Identification and Functional Characterization of a Novel Rhodopsin Mutation Associated with Autosomal Dominant CSNB

Abstract

purpose. Mutations in RHO, PDE6B, and GNAT1 can lead to autosomal dominant congenital stationary night blindness (adCSNB). The study was conducted to identify the genetic defect in a large Swiss family affected with adCSNB and to investigate the pathogenic mechanism of the mutation.

methods. Two affected cousins of a large Swiss family were examined clinically by standard methods: funduscopy, EOG, ERG, and dark adaptometry. Twelve family members were screened for mutations in RHO. The ability of mutant rhodopsin to activate transducin constitutively was monitored by measuring the catalytic exchange of bound GDP for radiolabeled [³⁵S]GTPγS in transducin.

results. A novel mutation was identified in RHO (c.884C>T, p.Ala295Val) in patients with adCSNB. They had full vision under photopic conditions, showed no fundus abnormalities, revealed EOG results in the normal range, but presented night blindness with an altered scotopic ERG. In the presence of 11-cis retinal, the mutant rhodopsin is inactive, similar to wild-type, responding only when exposed to light. However, in the absence of 11-cis-retinal, unlike wild-type opsin, the mutant opsin constitutively activates transducin.

conclusions. The study adds a fourth rhodopsin mutation associated with CSNB. Although the phenotype of autosomal dominant CSNB may vary slightly in patients showing mutations in RHO, PDE6B, or GNAT1, the disease course seems to be stationary with only scotopic vision being affected. The data indicate that the mutant opsin activates transducin constitutively, which is a consistent and common feature of all four CSNB-associated rhodopsin mutations reported to date.

Congenital stationary night blindness (CSNB) comprises a group of genetically and clinically heterogeneous nonprogressive retinal disorders, mainly due to defects in rod photoreceptor signal transmission. Mutations in genes coding for proteins of the phototransduction cascade (RHO, GNAT1, PDE6B, RHOK, SAG, and RDH5) or genes associated with the transmission of the signals from the photoreceptors to the adjacent bipolar cells (NYX, GRM6, CANCA1F, and CABP4) can lead to this disease (reviewed by Zeitz ¹ ). 

So far, three mutations in RHO have been described that cause nonprogressive autosomal dominant (ad)CSNB. ^{2 3 4 5} Patients carrying these mutations show relatively mild phenotypes that result only in the loss of rod-mediated vision under dim-light conditions without retinal degeneration. In contrast, more than 100 different mutations in RHO have been described that lead to retinitis pigmentosa (RP) (http://www.sph.uth.tmc.edu.Retnet/ provided in the public domain by the University of Texas Houston Health Science Center, Houston, TX), which results in a more severe phenotype. Initially, RP resembles CSNB, since it starts with night blindness, but RP gradually leads to the peripheral death of rods, eventually resulting in total vision loss. 

Rhodopsin, which can be mutated in patients with both CSNB and RP, is a member of the G-protein-coupled receptor superfamily. This visual pigment is highly expressed in vertebrate retinal rod cells that mediate vision in dim-light conditions. It is composed of an apoprotein opsin and a chromophore, 11-cis retinal, that is covalently bound to opsin via a protonated Schiff base linkage to the ε-amino group of a lysine at position 296 located in the seventh transmembrane helix. ^{6 7} The chromophore acts as an inverse agonist that prevents G-protein activation in the dark. On absorption of a photon, the chromophore isomerizes to its all-trans form, thereby causing a series of conformational changes resulting in the activated form of the receptor, metarhodopsin II (MII). It is this form that binds the α-subunit of the G-protein transducin, initiating the signaling cascade that ultimately results in hyperpolarization of the rod cell. 

In vitro and in vivo studies have shown that all three previously described rhodopsin mutations (p.Gly90Asp, p.Thr94Ile, and p.Ala292Glu) that lead to adCSNB result in constitutive activity of the mutant protein. ^{2 3 8 9} Based on these observations, one might speculate that all rhodopsin mutations that cause adCSNB would lead to constitutively active protein. The purpose of this study was to investigate the adCSNB phenotype in a large Swiss family, to analyze whether these patients also carry a RHO mutation, and to test whether the pathogenesis is also due to constitutive activation of transducin. 

