Death of a Cell . . . by Suicide

By William Allstetter Most people have seen a cell divide. Short movies showing the sequential phases of mitosis are a mainstay of science education. But few have seen a cell die. That may be because, until recently, few people paid much attention to cell death. Most scientists considered it the final, relatively dull endgame of a cell. s existence. They believed that once toxic insult or injury destroys a cell. s ability to carry on metabolism and maintain homeostasis, there is nothing to be done. The cell and its organelles swell, their membranes break apart, and the cell. s contents spill into the intercellular space. Immune system cells and molecules sweep in to clean up the mess, causing inflammation and additional damage in the process. Necrosis, as it is called, was described long ago and offered little of interest to researchers. In the 1990s, however, it has become abundantly clear that cell death is considerably more complex and interesting. Cells don. t always wait for deadly forces to do them in. They commit suicide, or, in a more scientific term, apoptosis. In fact, each cell in our bodies contains a complex machinery of death. A host of genes, receptors, and enzymes carefully controls when and under what conditions a cell will kill itself. Some cells initiate apoptosis after chemical messengers bind to their "death receptors." In others, suicide is the default, preventable only by a constant stream of reassuring messages from other cells. Cells also respond to internal signals, condemning to death any cell that poses a threat to the larger organism. "Apoptosis is a normal developmental and safety process," says Dr. Michael Shelanski, Delafield Professor and Chairman of Pathology. "It gets sick cells out of the way and makes sure we develop properly." The hand of a developing fetus begins as a flat paddle. Apoptosis sculpts it into individual fingers. A developing brain makes more than twice the neurons it will eventually employ; ones that fail to make the right connections are eliminated. When genetic damage occurs, internal sentries, such as p53, halt cell division until repairs can be made. If the damage is beyond repair, suicide is invoked. Cells in the gut, skin, and elsewhere commit suicide every day as part of the normal maintenance of tissue. But when the death machinery goes awry, disease can result. In neurodegenerative disorders, autoimmune disorders, and stroke, cells die prematurely. In cancer, cells fail to die when they should. Today, apoptosis is one of the hottest topics in biomedical research. Basic scientists have a whole new phenomenon to examine and describe. Clinical scientists are hoping to cure several diseases by preventing or inducing apoptosis. Thousands of papers on apoptosis are published every year, and entire conferences are organized around individual elements of the process. "Something about it intuitively appeals," says Dr. Beth Levine, assistant professor of medicine. "It has implications for so many aspects of biology." Some complain that apoptosis has become the latest research bandwagon, with scientists seeing it where it doesn. t exist and scrambling to include it in their grant applications. A handful of P&S researchers, however, were ahead of the curve, investigating apoptosis before it became the latest biological buzzword. They helped reveal basic elements of apoptosis in all cells. They have discovered how and where it occurs in both healthy and diseased tissue. They have even demonstrated its role in interspecies competition. Several are now developing therapies based on apoptosis. Apoptosis was first described in a 1972 paper by John Kerr, Andrew Wyllie, and Alastair Currie, three researchers at the Univer-sity of Aberdeen. The researchers also coined the term, which comes from the Greek word that describes the falling of leaves from a tree or petals from a flower. In contrast to the messy process of necrosis, apoptosis is quick and neat. Instead of swelling, a cell undergoing apoptosis shrivels and separates from its neighbors. Its DNA and organelles condense. The cell then "blebs," dividing into several small vesicles, which are consumed by neighboring cells. Apoptosis can occur in as little as 20 minutes and leaves no trace behind, which may explain why biologists failed for so long to see it. Although the paper by Kerr, Wyllie, and Currie is now considered a classic, their concept of apoptosis languished, largely unnoticed, for many years. Then, in the mid-1980s Robert Horvitz at the Massachusetts Institute of Technology began investigating cell death in the nematode worm, C. elegans. He learned that precisely 131 of the worm. s 1,092 cells disappear during development. In 1986 he published a paper identifying two genes, ced-3 and ced-4, involved in their disappearance. Then, in 1992, he found a third apoptosis gene, ced-9, which was nearly identical to bcl-2, a human gene implicated in lymphatic cancer. At that point, many biologists sat up and took notice of apoptosis. Clearly, if a gene controlling apoptosis in a worm had persisted through the eons of evolution to humans, then apoptosis must be a universally important biological function. Columbia. s Dr. Taka-Aki Sato and Dr. Thomas Franke have helped biologists understand pathways that transmit apoptotic signals from a receptor on the cell surface usually to the mitochondrion. The mitochondrion releases cytochrome c, which leads to the activation of caspases, the enzymatic executioners. The caspases, about a dozen of which have so far been identified, chew up the cell. s DNA, destroy molecules that attach a cell to its neighbors, and dismantle structural proteins within the cell.

