The vast majority of cells are formed by cell division. Cell division may be common in certain tissues but especially so in early development, when rapid cell proliferation underlies the growth of multicellular organisms and the progression toward maturity. Throughout development, cell death is common, and by the time of maturity, an equilibrium is reached between cell proliferation and cell death.
While some cell loss is accidental—the result of injury or disease—planned or programmed cell death is very common and is functionally important. Like the organisms that contain them, cells age and the aging process, cell senescence, is related to various factors, including the frequency of cell division.
Most of the cells in mature animals are nondividing cells, but some tissues and cells turn over rapidly
The number of cells in a multicellular organism is determined by the balance between the rates of cell proliferation (which depend, in turn, on continued cell division) and of cell death. Tracking the birth and death of mammalian cells in vivo is problematic. Most of our knowledge about rates of mammalian cell proliferation has therefore come from cultured cells where, under normal circumstances, one turn of the mammalian cell cycle lasts approximately 20–30 hours.
M phase (mitosis and cytokinesis) lasts only about 1 hour and cells spend the great majority of their time (and do most of their work) in interphase. Interphase comprises three cell cycle phases: S phase (DNA synthesis) and two intervening, or gap, phases that separate it from M phase, G1 phase (cell growth, centrosome duplication, and so on) and G2 phase (preparation of factors needed for mitosis). Of the two gap phases, G1 is particularly important and its length can vary greatly, under the control of several regulatory factors. Furthermore, when the supply of nutrients is poor, progress through the G1 phase of the cycle may be delayed.
If the cells receive an antiproliferative stimulus they may exit the cell cycle altogether to enter a modified G1 phase called the G0 phase (Figure 1A). Cells in the G0 phase are in a prolonged nondividing state. But they are not dormant. They can become terminally differentiated; that is, irreversibly committed to serve a specialized function. Most cells in the body are in this state, but they often actively synthesize and secrete proteins and may be highly motile.
Fig1. The cell cycle showing the three major checkpoints. (A) Passage through the cell cycle is controlled by checkpoints preceding transitions between phases. For example, cells can only leave G2 phase to proceed with mitosis (M phase) if DNA has been replicated and conditions are conducive for cell division. Cells in G1 phase may enter either a quiescent G0 phase, from which they can re-enter the cell cycle, or a terminally differentiated state. (B) Each checkpoint is regulated by a specific cyclin-dependent kinase (Cdk) (shown here by a rectangle) bound to a cyclin protein (represented here by the gray circle). Cyclins include cyclin E (important for G1 phase), cyclin A (important for S phase), and cyclin B (important for M phase), and are synthesized and degraded at specific times within the cell cycle, limiting the availability of each cyclin–Cdk complex.
G0 cells can also continue to grow. For example, after withdrawing from the cell cycle, neurons become progressively larger as they project long axons that continue to lengthen until growth stops at maturity. For some neurons, the cytoplasm–nucleus ratio increases by more than 100,000 times during this period. Some G0 cells do not become terminally differentiated but are quiescent. In response to certain external stimuli they can rejoin the cell cycle and start dividing again in order to replace cells lost through accidental cell death or tissue injury.
Mature multicellular animals do contain some dividing cells that are needed to replace cells that naturally undergo a high turnover. Sperm cells are continuously being manufactured in male mammals. There is also a high turnover of blood cells, and gut and skin epithelial cells are highly proliferative to compensate for the continuous shed ding of cells from these organs. Even the adult mammalian brain, long believed to be unable to make new neurons, is now known to have three regions where new neurons are born.
Mitogens promote cell proliferation by overcoming braking mechanisms that restrain cell cycle progression in G1
Cell proliferation is regulated by intrinsic (intracellular) factors and by extracellular signals. The intrinsic signals that regulate the cell cycle generally hold back (restrain) the cell cycle in response to sensors that indicate some fault, or unfavorable circumstance, at certain cell cycle checkpoints (Figure 1A).
The transition from one phase of the cell cycle to the next one is regulated by different cyclin-dependent kinases (Cdk). For example, Cdk1 and Cdk2 regulate entry into mitosis and S phase, respectively. Cdk concentrations are generally constant throughout the cell cycle but the Cdks are only active when they are bound by a cyclin protein. Different cyclins are synthesized and degraded at specific points in the cell cycle (Figure 1B). Thus, the amounts of individual cyclin–Cdk complexes, and ultimately of activated Cdks, parallels that of the amounts of the cyclins that they bind.
Unicellular organisms tend to grow and divide as rapidly as they can, but in multicellular organisms, cells divide only when the organism needs more cells. The start check point at the G1 /S boundary is a major target for regulators that prevent cell division. For example, in mammalian cells, the Rb retinoblastoma protein and the p53 protein cause cells to arrest in G1 if they contain damaged DNA.
