ature 438, 745-746 (8 December 2005) | doi:10.1038/438745a
Genomics: The dog has its day
Hans Ellegren1
Top of page
Abstract
Domestication and selective breeding have transformed wolves into the diversity of dogs we see today. The sequence of the genome of one breed adds to our understanding of mammalian biology and genome evolution.
Dogs have a special place in our society. Man's best friend is not just a valuable hunting partner, guard and herd manager — most of the world's estimated 400 million dogs1 are pets. Dogs were the first animals to be domesticated (at least 15,000 years ago)2, 3, 4. They all originate from a single and relatively homogeneous species — the wolf — but modern breeds display an extraordinary diversity of traits (or phenotypes). The hundreds of years of careful inbreeding to produce the many kinds of dog have delivered a geneticist's dream model of human genetic disease (Box 1). But to unlock the full potential of this model, we need to understand the genetic basis for the unprecedented diversity and how it has evolved5. The high-quality draft sequence of the dog genome described on page 803 of this issue6 is a good starting-point for that research.
C. COLLINS/CORBIS
Lindblad-Toh and colleagues6 invited breed clubs and veterinary schools to suggest an individual dog suitable for genome sequencing. The idea was to identify a highly inbred dog; this was based on the thinking that the animal's genetic homogeneity would simplify the gigantic jigsaw puzzle of assembling millions of sequence reads into a genome sequence. After testing certain genetic markers in a host of dogs, the sequencers settled on a female boxer called Tasha (so there is no Y chromosome in the current sequence).
The assembled sequence from Tasha's DNA spans 2.4109 base pairs (Gbp), which corresponds to an estimated 99% coverage of the canine genome (excluding highly repetitive regions). So, although dogs have 39 pairs of chromosomes (compared with 23 pairs in humans), their genome contains almost 0.5 Gbp less DNA than ours. The difference can be explained mainly by the existence of fewer repetitive elements in the dog lineage, and to some extent by deletion of sequences that were present in an early common mammalian ancestor. Dogs seem to have fewer genes than humans, but the actual numbers might be a bit out for both genomes because identifying genes across whole genomes continues to be a difficult task7.
The current work is not the first canine genome project. Sequencing of a male poodle (at a lower sequence coverage) recently characterized about 75% of its genome, although with much of the assembled sequence interleaved with gaps of undefined length8. However, by comparing it with the boxer genome, the poodle sequence is a useful tool for identifying genetic variants — single nucleotide polymorphisms (SNPs) — in dog populations. Augmented with SNPs identified in the boxer and by limited sequencing of many other dog breeds, 2.5 million variable sites have now been discovered6. Comparisons of the different breeds show that there is an average of around 1 SNP per 1,000 base pairs — a similar value to that in human populations.
The SNP data give several evolutionary insights. For instance, analysis of DNA from mitochondria (cellular organelles that have their own genome) has suggested that domestication is associated with a narrow genetic bottleneck where only a few wild ancestors contributed to the domestic gene pool9. However, the large genetic diversity seen among dogs is at odds with this hypothesis, and work on other domestic animals shows that they, too, have high levels of variability in their nuclear genes. This implies that, in many cases, back-crosses with wild relatives introduced additional genetic diversity into domesticated animals well after domestication began10. The genetic traces of such interbreeding may not be picked up by studies of mitochondrial DNA if the back-crossing occurred mainly between wild males and domestic females, because mitochondrial DNA is inherited only from mothers11.
The physical positions of the genetic variations within and among breeds create patterns in the genome that give a more detailed perspective on domestication and breed formation. Within breeds, most chromosomes are mosaics of alternating regions of homogeneous sequences — reflecting the recent common ancestry shared by individual dogs of the same breed — and heterogeneous sequences6, 12, 13. Mathematical simulations can be used to model the way in which population history might be expected to affect genetic diversity and its structural patterns. The model that best fits the observed pattern of SNPs is one that assumes an ancient bottleneck some 9,000 generations ago (domestication), followed by breed-specific bottlenecks 30–90 generations ago (breed formation). However, if repeated back-crossing has occurred, this model would have to be revised.
The dog adds to a growing list of vertebrate species that have had their genome sequenced14. A comparative analysis of the human, mouse and dog by Lindblad-Toh et al.6 showed that about 5% of the human genome is being maintained by natural selection — suggesting that it has some essential function. Almost all of this sequence is also present in the dog genome. Only 1–2% of the genomes encodes proteins, so there would seem to be an additional common set (about 3%) of functional elements in mammalian non-coding DNA. These common sequences may constitute, for example, regulatory elements, structural elements or RNA genes. Notably, such regions are found mostly within the 0.8 Gbp of ancestral sequence common to human, mouse and dog.