Taka-Aki Sato Dr.Taka-Aki Sato, associate professor of molecular oncology (in otolaryngology/head & neck surgery and pathology), did much of his early work on the bcl-2 family of genes. The bcl-2 protein plays a central role in apoptosis. It sits on the outer membrane of the mitochondrion and inhibits apoptosis. Several closely related proteins, such as BAD, Bax, and bcl-xl, interact with bcl-2 and each other to either promote or inhibit apoptosis. Dr. Sato, in collaboration with Dr. John Reed at the Burnham Institute, identified, among other genes, BAG-1, a bcl-2 family member that inhibits apoptosis. Several years ago Dr. Sato made the surprising discovery that yeast can serve as a powerful model for studying apoptosis. A single-celled organism would not be likely to have a suicide program, and yeast does not. But when two genes, antiapoptotic bcl-2 and its proapoptotic cousin Bax, are transfected into yeast, they establish a primitive form of the death machinery. This allowed Dr. Sato to use a clever system, known as the yeast two-hybrid system, to find several other genes involved in apoptosis. When Dr. Sato has one protein in a pathway and wants to know what it interacts with, he attaches half the amino acids of a transcription factor to the protein he already knows. He then attaches the other half of the transcription factor to proteins that might interact with the original protein. If two proteins bind in their normal interaction, the two halves of the transcription also join together. The unified transcription factor turns on a marker gene, beta-galactosidase, which turns the yeast blue. "Using a yeast system is a very powerful way to pull out proteins," says Dr. Sato. P75NTR is a neural receptor whose stimulation can either induce apoptosis or prevent it, depending upon the tissue in which it is expressed. Dr. Sato recently discovered the gene for the first protein known to interact with P75NTR. The gene, which Dr. Sato named NADE, promotes apoptosis. Dr. Sato hopes eventually to induce apoptosis in cancer cells. Early in his career, he worked for a company trying to develop Interluekin-2 as a cure for cancer. Interleukin-2 is a growth hormone-like substance that binds to receptors on cell surfaces. But interleukin-2. s actions were non-specific, causing both positive and negative effects, depending upon the type of cell. Dr. Sato believed that the various apoptosis pathways offered more specific and promising targets for therapy. "I thought apoptosis offered the fastest way to make a new drug against cancer," says Dr. Sato. He has been able to induce apoptosis in colon cancer cells by injecting the cells with a tiny molecule that binds to FAP-1, a protein he discovered in the Fas signaling pathway. By binding to FAP-1, the tripeptide molecule kills colon-cancer cells by turning on their apoptotic machinery.

