Lessons learned from the dog genome

Extensive genetic resources and a high-quality genome sequence position the dog as an important model species for understanding genome evolution, population genetics and genes underlying complex phenotypic traits. Newly developed genomic resources have expanded our understanding of canine evolutionary history and dog origins. Domestication involved genetic contributions from multiple populations of gray wolves probably through backcrossing. More recently, the advent of controlled breeding practices has segregated genetic variability into distinct dog breeds that possess specific phenotypic traits. Consequently, genome-wide association and selective sweep scans now allow the discovery of genes underlying breed-specific characteristics. The dog is finally emerging as a novel resource for studying the genetic basis of complex traits, including behavior.     

Canine epilepsy genetics

There has been much interest in utilizing the dog as a genetic model for common human diseases. Both dogs and humans suffer from naturally occurring epilepsies that share many clinical characteristics. Investigations of inherited human epilepsies have led to the discovery of several mutated genes involved in this disease; however, the vast majority of human epilepsies remain unexplained. Mouse models of epilepsy exist, including single-gene spontaneous and knockout models, but, similar to humans, other, polygenic models have been more difficult to discern. This appears to also be the case in canine epilepsy genetics. There are two forms of canine epilepsies for which gene mutations have been described to date: the progressive myoclonic epilepsies (PMEs) and idiopathic epilepsy (IE). Gene discovery in the PMEs has been more successful, with eight known genes; six of these are orthologous to corresponding human disorders, while two are novel genes that can now be used as candidates for human studies. Only one IE gene has been described in dogs, an LGI2 mutation in Lagotto Romagnolos with a focal, juvenile remitting epilepsy. This gene is also a novel candidate for human remitting childhood epilepsy studies. The majority of studies of dog breeds with IE, however, have either failed to identify any genes or loci of interest, or, as in complex mouse and human IEs, have identified multiple QTLs. There is still tremendous promise in the ongoing canine epilepsy studies, but if canine IEs prove to be as genetically complex as human and murine IEs, then deciphering the bases of these canine epilepsies will continue to be challenging.     

The dog: A powerful model for studying genotype–phenotype relationships

Within the last two years, series of studies have focused on the structure of the dog genome (Canis familiaris) and the characteristics of the dog population as it evolved since being domesticated from wolves about 14,000 years ago. In this review, we explain why the dog is a unique and promising model for determining genotype/phenotype relationships and why it should be easier with this model to identify the genes responsible for many genetic diseases. We also revisit the last ten years of developments in canine molecular genetics that culminated in the release of the entire genome sequence.     

Dog as a mammalian genetic model

Up to recently, studies on dog genetics were rather scare notwithstanding the enormous potential that the canine model can offer in the study of the genotype/phenotype relationship and the analysis of the causes of many genetic diseases, with simple or complex inheritance, that affect dogs but also the human population. This potentiality is essentially due to the natural history of dogs whose domestication from wolves dated back 15,000 years, at least. All modern dogs originated from a limited number of female wolves from Eastern Asia. By applying a combination of selections and strong inbreeding practices, humans have created over 350 breeds, each of them corresponding to a genetic isolate and altogether offering a unique panel of polymorphism never encountered in any other mammals. In this review we summarized what makes dogs an unavoidable model. Contrary to the classical models like the two yeasts, nematode, fish, fly, mouse, or rat mainly used to understand the function of genes, dog with the creation across the centuries of numerous breeds offers a unique opportunity to study the role of their alleles. We report recent data on the construction of genomic maps and on the sequencing program of the dog genome launched by the National Institute of Health (NIH). To take fully advantage of the canine model, we advocate for the systematic construction of a rich canine single nucleotide polymorphisms (SNP) ressource to perform linkage desiquilibrium studies of normal or pathological traits as well as to get insight into the genetic diversity of the canine species.     

