Conformational Changes in Guanylyl Cyclase-activating Protein 1 ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 1999 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 274, No. 28, Issue of July 9, pp. 19829 –19837, 1999 Printed in U.S.A.

Conformational Changes in Guanylyl Cyclase-activating Protein 1 (GCAP1) and Its Tryptophan Mutants as a Function of Calcium Concentration* (Received for publication, March 4, 1999)

Izabela Sokal‡§, Annie E. Otto-Bruc‡§, Irina Surgucheva§¶, Christophe L. M. J. Verlindei, Chien-Kao Wang**, Wolfgang Baehr¶‡‡, and Krzysztof Palczewski‡§§¶¶ii From the Departments of ‡Ophthalmology, §§Chemistry, ¶¶Pharmacology, iBiological Structure, and **Physiology and Biophysics, University of Washington, Seattle, Washington 98195 and the ¶Moran Eye Center, University of Utah Health Science Center, Salt Lake City, Utah 84132

Guanylyl cyclase-activating proteins (GCAPs are 23kDa Ca21-binding proteins belonging to the calmodulin superfamily. Ca21-free GCAPs are responsible for activation of photoreceptor guanylyl cyclase during light adaptation. In this study, we characterized GCAP1 mutants in which three endogenous nonessential Trp residues were replaced by Phe residues, eliminating intrinsic fluorescence. Subsequently, hydrophobic amino acids adjacent to each of the three functional Ca21-binding loops were replaced by reporter Trp residues. Using fluorescence spectroscopy and biochemical assays, we found that binding of Ca21 to GCAP1 causes a major conformational change especially in the region around the EF3-hand motif. This transition of GCAP1 from an activator to an inhibitor of GC requires an activation energy Ea 5 9.3 kcal/mol. When Tyr99 adjacent to the EF3-hand motif was replaced by Cys, a mutation linked to autosomal dominant cone dystrophy in humans, Cys99 is unable to stabilize the inactive GCAP1-Ca21 complex. Stopped-flow kinetic measurements indicated that GCAP1 rapidly loses its bound Ca21 (k21 5 72 s21 at 37 °C) and was estimated to associate with Ca21 at a rate (k1 > 2 3 108 M21 s21) close to the diffusion limit. Thus, GCAP1 displays thermodynamic and kinetic properties that are compatible with its involvement early in the phototransduction response.

The Ca21-binding motif termed EF-hand, introduced by Nockolds et al. (1), refers to the helix-loop-helix structure responsible for selective high affinity (Kd,1025 M) Ca21 binding. EF-hand motifs, reliably predictable based on primary sequence and present in one to eight copies in some polypeptides, have been identified in over 500 Ca21-binding proteins. Neuronal Ca21-binding proteins (NCBPs)1 are a subset of the * This work was supported by United States Public Health Service Research Grants EY08061 and EY08123, unrestricted grants from Research to Prevent Blindness (New York) to the Departments of Ophthalmology at the University of Washington and the University of Utah. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § These authors contributed equally to this work. ‡‡ Recipient of a Senior Investigator Award from Research to Prevent Blindness. ii Recipient of a Jules and Doris Stein Research to Prevent Blindness Professorship. To whom correspondence should be addressed: University of Washington, Dept. of Ophthalmology, Box 356485, Seattle, WA 98195-6485. Tel.: 206-543-9074; Fax: 206-543-4414; E-mail: palczews @u.washington.edu. 1 The abbreviations used are: NCBP, Neuronal Ca21-binding protein; This paper is available on line at http://www.jbc.org

EF-hand-containing proteins that are found predominantly in neurons. The function of most of these proteins is largely unknown. The topography of NCBPs is based on four EFhand motifs, some of which may be functional and some nonfunctional in Ca21 coordination (2). The sequence similarity among members of the NCBP family varies from ;25% between calmodulin (CaM) and visinin to ;60% between GCAP1 and GCAP3 (3). NCBPs are acidic and similar in length, with CaM and CaM-like proteins being the shortest (149 –150 amino acids; Mr 16,837), and other members of this family being ;200 amino acids in length (Mr ;23,000) (2, 4). GCAPs, a subgroup of NCBPs, activate guanylyl cyclase (GC) in their Ca21-free form. Three mammalian GCAPs, GCAP1 (5), GCAP2 (6, 7), and GCAP3 (3), have been characterized to date. Recently, we have identified a fourth photoreceptor Ca21-binding protein closely related to GCAPs (8). This novel protein does not stimulate GC at nanomolar [Ca21], but inhibits GC at micromolar [Ca21], and is therefore termed guanylyl cyclase inhibitory protein. A defect in exon 3 of the GCAP1 gene leading to a Y99C missense mutation has recently been linked to autosomal dominant cone dystrophy (9). The Y99C mutation has been shown to alter Ca21 sensitivity of GCAP1, leading to the constitutive activity of GC1 at high [Ca21] where normal GCAP1 is an inhibitor (10, 11). The mutant residue is adjacent (position 21) to the EF3-hand motif, which is important in inactivation of GCAP1 by Ca21 (12). Ca21-binding proteins may significantly change their conformation upon Ca21 coordination. Multiple EF-hand motifs may allow the Ca21-binding protein to respond cooperatively to changes in [Ca21]. The three-dimensional structures of NCBPs are known for both Ca21-free and Ca21-bound CaM and recoverin (13, 14), Ca21-bound GCAP2,2 and Ca21-bound unmyristoylated neurocalcin (16). GCAP2, recoverin, and neurocalcin are compact proteins (radius of gyration 16 –18 Å) made of two domains separated by a variable linker and are different from a dumbbell arrangement found in CaM and troponin C (reviewed in Ref. 17). The N- and C-terminal domains of NCBPs contain a pair of EF-hands each composed of 29 amino acids. Whereas structural studies give precise answers on conformational changes upon Ca21 chelation, much less is known about the dynamics of individual EF-hand motifs during this transition. Ca21 (occasionally also Mg21) is coordinated via oxygen GC, guanylyl cyclase; GCAP, GC-activating protein; GCAP1(W21F,W51F,W94F), rGCAP1(w2); ROS, rod outer segment(s); CaM, calmodulin; PCR, polymerase chain reaction; bp, base pair(s); MOPS, 4-morpholinopropanesulfonic acid. 2 Ames, J. B., Dizhoor, A. M., Ikura, M., Palczewski, K., and Stryer, L. (1999) J. Biol. Chem. 274, in press.

