HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 46, 43987-43996, November 15, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/46/43987    most recent M206211200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schröder, K. Articles by Hofmann, K. P. Search for Related Content PubMed PubMed Citation Articles by Schröder, K. Articles by Hofmann, K. P. Social Bookmarking                      What's this?

Arrestin and Its Splice Variant Arr^1-370A (p⁴⁴)

MECHANISM AND BIOLOGICAL ROLE OF THEIR INTERACTION WITH RHODOPSIN^*

Katrin Schröder, Alexander Pulvermüller, and Klaus Peter Hofmann

From the Institut für Medizinische Physik und Biophysik, Humboldt-Universität zu Berlin, Universitätsklinikum Charité, Schumannstrasse 20-21, Berlin 10098, Germany

Received for publication, June 21, 2002, and in revised form, August 13, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Deactivation of G-protein-coupled receptors relies on a timely blockade by arrestin. However, under dim light conditions, virtually all arrestin is in the rod inner segment, and the splice variant p⁴⁴ (Arr^1-370A) is the stop protein responsible for receptor deactivation. Using size exclusion chromatography and biophysical assays for membrane-bound protein-protein interaction, membrane binding, and G-protein activation, we have investigated the interactions of Arr^1-370A and proteolytically truncated Arr^3-367 with rhodopsin. We find that these short arrestins do not only interact with the phosphorylated active receptor but also with inactive phosphorylated rhodopsin or opsin in membranes or solution. Because of the latter interaction they are not soluble (like arrestin) but membrane-bound in the dark. Upon photoexcitation, Arr^3-367 and Arr^1-370A interact with prephosphorylated rhodopsin faster than arrestin and start to quench G_t activation on a subsecond time scale. The data indicate that in the course of rhodopsin deactivation, Arr^1-370A is handed over from inactive to active phosphorylated rhodopsin. This mechanism could provide a new aspect of receptor shutoff in the single photon operating range of the rod cell.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Arrestins are involved in the regulation of numerous signal transduction pathways through G-protein¹-coupled receptors. Arrestins bind tightly to the receptors when they are activated by chemical (e.g. diffusible ligands such as hormones) or physical (such as light) stimuli and phosphorylated through G-protein-coupled receptor kinases (1). Catalytic interaction with the G-protein, the primary transduction event, is thereby terminated.

Signal transduction in vertebrate rod cells (2, 3) starts with the light-induced formation of active rhodopsin (R*), which interacts with the G-protein (G_t) and catalyzes nucleotide exchange in the G_t-subunit. In its GTP-bound form, G_t activates its effector cGMP phosphodiesterase, which in turn hydrolyzes cGMP to 5'-GMP, leading to the closure of the cGMP-gated cation channels in the plasma membrane (4, 5). The specific R* conformation that is required for the interaction with G_t can only develop when the photoproduct MII is formed before. The MII state arises in turn from the replacement, by light-induced isomerization, of rhodopsins covalently attached antagonist 11-cis-retinal with all-trans-retinal, and subsequent relaxation and proton transfer reactions. There are analogies between the MII state and the high affinity states known from other G-protein-coupled receptors (6).

Interaction with visual arrestin requires not only the MII conformation of rhodopsin (7) but also the presence of phosphate groups at C-terminal sites (see Ref. 8). Phosphorylation is mediated by rhodopsin kinase (1, 9, 10), a member of the GRK1 family of G-protein-coupled receptor kinases. The subsequent binding of arrestin deactivates the transduction cascade by direct competition with the G-protein (11, 12). In contrast to the ubiquitously expressed -arrestins, visual arrestins are exclusively expressed in rod and cone photoreceptor cells of the vertebrate retina. Evidence has been presented that proper termination of the light signal depends crucially on a conformational switch in arrestin (7, 13), which is operated by the contact with the phosphorylated C terminus of the receptor (7, 8, 14-17) and controlled by the arrestin C terminus (13, 18). This mechanism may enhance the specificity and strength of interaction, but even more importantly, it serves to avoid interference of arrestin with G-protein activation before rhodopsin kinase-catalyzed phosphorylation of the active receptor has occurred. It thus leaves a time window for fast undisturbed G_t activation, in which arrestin cannot interfere with the G-protein because it has very low if any affinity to nonphosphorylated rhodopsin.

In view of this well established and sensible mechanism, it is surprising that short variants of arrestin are present in the rod cell which interact with both phosphorylated and nonphosphorylated forms of R* (pR* and R*, respectively (17, 19)). Bovine rods express a splice variant of arrestin, p⁴⁴ (Arr^1-370A), in which the last 35 amino acids are replaced by a single alanine (20). Other variants may arise from proteolytic truncation, such as the protein resulting from calpain proteolysis in vitro (21). The splice variant (Arr^1-370A) is present at 10% the amount of the full-length arrestin and thus at 1% of rhodopsin. Arr^1-370A is partially preactivated, and has even some affinity to membranes that contain the inactive prephosphorylated receptor (pR). The activation-phosphorylation scheme of interaction does therefore not apply, and the conformational switch appears to be lacking in Arr^1-370A. Intriguingly however, available evidence argues for Arr^1-370A, and not for full-length arrestin, as the actual stop protein that terminates signal transduction at low levels of light excitation, i.e. in the actual working range of the rod cell. Langlois and co-workers (22) performed time-resolved measurements of effector activity in a calorimetric phosphodiesterase assay. Pulses of phosphodiesterase activity evoked by small flashes of light were shortened by both arrestin and Arr^1-370A, but Arr^1-370A was five times more efficient than full-length arrestin. To resolve this apparent discrepancy, we have investigated Arr^1-370A further, in comparison with full-length arrestin, on purified preparations. We also draw on proteolytic forms of arrestin, which are available in quantity and were identified by mass spectroscopic analysis as Arr^3-382 and Arr^3-367, respectively (19). Arr^3-382 interacts like native arrestin only with pR*, whereas Arr^3-367 and Arr^1-370A interact with both R* and pR*. Moreover, Arr^3-367 binds like Arr^1-370A to phosphorylated membranes regardless of the bleaching status (19).

