Food Fight

by Mckay Jenkins

  Meyers and his lab assistants begin sequencing genes by taking a leaf from a soybean plant, for example; freezing it in liquid nitrogen; then grinding it up to break up the leaf’s cells. Then, using a series of chemical processes that shatter cell walls, the team extracts the plant’s DNA or RNA, and loads them—in chains of fifty to two hundred nucleotides—into a three-inch “flow cell,” a kind of glass slide with eight hair-thin capillaries running through it. The slides are then snapped into an Illumina HiSeq gene sequencer.

  Depending on whether the team is working on fragments of DNA or short strands of RNA, Meyers’s team then needs to code their samples by pumping tiny chains of nucleotides through the flow cells, where they are “amplified” in a series of chain reactions that replicate the sequence millions of times.

  Fluorescently charged nucleotides are hit with laser light and photographed at four different wavelengths, showing up on a computer screen as innumerable pinhead-sized “spots” of DNA—the spots that, to my eye, looked like clusters of stars. The microscopic scale of these spots is hard to comprehend. The images are actually measured in microns—that is, each spot is a thousand times smaller than a millimeter—yet each one contains up to 5,000 copies of the same short fragment of genes. And in any series of images, there might be 200 million spots, each one containing a different version of the original piece of DNA or RNA. On each slide, the instruments used by Meyers (and his computers) can detect 1.2 billion sequences of DNA fragments.

  The process requires ten terabytes of computer space for each two-week run of the instrument, about what it would take to store the entire printed collection of the Library of Congress.

  “Sequencing a genome is like putting together War and Peace,” Meyers said. “It’s long, and content rich, but these machines can only get you words or paragraphs at a time. The machine generating shorter reads of DNA can find you all the times the word ‘the’ appears, so then it’s up to you to find all the places where ‘the’ appears in the book. But some machines can now give you sentences or entire paragraphs, so you can make much quicker sense of the whole.

  “This is like making millions of copies of War and Peace. The computer can overlap the fragmented text and look for places where Paragraph 1001 ends and Paragraph 1002 begins, and repeat that process millions of times, analyzing millions of short stretches of words, and incrementally assembling sentences, paragraphs, and whole chapters. Eventually this can help you figure out the sequence of the entire text.”

  Meyers’s team needs so many copies because, statistically, there is always the chance they might find a glitch—a missing paragraph (or ten paragraphs). The longer the strings of text, the easier it is to get recognizable, repeated, and accurate strings.

  “As biologists, we’re designing experiments to take advantage of these sequences to look into solving practical problems, like drought resistance,” Meyers said. “What genes are expressed when plants are distressed? We can compare stressed plants to non-stressed plants. These transcripts are like short-term memory. If the plant were a computer, we’d want to know: Under these conditions, what software has it been running? It’s like asking an iPhone, ‘What apps have you been using to solve the problems you faced today?’ If we looked at your iPhone and saw you’d been using the Lonely Planet app for New Delhi, or looking at the Urban Spoon app, from that information, from that software, we could tell what you were up to. It’s the same with plants, looking at their patterns of gene activity.”

  Tinkering with the Genetic Machine

  If understanding a genome is like learning a language, Meyers told me, then using genetics to work on plants is like using a repair manual to fix a faulty wire in a car.

  “A cell is like a big machine,” Meyers said. “Genes are the blueprints, and proteins are like the different parts of a car. So let’s say you have a warning light in your car, and that light means you have a loose wire. You touch the wire and it shocks you. It’s not healthy to have that loose wire. In the old days, you would go in there and add a plastic cap. The anti-GMO argument is that this is not the same car; it’s been modified by this protective cap. Even though the phenotype itself is better—it’s preventing you from getting shocked.

  “There are new methods that are more like this analogy: ‘Let’s go in there with a tool, and remove that wire altogether, or unscrew that bolt. Now you’ve got the car, but removed the offending wire, and you’ve taken the tool with you. Is that modified car something that you are unhappy with? It’s better than having that protective cap.”

  In the early days of genetic engineering, scientists used .22 caliber rounds—a “gene gun”—to literally blast gene sequences into a plant’s cells. Scientists would coat tiny particles of gold with thousands (or millions) of copies of a specific gene sequence they knew would confer a phenotype of interest (for example, making a plant resistant to an herbicide like glyphosate). They would then shoot the gold particles into a group of plant cells. The chances that the gene sequence would actually integrate into the genome in a way that was functional were less than a million to one, but do it enough times (or with enough copies of the gene) and you’ll eventually find one cell that lives and has an integrated and functional copy of the gene.

  The group of cells would then be treated with glyphosate (aka Roundup), and all of them would die except for a select few. Once a surviving cell was found to have absorbed the foreign gene sequence, it would be allowed to recover and encouraged to proliferate. These cells, now structurally Roundup resistant, could be regenerated by a process of tissue culture into an entire plant.

