GENETIC CHANGES IN CANCER - ETIOLOGY AND PATHOGENESIS
Dr Bruce F. Burns, Dept of Pathology and Laboratory Medicine
Objectives
Describe what the term "clonal" means and it's implications in the biology of neoplasia as compared to hyperplasias. What are the features common to tumor cells of all types, in terms of kinetics and relationships to normal cells?
Describe the concept of tumor progression and its' implications in the clinical evolution of a cancer and its treatment.
Describe the common types of cancer-associated genes in terms of their normal functions and their effects when normal function or control is lost.
Describe the common types of genetic changes that can affect cancer-associated genes, causing their malfunctioning
Suggested Reading:
Robbins Pathologic Basis of Disease, 6’Th Ed., Cotran et al, 1999, p.276-301.
A. INTRODUCTION TO THE BIOLOGY OF NEOPLASIA
Neoplasms, both benign and malignant, are the result of genetic mutations. Usually these mutations are acquired, but occasional tumours arise because of inherited mutations passed on through the germ line from one or both parents. The mutations can be thought of as broadly affecting genes that regulate the growth, differentiation and normal turn-over of a variety of cells. Collectively these genes will be referred to as cancer-associated genes. Some of these mutations result in the overactivity of cancer-associated genes that promote growth (oncogenes), while others inactivate genes that normally suppress the growth and division of cells (tumor suppressor genes). Still other mutations stop cells from dying off in a normal fashion (affecting apoptosis genes). In the past couple of decades we have come to realise that many forms of cancer share similar genetic defects and that these cancers are the result of a series of mutations producing a cumulative effect that we recognize as cancerous behaviour.
B. NEOPLASMS ORIGINATE FROM A SINGLE MUTATED CELL
Most neoplasms are believed to originate from a single parental cell that undergoes a mutational event leading to an altered pattern of growth. All of the progeny of this cell are said to be clonal (or monoclonal) in origin. This feature of monoclonality is most easily recognized in tumors of plasma cells, called myelomas, where all of the tumor cells produce an immunoglobulin of an identical class and idiotype. As will be seen, however, the initial cellular mutation is usually not sufficient for the development of full malignant potential. Rather, the mutated clone of cells acquires a growth advantage over its neighbouring normal cells and is typically "immortalized". Occasionally an organ will develop multiple, separate tumours, each derived from a different clone of cells. This often occurs in the setting of a carcinogenic stimulus that affects the entire organ for long periods of time. This is sometimes referred to as a "field effect" or "field change" and a commonly cited example is that of multiple bladder tumors arising in people exposed to chemical carcinogens, such as aniline dye workers. In this case the carcinogen, excreted in an active form in the urine, affects the entire urothelium from renal pelvis to urethra resulting in multiple, independently arising tumors.
C. CELL KINETICS
Although it was originally thought that tumor cells proliferated at a faster rate than normal cells, it is now clear that many normal tissues, such as bone marrow, actually have cell cycling times which are faster than even the most aggressive malignancies. The difference between malignant tumors and these rapidly proliferating normal cells, however, is that the tumor cells do not die at the same rate as they are replicating, they are effectively immortalized. Thus, there is a progressive accumulation of tumor cells, which is not the case in normal tissues where cell death balances cell replication. Another important factor is the proportion of cells in an actively replicating form (the growth fraction). Tumors often have a higher proportion of cells which are in G1, S, G2 or M phases compared to G0 or resting phase when compared to normal cells of the same type. The growth fraction in some rapidly proliferating normal tissues such as small bowel epithelium and hair follicles, however, is higher than in most tumors. This has implications when treating cancer with chemotherapy agents that non-selectively affect all cells in the growth phase. Tumors with a low growth fraction will be slowly growing (which is good), but they will usually also be less susceptible to chemotherapy or radiotherapy (which is not so good). Most of the side-effects of chemotherapy/radiotherapy are due to the (temporary) damage to rapidly growing normal tissues, resulting in hair loss, bone marrow suppression (anemia, infections, bleeding) and loss of mucosal cells in the gut, etc.
