A Novel GCAPl Missense Mutation (L151F) in a Large Family with Autosomal Dominant Cone-Rod Dystrophy (adCORD)

Izabela Sokal,1 William J. Dupps,2 Michael A. Grassi,2 Jeremiah Bro wn, Jr,2 3 Louisa M. A ffatigato2 Nirmalya RoychowdhutyLili Yang,' Sfawomir Filipek,5 Krzysztof Palczewski,1 67 Edwin M. Stone,28 and Wolfgang Baehr'310

Purpose. To elucidate the phenotypic and biochemical charac­ ological [Ca2 f], consistent with a lowered affinity for Ca2 f- teristics of a novel mutation associated with autosomal domi­ binding to EF4. nant cone-rod dystrophy (adCORD). Conclusions. A novel L151F mutation in the EF4 hand domain Methods. Twenty-three family members of a CORD pedigree of GCAPl is associated with adCORD. The clinical phenotype underwent clinical examinations, including visual acuity tests, is characterized by early cone dysfunction and a progressive standardized full-field ERG, and fundus photography. Genomic loss of rod function. The biochemical phenotype is best de­ DNA was screened for mutations in GCAPl exons using DNA scribed as persistent stimulation of photoreceptor guanylate sequencing and single-strand conformational polymorphism cyclase, representing a gain of function of mutant GCAPl. (SSCP) analysis. Function and stability of recombinant GCAP1- Although a conservative substitution, molecular dynamics sug­ L151F were tested as a function of [Ca2 f], and its structure was gests a significant change in Ca2 ^-binding to EF4 and EF2 and changes in the shape of L151F-GCAP1. {Invest Ophthalmol Vis probed by molecular dynamics. Sci. 2005;46:1124-1132) DOklO.l l67/iovs.04-1431 Results. Affected family members experienced dyschromatop- sia, hemeralopia, and reduced visual acuity by the second to third decade of life. Electrophysiology revealed a nonrecord- T n rod and cone photoreceptors, two diffusible secondary able photopic response with later attenuation of the scotopic JLmessengers, cGMP1 and Ca2 regulate phototransduc­ response. Affected family members harbored a C—»T transition tion and recovery of photoreceptors from photobleaching. In in exon 4 of the GCAPl gene, resulting in an L151F missense the dark-adapted photoreceptor cells, concentrations of both mutation affecting the EF hand motif 4 (EF4). This change was cGMP and Ca2 f are high (1-10 and 0.5-1 /xM, respectively). absent in 11 unaffected family members and in 100 unrelated High levels of cGMP keep a portion of cGMP-gated cation normal subjects. GCAP1-L151F stimulation of photoreceptor channels in the open state, and photoreceptors are depolar­ ized. The effect of phototransduction is to lower [cGMP] rap­ guanylate cyclase was not completely inhibited at high physi- idly by activation of rod- and cone-specific cGMP phosphodi­ esterases (PDE6s), an event that closes channels located in the plasma membrane, causing hyperpolarization of the cell. In the From the Departments of ‘Ophthalmology. '’Pharmacology, and recovery phase, after shut-off of other activated phototransduc­ •'Chemistry. University of Washington. Seattle. Washington: the ■de­ tion components, cGMP must be replenished by a membrane- partment of Ophthalmology and Visual Sciences, and the "Howard associated guanylate cyclase (GC). GC is activated when Ca2 f Hughes Medical Institute. University of Iowa Carver College of Medi­ decreases after photobleaching, as a consequence of the clo­ cine. Iowa City. Towa: the Departments of 4Ophthalmology and Visual sure of cation channels and the continued extrusion of Ca2 f by Sciences. ‘■’Biology, and ll,Neurobiology and Anatomy. University of a light-insensitive Na VCa2 f-K *" exchanger.7 The Ca2 f sensi­ Utah. Salt Lake City. Utah: and the ’International Institute of Molecular tivity of GC is mediated by a set of calmodulin-like Ca2 f- and Cell Biology. Warsaw. Poland. Modeling was partly done at the TCM Computer Centre. Warsaw binding proteins termed guanylate cyclase-activating proteins University. Warsaw. Poland. (GCAPs).9,10 When the Ca2 f concentration decreases below ^Present affiliation: Ophthalmology Associates. San Antonio. 200 nM, Ca2 f dissociates from GCAP, converting it into a GC Texas. activator. Once cGMP levels are replenished to normal dark Supported by National Eye Institute Grants FY09339 (KP). levels, cation channels open again, Ca2 f levels are restored, FY08123 (WB): a grant from Research to Prevent Blindness (RPB) to and GCAP reassociates with Ca2 f and reverts into an inhibitor, the Departments of Ophthalmology at the University of Washington terminating GC stimulation. and the University of Utah: a grant from the Macular Vision Research In human retina, two GCs (GCl and -2) and three GCAPs Foundation (WTO: a Center grant of the Foundation Fighting Blindness (GCAPl to -3) have been identified.9"15 GCl (gene symbol (Owings Mills. MD) to the University of Utah: and a grant from the F. K. Bishop Foundation (KP). F.MS is an investigator at the Howard Hughes G U C Y2D ) is located on chromosome 17 at p i 3.1 and GC2 Medical Institute. {G U CY2F) at Xp22. Both cyclases are closely related in struc­ Submitted for publication December 6. 2004: accepted January 2. ture and function, and both are expressed specifically in pho­ 2005. toreceptors.1” Only GCl gene defects have been associated Disclosure: I. Sokal. None: W.J. Dupps. N one: M.A. Grassi. with Leber congenital amaurosis (LCA type i)17-ls and domi­ N one: J. Brown, Jr. None: I.M. Affatigato. None: N. Roy- nant cone-rod dystrophy (CORD l),19 whereas pathogenic mu­ chovvdluiry. N one: I,. Yang. None: S. Filipek. None: K. Palczewski. tations in the GC2 gene have not been identified. GCAPl and N one: E.M. Stone. None: W. Baehr. N one -2 are arranged on opposite strands in a tail-to-tail gene array The publication costs of this article were defrayed in part by page {GUCA1A and G U C A1B) on chromosome 6 at p21.1,2° sepa­ charge payment. This article must therefore be marked ‘advertise­ m en t* in accordance with 18 U.S.C. §1734 solely to indicate this fact. rated by a 5-kb intergenic region. The GCAP3 gene (G U CA1C), Corresponding author: W'olfgang Baehr. Moran Fye Center. Uni­ structurally identical with the GCAP1/2 genes, is located on versity of Utah Health Science Center. 15N/2030F FTHG. Salt Lake City. chromosome 3 at q13.1.lj Several missense mutations in two UT 84112: [email protected]. of three functional EF hands of GCAPl (Y99C, E155G, and

Investigative Ophthalmology & Visual Science, April 2005, Vol. 46, No. 4 1124 Copyright © Association for Research in Vision and Ophthalmology IOVS, April 2005, Vol. 46, No. 4 GCAP1-L151F and adCORD 1125

