Movement of the Retinylidene Schiff Base Counterion in Rhodopsin by ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 7, Issue of March 5,pp. 4661-4667, 1993 Printed in U.S.A.

Movement of the Retinylidene Schiff Base Counterion in Rhodopsin by One Helix Turn Reverses the pH Dependence of the MetarhodopsinI to Metarhodopsin I1 Transition* (Received for publication, October 9,

1992)

Tatyana A. Zvyaga$, K. Christopher Ming, Mareike Beck$, and ThomasP. SakmarSlI From the $Howard Hughes Medical Institute, §Rockefeller University, New York, New York10021

phore that is covalently linked via a protonated Schiff base at LysZg6.Photoisomerization of the chromophore to thealltrans form results in a series of conformational intermediates leading to the formation of metarhodopsin I1 (MII), which catalyzesguanine nucleotideexchange by transducin. The = 480) to MI1 ,X ,(, transition of metarhodopsin I (MI) (X, = 380 nm) requires Schiff base deprotonation (1).The Schiff base proton might be transferred to a carboxylate side chain (2), as is the case in bacteriorhodopsin (3,4). But as of yet, it has not beenshown that a specific Schiff baseproton acceptor residue is required in rhodopsin. However, the equilibrium between the MI form, which contains a protonated Schiff base, and the MI1 form of rhodopsin has been shown to be influenced by specific histidine residues (5). Site-specific mutagenesisstudieswith bovinerhodopsin have identified Glu113 as the counterion to the retinylidene Schiff base(6-8). These resultswere supported by a resonance Raman study of rhodopsin mutants in which a model of the Schiff base environment was proposed (9). The predominant effect of replacement of Glu113by a neutral aminoacid residue such as glutamineor alanine was a loweringof the Schiff base pK, from > 8.5 toabout 6 (6, 7). Thesemutants havea number of other interesting properties suchhydroxylamine as reactivity in darkness, theability of an exogenously supplied solute anion to act as a surrogate counterion to compensate for the positive charge of the protonated Schiff base (6, 7, 9, lo), and the ability to regenerate with all-trans retinal and activate transducin in darkness (6). However, results of experiments related to light-dependent Schiff base deprotonation in thesemutantsare difficult toapplyto rhodopsin because their Schiff base pK, values are reduced by about 3 orders of magnitude. The transmembrane domains of rhodopsin have been postulated to have a-helical secondary structure(11).This view of the third transmembrane segment was supported by several lines of evidence, including the assignment by site-directed Rhodopsin, the visual photoreceptor of the rod cell, is a mutagenesis of helix borders (6, 7), and structuralanalogy to member of the superfamily of seven transmembrane helix bacteriorhodopsin,in which a high resolution structureis available (12). It was postulated thata helical structure of the receptors that activate guanine nucleotide-binding regulatory proteins (G proteins).’ It contains an 11-cis-retinal chromo- third transmembrane segment might allow the repositioning of the counterion by one helix turn so that it fqces the same * This research was supported by the Howard Hughes Medical general direction but is moved by about 5.4 A toward the Institute and National Institutes of Health Medical Scientist Train- cytoplasmic end of the helix. Characterization of such a ing Program Grant 5-T32-GM-07739(to C . M.). The costs of publi- mutant would indirectly test the hypothesis that the third cation of this article were defrayed in part by the payment of page transmembrane domain was helical, but more importantly it charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate might provide a basis for further study of the chemical environment of the Schiff base. this fact. Here, we describe the characterization of a triple replace7l To whom correspondence should be addressed: Box 284, Rockefeller University, 1230 York Ave., New York, NY 10021. Tel.: 212- ment mutant (E113A/A117E/E122Q)? The mutant formed a

The environment of the retinylidene Schiff base in bovine rhodopsin has been studied by movementof its carboxylic acid counterion from position 113 to position 117 by site-specific mutagenesis. Replacement of the counterionat position 113 by a neutral amino acid residue has been shown to produce a lowering of the Schiff base acidity constant (pK,) from > 8.5 to about 6. The aim of the present work was to change the a significant position of the counterion without causing effect on the Schiff base pK,,. A triple replacement mutant (GlullS + Ala/Ala”’ + G1u/G1u122+ Gln) was designed to move the position of the counterionby one helix turn in the third putative transmembrane helix (helix C). The mutant bound 11-cis-retinal to form a chromophorewith a visible absorbance maximum (Xmax) of 490 nm which was independent of pH in the range of about 5-8.5. Upon illumination under conditions in which rhodopsin was converted to the active metarhodopsin I1 (MII) photoproduct, the mutant was converted to a metarhodopsin I (MI)-like species (X, = 475 nm). Furthermore, the effect of pH on the photobleaching behavior of the mutant was the reverse of that reported for rhodopsin. In the mutant, acidic pH favored the formation of the MI-like photoproduct,and basic pH favored the formationof an MII-like photoproduct,,X,( = 380 nm). The MII-like photoproductof the mutant pigment was able to activate the guanine nucleotide-binding protein, transducin. We conclude that the Schiff base counterion in rhodopsin can be repositioned to form a pigment with an apparently unperturbed Schiff base pK,. Furthermore, a specific amino acid residue that acts as a Schiff base proton acceptor is not strictly required for photoconversion of rhodopsin to its active MI1 form.