Methods

Clinical Examination of Patients

Electro-oculogram.

The EOG Arden test was performed with maximum mydriasis, applying the diagnostic system by Roland Consult (Wiesbaden, Germany) according to the manufacturer’s guidelines. In brief, skin electrodes were attached medially and laterally of both eyes, and the ground electrode was positioned on the forehead. A dark period of 15 minutes was followed by a 20-minute photopic phase. Fixation stimuli were separated by a 30° angle, generating saccades of the ocular dipole with a duration of 1.5 seconds. The Arden ratio was calculated, and the latency of the photopic increase was noted. 

Electroretinogram.

Standardized full-field ERGs were elicited with Ganzfeld stimuli using the commercial ERG system (Retiport32; Roland Consult) according to ISCEV (International Society for Clinical Electrophysiology of Vision) guidelines (standard flash, 3.0 cd · sec · m⁻²). For the recordings, pupils of both eyes were maximally dilated with 0.5% tropicamide and 0.5% phenylephrine, and the other eye was occluded. A DTL-electrode was applied, and the ground electrode was attached to the forehead. The scotopic ERG was recorded after 30 minutes of dark adaption. For the photopic ERG, the background luminance was set at 34 cd/m². 

Dark Adaption.

Dark adaption was registered according to the manufacturer’s guidelines (DARKadaptometer; Roland Consult). In brief, after maximum pupil dilation as described for ERG, bleaching of the retinal pigment epithelium (RPE) was achieved by exposure to 7000 cd/m² for 5 minutes in a 400-mm Ganzfeld bowl. Green light testing was started on a 4 × 16-dot illuminated display with a stimulus of 250 cd/m². For the next 35 minutes, stimuli with a 1-second length and decreasing intensity resulted in an overlapping photopic–scotopic answer, whereas red light testing almost selectively produced a cone-dependent graph. 

Mutational Analysis

Patient blood samples were sent to us for diagnostic testing, and informed consent was obtained from all family members examined. The protocol of the study adhered to the tenets of the Declaration of Helsinki. Genomic DNA from 5 unaffected and 7 affected members of a large Swiss adCSNB family was extracted from peripheral blood (Magnetic Separation Module 1; Chemagen AG, Baesweiler, Germany). Five PCR fragments, spanning the entire coding region and splice sites of RHO (RefSeq number NM000539.2; www.ncbi.nlm.nih.gov/locuslink/refseq/ provided in the public domain by the National Center for Biotechnology information [NCBI], Bethesda, MD) were amplified and directly sequenced as previously described. ¹⁰ In addition, PCR and sequencing of exon 4 was performed in 78 Swiss patients with RP and 168 ethnically matched control individuals. 

In Silico Prediction

Single nucleotide polymorphism (SNP) databases used were as follows (all available in the public domain):

•  
  http://www.genome.ucsc.edu/ (University of California at Santa Cruz)
•  
  http://www.ncbi.nlm.nih.gov/snp (NCBI)
•  
  http://www.ensembl.org/homo_sapiens ¹¹

Pathogenic impact prediction:

•  
  PolyPhen (polymorphism phenotyping): http://tux.embl-heidelberg.de/ramensky/ (European Molecular Biology Laboratory, Heidelberg, Germany)
•  
  SIFT: http://blocks.fhcrc.org/sift/SIFT.html (Fred Hutchinson Cancer Research Center, Seattle, WA)
•  
  Swiss Model: http://swissmodel.expasy.org/ (Swiss Institute of Bioinformatics, Geneva, Switzerland)
•  
  Expasy: http://www.expasy.org/spdbv/ (Swiss Institute of Bioinformatics, Geneva, Switzerland)

Homology Modeling

The 3D-model of the rhodopsin protein was generated as previously described. ¹⁰ using the PDB entry 1U19 and software provided by the Swiss Model and Expasy. ¹²  

