Cell Death as Cell Fate: Historical Context
Although apoptosis is often described as a "hot topic" or a "new and exploding" area of biological research, the concept of cell death as a normal cell fate was articulated only three years after Schleiden and Schwann introduced the Cell Theory when, in 1874, Vogt described natural cell death as an integral part of toad development (Cotter and Curtin, 2003). Since these early observations, natural cell death has been described "anew" several times. In 1885 Flemming provided the first morphological description of a natural cell death process, which we now label "apoptosis", a term coined by Kerr and colleagues to describe the unique morphology associated with a cell death that differs from necrosis (Kerr et al. 1972). The revolution that has occurred in apoptosis research is a direct result of a better understanding of the genetic program and biochemical mechanisms of apoptosis.
In the 1970s and 1980s, studies revealed that apoptosis not only had specific morphological characteristics but that it was also a tightly regulated process with specific biochemical characteristics. Studies of cell lineage in the nematode, Caenorhabditis elegans, showed that apoptosis was a normal feature of the nematode's invariant developmental program. Of the 1,090 somatic cells of the C. elegans adult hermaphrodite, 131 die during normal development (Hengartner, 1997). By documenting every cell division from the zygote to the adult, researchers discovered that the lineage and the timing of apoptosis for each of these 131 cells were constant, demonstrating that apoptosis was a tightly regulated process, presumably genetically programmed (i.e., programmed cell death). At the biochemical level, Wyllie showed that DNA degradation by a specific endonuclease during apoptosis resulted in a DNA ladder composed of mono- and oligonucleosomal-sized fragments (Wyllie, 1980).
Morphology and Overview of Apoptosis
Morphologically, apoptosis is first characterized by a change in the refractive index of the cell (Hengartner, 1997) followed by cytoplasmic shrinkage and nuclear condensation. The cell membrane begins to show blebs or spikes (protrusions of the cell membrane), depending on cell type (Figure 3.1), and eventually these blebs and spikes separate from the dying cell and form "apoptotic bodies". Apoptotic cells also cease to maintain phospholipid asymmetry in the cell membrane, and phosphotidylserine (PS) appears on the outer leaflet (Williamson, 2000). The mitochondrial outer membrane (MOM) also undergoes changes that include loss of its electrochemical gradient, possibly by the formation of pores in the MOM, and substances such as cytochrome c leak from the MOM into the cytoplasm. Finally, adjacent cells or macrophages phagocytose apoptotic bodies and the dying cell. The apoptotic cell does not provoke an inflammatory response, and only individual cells are affected by apoptosis in vivo.
Figure 3.1. Morphology changes during apoptosis.
The cell membrane begins to show blebs or spikes, depending on cell type. Eventually these separate from the dying cell and form "apoptotic bodies" that are phagocytosed by neighboring cells.
The events of apoptosis stand in contrast to necrosis, which is first marked by a loss of cell membrane integrity. The cytoplasm and mitochondria of the necrotic cell swell, and ultimately the cell and many of its internal organelles lyse. There is no vesicle or apoptotic body formation, and often necrosis affects groups of adjacent cells. The necrotic cell remnants are phagocytosed by macrophages, and inflammatory responses are provoked in vivo.
Apoptosis and necrosis represent two extremes of a continuum of cell death. This continuum includes many variations. "Apoptosis-like programmed cell death" refers to a cell death process that has some of the hallmarks of apoptosis such as chromatin condensation and the appearance of PS on the outer leaflet of the cell membrane but does not necessarily require caspase activity (Leist and Jäättelä, 2001). "Necrosis-like programmed cell death" describes programmed cell death that does not include chromatin condensation and has varying degrees of other apoptotic features. Caspase-1 and caspase-8 have been implicated in some cases of this type of programmed cell death (Leist and Jäättelä, 2001). "Paraptosis" describes a cell death that requires gene expression but morphologically does not resemble either apoptosis or necrosis (Sperandio et al. 2000).
In addition, apoptotic cells cultured in vitro will eventually undergo "secondary necrosis". After extended incubation, apoptotic cells ultimately shut down metabolism, lose membrane integrity and release their cytoplasmic contents into the culture medium (Riss and Moravec, 2004). Therefore, cells that have initiated apoptosis may exhibit some of the morphological phenotypes associated with necrosis. Because programmed cell death takes many forms, both morphologically and biochemically, researchers need to examine multiple biochemical markers at carefully selected time points to determine the mechanism of cell death in their experimental system.
Molecular Players in Apoptosis
Large-scale mutagenesis experiments in the nematode C. elegans identified mutations that disrupted the programmed cell death fates during development, the cell death abnormal (ced) genes (Hedgecock et al. 1983; Ellis and Horvitz, 1986). The gene ced-3 was cloned and found to encode a protease that contained a cysteine residue at the active site and cleaved its substrates after the amino acid aspartate (caspase; Yuan, J. et al. 1993). Genetic analysis showed that ced-3 was absolutely required for apoptosis in C. elegans.