327-8288;Fax: 212-327-8370. The abbreviations used are: G protein, guanine nucleotide-binding regulatory protein; MI, metarhodopsin I; MII, metarhodopsin 11; GTP-yS, guanosine 5’-0-(3-thiotriphosphate); dATPcuS, deoxyadenosine 5’-O-(l-thiotriphosphate).

Mutants are designated by the native amino acid residue (single letter code) and its position number followed by the substituted amino acid residue. For example, in mutant E113A, Glu113 is replaced by Ala.

4661

Rhodopsin Schiff Base Counterion

4662

pigment with a Xmax value of 490 nm andwithout an apparent effect on the Schiff base pK,. The mutant also displayed altered photobleaching properties with an MI/MII equilibrium that strongly favored MI. The pH dependence of the equilibrium was altered as well with acidic pH favoring MI formation and basic pH favoring MI1 formation. Nevertheless, the mutantpigment could activate transducinin response to light at a slightly reduced level. These results imply that the mutation may have uncoupled Schiff base deprotonation from a specific pathway and therefore, that a specific Schiff base proton acceptor residue may not be required for the formation of an active MI1species. The mutant described should prove useful in further chemical and spectroscopic studies of the mechanism of Schiff base deprotonation in the photoactivation pathway of rhodopsin. EXPERIMENTALPROCEDURES

Materials-Sources of most materials have been reported previously (6, 10, 13). Dodecyl maltoside detergent was purchased from Anatrace, Inc. (Cleveland, OH). Radionuclides were from Du PontNew England Nuclear, and BA85 nitrocellulose filters were from Schleicher & Schuell. Oligonucleotide synthesis was carried out on an Applied Biosystems model 392 synthesizer, and purification of synthetic DNA was carried out essentially as described previously (14, 15). Construction of Rhodopsin Mutants-Site-directed mutagenesis was performed using restriction fragment replacement "cassette mutagenesis" (16) in a synthetic gene for rhodopsin (14) which had been cloned into the expression vector as described previously (17). Each mutant was prepared by replacement of a 69-base pair RsrII-Spe I restriction fragment with a synthetic duplex containing the desired codon alteration(s). The nucleotide sequences of all cloned synthetic duplexes described were confirmed by the chain terminator method for DNA sequencing of purified plasmid DNA using [35S]dATPcuS (18). Expression of Rhodopsin Mutants-The mutant opsin genes were expressed in COS-1 cells following transient transfection by a DEAEdextran procedure as described (19). The rhodopsin used in this report was prepared exclusively from COS cells. Reconstitution and Purification of Mutant Pigments-COS cells expressing the mutantapoproteins were harvested and thenincubated in the presence of 11-cis-retinal under dim red light to reconstitute pigments as described (6, 13). The purification procedure employed was based on the immunoaffinity procedure of Oprian et al. (19) which was modified as described (6, 13, 20). UV-visible Absorption Spectroscopy-Spectroscopy was performed on a X-19 Perkin-Elmer spectrophotometer at 20 "C on purified detergent-solubilized samples in a cuvette with a 1-cm path length (21). All pigment samples were studied in solutions of 0.1% dodecyl maltoside detergent. The spectrum in Fig. 2 is presented without alteration of data by averaging or smoothing algorithms. Other spectra are displayed after smoothing using the Perkin-Elmer UVCSS data handling software package. Spectrophotometric Titrations of Mutant Rhodopsin PigmentsThe effect of pH value on spectral propertiesof pigments was carried out as described previously (10). Photobleaching of Mutant Rhodopsin Pigments-Samples were illuminated directly in the spectrophotometer cuvette using a fiber optic light guide (Dolan-Jenner, Inc.). The light source was a 150watt projector lamp fitted with a 515-nm long pass filter (Oriel, Inc.). For experiments to determine the pHdependence of photobleaching, mutants were prepared in 100 mM NaCl as described previously (lo), and 10 X buffer was added to give a final solution of 50 mM Tris maleate, 100 mM NaCI, 0.1% dodecyl maltoside. The Tris maleate buffers over a wide pH range (~5.5-8.5).A pH value was recorded using a micro-pH electrode in the cuvette before and after illumination. Transducin ActiuationAssay-Bovine transducin was purified from rod outer segment extracts by hexyl-agarose chromatography (22). The assay was based on the property that activated transducin containing GTPyS binds to nitrocellulose membranes without releasing its nucleotide. It was carried out essentially as described (23). Purified pigments were stored frozen at -20 "C and were used only once after thawing. The assay mixture consisted of purified pigment (1-10nM), purified transducin (4 p ~ ) and , [35S]GTPrS(20 +M) in