Using size exclusion chromatography, centrifugation, kinetic light scattering, and extra-MII experiments, we will specifically analyze membrane binding, receptor interactions, and influence on catalytic G_t activation of these arrestins. The results are relevant for the mechanism of receptor interaction and give a more precise pattern of the binding interactions in the different modes of signal transduction in the rod cell. We will substantiate the notion put forward by others (19, 22) that p⁴⁴ (Arr^1-370A) is the stop protein for signal transduction in the single quantum regime of rod operation, whereas full-length arrestin comes into play when, in bright light, a large amount of stop protein is needed.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- All chemicals were purchased from Merck, Roche Molecular Biochemicals, or Sigma. Radioactive [-³²P]ATP was purchased from PerkinElmer Life Sciences. 11-cis-Retinal was generously provided by Dr. R. K. Crouch, Medical University of South Carolina.

Bovine rod outer segments were isolated under dim red illumination from fresh, dark-adapted bovine retinas obtained from a local slaughterhouse using the discontinuous sucrose gradient method (23). Rhodopsin was prepared by removing the soluble and membrane-associated proteins from the disc membrane by repetitive washes with a low ionic strength buffer (24). Phosphorylated opsin was prepared from washed disc membranes as described previously by Wilden and Kühn (25). To remove retinaloxime from the membrane-bound phosphorylated opsin, the membranes were treated with urea and fatty acid-free bovine serum albumin (26). An average stoichiometry of ~1.5 phosphates/opsin was determined using radioactive [-³²P]ATP as a tracer.

Phosphorylated rhodopsin was prepared by regeneration of phosphorylated opsin with 11-cis-retinal (27). Phosphorylated opsin was suspended in 10 mM BTP (pH 7.5) containing 100 mM NaCl. A 3-fold molar excess of 11-cis-retinal was added in the dark to the sample, followed by incubation for 1 h at room temperature and then overnight at 4 °C. After regeneration, phosphorylated membranes were centrifuged (45,000 × g for 20 min) and washed four times with 10 mM BTP (pH 7.5) containing 100 mM NaCl to remove excess 11-cis-retinal. The concentration of rhodopsin and phosphorylated rhodopsin was determined spectrophotometrically at 498 nm (17). The membranes, containing rhodopsin and phosphorylated rhodopsin, were stored at 80 °C until use.

Solubilized rhodopsin and opsin in their phosphorylated and nonphosphorylated forms were prepared by solubilizing the respective membranes with dodecyl maltoside (3% w/v, final concentration) and purified by affinity chromatography using concanavalin A (28).

Arrestin was purified from frozen dark-adapted bovine retinas as described (29, 30). Purified arrestin was determined spectrophotometrically at 278 nm, assuming a molar absorption coefficient of E = 0.638 (31) and a molecular mass of 45,300 Da.

Arr^1-370A was isolated under dim red light from bovine rod outer segments as described (17). Purified Arr^1-370A was quantified as described above for arrestin purification.

Proteolytic forms of arrestin, Arr^3-382, and Arr^3-367 were isolated and quantified as described by Palczewski et al. (19). Arrestin was diluted in 100 mM BTP (pH 7.5) containing 0.1 mM CaCl₂ and 1 mM dithiothreitol and digested with trypsin (200:1, 20-21 °C) for 10 min. Proteolysis was stopped with 10-fold excess of trypsin inhibitor to added trypsin, and the arrestin fragments were applied onto a TSK-heparin steel column (Tosohaas; 0.75 × 7.5 cm, 10-µm particle size, 5-ml bed volume, equilibrated with 10 mM BTP (pH 8.4), flow 0.1 ml/min) and incubated on the column for 6 h. The column was washed with 10 mM BTP (pH 8.4) before being eluted with an NaCl gradient (0-1 M) in the same buffer. Concentrations of the fragments were determined as described for the arrestin purification.

Purity of the preparations of all arrestin variants was analyzed by SDS-PAGE (see Fig. 2A).

Transducin was purified from frozen dark-adapted bovine retinas (32). Purified transducin (G_t) concentration was determined using the Bradford method (33).

Centrifugation Assay-- The binding of arrestin, Arr^1-370A (p⁴⁴) and the proteolytic form Arr^3-367 to membrane suspensions of opsin, p-opsin, rhodopsin, and p-rhodopsin was determined using a centrifugation assay (34). Samples (2 µM arrestins and 5 µM receptors) were incubated in 10 mM BTP (pH 7.0) containing 130 mM NaCl and 1 mM MgCl₂. 100-µl aliquots of these samples were either kept in the dark or illuminated with a 150-watt fiberoptic light source filtered through a heat filter (Schott KG2) and a 495-nm long pass filter for 20 min on ice and pelleted by ultracentrifugation (45 min; 84,400 × g; 4 °C). After removal of the supernatant, the pellet was resuspended in 100 µl of buffer. The amount of arrestin and Arr^3-367 either bound to the membrane pellet or present in the supernatant was analyzed by densitometry on Coomassie Blue-stained SDS-PAGE. All pellet samples were heated to 95 °C for 10 min in the presence of SDS to aggregate most of rhodopsin.

Size Exclusion Chromatography-- Size exclusion chromatography was used to characterize membrane-independent, direct complex formation between the different forms of the solubilized receptor (opsin, p-opsin, rhodopsin, and p-rhodopsin) and arrestin and its proteolytic form Arr^3-367. 10 or 5 µg of each arrestin and 10 µg of each rhodopsin or opsin (freshly prepared) were incubated in buffer containing 10 mM BTP (pH 7.0) containing 130 mM NaCl, 1 mM MgCl₂, and 0.02% dodecyl maltoside for 5 min at room temperature. As controls, all samples (arrestin, Arr^3-367, opsin, p-opsin, rhodopsin, and p-rhodopsin) were incubated alone. The reaction mixtures were loaded on a Superose TM 12 column (Amersham Biosciences), equilibrated with buffer, and analyzed on a Smart System (Amersham Biosciences; flow rate, 40 µl/min), monitoring the elution by the absorbance at 280 nm.

UV/Visual Spectroscopy-- Formation of the photoproduct MII (_max = 380 nm) was assayed using the two-wavelength technique (7, 35). This technique minimizes scattering artifacts by comparing the flash-induced changes in the absorbance at 380 and 417 nm. The absorbance change at 417 nm (MI isosbestic to MII) serves as a reference for determining the level of MII. The two-wavelength spectrophotometer (UV 3000, Shimadzu Scientific Instruments, Inc., Kyoto, Japan; 2-nm slit width) is equipped with thermostated cuvettes (2-mm path), temperature regulation (Circulator G/D8, Haake GmbH, Karlsruhe, Germany), and a green photoflash (filtered to 500 ± 20 nm).

