Gene therapy restores vision in a canine model of childhood blindness.

The relationship between the neurosensory photoreceptors and the adjacent retinal pigment epithelium (RPE) controls not only normal retinal function, but also the pathogenesis of hereditary retinal degenerations. The molecular bases for both primary photoreceptor and RPE diseases that cause blindness have been identified. Gene therapy has been used successfully to slow degeneration in rodent models of primary photoreceptor diseases, but efficacy of gene therapy directed at photoreceptors and RPE in a large-animal model of human disease has not been reported. Here we study one of the most clinically severe retinal degenerations, Leber congenital amaurosis (LCA). LCA causes near total blindness in infancy and can result from mutations in RPE65 (LCA, type II; MIM 180069 and 204100). A naturally occurring animal model, the RPE65-/- dog, suffers from early and severe visual impairment similar to that seen in human LCA. We used a recombinant adeno-associated virus (AAV) carrying wild-type RPE65 (AAV-RPE65) to test the efficacy of gene therapy in this model. Our results indicate that visual function was restored in this large animal model of childhood blindness.

Identification of genetic variation and haplotype structure of the canine ABCA4 gene for retinal disease association studies.

Over 200 mutations in the retina specific member of the ATP-binding cassette transporter superfamily (ABCA4) have been associated with a diverse group of human retinal diseases. The disease mechanisms, and genotype-phenotype associations, nonetheless, remain elusive in many cases. As orthologous genes are commonly mutated in canine models of human blinding disorders, canine ABCA4 appears to be an ideal candidate gene to identify and study sequence changes in dogs affected by various forms of inherited retinal degeneration. However, the size of the gene and lack of haplotype assignment significantly limit targeted association and/or linkage approaches. This study assessed the naturally observed sequence diversity of ABCA4 in the dog, identifying 80% of novel variations. While none of the observed polymorphisms have been associated with blinding disorders to date, breed and potentially disease specific haplotypes have been identified. Moreover, a tag SNP map of 17 (15) markers has been established that accurately predicts common ABCA4 haplotypes (frequency > 5%) explaining >85% (>80%) of the observed genetic diversity and will considerably advance future studies. Our sequence analysis of the complete canine ABCA4 coding region will clearly provide a baseline and tools for future association studies and comparative genomics to further delineate the role of ABCA4 in canine blinding disorders.

Comparative genomic mapping of uncharacterized canine retinal ESTs to identify novel candidate genes for hereditary retinal disorders.

PURPOSE: To identify the genomic location of previously uncharacterized canine retina-expressed expressed sequence tags (ESTs), and thus identify potential candidate genes for heritable retinal disorders. METHODS: A set of over 500 retinal canine ESTs were mapped onto the canine genome using the RHDF(5000-2) radiation hybrid (RH) panel, and the resulting map positions were compared to their respective localization in the CanFam2 assembly of the canine genome sequence. RESULTS: Unique map positions could be assigned for 99% of the mapped clones, of which only 29% showed significant homology to known RefSeq sequences. A comparison between RH map and sequence assembly indicated some areas of discrepancy. Retinal expressed genes were not concentrated in particular areas of the canine genome, and also were located on the canine Y chromosome (CFAY). Several of the EST clones were located within areas of conserved synteny to human retinal disease loci. CONCLUSIONS: RH mapping of canine retinal ESTs provides insight into the location of potential candidate genes for hereditary retinal disorders, and, by comparison with the assembled canine genome sequence, highlights inconsistencies with the current assembly. Regions of conserved synteny between the canine and the human genomes allow this information to be extrapolated to identify potential positional candidate genes for mapped human retinal disorders. Furthermore, these ESTs can help identify novel or uncharacterized genes of significance for better understanding of retinal morphology, physiology, and pathology.

Targeting gene expression to cones with human cone opsin promoters in recombinant AAV.

