Cranial cruciate ligament rupture is considered a common cause of hind limb lameness in dogs. In a study1 of Newfoundland dogs that were evaluated at a veterinary teaching institution in a 6-year period, 36 of 163 (22.1%) had CCLR. The economic impact of medical and surgical treatment for CCLR in dogs in 2003 was estimated at > $1.3 billion.2 The mechanical instability created by loss of the CCL inevitably leads to osteoarthritis3–5; however, the severity of osteoarthritis does not correlate with lameness grade.6 Although multiple different surgical techniques are described as treatments for CCLR, no technique is considered superior to another for that purpose.7 The expected outcome of surgical treatment is clinical improvement in 85% to 90% of dogs. Unfortunately, < 50% of dogs achieve complete structural soundness8,9; in 1 study,7 < 15% of 131 dogs regained 80% of normal limb function within 6 months after surgery.
Specific breeds of dogs have an increased incidence of CCLR, whereas other breeds appear to be protected against development of CCLR. This suggests a genetic basis for CCLR in this species. Results of a previous study1 indicated that CCLR in Newfoundlands likely has an autosomal recessive mode of inheritance with 51% penetrance; the frequency of the recessive allele was 0.60. On the basis of these findings, determination of genetic susceptibility to CCLR in individual dogs is difficult. Also, in that study,1 heritability was estimated to be 0.27, which indicates that environmental effects have a moderate to high impact on the expression of CCLR in those dogs. A dog that has the genetic predisposition to rupture of the CCL must also have the contributing environmental conditions to express the phenotype. In addition, the age at which CCLR occurs varies among dogs, and the gold standard for diagnosis of CCLR is confirmation during surgery, which is an invasive procedure. Thus, Newfoundlands may be bred before onset of clinical signs or confirmation of diagnosis, thereby allowing genetic transmission of the trait to offspring and contributing to maintenance of the overall prevalence of the disease in this breed of dog.
Cruciate ligament rupture follows a similar clinical course in young female human athletes. It is estimated that, on a yearly basis, 38,000 women sustain injury to an anterior cruciate ligament.10 The importance of this problem becomes more evident when one considers that athletically active females are 2 to 8 times as likely to injure an anterior cruciate ligament as are athletically active males.11–13 Furthermore, a familial predisposition, regardless of gender, has been detected among athletes with rupture of the anterior cruciate ligament.14 The ability to identify an underlying genetic basis for CCLR in dogs would allow possible advances in understanding the etiopathogenesis of the disorder in humans as well as provide a large number of individuals in which to assess treatment options and preventative measures before their potential application in humans.
Although phenotypic causes for CCLR in both humans and dogs have been investigated, no comparative candidate genes have been identified in association with the condition. The purpose of the study reported here was to identify chromosomal regions that are associated with CCLR in a population of Newfoundlands. Our hypothesis was that there is 1 or more chromosomal regions associated with CCLR status in this breed of dog.