When photolyzed rhodopsin in its native disc membrane is cooled to temperatures at which the equilibrium is on the MI side (below 5 °C and pH 8.0) (36), any specific binding of protein or peptide to MII causes enhanced formation of MII (extra MII). Extra MII provides a kinetic and stoichiometric measure for the complex between photoactivated rhodopsin and the interactive polypeptide (37, 38).

Kinetic Light Scattering-- The gain or loss of membrane-bound protein mass can be measured readily by light scattering (LS) changes using a setup described in detail by Heck et al. (30). All measurements were performed in 10-mm path cuvettes with 300-µl volumes in hypotonic buffer (20 mM BTP (pH 7.5), 130 mM NaCl, 5 mM MgCl₂) at 20 °C. Reactions were triggered by flash photolysis of rhodopsin with a green (500 ± 20 nm) flash, attenuated by appropriate neutral density filters. The flash intensity is quantified photometrically by the amount of rhodopsin bleached and expressed as the mol fraction of photoexcited rhodopsin (R*/R). LS binding signals (R*/R = 32%) were corrected by a reference signal (N signal) measured on a sample without added protein as described by Pulvermüller et al. (9). LS dissociation signals (R*/R = 0.5%) were recorded with a 0.5-5 ms dwell time of the A/D converter (Nicolet 400, Madison, WI). To suppress base-line activation, 2.5 mM NH₂OH was added to the sample. The LS binding signal is interpreted as a gain of protein mass bound to the disc membranes and the LS dissociation signal as loss of protein mass from the disc vesicle (30).

Binding signals are large changes of LS, reflecting the binding of a protein from solution; small additional binding signals (see Fig. 2B, c and d) are likely to reflect the transition from the membrane binding sites to the receptor; LS signals can indeed arise from membrane-bound reactions, as was demonstrated for the binding of G_t to its effector phosphodiesterase (32).

Simulation of G_t Activation-- To fit the experimental data of G_t activation, the time course of G_t in its activated, GTP-binding form (G_t*) was simulated by a model comprising Reactions 1-4.

All simulations assume the same G_t* activation and inactivation rates (39-41), the same temperature-corrected rate constants of Reaction 4 (k_on = 5 × 10⁵ nM¹ s¹, k_off = 10³ s¹; extra MII measurements), and the same total concentration of R* and G_t. The Ksim software was applied for numerical integration of the rate equations, provided by Helmut Gutfreund.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Light-induced Interaction of Arrestin and Its Variants with Photoactivated Rhodopsin-- The biophysical assays separate receptor interaction and membrane binding of the arrestins from each other. The spectrophotometric "extra MII" assay is specific for the receptor interaction step in that it follows the time-dependent generation of the MII intermediate that is formed at the expense of the tautomeric MI. Arrestin, Arr^3-382, Arr^3-367, and Arr^1-370A (p⁴⁴) all enhance the formation of MII in prephosphorylated membranes (Fig. 1A). However, Arr^3-382, like the full-length protein, does not show this effect for native nonphosphorylated rhodopsin (Fig. 1B, a and b traces). It can be concluded that both proteins are sensitive to the presence of the phosphate groups at the C terminus of rhodopsin. This shows that the last 22 and the first 2 residues of arrestin (which are truncated in Arr^3-382) are not essential for the normal phosphate-dependent interaction. The observation is different for Arr^3-367, i.e. when an additional 15 amino acids are clipped off from the C terminus of arrestin. Arr^3-367 enhances the formation of MII for both pre- and nonphosphorylated membranes. For Arr^1-370A the same result was obtained (Fig. 1, c and d traces; (17)), indicating that it is indeed the lack of the C-terminal stretch and not of the N-terminal residues, which makes the protein-protein interaction insensitive to receptor phosphorylation.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Interaction of arrestins with photoactivated rhodopsin from extra MII formation. The signals represent the flash-induced formation of active MII in disc membranes, measured by A380 nm minus the change at A417 nm. Extra MII, which monitors the interaction with arrestins, is obtained by subtracting the control (not shown) from the absorbance change measured with the protein. A, extra MII formation from prephosphorylated rhodopsin in the presence of arrestin (a), Arr3-382 (b), Arr3-367 (c), and Arr1-370A (d) (splice variant p44). B, extra MII formation of nonphosphorylated rhodopsin under conditions identical to those in A. Solid lines through the data represent bimolecular reaction fit as described by Pulvermüller et al. (17). The final concentrations in all measurements were 10 µM rhodopsin and 1.5 µM arrestin. Experimental conditions were 100 mM HEPES (pH 8.0) at 1 °C; sample volume, 200 µl; cuvette path length, 2 mm; 12% of the rhodopsin was photolyzed/flash.

Membrane Binding of Arrestin and Its Variants-- A centrifugation assay was used to characterize qualitatively the binding of arrestin, Arr^1-370A (p⁴⁴), or Arr^3-367 to disc membranes containing opsin, rhodopsin, or their prephosphorylated forms. As shown in Fig. 2B, Arr^1-370A and Arr^3-367 (but not the full-length arrestin) bind to both p-opsin or pR in the dark and after illumination. The same line of experiments with arrestin shows that membrane binding is only observed with pR*.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Membrane binding of arrestin and its variants, Arr3-382, Arr3-367, and Arr1-370A (p44). A, SDS-PAGE demonstrating the quality of the arrestin preparations used in this study. 4 µg of each protein was applied per lane; lane 1, Arr1-370A; lane 2, Arr3-367; lane 3, Arr3-382; and lane 4, arrestin. Molecular mass standards in kDa are indicated on the left. B, centrifugation assay of the receptor interaction and membrane binding of arrestin, Arr1-370A and Arr3-367. Aliquots of 5 µM membrane suspensions were incubated with 2 µM arrestin or Arr3-367 in the dark or illuminated. The pellets (p) and supernatants (s) were analyzed using SDS-PAGE. For measuring conditions, see "Experimental Procedures." C and D, the binding signals are the LS change arising from the binding of interactive protein to disc membranes. Controls (LS change without protein) are subtracted from each signal. To induce the binding reaction, 32% of the rhodopsin in the sample is photolyzed by the flash. C, binding signals with prephosphorylated membranes and arrestin (a), Arr3-382 (b), Arr3-367 (c), and Arr1-370A (d). D, binding signals with nonphosphorylated membranes under conditions identical to those in C. The final concentrations in all measurements were 3 µM rhodopsin and 1.5 µM arrestin, Arr3-382, Arr3-367, or Arr1-370A. Experimental conditions were 20 mM BTP (pH 7.5), 130 mM NaCl and 1 mM MgCl2 at 25 °C, 300-µl sample volume, and 10-mm cuvette path length.

