Longtime radical activist and biologist Barry Commoner has just published a preposterous article in the February issue of Harper's that purports to call into question the "foundation of genetic engineering." Commoner claims to have uncovered a scandal at the heart of biology and biotechnology that the leading researchers and greedy corporations are keeping hidden from an unwitting public.
What is this dark secret? That a gene can specify more than one protein through alternative splicing. According to Commoner, alternative splicing contradicts what biologists often call the Central Dogma in which a DNA gene is transcribed into messenger RNA (mRNA) that is then translated into a string of amino acids that folds to form proteins. Proteins are the molecular machines at the heart of most processes in living cells.
The Human Genome Project's findings highlighted the importance of alternative splicing, according to Commoner. Recall that last year two teams, one private led by Craig Venter and the other government-run led by Francis Collins, jointly announced that they had decoded the human genome, the 3.2 billion DNA base pair recipe for making human beings. Preliminary analyses indicate that the recipe for making a human being is composed of 30,000 to 40,000 genes. But it is estimated that there are at least 100,000 different proteins in human bodies, so that means that some genes must be recipes for more than one protein.
To understand how this happens, please bear with me for a short lesson in molecular biology. Genes are not neat orderly sequences of DNA bases that are simply read off one by one. Instead, the DNA bases that make up a gene--called exons--are often interrupted by other DNA bases called introns that have nothing to do with the gene. In the first step in transcribing DNA into RNA, both exons and introns are read off to produce pre-messenger RNA. To get the proper recipe for a protein, the introns must be removed. That feat is accomplished by an editing machine composed of RNA and protein called the spliceosome that removes the introns and splices together the exons into mature messenger RNA that now embodies the proper recipe for a specific protein.
Alternative splicing occurs when regulatory elements in the genome perhaps tell the spliceosome to treat some introns as exons or some exons as introns, thus changing the protein recipe. As University of Georgia biologist Wayne Parrott notes, to a certain extent this is all a matter of nomenclature -- is it the "same" gene that is specifying different proteins or are they really different genes that happen to share overlapping DNA sequences? The fact is "there is still one DNA sequence per protein," says Parrott.
In this quibble over nomenclature, Barry Commoner sees far darker implications. "The discovery of alternative splicing...bluntly contradicts the precept that motivated the genome project," writes Commoner. "It nullifies the exclusiveness of the gene's hold on the molecular process of inheritance and disproves that by counting genes one can specify the array of proteins that define the scope of human inheritance. The gene's effect on inheritance thus cannot be predicted simply from its nucleotide sequence...."
First, the genome project is not just about "counting" genes. Second, researchers must find all the genes in order to have any idea what protein(s) they might express, alternatively spliced or not. The human genome (the set of all genes) is the gateway to the human proteome (the set of all proteins). Commoner wants to claim that researchers have ignored what he believes is the crucial role that protein and RNA molecular "machines" play in expressing inheritable traits.
But that is sheer nonsense as a quick glance at any freshman-level biology textbook shows. For example, Molecular Cell Biology 4th Edition by Lodish et alia, states: "A more accurate way of representing the relationship between the synthesis of DNA, RNA, and proteins in all cells would look like [the figure below], indicating that special proteins catalyze the synthesis of both RNA and DNA."
Next, Commoner asserts that alternative splicing "destroys the theoretical foundation of a multibillion-dollar industry, the genetic engineering of food crops." Because of alternative splicing, genetically enhanced crops "represent a massive uncontrolled experiment whose outcome is inherently unpredictable. The results could be catastrophic." Commoner is canny enough to know that his unusual views on protein synthesis and alternative splicing would be simply ignored by modern researchers unless he can garner attention for them. So he links his views to the activist campaign against plant biotechnology and publishes them in a popular magazine. (One may be forgiven for thinking that Commoner's article appeared in Harper's because no reputable peer-reviewed scientific journal would be likely to publish it.)