19829

Ca21 Binding to GCAP1

19830

atoms typically located on a consecutive sequence of 12 amino acids of the loop within helix-loop-helix structure (18). Importantly, the 21 positions of the EF-hand loops are most frequently occupied by a large hydrophobic amino acid (Ile, Leu, Val, Trp, Tyr, and Phe) (19). This unique position close to the EF-hand loop allows substitution by Trp without major disruption of conformation. Trp fluorescence techniques can then be used to monitor contributions of individual loops to conformational changes that occur during Ca21 coordination. In this study we used Trp-GCAP1 mutants to identify conformational changes in GCAP1 from the Ca21-bound to the Ca21-free form. We asked what role do the endogenous Trp residues play in the function of GCAP1, particularly in its conformational changes coupled to Ca21 binding and switching from activator to inhibitor of GC? Can functional protein variants be produced that provide spectroscopic signals specifically coupled to individual Ca21-binding sites? If so, what are the contributions of each of these binding sites to the functionally important conformational changes? What are the kinetics of Ca21 binding and release and associated conformational and functional switching? We found that Ca21 binding at the EF3hand motif evokes the largest conformational change, compared with other EF-hand motifs. Thus, EF3-hand region may act as a molecular [Ca21] switch between activating or inhibiting GC conformations. These findings provide mechanistic insights on how the mutation GCAP1(Y99C) is linked to autosomal dominant cone dystrophy. MATERIALS AND METHODS

Site-directed Mutagenesis, Removal of Intrinsic Trp from Wild-type GCAP1—The three mutants, GCAP1(W21F), GCAP1(W51F), and GCAP1(W94F) were generated by inserting mutant cassettes into wildtype pVL941bGCAP1 expressing GCAP1 in insect cells. For GCAP1(W21F), sense primer T1 (59-AGC AGC ACC GAG TGC CAC CAG TTC TAC AAG) carrying a DraIII site (underlined) and the TGG to TTC mutation and antisense primer W233A (59-CCA GTG AAA CAG CAG GCA CCA CCG TAC ACA C) were used to amplify the mutant GCAP1 product with pVL941bGCAP1 as template (2 ng/50 ml PCR mixture, 100 ng of each primer). Standard concentrations of MgCl2 and dNTP and regular Taq polymerase (Promega) were used. PCR conditions were 94 °C for 1 min, 69 °C for 1 min, 72 °C for 2 min, cycled 35 times. The product of 638 bp was purified (Qiagen) and digested with DraIII at 37 °C for 1 h and then SfiI was added and restriction continued for an additional hour at 50 °C. For ligation, pVL941bGCAP1 was digested with DraIII and SfiI. For generation of GCAP1(W51F), sense primer T2 (59-CTC CCG CAG GCC TGA GCG ATG) and antisense primer T2 (59-GCT CCA CGT ACT GGC TGG CGA ACG GGC TCA) were used to amplify a 188-bp PCR product. For ligation, pVL941bGCAP1 was digested with PrmI/StuI for 4 h at 37 °C. For generation of GCAP1(W94F), the intrinsic AlwNI site in pVL941 had to be deleted first. Briefly, pVL941 was linearized by digestion with AlwNI, blunt-ended, and re-ligated. Then bGCAP1 in pBluescript-SK was digested with EcoRI/XhoI and the fragment introduced into pVL941(AlwNI2) and digested with the same enzymes. A 418-bp PCR product was amplified with sense primer T3 (59-TGG AAC AGA AGC TGC GTT TCT ACT TCA-39) (carrying the mutation) and antisense primer W233A. The PCR product and vector pVL941bGCAP1(AlwNI2) were cut with AlwNI at 37 °C and then SfiI at 50 °C. To isolate the mutant constructs, mutant PCR cassettes were ligated (T4 ligase at 16 °C for 12 h) to the corresponding vectors from which the nonmutant product had been removed by gel electrophoresis. The ligation mixture was transformed in XL1-Blue competent cells. Clones and DNA samples were isolated by standard procedures, and inserts were completely sequenced. DNA vectors containing the correct mutation in GCAP1 were purified using CsCl centrifugation, verified by sequencing, and transfected into HighFive insect cells (PharMingen). Generation of GCAP1(W21F,W51F,W94F) or GCAP1(w2)—To obtain this construct, first, a double mutant GCAP1(W21F,W51F) fragment was generated. T1 sense and W233A antisense primers were used to generate a mutant fragment by PCR on pVL941bGCAP1(W51F) as template (see description for W21F). The resulting 650-bp PCR product carrying W21F and W51F mutations was ligated into a vector carrying the W94F mutation. The vector pVL941bGCAP1(W94F) was treated

FIG. 1. A representation of the primary amino acid sequence of GCAP1. Trp residues of native GCAP1 are located in positions Trp21, Trp51, and Trp94. Residues Phe63, Tyr99, Ile143 that precede EF2, EF3, and EF4-hand motifs, respectively, were mutated individually or in combination with Trp residues. White E letters on black background (E75, E111, and E155) mark Glu residues that are involved in Ca21 coordination within EF-hand motifs (position 12 within the Ca21-binding loop). C-terminal shaded residues are deleted in GCAP1 truncation mutants GCAP1-(1–183) and GCAP1-(1–164). with AlwNI and DraIII, and a 335-bp piece was excised and replaced by the (W21F,W51F) fragment. Ligations, transformations, sequencing, and transfections were performed as described above. Introduction of Trp into Positions Adjacent to EF2, EF3, and EF4 in GCAP1(w2)—These mutations were generated with a site-directed mutagenesis kit (Qickchange, Stratagene). For generation of W2GCAP1(w2), the two complementary W2s primer (59-TTG AGA CCT GGG ACT TCA) and W2a primer (59-TGA AGT CCC AGG TCT CAA) were used. For generation of W3-GCAP1(w2), the complementary primers W3s (59-AAG CTC TGG GAC GTG GAC and W3a and 59-GTC CAC GTC CCA GAG CTT) were used. For W4-GCAP1(w2), the primer pair W4s (59-TTC TCC AAG TGG GAC GTCA) and W4a (59-TGA CGT CCC ACT TGG AGA A) was employed. The PCR conditions were 95 °C for 30 s, 49 °C for 1 min, 68 °C for 8 min repeated 16 times. pVL941bGCAP1(w2) (10 ng) was used as a template. To clone the mutant vectors, pVL941bGCAP1 and the three mutant vectors in pBluescriptSK (deleted native Trp, and carrying Trp63, Trp99, and Trp143, respectively) were digested with StuI at 37 °C and SfiI at 50 °C. The linearized pVL941bGCAP1 vector and the mutant fragments were ligated, transformed, purified, and transfected into HighFive insect cells as described above. Generation of W2W3W4-GCAP1(W21F,W51F,W94F) or W2W3W4GCAP1(w2)—To obtain this construct, pBluescript-SKbGCAP1(W21F,W51F,W94F,F63W) was used as template (10 ng) in a PCR reaction in two steps. In the first step, primers W3s and W3a (to introduce Y99W) were used. In the second step, the PCR product from this amplification step was used as template with primers W4s and W4a (to introduce I143W). pVL941bGCAP1 and the mutant constructs after PCR were digested with StuI/NheI, religated, transformed, purified, and finally transfected into HighFive insect cells.