LS binding signals provide a more quantitative assay of membrane binding. They arise from flashes of light that generate photoactivated rhodopsin. One can distinguish between direct and indirect mechanisms that generate such signals. A protein can bind directly from solution or when binding sites at the membrane become available after transition of a protein from membrane sites to the receptor. In the present study, the binding signals are used as a tool to determine the state of membrane binding (prior to the flash) of a protein. The shift of protein mass from solution to the membrane becomes the larger the less of the protein that is bound to the membrane before receptor activation (30, 42). No binding signal will be seen when the protein is completely bound to the membrane.

The experimental data are shown in Fig. 2, C and D. With both arrestin and Arr^3-382, flash excitation of rhodopsin leads to large binding signals; a second flash (data not shown) produces only a small residual signal arising from excess arrestin or Arr^3-382 that was not bound to the active rhodopsin formed by the first flash. Consistent with previous analyses (17), this is interpreted as a stoichiometric, light-induced binding of arrestin or Arr^3-382 to the phosphorylated rhodopsin, which leads to the observed shift of protein mass to the membrane. We can also conclude that the truncation per se does not disturb the binding signal and that the lack of the extreme C terminus of arrestin leaves membrane binding and (in agreement with Fig. 1) interaction with pR* undisturbed. In sharp contrast to Arr^3-382, both Arr^1-370A and Arr^3-367 show only a very small, if any, binding signal (Fig. 2C, c and d traces). We know from the results shown in Fig. 1 that Arr^1-370A and Arr^3-367 interact vigorously with phosphorylated rhodopsin after photoactivation. In agreement with the analyses through centrifugation assays (see above and Ref. 19), we interpret the absence of the binding signal as membrane binding in the dark. Observations of the time it takes until this stable equilibrium is reached (in the order of 200 s, in which time binding signals of decreasing amplitude are seen; data not shown) allow estimation of the time it takes to form the pR·Arr^1-370A complex from solution. A lower limit for this parameter arises from the binding signal itself (which reflects formation of pR*·Arr^1-370A); this would yield a reaction time in the order of 10 s.

When the protocol in Fig. 2C is repeated with nonphosphorylated rhodopsin, the binding signal is absent with arrestin or Arr^3-382, but a large, although slower signal is seen with Arr^1-370A and Arr^3-367 (Fig. 2D, c and d traces). This "mirror image" of the behavior of phosphorylated membranes can be readily understood by the simple assumption that membrane binding in the dark does occur exclusively when the rhodopsin is prephosphorylated. Evidence that this occurs by direct interaction of Arr^1-370A and Arr^3-367 with inactive but phosphorylated rhodopsin will be presented in the next section.

These data allow an estimation of the dissociation constant of the pR·Arr^3-367 complex. If we assume that the small residual binding signal seen on photoexcitation of pR membranes reflects the presence of 5% of the added Arr^3-367 in solution (before the flash), the dissociation constant would be as shown in Equation 1.

(Eq. 1)

This is an upper limit for the K_D because the residual LS signal is likely to arise at least in part from physical effects other than binding of soluble protein (see "Experimental Procedures"). From the K_D and the k_on (0.003 µM¹ s¹) of equilibration (see above), the off-rate can be estimated with k_off = K_D * k_on  0.0003 s¹. Using the k_on of the binding signal (0.07 µM¹ s¹) yields the most conservative estimation, namely k_off  0.005 s¹. This means that it takes 200 s before an average pR·Arr^1-370A complex dissociates in thermodynamic equilibrium.

Direct Interaction of Arr^3-367 with Phosphorylated Opsin or Rhodopsin in the Dark-- The membrane-independent binding in solution of purified arrestin or Arr^3-367 to the different forms of opsin and rhodopsin was investigated by size exclusion chromatography. The elution profiles in Fig. 3, G-L, demonstrate the presence of complexes between Arr^3-367 and all phosphorylated opsins or rhodopsins by a shift of the elution peaks to higher molecular weights. The peaks are compared with a theoretical peak (black dotted line), calculated for the superposition of the single component profiles. The weak interaction with nonphosphorylated active R* (cf. Fig. 1Bc) is also reflected in these data (Fig. 3J). As expected, full-length arrestin binds only to pR* (Fig. 3F).

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Interaction of arrestin and Arr3-367 with the different forms of receptor. Binding reactions under different conditions were analyzed by size exclusion chromatography. A-F represent the interactions of arrestin, and G-L show the interactions of Arr3-367 with the different receptors (rhodopsin and opsin) as indicated. Elution profiles of arrestin and Arr3-367 (green), receptors alone (blue), and the mixture of the arrestins with the special receptors (red) are shown. The dotted lines are the calculated superpositions of the respective single component profiles yielding the predicted profiles for the mixture of the two noninteracting components. 10 µg (A and D-F) or 5 µg (B and C) of arrestin and 10 µg of Arr3-367 (G-L) were applied.

Inhibition of Transducin Activation-- Interaction of rhodopsin with transducin catalyzes nucleotide exchange in the -subunit of G-protein transducin. The GTP-bound -subunit dissociates rapidly from the membrane, providing a real time monitor of the activation rate (32). The loss of the G_t mass is measured as a kinetic LS change, the so-called dissociation signal (30, 43). The competition of arrestin and its variants with G_t for the activated receptor is best measured when G_t interaction is weakened by an addition of exogenous GDP. Under these specific conditions, arrestin or its variants can win the competition with G_t for R* even at the low concentrations in vitro. The competition becomes visible in a reduced slope of the dissociation signal. All arrestin variants compete with G_t, for pR* (Fig. 4). Arr^3-367 or Arr^1-370A are more potent inhibitors of G_t activation; the inhibition occurs with earlier onset and more efficiently (Fig. 4A). Given the apparent K_D of the pR·Arr^1-370A complex, even at the smallest concentration of the stop protein, half of its amount is membrane-bound. Under these conditions, the quench of the dissociation signal is already very efficient; in its fast initial phase, inhibition occurs with a reaction time in the range of 1 s; the slow component of the deactivation is kinetically similar to the binding signal, indicating a fraction of Arr^3-367 which goes through a membrane binding step before it can compete with G_t activation. It is important to note that under conditions of the b trace of Fig. 4C, only ca. 1/30 of the pR* formed (15 nM) carries Arr^3-367 in a preformed complex (0.1 µM Arr^3-367 versus 3 µM pR). Therefore, the probability is low that a given molecule of Arr^3-367 deactivates the same receptor to which it was bound before it became photoactivated. The slow phase is virtually absent in the d and e traces, indicating that the quench no longer involves any soluble protein. Even in this case, the probability of preformed complexes is < 20%. We conclude that the deactivation occurs from the pR-bound state of Arr^3-367 via a direct allosteric mechanism (see under "Discussion").

