What is Apoptosis?

Unlike most terms used in biomedical science, the term apoptosis is not simple to define, and this has led to some confusion and controversy. The proper pronounciation is also controversial (if in doubt try ap-a-tow'-sis). In any case there is intense current interest in this area, with a recent exponential expansion in PubMed textword citations. There are exciting and powerful ideas emerging from these studies, and I will try to give here a thumbnail description of the basic phenomenon, based largely on the diagram above.

The term apoptosis was coined in a now-classic paper by Kerr, Wyllie, and Currie (Brit J. Cancer 26:239) in 1972 as a means of distinguishing a morphologically distinctive form of cell death which was associated with normal physiology. Apoptosis was distinguished from necrosis, which was associated with acute injury to cells. Apoptosis is characterized by nuclear chromatin condensation, cytoplasmic shrinking, dilated endoplasmic reticulum, and membrane blebbing. Mitochondria remain unchanged morphologically. This type of cell death is often hard to observe in vivo because the dying cells are rapidly phagocytosed by tissue macrophages, and this phagocytosis is clearly different from that seen in inflammation, when activated macrophages are recruited from outside the immediate area of death. The above diagram (top line) illustrates the morphological changes associated with apoptosis. The simplest way to observe this phenomenon in vitro is to use a cell permeant DNA-staining fluorescent dye such as Hoechst 33342, which allows a striking visualization of the chromatin condensation.

Apoptotic death can be triggered by a wide variety of stimuli, and not all cells necessarily will die in response to the same stimulus. Among the more studied death stimuli is DNA damage (by irradiation or drugs used for cancer chemotherapy), which in many cells leads to apoptotic death via a pathway dependent on p53. Some hormones such as corticosteroids lead to death in particular cells (e.g., thymocytes), although other cell types may be stimulated. Some cells types express Fas, a surface protein which initiates an intracellular death signal in response to crosslinking. In other cases cells appear to have a default death pathway which must be actively blocked by a survival factor in order to allow cell survival. This is the case illustrated in the diagram, where the small black circles (top left) represent a survival factor which normally binds to its cell surface receptor. When the survival factor is removed, the default apoptotic death program is triggered (top center).

Biochemical correlates of these morphological features have emerged during the subsequent years of study of this phenomenon. The first and most dramatic is DNA fragmentation, which was described by Wyllie in 1980. When DNA from apoptotically dying cells was subjected to agarose gel electrophoresis, ladders with ~200 bp repeats were observed, corresponding histone protection in the nucleosomes of native chromatin. Subsequent pulsed field gel techniques have revealed earlier DNA cleavage patterns into larger fragments. Since even a few double stranded DNA breaks will render the cell unable to undergo mitosis successfully, such DNA fragmentation can be regarded as a biochemical definition of death. However, in some apoptotic systems (e.g., Fas killing of tumor cells) artificially enucleated cells lacking a nucleus still die, showing that the nucleus is not always necessary for apoptotic cell death. This DNA cleavage is depicted in the central blow-up at the bottom of the diagram above.

The changes in the apoptotic cell which trigger phagocytosis by non-activated macrophages have been investigated by several groups. Macrophages appear to recognize apoptotic cells via several different recognition systems, which seem to be used preferentially by different macrophage subpopulations. There is good evidence that apoptotic cells lose the normal phospholipid asymmetry in their plasma membrane, as manifested by the exposure of normally inward-facing phosphatidyl serine on the external face of the bilayer. Macrophages can recognize this exposed lipid headgroup via an unknown receptor, triggering phagocytosis. Exposure of phosphatidyl serine on the surface of apoptotic cells is depicted in the right blow-up at the bottom of the diagram above.

Another biochemical hallmark of apoptotic death which increasingly appears general is the activation of caspases, which are cysteine proteases related to ced-3, the "death gene" of the nematode Caenorhabditis elegans.  Caspases seem to be widely expressed in an inactive proenzyme form in most cells.  Their proteolytic activity is characterized by their unusual ability to cleave proteins at aspartic acid residues, although different caspases have different fine specificities involving recognition of neighboring amino acids. Active caspases can often activate other pro-caspases, allowing initiation of a protease cascade. While several protein substrates have been shown to be cleaved by caspases during apoptotic death, the functionally important substrates are not yet clearly defined.  Protein degradation by the major effector caspase Caspase-3 is shown in the left blow-up at the bottom of the diagram above.  Persuasive evidence that these proteases are involved in most examples of apoptotic cell death has come from the ability of specific caspase inhibitors to block cell death, as well as the demonstration that knockout mice lacking caspase 3, 8 and 9 fail to complete normal embryonic development..

A critical issue is how caspases become initially activated, which seems to be an irreversible commitment towards death. It seems that aggregation of some pro-caspases (those with large pro-domains) allows them to autoactivate.  Recent experiments make it clear that mitochondria are involved in one major pathway involving activation of pro-caspase-9.  Other experiments show that ligands crosslinking death receptors such as Fas trigger formation of a cytoplasmic complex in which pro-caspase-8 is aggregated and activated.  In both cases these initiator caspases in turn activate a cascade of other pro-caspases leading to death.

While there is much to be learned about the molecular pathways leading to apoptotic cell death, it is increasingly clear that cell death is a normal part of normal biological processes.  This had not been appreciated until relatively recently, and our understanding of such death, and our ability to manipulate it, could allow therapeutic intervention in major diseases such as cancer, heart disease, stroke, AIDS, autoimmunity, degenerative diseases, and others.

Last modified by Pierre Henkart, February 1, 1999