retinitis pigmentosa associated with rhodopsin …rader, dhiman, klein-seetharaman, & bahar,...

Vision Research 46 (2006) 4556–4567www.elsevier.com/locate/visres

Retinitis pigmentosa associated with rhodopsin mutations: Correlation between phenotypic variability and molecular eVects

Alessandro Iannaccone a,¤, David Man b, Naushin Waseem c, Barbara J. Jennings a, Madhavi Ganapathiraju d, Kevin Gallaher a, Elisheva Reese a,

Shomi S. Bhattacharya c, Judith Klein-Seetharaman b,d

a Hamilton Eye Institute, Department of Ophthalmology, Retinal Degeneration and Ophthalmic Genetics Service, University of Tennessee Health Science Center, 930 Madison Avenue, Suit 731, Memphis, TN 38163, USA

b Department of Structural Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USAc Institute of Ophthalmology, Department of Molecular Genetics, University College London, London, UK

d School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA

Received 19 June 2006; received in revised form 1 August 2006

Abstract

Similar retinitis pigmentosa (RP) phenotypes can result from mutations aVecting diVerent rhodopsin regions, and distinct amino acid substi-tutions can cause diVerent RP severity and progression rates. SpeciWcally, both the R135L and R135W mutations (cytoplasmic end of H3)result in diVuse, severe disease (class A), but R135W causes more severe and more rapidly progressive RP than R135L. The P180A and G188Rmutations (second intradiscal loop) exhibit a mild phenotype with regional variability (class B1) and diVuse disease of moderate severity (classB2), respectively. Computational and in vitro studies of these mutants provide molecular insights into this phenotypic variability.© 2006 Elsevier Ltd. All rights reserved.

Keywords: Retinitis pigmentosa; Rhodopsin; Visual function; Phenotype; Protein stability prediction; Membrane protein misfolding

1. Introduction

RHO mutations aVecting the amino acidic sequence ofthe rod-speciWc protein rhodopsin are responsible primarilyfor autosomal dominant retinitis pigmentosa (ADRP) andaccount for 30–40% of this form of RP (source: RetNet:http://www.sph.uth.tmc.edu/Retnet/ disease.htm#03.202d)(RetNet—Retinal Information Network). Uncertaintyremains regarding the disease severity, rate of progression,and clinical–functional phenotypes associated with speciWcRHO mutations. Accordingly, the determinants of diseaseexpression and severity remain incompletely characterized.

Many RHO mutations have been studied in vitro, andvaluable classiWcation schemes of the in vitro behavior of

* Corresponding author. Fax: +1 901 448 5028.E-mail address: [email protected] (A. Iannaccone).

0042-6989/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.visres.2006.08.018

the mutants studied have been proposed (Kaushal & Khor-ana, 1994; Sung, Davenport, & Nathans, 1993; Sung,Schneider, Agarwal, Papermaster, & Nathans, 1991). How-ever, not all clinically relevant mutations have yet beenstudied, and these classiWcation schemes do not alwayscorrelate well with clinical disease expression or severity.SigniWcant progress in understanding the disease sequenceand phenotypic characteristics were made when a classiWca-tion of rhodopsin phenotypes based on the combination ofvarious clinical and functional criteria was developed(Cideciyan et al., 1998). However, to date, neither has thisproposed classiWcation system been corroborated by inde-pendent investigations, nor has it yet been routinely appliedto families harboring other mutations.

The Wrst objective of this report is to illustrate the clini-cal and functional manifestations and the variability inseverity and disease progression rates of RP associated withdiVerent rhodopsin changes aVecting distinct portions of

mailto: [email protected]

mailto: [email protected]

http://www.sph.uth.tmc.edu/Retnet/


A. Iannaccone et al. / Vision Research 46 (2006) 4556–4567 4557

the molecule. We compared the following four mutations(Fig. 1): two distinct mutations aVecting codon 135, thearginine to tryptophan (Arg-135-Trp, or R135W) and theargine to lysine (Arg-135-Lys, or R135L) changes, aVectingan amino acid at the cytoplasmic edge of the third rhodop-sin transmembrane helix (H3); and two changes in the sec-ond intradiscal loop (E2), the proline to alanine change atcodon 180 (Pro-180-Ala, or P180A) and the glycine toarginine change at codon 188 (Gly-188-Arg, or G188R).

Fig. 1. Secondary structure model of rhodopsin illustrating the location ofthe four mutations investigated in this manuscript: R135L and R135W,aVecting the cytoplasmic edge of H3; and P180A and G188R, involvingthe second intradiscal loop (E2). In red and blue are highlighted the twolargest clusters of mutually rigid residues belonging to the core of stabilityidentiWed by FIRST analysis.

All sequence changes have been previously associated withADRP, but P180A has been reported only in a seeminglyisolated case of RP (Sohocki et al., 2001). Our Wndings dem-onstrate that similar clinical–functional phenotypes canresult from mutations aVecting diVerent regions of rhodop-sin, and that distinct amino acid changes at the same codoncan result in identical overall phenotypic classes but diVer-ent RP severity and progression rates.

Second, to test the hypothesis that these clinical Wndingscould be explained at least in part by structural and func-tional diVerences between these mutants, and to conWrmthe pathogenicity of the P180A change, computationalanalyses and in vitro experiments have been performed. Weshow that the P180A change has characteristics consistentwith pathogenicity, and that misfolding and stability pre-dictions and the corresponding characteristics experimen-tally observed in puriWed mutant proteins as well as theirglycosylation and aggregation patterns correlate well withthe gradient of severity observed among these fourmutants, providing insight into the observed phenotypicvariability.

2. Materials and methods

Data from four families (Fig. 2) are included in this report. Onset andtype of symptoms, best corrected visual acuity to either Snellen or ETDRScharts, clinical presentation, Goldmann visual Welds, and Xash ERGs wereanalyzed in 16 subjects (age: 3–64). To characterize disease expression inour families, we utilized the criteria set forth by Cideciyan et al. (1998). Allprocedures were conducted in accordance with the declaration of Helsinki.Blood samples were collected after obtaining signed informed consentfrom all subjects participating in the molecular genetic research studies,which were approved by the Institutional Review Boards of all participat-ing institutions.

Fig. 2. Pedigrees of the four families investigated in this study. Arrows identify the probands from each family. The horizontal bar above the symbols cor-respond to examined subjects. Filled symbols identify aVected individuals. In the P180A pedigree, the half-Wlled symbol (III:1) identiWes a subject whoreportedly suVered from glaucoma, and the “+” symbols indicate obesity (BMI>30).

4558 A. Iannaccone et al. / Vision Research 46 (2006) 4556–4567

2.1. Visual Weld and ERG recording methodology

Goldmann visual Welds to I4e and V4e targets were performed andtheir areas were measured in all patients using previously reported tech-niques (Iannaccone et al., 1995, 2004; Iannaccone, 2003). ERGs wererecorded at diVerent laboratories with monopolar electrodes according tostandard procedures in compliance with the ISCEV standards (Marmor,Holder, Seeliger, & Yamamoto, 2004). The ERGs performed on the twoItalian families were recorded according to methods also reported previ-ously in detail (Del Porto et al., 1993; Iannaccone et al., 1995; Koenekoopet al., 2002; Pannarale et al., 1996; Rispoli, Iannaccone, & Vingolo, 1994).To allow for comparison of data collected with diVerent instruments and/or electrodes at distinct laboratories, amplitudes have been normalized aspercent of the lowest limit of normal for each setting. When available, alsolight-(LA) and dark-adapted (DA) monochromatic automated perimetry(MAP) data were included in the analyses. The latter data were gatheredaccording to the technique developed by Jacobson co-workers (Apáthy,Jacobson, Nghiem-Phu, Knighton, & Parel, 1987; Jacobson & Apáthy,1988; Jacobson et al., 1986). Further methodological details are providedin Appendix A. Normative ranges for the ERG responses recorded atUTHSC are also provided in Supplementary Table 1.