19831

FIG. 2. Characterization of GCAP1 and GCAP1(w2). A, fluorescence emission spectra of GCAP1 and GCAP1(w2). Fluorescence spectra were measured as described under “Materials and Methods” at [Ca21] , 10210 M (solid line) and at [Ca21] 5 1025 M (dashed line). The emission spectra of GCAP1 and the mutant were recorded using excitation at 280 and 290 nm (inset). There is no significant change in spectral shape for GCAP1 in different [Ca21]. B, Ca21 titration of GC activity in washed ROS membranes in the presence of GCAP1 (;2 mg). Inset, the activity of ROS-GC (a) without GCAP1, (b) the activity of ROS-GC in the presence of GCAP1 (;2 mg) and [Ca21] 5 2 3 1026 M, and (c) the activity of ROS-GC in the presence of GCAP1 (;2 mg) and [Ca21] 5 2 3 1028 M. C, Ca21 titration of GC activity in washed ROS membranes in the presence of GCAP1(w2) (;2 mg). The dotted line represents the Ca21 titration profile of GC activity in washed ROS membranes in the presence of GCAP1 (fitted to the scale). Inset, the activity of ROS-GC (a) without GCAP1(w2), (b) in the presence of GCAP1(w2) (;2 mg) and [Ca21] 5 2 3 1026 M, and (c) in the presence of GCAP1(w2) (;2 mg) and [Ca21] 5 2 3 1028 M. Expression of GCAP1 Mutants in Insect Cells—The transfer vector pVLbGCAP1 was constructed by subcloning a full-length DNA fragment encoding bovine GCAP1 mutants into pVL1393 vector (Invitrogen). HighFive insect cells (2–3 3 106) derived from the cabbage looper (PharMingen) were cotransfected with 0.5 mg of BaculoGold DNA and 5 mg of pVLGCAP1 mutants in 25-cm2 tissue culture flasks. Mutants were purified using G2 affinity chromatography (7). The final protein preparation shows characteristic Ca21-dependent mobility shift on SDS-polyacrylamide gel electrophoresis and was .90% pure as determined by SDS-polyacrylamide gel electrophoresis and Coomassie staining (see comparable gels in Ref. 12). GC Activity Assay—Washed bovine ROS were prepared, reconstituted with GCAP1 and its mutants, and assayed as described previously (7). Sixty-four ml of the reaction mixture contained 50 mM HEPES, pH 7.8, containing 60 mM KCl, 20 mM NaCl, 10 mM MgCl2, 0.4 mM EGTA, 2 mM isobutylmethylxanthine, 0.4 mM ATP, 0.16 mM CaCl2 (45 nM [Ca21]), and washed ROS (80 mg of rhodopsin). [Ca21] was calculated using the computer program “Chelator 1.00” (20) and adjusted to higher concentrations in some assays by increasing the amount of CaCl2. The reaction was initiated by the addition of 1.3 mM [a-32P]GTP (8,000 – 9,000 cpm/nmol). The activity is expressed as the amount of cGMP formed per min per 100 mg of ROS proteins used in the assay. All assays were repeated at least twice. Fluorescence Measurement—Fluorescence measurements of GCAP1 and its mutants were carried out on a Perkin-Elmer LS 50B spectrofluo-

rimeter using a 1 3 1-cm quartz cuvette. Emission spectra were recorded with excitation at 280 nm or 290 nm and at 5-nm slit widths. Spectra were determined in 50 mM HEPES, pH 7.8, containing 60 mM KCl, 20 mM NaCl, 1 mM EGTA, 1 mM dithiothreitol, and 10210 to 1025 21 M CaCl2. [Ca ]free was calculated as noted previously (20). Kinetic Measurements—The kinetics of Ca21 binding to GCAP1 mutants were performed on an Applied Photophysics stopped-flow spectrophotometer. The dead time of the stopped-flow apparatus was determined to be 1.1 ms. The light source consisted of a 150-watt xenon lamp. Excitation and emission wavelengths were 290 and 340 nm, respectively. The slit widths were 5 and 10 nm for excitation and emission lights, respectively. Measurements were done in 20 mM MOPS, pH 7.0, containing 70 mM KCl, 30 mM NaCl, 1 mM dithiothreitol, and 0.1 mM EGTA. [Ca21] was adjusted by the addition of CaCl2 and EGTA (20). In a typical experiment, dissociation of Ca21 from GCAP1 was measured at 1–3 mM protein in the presence of 0.15 mM CaCl2. This solution (50 ml) was mixed in the stopped-flow apparatus with an equal volume of buffer containing 0.5–2.1 mM EGTA. Data were expressed as a function of free (EGTA) just after mixing, considering that each protein molecule binds three Ca21 molecules, and each EGTA molecule binds one Ca21 molecule in this buffer. Fluorescence Quenching—Steady state fluorescence measurements were carried out in an SLM 8000C fluorescence spectrometer in the ratio mode by photon counting. Fluorescence quenching was measured after addition of small aliquots of an acrylamide solution to the protein


19832

TABLE I Ca21 sensitivity and the maximal stimulation of ROS GC activity by GCAP1 and its mutants The activity of GCAP1 and its mutants were measured using rod outer segment preparations as a source of photoreceptor GC(s) (see “Materials and Methods”). Mutants

