The Harvard University geneticist Richard Lewontin's book consists of review essays originally published in the New York Review of Books between 1981 and 1993, with epilogues to bring them up to date or to answer his critics. They range from developmental biology to intelligence tests, human sexual behaviour, cloning and the human genome. They are erudite, readable and they debunk a great deal of hype. They are also full of factual errors, prejudiced misjudgements and vital omissions. <P class=MsoNormal> As an early example of media hype, Lewontin quotes newspaper reports in 1899 that the biologist Jacques Loeb had created life and explained the Virgin Birth of Mary when he induced unfertilised sea urchin eggs to divide and develop. However, Lewontin fails to mention one important reason why the hype was misplaced: these eggs did not develop into adult sea urchins, but stopped short at the larval stage. Unfertilised frogs' eggs can also be induced to divide and develop, but development stops at an early tadpole stage, because cells grown from unfertilised eggs contain only one set of chromosomes and lack the essential complementary set contributed by the sperm. The same fundamental problem would have beset the Virgin Birth.
<P class=MsoNormal> Lewontin condemns as biological determinism all attempts to link human intelligence and behaviour to race, class, physical characteristics or genes. He pokes fun at phrenology, the pseudoscience much in vogue in the 19th century that linked the size and shape of people's heads to their intelligence and character. Lewontin comments: "Since acquisitiveness is a product of a material organ, the brain, then highly developed acquisitiveness should be the manifestation of the enlargement of one region of the brain. On the not unreasonable (although factually incorrect) assumption that the skull will bulge a bit to accommodate a bulge in the cerebral hemisphere, we might well expect an enlarged 'bump of acquisitiveness' among the more successful members of the Exchange, not to mention Jews in general."
<P class=MsoNormal> This tasteless joke ill matches his injunction against racial determinism. Recent studies have shown a correlation between IQ scores and brain size, but it is weak, and one of the most intelligent men I know also has the smallest head.
<P class=MsoNormal> As a more recent form of biological determinism Lewontin attacks intelligence tests. He accuses the tests' practitioners of claiming that they "measure a single underlying innate thing which does not develop during the lifetime of the individual, but merely becomes crystallised by education". "It is the ability to learn, a fixed feature immanent to different degrees in every fertilised egg." "This biological determinism is the conjunction of political necessity with an ideologically formed view of nature, both of which arise out of the bourgeois revolution of the 17th and 18th centuries."
<P class=MsoNormal> These strictures may have been justified in the past, but are they still true today? To find out, I turned to some of the recent literature on IQs (articles and books by U. Neisser, C. Jencks, N. J. Mackintosh, I. J. Deary, among others), and was surprised to find it open-minded, undogmatic and thoughtful. For example, studies of identical and non-identical twins brought up either together or separately have shown that no more than half the variation in IQ scores of different groups are genetically determined. Identical twins can actually differ in intelligence because of different blood supply in the womb. Again, the mean IQ scores of African-American children are typically 15 points below those of white American children, but investigators have found no evidence of their being genetically determined; they confess that they cannot explain them. Another example of the limitations of IQ tests comes from observations of Chinese- and Japanese-American children. Their average IQs scored slightly below the general average of 100, at about 97 or 98, yet the later occupational success of these children has turned out to be equivalent to those of white American children with an average IQ of 120. The authors can offer no explanation for this anomaly.
<P class=MsoNormal> Neither Lewontin nor his wordy critics ask the crucial practical question: what predictive value do IQ scores of children have for their future performance?
<P class=MsoNormal> The correspondence between the IQs of the same groups of children at different ages is expressed statistically as a correlation coefficient. Coefficients of one imply complete correspondence, of zero none. Intermediate values signify different degrees of scatter of individual IQs from the mean; the lower the coefficient the greater the scatter. In Britain, the correlation coefficient between the IQs of groups of 11-year-olds and the results of school examination of the same groups at 16 is about 0.5, which means that there is a significant correlation, but with so large a scatter of individual values from the mean of the groups that they have little predictive value for the individual child: they should never have been used to assign 11-year-olds to either grammar or secondary modern schools. In America, a sample of children was tested repeatedly between the ages of three and 16. IQs at three and six showed only negligible correlation with later educational attainments. IQs between eight and 16 showed correlation coefficients of between 0.45 and 0.50, in agreement with British studies. About two-thirds of the differences in IQ scores were found to arise from causes other than family background. Children's IQ scores measured between the ages of 12 and 17 affected their future occupational status, but 60 to 80 per cent of that effect arose because most children with higher IQs also sought more education; those who failed to do this did no better as adults than children with lower IQs.
<P class=MsoNormal> The authors of one of the studies conclude that "even if IQ scores were entirely explained by genes, which they almost certainly are not, the genes that do affect IQ scores have rather modest effects on occupational success, even though mental ability differences remain remarkably unchanged from childhood to old age". Ian Deary and others confirmed this when they compared the IQ scores at age 11 of Scottish children born in 1921 with scores of the identical tests administered to 101 of their survivors at the age of 77. I could find no biological determinism in any of this work.