Mutagenesis, Expression, Reconstitution, and Purification of the Receptors

Opsin mutants were constructed with a site-directed mutagenesis kit (QuikChange; Stratagene, La Jolla, CA) on a synthetic bovine opsin gene. ^{13 14} The expression and purification of the visual receptors was performed as described. ^{15 16} Briefly, plasmids encoding the wild-type or mutant rhodopsin genes were transfected into 90% to 100% confluent COS cells, by using DEAE-dextran ¹⁵ with 6 μg plasmid DNA/15-cm plate. The cells were harvested 72 hours after transfection. The pigments were reconstituted with 11-cis retinal in intact cells and then solubilized with 1% (wt/vol) n-dodecyl-β-d-maltopyranoside (DDM; Anatrace, Maumee, OH) in PBS (10 mM sodium phosphate buffer [pH 7.5], containing 150 mM NaCl). The pigments and apoproteins were purified by immunoaffinity chromatography on 1D4-Sepharose 4B in PBS containing 0.1% (wt/vol) DDM. All procedures up to and including binding of the receptor to the immunoaffinity matrix were performed at 4°C, whereas subsequent washes and elutions were performed at room temperature. The receptors were washed on the column five times with 2 mM sodium phosphate (pH 6.4), 150 mM NaCl in 0.1% (wt/vol) DDM, followed by five times with 2 mM sodium phosphate (pH 6.4) in 0.1% (wt/vol) DDM. The receptors were eluted in 2 mM sodium phosphate (pH 6.4), 0.1% (wt/vol) DDM containing 1.4 mg/mL 1D4 9-mer peptide (TETSQVAPA). 

Preparation of COS Cell Membranes

Membranes from five 15-cm plates of COS cells were isolated 72 hours after transfection, essentially as previously described. ⁸ The final membrane pellet was resuspended in 500 μL 10 mM BTP (pH 7.4) and homogenized with a 25-gauge needle before flash freezing on dry ice. 

Absorbance Spectrophotometry and MII Decay Measurements

UV-visible absorption spectra of purified receptors were recorded with a spectrophotometer (Cary50; Varian, Palo Alto, CA) modified for darkroom use. A previously published method of measuring the rate of MII decay ⁸ was used, with slight changes. To prevent denaturation of opsin, these experiments were performed on the highly stable rhodopsin mutant that carries a p.Asn2Cys and p.Asp282Cys substitution, which is functionally similar to wild-type rhodopsin. ¹⁶  

The decay rate was determined in light of the fact that 11-cis retinal binds to opsin at a faster rate than MII decays. In the presence of twofold molar equivalents of 11-cis retinal, samples were subjected to a light flash from a 300-W bulb passed through a 455-nm long-pass filter for 1 second, to activate rhodopsin selectively in the sample, while not photoactivating the excess 11-cis retinal. Successive spectra were taken after the flash until no further increase in absorbance at 500 nm (482 nm for Val295) was detected. Rate constants for the reactions were determined as previously reported, ⁸ by using the following equation: ΔA = (A − A _∞)/(A ₀ − A _∞), where A is the absorbance recorded at 500 nm for wild-type and at 485 nm for Val295, A _∞ is the absorbance at time ∞, and A ₀ is the absorbance at time 0. The MII half-life is the time at which ΔA is 50% of A _∞. 

Transducin Activation Assays

The ability of the receptors to activate transducin was monitored by measuring the binding of [³⁵S]GTPγS, as described ⁸ with minor changes as follows. The assays were performed in COS cell membranes at a concentration of approximately 1 nM rhodopsin (or mutant), and they contained 10 mM BTP (pH 6.7), 150 mM NaCl, 1 mM MgCl₂, 1 mM CaCl₂, 0.5 mM EDTA, 1 mM dithiothreitol, 2 μM transducin (prepared as described elsewhere ⁸ ), and 3 μM [³⁵S]GTPγS (5 Ci/mmol). The reactions were performed in the presence and absence of 11-cis retinal. For the reactions with retinal, the membranes were incubated on ice at 4°C with 100 μM 11-cis retinal for 2 hours before use. Reactions were initiated by addition of [³⁵S]GTPγS in the dark, and aliquots were removed at 1-minute intervals for filtration, to determine the amount of bound [³⁵S]GTPγS. After incubation in the dark, samples were exposed to light and the reaction continued. For wild-type receptors, light was initiated after 7 minutes 30 seconds, whereas it was initiated after 8 minutes 15 seconds for mutant receptors. 