Specific cone-directed therapy is of high priority in the treatment of human hereditary retinal diseases. However, not much information exists about the specific targeting of photoreceptor subclasses. Three versions of the human red cone opsin promoter (PR0.5, 3LCR-PR0.5 and PR2.1), and the human blue cone opsin promoter HB569, were evaluated for their specificity and robustness in targeting green fluorescent protein (GFP) gene expression to subclasses of cones in the canine retina when used in recombinant adeno-associated viral vectors of serotype 5. The vectors were administered by subretinal injection. The promoter PR2.1 led to most effective and specific expression of GFP in the long- and medium-wavelength-absorbing cones (L/M cones) of normal and diseased retinas. The PR0.5 promoter was not effective. Adding three copies of the 35-bp LCR in front of PR0.5 lead to weak GFP expression in L/M cones. The HB569 promoter was not specific, and GFP was expressed in a few L/M cones, some rods and the retinal pigment epithelium. These results suggest that L/M cones, the predominant class of cone photoreceptors in the retinas of dogs and most mammalian species can be successfully targeted using the human red cone opsin promoter.

Mapping of KIT adjacent sequences on canid autosomes and B chromosomes.

B chromosomes are often considered to be one of the most mysterious elements of karyotypes (Camacho, 2004). It is generally believed that mammalian B chromosomes do not contain any protein coding genes. The discovery of a conserved KIT gene in Canidae B chromosomes has changed this view. Here we performed analysis of sequences surrounding KIT in B chromosomes of the fox and raccoon dog. The presence of the RPL23A pseudogene was shown in canid B chromosomes. The 3' end fragment of the KDR gene was found in raccoon dog B chromosomes. The size of the B-specific fragment homologous to the autosome fragment was estimated to be a minimum of 480 kbp in both species. The origin and evolution of B chromosomes in Canidae are discussed.

A second-generation genetic linkage map of the domestic dog, Canis familiaris.

Purebred strains, pronounced phenotypic variation, and a high incidence of heritable disease make the domestic dog uniquely suited to complement genetic analyses in humans and mice. A comprehensive genetic linkage map would afford many opportunities in dogs, ranging from the positional cloning of disease genes to the dissection of quantitative differences in size, shape, and behavior. Here we report a canine linkage map with the number of mapped loci expanded to 276 and 10-cM coverage extended to 75-90% of the genome. Most of the 38 canine autosomes are likely represented in the collection of 39 autosomal linkage groups. Eight markers were sufficiently informative to detect linkage at distances of 10-13 cM, yet remained unlinked to any other marker. Taken together, the results suggested a genome size of about 27 M. As in other species, the genetic length varied between sexes, with the female autosomal distance being approximately 1.4-fold greater than that of male meioses. Fifteen markers anchored well-described genes on the map, thereby serving as landmarks for comparative mapping in dogs. We discuss the utility of the current map and outline steps necessary for future map improvement.

A set of canine interrepeat sequence PCR markers for high-throughput genotyping.

One hundred and sixteen interspersed repetitive DNA sequence (IRS)-PCR markers have been developed and characterized from Canis familiaris for high-throughput filter-based genotyping. We present a detailed analysis of markers produced by amplification using primers directed to the conserved regions of the C. familiaris short interspersed nuclear element (Can-SINE). The majority of IRS-PCR markers developed were moderately to highly polymorphic with mean heterozygosity (HET) and polymorphism information content (PIC) values of approximately 0.6. The HET value for 22.3% of the markers exceeded 0.7. We also demonstrate that sequence variation of Can-SINEs between breeds is significant and also represents a rich source of polymorphisms. Mapping of 73 of the markers to the existing integrated linkage-radiation hybrid map enriches the map as well as establishes the utility of the markers. The significance and utility of this new class of IRS-PCR Can-SINE-based markers for high-throughput genotyping is discussed. This method can also be extended to other species that are currently map-poor but have a sufficiently high density of SINEs to allow IRS-PCR.

Cloning and expression of type II collagen mRNA: evaluation as a candidate for canine oculo-skeletal dysplasia.

