Programmed Cell Death

Published on 28/02/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 3926 times

CHAPTER 46 Programmed Cell Death

The Necessity for Cell Death in Multicellular Organisms

The ability to undergo programmed cell death (Box 46-1) is a built-in latent capacity in virtually all cells of multicellular organisms. Cell death is important for embryonic development, maintenance of tissue homeostasis, establishment of immune self-tolerance, killing by immune effector cells, and regulation of cell viability by hormones and growth factors. It has been proposed that most metazoan cells will die if they fail to receive survival signals from other cells. Abnormalities of the cell death program contribute to a number of diseases, including cancer, Alzheimer’s disease, and acquired immune deficiency syndrome (AIDS).

BOX 46-1 Key Terms

Programmed Cell Death: An active cellular process that culminates in cell death. This may occur in response to developmental or environmental cues or as a response to physiological damage detected by the cell’s internal surveillance networks.

Necrosis (Accidental Cell Death): Cell death that results from irreversible injury to the cell. Cell membranes swell and become permeable. Lytic enzymes destroy the cellular contents, which then leak out into the intercellular space, leading to the mounting of an inflammatory response.

Apoptosis: One type of programmed cell death that initially was characterized by a particular pattern of morphologic changes but now is defined by the action of molecular pathways involving cell surface receptors or mitochondria and resulting in the activation of specialized proteases. The name comes from the ancient Greek, referring to shedding of the petals from flowers or leaves from trees. Apoptosis is observed in all metazoans, including both plants and animals.

Apoptotic death occurs in two phases. During the latent phase, the cell looks morphologically normal but is actively making preparations for death. The execution phase is characterized by a series of dramatic structural and biochemical changes that culminate in the fragmentation of the cell into membrane-enclosed apoptotic bodies. Activities that cause cells to undergo apoptosis are said to be pro-apoptotic. Activities that protect cells from apoptosis are said to be anti-apoptotic.

Programmed Cell Death versus Accidental Cell Death: Apoptosis versus Necrosis

Although cells die in many ways, it is useful to focus on the two poles of this spectrum: apoptosis and necrosis. Apoptosis is the most commonly described pathway for programmed cell death, which is cellular suicide resulting from activation of a dedicated intracellular program (Fig. 46-1). Often, these cells appear completely healthy prior to committing suicide. At the other end of the spectrum is necrosis, also called accidental cell death, which occurs when cells receive a structural or chemical insult that kills them outright (Fig. 46-2). Examples of such insults include extremes of temperature and physical trauma. The cell itself can also initiate necrosis in response to certain stimuli, particularly when induction of apoptosis is inhibited. In contrast to the orderly biochemical pathways of apoptosis, which involve the action of enzyme cascades and the consumption of ATP, necrosis typically involves a collapse of normal cell physiology as a result of ATP depletion.

Figure 46-1 apoptosis—active cellular suicide—typically affects single cells. Neighboring cells remain healthy. Apoptotic cell death usually does not lead to an inflammatory response.

Figure 46-2 necrosis is a result of injury to cells. Typically, groups of cells are affected. In most cases, necrotic cell death leads to an inflammatory response (red “angry” macrophages).

Necrosis corresponds to what most of us naively imagine cell death would be like. Owing to lack of cellular homeostasis, water rushes into the dying cell, causing it to swell greatly so that the plasma and organelle membranes burst. As a result, the cell undergoes a generalized process of autodigestion and dissolution, culminating in the spilling of the cytoplasmic contents out into the surroundings (Fig. 46-2). This, in turn, produces local inflammation as phagocytic cells are activated, flock to the site, and ingest the debris (see Chapter 22). Because agents that damage cells act over areas that are large in comparison to the size of a single cell, necrosis often involves large groups of neighboring cells.

In contrast to necrosis, apoptotic cells shrink rather than swelling, as part of a reproducible pattern of structural alterations of both the nucleus and cytoplasm (Fig. 46-1). Apoptosis is a two-stage process. On receipt of the pro-apoptotic signal that triggers the pathway to death, cells enter a latent phase of apoptosis (Fig. 46-3). Although committed to a pathway that leads to their inevitable demise at some later time, cells in the latent phase look as healthy as their neighbors. The duration of the latent phase of apoptosis is extremely variable, ranging from a few hours to several days. The reason for this variability is not known.

Figure 46-3 the two phases of apoptosis. Note that the latent phase can be subdivided into two stages: a condemned stage, during which the cell is proceeding on a pathway toward death but can still be rescued if it is exposed to anti-apoptotic activities, and a committed stage, beyond which rescue is impossible.

Ultimately, the cells enter the execution phase of apoptosis, lasting about an hour, during which they undergo dramatic morphologic and physiological changes. These include (1) loss of microvilli and intercellular junctions (Fig. 46-4); (2) shrinkage of the cytoplasm; (3) dramatic changes in cytoplasmic motility with activation of violent blebbing (Fig. 46-5); (4) loss of plasma membrane asymmetry, with the distribution of phosphatidylserine being randomized so that it appears in the outer membrane leaflet; (5) hypercondensation of the chromatin and its collapse against the nuclear periphery; and (6) the “explosive” fragmentation of the cell into membrane-enclosed apoptotic bodies that contain remnants of the nucleus, mitochondria, and other organelles. The plasma membrane retains its integrity throughout the entire process. All of these changes are instigated by the action of a specific set of death-inducing proteases, discussed at length later.

Figure 46-4 scanning electron micrograph of intact and apoptotic mouse sarcoma cells. Intact cells are covered with microvilli, whereas apoptotic cells have numerous smooth blebs. These cells were stimulated to undergo apoptosis as a result of interference with RNA metabolism.

(From Wyllie AH, Kerr JFR, Currie AR: Cell death: The significance of apoptosis. Int Rev Cytol 68:251–305, 1980.)

Figure 46-5 apoptosis of a transformed pig kidney cell following exposure to etoposide, a drug used in cancer chemotherapy. The dramatic cytoplasmic blebbing results in the disassembly of the cell into membrane-enclosed vesicles.

(Courtesy of L. M. Martins and K. Samejima, Wellcome Trust Institute for Cell Biology, University of Edinburgh, Scotland.)

In tissues, apoptotic bodies are rapidly phagocytosed by surrounding cells that recognize the phosphatidylserine and other markers exposed on their surface (Fig. 46-6). Apoptosis can thus be considered to be the disassembly of the cell into “bite-sized” vesicles. Because these vesicles remain membrane bound, the cellular contents are not released into the environment. It is important to note that surface markers on apoptotic bodies cause cells that ingest them to secrete anti-inflammatory cytokines. As a result, apoptotic death does not lead to an inflammatory response.