F ig ii r f 1. Clinical data. (A) Pedi­ gree of a family with autosomal dom­ inant cone-rod dystrophy (adCORD). A C—*T transition in exon 4 of GCAPl that, would be expected to result in an 1.151F missense mutation in GCAPl w as identified by SSCP and DNA sequencing. Further analysis re­ vealed that this mutation created an 7?coRI restriction site in th e gene. A 200-bp PCR product, containing this site was generated for all family m em bers and w as subjected to 7?coRI digestion. Unaffected family mem­ bers demonstrated a single uncut, band, whereas affected individuals showed a diagnostic second band {arrow). Individuals from whom blood samples were not. acquired are not. depicted. (B) Examples of fundu- scopic features in adCORD associ­ ated with the GCAPl -1.151F muta­ tion. labels correspond to patient, num bers in Table 1. Top row. Patient. 2 had mild central pigmentary macu- lopathy at. age 40, more evident, with perifoveal RPE transmission defects on angiography {rightmost panel). Middle row. Patient. 2, 17 years later, had marked central atrophy. Bottom row. Patients 4, 19 (33 and 26 years of age, respectively) showed extra- macular pigmentary abnormalities.

1143NT) have been linked to autosomal dominant cone dystro­ C.CAP gene array located at 6p21.l'° and identified a novel phy (adCD).~ A transgenic mouse model expressing missense mutation which cosegregated with affected family C.CAP1-Y99C has recently been shown to produce cone-rod members. The mutation affected C a'+-binding to KF4 and com­ degeneration.'' A fourth mutation (P50L), affecting a variable promised inhibition of activated C.CAPl, leading to persistent Pro residue present only in some C.CAPIs of various species, C.C 1 stimulation in the dark, and presumably elevated levels of has been found to be associated with adCORD.'5 '0' Several cGMP. polymorphisms not linked to disease were discovered in the C.CAP2 gene,27 and recently, a C. 157R missense mutation in the G U C AIJi gene (C.CAP2) has been suggested to be causative of M a t e r ia l s a n d M e t h o d s recessive RP.'X However, one family member carrying the mutation had an asymptomatic normal phenotype; thus, the P a ti e n ts pathogenicity of this mutation is unproved. To date, no patho­ The study was approved by the institutional review board of the genic mutations have been identified in the human C.CAP3 University of Iowa Carver College of Medicine and adhered to the gene. tenets of the Declaration of Helsinki. Patients comprised a single large Null alleles of C.C1 in humans29 and the deletion of family. There were 11 branches of the family, with 4 branches having C.C1 in mouse30 and other animals31 cause severe retinal dys­ affected family members. We use the term ‘ family” to refer to the trophies. Deletion of C.CAPl and -2 genes in mice, however, descendants of a mult.igenerat.ion pedigree of known related individu­ affects only recovery from bleaching, whereas the retina stays als (Fig. 1). Autosomal dominant, inheritance was evident, with the intact,3' suggesting that C.CAP gene expression is not essential presence of multiple affected individuals in each generation and male- for development or survival of photoreceptor cells. In to-male transmission. Twenty-six individuals participated in this study. GCAP- retinas, Ca- sensitivity of C.C is abolished and re­ In all patients, phenotypic characterization included an ophthalmic covery of both the cone driven b- and a-wave is delayed. history, assessment, of visual acuity, Farnswell-Munsworth color vision Transgenic C.CAPl, but not C.CAP2, could restore normal rod testing, and a funduscopic examination. Selected patients (Table 1) and cone response recovery.3'~3J Taken together, these ge­ received additional psychophysical testing, which included a Ganzfeld netic results suggest that C.CAPl and C.C1 form a C a '+-sensi- electroretinogram (ERG) in accordance with the recommendations of tive regulatory complex indispensable for regulation of photo­ the International Society for Clinical Electrophysiology of Vision, as transduction in rods and cones. well as a Goldmann visual field. Patients were considered affected if In this communication, we analyzed a 23-member family of they had bilateral central visual loss associated with macular retinal an autosomal dominant cone-rod dystrophy (CORD) pedigree pigment, epithelial (RPE) abnormalities and poor color vision. In all for pathogenic mutations. We focused our attention on the cases, the disease status was determined before genotyping. In cases in 1126 Sokaletal. JO KS, April 2005, Vol. 46, No. 4

T a b l e 1 . Clinical and Flectrophysiologic Findings in Affected Family Members

Age BCVA Photopic ERG Scotopic ERG Color Patient Cy) (OD, OS) (% amp)* (% am p)f Field Defect Hemeralopia DeflcitJ

1 10 20/25. 20/25 Minimal Borderline -1 -1 -| 2320/40. 20/40 NR 62% 2 40 20/60. 20/200 Minimal <50% Central 57 20/400. 20/400 NR NR 4 33 20/160. 20/50 20% 37% Central. 38 20/200. 20/50 NR 73% Paracentral 11 58 20/400. 20/400 16 27 20/40. 20/30 35 20/50. 20/50 17 85 I,P. I.P U nable 18 36 20/40. 20/40 19 26 20/40. 20/50 Slightly reduced Borderline Central - 37 20/63. 20/80 20 40 20/60. 20/50 No red Normal N one 4 4 61 20/125. 20/125 22 17 20/30. 20/40 No red 13%: no flicker Normal Paracentral 4 4 38 20/50. 20/50 78% 4 23 18 20/25. 20/25 4 4 24 17 20/20. 20/20 Delayed implicit, time Normal

