Vertebrale rhodopsin kinase has been biochemically well characterized [2, 24-27], while little is known about invertebrate rhodopsin kinase [3]. Difficulties in studying the latter kinase arise from association of the kinase with its membranous rhodopsin substrate and the limited availability of biological material that hinders extensive purification and resolution of these two components. To overcome these difficulties, we have depleted the rhodopsin-containing membranes of rhodopsin kinase activity by urea treatment and reconstituted it under controlled conditions with partially purified rhodopsin kinase. This system was used to characterize the tissue and substrate specificity of rhodopsin kinase, as well as its kinetic parameters and the effect of light on its interaction with the mein brane. ~
Correspondence to Z. Selinger, Department of Biological Chemistry, The Institute of Life Science, The Hebrcw University oflerusalem, J L-9 1904 Jerusalem, Israel F a : + 972 2521427. Abbreviutions. I'hMeS02Ft phenylmethylsulfonyl fluoride: G protein, guanine-nucleotidc-binding regulatory protein. Enzyme. Rhodopsin kinase (EC 2.7.1.99).
MATERIALS AND METHODS [ Y - ~ ~ P ] A(3000 T P Cijmmol, 10 Cijl) was purchased from Amersham. ATP (grade l), EGTA, histone type 2-A, phenylmethylsulfonyl fluoride (PhMeSO,F), heparin (grade 2 ) ; 153 Ujmg) and the protein inhibitor of CAMP-dependent protein kinase were purchased from Sigma. Leupeptin wa5 from the Peptide Institute (Osaka, Japan), carboxymethyl-cellulose (CM52), DEAE-cellulose (DE23) and P-81 paper were from Whatman. The flies used were white-eyed Musca dornestica, whiteeyed Lucilia cuprina and white-eyed Oregon R Drosophzlu rnelanogaster. Preparation of Muscu eye membranes Musca eye membranes were prepared as previously described [2S]. Briefly, 100 eyes were dissected with a razor blade under white light and collected in 200 p1 ice-cold homogenization buffer, containing 12 mM phosphate, pH 7.5, 5 mM MgS04, 2 mM 2-mercaptoethanol, 60 pg/ml leupeptin and 2 mM PhMeS02F. All further operations were carried out at 0-4'C, under dim red light. Eyes were dark-adapted for 30 min. Unless otherwise indicated, eyes were illuminated for 1 Inin with orange light, prior to homogenization by 10 strokes of a Teflon pestle in a 1.5-ml Eppendorf tube. The homogenate was transferred to a new Eppendorf tube, leaving out the corneas, and reilluminated with orange light for 30 s. The membrane fraction was obtained by centrifugation (14000 g for 15 min at 4 "C). The homogenization supernatant, containing most of the rhodopsin kinase activity, was further purified as described. The pellet, containing all the rhodopsin, was resuspended in 200 pl. Eye membranes from other flies were obtained as described above, using SO Lucilia eyes and 350 Drosopkila heads, each in 200 p1 homogenization buffer.
Rhodopsin-kinase-depleted eye membranes Eye membranes suspended in 200 p1 homogenization buffer were mixed with 200 pl freshly prepared 8 M urea and
1036 2 inM 2-mercaptoethanol. Membranes were illuminated for 30 s with orange light, incubated for 5 min at 4°C and pelleted as before. Urea was removed by two centrifugal washes with 400 pI homogenization buffer, followed by resuspension in the initial homogenization volume. Urea treatcd membranes can be stored frozen at - 70°C in the dark for at least 3 months without significant loss of efficiency as substrate. Partial purification of rhodopsin kinase DEAE-cellulose was equilibrated in 12 mM phosphate, pH 7.5, and 5 mM MgS04. 200 pl swollen DEAE-cellulose/ buffer (I : 1) was centrifuged (1300 g for 4 min at 4'C). The supernatant was removed and the DEAE-cellulose resuspended in 200 pl homogenization supernatant. After incubation for 10 min on ice, the suspension was centrifuged as before. I'he supernatant containing rhodopsin kinase was carefully removed without taking any DEAE-cellulose fibers. Table 1. Purification of rhodopsin kinase. Activity of rhodopsin kinase in diffcrent fractions is given as a percentage of the activity measured by scanning densitometry of rhodopsin bands from a reconstitution assay (Fig. IC). Reconstitution with eye-homogenate supernatant is 100% (note that in DEAE-cellulose and carboxymethyl-cellulose supernatanls, the activity is higher than 100%; see text). The protein amounl of each fraction was determined by the Bradford assay, as described in Materials and Methods. Cm, carboxymcthyl. Supernatant
Homogenate DEAE-cellulose Cm-cellulose
Protein
Activity
Purification
Pg:eYe
%
-fold
6.14 z .33 0.16
100 145 143
1 6.7 55
This supernatant can be stored at --70°C for at least 3 months without significant loss of activity. A similar procedure was previously described to inactivate rhodopsin kinase of bovine rod outer segment [29] Further purification was carried out on a 1-ml carboxymethyl-cellulose column. The DEAE-cellulose supernatant equivalent of 200 eyes (25 pg protein in 400 pl) was loaded onto the carboxymethyl-cellulose column equilibrated with homogenization buffer. The column was washed with 5 ml homogenization buffer, and rhodopsin kinase was eluted by stepwise addition of 1 ml homogenization buffer containing 0.1 M, 0.3 M, 0.5 M and 1 M NaC1. Fractions of 800 pl were collected. Reconstitution of rhodopsin kinase activity Urea-treated membrane substrate (3 pg protein) in 10 pl and rhodopsin kinase in 10 pl DEAE-cellulose or 20 pl carboxymethyl-cellulose fraction containing (1.5 pg or 0.1 pg protein, respectively) as indicated, were mixed and illuminated for 30 s at 20°C with blue light. The reaction was initiated by addition of 25 pl reaction mixture containing 12 mM phosphate, pH 7.5, 5 mM MgS04, 18 pM ATP, 3 pCi [ Y - ~ ~ P I A T P and 3.6 mM EGTA, then incubated at 20°C. The reaction was terminated at the indicated times by addition of 11 111 SDSi PAGE loading buffer. Peptide synthesis and phosphorylation The peptide Ac-Arg-Arg-Arg-Ala-Ala-Ala-Ala-Ala-SerGlu-Glu-Glu-NH2, was synthesized on a Applied Biosystems 430A synthesizer. N-Butyloxycarbonyl amino acid were used for protection of the a-amino group. Side-chain protection was 0-benzyl for serine and glutamic acid and p-toluensulfonyl for arginine. Each arginine residue was introduced by a double-coupling cycle.
1
2
3
4
5
6
7
8
9
Fig. 1. Purification of rhodopsinkinase. (A) Coomassie-blue-stained 10% SDS/PAGE of homogcnate supkrnatant in lanc 1, and the supernatant from the DEAE-cellulose-purification step (as described in Materials and Methods) in lane 2, each containing an cquivalcnt amount of protein derived from six Muscu eyes. (B) Autoradiogram of reconstitution assay containing 10 pl homogenate supernatant or DEAE-cellulose fraction, and 20-pl carboxymethyl-cellulose column fractions, with 10 p1 urea-treated membranes and 25 p1 reaction mixture, as described in Materials and Methods. In each reaction, the salt conccntration was adjusted to 0.5 M NaCl. Lane 1, urea-trcated membranes alone. Other lanes also contain the following: 2, homogenization supernatant; 3, DEAE-cellulose supernatant; 4, fraction of the loading of DEAE-cellulose fraction containing rhodopsin kinase; 5 , the first fraction of the wash; 6, fraction from elution with 0.1 M NaC1; 7, fraction from elution with 0.3 M NaCl; 8, fraction from elution with 0.5 M NaC1; 9, fraction from clution with 1 M NaC1. (C) Quantitalion or rhodopsin kinase activity by densitometry of the rhodopsin lanes from (B). The percentage activity is written above each column.
1037
0
B
20
40
60
80
100
time (min)
Fig. 3. Peptide phosphorylation by rhodopsin kinase. Phosphate incorporation by rhodopsin kinase in the presence ( O ) ,or absence (0),of pcptidc substrate is shown. The reaction contained 1 mM peptide. 200 pM ATP, 3 pg protein from the DEAE-cellulose fraction of rhodopsin kinase in 200 pl, at 30"C, as described in Materials and Methods. At each time, 40 pl reaction was absorbed onto 2 cm x 2 cm P-81 papers, and immersed immediately in 0.6% phosphoric acid. The radioactivity on the P-81 papers was detcrmined by measuring Cherenkov radiation.
Protein quantitation
0
2
4
6
8
10
12
time (min)
Fig. 2. Time course of rhodopsin kinase activity. Assay of rhodopsin kinase was carried out as described in Materials and Methods with a carboxymethyl-cellulose fraction of rhodopsin kinase a t 100 pM ATP. (A) Autoradiogram of phosphorylated proteins in the rhodopsin kinase assay, separated by SDSIPAGE. Rhodopsin is indicated by an arrow. Lanc 1, urea-treated membranes phosphorylated for 10 min without addition of rhodopsin kinase; lanes 2 - 5, reconstitution of rhodopsin kinase activity, incubated for 1 , 2 , 5 and 10 min, respectively. (B) Quantitation o r the rhodopsin kinase activity shown in (A). Each rhodopsin band was cut from SDS/PAGE, and Cherenkov radiation determined.
Phosphorylation of the pcptide was carried out as described in [27], with the following modifications: the incubation was carried out in the reaction mixture described above, containing 1 mM peptide, 200 pM ATP and 3 pg protein from a DEAE-cellulose fraction of rhodopsin kinase.
SDS/PAGE This was performed according to Laemmli [30], using 1096 acrylamide gels in an SE-600 Hoeffer apparatus. Dried gels were autoradiographed at - 70°C using Agfa Curix film. Quantitation was performed by cutting the bands corresponding to rhodopsin from dry gels and measuring Cherenkov radiation, or by scanning densitometry of the rhodopsin band rrom autoradiographs with a Bio Instruments model SLTKFF scanning densitometer.
Protein quantitation was carried out in 96-well microplates, taking 5O-pl samples and 200 p1 Bradford reagent. The results were read with 595 filter on a Microplate EL309 reader. Bradford reagent was prepared according to [31]. Illumination with lights of different wavelength An optical bench, with a quartz iodide lamp of 100 W and 12 V, was used. An orange-edge filter (Schott OG-590) or a blue-light filter (Schott BG-28) and two KG3 heat-reducing filtcrs from Schott were placed in the light path. The light was focused and delivered to the tube containing the sample via a light guide (4 mm, Schott). The energy of the unattenuated orange or blue light, used at the level of the sample, was 12.7 mW/cm2 or 2.6 mW/cm2, respectively.
RESULTS AND DISCUSSION Purification of rhodopsin kinase was based on the fact that rhodopsin kinase is a basic protein. It dose not bind at all to DEAE-cellulose (in contrast to most of the other proteins of fly eye). Purification of rhodopsin kinase from the soluble fraction of orange-illuminated eye homogenate, by batch adsorption to DEAE-cellulose, resulted in a sevenfold increase in specific activity of rhodopsin kinase (Table 1, Fig. 1A). Further purification of rhodopsin kinase was carried out on a 1-ml carboxymethyl-cellulose column. All rhodopsin kinase activity of the DEAE-cellulose supernatant was absorbed to the carboxymethyl-cellulose column. The column was washed with 5 ml homogenization buffer, and rhodopsin kinase was eluted by stepwise wash with increasing NaCl concentrations in homogenization buffer (see Fig. 1 and Materials and Methods). The overall purification of these two steps was 55[old, but the rhodopsin kinase fraction still contained several
1038
0.0 -I 0
1
2
IlATP ( p M
8000
E
Q
.-0
3
)
Fig. 5. Double-reciprocal plot of rhodopsin kinase activity in reconstitution assay using different concentrations of ATP. Rhodopsin kinase activity was determined by a 2.S-min assay, as described under Materials and Methods with a DEAE-ccllulose fraction of rhodopsin kinase. The reactions were separated by SDSiPAGE; bands corresponding to rhodopsin were cut out and their Cherenkov radiation determincd. At low concentration of ATP (0.5 pM, 0.75 pM). thc reaction volume was 100 pl, keeping the same concentration of all other components except those or rhodopsin kinase and membrane substrate. The rcsults are means of five experirncnts.
B
-
-'
6000
c
E
P 0
.C
-
4000
0
mQ 8 c 0 2000
a N
m
0 0
1
2
3
No.
4
5
6
of wash
Fig. 4. Association of rhodopsin kinase with eye membranes containing either metarhodopsin or rhodopsin. Eye-membranc suspensions wcrc illuminated for 30 s with either blue or orange light. After each illumination, membranes were precipitated by centrifugation (14000 g for 15 min at 4°C). The supernatant was discardcd and membrancs were resuspended in fresh homogenization buffer. Aliquots of 10 pl were 500 removed for rhodopsin kinasc assay of the mcmbranc bound kinasc 0 2 4 6 8 10 12 after each successive centrifugation. (A) Autoradiogram of rhodopsin kinase activity of a 5-min assay associated with the membrane. MemHeparin ( Lr g h l ) branes washed 1- 6 times arc shown in lanes 1 and 7 , 2 and 8 , 3 and Fig. 6. Inhibition of rhodopsin kinase activity by heparin. Rhodopsin 9 , 4 and 10, 5 and 11 and 6 and 12, respectively. Lanes 1 - 6 represent membranes illuminated with blue light, lanes 7 - 12 represent mem- kiriase reconstitution assay was performed as dcscribed in Matcrials and Methods, using a DEAE-cellulose fraction of rhodopsin kinase branes illurninatcd with orange light. The rhodopsin band is indicated and diffcrcnt concentrations of heparin, as indicated. The reactions by an arrow. (€3) Quantitation of rhodopsin kinase activity was carried out as in Fig. 2R ( 0 ) .Blue-light illumination between washes; (A) were separated by SDSjPAGE; bands corresponding to rhodopsin were cut out and Cherenkov radiation was determined. orange-light illumination between washes.
proteins, as seen in Coomassie-blue-stained and silver-stained gels (not shown). The activity of rhodopsin kinase was monitored by a reconstitution assay of each fraction with urea-treated eye membranes (Fig. l B , C). As observed, the activity in DEAE-cellulose and carboxymethyl-cellulose fractions is higher than the activity in the unpurified supernatant (Fig. lB, Table 1). We suppose that DEAE-cellulose absorbs enzymes that inhibit the activity of rhodopsin kinase, like G-protein o r arrestin which can compete for binding to rhodopsin, or phosphatases. The activity of the fractions from the carboxymethyl-cellulose column show sharp elution of rhodopsin kinase in the elution volume containing 0.5 M NaC1.
Urea treatment of Musca eye membranes decreased the endogenous phosphorylation activity of membranes containing metarhodopsin by more than 90% (Fig. 2A). Routinely, urea-treated membranes were used as substrate in reconstitution assays of rhodopsin kinase activity. A time course of rhodopsin kinase activity at 100 FM ATP, in a reconstitution assay with carboxymethyl-cellulose fraction of rhodopsin kinase, is shown in Fig. 2A. Phosphorylation was stringently dependent on conversion of rhodopsin to metarhodopsin, and maximum phosphorylation was obtained in blue-illuminated membranes in the presence of EGTA (not shown) [ 3 ] .The first lane of Fig. 2A shows the background activity of urea-washed membranes. The amount of rhodopsin phosphorylation on
1039 Rhodopsin phosphorylation 3000
A
Histone phosphorylation
woo
B
5000
4000
3000
zoo0 1000
0 0
Time (min)
10
20
30
Time (min)
Fig. 7. Enzyme and substrate specificity of rhodopsin kinase activity. Phosphorylation of rhodopsin (A) and histone (B) by extract of Muscu eye, brain or abdomen. (A) Reconstitution assay of rhodopsin phosphorylation was performed as described in Materials and Methods with equivalent amounts of protein from extract of MUSCU brain, abdomen and eye. The equivalent amount of eye extract was purified with DEAEcellulose as described in Materials and Methods. Thc rhodopsin bands were cut out and the Cherenkov radiation was measured. (B) Kinase assay using histone. Each assay system contained 40 pg protein from Muscu brain, abdomen or cye extract. 40 pg protein from eye extracl was purified with DEAE-cellulose as described in Materials and Methods. Phosphorylation reactions were as described in rhodopsin kinase assay, except that histone (2 mg/ml) was addcd in place of eye membranes. The reaction volume was 100 fi/assay. At the indicated times, aliquots were spotted onto 2 cm x 2 cm 3MiM Whatman paper, which was immediately immersed in 10% ice-cold trichloroacctic acid. The papcrs were washed four times with 10% trichloroacetic acid, overnight at 4°C. Cherenkov radiation of the filters was measured. (@) Eye; (m)brain; (A)abdomen.
10-min incubation is negligible, less than the activity after 1 min in the reconstitution system containing the carboxymethyl-cellulose fraction of rhodopsin kinase (Fig. 2A, lane 2). At the end of the 10-min reaction, 60% of the A'TP remained intact. Quantitation of the extent of rhodopsin phosphorylation (Fig. 2B) shows that rhodopsin kinase activity in the reconstitution assay is linear for 5 min then decreases slightly after 5 - 10 min (Fig. 2A, B). Comparing the endogenous rhodopsin kinase activity in urea washed membranes with the activity ofa 10-min reconstitution assay (Fig. IB, cf. lanes 1 and 6), reveals that addition of rhodopsin kinase from the carboxymethyl-cellulose fraction to urea-washed membranes, specifically increases the phosphorylation of rhodopsin. The prominent band that undergoes phosphorylation is a 31-kDa protein. This protein has the same electrophoretic mobility and undergoes light-dependent phosphorylation as the D.rnelunoguster rhodopsin, identified by a monoclonal antibody specific to Drosophila rhodopsin [32]. Similar experiments with less-pure rhodopsin kinase from DEAE-cellulose fraction gave the same results. Further purification on a carboxymethyl-cellulose column did not increase the specificity of phosphorylation in the reconstitution assay (compare lanes 3 and 8 in Fig. 1B). Additional characterization of rhodopsin kinase was carried out with the DEAEcellulose fraction of rhodopsin kinase, because the carboxymethyl-cellulose fractions of rhodopsin kinase were very labile (only fresh fractions were used). Phosphorylation of other protein bands is considerably less affected by the addition of the rhodopsin kinase carboxymeth yl-cellulosc fraction. Phosphorylation of high-molecularmass proteins is seen in all experiments. This activity ir not related to rhodopsin kinase, since phosphorylation of proteins of high molecular mass also appears in urea-washed membranes, to a greater extent than in the reconstitution assays. To show the activity of rhodopsin kinase in the DEAE-cellulose fraction, a peptide resembling the recognition sequence of
bovine rhodopsin kinase [27] was used as substrate. In Fig. 3, the activity of rhodopsin kinase in the DEAE-cellulose fraction toward the peptide substrate is shown. The activity is slow (as in the case of bovine rhodopsin or p-adrenergic receptor kinase) [27], and linear for a long time (95 min in this experiment). To test the interaction between rhodopsin kinase and substrate in native membranes, membranes prepared from either blue- or orange-illuminated eyes were subjected to successive extraction with homogenization buffer followed by centrifugation. The amount of rhodopsin kinase bound to metarliodopsin or rhodopsin-containing membranes was assayed after each repeated centrifugal wash. It is shown that rhodopsin kinase is retained by metarhodopsin-containing membranes (Fig. 4). In contrast, rhodopsin kinase was gradually released from rhodopsin-containing membranes. The amount of rhodopsin kinase released from rhodopsin-containing membranes by each centrifugal wash was much less than one would expect from equilibration of a soluble protein with the aqueous medium. This result indicates that rhodopsin kinase binds tightly to its substrate, metarhodopsin, and partially dissociates from rhodopsin (Fig. 4). However, even under resting conditions, in rhodopsin-containing membranes, a substantial part of rhodopsin kinase is associated with the membranes. Perhaps fly rhodopsin kinase is post-translationally modified, as is bovine rhodopsin kinase which is isoprenylated [33]. Presumably, this localization of rhodopsin kinase enables it to quench immediately the activated form of the photopigment. Preferential interaction of other proteins with metarhodopsin has been described for preparations of bovine rod outer segments [4] and fly rhabdomere [3]. Determination of the K , for ATP using a reconstitution assay with a DEAE-cellulose fraction of rhodopsin kinase gave a value of 9 1.5 pM (Fig. 5). This result is not very difftrent from the K , for ATP of bovine rhodopsin kinase ~31.
1040 REFERENCES Kuhn, H. (1974) Nature 250, 588 590. Kuhn, H. (1978) Biochemistry 17,4389-4395. Bentrop, J. & Paulsen, R. (1986) Eur. J . Biochem. 161, 61 -67. Wilden, U., Hall, S. & Kuhn, H. (1986) Proc. Nut1 Acad. Sci. USA 83, 1174- 1178. 5. Miller, J. L. & Dratz, E. A. (1984) Vision Res. 24, 1509- 1521. 6. Sitaramayya, A. & Liebman, P. A. (1983) J . Bid. Chem. 258, 1205- 1209. 7. Bouvier, M., Hausdorff, W. P., DeBlasi, A., O’Dowd, R. F. Kobilka, B. K., Caron, M. G. & Lefkowitz, R. J. (1988) Xaturr 333, 370 - 373. 8. Roht, N . S., Campbell, P. T., Caron, M. G., Lefkowitz, R. J. & Lohse, M. J. (1991) Proc Natl Acad. Sci. USA 88,6201 -6204. 9. Lohse, M. J., Lefkowitz, R. J., Caron, M. G. & Benovic, J. L. (1989) Proc. Natl Acad. Sci. USA 86,301 1-3015. 10. Hausdorff, W. P., Bouvier, M., O’Dowd, B. F., Irons, G., Caron. M. G. & Lefkowitz, R. J. (1989) J . Bid. Chem. 264, 1265712665. 11. Hausdorff, W. P., Campbell, P. T., Ostrowski, J., Yu, S. S., Caron, M. G. & Lefkowitz, R. J. (1991) Proc. Natl Acad. Sci. USA 88, 2979 - 2983. 12. Huganir, R. L. & Greengard, P. (1990) Neuron 5 , 555 - 567. 13. Hausdorff, W. P., Caron, M. G. & Lefkowitz, R. J. (1990) FASEB J . 4,2881 -2889. 14. Reneke., J. F., Blumer, K. J., Courchesne, W. E. & Thorner, J. (1988) Cell 55, 221 -234. 15. Meier, K. & Klein, C. (1988) Proc. Null Acad. Sci. USA 85, 2181 -2185. 16. Vaughan, R. A. & Devreotes, P. N. (1988) J . Biol. Chem. 263, 14 538 - 14543. 17. Benovic, J. L., DeBlasi, A,, Stone, W. C., Caron, M. G. & Lefkowitz, R. J. (1989) Science 246, 235-240. 18. Lorenz, W., Inglese, J., Palczewski, K., Onorato, J. J., Caron, M. G. & Lefkowitz, R. J. (1991) Proc. Natl Acad. Sci. USA 88, 8715 - 8719. 19. Benovic, J . L.; Onorato, J. J., Arriza, J. L., Stone, W. C., Lohse, M.; Jenkins, N. A., Gilbert, D. J., Copeland, ?I. G., Caron, M. G. & Lefkowitz, R. J. (3991)J. Biol. Chem. 266,14939- 14946. 20. Strasser, R. €I., Benovic, J. L., Caron, M. G. & Lefkowitz, R. J . (1986) Proc. Natl Acad. Sci. USA 83,6362-6366. 21. Mayor, F. Jr, Benovic, J. L., Caron, M. G. & Lefkowitz, R. J. (1989) Science 246, 235-240. 22. Chuang, T. T., Sallese, M., Ambrosini, G., Parruti, G. & DeBlasi, A. (1 992) J. B i d . Chem. 267,6886 - 6892. 23. Palczewski, K., McDowell, J. H. d Hargrave, P. A. (1Y88) J . Biol. Chem. 263,14067-14073. 24. Palczewski, K., Kahn, N. & Hargrave, P. A. (1990) Biochemistry 29, 6276 - 6282. 25. Palczewski, K., Buczylko, J., Kaplan, M. W., Polans, A. S. & Crabb, J. W. (1991) J. Biol. Chem. 266,12949-12955. 26. Kelleher, D. J. & Jhonson, G. L. (1990) J . Biol. Chem. 265,26322639. 27. Onorato, J. J., Palczewski, K., Regan, J. W., Caron, M. G., Lefkowitz, R. J. & Benovic, J. L. (1991) Biochemlrtry 30,5118 5125. 28. Blumenfeld, A., Erusalimsky, J., Heichal, O., Selinger, Z. & Minke, B. (1985) Proc. Nut1 Acad. Sci. USA 82, 7116-7120, 29. Shichi, H. & Somers, R. L. (1978) J . Biol. Chem. 253, 70407046. 30. Lacmmli, U. K. (1970) Nature 227, 680-685. 31. Bradrord, M. M. (1976) Anal. Biuchem. 72, 248-254. 32. Couet, H. & Tanimura, T. (1987) J. Cell Biol. 44, 50- 56. 33. Inglese, J., Glickman, J . F., Lorenz, W., Caron, G. M. & Lefkowitz, R. J. (1992) .J. Biol. Chem. 267, 1422-1425. 34. Lohse, M. J., Benovic, J. L., Caron, M. G. & Lefkowitz, R. J. (lYY0) J . Biol. Chem. 265, 3202-3209. 1. 2. 3. 4.
Fig. 8. Musca rhodopsin kinase activity with rhodopsin from different fly species. Eye membranes from Muscu, Luciliu, and head membranes from Drosoplzila were prepared as described in Materials and Methods. The reconstitution assay contained a DEAE-cellulose fraction of rhodopsin kinase from five Musca eyes and urea-washed membranes of five Musca eyes (lane 2), four Lucilia eyes (lane 4), and 20 Drosophila heads (lane 6). As a control, phosphorylation of an cqual amount of urea-washed membranes alone, from each fly, is shown. Musca membranes in lane 1; Lucilia membranes in lane 3; Drosophila membranes in lane 5.
Heparin, an inhibitor of the fl-adrenergic receptor kinase [34], was tested in the reconstitution assay with DEAE-cellulose fraction of rhodopsin kinase, and was found to be a potent inhibitor of fly rhodopsin kinase with a median effective concentration of 2.5 pg/ml (Fig. 6). A protein inhibitor of CAMP-dependent kinase had no effect (data not shown). To test the specificity of rhodopsin kinase activity toward rhodopsin, extracts of different fly tissues were assayed using urea-treated membranes with histone as a general substrate. The DEAE-cellulose fraction of partially purified rhodopsin kinase has high activity with metarhodopsin (Fig. 7A), while it is almost inactive toward histone (Fig. 7B). In contrast, Musca brain cytosol contains high kinase activity with histone and has little activity with metarhodopsin. While the abdomen extract has lower kinase activity than brain, it also preferentially phosphorylates histone over metarhodopsin. These results indicate that rhodopsin kinase is specifically expressed in fly eye, while other Musca tissues lack rhodopsin kinase activity. Furthermore, metarhodopsin is poorly phosphorylated by kinases from other fly tissues that are highly active toward histone. It is concluded that the fly rhodopsin kinase is selective for metarhodopsin, like bovine rhodopsin kinase and 8-adrenergic receptor kinase which are selective for their corresponding activated receptors [2, 141. To test whether rhodopsin kinase also recognizes rhodopsin of other photoreceptors, the activity of Musca rhodopsin kinase was tested with several other fly metarhodopsins. As seen in Fig. 8, Musca rhodopsin kinase effectively phosphorylates Lucilia and Drosophila metarhodopsins. The resolution and reconstitution of rhodopsin kinase and rhodopsin can be used in analysis of the functional consequences of rhodopsin phosphorylation and whether other kinases take part in this modification. The large number of Drosophilu mutants should make it possible to test the role of phosphorylation in termination of light-dependent signal transduction. We thank Dr D. Blest, L)r S. Stowe and Dr S. Trowel1 for the R16 Drosophila opsin monoclonal antibodies which were prepared by Dr H. Couet and Dr T. Tanimurd in the laboratory of Dr n.Blest. This work was supported by Grants from the National Institute of Health (EY03529) and the US-Israel Binational Science Foundation (BSF).
~
c)
Characterization of fly rhodopsin kinase Yair N. DOZA', Baruch MJNKE', Michael CHOREV3 and Zvi SELINGER' Departments of Biologcal Chemistry Physiology' and Pharmaceutical Chemistry3, The Hebrew Univcrsity of Jerusalem and the Minerva center for studies of visual transduction, Jerusalem, Israel (Received July 20, 1992) - EJB 92 1033
Rhodopsin kinase activity of Musca dornestica was Characterized in a reconstitution assay, using urea-treated eye membranes as substrate and a purified fraction of eye cytosol as the enzymc. Analysis of kinase activity in fly eye, brain and abdomen extracts by reconstitution assays revealed that fly rhodopsin kinase is an eye-specific enzyme. It preferentially phosphorylates the light-activated form of rhodopsin (metarhodopsin) and has little activity with other protein substrates. Rhodopsin kinase binds to metarhodopsin and is released from rhodopsin-containing membranes. Metarhodopsin is a poor substrate for kinases from tissues other than the eye, making it a unique substrate for rhodopsin kinase. Rhodopsin kinase is inhibited by heparin, but not by the protein inhibitor of cAMP-dependent protein kinase. Its K,,, for ATP is 9 pM. Since fly rhodopsin is coupled to phospholipase C, studies of the interaction of rhodopsin with rhodopsin kinase can be useful in analysis of the reactions that lead to termination of thc inositol-phospholipid-signalingpathway.
'The photoreceptor protein, rhodopsin, is a major protein component in the retina of the compound eye of the fly. Like rhodopsin of vertebrates [I, 21, fly rhodopsin undergoes lightdependent phosphorylation [3]. Phosphorylation of vertebrate rhodopsin is the Grst event in quenching signal transduction initiated by light [4 - 61, and phosphorylation is important for desensitization of the P-adrenergic receptor [7 - 111 (for reviews see [12, 131). The family of kinases which phosphorylates receptors coupled to guanine-nucleotide-binding regulatory proteins (G-proteins) has been recently extended [14 - 161, and three members of this family have been cloned [I 7 191. A hallmark of these kinases is their specificity toward the active form of the receptor and their rather narrow range of substrate specificity. It has also been shown that these kinases translocate between the cytosol in unstimulated cells and the membranes of receptor-stimulated cells 120 - 231. Vertebrale rhodopsin kinase has been biochemically well characterized [2, 24-27], while little is known about invertebrate rhodopsin kinase [3]. Difficulties in studying the latter kinase arise from association of the kinase with its membranous rhodopsin substrate and the limited availability of biological material that hinders extensive purification and resolution of these two components. To overcome these difficulties, we have depleted the rhodopsin-containing membranes of rhodopsin kinase activity by urea treatment and reconstituted it under controlled conditions with partially purified rhodopsin kinase. This system was used to characterize the tissue and substrate specificity of rhodopsin kinase, as well as its kinetic parameters and the effect of light on its interaction with the mein brane. ~
Correspondence to Z. Selinger, Department of Biological Chemistry, The Institute of Life Science, The Hebrcw University oflerusalem, J L-9 1904 Jerusalem, Israel F a : + 972 2521427. Abbreviutions. I'hMeS02Ft phenylmethylsulfonyl fluoride: G protein, guanine-nucleotidc-binding regulatory protein. Enzyme. Rhodopsin kinase (EC 2.7.1.99).
MATERIALS AND METHODS [ Y - ~ ~ P ] A(3000 T P Cijmmol, 10 Cijl) was purchased from Amersham. ATP (grade l), EGTA, histone type 2-A, phenylmethylsulfonyl fluoride (PhMeSO,F), heparin (grade 2 ) ; 153 Ujmg) and the protein inhibitor of CAMP-dependent protein kinase were purchased from Sigma. Leupeptin wa5 from the Peptide Institute (Osaka, Japan), carboxymethyl-cellulose (CM52), DEAE-cellulose (DE23) and P-81 paper were from Whatman. The flies used were white-eyed Musca dornestica, whiteeyed Lucilia cuprina and white-eyed Oregon R Drosophzlu rnelanogaster. Preparation of Muscu eye membranes Musca eye membranes were prepared as previously described [2S]. Briefly, 100 eyes were dissected with a razor blade under white light and collected in 200 p1 ice-cold homogenization buffer, containing 12 mM phosphate, pH 7.5, 5 mM MgS04, 2 mM 2-mercaptoethanol, 60 pg/ml leupeptin and 2 mM PhMeS02F. All further operations were carried out at 0-4'C, under dim red light. Eyes were dark-adapted for 30 min. Unless otherwise indicated, eyes were illuminated for 1 Inin with orange light, prior to homogenization by 10 strokes of a Teflon pestle in a 1.5-ml Eppendorf tube. The homogenate was transferred to a new Eppendorf tube, leaving out the corneas, and reilluminated with orange light for 30 s. The membrane fraction was obtained by centrifugation (14000 g for 15 min at 4 "C). The homogenization supernatant, containing most of the rhodopsin kinase activity, was further purified as described. The pellet, containing all the rhodopsin, was resuspended in 200 pl. Eye membranes from other flies were obtained as described above, using SO Lucilia eyes and 350 Drosopkila heads, each in 200 p1 homogenization buffer.
Rhodopsin-kinase-depleted eye membranes Eye membranes suspended in 200 p1 homogenization buffer were mixed with 200 pl freshly prepared 8 M urea and
1036 2 inM 2-mercaptoethanol. Membranes were illuminated for 30 s with orange light, incubated for 5 min at 4°C and pelleted as before. Urea was removed by two centrifugal washes with 400 pI homogenization buffer, followed by resuspension in the initial homogenization volume. Urea treatcd membranes can be stored frozen at - 70°C in the dark for at least 3 months without significant loss of efficiency as substrate. Partial purification of rhodopsin kinase DEAE-cellulose was equilibrated in 12 mM phosphate, pH 7.5, and 5 mM MgS04. 200 pl swollen DEAE-cellulose/ buffer (I : 1) was centrifuged (1300 g for 4 min at 4'C). The supernatant was removed and the DEAE-cellulose resuspended in 200 pl homogenization supernatant. After incubation for 10 min on ice, the suspension was centrifuged as before. I'he supernatant containing rhodopsin kinase was carefully removed without taking any DEAE-cellulose fibers. Table 1. Purification of rhodopsin kinase. Activity of rhodopsin kinase in diffcrent fractions is given as a percentage of the activity measured by scanning densitometry of rhodopsin bands from a reconstitution assay (Fig. IC). Reconstitution with eye-homogenate supernatant is 100% (note that in DEAE-cellulose and carboxymethyl-cellulose supernatanls, the activity is higher than 100%; see text). The protein amounl of each fraction was determined by the Bradford assay, as described in Materials and Methods. Cm, carboxymcthyl. Supernatant
Homogenate DEAE-cellulose Cm-cellulose
Protein
Activity
Purification
Pg:eYe
%
-fold
6.14 z .33 0.16
100 145 143
1 6.7 55
This supernatant can be stored at --70°C for at least 3 months without significant loss of activity. A similar procedure was previously described to inactivate rhodopsin kinase of bovine rod outer segment [29] Further purification was carried out on a 1-ml carboxymethyl-cellulose column. The DEAE-cellulose supernatant equivalent of 200 eyes (25 pg protein in 400 pl) was loaded onto the carboxymethyl-cellulose column equilibrated with homogenization buffer. The column was washed with 5 ml homogenization buffer, and rhodopsin kinase was eluted by stepwise addition of 1 ml homogenization buffer containing 0.1 M, 0.3 M, 0.5 M and 1 M NaC1. Fractions of 800 pl were collected. Reconstitution of rhodopsin kinase activity Urea-treated membrane substrate (3 pg protein) in 10 pl and rhodopsin kinase in 10 pl DEAE-cellulose or 20 pl carboxymethyl-cellulose fraction containing (1.5 pg or 0.1 pg protein, respectively) as indicated, were mixed and illuminated for 30 s at 20°C with blue light. The reaction was initiated by addition of 25 pl reaction mixture containing 12 mM phosphate, pH 7.5, 5 mM MgS04, 18 pM ATP, 3 pCi [ Y - ~ ~ P I A T P and 3.6 mM EGTA, then incubated at 20°C. The reaction was terminated at the indicated times by addition of 11 111 SDSi PAGE loading buffer. Peptide synthesis and phosphorylation The peptide Ac-Arg-Arg-Arg-Ala-Ala-Ala-Ala-Ala-SerGlu-Glu-Glu-NH2, was synthesized on a Applied Biosystems 430A synthesizer. N-Butyloxycarbonyl amino acid were used for protection of the a-amino group. Side-chain protection was 0-benzyl for serine and glutamic acid and p-toluensulfonyl for arginine. Each arginine residue was introduced by a double-coupling cycle.
1
2
3
4
5
6
7
8
9
Fig. 1. Purification of rhodopsinkinase. (A) Coomassie-blue-stained 10% SDS/PAGE of homogcnate supkrnatant in lanc 1, and the supernatant from the DEAE-cellulose-purification step (as described in Materials and Methods) in lane 2, each containing an cquivalcnt amount of protein derived from six Muscu eyes. (B) Autoradiogram of reconstitution assay containing 10 pl homogenate supernatant or DEAE-cellulose fraction, and 20-pl carboxymethyl-cellulose column fractions, with 10 p1 urea-treated membranes and 25 p1 reaction mixture, as described in Materials and Methods. In each reaction, the salt conccntration was adjusted to 0.5 M NaCl. Lane 1, urea-trcated membranes alone. Other lanes also contain the following: 2, homogenization supernatant; 3, DEAE-cellulose supernatant; 4, fraction of the loading of DEAE-cellulose fraction containing rhodopsin kinase; 5 , the first fraction of the wash; 6, fraction from elution with 0.1 M NaC1; 7, fraction from elution with 0.3 M NaCl; 8, fraction from elution with 0.5 M NaC1; 9, fraction from clution with 1 M NaC1. (C) Quantitalion or rhodopsin kinase activity by densitometry of the rhodopsin lanes from (B). The percentage activity is written above each column.
1037
0
B
20
40
60
80
100
time (min)
Fig. 3. Peptide phosphorylation by rhodopsin kinase. Phosphate incorporation by rhodopsin kinase in the presence ( O ) ,or absence (0),of pcptidc substrate is shown. The reaction contained 1 mM peptide. 200 pM ATP, 3 pg protein from the DEAE-cellulose fraction of rhodopsin kinase in 200 pl, at 30"C, as described in Materials and Methods. At each time, 40 pl reaction was absorbed onto 2 cm x 2 cm P-81 papers, and immersed immediately in 0.6% phosphoric acid. The radioactivity on the P-81 papers was detcrmined by measuring Cherenkov radiation.
Protein quantitation
0
2
4
6
8
10
12
time (min)
Fig. 2. Time course of rhodopsin kinase activity. Assay of rhodopsin kinase was carried out as described in Materials and Methods with a carboxymethyl-cellulose fraction of rhodopsin kinase a t 100 pM ATP. (A) Autoradiogram of phosphorylated proteins in the rhodopsin kinase assay, separated by SDSIPAGE. Rhodopsin is indicated by an arrow. Lanc 1, urea-treated membranes phosphorylated for 10 min without addition of rhodopsin kinase; lanes 2 - 5, reconstitution of rhodopsin kinase activity, incubated for 1 , 2 , 5 and 10 min, respectively. (B) Quantitation o r the rhodopsin kinase activity shown in (A). Each rhodopsin band was cut from SDS/PAGE, and Cherenkov radiation determined.
Phosphorylation of the pcptide was carried out as described in [27], with the following modifications: the incubation was carried out in the reaction mixture described above, containing 1 mM peptide, 200 pM ATP and 3 pg protein from a DEAE-cellulose fraction of rhodopsin kinase.
SDS/PAGE This was performed according to Laemmli [30], using 1096 acrylamide gels in an SE-600 Hoeffer apparatus. Dried gels were autoradiographed at - 70°C using Agfa Curix film. Quantitation was performed by cutting the bands corresponding to rhodopsin from dry gels and measuring Cherenkov radiation, or by scanning densitometry of the rhodopsin band rrom autoradiographs with a Bio Instruments model SLTKFF scanning densitometer.
Protein quantitation was carried out in 96-well microplates, taking 5O-pl samples and 200 p1 Bradford reagent. The results were read with 595 filter on a Microplate EL309 reader. Bradford reagent was prepared according to [31]. Illumination with lights of different wavelength An optical bench, with a quartz iodide lamp of 100 W and 12 V, was used. An orange-edge filter (Schott OG-590) or a blue-light filter (Schott BG-28) and two KG3 heat-reducing filtcrs from Schott were placed in the light path. The light was focused and delivered to the tube containing the sample via a light guide (4 mm, Schott). The energy of the unattenuated orange or blue light, used at the level of the sample, was 12.7 mW/cm2 or 2.6 mW/cm2, respectively.
RESULTS AND DISCUSSION Purification of rhodopsin kinase was based on the fact that rhodopsin kinase is a basic protein. It dose not bind at all to DEAE-cellulose (in contrast to most of the other proteins of fly eye). Purification of rhodopsin kinase from the soluble fraction of orange-illuminated eye homogenate, by batch adsorption to DEAE-cellulose, resulted in a sevenfold increase in specific activity of rhodopsin kinase (Table 1, Fig. 1A). Further purification of rhodopsin kinase was carried out on a 1-ml carboxymethyl-cellulose column. All rhodopsin kinase activity of the DEAE-cellulose supernatant was absorbed to the carboxymethyl-cellulose column. The column was washed with 5 ml homogenization buffer, and rhodopsin kinase was eluted by stepwise wash with increasing NaCl concentrations in homogenization buffer (see Fig. 1 and Materials and Methods). The overall purification of these two steps was 55[old, but the rhodopsin kinase fraction still contained several
1038
0.0 -I 0
1
2
IlATP ( p M
8000
E
Q
.-0
3
)
Fig. 5. Double-reciprocal plot of rhodopsin kinase activity in reconstitution assay using different concentrations of ATP. Rhodopsin kinase activity was determined by a 2.S-min assay, as described under Materials and Methods with a DEAE-ccllulose fraction of rhodopsin kinase. The reactions were separated by SDSiPAGE; bands corresponding to rhodopsin were cut out and their Cherenkov radiation determincd. At low concentration of ATP (0.5 pM, 0.75 pM). thc reaction volume was 100 pl, keeping the same concentration of all other components except those or rhodopsin kinase and membrane substrate. The rcsults are means of five experirncnts.
B
-
-'
6000
c
E
P 0
.C
-
4000
0
mQ 8 c 0 2000
a N
m
0 0
1
2
3
No.
4
5
6
of wash
Fig. 4. Association of rhodopsin kinase with eye membranes containing either metarhodopsin or rhodopsin. Eye-membranc suspensions wcrc illuminated for 30 s with either blue or orange light. After each illumination, membranes were precipitated by centrifugation (14000 g for 15 min at 4°C). The supernatant was discardcd and membrancs were resuspended in fresh homogenization buffer. Aliquots of 10 pl were 500 removed for rhodopsin kinasc assay of the mcmbranc bound kinasc 0 2 4 6 8 10 12 after each successive centrifugation. (A) Autoradiogram of rhodopsin kinase activity of a 5-min assay associated with the membrane. MemHeparin ( Lr g h l ) branes washed 1- 6 times arc shown in lanes 1 and 7 , 2 and 8 , 3 and Fig. 6. Inhibition of rhodopsin kinase activity by heparin. Rhodopsin 9 , 4 and 10, 5 and 11 and 6 and 12, respectively. Lanes 1 - 6 represent membranes illuminated with blue light, lanes 7 - 12 represent mem- kiriase reconstitution assay was performed as dcscribed in Matcrials and Methods, using a DEAE-cellulose fraction of rhodopsin kinase branes illurninatcd with orange light. The rhodopsin band is indicated and diffcrcnt concentrations of heparin, as indicated. The reactions by an arrow. (€3) Quantitation of rhodopsin kinase activity was carried out as in Fig. 2R ( 0 ) .Blue-light illumination between washes; (A) were separated by SDSjPAGE; bands corresponding to rhodopsin were cut out and Cherenkov radiation was determined. orange-light illumination between washes.
proteins, as seen in Coomassie-blue-stained and silver-stained gels (not shown). The activity of rhodopsin kinase was monitored by a reconstitution assay of each fraction with urea-treated eye membranes (Fig. l B , C). As observed, the activity in DEAE-cellulose and carboxymethyl-cellulose fractions is higher than the activity in the unpurified supernatant (Fig. lB, Table 1). We suppose that DEAE-cellulose absorbs enzymes that inhibit the activity of rhodopsin kinase, like G-protein o r arrestin which can compete for binding to rhodopsin, or phosphatases. The activity of the fractions from the carboxymethyl-cellulose column show sharp elution of rhodopsin kinase in the elution volume containing 0.5 M NaC1.
Urea treatment of Musca eye membranes decreased the endogenous phosphorylation activity of membranes containing metarhodopsin by more than 90% (Fig. 2A). Routinely, urea-treated membranes were used as substrate in reconstitution assays of rhodopsin kinase activity. A time course of rhodopsin kinase activity at 100 FM ATP, in a reconstitution assay with carboxymethyl-cellulose fraction of rhodopsin kinase, is shown in Fig. 2A. Phosphorylation was stringently dependent on conversion of rhodopsin to metarhodopsin, and maximum phosphorylation was obtained in blue-illuminated membranes in the presence of EGTA (not shown) [ 3 ] .The first lane of Fig. 2A shows the background activity of urea-washed membranes. The amount of rhodopsin phosphorylation on
1039 Rhodopsin phosphorylation 3000
A
Histone phosphorylation
woo
B
5000
4000
3000
zoo0 1000
0 0
Time (min)
10
20
30
Time (min)
Fig. 7. Enzyme and substrate specificity of rhodopsin kinase activity. Phosphorylation of rhodopsin (A) and histone (B) by extract of Muscu eye, brain or abdomen. (A) Reconstitution assay of rhodopsin phosphorylation was performed as described in Materials and Methods with equivalent amounts of protein from extract of MUSCU brain, abdomen and eye. The equivalent amount of eye extract was purified with DEAEcellulose as described in Materials and Methods. Thc rhodopsin bands were cut out and the Cherenkov radiation was measured. (B) Kinase assay using histone. Each assay system contained 40 pg protein from Muscu brain, abdomen or cye extract. 40 pg protein from eye extracl was purified with DEAE-cellulose as described in Materials and Methods. Phosphorylation reactions were as described in rhodopsin kinase assay, except that histone (2 mg/ml) was addcd in place of eye membranes. The reaction volume was 100 fi/assay. At the indicated times, aliquots were spotted onto 2 cm x 2 cm 3MiM Whatman paper, which was immediately immersed in 10% ice-cold trichloroacctic acid. The papcrs were washed four times with 10% trichloroacetic acid, overnight at 4°C. Cherenkov radiation of the filters was measured. (@) Eye; (m)brain; (A)abdomen.
10-min incubation is negligible, less than the activity after 1 min in the reconstitution system containing the carboxymethyl-cellulose fraction of rhodopsin kinase (Fig. 2A, lane 2). At the end of the 10-min reaction, 60% of the A'TP remained intact. Quantitation of the extent of rhodopsin phosphorylation (Fig. 2B) shows that rhodopsin kinase activity in the reconstitution assay is linear for 5 min then decreases slightly after 5 - 10 min (Fig. 2A, B). Comparing the endogenous rhodopsin kinase activity in urea washed membranes with the activity ofa 10-min reconstitution assay (Fig. IB, cf. lanes 1 and 6), reveals that addition of rhodopsin kinase from the carboxymethyl-cellulose fraction to urea-washed membranes, specifically increases the phosphorylation of rhodopsin. The prominent band that undergoes phosphorylation is a 31-kDa protein. This protein has the same electrophoretic mobility and undergoes light-dependent phosphorylation as the D.rnelunoguster rhodopsin, identified by a monoclonal antibody specific to Drosophila rhodopsin [32]. Similar experiments with less-pure rhodopsin kinase from DEAE-cellulose fraction gave the same results. Further purification on a carboxymethyl-cellulose column did not increase the specificity of phosphorylation in the reconstitution assay (compare lanes 3 and 8 in Fig. 1B). Additional characterization of rhodopsin kinase was carried out with the DEAEcellulose fraction of rhodopsin kinase, because the carboxymethyl-cellulose fractions of rhodopsin kinase were very labile (only fresh fractions were used). Phosphorylation of other protein bands is considerably less affected by the addition of the rhodopsin kinase carboxymeth yl-cellulosc fraction. Phosphorylation of high-molecularmass proteins is seen in all experiments. This activity ir not related to rhodopsin kinase, since phosphorylation of proteins of high molecular mass also appears in urea-washed membranes, to a greater extent than in the reconstitution assays. To show the activity of rhodopsin kinase in the DEAE-cellulose fraction, a peptide resembling the recognition sequence of
bovine rhodopsin kinase [27] was used as substrate. In Fig. 3, the activity of rhodopsin kinase in the DEAE-cellulose fraction toward the peptide substrate is shown. The activity is slow (as in the case of bovine rhodopsin or p-adrenergic receptor kinase) [27], and linear for a long time (95 min in this experiment). To test the interaction between rhodopsin kinase and substrate in native membranes, membranes prepared from either blue- or orange-illuminated eyes were subjected to successive extraction with homogenization buffer followed by centrifugation. The amount of rhodopsin kinase bound to metarliodopsin or rhodopsin-containing membranes was assayed after each repeated centrifugal wash. It is shown that rhodopsin kinase is retained by metarhodopsin-containing membranes (Fig. 4). In contrast, rhodopsin kinase was gradually released from rhodopsin-containing membranes. The amount of rhodopsin kinase released from rhodopsin-containing membranes by each centrifugal wash was much less than one would expect from equilibration of a soluble protein with the aqueous medium. This result indicates that rhodopsin kinase binds tightly to its substrate, metarhodopsin, and partially dissociates from rhodopsin (Fig. 4). However, even under resting conditions, in rhodopsin-containing membranes, a substantial part of rhodopsin kinase is associated with the membranes. Perhaps fly rhodopsin kinase is post-translationally modified, as is bovine rhodopsin kinase which is isoprenylated [33]. Presumably, this localization of rhodopsin kinase enables it to quench immediately the activated form of the photopigment. Preferential interaction of other proteins with metarhodopsin has been described for preparations of bovine rod outer segments [4] and fly rhabdomere [3]. Determination of the K , for ATP using a reconstitution assay with a DEAE-cellulose fraction of rhodopsin kinase gave a value of 9 1.5 pM (Fig. 5). This result is not very difftrent from the K , for ATP of bovine rhodopsin kinase ~31.
1040 REFERENCES Kuhn, H. (1974) Nature 250, 588 590. Kuhn, H. (1978) Biochemistry 17,4389-4395. Bentrop, J. & Paulsen, R. (1986) Eur. J . Biochem. 161, 61 -67. Wilden, U., Hall, S. & Kuhn, H. (1986) Proc. Nut1 Acad. Sci. USA 83, 1174- 1178. 5. Miller, J. L. & Dratz, E. A. (1984) Vision Res. 24, 1509- 1521. 6. Sitaramayya, A. & Liebman, P. A. (1983) J . Bid. Chem. 258, 1205- 1209. 7. Bouvier, M., Hausdorff, W. P., DeBlasi, A., O’Dowd, R. F. Kobilka, B. K., Caron, M. G. & Lefkowitz, R. J. (1988) Xaturr 333, 370 - 373. 8. Roht, N . S., Campbell, P. T., Caron, M. G., Lefkowitz, R. J. & Lohse, M. J. (1991) Proc Natl Acad. Sci. USA 88,6201 -6204. 9. Lohse, M. J., Lefkowitz, R. J., Caron, M. G. & Benovic, J. L. (1989) Proc. Natl Acad. Sci. USA 86,301 1-3015. 10. Hausdorff, W. P., Bouvier, M., O’Dowd, B. F., Irons, G., Caron. M. G. & Lefkowitz, R. J. (1989) J . Bid. Chem. 264, 1265712665. 11. Hausdorff, W. P., Campbell, P. T., Ostrowski, J., Yu, S. S., Caron, M. G. & Lefkowitz, R. J. (1991) Proc. Natl Acad. Sci. USA 88, 2979 - 2983. 12. Huganir, R. L. & Greengard, P. (1990) Neuron 5 , 555 - 567. 13. Hausdorff, W. P., Caron, M. G. & Lefkowitz, R. J. (1990) FASEB J . 4,2881 -2889. 14. Reneke., J. F., Blumer, K. J., Courchesne, W. E. & Thorner, J. (1988) Cell 55, 221 -234. 15. Meier, K. & Klein, C. (1988) Proc. Null Acad. Sci. USA 85, 2181 -2185. 16. Vaughan, R. A. & Devreotes, P. N. (1988) J . Biol. Chem. 263, 14 538 - 14543. 17. Benovic, J. L., DeBlasi, A,, Stone, W. C., Caron, M. G. & Lefkowitz, R. J. (1989) Science 246, 235-240. 18. Lorenz, W., Inglese, J., Palczewski, K., Onorato, J. J., Caron, M. G. & Lefkowitz, R. J. (1991) Proc. Natl Acad. Sci. USA 88, 8715 - 8719. 19. Benovic, J . L.; Onorato, J. J., Arriza, J. L., Stone, W. C., Lohse, M.; Jenkins, N. A., Gilbert, D. J., Copeland, ?I. G., Caron, M. G. & Lefkowitz, R. J. (3991)J. Biol. Chem. 266,14939- 14946. 20. Strasser, R. €I., Benovic, J. L., Caron, M. G. & Lefkowitz, R. J . (1986) Proc. Natl Acad. Sci. USA 83,6362-6366. 21. Mayor, F. Jr, Benovic, J. L., Caron, M. G. & Lefkowitz, R. J. (1989) Science 246, 235-240. 22. Chuang, T. T., Sallese, M., Ambrosini, G., Parruti, G. & DeBlasi, A. (1 992) J. B i d . Chem. 267,6886 - 6892. 23. Palczewski, K., McDowell, J. H. d Hargrave, P. A. (1Y88) J . Biol. Chem. 263,14067-14073. 24. Palczewski, K., Kahn, N. & Hargrave, P. A. (1990) Biochemistry 29, 6276 - 6282. 25. Palczewski, K., Buczylko, J., Kaplan, M. W., Polans, A. S. & Crabb, J. W. (1991) J. Biol. Chem. 266,12949-12955. 26. Kelleher, D. J. & Jhonson, G. L. (1990) J . Biol. Chem. 265,26322639. 27. Onorato, J. J., Palczewski, K., Regan, J. W., Caron, M. G., Lefkowitz, R. J. & Benovic, J. L. (1991) Biochemlrtry 30,5118 5125. 28. Blumenfeld, A., Erusalimsky, J., Heichal, O., Selinger, Z. & Minke, B. (1985) Proc. Nut1 Acad. Sci. USA 82, 7116-7120, 29. Shichi, H. & Somers, R. L. (1978) J . Biol. Chem. 253, 70407046. 30. Lacmmli, U. K. (1970) Nature 227, 680-685. 31. Bradrord, M. M. (1976) Anal. Biuchem. 72, 248-254. 32. Couet, H. & Tanimura, T. (1987) J. Cell Biol. 44, 50- 56. 33. Inglese, J., Glickman, J . F., Lorenz, W., Caron, G. M. & Lefkowitz, R. J. (1992) .J. Biol. Chem. 267, 1422-1425. 34. Lohse, M. J., Benovic, J. L., Caron, M. G. & Lefkowitz, R. J. (lYY0) J . Biol. Chem. 265, 3202-3209. 1. 2. 3. 4.
Fig. 8. Musca rhodopsin kinase activity with rhodopsin from different fly species. Eye membranes from Muscu, Luciliu, and head membranes from Drosoplzila were prepared as described in Materials and Methods. The reconstitution assay contained a DEAE-cellulose fraction of rhodopsin kinase from five Musca eyes and urea-washed membranes of five Musca eyes (lane 2), four Lucilia eyes (lane 4), and 20 Drosophila heads (lane 6). As a control, phosphorylation of an cqual amount of urea-washed membranes alone, from each fly, is shown. Musca membranes in lane 1; Lucilia membranes in lane 3; Drosophila membranes in lane 5.
Heparin, an inhibitor of the fl-adrenergic receptor kinase [34], was tested in the reconstitution assay with DEAE-cellulose fraction of rhodopsin kinase, and was found to be a potent inhibitor of fly rhodopsin kinase with a median effective concentration of 2.5 pg/ml (Fig. 6). A protein inhibitor of CAMP-dependent kinase had no effect (data not shown). To test the specificity of rhodopsin kinase activity toward rhodopsin, extracts of different fly tissues were assayed using urea-treated membranes with histone as a general substrate. The DEAE-cellulose fraction of partially purified rhodopsin kinase has high activity with metarhodopsin (Fig. 7A), while it is almost inactive toward histone (Fig. 7B). In contrast, Musca brain cytosol contains high kinase activity with histone and has little activity with metarhodopsin. While the abdomen extract has lower kinase activity than brain, it also preferentially phosphorylates histone over metarhodopsin. These results indicate that rhodopsin kinase is specifically expressed in fly eye, while other Musca tissues lack rhodopsin kinase activity. Furthermore, metarhodopsin is poorly phosphorylated by kinases from other fly tissues that are highly active toward histone. It is concluded that the fly rhodopsin kinase is selective for metarhodopsin, like bovine rhodopsin kinase and 8-adrenergic receptor kinase which are selective for their corresponding activated receptors [2, 141. To test whether rhodopsin kinase also recognizes rhodopsin of other photoreceptors, the activity of Musca rhodopsin kinase was tested with several other fly metarhodopsins. As seen in Fig. 8, Musca rhodopsin kinase effectively phosphorylates Lucilia and Drosophila metarhodopsins. The resolution and reconstitution of rhodopsin kinase and rhodopsin can be used in analysis of the functional consequences of rhodopsin phosphorylation and whether other kinases take part in this modification. The large number of Drosophilu mutants should make it possible to test the role of phosphorylation in termination of light-dependent signal transduction. We thank Dr D. Blest, L)r S. Stowe and Dr S. Trowel1 for the R16 Drosophila opsin monoclonal antibodies which were prepared by Dr H. Couet and Dr T. Tanimurd in the laboratory of Dr n.Blest. This work was supported by Grants from the National Institute of Health (EY03529) and the US-Israel Binational Science Foundation (BSF).
~