D. MULTISTEP ACQUISITION OF THE NEOPLASTIC PHENOTYPE
The common denominator in the various chemicals, viruses and radiation that cause cancer is the activation or mutation of one or more cancer-associated genes. Whatever the initial step was it rarely, if ever, confers the complete neoplastic phenotype to the cell. Rather the groups of features that characterize a "fully malignant" tumor (immortality, invasive growth, and the ability to promote neovascularization and to metastasize) require a series of progressive alterations involving multiple genes. It appears that a cell that has started on the slippery slope towards being a tumor is at increased susceptibility for the development of additional mutations affecting other oncogenes, probably because DNA repair mechanisms are faulty. These additional mutations other cancer-associated genes, then result in a selective growth advantage of the mutated sub-clone of cells over the original tumor cells. This scenario is consistent with what we know of "tumor progression" or the tendency for progressive mutations to occur during the lifespan of a tumour, which result in increasingly aggressive biologic behaviour over time. The other corollary to this is that tumours become progressively more heterogeneous in terms of the cells that make up the tumor and less likely to behave in a uniform manner when it comes to response to chemotherapy or radiotherapy. This cardinal feature of malignancies, that is "tumor cell heterogeneity", is the single greatest stumbling block to developing a "cure" for cancer.
E. CANCER-ASSOCIATED GENES - WHAT ARE THEY?
Cancer-associated genes include a very diverse group of genetic elements, which when over-expressed, under-expressed or mutated result in a cell with an abnormal growth characteristic. The first forms of these genes were called oncogenes and were originally found in RNA retroviruses. They were referred to as viral oncogenes (these have a prefix v - followed by the specific name of the oncogene, e.g. v-myc). It was later recognized that these oncogenes are actually normal genes found in all cells, where they function to regulate cell growth and differentiation. These genes are referred to as proto-oncogenes in the cell when they are functioning normally or as cellular oncogenes when they are mutated or deregulated in cancer cells. Cellular oncogenes often have the prefix c- followed by the name of the oncogene (e.g. c-myc). Later it was appreciated that another class of cancer-associated genes was responsible for repressing tumor-like behavior and they were called tumor suppressor genes. Later still it was recognised that when genes that affect normal cell death (apoptosis) and DNA repair malfunction tumors may develop as well.
TYPES OF CANCER-ASSOCIATED GENES
- Proto-oncogenes
- Tumor suppressor genes
- Apoptosis genes
- DNA repair genes
1. NORMAL FUNCTIONS OF PROTO-ONCOGENES
There have been more than one hundred distinct cancer-associated genes of viral and cellular origin identified to date. These oncogenes may be conveniently divided into those that have an effect at the cell surface or in the cytoplasm, functioning as hormone receptors or proteins that regulate the levels of critical second messenger molecules. The other site of action of another set of oncogenes is in the nucleus, where the oncogene proteins may modulate the activity of the cells transcriptional machinery. Some of these oncogenes are expressed only during the growth and differentiation phase of a given cell and then become inactive in the adult cell. Others have brief periods of activity related to response to stimuli but are under strict regulatory controls. The following table is a summary of the known mechanisms of action of some of the better-studied oncogenes.
Table of Oncogene Classes
Function Mechanism of Action Examples Growth Factors Overexpression sis Growth Factor Receptors Overexpression or amplification erb-B1 or erb-B2 in lung, breast, ovary, stomach cancers Signal Transduction - -Protein phosphorylation Translocation leading to "fusion protein" abl in chronic myelogenous leukemia -GTP binding proteins Point mutations leading to deregulated overactivity ras in many common cancers, lung, colon, pancreas Nuclear transcription factors Translocation leading to overexpression c-myc in Burkitt lymphoma Amplification n-myc in neuroblastoma Cell Cycle Regulators Translocation, amplification or point mutations Cyclin D1 or CDK4
2. Tumor suppressor genes
Tumor suppressor genes are, in a sense, the opposite of proto-oncogenes. The "grand-daddy" of these is the gene associated with retinoblastoma (Rb). Tumor suppressor genes appear to function in normal cells to suppress cellular proliferation. Rb is such a gene and its mutation is responsible for a hereditary predisposition to developing retinoblastomas of the eye. These malignant retinal tumours arise in infancy in 90% of children who carry the mutant allele of the Rb gene. Interestingly this lesion is inherited as an autosomal dominant although the effects, as one might expect, have to be the result of both alleles being mutated (the "two hit" hypothesis first proposed by Knudson to explain tumor suppressor gene actions). It appears the mechanism for this is that, in the patient who already has one mutant allele, the chance of a single cell, amongst the millions of retinal cells replicating during fetal growth, developing a second mutation or deletion involving the other normal allele is virtually 100%. Thus, a clone of cells arises, with no functioning Rb alleles and these cells go on to form the retinoblastoma tumor because of the complete absence of the Rb gene product. Predictably, there is a very high incidence of bilateral, independently arising, retinoblastomas in the inherited form of this condition. Another non-inherited form of the same tumor arises when a developing retinal cell is hit by two sequential mutations that knock out both of its normal Rb alleles.