2.2. Patient population

In the Caucasian American family in which a R135L RHO mutationwas identiWed (Fig. 2), we examined a 36-year-old Caucasian male (V:3)and his symptomatic daughter (VI:2, the proband), who could be evalu-ated at 3, 6, and 10 years of age. In this family, there was an overt historyof RP, although with no known male-to-male transmission.

The Italian family in which the R135W RHO mutation co-segregatedwith RP demonstrated 24 aVected individuals (Fig. 2). Of them, 7 subjectswere available for investigation (subjects IV:5, IV:9, V:7, V:8, V:10, the pro-band, V:13, and V:14), ranging in age from 6 to 41 years. Information of thisfamily has been presented in part previously (Pannarale et al., 1996).

In the African–American family harboring the P180A RHO change,two individuals were investigated (Fig. 2), a 22-year-old male (IV:4, theproband) and his 47-year-old mother (III:2). Neither subject was aware ofbeing aVected at the time of examination, the state of aVection having beendiscovered in subject IV:4 in the course of a routine eye exam following atrauma to the head region. There was, however, history of RP in thisfamily in subjects I:2 and II:2.

Lastly, 14 aVected individuals were identiWed from the Italian family inwhich the G188R RHO mutation was found (Fig. 2), 13 of whom by his-tory. Of these, Wve were available for evaluation (III:11, IV:6, IV:10, IV:16,and IV:19, the proband). A sixth examined subject, an 8-year-old female(V:2), was asymptomatic but found to be aVected upon ERG testing.Investigated subjects from this pedigree ranged in age from 8 to 64 years.A report on the fundus features of this family has been previouslypublished (Del Porto et al., 1993).

2.3. Molecular genetic methods

PCR ampliWcation and sequencing of the RHO gene was performed ongenomic DNA extracted from whole blood samples according to manu-facturer’s speciWcations with DNA QiAmp Blood Maxi kits (Qiagen,Valencia, CA) or, for the R135W and G188R families, as previouslyreported (Del Porto et al., 1993; Pannarale et al., 1996). The primers usedfor the ampliWcation of all the Wve exons of rhodopsin have been reportedbefore (Inglehearn et al., 1991). The PCR ampliWcation was carried in25 ml reaction volume using Extensor Hi-Fidelity PCR Master Mix(ABgene, Surrey, UK) according to manufacturer’s instructions. The cycleconditions were 94 °C for 5 min, followed by 30 cycles of 94 °C for 30 s,60C° for 30 s and 72 °C for 30 s. The PCR products were treated with Exo-SAP-IT (Usb Corporation) and sequenced with Big Dye v1.1 according tomanufacturer’s instructions. The sequencing reaction was run on ABIPrism 3100 DNA Sequencer and the sequences were analysed withSequencher v 4.1.4 (Gene Codes).

2.4. Computational modeling

2.4.1. Mutational dataWe have recently compiled two types of experimental results for stabil-

ity changes upon single-point mutations for mammalian rhodopsin (Isin,Rader, Dhiman, Klein-Seetharaman, & Bahar, 2006; Rader et al., 2004;Tastan et al., submitted). First, the percentages of folding as deWned by theability to bind 11-cis retinal by the mutant opsins were collected. Retinalbinding was estimated from the ratio between absorbances at 280 and500 nm. If the mutant showed at least 70% or more of the wild type (WT)reference, it was considered correctly folded. Second, the eVects on Meta-rhodopsin II stability as assayed by the half-life of the Xuorescence decayof light-activated rhodopsin of protein were compiled. This was donebased on the assumption that factors contributing to stability of theground state should also do so for the light-activated state. If a decrease inrate by 25% or more was observed, the mutant was considered to be morestable. If an increase in rate by 25% or more was observed, the mutant wasconsidered to be less stable. In the case of mammalian rhodopsin, thisresulted in information on 147 single-point mutations for percent foldingand 159 mutations for Metarhodopsin II stability. Thus, for a total of 279there is information related to folding and stability jointly.

2.4.2. Stability predictions and simulated thermal unfoldingMammalian rhodopsin (PDB ID code: 1u19) (Okada et al., 2004)

was used as input for stability predictions by three publicly availablealgorithms: I-Mutant2.0 (Capriotti, Fariselli, & Casadio, 2005), DMU-TANT (Zhou & Zhou, 2002), and FOLDX (Guerois, Nielsen, & Ser-rano, 2002). In predictions, no distinction was made between increaseand no change in stability as compared to WT. In addition, we used theFIRST software (Jacobs, Rader, Kuhn, & Thorpe, 2001) to make pre-dictions of the rhodopsin folding core as previously described (Raderet al., 2004).

2.5. In vitro experimental methodology

2.5.1. Construction of rhodopsin mutant genesSingle amino acid replacement mutants R135L, R135W, P180A, and

G188R were prepared by a two-step PCR mutagenesis technique asdescribed in the Supplementary Methods section. The primers used arealso shown in Supplementary Table 2.

2.5.2. Cell culture and transient transfectionTransient transfection of COS-1 cells was as described (Oprian, Mol-

day, Kaufman, & Khorana, 1987) with the exception that cells were har-vested after 72 h. Proteins were puriWed by 1D4 immunoaYnitychromatography in 0.05% DM as described (Hwa, Reeves, Klein-Seethar-aman, Davidson, & Khorana, 1999). The proteins were eluted in PBScontaining 70 �M C-terminal nonapeptide and 0.05% (w/v) DM.

2.5.3. UV/Vis absorption spectroscopyUV/Vis absorption spectra were recorded using a Perkin–Elmer

Lambda 25 UV/Vis spectrophotometer with a bandwidth of 2 nm, aresponse time of 1 s, and a scan speed of 240 nm/min at 20 °C. The molarextinction value (�500) used was 40,600 M¡1 cm¡1.

2.5.4. Deglycosylation with PNGase FFor cleavage of oligosaccharide chains from rhodopsin glycosyla-

tion sites, proteins were deglycosylated with 0.5 U/�g of rhodopsin ofPNGase F (New England Biolabs). The puriWed proteins were Wrstincubated for 1 h at room temperature in denaturing buVer (0.5 Msodium phosphate, pH 7.5, 0.5% SDS, 40 mM DTT). PNGase F and 1%NP-40 were then added and the reaction mixture was further incubatedfor 6 h.

2.5.5. Polyacrylamide gel electrophoresis and immunoblottingProtein samples were resolved by SDS–PAGE on 15% polyacryl-

amide–Tris vertical slab gels and transferred onto nitrocellulosemembrane (Bio-Rad) according to standard protocols.


