Central nervous system
In the developing organism there is a specialised form of cell death previously called apoptosis but now known as ‘programmed cell death’ (PCD) . During embryonic development of the nervous system a surplus of cells is produced. PCD eliminates those neurons whose axons fail to reach the target. It occurs with the withdrawal of trophic substances, such as nerve growth factor, or with a loss of synaptic contact or afferent input. Cytokines (e.g. TNF-α) and ROS may trigger PCD .
Oxidative stress, glutamate excitotoxicity and calcium influx can induce apoptosis in the mature central nervous system. Excessive production of ROS causes ‘oxidative stress’, damaging lipid membranes, proteins, nucleic acids and extracellular matrix glycosaminoglycans. At low levels of ROS or depletion of antioxidants (e.g. superoxide dismutase, catalase, glutathione peroxidase) apoptosis occurs. At high levels ROS produce more damage and cause necrosis .
Glutamate receptor-mediated neuronal injury is an important cause of ‘excitotoxic‘ neuronal death following ischaemia, trauma, epileptic seizures or neurodegeneration . Glutamate produces either necrosis associated with influx of Na+, Cl− and water leading to cell swelling or delayed neuronal death (DND) which appears to be apoptotic, occurring several hours after exposure and associated with calcium influx via channels linked to glutamate receptors [69–71].
Calcium is an important second messenger, instrumental in inducing apoptosis by stimulating neurotransmitter release, gene induction and the activation of enzymes (proteases, phosphatases, protein kinases, endonucleases, phospholipases and nitric oxide synthase) . Phospholipase A2 produces superoxide anion, and nitric oxide synthase produces nitric oxide, both of which can cause oxidative stress leading to apoptosis.
Increased apoptosis caused by excessive intracellular Ca2+ has been implicated in cerebral ischaemia, traumatic brain injury, epilepsy and neurodegenerative disease. Cerebral ischaemia results in both necrosis and DND. Cells undergoing DND as a result of less severe insult are seen on the periphery of the infarct within the penumbra. There is accumulating evidence that apoptosis plays a role in DND . Gene transfer using herpes simplex viruses containing a Bcl-2 vector may offer neuroprotection against ischaemia . In traumatic brain injury apoptosis occurs in ≈ 10% of dying neurons, peaking 24–48 h after injury . With severe trauma, the proportion of necrotic cell death increases and antiapoptotic agents may minimise cell death following trauma .
Enhanced apoptosis is implicated in several neurodegenerative diseases. In Alzheimer’s disease accumulation of β-amyloid peptide in plaques in the brain and cerebral vasculature causes apoptosis of cortical neurones , possibly via nitric oxide (NO) . Parkinson's disease is associated with the loss of dopaminergic neurones in the substantia nigra, and dopamine is thought to cause apoptosis in exposed neurons . Some forms of familial amyotrophic lateral sclerosis may be caused by a mutant gene coding for the antioxidant superoxide dismutase resulting in motor neuron loss via apoptosis. Future therapy may involve the use of growth factors or inhibitors of macromolecular synthesis to block apoptosis .
Defective apoptosis caused by mutant apoptotic genes may contribute to the development of neural tumours. Mutated p53 is seen in astrocytic tumours [77, 78]. Protein kinase C inhibitors (e.g. hypericin and calphostin) can cause apoptosis in glioma cell lines. Gene transfer techniques have also been used to introduce Bcl-xs into neuroblastoma cells and ICE retrovirus into gliosarcoma cells with induction of apoptosis in these tumours [79, 80].
Recently it has been shown that myocyte death with cardiac disease occurs by both apoptosis and necrosis in response to hypoxia and ischaemia . Apoptosis has been documented in myocardial ischaemia, reperfusion and infarction, heart failure induced by rapid ventricular pacing or coronary microemboli, several models of pressure overload hypertrophy (by passive load on papillary muscle, aortic banding)  and ageing .