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Flash-induced Gt dissociation from disc membranes: inhibition by arrestin and its variants. Dissociation signals are LS changes arising from the dissociation of activated Gt from disc membranes (see "Experimental Procedures"). To activate Gt, a catalytic amount (fraction of R*/R = 5 × 103) of the rhodopsin in the disc membranes was activated in the presence of Gt, GTP, and GDP. A shows the dissociation signal with phosphorylated membranes (control) and the inhibitory effect of 1.5 µM arrestin, Arr1-370A, and Arr3-367. B same as A but with nonphosphorylated instead of phosphorylated membranes. Note that arrestin does not interact with nonphosphorylated rhodopsin. C, titration of the dissociation signal from phosphorylated membranes with increasing Arr3-367 concentrations. From the bottom to the top trace the amount of Arr3-367 increased by 0 (a), 0.1 (b), 0.3 (c), 0.5 (d), and 1.0 µM (e). Trace f represents the control without Gt. D, same as C but with nonphosphorylated membranes and higher Arr3-367 concentrations of 0 (a), 0.5 (b), 1.0 (c), 1.5 (d), and 2.0 µM (e). The inset shows computer simulations of Gt activation in the presence of increasing Arr3-367 concentrations as indicated in D. The individual traces differ only in the assumed Arr3-367 concentration (same as in D). The final concentrations in all measurements were 3 µM rhodopsin and 0.5 µM Gt, 200 µM GDP, 5 µM GTP. The experiments were otherwise performed as described in the legend of Fig. 2.

As expected from the specific interaction properties of Arr^3-367 and Arr^1-370A, these proteins also inhibit G_t activation with nonphosphorylated rhodopsin (Fig. 4, B and D). Compared with phosphorylated rhodopsin, the onset of competition is characteristically delayed, leaving a time window for fast and undisturbed G_t activation (Fig. 4D). During the initial activation phase, the slope of the signal is steeper, reflecting the better capacity of the nonphosphorylated R* to catalyze nucleotide exchange.

The return of the signal to the base line reflects the shift of G_t back to the inactive membrane-bound state. Its kinetics arise from a convolution between increasing successful interaction of Arr^1-370A with R* and the return of G_t into the deactivated state by GTP hydrolysis. With pR, the competition is so rapid and efficient that the level of activated G_t and thus the signal amplitude remain small; a falling phase does not occur because the GTP in the sample is not used up. The simulations (inset of Fig. 4D), which are based on a reaction model using linear differential equations and the simplifying assumption that Arr^1-370A and G_t compete for R* from a homogeneous pool of protein, reproduce the general features of the measured curves surprisingly well. Note however that the data obtained with prephosphorylated rhodopsin cannot reflect the real mechanism, which involves phosphorylation by rhodopsin kinase, but rather the slow binding of the arrestins from solution (all arrestins in Fig. 4 are in excess of activated rhodopsins).

Reaction Order of Light-induced Interaction Steps-- Variation of the protein concentrations and evaluation of the extra MII signal (7) allow determination of the dissociation constants (K_D) (titration data not shown). With prephosphorylated rhodopsin, both Arr^3-382 and Arr^3-367 interact even faster and tighter than arrestin; the affinity for Arr^3-367 is similar to Arr^1-370A (17). The interaction with nonphosphorylated rhodopsin is generally much weaker and slower (Fig. 1B). From the time course of MII enhancement the kinetics of complex formation can be determined. Applying the bimolecular reaction scheme (17), the data in Fig. 1A can be evaluated in terms of the on-rate of interaction. Because the on-rate (k_on) is conventionally measured in units of µM¹ s¹, a constant k_on means a normal bimolecular interaction mechanism. This is seen with Arr^3-382 (filled triangles) as with the parent arrestin (Fig. 5, filled diamonds), whereas Arr^3-367 (filled squares) and Arr^1-370A (filled cycles), interacts with a generally faster but concentration-dependent on-rate. The curve, similar to that measured with Arr^1-370A, indicates that the apparent time constant (k_on* concentration) does not depend on concentration; interestingly, the rates of arrestin and Arr^3-367 approach one another for sufficiently high concentration. The on-rate (k_on in units of µM¹ s¹) decreases by about 3-fold for a 3-fold increase of concentration. The resulting constancy of the absolute reaction rate (in units of s¹) indicates that an intramolecular conversion is rate-limiting. It cannot be the formation of MII itself, which is nearly five times faster under these conditions (data not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Extra MII formation as a function of arrestin concentration. Extra MII measurements were performed as described in the legend of Fig. 1. On-rates of protein-protein interaction are plotted for each arrestin as a function of protein concentration; filled diamonds, full-length arrestin; filled triangles, Arr3-382; filled squares, Arr3-367; filled circles, Arr1-370A (p44), each with prephosphorylated rhodopsin; open squares and circles, interactions of Arr3-367 and Arr1-370A with nonphosphorylated rhodopsin, respectively. The inset shows examples of original recordings for Arr3-367.

Similarity between Arr^3-367 and Arr^1-370A (p⁴⁴)-- For the discussion of the data, it will be most important that the truncated form Arr^3-367 and the splice variant Arr^1-370A (p⁴⁴) were qualitatively similar or identical in properties such as interaction with the activated receptor, membrane binding, and inhibition of G_t activation. The direct interaction in solution with the different rhodopsins (Fig. 3) could only demonstrated for the truncated form, which is available in the necessary quantities.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The salient result of this study is that the presence of the splice variant of arrestin, Arr^1-370A (p⁴⁴) in rod outer segments makes biological sense. Although the binding sites are not, as in full-length arrestin, completely masked in an inactive conformation of the protein, pR·Arr^1-370A complexes provide an inactive membrane-bound storage form of the arrestin variant. With the proposal that Arr^1-370A is bound to phosphorylated inactive rhodopsin or opsin, ready to interact with active phosphorylated rhodopsin, we have addressed the long-standing question about the role of the rhodopsin phosphorylated in excess of the light single molecules activated in the single photon regime.