Results

Clinical Findings

The CSNB phenotype of male patient I-1 (born 1809) from a large Swiss family was determined in 1847 by the first church warden from Donath, Switzerland. The document states that the patient has an immedicable “error of nature” that made it impossible for him to go out during dim-light conditions without assistance. In dim light, his visual field was completely dysfunctional, which made it impossible for him to work. Analysis of this patient’s descendents revealed an autosomal dominant inheritance of this phene. In seven generations, at least 16 family members had night blindness (Fig. 1) . They claimed an inability to orient themselves in dim-light conditions in dark places. 

For the study described herein, two of the affected family members (patients V-2: age 73 and V-19: age 67) had been clinically investigated. Patient V-2 was emmetropic and achieved full vision under photopic conditions. In 1999, he had experienced branch retinal venous occlusion (BRVO) of the upper temporal retinal vein in his right eye (RE). Intraocular pressure was always within the normal range, while systemic blood pressure was repeatedly elevated and mild arteriosclerotic changes of the retinal vessels were evident (Fig. 2A) . He received argon laser treatment of the upper temporal retinal quadrant, as well as systemic medication to control the blood pressure and to inhibit thrombocyte aggregation. This condition resulted in discrete retinal scarring and pigment irregularities of the right fundus (Fig. 2A , arrow) as well as a slight paracentral visual field defect (Fig. 2B , arrow). Mild age-related maculopathy of both eyes with a few serous drusen was noted, but no other ocular disease was present in the anterior or posterior segment. In addition, neither nystagmus nor squinting was detected. Patient V-19 was initially seen by an ophthalmologist due to recurrent chalazia of both eyes. He also was emmetropic and achieved full vision under photopic conditions. Mild retinal arteriosclerotic changes were evident as well as a small retinal pigment epithelium (RPE) defect close to the upper temporal vein of his left eye (LE). Additional ocular symptoms of the anterior or posterior segment, nystagmus, and squinting were not detectable (Fig. 2C) . The EOG of both patients was within the normal range (Fig. 3) . 

Scotopic ERGs of examined patients and a representative control subject are shown in Figure 4A . Patient V-2 exhibited an absent rod response (Fig. 4A , line 1; Table 1 ), as well as a negative-shaped ERG from the mixed maximum rod and cone response (Fig. 4A , line 5; Table 1 ). In patient V-19, the rod response was severely reduced (Fig. 4A , line 1; Table 1 ), but the mixed maximum rod and cone responses did not result in a negative ERG (Fig. 4A , line 5; Table 1 ). The scotopic a-wave response was within the normal range. The implicit times under scotopic conditions were not increased. The oscillatory potentials of both patients were severely diminished (Fig. 4B) . Under photopic conditions, only the amplitudes of the single photopic standard flash were slightly reduced in both patients, whereas the 30-Hz-flicker ERG did not show any alterations compared with the control (Fig. 4C ; Table 2 ). Formal dark adaptometry was performed and revealed an overlapping scotopic–photopic answer, which was reduced by three logarithmic units lacking the typical Kohlrausch’s break and resulting in a more or less monophasic configuration (Fig. 5) . These findings showed that the patients had full vision and a normal visual field under photopic conditions. Although the phenotype of both examined patients varied slightly, the disease course seemed to be stationary, with a main defect in scotopic vision. 

Mutation Analysis

Mutation analysis in RHO revealed a heterozygous c.884C>T mutation in exon 4 in the DNA of affected members of the Swiss adCSNB family (Fig. 6)resulting in a p.Ala295Val substitution. The mutation segregated with the disease in seven affected family members (IV-9, V-2, V-4, V-19, VI-6, VI-18, and VII-5), while five unaffected members did not show this exchange (VI-1, VI-2, VI-3, VI-5, and VI-20). The mutation was not identified in 78 Swiss patients with RP, ¹⁰ in 168 ethnically matched control individuals, nor reported as described SNP. 

The p.Ala295Val substitution was further investigated using different bioinformatic tools to predict a possible impact of the amino acid substitution on the structure and function of the protein. Polyphen ^{17 18} predicted a possible damaging effect (range, 1.5–2), with a value of 1.747, and SIFT (sorts intolerant from tolerant amino acid substitutions) ¹⁹ suggested a pathogenic variant with a value of 0.00 (<0.5 is predicted to be deleterious). 