3. Results

The clinical and functional features of the 16 subjectsincluded in this investigation are illustrated in detail in Sup-plementary Table 3, and the overall phenotypic featuresassociated with the four RHO mutations herein reportedare summarized in Supplementary Table 4. The followingsections summarize the most salient Wndings for eachmutation.

3.1. R135W and R135L: Two class A mutants resulting in diVerent disease severity and progression rates

The R135L and R135W mutations resulted in early-onset (symptomatic and clinically visible within the Wrstdecade of life, Fig. 3A and B), diVuse, and severe disease,and rod function was undetectable by ERG criteria sinceearly childhood (class A) (Cideciyan et al., 1998). VisualWeld loss of peripheral sensitivity and absolute constrictionwere already advanced by the teenage years (Fig. 4A andB), and ERGs were severely attenuated (always <10%) at

all ages (Fig. 4C and D). These Wndings demonstrate thatthe retinas of these patients do not ever develop a normalcomplement of rods and rod outer segments, and suggestthat cone degeneration is already ongoing at birth. OurR135W patients also experienced a more severe reductionin visual acuity then those with the R135L mutation.Fig. 4E shows that the overall kinetics of ERG b-waveamplitude loss deduced from cross-sectional data appearsimilar between our two families, being both Wt well by log-arithmic functions. However, unlike R135W patients,recordable ERGs were retained by our R135L patientthrough the late 30s, and their ERGs were signiWcantlylarger at comparable ages. A substantially slower rate ofERG deterioration from baseline was seen in the R135Lchild tested serially (a progressive decline to 75% of base-line over a 7-year period, equivalent to, on average, a 3.5%yearly rate) compared to older children with the R135Wmutation, in whom at a 50% year-to-year deterioration ratewas documented. These longitudinal Wndings suggest alsopotential for faster disease progression for the R135W-associated phenotype, at least in teenage years, and for an

Fig. 3. The composites in (A and B) The ophthalmoscopic Wndings for the right and left eye, respectively, of subject VI:2 from the R135L pedigree at age 6,an example of a class A phenotype. Although bone spicule-like deposits at this young age are very sparse, they are already visible in both the nasal andtemporal periphery. A moth-eaten appearance of the retinal tissue and dropout of the retinal pigment epithelium are the most prominent features andaVect both the mid-periphery in all four quadrants and the arcades. The retinal vasculature is already clearly attenuated, and waxy disc pallor is alreadyvisible, despite the high hyperopic refractive defect of this child (+8.00 sph in both eyes). At later age (not shown) prominent bone spicule-like depositsbecome apparent and abundant in all four quadrants, and atrophic macular changes appear in these mutants, more notably in association to the R135Wmutation. This presentation contrasts strikingly with that of class B1 phenotypes, exempliWed by the proband harboring the P180A mutation (right andleft eye in (C and D), respectively). The superior retina is essentially normal in appearance and so is the retinal vasculature. A sharply demarcated area ofdegeneration is apparent in the inferior retina, characterized by bone spicule-like deposits and whitish Xecks. In this region, mild vascular attenuation canbe noted. The macular area appears free of degenerative changes.


overall resulting more severe phenotype, although it is pos-sible that the non-linearity of the disease progression kinet-ics accounted for unequal rates in ERG decline at diVerentages. Further phenotypic details are provided in Supple-mentary Table 3.

3.2. P180A and G188R: Similarities and heterogeneity within class B mutants aVecting the second intradiscal loop

3.2.1. P180AUnlike the Class A phenotypes, the asymptomatic pro-

band (subject IV:4) exhibiting a C3820G RHO change,predicting the P180A amino acid substitution, had clear-cut altitudinal disease at age 21 (Fig. 3C and D). This wasaccompanied by minimal superior Goldmann visual Weldloss (Fig. 5A), essentially normal cone-mediated and onlymildly abnormal rod-mediated function in the inferiorhemiWeld (Fig. 5B), and fair preservation of rod-driven,mixed, and cone-driven ERGs (black traces in Fig. 5C).Sixteen months later, some deterioration of the ERGresponses (blue traces) was observed. SpeciWcally, themixed b-wave was 8% smaller from baseline (Fig. 5C andD). In addition, progression of the disease could beinferred by the simultaneous reduction in amplitude andincreased delay of the 30-Hz Xicker ERG, an increaseddelay in both rod- and cone-driven ERG, as well as otherdistinctive changes in DA responses to ¡0.31 cd s/m2

stimuli discussed below.

The mother of the proband was also unaware of beingaVected. Although she was not available for fundus photog-raphy or functional testing, she had clinical features of mildyet diVuse disease without regional predilection, which wasmost compatible with a class B2 phenotype (Cideciyanet al., 1998). Of likely relevance to this intrafamilial pheno-typic heterogeneity, and possibly to the sharply altitudinaldisease observed in the proband, the latter had a history ofunusually high exposure to sunlight for occupationalreasons (farm worker). Additional phenotypic details arepresented in Supplementary Table 3.

3.2.2. G188RSimilar to patients with mutations aVecting codon 135,

also the G188R mutation was associated with a diVuseretinopathy (Del Porto et al., 1993) (see also Supplemen-tary Table 3). However, the G188R mutation yielded amuch milder phenotype, with symptoms of night blind-ness as late as age 30. Accordingly, visual Weld constric-tion associated with this genotype was variable.Compared to the R135W mutation, twice as large visualWelds could still be measured in patients at the age of 42(Fig. 6A). In the Wrst decade of life, no fundus changeswere seen (Del Porto et al., 1993), patients were asymtom-atic, and mixed ERG amplitude was still 50–60% of nor-mal (Figs. 5D and 6B). Although the waveform andtiming of the response of subject V:2 suggest that thisb-wave was mainly driven by cone bipolars, and that

Fig. 4. Functional Wndings associated with mutations at codon 135 (class A phenotype). (A and B) Examples of visual Welds of children harboring theR135L and R135W mutations, respectively. By late teenage years, severe concentric visual Weld constriction is already present. (C and D) Mixed dark-adapted ERG responses recorded in subjects with these mutations. All responses are severely diminished (normal amplitude scale for these responses: 80�V/division). (E) The plot of normalized amplitude of mixed DA ERG b-wave (expressed as percent of the lowest limit of normal) vs. age for patients withthese mutations. The mixed ERGs of patients harboring the R135L mutation remain signiWcantly better preserved than those with the R135W mutation.The overall kinetics of ERG decay deduced from this cross-sectional data appear comparable for the two mutations. However, longitudinal measurementsfrom the R135L proband and two children harboring the R135W mutations (shown encircled) suggest diVerent progression rates vs. baseline between thetwo mutations: a 25% loss vs. baseline (age 3) was seen over a 7-year period for the R135L proband (on average, 3.5% per year), whereas a 50% loss 1 yearfrom baseline was observed for the two R135W children.


perhaps only the second, abnormally delayed peakreXected rod bipolar activation (see below), this Wndingprovides evidence for abnormal and reduced yet persis-tent and measurable rod-driven activity in childrenaVected by the G188R mutation.

In subsequent decades, a progressive age-dependentdecline of mixed ERG b-wave amplitudes was seen(Fig. 6B). By age 60, the visual Weld was severely constricted(Fig. 6A), the mixed ERG of G188R patients was stillrecordable although reduced to 2% of normal (Fig. 5D),and good visual acuity was still preserved, in line with theobserved preservation of macular integrity documentedophthalmoscopically (see Supplementary Table 3 for fur-ther details). The integrity of foveal cones is therefore main-tained until late in the course of this form of ADRP. Insummary, these features characterize the phenotype associated

with the G188R mutation as a class B2 mutant (Cideciyanet al., 1998).