Analysis of the unassembled part of the dog genome sequence: chromosomal localization of 115 genes inferred from multispecies comparative genomics

The identification of dog genes and their accurate localization to chromosomes remain a major challenge in the postgenomics era. The 132 annotated canine genes with human orthologs remaining in the unassembled part (chrUnknown) of the dog sequence assembly (CanFam1) are of limited use for candidate gene approaches or comparative mapping studies. We used a two-step comparative analysis to infer a canine chromosomal interval for localization of the chrUn genes. We first constructed a human-dog synteny map, using 14,456 gene-based comparative anchors. We then mapped the 132 chrUn genes onto the reference (human) synteny map and identified the corresponding, orthologous segment on the canine map, based on conserved gene order. Our results show that 110 chrUn genes could be localized to short intervals on 18 dog chromosomes, whereas 22 genes remained assigned to 2 possible intervals. We extended this comparative analysis to multiple species, using the chimpanzee, mouse, and rat genome sequences. This made it possible to narrow down the intervals concerned and to increase the number of canine chrUn genes with an inferred chromosome location to 115. This study demonstrates that dog chromosomal intervals for chrUn genes can be rapidly inferred, using a reference species, and indicates that comparative strategies based on larger numbers of species may be even more effective.     

Annotation of the domestic dog genome sequence: finding the missing genes

There are over 350 genetically distinct breeds of domestic dog that present considerable variation in morphology, physiology, and disease susceptibility. The genome sequence of the domestic dog was assembled and released in 2005, providing an estimated 20,000 protein-coding genes that are a great asset to the scientific community that uses the dog system as a genetic biomedical model and for comparative and evolutionary studies. Although the canine gene set had been predicted using a combination of ab initio methods, homology studies, motif analysis, and similarity-based programs, it still requires a deep annotation of noncoding genes, alternative splicing, pseudogenes, regulatory regions, and gain and loss events. Such analyses could benefit from new sequencing technologies (RNA-Seq) to better exploit the advantages of the canine genetic system in tracking disease genes. Here, we review the catalog of canine protein-coding genes and the search for missing genes, and we propose rationales for an accurate identification of noncoding genes though next-generation sequencing.     

Characterization of variation in the canine suppressor of cytokine signaling-2 (SOCS2) gene

Suppressor of cytokine signaling 2 (SOCS2) is a negative regulator of growth hormone signaling. The deletion of SOCS2 in mice results in a 30-50% increase in post-natal growth. In an effort to identify polymorphisms in the SOCS2 gene that may be associated with body size in dogs, we characterized the canine SOCS2 gene and analyzed its genetic diversity among small and large dog breeds. The study was carried out on a total of 520 dogs from 66 different breeds. Dogs were classified as large or small based on height and weight as determined by their respective American Kennel Club breed standards. The SH2 and SOCS domains of the canine SOCS2 gene were sequenced in 32 dogs from different breeds. Only one non-synonymous sequence variant (DQ415457:g.326G>T) was detected which corresponds to an amino acid change (Asp127Tyr). All samples were genotyped by PCR/RFLP and the allele frequencies were determined for each dog breed. The T allele was distributed primarily among European large dog breeds with a gene frequency ranging from 0.72 to 0.04. The nature of the nucleotide change and the effect on the protein together with the finding of a QTL related to body size in the same CFA15 region by other researchers suggest canine SOCS2 as a potential candidate gene for body size in dogs. Future studies will be needed to clarify the role of the 326G>T polymorphism and its interaction with genes like growth hormone and insulin-like growth factor 1.     