Does alternative splicing really destroy the biotech industry and threaten human health?
Commoner begins his article by citing problems with gene-transfer experiments in animals and with cloning (in which no genes are transferred). He then blithely notes in the same paragraph that some non-plant genes have been transferred into commercial crops. Commoner clearly hopes that readers will make the leap of assuming that if there are problems in animal biotech experiments then there must also be problems in commercial plant biotech. He has to use this sneaky rhetorical technique because there simply aren't any "failures" in commercial plant biotechnology he can cite.
Commoner's chief claim is that adding genes for traits such as bacterial pesticides to crops such as corn and soybeans is dangerous because in an "alien genetic environment, alternative splicing of the bacterial gene might give rise to multiple variants of the intended protein -- or even to proteins bearing little structural relationship to the original one, with unpredictable effects on ecosystems and human health." He argues that the plant protein complexes that guide RNA and protein synthesis will not be able to properly transcribe and translate bacterial genes, so unintended proteins will be produced.
Has Commoner any evidence that this occurs frequently or at all in commercial biotech crops? If he does, he doesn't cite it in Harper's. In fact, plant biotechnologists are quite adept at adapting bacterial genes to plant genomes so that their protein synthesis machinery functions normally. If it didn't function normally, the plants wouldn't produce the traits such as pest resistance that commercial biotech companies are so successfully selling to farmers around the world. To produce a commercial biotech crop variety, biotechnologists typically begin by producing hundreds and thousands of plants in which they are trying to insert a particular gene. Over the years they grow and select the ones in which the trait they are seeking -- say, pest resistance--is stable. Only after years of testing and research will they commercialize the selected crop variety.
Commoner also claims, "The degree to which disruptions (caused by gene transfers) do occur in genetically modified crops is not known at present, because the biotechnology industry is not required to provide even the most basic information about the actual composition of the transgenic plants to the regulatory agencies."
"Outright false," says biologist Parrott. "Transgenics cannot be different from conventional varieties." Biotech crops must be "substantially equivalent" to conventional varieties before they can be marketed. In every case, biotech companies have submitted reams of information to the Food and Drug Administration (FDA) on things like nutrient profiles and feeding values before marketing genetically enhanced crops.
Conventional plant breeders don't need FDA approval to market their crops, even though they often involve far more genetic changes (through techniques like crossbreeding) than the single-gene transfers common in genetically enhanced crops. "Every single differently shaped leaf of lettuce, every different color of bell pepper, every new variety of citrus fruit, is the result of genetic mutations that produce different proteins which were noticed and then selected by conventional plant breeders," says Parrott. Yet no one worries about being poisoned by these far more massive genetic alterations in crops. Interestingly, scores of varieties of crops being grown today were produced through mutations induced by radiation and caustic chemicals in the 1940s and 1950s. No one knows what proteins these random genetic mutations produced, but people have been eating them for half a century without ill effects.
Parrott points out that plant genomes are filled with DNA fragments called retrotransposons that naturally jump randomly from one part of a plant's genome to another. These jumps occur billions of times every growing season. They often disrupt gene expression in plants and may well sometimes induce the production of novel proteins. But this is no cause for alarm, since people have been eating these crops with their jumping genomes for centuries. It is evident that such disruptions in plant genomes have an extremely low probability of producing any dangerous proteins.
Why is Commoner making claims that can be refuted simply by referring to college biology textbooks? Partly because he is still smarting from his intellectual defeat at the hands of James Watson and Francis Crick (the co-discoverers of the structure of DNA) in the 1950s and 1960s. At that time, Commoner, besides being a campaigner against aboveground nuclear testing, was one of the leading advocates of the theory that proteins carried inheritable traits. To some extent, Commoner appears to be trying to reinterpret alternative splicing as a way to rescue at least a portion of his old thesis. It turns out that Commoner may be a perfect example of Thomas Kuhn's contention in The Structure of Scientific Revolutions that old theories never die until old theorists do.