Figure 46-6 phagocytosis of apoptotic cells. a–c, Phagocytosis that occurs when cells express “eat me” signals results in the production of anti-inflammatory cytokines. D, Electron micrograph of a phagocytosed apoptotic body containing a nuclear fragment. The nucleus of the epithelial cell that engulfed this apoptotic body is shown at left. In this case, apoptosis occurred during allo-graft rejection in a pig.

(A, Based on Lauber K, Blumenthal SG, Waibel M, Wesselborg S: Clearance of apoptotic cells: Getting rid of the corpses. Mol Cell 12:277–287, 2004. B, From Wyllie AH, Kerr JFR, Currie AR: Cell death: The significance of apoptosis. Int Rev Cytol 68:251–305, 1980.)

Nonapoptotic Programmed Cell Death

The terms apoptosis and programmed cell death are sometimes viewed as synonymous. However, in a number of well-documented systems, cells undergo programmed cell death without the dramatic structural changes that classically define apoptosis. Thus, all apoptosis is programmed cell death, but the converse is not necessarily true.

When the adult tobacco hawkmoth emerges from its cocoon, its intersegmental muscles undergo programmed cell death that differs in several ways from apoptosis as described earlier. The chromatin does not condense; DNA is not digested; and cytoplasm does not “boil.” Instead, a polyubiquitin gene is induced and plays an important role in intracellular protein degradation (see Chapter 23). Thus, although these muscle cells unquestionably undergo programmed cell death, they apparently do not use the apoptosis pathway.

Classes of Cells That Undergo Programmed Cell Death

At least six distinct classes of cells undergo programmed cell death (examples are given in Fig. 46-7).

Figure 46-7 a, Types of cells that undergo programmed cell death. B–D, Programmed cell death in the embryonic mouse paw. At day 12.5 of development, the digits are fully connected by webbing. By day 13.5, the webbing has started to die, and by day 14.5, all of the webbing cells are gone. E, Nuclei of cells undergoing programmed cell death take up acridine orange, whereas cells of the surrounding healthy tissue do not.

(Micrographs courtesy of William Wood and Paul Martin, Department of Anatomy and Developmental Biology, University College of London, England.)

Developmentally Defective Cells

During molecular maturation of T-lymphocyte antigen receptors (see Figs. 27-8 and 28-8), immature T cells in the thymus (known as thymocytes) rearrange the genes encoding the receptor a and b chains. Many newly created receptors bind to foreign antigens, but others interact with self-antigens. Cells with receptors recognizing self-antigens are potentially harmful and are eliminated through apoptosis in a process known as negative selection (Fig. 46-8). The drug cyclosporin A, which inhibits apoptosis in thymocytes, can cause autoimmune disease.

Figure 46-8 examples of signals that promote differentiation or programmed cell death of immature thymocytes in the thymus and mature t cells in the periphery. Thymocytes that make functional T-cell receptors and do not recognize self-antigens mature, provided that they receive survival signals, such as interleukin-7. Thymocytes undergo apoptosis if they produce defective T-cell receptor, recognize self-antigens, suffer DNA damage, or receive a death stimulus (glucocorticoid hormone). More than 95% of immature thymocytes die without leaving the thymus.

(Based on Strasser A: The role of BH3-only proteins in the immune system. Nat Rev Immunol 5:189–200, 2005.)

To function properly, the T-cell receptor must recognize major histocompatibility complex (MHC) glycoproteins on other cells during antigen presentation (see Fig. 27-8). T lymphocytes whose T-cell receptors cannot interact with the spectrum of MHC glycoproteins expressed in a given individual are ineffective in the immune response. These cells die by apoptosis in a process known as positive selection (Fig. 46-8). Overall, defects in T-cell receptor assembly are extremely common, and up to 95% of immature T cells die by apoptosis without leaving the thymus.

Similar positive and negative selection steps occur during the maturation of B lymphocytes (see Fig. 28-8), which is accomplished by a combination of gene rearrangements and facilitated mutagenesis. B lymphocytes expressing antibodies directed against self-antigens or producing antibodies whose affinity for antigen is below a critical threshold are eliminated through apoptosis.

Excess Cells

The use of programmed cell death for quality control during development is not limited to the immune system but is also extremely important during brain development. Embryonic ganglia often have many more neurons than are required to enervate their target muscles. Production of excess cells is part of a Darwinian strategy to ensure that a sufficient number of axons reach their targets. Excess neurons that fail to make appropriate connections have no function and are eliminated by programmed cell death. Up to 80% of neurons in certain developing ganglia die in this way. Because of the importance of apoptosis during its development, the brain is often seriously affected in mice that are engineered to lack components of the apoptotic pathway.

Cells That Serve No Function

The elimination of obsolete cells whose function has been completed is most evident in organisms, such as insects and amphibians, that undergo metamorphosis during development. For example, programmed cell death initiated by a burst of thyroid hormone is responsible for resorption of the tadpole tail.

Mammals also use programmed cell death to eliminate obsolete tissues during development. For example, in humans, the digits of hands and feet are connected by a tissue webbing during embryogenesis. Cells in this webbing serve no purpose in the adult and are eliminated by programmed cell death (Fig. 46-7).

During craniofacial development, the hard palate develops from two lateral precursors, each covered in a protective layer of epithelial cells. As the two halves grow together at the midline of the nasopharynx, they remain separated by this covering of epithelium until, in response to a developmental cue, the epithelial cells at the midline undergo programmed cell death. Then the two halves of the palate can fuse. Failure of the epithelial cells to die at the appropriate time can interfere with the fusion of the bone, causing cleft palate.

Populations of cells that are fully functional may become obsolete as a result of physiological changes in the status of an organism. For example, in male mammals, certain accessory glands of the reproductive system are regulated by the levels of circulating male hormone. If hormone levels fall below a critical threshold, these organs, including the prostate, virtually disappear in a very brief time as their constituent cells undergo massive apoptotic death. Should levels of circulating androgens rise again, the remaining prostatic stem cells proliferate and reconstruct the gland. A similar cycle of growth and involution is seen in the mammary gland of female mammals, which exhibits substantial differences in size and cellular composition in the lactating and nonlactating states. Interference with survival signaling by sex hormones is one important strategy that is commonly used in the treatment of breast and prostate cancer.

Programmed cell death is also used to eliminate certain populations of cells that never served any function to begin with. The Müllerian ducts develop into the female oviduct. In male embryos, progenitors of the Müllerian ducts develop, even though they have no function. Programmed cell death eliminates the constituent cells of these embryonic ducts.