A marker set for construction of a genetic map of the silver fox (Vulpes vulpes)

The silver fox, a variant of the red fox (Vulpes vulpes), is a close relative of the dog (Canis familiaris). Cytogenetic differences and similarities between these species are well understood, but their genomic organizations have not been compared at higher resolution. Differences in their behavior also remain unexplained. Two silver fox strains demonstrating markedly different behavior have been generated at the Institute of Cytology and Genetics of the Russian Academy of Sciences. Foxes selected for tameness are friendly, like domestic dogs, while foxes selected for aggression resist human contact. To refine our understanding of the comparative genomic organization of dogs and foxes, and enable a study of the genetic basis of behavior in these fox strains, we need a meiotic linkage map of the fox. Towards this goal we generated a primary set of fox microsatellite markers. Four hundred canine microsatellites, evenly distributed throughout the canine genome, have been identified that amplify robustly from fox DNA. Polymorphism information content (PIC) values were calculated for a representative subset of these markers and population inbreeding coefficients were determined for tame and aggressive foxes. To begin to identify fox-specific single nucleotide polymorphisms (SNPs) in genes involved in the neurobiology of behavior, fox and dog orthologs of serotonin 5-HT1A and 5-HT1B receptor genes have been cloned. Sequence comparison of these genes from tame and aggressive foxes reveal several SNPs. The close relationship of the fox and dog enables canine genomic tools to be utilized in developing a fox meiotic map and mapping behavioral traits in the fox.     

Leader of the pack: gene mapping in dogs and other model organisms

The domestic dog offers a unique opportunity to explore the genetic basis of disease, morphology and behaviour. We share many diseases with our canine companions, including cancer, diabetes and epilepsy, making the dog an ideal model organism for comparative disease genetics. Using newly developed resources, whole-genome association in dog breeds is proving to be exceptionally powerful. Here, we review the different trait-mapping strategies, some key biological findings emerging from recent studies and the implications for human health. We also discuss the development of similar resources for other vertebrate organisms.     

Dog keeping in Taiwan: its contribution to the problem of free-roaming dogs

This study conducted a quantitative ethnographic analysis of the influence of demographic factors and early experience (childhood exposure to dogs) on Taiwanese dog-keeping practices and behavior. A telephone survey of a randomly selected sample of 2,001 Taiwan residents determined their dog ownership histories, current patterns of dog ownership and disposal, and other dog-related activities. The results suggest that low rates of neutering, easy availability of low- or no-cost puppies, a tendency to allow owned dogs free access to the outdoors, unrealistic expectations of dog ownership, canine behavioral problems, and religious and cultural taboos against euthanasia and shelter relinquishment have contributed to the recent increase in the numbers of free-roaming dogs in Taiwan. Logistic regression analyses determined that a relatively small number of demographic and experiential variables predicted dog ownership and disposal patterns. The most important of these was childhood experience of living with household dogs. In light of these findings, future efforts to reduce the stray dog problem should focus on enforcing registration fees, particularly for unsterilized animals; low-cost neutering schemes; and educational programs designed to promote neutering, improve knowledge of canine behavior and behavior problems, and develop more realistic expectations and attitudes toward dog ownership. Marked Taiwanese resistance to canine euthanasia and shelter relinquishment suggests a need for alternative methods of managing the existing free-roaming dog population.     

The canine ERBB2 gene maps to a chromosome region frequently affected by aberrations in tumors of the dog (Canis familiaris)

The dog offers an increasingly important model for several human diseases, including cancer. Accordingly, the results of canine gene mapping studies will be of considerable significance. Herein, we have addressed the mapping of the canine gene ERBB2 (alias HER2, NEU). ERBB2 is a protooncogene encoding a tyrosine kinase receptor protein, the overexpression of which correlates with a more rapid progression and a worse prognosis in breast cancer. In addition, it apparently plays a role in the development of other tumors as well. By fluorescence in situ hybridization (FISH), we have mapped the canine ERBB2 to 1q13.1. Cytogenetic studies of canine tumors revealed that this region is very often affected by clonal chromosome aberrations in tumors of the dog.     

Evaluation of candidate genes in the absence of positional information: a poor bet on a blind dog!