Fig. 5D also shows that subjects harboring the G188Rmutation have signiWcantly smaller responses than the sub-ject with the P180A mutation. In addition, the estimatedyearly ERG loss progression rate seen in two patients withthe G188R mutation was signiWcantly higher (on average,21.5%) than the 8% observed with the P180A mutation.This correlates well with the regional nature of disease inthe latter case. Of relevance to the in vitro and computa-tional data presented later, the observed G188R yearlyaverage rate of disease progression was less than half thatof R135W younger patients, but signiWcantly greater thanthe R135L child followed serially. Hence, in addition to thegreater severity from onset of disease associated to codon135 mutations, this observation also suggests potential for

Fig. 5. Functional Wndings associated with the P180A RHO mutation. (A) Goldmann visual Welds demonstrating modest loss of the superior Weld and rel-ative scotomas in the superior hemiWeld, consistent with the altitudinal disease observed ophthalmoscopically. The regional predilection of the disease forthe inferior retina is conWrmed by the monochromatic automated perimetry plots shown in (B). These plots also show some loci exhibiting rod sensitivitylosses (RSL) also in the inferior hemiWeld at loci where no cone sensitivity loss (CSL) is seen, a phenomenon that is typical of class B1 phenotypes. (C)Representative pairs of ISCEV-standard ERG responses (black traces: baseline; blue traces: 16 months later). The Xash intensities in cd s/m2 are given tothe left of each response. DA D dark-adapted; LA D light-adapted. A response at ¡0.31 cd s/m2 at which the b-wave splitting (arrow) discussed further inthe text and in Appendix A is also shown. (D) The plot of normalized amplitude of mixed DA ERG b-wave (expressed as percent of the lowest limit ofnormal) vs. age for patients with RHO mutations P180A and G188R. The mixed ERG of the proband with the P180A mutation is signiWcantly better pre-served than that of patients with the G188R mutation, whose response decline is Wt well by an exponential function. The observed average yearly rate ofERG decline deduced from longitudinal observations (encircled data points) was 21.5% for the G188R mutation and 8% for the P180A.


an overall slower progression rate of the disease forpatients harboring the G188R mutation compared to thosewith R135W.

3.2.3. Electroretinographic evidence for anomalous rod-driven responses in patients with class B phenotypes

Subject IV:4 from the P180A pedigree also showed a dis-tinctive splitting of the DA b-wave at ¡0.31 cd s/m2 stimuli,which is approximately 0.5 log units below the ISCEV stan-dard Xash range (Fig. 5C, arrow). The second, delayed b-wave peak (75 ms) was shown to be genuinely rod-drivenby its disappearance when these responses were re-recordedimmediately following the LA series and by its progressivere-appearance as DA responses were allowed to recover(illustrated in the Supplementary Wgure). A similar phe-nomenon was documented also for the asymptomatic child(subject V:2) harboring the G188R mutation in response toa standard mixed Xash of 0.18 cd s/m2 (also illustrated in theSupplementary Wgure).

It appears reasonable to conclude that this distinctiveDA ERG b-wave splitting was consistent with the activityof rod-driven bipolar cells from areas of greater rod diseaseseverity, responding more slowly than the bipolar cells fromareas of well preserved rod- (and cone-) mediated activity.Furthermore, not only was the Wrst—and normal in tim-ing—peak diminished at follow up in the P180A proband,but the second delayed b-wave peak was also markedlyattenuated at follow-up testing (not shown). This Wnding issuggestive of greater disease progression for rods withinitially more severe disease.

3.3. Computational studies

The crystal structure of rhodopsin is known (Palczewskiet al., 2000) and can thus be used to infer likely eVects ofamino acid substitutions on stability based on the localenvironment of each residue provided by the structure. Twomeasures of stability were investigated, misfolding and thedecay rate of the light-activated, Metarhodopsin II state.Three diVerent algorithms were compared, I-Mutant2.0(Capriotti et al., 2005), DFIRE (Zhou & Zhou, 2002), andFOLDX (Guerois et al., 2002).

The results of these computations are shown in detail inSupplementary Table 5. The overall accuracies of all threemethods were relatively low. This indicates that the localstructure surrounding the mutation site alone is not suYcientto account for the experimentally observed eVects on rho-dopsin stability and folding. Usually, this is interpreted aslong-range eVects by the mutations, altering the folding orconformation of the protein. To investigate such long-rangeeVects, we utilized the previously applied FIRST analysismethod (Rader et al., 2004). In this analysis, we found thatthe core of stability includes several clusters, the largest ofwhich is located surrounding the conserved disulWde bondbetween residues Cys110 and Cys187 lining the retinal bindingpocket. The residues of the largest cluster are circled in red inFig. 1. The two mutations in the second intradiscal loop,Gly188 and Pro180, are both part of this cluster. It was previ-ously shown that >90% of the residues in this cluster showmisfolding upon mutation, indicating a role of this cluster inthe folding of rhodopsin (Rader et al., 2004). Thus, it is

Fig. 6. Functional Wndings associated with the G188R RHO mutation. (A) Examples of visual Welds in a 42- and a 64-year-old subject. Despite the markedconstriction eventually ensuing in these patients, subject III:7 retained 20/20 vision in the left eye (shown here). (B) Examples of mixed ERG responsesfrom this family, showing the far better preservation of the responses compared to codon 135 mutations, but the greater disease severity compared to theP180A mutation. The response of case V:2 also displays a b-wave splitting reminiscent of that seen in the P180A proband, and shown to have the same ori-gin (i.e., rod-driven) in the Supplementary Wgure.


highly likely that Gly188 and Pro180 mutations will also resultin misfolding of rhodopsin. The FIRST simulated unfoldingalso predicted several other, smaller regions of structural sta-bility. The largest of these is a cluster of residues at the cyto-plasmic end of helix 3 and the second cytoplasmic loop,highlighted by blue circles in Fig. 1. Arg135 is part of thiscluster, strongly suggesting a role of this residue in structuralstability of rhodopsin. Mutation at this site would thereforebe expected to cause misfolding.

3.4. Stability, misfolding, aggregation, and glycosylation of rhodopsin mutants

Expression of the four RHO mutations presented hereinby transient transfection in COS-1 cells, reconstitution with11-cis retinal and puriWcation on 1D4-antibody aYnitychromatography showed that the yield of fully folded,chromophore-containing rhodopsin was low in all fourmutants indicating a tendency to misfold in all four cases.Representative absorption spectra of the four mutants andthe WT are shown in Fig. 7A. Spectra are normalized to theabsorbance at 280 nm (opsin). The ratios between theabsorbance at 500 nm as compared to that at 280 nm, adirect measure of the ratio between folded, retinal-bound

protein and misfolded protein are summarized in Table 1.In relative terms to one another, the degree of misfoldingvaried considerably. The P180A and the R135L mutationscaused the least misfolding, whereas G188R and R135Wwere both severely misfolded. The Wnding that all fourmutants cause misfolding is in very good agreement withthe predictions by the FIRST method.

