No one who reads the newspapers or scientific journals can have missed the fact that this is the 50th anniversary of the publication of the correct three-dimensional structure of DNA. That structure, a double helix of two chains of nucleotides, has become a popular icon and the very phrase, “double helix” has been spoken and written so often as to become part of ordinary discourse.

The fact that genes were composed of DNA had already been established nine years before the publication of Watson and Crick’s paper on its structure, and the chemical, as opposed to the spatial, configuration of DNA was also well known before 1953. Yet, despite the obvious importance of DNA in understanding the molecular details of both heredity and development, it was not until after the publication of the proposed double helical structure that DNA started increasingly to occupy the interest of biologists and finally became the focus of the study of genetics and development. The last fifty years have seen the reorganization of most of biology around DNA as the central molecule of heredity, development, cell function and evolution. Nor is this reorganization only a reorientation of experiment. It informs the entire structure of explanation of living processes and has become the center of the general narrative of life and its evolution. An entire ideology has been created in which DNA is the “Secret of Life”, the “Master Molecule”, the “Holy Grail” of biology, a narrative in which we are “lumbering robots created, body and mind” by our DNA. This ideology has implications, not only for our understanding of biology, but for our attempts to manipulate and control biological processes in the interests of human health and welfare, and for the situation of the rest of the living world.

The first step in building the claim for the dominance of DNA over all living processes has been the assignment of two special properties to DNA, properties that are asserted over and over again, not only in popular expositions but in textbooks. On the one hand, it is said that DNA is self-replicating; on the other, that DNA makes proteins, the molecular building blocks of cells. But both of these assertions are false — and what is sodisturbing is that every biologist knows they are false.

First, DNA is not self replicating. It is manufactured out of small molecular bits and pieces by an elaborate cell machinery made up of proteins. If DNA is put in the presence of all the pieces that will be assembled into new DNA, but without the protein machinery, nothing happens. What actually happens is that the already present DNA is copied by the cellular machinery so that new DNA strands are replicas of the old ones. The process is analogous to the production of copies of a document by an office copying machine, a process that would never be described as “self-replication”. In fact, many errors are made in the DNA copying process; there is protein proofreading machinery devoted to comparing the newly manufactured strands to the old ones and correcting the errors. An office copier that made such mistakes would soon be discarded.

Second, DNA does not make anything, certainly not proteins. New proteins are made by a protein synthesis machinery that is itself made up of proteins. The role of the DNA is to provide a specification of the serial order of amino acids that are to be strung together by the synthetic machinery. But this string of amino acids is not yet a protein. To become a protein with physiological and structural functions, it must be folded into a three dimensional configuration that is partly a function of the amino acid sequence, but is also determined by the cellular environment and by special processing proteins that, among other things, may cut out parts of the amino acid chain and splice what remains back together again.

The other function of DNA is to provide a set of “on-off” switches that are responsive to cellular conditions so that different cells at different times will produce different proteins. When the conditions of the cell set a switch associated with a particular gene to the “on” position, then the protein manufacturing machinery of the cell will read that gene. Otherwise the cell will ignore it.

In this mechanical description of the relation of DNA to the rest of the cellular machinery there is no “master molecule”, no “secret of life.” The DNA is an archive of information about amino acid sequences to which the synthetic machinery of the cell needs to refer when a new protein molecule is to be produced. When and where in the organism that information is read depends on the physiological state of the cells. An organism cannot develop without its DNA, but it cannot develop without its already existing protein machinery (unless it is a parasite like a virus that has no synthetic power of its own but gets a free ride on its host’s protein machinery).

The unjustified claim for special autonomous powers of DNA is the prelude to the next step in building a picture of a DNA-dominated world. This picture is simply the molecular version of a biological determinism that has dominated explanations of the properties of organisms, and especially of humans, since the nineteenth century. Differences in temperament, talents, social status, wealth, and power were all said to reside “in the blood.” The physical manifestations of these claimed hereditary differences could be seen by criminal and racial anthropologists in the shapes of noses and heads and the color of skins. With the rise of Mendelian genetics, genes were substituted for blood in the explanations, but they remained, for the fifty years of genetics, merely formal entities with no concrete description beyond the fact that they were some bit of a chromosome. The discovery that DNA is the material of the gene, and the subsequent determination of the correspondence between nucleotide sequences of genes and amino acid sequences of proteins, then provided a concrete molecular basis for a total scheme of explanation of the organism. The fact that organisms are built primarily of proteins and that DNA carries the archive of information for the amino acid sequence of the proteins gave an immense weight to the conclusion that the organism as a whole is coded in its DNA. A manifestation of this view is the claim made, at a symposium in commemoration of the 100th anniversary of the death of Darwin, by a founder of the molecular biology of the gene: that if he were given the DNA sequence of an organism and a large enough computer, he could compute the organism. One is reminded of Archimedes’ claim that, given a long enough lever and a place to stand, he could move the earth. But while Archimedes may have at least been right in principle, the molecular biologist was not. An organism cannot be computed from its DNA because the organism does not compute itself from its own DNA.