Materials and Methods
Selection of dogs—Pedigree data and samples of DNA were collected from Newfoundlands in a study1 performed at Iowa State University. The study protocol was approved by the Iowa State University Committee on Animal Care, and written consent was obtained for all owners of study participants. For each of those dogs, determination of CCL status (ie, unaffected or affected with CCLR) was performed. The dogs were classified as affected with CCLR on the basis of results of stifle joint examination, including signs of pain on hyperextension, joint effusion, decreased range of motion, positive response during a cranial drawer sign test or cranial tibial thrust assessment, radiographic evidence of effusion and osteoarthritis, and confirmation of a ruptured CCL during surgery.15 From this pedigree, affected and unaffected dogs were selected to undergo genotyping on the basis of their predicted statistical likelihood of being homozygous CCLR-unaffected or homozygous CCLR-affected dogs. The statistical method for this selection procedure has been reported elsewhere.16
Genotyping and selection of markers—Two complementary genome-wide association investigations were conducted. An initial investigation was performed involving 130 MSATs. Most of the MSATs were selected from a minimal screening set (MSS-2) derived in another study.17 Initially, primer optimization for the 130 MSATs was performed for 8 additional randomly selected Newfoundlands; data generated from those 8 dogs were not used in the final analyses of the study. The procedure allowed selection of 97 of the 130 MSATs for use in the broad genome scan; selection was based on ease of scoring and informativeness of marker, as determined by polymorphic content of alleles. A second phase allowed the number of markers in the genome-wide investigation to be increased by the inclusion of an additional 425 MSATs. This list was generated from MSATs that are currently used in the Veterinary Genetics Laboratory, University of California, Davis. Similarly, markers were selected for this second phase of the genome-wide investigation on the basis of ease of scoring and ability to amplify. Genotyping of all MSATs was performed by use of standard laboratory procedures; the MSATs were multiplexed by use of established protocols of the respective laboratory performing the genotyping.a
Statistical analysis—Initial analysis of the MSATs included assessment of the number and frequency of alleles and genotypes and calculation of heterozygosity and PIC for each marker. These analyses were performed separately for the CCLR-affected and CCLR-unaffected dogs.18 A C2 analysis was then performed to compare the allelic and genotypic frequencies for each marker for each group of dogs (ie, those assumed to be homozygous CCLR-unaffected and those assumed to be homozygous CCLR-affected). Some of the contingency tables (CCLR status by allele and CCLR status by genotype) had cell counts (observations) < 5, which is a violation of an assumption of the classic C2 test. Thus, P values of the C2 test for each marker were obtained by use of permutation testing with Monte Carlo simulations. Briefly, the null hypothesis of the permutation test was that CCLR status is independent of allele or genotype; therefore, the C2 statistic derived from the data should be consistent with C2 statistics generated after randomly shuffling the CCLR status relative to the markers. If the original C2 statistic was an outlier (infrequent) relative to the distribution of randomly generated C2 statistics (< 5% are greater than the original statistic [ie, P < 0.05]), then the statistical test result was considered nominally significant. For each marker, 5,000 random C2 statistics gave a stable distribution for comparison with the original C2 statistic. To adjust for simultaneous multiple testings, the Storey and Tibshirani method19 was used to determine the threshold for significance by controlling the false discovery rate.
Results
Dogs—Of 205 Newfoundlands for which DNA samples were available and pedigree and cruciate status were known, 90 were included in the study. As previously stated, each dog was selected for genotyping on the basis of its predicted statistical likelihood of being genetically homozygous CCLR-unaffected or homozygous CCLR-affected.16 Fifty-three dogs were considered unaffected (16 males and 37 females); mean age of these dogs was 6.73 years. Thirty-seven dogs were considered affected (16 males and 21 females); mean age of these dogs was 7.15 years, and the mean age at which CCLR was diagnosed was 4.7 years.
The original optimization results revealed that 107 of the 130 selected MSATs could be reliably scored. Of the 107 MSATs, 97 were polymorphic. These 97 MSATs provided genome coverage of the 38 autosomes in the canine genome (mean interval, 28 cM). Initial C2 analyses of the genotyping results for the 90 Newfoundland dogs revealed no association of any chromosomal region and CCLR status. For the expanded genome-wide study, 435 MSATs were selected for genotyping on the basis of ease of scoring and ability to amplify. This provided additional genome coverage with informative markers at intervals of approximately 5.5 cM. Cumulatively, genotyping was performed for 532 MSATs, of which 10 were duplicated to allow for error checking. Ten markers failed to amplify in most dogs, and 17 markers were essentially monomorphic; thus, there were 495 informative markers. Of these 495 markers, the mean number of alleles per marker was 5 (range, 2 to 20 alleles/marker). Median heterozygosity for all markers for the CCLR-unaffected dogs was 0.55 (range, 0 to 0.91), and median heterozygosity for the CCLR-affected dogs was 0.54 (range, 0 to 0.92). Median PIC value for all markers for the CCLR-unaffected dogs was 0.49 (range, 0.02 to 0.90), and median PIC value for the CCLR-affected dogs was 0.48 (range, 0.03 to 0.91). Markers with a PIC value ≥ 0.3 but ≤ 0.59 are considered moderately informative.20 Informativeness is important for determining the marker density necessary for a genome scan.