Ala295 is located in the seventh transmembrane domain of rhodopsin (Fig. 7A) . The side chain of this alanine is partly oriented toward the neighboring sixth transmembrane domain facing toward the center of rhodopsin. The p.Ala295Val substitution introduces a nonpolar, β-branched amino acid in place of an alanine, which restricts its rotational freedom on the polypeptide backbone. This can introduce steric hindrance, leading to local structural perturbations between the valine side chain and the chromophore within the protein (Figs. 7B 7C) . Based on these predictions, we expressed the wild-type and mutant proteins in COS cells and compared their biochemical properties. 

In Vitro Properties of Val295 Rhodopsin

To gain a better understanding of the relationship between the pathophysiology of the disease and the role of the mutant protein in these patients, we expressed Val295 in COS cells and analyzed the biochemical properties of the mutant protein as previously described for other mutations. ^{2 3 8} We first examined the lifetime of the active form of rhodopsin, MII. When 11-cis retinal was bound to Val295 opsin, the absorbance maximum blue shifted by approximately 18 nm, compared with wild-type rhodopsin (Fig. 8A) . This indicates that the retinal binding pocket of the mutant form has been perturbed, most likely by the possible steric strain of the chromophore (Fig. 7) . To test whether this mutation affects the lifetime of the activated state, the rate of MII decay was determined by activating rhodopsin with light in the presence of excess 11-cis retinal. Reconstitution of the bleached photopigment was monitored by recording absorbance at 500 nm (482 nm for Val295) over time to determine the MII lifetime. Val295 rhodopsin has an increased MII lifetime compared with wild-type (Fig. 8B) . We found that wild-type MII decayed with a half-life of 5.8 ± 0.9 minutes at 23°C (n = 3), whereas the MII half-life of the Val295 was 9.7 ± 2.6 minutes (n = 4). This corresponds to a 1.7- fold increase in MII lifetime compared with wild-type. This slowing may not have deleterious effects on night vision in vivo, however, as under scotopic conditions, phosphorylation of the activated receptor followed by arrestin binding occurs much more rapidly than the decay of MII. ^{20 21}  

Since the other three known rhodopsin mutations associated with CSNB constitutively activate transducin, we performed transducin-activation assays with Val295 opsin. When 11-cis retinal was bound, Val295 rhodopsin activated transducin with kinetics similar to that of wild-type rhodopsin (Fig. 9 , open and filled circles). However, a striking difference was observed in the absence of the chromophore: Whereas wild-type opsin did not activate transducin (Fig. 9 , filled triangles, top panel), the mutant opsin constitutively activated it (Fig. 9 , filled triangles, bottom panel). The constitutive activation of the mutant fell within the range of the other known CSNB-associated rhodopsin mutants. ⁸  

Discussion

Genotype–Phenotype Correlation of Patients with adCSNB

This study was performed to identify the genetic defect and the pathogenic mechanism in a large Swiss family with autosomal dominant (ad)CSNB. This form of CSNB has been associated with mutations in RHO, GNAT1, or PDE6B (reviewed by Zeitz ¹ ). Our analyses revealed a heterozygous c.884C>T mutation, resulting in a p.Ala295Val substitution in rhodopsin. 

Although X-linked inherited CSNB has been described in variable degrees of myopia and nystagmus, ²² these features do not seem to be part of adCSNB. ²³ Thus, myopia and/or nystagmus do not belong to the common phenotypic repertoire in patients with adCSNB and RHO mutations, but additional genotype–phenotype correlation studies are needed to confirm our hypothesis. As observed in other inherited forms of CSNB, to date, no degenerative signs of the retina, the lens, or other ocular tissue in patients with adCSNB carrying RHO mutations have been described. We report a similar observation in the present study. The two patients were emmetropic, had normal orthoptic findings, and no RP-typical signs of degenerative retina, such as mottling of the retinal pigment epithelium with bone spicule pigmentation. Notably, in contrast to patients with RP, they had no severe visual impairment in photopic conditions (i.e., normal visual acuity and no ring- or donut-shaped mid-peripheral visual field defects). Slight arteriosclerotic retinal vasculature and some mild degenerative retinal signs in the macula, such as drusen and slight pigment irregularities, were attributed to the patients’ age. 

