Opinion, arguments & analyses from the editors of Scientific American

Fruit Flies Aid Efforts to Develop Personalized Cancer Treatments


Credit: Thomas Wydra/Wikimedia, Creative Commons

For years clinicians have puzzled over the observation that people with type 2 diabetes are more likely to develop certain malignancies, such as pancreatic, breast and liver cancers. The reason behind their confusion: standard biological principles predict that, if anything, folks with diabetes should suffer fewer tumors, not more of them. However, new research on fruit flies, published this month in the scientific journal Cell, may have finally cracked the long-standing mystery. If confirmed in humans, the findings could one day prove beneficial not just in the treatment of diabetes and cancer, but they may also aid efforts to develop customized treatments based on an individual’s own genetic profile.

To learn more, I phoned Ross Cagan, a professor and associate dean at the Icahn School of Medicine at Mt. Sinai in New York City and a senior author on the paper. We spent about two-thirds of our time talking about the general principles behind the experiment before we got into the specific details of the article in Cell. This sort of wide-ranging discussion is one of my favorite parts of being a science journalist. It gives me a sense of how biologists are starting to weave together insights from genetics, developmental biology and physiology to create a better picture of what really happens inside a whole organism—whether it’s a fruit fly or a person.

Made-to-order flies

Cagan told me there is a small but growing number of scientists who are turning to the most well-studied insect in the world, Drosophila melanogaster (a. k. a. the common fruit fly), to get a better sense of how tumors develop in the real world—that is to say in the body over time—as opposed to the cell cultures in a petri dish that have been the basis for so much cancer research over the past several decades.

Fruit flies offer lots of advantages as lab animals. They are tiny, inexpensive to feed and house and have very short life spans—on the order of days to weeks, depending on the temperature—which means you can do lots of experiments. (Replication and confirmation being the lifeblood of science.)

By contrast, white lab mice, which are more often used as stand-ins for Homo sapiens, typically weigh about 20 grams apiece (on average), can live for years, need plenty of food and water and produce lots of smelly wastes.

Despite the fact that the last common ancestor of insects and people lived around half a billion years ago, evolution has made relatively few changes in the genes that are most critical to their growth and development. (If it ain’t broke, don’t fix it.)

These similarities among key genes in fruit flies and people also means you can learn quite a lot about how genes interact with each other and their environment. You can follow the many interlocking signals between whole networks of genes that eventually produce, for example, the body plan of an embryo or the intricate shape of an eye. The genes that are important for such developments in fruit flies are close enough on the molecular level to those found in people that studying the pathways in the insect will give you a lot of insight about what’s going in people. And you can always check your best ideas out by testing them out in mice.

Indeed, biologists have gotten so good at producing fruit flies with specific genetic mutations that they can now order their own custom-designed insects from various supply houses via computer and have them delivered straight to the laboratory door.

This is where things really start to get interesting. Thanks to the genomics revolution of the past decade, clinicians can now take a genetic snapshot of a person’s malignant tumor and find out which genes are acting differently than usual—giving rise to more or fewer proteins than normal, for example. “About a third of genes are overexpressed in cancer and about a third of genes are underexpressed,” Cagan says. But all these genetic changes are not necessarily critical to understanding how the tumor got started in the first place—or how to treat it.

In other words, some of the mutated genes found in a tumor are acting as “drivers” of cancer growth and spread while others are “passengers” that pop up as the cells becomes more and more disorganized and mutations start to accumulate. The trouble comes when clinicians find that an individual patient’s tumor has 200 or more mutated genes—which ones should they be focusing their attention on and which can they safely ignore?

Changes in some genes, with names like Ras and Src, have long been linked to particular types of tumors. So if you find they have been altered in a particular cancer patient, you know that you want to pay attention to them. But the latest research suggests that most cancers in people probably result from the combination of a few common mutations and several rare ones. And the rare ones may in fact be the ones that determine how aggressive a cancer is, or how likely it is to spread.

That level of complexity means that it’s important to identify the rare variants that are driving any particular individual’s cancer.

So back to those 200 or so mutated genes that pop up in a profile of a human cancer patient. Cagan and his colleagues compare those genes to the genes that are found in fruit flies. On average, they find 180 matching genes in the flies. Then they go to a computer and order up 180 fruit fly lines—each one of which is specifically bred to have the same Ras and Src mutations plus one rare variant, based on the genetic profile of the human patient’s tumor.

(The mutations are designed to show up in the tissues of the fly’s eye because changes there are easy to spot and analyze—and the epithelial cells of the fly eye are very much like the epithelial cells that give rise to the majority of cancers in people.)

Once the flies become adults, the scientists check how each of those 180 groups of flies are doing—whether or not they develop abnormal growths, how quickly the growths emerge. Eventually they whittle the number of genes down to about ten that seem to matter. Those ten genes (including the Ras and Src genes) produce a cancerous growth in the fly that most closely resembles the one in the human being. In other words, as Cagan says, “We’re building personalized flies.”