In multicellular organisms, the cells that do divide must receive extracellular signals called mitogens that stimulate them to divide. Mitogens typically regulate cell division by overcoming intracellular braking mechanisms operating in the G1 phase that restrain progress through the cell cycle.
The braking mechanisms in G1 are naturally overcome by factors promoting S phase, such as E2F, a regulator that controls the synthesis of many proteins needed for S phase. During G1 , E2F is initially inhibited by being bound by the negative regulator Rb. As G1 progresses, regulatory protein complexes (cyclin D–Cdk4 and cyclin E–Cdk2) accumulate, resulting in phosphorylation of Rb. Phosphorylated Rb has much lower affinity for E2F, freeing it to promote the synthesis of factors needed for S phase.
To ease the normal restraints on passage through G1 , mitogens must first bind to transmembrane receptor tyrosine kinases, stimulating a signal-transduction pathway that includes a small GTPase known as Ras and a MAP (mitogen-activated protein) kinase cascade. Ultimately transcription factors are activated that promote the transition to S phase (Figure 2). Downstream targets include proteins such as MYC, which stimulates production of both E2F and of cyclin–Cdk complexes that phosphorylate Rb and so liberate E2F.
Fig2. Mitogens promote cell proliferation through MAP kinase pathways. Mitogens bind to tyrosine receptor kinases causing the monomers to dimerize and cross-phosphorylate each other. The activated receptor relays the signal through an accessory protein leading ultimately to activation of a MAP kinase and subsequently the activation of transcription of target genes, such as the gene encoding the MYC transcription factor. MYC in turn activates various genes, including some that lead to increased G1-Cdk activity, which in turn causes phosphorylation of the retinoblastoma protein Rb. Unphosphorylated Rb normally binds the transcription factor E2F and keeps it in an inactivate state, but phosphorylation of Rb causes a conformational change so that it releases E2F. The activated E2F transcription factor then activates the transcription of genes that promote S phase, notably the cyclin A gene. Red spikes on the phosphate groups of the tyrosine kinase receptor and on the Ras, MAP kinase, G1-Cdk and E2F proteins signify activation of the relevant protein; black arrows indicate transcriptional activation.
As detailed in Chapter 16, cancer cells find ways of avoiding restrictions on the cell cycle, sometimes by mutating genes that code for checkpoint control proteins.
Cell proliferation limits and the concept of cell senescence
Intracellular mechanisms limit cell proliferation when it is not required. As organ isms age (senesce), physiological deficits accumulate that undoubtedly have a cellular basis, including progressive oxidative damage to macromolecules. Cell senescence cannot easily be studied in vivo; instead, cell culture models have been used. Fibroblasts grown in culture from surgically removed human tissue typically achieve about only 30–50 population doublings in standard medium—the Hayflick limit. Proliferation rates are initially normal but gradually decline and then halt as the cells become arrested in G1 . They enter a terminal, nondividing state where they remain metabolically active for a while, but eventually die.
How this replicative cell senescence relates to cells in vivo and to organismal aging is not fully understood. Links between cellular and organismal senescence were postulated and for many years the Hayflick limit was viewed to be age-dependent. The modern consensus, however, is that there is no compelling evidence to support a relationship between replicative capacity in vitro and the age of the donors that provided the fibroblasts.
The phenomenon of cell senescence described above suggested the existence of a cellular “biological clock’’ or, more accurately, replication counter. In cells undergoing senescence, the telomeres (chromosome ends) progressively shorten at each cell division (because of the problem of replicating chromosome ends). Eventually, the erosion of the telomeres destabilizes the telomeric T-loops, leading to removal of the protective telomere cap. When this happens, the uncapped DNA at the end of the chromosomes is no different from the double-stranded DNA breaks resulting from DNA damage. As a result, cell surveillance systems monitoring DNA integrity likely induce the cells to enter a state of senescence. Cell senescence effectively appears to be a type of DNA damage response.
Not all cells are subject to cell senescence. Certain cells from embryos can be propagated for very long periods in culture and are effectively immortal, as are tumor cells, which subvert normal cell cycle controls. The idea that lack of telomere integrity is important in cell senescence is supported by studies of telomerase, an enzyme that counteracts telomere shortening by re-elongating telomeres. While most normal human somatic cells have negligible or tightly regulated telomerase activity, immortal cells have constitutively high telomerase levels. Fibroblasts and other somatic cells can be artificially immortalized by transfecting them with a gene encoding the catalytic sub unit of telomerase.
