2.2. Common mechanisms of apoptosis
Apoptosis-regulating genes have been found in every metazoan organism. These genes have proven to been exceptionally well conserved, throughout evolution (reviewed in Liu & Hengartner 1999). While it has been thought that apoptosis is a mechanism that would be confined to multicellular animals, surprisingly even some bacteria show hints of a similar procedure to control homeostasis (Lewis 2000) (Fig. 2).
While a large number of important observations have been made in this field, perhaps the greatest impact on the study of apoptosis has been the identification of the different genes that encode the proteins which are responsible for initiation, processing and execution of cell death. Much of what we know about the genetic basis of apoptosis in mammalian cells has been derived from studies using the nematode Caenorhabditis elegans, which has proven to be a Rosetta stone in deciphering the cell death programme. In the C. elegans hermaphrodite, 131 of the 1090 somatic cells that are generated during development undergo apoptosis. These deaths are dependent on the actions of at least three key genes: ced-3, ced-4 and ced-9 (reviewed in Hengartner 1996). The cloning of these three genes indicated the existence of homologues in other species, including humans (Fig. 2).
2.2.1. Regulation of apoptosis
A wide variety of stimuli are capable of inducing apoptosis. Some are universal and can produce apoptosis in almost any cell, while most apoptosis-inducing factors show some selection of their targets (Rich et al. 2000). As a result of the profusion of apoptosis-inducing mechanisms and the fact that virtually all eukaryotic cells can be induced to undergo apoptosis, such a massacre has to be under tight lock and key. Apoptosis-inducing signals are carefully processed and evaluated against anti-apoptotic factors in the target cells. If the pro-apoptotic elements beat their counterparts, a dedicated death program is then activated and the cell will undergo apoptosis (Fig. 3).
As apoptosis plays a part in the regulation of tissue homeostasis and a variety of other physiological functions, it is controlled by physiological mechanisms. Death by neglect is a classical example of this. Just as cells require extracellular growth factors and mitogens to grow and divide, they also need survival factors to escape cell death. Failure to supply adequate levels of survival factors leads to activation of apoptosis (reviewed in Raff 1992, Conlon & Raff 1999). Furthermore, cells can also be driven into apoptosis by conflicting signals that scramble the normal status of a cell (Wyllie 1997b.)
In normal physiological cell turnover the apoptotic stimuli can be represented by cytokines, or death factors, such as Fas ligand (FasL) (reviewed in Nagata 1994) or TNF-_ which is present in the ovary and endometrium (Tabibzadeh et al. 1999). Both of these proteins belong to the same transmembrane protein subfamily that can be generated into soluble forms by metalloproteinase-mediated cleavage (reviewed in Krammer 1999, Schmitz et al. 2000).
In addition to physiological control mechanisms of apoptosis, a variety of pathological insults can trigger apoptosis. Factors that are capable of causing DNA damage, such as radiation, cytostatic drugs or genotoxic compounds, can also induce apoptosis (reviewed in Wahl & Carr 2001, Bratton & Cohen 2001). DNA breaks are detected by transcription factor p53, which is subsequently activated. Depending on the damage and cell type, p53 will either cause an arrest in the cell cycle or activate the apoptotic self-destruction sequence (reviewed in Balint & Vousden 2001). The antimicroboid drug staurosporin functions as a general kinase inhibitor and activates apoptosis via an unknown pathway (Ojeda et al. 1995). Some chemical agents, such as hydrogen peroxide, can also trigger the apoptotic pathway in several cell types (Madesh & Hajnoczky 2001, Gorman et al. 1997).
2.2.1.1. Death receptors
Death factors induce apoptosis through activation of specific death receptors that belong to the growing superfamily of TNF/NGF receptors (reviewed in Schmitz et al. 2000). These death receptors are characterized by a unique intracellular death domain (DD), which is crucial for death ligand-induced apoptosis (Huang et al. 1996). The binding of death ligand to its receptor leads to trimerization of the receptors. In concordance with this, functional soluble forms of FasL and TNF-_ exist as trimers (Nagata 1994).