The disease phenotype of oculo-skeletal dysplasia (OSD) detected in Labrador retrievers and Samoyeds shows a large degree of similarity with human Stickler and Kniest dysplasia. Type II collagen (COL2A1) mRNA, which is defective in a larger number of Stickler and Kniest patients, has been cloned and characterized from normal dog. The amino acid sequence of the canine type II procollagen is predicted to contain 1487 residues, with high degree of homology with its human homologue, and maintains all the characteristic structural domains. In addition to cartilage, expression of COL2A1 has also been detected in canine retina and testes. In testes, the N-propeptide region of COL2A1 displayed differential splicing and expressed both splice variants, IIA (with exon 2) and IIB (without exon 2), suggesting the importance of both forms in testis maturation and maintenance. Despite a severe decrease of type II collagen protein in the vitreous of OSD affected Labrador retrievers, COL2A1 gene has been excluded from having any causal association with the disease locus by linkage analysis. Using an intragenic RFLP marker, COL2A1 gene has also been tested as a candidate gene for the non-allelic form of the other canine OSD identified in Samoyeds, and excluded by linkage analysis. Oculo-skeletal dysplastic Labrador retriever and Samoyed provide two animal models for chondrodysplasia with genetic heterogeneity.

TIMP-1 expression is increased in X-linked progressive retinal atrophy despite its exclusion as a candidate gene.

X-linked progressive retinal atrophy (XLPRA) is the only known natural animal model for X-linked retinitis pigmentosa (XLRP), a blinding disorder in man. The tissue inhibitor metalloproteinase 1 gene (TIMP-1), present in close proximity to one of the two XLRP loci, was tested as a candidate for XLPRA, by first characterizing the cDNA and gene from a normal dog. The cloned canine TIMP-1 cDNA is predicted to encode a protein of 207 amino acids with 66-83% identity in the deduced aa sequence with homologous mammalian genes. No sequence difference in the coding sequence of TIMP-1 was observed between normal and XLPRA-affected dogs. TIMP-1 was found to be expressed in all of the canine tissues examined by reverse transcription and polymerase chain reaction. The canine TIMP-1 spans 3.5kb and is interrupted by five introns with sizes comparable to those observed in the human and mouse homologues of the gene. The proximal promoter region of canine TIMP-1 contains sequence motifs shown to have regulatory significance in transcription of human TIMP-1. Linkage analysis between XLPRA and TIMP-1 using a newly identified intragenic polymorphism identified recombinants, which conclusively excluded the gene as a candidate for the disease. TIMP-1 is overexpressed several months before retinal degeneration is histologically evident in XLPRA dogs, implying that alterations in interphotoreceptor matrix composition precede retinal degeneration by a significant time period.

Cloning and characterization of the cDNA and gene encoding the gamma-subunit of cGMP-phosphodiesterase in canine retinal rod photoreceptor cells.

Rod photoreceptor cyclic GMP-phosphodiesterase (cGMP-PDE) is one of the key enzymes of the visual phototransduction cascade in the vertebrate retina. The enzyme is composed of alpha- and beta-catalytic subunits and two identical inhibitory gamma-subunits. A defect in any of the subunits may potentially alter the activity of the enzyme, leading to aberration in the visual phototransduction. We have cloned and sequenced both the cDNA and gene for the canine gamma-subunit of cGMP-PDE (PDE gamma). The 952-bp cDNA has a coding region of 261 bp which is very similar to those of the PDE gamma cDNAs from human, mouse and bovine retinas. Among the 87 amino acids encoded by the transcribed region, differences in only three residues located within the first 17 amino acids were identified. The carboxyl terminus of PDE gamma, involved in interaction with the catalytic subunits of cGMP-PDE and the alpha-subunit of transducin, is conserved through evolution. The single polyadenylation signal (AATAAA) present in human and bovine PDE gamma cDNAs is replaced by AGTAAA in the canine sequence. The canine gene (2.8 kb) consists of four exons and is much smaller than the human gene (6 kb). The larger size of the human gene is primarily due to the presence of AluI repetitive elements in its first two introns.

Characterization of canine photoreceptor phosducin cDNA and identification of a sequence variant in dogs with photoreceptor dysplasia.

