Statement of Mary-Claire King to the National Academy of Sciences committee charged with evaluating the Human Genome Diversity Project, September 16, 1996
Good afternoon. Thank you for permitting me to speak in this public meeting. My colleague John Moore and I represent the North American Regional Committee of the Human Genome Diversity Project. I am American Cancer Society Professor of Genetics and Medicine at the University of Washington. My research and teaching interests lie in human genetics, in particular how molecular genetics and genomics can be integrated with approaches from population genetics and epidemiology to address problems of complex human diseases and questions of human evolution and diversity. I have been interested in these questions for 25 years, since I was a graduate student in Allan Wilson’s laboratory at UC Berkeley, where our work focused on human evolution at the molecular level.
Since then, my lab has studied problems of complex human disease traits, especially breast and ovarian cancer, inherited deafness, systemic lupus erythematosus, and AIDS. We have tried to contribute to and to use information about the human genome to identify genes critical to the development of diseases such as breast cancer, then to take the normal alleles of the same genes as the basis for developing cures. I’ll return to this theme of complex diseases in a moment.
In parallel with our work on specific diseases, we have had the opportunity to apply molecular genetics to problems of human rights. We have worked since 1983 with the Abuelas de la Plaza de Mayo to identify their grandchildren kidnapped by military squads in Argentina during the Dirty War, to learn who these children are, to return them to their families, and–at least in some cases–to bring their kidnappers to justice. As soon as PCR was developed, we applied mitochondrial DNA sequencing, in the context of human diversity, specifically to the project with the Abuelas. This was an outgrowth, of course, of the evolutionary studies based on mtDNA sequencing of PCR products in the Wilson lab. This human rights work in Argentina was one of the first projects supported by the then-very-new ELSI Committee of the Human Genome project. It has also been supported by Amnesty International. The project continues to the present, with identification of now-adult children who present themselves as possible members of the disappeared.
We have applied the approaches we developed with the Abuelas to other problems of human rights, including identification of victims of military murders in Chiapas, in Salvador, and in Somalia, and to the identification of MIAs from World War II, Korea, and Vietnam. We are now working on behalf of the United Nations War Crimes Tribunal to identify murder victims from Bosnia and Rwanda. This work is supported by the United Nations through Physicians for Human Rights.
What does all this have to do with the Human Genome Diversity Project? The connection is that we can use the tools of molecular genomics and population genetics to answer questions about ourselves. An important class of these questions are those of complex diseases of people. In recent years, there has been a great deal of progress in the identification of genes that influence diseases like cystic fibrosis, some rare familial cancers, some common cancers, some inherited blindness and deafness, and many others. We are beginning to learn how to use these genes for prevention and cure.
The biggest challenges are the complex diseases: that is, diseases for which more than one gene, as well as environmental exposures, are likely to influence the appearance of symptoms in each ill person. Diseases like this include diabetes, rheumatoid arthritis, and hypertension, to name just three. How are we to disentangle the multiple causes of these diseases? An approach that is proving successful is to work with relatively isolated groups of people in whom the disease is common, with the hope that the picture may be clearer in a group in whom relatively few, relatively ancient mutations may be responsible for a portion of the genetic influence on the disease.
Of many current examples of this approach, one published last week is a study of diabetes in a linguistically isolated community in the Botnian region of western Finland. In a subset of families with diabetes in this geographic area, a segment of chromosome 12 about 20 million basepairs in length appears to be shared by relatives with the disease. Now the problem is to close in on this gene and isolate it. However, for this and other complex diseases, isolating genes will be difficult, because no informative recombination events remain to decrease the size of the linked region.
To identify these complex disease genes, then, one must rely on the association of genetic markers with diseases, or genetic disequilibrium. Genetic disequilibrium is the association of specific DNA sequences, or markers, with one another in populations. Genetic disequilibrium depends on (1) the physical distance between the marker and the disease allele; (2) polymorphism of the marker, or how variable the marker is; (3) mutation rate at the marker, or how fast the variation changes; and (4) recombination rate as a function of physical distance for this specific part of the genome. Some of these parameters depend, in turn, on features of populations; specifically (1) effective population size; (2) degree of endogamy; and (3) age of the disease allele in the population.
Another way of phrasing this question is: given linkage and disequilibrium between markers hand a hypothetical disease allele over a large genomic region, where is the best place in that genomic region to start looking for the disease gene? Twenty million base pairs is very large to scan for individual mutations, even when the genome sequence for one amalgamated person is known.
In order to narrow the search, it will be important to know more about genomic structure at the population level. The way to learn genomic structure at the population level is to study various genomic regions in different populations, with each genomic region evaluated at markers very densely distributed. The goal of this approach is to evaluate disequilibrium among markers (at different distances apart and with different mutation rates) across the genome in populations differing in size, endogamy, and history. In other words, how is disequilibrium distributed in the genome, as a function of (1) features of individual variant sites, (2) chromosomal structure, and (3) organization of populations?
How does the Human Genome Diversity Project contribute to this understanding? Wouldn’t it be possible to carry out this analysis one population at a time, one chromosomal region at a time: for example, for chromosome 12 in western Finland? Of course. This is analogous to sequencing a part of the genome from one person with disease at a time. Indeed, this is where we stand right now in the Human Genome Project. Individual studies are a necessary interim solution. However, much more is to be learned, and the information will be more universally useful, if disequilibrium structure is studied for the genome as a whole and in populations with a range of historic features. Small numbers of people from many historically coherent populations, specifically including those without disease, are the best source of this information.
Why should anyone from any population–particularly indigenous populations exploited for lifetimes by outsiders–be willing to participate in such research? Good question. A fundamental tenet of those of us working on the North American Regional Committee of the HGDP is that any community, including in particular indigenous communities, should control research access to the own community, to their DNA, and to any possible commercial value from research with these resources, however remote the possibility of commercial value might be.
In our research on breast cancer in families, many families choose to participate in our project; others decline. Similarly, in the Human Genome Diversity Project, it is to be expected that some communities would participate and others decline. It is obvious to me that individuals, families, and communities have the wisdom and intelligence to make these decisions for themselves, and the right to be provided with information that is useful for doing so. With these principles in mind, the North American Committee of the HGDP has developed a draft of a model ethical protocol for the collection of DNA samples for research. Your committee has, of course, obtained and discussed this draft protocol. We hope that the HGDP will break new ground in recognizing the autonomy of communities in making such decisions, and that this standard will be applied beyond HGDP to biomedical research generally.