The trimerization of death receptors and subsequent association of three death domains lead to the formation of a death-inducing signalling complex (DISC) which leads into activation of pro-caspase-8 (Fig. 4). Activation of a death receptor can also lead to generation of additional death signals such as ceramide (reviewed in Kronke 1999).
The message to induce apoptosis can be modulated directly at the death receptor level. For example, glycosylation of Fas receptor has been shown to regulate the FasL-induced apoptosis pathway (Peter et al. 1997, Keppler et al. 1999). In addition, death receptor signalling can obviously be regulated on a transcriptional level. It has been observed that activation of tumour suppressor p53 upregulates Fas expression (Muller et al. 1998). Furthermore, a class of proteins named FLIPs (Thome et al. 1997) can block apoptosis by directly interacting with the death receptor pathway (Hu et al. 1997, Bertin et al. 1997).
Mitogens and growth factors can also inhibit death receptor induced apoptosis. Mitogen-activated protein kinase (MAPK) pathways can be activated by mitogens, growth factors and environmental stress (Seger & Krebs 1995, Robinson & Cob 1997, Lewis et al. 1998). The MAPK family consists of at least three different signalling cascades: the ERK1/2, JNK and p38 kinase pathways. ERK1/2 has been shown to directly inhibit death receptor-induced apoptosis by preventing caspase activation via unknown mechanism (Holmström et al. 1998, 1999, Tran et al. 2001). Furthermore, activation of the MAPK cascade can inactivate the pro-apoptotic Bcl-2 family member BAD by phosphorylating BAD (Bonni et al. 1999, Scheid et al. 1999). Another proliferation stimulating signalling cascade, PI-3K pathway, has also been shown to inhibit apoptosis by phosphorylating BAD (Craddock et al. 1999, Wolf et al. 2001).
2.2.1.2. NF-kB
NF-κ B is a transcription factor with a wide variety of functions ranging from inflammatory reaction and development to cellular survival and oncogenesis (Chen et al. 2001). In their inactive forms NF-κ B proteins are bound to their inhibitor Iκ B and sequestered into the cytoplasm (Magnani et al. 2000). A response to different extracellular signals leads to phosphorylation of Iκ B, breakdown of Iκ B/NF-κ B complexes and subsequent translocation of NF-κ B dimer into the nucleus where it activates transcription (Fig. 4).
There is growing evidence that the NF-κ B pathway is involved in regulation of apoptosis. NF-κ B targets a variety of apoptosis regulating genes including those for TP53, RB1, TNF-_, TRAF-1 and TRAF-2 (Webster & Perkins 1999, reviewed in Pahl 1999, Foo & Nolan 1999). Furthermore, NF-κ B may compete with p53 transcriptional activity, providing a second potential mechanism for the NF-κ B to regulate the cell death program (Ravi et al. 1998, Wadgaonkar et al. 1999). Thirdly, since NF-κ B can be activated by TNFR pathway, it has been suspected that NF-κ B might be particularly involved in interfering with TNF-_-induced apoptosis.
2.2.1.3. Tumour suppressor p53
One of the main envisaged missions of apoptosis is deletion of potentially harmful cells after DNA damage. Tumour suppressor p53 has a key role in DNA damage recognition, DNA repair, cell cycle regulation and particularly in triggering apoptosis after genetic injury (Robles & Harris 2001). The observation that p53 is at the crossroads of multiple pathways of fundamental importance in the development of cancer partially explain why its gene is the most commonly mutated one in human malignancies (reviewed in Sigal & Rotter 2000). Furthermore, the gene encoding p53, TP53, is highly vulnerable to even a single base change in the coding sequence and loss of a single allele of TP53 can result in a reduction of p53 function (Sigal & Rotter 2000).