More than 350 inherited diseases have been reported in dogs and at least 50% of them have human counterparts. To remove the diseases from dog breeds and to identify canine models for human diseases, it is necessary to find the mutations underlying them. To this end, two methods have been used: the functional candidate gene approach and linkage analysis. Here we present an evaluation of these in canine retinal diseases, which have been the subject of a large number of molecular genetic studies, and we show the contrasting outcomes of these approaches when dealing with genetically heterogeneous diseases. The candidate gene approach has led to 377 published results with 23 genes. Most of the results (66.6%) excluded the presence of a mutation in a gene or its coding region, while only 3.4% of the results identified the mutation causing the disease. On the other hand, five linkage analysis studies have been done on retinal diseases, resulting in three identified mutations and two mapped disease loci. Mapping studies have relied on dog research colonies. If this favorable application of linkage analysis can be extended to dogs in the pet population, success in identifying canine mutations could increase, with advantages to veterinary and human medicine.     

Canine population structure: assessment and impact of intra-breed stratification on SNP-based association studies

Background

In canine genetics, the impact of population structure on whole genome association studies is typically addressed by sampling approximately equal numbers of cases and controls from dogs of a single breed, usually from the same country or geographic area. However one way to increase the power of genetic studies is to sample individuals of the same breed but from different geographic areas, with the expectation that independent meiotic events will have shortened the presumed ancestral haplotype around the mutation differently. Little is known, however, about genetic variation among dogs of the same breed collected from different geographic regions.

Methodology/Principal Findings

In this report, we address the magnitude and impact of genetic diversity among common breeds sampled in the U.S. and Europe. The breeds selected, including the Rottweiler, Bernese mountain dog, flat-coated retriever, and golden retriever, share susceptibility to a class of soft tissue cancers typified by malignant histiocytosis in the Bernese mountain dog. We genotyped 722 SNPs at four unlinked loci (between 95 and 271 per locus) on canine chromosome 1 (CFA1). We showed that each population is characterized by distinct genetic diversity that can be correlated with breed history. When the breed studied has a reduced intra-breed diversity, the combination of dogs from international locations does not increase the rate of false positives and potentially increases the power of association studies. However, over-sampling cases from one geographic location is more likely to lead to false positive results in breeds with significant genetic diversity.

Conclusions

These data provide new guidelines for association studies using purebred dogs that take into account population structure.     

Genome analysis of the domestic dog (Korean Jindo) by massively parallel sequencing

Although pioneering sequencing projects have shed light on the boxer and poodle genomes, a number of challenges need to be met before the sequencing and annotation of the dog genome can be considered complete. Here, we present the DNA sequence of the Jindo dog genome, sequenced to 45-fold average coverage using Illumina massively parallel sequencing technology. A comparison of the sequence to the reference boxer genome led to the identification of 4 675 437 single nucleotide polymorphisms (SNPs, including 3 346 058 novel SNPs), 71 642 indels and 8131 structural variations. Of these, 339 non-synonymous SNPs and 3 indels are located within coding sequences (CDS). In particular, 3 non-synonymous SNPs and a 26-bp deletion occur in the TCOF1 locus, implying that the difference observed in cranial facial morphology between Jindo and boxer dogs might be influenced by those variations. Through the annotation of the Jindo olfactory receptor gene family, we found 2 unique olfactory receptor genes and 236 olfactory receptor genes harbouring non-synonymous homozygous SNPs that are likely to affect smelling capability. In addition, we determined the DNA sequence of the Jindo dog mitochondrial genome and identified Jindo dog-specific mtDNA genotypes. This Jindo genome data upgrade our understanding of dog genomic architecture and will be a very valuable resource for investigating not only dog genetics and genomics but also human and dog disease genetics and comparative genomics.     

A novel retinal degeneration locus identified by linkage and comparative mapping of canine early retinal degeneration.