BCVA. best co rrected visual acuity: I,P. light, perception only: NR. nonrecordable FRG. * b-Wave amplitude for 10-minut.e light-adapted, bright, white flash (as percentage of the minimum normal response, average of both eyes). Textual descriptions were adapted directly from summary reports when original tracings were not available. t b-Wave amplitude for 30-minute dark-adapted, dim (-2 4 dB) white flash (calculated as described for photopic response). t (4 4 ). severe dyschromatopsia based on FM-100 error score >300 or 0/14 pseudo-isochromatic plates: (4 ). FM-100 error score >200 or red dyschromatopsia by Nagel anomaloscope. which the patient had died, the disease status was inferred by clinical Expression Constructs of Mutant GCAPls history obtained from family members and from medical records, when available. The mutations were introduced into a human GCAPl plasmid (hG- CAP1'5) by site-directed mutagenesis, and the mutant. GCAPs were expressed in insect, cells. Briefly, the i'coRI-digested insert, of hGCAPl Genotyping and Mutation Analysis was cloned into pFast.Bac-1 (Invit.rogen-C.ibco, Grand Island. NY) and transformed into the XZi-Gold strain. The orientation Genomic DNA was extracted from peripheral blood by using standard Escherichia coli of the resultant, plasmid was confirmed with digests, and the techniques. For both genotyping and mutation screening. 12.4 ng of Pstl mutations were verified by DNA sequencing with the primers 5'- each patient’s DNA were used as a template in an 8.35 ptl. polymerase GTTGGCTACGTATACTCCGG and 5' -GT A AAACCT CT AC A AAT GT GG. chain reaction (PCR)containing: 1.25 ptl. 10X buffer(100 mMTris-HCl Mutagenesis to generate GCAPl-1151F was performed in this plasmid [pH 8.3]. 500 mM KC1. and 15 mM MgCU): 300 fj,M each o f dCTP. with the sense primer 5'-AACGGGGATGGGGAAXTCTCCCTGGAA- dATP. dGTP. and dTTP: 1 picomole of each primer, and 0.25 units GAG and the antisense primer 5'-CTCTTCCAGGGAGA^TTCCCCATC- polymerase (Biolase. San Clemente. CA). Samples were denatured for 5 CCCGTT (italic nucleotides introduce the mutation). Wild-type human minutes at 94°C and incubated for 35 cycles under the following GCAPl. GCAP1-F155G. and GCAP1-I143NT plasmids (pFastBac) were conditions: 94°C for 30 seconds. 55°C for 30 seconds, and 72°C for 30 generated as previously described.22 seconds in a DNA thermocycler (Omnigene: ThermoHybaid. Ashton. UK). After amplification. 5 /xl, of stop solution (95% formamide. 10 mM NaOH. 0.05% bromophenol blue, and 0.05% xylene cyanol) were SDS-PAGE and Im m unoblot Analysis added to each sample. For analysis of short, tandem repeat, polymor­ phisms (STRPs). the PCR amplification products were denatured for 3 The expression of GCAPl proteins was confirmed with UW101 anti­ minutes at. 94°C and electrophoresed on 0.4 mm denaturing gels (6% body using SDS-PAGF and immunoblot. analysis, as described.5’ AH 19:1 acrylamide-bis. and 7 M urea) with a running buffer of 1.0% TBF purification procedures were performed at. 4°C. After homogenization. at. 65 W for 3 hours at. room temperature. After electrophoresis, gels cell suspensions were centrifuged at. lOO.OOOg for 20 minutes. Super- were stained with a silver nitrate solution. For each marker, samples natants were used as a source of recombinant, proteins for protein were labeled according to allele pair size and analyzed for segregation purification. GCAPs were purified using monoclonal antibody (G2) within the family. Pair-wise linkage analysis was performed with coupled to CNBr-act.ivat.ed Sepharose resin ( —5 mg of IgG per 1 ml, of MI.INK and I.ODSCORF programs as implemented in the FASTUNK the gel) as described previously.5'’ Purified proteins were dialyzed (ver. 2.3) version of the UNKAGF program package (http:w w w .h g m p . overnight, against. 10 mM BTP (pH 7.5). containing 100 mM NaCl. The mrc.ac.uk/: provided in the public domain by the Human Genome purity of the proteins was estimated by SDS-PAGF and Coomassie Mapping Project. Resources Centre. Cambridge. UK). For single strand staining. SDS-PAGF in the presence and absence of Ca21 and limited conformational polymorphism (SSCP) analysis, amplification products proteolysis of purified GCAPl and GCAPl mutants were performed as were denatured for 3 minutes at. 94°C and electrophoresed on 6% described previously.22 5 polyacrylamide. 5% glycerol gels at. 25 W for approximately 3 hours at. room temperature. After electrophoresis, gels were stained with silver GC Activity Assay nitrate. Abnormal PCR products identified by SSCP analysis were se­ quenced by using fluorescent, dideoxynucleotides on an automated Washed rod outer segment. (ROS) membranes58 were prepared from sequencer (model 377: Applied Biosystems Inc.. Foster City. CA). AH fresh bovine retinas reconstituted with recombinant. GCAPs and as­ sequencing was bidirectional. sayed as described.5'’ [Ca2 1 ] was calculated using the computer pro­ IOVS, April 2005, Vol. 46, No. 4 GCAP1-L151F and adCORD 1127 gram Chelator 1.00*° and adjusted to higher concentrations by increas­ revealed central atrophy with marked arteriolar attenuation ing the amount of CaCK. All assays were repeated at least twice. and bone-spicule-like pigmentation in the periphery.