<P class=MsoNormal> Much that has been discovered about the brain's anatomical development is consistent with the psychologists' finding that nature and nurture are closely interwoven in the development of mental abilities. The figure on page 25 shows sections through the same area of the brain at birth and at six years of age. The main nerve fibres seen at six have been laid down already before birth because they are genetically determined, but the branches which connect them to other nerves have multiplied. The growth of the branches and their different functions are also genetically controlled, but the connections they make with other nerves, or the strength of these connections, seem to depend at least in part on external stimuli which include learning. Lewontin gives an interesting summary of the widely accepted theory that in the absence of stimuli, these branches make a multitude of random connections and that later only those survive that are strengthened by external stimuli. We do not yet know exactly what controls the connections, but the vital importance of external stimuli and nutrition for the development, strength and maintenance of the right connections is well established.
<P class=MsoNormal> Lewontin's black sheep include molecular biology, the human genome and gene therapy. He attacks Max DelbrŸck, the physicist who pioneered the immensely fruitful genetics of bacterial viruses, and whom he wrongly describes as a pupil of Schršdinger. He brands the phage group, the enthusiastic band of young people whom DelbrŸck assembled around himself, as "a political apparatus", and molecular biology as "a religion", which is absurd. He is right when he ridicules the molecular biologist Walter Gilbert's vision of the human genome as its "holy grail" that will change our philosophic understanding of ourselves, but he then continues: "It is a sure sign of their alienation from revealed religion that a scientific community with a high concentration of Eastern European Jews and atheists has chosen for its central metaphor the most mystery-laden object of medieval Christianity." Remarks about people's race, religion and origins have no place in a book about science.
<P class=MsoNormal> What does the human genome really tell us, now that it is almost completely known? Genes' only function is to code for the synthesis of the proteins and nucleic acids that make up most of our bodies and perform nearly all their chemical functions. It takes about 31,000 genes to make a human, fewer than the 100,000 people guessed earlier. This implies that it takes at least that number of different proteins to make a human; probably even more, because the same genes can be spliced in different ways to make more than one protein. The mouse genome contains about the same number of genes as the human, and 90 per cent of mouse genes are the same as human ones. I agree with Lewontin that this does not increase our philosophical understanding of what makes us different from mice. It takes as many as 19,000 genes to make a nematode worm that contains only a thousand cells and is only a millimetre long; 14,000 genes to make a fruitfly and 26,000 to make thale cress ( Arabidopsis thaliana ), a weed of the mustard family. Judging by the number of genes, plant organisms must be almost as complex on the molecular scale as those of mammals. A yeast cell needs about 6,000 genes and the humble coli bacterium 5,416, which makes one realise the staggering complexity of even the simplest single-cell organisms. The functions of sizeable proportions of the genes of all these organisms are still unknown. Two-fifths of the worm's, about half the fruitfly's, and 15 per cent of the weed's genes have human homologues, which testifies to the remarkable unity of life on the molecular scale.
<P class=MsoNormal> David Baltimore writes in the issue of Nature (15 February 2001) announcing the completion of the human genome: "Understanding what does give us our complexity - our enormous behavioural repertoire, ability to produce conscious action, remarkable physical coordination (shared with other vertebrates), precisely tuned alterations in response to external variations of the environment, learning, memory... need I go on? - remains a challenge for the future.
<P class=MsoNormal> "We wait with bated breath to see the chimpanzee genome. But knowing now how few genes humans have, I wonder if we will learn much about the origins of speech, the elaboration of the frontal lobes and the opposable thumb, the advent of upright posture or the sources of abstract reasoning ability, from a simple genomic comparison of human and chimp. It seems likely that these features and abilities have mainly come from subtle changes... that are not now easily visible to our computers and will require much more experimental study to tease out. Another half-century of work by armies of biologists may be needed before this key step of evolution is fully elucidated."
<P class=MsoNormal> Only a small fraction of the diseases that affect us are inherited, or are due to inherited susceptibilities. Susceptibility genes for breast or colon cancer or Alzheimer's disease account for less than 3 per cent of all cases, and no preventive measures against them have so far proved safe and effective. The greatest medical advances from the human genome are expected in diagnosis of inherited conditions, in our knowledge of their prevalence in different populations, and of inherited predispositions in patients' response to drugs.
<P class=MsoNormal> The genes for the most common inherited diseases that are due to mutations in single genes were already identified before scientists thought of sequencing the entire genome, but its analysis has already led to the identification of the genes for about 30 additional congenital diseases, among them one responsible for susceptibility to breast cancer, in addition to the ones already known, and another that causes a devastating lung disease. The causes of these diseases are still unknown, but knowledge of the gene for a disease can lead quickly to discovery of its molecular mechanism, which is the first essential stop towards a therapy.