Although transcription of TP53 is regulated by a number of genes including NF-kB (Webster& Perkins 1999), the main mechanisms that govern p53 action are exerted at the protein level (reviewed in Woods & Vousden 2001, Pluquet & Hainaut 2001). Mdm2 is a protein that has a central role in suppressing p53 action by binding and targeting p53 for degradation (Kussie et al. 1996, Kubbutat et al. 1997). The action of p53 induces Mdm2 expression, creating a negative feedback loop (Wu et al. 1993). Hyperproliferative signals and activation of oncogenes induce stablization of p53 through p14ARF, which prevents Mdm2-mediated degradation of p53 (Zhang et al. 1998, Stott et al. 1998, Honda & Yasuda 1999). Other signals that can lead to stabilization and activation of p53 via a variety of pathways include cytokines (Eizenberg et al. 1995), cytokine deprivation (Canman et al. 1995), hypoxia and heat shock (Graeber et al. 1994), and most importantly, DNA damage (Robles & Harris 2001). Activation of p53 induces expression of p21WAF-1, which inhibits the function of cyclin-dependent kinases (Cdk) and consequently induces cell-cycle arrest, providing more time for the cell to repair the possible genetic damage before mitosis (Pluquet & Hainaut 2001). P53 also induces transcription of a number of pro-apoptotic genes, including bax and fas (Miyashita & Reed 1995, Owen-Schaub et al. 1995).
While the control of p53 stability is critical in the regulation of its function, its transcriptional activity is regulated by post-translational modifications of the p53 protein (reviewed in Woods & Vousden 2001, Pluquet & Hainaut 2001). The different patterns of these modifications may determine p53 target genes, and explain how p53 selects between different cellular responses, such as cell-cycle arrest and apoptosis.
2.2.1.4. GATA transcription factors
GATA transcription factors form a family of zinc finger proteins that participate in the regulation of a large number of genes involved in differentiation and proliferation in a variety of tissues (Molkentin 2000). GATA-binding proteins act according to their name by binding to a consensus GATA motif, (A/T)GATA(A/G), in the promoter and enhancer regions of target genes (reviewed in Orkin 1992, Yang & Evans 1995). GATA-1, the founding member of the family, is a fundamental regulator of gene expression in haematopoietic cell lineages (reviewed in Orkin 1992), where it has been reported to protect these cells from apoptosis (Blobel & Orkin 1996, De Maria et al. 1999).
GATA-4 and GATA-6 transcription factors are expressed in mouse (Heikinheimo et al. 1997, Viger et al. 1998, Ketola et al. 1999) and human ovary and testis (Laitinen et al. 2000, Ketola et al. 2000). In addition, GATA-2 has also been recently associated with ovarian development (Siggers et al. 2002). In mouse ovary, GATA-4 and GATA-6 have overlapping expression patterns. GATA-4 is associated with follicular development and its downregulation precedes ovulation and follicular atresia, i.e. granulosa cell apoptosis. In contrast, GATA-6 expression is unaffected by ovulation or atresia and the corpus luteum (CL) has abundant GATA-6 expression (Heikinheimo et al. 1997). GATA-4 expression is also seen in sex-cord-derived ovarian tumours, indicating a possible role in suppressing apoptosis (Laitinen et al. 2000). Furthermore, pituitary gonadotrophins, which are known to suppress ovarian apoptosis, induce GATA-4 expression in gonadal tumour cell lines (Heikinheimo et al. 1997, Ketola et al. 2000). These observations indicate a possible role for GATA-4 in regulation of granulosa cell apoptosis.
2.2.1.5. The Bcl-2 family
The mitochondrion is not just the cell’s powerhouse; it is also intimately involved in the delicate network of apoptosis-regulating pathways. In this web, the mitochondrion is positioned in the middle, where enhancing or silencing of apoptotic signals take place. The key regulators that operate this cellular rheostat are members of Bcl-2 family.