Hypoxia, nutritional deficiency and toxins (e.g. chemotherapy) that can cause necrosis may induce apoptosis at lower doses [45, 52]. Infarcted areas have a central area of necrosis with surrounding apoptosis in the peri-infarct border zone . Autopsy studies of myocardial infarction have found apoptotic cells scattered between the necrotic cells centrally and the normal cells around the periphery. Coronary artery ligation induces marked increases in proto-oncogenes (Bcl-2 and Fas) in myocytes. The initial event during the early hours of ischaemia appears to be apoptosis. As myocardial cells are overwhelmed by severe ischaemia, apoptotic genes can no longer be expressed and necrosis supervenes . In in vivo reperfusion studies in rabbit models, apoptosis was seen predominantly in reperfused hearts, whereas necrosis was seen mainly with persistent occlusion . Reperfusion increases free radical production and intracellular calcium, which are potent inducers of apoptosis.
Apoptosis in vascular smooth muscle cells in atherosclerotic plaques in coronary arteries may be involved in plaque destabilisation and rupture . Apoptosis is increased by 43% in primary atherosclerotic lesions and 93% in restenoses [84, 85]. It may be triggered by NO and is associated with the expression of caspase 1 .
Arrhythmias may be associated with deficient apoptosis during cardiac development. Normally after birth there is apoptosis of the small round pacemaker cells in the central and lower part of the AV node. Delay or failure of apoptosis in these cells may lead to life-threatening arrhythmias that may resolve spontaneously. Excessive apoptosis can lead to bradyarrhythmias and sudden death . Patients with long QT syndrome have increased rates of apoptosis in the sinus node .
A common end-stage pathway of various cardiovascular diseases results in deteriorating function and cardiac failure associated with dilated cardiomyopathy [45, 46]. Myocardial hypertrophy may be an initial compensatory response to overloading of adult myocytes. This is associated with c-myc, c-fos, and transforming growth factor, which are also involved in initiating apoptosis. Dilated cardiomyopathy can be induced by chronic infusion of TNF-α, which causes apoptosis . A progressive decline in function in end-stage cardiac failure may be due to apoptosis and this is supported by the finding of increased levels of Bcl-2 in the myocytes of patients with cardiac failure. Although apoptosis itself is irreversible, minimising it with growth factors and cytokines may prevent the progressive decline in left ventricular function .
Arrhythmogenic right ventricular dysplasia is a cardiomyopathy associated with sudden death. It is postulated that repetitive episodes of ischaemia and reperfusion associated with ventricular arrhythmias produce apoptosis, which is found in myocardial biopsies in these patients .
Dysfunction of the apoptotic pathway causes autoimmune disease, immunodeficiencies and lymphoid malignancies. During development large numbers of precursor cells from the bone marrow migrate to the thymus. The majority (90–95%) fail to produce T-cell receptor (TCR) and die via the apoptotic pathway . Thymocyte apoptosis can be induced by endogenously produced glucocorticoids and a lack of TCR [87, 88]. Death of a cell clone (clonal deletion) due to TCR-induced apoptosis takes place in immature B cells that produce antiself antibodies in the fetal liver and later in the bone marrow [89, 90]. Animal studies indicate that autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematosis, inflammatory bowel disease and insulin-dependent diabetes, may be caused by dysfunctional apoptosis [91, 92]. There is evidence that cytotoxic T cells kill their target cells by the induction of apoptosis in the target . Activated peripheral cytotoxic T cells and peripheral B cells are removed by apoptosis .
Granulocyte apoptosis is also important for the resolution of the inflammatory response. Inflammatory agents such as lipopolysaccharide (LPS) and granulocyte colony-stimulating factor (G-CSF) inhibit neutrophil apoptosis. During apoptosis induced by TNF-α and NO donors the neutrophil loses its ability to degranulate, thus limiting the inflammation. Intact senescent neutrophils are engulfed by macrophages and degraded within minutes, with no release of pro-inflammatory mediators. When necrotic granulocytes are ingested, the macrophages release mediators leading to inflammation .
Haematological diseases, such as myelodysplastic syndromes, aplastic anaemia, chronic neutropenia or severe β-thalassaemia, are associated with increased apoptosis within the bone marrow .
Many viruses inhibit apoptosis in their target cells to prolong host cell life and permit viral replication. Viruses may encode anti-apoptotic proteins, such as the baculovirus IAPs, the baculovirus p35 and cowpox serpin protein crmA, to promote the development of certain cancers . DNA viruses also contain anti-apoptotic genes. For example, the papilloma virus and adenovirus can encode a p53 inhibitor. Ribonucleic acid (RNA) viruses may also contain anti-apoptotic genes .