With the dog genome sequence available, it will be exciting to follow the forthcoming search for associations between certain phenotypes in different breeds and the genes responsible for them (Box 1). It will now be possible, using various genomic approaches, to map breed-specific traits related to morphology, physiology and behaviour15, which will provide insight into key features of mammalian biology and disease.
Top of page
References
Coppinger, R. & Coppinger, L. Dogs: A Startling New Understanding of Canine Origin, Behaviour & Evolution (Scribner, New York, 2001).
Vilà, C. et al. Science 276, 1687–1689 (1997). | Article | PubMed | ISI | ChemPort |
Savolainen, P. et al. Science 298, 1610–1613 (2002). | Article | PubMed | ISI | ChemPort |
Leonard, J. A. et al. Science 298, 1613–1616 (2002). | Article | PubMed | ISI | ChemPort |
Sutter, N. B. & Ostrander, E. Nature Rev. Genet. 5, 900–910 (2004). | Article |
Lindblad-Toh, K. et al. Nature 438, 803–819 (2005). | Article |
International Human Genome Sequencing Consortium Nature 431, 931–945 (2004). | Article |
Kirkness, E. F. et al. Science 301, 1898–1903 (2003). | Article | PubMed | ISI |
Bruford, M. et al. Nature Rev. Genet. 4, 900–910 (2003). | Article |
Vilà, C. et al. Trends Genet. 21, 214–218 (2005). | Article | PubMed | ISI | ChemPort |
Götherström, A. et al. Proc. R. Soc. Lond. B 272, 2337–2344 (2005).
Sutter, N. B. et al. Genome Res. 14, 2388–2396 (2004). | Article | PubMed | ISI | ChemPort |
Parker, H. G. et al. Science 304, 1160–1164 (2004). | Article | PubMed | ISI | ChemPort |
http://www.ncbi.nih.gov/Genomes
Pollinger, J. P. et al. Genome Res. (in the press).
Lin, L. et al. Cell 98, 365–376 (1999). | Article | PubMed | ISI | ChemPort |
Lohi, H. et al. Science 307, 81 (2005). | Article | PubMed | ISI | ChemPort |
Genomics: The dog has its day
Hans Ellegren1
Top of page
Abstract
Domestication and selective breeding have transformed wolves into the diversity of dogs we see today. The sequence of the genome of one breed adds to our understanding of mammalian biology and genome evolution.
Dogs have a special place in our society. Man's best friend is not just a valuable hunting partner, guard and herd manager — most of the world's estimated 400 million dogs1 are pets. Dogs were the first animals to be domesticated (at least 15,000 years ago)2, 3, 4. They all originate from a single and relatively homogeneous species — the wolf — but modern breeds display an extraordinary diversity of traits (or phenotypes). The hundreds of years of careful inbreeding to produce the many kinds of dog have delivered a geneticist's dream model of human genetic disease (Box 1). But to unlock the full potential of this model, we need to understand the genetic basis for the unprecedented diversity and how it has evolved5. The high-quality draft sequence of the dog genome described on page 803 of this issue6 is a good starting-point for that research.
C. COLLINS/CORBIS
Lindblad-Toh and colleagues6 invited breed clubs and veterinary schools to suggest an individual dog suitable for genome sequencing. The idea was to identify a highly inbred dog; this was based on the thinking that the animal's genetic homogeneity would simplify the gigantic jigsaw puzzle of assembling millions of sequence reads into a genome sequence. After testing certain genetic markers in a host of dogs, the sequencers settled on a female boxer called Tasha (so there is no Y chromosome in the current sequence).
The assembled sequence from Tasha's DNA spans 2.4109 base pairs (Gbp), which corresponds to an estimated 99% coverage of the canine genome (excluding highly repetitive regions). So, although dogs have 39 pairs of chromosomes (compared with 23 pairs in humans), their genome contains almost 0.5 Gbp less DNA than ours. The difference can be explained mainly by the existence of fewer repetitive elements in the dog lineage, and to some extent by deletion of sequences that were present in an early common mammalian ancestor. Dogs seem to have fewer genes than humans, but the actual numbers might be a bit out for both genomes because identifying genes across whole genomes continues to be a difficult task7.