Bcl-2 was discovered in B-cell neoplasms where chromosomal translocation juxtaposed the location of the bcl-2 gene locus at chromosome segment 18q21 with the Ig heavy chain locus at 14q32, resulting in overexpression of Bcl-2 (Tsujimoto et al. 1985, Bakhshi et al. 1985, Cleary & Sklar 1985). It was soon realized that bcl-2 is a homologue of the ced-9 gene, which has a key role in regulation of apoptosis in the nematode C. elegans (Fig. 2). While C. elegans has only one ced-9-type gene, in humans the situation proved to be different. Reports of the first two bcl-2 homologues bax and bcl-X were published at the same time. Although Bax was found to be highly homologous to Bcl-2, it proved to have an apoptosis-promoting action (Oltvai et al. 1993). Bcl-X was found to be present in two splice variants that repress (Bcl-XL) or augment (Bcl-XS) cell death (Boise et al. 1993). These observations made it evident that in humans a whole family of anti- and pro-apoptotic proteins contributes to regulation of apoptosis. Today the Bcl-2 family consists of at least 24 members (reviewed in Adams & Cory 1998, Chao & Korsmeyer 1998, Hsu & Hsueh 2000) (Fig. 5).
2.2.1.6. The mitochondrial forum
The mitochondrion seems to be the focus of the actions of the Bcl-2 family. It is also the source of reactive oxygen species (ROS) that were initially suspected to be linked to Bcl-2 and apoptotic cell death (Kane et al. 1993). Although ROS are hazardous for the cell, the subsequent reports of apoptosis in the absence of oxygen suggested that they are not essential for apoptosis (Shimizu et al. 1995, Jacobson & Raff 1995). The fog surrounding mitochondria and apoptosis started clearing when the human Ced-4 homologue, apoptotic protease-activating factor-1 (Apaf-1), was identified (Zou et al. 1997). Surprisingly a factor initially termed Apaf-2 was recognized as cytochrome c, which is also intimately involved in the mitochondrial respiratory chain (Liu et al. 1996). The participation of cytochrome c in apoptosis was supported by the results of experiments where injected cytochrome c induced caspase activation and apoptosis in various cell types (Zhivotovsky et al. 1998). Although it has been observed that during apoptosis many different mitochondrial proteins that might be potentially involved with apoptosis are released into the cytoplasm, cytochrome c remains as the most probable key inducer of apoptosis.
Three plausible models for the mechanism of how Bcl-2 family members regulate cytochrome c release have been presented. First, Bcl-2 proteins have been shown to have pore-forming capabilities (Muchmore et al. 1996) (Fig. 6). Following a conformational change, they could form channels or even holes in the outer mitochondrial membrane (Hengartner 2000). However, it is still unclear whether these channels would ever be big enough for proteins to pass through. The second model suggests that Bcl-2 family members may interact with other proteins on the mitochondrial membrane to form large pore channels. A particular candidate for a such function is the voltage-dependent anion channel (VDAC), as several Bcl-2 proteins can bind to it and regulate its activity (Shimizu et al. 1999). However, the pore size of VDAC is not large enough for cytochrome c to pass through (Shimizu et al. 1999) and this model must assume that VDAC undergoes significant conformational change upon binding to Bcl-2 proteins. The third model proposes that Bcl-2 members induce rupturing of the mitochondrial membrane, which subsequently triggers caspase activation and apoptosis (Hengartner 2000).
Although it has been observed that deleting the TM domain renders both Bcl-2 and Bax unable to complete their anti- or pro-apoptotic function (Tanaka et al. 1993, Zha et al. 1996), some members of the Bcl-2 family, e.g. Bcl-XL/Bcl-2-associated death promoter (BAD) and BID, do not possess a TM and their function does not seem to be dependent on localization at the lipid membranes. This suggests that they exert their function solely by dimerisation with other active members of the Bcl-2 family.
2.2.1.7. The Apaf-1 complex or apoptosome
If the apoptotic-inducing fraction of Bcl-2 family outbalances their anti-apoptotic relatives, an array of molecules is released from the mitochondrial compartment. The principal actor in the event that follows is cytochrome c, which associates with apoptosis protease-inducing factor-1 (Apaf-1) and procaspase-9 (Zou et al. 1997). Together these three factors create a holoenzyme termed apoptosome, which is a key connection between mitochondria and caspase activation (Fig. 7). The apoptosome proceeds to activate caspase-3 and other death effector caspases that are required for the final stages of apoptotic cell death (reviewed in Adrain & Martin 2001).