  Meyers doesn’t use the gene gun in his lab. Like most of his peers, he uses a bacterium (known as Agrobacterium tumefaciens) to do the work instead. With a natural ability to insert their own DNA into plant cells, this soil bacterium can be outfitted with gene sequences scientists want to see integrated into a plant’s own genome. Plants are dipped in a solution full of the bacteria, then covered up. The bacterium, a plant pathogen, is effective at infecting the plant and will find its way into the plant stem cells, transferring the foreign DNA into the plant genome just as it evolved to do, and creating a stable transgenic plant.

  In his laboratory’s “green vaults”—growth chambers that resemble walk-in food coolers—Meyers showed me plants maintained with high humidity and variable light and temperature. Inside were hundreds of examples of two plant species that have become the workhorses for plant geneticists: tobacco and Arabidopsis, the latter a member of the brassica family that includes things like broccoli.

  These plants are like lab mice for plants—they flower, they reproduce, they respond to stressors. And the research Meyers is doing is similar to the basic research biomedical researchers do with mice: both are trying to unlock the mysteries of the ways organisms function.

  The Arabidopsis genome is a fraction—perhaps 5 percent—the size of the corn genome. It is small enough to grow several plants to maturity in a single coffee cup, which makes it much easier (and cheaper) to work with than corn, which is happiest in a field, well separated from its neighbors. And since “core responses” are similar across all sorts of plants, scientists can play around with an Arabidopsis plant and its genome and extrapolate conclusions that will likely hold true for corn or soybeans.

  “Let’s say you take two varieties of Arabidopsis,” Meyers said. “One grows well in moist climate like Germany. One grows in dry climate like Utah. You can make a cross between them, then use the progeny and traits segregating in those progeny to map onto the chromosomes the loci controlling responses to these climates. You can then identify the locations in the genome that contribute genes important in the line for Utah for drought resistance that may be missing in the line from Germany.

  “Say that you’ve found gene X that confers the ability to survive in drought conditions. As a geneticist, to test that function and demonstrate causality, you may want to break gene X, to see what happe
ns when the plant normally happy in Utah loses the gene you think contributes to fitness under dry conditions. Or you can misregulate gene X, or modify key parts of the protein that gene X produces. With the resulting data, we can make insights into how gene X functions and confers the phenotype that had attributed to it through a standard genetic approach. For all this, your work is greatly facilitated by having a plant into which you can easily introduce the gene. With Arabidopsis, you have a generation time of eight weeks, lots of molecular tools and preexisting data, and great toolkit for molecular biology. It’s really easy to work with, particularly relative to most crops.”

  —

  THE WORLD INSIDE a plant laboratory is so ordered, so controlled, that scientists can be forgiven for their frustration at the screaming debate that has developed over the work they do. “Forty years ago when I was in school, we’d see farms using chemicals to prevent diseases, where the soil was sterile because of the methyl bromide used as a fungicide and biocide for strawberries,” Jim Carrington of the Danforth Center in St. Louis told me. “We would visit strawberry fields in California and say, ‘Wouldn’t it be great if there was a different way to do this, to use a plant’s own genetics to fight off diseases more effectively, so we wouldn’t have to douse them with fungicides—some of which do have adverse human health effects? And wouldn’t it be great if we had plants that help build the soil, rather than forcing us to lose carbon and other organic matter?’ We’d imagine we could improve plant genetics, and now some of that has come to fruition. We have low- (or no) till farming, doing all those things that fall under desired practices to promote sustainability. As a scientist who envisioned many of those things—to see all those aims co-opted by groups that are antagonistic to science—is really frustrating.”

  Scientists are trained to discuss data, not to make political arguments, Blake Meyers said. “We aren’t trained in arguments about whether you prefer organic or nonorganic food,” Meyers said. “If you tell me this gene is bad for you, we can have that discussion. Show me the data, I might say. But those aren’t the arguments that we are often having. It’s like a discussion about evolution versus intelligent design. There is no science supporting intelligent design, so to a scientist this debate will be fruitless, as it’s irrational. There’s no common ground between the rational scientist and the passionate believer. With GMOs, you get the same sort of situation.

  “On the public side, arguments against GMOs are often not grounded in science,” he said. “They may be based on ‘the way things should be,’ or ‘the way things used to be,’ or based on someone’s individual opinion. This can be very frustrating for scientists. Those are arguments that don’t fit the data. Show me an argument based on data, and we can have a reasonable or at least scientific discussion.”

  The frustration felt by Carrington and Meyers—that arguments mounted by anti-GMO skeptics are not based on data or are even “antagonistic to science”—is a common refrain that, to the ears of skeptics, sounds ideological in its own right. Scientists have a way of claiming that their field—or the scientific method itself—is somehow beyond reproach. But no matter the natural laws they seek to discover, scientists are people too—and are thus hamstrung by their own preconceptions, desires, and intellectual parameters. John Fagan is a molecular biologist whose professional change of heart on GMOs has made him a leading figure in the scientific debate. In 1994, Fagan became so concerned about the direction of genetic engineering he returned more than $613,000 in grant money to the National Institutes of Health. He quit an academic job to found the Global ID Group, which developed tools for testing genetically modified food, and now directs Earth Open Source, a leading anti-GMO clearinghouse whose publication GMO Myths and Truths has become a bible for GMO critics.