It is a basic principle of biology, known to all biologists but ignored by most of them as inconvenient, that the development of an organism is the unique consequence of its genes and the temporal sequence of environments in which it developed. The current fascination of developmental genetics is with the way in which information from different genes enters into the formation of the major features of an organism. How does the front end of the animal become differentiated from the back end? Why does the egg of a horse develop into an animal with four legs while the egg of a bird produces an organism with two legs and two wings, and the egg of a butterfly results in an animal with six legs and two sets of wings? This concentration on the major differences and similarities between different species has resulted in a genetically determinist view of development that ignores the actual variation among individuals. There is an immense experimental literature in plants and animals showing that individuals of the same genetic constitution differ widely from each other in physical characteristics if they develop in different environments. Moreover, the relative ranking in some physical trait of individuals of different genotypes changes from environment to environment. Thus, a genetic type that is the fastest growing at one temperature may be the slowest at another. But even genes and environment together do not determine the organism. All “symmetrical” organisms show a fluctuating asymmetry between their two sides and the variation between left and right sides is often as great as the difference between individuals. For example, the fingerprint pattern on the left and right hands of a human individual are not identical; on some fingers, they may be extremely dissimilar. This variation is the manifestation of random growth differences that arise from small differences in the local tissue and cell conditions in different parts of the body, and from the fact that there is random variation in the number of copies of particular molecules in different cells. A consequence is that two individuals with identical genes and identical environments will not develop identically. If we want to understand human variation, we need to ask far more subtle and complex questions than is the rule in DNA-dominated biology.

The other side of the movement of DNA to the center of attention in biology has been the development of tools for the automated reading of DNA sequences, for the laboratory replication and alteration of DNA sequences and for the insertion of pieces of DNA into an organism’s genome. Taken together, these techniques provide the power to manipulate an organism’s DNA to order. The three obvious implications of this power are in the detection and possible treatment of diseases, the use of organisms as productive machines for the manufacture of specific biological molecules, and the breeding of agricultural species with novel properties.

The Human Genome Project has been largely justified by the promise that it will now be possible to locate genes that cause human disease by comparing the DNA sequences of affected and unaffected individuals. Once the nucleotide difference has been established, that difference can be used as a diagnostic criterion, as a predictor of a future onset of the disease, and as a basis for a cure by gene replacement therapy. It is undoubtedly true that some fraction of human ill health is a consequence of deleterious mutations. However, while family studies can strongly suggest that a disease is being inherited as a single Mendelian gene difference, the determination that it is a consequence of mutation of a particular gene is not a trivial problem. A blind search for a genetic difference that is common to all affected individuals is impractical given that, on the average, any two humans differ from each other at 3 million nucleotide sites. On the other hand, if the biochemistry of the disease is sufficiently well understood, it may be that a few candidate genes can be singled out for investigation. Alternatively, studies of the pattern of inheritance may show that the disorder is inherited coordinately with an associated gene of known location in the genome, greatly narrowing down the search for the DNA variation implicated in the disease.

As in all other species, for any given gene, human mutations with deleterious effects almost always occur in low frequency. Hence specific genetic diseases are rare. Even in the aggregate, genes do not account for most of human ill health. Given the cost and expenditure of energy that would be required to locate, diagnose and genetically repair any single disease, there is no realistic prospect of such genetic fixes as a general approach for this class of diseases. There are exceptions, such as sickle cell anemia and conditions associated with other abnormal hemoglobins, in which a non-negligible fraction of a population may be affected, so that these might be considered as candidates for gene therapy. But for most disease that represents a substantial fraction of ill health and for which some evidence of genetic influence has been found, the relation between disease and DNA is much more complex and ambiguous. Claims for the discovery of “genes for” schizophrenia and bipolar syndrome have repeatedly been made and retracted. It is generally accepted that cancer is a consequence of mutations in a variety of genes related to the control of cell division, but even in the strongest individual case, the breast cancer-inducing BRCA1 mutations, only about 5% of such cancers are linked to these specific mutations.

Up to the present we do not have a single case of a successful cure for a disease by means of gene therapy. All successful interventions, whether in genetically simple disorders like phenylketonuria or in complex cases like diabetes, have been at the level of biochemistry and were in place well before anything was known about DNA. Of course, a successful gene therapy for some disease may be produced in the future, but the claim that the manipulation of DNA is the path to general health is unfounded. In fact, on a world scale, most ill-health and premature death is caused by a combination of infectious disease and undernourishment — factors which genetic manipulation will never solve.

The second implication, the possibility of using genetically transformed organisms as factories for the commercial production of biologically useful molecules, has been realized in practice. The most famous case, the mass production of human insulin by bacteria, is particularly instructive. Insulin for diabetics was originally extracted from cow and pig pancreases. This molecule, however, differed in a couple of amino acids from human insulin. Recently, the DNA coding sequence for human insulin has been inserted into bacteria, which are then grown in large fermenters; a protein with the amino acid sequence of human insulin is extracted from the liquid culture medium. But amino acid sequence does not determine the shape of a protein. The first proteins harvested through this process, though they possessed the correct amino acid sequence, were physiologically inactive. The bacterial cell had folded the protein incorrectly.