The maximal activity of ROS GC

The maximal activity of ROS GC

nmol/min

%

nM

GCAP1 GCAP1(W21F) GCAP1(W51F) GCAP1(W94F) GCAP1(W94Y) GCAP1(w2) W2W3W4-GCAP1(w2) W2-GCAP1(w2) W3-GCAP1(w2) W4-GCAP1(w2)

115 6 21 51 6 15 96 6 12 65 6 18 60 6 20 32 6 6 39 6 10 60 6 14 30 6 8 52 6 2

100 44 84 56 52 28 34 52 26 45

260 6 25 360 6 11 290 6 20 320 6 21 330 6 15 190 6 20 210 6 50 310 6 52 290 6 43 280 6 32

IC50 for Ca21

sample. Protein samples were excited at 290 nm, and emission intensities were recorded at 340 nm. These fluorescence intensities were corrected for the inner filter effect resulting from the addition of acrylamide to the protein solution (21). Fluorescence quenching data were then analyzed according to the following modified Stern-Volmer equation, F0/F > 1 1 ~KSV 1 V!@Q# 5 1 1 Kapp@Q#

(Eq. 1)

where F0 and F are the fluorescence intensities in the absence and presence of quencher, Q, respectively. KSV is the Stern-Volmer quenching constant (also called dynamic quenching constant) for the collisional quenching process. V is the static quenching constant in relation to a certain effective volume in the immediate vicinity of the fluorophore. The above modified Stern-Volmer equation is derived from SternVolmer equation F0/F 5 (1 1 KSV[Q])eV[Q], as [Q] becomes very small. Fluorescence Anisotropy—Fluorescence anisotropy was determined with the SLM 8000C fluorimeter equipped with a Glan-Thompson polarizer assembly and a micro-magnetic stirrer. The anisotropy is defined as follows, A 5 ~I\ 2 I'!/~I\ 1 2I'!

(Eq. 2)

where Ii and I' are the fluorescence intensities observed parallel and perpendicular to the direction of polarization of the exciting light, respectively. Standard methods were applied to correct for unequal sensitivities of the detector system for vertically and horizontally polarized light (22). The measurements were done in 20 mM MOPS, pH 7.0, containing 50 mM KCl, 30 mM NaCl, and 3 mM W3-GCAP1(w2) at 15 6 0.1 °C (lex 5 290 nm, lem 5 340 nm). [Mg21] was 0.4 mM, and [Ca21] was 0.5 mM in the sample with cations. The calculations of the rotational correlation time was carried out as described previously (23). Homology Model of GCAP1—A homology model of GCAP1 was created on the basis of the crystal structure of unmyristoylated recoverin (Protein Data Bank entry: 1REC (13)), taking advantage of the sequence alignments (5, 7). The model was generated with the HOMOLOGY module of the INSIGHTII software (Molecular Simulations, Inc., San Diego, CA) using established homology modeling protocols (24). In short, protein backbone coordinates were taken from recoverin for all helices, strands, EF-hand motifs, and loops with identical lengths. Coordinates for other loops were transplanted from appropriate Protein Data Bank entries. Coordinates of conserved side chains were kept. Nonconserved side chains were built from a rotamer data base. Finally, 2,000 steps of conjugate gradient energy minimization were executed to alleviate small irregularities in the structure. This model is C-terminally truncated after Thr170 because of the lack of sequence identity with the recoverin structure. Although the structure of GCAP1 is unknown, superposition of the main chain structures of unmyristoylated, Ca21-bound GCAP-2, recoverin, and neurocalcin indicates that the root mean square deviation of the main chain atoms (in the EF-hand motifs) is 2.2 Å in comparing GCAP-2 to recoverin and 2.0 Å in comparing GCAP-2 to neurocalcin.2 RESULTS AND DISCUSSION

Trp Residues in Native and Mutant GCAP1—Trp fluorescence of GCAP1 and its mutants was employed to explore conforma-

FIG. 3. Characterization of GCAP1s with individually mutated Trp residues. A, C, and E, Ca21 titration of GC activity in washed ROS membranes in the presence of ;2 mg of GCAP1(W21F), GCAP1(W51F), or GCAP1(W94F) (solid line), respectively. The dotted line represents the Ca21 titration profile of GC activity in washed ROS membranes in the presence of GCAP1. The insets show bar graphs depicting GC activity with and without respective mutant GCAP1 at distinct Ca21 concentrations as described in the legend to Fig. 2. B, D, and F, fluorescence emission spectra of GCAP1(W21F), GCAP1(W51F), and GCAP1(W94F), respectively, at Ca21 concentrations as indicated in the legend to Fig. 2. The emission spectra of GCAP1(W21F) were recorded using excitation at 290 nm. There is no significant change in spectral shape for GCAP1 mutants in different [Ca21].

tional changes within these proteins evoked by Ca21 binding. Native GCAP1 contains three Trp residues in positions 21, 51, and 94 (Fig. 1) that would interfere with fluorescence measurements of mutant GCAP1. These residues were replaced by Phe or Tyr residues in GCAP1(W21F), GCAP1(W51F), GCAP1(W94F), and GCAP1(W94Y) and in a triple mutant (GCAP1(W21F,W51F, W94F) or GCAP1(w2)) lacking all Trp residues (Fig. 1). We investigated the fluorescence properties of GCAP1 mutants (Fig. 1) lacking endogenous Trp residues in which reporter Trp residues were placed in front of EF2-EF4 (W2W3W4-GCAP1(w2)), EF2 (W2-GCAP1(w2)), EF3 (W3-GCAP1(w2)), and EF4 (W4GCAP1(w2)). GCAP1 and its mutants were expressed in insect


19833

TABLE II Fluorescence properties of GCAP1 and its mutants The fluorescence measurements were carried out as described under “Materials and Methods.” lmaxa

Mutants 10

210

1025

M

nm

GCAP1 GCAP1(W21F) GCAP1(W51F) GCAP1(W94F) GCAP1(W94Y) GCAP1(w2) W2W3W4-GCAP1(w2) W2-GCAP1(w2) W3-GCAP1(w2) W4-GCAP1(w2)

345 6 0.5 345 6 0.5 344 6 1.0 346 6 1.0 343 6 0.5 NDd 345 6 0.5 343 6 0.5 345 6 1.0 345 6 1.0