Cells Whose Cell Cycle Is Perturbed

Chapters 40 to 43 describe how biochemical circuits called checkpoints regulate the cell cycle. If DNA is damaged, checkpoint activation blocks cell-cycle progression while repair processes operate. An important downstream effector of checkpoints, the p53 transcription factor, induces the expression of genes encoding proteins that arrest the cell cycle as well as genes encoding proteins that induce cell death. It is generally thought that if the damage cannot be repaired quickly, the pro-death factors win out, and the outcome is apoptosis. Types of DNA damage that commonly trigger cell death are double-strand breaks induced by ionizing radiation and DNA breaks or other damage induced by chemotherapeutic agents.

A second important cell-cycle checkpoint regulates the transition from the G1 phase to the S phase. Passage of the restriction point (see Fig. 41-7) represents the commitment of the cell to undergo another cycle of DNA replication and division. Restriction point control centers on the regulation of the E2F family of transcription factors. However, E2F not only regulates genes that promote cell-cycle progression; it also induces the expression of genes that promote apoptosis. It is now thought that if E2F is activated too strongly, as, for example, where restriction point control has broken down (see Fig. 41-10), its function as a death inducer takes over, and the cells undergo apoptosis. Cells that die in response to inappropriate signals to proliferate include those that are infected by certain viruses or overexpress genes involved in cell proliferation (such as c-myc and c-fos [Fig. 46-14]). This ability to recognize an inappropriate stimulus to proliferate and respond to it by undergoing apoptosis may be an important defense against cancer.

Figure 46-14 Bcl-2 protects cells from oncogene-induced cell death. Chinese hamster ovary cells die when they are induced to express abnormally high levels of the c-myc protein, but simultaneous expression of the Bcl-2 protein rescues them from this effect. These cells contain many copies of the c-myc gene under control of a promoter that is activated when the cells are briefly exposed to high temperature (43°C for 90 minutes). The curves show the percentage of viable cells remaining at various times following the induction of c-myc expression. The Bcl-2 gene was introduced into these cells on a plasmid molecule under the control of a viral promoter, which is always active. The blue line represents the parental cell line lacking either the cloned c-myc or Bcl-2 genes. (Note that about 40% of these cells die following the heat treatment used to induce c-myc expression.) The yellow line shows that the cells that produce high levels of c-myc protein alone rapidly die by apoptosis. The red line shows that cells expressing both the c-myc and Bcl-2 proteins survive the treatment almost as well as the parental cells.

(From Bissonnette RP, Echeverri F, Mahboubi A, et al: Apoptotic cell death induced by c-myc is inhibited by bcl-2. Nature 359:552–556, 1992.)

Virus-Infected Cells

Cells that harbor infectious agents, such as viruses, are harmful to the organism. Cytotoxic T lymphocytes eliminate virus-infected cells by causing them to undergo programmed cell death either by apoptosis or by a second related pathway.

At least part of the loss of mature CD4⁺ T helper cells (see Fig. 28-8) in people who are infected with HIV-1 results from programmed cell death. When exposed to agents that normally stimulate cell proliferation, these cells instead undergo apoptosis. Paradoxically, it ap-pears that many of these dying cells are not themselves infected with HIV.

Chemotherapeutic Killing of Cells

Exposure of cancer cells to many of the agents that are used in chemotherapy does not kill the cells outright. Instead, they die because the drugs cause intracellular damage that acts as a signal for the induction of apoptotic cell death.

Genetic Analysis of Programmed Cell Death

Several key components that are involved in the apoptotic execution of mammalian cells were first identified by a genetic analysis of the nematode worm Caenorhabditis elegans. Because C. elegans is optically clear, it is possible to see every cell in a developing worm by using differential interference contrast optics (see Fig. 6-2). This enabled investigators to develop a complete fate map for C. elegans that traces the lineage of each cell in an adult worm back to the fertilized egg. These studies led to the surprising discovery that programmed cell death is one of the most common fates for newborn C. elegans cells. Of the 1090 somatic cells that are produced during embryogenesis of the C. elegans hermaphrodite, 131 undergo programmed cell death at reproducible locations and times.

Mutations in at least 14 C. elegans genes affect programmed cell death (Fig. 46-9). These may be divided into three classes: (1) genes that mark cells for subsequent programmed death, (2) genes that are involved in cell killing and its regulation, and (3) genes that are involved in the phagocytosis and subsequent processing of the cell corpses. These mutants are collectively known as “cell death abnormal” (ced) mutants.

Figure 46-9 genetic dissection of programmed cell death. The ced (cell death abnormal) mutants of the nematode worm C. elegans affect the killing, engulfment, and degradation stages of programmed cell death.

The three best-known cell death genes are ced-3, ced-4, and ced-9. Ced-3 and ced-4 are required for cells to undergo apoptotic programmed cell death. If either gene is inactivated, all cells throughout the organism that should die by apoptosis are reprieved. These cells remain alive and are apparently functional. Interestingly, these worms have normal life spans. This suggests that programmed cell death is not involved in the normal aging process, at least not in C. elegans. Ced-9 regulates ced-3 and ced-4. In ced-9 loss-of-function mutants, many cells die that should normally stay alive. This is deleterious for the organism, and ced-9 mutants die.

These genes all have mammalian counterparts (discussed more fully later). Ced-3 is a member of a specialized family of cell death proteases called caspases. Ced-4 is a scaffolding/adapter protein that plays an essential role in the activation of Ced-3 from its zymogen precursor. Its mammalian counterpart is apoptotic protease-activating factor-1 (Apaf-1). Ced-9 is a member of the Bcl-2 family of cell death regulators. In mammals, some Bcl-2 family members protect against cell death, whereas others actively promote cell death.

Eight C. elegans genes encode proteins that are involved in the phagocytosis and processing of cell corpses. Several are signaling proteins with roles in reorganizing the cytoskeleton to permit the cell to move toward and engulf its target. Another, the nuc-1 gene, encodes one of several nucleases that digest the DNA of the dead cell. In the worm, digestion of the DNA occurs in lysosomes of cells that ingest the corpse. In mammals, this digestion typically is initiated within the dying cell itself. The process of phagocytosis turned out to be surprisingly complex, probably involving both ligand-receptor interactions and directed cell motility.

