58 Etiopathogenesis of Cruciate Ligament Rupture
CR at 0.28 in a prospective study of 414 Boxer
litters. Both of these studies used restricted
maximum likelihood (REML) analysis for cal-
culating heritability estimates. In a study of 237
Labrador Retrievers, the present authors’ group
employed a Bayesian method (Perez & de los ́
Campos 2014) to estimate narrow-sense heri-
tability from single nucleotide polymorphism
(SNP) markers, as well as from pedigrees. Both
methods gave similar results, with a heritabil-
ity of CR in Labrador Retrievers estimated at
0.49 and 0.48, respectively (Bakeret al. 2017).
It is unclear whether the higher heritability
indicates that CR is truly more heritable in the
Labrador Retriever, or if it is a reflection of the
Bayesian method used.
Taken together, these results suggest that
canine CR is moderately heritable within the
range of 0.3–0.5. Heritability estimates of this
level indicate that a reasonable reduction can be
expected in overall prevalence of the condition
through genetic screening applied to a selective
breeding program (Nielenet al. 2001). While it
has previously been suggested that CR has an
autosomal recessive mode of inheritance (Wilke
et al. 2006), recent research suggests that CR is
a complex polygenic trait (Bairdet al. 2014a,b;
Bakeret al. 2017).
Complex trait genetics
Complex traits or diseases are simply defined
as phenotypic traits that are determined by
both genetic and non-genetic (environmen-
tal) factors. The genetic factors that influence
complex traits are typically composed of many
small-effect mutations and fewer large-effect
mutations that contribute predominantly addi-
tively to phenotypic variation (Robinsonet al.
2014). These mutations are expected to be
large in number, and so their contribution to
variance on a population level is quite small
(Robinsonet al. 2014). The relatively small con-
tribution of individual mutations makes all but
the largest effect mutations difficult to detect
without very large sample sizes. Complete dis-
covery of genetic variants contributing to CR
(and other complex traits) will likely require
a combination of complementary analyses
including candidate gene analyses, large-scale
genome-wide association studies (GWAS), and
next-generation sequencing.Candidate gene analyses
For many, the candidate gene approach rep-
resents the first logical step for genetic asso-
ciation testing. Prior knowledge of gene func-
tion is used to choose candidate genes to test
for association, based on their potential role in
disease etiology. Candidate gene studies have
been successful in identifying large-effect muta-
tions, and mutations associated with Mendelian
(monogenic) traits. However, this approach has
faced criticism as current knowledge of gene
functions is limited, and thus disease associa-
tions may be missed when they are not within
‘obvious’ genes (Taboret al. 2002; Bairdet al.
2014a). Additionally, the genetic contribution to
complex traits, such as CR, likely includes many
small-effect variants. Therefore, it is doubtful
that associations identified in a candidate gene
study would explain a large proportion of the
genetic variance for the trait. With that being
said, associations identified in these studies
are still valuable as part of genetic discovery,
andshouldbeconsideredcomplementaryto
genome-wide approaches.
In a study of a population of Newfound-
land dogs, several genes were selected for anal-
ysis based on their known association with
joint hypermobility in cattle or development
of primary arthritis in humans. These included
Cartilage Oligomeric Matrix Protein (COMP),
Matrilin-3(MATN),Collagen Type 9 alpha 1, 2 ,
and 3 (COL9A1,COL9A2,COL9A3),Fibrillin-1
(FBN1), andInterleukin Receptor 4(IL4R).Single
nucleotide polymorphisms (SNPs) were iden-
tified inCOMP,COL9A1,COL9A2,andFBN1.
Based on chi-square analyses, there was no sig-
nificant association between the SNPs and CR
status (affected and unaffected), although some
suggestion of an association was found between
COL9A1(located on CFA 12) and CR affected
status (P=0.10) (Wilkeet al. 2005).
Another study evaluated microsatellite
markers closely located to the genesCollagen
Type 9 alpha 1,2, and 3(COL9A1,COL9A2,
andCOL9A3) in a population of Boxers with a
high incidence of CR (Temwichitret al. 2007).
Data arising from this study suggested that the