 Hi, my name is Elaine Ostrenner and I'm head of the Comparative Genomics Section and I'm Chief of the Cancer Genetics Branch at the National Human Genome Research Institute at the National Institutes of Health. I'm going to be talking today about comparative genomics, that is, the comparison of genomes from different organisms in order to better inform us about the 20 to 25,000 genes that have been described in the human genome. To do this kind of work, one has to take into account the genome sequence of everything from humans to small rodents to insects like flies and worms to single-celled organisms like bacteria and yeast. Comparing genomes is a lot like cryptography. You line up the DNA sequence as best as you possibly can and then you see what is similar and what is different between the genomes. This involves doing many two-way comparisons between the genome sequence of, say, species A, in this case humans, and species B, in this case mouse. You then generate a consensus document that highlights the regions that are the same and those that are different between the two organisms in question. This gives you guidance as to what portions of the genome are likely to encode genes or regulate genes that have a common and vital function in all of mammals or actually in all of the animal kingdom versus those that are diverged, that is, their function has changed or evolved over time, becoming more specialized in order to meet the needs of that organism. These are some of the vertebrate genomes that either have been sequenced or are currently being sequenced and you can see that they include everything from humans and chickens down to wallabies and puffer fish. The greater the divergence of organisms that we can sequence, the better job we can do finding genes that are well-conserved, performing functions that are both common and important in all organisms in the planet. These are some of the genomes that are actually under discussion to sequence and again you can see that they include a wide variety of species, everything from ostrich to bat. It'll be interesting to see which actually make the final shortlist. Now my own lab has largely been interested in the domestic dog. There are about 150 breeds of dog recognized in the United States by the American Kennel Club and they display an amazing amount of variation in phenotype, that is, differences in body type, in behavior, and in disease susceptibility. The comparison is very nicely made just by looking at this great Dane who's about 40 fold larger in size than this Yorkshire Terrier. Keep in mind that they're both members of the same species, Canis familiaris. Now the process of creating dog breeds was done by dog fanciers largely in the last two to three hundred years and it often involved crossing closely related individuals and as a result we see not only fixation of specific traits, body size, leg length, coat color, the curl of the coat, but we also see an excess of genetic diseases in many purebred dogs. This offers us a chance to find the genes that are important both for human health as well as for companion animal health and in my lab at NHGRI who are actually interested in both. Now the dog has 39 pairs of chromosomes, 38 of those are autosomes, and then there are the sex chromosomes X and Y. Humans have a little bit less. There are a total of 23 pairs, 22 autosomes in addition to the X and the Y. Now one of the first things that we did in my lab was to make comparative maps of the dog and the human genome. So for instance we know that canine chromosome 30 contains the same genes and largely the same order as a portion of human chromosome 15. Canine chromosome 1 contains chunks of human chromosomes 18, 6, 9 and 19 with largely the same genes and the same order in spacing. Now the dog genome sequence was recently completed and my lab had the opportunity to pick which dog was actually going to get sequence and we chose a boxer named Tasha because after scanning the DNA from hundreds of different dogs we found that Tasha had a very low level of genetic variation and that make the process of assembling her genome after the actual sequencing exercise go much better. The dog genome is actually a bit smaller than the human genome but to a first approximation it contains all the same genes. The sequence was completed very quickly taking about a year from start to finish. A very large number of people from all over the world were involved in different aspects of the project from preparing DNA to actually generating DNA to finally analyzing the data. There were even some high school students in our lab that had a chance to participate but most of the large scale data generation was actually done at the Broad Center for Genome Research at MIT and Harvard and you can see in this picture a small number of people in a very large number of robots working on the genome sequence of the dog. This data is all publicly available and it's very easy to manipulate if you go to the website shown in the slide. You can look for the sequence of your favorite gene or you can just play around with the sequence if you'd like to. Now in my lab we've been particularly interested in genetic variation. The American Kennel Club has divided the 150 or so breeds that they recognize into seven major groups. Members of a group share either a common heritage, a common phenotype or a set of appearances or a common set of behaviors. When we began the dog genome we proposed that it would allow us to find genes that were important in morphology that is the shape and the size of a dog. Here that is why to herding dogs herd, why to pointing dogs point, as well as disease susceptibility. Why are some diseases so prevalent in some breeds while they're essentially absent in other breeds? Now worldwide all of these things are going on but in the next few moments I'm only going to talk about the first. It's important to know that we obtain samples from dogs by going to dog shows and then just collecting a simple cheek swab. We don't keep any dogs on kennels and we don't breed any dogs. But we partner with people and families like yours, dog owners, dog breeders and the American Kennel Club and the AKC K9 Health Foundation in order to obtain