Today, as genetics morphs into genomics, certain populations are again playing prominent roles in gene discovery. These "Founder populations" are modern groups that descend from a few individuals who left one area to settle another for political, religious, or social reasons. Founder populations are sought after for association studies that search for links between health-related gene variants and SNP patterns (profiles of single nucleotide polymorphisms, which are sites within the genome where a single base can differ among individuals).
At the Genome 2001 Tri-conference held in San Francisco this past March, Philippe Douville, vice president and chief business officer of Galileo Genomics Inc. of Montreal explained that "The main recognized Founder populations in the world are those of Quebec, Finland, Sardinia, Iceland, Costa Rica, the northern Netherlands, Newfoundland, and several discrete ethic groups, including the Ashkenazi Jews." In addition to providing useful tools to predict disease susceptibilities, studying Founder groups often turns up fascinating historical facts.
From Monogenic to Polygenic Disorders
Founder populations amplify certain gene variants while maintaining great stretches of uniformity in other DNA sequences. This manifests as increased disease frequencies. Among an oft-studied population of 18,000 who live in northeastern Finland, for example, the lifetime risk for developing schizophrenia is 3.2 percent, nearly three times the national-average risk. This subpopulation traces its ancestry to 40 families who settled in the region at the end of the 17th century.1 The population has been easy to study because the Finnish church has records of births, deaths, marriages, and moves, and hospital records are available. In some countries, such as Iceland,2 Estonia, and the South Pacific archipelago of Tonga,3 large-scale projects have already begun to connect various health conditions to DNA patterns.
Geneticists have long used Founder populations, with their skewed gene frequencies and meticulously kept genealogies, to study rare, single-gene disorders. In the United States, the Mormons helped reveal the roots of hereditary colon cancer, and the Amish helped demonstrated a suite of metabolic disorders and other characteristics. All of the cases of polydactyly (extra digits) among the Old Order Amish in Lancaster, Pa., stem from one founder. Postgenomic era biosleuths, however, are pursuing the more common polygenic conditions, where individual genes contribute incrementally to the phenotype.
A Signal-to-Noise Challenge
In association studies, researchers see if people with a particular condition consistently have a particular SNP pattern, and if matched controls do not share this pattern. The disease-associated SNPs might be parts of genes that increase susceptibility to the condition. Clearly, the larger the samples compared, the more powerful the association. Hence, the budding alliances between biotech start-ups as well as more established firms with clinical partners that can provide such information.
Underlying both family and population-based gene discovery is the phenomenon of linkage: the idea that genes that are located close together on the same chromosome are transmitted together. In fact, the first genetic maps, in Drosophila, were assembled by determining how often two genes cross over from one chromosome to its mate--the farther apart two genes are, the more often this can happen.
Genes that are very close together exhibit extremely tight linkage, called linkage disequilibrium (LD), which is particularly pronounced in isolated populations.4 Association studies on such peoples tap into this LD, seeking SNP patterns that herald the presence of health-related gene variants against a backdrop of genetic uniformity that persists because few, if any, new variants are introduced. Ariel Darvasi, president and chief scientific officer at idGENE Pharmaceuticals Ltd. in Jerusalem explains, "Gene discovery is a signal-to-noise problem. Using isolated populations reduces the genetic noise."
Each Founder population has its own intriguing tale. Here are two.
The Quebec Founder Population
The French Canadian population in Quebec has what Douville calls "optimum characteristics for gene discovery in each of the key criteria." These include a large number of generations since founding (14 generations), a small number of founders (about 2,500), a high rate of population expansion (74 percent per generation), a large population today (6 million), and minimal intermarriage.
The French Canadians have remained a founder population because it has stayed pretty much to itself, within a larger population. The French founded Quebec City in 1608. Until 1660, the population grew from continuing immigration from several parts of France but then began to increase from reproduction. By the time the British conquered the area in 1759, 10,000 French had immigrated, but many of them had headed westward, taking their genes with them. Meanwhile, back home in Quebec, language and religious differences kept the French and English from mingling their gene pools to any great extent. Douville concludes "The Quebec population has grown, by natural increase, from some 2,000 to 4,000 founding genotypes to some 6 million today."
In the 19th century, agricultural lands opened up about 150 miles north of Quebec, in the Charlevoix and Saguenay, Lac-St.-Jean (SLSJ) regions. Families migrated north, and their descendants form an incredibly genetically homogeneous subpopulation of founders split off from the original set of founders. "It is estimated that the population of SLSJ of approximately 300,000 people, roughly the same size as the Icelandic population, has descended from some 600 effective founders of the Charlevoix region, compared to 8,000 to 20,000 founders of Iceland," says Douville.