Signals and Pathways of Apoptosis

Two principal pathways lead to cell death by apoptosis. These are introduced only briefly here. The intrinsic pathway (Fig. 46-16) is activated by internal surveillance mechanisms or signals sent (or not sent) by other cells. Signals that induce this pathway include DNA damage, exposure to chemicals that interfere with a variety of cellular pathways, excessive activation of factors that promote cell-cycle progression, and receipt of certain pro-apoptotic stimuli from the surrounding medium. Withdrawal of nutrients or of nurturing signals from the environment also activates the intrinsic pathway. Survival signals include lymphokines, such as interleukin-2 and interleukin-3, which are essential for survival of thymocytes; nerve growth factor, which is required for survival of many neurons; and extracellular matrix, which is required for survival of epithelial cells. Signals that activate the intrinsic pathway converge on mitochondria, which release key factors that drive the apoptotic response.

Figure 46-16 the intrinsic cell death pathway. a–b, This drawing shows one of several models for how the death-promoting factors, such as cytochrome c, are released by mitochondria. In this model, Bax and Bak form a pore that releases cytochrome c. Released cytochrome c binds to Apaf-1, inducing formation of the apoptosome, which binds and activates procaspase 9 via scaffold-induced oligomerization. Activated caspase 9 subsequently activates downstream effector caspases, leading to the death of the cell. C, Reconstruction of the apoptosome from cryoelectron microscopy (top) with crystal structure of Apaf-1 superimposed (bottom).

(C, top, Reprinted from Acehan D, Jiang X, Morgan DG, et al: Three-dimensional structure of the apoptosome: Implications for assembly, procaspase-9 binding, and activation. Mol Cell 9:423–432, 2002. Copyright 2002, with permission from Elsevier. C, bottom, Based on Riedl SJ, Li W, Chao Y, et al: Structure of the apoptotic protease-activating factor 1 bound to ADP. Nature 434:926–933, 2005. PDB file: 1Z6T.)

Signals from other cells are the primary triggers of the extrinsic pathway (Fig. 46-17). Direct contact with the target cell activates specific receptors that initiate this pathway, starting on the inner surface of the plasma membrane. Activation of the extrinsic pathway is one strategy that cytotoxic T lymphocytes use to kill cells that are recognized as foreign (or as harboring foreign pathogens). This pathway is also widely used to control cell populations in the immune system.

Figure 46-17 the extrinsic cell death pathway. The pathways shown here are downstream of the Fas cell death receptor. A, Preligation. B, Ligand docked on a trimerized receptor. C, Release of active caspase. D, Release of effector caspases. E, Death.

(See the text for a detailed description.)

Protein Regulators and Effectors of Apoptosis

Since the penalty for misregulation of apoptosis is inappropriate cell death, it is not surprising that the process is carefully regulated. This is essential for cells but complicates matters for students. This section first lays out the overall strategy in generic terms and then fills in some important details.

A cascade of proteases called caspases drives apoptosis. Each caspase is harmless until activated (usually by proteolytic cleavage). The cascade starts with the activation of a small number of initiator caspases, which activate numerous effector caspases. The ability of effector caspases to activate further initiators and effectors further amplifies the cascade.

This strategy of employing amplification and positive feedback has two powerful advantages. First, it can provide a very rapid change in the state of the cytoplasm, from pro-life to pro-death within seconds. Second, because a relatively small number of initiator caspases initiate the cascade, these enzymes are feasible targets for negative regulators that can rapidly quell responses that are initiated under borderline conditions or by mistake. This is beneficial but also complicates the overall system. If initiator caspases start apoptosis and are then inactivated by suppressers, how does the response ever take hold? The answer is at least one more level of regulation: inhibitors of the inhibitors.

The following sections discuss the workhorses of apoptosis—the caspases—followed by regulation of the response.

Caspases

Caspases (cysteine aspartases) are specialized proteases with a cysteine in their active site that cleave on the C-terminal side of aspartate residues. Caspases inactivate cellular survival pathways and specifically activate other factors that promote cell death.

C. elegans has three caspases, one of which (Ced-3) is essential for cell death. In contrast, mammals have at least 13 caspase genes (Fig. 46-10A). Analysis based on sequence comparisons divides caspases into two major subfamilies. The caspase 1 subfamily encodes enzymes that process pro-interleukin-1b to yield mature interleukin-1b. Macrophages secrete this cytokine, which is involved in causing inflammation. In contrast, the caspase 3 subfamily of enzymes participates almost exclusively in apoptotic cell death.

Figure 46-10 introduction to caspases. A, The 13 mammalian caspases fall into three groups. Where it has been determined, the portions of the zymogens that give rise to the large (blue) and small (yellow) subunits are shown. B, Initiator caspases have large prodomains that participate in subcellular targeting. Two procaspases come together to form the active enzyme. C–D, Crystal structure of caspases 1 and 3. The catalytic residues come primarily from the large subunit (blue). In the pro-inflammatory caspase 1–like enzymes, the catalytic site is relatively open. In the caspase 3–like enzymes (involved in apoptosis), the small subunit (yellow) forms a “hood” that limits access to the active site. The space-filling structure (red) represents a peptide inhibitor covalently bound in the active site of the enzyme.

Like many proteases, caspases are synthesized as inactive zymogens. All living vertebrate cells apparently synthesize these zymogens constitutively. The caspase zymogen consists of three domains and an N-terminal prodomain followed by the large and small subunits of the mature enzyme (Fig. 46-10A-B). These three domains are separated by aspartate residues, the cleavage target for caspases. Caspase zymogens are usually activated by zymogen cleavage and release of the prodomains. Following cleavage, the two large and two small subunits associate in a compact, block-like heterotetrameric molecule (Fig. 46-10B-D). Cleavage of the zymogens permits a major conformational change in the polypeptide, creating two stable active site pockets between the large and small subunits.

Two classes of caspases are involved in cell death. Initiator caspases have long prodomains (Fig. 46-10A). These zymogens exist as monomers in cells and become autoactivated when scaffolding cofactors promote their aggregation. Activation is thought to involve dimerization of the zymogens and might not necessarily require zymogen cleavage. Sequences within the extended prodomains are involved in targeting the initiator caspase zymogens to the appropriate cellular locations and in interactions with scaffolding factors.

Effector caspase zymogens are monomers with short prodomains in healthy cells. These inactive enzymes are incapable of autoactivation under normal circumstances. Instead, they are activated through cleavage by initiator caspases.