On the basis of the nominal value of P < 0.05, 86 markers were considered nominally significantly associated with the CCLR trait. The 86 markers were located on 31 autosomes. By use of the Storey and Tibshirani method of false discovery rate correction (with significance set at a q value of < 0.05), 4 markers (located on Canis familiaris chromosomes 3, 5, 13, and 24) were considered to be significantly associated with CCLR (Table 1).21 These markers were validated on the basis of genotyping additional closely located MSATs. The significant trait association of these 3 markers or regions was confirmed on the basis of both genotype and allele frequency or on the basis of allele frequency determined via C2 analyses (Table 2).
Results of genotyping of 90 Newfoundlands for MSATs that were significantly (P < 0.001) associated with CCLR status (on the basis of permutation testing and a false discovery rate of 0.05). Genotyping was performed for 532 MSATs in 53 homozygous CCLR-unaffected and 37 homozygous CCLR-affected dogs; 495 MSATs were informative, and 4 markers (located on 4 chromosomes) were significantly associated with the trait.
| Variable | MSAT | |||
|---|---|---|---|---|
| CPH19 | FH3702 | REN147D07 | FH3750 | |
| No. of alleles | 4 | 4 | 6 | 6 |
| PIC | ||||
| Affected dogs | 0.24 | 0.53 | 0.56 | 0.74 |
| Unaffected dogs | 0.28 | 0.59 | 0.43 | 0.74 |
| Heterozygosity | ||||
| Affected dogs | 0.11 | 0.61 | 0.73 | 0.73 |
| Unaffected dogs | 0.50 | 0.62 | 0.76 | 0.73 |
| Canis familiaris chromosome | 3 | 5 | 13 | 24 |
| Location (Mb)21 | 68.2 | 32.9 | 32.8 | 3.6 |
Results of χ2 analyses to validate the 4 MSATs that were significantly associated with the CCLR trait in the study population of 90 Newfoundlands.
| Canis familaris chromosome (location [Mb])21 | MSAT | CCLR status by allele | CCLR status by genotype | ||
|---|---|---|---|---|---|
| χ2 | Probability value | χ2 | Probability value | ||
| 3 (68.3) | 03_068_CT | 21.292 | < 0.001 | 18.917 | 0.002 |
| 5 (32.7) | 05_032_B_CAAA | 7.571 | 0.023 | 8.274 | 0.142 |
| 5 (32.9) | 05_032F_CT | 4.086 | 0.252 | 5.961 | 0.544 |
| 5 (33.2) | 05_033A_CA | 8.448 | 0.077 | 15.072 | 0.089 |
| 13 (32.6) | 13_032F_CT | 6.792 | 0.147 | 10.639 | 0.301 |
| 13 (32.7) | 13_032I_CA | 6.574 | 0.087 | 10.599 | 0.157 |
| 13 (33.0) | 13_032K_CA | 6.866 | 0.032 | 6.898 | 0.141 |
| 13 (33.1) | 13_033A_CA | 5.011 | 0.025 | 4.844 | 0.089 |
| 24 (3.5) | 24_003B_CA | 2.972 | 0.396 | 8.161 | 0.518 |
| 24 (3.7) | 24_003C_CT | 0.461 | 0.794 | 1.807 | 0.875 |
| 24 (3.8) | 24_003D_CT | 0.032 | 0.858 | 0.748 | 0.688 |
| 24 (3.8) | 24_003D_CA | 12.234 | 0.057 | 14.836 | 0.251 |
Discussion
In the present study, an initial MSAT-based genome scan was performed to identify chromosomal regions that are associated with the CCLR trait in Newfoundlands. It was anticipated that information derived from the study would facilitate identification of a causative or associated mutation for CCLR. By use of a positional candidate gene approach, a chromosomal region that is associated with a disorder is first identified, and then all genes located in that region are identified. The genes are then organized according to their roles (eg, cellular component, molecular function, or physiologic function) on the basis of gene ontology and investigated for an association to the trait. Because the canine genome has recently been sequenced, one can use that information, in conjunction with information regarding the human genome and the comparative map, to choose the best positional candidates to further study for possible association with a trait.