Stabilization of the Inactive Ground State of Arrestin-- We have extended previous investigations on the interactions of visual arrestin and its splice variant p⁴⁴ (Arr^1-370A), with different forms of the rod photoreceptor rhodopsin. Proteins under investigation included the inactive receptor (R), its activated (R*), and their prephosphorylated forms (pR and pR*), and arrestin, Arr^1-370A, and the truncated forms Arr^3-382 and Arr^3-367. Truncated forms of visual arrestin arise from partial digestion (as used in this study) and may be present in the rod in vivo. In general, the behavior of truncated arrestins would be hard to predict (21), but the truncations applied in this study will help us to draw conclusions on two different, although related, topics, namely the mechanism of activation of the arrestins and their role in phototransduction. Current models of arrestin function are based on the concept of a conformational switch, which keeps the protein inactive during the period of catalytic interaction of the active receptor with the G-protein and converts it into an active molecule on contact with the phosphorylated and activated receptor (pR*). The inactive arrestin conformation is stabilized mainly by the ionic interactions within the so-called polar core and by hydrophobic interaction of the N terminus with the C terminus and the short -helix found in the structure (see Refs. 8 and 44).

Our results add some new information to proposed mechanisms of stabilization in the polar core (Fig. 6). When the extreme arrestin C terminus is lacking (Arr^3-382), the interaction with photoactivated phosphorylated rhodopsin (pR*) becomes faster (see Figs. 1A and 2B). This shows that even the extreme C terminus (residues 383-404) is important for the stability of the inactive arrestin conformation and thus for the conformational switch. Consistently, Arr^3-382, like full-length arrestin, can only interact with prephosphorylated R*. The decisive change in the function occurs when the C terminus is shortened by another 13 or 15 residues: both Arr^1-370A and Arr^3-367 interact with nonphosphorylated R* and pR. This result is consistent with an assignment of Arg³⁸² to the polar core (Fig. 6) and the 375-377 stretch to a hydrophobic interaction site with the N terminus (residues 11-13) and the -helix (involving residues 103, 107, and 111; (8)). However, our results do not confirm Lys² as part of the polar core (45). Disruption of the ionic interaction should destabilize the inactive arrestin conformation, and both Arr^3-382 and Arr^3-367 should interact with nonphosphorylated R*, which is not observed.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Polar cores. Residues of the polar core A, as identified by Granzin et al. (45), and B, as identified by Hirsch et al. (14) are shown. The color of the backbone changes from light green (N terminus) to light blue (C terminus) via turquoise (middle of sequence). The different amino acids (Lys2 and Arg382) are marked in red. The dashed lines represent the hydrogen bonds according to related publications.

Mechanism of the Conformational Switch-- The first approach to the conformational switch arose from time-resolved experimental data (7) and partial digestion of arrestin (13). A reaction scheme of R*-arrestin interaction was developed in which arrestin exists in an inactive conformation (A_i) which is different from the pR*-bound conformation (A_b). The kinetic data in the present study suggest to extend this general concept as shown in Reactions 5 and 6.

The interaction between pR* and arrestin proceeds through (at least) two individual reaction steps. In Reaction 5, pR* interacts to stabilize a "prebound" transitory conformation of arrestin (A_pb), and Reaction 6 represents a structural rearrangement leading to the stable arrestin-receptor complex.

The H of Reaction 5 shows up in the measured activation energy of the overall binding reaction. It is this conversion that is lacking in Arr^1-370A or Arr^3-367; in these arrestin variants, the conformational switch is absent, and the protein is partially preactivated, leading to a lower apparent activation energy of the binding reaction (70 kJ/mol instead of 140 kJ/mol (17)). We therefore assume that the conformational change in Reaction 5 involves the movement of the C terminus. A sequential scheme for pR*-arrestin interaction was proposed on the basis of site-directed mutagenesis (18); it involves the contact of the phosphorylated region of rhodopsin (P site) as a trigger (46), switching arrestin into its active conformation and allowing it to interact with the binding sites exposed on rhodopsin by photoactivation (M site). These two steps may be identified with the Reactions 5 and 6 arising from kinetic analysis. According to recent mutational analyses (8), the phosphorylated sites on pR* make contact with arrestin through a stepwise replacement of elements stabilizing the A_i structure by the phosphates of pR*. Identified elements in arrestin include ionic interactions within the polar core and the hydrophobic interaction between the N and C terminus and the -helix (8, 44). The first contact between arrestin and pR* is made through the Lys¹⁴/Lys¹⁵ tandem and its interaction with one of the phosphates on the C terminus of pR*. In the language of Reactions 5 and 6, this trigger event is a first substep of Reaction 5. It remains to be elucidated whether the trigger makes use of a thermodynamic instability of the ground state, as assumed in our previous reaction model; in its original form, it assumes a "pre-equilibrium," in which the conformation A_pb is always present, and is only enhanced at the expense of A_i by interaction with pR* (13, 47).

The evidence for introducing Reaction 6 comes from the kinetics of interaction of the short forms of arrestin. Applying the bimolecular reaction scheme, we have found that Arr^3-367 like Arr^1-370A (Fig. 5) interacts with a generally faster on-rate than full-length arrestin. This is consistent with the conclusions above, namely that the shorter variants do not need to go through Reaction 5, which limits the speed at which the proteins can interact. The data in Fig. 5 (constancy of the absolute reaction rate) have indicated an intramolecular conversion, which limits the overall rate under the conditions. In Reactions 5 and 6, we give the most straightforward interpretation, namely by Reaction 6, which occurs when the complex between the collisional complex is already formed. Although Reaction 6 is not resolved with full-length arrestin (because Reaction 5 is rate-limiting), we assume that it also occurs in this case. It is possible that only with Reaction 6 is MII stabilized.