Scaffolding proteins and adapters play an essential role in the activation of initiator caspases. For the intrinsic pathway of cell death, factors released from mitochondria (discussed later) activate the scaffold protein apoptotic protease activating factor 1 (Apaf-1). Active Apaf-1 forms a seven-spoked ring-like structure called the apoptosome (Fig. 46-16). Binding of the procaspase 9 zymogen to this structure promotes dimerization and activation of the enzyme, which appears to achieve full activity without the necessity of zymogen cleavage. The scaffold proteins for the extrinsic death pathway are cytoplasmic domains of cell surface receptors. When these receptors bind their ligands on the surface of other cells, they form stable trimeric complexes that recruit adapter proteins, which have multiple protein-protein interaction motifs and link the procaspase 8 zymogen to the receptor complex (Box 46-2). This leads to the dimerization, activation, self-cleavage, and release of active caspase 8, thereby starting the apoptotic cascade.

BOX 46-2 Matchmaking for Cell Death: The Key Is in the Domains

There are so many different proteins involved in apoptosis that even experts have a difficult time keeping them all straight. However, an understanding of the general principles is much simplified if the following principle is kept in mind. Most proteins that are involved in apoptosis regulation are built from a relatively limited number of modules, many of which act as sites for protein-protein interactions. The following are the most important modules for apoptosis.

Bcl-2 family members are defined by the presence of short regions of conserved sequence, referred to as BH domains (Bcl-2 homology). One of these, the approximately 20-residue BH3 domain, found in all Bcl-2 family members, is thought to promote complex formation between Bcl-2 family members.

The process of caspase targeting and activation is regulated by three domains that, although they are not significantly related to one another at the level of amino acid sequence, all adopt a similar structure in solution. All are regions of approximately 80 to 90 residues that form a characteristic arrangement of six α-helix bundles.

1. The death domain (DD) is found in many proteins that are involved in signaling pathways related to cell death. These include cell death receptors, such as Fas (and others), and adapter molecules, such as FADD.

2. The death effector domain (DED) is found in adapters such as FADD, the prodomains of caspases 8 and 10, and certain inhibitors of apoptosis.

3. The caspase recruitment domain (CARD) is found both in a number of adapter proteins involved in cell death, including CED-4 and Apaf-1, and in caspases 1, 2, and 9.

In general, these domains prefer to interact with themselves (i.e., DD-DD, DED-DED, and CARD-CARD). Such interactions are said to be homophilic. As a result, when a new apoptosis effector protein is cloned, it is possible to predict from an analysis of the sequence which of the known proteins it is most likely to inter-act with.

Caspases are selective enzymes that cleave a relatively limited subset of cellular proteins (Fig. 46-11). Some targets are structural proteins, but many are involved in cellular signaling. For example, caspases cleave several protein kinases. Many kinases have autoregulatory domains that enable them to be switched on and off in response to physiological stimuli (see Fig. 25-4). Caspase cleavage often neatly removes these regulatory domains, thereby producing constitutively active enzymes. Presumably, these unregulated kinases then activate factors that promote cell death. Caspases also cleave and inactivate a number of proteins that normally function in the detection and repair of DNA damage.

Figure 46-11 some ways that caspases promote cell death. A, Some of the many proteins cleaved by caspases in apoptotic cell death. Proteins shown in green normally have a role in keeping the cell alive and are inactivated by caspases. Proteins shown in red are turned into active death-promoting factors as a result of caspase cleavage. Proteins shown in black are not cleaved and are included to show the pathways that are affected by cleavage. Caspases inactivate a number of pathways that promote cell survival, thereby strongly reinforcing the decision of the cell to die. B, Some of the roles of caspases in disassembly of the nucleus.

Caspases also act on a number of targets that directly promote cell death. The most obvious example of this is caspase activation of other caspases on the death cascade. Caspases also act indirectly to cause the release from mitochondria of factors that promote cell death through the intrinsic pathway. Caspase cleavage of an inhibitory chaperone is responsible for activation of the nuclease that ultimately destroys the chromosomal DNA of most cells undergoing apoptosis (see later discussion).

Natural Caspase Inhibitors

Because most healthy cells express initiator procaspases with the potential to oligomerize by mistake and kill the cell, it is important to have a mechanism that dampens this “noise” in the pro-apoptotic pathway. The inhibitor of apoptosis protein (IAP) family is defined by the presence of a motif of approximately 80 amino acids known as a baculovirus IAP repeat domain. This is a type of Zn²⁺ finger (Fig. 15-17) that mediates protein-protein interactions. IAP proteins inhibit caspases in two ways. First, they bind the caspase and invade the active site, thereby blocking its access to substrates. Second, several IAPs are also E3 ubiquitin ligases (see Fig. 23-8). When they bind caspases, they ubiquitinate them, thereby tagging them for destruction by proteasomes.

If IAP proteins inactivate caspases, then how is the apoptotic response ever initiated? Cells also express an antidote for the IAPs. This protein, known as second mitochondrial activator of caspases (Smac or DIABLO), is normally sequestered in mitochondria. It is released when the intrinsic pathway of apoptosis is initiated.

IAPs were discovered in studies of the mechanisms viruses use to avoid being eliminated by cell death. When viruses infect cells and disassemble their capsids, they become vulnerable to suicide defense mechanisms: If cells can kill themselves before the virus has had time to complete its life cycle, they will take the virus with them, and the organism will survive. To defend against this, viruses pilfer cellular proteins and adapt them for their own means. For example, insect baculoviruses make two proteins that inhibit apoptosis, keeping the cell alive long enough for the virus to reproduce. One of these, IAP, was derived from a cellular gene. The origin of the second, p35, is less clear. p35 is a broad-spectrum caspase inhibitor that is thought to work by a serpin-like mechanism. Serpins are special protease substrates that, on cleavage, form a tight complex with the enzyme, thereby inactivating it. Several mammalian pox viruses also make a serpin-like inhibitor of certain caspases called CrmA.

CAD Nuclease and Its Chaperone ICAD

During apoptosis, the chromosomal DNA is destroyed. The many nucleases involved in cleaving the cellular DNA during (and after) apoptotic cell death fall into two classes. Cell autonomous nucleases degrade the DNA from within the dying cell (Fig. 46-12A). The best known is the caspase-activated DNase (CAD; see later discussion). In some cell types, a mitochondrial nuclease known as endonuclease G may also be involved. Cell autonomous nucleases are dispensable for cell death and for the life of the organism. They might have evolved to eliminate viral DNA as part of the suicide defense response described in the previous section.

Figure 46-12 the nucleases that digest the cellular dna during apoptosis. A, In apoptosis, the DNA is digested first to large fragments and later to nucleosome-sized pieces (see Fig. 13-1) by cell autonomous nucleases expressed within the dying cell. Waste management nucleases made by other cells also have an essential role in cleaning up apoptotic and necrotic debris. B, The predominant cell autonomous nuclease (CAD) has a scissors-like structure. C, ICAD is an inhibitory chaperone for CAD, promoting its folding on the ribosome and continuing as an inhibitor when CAD is stored in the nucleus. ICAD cleavage leads to CAD activation. D, Cleavage of the chromosomal DNA by CAD during chemotherapy-induced apoptosis of a leukemia cell line. DNA separated according to size by electrophoresis on an agarose gel was stained with ethidium bromide.