Testing Treatments

Not only are these personalized flies likely to provide greater insight into how tumors form and grow, they offer a radically new way to test potential cancer treatments.

Now researchers can screen several different drugs, or combinations of drugs, against the specially designed fruit flies that share the same 10 key mutated genes as are found in a particular human being’s malignant growth. (Technically the fruit fly and human genes are orthologs, genes from different species that can be traced backed to the same gene in a common ancestor.) The fruit flies are kind of like stand-ins for the people. “When you put these rare variants in, they strongly change the way these patients—in this case what I mean are these particular flies—respond to drug treatment,” Cagan says.

This crazy-sounding approach has already started to pay off. By creating fruit flies with the matching genetic profiles to the tumors found in folks with an unusual disorder called multiple endocrine neoplasia, type 2, Cagan and his colleagues discovered that a drug called ZD6474 was likely to work against certain of their cancerous growths. The pharmaceutical company AstraZeneca followed up on that insight and in 2011, the Food and Drug Administration approved the drug, now called vandetanib, for the treatment of advanced thyroid cancer.

The beauty of these highly detailed fruit fly experiments is that they allow researchers to start tackle the real-world complexity of malignant tumors rather than having to simplify everything, treating all breast cancers or all colon cancers alike and being disappointed when the results aren’t more predictable.

In the past few years, Cagan has joined with Thomas Baranski at the Washington University School of Medicine to try to peel back another layer of cancer complexity using fruit flies. (The two men also founded a company in 2006 to try to find commercial applications for some of their basic research.) They want to find out more about how cancers grow in people with diabetes.

If you’ve stayed with me this far, you’ll recognize that this where we started—with a mystery about diabetes and cancer, namely, why are people with diabetes more likely to develop certain cancers?

Diabetes and Cancer

At first glance, you might think that this relationship between diabetes, specifically type 2 diabetes, and cancer makes sense. After all, people with type 2 diabetes have a lot of sugar (in the form of glucose) circulating in their blood. And their bodies also produce a lot of insulin, which is a hormone that allows cells to use sugar. Tumors need to grow and insulin and sugar are two of the best growth promoters around. So it’s obvious, type 2 diabetes must promote cancer growth.

Such a straightforward explanation quickly falls apart, however, when you consider why there’s so much sugar and insulin floating around in the blood of people with type 2 diabetes. Their condition actually makes them resistant to the effects of insulin. That means that all the glucose in their blood is stuck there because it cannot get into the cells of their body, including cells that might be prone to becoming malignant. No sugar, no growth. No growth, no cancer. Or at least, a lot less cancer.

And yet, we know from epidemiological studies that people with diabetes are in fact at greater risk of developing certain cancers, hence, the mystery.

As it happened, one of the human cancer patients Cagan and Baranski wanted to study also had diabetes. So, developing a personalized fruit fly that more closely mimicked the conditions under which that patient’s tumor developed required not just copying the right combination of genetic changes but also coming up with a similar metabolic environment. In other words, they had to make the fly diabetic.

It turns out to be pretty easy to give fruit flies a condition—call it flyabetes—that looks a lot like diabetes in people. You just keep feeding the flies bananas and nothing else. So Cagan and Baranksi ordered up a group of fruit flies that had the same genetic changes as their cancer patient, checked to see which variations, in addition to Ras and Src, most closely matched their human patient’s tumor, then fed the flies bananas round the clock.

What the researchers discovered is that the alterations in the Ras/Src genes had an unexpected benefit—at least for wannabe cancer cells. In the presence of dietary sugar, the mutated Ras/Src combination enables the putative cancer cells to reassert their sensitivity to insulin. Suddenly, they and only they have greater access to all that sugar in the blood. The cancer cells become like a sink for all that extra sugar, which allows them to grow faster than they would otherwise. By contrast, the authors of the August 1 Cell paper reported, that the Ras/Src combination in this particular line of customized flies could not sustain a malignant growth in the presence of a low-sugar, high-fat diet.

The next step would be to confirm these findings in mice and eventually people. Cagan and Baranksi are now putting together a combination of three different drugs that they are now testing in rodent and human cells lines. The idea is to try to come up with a treatment that is tailor-made to be more effective against tumors that have developed in people with diabetes.

There is still a long way to go, of course, to get from Drosophila melanogaster to Homo sapiens. But the main point to take away from this research is that complexity matters when it comes to studying cancer—and eventually developing much more effect treatments for it. How much of that complexity you need to recreate in the lab is still a big unknown. But using whole organisms—in this case fruit flies—with multiple genetic and metabolic changes may give us more answers than were ever before possible.


The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions
Share this Article:


You must sign in or register as a member to submit a comment.

Starting Thanksgiving

Enter code: HOLIDAY 2015
at checkout

Get 20% off now! >


Email this Article