Dlmaxb

DIc

nm

%

2.0 0 1.0 0.0 1.0 NDd 22.0 22.0 29.0 22.0

2 220 13 227 244 NDd 213 211 32 215

M

347 6 1.0 345 6 0.5 345 6 0.5 346 6 0.5 344 6 0.5 NDd 343 6 0.5 341 6 1.0 336 6 0.5 343 6 1.0

a

The wavelength of the maximum fluorescence at the indicated [Ca21]. Excitation at 290 nm. The difference between the maximum fluorescence at [Ca21] 5 1025 M minus the maximum fluorescence at [Ca21] 5 10210 M. Excitation at 290 nm. c The change in the intensity of fluorescence was defined as DI 5 [(I at [Ca21] 5 1025 M) 2 I at [Ca21] 5 10210 M)/(I at [Ca21] 5 10210 M)] 3 100. Excitation at 290 nm. d Not determined. b

FIG. 4. Characterization of W2GCAP1(w2). A, Ca21 titration of GC activity in washed ROS membranes in the presence of W2-GCAP1(w2) (;2 mg). The dotted line represents the Ca21 titration profile of GC activity in washed ROS membranes in the presence of GCAP1. The inset shows a bar graph depicting GC activity with and without mutant GCAP1 at distinct Ca21 concentrations as described in the legend to Fig. 2. B, fluorescence emission spectra of W2-GCAP1(w2) using excitation at 290 nm at [Ca21] as indicated in Fig. 2. C, changes in fluorescence at lex 5 290 nm and lem 5 343 6 3 nm as a function of [Ca21]. D, fluorescence difference spectra for W2GCAP1(w2) measured at [Ca21] 5 1025 M minus 10210 M.

cells, immunoaffinity-purified to apparent homogeneity, and their properties stringently characterized in enzymatic assays and by fluorescence spectroscopy. Characterization of GCAP1 and GCAP1(w2)—The fluorescence spectra of GCAP1 displayed complex changes as a result of Ca21 addition. First, addition of Ca21 caused a decrease in fluorescence intensity with a minimum at 100 –200 nM [Ca21]. Further addition of Ca21 led to a 2% increase in fluorescence intensity (Ref. 25; Table II). This subsequent change in the fluorescence intensity correlates with a transition of GCAP1 from an activator to an inhibitor of photoreceptor GC. Ca21 also caused a 2-nm bathochromic shift in fluorescence lmax (Table II). In addition to 3 Trp residues, GCAP1 contains 7 Tyr and 12 Phe residues (;10% of all amino acids) located mostly in the N-terminal part of the protein (5). Tyr and Phe residues contribute to the suppressed fluorescence spectrum of GCAP1(w2),

which displayed at lex 5 280 nm maximal emission at 305 nm, typical for proteins lacking Trp residues. Relative to the signal from normal GCAP1, the fluorescence of this mutant further decreased in its intensity at lex 5 290 nm. Ca21 had no effect on fluorescence properties of the mutant at either excitation wavelength (Fig. 2A). GCAP1 activated photoreceptor GC at [Ca21] , 100 nM (Fig. 2B, inset, c) and inhibited at micromolar [Ca21] (Fig. 2B, inset, b; see also Ref. 12; for GCAP2, see Ref. 26). Maximal stimulation of ROS GC by GCAP1 was 115 6 21 nmol/min and IC50 for Ca21 was observed at 260 6 25 nM (Table I, Fig. 2B). The maximal stimulation of ROS GC activity by GCAP1(w2) was about one-third as high as that of GCAP1. This mutant inactivated GC at micromolar [Ca21], and its IC50 for Ca21 was lowered to ;190 nM (Fig. 2C, Table I). These data suggest that endogenous Trp residues, to some degree, affect overall inter-

19834


FIG. 5. Characterization of W3GCAP1(w2). A, Ca21 titration of GC activity as described in the legend to Fig. 4A. The inset shows a bar graph depicting GC activity with and without mutant GCAP1 at distinct Ca21 concentrations as described in the legend to Fig. 2. B, fluorescence emission spectra of W3GCAP1(w2), experimental conditions are as in the legend to Fig. 4B. C, changes in fluorescence at lex 5 290 nm and lem 5 340 6 5 nm as a function of [Ca21]. D, fluorescence difference for W3GCAP1(w2) measured at [Ca21] 5 1025 M minus 10210 M. E, fluorescence quenching of W3-GCAP1(w2) by acrylamide. The measurements were done in 20 mM MOPS, pH 7.0, containing 50 mM KCl, 30 mM NaCl, and 3 mM W3-GCAP1(w2) in 15 6 0.1 °C. lex 5 290 nm, lem 5 340 nm. The samples were with or without 0.4 mM [Mg21] plus 0.5 mM [Ca21].

action with GC and/or affinity for Ca21, and when these residues are replaced by Phe, the mutant GCAP1 appears to be more efficiently inactivated by Ca21 (no stimulation of GC above 300 nM). Importantly, endogenous Trp residues are nonessential for overall activity/structure of GCAP1, allowing kinetic and steady state analysis of conformational changes induced by Ca21 in GCAP1 mutants lacking native Trp residues. Characterization of GCAP1 Mutants in Which Single Trp Residues Were Replaced by Phe—Mutation of Trp21 to Phe in GCAP1 produced a protein that was similar to GCAP1, but exhibited slightly modified Ca21 sensitivity, decreased the maximal stimulation of ROS GC (;50% of normal GCAP1) (Fig. 3A, Table I), similar affinity for ROS GC (Fig. 3A, inset), and decreased fluorescence without a shift in lmax upon addition of Ca21 (Table II and Fig. 3B). Changes in fluorescence at lex 5 290 nm were similar to fluorescence changes at lex 5 280 nm (data not shown). Assuming that the fluorescence of Trp residues in GCAP1 is additive, these results suggest that Trp21 increases GCAP1’s fluorescence upon addition of Ca21 and blue-shifts its fluorescence at .1027 M [Ca21]. One of several interpretations would be that the N terminus around residue 21 is exposed to a more hydrophobic environment after conformational rearrangement upon addition of .1027 M [Ca21]. This change in the N-terminal portion could be a part of inactivation of GCAP1 by Ca21. This hypothesis is consistent with findings that deletion of the N-terminal 10 and 20 residues in GCAP1 leads to only partial inactivation by [Ca21] . 1027 M (25). Mutation of Trp51 to Phe in GCAP1, located close to the EF2-hand motif, produced protein that was similar to GCAP1 in stimulation of GC and had a small increase in fluorescence upon Ca21 addition (Tables I and II and Fig. 3, C and D). In contrast to other single mutants, Trp51 may decrease its fluorescence intensity upon addition of Ca21 because it may become exposed. In normal GCAP1, this decrease may be compensated by an increase in fluorescence due to movements of Trp21 and Trp94 to more hydrophobic environments. Consistent with this idea is property of the next mutant. GCAP1(W94F) had Ca21 sensitivity similar to GCAP1, a ;50% decreased maximal stimulation of ROS GC activity, and a ;27% decreased fluorescence intensity with a lmax that does not change upon addition of Ca21 (Tables I and II and Fig. 3, E and F).