Thomas Franke and the Akt signaling pathway  Dr. Thomas Franke, assistant professor of pharmacology, has focused most of his work in apoptosis on the Akt pathway, so named for one of its central characters. Originally discovered as a viral oncogene, Dr. Franke cloned and characterized the cellular Akt gene in mice in 1993. Although he knew that the human homologue is overexpressed in some forms of ovarian cancer, it was considered an "orphan" because no one knew what function the Akt protein served or what molecules it interacts with. In 1995, Dr. Franke linked Akt to a growth-factor receptor, PDGF, and an important enzyme, PI3-kinase, which regulates numerous physiological processes. In 1997, Dr. Franke outlined the explicit biochemical steps leading from growth-factor receptors through PI3-kinase to the activation of Akt protein and showed that the pathway controlled cell survival in neurons and hematopoietic cells. "The . orphan. kinase has now moved to center stage as a crucial regulator of life and death decisions emanating from the cell membrane," wrote Dr. Brian A. Hemmings in a commentary accompanying two papers in Science. Since joining the P&S faculty in August 1997, Dr. Franke has helped identify many of the downstream targets of Akt, of which about a dozen have been identified. Last November he helped show that Akt inactivates caspase 9, the "effector" caspase that activates other caspases in the final march to death. In March he helped show that Akt promotes the production of nitric oxide in blood vessels. And in April he contributed to a paper showing that calcium influx induces apoptosis by triggering phosphorylation of BAD, a downstream target of Akt. "It has been wonderful to see how this pathway has fallen into place," says Dr. Franke. Dr. Franke is now turning his attention to the role of Akt in cancer and heart disease. He is studying how the expression of caspase 9 is altered in patients with ovarian cancer. He is also beginning a study of the role Akt plays in prevention of apoptosis in cardiac myocytes. "You want your research to have human implications," says Dr. Franke. "Columbia is a very good environment for connecting basic research to clinical research." The signaling pathways elucidated by Drs. Sato and Franke trigger apoptosis in many cells. But the biological role that apoptosis plays and the signals that trigger it vary from tissue to tissue. Several P&S researchers have studied the biology of apoptosis in varied tissues ranging from lymph nodes to neurons and the prostate.

Seth Lederman and B-cell selection  Dr. Seth Lederman, associate professor of medicine, showed how apoptosis helps select B cells that make infection-fighting antibodies. Each B cell is capable of making a unique antibody. As with many other cell types, apoptosis is the default program for B cells; they commit suicide before maturing unless they receive a signal that saves them. Helper T cells deliver the signal that saves B cells from apoptosis. Dr. Lederman described the molecules on the surfaces of B cells and T cells, known as CD40 and CD40-ligand, respectively, that interact to save B cells. "Our studies showed how the antibody response develops through apoptosis and the rescue of selected cells," says Dr. Lederman. The selection of B cells occurs in the germinal centers of lymph nodes. In those centers dendritic cells release a limited number of antigens, which are fragments of the invading organism. B cells compete to bind with the antigens. The B cells whose antibodies bind most strongly to the antigen, consume it, and present it on their surface. Helper T cells bind to the B cell presenting those antigens and rescue it from death. The saved B cell then begins to proliferate. But the antibody genes in B cells mutate frequently. As a result, many of the daughter cells produce antibodies slightly different from the parent B cell. The daughter cells once again compete to bind the antigen. The one making the antibody with the best fit, slightly better than the parent cell, is saved by the T cell. The B cells go through several rounds of this accelerated evolutionary process in which the "fittest" B cells survive to produce another generation of B cells. It eventually produces a B cell whose antibodies bind tightly to the antigen and neutralize the foreign body. At that point, it is released from the germinal center and begins producing massive numbers of antibodies. "The process of selecting B cells and refining the structure of their antibody molecules is iterative," says Dr. Lederman. "Like other forms of evolution, it tinkers with the antibody until the structure is best." Dr. Lederman believes apoptosis has transformed the field of physiology. Physiologists used to define homeostasis largely in terms of the plasma levels of various molecules, such as electrolytes or hormones. Now they take a more cellular approach. "Apoptosis has focused physiologists. attention on homeostatic mechanisms that balance cell division and death."