To date, only three RHO mutations have been described in patients with CSNB. Dryja et al. ² reported a p.Ala292Glu mutation, which was associated with unrecordable rod responses and reduced amplitude responses to white flashes, resulting in a nonnegative ERG. The implicit times for the b-wave were shortened. Sieving et al. ²⁴ described an adCSNB-like family with a p.Gly90Asp mutation. None of these patients showed a recordable rod response, nor did they exhibit rod a-waves in response to white flashes. Unrecordable rod responses were also described by al Jandal et al. ⁴ in patients from a large Irish family with a p.Thr94Ile mutation. The maximal dark-adapted response revealed a negative ERG due to a reduction in the a- and b-wave. The implicit time of the b-wave was accelerated. The oscillatory potentials were well preserved. Kabanarou et al. ²⁵ found, in addition to a slightly reduced a-wave, a negative ERG in a family with autosomal dominant inheritance. However, from their results, it was not clear whether RHO mutations were involved. 

Of interest, the different phenotypes observed under scotopic conditions segregated in the family described herein: Patient V-2 clearly revealed a negative ERG in the mixed maximum rod and cone response, whereas patient V-19 did not. In patient V-2, no rod response was detected after dim stimuli testing, whereas in patient V-19 a slight rod response was still detectable. However, in both patients the oscillatory potentials were severely diminished, whereas the b-wave implicit times were normal. Dark adaptation revealed a monophasic configuration in both patients. This finding is in agreement with the patients described by al Jandal et al. ⁴ Strikingly, in our study, the a-wave amplitudes were in the normal range, whereas they were not detectable in patients with p.Gly90Asp and were diminished in patients with a p.Thr94Ile mutation. A possible explanation could be that the wild-type allele is more highly expressed in the patients in our study, leading to a-wave amplitudes in the normal range in contrast to patients with p.Gly90Asp or p.Thr94Ile substitutions showing reduced a-wave amplitudes. This phenomenon has been suggested to occur in a large Japanese family with an adCSNB pedigree ²⁶ and in a mouse model carrying the p.Gly90Asp mutation in Rho. In the latter study it has been shown that the reduction of the a-wave depends on the relative expression of the p.Gly90Asp and wild-type rhodopsin allele with a reduction of the b-waves greater than that of the a-waves. ²⁷ In other autosomal dominant inherited retinal diseases, such as RP11, mutations in PRPF31 can even lead to partial penetrance. ²⁸ It would be interesting to test whether other family members of our Swiss family with adCSNB also have a selective b-wave reduction with the maximum flash under scotopic conditions. 

Another possible explanation is that the phenotype is influenced by genetic factors other than rhodopsin. Of note, both patients and a mouse model with a p.His258Asn mutation in PDE6B leading to another form of adCSNB ^{29 30} demonstrated a selective loss of b-wave with relative intact a-waves. ³¹ However, in mice, when the His258Asn allele was introduced into another genetic background, there was no evidence of selective reduction in b-wave amplitudes; rather a- and b-wave amplitudes were both reduced. Based on this example, the genetic background is an important factor contributing to variation of an electrophysiological response. 

In Vitro Properties of the Mutated Val295 Opsin

A unifying feature of all RHO mutations associated with night blindness to date is that the mutant apoproteins (opsins) constitutively activate the G-protein transducin. This process is thought to occur by disruption of the essential salt bridge between Glu113 and Lys296, which helps to maintain the protein in an inactive conformation. ^{2 3 32} Persistent activation of the phototransduction cascade is not limited to mutants associated with CSNB; there are two mutations found in patients with adRP that are also constitutively active: p.Lys296Met and p.Lys296Glu. ^{33 34} The phenotype in the patients with adRP is more severe, because 11-cis retinal acts as an inverse agonist, shutting off signaling once bound. Since 11-cis retinal cannot bind opsin in the constitutively active adRP mutants, it cannot serve as an inverse agonist in this manner. ⁵  