0.1 ml of assay buffer (50 mM Tris-C1, pH 7.2, 100 mM NaCl, 4 mM MgC4,l mM dithiothreitol, and 0.01% dodecylmaltoside). The assay was carried out at 10 "C. The time course was begun in darkness by the addition of nucleotide. After a 2-min incubation, the assay mixture was illuminated with a 150-watt projector lamp (Dolan-Jenner) afixed with a 495-nm long pass filter. At 30-5 intervals in darkness and 15s intervals after illumination, 10-+1aliquots were removed and transferred to thenitrocellulose filters. The filters were washed and dried, and bound nucleotide was quantitated using a PhosphorImager System (Molecular Dynamics, Inc.). An assay was carried out in darkness in parallel. Control experiments in the absence of transducin revealed that less than 0.03% of the nucleotide bound to the filter. Reaction of Mutant Rhodopsin Pigments with Hydroxylamine in Darkness-Samples were prepared in 50 mM Tris maleate, pH 6.5 or 6.7, 100 mM NaC1, 0.1% dodecyl maltoside. After an initial spectrum was recorded, a solution of hydroxylamine hydrochloride adjusted to the pH of the reaction with NaOH was added to a final concentration of 25 mM to start thereaction. Scanning spectroscopy was repeated periodically to monitor the loss of pigment and formation of retinal oxime. The pHwas measured again at the endof the reaction. Reaction of Mutant Rhodopsin Pigments with Hydroxylamineunder Continuous Illumination-The experimental design was identical to that in darkness as described above except that immediately after the addition of hydroxylamine the sample was illuminated. Illumination was continued during scanning spectrophotometry. The use of a fiber optic light guide, spectral beam masks, and black cuvettes prevented spectral artifacts during simultaneous scanning and illumination. RESULTS

Preparation of Mutant Pigments-A schematic representation of putative transmembranehelix C of bovine rhodopsin is shown in Fig. 1. A triple amino acid replacement mutant was prepared (E113A/A117E/E122Q) to reposition the carboxylate Schiff base counterion by one helix turn within helix C. In addition, three single amino acid replacement mutants were prepared (E113D, E113A, and E122Q) to use as controls in evaluating the properties of the triple replacement mutant. The mutant genes were expressed in COS-1 cells in tissue culture (19). The mutantopsins were regenerated with ll-cisretinal and purified by an immunoaffinity absorption procedure (6, 13, 19). Each of the mutant opsins bound ll-cisretinal to form a visible chromophore. UV-visibleSpectroscopy of Mutant Pigments-The UVvisible spectral properties of the mutant pigments are summarized in Table 1. Rhodopsin purified fromCOScellsin value of 500 nm. dodecyl maltoside detergent showed a, , ,X The UV-visible absorption spectrum of the triple replacement pigment (E113A/A117E/E122Q) is shown in Fig. 2. It displayed a visible, , ,X value of 490 nm. The UV-visible spectral properties of the single amino acid replacement mutants (E113D, E113A, and E122Q)were essentially identical to those reported previously (6, 10). The effect of pH on the spectral properties of each of the mutant pigments is shown in Fig. 3. Mutant pigment E113A existed as a pH-dependent equilibrium mixture of 506 and 386 nm forms as reported previously (10). The pK, value for the Schiff base in mutant E113Awas reported to be 5.7 in 100 mM NaCl (10). The spectrum of the triple mutant E113A/A117E/E122Q was independent of pH over the pHrange of about 5-8.5. The same pH stability was shared by rhodopsin and mutant pigments E113D and E122Q. Photobleaching Properties of Mutant Pigments-The photobleaching properties of the mutant E113A/A117E/E122Q were evaluated and compared with rhodopsin and the single replacement mutants. Rhodopsin solubilized in dodecyl maltoside detergent is rapidly converted to its MI1 form upon photoisomerization of the 11-cis-retinal chromophore to the all-trans-retinal form. The MI photoproduct is not observed in thisdetergent at temperatures above freezing (13). As shown in Fig. 4, this photobleaching property of rhodopsin is

Rhodopsin Schiff Base Counterion

4663

10.07

0.06

4

0.02 0.01

I

0.00

300

400

500

600

Wavelength ( n m ) FIG. 2. UV-visible absorption spectrum of mutant pigment E l 13A/Ali7E/E122&. A spectrum for the triple replacement mutant pigment regenerated with ll-cis-retinal and purified in dodecyl maltoside detergent in darkness is shown. The mutant formed a visible chromophore with a Amax value of 490 nm. A comparison of , , X values and extinction coefficients among rhodopsin and thesingle replacement mutants E113D,E122Q. and E113A is presented in Table I.

