How and when did you first start studying the myc protein, and what makes this particular protein of such interest?

In the mid-1980s, I worked on trying to understand the genetic mechanisms that make cells proliferate. In particular, I worked on one gene called c-myc, which seems to be pretty fundamental to cell growth. The name "myc" is an acronym of a myelocytic leukemia, a neoplastic disease of chicken blood cells where the gene was first discovered. The myc gene, like most of the now well-described oncogenes, turns out to be part of the normal machinery of cell growth which, when disrupted and turned on at the wrong place and the wrong time, make cells proliferate uncontrollably and cause cancer.

  How did you approach the problem of elucidating the role of myc in cancer cells?

The issue in my mind at the time—and this was the late 1980s—was that there was a real problem with genes like myc. We knew that when such genes get turned on, they make cells replicate. Now, there are hundreds of thousands of billions of cells in our bodies. If we imagine myc gets turned on in even one cell, something we imagine happens hundreds of times a day, that cell should start to replicate. As it does so, it makes daughter cells and they in turn replicate and make daughter cells that replicate and, bingo, an instant cancer. Now, mutations like myc activation must be relatively common within our lifetimes. Yet, if they are, and if every one of those mutations makes a tumor, we would be a mess of cancers almost immediately. Cancers should therefore be very common whereas, in fact, they are amazingly rare: we know that it takes almost a lifetime for even one in three people to get cancer. Thus the great mystery is why is cancer so incredibly rare? There must be something that prevents genes like myc, when they get activated, from inevitably and immediately forming cancers.

  So how did you find the solution to that mystery?

We already knew from the beautiful work of people like Robert Weinberg in the early 1980s that oncogenes cooperate. That is, you need more than one type of oncogene to make a cell into a tumor cell. It seems to be that there are different flavors of cancerous mutations and that you need to get an ensemble of flavors to get a cancer. The question was what are these flavors, and what do they do? In experiments we undertook in 1989 and 1990, we turned on the myc oncogene and looked to see what happened. As expected, the cells in which myc were turned on started to replicate and proliferate uncontrollably. But the interesting thing was when we came back a couple of days later to see if there were more cells than before, there weren't. In fact, there were fewer! This was quite a mystery. We could come up with only two possible explanations for the loss of cells: one was that the cells escaped in the night, which was fortunately not the case. The other was that the cells died. It turned out that when you turn on myc, you not only turned on the cell’s capacity to replicate but also turned on the ability of these cells to commit suicide very efficiently.

  And that's what you were reporting in your highly cited 1992 Cell paper "Induction of apoptosis in fibroblasts by c-myc protein"?

Yes. That comprised the work that became this highly cited paper. It was the depiction of the fact that oncogenes can do completely opposing things in cells. Not only can they increase cell number by driving cell proliferation, but they also have a dark side. They can decrease cell number by triggering programmed cell death, this process now termed apoptosis.

  Is that true for all oncogenes?

It looks like it's a general property that oncogenes that make cells replicate also have the capacity to kill them. That's become clear over the last nine or so years.

  How does the cell know when to switch from proliferating to suicide?

When a cell replicates in its normal environment in the body, its neighbors provide it with survival signals that shut down the death program. This means that cells in the body can only replicate and survive if they're in the right place and right time, as defined by the fact that they've got the correct neighbors that provide them with appropriate survival signals. If a cell, through whatever mutation, begins to replicate uncontrollably and spills into a part of the body where it doesn't belong, it no longer gets these survival signals and the same mutation that's making it replicate now kills it.

  When you published the Cell paper, did you expect it to have such enormous influence?

Well, to me it offered an answer to that very fundamental question: why are cancers so rare and how are potential cancer cells reined in once their oncogenes are mutated? It provided a very simple and robust solution to the problem of how to allow cells to replicate with ease when necessary, but only in the right place and the right time.

  So you had little doubt you had the correct answer to the mystery?

I had no doubt this was a pretty important mechanism, and later that year we wrote a review called "Oncogenes and cell death," in which we laid out this concept. Our idea was that not just myc, but other oncogenes also had their own dark side. However, their dark side might be different. It might not involve cell death, for instance, but perhaps cause cells to arrest their growth so they could never go on to become tumors. We suggested that when several oncogenes are mutated they cooperate because each counteracts the other one's dark side. If the oncogene is by itself, its dark side wins and the cells shut down or die so that no tumors develop. But when several oncogenes are turned on together, they can counteract each other's dark side and that allows the cells to expand uncontrollably into a tumor.

  Can you give us an example of such counteracting oncogenes?

In the same year, for instance, we published a paper about an oncogene that blocks cell death called bcl-2. About four years earlier, Suzanne Cory and Jerry Adam’s labs in Australia had showed that bcl-2 cooperates with myc. That is, neither myc nor bcl-2 alone is very tumorogenic but the two together make a potent tumor inducer. A little later on it was shown by Cory's lab and also by Stan Korsmeyer that bcl-2 is an oncogene because it blocks cell death. We showed that myc by itself made cells replicate but then killed them, bcl-2 by itself blocked cell death but also suppressed cell proliferation, but the two together did something magical that neither could do alone. Myc proliferation overcame the bcl-2 stop command and bcl-2 survival overcame the myc death signal, generating cells that can both proliferate and survive— a neat exposition of how different flavors of oncogenicity can interlock together and cause cancer.

  Why did you choose to send the myc-induced apoptosis paper to Cell?

I submitted it to Cell because I felt that the paradigm that would come out of our work—that oncogenes have intrinsically antagonistic functions—was a very important concept and if I could get it into Cell everybody would read about it, which they did. After all, Cell is one of the very top journals, along with Science and Nature. In fact, at the time we were submitting our work, there were already hints in the literature that myc might induce apoptosis. A few months earlier a very good friend of mine, John Cleveland, had published that myc could promote apoptosis in a bone marrow cell line. We were, not surprisingly, a bit distressed when this came out. But it made us do every experiment we could think of to absolutely nail that the apoptosis we observed was myc-specific. We made myc mutants and switchable forms of the myc protein. I’m now told that we absolutely nailed the concept to the wall so that no one at the end of the day could say, "I think there's some sort of artifact or I'm not sure I believe it." In general, Science or Nature don’t really lend themselves to that sort of paper. However, a journal like Cell is a good vehicle for thorough and careful, if sometimes mundane, experimentation.

However, the world of top journals can be unpredictable. We published a paper in 1994 that I think is probably the best work we ever did and looked at the relationship between myc and survival signals. We showed that when you activate various survival-signaling pathways, myc-induced apoptosis is shut down. This was the final piece of this puzzle that I described earlier. However, when we submitted the work to Cell it got bounced. In the end, it went into EMBO Journal, a very good journal, but it never had the impact the 1992 Cell paper did, even though in my mind it actually completed the puzzle. I suspect that it wasn't jazzy in that there were no new genes, for example. Journals love new genes! Our paper just put the puzzle together and new synthesis is often not considered as that important.

  What were the biggest challenges to unraveling these oncogenic programs?

So far, it's been that most of this work has been done in cell culture, and it’s been extremely difficult to demonstrate that myc-induced apoptosis really does suppress malignancy in live animals. Lately, we've carried out experiments in transgenic mice that show that when we switch on myc, some tissues just disappear. They undergo massive apoptosis—clear evidence that the dark side of myc wins out in real tissues in vivo. In addition, we’ve now shown that when we block myc-induced apoptosis we can trigger instant malignancy. Satisfyingly, these studies have also just been accepted for publication by Cell.

Gerard Evan, Ph.D.
University of California, San Francisco
Cancer Research Institute
San Francisco, CA, USA