The type and location of the mutations within rhodopsin influence the level of transducin activation as follows: p.Thr94Ile < p.Gly90Asp ≈ p.Ala295Val < p.Ala292Glu. The relatively high level of constitutive activity of the Val295 opsin could be due to the combination of the specific amino acids involved and the position of the mutation. The substitution p.Ala295Val introduces a nonpolar β-branched amino acid that restricts the rotational freedom on the polypeptide backbone. This restriction has been shown to destabilize α helices, as the α helix is induced to adopt a more confined conformation with unfavored torsional angles relative to non-β-branched amino acids. ³⁵ Hence, the position of the mutation appears to be critical, because the β-branched valine at position 295 is one residue away from the active-site lysine at 296. 

Mutations at position 295 have been studied previously. ^{36 37} A serine at position 295 shifts the absorption maximum of rhodopsin of 5 nm to the blue ³⁷ , much less than the 18-nm shift seen with valine at position 295 (Fig. 8) . Ser295 rhodopsin activates transducin four times more slowly than does wild-type rhodopsin and releases retinal threefold more rapidly from MII. ³⁶ The differences between the two mutants could be due to the nature of the mutation; the introduction of a serine could destabilize the active MII conformation by increasing the rate of hydrolysis of the Schiff base due to stabilization of the tetrahedral carbinolamine intermediate of the reaction. It would then be predicted that a valine at position 295 would not stabilize the reaction intermediate in this manner. The increased rate of MII decay in Ser295 could account for the decreased initial rates of transducin activation, simply due to the reduction in active MII. It would be interesting to test whether the p.Ala295Ser mutation also leads to constitutive activity. 

Correlation of the Clinical Features with the Degree of Constitutive Activation In Vitro

In vitro studies of the known CSNB-associated rhodopsin mutations have revealed differences in the degree of constitutive activation. Patients with the p.Gly90Asp have reduced b- and undetectable rod a-waves, but show a degree of constitutive activation similar to that of patients investigated in this study with selective b-wave reduction and normal a-wave amplitudes. Constitutive activation of the phototransduction cascade has also been suggested to be the underlying pathogenic mechanism in patients carrying a p.His258Asn mutation in PDE6B. Similar to the patients in this study, the p.His258Asn mutation caused only a selective loss of b-wave in the absence of a-wave reduction. ^{30 31} Whether preferential expression of the wild-type allele may be responsible for the differences in electrophysiological observations cannot be addressed in our in vitro assays. 

In the present study, we identified the fourth RHO mutation associated with adCSNB that leads to constitutive activation of transducin. Clinical findings indicate that all patients described to date have night blindness based on ERG and subjective self-assessment but show neither myopia nor nystagmus. The mixed maximum rod and cone response does not necessarily result in a reduced a-wave or negative ERG, but a uniform feature in all patients with RHO mutations is the severely reduced rod response. Since in most patients X-linked CSNB is associated with an electronegative ERG and ocular symptoms such as myopia and nystagmus, our study brought up differences between the different forms of CSNB that will help to improve the classification of them, findings important for diagnostic purposes. Future studies will show whether novel RHO mutations in patients with adCSNB exhibit a similar phenotype and always lead to constitutive activation of the phototransduction cascade. 

³  Contributed equally to the work and therefore should be considered equivalent authors.

 Supported by the Forschungskredit of the University of Zurich, the Fondation Voir et Entendre (CZ), and the EyeSight Foundation of Alabama (AKG).

 Submitted for publication January 10, 2008; revised April 11 and May 5, 2008; accepted July 21, 2008.

 Disclosure: C. Zeitz, None; A. K. Gross, None; D. Leifert, None; B. Kloeckener-Gruissem, None; S.D. McAlear, None; J. Lemke, None; J. Neidhardt, None; W. Berger, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

A seven-generation pedigree of a Swiss family with CSNB. Filled symbols: all affected members of the family; square symbols: males; circles: females; slashed symbols: deceased persons; arrows: family members who were screened for mutations in RHO; question marks: family members with no information about the phenotype available.

Figure 2.