1""

/

RHO

FIG. 1. Schematic representation of the third transmembrane helix (helix C) in bovine rhodopsin. Amino acids replaced in putative helix C in rhodopsin are boxed and numbered. The mutant pigments prepared are listed in Table I. CYS"~ has been reported to form a disulfide bond with CysIs7and may form the intradiscal border of helix C (20, 36). The charged pair G l ~ ' ~ ~ - Ahas r g 'been ~ ~ shown to be involved in transducin binding (25). Mutations of this charged pair do not affect UV-visible spectral properties (6, 13).

TABLE I Summary of, , X

values and extinction coefficients of rhodopsin

mutants M" cm"

nm

42,700 Rhodopsin 500 E113D 43,500 506 E122Q 43,500 485 E113A E113A/A117E/E122Q 39.000 490 Amax values were determined from the peaks of dark spectra from at leastthree independent preparations (see Figs. 2 and 3). The precision is estimated to be 2 2 nm. *Values for c (extinction coefficient) were determined by acid denaturation of each mutant opsin as described (6). The c value for rhodopsin was assumed to be 42,700 M" cm" at 500 nm (35). The values for mutant pigments E113D, E122Q,and thetriple mutant are , ,X values at pH 7.1 and represent given a t their respective visible the average of at least two independent experiments. The c value for mutant E113A was determined based on the absorbance a t 506 nm after complete conversion to the acidic form of the pigment at about pH 4.5 (see Fig. 3). All values were rounded to the nearest 500 M" cm". e Mutant E113A existed as a pH-dependent mixture of 386 and 506 nm forms. ~

