Better Living through Cell Death: HHMI's Herman Steller on Apoptosis

Better Living through Cell Death:
HHMI's Herman Steller on Apoptosis

onsidering that the human body is composed of more cells than the Milky Way has stars, it’s not surprising that 100,000 or so are dying off every second, while an equal number are being created to replace them. Since 1972 this natural process of cell death has been known as "apoptosis," from the Greek word for "falling away." It has been likened to the process of deciduous trees shedding their leaves every autumn. Gradually, researchers have come to understand that cell death and cell suicide are not just common in organisms–business as usual–but are crucial to survival, development, and health. And laboratory researchers are not the only ones to realize the benefits of understanding and learning to control apoptosis. As Fortune magazine put it in 1995, "apoptosis looks to be the richest drug vein struck in the two-decade history of biotech."

Hermann Steller of the Howard Hughes Medical Institute, MIT

"I strongly believe that there will be many diseases where blocking apoptosis will improve the condition of patients," says Hermann Steller of the Howard Hughes Medical Institute, MIT.

The seminal work in elucidating the death program in cells came in the early 1990s from Robert Horvitz’s laboratory at the Massachusetts Institute of Technology, based on work with the roundworm C. elegans. That research has been extended using the fruit fly system by Horvitz’s MIT colleague Hermann Steller, who has produced one breakthrough after another on the regulatory control mechanisms of the cell death program. Steller’s impact is aptly demonstrated by his 1995 article in Science, "Mechanisms and genes of cellular suicide," which has already been cited more than 900 times in the five years since publication (see the table on page 4, paper #1).

Steller, 42, did his undergraduate work at the University of Frankfurt in Germany, where he majored in microbiology and molecular genetics, and then moved on to the European Molecular Biology Laboratory and the University of Heidelberg for his Ph.D. in molecular biology. He spent three years working on Drosophila development in Gerald Rubin’s lab at the University of California at Berkeley before joining the MIT faculty in 1987 and launching his studies of apoptosis. He is now a professor in the Department of Biology and the Department of Brain and Cognitive Sciences as well as an investigator with the Howard Hughes Medical Institute.

Steller spoke to Science Watch correspondent Gary Taubes from his office at MIT.

What would you say is the major focus of your research?

We're interested in understanding the mechanism by which cells self-destruct–how the decision of whether a cell lives or dies is regulated by a variety of different and distinct signaling pathways. This turns out to be a process that is very conserved in animals: the basic machinery that cells use to kill themselves is identical, more or less, in worms, flies, and us. The process is important because it is a very stringent quality-control mechanism that makes sure there are no unwanted, potentially dangerous cells in the body. As we're talking right now, cells are dying in us at a rate of about 100,000 cells per second. And it's precisely controlled: for each cell we're losing, somewhere cell division is occurring to replace it. But any disturbance in this degree of cell turnover or cell death can have dramatic consequences. All kinds of diseases can develop. There are a huge number of degenerative diseases, including stroke, heart attack, and spinal cord injuries, in which a lot of cells die by committing suicide. Too much cell death is also associated with neurodegenerative and muscular diseases–Alzheimer's, Parkinson's, and Huntington's, for instance. Too little death is also very bad. If you inhibit the efficiency with which cells commit suicide, it exposes you to an extremely high risk of cancer. In general, if there is a lot of damage and not very well behaved cells, you have a breeding ground for additional mutational changes that may lead to tumors.

Considering that you did your postdoc work on the development of the retina in Drosophila, how did you get into the apoptosis business?