Ralph Buttyan: apoptosis in prostate cancer  Dr. Ralph Buttyan, associate professor of pathology (in urology), was probably the first person at P&S to study apoptosis, although he didn. t recognize its signature when he first observed it in 1984. Apoptosis was still an obscure subject at that time, but it was well known that prostate cells require the male steroid hormones known as androgens to survive. That is why various forms of castration therapy are used to treat advanced prostate cancers. When Dr. Buttyan castrated rats with prostate cancer, most of the prostate cells did dutifully die. But Dr. Buttyan made the enigmatic observation that the prostate cells were becoming more active just before their demise, synthesizing new gene products that are often made by cells in the process of division. "This was such an unexpected observation," says Dr. Buttyan 15 years later. "That. s why I decided to pursue it." A short time later, Dr. Buttyan came across the literature of Kerr and his colleagues describing apoptosis, which helped him understand what was occurring in those prostate cells and led to the publication of his findings in 1988. Several of the gene products he saw in the dying prostate cells, such as c-myc, c-fos, and p53, have since been proved to participate in apoptotic signaling and regulation. As Dr. Buttyan and others have since discovered, apoptosis is closely linked to mitosis. "Proliferation is a good time to weed out bad cells," says Dr. Buttyan. "It presents the perfect opportunity for a cell to recognize internal genetic problems so that they can be eliminated before they are passed on to progenitors." Dr. Buttyan has continued to study the role of apoptosis in the prostate, especially in prostate cancers. His recent studies have shown that castration induces apoptosis in endothelial cells lining the blood vessels of the prostate. When those cells die, the blood supply to the cancerous tissue is disrupted, depriving other cells in the prostate of oxygen. While oxygen deprivation eventually causes such damage that cells undergo necrosis, it appears that the prostate cells sense this impending disaster and trigger their own suicidal program, thus removing those cells in a more controlled and cleaner process. "I now tell urologists, . In your use of castration therapy for prostate cancer, you have been using antiangiogenic ther-apy for 50 years,. " says Dr. Buttyan. Antiangiogenic therapy, in which tumors are starved of their blood supply, has received attention recently as a promising anticancer therapy. Some prostate cancer cells inevitably survive castration. Dr. Buttyan has shown that these cells express high levels of the antiapoptotic protein bcl-2, making them resistant not only to castration but also to other important cancer therapies. Dr. Buttyan and associate research scientist Thambi Dorai have developed an antisense ribozyme that degrades bcl-2 messenger RNA in cells and destroys the ability of prostate cancer cells to make bcl-2 protein. They are now testing the approach in animal models and believe that it holds promise as a treatment for prostate cancer cells that resist other forms of therapy. Dr. Bernard Weinstein, the Frode Jensen Professor of Medicine, says the concept of apoptosis has completely transformed the study of cancer. Until recently, cancer was viewed as a disease that results from the uncontrolled proliferation of cells. But today, it has become clear that the failure of cells to die at the appropriate time is just as important in the development of cancer. "Apoptosis is a dominant theme in cancer research," says Dr. Weinstein. "The concept has provided insight into why drugs work or don. t work and has opened up whole new lines of research. Investigators are looking for new forms of cancer therapy by screening for drugs that cause apoptosis."