E, Activated CAD causes chromatin condensation and appearance of an apoptotic morphology in isolated cell nuclei. Cloned CAD and ICAD were expressed together in E. coli (the expression vector is diagrammed at right) and incubated with nuclei. ICAD cleavage with caspase 3 released active CAD, which degrades the nuclear DNA (Lane 3). Other lanes: DNA gel size markers (left); nuclei incubated with buffer or caspase 3 alone (Lanes 1 and 2, respectively); same experiment as in Lane 3 but performed by using a mutant ICAD that could not be cleaved by caspase 3 (Lane 4). To the right is an electron micrograph of a thin section of one nucleus with condensed chromatin at the nuclear periphery.

(A, Based on Samejima K, Earnshaw WC: Trashing the genome: The role of nucleases during apoptosis. Nat Rev Mol Cell Biol 6:677–688, 2005. B, PDB file: 1V0D. Structure described in Woo EJ, Kim YG, Kim MS, et al: Structural mechanism for inactivation and activation of CAD/DFF40 in the apoptotic pathway. Mol Cell 14:531–539, 2004. D, From Kaufmann SH: Induction of endonucleolytic DNA cleavage in human acute myelogenous leukemia cells by etoposide, camptothecin, and other cytotoxic anticancer drugs: A cautionary note. Cancer Res 49:5870–5878, 1989. E, Courtesy of K. Samejima, Wellcome Trust Institute for Cell Biology, University of Edinburgh, Scotland.)

Waste management nucleases clean up the debris after cells die. They either function within lysosomes of cells that have phagocytosed apoptotic cell fragments or are secreted and function in the extracellular space. DNase II, one of the most important waste management nucleases, is essential for life. Mouse embryos that lack DNase II become overwhelmed with undegraded DNA and die.

Cell autonomous nucleases act in two stages. After an initial cleavage of the chromosomes into fragments of roughly 50,000 base pairs, DNA is usually (but not always) cleaved between nucleosomes, producing a characteristic “ladder” of DNA fragments with a periodicity of about 200 base pairs. This ladder is seen when DNA isolated from apoptotic cells is subjected to gel electrophoresis. The responsible nuclease is CAD. CAD is normally present in a complex with ICAD (inhibitor of CAD [Fig. 46-12C]). The complex of CAD and ICAD is also known as DNA fragmentation factor (DFF). ICAD is a chaperone that must be present for CAD to fold into an active conformation as it is being translated on the ribosome. However, ICAD also inhibits the nuclease activity of CAD. This dual function of ICAD guarantees that only inactive CAD can be synthesized in healthy cells. During apoptosis, caspase 3 cleaves ICAD and releases active CAD nuclease.

Bcl-2 Proteins and the Intrinsic Pathway of Apoptotic Cell Death

As was mentioned previously, mitochondria are key players in a pathway to cell death that is triggered by a variety of toxic insults (Fig. 46-16). These mitochondrial events are regulated by the Bcl-2 family of proteins. The following sections describe this important protein family and their regulation of the intrinsic pathway of apoptosis.

Bcl-2 Proteins

Bcl-2 proteins can be grouped into three subfamilies (Fig. 46-13). Bcl-2 protectors protect cells against apoptosis. Bcl-2 killers (e.g., Bax and Bak) are pro-apoptotic proteins that actively kill cells. Bcl-2 regulators promote cell killing by either interfering with the protectors or activating the killers. These proteins primarily regulate the release of death-promoting factors from mitochondria when cells receive signals that activate the intrinsic pathway.

Figure 46-13 Introduction to the Bcl-2 family of proteins.

C. elegans genetics identified a gene, ced-9, that protects cells against apoptosis. In ced-9 mutants, many cells that normally survive into the adulthood of the organism die during development. This kills the worm. Human Bcl-2 is functionally and structurally homologous to C. elegans Ced-9 and can substitute for it in living worms. This ability of a human gene to protect nematode cells is just one of many examples showing that the fundamental mechanisms that are involved in apoptotic cell death have been conserved over great evolutionary distances.

Bcl-2 family members are defined by the presence of one to four short blocks of conserved protein sequence called BH (Bcl-2 homology) domains. Anti-apoptotic Bcl-2 protectors typically have four of the domains. Pro-apoptotic Bcl-2 killers typically have three of these domains, while the Bcl-2 pro-apoptotic regulators have only the BH3 domain. The BH3 domain is a short segment of helix that fits into a groove on the surface of both Bcl-2 protectors and killers, forming a complex that regulates their activity. It is now believed that the Bcl-2 protectors regulate the behavior of Bcl-2 killers by a similar interaction. For example, Bcl-2 protein forms a complex with a pro-apoptotic Bcl-2 killer called Bax, thereby interfering with the ability of Bax to kill cells.

Genetic experiments in mice revealed several different functions for Bcl-2 family members. Mice that are born without Bcl-2 have deficiencies of the immune system that are best understood if one role of this protein in vivo is to render lymphocytes resistant to pro-apoptotic signals during immune system maturation. In contrast, loss of another pro-life family member, Bcl-x_L, is lethal. Embryos die, apparently as a result of widespread death of neurons in the central and peripheral nervous systems and hematopoietic cells of the liver. In contrast, loss of the killers Bax plus Bak makes cells highly resistant to apoptosis by a wide variety of intrinsic pathway stimuli.

Bcl-2 Family Members and Cancer

A gene that prevents cells from dying poses a potential danger in multicellular organisms, in which rates of cell proliferation and death must be balanced carefully. In fact, the name Bcl-2 comes from the discovery that this gene is the culprit responsible for certain types of B-cell lymphoma. These particular lymphomas arise when a chromosome translocation, involving chromosomes 14 and 18, moves the Bcl-2 gene into the immunoglobulin heavy-chain gene cluster, a site of very active gene transcription in B lymphocytes. Elevated transcription of Bcl-2 is thought to be directly responsible for the cancerous phenotype in these patients, making Bcl-2 a cancer-promoting oncogene (see Chapter 41).

