Hunting down the mutant gene for a specific inherited disease ishard enough. It took 10 years to pinpoint the sequence responsiblefor Huntington's disease, a classic instance of a malady passeddown from parent to progeny. Gene-hunters finally zeroed in onHuntington's DNA address early last year. (See BioWorld Today,March 23, 1933, p. 1.)In exploring human chromosomes for the genomic instigators ofinherited ailments, they followed the Mendelian rule of thumb: onegene per malady. This has led to genetic mapping of more than 400of the 3,000 on the recognized roster of monogenic diseases. Thefounder and keeper of this roster is medical geneticist VictorMcKusick of Johns Hopkins University, author of MendelianInheritance in Man, first published in 1956. "We usually say3,000," McKusick told BioWorld Today. "It's a softish figure,fuzzy around the edges." He added, "All disease tends to have agenetic component, even automobile accidents. It may be the otherguy's genotype rather than your own."Now the rules of the gene-hunting game are changing, its playingfield widening. Thus, though most cancers don't appear on thatmaster list of inherited diseases, they have become fair game forseeking sequences implicated in malignancy.Since 1990, genetic detectives have tracked down genes thataccount for a fraction _ 10 to 15 percent _ of cancers that "run inthe family," notably in patients diagnosed with familial colon andbreast tumors. What that minority inherits is a strong susceptibilityor predisposition to those cancers. (See BioWorld Today, Sept. 17,18 and 19, 1994, all p. 1.)This leaves the majority, 80 to 85 percent or so, contracting theirtumors primarily from environmental causes, such as smoking,while the genetic co-component, if any, remains occult.Nowadays, any and every disorder that flesh is heir to, rather thanjust the hereditary ones, is on the gene-hunters' hit list. Diabetes,hypertension, heart disease, infection, plus cancer in all its non-familial forms, are examples of diseases that don't follow simpleinheritance patterns. These, and many other afflictions, answer tothe genetic description of "complex," because their etiologyimplicates more than a single gene.Complex Traits: Guide to a GuideThe Sept. 30 issue of Science is largely devoted to the theme ofgenetic mapping. Its lead article, "Genetic Dissection of ComplexTraits," is virtually a users' guide to the gene-hunters' expandingkit of high-tech tools for charting the chromosomes of affectedindividuals and their relatives."Human genetics," declares the paper's first author, Eric Lander,"has sparked a revolution in medical science, on the basis of theimprobable notion that one can systematically discover the genescausing inherited diseases without any prior biological clue as tohow they function."Lander, who is director of the Whitehead/ Massachusetts Instituteof Technology Center for Genomic Research, credits the advent ofrecombinant DNA for "carrying genetic mapping to its logicalconclusion with the development of positional cloning." Thistechnique he defines as "the isolation of a gene solely on the basisof its chromosomal location, without regard to its biochemicalfunction."The "key breakthrough," he dates to 1980, with "the recognitionthat naturally occurring DNA sequence variation provided avirtually unlimited supply of genetic markers" for human families.`The Fourfold Way' To A Gene's LocusLander predicts that this "new genetic frontier" of dissectingcomplex traits "is ready to explode." So far, its arsenal includeswhat he calls "the fourfold way" _ four major approaches: linkageanalysis, allele-sharing methods, association studies, and polygenicanalysis of experimental crosses.These are not exact sciences, but numbers games. Rather thandetermining absolute correlations between the data they obtain andthe target traits they seek to scope, their findings merely lay downdegrees of likelihood. These the geneticists call "lod scores."("Lod" stands for "logarithm of odds.")A salient, very recent, example of linkage analysis is its use tolocate the first gene for early-onset breast cancer, BRCA1, on aregion of chromosome 17's long arm. (See BioWorld Today, Sept.18, 1994, p.1.) The researchers kept adding susceptible families inthe order of their average age at onset of the disease. The lod scorekept rising until age 47, then dropped off for older ones. Thatdefined the upper age-limit for familial, early-onset disease, incontrast to most breast cancer, which affects post-menopausalwomen.Linkage analysis can also ferret out multigene complexities bysimultaneously searching pairs of chromosomal regions. Landercites genes for multiple