In addition to cytochrome c, the mitochondrion seems to contain redundant mechanisms for induction of apoptosis. Apoptosis-inducing factor (AIF) is encoded by a single gene located on the X chromosome (Susin et al. 1999). AIF protein is normally contained in the mitochondria, but once dislocated from there to the cytoplasm, it induces apoptosis via an as yet unknown pathway (reviewed by Daugas et al. 2000). However, AIF-induced apoptosis appears to be independent of caspase activation and it cannot be inhibited by Bcl-2 overexpression (Susin et al. 1999, Zamzami & Kroemer 1999).
2.2.2. Execution of apoptosis
Proteolytic enzymes, caspases, are responsible for the actual physical labour that is involved in apoptosis. However, there are still fail-safe mechanisms that are designed to keep caspases in check, and of course a counterforce to regulate this action.
2.2.2.1. Caspases
Most of the morphological changes in apoptotic cells that were observed by Kerr et al. (1972) are the result of cleavage of cytoskeletal proteins and nuclear laminins by a family of cysteine proteases, caspases, which are activated in cell death (Kothakota et al. 1997, Rao et al. 1996, Buendia et al. 1999). Caspases have been conserved through evolution and their homologues can be found in insects and nematodes (Miura et al. 1993). In mammals these proteases form a large family that consists of at least fourteen members (reviewed in Budihardjo et al. 1999). They all show a high degree of specificity, which is important in apoptotic cell death as the process involves cleavage of a particular group of proteins in a coordinated manner, rather than random proteolysis (reviewed in Grutter 2000). In most cases, caspase-mediated protein surgery results in inactivation of the target proteins. However, caspases can also activate proteins by cleaving off an inhibitory domain or inactivating a subunit that regulates enzyme activation.
Caspases are initially present as inactive zymogens, procaspases (reviewed in Earnshaw et al. 1999). It is notable that procaspases themselves can be activated by proteolysis. Consequently, once caspase activation is triggered, the effect can be exponentially multiplied by processing of other procaspases to their active state. Thus the most obvious way to activate a procaspase is to expose it to an activated caspase molecule. This type of activation has been termed the caspase cascade, and it is used extensively for activation of downstream effector caspases-3, -6 and -7. These three short prodomain caspases are considered to be the workhorses of apoptotic cell death and they are usually more abundant than their long prodomain cousins (Budihardjo et al. 1999).
One of the end points in apoptosis is fragmentation of DNA into multiples of approximately 180 bp (Wyllie et al. 1984). Recently, the enzyme responsible for this action has been found, and it is now termed caspase-activated DNase (CAD) (Enari et al. 1998). CAD is found as an inactive complex, which is bound to an inhibitory subunit, inhibitor of CAD (ICAD) (Sakahira et al. 1998). The finding that activation of CAD is dependent on caspase-3-mediated cleavage of ICAD provides the final link between the programmed cell death pathway and internucleosomal DNA cleavage. The active caspase-3 molecule cleaves the inhibitory subunit, which then results in the release of the catalytic enzyme. Subsequently, activated CAD proceeds with its intended mission to cleave the genomic DNA.
2.2.2.2. Inhibitors of apoptosis proteins
When activation of one caspase can easily escalate into recruitment of the caspase cascade and cell death, it is easy to imagine that there must exist pre-emptive measures to prevent accidental cell death due to adventitious caspase activation (reviewed in Deveraux & Reed 1999). A family of proteins named inhibitors-of-apoptosis (IAP), have the ability to bind procaspases and activated caspases, blocking their processing and activity. IAPs were first found in baculoviruses as proteins that suppress apoptosis, so allowing the virus to replicate in infected cells (Crook et al. 1993, Birnbaum et al. 1994). They are characterised by a novel domain of 70 amino acids, termed the baculoviral IAP repeat (BIR) domain, which is essential for suppression of apoptosis (reviewed in Deveraux & Reed 1999).