The current work is not the first canine genome project. Sequencing of a male poodle (at a lower sequence coverage) recently characterized about 75% of its genome, although with much of the assembled sequence interleaved with gaps of undefined length8. However, by comparing it with the boxer genome, the poodle sequence is a useful tool for identifying genetic variants — single nucleotide polymorphisms (SNPs) — in dog populations. Augmented with SNPs identified in the boxer and by limited sequencing of many other dog breeds, 2.5 million variable sites have now been discovered6. Comparisons of the different breeds show that there is an average of around 1 SNP per 1,000 base pairs — a similar value to that in human populations.
The SNP data give several evolutionary insights. For instance, analysis of DNA from mitochondria (cellular organelles that have their own genome) has suggested that domestication is associated with a narrow genetic bottleneck where only a few wild ancestors contributed to the domestic gene pool9. However, the large genetic diversity seen among dogs is at odds with this hypothesis, and work on other domestic animals shows that they, too, have high levels of variability in their nuclear genes. This implies that, in many cases, back-crosses with wild relatives introduced additional genetic diversity into domesticated animals well after domestication began10. The genetic traces of such interbreeding may not be picked up by studies of mitochondrial DNA if the back-crossing occurred mainly between wild males and domestic females, because mitochondrial DNA is inherited only from mothers11.
The physical positions of the genetic variations within and among breeds create patterns in the genome that give a more detailed perspective on domestication and breed formation. Within breeds, most chromosomes are mosaics of alternating regions of homogeneous sequences — reflecting the recent common ancestry shared by individual dogs of the same breed — and heterogeneous sequences6, 12, 13. Mathematical simulations can be used to model the way in which population history might be expected to affect genetic diversity and its structural patterns. The model that best fits the observed pattern of SNPs is one that assumes an ancient bottleneck some 9,000 generations ago (domestication), followed by breed-specific bottlenecks 30–90 generations ago (breed formation). However, if repeated back-crossing has occurred, this model would have to be revised.
The dog adds to a growing list of vertebrate species that have had their genome sequenced14. A comparative analysis of the human, mouse and dog by Lindblad-Toh et al.6 showed that about 5% of the human genome is being maintained by natural selection — suggesting that it has some essential function. Almost all of this sequence is also present in the dog genome. Only 1–2% of the genomes encodes proteins, so there would seem to be an additional common set (about 3%) of functional elements in mammalian non-coding DNA. These common sequences may constitute, for example, regulatory elements, structural elements or RNA genes. Notably, such regions are found mostly within the 0.8 Gbp of ancestral sequence common to human, mouse and dog.
With the dog genome sequence available, it will be exciting to follow the forthcoming search for associations between certain phenotypes in different breeds and the genes responsible for them (Box 1). It will now be possible, using various genomic approaches, to map breed-specific traits related to morphology, physiology and behaviour15, which will provide insight into key features of mammalian biology and disease.
Top of page
References
Coppinger, R. & Coppinger, L. Dogs: A Startling New Understanding of Canine Origin, Behaviour & Evolution (Scribner, New York, 2001).
Vilà, C. et al. Science 276, 1687–1689 (1997). | Article | PubMed | ISI | ChemPort |
Savolainen, P. et al. Science 298, 1610–1613 (2002). | Article | PubMed | ISI | ChemPort |
Leonard, J. A. et al. Science 298, 1613–1616 (2002). | Article | PubMed | ISI | ChemPort |
Sutter, N. B. & Ostrander, E. Nature Rev. Genet. 5, 900–910 (2004). | Article |
Lindblad-Toh, K. et al. Nature 438, 803–819 (2005). | Article |
International Human Genome Sequencing Consortium Nature 431, 931–945 (2004). | Article |
Kirkness, E. F. et al. Science 301, 1898–1903 (2003). | Article | PubMed | ISI |
Bruford, M. et al. Nature Rev. Genet. 4, 900–910 (2003). | Article |
Vilà, C. et al. Trends Genet. 21, 214–218 (2005). | Article | PubMed | ISI | ChemPort |
Götherström, A. et al. Proc. R. Soc. Lond. B 272, 2337–2344 (2005).
Sutter, N. B. et al. Genome Res. 14, 2388–2396 (2004). | Article | PubMed | ISI | ChemPort |
Parker, H. G. et al. Science 304, 1160–1164 (2004). | Article | PubMed | ISI | ChemPort |
http://www.ncbi.nih.gov/Genomes
Pollinger, J. P. et al. Genome Res. (in the press).
Lin, L. et al. Cell 98, 365–376 (1999). | Article | PubMed | ISI | ChemPort |
Lohi, H. et al. Science 307, 81 (2005). | Article | PubMed | ISI | ChemPort |
