An article by Pemberton and colleagues1 in this month’s issue informs readers about the potential for advances in genomics to bring about a revolution in dental research. The authors present a textbook case of how modern genomic methods have led to outstanding successes in identifying genes that cause congenitally missing teeth. Yet they correctly point out that such successes in dentistry have been the exception and not the rule. Compared with many areas of medicine, progress in understanding oral and dental disorders using genetic approaches such as linkage analysis has been limited. The authors call for increased focus on family history when evaluating dental conditions and for enhanced communication and collaboration between dentists and experts in the field of human genetics.
Borrowing from Shakespeare’s Hamlet, this surely is "a consummation devoutly to be wished." But, in reality, it may take something like a Cultural Revolution (borrowing from Chairman Mao Tse-tung) to bridge the wide gap that exists in many areas between basic biological sciences and dental practice and education.2 Of 264 dental hygiene programs surveyed recently, not one required a formal genetics course as a prerequisite for admission or as part of the curriculum.3 Similarly, only eight of 53 schools provide dental students with a formal genetics course.4 Unless major changes are made, neither today’s students nor tomorrow’s faculty and practitioners will be equipped with the necessary foundation of knowledge and experience in the area of genetics to break down the barriers that have until now prevented dental research from fully capitalizing on the new genomics tools.
Even when carefully clinically evaluated families for various craniofacial, oral and dental conditions are recruited, causative disease gene mutations may not found easily or quickly. The "simple" genetic etiology of the hypodontia example focused on by Pemberton and colleagues1 in their "primer" article may be relevant for only a small portion of dental disorders. When only a single gene controls the development of a disease, when a high proportion of the subjects who carry the high-risk mutation actually develop the disease and when the disease is otherwise rare in the population, then the linkage analysis approach that was highlighted can lead rapidly to successful gene identification. These kinds of diseases are sometimes referred to as "the low-hanging fruit" of human genetics. In medicine, prizes such as the cystic fibrosis and muscular dystrophy disease genes were harvested rapidly after the first generation of whole genome polymorphism maps were developed nearly two decades ago. It is highly uncertain, however, how many such easily found dental disease genes lie out there, easy to discover by gathering only a small number of families suitable for the linkage approach. The only real way to find out is to look, and we should look hard; but we also should guard against building up hopes that are too high, so that we don’t abandon the genomics approach too quickly if overnight successes aren’t always the rule.
Twin studies are a powerful design for assessing the relative contribution of genes versus environment to disease risk. Such studies indicate that 50 percent or more of individual differences in risk of developing chronic periodontitis5 and dental caries6 is genetically based. So we can conclude with high confidence that genes are important in these major dental disorders. However, it is likely that a complex or multifactorial etiology underlies these and many other conditions that are a primary focus of dental practice. In a complex disease, inherited genetic differences control a large portion of the differences among people in disease risk; but this genetic risk is spread over a dozen or more genes located at different positions in the human genome, with each individual gene having only a modest effect (for instance, twofold or fourfold increase in risk of developing the disease).
Although each gene has a small effect, the total effect of all genes may be substantial. Furthermore, even genes that have small effects on risk in a person may be responsible for a large portion of the disease in the population. This is because some of these DNA variants may be very common. Consider a Huntington’s disease mutation. This acts as a virtual "light switch" in turning on the person’s chance of developing this terrible and debilitating disorder with certainty by late middle age. But since this mutation fortunately is uncommon, it doesn’t cause much disease burden for the overall population.7 Contrast this with the much more common E4 variant of the apolipoprotein E gene, which is associated with decreased longevity, increased cholesterol levels, cardiovascular disease and Alzheimer’s disease.8 Although the variant has relatively small effects on a person’s risk, its high frequency in populations makes it a major contributor to disease in the population.
The linkage analysis approach, with few exceptions, has failed at the task of mapping genes for such complex diseases. This failure has been explained by a mathematical analysis that has shown that thousands of families would be required to provide sufficient statistical power when a gene only increases risk by two- or fourfold.9 Obviously, recruitment of such large clinical samples rarely can be achieved. Fortunately, this mathematical analysis suggested that association-based methods sometimes can provide adequate statistical power with much smaller numbers of patients, using either families or unrelated cases and controls. The only catch is that instead of needing only a few hundred polymorphic DNA markers to cover the entire human genome, the association strategy requires several hundred thousand such markers assayed on each subject.
Fortunately, molecular genetic technologies have been developed that can meet the challenge of producing these massive amounts of data, and the International HapMap Project ("HapMap" being an abbreviation of "haplotype map") has catalogued this variation in several human populations and made it freely available online for researchers wishing to tap into this rich genomic treasure chest.10 In the first phase of this project, the frequency of DNA variants was measured at more than 1 million locations distributed across the human genome in European, African and Asian subjects. DNA variants that are located near each other on the chromosome often are correlated with each other, so if an investigator determines the DNA sequence for a subject at one position, the DNA bases at the neighboring variant positions often can be predicted with a high degree of confidence. This phenomenon is known as "linkage disequilibrium," and what the HapMap Project has done is to determine these patterns of correlation among neighboring DNA variants for a large portion of the human genome. Armed with this information, investigators interested in studying inherited variation at a candidate gene for their disease (or searching through all genes in the entire genome) need not undertake the large effort of conducting assays for all known variants in their clinical subjects. Instead, they can use computer algorithms on data derived from the HapMap Project to measure most of the inherited variation present in the genome at a substantially reduced cost by identifying an optimized subset of DNA variants that serve as statistical "tags" for many other variants that are not actually assayed in the laboratory.
The association approach, however, is not a panacea. It also will fail in some circumstances that exist in human populations. It does, however, extend the range of dental disease genes that can be identified beyond that within the limited reach of the linkage approach. But by no means is this intended to suggest that linkage analysis is useless and should be abandoned. Both linkage and association can be powerful strategies, and the choice of which to employ will depend on researchers’ best guess about the genetic architecture (simple versus complex) of the disease they are interested in studying.
A final important point is that the HapMap Project approach has promise not only for identifying disease susceptibility genes, as was focused on here, but also for understanding the cause of individual differences in response to various kinds of clinical treatments, including both drug and nondrug therapies.11 In dentistry, for example, this knowledge may be used one day to guide the choice of implant or orthodontic techniques used to treat patients based on the patients’ genetic polymorphism profiles. Such research studies are only recently getting started, but if successful, these approaches may truly bring about a revolution in the practice of dentistry.
TOP
REFERENCES