Photoreceptor dysplasia (pd) is an autosomal recessive disease of miniature schnauzer dogs causing retinal degeneration. The disease is a homologue of retinitis pigmentosa, a group of genetically heterogeneous diseases, causing blindness in humans. A subtraction library was prepared from retinas of pd affected and age-matched normal control dogs to isolate de novo candidate genes for further examination. From the subtraction library, cDNA for phosducin (PDC), a member of the phototransduction pathway, was isolated as a transcript expressed at a higher level in the affected retina. First, the normal canine PDC cDNA was characterized to evaluate the PDC gene in the pd-affected retina. The characterized region of normal PDC cDNA spans 1258 nucleotides (nt) that include 738 nt of coding sequence predicted to encode a protein (Mr=28 209) of 245 amino acids (aa). Over the coding region, PDC shares 86-95% nt sequence identity and 90-95% identity in the deduced aa sequence with homologous mammalian sequences. A major transcript (1.9 kb) was observed only in retina by Northern analysis, but low levels of transcript were detected in brain, liver and kidney by reverse transcription and polymerase chain reaction. Retinal immunocytochemistry showed that PDC was detected only in rod photoreceptors, mainly in the inner segment and perinuclear region. By Northern blot analysis, increased PDC expression was observed in pre-degenerate affected retina relative to the age-matched normal. In pd- affected miniature schnauzer pedigree, a missense mutation was detected in codon 82 (CGA to GGA) that would create a non-conservative substitution (Arg to Gly) in close vicinity to the residue (Glu 85) which directly interacts with the betagamma-subunits of transducin. Only pd-affected dogs were found to be homozygous for the mutant allele, and none among 48 dogs tested from 20 other dog breeds had this allele, suggesting that the mutation could be causally associated with pd in miniature schnauzers. However, since some affected dogs are heterozygous for the mutant allele, and some are homozygous for the wild-type allele, this putative PDC missense mutation, if it is indeed a disease causing mutation, does not account entirely for the genetics of inherited retinal degeneration in the miniature schnauzer breed.

Molecular cloning, characterization and expression of a novel retinal clusterin-like protein cDNA.

A novel gene expressed predominantly in retina, but detected at a conspicuously lower level in retina of canine progressive rod cone degeneration (prcd), has been identified by suppression subtractive hybridization and retinal cDNA library screening. The characterized region of cDNA of the novel gene includes 1017 nucleotides of coding sequence predicted to encode a protein of 338 amino acids (M(r) 39389), 791 nucleotides of 5'-untranslated region (UTR), and 300 nucleotides of 3'-UTR including the poly(A)(+) tail. Multiple transcripts were detected in retina by Northern blot analysis, and a lower level of expression was observed in brain and liver by RT-PCR. The transcript appears to be developmentally regulated with a burst in gene expression at a time period (34 postnatal days) that coincides with the photoreceptor differentiation phase of retinal development. The deduced amino acid sequence from the cDNA of the novel gene has 24% identity and 48% similarity with the multifunctional glycoprotein clusterin. Hence, the putative gene product from the novel transcript has been named clusterin-like protein 1 (CLUL1). The human homologue of CLUL1 cDNA has 84 and 70% identity at the level of nucleotides and amino acids, respectively, with the characterized canine cDNA. The presence of a stretch of 128 amino acids in the putative human CLUL1, not detected in canine CLUL1, suggests alternate splicing events. An STS database search revealed that the human homologue of CLUL1 maps to chromosome 18p, a location not yet reported to harbor an RP locus. Tissue-specific expression of CLUL1 in retina, and its lower abundance in different forms of PRA suggest that this novel gene may represent an as-yet unidentified locus for a retinal disorder.

Molecular characterization and mapping of canine cGMP-phosphodiesterase delta subunit (PDE6D).

cGMP-phosphodiesterase (PDE) is composed of two catalytic (alpha and beta) and two identical inhibitory (gamma) subunits. The human gene (PDE6D) encoding a new subunit (delta) has been characterized and mapped to the long arm of chromosome 2 (HSA2q35-q36) where a new autosomal recessive retinitis pigmentosa (arRP) locus (RP26) has been localized. Characterization of the canine PDE6D shows the gene is about 4.2kb containing four exons interrupted by three introns; the size of the cDNA is 1059bp with an open reading frame (ORF) of 453bp. A single transcript of identical size (1.43kb) was detected in all tissues examined (liver, lung, spleen, kidney, heart, brain and retina), with the highest abundance in the retina. Canine PDE6D has been localized to canine radiation hybrid group 14-a, which extends conserved synteny between the dog, human chromosome 2q and mouse chromosome 1. The characterization of the canine PDE6D gene and its mapping provide important information for testing causal association of the gene with canine retinal degenerations, in particular rod-cone dysplasia 2 (rcd2) in collie dogs. This disease is characterized by abnormal retinal cGMP metabolism due to a deficiency in cGMP-PDE activity, yet the alpha, beta and gamma subunits of PDE have been excluded as candidate gene loci.

