Gene hunters have used restriction fragment length polymorphisms(RFLPs _ pronounced "ree-flips") as place-names on genetic linkagemaps, pointing the way to gene sequences suspected of causing specificdiseases. RFLPs can also help forensic specialists match criminals tothe scene of their crime, or to exclude innocent suspects, on the basis oftheir unique genomic DNA configurations.RFLPs reflect the pattern of sites on the genome that can be cleaved bya specific restriction enzyme. These patterns vary not only from oneperson to another, but from generation to generation and millennium tomillennium, all the way back to the origin of life on earth.So the tool-kits of evolutionary geneticists also contain gene probes forfinding variations in RFLP-defined alleles _ mutant gene variants _back through the ages.One such evolutionary gene mapper is Stanford University geneticistLuca Cavalli-Sforza. He and his graduate student, Joanna LouiseMountain, have a paper in the current Proceedings of the NationalAcademy of Sciences (PNAS) titled "Inference of human evolutionthrough cladistic analysis of nuclear DNA restriction polymorphisms."Cladistics is the technique of tracking forms of life back throughevolution, using complex statistical inferences to construct two-dimensional "trees." These diagram how a given species branched offin times past from earlier and earlier branches, to the tree's "root" in afounder proto-species.Cavalli-Sforza and his co-author used RFLP variations to pinpointancestral similarities between modern Homo Sapiens and our collateralprimate relatives, the chimpanzees (Pan troglodytes) of Africa, gorillas(Gorilla gorilla), also African, and orangutans (Pongo pygmaeus) ofBorneo.For many years, Cavalli-Sforza has been studying two separate tribesof African pygmies in their native habitats including analysis of theirblood types. More recently, as reported in PNAS, he has comparedtheir RFLPs with those of six other existing human populations _Chinese, Japanese, Melanesians, Australians, New Guineans andEuropeans.From the first five groups, he obtained random DNA fragments andknown genes by culturing transformed lymphocytes. The Australianand New Guinean samples came from a scientist colleague, nowdeceased. He obtained most of the European material "fromopportunistic samples or literature data."Whereby hangs this tale.To begin with, Cavalli-Sforza told BioWorld Today, "We realized thatRFLPs lend themselves to one thing, which you could hardly dowithout them. This is, you can see if the same polymorphisms arepresent or not in other species, namely, chimps and other primates."He found that to do this required using polymorphisms with a smallnumber of alleles. "The simple reason for this," he explained, "is thatthese alleles are produced at a low mutation rate. If the rate were veryhigh, you would have many alleles, which means that their replacementis also faster. So you don't stand a good chance of finding them inprimates that separated from us millions of years ago, but are still theclosest animals to humans."He knew that the genes from the eight worldwide human populationsare polymorphic. But he almost never found the same polymorphism inother primates, so concluded that the human ones probably arose aftertheir separation from the primates, millions of years back.Chimps and humans diverged about 5 million years ago, gorillas 10million, orangutans, 15. That left more time to replace the one allelewith another one in the species further away from humans.Humans from all eight populations share an allele with chimpanzeesfor 62 of the 79 polymorphisms examined, the PNAS paper reports.In almost all of the populations that Cavalli-Sforza looked at, the mostfrequent polymorphisms are those in which the unmutated, ancestralallele is more frequent than the non-ancestral, mutated one, whichappeared more recently."This is true," he said, "of all populations except one _ the European.The reason for that," he allowed, "is a bias in our ascertainment ofpolymorphisms."Specifically, the European samples available contained polymorphismsin their RFLPs selected originally to find good markers for geneticlinkage, for positional cloning, for discovering genes responsible fordisease entities. "Molecular biologists," the Stanford geneticistobserved, "have inadvertently specialized in finding suchpolymorphisms in European populations. They use blood samples frompeople at hand in their labs, nearly all of whom are of European origin.Ideally these European samples, when there are only two alleles,display a 50 percent frequency of each, paternal and maternal. But notin the non-European populations, which were selected in other ways.So we generated a bias in the distribution, which we now see veryclearly."And that," he continued, "generates problems of data interpretation.For one, Europeans look as if they have never evolved at all."So we have to change that perspective, because it is extremely likelyto be wrong. This we can do by selecting in different ways. WithRFLPs we would have to start from scratch, but there is no need to dothat, because other markers are coming out, like microsatellites, to givea more correct perception."The PNAS paper reports estimated allelic mutation rates due tomutations that appeared on average 600,000 or 700,000 years ago. Theauthors arrived at that number "by constructing a cladistic tree in adifferent way from how it was done before."Using primates, they computed times at which nonancestral allelesoriginated, and found all of these polymorphisms in practically all theraces that they examined, in every corner of the world. That meansthey must have been there before humans started to spread from Africa,about 100,000 years ago. Their calculations suggested that mutationson average must have appeared 600,000 or 700,000 years _ 30,000 to35,000 human generations _ before that.Anatomically modern humans entered Europe about 40,000 years agofrom Western Asia, Cavalli-Sforza said. They appeared in Africa about100,000 years ago."We have now corrected the mistake that was made of having thegreatest variation in Europeans," he went on. "That will cancel theerroneous impression that Europeans have evolved less than theyactually have."He concluded: "Evolution doesn't make much money, but I think it hasa lot of wonderful problems. Biotechnologists who have an interest inscientific challenge, not only in technological challenge, will beattracted more and more by evolution." n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.