  “There are bona fide scientists who are doing genetic engineering of crops in one way or another and they really sincerely believe that there is not a problem,” Fagan said. “You can get all the way through your PhD without ever having a course in the philosophy of science, or a course that discusses the social or environmental impact of technology. The training is very focused on technical aspects of doing molecular biology in one area or another, and as a result you end up with scientists who are really experts in their own area but oftentimes do not understand the relationship between their work and the world out there. Many of them have the attitude that it would be compromising for them to think about or be involved in a debate about larger issues. They feel that they need to be scientific about what they do, and impartial, and true to numbers they get in a lab. There is some merit in that, but on the other side, to have only that perspective on whether a technology is commercialized on a large scale in the world is a very risky thing.”

  The safety of genetic engineering is not nearly as settled as the majority view claims it is, Fagan maintains.

  “The evolution of the debate on GMOs has really evolved over the last twenty years,” Fagan said. “Early on we were saying, ‘Based on what we know about how genes function, and what we know about the process, we feel this is a very sloppy and imprecise process that could lead to unexpected problems.’ Today, there is lots of evidence that says GMOs do not function the way we predicted, and there are a lot of unintended side effects that have come up.”

  The prevailing idea, the Central Dogma—that inserting a single gene into the DNA of another organism will cause a single, predictable change in a single protein, followed by a single change at the cellular level of the organism, a single change at the tissue and organ levels, and a single change at the level of the plant as a whole—fails to recognize the complexity and interconnectedness of the many components of living organisms, according to Fagan.

  It is now thought that most genes encode not just one protein but two, three, four, or more, and that the regulatory sequences associated with one gene can influence the expression of neighboring genes. When a new gene is inserted into the DNA of an organism, that gene is likely to influence the expression not just of one gene, but several. Likewise when the newly inserted gene is expressed as a protein, that protein will not have just a single effect, but several. It will influence multiple cellular processes and, subsequently, multiple processes in tissues, organs, and the entire plant. In other words, instead of a single, predictable effect, the insertion of a single gene can result in multiple effects, which can themselves affect many other processes, from the cellular level on up. The more effects, the more unpredictability.

  “There are spatial and temporal aspects of this,” Richard Manshardt, a plant virologist who helped develop the GM papaya in Hawaii, told me. “The idea that a gene occupies a particular location, and makes a particular protein, and that protein has a single function, is long outdated. Genes are complicated, and they can interact with different parts of the genome in different ways and at different times. It’s a much more dynamic system, and this is even before RNA. This is just in the coding sequence of DNA. If there are two functional units, one might interact with a different gene on a different part of the chromosome. For sure, science is always finding out how ignorant we are.”

  And consider that there are over 20,000 genes in the human genome, but in excess of 200,000 proteins—yet only a small fraction of DNA actually codes for proteins. What is the rest of DNA doing? We don’t really know. What was once considered “junk DNA” is only now beginning to be understood—which is further reason for caution when spreading engineered genes across the globe, Fagan said.

  “My belief is that nature is parsimonious in what it does,” he said. “It doesn’t waste anything. There are those hubristic opinions that say, ‘If we don’t understand it, it doesn’t exist, or it’s superfluous.’ But that’s the kind of thinking that allows people to be comfortable with the idea of going in and manipulating new genes in very sloppy ways and being so confident in putting them on millions of acres, and for decades. That kind of logic is really risky.”

  Arguments abo
ut GMO technology are one thing, in other words. The real anxiety arrives when the technology is applied systematically, across wide swaths of the continent and the globe—almost all of it in the service of industrial food. A big part of the problem is that a great deal of university science is funded by industry, or by a federal government in full-throated support of industry, Larry Bohlen told me. Bohlen is a veteran environmental activist who made international headlines fifteen years ago when he discovered that a GM corn (known as StarLink and approved only for animal feed) had made it into the human food supply.

  Consider the funding that flows from the USDA into research on GMOs. “When I looked at it in 2002, it was $193 million for GMO research, of which $3 million was to look at potential environmental problems—or about 1.5 percent—and zero for health effects,” Bohlen said. “That would be $2 billion over the last ten years. If I want to survive in academia, of course I’m going to go after the piece of the pie that is 98.5 percent of the budget. There’s no real blame in that—you can just look at it objectively and see that most of the money is going to the promotion of GMOs, so that’s where the scientists are going. And somebody is setting that budget.”

  Blake Meyers is matter-of-fact when it comes to the funding of his—and all—scientific research. His work is supported by the NSF and the USDA, but he has also gotten money from the big industrial players: DuPont, Dow, Syngenta, BASF. The building that houses his lab was constructed in the mid-1990s, with funds from the University of Delaware, the state of Delaware, and DuPont, whose world headquarters are just up the road in Wilmington. The institute was designed as a hub for research and teaching in the life sciences, but also to support the development of start-up biotech companies.