Bi156 Lecture 1/13/12 Dog Genetics
The radiation of the family Canidae occurred about 100 million years ago. Dogs are most closely related to wolves, from which they diverged through domestication about 10-20,000 years ago.
Advantages of studying evolution of complex behavior in dogs vs. apes Unlike humans and apes, dogs have a living ancestor to which they can be directly compared. Also, dogs and humans exhibit convergent behaviors resulting from socialization together. We understand dog behavior in much more detail than ape behavior since we live with dogs.
Tree of dog breeds There are 157 recognized dog breeds. Most were only created during the last 200 years, so they are very closely related. Some Asian dogs split off earlier before the selection for European breeds.
Haplotypes Ancestral mutation Ancestral chromosome haplotype If a mutation is in linkage disequilibrium with DNA markers surrounding it (purple patch) derived from the original chromosome, these markers will be found associated with the trait caused by the mutation at a higher than expected frequency. The region right next to the mutation may have never been separated from it by recombination; but this region may be so small that it contains no polymorphic markers.
Linkage disequilibrium Alleles in two segments of the genome are in linkage disequilibrium with one another when their frequency together is higher than the frequencies expected based on random association. LD can be used to map loci without having to first find the causative variant (the actual trait allele). Requires analysis of many single nucleotide polymorphisms (SNPs) in affecteds and controls. If one had a SNP for each LD region in the genome, one could use only this SNP subset for analysis. But: the extent of LD varies greatly within the human (and other) genomes. HLA region: 1-2 Mb Lipoprotein lipase gene: a few kb
Linkage disequilibrium exists over large distances in dogs Because dog populations are derived from a few recent founders, alleles of genes are inherited together over a much larger distance than in humans. This means that fewer SNPs need to be used to map a trait in dogs than in humans. (label on x axis of human graph is incorrect; should be kb, not Mb)
Haplotype structure in dogs
Two-stage strategy for gene mapping Step 1: use long LD blocks within a breed to rapidly map trait to a ~1 Mb region. Step 2: find shared haplotypes within this region between two breeds that exhibit the same phenotype (short LD). This is often successful because the mutations that produce traits predate the split between breeds, and thus more than one breed often exhibits the same trait.
An example of the two-stage strategy White spotting mapped to 800 kb in boxers. Then, 800 kb region narrowed down in boxers and terriers. One or two of the changes near the MITF gene may cause the phenotype.
Genes involved in size determination Genes involved in morphology have been successfully identified in dogs, because dogs with a particular characteristic usually seem to share a common ancestry. Small size in dogs is determined by variation in the IGF-1 gene, which encodes a soluble growth factor.
Short vs. long legs Short legs (chondrodysplasia), which is related to dwarfism in humans, maps to a single chromosomal locus. The responsible gene is a retrogene copy of FGF-4.
Short dogs all have a new copy of the FGF4 gene, which is flanked by short tandem repeats. Presumably, it is expressed in a different pattern than the normal FGF4 gene, possibly due to the flanking sequence in which it landed.
Coat characteristics are determined by three genes Dogs can be defined as short or longhaired, straight or curly-haired, and with or without furnishings. These traits map to three loci.
FGF5 and keratin affect hair length and curliness
Explaining all hair traits with 3 genes The relevant genes are FGF5, R-spondin, and a keratin. The keratin differentiates straight from curly and wire; R-spondin affects wire and furnishings, and FGF5 differentiates short from long.
Why is there so much morphological variation in dogs? Many of the remarkable variations in size, appearance, hair, etc., probably arise from genetic variants that are present in the wolf genome as well, even though wolves don t exhibit any of this variation in appearance. The split from wolves is so recent that it seems unlikely that many new mutations could have occurred. Perhaps visible variation appears now because breeds are often homozygous for recessive mutations (however, many dog mutations are dominant). Also, odd characteristics were picked out by humans and isolated by inbreeding. But, why do other domesticated animals that are bred by humans not exhibit the same degree of variation? Possible explanations include increased rates of replication slippage, leading to changes in microsatellites, and active transposable elements (SINEs). Identified mutations causing traits are often SINEs, duplications, etc.
Contrasts between genetic control of traits in dogs and humans In humans, we have been mostly unsuccessful in finding the loci that control traits (this is called missing heritability). In dogs, however, alterations in size, leg length, etc. are caused by single gene changes, and we ve been very successful in finding these. These differences are probably due to the outbred nature of the human population relative to dog breeds.
Dogs exhibit complex behaviors
Information on video section of lecture: Go onto YouTube, search for Secret Life of the Dog (BBC program). It is divided into 6 parts. The section on fox domestication is in part 4, starting at about 8 min (2 min section); then the first ~7 min. of part 5. Material on dog intelligence is in part 3, starting at about 8 min., then the first ~3 min. of part 4.
Domestication in many animals involves common morphological traits
Fox genome The fox genome is being analyzed in a project at Cornell. It is much like the dog but has many rearrangements This should allow mapping of traits in the tame vs aggressive foxes
Reviews *Canine Morphology: Hunting for Genes and Tracking Mutations. A. L. Shearin and E. A Ostrander, PLoS Biology 8, e1000310 (2010). *Leader of the pack: gene mapping in dogs and other model organisms. E. K. Karlsson and K. Lindblad-Toh. Nature Rev. Genetics 9, 713-725 (2008). *Animal evolution during domestication: the domesticated fox as a model. Trut, L., et al. (2009). Bioessays 31, 349-360. Science News and Views, available on Bi156 website. *Required reading
Papers for student presentations LGI2 Truncation Causes a Remitting Focal Epilepsy in Dogs. E. H. Seppala et al., PLoS Genetics 7, e1002194 (2011). Differential Sensitivity to Human Communication in Dogs, Wolves, and Human Infants. J. Topál, et al. Science 325, 1269 (2009).