Fluorescence Measurements Genetic Screening and Identification of a Missense Mutation in GCAPl Fluorescence measurements of GCAPl and its mutants were per­ formed on a spectrofluorometer (IS 50B; PerkinFlmer. Wellesley. MA) Linkage analysis using short tandem repeat polymorphisms using a 1 X 1-cm quartz cuvette. Emission spectra were recorded with revealed significant linkage between the CORD phenotype and excitation at 280 nm at 5-tun slit widths in 50 mM HFPFS (pH 7.8). genetic markers from 6p21, near the locus of the peripherin-2 containing 60 mM KC1. 20 mMNaCl. 1 mMFGTA. 1 mM dithiothreitol. gene (RDS) (6p21.2-pl2.3). The maximum LOD score of 3.38 and 10 111 to 10 s M CaCl,. was obtained with marker D 6 S 1 6 5 0 . Analysis of recombination events in the pedigree revealed the limits of the disease interval M odeling GCAPl and GCAF1-L151F to lie between markers D 6 S 1 5 4 9 and D 6 S 4 5 9 , an interval that contains the RDS, GCAPl, and GCAP2 genes. However, SSCP A model of GCAPl was built using the homology modeling method in analysis and direct DNA sequencing of the coding sequences of the program Modeler*1 from the Homology Module of the Insight IT the RDS gene revealed no disease-causing sequence variations. software (ver. 2000; Accelrys Inc.. San Diego. CA) based on the crystal We therefore analyzed the coding regions of the GCAPl and structure of unmyristoylated GCAP2 with three calcium ions bound GCAP2 genes and identified a novel C—»T transition at the 5' (1JBA accession code in Protein Data Bank; pdbbeta.rcsb.org/pdb; end of exon 4 in the GCAPl gene in all affected family mem­ provided in the public domain by Research Collaborator for Structural bers (Fig. 1A). This variation changes the normal leucine resi­ Bioinformatics)/2*3 The quality of the calculated structure was due at position 151 to phenylalanine (LI5IF). It was not checked by the Profile-3D program.*1 The C-terminal region, encom­ present in any unaffected family members nor in any of more passing residues 177-201 of GCAPl extending beyond residues in the than 100 additional control subjects. Residue LI51 is located crystal structure of GCAP2. was not modeled. This 177-201 fragment within a high-affinity, C a'+-binding domain (KF4) in the C- does not contain FF hand domains and is not essential for GC-stimu- terminal domain of GCAPl. Replacement of LI51 by F is con­ lating activity.37 sidered a conservative substitution, and its linkage to disease is Homology modeling was applied for the structure of WT and surprising. I.151F-GCAP1. After energy minimization, molecular dynamic (MD) GCAPs have three high-affinity C a'+-binding sites consisting simulations were performed for WT and I.151F-GCAP1. All MD tasks of helix-loop-helix KF hand motifs,17 termed KF2, -3, and -4. The were conducted in a periodic box filled with TIPJP-type water at a KF hand is a common Ca'+-binding motif consisting of a 12- constant pressure of 1 atm and a temperature of 300 K. During the first amino-acid loop flanked by a 12-amino-acid a helix at either 100 ps of equilibration phase, the protein was frozen, and water with end, providing heptahedral coordination forthe Ca'+ ion. ,H Of sodium and chloride counterions was allowed to move. Then a 1-ns note, residues 1 and 12 of the Ca'+-binding loop are invariantly production phase of MD was applied. To double-check results, we acidic (1) or K), and residues flanking the binding loop are used Yasara (‘yet another scientific artificial reality application' ver. invariantly hydrophobic (1, L, F, Y).19 In all GCAPs, the KF1 4.9; Yasara Biosciences, http://www.yasara.org/index.html) with the motif is incompatible with Ca'+-binding because of the ab­ Yamber2 force field (modification of the Amber99 force field) and sence of key residues essential for Ca'+ coordination. Namd2 (http://www.ks.uiuc.edu/research/namd/ provided in the pub­ lic domain by Theoretical Biophysics Group. Beckman Institute. Uni­ Ca2+-Binding Characteristics o f GCAP1-L151F versity of Illinois. Chicago. II,)*5 with Charmm 27 force field (http:// www.charmm.org) for both minimization and MD. In both programs, Cosegregation of GCAP1-L151F with CORD in 16 of 26 family a particle mesh Fwald (PMF) algorithm*0 for treatment of long-range members represents a strong indication of pathogenicity. We electrostatic interactions was used. hypothesized that the mutation, although a conservative sub­ stitution, may affect Ca'+-binding to KF4, because LI5IF is located at a central position in KF4 (Fig. 2A). Residues close to R e s u l t s L151—namely, HI55 and 1143— have been found to be mu­ tated in families with cone dystrophy.'' '5 However, the mu­ Phenotypic Appearance of GCAP1-L151F Patients tations K155G and I143NT, are nonconservative and are Common symptoms in affected family members included pho­ thought to affect Ca'+-binding severely. To verify the effects of tophobia, depressed central vision, and poor color vision. In LI5IF on Ca'+-binding, we further analyzed recombinant the first decade of life, the funduscopic changes w ere minimal, GCAP1-L151F by biochemical methods. consisting of a subtle, but definite, granularity of the retinal Ca2+-Dependent Conformational Changes in WT pigment epithelium (RPB). Visualization of the RPB abnormal­ ity was enhanced by fluorescein angiography. These minimal and Mutant GCAPl findings were consistent with the patients' vision, which was Ca'+-binding proteins of the calmodulin supergene family as good as 20/20. However, color vision was consistently show a characteristic Ca'+-dependent change in conforma­ abnormal, even in patients with relatively good visual acuity. tion. In its Ca'+-bound form (KF2 to -4 occupied by Ca'+), For example, patient 1 had a complete inability to recognize GCAPl assumes a compact structure with a higher apparent any of the color plates at the age of 11. Moreover, the photopic mobility on gel electrophoresis. In its Ca'+-free form (presence KRG response was severely depressed (Table 1). later changes of 1 mM KGTA), GCAPl folds into a relaxed open structure included a progressive decline in best corrected visual acuity with a lower apparent mobility. Affinity-purified recombinant associated with the development of more pronounced macular GCAP1-L15 IF showed a C a'+-shift very similar to wild-type and atrophy (Fig. 1B). The photopic electroretinogram was nonre- GCAP1-I143NT (Fig. 2 B )'' when electrophoresed in high- and cordable in patients in the second and third decades of life. low-Ca'+ buffers. These results show that, at least at extreme Some patients noted precipitous vision loss when the macular Ca'+ buffer conditions (0 Ca'+ and 1 mM Ca'+), mutant and atrophy extended through fixation. Patients older than 55 had WT- GCAP 1 behave indistinguishably. very poor vision (Table 1). The eldest patient (no. 17) was 85 We have noted57 that in the absence of Ca'+, the open at the time of examination and had only light perception vision conformation of GCAPl is readily susceptible to proteolysis by in both eyes. Funduscopic examination of the latter individual trypsin. This is demonstrated in Figure 2C, w here it is shown 1128 Sokal et al. JOVS, April 2005, Vol. 46, No. 4

F ig u r e 2 . Location of the I.151F missense mutation and biochemical data of GCAP 1-1,15 IF. (A) I.151F is located in the Ca2 ' -binding domain FF4. An amino acid sequence alignment of vertebrate GCAPl s encompassing FF3 and -4 (C-terminal half of the molecule) is shown. This part of the molecule contains the Ca2 1 -binding sites FF3 and -4 which form hclix-loop-helix structures. FF3 and -4 are relevant for converting the Ca2 1 -bound GCAPl inhibitor of photoreceptor GC into an activator (Ca2 1 -free). Three disease-causing missense mutations (Y 9 9 C , I143NT, F155G in the human sequence) are located in this region (arrows). Conserved residues are printed on a black background. Residues conserved in six of seven sequences are printed a gray background. Prefix h, human; prefix m, mouse; prefix b, bovine; prefix f, frog (Sana pipiensy, prefix c, chicken; prefix fugu, puffer fish (Fugu rubripes)-, prefix z, zebrafish (Danio rerio). (B) Ca21-dependent mobility shift of wild-type and mutant GCAPl polypeptides. Immunoblotting of GCAPl, GCAP1-I.151F, and GCAPl (II43NT) in the presence (H ) or absence (-) of Ca2 1. The antibody used was UW14. (C, D) Limited proteolysis of GCAPl (C) and GCAPl-LI 51F (D) by trypsin. The digestions were performed at 30°C at a ratio of GCAPl/trypsin 300:1, and the digest was analyzed by SDS-PAGF at 0, 5, 10, and 20 minutes; H Ca2 1 represents 2 fxM [Ca2+] and -C a2 1 indicates 30 nM [Ca2+], Note that after 10 minutes' digestion, the high-molecular-weight components were nearly completely digested in GCAP1-L151F. that exposure of C.CAPl in the absence of Ca2 led to com­ previously characterized mutants C.CAP1-I143NT and GCAP1- plete digestion after just 5 minutes (last two lanes). In the K155C., were assayed for C.C 1 stimulation as a function of Ca2+. presence of 2 /niM Ca2+, however, proteolysis was restricted Due to previous analyses with KF3 mutations5'1,19 and KF4 because GCAPl assumed a much more compact structure mutations,22,23 we expected incomplete inactivation of C.C 1 at inaccessible to trypsin (Fig. 2C, left lanes). The compact core 700 mM Ca2+ levels found in dark-adapted photoreceptors. We of C.CAPl remained intact, even at 20 minutes of digestion. In found that C.CAPl-LI 5 IF was active at low [Ca2+J (<200 nM) contrast, C.CAP1-L151F at 2 /nM Ca2+ is much more susceptible and remained active even at levels (—700 mM) that inhibit to proteolysis. These biochemical results illustrate that the WT-C.CAP 1. This inhibition is physiologically relevant since it C.CAP1-L151F mutation prevents the protein from folding into terminates production of cGMP by C.C when Ca2+ levels have a compact structure. recovered to dark levels. The mutants C.CAP1-K155C. and C.CAP1-I143NT exhibit similar activation-inactivation kinetics. WT and Mutant GCAPl Conformations Probed Together, the findings indicate that mutations affecting KF4 by Fluorescence and -3 partially impaired C.C inhibition, leading to persistent An increase of Ca2 bound to C.CAPl causes a decrease in stimulation in dark-adapted photoreceptors. fluorescence intensity, with a minimum occurring at 200 to 300 mM ICa2+J(rrt.. Further increases in Ca2+ levels reverse this The Effect of the L151F Mutation on trend and cause an increase in fluorescence intensity (Fig. 3A). GCAPl Structure These changes in the fluorescence correlate with a structural The N- and C-terminal domains of C.CAPl each contain pairs of rearrangement of C.CAPl similarly to the Ca2+-shift experi­ KF hands that form a compact structure in the Ca2+-bound ment, and reflect the transition from an activator to an inhibitor state. The arrangement of amino acids in KF1 is incompatible of photoreceptor C.C, C.CAP1-L151F showed an incomplete with Ca2+-binding,9 whereas KF2, -3, and -4 are known to each reversal (Fig. 3B), suggesting a Ca2+-dependent structural re­ bind one Ca2+ ion, confirmed by numerous site-directed mu­ arrangement, perhaps a defect in Ca2 -binding at KF4 (caused tagenesis experiments.37,39,50,52-51 The mutant residue L151F by the L151F mutation). is located in the j3-sheet of the last KF hand domain, w here it is in a position to influence nearby Ca2+-binding sites. 1,15IF is a Persistent Stim ulation o f ROS GC by GCAP1-L151F conservative replacement and it is not obvious why affinity for We next compared the ability of C.CAP1-L151F and WT-GCAP1 Ca2+ is affected. To identify potentially significant changes in to stimulate C.C1 in vitro (Fig. 3C). Both C.CAPs, together with protein folding and Ca2+-binding at the molecular level, and to 10VS, April 2005, Vol. 46, No. 4 GCAP1-L151F and adCORD 1129