Ashkenazi Ups and Downs
idGENE Pharmaceuticals is analyzing the 2.5 million Ashkenazi Jewish population in Israel. "This homogeneous population came from a small area in Germany, about 500 years ago. The core population was only about 1,000 to 2,000. This is a great advantage for identifying genes affecting common diseases today," comments Darvasi. The single-gene disorders that are more prevalent among the Ashkenazim include Gaucher disease, factor IX deficiency, Bloom syndrome, Canavan disease, and Niemann-Pick disease. idGENE will select the complex disorders it will follow from a long list of candidates which includes diabetes, osteoporosis, asthma, Alzheimer's disease, and multiple sclerosis.
The Founder populations whose genomes are being combed for SNP patterns that serve as beacons for disease will directly benefit from the investigations in the form of tests that can reveal inherited susceptibilities. Individuals can then alter their lifestyles to control environmental factors, or begin diagnostic testing or treatment early in the course of pathogenesis. But the bigger picture being painted, or at least envisioned, is that such studies on Founder populations will enable researchers to dissect the interactions of genes that lie behind the major illnesses of humans--information that will ultimately benefit us all.
1. I. Hovatta et al., "A genomewide screen for schizophrenia genes in an isolated Finnish subpopulation, suggesting multiple susceptibility loci," American Journal of Human Genetics, 65:1114-24, 1999.
2. R. Lewis, "Iceland's public supports database, but scientists object," The Scientist, 12:1, July 19, 1999.
3. T. Hollon, "Gene pool expeditions," The Scientist, 15:1, Feb. 19, 2001.
4. G.P. Page, C.I. Amos, "Comparison of linkage-disequilibrium models for localization of genes influencing quantitative traits in humans," American Journal of Human Genetics, 64:1194-1205, 1999.
2001 Tri-Conference Celebrates Human Genome Project
The Fairmont Hotel atop San Francisco hosted three two-day genomefests in March, sponsored by the Cambridge Healthtech Institute: "Genomic Partnering: Emerging and Early-Stage Companies," "Human Genome Discovery: Commercial Implications," and "Gene Functional Analysis." An eclectic roster of speakers tried to encapsulate their excitement in 15-minute speed-of-light Powerpoint presentations, while the exhibition featured a smorgasbord of bioinformatics companies old and new, with ever more clever names containing the letters "gen." Many were interested in assembling SNP maps, either starting with Founder populations, or with patient populations plagued by a particular complex trait, from cardiovascular diseases to eating disorders.
The jargon was a curious mix of acronyms heralding the dark ages when genomics was genetics--from BACs and ESTs, to RFLPs (probably the worst of the bunch) to SNPs (probably the best)--all mixed with MBA-speak. At one point, when yet another presenter evoked the tired image of mining for data, promising his start-up would provide "the picks and shovels for genome prospectors," a collective groan arose from the audience. And so quickly has this new field evolved that some journalists were taken by surprise, expressing astonishment that a biomedical meeting would attract so many computer companies.
Yet any genome conference nowadays is a heady experience, especially for anyone old enough to remember the mid-to-late '80s meetings where the idea to sequence the human genome was first batted around, amid much skepticism.
Ari Patrinos, associate director of science for biological and environmental research at the U.S. Department of Energy in Germantown, Md., succinctly summed up the results of the project. "We have 30,000 to 35,000 genes that encode more than 100,000 proteins. Our complexity arises from how we use our genome." (Patrinos can be seen in the human genome issue of Science standing between a grinning J. Craig Venter and Francis S. Collins,1 perhaps a greater feat than the actual sequencing.) He also eloquently captured the general ambiance of amazement. "It seems like the morning after. Now we are sobering up, picking up the pieces, and realizing two fundamental things: One, there is still very much left to be done in completing the sequence. What we celebrated [in February] was a rough draft. Some parts are extremely difficult to sequence. Two, there is reason to be very excited, in delicious anticipation of what lies ahead. We are at the beginning of uncharted territory."
But it was a comment by Noubar Afeyan, president and CEO of NewcoGen Group of Cambridge, Mass., that brought down the house. Said he, "Those of you who still think junk DNA is junk, I invite you to take it out of your genome, and see what happens."
1. E. Marshall. "Sharing the glory, not the credit." Science, 291:1189, Feb. 16, 2001.