<P class=MsoNormal> Unfortunately not a sure step. Sickle-cell anaemia is a hereditary disease that affects inhabitants of malarial regions of the world. It is a disease of haemoglobin, the protein of the red blood cell, a molecule on which I spent most of my life. As a result, we know in atomic detail the alteration of atomic structure caused by the sickle-cell mutation and understand exactly why that alteration causes anaemia, yet all our efforts to find a drug that would remedy this condition have failed. On the other hand, identification of the responsible gene has led to a successful therapy for haemophilia, which is caused by mutations in the genes for proteins required for blood clotting. Its most common form used to be treated by regular injections of the healthy protein isolated from donated blood, but there have been tragic cases of infection with HIV and hepatitis virus; knowledge of the gene has enabled scientists to manufacture the protein by recombinant DNA technology without any risk of infection.
<P class=MsoNormal> In principle, many commonly inherited diseases should become curable by gene therapy either by introducing the healthy gene into the fertilised egg or by administering it to the patient; Lewontin condemns introduction into the egg as too risky and the recent creation of the first genetically modified monkey bears this out. In order to know whether the gene they introduced had taken, scientists chose one that codes for a fluorescent protein made in a jellyfish, believing that successful introduction would make the monkey fluoresce. They coupled the gene to a harmless virus which they injected into 222 monkey eggs; they then fertilised them with monkey sperm and incubated them. They implanted two each of 40 early embryos into the wombs of 20 surrogate mother monkeys.
<P class=MsoNormal> Only five of these resulted in pregnancies, one of them of twins. Only three monkey babies were born alive and only one carried the jellyfish gene, but the monkey does not fluoresce, because the gene, though spliced into the monkey's chromosomes, fails to express the protein for which it codes. This kind of gene transfer has now been practised in mice for several years with no greater success rate. It would be criminal to try it in humans and I can see no good reason why it needed to be tried on our nearest relative, an intelligent monkey. Human cloning carries the same risks.
<P class=MsoNormal> After many failures, gene therapy in patients succeeded last year for the first time. A French team has managed to cure a fatal inherited immune-deficiency disease. They took bone marrow from two baby boys who suffered from the disease and infected it with a harmless virus that carried the healthy gene. They then re-injected the infected marrow into the boys' bones. It restored complete immune function that was still active nine months later.
<P class=MsoNormal> Another hopeful development has sprung from the identification, before the completion of the genome, of the gene for muscular dystrophy. Injection of fragments of the gene into a muscle of dystrophic mice has restored the muscle's normal function. This might, in due course, lead to successful treatment of human patients; both these successes are really great news. On the other hand, it has so far not been possible to cure one of the most common inherited disorders among Western Europeans, cystic fibrosis, which affects the lining of the lungs and airways. Patients who inhaled the healthy gene, carried either by a virus or contained in tiny fat droplets, derived little benefit; the treatment temporarily cured the lining of their noses, but it had no effect on their lungs. It is often proving extremely difficult to incorporate a gene in the correct place of patients' chromosomes and express the required protein in sufficient quantity in the right tissues.
<P class=MsoNormal> Before the completion of the Human Genome Project, identification of some of these genes required truly heroic efforts. The search for the Huntington's disease gene occupied up to a hundred people for about ten years. The same work could now be accomplished by few people in a fraction of the time. This is one of the Human Genome Project's important medical benefits. Another may be the rapid identification of promising new drug targets against diseases ranging from high blood pressure to a variety of cancers.
<P class=MsoNormal> The title of Lewontin's book is misleading. The human genome is proving no illusion either medically or biologically. It has not been, as Lewontin alleges, initiated by financially interested scientists in order to extract money from the public purse for their own pockets. Some commercial companies have stepped in, but as John Sulston, the head of the Sanger Centre near Cambridge where a large part of the genome was sequenced, has written, the hundreds of devoted people who have contributed to it and made it speedy, efficient and effective have done this not for wealth, nor for unusual recognition, but to benefit mankind. I would add, also out of curiosity for the working of nature. It is, by any measure, a magnificent achievement. The man who invented the chemical method used for sequencing the genome was Fred Sanger, not Allan Maxam and Walter Gilbert whom Lewontin wrongly credits with it. Sanger never patented his method, but it won him his second Nobel prize.
<P class=MsoNormal> The greatest threats to health come not from our genes, but from infectious diseases. Aids remains incurable and threatens to wipe out entire populations. In many countries of Africa and Asia abuse of antibiotics has bred tubercle bacilli that are resistant to all known drugs. Travellers import them. They may make us once more helpless against the "white plague" of tuberculosis that killed Chekhov, the Bront‘s, Keats, Chopin, D. H. Lawrence, George Orwell and many thousands of others in the prime of life.
<P class=MsoNormal>
<P class=MsoNormal> Max Perutz, OM, Nobel laureate, is at the Medical Research Council Laboratory of Molecular Biology, Cambridge.
<P class=MsoNormal>