Interestingly, IAPs also have a controlling factor termed second mitochondria-derived activator of caspases (Smac) or Diablo that can prevent IAPs from binding caspases, allowing them to be activated and perform their part in apoptotic program (Verhagen et al. 2000). Smac/Diablo is normally confined to the mitochondria, but once released, it binds to IAPs and removes this block in the cell death pathway. Furthermore, Smac/Diablo possesses an amino-terminal sequence that is capable of procaspase-3 activation (Chai et al. 2000).
2.2.3. Role of apoptosis in development
During the ontogeny of many organs, cells are over-produced only to be carved away. Classical examples of apoptosis during development are found in the development of the nervous and immune systems. Both rely on a three-phased design, which incorporates proliferation, differentiation and cell death. The central nervous system assembles itself according to a genetic programme, which does not contain a cellular map for placement and connection for every neuron. Once the pieces (neurons) are in place, the set-up is tested. Any orphan cells without adequate connections to neighbouring cells are removed by the mechanism of apoptosis, because their survival depends on the availability of neurotrophic factors secreted by the target cells they innervate (reviewed in Meier et al. 2000). It has been estimated that through this process more than 80% of the ganglion cells in the cat retina and optic nerve die shortly after they are born (reviewed in Barres & Raff 1999, Meier et al. 2000). In the immune system the wastage of cells is equally profound (Debatin 2001).
A similar course of action takes place in many other places in the developing embryo. Apoptosis seems to be an essential component for fusion of epithelial sheets that merge to form the neural tube. If explanted chick embryos are treated with apoptosis inhibitors the epithelial sheets still meet but fail to fuse to form the neural tube (Weil et al. 1997). In the reproductive system apoptosis is responsible for removing the Wolffian duct in females and the Müllerian duct in males (reviewed in Capel 2000). During gonadal development a large number of germ cells is culled from the developing male and female gonads. Germ cells migrate to the human ovary during the fifth and sixth weeks of development and begin to multiply through mitotic divisions. In the testis, this process also continues in postnatal life, renewing the germ cell reserve continuously, decreasing the effect of apoptotic cell death in the male gonads. However, in the ovary mitotic divisions of germ cells cease before birth and apoptotic death of the remaining oocytes has a profound effect on the reproductive lifespan (reviewed in Morita & Tilly 1999).
Based on previous findings, it has been postulated that cell death is the default state during the development of a metazoan organism, and that cells need to compete for adequate survival factors (Raff 1992, Raff et al. 1993). This suggests that cells are as a rule trapped in their specific microenvironments with sufficient levels of survival factors. The observation that cells generally commit suicide when they are separated from their neighbours or basal stroma support this notion (Zhang et al. 1995).
2.2.4. Apoptosis and tissue homeostasis
The maintenance of tissue homeostasis is finely tuned between cell proliferation and cell death i.e. apoptosis. The maintenance of this balance is crucial to any multicellular organism. Too much proliferation leads to hyperplasia and to anatomical and physiological problems that are associated with it. The worst-case scenario is a total loss of homeostatic control and development of cancer (reviewed in Lyons & Clarke 1997). If apoptosis supersedes proliferation, the result is a reduction of the tissue mass. If the process runs rampant, it eventually reaches a point where physiological function is no longer possible (Thompson 1995). Apoptosis has been shown to function as a limiting factor of tumour growth in early stages, when the angiogenesis is limiting the tumour progression (Naik et al. 1996, O"Reilly et al. 1996). Furthermore, a tumour’s resistance to chemotherapeutic agents has often been suggested to be dependent on expression of anti-apoptotic genes, such as members of the Bcl-2 family, or loss of apoptosis-inducing genes, such as TP53 (Minn et al. 1996, Lowe et al. 1993).
Tissues that have constant cellular proliferation, such as haematopoietic cells, epithelium lining the intestinal crypts and male germ cells, also have a high rate of apoptosis (Wyllie 1987, Billig et al. 1995). Similarly, tissues that have a minimal rate of cell proliferation, such as the nervous system, heart, liver and kidney, exhibit only a very little apoptosis (Benedetti et al. 1988). The ovary is an exception to this rule. While a high level of granulosa cell proliferation is matched by a high rate of apoptosis, the oocytes increase their number through mitosis only during the early fetal life, whereafter the oocyte population is only reduced through the mechanism of apoptosis. Eventually, the pool of resting follicles, i.e. oocytes, is depleted and menopause ensues (reviewed in Morita & Tilly 1999).