reveal how this mutation affects the structure, we performed 126 GCAP1 s 25 nM 1-ns Ml) simulations on WT and mutant C.CAPl based on the -100 50 nM nuclear magnetic resonance (NMR) structure of GCAP2.1? Ml) 150 nM allows prediction of molecule behavior through the use of • 250 nM c I 119 classic mechanics. It numerically solves Newton’s equations of D >■ \ I-/\ 300 nM motion for an atomistic model of a polypeptide such as C.CAPl f I \ 1000 nM to obtain information about its time-dependent structural prop­ i 2500 nM \\ 300 350 400 erties. We used two programs for Ml) simulations, Yasara and 1ii 1 ® 112 Wavelength (nm) Namd2, and obtained nearly identical results. Yasara is a mo- \/ lecular-graphics, -modeling and -simulation package in which autostereoscopic displays are used to visualize predicted struc­

10 ' 1CT 10’ IQ- Ca2*]free (M) tures. Namd2, described by Kale et al.,15 is specifically tailored to parallel computing platforms. This program uses spatial GCAP1(L151F) 5 200 B 216 decomposition combined with force decomposition to en­ A \ hance scalability. When predicted structures of GCAPl and * : \ GCAP1-L151F were superimposed with their C-terminal BF- \ i 100 \ hand pair domains, greater changes occurred during simulation C i \ of the L151F mutant. Direct comparison between WT and o c* 207 SG 2 \ ‘ I N LI 51F structures (Fig. 4A) revealed that there was a rotation of ^ 300 350 400 approximately 20° of the N-terminal BF-hand pair in relation to k J* Wavelength (nm) itii ^ the C-terminal pair about a long axis of the molecule. However, 198 hydrophobic interactions between interacting helices (Fig. 4A) w ere preserved. Furthermore, the mutated BF-hand domain 10'7 10-6 10-5 10"4 [Ca ,(M) was moved by approximately 0.1 nm (1 A) along the |3-sheet in - • - GCAP1 relation to WT-GCAP1. f - L • GCAP1-L151F In WT-GCAP1, five amino acids participate in the binding of 04- Ku —o—GCAP1-I143NT GCAP1-E155G Ca2+ in KF4:1)144, N 146,1)148, HI 50, and HI 55. In the mutant "c" x X structure, N146 was moved apart and 1)148 accommodated to | A form a stronger bond with Ca2+, using both acidic oxygen "5 1 \w E t \1 atoms. We calculated the distance between C:r and Asnl46 £ 0.2-I I VI (side chain oxygen Osl atom) in both mutant and WT structure • ^ during Ml). Figure 4B shows that the preferred position of O \ ^ L — o \ ■ Asnl46 in WT-GCAP1 was close to the bound C:r (average distance 0.23 nm), although during the first 300 ps it was ■ I • • • 00- separated from the calcium ion with an average distance of 10-7 10* 0.44 nm. This is contrary to the LI 51F mutant, w here the initial distance of 0.23 nm increased shortly after 100 ps to 0.57 nm F ig to f . 3. Ca2*'-dependence of fluorescence emission and biological and after 600 ps to 0.68 nm. Fluctuations in this distant posi­ activity of wild-type GC.AP1 and its mutants. Fluorescence emission of (A) WT-GC.AP1 and (B) GC.AP1-I.151F as a function o f Ca2 '. Fluores­ tion are much greater than in proximity to the C:r ion in cence was excited at Aex = 290 nm. and emission was measured at A^,, WT-GCAPl. This finding is consistent with the weakening of = 343 nm. Insets-, fluorescence emission spectra of GCAPs using Ca2+-binding to BF4, which is the molecular reason for persis­ excitation at 280 nm from 4.6 X 10 s M to 2.4 X 10 0 M Ca2'. (C) tent stimulation of GCl at physiological Ca2+. These theoreti­ Stimulation of GC activity in ROS membranes by normal and mutant cal predictions are in excellent agreement with biochemical GCAPs as a function of [Ca2 ' ]. The dark shaded area above the jc-axis and fluorescence titration experiments described earlier. indicates low [Ca2 ' expected in the light-adapted photoreceptors (~50 nM), the gray shaded area reflects high [Ca2' expected in the dark-adapted photoreceptors (500-700 nM). Ca2 ' dependence of D is c u s s io n GCAPM.151F is similar to that of GCAPl-T143NT and GCAPl-F.155G. which also carry mutations in F.F4. Autosomal dominant cone-rod dystrophy is a heterogeneous disease caused by several diverse gene defects. At least 13 loci

F ig to f . 4. Molecular modeling of WT-GCAPl and GCAPl-1.151F w ith bound Ca2'. (A) Comparison of WT-GCAPl and GCA Pl-1.151F struc­ L151F tures after 1 nsec of molecular dy­ namics. Red\ WT-GCAPl: green\ GCAPl-1.151F. T he structures are su­ perimposed at the third and fourth AWTHVVV, L151F F.F hand (F.F3 and F.F4. left). Arrows'. residues 1.151 and FI 51. Calculated ft m positions of Ca2' ions in WT and LJ mutant GCAPl are depicted by red o r green circles, respectively. (B) Dis­ tance betw een Ca2 ' in m utant F.F4 1N146 - -_ ™ ______and Asnl46 (side chain oxygen 051 MU1 N14fe I ^ atom) during MDs. Red: W T-GCAPl: green-. GCAP1-T.151F. The preferred position of Asnl46 in WT was close to bound Ca2 ' (average distance 0.23 ran), although during the first 300 ps it separated from the Ca2 ' ion by an average distance of 0.44 nm. In the GCAPl- T.151F mutant, the initial distance of 0.23 nm increased shortly after 100 ps to 0.57 nm and after 600 ps to 0.68 nm. 1130 Sokal et al. JOVS, April 2005, Vol. 46, No. 4

T a b l e 2. Genes Linked to Autosomal Dominant CORD

Gene Locus Protein Mutation Disease Ref.

