I. The Basics A. “Transformation” from a normal to a cancerous state occurs at the level of individual cells… metastatic cancer cells resemble the primary cancer. cancers are usually clonal in origin (X-inactivation studies in females; unique Ig or TCR rearrangements in lymphoid cancers). single cell transformation can be observed in culture systems (decreased adherence, anchorage independence for growth, loss of contact inhibition, decreased growth factor requirement, increased nutrient uptake and membrane ruffling).
Implications: Cancer arises at the level of the single cell and therefore must be understood at that level. This is different than other common diseases such as hypertension or diabetes that are due to “systemic” perturbations in physiology. B. Cancer is primarily a genetic disease… somatic mutations occur in common “sporadic” (non-familial) cancers. inherited germline mutations occur in rare familial cancer syndromes.
increases in the mutation rate or genomic instability correlate with increased risk of cancer. selection for mutations in cancer occur at the level of the single cell, not at the level of organismal survival. C. Cells must make critical decisions in a multi-cellular organism… stem cell renewal; growth / quiescence; differentiation; cell death. adult humans have a remarkable homeostasis of cell number.
communication between different cells is critical for this homeostasis. failure at any of these decision points can lead to overgrowth D. The good, the bad, and the ugly… tumor suppressor genes prevent transformation [good]. oncogenes cause transformation [bad]. loss of genomic integrity causes mutations in both [ugly] II.
Oncogenes A. Oncogenes are dominantly acting agents of cellular transformation. Note that a small number of different oncogenes explains a very large number of different types of cancer. This is a great simplification of the problem. B. Independent lines of cancer research identify the same set of oncogenes.
1. Acutely transforming retroviruses. why study tumor viruses? rapid and reliable oncogenesis can be readily quantit ated genetically simple: only 4 genes instead of 60, 000. retroviral oncogenes required for transformation but not viral growth.
retroviral oncogenes (v-onc) arise from highly conserved cellular proto-oncogenes (c-onc) [examples: v-myc in avian myelocytomatosis virus; v-ras in rat sarcoma virus; v-src in Rous sarcoma virus] Note that Rous sarcoma virus (RSV) is unusual because it has all three essential viral genes (gag, pol, env) and also contains a viral oncogene of cellular origin (src). Some transformation defective (td) mutants of RSV have completely lost the src gene but still replicate. All other acutely transforming retroviruses are replication defective because essential viral genes have been replaced by an oncogene (e. g. MC 29). Such defective viruses need a non-defective “helper” virus like AL to provide essential replication functions.
2. Retroviruses without oncogenes. retroviruses are transposon’s which must insert into host DNA. initial random integration sites = > clonal selection by cell growth = > specific common integration sites in tumor cells. insertion al mutagenesis activates an adjacent proto-oncogene [example: activation of c-myc in avian bursal (B-cell) lymphoma]. acutely transforming retroviruses arise by rare recombination of integrated virus with adjacent c-onc Note that the viral oncogene is an intron-less cDNA copy of the original cellular proto-oncogene.
This is because reverse transcriptase copies the already spliced RNA back into DNA. Acutely Transforming Long Latency Retroviruses Frequency of Tumors ~ every animal occasional animal Latency of Tumors very short (weeks) long (months – years) Clonal ity of Tumors polyclonal mono / oligo clonal Viral Oncogene yes no Transformation in Culture yes no Integration Sites in Tumors random common; near proto-oncogene 3. Gene amplification. homogeneously staining regions (Hrs) and double minute chromosomes (DMs) are common in cancer cells. these cytogenetic abnormalities represent gene amplification. amplified regions often contain proto-oncogenes [example: N-MYC in neuroblastoma’s] 4.
Chromosomal rearrangement. most cancer cells have chromosomal abnormalities. specific types of cancer often have consistent translocations. these translocations activate proto-oncogenes at or near breakpoints 5.
Transfection of human tumor DNA. “immortalized” rodent fibroblasts in culture can be transformed to a cancerous phenotype by introduction of human tumor DNA. such transformation requires a single dominantly acting oncogene. activated RAS oncogenes are found in many human tumors [example: 50% of human colon cancers contain mutated N-RAS]. caveats: assay appears biased towards RAS; genes can be “activated” by rearrangement during transfection rather than by mutation in the original tumor; truly normal cells need more than one genetic change 6.
Chemical carcinogenesis. induction of specific cancers in rodents results in consistent mutations in the same proto-oncogene [example: specific point mutation in RAS in NMU-induced mammary carcinomas in rats]. different chemical carcinogens cause cancers with consistent mutations in specific oncogenes [ = > useful for epidemiology of cancer? ] 7. Rare familial cancer syndromes. dominant inheritance.
germline mutation of proven oncogene [example: activating mutations of RET oncogene in human multiple endocrine neoplasia type 2 and familial medullary thyroid carcinoma]. note: most familial cancer syndromes are caused by germline inactivation of a tumor suppressor gene (see below) rather than by germline activation of an oncogene C. Protein products of proto-oncogenes are essential for normal growth and differentiation and are conserved in evolution. 1. Key components of various signaling pathways [RAS] 2. Homologs of genes which determine cell identity [HOX] 3.
Inhibitors of cell death [BCL 2] D. Mechanisms of proto-oncogene activation 1. Normal protein at wrong level, place, or time [example: activation of MYC by chromosomal translocations that cause transcriptional deregulation] 2. Mutated protein [example: RAS mutations cause specific amino acid substitutions] In response to an upstream signal, the inactive GDP-RAS undergoes exchange of GTP for GDP. The active GTP-RAS then sends a signal downstream. However, the intrinsic GTPase activity of the RAS protein limits the duration of this signal by converting the active GTP-RAS to an inactive GDP-RAS.
