Tumor Suppressor Genes

When functioning normally, a tumor suppressor gene prevents the formation of one or more types of cancer. Mutations in tumor suppressor genes that interfere with their function can be inherited in the germ line or can occur in somatic cells; the accumulation of such mutations can allow cancer to develop.

When functioning normally, a tumor suppressor gene prevents the formation of one or more types of cancer. Mutations in tumor suppressor genes that interfere with their function can be inherited in the germ line or can occur in somatic cells; the accumulation of such mutations can allow cancer to develop.

Cancer as a Multistep Process
For cancer to develop, defects must arise in the regulatory processes that normally ensure that cells in our bodies divide, differentiate and die at exactly the right time and place. The number of regulatory circuits that must be disrupted for a cell to become cancerous depends on the cell type. But in all situations the change from a normal cell to a cancerous cell -'carcinogenesis' -is a multistep process with each step reflecting a genetic alteration. Although the genes that must be mutated for cancer to develop differ between cell types, Hanahan and Weinberg have suggested that the catalog of cancer cell genotypes reflects six essential perturbations of cell physiology, some or all of which are necessary: Acquisition of self-sufficiency in growth signals. Normal cells only move from a quiescent state into an actively proliferating state in the presence of mitogenic growth signals, which are transmitted into the cell when signaling molecules bind to transmembrane receptors. Cancer cells acquire genetic alterations that enable them to generate their own growth signals. Loss of sensitivity to antiproliferative signals. Antiproliferative signals can prevent normal cell growth, either by driving cells reversibly into a quiescent state or by inducing them to undergo an irreversible process of differentiation in which they lose their capacity to divide. Cancer cells acquire changes that block these signals or their effects. Evasion of apoptosis. Programmed cell death by the process known as apoptosis is a mechanism that protects against uncontrolled cell growth in most if not all cell types. Mutations that disrupt this process commonly occur in cancer. Acquisition of limitless replicative potential. Once normal cells have undergone a certain number of divisions they stop growing -a process that is known as 'senescence'. Mutations in cancer cells can disrupt this inbuilt program, conferring unlimited replicative potential or 'immortalization'. Acquisition of capacity for sustained angiogenesis. The oxygen and nutrient supply necessary for normal cell function is ensured by precise regulation of angiogenesis -the formation and growth of blood vessels. When tumors begin to grow they are at first limited in size by the availability of already existing blood vessels and to continue to grow they must develop the ability to induce new blood vessel growth. Acquisition of invasive and metastatic capacity. Most deaths from cancer in humans are not due directly to the growth of the primary tumor but to the formation of secondary tumors (metastases) at distant sites in the body. The process of metastasis requires that the cells become capable of invading into the tissue that surrounds the primary tumor, reaching the circulation and initiating new growth in sites where their normal parent cells cannot grow.

Oncogenes and Tumor Suppressor Genes
The types of genetic change listed in the previous section fall into two broad classes. The first class results in altered genes that have acquired new functions that the normal version of the gene does not possess: these are gain-of-function mutations. The second class results in genes that have become inactivated: these are loss-of-function mutations. When gain-of-function mutations are involved in carcinogenesis, the activated mutant gene is termed an oncogene and the normal version of the gene a proto-oncogene. When loss-offunction mutations are involved in carcinogenesis, the gene is termed a tumor suppressor gene.
Thus, when a tumor suppressor gene is functioning normally it acts to prevent cancer formation. Of the six types of perturbation listed in the previous section, tumor suppressor gene inactivation is involved most commonly in loss of sensitivity to antiproliferative signals and in evasion of apoptosis (see below

Germ Line and Somatic Events
In general, activation of one allele of a proto-oncogene is sufficient to confer altered properties on a cell, whereas it is necessary to inactivate both alleles of a tumor suppressor gene to alter cellular properties. Therefore, when oncogenes are involved in cancer formation, they are usually activated in somatic cells rather than inherited through the germ line, because the presence of the activated oncogene in the germ line normally affects embryonic development so severely that it causes embryonic lethality.
By contrast, an inactivating mutation in one allele of a tumor suppressor gene may be inherited through the germ line, although mutations can also occur in somatic cells. Individuals heterozygous for such a mutation are at increased risk of developing specific types of cancer because they require fewer somatic mutations for this to happen, and they transmit this risk to offspring that inherit the mutant allele. The same gene can be affected by both germ line and somatic mutations. This explains why similar cancers run in families in some cases and occur sporadically in others, owing completely to somatic mutation.