GUCAIA 6p21.1 GCAPl (guanylate cyclase activating protein 1) P50T. adCORD 25 Y99C adCD 21 Y99C adCORD in mouse model 24 El 55C. adCD 23 1143NT adCD 22 T.151F adCORD This paper RDS 6p21.2 Peripherin 2 V200E adCORD 55 RIMS I 6 q l3 Regulating of synaptic membrane exocytosis R844H adCORD 56 protein 1 or rab3A-int.eract.ing molecule AIPT.l 17 p l3 .2 ATPI.l (arylhydrocarbon-int.eract.ing receptor P351A12 adCORD 57 protein-like 1) GUCY2D 17pl3.1 Guanylate cyclase 1 (GC-E in mouse) R838C adCORD 29. 58 E837D T839M UNCI 19 17(jl 1.2 HRG4 human homolog of C. elegans u n ci 19 T.57t.er adCORD 59 protein CRX 19ql3.32 Cone-rod otx-like photoreceptor homeobox E60A adCORD 60 transcription factor A196 H 1 Arg41Trp

The seven CORD genes that, were identified and cloned are listed. Only mutations leading to autosomal dominant. CORD phenotype are shown. The list, of mutations is not. comprehensive; additional mutations may exist.. linked to cone-rod dystrophy are listed in RetNet (http:// the Ca2+-binding loop of KF4, by another hydrophobic residue www.sph.uth.tmc.edu/RetNet/ provided in the public domain (F), which is not much bulkier than L. This is surprising, and by the University of Texas Houston Health Science Center, for this reason we investigated the consequences of this mu­ Houston, TX), 6 of which show an autosomal dominant inher­ tation rigorously by biochemical, biophysical, and molecular itance (Table 2). Two of these genes (GUCA 1A and G U C Y2D ) modeling techniques. The pathogenic properties of the encode C.CAPl and C.C1, respectively, which are directly in­ C.CAP1-L151F mutations described in this article are supported volved in synthesis of cGMP and phototransduetion. Other by several independent observations. First, the mutation de­ CORD genes encode a transcription factor (CRX), a PDK chap­ creases the Ca2+ sensitivity of C.C stimulation, an effect also erone (AIPL1), a structural protein (peripherin/RDS), a protein seen in Y99C, K155C., and I143NT mutations.22 The change in regulating synaptic exocytosis (RIMSl), or a protein with un­ sensitivity leads to persistent stimulation of C.C1 at high “dark” known function (HRC.4), In the C.CAPl gene, three mutations Ca2+ and to disease due to gain of function. Second, recombi­ (Y99C, K155C., and I143NT) have been linked to adCD, and nant C.CAP1-L151F is susceptible to proteolysis, because it is one (P50L) has been suggested to cause a more severe form of unable to assume a compact Ca2+-bound conformation essen­ retinal degeneration (adCORD). In C.CAP-Y99C,21 a hydropho­ tial for inactivity at dark Ca2+ (Fig. 2D). Third, MDs of WT- bic residue O') flanking the KF3 Ca2+-binding loop was re­ C.CAP1 and L151F-C.CAP1 confirmed that a significant change placed by a polar amino acid (Cys), thus distorting the KF3 in the structure of mutant C.CAPl influences the binding of motif.50-51 Transgenic mice expressing C.CAPl-Y99C w ere re­ Ca2+ in KF4 and KF2 (Fig. 4). Although residue 151 in KF4 is cently generated,21 but displayed a dominant CORD pheno­ forming a |3-sheet with KF3, no hydrogen bond is broken after type rather than a CD, as observed in humans. In the C.CAP1- mutation and during the entire simulation. Fourth, the L151F K155C. mutant,19 an acidic residue (K), 100% conserved in all mutations have been independently identified in a large Utah KF hand motifs, was replaced in KF4 by a neutral residue, pedigree with dominant cone dystrophy.05 The reason for the essentially disabling Ca2+-binding. Recently, the hydrophobic discrepancy in phenotype (adCD versus adCORD) is unclear, residue (He) flanking KF4 (1143) was found to be replaced by but anomalies in the rod response may be slow in developing two polar residues (Asn, Thr) in another family with adCD, and may depend on the genetic background. again distorting KF4 and altering C;r -binding (GCAPl- In summary, all C.CAPl mutations identified so far and 1143 NT) ” linked to retinal disease (Y99C, K155C., I143NT) are inherited In contrast to these KF-hand mutations, recombinant in a dominant fashion, and affect Ca2+ coordination and Ca2+ C.CAP1-P50L does not influence the C;r sensitivity of C.C1 in sensitivity (exception P50L, which may be a rare polymor­ vitro,01 which is consistent with the location of P50 in the phism). The KF hand mutants display a gain of function at N-terminal half between KF1 and -2. However, C.CAPl-P50L physiological high (dark) [Ca2+J, at which WT-C.CAP1 nor­ shows a marked increase in susceptibility to protease degrada­ mally would be inactive. The gain of function consists of tion and a reduction in thermal stability, as observed by CD persistent stimulation of photoreceptor guanylate cyclase lead­ spectroscopy.25 Its lower stability could reduce its cellular ing to elevated levels of cGMP in the cytoplasm of rod and cone concentration, as has been observed for C.CAPl in GC 1 / outer segments. It is not surprising that C.CAPl mutants dis­ animals.52 02 Reduction of C.CAPl levels, however, should have rupting the interface between C.CAPl and guanylate cyclase no impact on the survival of retinas, since C.CAPl+/_ and associated with retina disease have not been identified, since C.CAPlmice have morphologically normal retinas. Kither mice lacking C.CAPl and -2 display no measurable retina de­ C.CAP1-P50L causes adCORD by an entirely different mecha­ generation.?2?1 It is currently not understood why some mu­ nism involving an unknown dominant negative effect of par­ tations (K155C., Y99C) lead to adCD, whereas L151F causes a tially degraded mutant C.CAPl-P50L, or the mutation repre­ more severe adCORD phenotype. This may depend on the sents a nonpathogenic rare polymorphism. degree of stimulation in the dark-adapted photoreceptors, and We describe an adCORD phenotype based on a conserva­ the level of cytoplasmic cGMP that is established in the dark. tive substitution of a hydrophobic residue (LI 51), located in Higher levels would be considered more cytotoxic, but not IOVS, April 2005, Vol. 46, No. 4 GCAP1-L151F and adCORD 1131 quite as high levels as are thought to lead to milder pheno­ autosomal dominant cone degeneration. Invest Ophthalmol Vis types. Transgenic mice expressing mutant GCAPs and mimick­ Sci. 2004;45:3863-3870. ing the human disease may be useful in answering these ques­ 23- W ilkie SF, Li Y, Deery FC, et al. Identification and functional tions. consequences of a new mutation (F155G) in GCAPl causing au­ tosomal dominant cone dystrophy. Am J Hum Genet. 2001;69: 47 1-480. References 24. Olshevskaya FV, Calvert PD, Woodruff ML, et al. The Y99C muta­ tion in guanylyl cyclase-activating protein 1 increases intracellular 1. Lucas ISA, Pitari GM, Kazerounian S, et al. Guanylyl cyclases and C'.a2+ and causes photoreceptor degeneration in transgenic mice. signaling by cyclic GMP. Pharmacol Rev. 2000;52:375-414. J Neurosci. 2004;24:6078 - 6085. 