February 8, 2005
Message from Dr. Gregory Acland posted to ACVO Diplomate List and VOPH List
Collie Eye Anomaly Genetics
As genetic testing for CEA is now a reality, it has been suggested to me that I should bring folks up to speed on what we know about Collie Eye Anomaly genetics.
In 2003, we published a mapping paper, locating the primary CEA locus to canine chromosome 37. (see: Lowe JK, Kukekova AV, Kirkness EF, Langlois MC, Aguirre GD, Acland GM, Ostrander EA. Linkage Mapping of the Primary Disease Locus for Collie Eye Anomaly. Genomics 2003 Jul;82(1):86-95). (Ed.note: Abstract is appended below.)
In that paper, apart from the central result, we also clarified some of the basic aspects of CEA genetics. Some of these issues are still not widely appreciated, and may cause confusion as CEA genetic testing becomes more frequent.
Prior to the work we undertook, 2 aspects of CEA genetics had become abundantly clear. First, some dogs that were clinically affected (by ophthalmoscopic exam) at a young age, subsequently appeared to be nonaffected. These dogs, generally referred to as "go-normals", have been clinically recognized since at least 1989 (see, e.g. L.F. Rubin, Inherited Eye Diseases in Purebred Dogs, Lippincott, Williams & Wilkins, Philadelphia, 1989.) It is fair to say that there has been an informal consensus that these dogs seem to behave genetically as affecteds. Second, there is strong data documenting segregation distortion in the transmission of CEA (see: B. Wallin-Hakanson, N. Wallin-Hakanson, A. Hedhammar, Collie eye anomaly in the rough collie in Sweden: genetic transmission and influence on offspring vitality, J. Small Anim. Pract. 41 (2000) 254-258; and B. Wallin-Hakanson, N. Wallin-Hakanson, A. Hedhammar, Influence of selective breeding on the prevalence of chorioretinal dysplasia and coloboma in the rough collie in Sweden, J. Small Anim. Pract. 41 (2000) 56-59). (Ed. note: Abstracts of these two reports are appended below.)
In the work we undertook prior to the 2003 mapping paper, I investigated these 2 questions, and established that the “go-normal” problem was the primary cause of the segregation distortion. (Ed.note: “Segregation distortion” means that the proportion of affected to non-affected is not as expected according to rules of Mendelian genetics.) However, first of all, you have to examine pups very carefully from a very young age (starting by 5-6 weeks postnatal) to detect all the ones that "go normal", as some pups exhibiting unmistakeable choroidal hypoplasia by 6 weeks old, will have no ophthalmoscopically detectable lesions by 9 weeks old. Having done that, if you then count all dogs clinically exhibiting CH, even temporarily, as affecteds, the segregation distortion problem goes away, and we were able to map the CH locus. Having done that, we found that a small percentage of the “go normals” were in fact not homozygous for the CH mutant allele, but were heterozygous. This data was published in the 2003 Genomics paper.
Now, as a reliable CEA DNA test becomes more widely implemented, several issues will impact veterinary ophthalmologists. The test classifies dogs as affected, carrier or normal, based on their genotype at the CH locus. However this can cause some unexpected results and reactions.
- The test will detect all homozygous mutant dogs, regardless of whether they were ever observed to be affected clinically. That is, it resolves false negatives. A significant percentage of dogs thought to be nonaffected (based on clinical exam at ages greater than 9 wks), will prove to be homozygous for the mutation.
- The test will discriminate true affecteds from false positives. The most common cause for a false positive diagnosis is clinical confusion between an albinotic or subalbinotic fundus (most usually merle) and choroidal hypoplasia.
- In rare cases, it is possible that a dog diagnosed correctly as clinically affected (or “go normal”) may prove to be a carrier, i.e. heterozygous for the CH mutation.
Linkage mapping of the primary disease locus for collie eye anomaly. Genomics. 2003 Jul:82(1):86-95. Lowe JK, Kukekova AV, Kirkness EF, Langlois MC, Aguirre GD, Acland GM, Ostrander EA. Division of Human Biology, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, D4-100, Seattle, WA 98109-1024, USA.
Collie eye anomaly (CEA) is a hereditary ocular disorder affecting development of the choroid and sclera segregating in several breeds of dog, including rough, smooth, and Border collies and Australian shepherds. The disease is reminiscent of the choroidal hypoplasia phenotype observed in humans in conjunction with craniofacial or renal abnormalities. In dogs, however, the clinical phenotype can vary significantly; many dogs exhibit no obvious clinical consequences and retain apparently normal vision throughout life, while severely affected animals develop secondary retinal detachment, intraocular hemorrhage, and blindness. We report genetic studies establishing that the primary cea phenotype, choroidal hypoplasia, segregates as an autosomal recessive trait with nearly 100% penetrance. We further report linkage mapping of the primary cea locus to a 3.9-cM region of canine chromosome 37 (LOD = 22.17 at theta = 0.076), in a region corresponding to human chromosome 2q35. These results suggest the presence of a developmental regulatory gene important in ocular embryogenesis, with potential implications for other disorders of ocular vascularization.
Collie eye anomaly in the rough collie in Sweden: genetic transmission and influence on offspring vitality. J. Small Anim. Pract. 2000 Jun;41(6):254-8. Comment in: J. Small Anim. Pract. 2001 Apr;42(4):204. Wallin-Hakanson B, Wallin-Hakanson N, Hedhammar A. Referral Animal Hospital Bagarmossen, Sweden.
Between 1989 and 1997, 8204 rough collies were examined for collie eye anomaly (CEA) at up to 10 weeks of age. All dogs were positively identified and the results were registered under the Swedish Kennel Club genetic health programme. A significant decrease in litter size occurred if one of two affected parents had coloboma (3.8 pups) compared with litters from two chorioretinal dysplasia- (CRD-) affected collies (5.2 pups) or litters by two normal collies (5.0 pups), indicating an influence of the coloboma genotype on offspring vitality. The prevalence of CRD in pups from normal x normal matings and CRD x CRD matings was significantly different from that expected under simple autosomal recessive inheritance (43 per cent versus 25 per cent and 85 per cent versus 100 per cent). The results are compatible with polygenic inheritance but not with simple autosomal recessive inheritance. CRD prevalence in offspring of CRD x coloboma matings was significantly lower than in pups of CRD x CRD matings, reflecting effects of the coloboma genotype on vitality. These results have important implications for breeding programmes and the genetic control of CEA.
Influence of selective breeding on the prevalence of chorioretinal dysplasia and coloboma in the rough collie in Sweden. J. Small Anim. Pract. 2000 Feb;41(2):56-9. Comment in: J Small Anim Pract. 2001 Apr;42(4)204. Wallin-Hakanson B, Wallin-Hakanson N, Hedhammar A. Bagarmossen Referral Animal Hospital, Sweden.
A total of 8204 rough collies, representing 76 per cent of all collies registered by the Swedish Kennel Club between 1989 and 1997, were examined before 10 weeks of age for collie eye anomaly (CEA). All dogs were permanently identified and examination results were registered and computerised. The policy of breeders during the study period was to select against coloboma in breeding stock, but to allow breeding of chorioretinal dysplasia (CRD) affected animals. The prevalence of CRD increased significantly from 54.2 per cent to 68.1 per cent (P < 0.001) from 1989 to 1997, while the prevalence of coloboma did not (8.3 per cent to 8.5 per cent, P = 0.4). These results are not compatible with a simple, recessive, autosomal inheritance for the entire CEA complex.