Unlike many other oncogenes, Bcl-2 overexpression does not cause cell proliferation. Instead, it disrupts the balance of regulation between life and death of the affected cells. Cells that overexpress Bcl-2 protein actually grow, if anything, more slowly than do their normal counterparts. However, Bcl-2 overexpressers are highly resistant to many stimuli that normally promote cell death. The net result is an accumulation of B cells: a lymphoma. Figure 46-14 shows an example of Bcl-2 conferring resistance to death when the cell cycle is perturbed by expression of an oncogene.

The Intrinsic Pathway of Apoptotic Death

In addition to their role in energy production, mitochondria have an essential role as sensors of the health of the cell. If cells sense insults from which they cannot recover, mitochondria trigger the intrinsic pathway of cell death (Fig. 46-16). This pathway is regulated by Bcl-2 family members. The regulation seems straightforward in C. elegans, in which the protector CED-9 (Bcl-2–like) binds to the CED-4 scaffolding protein (Apaf-1-like) and interferes with its activation of the CED-3 caspase. Apoptosis is induced when the regulator BH3-only protein EGL-1 binds to CED-9 and blocks it from inactivating CED-4.

In mammals, the situation is more complex, partly because Bcl-2 family members are more numerous and partly because they do not interact in such a straightforward fashion. In mammals, two of the killer proteins, Bax and Bak, are essential for activation of the intrinsic pathway. In healthy cells, Bak is loosely associated with the mitochondrial outer membrane, and Bax is in the cytoplasm (Fig. 46-16). On receipt of a pro-apoptotic stimulus, Bax and Bak insert deeply into the mitochondrial outer membrane, form oligomers, and somehow (not yet known but possibly involving the formation of membrane pores) cause the release of pro-apoptotic factors from the mitochondrial intermembrane space. Binding of anti-apoptotic Bcl-2 family members to Bax/Bak somehow prevents the release of pro-apoptotic factors from mitochondria. Various BH3-only family members either facilitate Bax/Bak oligomerization or bind and neutralize anti-apoptotic Bcl-2 family members.

The pro-apoptotic factors that are released from the mitochondrial intermembrane space by Bax and Bak include the electron transport protein cytochrome c (see Fig. 19-5), Smac, and endonuclease G (Fig. 46-15). These mitochondrial proteins actively promote apoptotic cell death. In the cytoplasm, cytochrome c binds to the scaffolding protein Apaf-1, a mammalian homologue of C. elegans CED-4 protein, causing it to form a seven-spoked wheel-like structure called the apoptosome (Fig. 46-16). Apaf-1 in the apoptosome binds caspase 9 through an N-terminal caspase recruitment domain.

Figure 46-15 mitochondria as integrators of a cell’s life and death decisions. A, In healthy cells, a number of factors that promote apoptosis are stored in the intermembrane space of the mitochondria. B, In cells undergoing apoptosis, caspases trigger the pro-apoptotic Bcl-2 family members to induce the release of these death-promoting factors; this initiates an amplifying cycle that ultimately leads to cell death. AIF, apoptosis-inducing factor.

The C-terminal portion of Apaf-1 acts as an autoinhibitor of Apaf-1 function. Binding of cytochrome c and deoxyadenosine triphosphate induces a conformational change that turns off the autoinhibition, thereby permitting binding and autoactivation of seven procaspase 9 monomers. Binding to the apoptosome elevates the catalytic activity of the procaspase 9 zymogen approximately 2000-fold without the need for its cleavage. Thus, the active form of caspase 9 is an oligomeric complex of the zymogen with the apoptosome. Activated cas-pase 9 then cleaves multiple procaspase 3 zymogens, amplifying the cell death cascade. This cascade can be further amplified in at least two ways. First, caspase 3 cleaves other effector caspases, directly amplifying the cascade. In addition, active caspases cleave the BH3-only protein Bid, which then activates more Bax and Bak in a feedback loop, thereby promoting the release of more cytochrome c and Smac, and enhancing caspase 9 activation.

It was extremely surprising to find that an essential metabolic protein such as cytochrome c has a second function that is essential for death. Among the studies supporting the Jekyll-and-Hyde-like nature of this protein in life and death was the engineering of mice whose cytochrome c can function in electron trans-port but cannot bind Apaf-1. These mice die as a re-sult of brain abnormalities caused by insufficient cell death.

The Extrinsic Pathway of Apoptotic Death

Cells express at least six different cell surface molecules, collectively termed death receptors, that can trigger apoptotic death. These receptors generally bind protein ligands that are expressed on the surface of other cells. This binding activates the receptor, turning on a pathway that leads to apoptotic death.

One well-characterized death receptor is called Fas (also known as Apo1 or CD95), a member of the tumor necrosis factor receptor family (see Fig. 24-10). Fas is a type I membrane protein whose extracellular domain consists of three cysteine-rich domains (see Fig. 24-11, which shows the atomic structure of the related trimeric tumor necrosis factor receptor with bound ligand). The cytoplasmic domain of Fas contains a death domain of about 80 residues, which is shared by all of the death receptors (Box 46-2).

The Fas ligand is a trimeric 40-kD intrinsic membrane protein found on the surface of cells. Cytotoxic T lymphocytes use Fas ligand to rid the body of virally infected cells. When a cytotoxic T lymphocyte contacts a target cell, the Fas ligand on the lymphocyte surface binds to Fas on the target cell and initiates the extrinsic pathway of apoptotic death (Fig. 46-17). Ligand binding activates signaling from the intracellular death domain of Fas, possibly by stabilizing Fas trimers or by altering their conformation. Activated Fas binds an adapter protein called FADD (Fas-associated protein with a death domain). The Fas-FADD complex binds procaspase 8 through interactions involving another type of motif called the death effector domain, which is present on both FADD and the prodomain of procaspase 8. On this molecular scaffold, procaspase 8 monomers dimerize and acquire catalytic activity. These dimers can cleave neighboring dimers, creating and releasing heterotetrameric active caspase 8, which initiates the caspase cascade by activating downstream effector caspases.

This pathway poses considerable risk for the cell. Fas is constitutively present in the cell membrane and appears to have the ability to form at least transient trimers in the absence of binding by its ligand. How do cells avoid the accidental activation of apoptosis caused by chance binding of procaspase 8 zymogens to naturally occurring transient Fas trimers?

Cells express a protein called FLIP (FLICE-like inhibitory protein; FLICE is another name for caspase 8) that looks very much like a catalytically dead version of procaspase 8. When expressed at high levels, FLIP competes with procaspase 8 monomer for binding to FADD, thereby inhibiting the autoactivation of the caspase. This role of FLIP may be to dampen the Fas response locally to ensure that the cascade does not get activated by mistake. When expressed at low levels, FLIP can have the opposite function, facilitating caspase 8 activation by forming a heterodimer with procaspase 8 monomers and causing a conformational change that activates the protease.