~

~

0.03

~~~~

independent of pH indodecyl maltoside detergent-solubilized preparations. Under other conditions such as in rod outer segment disc membranes (24, 25), in "doped" micelles (26), or in digitonin-solubilized preparations ( 5 , 13), MI is observed

0.02 0.01

0.00

I

0.00

J i . 400

500

Wavelength ( n m ) FIG. 3. Effect of pH on the visible absorption properties of rhodopsin and mutant pigments. Visible spectra are shown for rhodopsin, E113D, E122Q, and E113A/A117E/E122Q in darkness at four pH values: 5.9, 6.8, 7.5, and 8.4. For mutant E113A, spectra are shown at five pH values as noted in the appropriate panel: 5.9, 6.2, 6.5, 6.8, 7.1, Only the spectrum of mutant pigment E113A displayed a significant pH dependence as reported previously (10). Visible, ,X values of rhodopsin and mutants at pH7.1 are presented in Table I.

4664

Counterion Base

1 " ' '

RHO

SchiffRhodopsin

t

E113A/Al17E/E122Q~

0.04

0.00 0'02

0.03

m

0.02

0.01

s rd

P

0.00

F.

$

0.04

n

0.01

4

0.03 0.00

0.02

0.01

-0.01 0.01

0.00

400

500

400

500

Wavelength(nm) FIG. 4. Effect of pH on the photobleaching properties of rhodopsin and mutant E113A/A117E/E122Q. Visible spectra are shown in the two upper panelsfor rhodopsin and E113A/A117E/ E1226 in darkness and after 1, 2, 3, 4, and 5 min of illumination at pH 5.9. The results of the same experiment at pH 8.4 are shown in the two lower panels. Thesespectra show the results of the pH extremes tested. Asummary of experiments at intermediate pH values is presented in Fig. 5. Rhodopsin formed exclusively MI1(Xmax = 380 nm) upon illumination under all conditions tested. No MI (Xmax = 480 nm) was formed under the conditions tested. For rhodopsin, the rate of MI1 formation was essentially independent of pH. The mutant E113A/A117E/E122Q upon illumination formed an MI-like species with a Xmax value of 475 nm at pH 5.9. However, at pH 8.4 an MIIlike species with a, , ,X value of 380 nm was formed. The results of experiments at intermediate pH values are given in Fig. 5.

upon illumination with low temperature or increasing pH tending to stabilize the MI form. At pH 5.9 under conditions in which illumination of rhodopsin caused the rapid formation of MI1 at 380 nm, illumination of the triple mutant E113A/A117E/E122Q resulted in the shift of the visible absorbance peak from a Amax value of 490 nm to a value of about 475 nm (Fig. 4). Only a small amount of 380 nm species was formed with continued illumination. Furthermore, the photobleaching behavior of the triple mutant was pH-dependent. At basic pH of 8.4, the 475 nm species was absent. Illumination caused the rapid conversion of the 490 nm peak to a peak in thenear UV range. This pH dependence in mutant E113A/A117E/E122Q with decreasing pH favoring an MI-like form and increasing pH favoring an MII-like form is the inverse of that reported for rhodopsin (5, 24, 27). A comparison of the pHdependence of photobleaching for rhodopsin and each of the mutant pigments is presented in Fig. 5. Rhodopsin and mutant pigment E122Q were insensitive to pH effects and showed a rapid conversion to MI1 and MII-like forms, respectively, upon illumination. As reported previously (6), mutantpigment E113D displayed a mixture of MI-like and MII-like spectral forms upon illumination. This photobleaching behavior showed no significant pH dependence. Only at basic pH of 8.4 was a slight shift toward the MII-like form noted. The pHdependence of the photobleaching of E113A/A117E/E122Q is clearly demonstrated over a range of pH values. The MI-like form of pigment E113A/ A117E/E122Q could be converted directly to its respective MII-like form by raising the pH after illumination at acidic pH (Fig. 6).

0.00 -0.01

0.01

400

500

Wavelength(nm) FIG. 5. Effect of pH on the photobleaching properties of rhodopsin and mutants. Photobleaching difference spectraare shown for rhodopsin and each mutant in darkness and after 1min of illumination at four different pH values. The pH values used for rhodopsin and mutantpigment E122Q were 5.9,6.8,7.5, and 8.4. The photobleaching difference spectra of rhodopsin and E122Q were not sensitive to pH. The pH values used for the other mutant pigments are listed in the figure. Mutant pigment E113D formed a mixture of MI-like (490 nm) and MII-like species at all pH values, but at the most basic pH value tested, the MII-like form was somewhat favored. Mutant pigment E113A/A117E/E122Q displayed a marked pH dependence and formed predominately an MII-like species at basic pH values and a predominately MI-like species at acidic pH values (see Fig. 4). Mutant pigment E113A formed only an MII-like species upon illumination, but the rate ofMI1 formation was dependent on the amount of visible species present at a given pH (see Fig. 3).

There was also a marked pH dependence of photobleaching of mutant pigment E113A (Fig. 5). This photobleaching pH dependence was directly related to the pHdependence of the spectral forms of the pigment in the dark. At acidic pH, proportionally more of the visible form of the pigment (506 nm) was present than atbasic pH. The effect of illumination with long wavelength light (>515 nm) was therefore markedly pH-dependent. However, at all pH values, the 506 nm form was photoconverted to an MII-like 380 nm form. A MI-like form was not observed upon illumination of mutant pigment E113A. Transducin Activation by Mutant Pigments-Assays were performed to determine whether or not the MII-like spectral form of mutant pigment E113A/A117E/E122Q couldactivate transducin. The results of a single experiment are presented in Fig. 7. In the time course shown, mutant E113A/A117E/ E122Qshowed no activity in darkness. Inthe light, the mutant catalyzed the exchange of GDP for GTPyS by purified bovine transducin. The activation rate of the mutant was slightly less than thatof rhodopsin. A summary of transducin