Gatekeeper and Caretaker Functions
Inactivation of a tumor suppressor gene can directly affect the processes of cell proliferation and apoptosis if the gene has a direct role in antiproliferative signaling or in the cell death program. But there is another mechanism by which inactivation of tumor suppressor genes can function, that is, by making the genome less stable such that mutations or altered numbers of copies of other, directly acting genes occur more frequently. Kinzler and Vogelstein introduced the term 'caretaker' to describe genes that act in this more indirect way, distinguishing them from the 'gatekeeper' genes that directly affect cell proliferation or apoptosis.
The normal function of caretaker genes is to ensure the stable propagation of the genome when cells divide. Two types of genomic instability that arise from caretaker gene inactivation are important in carcinogenesis. The first, chromosome instability, affects the accuracy of chromosome disjunction and results in an increased incidence of aneuploid cells. The second, microsatellite instability, leads to alterations in the number of repeats of microsatellite sequences and therefore has a selective effect on genes that contain microsatellites at sites that are crucial to their function. Microsatellite instability results from mutations in deoxyribonucleic acid (DNA) repair genes. (See Chromosomal Instability (CIN) in Cancer; Mismatch Repair Genes.) Four subdivisions of gatekeeper genes can be made: first, genes encoding intracellular proteins that regulate progression through a specific stage of the cell cycle; second, genes encoding receptors for secreted molecules that function to inhibit cell proliferation; third, genes encoding proteins that regulate cell differentiation; and last, genes encoding proteins that signal or are otherwise involved in apoptosis. The classification of tumor suppressor genes into caretakers and gatekeepers is useful but should not be applied too rigidly as some genes can function in both capacities. The p53 (tumor protein p53; TP53) gene, for example, functions as a caretaker in ensuring chromosome stability and as a gatekeeper in signaling apoptosis in response to DNA damage. The concept of caretaker and gatekeeper functions is perhaps more useful than that of caretaker and gatekeeper genes. (See Caretakers and Gatekeepers.)

RB1 Gene and Retinoblastoma
As is often the case in science, our understanding of tumor suppressor genes was advanced by finding an example that was particularly straightforward to study and then applying the lessons to progressively more complex situations. The straightforward example emerged from a study of a relatively rare cancer called childhood retinoblastoma. Almost half of the cases of this cancer are familial and the remainder are sporadic. In familial retinoblastoma, tumors develop in the retina, usually before the individual is 5 years old, and in general affect both eyes. If diagnosed early enough, the tumors respond well to radiation or other treatments that preserve useful vision; if they have progressed beyond this point, however, surgical removal of the eye is necessary to prevent further spread. In sporadic retinoblastoma, tumor formation occurs later and there is usually only a single tumor in one eye. (See Retinoblastoma.) In the early 1970s, Knudson found that the age distribution at diagnosis in individuals with retinoblastoma in both eyes (bilateral retinoblastoma) corresponded to what was predicted if each tumor arose as a result of a single event that affected a retinal cell. But for unilateral retinoblastoma, in which only one eye is affected, the data corresponded to what was predicted if two events affecting a single retinal cell were necessary for tumor development. As the resulting tumors are similar in both groups of individuals, he drew the inspired conclusion that the 'events' correspond to genetic mutations and that one mutation is already present in the germ line of individuals with bilateral retinoblastoma.
Comings suggested shortly thereafter that the two mutations involved the two alleles of the same gene.
But it was another 20 years before the technology was developed that allowed these predictions to be confirmed at the molecular level. The gene responsible was identified and named retinoblastoma 1 (RB1), and mapped to the long arm of chromosome 13.
Individuals with familial retinoblastoma are therefore heterozygous for a mutation that inactivates RB1. They have an increased risk (more than 90%) of developing retinoblastoma because only one allele remains to be mutated in each retinal cell and the chance of this happening in at least one cell is very high. By contrast, in the rest of the population, this chance is low because two mutations have to happen in the same cell. Although both RB1 alleles have to be inactivated for a cell to grow into a tumor, it is heterozygous individuals, who inherit only one mutant gene, that are at risk. Another way of saying this is that susceptibility to retinoblastoma is inherited as a dominant trait (to be precise, an autosomal dominant trait). This is an example of the more general phenomenon that the same mutation can have both a recessive and a dominant effect depending on the phenotype under study. In cells, where the phenotype is the growth of a cell into a tumor, the mutation is recessive. In the individual, where the phenotype is susceptibility to retinoblastoma, the mutation is dominant, and of course this is what is clinically significant.