Caspases constitute a large protein family that is highly conserved among multicellular organisms. The family can be divided into two major subfamilies: caspases that are involved primarily in inflammation and have homology to caspase-1 (Interleukin-1β-Converting Enzyme), and those caspases that are related to CED-3 and are primarily involved in apoptosis. Caspases are constitutively expressed in most cell types as inactive zymogens that are proteolytically processed before they gain full activity. The caspase zymogens contain several domains including an N-terminal prodomain, a large subunit and a small subunit. Caspase activation involves cleaving the zymogen at a specific aspartic acid in the region between the large and small subunits and removing the prodomain. The active site is formed by a heterodimer containing one large and one small subunit, and the fully active caspase protein is a tetramer composed of two heterodimers (Figure 3.2). Because caspases exist as zymogens, their activity is thought to be regulated primarily post-translationally. However recent studies indicate that expression of the caspase-9 gene is regulated transcriptionally as well (Csiszar, 2003), and endoplasmic reticulum (ER) stress can induce expression of mouse caspase-12 in transfected cells (Rao et al. 2001).
Figure 3.2. The active caspase enzyme.
Caspase zymogens are cleaved between the large and small subunits, and the prodomains are removed. The active site is formed by a heterodimer that contains one large and one small subunit. Two heterodimers associate to form the fully active tetramer.
Human caspase-8 and caspase-9 are involved in initiating apoptosis through two different signaling mechanisms and are known as "initiator caspases." They can activate the effector caspases, including caspase-3, by proteolytic processing. In turn, caspase-3 cleaves downstream targets and irreversibly commits the cell to the apoptotic fate.
Bcl-2 Family Proteins
The gene ced-9 protects against apoptosis in C. elegans, and genetic loss of function of ced-9 leads to increased apoptosis (Hengartner et al. 1992). The ced-9-encoded protein is homologous with the bcl-2 gene, a proto-oncogene involved in human lymphoma (Tsujimoto and Croce, 1986). The conservation of function of ced-9 and bcl-2 is demonstrated by transgenic experiments in which the human bcl-2 gene rescues the ced-9 loss-of-function phenotype in C. elegans mutants (Hengartner and Horvitz, 1994). The Bcl-2 protein family includes a large number of proteins that share common Bcl-2 homology (BH) domains. Structurally, the Bcl-2 proteins can be divided into three groups. Group I proteins include Bcl-2, and these proteins are anti-apoptotic. Group II and III family members are pro-apoptotic. The group II family members contain all three of the BH domains; the group III family members contain only the BH-3 domain. The pro-apoptotic members of the Bcl-2 family are implicated in permeabilizing the MOM and allowing leakage of mitochondrial proteins such as cytochrome c. The anti-apoptotic members of the protein family, such as Bcl-2, appear to protect cells from apoptosis by sequestering pro-apoptotic proteins or interfering with their activity (Daniel et al. 2003).
Apoptosis can be induced in response to many external stimuli (extrinsic pathway) including activation of cell surface receptors such as Fas, TNFR1 (tumor necrosis factor receptor 1), TRAIL-R1 (TNF-related apoptosis-inducing ligand receptor 1), TRAIL-R2, p75-NGFR (p75-nerve growth factor receptor) and others (Wajant et al. 2003). These "death receptors" have two distinct signaling motifs: death domains (DD) and death effector domains (DED) that allow them to interact with other proteins involved in the apoptosis cascade. Typically the extrinsic pathway involves activating the initiator caspase, caspase-8, which in turn either activates caspase-3 or cleaves the Bcl-2 family member, Bid, leading to the formation of the apoptosome and activation of caspase-9 (Hengartner, 2000). The alternative mitochondrial pathway (intrinsic pathway) can be activated by events such as DNA damage (Rich et al. 1999). The mitochondrial pathway involves members of the Bcl-2 family that regulate cytochrome c release from the mitochondria.