p53 is another example of a tumour suppressor gene and is probably the most frequently implicated gene in "cancer" as a whole. This gene functions to block cell division in genetically damaged cells, allowing time for repair. If such repairs are not possible the gene triggers the cell into apoptosis, thus preventing propagation of genetically damaged cells. It has thus been referred to as the "Guardian of the Genome". If the gene is knocked out the affected cells are more prone to further mutations. Defects in this gene are found in most colon cancers and almost half of breast and lung cancers as well as non-epithelial tumors such as lymphomas and leukemias. An inherited defect of this gene results in the Li-Fraumeni syndrome, a multicancer predisposition syndrome.
Recently another example of defective tumour suppressing gene has been described in a very common tumor, carcinoma of the colon. The gene, which is deleted in a high percentage of colonic carcinomas, is believed to have functions related to cell-cell adhesion. This new tumour-suppressing gene has been termed the DCC gene (deleted in colorectal carcinomas).
The characteristic feature of all tumor-suppressing genes is that they act as recessive genes, in that both normal alleles must be mutated before the cell will miss them. This is contrasted with the effects of oncogene mutations that typically behave in a dominant fashion. The strict specificity of cell type that these tumor suppressor genes were thought to be expressed in is starting to be doubted as well. It appears, for instance, that the DCC gene may be affected in lung carcinomas as well in the colon.
3. Apoptosis related genes
Another novel family of genes, called bcl-2, which has recently been linked to a form of malignant lymphoma has been found to prevent apoptosis in cells (also known as "programmed cell death", i.e. the normal death of cells). When the chromosome containing bcl-2 is translocated to another chromosome it loses it's normally regulatory influences and is constituitively (continuously) expressed, resulting in the indefinite lifespan of the mutated cells.
4. DNA repair enzyme defects
While not causing cancer directly defects in these proteins predispose affected individuals to a variety of cancers. Probably the best known of these is the hereditary condition known as xeroderma pigmentosum, where a defect in repair of radiation inducted DNA damage occurs. These patients are very sensitive to sunlight and develop a number of different skin tumors, including malignant melanomas.
The hereditary non-polyposis colon cancer syndrome is another example where an inheritance of one defective mismatch repair enzyme allele is associated with an acquired loss of the second normal allele in random colonic epithelial cells (similar to the way the hereditary form of retinoblastoma works genetically). These cells are then vulnerable to genetic mutations of cancer-associated genes and the development of carcinoma of the colon.
F. GENETIC ALTERATIONS IN NEOPLASIA
The genetic alterations that may result in neoplasia include:
1. Point Mutations of oncogenes resulting in oncogene products which are abnormally active or lack normal regulatory controls such as degradation or feedback inhibition. These mutations may be as subtle as a single base pair alteration or deletion. The ras gene is frequently mutated in this fashion, resulting in a protein that remains active rather than cycling back to an inactive state normally.
2. Amplification of oncogenes. This appears to be a relatively common phenomenon in which the number of gene copies/cell of a given oncogene is increased in the tumour cells. Karyotype analysis of the tumor often reveals either small chromosome fragments called "double minutes" or elongated stretches of a chromosome called "homogeneous staining regions". A number of studies have shown that increasing degrees of gene amplification, for instance of N-myc copy number, correlates with an increasingly worse prognosis in neuroblastomas, a malignant tumour of childhood.
3. Chromosomal translocations. The best known of these is the Philadelphia chromosome which is the result of a t(9;22) translocation. Found in chronic granulocytic leukemia, the translocation results in a "fusion protein" called bcr-abl, which behaves as a constantly active abl oncogene. Lymphomas and leukemias seem to arise preferentially from translocations and a number of these are specific to the sub-type of malignancy involved. Example of these can be found in the lecture notes on Malignant lymphomas.
4. Chromosomal or gene deletions may result in loss of activity of some of the tumour suppressing genes, such as Rb.
The Lecture slides are available as an PDF file
(271 Kb).