Receptor Interactions of Arr^3-367 and Arr^1-370A without Photoactivation-- With Arr^3-367 or Arr^1-370A and prephosphorylated membranes, the LS binding signal was not observed (see Fig. 2C). Because this light scattering change reflects the transition of a soluble protein to the membrane and because the extra MII assay showed light-induced interaction of these proteins with rhodopsin, we have concluded that Arr^3-367 and Arr^1-370A bind to the receptor from a membrane-bound state (dark binding). Because a large binding signal was seen with nonphosphorylated membranes, the membrane association in the dark must require phosphorylated rhodopsin. Our direct chromatographic analysis (Fig. 3) has confirmed the old surmise (19) that the dark binding occurs by direct protein-protein interaction with phosphorylated rhodopsin in the absence of photoactivation (pR state). Hydrophobic lipid-protein interaction of Arr^1-370A per se is apparently not sufficient to keep Arr^1-370A on the membrane. The hydrophobic helix I, which was suggested to be exposed in activated arrestin (15) and presumably also in preactivated forms such as Arr^1-370A, may only be functional in concert with interaction with the activated and/or phosphorylated receptor, R* and pR*. respectively.

Conceivably, the lack of the C terminus of arrestin leads to a complete loss of the stabilizing interaction between N and C terminus and to a relatively free movement of the N terminus, which may enable Arr^3-367 and Arr^1-370A to bind to pR (Fig. 7). Binding of both Arr^3-367 and Arr^1-370A to phosphorylated rhodopsin in the absence of photoactivation is also consistent with the proposed role of the N terminus of arrestin in making the first contact with the phosphate of rhodopsin (8, 15).

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Model for rhodopsin interactions of arrestin and its variants. A, arrestin and Arr3-382 interact only with the MII conformation of rhodopsin and only in its phosphorylated form (p-MII). In vitro, one starts from prephosphorylated rhodopsin (p-R; see "Experimental Procedures"). The phosphorylated C terminus of rhodopsin interacts with the N terminus of arrestin (indicated as red lines) thus breaking the stabilizing interactions between the N and C termini of arrestin (red and green lines). The resulting conformational switch in arrestin enables tight binding to p-MII. B, the lack of 37 C-terminal residues in Arr3-367 or 35 C-terminal residues in Arr1-370A results in the loss of the stabilizing interaction between the C and N termini. In these preactivated forms of arrestin, the N terminus of Arr3-367 is free to bind to inactive (dark-adapted) phosphorylated rhodopsin (p-R). After light absorption (another) phosphorylated rhodopsin switches to p-MII, and the conformational changes of Arr3-367 or Arr1-370A lead to complex stabilization via a bridging mechanism.

The affinity of Arr^1-370A to pR was estimated from the residual amplitude of the binding signal (Fig. 2C) with < 0.08 µM, which resulted in an off-rate of the pR·Arr^1-370A complex of 0.06 s¹. The presence of a full binding signal in the case of nonphosphorylated membranes (Fig. 2D) implies in turn that the affinity to inactive nonphosphorylated rhodopsin (R) is very small.

Competition with the G-protein for Photoactivated Rhodopsin-- The real time measurements of G_t activation by a catalytic amount of R* or pR* (Fig. 4) have shown that G_t activation is quenched by both Arr^1-370A or Arr^3-367 with similar efficiency. Although the delays in the onset of deactivation reflect nicely the strength and on-rate of competitive interaction, these measurements on nonphosphorylated or prephosphorylated membranes cannot reflect the biologically real case, in which single R* switches from R* to pR* through rhodopsin kinase activation. However, it is evident from the pR data (Fig. 4, A and C) that the membrane-bound fraction of truncated arrestin binds and inhibits pR* with a reaction time of less than 1 s, presumably because of its presence at the membrane and via its interaction with pR. We will now discuss how this mechanism fits into the known properties of signal transduction in vivo.

Extrapolation to the Case in Vivo-- Recent investigations (48) have confirmed the old notion (49-51) that arrestin undergoes translocation in bright light from the inner to the outer segment of the rod cell. The translocation occurs in opposite direction to transducin, which disappears from the outer segment under conditions of bright illumination (34, 52). It is not known if and how arrestin leaves the outer segment in periods of darkness, but recent immunofluorescence staining experiments failed to detect even small amounts of the protein under conditions of thorough dark adaptation in mice.² Although it is not yet clear which variants of arrestin take part in the translocation, Arr^1-370A protein seems to be most abundant in the outer segment (19). The preparation procedure for the protein starts from rod outer segment preparations under dim red light conditions.

For preparations of isolated rod outer segments, a role of Arr^1-370A in quenching phototransduction in the single photon regime could be demonstrated directly by measurements of effector activity in rod outer segment preparations (pulses of phosphodiesterase activity in a calorimetric phosphodiesterase assay (22)). Exogenously added proteins (Arr^1-370A much more efficiently than full-length arrestin) deactivated the phosphodiesterase signal; remarkably, however, neither arrestin nor Arr^1-370A inhibited the initial rising phase of effector activation. Deactivation cut in after about 0.2-0.5 s, roughly the time window seen in Fig. 4, A and C. This implies that during the initial phase, neither protein influences the activation phase of the G-protein before the activated single molecule of R* is phosphorylated. This calls for a mechanism that sequesters the arrestins from R* but presents it for interaction with pR*. In the case of full-length protein, this is the conformational switch. The mechanism is less obvious for Arr^1-370A, but our results may provide first insights into a possible mechanism. We will show that the time-ordered interactions of Arr^1-370A with pR, p-opsin, and pR* can provide a mechanism that replaces the inactivating conformational switch of full-length arrestin.

The presence of pR in dark-adapted rod cells has long been known; photoreceptor disc membranes contain about 1% pR, phosphorylated as a by-product of the rhodopsin kinase-mediated phosphorylation of rhodopsin in the single photon regime (see Ref. 53 and citations therein). This is approximately the same amount as was estimated for the amount of Arr^1-370A in rod outer segments (19). With the estimated apparent K_D and the concentration of Arr^1-370A in situ (about 30 µM), virtually all of the Arr^1-370A will be bound to pR with a low off-rate (<200 s, see above). On formation of a single copy of R* (note that less than one R* is formed per membrane in the single quantum regime of rod operation), Arr^1-370A will form the specific tight complex with this molecule after phosphorylation (pR*). The time in which this occurs can be extrapolated from the time of inhibition of the dissociation signal, by the following consideration.