With regard to CCLR in Newfoundlands, we previously predicted a simple recessive mode of inheritance.1 However, in the present study, several chromosomal regions were significantly associated with CCLR status. This may reflect false identification of certain regions, or given the heritability estimate of 0.27 for this trait,1 it is possible that expression of modifying genes may explain trait variability. Validation of the MSATs that were apparently significantly associated with the CCLR trait in Newfoundlands is necessary to determine the true association of the identified markers and identify chromosomal regions for further investigation.
Of the 4 MSATs that were significantly associated with the CCLR trait in the dogs of the present study, 3 were validated on the basis of genotyping additional closely located MSATs. Additional dogs should be genotyped to determine whether the results are repeatable in a similar population of dogs that has been stratified on the basis of age at the time of CCLR diagnosis (ie, dogs that are < 2 years old with rupture of the CCL bilaterally and dogs that are > 8 years old and do not have CCLR).
One potential limitation of the present study was CCLR status determination—classification of a dog was based on an assumption that its phenotype correctly correlated with its genotype. Rupture of the CCL is known to have a variable age of onset and incomplete penetrance,1 making it difficult to characterize a dog as unaffected with CCLR until it is at least 8 years old.22 Given the nature of the present study, it was impossible to include only dogs that were > 8 years old and, even then, older dogs could still be genotypically affected but not express clinical signs because of the incomplete penetrance of the disorder. In addition, it is possible that there are several different causes of CCLR, which may be breed or age specific. In a recent study,23 there was increased incidence of CCLR in young neutered dogs, compared with age-matched control dogs; however, classic investigation of CCLR has concentrated on degenerative changes of the CCL that are associated with aging.24 Therefore, even though dogs were selected for our genome-wide study on the basis of their statistical likelihood of being homozygous CCLR-affected or homozygous CCLR-unaffected individuals, an inaccurate classification of some dogs would decrease the opportunity to identify specific markers that were significantly associated with the trait or falsely increase the number of markers that appeared to be significantly associated with the trait. Ideally, one would rather identify all chromosomal regions potentially associated with the trait and validate them via further testing than erroneously eliminate areas that actually were associated with the trait.
Rupture of the CCL appears to have a genetic basis in Newfoundlands.1 Results of the present MSAT-based genome-wide association study in a population of Newfoundlands suggest that several chromosomal regions are associated with CCLR status. Fine mapping of these chromosomal regions should further narrow the list of potential CCLR candidate genes and allow for eventual identification of predisposed genotypes.
ABBREVIATIONS
| CCL | Cranial cruciate ligament |
| CCLR | Cranial cruciate ligament rupture |
| cM | Centimorgan |
| MSAT | Microsatellite marker |
| PIC | Polymorphic information content |
PCR conditions, UC-Davis Veterinary Genetics Laboratory, Davis, Calif. Available at: www.vgl.ucdavis.edu/dogset/jelp.jsp. Accessed May 11, 2009.
References
- 1.↑
Wilke VL, Conzemius MG, Kinghorn BP, et al. Inheritance of rupture of the cranial cruciate ligament in Newfoundlands. J Am Vet Med Assoc 2006;228:61–64.
- 2.↑
Wilke VL, Robinson DA, Evans RB, et al. Estimate of the annual economic impact of treatment of cranial cruciate ligament injury in dogs in the United States. J Am Vet Med Assoc 2005;227:1604–1607.
- 3.
Vasseur PB, Berry CR. Progression of stifle osteoarthrosis following reconstruction of the cranial cruciate ligament in 21 dogs. J Am Anim Hosp Assoc 1992;28:129–136.
- 4.
Elkins AD, Pechman R, Kearney MT, et al. A retrospective study evaluating the degree of degenerative joint disease in the stifle joint of dogs following surgical repair of anterior cruciate ligament rupture. J Am Anim Hosp Assoc 1991;27:533–540.