2. Jindrova II. Vertebrate phototransduction: activation, recovery, 25. Newbold RJ, Deery FC, Walker CF, et al. The destabilization of and adaptation. Physiol Res. 1998;47:155-168. human GCAPl by a proline to leucine mutation might cause 3. Burns ME, M endez A, C hen J, Baylor DA. Dynam ics o f cyclic GMP cone-rod dystrophy. Hum Mol Genet. 2001;10:47-54. synthesis in retinal rods. Neuron. 2002;36:81-91. 26. Downes SM, Holder GF, Fitzke ra', et al. Autosomal dominant 4. Krizaj D, Copenhagen DR. Calcium regulation in photoreceptors. cone and cone-rod dystrophy with mutations in the guanylate Front Biosci. 2002;7:D2023-D2044. cyclase activator 1A gene-encoding guanylate cyclase activating 5. Baehr W, Palczew ski K. Photoreceptors and Calcium. New York: protein-1. Arch Ophthalmol. 2001;119:96-105. Kluwer Academic/Plenum Publishers/Landes Bioscience/ 27. Payne AM, Downes SM, Bessant DA, et al. Genetic analysis of the Eurekah.com; 2002. guanylate cyclase activator IB (GUCA1B) gene in patients with 6. Koutalos Y, Yau KW. Regulation of sensitivity in vertebrate rod autosomal dominant retinal dystrophies. / M ed Genet. 199936: photoreceptors by calcium. Trends Neurosci. 1996;19:73-81. 6 9 1-693. 7. Palczewski K, Polans AS, Baehr W, Ames JB. Calcium binding 28. Sato M, Nakazawa M, Usui T, Tanimoto N, Abe II, Ohguro H. proteins in the retina: structure, function, and the etiology of Mutations in the gene coding for guanylate cyclase-activating pro­ human visual diseases. Bioessays. 2000;22:337-350. tein 2 (GUCA1B gene) in patients with autosomal dominant retinal 8. Polans A, Raehr W, Palczewski K. Turned on by Ca2 1! The phys­ dystrophies. Graefes Arch Clin Fxp Ophthalmol. In press. iology and pathology of Ca2 1 binding proteins in the retina. 29. Perrault I, Rozet JM, Gerber S, et al. A retGC-1 mutation in auto­ Trends Neurosci. 1996;19:547-554. somal dominant cone-rod dystrophy. Am J Hum Genet. 1998;63: 9. Palczewski K, Subbaraya I, Gorczyca WA, et al. Molecular cloning 6 5 1-654. and characterization of retinal photoreceptor guanylyl cyclase ac­ 30. Yang RB, Robinson SW, Xiong WI1, Yau KW, Birch DG, Garbers tivating protein (GCAP). Neuron. 1994;13:395-404. DL. Disruption of a retinal guanylyl cyclase gene leads to cone- 10. Gorczyca WA, Gray-Keller MP, Detwiler PB, Palczewski K. Purifi­ specific dystrophy and paradoxical rod behavior. J Neurosci. 1999; cation and physiological evaluation of a guanylate cyclase activat­ 19:5889-5897. ing protein from retinal rods. 1994;91: Proc Natl Acad Sci USA. 31. Semple-Rowland SL, Lee NR, Van-Ilooser JP, Palczewski K, Baehr 40 1 4 -4 0 1 8 . W. A null mutation in the photoreceptor guanylate cyclase gene 11. Shyjan AW, de Sauvage FJ, Gillett NA, G oeddel DV, Lowe DG. causes the retinal degeneration chicken phenotype. Proc Natl Molecular cloning of a retina-specific membrane guanylyl cyclase. Acad Sci USA. 1998;95:1271-1276. Neuron. 1992;9:727-737. 32. Mendez A, Burns MF, Sokal I, et al. Role of guanylate cyclase- 12. Lowe DG, D izhoor AM, Liu K, et al. Cloning and expression o f a activating proteins (GCAPs) in setting the flash sensitivity of rod second photoreceptor-specific membrane retina guanylyl cyclase photoreceptors. Proc N atl Acad Sci USA. 2001;98:9948-9953. (RetGC), RetGC-2. Proc N atl Acad Sci USA. 1995;92:5535-5539. 33. Pennesi MF, Howes KA, Baehr W, Wu SM. GCAPl rescues normal 13- Subbaraya I, Ruiz CC, H elekar BS, et al. M olecular characterization cone photoreceptor responses in GCAP1/GCAP2 knockout mice. of human and mouse photoreceptor guanylate cyclase activating Proc N atl Acad Sci USA. 2003;100:6783-6788. protein (GCAP) and chromosomal localization of the human gene. 34. Howes KA, Pennesi MF, Sokal I, et al. GCAPl rescues rod photo­ J Biol Chem. 1994;269:31080-31089. receptor response in GCAP1/GCAP2 knockout mice. FMBO J. 14. Haeseleer F, Sokal I, Li N, et al. Molecular characterization of a 2002;21:1545-1554. third member of the guanylyl cyclase-activating protein subfamily. 35. Sokal I, Alekseev A, Baehr W, Haeseleer F, Palczewski K. Soluble / Biol Chem. 1999;274:6526-6535. fusion proteins between single transmembrane photoreceptor 15. Dizhoor AM, Olshevskaya FV, Ilenzel WJ, et al. Cloning, sequenc­ guanylyl cyclases and their activators. Biochemistry. 2002;41:251- ing, and expression of a 24-kDa Ca2 ' -binding protein activating 257. photoreceptor guanylyl cyclase. / Biol Chem. 1995;270:25200- 25206. 36. Gorczyca WA, Polans AS, Surgucheva I, Subbaraya I, Baehr W, Palczewski K. Guanylyl cyclase activating protein: a calcium-sen­ 16. Yang R-B, Foster DC, G arbers DL, Fulle H-J. Tw o m em brane forms sitive regulator of phototransduction. / Biol Chem. 1995;270: of guanylyl cyclase found in the eye. Proc Natl Acad Sci USA. 22029-22036. 1995;92:602-606. 17. Perrault I, Rozet JM, Gerber S, et al. Spectrum of retGCl mutations 37. Rudnicka-Nawrot M, Surgucheva I, Hulmes JD, et al. Changes in biological activity and folding of guanylate cyclase-activating pro­ in Leber’s congenital amaurosis. Fur J Ilum Genet. 2000;8:578- 582. tein 1 as a function of calcium. Biochemistry. 1998;37:248-257. 18. Perrault I, Rozet J-M, Calvas P, et al. Retinal-specific guanylate 38. Papermaster DS. Preparation of retinal rod outer segments. M eth­ cyclase gene mutations in Leber’s congenital amaurosis. N at ods Fnzymol. 1982;81:48-52. Genet. 1996;14:461-464. 39. Otto-Bruc A, Buczylko J, Surgucheva I, et al. Functional reconsti­ 19. W ilkie SF, N ew bold RJ, Deery F, et al. Functional characterization tution of photoreceptor guanylate cyclase with native and mutant of missense mutations at codon 838 in retinal guanylate cyclase forms of guanylate cyclase activating protein 1. Biochemistrv. correlates with disease severity in patients with autosomal domi­ 1997;36:4295-4302. nant cone-rod dystrophy. Hum Mol Genet. 2000;9:3065-3073. 40. Schoenmakers TJ, Visser GJ, Flik G, Theuvenet AP. CHELATOR: an 20. Surguchov A, Bronson JD, Banerjee P, et al. The human GCAPl and improved method for computing metal ion concentrations in phys­ GC.AP2 genes are arranged in a tail-to-tail array on the short arm of iological solutions. Bio/Tech. 1992;12:870-879. chromosome 6 (p21.1). Genomics. 1997;39:312-322. 41. Sail A, Potterton L, Yuan F, van VII, Karplus M. Evaluation of 21. Payne AM, D ow nes SM, Bessant DA, et al. A m utation in guanylate comparative protein modeling by MODELLER. Proteins. 1995;23: cyclase activator 1A (GUCA1A) in an autosomal dominant cone 318 -326. dystrophy pedigree mapping to a new locus on chromosome 42. Berman IIM, Westbrook J, Feng et al. The Protein Data Bank. 6p21.1. Hum Mol Genet. 1998;7:273-277. Nucleic Acids Res 2000;23:235-242. 22. Nishiguchi KM, Sokal I, Yang L, et al. A novel mutation (I143NT) 43- Ames JB, Dizhoor AM, Ikura M, Palczewski K, Stryer L. Three­ in guanylate cyclase-activating protein 1 (GCAPl) associated with dimensional structure of guanylyl cyclase activating protein-2, a 1132 Sokal et al. JOVS, April 2005, Vol. 46, No. 4