Role of the Fas Death Receptor in Normal and Diseased Cells

Mouse mutants provide clear evidence for an important role of Fas in regulation of the immune system. Mice with mutated Fas (the lpr mutation) or Fas ligand (the gld mutation) accumulate excessive lymphocytes. In the appropriate genetic background, these mice tend to develop autoimmune disorders that, in some cases, resemble the human disease systemic lupus erythematosus. Evidence that Fas is involved in human systemic lupus erythematosus is still scant.

Fas is important in regulating the life span of activated tissue T and B lymphocytes. Normally, T cells die within a few days of their activation during an immune response. Activation initiates the expression of Fas ligand on the T cells themselves. This interacts by an unknown mechanism with Fas already on the cell surface, causing the cell to commit apoptotic suicide. T-cell activation also downregulates the expression of FLIP, thus permitting the more efficient activation of procaspase 8 by trimerized Fas. A similar mechanism (export of Fas and Fas ligand to the surface of the same cell) is responsible for some examples of p53-induced cell death and some instances of cell death following exposure to chemotherapeutic agents.

Expression of Fas ligand can protect tissues against immune system cells that express Fas. Some tissues, like the lens of the eye and the testis, avoid immune and inflammatory responses by expressing Fas ligand. Immune effector cells that enter these tissues encounter Fas ligand and die by apoptosis. These tissues are known as immune-privileged. Not surprisingly, certain tumor cells subvert this strategy as protection against the immune system. Melanoma cells expressing Fas ligand establish tumors particularly efficiently. Some tumor cells, especially colon and lung cancer cells, also defend themselves against immune surveillance with so-called decoy receptors. A secreted Fas decoy receptor blocks Fas ligand on cytotoxic cells. Other decoy receptors remain membrane bound but do not signal cell death when they bind ligand because their intracellular domains lack functional death domains.

Linking Apoptosis to the Cell Cycle by p53

No obligate link exists between particular cell-cycle phases and apoptosis. Noncycling G₀ cells can undergo apoptosis, and cycling cells appear able to do so from any cell-cycle phase. However, one link between apoptosis and the cell-cycle machinery has now been firmly established. This involves the p53 tumor suppresser and DNA damage.

The p53 transcription factor is one of the downstream effectors of the DNA damage response pathway (see Fig. 40-4). When cells sense DNA damage induced by agents such as ionizing radiation, levels of p53 rise dramatically (Fig. 46-18). When stabilized and activated by phosphorylation (see Fig. 41-13) p53 upregulates the expression of a number of genes, including the Cdk inhibitor p21, which blocks the entry into the S and M phases. p53 also can trigger an apoptotic response in instances in which the DNA damage is too severe to repair. This tumor suppressor protein is very important in the body’s defense against cancer. Mutations in the p53 gene/protein are found in about 50% of all human cancers.

Figure 46-18 a–b, survival of thymocytes from three strains of mice after exposure to glucocorticoids or irradiation. Cell death was due to apoptosis. The strains were as follows: wild-type mice (yellow), heterozygous mice having one good copy of the p53 gene and one defective copy (red), and mice lacking a functional copy of the p53 gene (blue). Thymocytes that lack p53 are resistant to radiation-induced apoptosis but show normal induction of apoptosis following exposure to glucocorticoid hormone. C–D, INDUCTION OF P53 ACCUMULATION FOLLOWING RADIATION OF THE SMALL INTESTINE. Black arrows indicate cells with increased levels of p53. Red arrows indicate apoptotic cells.

(A–B, From Lowe SW, Schmitt EM, Smith SW, et al: p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 362:847–849, 1993. C–D, Courtesy of John Hickman, Molecular and Cellular Pharmacology Group, University of Manchester, England.)

A direct connection between p53 and apoptosis was revealed by overexpression of the cloned p53 gene in different cell types. In most cells, overexpression of p53 arrests the cell cycle at the G₁/S boundary. How-ever, ectopic expression of cloned p53 in certain cancer-derived cell lines causes the cells to undergo apoptosis.

The role of p53 in apoptosis was confirmed in transgenic mice lacking a functional p53 gene (p53 knockout mice). These mice develop normally but are extremely prone to cancer at a very young age. Thus, although mice do not require p53 for programmed cell death during embryogenesis, p53 is critical for apoptosis of certain cells. Thymocytes isolated from p53 knockout mice are extremely resistant to the induction of apoptosis by ionizing radiation and other agents that cause DNA breaks (Fig. 46-18B). However, p53 is not involved in all types of apoptosis. For example, even thymocytes isolated from p53 knockout mice show normal induction of apoptosis following exposure to glucocorticoid hormone (Fig. 46-18).

p53 promotes apoptosis by functioning as a transcriptional activator. It controls, among others, the well-studied death-promoting genes Bax, Fas (CD95/APO-1), and APAF-1. However it now appears that the key target gene is PUMA (p53 modulated upregulator of apoptosis), a BH3-only protein that promotes apoptotic cell death by activating Bax and Bak. PUMA knockout mice show defects in cell death pathways that are essentially identical to those seen in p53 knockout mice and not seen in mice that lack Bax, Fas, or Apaf-1.

Importance of Apoptosis in Human Disease

Studies of apoptosis now account for a substantial fraction of cell biology research. Why has this field so caught the scientific eye? The most likely answer is that apoptosis is a point of intersection between cell signaling pathways, cell structure, the cell cycle, and, of course, human disease. This chapter has mentioned the roles that aberrations in apoptosis play in the etiology of autoimmunity, AIDS, and cancer. Apoptosis is also emerging as a key factor in neurodegenerative diseases, such as Huntington’s disease and Alzheimer’s disease, as well as in myocardial infarction and stroke (Fig. 46-19). At a practical level, the realization that many successful chemotherapeutic agents act by inducing cancer cells to undergo apoptosis has motivated searches for newer and better drugs that elicit this response. One promising approach is to combine reagents that either directly promote apoptosis or weaken cellular resistance to apoptosis with cancer chemotherapy. The goal is to find combinations that increase cell killing synergistically. It is hoped that this strategy will improve outcomes in cases where tumors are resistant to current chemotherapy. Conversely, the realization that a large fraction of the cell deaths in stroke are attributable to a wave of apoptosis that radiates outward from the original focus of ischemic death has led to the hunt for molecules that will prevent apoptosis during the critical period following the stroke or infarct. With such important practical problems to be solved, apoptosis will continue to occupy a prominent position in cell biology research over the coming years.