2.2.5. Apoptosis in disease
A deep look into the molecular pathogenesis of human diseases has revealed an apoptotic component that either contributes to disease progression or is responsible for the process in a growing number of pathological conditions. While apoptosis is an essential building block of physiological events, it seems to have an equally important function in pathological events. Autoimmune disorders are often associated with erroneously activated apoptosis in the target tissue (Greer et al. 2001). Ischaemic injury can trigger apoptosis in vascular tissue or in the surrounding nerve cells in the central nervous system (reviewed in Johnston et al. 2001). In graft versus host disease apoptosis plays a role in the disease process (reviewed in French & Tschopp 2000). In many neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, Charcot-Marie-Tooth type 1A demyelinating neuropathy and spinal muscular atrophy, an apoptotic component that significantly contributes to the pathogenesis of the disease has been found (Mazarakis et al. 1997).
Many pathogens have discovered a way to circumvent their host’s defences by manipulating apoptotic inducing factors. Many viruses encode genes that prevent apoptotic destruction of the host cell (reviewed in White 2001). On the other hand, HIV-infected lymphocytes can activate apoptosis by expressing Fas ligand (reviewed in Kaplan and Sieg 1998). During Ebola virus infection, massive apoptosis is observed in the cells of the vascular endothelium (Baize et al. 2000). Some parasites, such as Trypanoma cruzi can induce cell death in T cells, which consequently inhibits the macrophage-directed killing of parasites (Freire-de-Lima et al. 2000).
The universality of the genetic programme controlling apoptosis has helped to rapidly analyse the regulation of apoptosis in different tissues and diseases. While there are high expectations about possible therapeutics that might be targeted against apoptosis- regulating factors, the ubiquitous nature of the event also sets limitations to the possible measures.
2.2.6. Studying apoptosis
Apoptosis has many distinctive features that give away its presence to a trained eye. However, the apoptotic bodies are ultimately swallowed by neighbouring cells or macrophages (reviewed in Savill & Fadok 2000). Although the exact speed of the process depends on the cell type, in most case the whole episode is over within a few hours (reviewed by Wyllie 1997b, Cummings et al. 1997). Morphological changes of the apoptotic cells, including condensation of chromatin and cytoplasm, fragmentation of the cell and apoptotic body formation, can be detected by using light microscopy (Kerr et al. 1972). While it has been argued that the method can be as sensitive as biochemical methods, it is highly dependent on the observer. Furthermore, careful inspection of histological samples to detect these changes is very time consuming.
Electron microscopy can also be used in detecting apoptosis. In some respects it has been considered as the most convincing method for accurate identification of apoptosis (Kerr et al. 1994). However, for obvious reasons electron microscopy is the least feasible method for analysis of clinical samples.
2.2.6.1. DNA analysis
Breakdown of genomic DNA into multiples of approximately 180 bp is considered to be a hallmark of apoptosis (Wyllie et al. 1984). This cleavage of chromosomes produces a large number of DNA breaks, and subsequently a simultaneous amount of new 3’-OH DNA ends. In normal living cells, only an insignificant number of 3’-ends are present, making this a promising target for detection of apoptosis. The enzyme terminal deoxynucleotidyl transferase (TdT) has the capability to incorporate individual deoxyribonucleotide triphosphates to the 3’-end of double- or single-stranded DNA. This quality can be used to detect 3’-ends with nucleotides that have been labelled with radioactive, fluorescent or digoxigenin labels. Apoptosis can then be measured quantitatively by using gel electrophoresis, where apoptotic DNA is organized into a typical ladder pattern of multiples of 180 bp. In situ labelling of 3’-ends can be used to qualitatively recognise apoptotic cells in immuhistochemical tissue sections.