Large numbers of our cells are naturally programmed to die
Throughout the existence of a multicellular organism, individual cells are born and die. Cell death occasionally occurs because of irreversible accidental damage to cells (necrosis). Causes include physical trauma, exposure to extreme temperatures, and oxygen starvation. Typically, large groups of neighboring cells are simultaneously affected. The process leads to leakiness of the plasma membrane, and water rushes in and causes the cells to swell up until the cellular membranes burst. Thereafter, the cells undergo auto digestion, producing local inflammation and attracting macrophages that ingest the cell debris.
In addition to accidental cell death, very large numbers of cells are also deliberately and naturally selected to die throughout the existence of a multicellular organism, even at very early stages of embryonic development. Such programmed cell death (PCD) can occur in response to signals sent, or withheld, by other cells (during development or immune surveillance, for example), or it can arise after a cell’s internal monitoring systems sense major damage to vital cell components such as its DNA, mitochondria, and so on.
A variety of different types of programmed cell death are known. Of these, apoptosis, or type I PCD, has been extensively studied and is characterized by very specific changes in cell structure. Typically, individual cells (rather than groups of cells) undergo apoptosis and, as they die, the cells shrink rather than swell. A characteristic feature is that the chromatin condenses into compact patches that accumulate around the periphery of the nucleus. The nuclear DNA fragments and the nucleus breaks into discrete chromatin bodies. There is violent cellular movement (the cell appears to “boil”), and eventually the cell breaks apart into several membrane-lined vesicles called apoptotic bodies that are phagocytosed.
Another form of PCD is autophagy, a catabolic process where the cell’s components are degraded as a way of coping with adverse conditions such as nutrient starvation or infection by certain intracellular pathogens. Intracellular double-membrane structures engulf large sections of the cytoplasm and fuse with lysosomes, thereby targeting the enclosed proteins and organelles for degradation. Taken to an extreme, autophagy can lead to cell death. Other forms of PCD are known but are poorly characterized.
The importance of programmed cell death
Programmed cell death is crucially important in many aspects of the development of a multicellular organism and is also vital for the normal functioning of mature organisms. A quantitative illustration of its importance in development is provided by the nematode Caenorhabditis elegans, the only multicellular organism for which the lineage of all body cells is known. The adult worm has 959 somatic cells but is formed from a total of 1090 cells, 131 of which apoptose during embryonic development.
The selection of cells destined to die in C. elegans development is highly specific: the same 131 cells die in different individuals. Although 959 out of the 1090 cells survive, apoptosis is the default cell fate: all 1090 cells are programmed to die. The cells that survive do so by signaling to each other: they secrete proteins called survival factors that bind to cell surface receptors and override default apoptosis pathways. While our understanding of programmed cell death in early human development is incomplete, PCD is used to remove defective and excess or unwanted cells during the development of mammalian embryos and fetuses (see Figure 3 for an example).
Fig3. Programmed cell death sculpts fingers and toes during embryonic development. (A) The digits of a mouse paw are sculpted from the platelike structure seen at embryonic day 12.5 (E12.5); the digits are fully connected by webbing. The cells within the webbing are programmed to die and have disappeared by E14.5. The dying cells are identified by acridine orange staining on the right panel. (B) Incomplete programmed cell death during human hand development results in webbed fingers. (A, from Pollard TD & Earnshaw WC [2002] Cell Biology, 2nd edn. With permission from Elsevier.)
In complex multicellular organisms, cell death and cell proliferation need to be carefully balanced in the mature organism. There is a high cell turnover in some systems, such as for mammalian blood cells and epithelial cells in gut and skin. About 100,000 cells are programmed to die each second in adult humans but are replaced by mitosis. For some cells, such as B and T lymphocytes and neurons, special mechanisms are used to generate diversity, and here PCD is responsible for quality control, removing cells where the diversity-generating mechanisms have been unproductive (Table 1). As described in Section 3.4, PCD is also important in the immune system, where T cells and natural killer cells are employed to induce the death of body cells that are perceived to be harmful in some way, including virus-infected cells and tumor cells.
Table1. SOME IMPORTANT FUNCTIONS OF PROGRAMMED CELL DEATH
PCD has increasingly been recognized to be important in human disease. Aberrations in apoptosis play important parts in the etiology of autoimmune diseases, virally-induced diseases, and cancer. For example, helper T lymphocytes are key cells in immunosurveillance systems that we use to recognize and kill virally-infected cells.
Human immunodeficiency virus (HIV) proteins cause apoptosis of these key immune system cells, allowing disease progression toward acquired immunodeficiency syn drome (AIDS). Many cancer cells have devised strategies to oppose the immunosurveillance systems that normally induce apoptosis of cancer cells. Successful chemotherapy often relies on using chemicals that help induce cancer cells to apoptose. Other forms of PCD are important in neurodegenerative disease, such as in Huntington disease and Alzheimer’s disease, as well as in myocardial infarction and in stroke, where secondary PCD caused by oxygen deprivation in the area surrounding the initially affected cells greatly increases the size of the affected area.