Characterization of GCAP1 Mutants in Which Trp Residues Were Placed in Front of Ca21 Loops—Three Trp residues were introduced to the GCAP1(w2) molecule in front of each of the EF2-, EF3-, and EF4-hand motifs (residues 63, 99, and 143). W2W3W4-GCAP1(w2) had the maximal stimulation of ROS GC activity .one-third of that for GCAP1, but similar IC50 (210 nM) for Ca21 inactivation (Table I). Addition of Ca21 caused blue-shifted emission by 2 nm and ;13% decrease in fluorescence, opposite to that of GCAP1 (Table II). These results suggest that Trp residues placed in the front of EF-hand motifs do not inactivate GCAP1. To assess the contributions of individual EF-hand motifs into Ca21 inactivation of GCAP1, mutants with a single reporter Trp residue at positions 63, 99, 143, respectively, were characterized (on GCAP1(w2) background). W2-GCAP1(w2), with the Trp residue in front of the EF2hand motif, exhibited properties similar to GCAP1, but had decreased maximal stimulation of ROS GC activity (;50% of that for GCAP1), a ;11% decreased fluorescence intensity, and a 2-nm blue-shifted lmax upon addition of Ca21 (Tables I and II and Fig. 4, A, B, and D). The fluorescence change was reduced by ;50% at ;200 nM [Ca21] (Fig. 4C). W3-GCAP1(w2), with the Trp residue in front of the EF3hand motif, exhibited properties similar to GCAP1, but had a decreased maximal stimulation of ROS GC activity (;70% of that for GCAP1), a ;32% increased fluorescence intensity, and a large 9-nm blue-shifted lmax upon addition of Ca21 (Tables I and II and Fig. 5, A, B, and D). The fluorescence change was increased by ;50% at ;220 nM [Ca21] (Fig. 5C). These results suggest major conformational changes around the reporter Trp residue, consistent with this Trp moving into a more hydrophobic environment upon addition of Ca21. These results are also similar to data obtained for GCAP1(W51F) with Trp residues present in positions 94 and 21. W4-GCAP1(w2), with a Trp residue in front of the EF4-hand motif had decreased maximal stimulation of ROS GC activity (;50% of that for GCAP1), a ;15% decreased fluorescence intensity, and 2-nm blue-shifted lmax upon addition of Ca21 (Tables I and II and Fig. 6, A, B, and D). The fluorescence change was increased by ;50% at ;190 nM [Ca21] (Fig. 6C). These results suggest some conformational changes around the reporter Trp residue, consistent with this Trp moving into a


19835

FIG. 6. Characterization of W4GCAP1(w2). A, Ca21 titration of GC activity in washed ROS membranes and inset as in the legend to Fig. 4A. B, fluorescence emission spectra of W4GCAP1(w2) under conditions as described in the legend to Fig. 4B. C, changes in fluorescence at lex 5 290 nm and lem 5 343 6 3 nm as a function of [Ca21]. D, fluorescence difference for W4GCAP1(w2) measured at [Ca21] 5 1025 M minus 10210 M.

more hydrophilic environment upon addition of Ca21. They point to the importance of this region in the interaction with ROS GC in sensing changes in [Ca21]. This hypothesis is supported further by 1) the truncation mutant of GCAP1-(1–164), which displayed Ca21 binding properties very similar to normal GCAP1, but does not activate ROS GC (data not shown) and 2) by inactivation of EF-hand motifs by mutagenesis of GCAP1, which showed that only EF-hand motif 3 and 4 are essential in inactivation of GCAP1 (12). It appears that the 164 –183 segment of GCAP1 is an important region in GC activation, because a truncation mutant GCAP1-(1–183) (containing EF4) is active and has properties indistinguishable from GCAP1 (data not shown). The C-terminal region (184 – 204) is disordered in the NMR structure of GCAP2.2 Properties of Trp in W3-GCAP1(W21F,W51F,W94F) in the Presence and Absence of Cations—The emission anisotropy of fluorescence was measured for W3-GCAP1(w2) in the presence or absence of Mg21 and Ca21. The anisotropy for this mutant was significantly higher in the absence of cations (0.144) than in the presence of cations (0.091). A decreased anisotropy indicates a decreased molecular size (or a more compact structure of GCAP1-Ca21), or an increased mobility. This is also consistent with the shorter rotational correlation times in the presence of cations for W3-GCAP1(w2) (46.338 6 1.943 in Ca21free, 36.999 6 1.257 in Ca21-bound). Accessibility of the Trp residues to small, neutral acrylamide was addressed by steady state fluorescence quenching. The fluorescence emission and excitation maxima were unchanged in the presence of acrylamide (data not shown). The apparent Stern-Volmer quenching constant is a product of the lifetime of the fluorophore and collisional rate constant. Because, the Stern-Volmer constant appears to be very similar for W3GCAP1(w2) in the presence or absence of cations (Fig. 5E), the higher lifetime (intensity) in the presence of Ca21 suggests that under these conditions the collisional rate constant also must be lower. This interpretation is consistent with the view that Trp99 residue moves to a more hydrophobic environment, as observed by a ;32% increase in the fluorescence intensity, and