Loss of heterozygosity
The loss of the remaining normal allele from heterozygous cells can occur in individuals with familial retinoblastoma by one of many mechanisms (Figure 1). Whereas germ-line mutations are usually either point mutations or small deletions, somatic events that lead to the production of cells with no remaining wild-type allele often involve whole chromosomes or large chromosome segments. These more extensive genome changes are not compatible with normal embryonic development, which explains their absence from the spectrum of germ-line mutations, whereas the lesser requirements for uncontrolled growth as a tumor cell place fewer restrictions on the types of somatic event that are commonly seen.

Function of the RB1 gene product
The RB1 gene encodes a protein with a relative molecular mass of about 105 000 that regulates the passage of the cell through different phases of the cell cycle. This protein, retinoblastoma 1 (RB1), undergoes changes in phosphorylation mediated by cyclin-dependent kinases whose activity is itself dependent on cell-cycle phase. The unphosphorylated protein binds to and inhibits the function of other proteins that are required for progression through the G1 phase of the cell cycle and entry into S phase. During passage through G1, different residues of RB1 are phosphorylated at different time points and this releases the bound proteins in a progressive manner, allowing the cell cycle to proceed. This process is . Independent mutation of the wild-type allele (i) is possible but relatively rare. Aberrant cell division in which segregation of chromosomes into daughter cells does not occur faithfully can lead to loss of the whole chromosome carrying the wild-type allele (ii), and this may be combined with duplication of its homolog carrying the mutant allele (iii, iv). Alternatively, after DNA replication (v) exchange of material between chromatid arms can occur by mitotic recombination (vi). As the two copies of the mutant allele are now on different copies of the chromosome, they will segregate independently at the following cell division (vii), and some daughter cells will acquire two copies of the mutant allele and no wild-type allele. These pathways lead to four possible genotypes (1-4), all of which have no remaining wild-type allele. Not all of these pathways are important for all tumor suppressor genes; in particular, for some chromosomes genotype 2 may be inconsistent with cell survival because essential gene products are not produced in sufficient amounts. regulated in response to extracellular growth-promoting and growth-inhibiting molecules that transmit signals into the cell by binding to receptors on the cell surface. (See Evolution: Selectionist View.)

Mouse Rb mutants
The mouse gene equivalent to RB1 is designated Rb (sometimes Rb1). Mice heterozygous for an inactive allele of Rb do not develop retinoblastomas but instead develop tumors in the pituitary. The latter do not occur in human RB1 heterozygotes, probably because the structure of the mouse and human pituitaries is different. The reason for the absence of retinoblastomas in the mice is still under study and may help us to understand in more detail how retinoblastomas develop. Where the mice provide new information, however, is in allowing us to study homozygotes. No human RB1 mutant homozygotes have been reported, because families in which both parents are heterozygotes are extremely rare. In mice, however, planned matings of heterozygotes can be carried out. These matings have shown that homozygous embryos die before birth and have abnormal development of the blood and nervous system, which arises from increased apoptosis and overabundant or inappropriately located cell division. The tissues affected are those where cells normally differentiate and stop dividing earliest in embryonic development, which indicates that the embryos die because this process cannot occur normally. This provides evidence for a role for Rb in the process of maturation of dividing precursor cells to nondividing differentiated cells. This ties in with what we know about the role of RB1 in regulating the cell cycle (see above).

