23:  23.8 Cell Death and Its Regulation

Programmed Cell Death Occurs through Apoptosis

The demise of cells by programmed cell death is marked by a well-defined sequence of morphological changes, collectively referred to as apoptosis, a Greek word that means “dropping off” or “falling off” as in leaves from a tree. Dying cells shrink and condense and then fragment, releasing small membrane-bound apoptotic bodies, which generally are phagocytosed by other cells (Figure 23-45). Importantly, the intracellular constituents are not released into the extracellular milieu where they might have deleterious effects on neighboring cells. The highly stereotyped changes accompanying apoptosis suggested to early workers that this type of cell death was under the control of a strict cellular program.

In contrast, cells that die in response to tissue damage exhibit very different morphological changes. Typically, cells that undergo this process, called necrosis, swell and burst, releasing their intracellular contents, which can damage surrounding cells and frequently cause inflammation.

Neutrophins Promote Survival of Neurons

The earliest studies demonstrating the importance of trophic factors in cellular development came from analyses of the developing nervous system. In the early 1900s the number of neurons innervating the periphery was shown to depend upon the size of the target field. For instance, removal of limb buds from the developing chick embryo leads to a reduction in the number of sensory neurons and motoneurons innervating it, while grafting of ectopic limb tissue leads to an increase in the number of neurons in corresponding regions of the spinal cord and sensory ganglia. Indeed, incremental increases in target-field size is accompanied by commensurate incremental increases in the number of neurons innervating it. This relation was found to result from the selective survival of neurons rather than changes in their differentiation or proliferation. The observation that neurons die after reaching their target field in several regions of the nervous system suggested that neurons compete for survival factors produced by the target tissue.

Subsequent to the early observations that peripheral tissue promotes survival of sensory and motoneurons, scientists discovered that transplantation of a mouse sarcoma tumor into a chick led to a marked increase in cell number in sympathetic and spinal ganglion neurons. This finding implicated the tumor as a rich source of the presumed trophic factor. To isolate and purify this factor, known simply as nerve growth factor (NGF), scientists used an in vitro assay in which outgrowth of neurites from sensory ganglia was measured. The later discovery that the submaxillary gland in the mouse also produces large quantities of NGF enabled biochemists to purify it to homogeneity and to sequence it. A homodimer of two 118-residue polypeptides, NGF belongs to a family of structurally and functionally related trophic factors, collectively referred to as neutrophins. Brain-derived growth factor (BDNF) and neurotrophin-3 (NT- 3) also are members of this protein family.

Three Classes of Proteins Function in the Apoptotic Pathway

Key insights into the molecular mechanisms regulating cell death came from genetic studies in C. elegans. Scientists have traced the lineage of all the somatic cells in C. elegans from the fertilized egg to the mature worm simply by following the development of live worms under Nomarski optics. Of the 1090 somatic cells generated during development, 131 cells undergo programmed cell death. Specific mutations have identified a variety of genes whose encoded proteins play an essential role in controlling this process. For instance, in worms carrying mutations in the ced-3 or the ced-4 genes, programmed cell death does not occur, and all 1090 cells survive (Figure 23-48). In contrast, in ced-9 mutant animals, all 1090 cells die. These genetic studies indicate that the CED-3 and CED-4 proteins are required for cell death, that CED-9 suppresses apoptosis, and that the apoptotic pathway can be activated in all cells. Moreover, the finding that cell death does not occur in ced-9/ced-3 double mutants suggests that CED-9 acts upstream of CED-3 to suppress the apoptotic pathway.

That apoptosis involved an evolutionarily conserved pathway was first suggested by the confluence of genetic studies in worms and studies on human cancer cells (Figure 23-49). The first apoptotic gene to be cloned, bcl-2, was isolated as a breakpoint rearrangement in human follicular lymphomas and was shown to act as an oncogene that promoted cell survival rather than cell proliferation. Not only are the Bcl-2 and CED-9 proteins homologous, but a bcl-2 transgene can block the extensive cell death found in ced-9 mutant worms. Thus both proteins act as regulators that suppress the apoptotic pathway. In addition, both proteins contain a single transmembrane domain and are localized to the outer mitochondrial, nuclear, and endoplasmic reticulum membranes.

The effector molecules in the apoptotic pathway are a family of enzymes called caspases, so named because they are cysteine proteases that selectively cleave proteins at sites just C-terminal to aspartate residues. These proteases have specific intracellular targets such as proteins of the nuclear lamina and cytoskeleton. Cleavage of these substrates leads to the demise of a cell. Activation of caspases, discussed below, appears to be a common feature of most, if not all, cell-death programs. The principal effector protease in C. elegans is CED-3. Mammalian cells contain multiple caspases.

Cell-culture studies have yielded important insights into how the various CED proteins in C. elegans and the related mammalian proteins act together to control apoptosis (see Figure 23-49). Expression of C. elegans CED-4 in a human kidney cell line induces rapid apoptosis. This can be blocked by co-expression of CED-9 (or mammalian Bcl-2). CED-9 directly binds to CED-4 and relocalizes it from the cytosol to intracellular membranes. Thus the pro-apoptotic function of CED-4 is directly suppressed by the anti-apoptotic function of CED-9. CED-4 also binds directly to CED-3 (and related mammalian caspases) and promotes activation of its protease activity. Biochemical studies have shown that CED-4 can simultaneously bind both to CED-9 and CED-3.