Next, we investigated if the folded portions of themutants diVered in stability. Fig. 7B shows the decrease in500 nm chromophore at 55 °C, an indicator of thermal sta-bility of rhodopsin. WT rhodopsin shows a 10% decrease

Table 1Degrees of misfolding of rhodopsin mutants as compared to wild type

a A280, opsin absorbance; A500, absorbance after retinal binding. Sincemisfolded rhodopsin does not give rise to 500 nm absorbance peaks,higher ratios are indicative of greater misfolding. Values are the averagesobtained for six elution fractions.

A280:A500 absorbance ratioa (means § SD)

Wild type 2.0 § 0.2R135L 5.6 § 0.3R135W 10.2 § 1.8P180A 5.9 § 0.7G188R 12.7 § 3.0

Fig. 7. In vitro misfolding, stability, glycosylation, and aggregation state of rhodopsin mutants as compared to wild type. (A) Representative absorbancerhodopsin spectra normalized to the 280 nm absorbance: wild type (black), R135L (teal green), R135W (light green), P180A (red), and G188R (blue). (B)Rhodopsin stability measured by loss of 500 nm chromophore over time at 55 °C. Absorbance spectra were measured in time intervals and the decrease inabsorbance at 500 nm was expressed as percent of intact structure. (C) Western-blot of wild type (WT) and rhodopsin mutants (R135L, R135W, P180A,and G188R) expressed transiently in COS-1 cells. Protein concentrations were adjusted according to expression level and approximately 0.075 �g wasloaded. For each mutant or the wild type, equivalent amounts were analyzed before (¡) and after (+) treatment with PNGase F.


over a time-course of 13 h. L135R and L135W showed sim-ilar but lower stability than wild type, while P180A washighly unstable: within 60 min, greater than 95% of theoriginal chromophore was lost. These results are in goodagreement with the stability predictions. We predicted anincrease/no change in stability for both R135 mutants, andexperimentally observed a behavior similar to wild typerhodopsin, whereas we predicted a highly destabilizingeVect of the P180A mutation, which is what we observed inthe experiment. We could not carry out measurements forG188R due to the low expression yields of this mutant.

Finally, we compared the SDS–PAGE mobility of thefour RHO mutants with the wild type. A Western-blotusing the 1D4-antirhodopsin monoclonal antibody isshown in Fig. 7C. First, all four mutants display a markedlydiVerent appearance. Normal appearance is a “smear” of abroad band corresponding to heterogeneous glycosylationand monomeric aggregation state, as shown in Fig. 7C inthe WT control. The R135L and P180A mutants exhibiteda behavior the most similar to wild type, but with increaseddimeric and tetrameric conWgurations. In addition, P180Ashowed degradation products and lesser glycosylation thanthe wild type and the R135L mutant. Unlike these mutants,R135W and G188R were very diVerent from the wild type.R135W and G188R displayed defects in glycosylation andfor R135W much of the protein was aggregated (tetramericband at 100 kDa). G188R displayed mobility consistentwith a strong tendency to dimerize.

4. Discussion

We have reported the manifestations and the variabilityin severity and disease progression rates of ADRP associ-ated with RHO sequence changes involving diVerent por-tions of the rhodopsin molecule (the cytoplasmic edge ofH3: R135W and R135L; and the second intradiscal loop:P180A and G188R). We characterized the phenotype ofseveral new families and appraised the phenotype of thetwo previously reported Italian families within the classiW-cation system of Cideciyan et al. (1998). The G188R muta-tion is thus added to the list of those classiWed according tothis system, supplementing the previously published data(Del Porto et al., 1993) with additional longitudinal obser-vations. Lastly, the availability of data starting from thepediatric age range allowed us to perform a direct compari-son since the early stages of disease expression between theR135W (Pannarale et al., 1996) and the R135L mutations.We directly compared the eVects of the four amino acidsubstitutions on the folding, stability, and glycosylationpatterns of rhodopsin by computational predictions andin vitro studies of the rhodopsin mutants.

All residues involved in these four families (Arg135,Pro180, and Gly188) are highly conserved residues in rhodop-sins across all species. The four mutations reported herevaried with respect to the types of amino acid substitutions.The mutations at codon 135 both lead to a change in sizeand charge for this residue. The large and basic R amino

acid is replaced by a smaller non-polar one in the R135Lmutant, and by a non-polar, large and aromatic one in theR135W mutant. The P180A substitution replaces mediumsize hydrophobic residue (P) with a hydrophobic butsmaller (A). Unlike these mutations, G188R leads to thesubstitution of a small, non-polar amino acid (G) to a largeand basic one (R). These diVerences create the premises fordiVerent structural and, potentially, phenotypic eVects.

4.1. Computational and in vitro analyses correlate with greater disease severity associated with the R135W mutation

The most severe disease (class A) was associated with themutations at codon 135. This Wnding is consistent with pre-vious independent reports of ADRP associated with muta-tions aVecting the Arg135 residue (Jacobson, Kemp, Sung, &Nathans, 1991; Jacobson, Kemp, Cideciyan, & Nathans,1996; Ponjavic, Abrahamson, Andreasson, Ehinger, & Fex,1997). We could successfully document the persistence ofrecordable ERG responses in children from both theR135W and the R135L family. This data show that ERGsare signiWcantly larger and better preserved over time insubjects with the R135L mutation.

The consistently severe disease associated with these twomutations underscores the functional importance of thisresidue. The Glu134 and Arg135 residues are part of thehighly conserved D/E R Y motif at the edge of H3 in therhodopsin 3D structure (Li, Edwards, Burghammer, Villa,& Schertler, 2004; Palczewski et al., 2000) (see Fig. 1), animportant interaction site with transducin. In vitro studiesof bovine rhodopsin mutants have shown that both R135Land R135W bound retinal almost in a wild type-like fash-ion under the conditions used for reconstitution and puriW-cation, but were unable to activate the G protein (Shi et al.,1998). Prior studies of human rhodopsin had led to the clas-siWcation of both R135L and R135W as class IIb mutants(Sung et al., 1993), i.e., defective in their ability to bind11-cis retinal, folding, and stability, and partial transporta-tion to cell surface. By the criteria used in previous studies,these two mutants did not diVer in their properties. Ourcomputational and in vitro studies support the conclusionthat the folding in both of the mutants is severely impairedand show that the predicted misfolding of the R135Wmutant is signiWcantly greater than the R135L and that itsglycosylation state is the most defective of all four mutantsstudied. These Wndings provide, at least in part, a potentialexplanation at the molecular level for the greater diseaseseverity caused by the R135W amino acid substitution.

4.2. P180A: Evidence for pathogenicity of this amino acid substitution and rationale for the observed mild disease phenotype

Unlike the above mutations, the two changes aVectingthe second intradiscal loop were associated with signiW-cantly milder phenotypes. The P180A substitution led tothe least severe of the manifestations, with evidence for


class B1 behavior in the proband who had a occupationalhistory of high environmental light exposure, and a seem-ingly mild class B2 phenotype in his mother who did notshare this environmental exposure. The potential for classB2 manifestations in association with the P180A mutationappears similar to what has been reported for the P23H-associated phenotype (Cideciyan et al., 1998). In consider-ation of the recent evidence that dogs harboring the T4RRHO mutation are exquisitely sensitive to light damage(Cideciyan et al., 2005; Zhu et al., 2004), the predilection forinferior retinal disease in the proband may reXect a similarpredisposition in humans, as already suggested by clinical(Heckenlively, Rodriguez, & Daiger, 1991) and rat (Organ-isciak, Darrow, Barsalou, Kutty, & Wiggert, 2003; Walsh,van Driel, Lee, & Stone, 2004) data for the P23H mutation.