 

Clinical investigations of two patients (V-2 and V-19) of a Swiss family with CSNB. (A) Pigment irregularities (arrow) due to laser treatment after BRVO and mild age-related macular degeneration in otherwise normal right and left fundi of patient V-2. (B) Slight paracentral visual field defect due to retinal laser treatment after BRVO of the right eye of patient V-2. (C) Small RPE defect (left eye, arrow) and mild arteriosclerotic changes in otherwise normal left and right fundi of patient V-19.

Figure 3.

 

The EOGs of patients V-2 and V-19 showed no measurable differences compared with that of a representative control subject.

Figure 4.

 

Electrophysiology of patients V-2 and V-19. (A) Scotopic ERGs of a representative control subject and the patients. Intensities represented are 1, 0.01; 2, 0.03; 3, 0.3; 4, 1.0; and 5, 3.0 cd · s · m⁻². The rod response was severely diminished in both patients, whereas only patient V-2 showed an electronegative ERG. (B) Oscillatory potentials of both patients compared with those of a control subject. (C) Photopic ERGs of a representative control subject and patients V-2 and V-19; 30-Hz flicker ERG single photopic standard flash in both patients compared with that of a control subject.

Table 1.

 

Scotopic ERG

Control	Patient V-2	Patient V-19
1: 156.2–449.6 μV, 71–84 ms	1: 0.0 μV, N/A	1: 24.7 μV, 70 ms
2: 238.8–575.8 μV, 61–73.7 ms	2: 49.4 μV, 79 ms	2: 45 μV, 66 ms
3: 298.2–692.4 μV, 45–51.8 ms	3: 115 μV, 49 ms	3: 138 μV, 45 ms
4: 346.6–803.4 μV, 41.6–50 ms	4: 139 μV, 38 ms	4: 172 μV, 43 ms
5: 331.4–776.7 μV, 34–47.3 ms	5: 165 μV, 34 ms	5: 219 μV, 35 ms

 1–5, the light intensities depicted in Figure 4A .

Table 2.

 

Photopic ERG

Control	Patient V-2	Patient V-19
1: 65–205.6 μV, 27–32 ms	1: 89.6 μV, 29 ms	1: 74.5 μV, 27 ms
2: 117.5–250 μV, 30.2–34 ms	2: 92.9 μV, 32 ms	2: 109 μV, 31 ms

 1 and 2, light intensities depicted in Figure 4C .

Figure 5.

 

Dark-adaptation curves of both patients compared with that of a representative control subject. Crosses: rod-related response; dots: cone-dependent answer.

Figure 6.

 

Profile of control and patient sequences. Arrow: heterozygous c.884C>T nucleotide exchange.

Figure 7.

 

Structure of bovine rhodopsin highlighting key residues and a model of the mutant rhodopsin with a valine at amino acid position 295. (A) Side view of the crystal structure of rhodopsin (PDB coordinates 1U19). (*) Position of Ala295 within the RHO structure. (B) Top view of the active site in RHO showing 11-cis retinal (orange) and the active site Lys296 (red). Green: Ala295. (C) Top view of the mutated RHO showing valine at position 295. (B, C, arrowhead) marks the additional space occupied by Val295, suggesting steric clashes between the valine side chain and the chromophore.

Figure 8.

 

Spectra and mean MII half-lives of wild-type (WT) and Val295 rhodopsins. (A) Spectra of WT and Val295 rhodopsin were taken in the dark. The absorption maximum of Val295 was blue-shifted 18 nm compared with WT. (B) Rates of MII decay were measured, and the change in absorbance (500 nm for WT or 482 nm for Val295) was monitored after the pigments were bleached for 1 second in the presence of a twofold excess of 11-cis retinal. The mean MII half-life of Val295 rhodopsin increased 1.7-fold compared with WT (n = 3).

Figure 9.

 

Transducin activation by WT and Val295 opsin in the absence and presence of 11-cis retinal. Transducin activation was assayed in membranes from transfected COS cells. Opsins (▴) without added 11-cis retinal; (•) in the presence of retinal in the dark; and (○) with added 11-cis retinal after 30 seconds exposure to saturating light (hν). Error bars, SD (n = 3).

The authors thank patients and family members for their participation; Esther Glaus, Jaya Balakrishnan, and Philippe Reuge for DNA extraction; Gábor Mátyás for support in sequencing; and the National Institutes of Health and Rosalie Crouch for 11-cis retinal.