Early retinal degeneration (erd) is an early onset progressive retinal atrophy, a hereditary canine retinal disease phenotypically similar to human retinitis pigmentosa (RP). In previous efforts to identify the erd locus, canine homologs of genes causally associated with RP in humans, such as opsin (RHO), the beta-subunit gene for cyclic GMP phosphodiesterase (PDE6B), and RDS/peripherin, were excluded. A genome-wide screen was undertaken on canine families segregating the erd disease. Analysis of over 150 canine-specific markers has localized erd to a single linkage group comprising two previously identified canine linkage groups, 20 and 26, corresponding to canine radiation hybrid groups RH.34-a and RH.40-a. Multipoint analysis places erd in the interval between marker FH2289 (distance 23.6 cM) and FH2407 (5.9 cM) with a lod score of 12.23. Although the erd linkage group has not been assigned to an identified canine chromosome, conserved synteny of this linkage group with human 12p13-q13 suggests several candidates for erd and identifies a novel retinal degeneration locus. The rapid progress now occurring in canine genetics will expedite identification of the genes and molecular mechanisms underlying the inherited traits and diseases that make the dog a unique asset for study of mammalian traits.     

Multiple structural genes for the α chain of canine (Canis familiaris) hemoglobin

There are two forms of the alpha chain of canine hemoglobin differing only at residue 130. One form (taualpha) contains threonine at this position, the other (Aalpha) contains alanine. Studies of two Labrador dog families strongly support the existance of multiple alpha-chain structural genes as the basis of the alpha-chain heterogeneity. There must be at least one gene locus for Talpha and one for Aalpha; the exact number of loci has not been determined. Two other dog breeds, the Basenji and the Beagle, also have both Talpha and Aalpha chains.     

Structure and variation of three canine genes involved in serotonin binding and transport: the serotonin receptor 1A gene (htr1A), serotonin receptor 2A gene (htr2A), and serotonin transporter gene (slc6A4)

Aggressive behavior is the most frequently encountered behavioral problem in dogs. Abnormalities in brain serotonin metabolism have been described in aggressive dogs. We studied canine serotonergic genes to investigate genetic factors underlying canine aggression. Here, we describe the characterization of three genes of the canine serotonergic system: the serotonin receptor 1A and 2A gene (htr1A and htr2A) and the serotonin transporter gene (slc6A4). We isolated canine bacterial artificial chromosome clones containing these genes and designed oligonucleotides for genomic sequencing of coding regions and intron-exon boundaries. Golden retrievers were analyzed for DNA sequence variations. We found two nonsynonymous single nucleotide polymorphisms (SNPs) in the coding sequence of htr1A; one SNP close to a splice site in htr2A; and two SNPs in slc6A4, one in the coding sequence and one close to a splice site. In addition, we identified a polymorphic microsatellite marker for each gene. Htr1A is a strong candidate for involvement in the domestication of the dog. We genotyped the htr1A SNPs in 41 dogs of seven breeds with diverse behavioral characteristics. At least three SNP haplotypes were found. Our results do not support involvement of the gene in domestication.     

Evolution of the feline-subgroup parvoviruses and the control of canine host range in vivo.

A related group of parvoviruses infects members of many different carnivore families. Some of those viruses differ in host range or antigenic properties, but the true relationships are poorly understood. We examined 24 VP1/VP2 and 8 NS1 gene sequences from various parvovirus isolates to determine the phylogenetic relationships between viruses isolated from cats, dogs, Asiatic raccoon dogs, mink, raccoons, and foxes. There were about 1.3% pairwise sequence differences between the VP1/VP2 genes of viruses collected up to four decades apart. Viruses from cats, mink, foxes, and raccoons were not distinguished from each other phylogenetically, but the canine or Asiatic raccoon dog isolates formed a distinct clade. Characteristic antigenic, tissue culture host range, and other properties of the canine isolates have previously been shown to be determined by differences in the VP1/VP2 gene, and we show here that there are at least 10 nucleotide sequence differences which distinguish all canine isolates from any other virus. The VP1/VP2 gene sequences grouped roughly according to the time of virus isolation, and there were similar rates of sequence divergence among the canine isolates and those from the other species. A smaller number of differences were present in the NS1 gene sequences, but a similar phylogeny was revealed. Inoculation of mutants of a feline virus isolate into dogs showed that three or four CPV-specific differences in the VP1/VP2 gene controlled the in vivo canine host range.