Lloyd Greene and neuronal survival Dr. Lloyd Greene, professor of pathology (in the Center for Neurobiology and Behavior), has for many years studied the growth and differentiation of neurons. His primary tool has been a cell line, PC12, that he and Dr. Arthur Tischler developed in 1975. (Dr. Greene jokes that some of the PC12 cells in his laboratory are older than some of the people working there.) PC12 cells, originally derived from tumors in the adrenal glands of rats, develop into cells resembling sympathetic neurons when exposed to nerve growth factor (NGF). They are one of the few models of neuronal cells and are widely used in neuronal research. In the early 1990s Dr. Greene became interested in how NGF promotes cell survival. One of his students, Anna Batistatou, suggested that depriving neurons of NGF might trigger apoptosis. Although researchers at a recent conference reached a consensus that apoptosis probably does not occur in neurons, Dr. Greene agreed to look for it anyway. Dr. Green and Ms. Batistatou became one of the first groups to show that neurons do indeed undergo apoptosis when deprived of growth factors. In subsequent years, Dr. Greene, working with Drs. Carol Troy, Leonidas Stefanis, and Michael Shelanski, has described many features of neuronal apoptosis. The researchers learned that removal of NGF initiates changes in the expression and activity of molecules associated with the cell cycle. Mature neurons don't normally divide. "When you turn on the cell cycle in neurons, then they are in trouble and can die," says Dr. Greene. But Dr. Greene and his colleagues have uncovered evidence that the signal turning on the cell-cycle genes cannot initiate apoptosis by itself. It appears that a second signaling pathway, one normally associated with stress, also must be stimulated for neurons to undergo apoptosis. Dr. Greene believes neurons may need two independent signals to begin apoptosis because they are so important to an organism. s survival and quite difficult to replace. "You don. t want neurons to make a mistake and die if they don. t have to," says Dr. Greene. His "two-key" hypothesis equates neuronal apoptosis to nuclear missiles; both require two independent signals for their launch. Dr. Greene has turned his attention to the mitochondrion and the changes it undergoes before releasing cytochrome c. He is using a powerful technique developed by cancer researcher Bert Vogelstein to determine not only genes newly expressed during apoptosis, but also changing levels of gene expression. This technique, called SAGE, requires high-throughput sequencing of genes. To that end, Drs. Greene and Shelanski have purchased an AB prism 3700, the fastest gene sequencer available and the first one in New York City. Dr. Greene says he is most surprised by the fact that the three genes in C. elegans originally discovered by Robert Horvitz describe the basic process of apoptosis in higher animals. One gene codes for the enzyme that cleaves the cell. s proteins. One codes for a protein that activates that enzyme, and another codes for a protein that inhibits apoptosis. "In mammalian cells, it has become much more baroque," says Dr. Greene. "At the beginning I thought it would be a switch. You turn it on and the cell dies. It. s not that easy. The more I think and read about this the more complex it gets. It. s getting embarrassingly complex."

Robert Burke and Parkinson's disease Neurons undergo apoptosis in several neurodegenerative diseases, including Alzheimer' s disease, Huntington. s disease, and amyotrophic lateral sclerosis. But the exact role of apoptosis is unclear. "Is it part of the disease? Can you stop it and make cells survive? Or is the neuron so damaged that it must be removed?" asks Dr. Shelanski. Dr. Robert Burke, professor of neurology (in pathology), has focused on unraveling the role of apoptosis in Parkinson. s disease. In that movement disorder, dopamine-emitting neurons in the substantia nigra region of the brain die prematurely. In the early 1990s Dr. Burke discovered that the dopamine neurons undergo apoptosis during development if the target neurons are destroyed. The target neurons are believed to provide growth factors to prevent the default apoptosis. Dr. Burke has also shown that dopamine neurons undergo apoptosis in a rat model of Parkinson's disease. "There. s no question that apoptosis occurs in dopamine neurons," says Dr. Burke. The key will be showing definitively that the dopamine neurons in the substantia nigra undergo apoptosis in the brains of Parkinson. s patients. But it can be very hard to spot since apoptosis occurs rapidly and leaves no trace. Dr. Burke has discovered that those neurons undergoing apoptosis in rats express the caspase 3 gene. He plans to search for expression of that gene in the brains of Parkinson. s patients who have died.