a large 9-nm blue-shifted lmax upon addition of Ca21 (Table II). These results further strengthen the notion that GCAP1 undergoes a major conformational change around this Trp residue. It is likely that Trp99 moves into a more hydrophobic environment upon Ca21 binding. Kinetic Studies of the Conformational Changes in W3GCAP1(w2)—EGTA-induced conformational change was studied by mixing 0.5–1.5 mM W3-GCAP1(w2) with bound Ca21 with 0.1–1 mM EGTA (a typical result is shown in Fig. 7A) in a stopped-flow apparatus. The observed rates were independent of [EGTA] and [W3-GCAP1(w2)] in the studied range (Fig. 7B). The EGTA-induced dissociation of Ca21 was fitted to the equation for the appearance of B [Ca21-free form of W3GCAP1(w2)] in a first-order reaction: A f B, where A is the Ca21-loaded form of W3-GCAP1(w2). The average rate for the conformational change induced by EGTA, k21, was ;9 s21 at 7 °C, similar to that observed for Drosophila CaM (27). This conformational change most likely reflects Ca21 dissociation from W3-GCAP1(w2). Since Kd 5 290 nM (derived from the Ca21-dependent activity titration, Table I), and since k21 at 37 °C is 72 s21 (Fig. 7C), and the estimated association rate constant is k1 . 2 3 108 M21 s21 (with Kd 5 k21/k1), close to the rate limited by diffusion (28). These calculations should be considered approximate, as they do not take into account cooperativity of Ca21 binding. Activity tests and fluorescence measurements presented in this study indicated a Hill coefficient of 1.5–2.2. This approximate rate could even be underestimated if the conformational rearrangement, measured by fluorescence change, is slower than the actual dissociation of Ca21. The temperature dependence of the conformational change in W3-GCAP1(w2) was elucidated by plotting ln kobs (Fig. 7B) against 1/T (in kelvin) according to the Arrhenius equation (Fig. 7C). kobs likely represents the k21 rate constant of Ca21 dissociation. An approximately linear relationship was found, and from the slope (2Ea/R) of the Arrhenius plot, an activation energy of Ea 5 9.3 kcal/mol of the conformational change of W3-GCAP1(w2) could be calculated.


19836

FIG. 7. Kinetics of the W3GCAP1(w2) conformational changes as a consequence of Ca21 dissociation. A, an example of a measurement done at 25 °C after mixing of 1 mM W3GCAP1(w2), in the presence of 0.075 mM CaCl2 and 1.1 mM EGTA. The kobs is 37.8 6 1.1 s21. Plot of residuals is shown as warrant to the lack of systematic error. B, the values of kobs as a function of [EGTA] at three different concentrations of W3-GCAP1(w2). The average k21 is ;9 s21 at 7 °C at 0.5 mM EGTA. Closed circles, 0.5 mM W3-GCAP1(w2); open circles, 1 mM W3-GCAP1(w2), and closed squares, 1.5 mM W3-GCAP1(w2). C, Arrhenius plot of the conformational change of W3GCAP1(w2) measured as the kobs (see text). The extrapolated value of kobs at 37 °C is e4.278 or 72 s21.

FIG. 8. Environments of 6 Trp residues in native GCAP1 and its Trp mutants. The ribbon representation of the protein is colored as follows: EF-hands 1 through 4 in yellow, wild-type Trp residues in green, new Trp residues introduced by mutation in red, and the rest of the protein in purple. The model was generated as described under “Material and Methods.”

CONCLUSIONS

Binding of Ca21 converts GCAP1 from an activator to an inhibitor3 of photoreceptor GC (for GCAP1, Ref. 12; for GCAP2, Ref. 26). Using Trp fluorescence spectroscopy, we present evidence that GCAP1 undergoes major conformational changes during this transition. The region around the EF3-hand motif, specifically Tyr99, becomes more exposed to solvent in the Ca21-free form of GCAP1 on the basis of the following observations: 1) significant increase in the Trp fluorescence intensity of W3-GCAP1(w2) upon addition of Ca21; 2) hypsochromic shift of the maximum of the fluorescence upon addition of Ca21; 3) 3 The (minor) inhibitory properties of GCAP1/GCAP2 could be an artifact of biochemical procedures, because lack of GCAPs in GCAP1/ GCAP2 double knockout mice does not change the dark current. Under dark conditions, GCAPs would be in the Ca21-loaded form and inhibitory (see Footnote 4). Identical dark currents in normal and knockout mice precludes a physiological role inhibition.

similar accessibility to quenching of the fluorescence of this mutant by acrylamide in the presence and absence of cations, while the excited fluorophore in the presence of Ca21 produce a significantly higher intensity fluorescence; and 4) increased accessibility of the Ca21-free form to limited proteolysis in contrast to the GCAP1 Ca21-bound form (12). In the model generated based on the structure of recoverin, it is apparent that the Tyr99 mutant is buried within a C-terminal a-helical domain of GCAP1 (Fig. 8). The conformational changes in recoverin are more profound than those found around other Ca21-binding loops of GCAP1. A Y99C mutation in GCAP1 is associated with autosomal dominant cone dystrophy (9). The biochemical explanation of this disease is that the mutated GCAP1 is constitutively active at high [Ca21] in conditions when normal GCAP1 inhibits GC (10, 11). In other words, this mutation changes the affinity for Ca21 to the EF3-hand motif. From several other independent