Structure and analysis of the transducin beta3-subunit gene, a candidate for inherited cone degeneration (cd) in the dog.

The cDNA for the beta3-subunit of cone-specific transducin (Tbeta3) was cloned and characterized from wild type dogs, and used in linkage studies as a candidate gene for cone degeneration. Sequence analysis of the Tbeta3 cDNA revealed an open reading frame of 1020 bp, potentially coding for a protein of 340 amino acids (aa). The deduced aa sequence of canine Tbeta3 shares 97% identity with the previously identified human Tbeta3, and 82% identity with bovine rod-specific transducin (Tbeta1). RT-PCR and sequencing of the amplified products demonstrated that the retinal canine Tbeta3 gene is expressed in two different transcripts which can be generated by alternative splicing of the intron in the 3'-untranslated region (UTR). The short and the long mRNAs differ in the length of their 3'-UTR by 456 nt. We have also determined the genomic organization of the canine Tbeta3 gene; it consists of ten exons and the first exon is in the 5'-UTR. The cDNA encoding Tbeta3 from cd-affected dogs was also cloned and sequenced. We found no differences at the nucleotide level between the cDNAs isolated from normal and diseased retinas. The level of transcription of Tbeta3 mRNA in the cd dog retina appeared to be normal. Linkage analysis of a crossbred informative pedigree showed five obligate recombinants out of nine informative offspring. These results suggest that Tbeta3 is not a candidate gene for the cone degeneration of the cd mutant.

Nonallelism of erd and prcd and exclusion of the canine RDS/peripherin gene as a candidate for both retinal degeneration loci.

PURPOSE: To determine whether early retinal degeneration (erd) and progressive rod cone degeneration (prcd), two canine hereditary retinal degenerations, are caused by allelic mutations; to determine the cDNA sequence of the canine RDS/peripherin homolog (CFRDSP); and to test whether mutations(s) in CFRDSP cause(s) either erd or prcd. METHODS: Three erd-affected dogs were crossbred to three prcd-affected dogs, and their progeny were tested by electroretinography and retinal morphology for evidence of retinal degeneration. Canine RDS/peripherin cDNA was cloned and sequenced after reverse-transcription-polymerase chain reaction (RT-PCR) of total retinal RNA. A set of overlapping fragments of CFRDSP cDNA amplified from normal and prcd-affected retinal RNA was examined by double-stranded conformational polymorphism analysis for evidence of any mutation in prcd-affected dogs. RDS/peripherin-specific restriction fragment length polymorphism (RFLP) allelic differences within informative prcd and erd pedigrees were sought by digestion of amplified regions of the CFRDSP gene with different restriction enzymes. A Hinf I RFLP was identified with alleles segregating in a set of prcd and erd informative pedigrees. Linkage of CFRDSP to either prcd or erd was tested using the LINKAGE analysis package. RESULTS: All progeny from the erd x prcd cross were phenotypically normal at ages beyond the age of diagnosis for both parental disorders. The sequence of CFRDSP cDNA is reported (Genbank accession U27349). It is, overall, 79% identical at the nucleotide level with the corresponding human sequence. The coding region shares 89% and 93% nucleotide identity with the corresponding human and feline sequences, respectively. No mutation has been identified in the coding region of CFRDSP in prcd-affected dogs. In prcd pedigrees informative for both prcd and the CFRDSP Hinf I RFLP, a minimum of six obligate recombinants were identified. Similarly, in erd pedigrees, 14 of 29 progeny informative for both erd and this RFLP were obligate recombinants. CONCLUSIONS: The canine erd and prcd mutations are nonallelic. The canine RDS/peripherin gene (CFRDSP) has been excluded as a candidate for both prcd and erd. The demonstrated informativeness of the canine pedigrees on which these studies were based will enable testing other candidate genes for prcd and erd. Sequence information of CFRDSP will enable testing this locus as a candidate in other canine hereditary retinal degenerations.