Pro-Apoptotic Regulators Promote Caspase Activation

Having introduced the major participants in the apoptotic pathway, we now take a closer look at how the effector caspases are activated in mammalian cells. Although CED-9 and Bcl-2 suppress the cell-death pathway, other regulatory proteins act to promote apoptosis. The first pro-apoptotic regulator to be identified, named Bax, was found associated with Bcl-2 in extracts of cells expressing high levels of Bcl-2. Sequence analysis demonstrated that Bax is related in sequence to CED-9 and Bcl-2, but overexpression of Bax induces cell death rather than protecting cells from apoptosis, as CED-9 and Bcl-2 do. Thus this family of homologous regulatory proteins comprises both anti-apoptotic members (e.g., CED-9, Bcl-2) and pro-apoptotic members (e. g., Bax). All members of this family, which we refer to as the Bcl-2 family, are single-pass transmembrane proteins and can participate in oligomeric interactions. Thus the fate of a given cell—survival or death—may reflect the particular spectrum of Bcl-2 family members present in the cell and the intracellular signaling pathways regulating them.

Recent studies suggest that Bcl-2 family members can influence the subcellular distribution of cytochrome c; moreover, biochemical studies have implicated cytochrome c in caspase activation. The current model of caspase activation in mammalian cells, summarized in Figure 23-50a, accounts for the involvement of cytochrome c. In normal healthy cells, cytochrome c is localized between the inner and outer mitochrodrial membrane, but in cells undergoing apoptosis, cytochrome c is released into the cytosol. This release can be blocked by overexpression of Bcl-2; conversely, overexpression of Bax promotes release of cytochrome c into the cytosol and apoptosis. In the cytosol, binding of cytochrome c to the adapter protein Apaf-1 (i.e., mammalian CED-4) promotes activation of a caspase cascade. Bax homodimers, but not Bcl-2 homodimers or Bcl-2/Bax heterodimers, permit influx of ions through the mitochondrial membrane. It remains unclear how this ion influx acts to trigger the release of cytochrome c.

Gene knockout experiments have dramatically confirmed the importance of both pro-apoptotic and anti-apoptotic Bcl-2 family members in neuronal development. Mice lacking the bcl-xl gene, which encodes another anti-apoptotic protein, show massive defects in nervous system development with widespread cell death in the spinal cord, dorsal root ganglion, and brain of developing embryos. In contrast, bax knockouts exhibit a marked increase in neurons in some regions of the nervous system.

Some Trophic Factors Prevent Apoptosis by Inducing Inactivation of a Pro-Apoptotic Regulator

We saw earlier that neutrophins such as nerve growth factor (NGF) protect neurons from cell death. The intracellular signaling mechanisms linking such survival factors to inactivation of the cell-death machinery are not known in detail, but some intriguing clues are available. The finding that trophic factors appear to work largely independent of protein synthesis suggested that the activity of one or more components of the cell-death pathway is altered post-translationally in response to intracellular signals activated by binding of trophic factors to their receptors. Scientists demonstrated that in the absence of trophic factors, the nonphosphorylated form of Bad is associated with Bcl-2/Bcl-xl at the mitochondrial membrane (see Figure 23-50a). Binding of Bad inhibits the anti-apoptotic function of Bcl-2/Bcl-xl, thereby promoting cell death. Phosphorylated Bad, however, cannot bind to Bcl-2/Bcl-xl and is found in the cytosol complexed to the phosphoserine-binding protein 14-3-3 (Chapter 20). Hence, signaling pathways leading to Bad phosphorylation would be particularly attractive candidates for transmitting survival signals.

A number of trophic factors including NGF have been shown to activate PI-3 kinase, which in turn activates a downstream kinase called Akt (see Chapter 20). This kinase phosphorylates Bad at sites known to inhibit its pro-apoptotic activity. Moreover, a constitutively active form of Akt can rescue cultured neutrophin-deprived neurons, which otherwise would undergo apoptosis and die. These findings support the mechanism for the survival action of trophic factors depicted in Figure 23-50b. In other cell types, different trophic factors may promote cell survival through post-translational modification of other components of the cell-death machinery.

SUMMARY

• Cells die by murder or suicide through programmed cell death, often referred to as apoptosis.

• All cells require trophic factors to prevent apoptosis and thus survive.

• The best-characterized trophic factors are the neutrophins, including NGF, BDNF, and NT-3. During development, neurons compete for a limited supply of these trophic factors in their target fields. As a result, many cells undergo programmed cell death, so that the number of surviving neurons matches the target-field size.

• Genetic studies in C. elegans defined an evolutionarily conserved cell-death pathway with three major components (see Figure 23-49). The C. elegans anti-apoptotic protein, CED-9, is structurally and functionally homologous to Bcl-2 in vertebrates.

• The effectors of cell death are cysteine proteases, called caspases. Once activated, these proteases cleave specific intracellular substrates leading to the demise of a cell. Pro-apoptotic proteins promote caspase activation, and anti-apoptotic proteins suppress activation.

• Direct interactions between pro-apoptotic and anti-apoptotic proteins lead to cell death in the absence of trophic factors. Binding of extracellular trophic factors can trigger modulation of these interactions via phosphorylation or other post-translational modifications, resulting in cell survival (see Figure 23-50).