Apoptosis is carried out by caspases in response to death signals or sustained cell stress The key molecules that execute apoptosis are the caspase family of proteases. These enzymes have cysteine at their active site and cleave their substrates on the C-terminal side of aspartate residues. Inactive caspase precursors (procaspases) are synthesized naturally by all cells and have an N-terminal prodomain that needs to be cleaved off to activate the caspase. There are two classes of procaspases:
• Initiator procaspases, such as caspase 8 and caspase 9, have long prodomains and can undergo autoactivation. Their job is to start off cell death pathways after they have been activated by signals transmitted through cell surface receptors or from internal sensors;
• Effector procaspases, such as caspases 3, 6, and 7, have short prodomains that are cleaved by initiator caspases. Effector caspases cleave about 100 different target proteins, including nuclear lamins (causing breakdown of the nuclear envelope) and cytoskeletal proteins (destroying cell architecture). In addition, cells naturally possess a cytoplasmic caspase-activated DNase. The DNase is normally kept inactive by being bound by an inhibitor protein. Effector caspases can cleave the inhibitor protein, however, thereby releasing the DNase, which migrates to the nucleus and fragments cellular DNA.
Protein modifications such as phosphorylation and ubiquitylation are reversible. By contrast, proteolysis is irreversible: once a peptide bond is cleaved, cells cannot re-ligate the cleavage products. Once apoptosis has been initiated, therefore, there is no going back.
As described below, extrinsic and intrinsic apoptosis pathways use different initiator caspases, but they use the same types of effector caspases to cleave target proteins.
Extrinsic apoptosis pathways
During development, neighboring cells exchange death signals and survival fac tors and have specific receptors for both types of signal. Cells that receive enough of a survival factor will live, because they initiate cell survival pathways that suppress programmed cell death pathways; those that don’t receive enough survival factor will die. Competition between cells to receive enough survival factor is thought to control cell numbers both during development and in adulthood.
Many death receptors are members of the TNF (tumor necrosis factor) superfamily. Fas is a well-studied example. Its ligand (FasL) forms trimers and induces the Fas receptor to trimerize, causing clustering of death domains on the receptor’s cytoplasmic tail. The clustered death domains attract binding of an adaptor protein, FADD, that then recruits procaspase 8 to initiate a series of caspase cleavages (Figure 4).
Fig4. Apoptosis pathways. Extrinsic apoptosis pathways involve activation of a cell surface death receptor by a ligand on a neighboring cell. The Fas death receptor is normally found as a monomer but is induced to form a trimer by its trimeric ligand FasL. The resulting clustering of Fas receptors recruits the FADD adaptor protein. FADD acts as a scaffold to recruit procaspase 8, which undergoes autoactivation to initiate a caspase cascade. Intrinsic apoptosis pathways are activated when vital cell components are damaged or stressed, for example in response to harmful radiation, chemicals, hypoxia, and so on. These pathways are activated from within cells using mitochondrial and endoplasmic reticulum components. In the mitochondrial pathways, proapoptosis factors, such as Bax shown here, form oligomers in the mitochondrial outer membrane, forming pores that allow release of cytochrome c. In the cytosol, cytochrome c binds and activates the Apaf1 protein and induces the formation of an apoptosome that activates procaspase 9 and ultimately the same effector caspases (caspase 3, caspase 6, and caspase 7) as the death receptor pathways.
Intrinsic apoptosis pathways
Apaf1 intrinsic pathway Bax cytochrome c procaspase 9 caspase 9 mitochondrion Cells are induced to apoptose when they are sufficiently stressed that sensors indicate significant damage to certain key components. The integrity of genomic DNA needs to be protected, and mitochondria are vitally important energy producers. The endoplasmic reticulum is also crucial. As well as being required for protein and lipid synthesis, it is essential for correct protein folding and is the major intracellular depot for storing Ca2+, the most widely used second messenger in cell signaling. Prolonged changes in Ca2+ con centration in the endoplasmic reticulum or the accumulation of unfolded or misfolded proteins can lead to apoptosis.
The mitochondrial pathway of apoptosis is initiated when proapoptosis cytoplasmic proteins such as Bax are activated. Bax then binds to the mitochondrial outer membrane and forms oligomers, permitting release of cytochrome c, which, in turn, activates the cytoplasmic Apaf1 protein causing activation of procaspase-9 (see Figure 4). Bax belongs to a large family of apoptosis regulators that includes antiapoptosis factors (such as Bcl-2) as well as proapoptosis factors.
