Pathological and Electrophysiological Features of a Canine Cone–Rod Dystrophy in the Miniature Longhaired Dachshund

Abstract

purpose. To characterize the electrophysiological and histopathological features of a retinal degenerative disease in a colony of miniature longhaired dachshunds known to have a form of progressive retinal atrophy (PRA).

methods. Serial electroretinograms were recorded from affected homozygous (n = 36) and heterozygous (n = 15) dogs. Morphologic investigations including immunohistochemistry and lectin histochemistry were performed on selected homozygous animals (n = 15).

results. Clinical findings included loss of tapetal hyperreflectivity. The mode of inheritance was autosomal recessive. An early dramatic reduction of cone-specific ERG amplitude with a more modest reduction in rod b-wave amplitude was demonstrated. Progressively, rod specific responses diminished until there were no recordable responses to the ERG stimuli at 40 weeks of age. Morphologic changes confirmed early cone inner and outer segment loss. Other abnormalities included opsin mislocalization and outer nuclear layer thinning due to the subsequent loss of rod photoreceptors.

conclusions. A novel canine cone–rod dystrophy has been identified.

Retinitis pigmentosa (RP) is the clinical manifestation of a diverse group of genetically determined photoreceptor degenerations in humans. Currently, human retinal degenerations are reported to affect approximately 1 in 3700 people worldwide, ¹ and there is currently no clinically effective treatment for any of the various forms. In classic rod–cone dystrophy, the rod abnormalities precede those of the cones. Clinically, rod–cone dystrophies are characterized by the early onset of night blindness (nyctalopia), progressive constriction of visual fields and loss of visual acuity. Subsequently, there may be continued constriction of visual fields and complete blindness. In cone–rod dystrophies, the cone system is more affected, resulting in severe loss of central vision, color vision, and photophobia. There may be peripheral preservation, with significant central visual loss but only mildly restricted peripheral visual fields. The dominant cone disease can result in cone–rod dystrophies' being more clinically devastating than many rod–cone degenerations. Onset may occur in childhood or early adolescence, with rapid loss of vision within months in aggressive cases. 

A canine autosomal recessive, early-onset retinal atrophy that affects miniature longhaired dachshunds (MLHDs) has been reported. ² The earliest ophthalmoscopic signs appear at approximately 26 weeks of age, coinciding in some cases with the onset of nyctalopia and with changes in the granular appearance of the tapetal fundus. Histologic changes include thinning of the outer nuclear layer, irregularity and attenuation of the rod photoreceptor outer segments, and disorganization of the rod outer segment disc lamellae. This MLHD retinal atrophy has a later onset than does the rod–cone dysplasia of Irish setters, ^{3 4} but is significantly earlier than the progressive retinal atrophy (PRA) of Tibetan terriers and progressive rod–cone degeneration in miniature poodles. ^{5 6 7}  

To characterize further this canine model of retinal degeneration in the MLHD, extensive electrophysiological recordings were performed. These were compared with the morphologic lectin and immunohistochemical changes identified within the retina. The findings suggest that the PRA in these animals is a cone–rod dystrophy and not, as previously thought, ² a primary rod degeneration with secondary cone loss. 

Methods

Animals

All puppies studied electrophysiologically were bred from either a heterozygous animal crossed with a homozygous affected dog or from two homozygous affected dogs by natural mating or by artificial insemination. Hence, all studied dogs were either affected homozygous or heterozygous. Unaffected littermates were identified through genetic screening. At 2 weeks of age, when the pups were large enough for less than 15% blood volume to be taken, as required under the Home Office License, 1 mL of blood was taken using a 23-gauge, 5/8-in. needle from the jugular vein into EDTA tubes. The sample was then screened by the Animal Health Trust genetics department to determine the dogs' PRA status before they were entered into the program at 5 to 6 weeks of age. Electroretinograms (ERGs) were recorded in affected animals (n = 36) and heterozygous littermates (n = 15) for comparison, between the ages of 6 to 45 weeks. Some of the affected group (n = 15) and two additional cases that were not studied electrophysiologically, were used for postmortem histopathological examination. One normal dog, killed for an unrelated medical condition at 8 weeks of age, was included as a morphologic control animal. In this characterization study, most animals within the affected group had received sham intravitreal injections (n = 34) at some stage in conjunction with other studies. Because of ethical considerations, we were unable to use any wild-type MLHDs for electrophysiological or pathologic characterization. Similarly, heterozygous dogs were rehomed and thus were not available for histopathological study. The animals were bred and studied under the regulation of the U.K. Animals (Scientific Procedures) Act 1986, and all animal procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Clinical Observations

Direct and indirect ophthalmoscopy and slit lamp biomicroscopy were performed at frequent intervals. In addition, lens opacities were viewed against the tapetal reflex using distant direct ophthalmoscopy or a retinal camera. Photographs were taken with a hand-held fundus camera (RCΠ; Kowa, Tokyo, Japan). Pupils were dilated with tropicamide and phenylephrine (both from Chauvin Pharmaceuticals Ltd., Romford, UK) when photographed immediately after ERG and with tropicamide only if observations were performed at intervals between ERGs. 

Electroretinography

Bilateral ERG recordings were taken between 6 and 45 weeks of age in homozygous and 8 and 34 weeks of age in heterozygous animals. Homozygotes and heterozygotes were examined at 2- to 3-week intervals under general anesthesia, using in-house purpose-built electroretinography equipment. The number of observations per animal varied from 1 to 16 (median = 6). 

The animals were premedicated with acepromazine (0.02 mg/kg; C-Vet) and buprenorphine (0.01mg/kg; Vetergesic-Alstoe Animal Health, Ltd., Melton Mowbray, UK), administered by intramuscular injection approximately 1 hour before induction of general anesthetic. Anesthesia was induced by propofol (∼5 mg/kg; Rapinovet; Schering Plough Animal Health, Uxbridge, UK) administered slowly to effect, via a cephalic intravenous catheter. The animals were then intubated with a 5- to 6.5-mm cuffed endotracheal tube. Anesthesia was maintained with halothane (1%–1.5%) in oxygen (3 L/min) depending on the individual. During the ERG, body temperature of the animals was maintained by warming with a heated pad, and the animal was insulated with bubble wrap and a vet bed at a room temperature of approximately 25°C. Mydriasis was achieved with tropicamide (1%) and phenylephrine (2.5%) eye drops. Eye-opening and position were maintained with a conjunctival suture placed on the upper conjunctiva at the limbus, above the pupil. The two ends of the suture were pulled, lifting the eye as though looking straight ahead, to optimize pupillary–retinal illumination. Both ends of the suture were then taped to the brow (Transpore; 3M, St. Paul, MN). 

ERG corneal electrodes (Jet; LKC Technologies, Gaithersburg, MD) were used in the beginning of the study in a few of the animals (n = 7) but were later replaced by a custom-made corneal contact lens electrode for improved physical and electrical stability. The recording electrodes in the custom-made corneal lens were of the same diameter area and material (gold) as for the corneal electrode, and optically they were almost identical. No single animal was tested with the two different electrodes. A methylcellulose buffer diluted with saline solution was used between the eye and the corneal electrode. Platinum iridium subdermal ground and reference electrodes were attached at the crown and one centimeter behind the lateral canthus, respectively. 

Instrumentation included a specially built Ganzfeld ERG stimulator with a 50-cm bowl, powered by a photic stimulator (model PS22C; Grass-Telefactor, West Warwick, RI), and incorporating infrared closed-circuit television monitoring of the dog during recording. ERGs were recorded on computer-based data-acquisition system (CH Electronics, Kidderminster, UK). 

The recording protocol used was based on, and incorporated recommendations from, that for the human ERG compiled by the International Society for Clinical Electrophysiology of Vision (ISCEV) standard available at the time the study commenced. ⁸ We also included additional bright stimuli (up to 12.5 cd · s/m²) as subsequently recommended in the later 2004 update. ⁹ The protocol consisted of 20 minutes of dark adaptation, followed by an intensity series using discrete white stimuli (0.012–12.5 cd · s/m²). Dark-adapted 30-Hz flicker was performed by using a relatively dim stimulus. After a minimum of 8 minutes of light adaptation, 30-Hz flicker, single-flash photopic ERG recording was performed (3.125-cd · s/m² stimulus; 25-cd · s/m² background). 

There was a relatively marked variation in ERG parameters from animal to animal, and with the multiple recordings from many different animals, there was a need to summarize the data for each case. For implicit times, this was achieved by making box-and-whisker plots of the median for each animal in the statistical programming language R. ¹⁰ The upper and lower limits of the box define the upper and lower quartiles, respectively. The horizontal bar within the box gives the median. The whiskers show the upper and lower limits of the range of the data up to 1.5 times the interquartile range above or below the nearest quartile. If data lie outside this range, the whisker marks this boundary, and outliers are plotted individually. This method is the default one used for box and whisker plots in R. 

Comparisons between homozygotes and heterozygotes were made using the Mann-Whitney test and given the multiple comparisons that were being made, P < 0.001 was taken as the required significance level. Amplitude data were more complex to analyze, as amplitudes varied markedly with the stage of the disease. The decay was approximately exponential, and so using nonlinear mixed-effect regression, we fitted the parameters for an exponential decay (equation 1)fitted to each amplitude parameter for each animal.  

\[\mathrm{Amplitude}\ {=}\ B_{0}\ {\times}\ \mathrm{exp}(B_{1}\ {\times}\ t)\]

where B ₀ is an estimate of the extrapolated maximum amplitude, B ₁ is the decay constant, and t is time. The advantage of a mixed-effect analysis is that as well as deriving estimates of B ₀ and B ₁ for the group as a whole (fixed effects), it is possible to derive parameters for each individual animal (random effects). The analysis was performed in R ¹⁰ by using the nonlinear mixed-effects library. ¹¹ It was not possible to fit random effects for B ₀ and B ₁ simultaneously, and the analysis was therefore performed twice. Fixed-effect estimates of B ₀ and B ₁ were calculated each time and averaged and random effects were estimated individually for B ₀ and B ₁ in the two separate analyses. The motivation for this approach was to find descriptors for the evolution of ERG parameters in each case rather than to perform a detailed statistical hypothesis testing. Derived parameters from the statistical analyses were plotted as box and whisker plots in R, as described earlier. 

Microscopy

Eyes were fixed for histology by immersion in either 10% vol/vol formal saline or one half-strength Karnovsky's paraformaldehyde–glutaraldehyde fixative (10% paraformaldehyde 20 mL, 25% glutaraldehyde 10 mL, distilled water 40 mL, 0.2 M sodium cacodylate buffer [pH 7.4] 100 mL, and calcium chloride 0.1 g) for 48 hours. The inferior nasal quadrant from one eye of each animal was processed through ascending concentrations of alcohol into xylene and then embedded in paraffin. General morphology was assessed in hematoxylin and eosin–stained sections. 

Lectin Histochemistry

The histochemical distribution of peanut agglutinin (PNA) was investigated by using a biotinylated lectin (Vector Laboratories, Ltd., Peterborough, UK). The labeled lectin was visualized as the brown final reaction product of diaminobenzidine (DAB) after biotin detection with streptavidin–peroxidase. Pretreatment with 0.1% trypsin in 0.1% calcium chloride (pH 7.8) at 37°C for 15 minutes (dilution 1:200) was used. 

Immunohistochemistry

The immunohistochemical distribution of synaptophysin, glial fibrillary acidic protein (GFAP), cytochrome oxidase, and opsin was investigated by using a conventional biotin-streptavidin–peroxidase method (Immunostainer; Dako, Ltd., High Wycombe, UK). Associated immuno reagents were purchased from Dako, Ltd. The primary antibodies used are listed in Table 1 . Antigen retrieval, using 0.1% trypsin in 0.1% calcium chloride at pH 7.8 and performed for 15 minutes at 37°C, was necessary for the optimal demonstration of all the antigens studied. The peroxidase label was visualized as the brown final reaction product of DAB. Appropriate positive and negative procedural controls were used throughout. 

Outer Nuclear Cell Counts

Outer nuclear layer cell counts, made by investigators blind to animal status, were taken across the retina from three consistent locations at the posterior pole, equator, and anterior. These were positioned approximately 1, 4, and 6 mm from the optic nerve head. Paraffin-embedded sections cut at a nominal thickness of 5 μm were used. A 100-μm wide sample strip defined by grid lines on a 10 × 10 eye piece graticule grid (Graticules Pyser-SGI Ltd., Edenbridge, UK) was positioned across the outer nuclear layer, and all cells within counted with the proviso that cells overlapping a boundary line were measured on only one side. The mean of the three counts was then taken, the counts were checked by an independent observer and plotted against age, and an exponential curve was fitted to the data. As with the ERG amplitude data, the two parameters B ₀ and B ₁ were estimated (equation 1) . 

Results

Clinical Observations

All clinical observations up to the age of 21 weeks disclosed no significant abnormality. The first observable ophthalmoscopic sign in affected puppies was seen at 25 weeks and was of a bilateral patchy and mild hyperreflectivity in the tapetal fundus (Fig. 1A) . These signs were not seen in heterozygous littermates. At this stage, there was no evidence of defective vision or abnormalities of pupil size or reaction to light. Over the next few months, the intensity of hyperreflectivity increased, becoming generalized and marked and accompanied by obvious attenuation of the retinal blood vessels. Finally, pigmentary changes in the nontapetal fundus were apparent. These ophthalmoscopic signs are consistent with similar forms of PRA (rod–cone dysplasias and degenerations) in many other breeds. 

Bilateral suture line opacities were noted in many homozygous and heterozygous animals. These were temporary and noted in routine examinations after intravitreal injections. They were bilateral but not usually symmetrical and disappeared completely, usually after a few days to weeks, but sometimes as soon as 24 hours or as late as 8 weeks. The opacities could involve posterior or anterior sutures lines or both and although most common at the ends they could occupy the entire length of the suture (Fig. 2) . 

Electroretinography

There were no apparent abnormalities of the ERG arising from the sham injection that some of the animals received. 

The ERG of a 6-week-old affected MLHD showed significant reduction of the 30-Hz cone-specific flicker response. At 7 weeks of age, there was little difference in the rod responses between heterozygous and affected dogs (Fig. 3) . The maximum (combined) a-wave implicit times were no different between heterozygous and affected animals, although the a-wave amplitude was slightly reduced in the affected animals. The implicit time and amplitude of the b-wave and oscillatory potentials did not differ significantly between affected and heterozygous animals at this age. 

At 14 weeks both 30-Hz flicker and 2-Hz cone responses were severely reduced in amplitude in the homozygous dogs and in some cases were indistinguishable from noise. Rod responses in the homozygous animals were of slightly lower amplitude than in the heterozygous group. The rod-dominated maximum bright-white flash ERG, was not noticeably different from that in heterozygous animals. 

By 21 weeks (see Fig. 3 ), the ERG amplitudes in the homozygous group were further reduced. Rod-specific responses became severely reduced and sometimes indistinguishable from noise, whereas the maximum bright white flash a- and b-waves were delayed and severely reduced in amplitude. Cone responses were variable and ranged from undetectable to approximately 15% of heterozygous responses. The cone system was again more affected than the rod system. In heterozygous animals, there were slight changes in the rod and cone responses with increasing age. The a- and b-wave implicit times remained unchanged but rod and cone amplitudes were reduced. 

By 30 weeks, all ERG responses from most homozygous dogs were barely detectable, whereas in the heterozygous animals, there was at most only a minor, further reduction in amplitude from that recorded at 21 weeks. 

A panel of graphs of light-adapted 30-Hz flicker amplitudes is shown in Figure 4 . Note the clear exponential decay of amplitude in some cases and also the considerable variation from animal to animal. Summary box-and-whisker plot of dark- and light-adapted 30-Hz and light-adapted 2-Hz amplitudes are given in Figure 5 . Note how the absolute amplitudes (B ₀) were lower in affected cases and that the rate of decline (B ₁) was faster. The scotopic intensity series parameters are presented in Fig. 6 . Data are given for flash intensities of 0.789 cd · s/m² (approximately the 50% V/V _max stimulus intensity in these dogs), 3.125 cd · s/m² (standard flash), and 12.5 cd · s/m² (bright-flash stimulus dominated by rod responses, but with a cone component). In Figures 6A and 6Bare the B ₀ parameters for a- and b-wave amplitudes at increasing stimulus intensity. These values affectively represent a backward extrapolation of maximum amplitude before the degeneration starts, and there was very little difference between heterozygous and affected animals. By contrast, shown in Figures 6C and 6Dare the rate constants for the decay of the scotopic responses. The decay rates were much faster in the affected pups. Note, however, that the rates of decay were much gentler than those for photopic responses. The a- and b-wave implicit times for the scotopic series are given in Figures 7A 7B . There was a consistent, modest increase in median a-wave implicit times that is significant at all intensities (P < 0.0001; P = 0.06 and P = 0.0005 with increasing intensity). There was a significant increase in median b-wave implicit time only with the highest-intensity flash (P = 0.0001). With photopic stimuli (Figs. 7C 7D) , only the single flash median times increased significantly (P = 0.0009 and P = 0.006 for a- and b-wave, respectively), but in general the homozygotes showed greater variability. 

Microscopy

Examination of hematoxylin and eosin–stained sections revealed histologic changes from an early age. At 6 weeks, there was some shortening of inner and outer segments but photoreceptors with morphologic features of both cones and rods were present. Cells with the typical nuclear appearance of cones were much less apparent by 10 weeks of age, whereas rod nuclei at this age appeared to be unaltered. Later, rods became grossly degenerated, exhibiting short, rounded inner segments, with only residual amounts of outer segment remaining. By 43 weeks, only a single row of cells within the outer nuclear layer remained. 

Immunohistochemistry

Opsin was immunohistochemically identified in the rod outer segments of the normal control retina and was seen both there and mislocalized to the inner segments and outer nuclear layer at 6 weeks. By 23 weeks, in affected dogs (Fig. 8A)opsin was observed in these locations, and immunoreactivity extended as far as the outer plexiform layer. At later time points, first outer and then inner segment staining was lost. This mirrored the morphologic changes seen by conventional microscopy. Residual perikaryal opsin staining was still evident at 45 weeks of age (Fig. 8B)when no physical inner or outer segments remained. 

In the normal retina, immunostaining for cytochrome oxidase gave a more extensive pattern of immunoreactivity in the inner segments of both rods and cones than was seen in the affected eyes. By 45 weeks, there was little immunostaining associated with the photoreceptor remnants. 

Synaptophysin immunoreactivity was observed within both the outer and inner plexiform layers of the normal retina. In the affected dogs immunolabeling of the outer plexiform layer was seen to diminish with age in parallel with the reduction in number of photoreceptor cells. Synaptophysin staining of the inner plexiform layer remained essentially constant. 

Only astrocytes were immunoreactive for GFAP in the normal retina. In the affected retina, Müller cells increasingly expressed GFAP from as early as 6 weeks of age (Fig. 8C) . 

Lectin Histochemistry

Both the cone inner and outer segments of the normal retina were clearly identified by PNA staining. In a 6-week-old affected animal, the staining intensity was slightly reduced (Fig. 8D)and cone outer segments were less apparent than inner segments. By 23 weeks of age, there was no staining of the outer segments, and there was further reduced staining in inner segments, which no longer appeared morphologically normal (Fig. 8E) . By 45 weeks, when no outer or inner segments were seen, PNA staining presented as a thin, nonspecific line on the outer surface of what remained of the photoreceptor layer (Fig. 8F) . 

Outer Nuclear Layer Counts

Outer nuclear layer neuronal density measurements are given in Figure 9 . Note that, as with the ERG amplitudes, there was an approximately exponential decay in cell number. These counts obviously reflect both cone and rod loss. There is a suggestion that the initial decay was more rapid, perhaps reflecting the earlier cone loss. For comparison with the ERG data, the parameter B ₁ for the exponential fit is −0.0076. 

Discussion

The generalized form of PRA is the canine homologue of retinitis pigmentosa in man and is a relatively common and well-recognized inherited retinal degeneration in the dog. ^{12 13} The condition was first described by Magnusson ¹⁴ in the Gordon setter in Sweden in a group of closely in-bred dogs related to an imported sire from the United Kingdom. In the original paper, the condition was called retinitis pigmentosa, but later Magnusson changed the name to PRA. ¹⁵ In the 1930s and 1940s, many cases of PRA were described in the Irish Setter and later in the miniature and toy poodles. ¹⁶ Several breeds with an inherited PRA followed, and although all were thought to show the same condition, because of the similar ophthalmoscopic and clinical signs, distinct, well-characterized variants became recognized on the basis of, for instance, age of onset. The Irish setter was shown to have a rod–cone type-1 dysplasia and a disease-causing mutation in the cGMP phosphodiesterase β-subunit gene. ³ Breeds such as the miniature poodle, cocker spaniel, and papillon exhibit a retinal degeneration of later onset. Ophthalmoscopically, the first sign in all these conditions is increased tapetal reflectivity, due to retinal thinning, followed by vascular attenuation and pallor of the optic discs. On electroretinography, rod amplitude is reduced in comparison to cones with delayed b-wave implicit time by approximately 18 months of age. ^{5 17}  

Curtis and Barnett ² described what was thought to be a rod–cone dystrophy in the MLHD. In the later stages of the condition, the rod ERG was reduced in amplitude with prolonged implicit time of a- and b-waves. No cone-response studies were performed. Histopathological examination disclosed changes in the rod outer segments at 6 weeks of age, and by 11 weeks there was reduction in the thickness of the outer nuclear layer and loss of photoreceptor cells. 

In the present study, the findings show that PRA in these MLHDs manifested as a cone–rod dystrophy. At 6 weeks of age, the 30-Hz flicker cone-specific ERG was reduced in amplitude, indicating loss of cone function when compared with that in heterozygous littermates. Reduced histochemical staining with PNA provided the morphologic correlate for this observation. Rod-specific ERGs are effectively normal at this stage. At 23 weeks, PNA staining was only apparent on cone inner segments, by which point the cone-specific ERG was no longer detectable. By 40 weeks of age, the only morphologic evidence of surviving photoreceptors was perikaryal opsin staining around the nuclei in the outer retina. There was no electrophysiological indication of any residual function. With cone photoreceptors being more severely affected in the first instance, followed by progressive rod involvement, the findings are most in keeping with a cone–rod dystrophy. We have chosen to use the term dystrophy as opposed to dysplasia, as the dominant process appears to be degenerative with loss of cones and cone function. This degeneration is best seen in a subset of animals in which cone amplitudes were relatively well preserved in early disease (Fig. 4) . This preservation is in contradistinction to, for instance, the dysplasia in the Rdy Abyssinian cat, in which the ERG is essentially nonrecordable from birth. ¹⁸ Nevertheless, we cannot exclude the possibility that some of the cones are never normally formed. 

The development of transient lenticular opacities was noted throughout the characterization. These usually presented as well-defined opacities at the ends of the posterior suture lines, sometimes the anterior suture lines, and occasionally both. They always appeared in young dogs and have never been seen after 12 months of age. The cause of the changes is, to date, unknown. They have been reported ¹⁹ in clinically normal puppies in several breeds, including the MLHD but also the Shetland sheep dog, beagle, and Cardigan corgi, among others. 

Although rare, cone–rod dystrophies have been reported in animals. In 1976, Vainisi et al. ²⁰ described the histology and electrophysiology of the retina in a baboon with a cone–rod dystrophy. Ophthalmoscopically, there were signs of retinal vessel attenuation and optic disc pallor in advanced cases, with associated behavioral abnormalities. The MLHD appears to differ from recently described retinal degeneration in the pit bull terrier ²¹ in that the rod responses appear to be relatively better preserved early in disease. 

Few data are available that describe the normal evolution of canine ERG amplitude with age, but it appears that adult amplitudes are achieved by 8 weeks. ²² The negative values for the time constants for the exponential fits in the heterozygous animals indicated that there was a decrease in amplitude with age over the period of the present study. Such a decrease has been documented in normal animals, ²³ and, given the unavailability of wild-type animals in this study, we were unable to discern whether the reduction was a result of normal aging or a manifestation of disease relating to expression of a single abnormal gene, as has been seen in other retinal degenerations. ²⁴  

The variation in ERG amplitude from case to case of affected homozygous animals was particularly striking. For example, some animals at 16 weeks had a maximum 30-Hz flicker cone amplitude of 9 μV, whereas others at the same age had no recordable response. Individual animals tended to maintain their relatively high- or low-amplitude recordings over several weeks, suggesting that this variation is not technical. Of note, there appeared to be less variation among siblings, even from different litters of the same parents, than between individuals with different parentage (data not shown). This implies that there may be an interaction between the disease-causing gene and one or more other genes. 

Opsin immunolabeling was largely restricted to the rod outer segments (ROS) in a normal animal, as is the case in other species. In the affected MLHDs, as in the Rdy cat, ¹⁸ opsin was also present in the rod inner segments and around the cell perikarya. Similar findings have been observed in rodent models of retinal degeneration including rd mice, rds mice, RCS rats, and P347L trangenic pigs and mice with the human P23H rhodopsin transgene, ^{25 26 27 28 29} as well as in human retinitis pigmentosa. ³⁰ However, the opsin-positive rod neurites sprouting toward the ganglion cell layer in the Rdy cats ¹⁸ and in patients with RP ³⁰ were not seen in our MLHDs. It is not certain whether the neurites were truly absent or were not visualized, as the retina was severely degenerated by the age of 40 weeks. 

The immunohistochemistry of synaptophysin and GFAP is consistent with other models of retinal degeneration, ^{18 31} although it is perhaps surprising that more GFAP immunoreactivity is not seen in outer portions of Müller cells as seen in many retinal degenerations. ³²  

It has been suggested that in photoreceptor degenerations and in other causes of neuronal death, the risk of cell death remains constant during the disease course. ³³ The observation in this study that the ERG amplitudes in some animals and the ONL counts decay with age in an approximately exponential fashion is entirely consistent with this concept. 

This article reports in detail the electrophysiology and histopathology of what is now recognized as a novel cone–rod dystrophy in a nonhuman model and thus generates new opportunities in the investigation of therapeutic interventions for the human disease. As the clinical problems associated with the human photoreceptor degeneration mainly relate to cone rather than rod dysfunction, this model is potentially of great benefit. 

Note added in Proof

Since the submission of this article the genetic basis of this retinal degeneration has been established (Mellersh CS, Boursnell ME, Pettitt L, et al. Canine RPGRIP1 mutation establishes cone-rod dystrophy in miniature longhaired dachshunds as a homologue of human Leber congenital amaurosis. Genomics. 2006;88:293-301). 

 Supported by the Wellcome Trust.

 Submitted for publication June 23, 2004; revised May 6 and September 25, 2005; accepted July 16, 2007.

 Disclosure: C. Turney, None; N.H.V. Chong, None; R.A. Alexander, None; C.R. Hogg, None; L. Fleming, None; D. Flack, None; K.C. Barnett, None; A.C. Bird, None; G.E. Holder, None; P.J. Luthert, 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.

 

Table 1.

 

Antigens and Immunoreagents

Antigen	Host Species	Reaction	Antibody Supplier	Catalog No.	Dilution (for 1 h)
Synaptophysin	Rabbit	Demonstrable in retinal synapses	Dakocytomation, Glostrup, Denmark	A0010	1:75
GFAP	Rabbit	Astrocytes and reactive Müller cells	Dakocytomation	Z0334	1:4000
Cytochrome oxidase 1	Mouse	Differing amounts in all cells and marked in inner segments	Invitrogen-Molecular Probes, Eugene, OR	A6401	1:100
Opsin	Rabbit	Retinal outer segments	Dean Bok, University of California at Los Angeles School of Medicine		1:4000

Figure 1.

 

Fundus photographs of (A) a 26-week-old affected and (B) a 4-year-old affected MLHD showing tapetal hyperreflectivity, vessel attenuation, and progressive optic atrophy.

Figure 2.

 

Photograph of a lens in a 15-week-old MLHD, showing large, well-defined opacities along the posterior suture lines.

Figure 3.

 

ERG intensity series of a homozygous affected and a heterozygous MLHD at 7 and 21 weeks of age. Traces show rod and cone responses recorded in the right eyes of littermates for comparison. Some noise was apparent on the bottom traces from the 21-week-old affected animal.

Figure 4.

 

Montage of graphs of light-adapted 30-Hz photopic flicker amplitude data plotted against time for individual affected animals. Dotted line: the exponential decay fit. There was substantial variation from animal to animal.

Figure 5.

 

Box-and-whisker plots of parameters for the equation: amplitude [μV] = B ₀ × exp(B ₁ × age[weeks]). Data for heterozygotes are given on the left and for homozygotes on the right. A, dark-adapted 30-Hz flicker; B, light-adapted 30-Hz flicker; C, light-adapted single-flash a-wave amplitude; and D, light-adapted single-flash b-wave amplitude. B ₀, which represents the maximum backward extrapolated amplitude at the onset of the degeneration, was clearly very low in the affected animals. Furthermore, the rates of degeneration, reflected in larger negative values of B ₁, were substantially greater in the homozygotes. The overall picture was of early-onset cone degeneration.

Figure 6.

 

Box-and-whisker plots of parameters for the equation: amplitude [μV] = B ₀ × exp(B ₁ × age[weeks]). Data are given for (A , C) heterozygotes and (B, D) homozygotes. A, a-wave parameters; B, b-wave parameters. Suffix 1, 0.789; 2, 3.125; and 3, 12.5 cd · s/m² for increasing stimulation intensities. B ₀, which reflects the backward-extrapolated maximum amplitude at the beginning of the degeneration was relatively similar in heterozygotes and homozygotes but the rate of degeneration reflected in B ₁ was much faster in affected animals.

Figure 7.

 

Box-and-whisker plots of ERG implicit times (in milliseconds). For details of the plots see Methods. Data are given for (A, C) heterozygotes and (B, D) homozygotes. (A, B) Scotopic data A, a-wave measurements; B, b-wave. Suffix 1, 0.789; 2, 3.125; and 3, 12.5 cd · s/m² for increasing stimulation intensities. (C, D) Da and La 30 Hz, implicit time data for dark- and light- adapted 30-Hz flicker stimuli, respectively; Phota and Photb, implicit time data for photopic a- and b-waves.

Figure 8.

 

Light photomicrographs of retinas from homozygous, affected animals. All sections were nuclear counterstained with hematoxylin. (A) MLHD 23 weeks old. Opsin staining could be clearly seen surrounding the photoreceptor nuclei and extending toward the OPL. (B) MLHD 45 weeks old. Residual perikaryal opsin staining was present although no photoreceptor inner or outer segments remained. (C) MLHD 6 weeks old. GFAP-stained Müller cells are shown. (D) MLHD 6 weeks old. A near-normal number of cones were visualized with PNA staining. (E) MLHD 23 weeks old. Fewer photoreceptor nuclei were present; consequently, there was ONL thinning. Cone remnants were observed with PNA staining. (F) MLHD 45 weeks old. The retina had grossly degenerated with only a few photoreceptor nuclei remaining within the greatly thinned ONL. No inner or outer photoreceptor segments were demonstrated with PNA. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bars, 30 μm.

Figure 9.

 

Graphs of outer nuclear layer counts with time in homozygous MLHDs. Note the approximate exponential decay in cell count.

The authors thank Claire Wallace and Kirsty Mayers (Animal Health Trust [AHT]) for help and support in caring for the dogs; Jackie Brearley (AHT) for valuable assistance with anesthesia; and Dean Bok (University of California at Los Angeles School of Medicine) for the generous gift of the anti-opsin antibody.