View our other publications:

What exactly is genetically modified food?

Asking the right questions and understanding the answers require brushing up on your chemistry.

You’ve probably seen and heard a lot about genetically modified food in recent months: on the news, among friends, even as a label on the food you eat. But for all the information and opinions out there, there’s still very little explanation about what exactly genetically modified organisms are. To get a good answer, it’s helpful to talk with someone who actually makes them.

But when you do, be prepared to wrap your head around genetics, biochemistry, and microbiology, topics that many of us supposedly learned about in high school and then—maybe even before the test—promptly forgot. Which is unfortunate, because understanding how the world works at this smallest of scales is essential to understanding so many of the discoveries, the innovations, and the controversies now swirling around us. Especially when it comes to the food we eat. It’s the real test that no one told us about.

Matthew Willmann is director of the College of Agriculture and Life Sciences Plant Transformation Facility at Cornell University. He is someone who makes genetically engineered plants. Not for food and not for commercial purposes—Willmann currently uses biotechnology to modify crops for research only.

“It’s relatively time consuming and it’s a specialized skill,” he says. Which is why researchers at Cornell and other institutions contract with Willmann to do the grunt work of modifying the genes in a plant to produce a specific trait. And, according to Willmann, the facility does not receive funding from “Big Ag” companies to do this work.

Which makes him a very interesting person to talk with.


What are genetically modified foods?

“Answering your question requires more than a definition,” says Willmann. “It also requires a discussion of the relationship between humans and our food over time. I believe that all plants that we eat are genetically modified, because to me the words ‘genetically modified’ mean that humans have had a hand in altering the genetics of the organism. Humans have been altering the genetics of plants ever since the onset of plant domestication. By choosing which individuals of a species would be selected for food or for replanting, we changed the genetics of the population of all plants we domesticated.”

The biotech way of altering genetics is one more way to domesticate and breed the plants we eat, he says.


How is genetic modification different from traditional breeding?

“Variation occurs in nature because every time that a cell divides, it has to replicate its chromosomes,” says Willmann.

This replication process is not perfect; there are always going to be little errors that occur, little switches that DNA will make. They can happen during mitosis, when cells replicate to make more cells of exactly the same kind. Or the errors can happen during meiosis, when cells divide to make sex cells that will come together to create offspring. The DNA—the code in our cells that is the basis for every trait you can see, hear, smell, taste, or feel—must split apart and then recombine. And DNA won’t always recombine exactly like the original. If it did, we wouldn’t have any variation in plants or animals or humans. In fact, there wouldn’t even be humans.

“The replication error rate allows for diversity, which allows for evolution,” says Willmann. It’s these seemingly random errors in our DNA that can result in different physical traits that allow organisms to adapt and survive. When the environment changes or when a species moves into a new environment, there will hopefully be enough variation of genetics and physical traits that some will survive within those new conditions. “Those traits will then be seen in a larger percentage of the population because of natural selection.”

This natural variation is what humans have relied on for thousands of years to shape the crops we grow. We domesticated certain plants by specifically selecting and nurturing the variations we liked. Then we started exerting control over how plants would mate and reproduce, breeding them to exhibit certain traits. We would collect and replant the seeds from preferred plants, generation after generation, trying to keep traits exactly the same or trying to push them in a new direction. It’s how our ancestors took corn from being gnarly little stalks of grass and turned them into big, fat ears of summer goodness. It’s still how we breed plants in the twenty-first century. But it’s no longer the only process of breeding.

Because traditional breeding takes time. Even under the best circumstances, it can take years to selectively and determinately get the characteristics we might be after. That’s why humans have been looking for ways to speed up this process—to speed up variation—by leveraging the tools and technology we have available: cross-breeding, chemicals, radiation, marker-assisted selection, and, now, bioengineering.

“Let’s take an extremely simple example,” says Willmann. “Color.”

Let’s say you really like the blue color of a certain type of corn, but it doesn’t taste good. Obviously, you want to cross it with a variety of corn that is enjoyable to eat. With traditional breeding methods, you would transfer pollen between the blue corn (Line 1) and the tasty corn (Line 2), and within the first couple generations of offspring you would get a range of results for color and taste. But nothing that is really blue and really tasty. So you select the best individuals and continue crossing pollen; you might backcross really tasty offspring with Line 1, you might backcross really blue offspring with Line 2, or you might cross offspring with itself. You keep doing this over and over until you finally get corn that exhibits both the color and the taste you’re looking for—or as close as possible.

“This is why it gets complicated and why it takes longer,” says Willmann. “This is why doing transformation can speed up the process.” Because if you were to transform corn through bioengineering, you could take the color gene from Line 1 and transform it into individual cells of Line 2. Within a single generation, you could get the really blue and really tasty corn you’re after.

But remember: this is an extremely simple example of how breeding works and how bioengineering can speed things up. Breeders who use genetic engineering also rely on traditional breeding methods to produce different crop varieties. And unpredictable variation can occur with bioengineering because of the ongoing mystery of what each gene in a plant actually does.


How do we genetically modify the foods we eat?

“There are currently two major techniques in use for altering the genetics of plants,” says Willmann. “The first inserts DNA from another organism into the plant genome.” This is called transgenic modification. “The second technique is called gene editing; it changes the genotype at specific locations in the genome and does not require DNA from another organism to modify the genome of a plant.”

There are two big challenges with transgenic modification. The first is getting foreign genetic material through the cell wall: a tough outer layer of the cell that provides structural support to the whole plant and limits the entry of foreign molecules into the cell. The second challenge is getting that same genetic material into the nucleus where it can insert into the plant’s 

genome. But we humans have created several different tools for overcoming these challenges.

Agrobacterium tumefaciens is a natural plant pathogen that inserts DNA into plants,“ says Willmann.