Counterion Base

Schiff Rhodopsin

, l . dark

1

4665 TABLE I1 Summary of transducin activation ratesof rhodopsin mutants Activation rate" Relative Mutation($ (mean & S.E. (n)) activityb pmol/pmol/min

Wavelength ( n m ) FIG. 6 . Conversion of the MI-like form of mutant pigment E l 13A/A117E/E122Q to a n MII-like formby increasing pH. A visible spectrum is shown for E113A/A117E/E122Q in darkness (curve I ) and after 40 s of illumination at pH 5.6 (curve 2). A dilute solution of NaOH was then added to bring the pH to 8.7 (curve 3 ) . The MI-like form could be partially restored by the addition of dilute HCl. The presence of a Schiff base linkage at the end of the experiment was confirmed by the addition of dilute HCI to form the characteristic 440 nm species (not shown).

2 E a ,a

loo

75

1

1

A

RHO

0

E113A/A117E/E12ZQ A /

I

-1.0

-0.5

0.0

0.5

1.0

Time (min) FIG. 7. Transducinactivation by mutant E113A/A117E/ E122Q.The ability of rhodopsin and mutantE113A/A117E/E122Q t o activate transducin was evaluated by the use of a GTP-yS filter binding assay. Only the GTP-yS specifically bound to transducin binds to the filter. The amount of GTP-yS bound is plotted as a function of time. Under the conditions of the assay, the mutant pigment activated transducin a t a slightly reduced level when compared with rhodopsin. A summary of the results of transducin activation assays of all of the pigments is given in Table 11.

Rhodopsin 160 & 11 (5) 1.00 140 & 10 (4) 0.88 E113D 140 ElZZQ & 22 (4) 0.88 0.28 & 15 (4) 45 E113A 0.75 120 & 22 (4) E113A/A117E/E122Q Transducin activation rate refers to thelight-dependent exchange of GDP for GTP+ by transducin in the presence of a purified pigment. The activation rate is expressed in pmol of GTP-yS bound/ pmol of pigment/min. The values given are the mean & S.E. for n independent experiments. The amount of dark GTP-yS binding by transducin in darkness was insignificant in all cases (an average of 0.3 pmol/pmol/min). The rates were adjusted to account for the different extinction coefficients of mutant pigments (see Table I). *The values are normalized to the activity of rhodopsin purified from COS cells prepared in parallel with mutants.

absorbance changes at the, , ,X value of retinal oxime (365 nm) and at the visible Xmax value of the pigment. Only the decrease of the visible Xmax value is shown inFig. 8. The dark reactions were well characterized by a single exponential rate constant, and the half-time for the reaction (t112 value) for each pigment is given in Table 111. Rhodopsin and mutant pigments E113D and E122Q did not react significantly with hydroxylamine in darkness. The reaction half-time for recombinant rhodopsin (tl12= 261.5 min) was the same as thatfor bovine rhodopsinin dodecyl maltoside detergent.Mutant pigment E113A/A117E/E122Q reacted about 50 times faster than rhodopsin under identical conditions. Mutant pigment E113A reacted somewhat more slowly than pigment E113A/ A117E/E122Q. Light-induced Reaction of Mutant Pigments with Hydroxylamine-The kinetics of thereactions of rhodopsin and mutant pigments under continuous illumination shown are in Fig. 8B. The reactions were followed by monitoring absorbance changes at the, , ,X value of retinal oxime (365 nm) and at the visible, , ,X value of the pigment (Table I). Only the value is shown in Fig. 8. Unlike decrease of the visible, , ,X the dark reactions, under continuous illumination the kinetics of decrease in visible Xmax value (or of increase in 365 nm value) did not conform to those of a single order reaction at least in part because of the effect of photoactivation rate on the course of the reaction. Rhodopsin, mutant E113D, and mutant E113A/A117E/E122Q had qualitativelysimilar lightinduced reaction rates. Mutants E122Q and E113A appeared to have somewhatslower reaction rates.

activation datafor all of the mutants is presented in Table 11. The results of assays of E113D and E122Q relative to rhodopDISCUSSION sin agree well with previously published results obtained using G protein-coupled receptors share the structural motif of of a GTPase assay (6). Their levels of activity relative to that rhodopsin were slightly reduced (0.88 each). The activity of seven transmembrane segments. These transmembrane sega predominantly a-helical ments have been postulated to have E113A/A117E/E122Q was alsoreduced mutantpigment secondary structure (11). The borders of the third putative (0.75). The mutant pigment E113A displayed a more severe activation defect (0.28) that was most likely a result of its transmembrane helix (helix C) in rhodopsin are thought to at the intradiscal surface slow photoactivation kinetics at the pHof the assay (Fig. 5). be near an essential cysteine residue chargedpair atthe The results were notadjustedfor differences inrates of (Cys"') (20) and ahighlyconserved cytoplasmic surface ( G l ~ ' ~ ~ - A r(Fig. g l ~ ~1)) (6, 7). Glu113,the photoactivation