Like in Sohocki et al. (2001), the P180A change observedin our proband could not be conWrmed by segregation stud-ies in other family members. However, the Pro180 residue ishighly conserved, and it is in the proximity of Glu181, a resi-due that has been reported to cause ADRP likely to resultin class A disease (Cideciyan et al., 1998). Also, in vitro rho-dopsin deletion studies have shown that removal of thePro180 residue yields a structurally and functionally defec-tive rhodopsin molecule (Doi, Molday, & Khorana, 1990).Therefore, the introduction of a smaller amino acid,alanine, in place of the wild type proline likely induces adestabilization of the aforementioned network. Thepathogenicity of the P180A mutation was strongly corrob-orated by the computational and in vitro evidence hereinprovided.

4.3. G188R mutation: A rationale for intermediate disease severity

The G188R mutation was associated only with class B2features, without regional predilection. Our data show thatcone function is at least partially maintained for life inpatients with this mutation and that there may be room forearly rod-speciWc interventions. The in vitro behavior of theG188R mutant has been studied before (Liu, Garriga, &Khorana, 1996; Sung et al., 1991), conWrming our predic-tion that this mutant is misfolded (Del Porto et al., 1993).The mutation at codon 188 introduces a large and basicarginine residue that is likely to preclude the correct foldingof the intradiscal portion of the molecule mediated by theformation of a disulWde bond between Cys187 and Cys110

(Davidson, Loewen, & Khorana, 1994; Hwa et al., 1999;Karnik, Sakmar, Chen, & Khorana, 1988). Like Pro180, alsoGly188 is part of the folding core predicted by the FIRSTmethod. The in vitro experimental data presented hereinshow that the G188R mutation causes much more pro-nounced misfolding than that observed for the P180Amutation.

Although other diVerences are likely to exist in thein vitro behavior of these two mutants, the lesser degree ofmisfolding of the P180A mutant compared to the G188Rcorrelates well with the milder phenotype associated with

the former amino acid change. However, the G188Rmutant was as severely misfolded as the R135W, whichcaused the most severe disease of all in our series. Longitu-dinal measurements also suggest that the rate of diseaseprogression is about twice as fast in patients with theR135W mutation compared to patients harboring theG188R one. Likewise, the P180A and the R135L mutationswere comparable in severity of misfolding, but the latterresulted in a much more severe phenotype. Therefore,misfolding alone is an insuYcient criterion to account forphenotypic diVerences among these four mutants.

One plausible explanation is the known lack of trans-ducin activation by the Arg135 mutants (Shi et al., 1998).Thus, in the Arg135 mutations, misfolding and impairmentin signalling function may be acting together to cause astronger phenotype than the G188R and P180A mutations.It will be an important goal of future studies to characterizethe interactions of the G188R and P180A mutants withtransducin, arrestin, and rhodopsin kinase, and contrasttheir properties between themselves and with those of thetwo Arg135 mutants. It would also appear important tostudy these behaviors in the simultaneous presence of WTrhodopsin, since in humans aVected by disease the potentialfor expression of the WT allele remains, and the extent towhich this occurs may be directly correlated to the behaviorof the mutant allele. This is a particularly exciting avenuefor further research because evidence is mounting thatrhodopsin in the ROS is in a dimeric state (reviewed inPalczewski, 2006) and we found preliminary evidence thatG188R may have an increased dimer aYnity, providing apossible additional explanation for the less severe pheno-type associated with this mutant despite misfolding compa-rable to R135W.

5. Conclusions

In summary, we have shown that phenotypic diVerencesexist between mutations aVecting the same portion of themolecule. These diVerences pertain to disease severity andprogression rates, and in part also to phenotypic classes. Acombination of computational folding predictions andin vitro assays allowed us to provide potential explanationsfor the clinical disease gradient observed for these fourmutations (speciWcally, P180A>G188R>R135L>R135W).Caution must always be exerted, however, in drawingin vivo human conclusions from in vitro and animal data.For example, previous in vitro studies could not provideevidence for mislocalization of another mutation present-ing with a mixed class B1 and B2 phenotype, P23H. Rho-dopsin mutated at this codon mislocalizes to the synapticterminal of photoreceptors in mice with a human P23Hrhodopsin transgene (Roof, Adamian, & Hayes, 1994) butnot in mice expressing a murine transgene (Wu et al., 1998)or in Drosophila (Galy, Roux, Sahel, Leveillard, &Giangrande, 2005). The role of this aberrant localization indisease pathogenesis and progression of this and otherforms of RHO-linked ADRP is not well understood.


Recent evidence for a direct and light-independent degener-ative eVect of mislocalized rhodopsin has been provided inXenopus laevis retinas harboring a C-terminus mutantequivalent to the Q344X human counterpart (Tam, Xie,Oprian, & Moritz, 2006). Recent in vitro data also suggestthat the P23H mutant may result in formation of aggre-somes and generalized impairment of the ubiquitin protea-some system (Illing, Rajan, Bence, & Kopito, 2002; Saliba,Munro, Luthert, & Cheetham, 2002). Therefore, the in vivointracellular fate of mutant rhodopsin molecules may alsobe a very important variable to characterize to fully appre-ciate the consequences and mechanism(s) of disease linkedto RHO mutations.

IdentiWcation and characterization of RHO mutationsremains important because it continues to improve ourunderstanding of RP pathophysiology and yields clinicallyimportant prognostic information. The combined use ofclinical and functional data with computational andin vitro expression experiments show promise as anapproach to improve further the understanding of RPcaused by RHO mutations and test hypotheses that eachapproach, when used alone, cannot address eVectively.

Acknowledgments

We acknowledge the generous support of: Research toPrevent Blindness, New York, NY (Career DevelopmentAward to AI and unrestricted Grant to UTHSC Ophthal-mology); Sofya Kovalevskaya Prize of the HumboldtFoundation, Germany/Zukunftsinvestitionsprogramm derBundesregierung Deutschland (J.K.S.), National ScienceFoundation Grants EIA0225636, EIA0225656 andCAREER CC044917 (J.K.S.), and National Institutes ofHealth Grant NLM108730, USA (J.K.S.); the SpecialTrustees of MoorWeld’s Eye Hospital (S.S.B.). Moleculargenetic analyses for the G188R family were originally per-formed at the following labs: Institute of Medical Genetics,University La Sapienza, Rome, Italy; International Insti-tute of Genetics and Biophysics, Naples, Italy; and Institutfür Human Genetik, Universität Lübeck, Germany. TheR135L mutation was identiWed in the laboratory of Dr.Thaddeus P. Dryja, MD, Harvard University, Boston, MA,USA, where also the presence of the P180A change wasindependently veriWed.

Appendix A. Supplementary data

Supplementary data associated with this article canbe found, in the online version, at doi:10.1016/j.visres.2006.08.018.