Agrobacterium is found in soil. It causes crown gall disease, which causes those big, gnarly tumors you see on trees, grape vines, and other plants. The tumors are a result of A. tumefaciens hijacking a plant’s hormonal system and getting it to produce food for the bacteria, which it does by inserting some of its own DNA into the plant’s cells. When researchers figured this out, they started experimenting—they started using A. tumefaciens to insert DNA from other organisms instead.

The gene gun is a completely man-made method for getting genetic material through the cell wall. It is exactly what it sounds like. “The gene gun blasts gold particles covered with DNA through the cell wall and through the cell membrane,” says Willmann.

Created by two Cornell professors in the 1980s, the gene gun operates on a “shotgun principle”: if enough genetic material is blasted at plant cells, some will get inside and be incorporated into the plant’s own DNA.

A third approach is to use protoplasts, cells whose walls have been stripped away with an enzyme. “Then you can add DNA and a chemical that makes small holes in the cell membrane,” says Willmann. The DNA makes its way into the cell through those holes.

While these techniques manage to get genetic material into the cell, they also have the potential to cause the aforementioned unpredictable variation. Once foreign DNA gets inside a cell, you can’t be sure where it will insert into the host genome; depending on the location, the foreign genes might be expressed in very different ways than expected.

“The Agrobacterium method is more efficient because the DNA is targeted to the plant genome but it inserts at random locations,” says Willmann. “For the other two methods, DNA is inserted in random locations and at low efficiency.”

These transgenic tools work, but not in an ideal way all the time; researchers want to consistently make mutations and modifications at specific locations in a plant’s genome at high efficiency. Which is why Willmann and seemingly everyone else in his field are extremely interested in newer techniques known as gene editing. In particular, they’re interested in a technique known as CRISPR/Cas9.

While the letters in the acronym—clustered regularly interspaced palindromic repeats—might not make much sense on their own, the important thing to understand about CRISPR is that it’s based on the adaptive immune response found in certain bacteria. Researchers are figuring out ways to employ this process to modify our food. When prepared with a specific sequence of genetic information as a target, the CRISPR molecule will seek out that same sequence in a genome; it will unzip and cut the DNA at the target location. The DNA will then naturally mend itself at the cut, resulting in a modification by either cutting out a particular sequence or introducing a different sequence into the genome by providing a genetic template for the DNA to follow during repair. In either case, CRISPR has modified the genome of a cell—and the traits of the organism—without introducing foreign genetic material.

This is why CRISPR is so radically different from transgenic methods. And, according to Willmann, this is why researchers don’t want gene editing being associated with genetic modification of food as it has happened to date.

“We’re hoping it doesn’t get classified as a GMO.”

While CRISPR is the tool everyone wants to be using, transgenic approaches are still commonly used in both research and industry because of their effectiveness under certain circumstances. It all depends 

on which crop you’re working with and which trait you’re hoping to modify.

But regardless of how a cell is transformed, the process for regenerating it into a full plant is the same.

Regeneration is accomplished with a very structured application of two different plant hormones: auxin and cytokinin. In his lab, Willmann starts by adding equal parts of the two hormones to get callus: an explosion of plant cells,  dividing uncontrollably and undifferentiated, like cancer cells. Then he will apply more cytokinin than auxin, and cells begin differentiating; they become shoots and stems and leaves. When he adds more auxin than cytokinin, other cells start growing into roots. Through this process, Willmann ends up with numerous little plants. He will pot them into soil and return them to the researchers for continued study.


Are genetically modified foods safe?

When Willmann transforms crops, he doesn’t just modify one cell; he modifies dozens of cells. He regenerates dozens of plants to give back to the researcher. And there’s a very specific reason for this.

“The scientist doesn’t always know what characteristic he or she will actually see,” says Willmann. “It’s possible that the DNA could insert in a region of a genome that affects the expression of a different gene. If that happens you can create an unintentional side effect. You can get traits that are not related to the transgene itself.”

To be clear: “unintentional side effect” does not automatically mean “harmful to humans.” It just means something unexpected happened.

But the unexpected could be a whole series of downstream effects in a plant. Changing the expression of a single gene doesn’t always affect just a single trait; if you turn up the dial on the gene for the color blue, it may also turn up the dial on other genes for traits like plant height or yield or taste. And then those increased expressions of genes may increase the expression of still more genes. There is the possibility of a cascading effect.

“So one thing scientists do before they release a new line is to do an awful lot of testing to figure out whether there are unintended functions of this gene,” says Willmann. “You have to consider what all of those downstream genes are doing.”

This is why he returns a whole population of transformed plants to a researcher, not just a single plant from a single cell. The bigger the population, the better the chance of seeing unintended variation.

And, says Williman, “The more knowledge there is about a foreign gene in its own species, the less unpredictable the results.”

Willmann sees great potential in these biotech tools for modifying the crops we grow and the food we eat. But because we’re still learning about the function of all the various genes in all the various plants, there is still much we don’t know when it comes to making changes.

“Plants have a very complicated biochemistry. They have a very complicated…” he pauses, thoughtful. “…everything.” Which completely makes sense when you think about it. Plants can’t move like animals; once they take root, that’s where they live out their lives. And they need to be able to accommodate and adapt for a changing environment right where they are.

“We don’t know everything we need to know about every gene,” says Willmann, speaking of the plants he works with. “That’s where the mystery remains.”

So should you be concerned about this mystery when it comes to eating bioengineered food? So should you be concerned about this mystery when it comes to eating bioengineered food? The only way to get a good answer is to brush up on your biology and chemistry.  


Matt Kelly is a freelance writer and photographer based in the Finger Lakes. A complete portfolio of his work can be found at

Subscribe to our newsletter