Generalization of Knudson's Hypothesis
Although Knudson's studies are consistent with a requirement for only two mutations for retinoblastoma to occur, they do not exclude the possibility that other mutations might be needed. Most retinoblastomas do have some additional genetic abnormalities, but it is not clear whether these are required for the tumor to begin growing or whether they occur after tumor growth has begun and produce faster-growing cells that eventually dominate the population of cells in the tumor. In other types of cancer, it is well established that the situation is more complex and that inactivation of more than one tumor suppressor gene is required for cancer formation.
Several tumor suppressor genes are now known and some examples are listed in Table 1. In some cancers, tumor formation proceeds through well-defined intermediate stages and the requirement for mutation of a particular gene can be identified with a particular stage in the process. The process of developing colon cancer, for example, proceeds through the formation of a benign tumor known as a polyp or adenoma. Mutation of the adenomatous polyposis coli (APC) gene is a requirement for adenoma formation, whereas mutation of either p53 or one of a group of mismatch repair Notably, inactivation of RB1 is required in many cancers whose incidence is not noticeably elevated in individuals with familial retinoblastoma. This apparent anomaly occurs because in these cancers RB1 is merely one of a group of tumor suppressor genes whose inactivation is necessary; therefore, several somatic events still need to accumulate even when an RB1 germ-line mutation is already present, and this results in an increase in incidence that is too small to be distinguished from a chance occurrence.
Although it is a general rule that mutations in both alleles of a tumor suppressor gene are required for cancer to develop, there are exceptions to this rule. There are three common reasons for this: Haploinsufficiency. Usually the amount of gene product produced by one functioning allele is sufficient for a cell to have the normal properties of a cell with two functioning alleles. But this is not always so, and the reduction in the amount of gene product in a cell with only one functioning allele can have a role in cancer development. This phenomenon is termed haploinsufficiency. Dominant-negative mutations. Some mutations lead to the production of an altered protein that is not only inactive itself but can interact with the protein produced by the normal allele in such a way as to inactivate it. These mutations are described as dominant-negative and can reduce the functioning protein to very low amounts without the need for a mutation in the normal allele. Imprinting. For some genes, the allele inherited from one parent is inactivated without mutation, in some or all of the tissues in which it is usually expressed, by a normal developmental process called imprinting. For such genes, it is necessary to mutate only the copy of the gene inherited from the other parent to reduce the amount of functioning protein in the cell to zero. (See Genomic Imprinting at the Transcriptional Level.)

Tumor Suppressor Gene Pathways
In some cell types, cancer development requires the disruption of a regulatory pathway that involves more than one tumor suppressor gene. For example, the phosphorylation of RB1 is carried out by a cyclindependent kinase that can be inhibited by the protein produced by the p16 (cyclin-dependent kinase inhibitor 2A; CDKN2A) gene. If p16 is inactivated by mutation, RB1 is permanently inactivated by phosphorylation in cell types where there is no alternative mechanism to regulate the activity of the kinase. In these cell types, the same result can be achieved by inactivating either p16 or RB1. For example, this happens in pancreatic carcinomas, in which the abnormalities usually include mutations in either p16 or RB1, but not both. But individuals with germ-line mutations in p16 develop melanomas and not retinoblastomas, which emphasizes the fact that differences in regulatory circuits are present between different cell types because they express a different gene repertoire. (See Cell Cycle Checkpoint Genes and Cancer; Pancreatic Cancer.)

See also
Caretakers and Gatekeepers Oncogenes Retinoblastoma Tumor Formation: Number of Mutations Required