samples from families or breeds of dogs that have particular traits or diseases that we're interested in. Now shown here are some of the students and postdoc fellows in my lab but a dog show in Seattle and you can see that they're collecting cheek swabs from a number of interesting breeds. We do a lot of computational work in order to manipulate the data that we get and try and answer the questions that we've posed so everyone in my lab has to have some kind of computer skill and some facility for analyzing data. Shown here are results of an interesting experiment that demonstrates that the DNA of different dog breeds is really distinct. So if we examine a set of markers that span the genome we find that computer programs can very easily distinguish the five dogs in this experiment that were doxons from the DNA of the five that were golden retrievers, from the five that were Newfoundland, the five that were Pekinese and so on and so forth. We can also make statements about how related dog breeds are one to another. So we know for instance that the Asian breeds, the Akita, the Chow, the Pekinese, the Chiba Inu are all very close to related one to another and they're also pretty close related to wolves. We also know that the Mastiff-like breeds are also pretty close to related. So the Boxer and the Mastiff and the Bulldog, those breeds of dog. And finally we know that the herding dogs all share a common heritage as well. We can even take a cheek swab and now with better than 99% accuracy if it's a purebred dog we can tell you what breed of dog it actually is. Now here on the slide is an analysis that involves 414 dogs and just using cheek swab DNA, the computer correctly identified the breed of 400 of those dogs. So we're currently working on computer programs and computer algorithms to get that 99% up to 99.999%. Now this kind of work is actually important because it means that we can group dogs together that have similar diseases and in doing so we can really increase our statistical power even though they may be representing different breeds of dog. So dogs present with the same diseases that humans do and cancer, epilepsy, allergy, heart disease, eye diseases are all problems that are common to human health and biology as well as companion animal health. Cancer is really at the top of the list something like one in three dogs will get cancer at some point in their life compared to one in four humans. For that reason cancer has been a major focus of the work going on in my lab as well as that of my colleague Dr. Chan Khanna at the National Cancer Institute who's a veterinary oncologist. We've been for instance interested in a disease called this abbreviated RCND and it causes a type of kidney cancer in German shepherds. We've worked with colleagues at the Norwegian School of Veterinary Medicine to identify the gene that's responsible for RCND and found that the same gene is responsible for a very similar human cancer syndrome. In dogs the disease is caused by a mutation in a gene coding a protein called folliculin. Mutations in the same gene cause a disease in humans called Burt-Hogged Bay syndrome. When we compare the sequence of the folliculin protein from dozens of organisms everything from dogs and humans down to flies and worms and beasts we find that there's a very invariant region of the protein that when mutated is what causes the disease in dogs. So that's telling us that this must be a very important region of the protein for carrying out whatever its assigned function is. Dr. Khanna's lab works on osteosarcoma and this is a disease of the bone that's common in Rottweilers, Irish Wolfhounds, and Scottish Deerhounds. But it's also very prevalent in young people, more often males than females and it affects the long bones of the arm and of the leg. This is a picture of some bone cancer cells and Dr. Khanna's lab is interested in how different genes are turned on and turned off in cancer cells. When a gene is turned on researchers say the gene is being expressed. Shown here is a microarray and that helps us understand the different expression patterns that we see in cancer cells versus normal cells. And that's really key to understanding what proteins and genes we need to target in order to develop effective therapies for targeting various kinds of tumors. So there are lots of take home messages from this presentation. But the most important is that in thinking about how to design and experiment, try not to be limited by what those around you have been doing or how they think about the problem. Choose an interesting question and consider the smorgasbord of variation that nature has to offer in choosing the models that you're going to be working on for tackling whatever question it is that most interests you. Now my own lab has worked on everything from dogs to elephants to wolves to mink and ermine and sea otters. Try to think about something that's really interesting to you and then consider the array of choices that nature has to offer for how you're going to tackle that question and understand really at a deep level the genetics and the genomics of that question. You're sure to learn something interesting, not just about the biological system of choice, but about what biology and nature as a whole has to offer. By the way, for the cat lovers in the audience, sequencing of the cat genome is just beginning now. This is actually Cinnamon, the lucky cat whose DNA is being sequenced. So cats will soon be able to take their rightful place in comparative genomics right next to the dog. So in summary, comparative genomics is a powerful tool for understanding the human genome, and dogs are really a strategic part of that effort. Availability of the dog genome sequence is opening new avenues of research. Population genetics of dogs facilitates discovery of canine and human disease genes, including those involved in cancer. And the study of dog genomics is informative for understanding basic principles in biology. So thank you very much for listening, and thanks also to the many dog owners, breeders, and our colleagues, and especially the dogs who've been willing to share their DNA with us. Now I think I'll go get a cup of coffee. I just pulled.