Genetics In Labrador Retrievers Article

CLEVER ANCESTORS, HUMAN AND CANINE

Fifteen thousand years ago dog domestication was probably not so concentrated or focused an endeavor as the farm-fox experiment, but humans had to have been ready for it. The people who domesticated dogs must have had a society in which a long-term project of this kind could be supported and passed on to ensuing generations, even without the advantage of written records. They must have had some understanding that traits are inherited and must have been able to keep wild wolves from breeding with their proto-dogs. They also must have placed great value in the dog to invest their time and resources into its breeding.

Possibly they had some partially tame wolves to work with, such as self-selected wolves that found an ecological niche as scavengers on the outskirts of settlements. According to Dr. Raymond Coppinger (Hampshire College, Massachusetts), such wolves may have been less fearful of humans than most wild wolves, and this trait must have been inheritable, making these wolves more amenable to domestication. The wolves hypothetically separated into two populations, the village-oriented scavengers and the packs of hunters. The next steps have not been defined, but selective pressure must have been present to sustain the divergence of these populations. At some point, conditions under which social bonds could form between pups and humans, perhaps while the pups were still of nursing age, must have developed. That is an uncomfortable number of "must-haves," but research has not yet shed further light on this aspect of our remarkable pre-history.

The wolf-as-camp-follower hypothesis is unfortunately difficult to test and is supported primarily by observations that in some parts of the world, dogs live as locally adapted village scavengers. However, it may be of value to study other urbanized canids, such as urban coyotes that have become increasingly present as scavengers in Los Angeles and Vancouver for two or three decades. Anecdotal observations indicate that these animals are less fearful of humans than are coyotes remaining in the wild. Is this an example of individual adaptation, or are populations of animals undergoing genetic selection for tolerance to humans? Do they pass these traits on to their offspring? I have been unable to locate any research in this area. What an intriguing project for an evolutionary genomicist, an animal behaviorist, and an ecologist.

The fact remains that our human forebears and dogs' wolf forebears were both resourceful organisms who established a commensal relationship that has endured over millennia. The relationship is still evolving, and with recent revelations from canine genomics, it is evolving in unexpected ways. Not content with the basic dog, humans took to artificial selection for creating races or breeds of dogs with specific appearances, behaviors, and dispositions. More than four hundred breeds, one hundred and fifty of which are recognized by the American Kennel Club, have been created by generations of tightly regulated breeding to a breed standard, mostly in the last one hundred to five hundred years. No other animal seems to have been subject to as much intentional selection as the dog. While purebred dogs were initially bred for specific working behaviors, they are now more commonly bred for desired appearance and temperament, and the job title held by most purebred dogs is "companion and pet." However, their service to humans may expand after the publication in May 2004 of the first detailed genetic comparison of purebred dogs. A serendipitous offshoot of creating dog breeds seems to be new models for human hereditary diseases.

ADDING TO THE DOG'S RÉSUMÉ

Dr. Elaine A. Ostrander and Dr. Leonid Kruglyak (Science, Vol. 304, p. 1160; 2004) at the Fred Hutchinson Cancer Research Center in Seattle found that by comparing genetic information among eighty-five dog breeds, they could discover the genetic and evolutionary relationships among breeds. Breeds are revealed to have the same genes but with variants in DNA sequence that are breed-specific. In this respect, dog breeds are even more distinct than are human populations that have evolved on different continents. Surprisingly, the dogs with the closest genetic relationship to ancestral wolves are several ancient Asian breeds, including the unlikely Sharpei, a medium-sized, broad-muzzled dog that looks like it is wearing a much larger dog's skin. In contrast, the Pharoah Hound and Ibizan Hound, which resemble pictures on ancient Egyptian tomb walls, turn out to be modern recreations of the body type from other breeds. Dr. Ostrander's larger message, however, is the usefulness of purebred dogs in the study of hereditary diseases. Various breeds have distinct hereditary health problems. Breeding records have been kept on many breeds for decades. Therefore, purebred-dog populations may be useful for identifying specific genes associated with diseases that affect both dogs and humans, such as cancer, epilepsy, and heart disease.

Just when a purebred-dog's life was getting easy, along comes a new job responsibility. Most terriers no longer have to chase moles and rats, nor do most Labrador retrievers have to fetch fallen game birds. Few if any poodles have to serve their original job function, which, like Labradors, was to retrieve water birds for hunters. Now purebred dogs may have a new role: they are the living repositories of genetic goldmines.

Dr. Ostrander's work with dogs makes more plausible the long-standing belief by some scientists that dogs are good animal models for hereditary human diseases. The more convenient, commonly used, and fundable (through federal-granting agencies) models for genetic, molecular, and cellular aspects of disease are laboratory mice (Mus musculus) and rats (Ratus norvegicus albinus). This state of affairs exists for both practical and circuitous reasons. The practical reasons are that rats and mice are similar to humans at the cellular and molecular levels; develop some of the same diseases as humans, particularly certain cancers; and have very short generation times. Rats and mice have gestational lengths of about twenty days, are ready to breed at four months of age, and progress through aging and death in two years. In contrast, the gestational period of dogs is about sixty days; dogs can breed at seven to twelve months of age; and life spans vary from seven to fifteen years, depending on breed.

In addition, mice can be genetically manipulated to express designer genes, including human genes or mutated genes, or to have gene additions or deletions. This development of "designer mice" in laboratories follows from their short generation times and the wealth of information that scientists have gained about their reproductive biology and genetics. Rats are not used as widely for genetic manipulation, but they are commonly used for toxicologie and neurobehavioral studies. Mice and rats have been studied for nearly a hundred years in the laboratory, which brings us to the circuitous part of the reasoning. We have so much information about mice and rats that we tend to keep using them because other animal species are less well-characterized. However, though rodent models are here to stay for the foreseeable future, a major reevaluation of the value of canine models for human disease is about to take place.

Valuable though they might be to science, the idea of increased use of dogs as research animals is particularly antithetical to our view of them as man's best friend. However, the dog as a subject of study for inheritable diseases will be much more like the human than the mouse, with some of the same ethical considerations. Much can be learned from pet populations through veterinary records and regular health checks, as well as through the store of breeding information maintained by breed clubs and the American Kennel Club. Dr. Ostrander's study, for example, required cheek-swab samples for DNA from purebred dogs at shows or in their owners' residences. Purebred dogs may be much better populations for study than human families because data on groups of related canine individuals over several generations are more likely to be available. It will be interesting to see the next developments in canine genomics. Today dogs, tomorrow - cats? Stay tuned as science strides across this field in seven-league boots.

Evelyn Tiffany-Castiglioni, PhD, is associate dean for Undergraduate Education and head of the Department of Integrative Biosciences, College of Veterinary Medicine, Texas A&M University. She conducts research on the neurotoxicity of environmental contaminants. She recently edited In Vitro Neurotoxicology: Principles and Challenges (Humana Press, 2004).