A physiologically active molecule was finally produced by unfolding the bacterially produced protein and refolding it under conditions that are a trade secret known only to the manufacturer, Eli Lilly. This success, however, has a severely negative consequence. For some diabetics this “human” insulin produces the symptoms of insulin shock, including loss of consciousness. Whether this effect is caused by a manufacturing impurity, or because the insulin is not folded in the same way as in the human pancreas, or because the molecule is simply too physiologically active to be taken in large discrete doses rather than internal, continuously released amounts calibrated by a normal metabolism, is unknown.
The problem is that Eli Lilly, which holds the patent on the extraction of insulin from animal pancreases, no longer produces pig or cow insulin. Hypersensitive diabetics for whom Eli Lilly’s standard treatment is dangerous no longer have an easily obtainable alternative supply. The most widely known and contentious application of DNA technology to production is in agriculture. The introduction of DNA sequences derived from widely divergent species into agricultural varieties has resulted in a struggle of immense proportions in both North America and Europe. The proximate purpose of the creation of varieties with DNA introduced from other kinds of organisms is to produce agricultural crops with novel features that cannot be obtained by the usual methods of selection because the relevant genes are not present in the agricultural species. The benefits to farmers, consumers and commercial seed producers vary considerably from case to case, although in every case the ultimate goal of the commercial breeder is increased profit and the protection of their property rights. There are four cases to be distinguished. First there is the introduction of pest and disease resistance, as in the introduction of the BT protein from Bacillus thuringiensis into maize. This is intended to reduce the labor, chemicals and machinery needed by the farmer for pest control. Some of the cost reduction is lost in the higher price of the commercial seed, but saving labor is important to farmers. Second, there is creation of varieties that are resistant to herbicides used to control weeds. The best-known examples are the Roundup Ready varieties produced by Monsanto, designed to coerce farmers into purchasing Monsanto’s general herbicide (Roundup) as well as their seed. The supposed advantage to the farmer is a reduction in machinery and labor involved in tillage, but again the cost saving is reduced by the increased price of seed. The third case is the pure protection of property rights of the seed producers with no benefit to farmers or consumers. The most infamous example is the attempted introduction of “Terminator” technology by the Delta Pine and Land Company, which was later purchased by Monsanto. Terminator seed varieties will germinate and produce sterile crops, thus forcing farmers to purchase commercial seed anew every year. (It should be noted that this technology, of no advantage to farmers or consumers, was produced in cooperation with the U.S. Department of Agriculture). The fourth case is the introduction into mass produced field crops of DNA coding for particular compounds normally only produced by specialty species. This technology has the potential to destroy much of the economy of Third World countries that are dependent on the export of agriculturally produced commodities. An example is the transfer into rape seed, a widely grown crop in North America, of the DNA coding for palmitic acid oils that are used in industrial processes. Normally these oils are extracted from oil palm seed grown in Southeast Asia.

While much of the opposition to transgenic agriculture has been based on the “unnaturalness” of the process, this objection misses the point. No agricultural variety is ‘natural’,but is the product of centuries of gradual, cumulative genetic modification from its wild ancestors to produce varieties that are utterly different from the ancestral forms. Moreover, crosses between different species have been a standard method of plant breeding for more than a century. The real issue is that DNA technology provides a powerful tool for the control of agricultural production by monopolistic producers of the inputs into agriculture with no ultimate advantage either to farmers or consumers and with the possibility of destroying entire national agricultural economies. All of the elements that characterize the era of DNA have in common an underlying simplistic view of living organisms. By concentrating in practice and in theory on the properties and functions of a single molecule, biologists, both in their professional work and in their public statements, reduce the extraordinary complexity of life processes to the structure and metabolism of DNA. This emphasis ignores the intricate and multiple ways in which organisms are built and function. The intricacy is a consequence of the structural and metabolic functions of proteins and the interactions of those proteins with each other, with other molecules, and with the environment in the course of development.

Moreover, for human life, no account at all is taken of the role of social and economic processes in determining health and life activities and molding the processes of industrial and agricultural production. We cannot understand our size, shape and internal functioning except by a detailed understanding of the extremely complex web of interactions among the various molecules which form the body in concert with influences exerted by our environments. We cannot understand the origin and development of our mental states except by an understanding of the map of nervous connections and how that map is influenced by experience. We cannot understand why agricultural technology develops in particular directions if we do not understand the social, political and economic interactions that drive technological innovation. The bottom line is that life in all its manifestations is complex and messy and cannot be understood or influenced by concentrating attention on a particular molecule of rather restricted function.

***

Richard C. Lewontin is an evolutionary geneticist, philosopher of science, and social critic. An early pioneer in the development of molecular population genetics, his works include Biology as Ideology, The Triple Helix: Gene, Organism, and Environment, and Not in Our Genes, co-authored with Steven Rose and Leon Kamin. He is Alexander Agassiz Research Professor at Harvard University, and regularly writes for the New York Review of Books.