First, the data require that Arr^1-370A leave pR (present in large excess) and bind to pR* (formed at an amount of 0.5% of pR) in less than 1 s. In view of the low off-rate of the pR complex (about 0.005 s¹; see above), this cannot happen by spontaneous off-reaction of the bound Arr. It is thus that unlikely Arr^1-370A reaches pR* via a soluble intermediate.

Second, in view of the lack of evidence for a membrane-bound state of Arr^1-370A (Arr^1-370A does not bind to nonphosphorylated membranes), a membrane-bound intermediate is also very unlikely.

Third, there must be therefore some kind of direct or indirect interaction between pR* and Arr^1-370A and/or pR which induces the release of Arr^1-370A from pR, presumably by an allosteric mechanism. We may therefore assume that Arr^1-370A forms a bridge between pR and pR* (as assumed in Fig. 7B) and thus a pR-pR* dimer. This may be interpreted as an Arr^1-370A-mediated transitory dimer formation between the pR and pR* receptors. Dimerization of G-protein-coupled receptors plays a potential role in a variety of functions (see Ref. 54). In the case of rhodopsin, evidence for constitutive dimers or a dimerization in the course of catalytic G-protein activation is still lacking. However, the present data may support a pR-pR* dimer mediated by Arr^1-370A involved in the shutoff of G-protein-coupled signal transduction.

Although important details of the hypothetical interaction mechanism remain to be elucidated, we can say that the interaction is instantaneous on the time scale of seconds (Fig. 4, A and C), and the quench virtually complete, even at the relatively low concentration used in the experiment in vitro. Thus, the pacemaker in vivo is presumably the phosphorylation of R* because interaction with R* virtually does not occur as long as the R* is not phosphorylated (see above and Ref. 22). This allows for a time window of undisturbed catalytic interaction of the R* with G_t. It may also be relevant that Arr^1-370A "sites" on pR, thus saving the lipid surface for rapid diffusion of the G-protein.

Fig. 7A outlines the normal switch mechanism of arrestin, which interacts from solution with pR*, to switch into the A_pb and in a second step into the A_b conformation. Both Arr^1-370A and Arr^3-367 are in preactivated conformation A_pb (Fig. 7B) and interact therefore with pR in the dark. On photoexcitation, they bind to pR*, either to the same molecule (which typically occurs only at very high levels of illumination and/or phosphorylation) or by forming a bridge with the phosphorylated/activated surface of another rhodopsin molecule. Because the affinity of all arrestins to nonphosphorylated R* is much smaller than to pR*, neither Arr^1-370A nor Arr^3-367 is assumed to interact with R*.

Roles of Arr^1-370A and Arrestin in Different Regimes of Phototransduction-- The proposed mechanism specifies the notion (20, 22) that Arr^1-370A is the form of arrestin which inactivates phototransduction at low light levels of illumination. It is also consistent with the prolonged deactivation seen in arrestin knock out mice, when we assume that both full-length arrestin and the splice variant Arr^1-370A are knocked out (12). Arr^1-370A and full-length arrestin can thus be assigned specific functions in the different regimes of rod function. As long as the amount of R* is very small (single quantum regime), the Arr^1-370A bound to pR and present at a density of about 300/µm² will have a chance to collide quickly with the single R* formed by absorption events and to occupy it as soon as it is phosphorylated. A large amount of arrestin in the narrow interdiscal space would only hinder diffusion of G_t and thus the function of the visual amplifier, which might be one of the reasons why arrestin is only present in the outer segment of the rod when high levels of illumination have been applied before (49, 50).

At high light levels and thus at large (>1%) amounts of active rhodopsin, the amount of pR is low compared with pR* (53), and an alternative mechanism is needed to keep the arrestin inactive before pR* is formed. This is the A_i-A_b activation switch that operates by the contact of the arrestin with the phosphorylated residues on pR*. By the very large amount of arrestin needed to cap large amounts of R*, it would be impossible to host the stop protein at the membrane, where G_t already occupies half of the available space. However, the binding of a small amount of arresting Arr^1-370A protein at the surface of dormant phosphorylated receptors does not hinder G_t diffusion.

In the range between single photon detection and a light-adapted state of the rod (when a few R*/membrane·s are formed), shutoff of R* will depend on the precise amount of pR and on its balance with pR*. An additional factor is the phosphorylated apoprotein p-opsin, which also binds Arr^1-370A (Fig. 3H). The amount of p-opsin formed for a certain condition (see Ref. 55) depends on the interplay between kinase and phosphatase activities. Complexity is introduced by the fact that bound arrestins tend to stabilize MII and MII-like opsin-all-trans-retinal "pseudophotoproducts" (27) and that, in turn, bound arrestins inhibit the attack of phosphatase. We know (17) that the NH₂OH-induced loss of all-trans-retinal from the opsin apoprotein releases Arr^1-370A, like arrestin. A possible scenario would be that opsin-all-trans-retinal complexes are generated when MII decays and that the subsequent interaction of Arr^1-370A with the complex inhibits the attack of the phosphatase so that pR·Arr^1-370A-pR* survives. Thus the Arr^1-370A mechanism outlined here could also open new possibilities for regulation. It remains to be elucidated at which level of light adaptation the mechanism is set in action and when it is gradually replaced by the arrestin mechanism.

	ACKNOWLEDGEMENTS

We thank Martin Heck for help with the competition assay and Krzysztof Palczewski and Oliver Ernst for valuable discussions. Jana Engelmann and Ingrid Semjonow provided expert technical assistance.

	FOOTNOTES

^* This work was supported by Deutsche Forschungsgemeinschaft (SSP Molecular Sensory Physiology) (to A. P. and K. P. H.) and by the Fonds der Chemischen Industrie.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: IMPB, HU-Berlin, Charité, 5-9 Ziegelstrasse, Berlin 10117, Germany. Tel.: 49-30-450-524-176; Fax: 49-30-450-524-952; E-mail: alexander.pulvermueller@charite.de.

Published, JBC Papers in Press, August 22, 2002, DOI 10.1074/jbc.M206211200

² U. Wolfrum, personal communication.

	ABBREVIATIONS

The abbreviations used are: G-protein, guanine nucleotide-binding regulatory protein; Arr, arrestin; BTP, 1,3-bis[tris(hydroxymethyl)methylamino]propane; G_t, retinal G-protein, transducin; LS, light scattering; MI, metarhodopsin I; MII, metarhodopsin II; pR, prephosphorylated rhodopsin; pR*, phosphorylated active rhodopsin; R*, nonphosphorylated active rhodopsin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES