Development of retinal bullae in dogs with progressive retinal atrophy

Abstract

Objective:

To report the development of focal bullous retinal detachments (bullae) in dogs with different forms of progressive retinal atrophy (PRA).

Procedures:

Dogs with three distinct forms of PRA (PRA-affected Whippets, German Spitzes and CNGB1-mutant Papillon crosses) were examined by indirect ophthalmoscopy and spectral domain optical coherence tomography (SD-OCT). Retinal bullae were monitored over time. One CNGB1-mutant dog was treated with gene augmentation therapy. The canine BEST1 gene coding region and flanking intronic sequence was sequenced in at least one affected dog of each breed.

Results:

Multiple focal bullous retinal detachments (bullae) were identified in PRA-affected dogs of all three types. They developed in 4 of 5 PRA-affected Whippets, 3 of 8 PRA-affected Germans Spitzes and 15 of 20 CNGB1-mutant dogs. The bullae appeared prior to marked retinal degeneration and became less apparent as retinal degeneration progressed. Bullae were not seen in any heterozygous animals of any of the types of PRA. Screening of the coding region and flanking intronic regions of the canine BEST1 gene failed to reveal any associated pathogenic variants. Retinal gene augmentation therapy in one of the CNGB1-mutant dogs appeared to prevent formation of bullae.

Conclusions:

Retinal bullae were identified in dogs with three distinct forms of progressive retinal atrophy. The lesions develop prior to retinal thinning. This clinical change should be monitored for in dogs with PRA.

1 |. INTRODUCTION

Progressive retinal atrophy (PRA) is the commonest category of inherited retinal degeneration in dogs. The term describes a group of inherited degenerative disorders of the retina with shared clinical signs that result in vision impairment and blindness. PRA has been identified in more than one hundred breeds of dogs.^1,2 It is most commonly a rod-led degeneration (rod-cone dystrophy), with secondary and slower loss of cones that leads in many instances to complete blindness. The molecular genetic basis of the condition has been investigated in many dog breeds revealing the genetic heterogeneity of the condition, a feature shared with the analogous human condition, retinitis pigmentosa. In some forms of PRA studies have shown that cones are either involved prior to, or at the same time as rods, such pheno-types are classified as cone-rod dystrophies.^3,4 However, the funduscopic changes indicating a bilateral progressive retinal thinning are similar to those of the rod-cone dystrophies.

The typical clinical signs of PRA include a loss of night vision and on funduscopic examination indications of progressive retinal thinning (eg, tapetal hyperreflectivity) and accompanying retinal vascular attenuation. In some forms of PRA, tapetal hyporeflectivity may be an early change,⁵ perhaps indicating that prior to the death of photoreceptors and resulting retinal thinning, alterations are occurring that result in absorption of more of the examination light. Some forms of PRA have changes specific to that particular form; for example, dominant PRA due to a rhodopsin mutation renders the retina of affected dogs very sensitive to light damage. This can lead to patchy degeneration brought on by environmental light exposure.⁶ A form of PRA reported in Whippets by Somma et al.⁷ has been associated with the formation of multiple retinal bullae, a change that had not been previously reported in dogs with PRA. In addition, Dufour et al.⁸ recently reported that they occasionally observe transient retinal separation in dogs with X-linked PRA type 2 which is due to a mutation in the RPGR (Retinitis Pigmentosa GTPase Regulator) gene.

In this study, we provide further details of bullae formation in PRA-affected Whippets and report that retinal bullae are a feature of two other forms of PRA. In addition, in one of those forms of PRA we show preliminary data suggesting that gene therapy can prevent bullae formation.

2 |. MATERIALS AND METHODS

2.1 |. Animals

Procedures were conducted according to the ARVO statement for Use of Animals in Ophthalmic and Vision Research. The study at Michigan State University was approved by the Institutional Animal Care and Use Committee. Studies in Brazil were approved by the Animal Use Ethics Committee of the Federal University of Paraná.

Three groups of dogs with three different forms of autosomal recessive PRA were included. Only dogs examined at early disease stages were included. Heterozygous littermates were also examined for the two groups maintained in a colony. The first group consisted of 5 PRA-affected Whippets (4 males and one female) and 3 males heterozygous for PRA from a colony of Whippets established at Michigan State University derived from the previously reported PRA-affected Whippets.^7,9 These dogs were monitored from 3 months up to at least 18 months of age. The second group consisted of 8 PRA-affected (6 male and 2 females) privately-owned German Spitzes aged 3–12 months diagnosed and examined at the Veterinary Teaching Hospital of the Federal University of Paraná (Brazil). The final group consisted of 20 Papillon beagle crosses (11 males, 9 females) homozygous for a previously described mutation in rod cyclic nucleotide gated channel subunit beta 1 (CNGB1)⁵ followed from 3 to at least 18 months of age. One of these dogs underwent retinal gene augmentation therapy as previously reported.¹⁰ Six dogs heterozygous for the CNGB1 mutation (1 male, 5 females) were also examined. The Whippet and Papillon crosses were housed in the same facility under a 12/12 h light cycle and fed the same commercial dog food.

2.2 |. Eye examination and fundus photographs

For examination and spectral domain optical coherence tomography pupils were dilated with tropicamide (1% Tropicamide, Akorn Inc., or Alcon Laboratórios do Brasil). Ophthalmic examination included slit-lamp biomicroscopy (Kowa^® SL-17, Kowa American Corporation or Hawk Eye; Dioptrix), indirect ophthalmoscopy (Keller Vantage Plus wireless LED, Keeler Instruments and Heine Omega 200; Heine Instruments) and fundus color images were collected with a RetCam II (Clarity Medical Systems, Inc) or a ClearView Optical Imaging System (Optibrand).

2.3 |. Spectral domain optical coherence tomography (SD-OCT)

Retinal imaging was performed with dogs under general anesthesia. Briefly, the dogs were pre-medicated with subcutaneous or intramuscular acepromazine (0.02–0.1 mg/Kg; Aceprojet, Henry Schein Animal Health or Acepran, 0.2%—Vetnil) and in some cases with buprenorphine (0.01–0.03 mg/kg; Reckitt Benckiser Company). Anesthesia was induced with intravenous propofol (4–6 mg/kg; PropoFlo, Zoetis or Propovan 1%, Cristália), and maintained on 1.5%–3% Isoflurane (Isothesia Inc, Henry Schein Animal Health). Eyes were positioned in primary gaze using conjunctival stay sutures (4–0 or 6–0 silk, Ethicon, LLC, Johnson & Johnson Company). Imaging was performed using a combined confocal scanning laser ophthalmoscopy (cSLO) and spectral domain optical coherence tomography (SD-OCT) (Spectralis HRA/OCT, Heidelberg Engineering). Infrared (IR) and autofluorescence (AF) cSLO images of the fundus were acquired using a 55° lens. Horizontal and vertical SD-OCT single line scans as well as raster volume scans from the areas with bullae and from the rest of the retina were obtained using a 30° lens.

2.4 |. Screening canine BEST1 gene for variant

Canine multifocal retinopathy (CMR) is associated with retinal bullae formation and is due to homozygous mutations in the BEST1 gene.^11,12 To screen for the unlikely possibility that a BEST1 mutation was segregating in the PRA-affected dog pedigrees the entire coding region (including parts of the flanking introns) were Sanger sequenced in two affected Whippets and one CNGB1-mutant dog. The sequence for PRA-affected German Spitz was obtained from whole genome sequencing of 2 PRA-affected dogs. For Sanger sequencing blood DNA was extracted using standard protocols and polymerase chain reaction used to amplify the coding region and intronic regions flanking exons of the BEST1 gene (see Table S1 for the PCR primers used). Amplicons were submitted for Sanger sequencing. The resulting DNA sequence was aligned to the canine reference sequence using Canfam3.1 and screened for any polymorphisms. Whole genome sequencing was performed as previous described.¹³ Aligned sequence files were viewed using Integrated Genomics Viewer (version 2.4.9) and screened for variants from the Canfam3.1 genome in the coding exons and flanking intronic regions of BEST1.

2.5 |. Gene augmentation therapy

An adeno-associated viral construct (serotype 5) packaged with CNGB1 cDNA was delivered by two subretinal injections of about 200 μl of a titer of 5 × 10¹¹ vg/ml. Procedures have been previously described.¹⁰

3 |. RESULTS

3.1 |. Fundus imaging

Multiple retinal bullae were detected in dogs with all three types of PRA. The age they were first identified varied between breeds being as early as 3 months of age in the Whippets to about 6 months of age in the German Spitz. In addition, the localization, number and size of the bullae differed between the groups (Figures 1–3). In all three breeds the subretinal fluid appeared to be clear on indirect ophthalmoscopy and color fundus imaging. On cSLO IR imaging the bullae were clearly visible in the tapetal fundus. On AF imaging bullae within the non-tapetal fundus showed hyperfluorescence. SD-OCT imaging confirmed that the lesions were focal bullous retinal detachments and the subretinal fluid was hyporeflective. SD-OCT was found to be a more sensitive method for detecting the presence of bullae, particularly those that were relatively flat detachments and those that were in the non-tapetal area.

FIGURE 1.

Retinal bullae in a PRA-affected Whippet. (A) Wide-angle color fundus image of the right eye of a 6-month-old male Whippet showing 1 large bullae in the area centralis region and several smaller bullae. (B) cSLO infrared and (C) Autofluorescent image of the same dog. (D) IR image, the site of the magnified SD-OCT image in (E) is indicated by the white bracket. (E) SD-OCT high resolution cross-section image across the bulla in the area centralis region indicated in D. In purple are the measurements of the width and height of the bulla

FIGURE 3.

Multiple bullae in a homozygous CNGB1-mutant dog (Papillon derived). (A) Wide-angle fundus color image showing presence of multiple bullae across the tapetal fundus. (B) A cSLO infrared image of the tapetal region showing the multiple bullae. (C) A cSLO autofluorescent image of the non-tapetal fundus. Bullae show as hyperfluorescent focal regions. (D) IR image the site of the magnified SD-OCT image in (E) is indicated by the white bracket. (E) SD-OCT image across the bulla in the central retina indicated in (D). White arrows indicate the position of the bullae in (D and E)

Four of the 5 PRA-affected Whippets (2 males and 2 females) developed retinal bullae localized to the central tapetal fundus (Figure 1), but there were fewer bullae per eye compared to the other 2 groups. Bullae were detected from 3 months of age and were up to 1210 μm in diameter (Figure 1). They were first detectable by SD-OCT but most became large enough to be seen by indirect ophthalmoscopy and RetCam photography. Figure 4 shows serial SD-OCT examinations of the large bulla in a male PRA-affected Whippet. As the photoreceptor degeneration became apparent (ie, thinning of the photoreceptor layer became apparent on SD-OCT examination) the bullae tended to flatten, and the outer retina of the detached region degenerated more rapidly than the surrounding attached retina which can be clearly seen in the cross-sectional retinal images at 12 and 18 months ages (Figure 4). None of PRA carrier Whippets had bullae detected.

FIGURE 4.

Progression of the retinal bulla in the right eye of the male PRA-affected Whippet shown in Figure 1. The same region is imaged starting at 3 months of age in the top image where a small bulla is present. Note the expansion of the bulla by 6 months of age after which it decreases in size. Outer retinal thinning is apparent at the 12 and 18 month ages. Note that there is more severe degeneration in the region of the previous bulla. Red arrows indicate same region in all images

Three of the 8 PRA-affected German Spitz dogs developed multiple bullae (from 40 to 90 per eye) involving the whole tapetal area, with a tendency to be more concentrated near the dorsal blood vessels. Bullae diameter was always small, from one fourth to one sixth of the optic disk diameter (Figure 2). The precise age at development is not known but at 3 months of age bullae were not detected but when examined at 6 months of age they could be detected. Seven of the 8 PRA-affected German Spitz dogs were examined by SD-OCT as well as indirect ophthalmoscopy. One dog was only examined by indirect ophthalmoscopy and no bullae were detected in this animal.

FIGURE 2.

Multiple bullae in a PRA-affected German Spitz dog. (A and B) are right and left eyes of the same female dog at 7 months of age. (C) is a cSLO infrared fundus image of the right eye indicating the position of the SD-OCT scan shown in (D). (D) Multiple bullae are present across the retinal cross-sectional SD-OCT image (indicated by white arrows)

Bullae were identified in 15 of the 20 CNGB1-mutant dogs. They varied between individual animals in size, but they tended to be numerous and first detected in the peripheral tapetal and non-tapetal fundi, before developing in the more central tapetal regions (Figure 3). They appeared to become larger in the periphery with the largest measured reaching 3800 μm in diameter. As with the German Spitz the lesions were not present at the initial examinations of 2 – 3 months of age but appeared from about 4 months of age. They became more obvious at between 6 and 9 months of age then appeared to decrease from about 12−15 months of age. For 6 of the 15 with bullae these were detected by SD-OCT but were not apparent in the RetCam images and not noted on indirect ophthalmoscopy. The 5 CNGB1-mutant dogs in which bullae were not detected SD-OCT was not performed and they were only examined by indirect ophthalmoscopy and RetCam imaging. Bullae were not detected in any of the dogs heterozygous for the CNGB1 mutation.

The CNGB1-mutant dog that had undergone gene augmentation therapy did not develop bullae in the treated retinal regions but did develop multiple bullae in the surrounding untreated retina (Figure 5).

FIGURE 5.

Retinal gene augmentation therapy in a homozygous CNGB1-mutant dog appears to protect against retinal bullae formation. (A) Wide-angle fundus color photograph taken immediately after subretinal injection of an AAV vector delivering a normal copy of the cDNA for CNGB1. Two separate injection blebs have been created – dog was 3 months of age. The retinotomy site for one bleb is indicated by a white arrowhead. (B) The same eye as in A 6 months post therapy. Note that there are no bullae in the region of either treatment bleb (also note the slight change in tapetal color in the treated areas). The non-treated region has developed multiple bullae. The arrowhead indicates the scar from the injection retinotomy. (C) cSLO Infrared image showing the lack of bullae in the region of the gene augmentation therapy treatment. (D) A cSLO infrared image with a green line indicating the position of the SD-OCT cross-sectional image in (E). (E) SD-OCT across the treated/untreated retinal region. The treated region is indicated with the white line. The presence of bullae in the untreated region is shown by the arrows

3.2 |. Sequence of exons and adjacent portions of the introns of Best1

Sequencing of the exons and flanking intronic regions of BEST1 in dogs of each breed revealed the presence of SNPs in the coding regions of exons 2, 4, 7 and 8 (Table S2). In each case these were synonymous variants which were previously recorded in the canine SNP database.

4 |. DISCUSSION

Progressive retinal atrophy in dogs has been recognized since the 1900s.¹⁴ Today it is known that it is a genetically heterogeneous condition that varies between forms in mode of inheritance, age of onset and rate of progression (for a recent review of PRA see Petersen-Jones and Mowat 2021²). The different forms of PRA share clinical features including a progressive photoreceptor loss leading to retinal thinning, retinal vascular attenuation and optic nerve head atrophy. The precise molecular mechanism underlying different genetic forms of PRA may differ and can include mechanisms such as failure of phototransduction leading to accumulation of cyclic GMP in photoreceptors (eg, in PDE6B and PDE6A mutations^10,15).

We recently described a new form of PRA in Whippets in which retinal bullae formation occurred prior to retinal thinning.⁷ There was also an absence of the ERG b-wave suggesting impaired photoreceptor to bipolar cell synaptic transmission.⁹ In the current study 4 of the 5 PRA-affected Whippets examined were identified to have bullae formation. Here we also report retinal bullae formation prior to retinal degeneration in two additional groups of dogs with two distinct forms of PRA; German Spitz with an early-onset autosomal recessive form of PRA and in dogs with PRA due to a mutation in CNGB1 that we have previously identified and characterized.⁵ In PRA-affected German Spitz 3 out of 8 of the examined dogs (prior to 12 months of age) and in CNGB1-mutant dogs 15 of 20 dogs examined were found to develop bullae. Bullae were not identified in any of the heterozygotes for the Whippet form of PRA nor heterozygotes for CNGB1-PRA. This strongly suggests that the presence of diseased photoreceptors prior to extensive cell death is necessary for the formation of bullae.

The bullae represent focal small bullous retinal detachments characterized by subretinal fluid accumulation. While many of the lesions could be seen by indirect ophthalmoscopy, SD-OCT was a more sensitive method to detect small, flatter detachments and in particular those in the non-tapetal fundus. In dogs where the lesions were monitored over time, the detached retinal region tended to degenerate more rapidly than the adjacent attached region (see Figure 4). This is most likely because of the additional deleterious effect of separation of the diseased photoreceptors from the supporting RPE.

There were some differences in the bullae between the three breeds. The Whippets tended to have fewer bullae compared to PRA-affected German Spitz and the CNGB1-mutant dogs. Retinal gene therapy in one CNGB1 dog seemed to prevent bullae formation just in the treated retinal regions. The therapy restored normal retinal function and preserved structure (data not shown) suggesting that the presence of abnormal and degenerating photoreceptors was necessary for the formation of the bullae.

The development of retinal bullae in dogs with 3 genetically distinct forms of PRA suggest that their formation is not specific to the PRA-causing gene mutation. The causal mutations in Whippets and German Spitz will be the subject of separate publications and are not in genes related to CNGB1 which is mutated in the Papillon derived dogs in this study. They are all photoreceptor specific genes and not expressed in RPE. There is an additional previous report of bullae formation in dogs with X-linked PRA type 2.⁸ Retinal bullae formation has also been reported by three independent groups in dogs with sudden acquired retinal degeneration syndrome (SARDS).^16–18 As noted in the current study, the authors reporting retinal bullae formation in dogs with SARDS found SD-OCT a more sensitive tool to detect the presence of bullae. The development of bullae formation in several forms of PRA with different disease mechanisms and also in dogs with a non-inherited cause of photoreceptor degeneration suggests active photoreceptor degeneration is the common factor.

There are several possible mechanisms for the formation of subretinal fluid. These include disruption of the normal pump mechanism of the RPE or a loss of the integrity of the Blood Retinal Barrier (BRB). The RPE actively transports fluid out of the subretinal space, while keeping K+ and lactate levels in this compartment tightly controlled. This function of the RPE maintains a negative hydrostatic pressure, essential for the adhesion between RPE and photoreceptors; failure of this transport system leads to retinal edema and retinal detachment.¹⁹ Bestrophin, encoded by the BEST1 gene, plays an important role in this process (for a review see²⁰).

The BRB (inner/outer) plays an important role in the homeostatic regulation of the retinal microenvironment by controlling fluid and molecular movement between the ocular vascular meshwork and the retinal tissues and prevents leakage into the retina of macromolecules and other potentially noxious agents as previously reported. The inner BRB (iBRB) is formed by the tight junctions between the capillary endothelial cells and Muller cells. The outer BRB (oBRB) is consisted by tight junctions between cells of RPE and the Bruch’s membrane separates the neural retina from the choriocapillaris and is essential for transporting nutrients from the blood to the outer retina. Toxic products from the degenerating retina are suggested as a possible cause of breakdown of the BRB.^21,22

Retinal bullae formation has not been reported in humans with the analogous condition, retinitis pigmentosa as far as the authors can tell. However, cystoid macular edema does occur in up to 50% of RP patients further reducing visual function.²³ In contrast to the findings in dogs in this study human patients with cystoid macular edema have fluid accumulation within retinal layers rather in the subretinal space.^23,24

The fact that not all the PRA-affected dogs in the three groups developed bullae suggests the need for additional factors that could include environmental influences or background genetics. To investigate the latter possibility, we sequenced the BEST1 gene in affected dogs to screen for coding variants. Mutations in BEST1 cause canine multifocal retinopathy which is characterized by multiple retinal focal detachments.^11,12 No variants that changed predicted coding were detected. Further studies are required to understand the additional factors that lead to bullae formation and why it occurs in these three forms of PRA. Although not reported in other forms of PRA, it is possible that lesions might develop transiently in the early stages of PRA in other breeds of dog. Once retinal thinning is established the bullae seem to resolve meaning that many dogs diagnosed with PRA in the clinic have more advanced disease than the dogs in this study when they were identified with bullae. The authors have seen dogs of other breeds with multiple bullae that have gone on to develop generalized retinal degeneration (SPJ unpublished observations). It is also tempting to speculate whether dogs with more advanced PRA that have retinal regions with more advanced retinal degeneration could have previously had bullae in those regions.

In conclusion, veterinarians should be aware that retinal bullae can develop in some forms of PRA and occur prior to retinal thinning and that they are no longer apparent once retinal degeneration is established. The retina in the affected region appears to degenerate more rapidly than in the adjacent regions not affected by bullae. SD-OCT is a sensitive tool for detecting these bullae.

Funding information

Myers-Dunlap Endowment for canine health, MSU/Universidade do Parana/fundings, NIH R24EY027285.