calcium-sensitive modulator of photoreceptor guanylyl cyclases. phan mutants as a function of calcium concentration. J Biol Chem. JBiol Chem. 1999:274:19329-19337. 1999:274:19829-19837. 44. Bowie JU, T.uthy R, Eisenberg D. A method to identify protein 55. Nakazawa M, Naoi N, Wada Y, et al. Autosomal dominant cone-rod sequences that fold into a known three-dimensional structure. dystrophy associated with a Val200Glu mutation of the periph- Science. 1991:253:164-170. erin/RDS gene. Retina. 1996:16:405-410. 45. Kale I., Skeel R, Bhandarkar M, e t al. NAMD2: G reater scalability for 56. Johnson S, Halford S, Morris AG, et al. Genomic organisation and parallel molecular dynamics./Com put Phys. 1993:151:283-312. alternative splicing of human RTM1, a gene Implicated in autoso­ 46. Darden T, York D, Pedersen I.. An N log(N) method for Ewald mal dominant cone-rod dystrophy (C.ORD7). Genomics. 2003:81: sums in large systems./Chem Phys. 1993:98:10089-10092. 304-314. 47. Kawasaki H, Nakayama S, Kretsinger RH. Classification and evolu­ 57. Sohocki MM, Perrault T, Leroy BP, et al. Prevalence of ATPI.1 tion of EF-hand proteins. Biometals. 1998:11:277-295. mutations in inherited retinal degenerative disease. M ol Genet 48. Haeseleer F, Imanishi Y, Sokal T, Filipek S, Palczewski K. Calcium- Metab. 2000:70:142-150. binding proteins: intracellular sensors from the calmodulin super­ 58. T ucker CL, W oodcock SC, Kelsell RE, Ram am urthy V, H unt DM, family. Biochem Biophys Res Commun. 2002:290:615-623. Hurley JB. Biochemical analysis of a dlmerization domain mutation 49. Falke JJ, Drake SK, Hazard Al, Peerson OB. Molecular tuning of ion in RetGC-1 associated with dominant cone-rod dystrophy. Proc binding to calcium signaling proteins. Q Rev Biophys. 1994:27: Natl Acad Sci USA. 1999:96:9039-9044. 219-290. 59. Kobayashi A, Higashide T, Hamasaki D, Kubota S, Sakuma H, An W' 50. Sokal T, I.i N, Surgueheva T, et al. GCAPl (Y99Q mutant is consti- et al. HRG4 (UNC119) mutation found in cone-rod dystrophy tutivciy active in autosomal dominant cone dystrophy. M ol Cell. causes retinal degeneration in a transgenic model. Invest Ophthal­ 1998:2:129-133. m ol Vis Sci. 2000:41:3268-3277. 51. Dizhoor AM, Boikov SG, Olshevskaya E. Constitutive activation of 60. Sohocki MM, Sullivan I.S, Mlntz-Hittner HA, et al. A range of clinical photoreceptor guanylate cyclase by Y99C mutant of GCAP-1. phenotypes associated with mutations in CRX, a photoreceptor / Biol Chem. 1998:273:17311 -17314. transcription-factor gene. Am J Hum Genet. 1998:63:1307-1315. 52. I.i N, Sokal T, Bronson JD, Palczewski K, Baehr W. Identification 61. Sokal T, Li N, Verllnde CL, Haeseleer F, Baehr W', Palczewski K. and functional regions of guanylate cyclase-activating protein 1 Calcium-binding proteins in the retina: from discovery to etiology (GCAPl) using GCAP1/GCTP chimeras. Biol Chem. 2001:382: of human disease. Biochim Biophys Acta. 2000:1498:233-251. 1179-1188. 62. Coleman JE, Zhang Y, Brown GA, Semple-Rowland SL. Cone cell 53. Sokal T, I.i N, Klug CS, et al. Calcium-sensitive regions of GCAPl as survival and downregulation of GCAPl protein in the retinas of observed by chemical modifications, fluorescence and EPR spec­ GCl knockout mice. Invest O phthalmol Vis Sci. 2004:45:3397­ troscopies. J Biol Chem. 2001:276:43361-43373. 3403. 54. Sokal T, Otto-Brue AE, Surgueheva T, et al. Conformational changes 63. Jiang L, Katz BJ, Yang Z, et al. Autosomal dominant cone dystrophy in guanylyl cyclase-activating protein 1 (GCAPl) and its trypto­ caused by a novel mutation in GUCAIA. M ol Vis. In press.