References

Apáthy, P. P., Jacobson, S. G., Nghiem-Phu, L., Knighton, R. W., & Parel,J.-M. (1987). Computer-aided analysis in automated dark-adaptedstatic perimetry. In Seventh international visual Weld symposium (pp.278–284). Amsterdam, September 1986: Martinus NijhoV/Dr. W. JunkPublishers, Dordrecht.

Capriotti, E., Fariselli, P., & Casadio, R. (2005). I-Mutant2.0: predictingstability changes upon mutation from the protein sequence or structure.Nucleic Acids Research, 33(Web Server issue), W306–W310.

Cideciyan, A. V., Hood, D. C., Huang, Y., Banin, E., Li, Z.-L., Stone, E. M.,Milam, A. H., & Jacobson, S. G. (1998). Disease sequence from mutantrhodopsin allele to rod and cone photoreceptor degeneration in man.Proceedings of the National Academy of Sciences USA, 95, 7103–7108.

Cideciyan, A. V., Jacobson, S. G., Aleman, T. S., Gu, D., Pearce-Kelling, S.E., Sumaroka, A., Acland, G. M., & Aguirre, G. D. (2005). In vivodynamics of retinal injury and repair in the rhodopsin mutant dogmodel of human retinitis pigmentosa. Proceedings of the NationalAcademy of Sciences USA, 102(14), 5233–5238.

Davidson, F. F., Loewen, P. C., & Khorana, H. G. (1994). Structure andfunction in rhodopsin: replacement by alanine of cysteine residues 110and 187, components of a conserved disulWde bond in rhodopsin,aVects the light-activated metarhodopsin II state. Proceedings of theNational Academy of Sciences USA, 91(9), 4029–4033.

Del Porto, G., Vingolo, E. M., David, D., Steindl, K., Wedemann, H.,Forte, R., Iannccone, A., Gal, A., & Pannarale, M. R. (1993). Clinicalfeatures of autosomal dominant retinitis pigmentosa associated withthe Gly-188-Arg mutation of the rhodopsin gene. In J. G. HollyWeld,M. M. La Vail, & R. E. Anderson (Eds.), Retinal degeneration: Clinicaland laboratory applications (pp. 91–101). New York: Plenum Press.

Doi, T., Molday, R. S., & Khorana, H. G. (1990). Role of the intradiscaldomain in rhodopsin assembly and function. Proceedings of theNational Academy of Sciences USA, 87(13), 4991–4995.

Galy, A., Roux, M. J., Sahel, J. A., Leveillard, T., & Giangrande, A. (2005).Rhodopsin maturation defects induce photoreceptor death by apopto-sis: a Xy model for RhodopsinPro23His human retinitis pigmentosa.Human Molecular Genetics, 14(17), 2547–2557.

Guerois, R., Nielsen, J. E., & Serrano, L. (2002). Predicting changes in thestability of proteins and protein complexes: a study of more than 1000mutations. Journal of Molecular Biology, 320(2), 369–387.

Heckenlively, J. R., Rodriguez, J. A., & Daiger, S. P. (1991). Autosomaldominant sectoral retinitis pigmentosa. Two families with transversionmutation in codon 23 of rhodopsin. Archives of Ophthalmology, 109(1),84–91.

Hwa, J., Reeves, P. J., Klein-Seetharaman, J., Davidson, F., & Khorana, H.G. (1999). Structure and function in rhodopsin: further elucidation ofthe role of the intradiscal cysteines, Cys-110, -185, and -187, in rhodop-sin folding and function. Proceedings of the National Academy of Sci-ences USA, 96(5), 1932–1935.

Iannaccone, A. (2003). Usher syndrome: correlation between visual Weldsize and maximal ERG response b-wave amplitude. In M. M. LaVail, J.G. HollyWeld, & R. E. Anderson (Eds.), Retinal degenerations: Mecha-nisms and experimental therapy (vol. 533, pp. 123–131). New York: Ple-num Publishers.

Iannaccone, A., Kritchevsky, S. B., Ciccarelli, M. L., Tedesco, S. A., Maca-luso, C., Kimberling, W. J., & Somes, G. W. (2004). Kinetics of visualWeld loss in Usher syndrome Type II. Investigative Ophthalmology andVisual Science, 45(3), 784–792.

Iannaccone, A., Rispoli, E., Vingolo, E. M., Onori, P., Steindl, K., Rispoli,D., & Pannarale, M. R. (1995). Correlation between Goldmann perime-try and maximal electroretinogram response in retinitis pigmentosa.Documenta Ophthalmologica, 90, 129–142.

Illing, M. E., Rajan, R. S., Bence, N. F., & Kopito, R. R. (2002). A rhodop-sin mutant linked to autosomal dominant retinitis pigmentosa is proneto aggregate and interacts with the ubiquitin proteasome system. TheJournal of Biological Chemistry, 277(37), 34150–34160.

Inglehearn, C. F., Bashir, R., Lester, D. H., Jay, M., Bird, A. C., & Bhat-tacharya, S. S. (1991). A 3-bp deletion in the rhodopsin gene in a familywith autosomal dominant retinitis pigmentosa. American Journal ofHuman Genetics, 48(1), 26–30.

Isin, B., Rader, A. J., Dhiman, H. K., Klein-Seetharaman, J., & Bahar, I.(2006, in press). Predisposition of the dark state of rhodopsin to func-tional changes in structure. Proteins.

Jacobs, D. J., Rader, A. J., Kuhn, L. A., & Thorpe, M. F. (2001). ProteinXexibility predictions using graph theory. Proteins, 44(2), 150–165.

http://dx.doi.org/10.1016/j.visres.2006.08.018




Jacobson, S. G., & Apáthy, P. P. (1988). Automated rod and cone perime-try in retinitis pigmentosa. In J. L. Smith & R. S. Katz (Eds.), Neuro-ophthalmology enters the nineties (pp. 35–47). Hialeah: Dutton Press.

Jacobson, S. G., Kemp, C. M., Cideciyan, A. V., & Nathans, J. (1996). Rho-dopsin gene mutations causing retinitis pigmentosa. Functional pheno-types of codon 23 and codon 135 genotypes. In J. Robbins, A. Taylor,& M. B. A. Djamgoz (Eds.), Basic and clinical perspectives in visionresearch (pp. 53–62). New York: Plenum Press.

Jacobson, S. G., Kemp, C. M., Sung, C.-H., & Nathans, J. (1991). Retinalfunction and rhodopsin levels in autosomal dominant retinitis pigmen-tosa with rhodopsin mutations. American Journal of Ophthalmology,112, 256–271.

Jacobson, S. G., Voigt, W. J., Parel, J.-M., Apáthy, P. P., Nghiem-Phu, L.,Myers, S. W., & Patella, V. M. (1986). Automated light- and dark-adapted perimetry for evaluating retinitis pigmentosa. Ophthalmology,93(12), 1604–1611.

Karnik, S. S., Sakmar, T. P., Chen, H. B., & Khorana, H. G. (1988). Cys-teine residues 110 and 187 are essential for the formation of correctstructure in bovine rhodopsin. Proceedings of the National Academy ofSciences USA, 85(22), 8459–8463.

Kaushal, S., & Khorana, H. G. (1994). Structure and function in rhodop-sin. 7. Point mutations associated with autosomal dominant retinitispigmentosa. Biochemistry, 33, 6121–6128.