Auditory testing of dogs with inherited retinal degeneration.

Auditory function was tested by brainstem-evoked response (BSR) audiometry in dogs affected by hereditary retinal degeneration (HRD). Comparison of BSR thresholds and latency-intensity functions revealed no significant difference between progressive rod-cone degeneration (PRCD) affected and unaffected miniature poodles, and no evidence of sensorineural hearing loss in HRD-affected English cocker spaniels and miniature schnauzers. The authors conclude that hearing loss is not a feature of the retinal degenerations in these dogs.

Retinal pathology of canine X-linked progressive retinal atrophy, the locus homologue of RP3.

PURPOSE: To describe the course of photoreceptor disease in canine X-linked retinal degeneration. METHODS: Retinas from 55 dogs (44 males, 8 carrier females, 3 homozygous females) were obtained by enucleation under general anesthesia. After fixation and dehydration, tissues were embedded in epoxy resin, sectioned at 1 microm for light microscopy and stained with azure II/methylene blue and a paraphenylenediamine counterstain. For electron microscopy, regions identified by light microscopy were selected and cut at 60 nm. Sections were stained with uranyl acetate-lead citrate. Electroretinography from an additional group of normal males, affected males, and carrier females was performed and the rod and cone responses evaluated. RESULTS: The earliest lesion detectable by electron microscopy was vesiculation of rod discs, followed by disruption of outer segments and death of rods. Loss of cones and progressive atrophy of inner retinal layers followed. Lesions were most severe in the peripheral retina and advanced toward the optic disc with disease progression. Significant variation in disease severity was present in males despite the presence of the same disease allele in all affected dogs. Carrier females displayed generalized reduction in photoreceptor density as well as multifocal areas of complete rod loss. The electroretinogram (ERG) findings were compatible with the histopathologic abnormalities. Homozygous females had lesions similar to those seen in affected males. CONCLUSIONS: X-linked retinal degeneration is characterized by initial degeneration of rod photoreceptors, followed by loss of cones and progressive atrophy of the inner retina. Carrier females display a phenotype consistent with random X-chromosome inactivation. Variation in genetic background may alter expression of the disease allele in affected animals, thus accounting for variation in phenotypic expression of the disease.

Cosegregation of codon 807 mutation of the canine rod cGMP phosphodiesterase beta gene and rcd1.

PURPOSE: To determine if a previously reported nonsense mutation (G to A transition at nucleotide position 2420) in the canine rod cyclic GMP (cGMP) phosphodiesterase beta (PDEB) subunit gene cosegregates with the rod-cone dysplasia 1 disease allele (rcd1) in the rcd1-dog reference colony; to establish the prevalence of this mutation among rcd1-affected Irish setters in the United States; and to screen for this mutation in other forms of canine hereditary progressive retinal atrophy (PRA). METHODS: Exon 21 of canine PDEB, previously reported to contain a nonsense mutation in rcd1-affected dogs, was amplified by polymerase chain reaction from genomic DNA isolated from peripheral blood samples. The mutation was detected in amplified DNA by restriction enzyme digestion and double-stranded conformational polymorphism. Linkage between rcd1 and the PDEB mutation was tested using the computer program LIPED: RESULTS: Three different rcd1-informative canine pedigrees were tested for the PDEB nonsense mutation. The first was a multigenerational pedigree representing the rcd1 reference colony. The other two pedigrees represented purebred Irish setter breeding lines in which rcd1 was known to be segregating. In all three pedigrees, the same point mutation was present and segregated with no discordance with the rcd1 allele. Linkage analysis established a maximum logarithm of odds (LOD) score of 12.05 at a linkage distance (theta) of 0.0. In a representative sampling of Irish setters in the United States diagnosed clinically as affected with typical rcd1 phenotype, all dogs were demonstrated to have the same (codon 807) PDEB mutation. Three of four Irish setters affected with atypical, relatively slower disease also had this mutation, but one dog did not. This point mutation in the canine PDEB gene was absent in other forms of canine hereditary retinal degeneration. CONCLUSIONS: In three informative pedigrees, the codon 807 mutation in canine PDEB cosegregates with the rcd1 disease allele with zero discordance. A linkage distance (theta) of zero, with an LOD score of 12.05, indicates identity of this mutation and rcd1. This appears to be the only mutation causing rcd1 in the United States. In all other forms of canine hereditary retinal degeneration tested (cd, erd, prcd, rcd2, X-linked PRA, and in one Iris degeneration tested (cd, erd, prcd, rcd2, X-linked PRA, and in one Irish setter with late onset PRA), this PDEB point mutation was absent.