- 5.
Chauvet AE, Johnson AJ, Pijanowski GJ, et al. Evaluation of fibular head transposition, lateral fabellar suture, and conservative treatment of cranial cruciate ligament rupture in large dogs: a retrospective study. J Am Anim Hosp Assoc 1996;32:247–255.
- 6.↑
Gordon WJ, Conzemius MG, Riedesel E, et al. The relationship between limb function and radiographic osteoarthrosis in dogs with stifle osteoarthrosis. Vet Surg 2003;32:451–454.
- 7.↑
Conzemius MG, Evans RB, Besancon MF, et al. Effect of surgical technique on limb function after surgery for rupture of the cranial cruciate ligament in dogs. J Am Vet Med Assoc 2005;226:232–236.
- 8.
Doverspike M, Vasseur PB, Harb MF, et al. Contralateral cranial cruciate ligament rupture: incidence in 114 dogs. J Am Anim Hosp Assoc 1993;29:167–170.
- 9.
Smith GK, Torg JS. Fibular head transposition for repair of cruciate-deficient stifle in the dog. J Am Vet Med Assoc 1985;187:375–383.
- 10.↑
Toth AP, Cordasco FA. Anterior cruciate ligament injuries in the female athlete. J Gend Specif Med 2001;4:25–34.
- 11.
Arendt E, Dick R. Knee injury patterns among men and women in collegiate basketball and soccer. NCAA data and review of literature. Am J Sports Med 1995;23:694–701.
- 12.
Gwinn DE, Wilckens JH, McDevitt ER, et al. The relative incidence of anterior cruciate ligament injury in men and women at the United States Naval Academy. Am J Sports Med 2000;28:98–102.
- 13.
Lohmander LS, Ostenberg A, Englund M, et al. High prevalence of knee osteoarthritis, pain, and functional limitations in female soccer players twelve years after anterior cruciate ligament injury. Arthritis Rheum 2004;50:3145–3152.
- 14.↑
Flynn RK, Pedersen CL, Birmingham TB, et al. The familial predisposition toward tearing the anterior cruciate ligament: a case control study. Am J Sports Med 2005;33:23–28.
- 15.↑
Moore KW, Read RA. Rupture of the CCL in dogs—part I. Compend Contin Educ Pract Vet 1996;18:223–234.
- 16.↑
Macrossan PE, Kinghorn BP, Wilke VL, et al. Selective genotyping for determination of a major gene associated with cranial cruciate ligament disease in the Newfoundland dog, in Proceedings. Assoc Adv Anim Breed Genet, 2005;16:346–349.
- 17.↑
Guyon R, Lorentzen TD, Hitte C, et al. A 1Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A 2003;100:5296–5301.
- 18.↑
Todhunter RJ, Bliss SP, Casella G, et al. Genetic structure of susceptibility traits for hip dysplasia and microsatellite informativeness of an outcrossed pedigree. J Hered 2003;94:39–48.
- 19.↑
Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci U S A 2003;100:9440–9445.
- 20.↑
Botstein D, White R, Skolnik M, et al. Construction of a genetic linkage map using restriction fragment length polymorphisms. Am J Hum Genet 1980;32:314–331.
- 21.↑
Ensembl release 54 [database online]. Chromosomal locations and primer sequences. Hinxton, England: Ensembl Project. Available at: www.ensembl.org/index.html. Accessed May 11, 2009.
- 22.↑
Reif U, Probst CW. Comparison of tibial plateau angles in normal and cranial cruciate deficient stifles of Labrador retrievers. Vet Surg 2003;32:385–389.
- 23.↑
Duval JM, Budsberg SC, Flo GL, et al. Breed, sex, and body weight as risk factors for rupture of the cranial cruciate ligament in young dogs. J Am Vet Med Assoc 1999;215:811–814.
- 24.↑
Vasseur PB, Pool RR, Arnoczky SP, et al. Correlative biomechanical and histologic study of the cranial cruciate ligament in dogs. Am J Vet Res 1985;46:1842–1854.