Koenekoop, R. K., Fishman, G. A., Iannaccone, A., Ciccarelli, M. L.,Ezzeldin, H., Baldi, A., Sunness, J. S., Lotery, A., Jablonski, M. M., Pit-tler, S. J., & Maumenee, I. (2002). Electroretinographic and psycho-physical abnormalities in parents of Leber Congenital Amaurosispatients. Archives of Ophthalmology, 120, 1325–1330.

Li, J., Edwards, P. C., Burghammer, M., Villa, C., & Schertler, G. F. (2004).Structure of bovine rhodopsin in a trigonal crystal form. Journal ofMolecular Biology, 343(5), 1409–1438.

Liu, X., Garriga, P., & Khorana, H. G. (1996). Structure and function inrhodopsin: correct folding and misfolding in two point mutants in theintradiscal domain of rhodopsin identiWed in retinitis pigmentosa. Pro-ceedings of the National Academy of Sciences USA, 93(10), 4554–4559.

Marmor, M. F., Holder, G. E., Seeliger, M. W., & Yamamoto, S. (2004, inpress). Standard for clinical electroretinography (2003 update). Docu-menta Ophthalmologica.

Okada, T., Sugihara, M., Bondar, A. N., Elstner, M., Entel, P., & Buss, V.(2004). The retinal conformation and its environment in rhodopsin inlight of a new 2.2 A crystal structure. Journal of Molecular Biology,342(2), 571–583.

Oprian, D. D., Molday, R. S., Kaufman, R. J., & Khorana, H. G. (1987).Expression of a synthetic bovine rhodopsin gene in monkey kidneycells. Proceedings of the National Academy of Sciences USA, 84(24),8874–8878.

Organisciak, D. T., Darrow, R. M., Barsalou, L., Kutty, R. K., & Wiggert,B. (2003). Susceptibility to retinal light damage in transgenic rats withrhodopsin mutations. Investigative Ophthalmology and Visual Science,44(2), 486–492.

Palczewski, K. (2006). G protein-coupled receptor rhodopsin. AnnualReview of Biochemistry, 75, 743–767.

Palczewski, K., Kumasaka, T., Hori, T., Behnke, C. A., Motoshima, H.,Fox, B. A., Le Trong, I., Teller, D. C., Okado, T., Stenkamp, R. E.,Yamomato, M., & Miyano, M. (2000). Crystal structure of rhodopsin:A G protein-coupled receptor. Science, 289(5480), 739–745.

Pannarale, M. R., Grammatico, B., Iannaccone, A., Forte, R., De Ber-nardo, C., Flagiello, L., Vingolo, E. M., & Del Porto, G. (1996). Autoso-mal dominant retinitis pigmentosa associated with an Arg-135-Trppoint mutation of the rhodopsin gene: clinical features and longitudi-nal observations. Ophthalmology, 103, 1443–1452.

Ponjavic, V., Abrahamson, M., Andreasson, S., Ehinger, B., & Fex, G.(1997). Autosomal dominant retinitis pigmentosa with a rhodopsinmutation (Arg-135-Trp). Disease phenotype in a Swedish family. ActaOphthalmology Scandinavica, 75(2), 218–223.

Rader, A. J., Anderson, G., Isin, B., Khorana, H. G., Bahar, I., & Klein-See-tharaman, J. (2004). IdentiWcation of core amino acids stabilizing rho-dopsin. Proceedings of the National Academy of Sciences USA, 101(19),7246–7251.

RetNet—Retinal Information Network http://www.sph.uth.tmc.edu/Ret-net/.

Rispoli, E., Iannaccone, A., & Vingolo, E. M. (1994). Low-noise electroret-inogram recording techniques in retinitis pigmentosa. Documenta Oph-thalmologica, 88, 27–37.

Roof, D. J., Adamian, M., & Hayes, A. (1994). Rhodopsin accumulation atabnormal sites in retinas of mice with a human P23H rhodopsin trans-gene. Investigative Ophthalmology and Visual Science, 35(12), 4049–4062.

Saliba, R. S., Munro, P. M., Luthert, P. J., & Cheetham, M. E. (2002). Thecellular fate of mutant rhodopsin: quality control, degradation andaggresome formation. Journal of Cell Science, 115(Pt 14), 2907–2918.

Shi, W., Sports, C. D., Raman, D., Shirakawa, S., Osawa, S., & Weiss, E. R.(1998). Rhodopsin arginine-135 mutants are phosphorylated by rho-dopsin kinase and bind arrestin in the absence of 11-cis retinal. Bio-chemistry, 37(14), 4869–4874.

Sohocki, M. M., Daiger, S. P., Bowne, S. J., Rodriquez, J. A., Northrup, H.,Heckenlively, J. R., Birch, D. G., Mintz-Hittner, H., Ruiz, R. S., Lewis,R. A., Saperstein, D. A., & Sullivan, L. S. (2001). Prevalence of muta-tions causing retinitis pigmentosa and other inherited retinopathies.Human Mutation, 17(1), 42–51.

Sung, C.-H., Davenport, C. M., & Nathans, J. (1993). Rhodopsin muta-tions responsible for autosomal dominant retinitis pigmentosa. Clus-tering of functional classes along the polypeptide chain. The Journal ofBiological Chemistry, 268(35), 26645–26649.

Sung, C.-H., Schneider, B. G., Agarwal, N., Papermaster, D. S., & Nathans,J. (1991). Functional heterogeneity of mutant rhodopsins responsiblefor autosomal dominant retinitis pigmentosa. Proceedings of theNational Academy of Sciences USA, 88, 8840–8844.

Tam, B. M., Xie, G., Oprian, D. D., & Moritz, O. L. (2006). Mislocalizedrhodopsin does not require activation to cause retinal degenerationand neurite outgrowth in Xenopus laevis. Journal of Neuroscience,26(1), 203–209.

Tastan, O., Yu, E., Ganapathiraju, M., Aref, A., Rader, A. J., & Klein-See-tharaman, J. (submitted for publication). Comparison of stability pre-dictions and simulated unfolding of rhodopsin structures.Photochemistry and Photobiology.

Walsh, N., van Driel, D., Lee, D., & Stone, J. (2004). Multiple vulnerabilityof photoreceptors to mesopic ambient light in the P23H transgenic rat.Brain Research, 1013(2), 194–203.

Wu, T. H., Ting, T. D., Okajima, T. I., Pepperberg, D. R., Ho, Y. K., Ripps,H., & Naash, M. I. (1998). Opsin localization and rhodopsin photo-chemistry in a transgenic mouse model of retinitis pigmentosa. Neuro-science, 87, 709–717.

Zhou, H., & Zhou, Y. (2002). Distance-scaled, Wnite ideal-gas referencestate improves structure-derived potentials of mean force for struc-ture selection and stability prediction. Protein Science, 11(11), 2714–2726.

Zhu, L., Jang, G. F., Jastrzebska, B., Filipek, S., Pearce-Kelling, S. E., Agu-irre, G. D., Stenkamp, R. E., Acland, G. M., & Palczewski, K. (2004). Anaturally occurring mutation of the opsin gene (T4R) in dogs aVectsglycosylation and stability of the G protein-coupled receptor. TheJournal of Biological Chemistry, 279(51), 53828–53839.




retinitis pigmentosa associated with rhodopsin …rader, dhiman, klein-seetharaman, & bahar,...

Documents