Other studies have suggested that a third pathway for activating apoptosis may involve the endoplasmic reticulum. In mice, caspase-12 has been implicated in an ER stress pathway that induces apoptosis (Nakagawa et al. 2000). Caspase-12 in the mouse localizes to the ER and is cleaved in response to ER stress such as the accumulation of unfolded proteins in the ER (Nakagawa et al. 2000). Mouse caspase-12 activation appears to be mediated by calpain, and Ca2+ homeostasis may be an important indicator of cell health (Rao et al. 2001). Caspase-12-deficient mice show less sensitivity to amyloid beta (Aβ)-induced cell death (Nakagawa, 2000), suggesting that the ER may be involved in Aβ-induced cell death pathways. The amyloid plaques characteristic of Alzheimer’s disease contain Aβ fragments. These Aβ fragments are neurotoxic and are implicated in many neurodegenerative diseases (Yuan and Yankner, 2000). Enzymatic activity has not been demonstrated for human caspase-12, but neuroblastoma cells expressing mouse caspase-12 are more sensitive to ER stress (Hitomi et al. 2003). Screens for human genes closely related to caspase-12 have identified human caspase-4 as a potential candidate for activating apoptosis through an ER stress pathway (Hitomi et al. 2004). Caspase-4 is cleaved when cells are treated with ER stress-inducing agents, and caspase-4 localizes to the ER and mitochondria in SK-N-SH neuroblastoma and HeLa cells (Hitomi et al. 2004). Treating cells with caspase-4 siRNA increases their resistance to ER stress-induced apoptosis and also increases the resistance of SK-N-SH cells to Aβ-induced cell death (Hitomi et al. 2004).
Fas: An Example of the Death Receptor Pathway
Extrinsic signaling at the cell surface can be initiated by aggregation of Fas receptors when they bind to the multivalent Fas Ligand (FasL). This aggregation brings the cytoplasmic domains of the membrane receptors into close proximity and induces a conformational change that allows the assembly of a signaling complex, the death inducing signaling complex (DISC; Figure 3.3), at the cytoplasmic tail of the receptors. Some studies have suggested that the death receptors may be pre-aggregated in the membrane through interaction of pre-ligand-binding assembly domains (PLAD; Chan et al. 2000; Siegel, et al. 2000). The DISC comprises the receptors and ligand as well as an "adaptor" protein, Fas associated death domain protein (FADD), that binds through its C-terminal DD to the ligand-bound receptor and recruits procaspase-8. Procaspase-8 in turn binds to the DED of FADD via its own N-terminal DED domains. As a consequence of DISC formation at ligand-bound receptors, several molecules of procaspase-8 are brought into close proximity, resulting in high local concentration of procaspase-8. One hypothesis suggests that the low intrinsic activity of procaspase-8 allows the procaspase-8 zymogens to cleave and activate each other (induced proximity activation; Hengartner, 2000). Induced proximity activation has also been proposed for human caspase-2 and nematode CED-3 (Hengartner, 2000). However, other studies have suggested that the activation of caspase-8 requires dimerization (Boatright et al. 2003). Active caspase-8 heterotetramers are released from DISC and are free to cleave and activate the effector caspase, caspase-3. An animated presentation illustrating the death receptor pathway is available. In some cells caspase-8 leads to an amplification loop that involves caspase-8 cleavage of the Bcl-2 protein family member, Bid. When Bid is cleaved it can induce Bax-mediated release of cytochrome c from the mitochondria, further committing the cell to the apoptosis fate.
Figure 3.3. DISC Formation.
Aggregation of activated death receptors brings the cytoplasmic domains into close proximity and induces a conformational change that allows the assembly of the death inducing signaling complex (DISC) at the cytoplasmic tail of the receptors.
The Mitochondrial Pathway (Intrinsic)
The mitochondrial pathway involves members of the Bcl-2 family of proteins and can be activated by the death receptor pathway (Section I.E) or by other stimuli that are independent of death receptors including DNA damage, topoisomerase inhibition or withdrawal of trophic factors (Parone et al. 2003). Many of the Group II and Group III Bcl-2 family members, such as Bax, Bad and Bid, shuttle between the mitochondria and the other parts of the cell. Their activity is regulated by a variety of mechanisms including proteolytic processing, phosphorylation and sequestration by inhibitory proteins.
Pro-apoptotic signals direct the Group II and III Bcl-2 family proteins to the mitochondria where the pro-apoptotic members interact with anti-apoptotic Bcl-2 family members including Bcl-2 and Bcl-XL to determine whether or not apoptosis will be initiated. If the pro-apoptotic proteins "win," cytochrome c and other molecules are released from the MOM. Once cytochrome c is released from the mitochondria, it can interact with Apaf-1 (a mammalian homolog of C. elegans CED-4; Zou et al. 1997), dATP and procaspase-9 in a protein complex called the apoptosome. Caspase-9 is processed and activated when it is part of the apoptosome, where it can cleave and activate caspase-3. An animated presentation illustrating the mitochondrial pathway is available.
Clinical Applications of Apoptosis Research
Many diseases—cancers, autoimmune diseases and neurodegenerative diseases, including Alzheimer's Huntington's, and ALS—demonstrate either a failure of apoptosis to eliminate harmful cells or the inappropriate activation of apoptosis leading to loss of essential cells. The complexity of apoptosis regulation and the large numbers of molecular players in the apoptotic signaling pathways provide ample opportunity for developing therapeutics to modulate the pathway. Potential therapeutic strategies include small molecules that inhibit or activate specific proteins involved in the pathway, antisense oligos directed against specific genes involved in apoptosis, and antibodies that can oligomerize cell membrane receptors to modulate the pathway, among others (Murphy et al. 2003).