Beth Levine and apoptosis in viruses  Apoptosis plays a role in more than just the proper development and protection of a multicellular organism. It is a weapon in the conflict between species. When a cell becomes infected by a virus, it often undergoes apoptosis. "The cell commits this altruistic suicide for the good of the organism," says Dr. Beth Levine, assistant professor of medicine. "Apoptosis may have evolved as a defense mechanism." But in the arms race of interspecies conflict, organisms often evolve new ways to overcome the defenses. Viruses want a cell to survive long enough for the viruses inside to reproduce many times. Then the cell lyses, releasing the new viruses to infect other cells. "Viruses want to keep a cell alive. They have evolved a number of defenses to block apoptosis," says Dr. Levine. One of the first people to study apoptosis in connection with viruses, Dr. Levine explains that many viruses have picked up antiapoptotic genes from the cells they infect. For example, in 1997 Dr. Sato and Dr. Yuan Chang, associate professor of pathology, showed that the Kaposi. s sarcoma-associated herpesvirus expresses a bcl-2 gene that prevents infected cells from committing suicide. Most of Dr. Levine. s work has focused on the Sindbis virus, one of a family of RNA viruses that can cause acute encephalitis. The Sindbis virus is an evolutionarily simple virus with such a tiny genome that it has no gene to prevent apoptosis. But it has developed its own strategy for avoiding removal by apoptosis. "Sindbis virus preferentially infects neurons, which are more resistant to virus-induced apoptosis," says Dr. Levine. Most virally infected cells present a signal on their surface that tells killer T cells to destroy them. But neurons do not. As a result the immune system can. t kill the infected neurons, allowing the virus to persist in those cells. Dr. Levine showed that the Sindbis virus, previously believed to cause only acute infections, persists indefinitely in neurons. She believes other viruses also may establish persistent infections within neurons. "Our neurons are probably loaded with viruses," says Dr. Levine. Dr. Levine also showed that bcl-2 protects neurons from the apoptosis that occurs when Sindbis infects other cell types. Bcl-2 also slows the replication of the Sindbis virus inside neurons. Recently Dr. Levine discovered a new gene, beclin 1, whose protein product interacts with bcl-2 to protect neurons against apoptosis. Counterintuitively, it also appears to play a role in tumor suppression and is mutated in several breast-cancer cell lines. Her most recent work has identified beclin 1 as the first mammalian gene known to control the process of autophagy. "Autophagy is the process by which cells degrade their own cytoplasmic contents," says Dr. Levine. "That process is defective in tumor cells, which raises the interesting possibility that there are genetic links between the regulation of apoptosis, tumor suppression, and autophagy."

Future directions and therapy  Many more people at P&S are studying apoptosis in a variety of contexts. Among them are Dr. Jean Gautier, assistant professor of genetics and development (in dermatology), who has described apoptosis in the earliest stages of development in the frog, Xenopus. Dr. Ramon Parsons, assistant professor of pathol-ogy, has shown the tumor suppressor he co-discovered, PTEN, is an antiapoptic element in the Akt signaling pathway. Dr. Steven Moss, assistant professor of medicine, has shown that Helicobacter pylori, the bacterium associated with ulcers and other gastrointestinal disorders, induces apoptosis in the gut. Dr. Abraham Spector, the Malcolm P. Aldrich Research Professor of Ophthalmology, has shown that apoptosis contributes to the development of cataracts in the eyes. P&S researchers also have seen evidence of apoptosis in stroke, heart attacks, and diseases of the kidney. It is also becoming clear that several anticancer therapies work by promoting apoptosis. DNA damage caused by radiation causes apoptosis. Several established chemotherapies, such as taxol, also have been shown to induce apoptosis. New therapies developed as a result of new knowledge about apoptosis are mostly in the preclinical stages. Several researchers have shown promising results in cell-culture and animal studies. The drug sulindac sulfone, a derivative of the non-steroidal anti-inflammatory drug sulindac, is being tested in clinical trials as a therapy against prostate cancer. Dr. Weinstein first showed that the drug and related compounds induce apoptosis in human prostate cancer cells in culture. This occurs even if the cancer cells lack the important tumor suppressor gene p53, have increased expression of bcl-2, or have become androgen-independent. Based on those findings, Dr. Erik Goluboff, assistant professor of urology, has begun a Phase III trial of sulindac sulfone on 90 patients whose prostates have been removed but whose rising levels of prostate-specific antigen indicate that the cancer has not been vanquished. Apoptosis has brought a certain balance to biology. Scientists are now learning as much about the end of a cell. s life as they already know about its beginning. And in its complexity they see opportunities for therapy. They realize that a better understanding of death may one day save lives.