Human immunodeficiency virus (HIV) infection is characterised by a decreased proliferation of T cells with loss of CD4+ cells initially, and loss of CD8+ cells, natural killer cells and neurons later. Inappropriate induction of apoptosis in HIV-infected CD4+ cells is triggered by the virus . Another viral product, the HIV-1 transactivating protein, Tat, is produced by infected cells and taken up by uninfected T cells. It enhances apoptosis by reducing intracellular antioxidant levels, producing oxidative stress .
Sepsis and multiple organ dysfunction syndrome
Sepsis is often accompanied by MODS, which may caused by apoptosis . Systemic release of cytokines such as TNF-α and IL-1β by bacterial LPS, resulting in increased intracellular calcium and enhanced production of oxygen free radicals, is thought to produce apoptosis in various organs [101–104]. TNF-α may act by causing macrophage release of IL-1β. In a murine model of septic shock, direct application of TNF-α caused apoptosis in hepatocytes . Further support for the role of apoptosis in endotoxic shock is provided by the finding that mice deficient in ICE have a marked resistance to septic shock . It is postulated that apoptotic (as opposed to necrotic) cell death may confer advantages to the septic organism by limiting the inflammatory response . Apoptosis is an important cause of death of lymphocytes in the immune system (thymus, lymph nodes and spleen) during bacterial and viral infections [101, 108–113]. Furthermore, apoptotic cell death has been shown in the lung, ileum, colon and skeletal muscle during sepsis . Apoptosis may be an important regulator of a balance between pro- and anti-inflammatory factors and this balance is achieved through widespread lymphocyte apoptosis and subsequently diminished cytokine load .
Trauma patients with and without septic complications show decreased neutrophil apoptosis associated with increased tyrosine phosphorylation leading to excessive tissue damage [115, 116]. Patients with SIRS and following major elective aortic surgery also show decreased neutrophil apoptosis, possibly due to circulating pro-inflammatory anti-apoptotic factors such as LPS, TNF-α, interferon (IFN)-γ, G-CSF and granulocyte/monocyte colony-stimulating factor . The persistence of neutrophils in such conditions may contribute to the pathophysiology of MODS through the excessive release of toxic granule components and ROS leading to tissue destruction. Apoptosis can be partially restored by the addition of anti-TNF antibody and IL-10 . Macrophages show increased rates of apoptosis in polymicrobial sepsis, associated with NO and caspase activation, leading to loss of phagocytic function .
There is evidence that a failure to initiate apoptosis following DNA damage may cause cancer. Elevated levels of c-myc are found in many tumours [120–122]. In human follicular lymphoma, translocation of chromosomes 14 and 18 causes up-regulation of Bcl-2‘oncogene’ expression [48, 123]. Levels of Bcl-2 are elevated in various other human cancers, for example lymphomas, leukaemias, adenocarcinomas, renal and lung cancers, neuroblastomas and melanomas [123–126].
Other mutations may also be involved in carcinogenesis. A notable example involves the tumour suppressor p53 that represses Bcl-2 expression. The p53 gene is deficient in over half of human cancers [127–130]. For example, in Wilm's tumour p53 mutations occur in anaplastic areas.
Gastrointestinal diseases may be associated with excessive or defective apoptosis. Shigella dysenteriae causes excessive apoptosis of macrophages in the lamina propria of the intestine by the release of IL-1β. Mice expressing a mutant nonfunctional N-cadherin in intestinal villi develop changes similar to Crohn's disease and show increased rates of apoptosis in both villi and crypts, as well as higher rates of adenomas.
Progressive inhibition of apoptosis appears to be involved in the pathogenesis of gastrointestinal neoplasia, in particular colorectal cancer . Genes that regulate apoptosis are mutant in colonic and gastric cancers. Wild-type p53, when introduced into human colon cancer cell lines, inhibits cell growth and induces apoptosis . However, p53 expression is associated with a poorer prognosis in both colorectal and gastric cancers  possibly owing to mutant or nonfunctional p53 that fails to induce the usual apoptosis. The Bcl-2 protein can also be detected in human cancers and is highest in adenomas .
Hepatic cells develop apoptosis when infected with viruses as in chronic hepatitis . Abnormal activation of cytotoxic T cells may be involved in human fulminant hepatitis . Paracetamol stimulates increases in intracellular calcium which activates Ca2+-dependent nucleases . Apoptosis also appears to mediate allograft rejection in a model of liver transplantation .