   Well, at the time I was looking at mutants where the eye and brain did not properly connect. My interest was in understanding how neuronal connections are made; how the eye and brain get wired together. But in the absence of such neuronal connections, a lot of brain cells die, as do cells in the eye later on. And the reason they die is because of what is now referred to as the social control of cell survival, the idea that many cells–and in mammals, perhaps all cells–need survival signals from other cells to stay alive. When you take a cell away and isolate it from such signals, the cell kills itself. It's a great way to achieve a nice balance between different cell types, and it prevents any one type of cell from becoming overabundant.
   When I moved to MIT in late 1987 and set up my own lab, we pursued these early observations that I had made on the communication between eye and brain, following up on questions of how and under what circumstances this cell death is controlled and carried out. We then decided about ten years ago that what was really needed was a systematic genetic analysis of programmed cell death in the fly.
   Of course, we knew of and were influenced by the pioneering work that my colleague Bob Horvitz was doing on C. elegans. This defined the genetic pathway for programmed cell death–the core mechanism. What wasn't clear at the time was how this cell death program is regulated–what makes it turn on at the right time in the right cells, which is, of course, a critical decision. That's what we wanted to find out. We wanted insight into how this common death program is regulated by many different distinct signaling pathways. And we reasoned that we should first define a meeting point for different signaling pathways, a step at which many different death signals meet. We hoped that in this way we might be able to work out how a common death program is activated in many different cell types and different situations.

How do you get at this program using fruit flies?

The same programs used by flies to decide which cells die appear to be very similar to the controls seen in humans. The fly serves as a gene-discovery system. You define the regulation of the death program genetically in the fly, clone the genes, work out what the genes do and what the proteins they encode do, discover how the parts in the pathways interact; and then work out if what you’ve found in flies is conserved in humans. You can then clone homologous genes and try to manipulate the process ultimately by making drugs that selectively target different components in the regulation of death.

What are the key discoveries you’ve made?

A few years ago, we did a systematic screen to see if we could identify mutations that were required essentially for all cell deaths. So after looking for mutants in which cell death fails to occur, we cloned the corresponding genes. The initial gene we called reaper, which is the messenger or harbinger of death. It triggers the onset of death in many cells of flies in many different circumstances, including in response to radiation damage. Then we cloned two additional genes: the second one is hid, for "head involution defective." It has a defect in taking the head and moving it back into the body; there's not enough cell death during development in these cases to generate the space needed to allow these kind of tissue movements. The third gene is called grim, because it mapped to reaper. That was characterized by one of my former postdocs, John Abrams, after he set up his own laboratory at the University of Texas Southwestern Medical Center at Dallas.

Have you found homologues in humans?

Not yet, but we know mammalian cells respond to these genes. We can take the Drosophila genes and express them in mammalian cells, and they induce apoptosis in those cells. So the genes are interacting with other components in the death program in mammals to induce apoptosis. And we know that the "inhibitor of apoptosis proteins, " or IAPs–the targets for cell killing by reaper, hid, and grim– are found all the way up in humans. Therefore, this program is conserved between flies and humans. That makes us absolutely certain there will be mammalian homologues with equivalent function. They have not been found, although they will be, I'm sure, when the human genome is entirely sequenced, but we're still quite far away from that.

How do reaper, grim, and hid work to control cell death?

They're regulators, like the keys that start your car engine. These three genes are the key activators of apoptosis. And they all do rather similar things. They all lead to activation of cysteine proteases known as caspases, which are a particular type of protease that cleaves other proteins. The way they do this is by inhibiting the life-saving activity of IAPs, factors that are essential to prevent the inappropriate activation of caspases in cells that should live. In other words, hid and grim remove the "brake on death," a brake which prevents livings cells from committing suicide.
What's unusual about these genes, particularly reaper and grim, is that they are only expressed in cells doomed to die. The death program is everywhere, but reaper, grim, and to some extent hid are only turned on by transcriptional activation in cells that are doomed to die. That realization led us to a first answer of how the integration of many distinct death-inducing signals occurs: It turns out that reaper, which is the gene best analyzed from the transcriptional perspective, has many different regulatory elements that control expression of the gene. Reaper is a very small gene as far as the protein it expresses is concerned, but it has a very large regulatory control region. And many different death-regulating signals influence whether this gene should be expressed.
continued