Docosapentaenoic acid is converted to docosahexaenoic acid in the retinas of normal and prcd-affected miniature poodle dogs.

PURPOSE: Docosahexaenoic acid (DHA, 22:6n-3) is the major fatty acid of photoreceptor membranes that is necessary for optimal retinal function. Miniature poodle dogs with progressive rod-cone degeneration have lower plasma levels of DHA than normal dogs and higher 22:5n-3/22:6n-3 ratios. The purpose of this study was to test the hypothesis that the metabolic defect in dogs affected with progressive rod-cone degeneration was a reduced capacity for ocular synthesis of DHA from its precursor 22:5n-3. METHODS: The in vivo retinal conversion of [14C]22:5n-3 to [14C]22:6n-3 was investigated by injecting normal dogs and dogs affected with progressive rod-cone degeneration intravitreally with [14C]22:5n-3. After 72 hours, rod outer segments, remaining retina, and retinal pigment epithelium/choroid were analyzed for metabolic products. RESULTS: Using high-performance liquid chromatography, six radioactive peaks were detected in both normal and affected dogs: [14C]20:5n-3, [14C]22:6n-3, [14C]22:5n-3, [14C]24:6n-3, [14C]24:5n-3, and [14C]18:0. The majority of the label in each tissue was in 22:6n-3 and there was no difference in the dpm% of [14C]22:6n-3 in normal and affected animals in any of the three tissues. Voss et al (J Biol Chem 1991;266:19995-20000) proposed a new pathway for the synthesis of DHA that involves elongation of 22:5n-3 to 24:5n-3, desaturation to 24:6n-3, and beta-oxidation to 22:6n-3. Identification of the products [14C]24:5n-3 and [14C]24:6n-3 verified that these putative intermediates are present in the dog retina. CONCLUSIONS: The finding of large amounts of label in DHA shows that the normal and progressive rod-cone-degenerated retina and retinal pigment epithelium are capable of DHA synthesis from 22:5n-3. The presence of radioactivity in 24:5n-3 and 24:6n-3 suggests that the synthesis of DHA in the retina is similar to that described in the liver.

Synthesis and release of docosahexaenoic acid by the RPE cells of prcd-affected dogs.

PURPOSE: Dogs affected with progressive rod-cone degeneration (prcd) have reduced levels of docosahexaenoic acid (DHA, 22:6n-3) in their plasma and rod photoreceptor outer segments (ROS). Dietary supplementation of DHA has failed to increase the ROS DHA levels to that of unaffected control dogs. The present study was undertaken to test the hypothesis that prcd-affected dogs have a reduced capacity for the synthesis and/or release of DHA in retinal pigment epithelial (RPE) cells. METHODS: RPE cells (first passage cultures) from prcd-affected and normal dogs were incubated with [3H]eicosapentaenoic acid (EPA, 20:5n-3) for 24 and 72 hours. After incubation, the radiolabeled fatty acids in the cells and media were analyzed. RESULTS: DHA and all its metabolic intermediates were detected in RPE cells from prcd-affected and normal dogs. No significant difference was found in the amount of products (including DHA) synthesized between normal and affected RPE cells at either time point. In the culture media, RPE cells from prcd-affected dogs released significantly more DHA than cells from normal dogs after 72-hour incubation, but not after 24-hour incubation. CONCLUSIONS: RPE cells from prcd-affected dogs can synthesize and release DHA at least as efficiently as cells from normal dogs. Therefore, synthesis of DHA from its precursor and its release from RPE cells does not appear to contribute to the reduction in ROS DHA levels found in prcd-affected animals.