Ca21 Binding to GCAP1 lines of evidence, it is clear that the Ca21 binding properties of this loop are critical for reversal of GCAP1 stimulation (12). It is conceivable that the binding of Ca21 to this loop causes Tyr99 to move from a solvent-exposed environment into a hydrophobic pocket stabilized by van der Waals interactions, requiring Ea 5 9 kcal/mol as determined from the Arrhenius plot. When Tyr99 is replaced by Cys in the GCAP1(Y99C) mutant, the smaller and nonaromatic Cys may not provide sufficient contacts within this pocket, thus preventing stabilization of the Ca21-bound form of GCAP1. This phenomenon may be more general for this family of proteins, as typically this position is highly conserved, and large aromatic residues are present in this position. Indeed, such movement of a corresponding region was shown already for recoverin (29). Mutations of the endogenous three Trp residues in GCAP1 to Phe residues provide complementary information about specific regions of the molecule that undergo conformational changes during Ca21 binding. Because they are distributed evenly within the N-terminal domain, Trp21, Trp51 and Trp94, combined results from characterization of these mutants and W2- and W3-GCAP1(w2) suggest that the relatively large region of protein around residues 94 –99, and the N-terminal helix of EF3-hand motif, undergo structural rearrangement. Another important point of this study is the kinetics of these conformational changes in relation to phototransduction. Current studies on the kinetics of phototransduction responses in knockout mice that had GCAP1 and GCAP2 inactivated point out the importance of activation of photoreceptor GC very early in the phototransduction response.4 To elicit a fast response of GC, Ca21 dissociation from GCAP1 must be compatible with the rate of the initial phase of phototransduction. Indeed, GCAP1 loses its bound Ca21 rapidly enough to account for its importance in the early phase of phototransduction (less than 200 ms) (Ref. 30 and references cited therein). The association of Ca21 and inactivation of GCAP1 is largely limited by the diffusion rate. Thus, GCAP1 has thermodynamic and kinetic properties that fulfill a key role in photoreceptor GC activation during phototransduction response (15, 31). Acknowledgments—We thank Drs. Tomasz Heyduk, Ted Wensel, and Jack C. Saari for their valuable comments and advice during the course of these studies and Darin Bronson for expert technical assistance in sequencing DNA and primer synthesis. 4 A. Mendez, M. Burns, I. Sokal, W. Baehr, K. Palczewski, D. A. Baylor, and J. Chen, manuscript in preparation.

19837 REFERENCES

1. Nockolds, C. E., Kretsinger, R. H., Coffee, C. J., and Bradshaw, R. A. (1972) Proc. Natl. Acad. Sci. U. S. A. 69, 581–584 2. Polans, A., Baehr, W., and Palczewski, K. (1996) Trends. Neurosci. 19, 547–554 3. Haeseleer, F., Sokal, I., Li, N., Pettenati, M., Rao, P. N., Bronson, D., Wechter, R., Baehr, W., and Palczewski, K. (1999) J. Biol. Chem 274, 6526 – 6535 4. Braunewell, K.-H., and Gundelfinger, E. D. (1999) Cell Tissue Res. 295, 1–12 5. Palczewski, K., Subbaraya, I., Gorczyca, W. A., Helekar, B. S., Ruiz, C. C., Ohguro, H., Huang, J., Zhao, X., Crabb, J. W., Johnson, R. S., Walsh, K. A., Gray-Keller, M. P., Detwiler, P. B., and Baehr, W. (1994) Neuron 13, 395– 404 6. Dizhoor, A. M., Olshevskaya, E. V., Henzel, W. J., Wong, S. C., Stults, J. T., Ankoudinova, I., and Hurley, J. B. (1995) J. Biol. Chem. 270, 25200 –25206 7. Gorczyca, W. A., Polans, A. S., Surgucheva, I. G., Subbaraya, I., Baehr, W., and Palczewski, K. (1995) J. Biol. Chem. 270, 22029 –22036 8. Li, N., Fariss, R. N., Zhang, K., Otto-Bruc, A., Haeseleer, F., Bronson, J. D., Qin, N., Yamazaki, A., Subbaraya, I., Milam, A. H., Palczewski, K., and Baehr, W. (1998) Eur. J. Biochem. 252, 591–599 9. Payne, A. M., Downes, S. M., Bessant, D. A., Taylor, R., Holder, G. E., Warren, M. J., Bird, A. C., and Bhattacharya, S. S. (1998) Hum. Mol. Genet. 7, 273–277 10. Sokal, I., Li, N., Surgucheva, I., Warren, M. J., Payne, A. M., Bhattacharya, S. S., Baehr, W., and Palczewski, K. (1998) Mol. Cell 2, 129 –133 11. Dizhoor, A. M., Boikov, S. G., and Olshevskaya, E. V. (1998) J. Biol. Chem. 273, 17311–17314 12. Rudnicka-Nawrot, M., Surgucheva, I., Hulmes, J. D., Haeseleer, F., Sokal, I., Crabb, J. W., Baehr, W., and Palczewski, K. (1998) Biochemistry 37, 248 –257 13. Flaherty, K. M., Zozulya, S., Stryer, L., and McKay, D. B. (1993) Cell 75, 709 –716 14. Ames, J. B., Ishima, R., Tanaka, T., Gordon, J. I., Stryer, L., and Ikura, M. (1997) Nature 389, 198 –202 15. Koutalos, Y., and Yau, K. W. (1996) Trends Neurosci. 19, 73– 81 16. Vijay-Kumar, S., and Kumar, V. D. (1999) Nat. Struct. Biol. 6, 80 – 88 17. Crivici, A., and Ikura, M. (1995) Annu. Rev. Biophys. Biomol. Struct. 24, 85–116 18. Kretsinger, R. H. (1976) Annu. Rev. Biochem. 45, 239 –266 19. Falke, J. J., Drake, S. K., Hazard, A. L., and Peersen, O. B. (1994) Q. Rev. Biophys. 27, 219 –290 20. Schoenmakers, T. J., Visser, G. J., Flik, G., and Theuvenet, A. P. (1992) BioTechniques 12, 870 – 874 21. Parker, C. A. (1968) in Photoluminescence of Solution, Elsevier Science Publishing Co., Inc., New York 22. Lakowicz, J. R. (1983) in Principles of Fluorescence Spectroscopy, Plenum Press, New York 23. Wang, C. K., Mani, R. S., Kay, C. M., and Cheung, H. C. (1992) Biochemistry 31, 4289 – 4295 24. Ring, C. S., and Cohen, F. E. (1993) FASEB J. 7, 783–790 25. Otto-Bruc, A., Buczylko, J., Surgucheva, I., Subbaraya, I., Rudnicka-Nawrot, M., Crabb, J., Arendt, A., Hargrave, P. A., Baehr, W., and Palczewski, K. (1997) Biochemistry 36, 4295– 4302 26. Dizhoor, A. M., and Hurley, J. B. (1996) J. Biol. Chem. 271, 19346 –19350 27. Brown, S. E., Martin, S. R., and Bayley, P. M. (1997) J. Biol. Chem. 272, 3389 –3397 28. Eigen, M., and Hammes, G. G. (1963) Methods Enzymol. 25, 1–38 29. Tanaka, T., Ames, J. B., Harvey, T. S., Stryer, L., and Ikura, M. (1995) Nature 376, 444 – 447 30. Baylor, D. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 560 –565 31. Stryer, L. (1991) J. Biol. Chem. 266, 10711–10714