sclerosis, traced by two-locus inheritanceanalysis in large numbers of Finnish kindreds. The linkages led tothe myelin basic protein gene on chromosome 18 HLA (humanleukocyte antigen) and on chromosome 6."The more complex the trait," he concludes, "the harder it is to uselinkage analysis."Allele-sharing is the opposite of linkage analysis. Instead ofconstructing and testing a hypothetical model, it rejects a model. Itcompares the familial inheritance patterns of the trait in questionwith those of a completely random sampling of genomes in thepopulation, to see if "affected relatives inherit identical copies ofthe region more often than expected by chance." (Alleles reflect thevariant DNA sequence patterns on gene copies inherited fromone's father and mother.)In its simplest form, allele sharing compares the genomes of twoaffected siblings. It has found, Lander reports, "a locus onchromosome 11q [its long arm] pointing to a previouslyunidentified causal factor in type I diabetes." In another search,"brothers concordant for the trait of homosexual orientationshowed significant excess allele sharing" on the sex-linked Xchromosome.Association studies focus on the affected individual rather than hisor her family. If a given allele of a gene under scrutiny occursmuch more often on the genome of an affected individual than onjudiciously selected controls in the population at large, it is said tobe "associated" with the trait in question.This method shines in unraveling autoimmune diseases. Its classiccase is ankylosing spondylitis _ arthritis of the spine _ which ismarkedly associated with the HLA complex. The HLA-B27 alleleoccurs in 90 percent of these patients, compared with only 9percent in the general population.Lander cites HLA associations with Type I diabetes, rheumatoidarthritis and systemic lupus erythematosus (SLE). More recently,he noted that similar association studies implicate the apoliproteinE gene in late-onset Alzheimer's disease and heart disease.He cautions that in and of itself, association cannot conclusivelyidentify the cause of a disease. To do so, it must occur in allaffected populations."Most disturbingly," he cautions, it can also arise "as an artifact ofpopulation admixture." In a "light-hearted example" of such anartifact, he describes a geneticist who studies the "trait" for eatingwith chopsticks in San Francisco. Sure enough, he finds anassociated allele, presumably on a gene for manual dexterity. Inactual fact, that marker is simply more common among Asiansthan Caucasians.A genuine instance of such spurious association arose in studies ofwhy Pima Amerindians are more susceptible than Caucasians toType II diabetes. Researchers found a "protective allele" present inmost Caucasians, as well as in some Pimas spared the disease. Itlater turned out that the latter had higher degrees of Caucasianancestry.The jury is still out on ongoing studies testing the association ofalcoholism with an allele of the dopamine D2 receptor.Experimental crosses for mapping polygenic traits, the fourthapproach in Lander's tetralogy, is by definition limited tolaboratory animals. "With the opportunity to study hundreds ofmeioses [dividing up of chromosomes in a cell's nucleus], from asingle set of parents," he writes, "the problem of genomicheterogeneity disappears, and far more complex geneticinteractions can be probed than is possible for human families."But Lander adds the caveat that animal studies must be applicableto human diseases.And this counter-caveat: "Animal models that are poor models forpharmacologists seeking to evaluate a new human drug therapymay nonetheless be excellent for geneticists seeking to elucidate thepossible molecular mechanisms or pathways affected in a disease."For example, a study of hypertension in rats quickly led to parallelsin human high blood-pressure, which revealed the role ofangiotensinogen in essential hypertension.Cloning the actual human gene, after genetic analysis haspinpointed its location to a narrow chromosomal region is a wholeseparate challenge. That pinpoint region, Lander points out, maybe large enough to contain 500 genes.For now, he suggests, "the most powerful strategy may prove to belinkage disequilibrium mapping in a genetically isolatedpopulation, such as that of the Finns, founded 100 generations ago,or the Mennonites, 10 generations back."The Human Genome Project," Lander predicts "promises to makea tremendous contribution to the positional cloning of complextraits by eventually providing a complete catalog of all genes in arelevant region." n
-- David N. Leff Science Editor
(c) 1997 American Health Consultants. All rights reserved.