among the mutants. Reaction of Mutant Pigmentswith Hydroxylamine in Dark- retinylidene Schiff base counterion in bovine rhodopsin, is located near the intradiscal surface, and it is likely that the ness-The reactions of hydroxylamine with rhodopsin and mutant pigments in darkness were evaluated. Mutant pig- Schiff baselinkage is within the membrane-embedded domain ments E113A and E113A/A117E/E122Q showed significant (6, 7, 28, 29). This model of rhodopsin is indirectly supported reactivity to hydroxylamine in darkness. The kineticsof the by the high resolution cryoelectron microscopy structure of darkreactionsfor E113A and E113A/A117E/E122Q are bacteriorhodopsin, the retinal-based light-driven proton shown in Fig. 8A. The reactions were followed by monitoring pump of Halobacterium halobium, in which the seven trans-

Rhodopsin Schiff Base Counterion

4666

.

1 .o

A.-

DARK

E113A

Mutation(s)

0.8

E113A/A117E/E122Q E113A E122Q E113D Rhodopsin

10.3' 182.8 239.0

0

r: 0.6

k 0


8.5) (Fig. 3). These data, takentogether with data from the single mutants thatwere evaluated as controls, are consistent with the conclusion that theglutamic acid residue introduced at position 117 does function as theSchiff base counterion in the mutant E113A/A117E/E122Q. This result supports the presumption of a helical structure for the third transmembrane segment in rhodopsin. Recently, the same conclusion was obtained by analysis of a mutant rhodopsin (E113Q/ A117D) that had properties similar to those of mutant E113A/ A117E/E122Q (30). The photobleaching behavior of mutant E113A/A117E/ E122Q was dramatically altered from that of rhodopsin under the conditions tested. Upon illumination at acidic pH, the mutant displayed a stable MI-like form (Fig. 4). At basic pH, the MII-like form of the mutant was favored. This pH dependence of the MI toMI1 transition was the inverse of that reported for rhodopsin ( 5 , 24,25). The pH dependence of photoproduct formation in rhodopsin was recently reported to be influenced by specific histidine residues ( 5 ) .The altered photobleaching pH dependence in the mutant implies that the Schiff base deprotonation in the MI to MI1 transition has been uncoupled from the influence of these histidine residues and that the dissociation of the imine proton in the MI-like photoproduct of mutant E113A/A117E/E122Q is hampered at acidic pH. Although a specific Schiff base proton acceptor has not yet been identified in rhodopsin, these results could indicate that such a residue was removed or incapacitated by the mutation. Under the conditions tested in dodecyl maltoside detergent rhodopsin formed only MI1 (Fig. 4). The MI1 form of rhodopsin was so strongly favored over MI that an

SchiffRhodopsin

Base Counterion

4667

equilibrium state between the two forms could not be experi- receptor or hydroxylamine in the case of the rhodopsin mutant. mentally demonstrated. Given this tendency, the dramatic In conclusion, the mutant E113A/A117E/E122Q has propeffect of the mutation to stabilize an MI-like form was espeerties that are consistent with a repositioning of the Schiff cially striking. base counterion by one helix turn withouta significant change The MI-like formof mutant pigmentE113A/A117E/E122Q was in apparent equilibrium with its MII-like form (Fig. 6). in the apparent Schiff base pK,. Photobleaching studies inHowever, under illumination conditions thatfavored the for- dicate that the mutation hasa marked effect on the MI-like mation of the MI-like form, mutant pigment E113A/A117E/ to MII-like transition. The results support the hypothesis protonacceptor may E122Q was able to activate transducin, albeita t a somewhat thatalthough aspecificSchiffbase mediate the MI to MI1 transition in rhodopsin, it is not reduced level (Fig. 7). This level of activity a t a pH in which a significant amount of the mutant photoproduct would be strictly requiredfor light-dependent transducin activation. expected to exist in the MI-like form (see Figs. 4 and 5) again Detailed knowledge of the chemical environment of the Schiff of light-induced protonmovements is essential supports theexistence of a n equilibrium between MI-like and base imine and MII-like forms which is influenced by transducin binding as to understanding function in rhodopsin. With this goal in in the case of rhodopsin (26, 31). It is unlikely that the MI- mind, further biophysical and spectroscopic study of the mutant rhodopsin reportedis under way. like form of mutant pigment E113A/A117E/E122Q was responsible for transducin activation (1, 2, 32). Acknowledgments-We thank P. Deval, R. Franke, and K. Fahmy The data indicate that mutant E113A/A117E/E122Q can for helpful suggestions. M. Lee and L. Yelich provided outstanding form an active MII-like species in a light-dependent pathway. technical assistance. Of course, an MII-like species can also form in light-indeREFERENCES pendent pathways. Results of assays of rhodopsin mutant 1. Longstaff, C., Calhoon R. D., and Rando, R. R. (1986) Proc. Natl. Acad. E113Q regeneratedwithall-trans-retinal suggested that a Sci. U. S. A. 83,4206-4213 U. M., Schmid, E. D., Perez-Sala, D., Rando, R. R., and Siebert, F. neutral Schiff base, a neutral residue a t position 113, and an 2. Ganter, (1989) Biochemistry 28,5954-5962 3. Braiman, M. S., Mogi, T., Marti, T., Stern, L. J., Khorana, H. G., and all-transchromophore were theminimalrequirements for Rothschild, K. J. (1988) Biochemistry 27,8516-8520 active rhodopsin (6). Recently, it has been shown that rho4. Otto, H., Marti,T., Holz, M., Mogi, T., Stern, L. J., En el F , Khorana, H. G., and Heyn, M. P. (1990) Proc. Natl. Acad. Sci. U. 8.A.87, 1018-1022 dopsin mutants show constitutive activityeven in the absence 5. Weitz, C. J., and Nathans, J. (1992) Neuron 8,465-472 of a chromophore provided that the site of the Schiff base 6. Sakmar, T. P., Franke R. R., and Khorana, H. G. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,8369-8313 linkage is neutral (32). 7. Zhukovsky, E. A,, and Oprian, D. D. (1989) Science 246,928-930 Hydroxylamine reactivity has been used as a probe of Schiff 8. Nathans, J. (1990) Biochemistry 29,9746-9752 9. Lin, S. W., Sakmar, T. P., Franke, R. R., Khorana, H. G., and Mathies, R. base accessibility in bacteriorhodopsin (33) and visual pigA. (1992) Biochemistry 3 1 , 5105-5111 ments (24). Rhodopsin is stable to hydroxylamine in darkness10. Sakmar, T. P., Franke, R. R., and Khorana, H. G. (1991) Proc. Natl. Acad. Sci. U. S. A. 88,3079-3083 but reacts rapidly upon illumination. Somecone pigments are 11. Dratz, E. A., and Hargrave, P. A. (1983) Trends Biochem. Sci. 8 , 128-131 12. Henderson, R., Baldwin, J. M., Ceska,.T. A., Zemlin, F., Beckmann, E., reactive in darkness (24). Rhodopsin mutants with replaceand Downing, K. H. (1990) J. Mol. Brol. 213,899-929 ments of Glu113 have been reported to be reactive to hydrox- 13. Franke, R. R., Sakmar, T. P., Graham, R. M., and Khorana, H. G. (1992) J. Biol. Chem. 2 6 7 , 14767-14774 ylamine in darkness as well (6, 7, 10). One interpretation of 14. Ferretti, L., Karnik, S. S., Khorana, H. G., Nassal, M., and Oprian, D. D. this findingwas that thelow Schiff base pK, allowed reaction (1986) Proc. Natl. Acad. Sci. U. S. A. 8 3 , 599-603 15. Sa_k_m_ar,T. P., and Khorana, H. G. (1988) Nuclerc Acids Res. 1 6 , 6361withthe basic (380 nm) form of themutantpigmentin W'iZ darkness (6). Interestingly,thetriplereplacementmutant, 16. Lo, K.-M., Jones, S. S., Hackett, N. R., and Khorana, H. G. (1984) Proc. Natl. Acad. Sci. U. S. A. 81,2285-2289 with an apparently normal Schiff base pK,, reactedquite 17. Franke, R. R. Sakmar T. P Oprian, D. D., and Khorana, H. G. (1988) J. Biol. Chem.'263,21?9-21'i2 rapidly withhydroxylamine (t1/2= 5.6 min) in darkness under F., Nicklen, S., and Coulsen, A. R. (1977) Proc. Natl. Acad. Sci. conditions in which rhodopsin was unreactive ( t l l P= 261.5 18. Sanger, U. S. A. 74,5463-5467 G. (1987) min) (Fig. 8). The results taken together may indicate that 19. Oprian D. D. Molday R. S. Kaufman R. J., and Khorana, H. Procl Natl. head. Sc: U. S.'A. 84, 88?4-8878 the replacement of Glu113 by alanine makes the Schiff base 20. Karnik S. S. Sakmar, T. P. Chen H.-B., and Khorana, H. G. (1988) Proc. Natl.'Acad: Sei. U. S. A. 8 5 , 84i9-8463 linkage accessible to attack from the intradiscal surface and 21. Chan, T., Lee, M., and Sakmar, T. P. (1992) J. Biol. Chem. 2 6 7 , 94789480 that reaction with a protonated Schiff base is favored. The 22. Fung, B. K.-K., Hurley, J. B., and Stryer, L. (1981) Proc. Natl. Acad. Sci. rate of reaction under continuous illumination was similar for U. S. A. 7 5 , 152-156 all of the mutants (Fig. 8). A more detailed evaluation of 23. Wessling-Resnick, M., and Johnson,G. L. (1987) J. Biol. Chem. 262,36973705 hydroxylamine reactivity as a probe of Schiff base environ- 24. Wald, G., Brown, P. K., and Smith, P. H. (1955) J. Gen. Physiol. 3 8 , 623681 ment in these mutants is underway. 25. Hofmann, K. P. (1986) Photobiochem. Photobiophys. 13,309-338 The hydroxylaminereactivityresultsarerelevantto a 26. Franke, R. R., Konig, B., Sakmar, T. P., Khorana, H. G., and Hofmann, K. P. (1990) Science 2 5 0 , 123-125 comparison of adrenergic receptors andrhodopsin. An aspar- 27. Mathews, R. G., Hubbard, R., Brown, P. K., and Wald, G. (1963) J. Gen. Physiol. 47,215-240 tic acid residue AS^"^) in helix C of a &adrenergic receptor 28. Birge, R. R. (1990) Biochim. Biophys. Acta 1 0 1 6 , 293-327 has been shown to be the counterion to cationic amine ligands 29. NakExama, T. A., and Khorana, H. G. (1990) J. Biol. Chem. 2 6 6 , 426942 I O (34). This position alignswith position 117 inrhodopsin 30. Zhukovsky, E. A., Robinson, P. R., and Oprian, D. D. (1992) Biochemistry relative to the conserved cysteine residue at the extracellular 3 1 , 10400-10405 D., Kuhn, H., Reichert, J., and Hofmann, K. P. (1982) FEBS Lett. border and the conserved charged pair at the cytoplasmic 31. Emeis, 143,29-34 32. Robinson, P. R., Cohen, G. B., Zhukovsky, E. A., and ODrian, D. D. (1992) border of helix C (6, 7). The counterion in the adrenergic Neuron 9, 719-725 receptor may be positioned on the cytoplasmic side of the 33. Subramaniam S. Marti T. Rosselet, S. J. Rothschild K. J. and Khorana, H. G. (1991j P k . Nitl. Acad. Sci. U. S.'A. 88, 2589-258$ ligand binding pocket. This might be analogous to the coun34. Strader, C. D., Sigal, I. S., Candelore M. R. Rands E., Hill, W. S., and terion position in mutant E113A/A117E/E122Q. This posiDixon, R. A. F. (1988) J. Bid. Che;. 263,'10267-i0271 35. Hong, K., and Hubbel, W. L. (1972) Proc. Natl. Acad. Sci. U. S. A. 6 9 , tioning would allow accessibility from the extracellular sur361 7-3fi31 -". face for water-soluble ligands in the case of the adrenergic 36. Karnik, S. S., and Khorana, H.G. (1990) J . Bid. Chem. 2 6 5 , 17520-17524

_"_