Accelerated Publications Phosphorylation Stabilizes the Active Conformation of Rhodopsin†

Deactivation of many G protein coupled receptors (GPCRs) is now known to require phosphorylation of the activated receptor. The first such GPCR so analyzed was rhodopsin, which upon light activation forms an intramolecular equilibrium between the two conformers, metarhodopsin I and II (MI and MII). In this study, we find surprisingly that rhodopsin phosphorylation increases rather than diminishes the formation of MII, the conformation that activates G protein. The MI-MII equilibrium constant was progressively shifted toward MII as the experimental phosphorylation stoichiometry was increased from 0 to 6.4 phosphates per rhodopsin. Increasing phosphorylation both increased MII’s formation rate (k1) and decreased its rate of loss (k-1). The direct effect of cytoplasmic surface phosphorylation on intramolecular conformer equilibria observed here may be important to functional state modulation of other membrane proteins. G protein coupled receptors (GPCRs)1 are thought to exist in several intramolecular equilibrium configurations, one of which is selected by agonist binding: this state activates G proteins. The activity and lifetime of activated receptors are regulated by receptor phosphorylation (1-4), a reaction catalyzed by G protein coupled receptor kinases (GRKs) (5-7). Rhodopsin is perhaps the best-understood GPCR. Its active site contains the chromophore, 11-cis-retinal, which acts as resting state receptor antagonist (8, 9). Absorbed light isomerizes 11-cisto all-trans-retinal, the agonist for rhodopsin activation. The receptor then relaxes into a longlived pH and temperature-dependent equilibrium between the two spectroscopic species, metarhodopsin I and II (MI and MII). MII (λmax 387 nm), the dominant species present at body temperature, is the protein conformation responsible for activation of the retinal G protein, transducin (10-12). At lower temperatures or higher pH, the equilibrium is shifted toward MI (λmax 480 nm), a form that does not activate G protein. The remarkable change in light absorption properties accompanying formation of MI and MII provides a natural spectroscopic probe that has been used to monitor the distribution between these states of the receptor during their formation and subsequent dissipation. Activated rhodopsin is phosphorylated on C-terminal serines and threonines by rhodopsin kinase (GRK1) (1315). Phosphorylation is essential to rapid quenching of the visual transduction cascade (1, 2, 5). High phosphorylation stoichiometries, up to nine phosphates per rhodopsin found in vitro (16), can directly terminate signaling (17) or can be augmented by arrestin binding at lower ratios (18). Multiple † Supported by NIH Grants EY00012, EY01583, and EY07035. * Address correspondence to this author. Telephone: 215-898-6917. FAX: 215-898-4217. Email: [email protected]. 1 Abbreviations: GPCR, G protein coupled receptor; GRK, G protein receptor kinase; λmax, maximum absorbance wavelength; MI, metarhodopsin I; MII, metarhodopsin II; MIII, metarhodopsin III; RDM, rod disk membrane; EDTA, ethylenediaminetetraacetic acid; NADPH, â-nicotinamide adenine dinucleotide phosphate, reduced form; DTT, dithiothreitol; OD, optical density; POPC, 1-palmitoyl-2-oleoyl-snglycero-3 phosphocholine; PC, phosphatidylcholine; PS, phosphatidylserine. © Copyright 1998 by the American Chemical Society Volume 37, Number 33 August 18, 1998 S0006-2960(98)00933-7 CCC: $15.00 © 1998 American Chemical Society Published on Web 07/29/1998 phosphorylation is a candidate mechanism for explaining graded turnoff of activated receptors (19). In this report, we show that rhodopsin phosphorylation also modulates the MI-MII equilibrium constant itself, unexpectedly increasing the amount of MII formed with increasing receptor phosphorylation. Companion kinetic analysis showed phosphorylation to both increase the rate of MII formation (k1) and decrease the back rate from MII to MI (k-1). This effect might be important in accelerating multiple (cooperative) phosphorylation by enriching MII, favoring early binding of arrestin to phosphorylated MII, or enhancing loss of activity through more MIII formation at higher bleach levels. EXPERIMENTAL PROCEDURES Preparation of Rod Disk Membranes. Retinas were dissected under infrared light from bovine eyes freshly obtained from a local slaughterhouse (MOPAC; Souderton, PA). Rod disk membranes (RDM) were purified by sucrose density gradient centrifugation as previously described (20). Rhodopsin concentration was determined from the absorption spectrum of each RDM sample using the Dartnall correction for light scattering (21).2 Preparation of Phosphorylated Rhodopsin. Rhodopsin was phosphorylated according to the method of Wilden and Kühn (16). Briefly, RDM samples containing 12.5 μM rhodopsin, 3 mM ATP, 1 mM MgCl2, 1 mM DTT, and 100 mM NaH2PO4, pH 7.4, were phosphorylated in 12 × 75 mm optically transparent polystyrene tubes (Falcon; Lincoln Park, NJ) held in a 30 °C water bath. Phosphorylation was initiated by exposure to a 60 W tungsten lamp, placed 20 cm from the tubes. Reaction tubes were frequently agitated to ensure uniform illumination. Control, unphosphorylated samples were treated identically except for the absence of ATP. One identical RDM phosphorylation tube containing [γ-32P]ATP at 2.7 μCi/μmol (Dupont NEN; Boston, MA) was used to monitor phosphate incorporation progress and stoichiometry. At selected times, 100 μL aliquots were removed from it and transferred to 700 μL of ice-cold quenching solution (100 mM, pH 7.4, NaH2PO4, 20 mM EDTA). RDM were then pelleted by centrifugation, washed twice with 700 μL of 100 mM NaH2PO4, pH 7.4, and either counted in a Quick-Count benchtop radioisotope counter (Bioscan; Washington, DC) or solubilized in Ecolume Scintillation Cocktail (ICN; Costa Mesa, CA) and later counted in a liquid scintillation counter (Intertechnique; Fairfield, NJ). Nonradioactive reaction samples were stopped similarly and used in the spectroscopic experiments after pigment regeneration described below. Phosphorylation stoichiometry was calculated by dividing the moles of phosphate incorporated by the moles of rhodopsin. Phosphorylation stoichiometry was varied by increasing the incubation time from 0 to 160 min. Bleached phosphorylated and unphosphorylated control samples were regenerated on ice with a 3-fold excess of 11cis-retinal (a generous gift from R. K. Crouch and the National Eye Institute) for at least 12 h of dark incubation followed by a further 1 h incubation at 30 °C to ensure completion. Before determination of the regenerated rhodopsin concentration, the excess 11-cis-retinal (λmax 380 nm) was reduced to 11-cis-retinol (λmax ∼320 nm) by incubating with NADPH in 10-fold excess over the starting rhodopsin concentration in the presence of native retinal reductase for 1 h at 30 °C. Regeneration was confirmed to be 100 ( 5% complete by absorption spectroscopy. The phosphorylated, regenerated RDM were then hypotonically stripped to remove peripheral proteins (5). The final pellet was suspended in 10 mM KH2PO4, pH 7.0, 0.1 mM EDTA, 1 mM DTT, and 100 mM KCl to form a stock of RDM at approximately 250 μM rhodopsin. Stocks were purged with argon and held on ice in dark storage containers. Aliquots were diluted from the stock concentration for spectroscopy. Incremental Bleach Methodology. An incremental rhodopsin bleaching method (22, 23) was used to determine the MI-MII equilibrium constant and the fraction of rhodopsin bleached in RDM suspensions at several phosphorylation levels. Briefly, 19 serial light flashes of equal intensity were used to bleach 3-4% of the remaining rhodopsin per flash at 50 s intervals. MII formation was monitored by kinetically measuring the increase in absorbance at 390 nm while subtracting the accompanying light scattering change (24) at the nearby MI-MII isosbestic wavelength (426 nm). At pH 7.0, this resulted in an upward staircase of absorbance increments, individual steps diminishing exponentially in amplitude, analogous to radioactive decay, as the remaining unbleached rhodopsin was depleted with each flash (Figure 1). All spectral data were acquired using an SLM/Aminco DW2000 dual-wavelength spectrophotometer (Urbana, IL) equipped with an EG&G Electrooptics (Salem, MA) xenon flash unit (FX-199 tube, PS-302 power supply set at 500 V, with a 7 μF external capacitor) and a thermally jacketed cuvette holder connected to a constant temperature water circulator. Optical filters were used to limit the bleaching flash to wavelengths between 420 and 680 nm. Experimental samples contained about 10 μM rhodopsin (assayed exactly for each experimental run) in 10 mM KH2PO4, pH 7.00, 100 mM KCl, and 0.01 mM EDTA. Substitution of MOPS [3-(N-morpholino)propanesulfonic acid] for phosphate buffer did not alter the results. The measurements were made at 0.5 °C to minimize slow conversion of MII into MIII during the course of an experiment (see below). Determination of MI-MII Equilibrium Constant and Fraction Bleached. The reactions studied are depicted in Scheme 1: The MI and MII concentrations that occur after each flash and their Keq value can be determined as follows: if [Rh]0 is the initial rhodopsin concentration, f the fraction of rhodopsin bleached per flash, [Rh*] the concentration of bleached rhodopsin, and re a correction for photoregeneration, the fraction of rhodopsin activated by the nth flash is (23) 2 [Rh] ) [1.10(OD500) 0.77(OD600) 0.33(OD400)]/40000 cm-1 M-1. Scheme 1 Rh98 light MI y\z k1 k-1 MII 11394 Biochemistry, Vol. 37, No. 33, 1998 Accelerated Publications