One obvious target for modulating apoptosis is the caspase family of proteins. The natural delay in activation of the caspases after injury allows time for treatment, and molecules that target the caspases have shown therapeutic potential in preclinical animal models (Reed, 2002; Nicholson, 2000). In mouse models of ischemic injury, active site inhibitors of caspases have been used and result in decreased apoptosis and increased survival and organ function (Nicholson, 2000; Hayakawa et al. 2003). Caspase inhibitors have also been used to treat sepsis in mouse models. In these models, caspase inhibition decreased lymphocyte apoptosis and improved survival rates. One pharmaceutical company, Vertex, has a caspase inhibitor in preclinical trials for treating sepsis (Murphy et al. 2003).
Molecules called "inhibitors of apoptosis" or IAPs are also potential therapeutic targets. These proteins, which function to suppress apoptosis, are evolutionarily conserved. Some cancers overexpress IAPs, and IAP expression is associated with resistance to apoptosis (Reed, 2002). Survivin is an IAP that has been associated with many human cancers, including lung cancer and malignant melanoma (Nicholson, 2000). Eliminating survivin activity has the potential of rendering cancer cells more sensitive to drugs that initiate apoptosis. IAPs are also being investigated in gene therapy strategies as a way of preventing excessive cell loss after stroke (Reed, 2002).
Both the death receptor and mitochondrial pathways present potential therapeutic targets as well. Normal and cancer cells show different sensitivities to TRAIL-mediated apoptosis, with approximately 80 percent of human cancer cell lines being sensitive to TRAIL-mediated apoptosis (Nicholson, 200). In studies where TRAIL (Apo-2L) was administered with cisplatin or etoposide, cancer cells showed increased apoptosis (Nicholson, 2000). In experiments with SCID mice, recombinant TRAIL was able to slow the growth of tumors after transplantation or decrease the size of established tumors. Recombinant TRAIL also showed lower liver toxicity than CD95 ligand or TNF-α (Nicholson, 2002).
The Bcl-2 family members that play essential roles in the mitochondrial pathway are also being targeted by drug companies. Bcl-2 protein is upregulated in many cancer cells. An antisense Bcl-2 oligo has shown promise in preclinical trials in SCID mice and in Phase III clinical trails (Nicholson, 2000; Reed, 2002). Bad is a pro-apoptotic Bcl-2 family member that is implicated in neuronal apoptosis. It is a substrate of calcineurin/calmodulin-dependent phosphatase, and dephosphorylation of Bad allows Bad to bind and neutralize the anti-apoptotic protein Bcl-XL. Current therapeutics that target this part of the apoptotic pathway include active site inhibitors of calcineurin and compounds like the NMDA receptor antagonist, memantine, that prevent calcium influx. Memantine is in clinical trials for treatment of Alzheimer's disease and multi-infarct dementia (Reed 2002).
Many other regulators and players in the apoptotic signaling pathways are also being targeted for developing therapeutics. There are many signaling cascades in cells that influence the decision of a cell to undergo apoptosis. Modifying these signaling inputs is another way to influence cell fate. MAPK family members, JUN kinases, and AKT kinase pathways all provide ways for potentially modulating inputs into apoptosis pathways of target cells (Reed, 2002; Murphy et al. 2003; Nicholson, 2000).
Much remains to be understood about the precise regulation of natural cell death. Understanding these cell death pathways will provide opportunity to influence and modulate cell death signaling so that inappropriate cell death can be prevented or inappropriately dividing cells can be killed using the cell's own molecular machinery.
Methods and Technologies for Detecting Apoptosis
Apoptosis occurs via a complex signaling cascade that is tightly regulated at multiple points, providing many opportunities to evaluate the activity of the proteins involved. The initiator and effector caspases are particularly good targets for detecting apoptosis in cells. These ubiquitous enzymes exist as inactive zymogens in cells and are cleaved before forming active heterotetramers that drive apoptotic events. Luminescent and fluorescent substrates for specific caspases have allowed the development of homogeneous assays to detect their activity. Additionally, specific antibodies that recognize the active form of the caspases or the products of caspase cleavage can be used to detect apoptosis within cells. Fluorescently conjugated caspase inhibitors can also be used to label active caspases within cells.
In addition to monitoring caspase activity, many reagents exist for monitoring molecules in the mitochondria that are indicators of apoptosis, such as cytochrome c. Some of the biochemical features of apoptosis such as loss of membrane phospholipid asymmetry and DNA fragmentation can also be used to identify apoptosis. Cell viability assays can be combined with apoptosis assays to provide more information about mechanisms of cell death through multiplexing assays on a single sample. The remainder of this chapter will describe technologies, protocols and tools to allow you to detect apoptosis in a variety of experimental systems.