The most common oncogenic mutations in RAS greatly decrease the intrinsic GTPase activity. This results in a prolonged and inappropriate signal output. 3. Abnormal fusion protein after chromosomal translocation [example: BCR-ABL in chronic myelogenous leukemia] 4. Activation of normal cellular proteins by viral proteins [examples: papilloma virus E 5 protein activates growth factor receptors in the absence of growth factors] III.
Tumor Suppressor Genes (Anti-Oncogenes) A. Independent lines of research identify genes which prevent transformation. 1. Cell fusion experiments.
transformation is recessive in most fusions of normal and cancer cells. somatic cell hybrids identify specific chromosomal locations of tumor suppressors = > genetic cause 2. Familial cancer syndromes identify genes that cause “sporadic” cancer. a) Retinoblastoma as paradigm.
inherited form occurs earlier and in both eyes. sporadic form occurs later and usually in only one eye. Knudson’s hypothesis (two “hits” are required; familial cases inherit the first “hit” as a germline mutation). deletions of chromosome 13 correlate with disease.
re introduction of Rb protein reverts transformed phenotype b) Neurofibromatosis (NF-1 is a negative regulator of RAS) The intrinsic GTPase activity of RAS is normally increased by GAP (GTPase activating protein) activity. Loss of GAP function due to NF-1 mutations therefore causes prolonged signal output from RAS even without oncogenic mutations in RAS. This results in multiple benign tumors which only rarely (in ~3% of patients) become malignant. 3. Loss of heterozygosity (LOH) in tumor cells. tumor susceptibility genes usually show reduction to homo- or hemizygosity in cancer cells in familial cancer syndromes.
consistent loss of heterozygosity at a specific locus in sporadic, non-familial cancers suggests the presence of a tumor suppressor gene = > major gene-hunting strategy in non-familial cancers 4. DNA tumor viruses. some DNA viruses cause cancer in “non-permissive” and some natural hosts [example: SV 40, adenovirus, human papilloma viruses]. these viruses “jump start” quiescent cells to permit viral replication.
viral oncogenes are required both for the viral life cycle and for malignant transformation (note: unlike retroviral oncogenes, these genes do not appear to be of recent cellular origin). the protein products of many of these viral genes often bind to and functionally inactivate cellular tumor suppressor proteins B. Normal functions of tumor suppressor proteins 1. Negative regulators of signalling pathways.
NF-1 negatively regulates RAS proteins (see above) 2. Negative regulators of cell cycle progression. Rb inhibits G 1/S progression 3. Positive regulators of adherence.
APC functions in cell-cell adhesion 4. Components of DNA damage sensors and repair pathways. p 53 prevents cell division in the presence of DNA damage IV. Putting It All Together The “clock” represents a typical cell cycle. The pointed arrows represent activation. The blocked arrows represent inhibition.
Note that although the action of cyclin D directly inhibits the immediate downstream component (Rb), this downstream component in turn negatively regulates the cycle. Therefore, cyclin D is an activator in the overall scheme of the cell cycle (remember, a “double negative” equals a “positive”). Growth factors, their receptors, the RAS pathway, and cyclin D all act positively to promote the cell cycle. Activated forms of these components can function as dominant oncogenes. NF-1, Rb, and p 53 all act negatively to inhibit the cell cycle if everything isn’t OK. These components function as tumor suppressors (STOP signs).
Their absence contributes to cancer. V. Development of Cancer is a Multi-Step Process Studies of human populations imply that multiple “hits” are required for most cancers to develop. Multiple exposures to carcinogens increase the incidence and decrease the latency of onset of the disease. Studies of progression of specific types of human cancer imply that these multiple “hits” represent different mutations that accumulate over time. The order of accumulation of these mutations in a particular type of cancer is generally similar, but not invariant.
Additional reading (reviews): . Cooper, G. M. , Oncogenes, 2 nd Ed.
, Jones and Bartlett Publishers, Sudbury, MA, 1995… Bishop, J. M. “Cancer: the rise of the genetic paradigm.” Genes and Development, 9: 1309-15 (1995). A vivid overview of the history and major concepts of oncogenes and tumor suppressor genes… Kitzler and Vogel stein, “Lessons from Hereditary Colon Cancer”, Cell 87: 159-170 (1996).
A review of the evidence that multiple genetic changes are required for human cancer, focusing on colon cancer as a model… Varmus, H. E. Nobel lecture. “Retroviruses and oncogenes. I.” Bioscience Reports, 10: 413-30 (1990).
A personal account of the work that led to the discovery of cellular proto-oncogenes and a Nobel Prize for Bishop and Varmus. Additional reading (original articles): . Stehelin, D. , et al.
, “DNA related to the transforming gene (s) of avian sarcoma viruses is present in normal avian DNA.” Nature 260: 170-3 (1976)… Cavenee, W. K. , et al. “Expression of recessive alleles by chromosomal mechanisms in retinoblastoma.” Nature 305: 779-84 (1983)… Ionov, et al.
, “Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis.” Nature 363: 558-61 (1993). Additional reading (original articles): . Stehelin, D. , et al. , “DNA related to the transforming gene (s) of avian sarcoma viruses is present in normal avian DNA.” Nature 260: 170-3 (1976)… Cavenee, W.
K. , et al. “Expression of recessive alleles by chromosomal mechanisms in retinoblastoma.” Nature 305: 779-84 (1983)… Ionov, et al. , “Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis.” Nature 363: 558-61 (1993)..