Early detection of malignancy
What is known about growth factors, oncogenes, and tumor suppressor genes in the evolution of Barrett's dysplasia and carcinoma?
J.J.B. van Lanschot, W. Polkowski, H. Obertop, G.J.A. Offerhaus (Amsterdam)
The development of a cancer generally requires several molecular genetic alterations occurring in subsequent generations of cells. This was first suggested from epidemiologic data on the incidence of cancer as a function of age. For most types of cancer the chance of developing cancer increases very steeply with age. It has been estimated, that an accumulation of five to seven genetic events are required to turn a normal cell into a malignant cell . The best established model in this regard is the tumorigenesis that occurs during the adenoma-carcinoma sequence in the colorectum. The typical sequence of genetic mutations underlying the development of a colorectal carcinoma has now been unraveled, although these mutations do not always occur in the same sequence, nor are they the only route to this disease.
Cancer is caused by an accumulation of molecular genetic alterations, characterized by an increasing dysregulation of cell proliferation. Under normal circumstances cell proliferation is influenced by external growth stimuli which are transduced to the nucleus through various intermediate steps. The transduction of external growth stimuli to the nucleus can be disturbed at different levels, which are schematically depicted in Figure 1: growth factors (I), growth factor-receptors (II), cytoplasmatic mediators (III), and nuclear proteins (IV). In general, there are two mutational routes toward uncontrolled cell proliferation. The first is to make a normal proto-oncogene hyperactive by mutation into an oncogene. The second is to make an inhibitory tumor suppressor gene inactive. Activation of dominantly acting oncogenes and inhibition of recessive tumor suppressor genes lead to aberrant growth and thereby clonal expansion. A generalized genetic instability, e.g. through mismatch repair gene defects, may underlie an accumulation of molecular genetic alterations [2, 3].
The dysplasia-carcinoma sequence in Barrett's mucosa has many features in common with the carcinogenesis in the colorectum, but it is as yet less well-defined. The most important factors, that have been recently studied in the carcinogenesis of Barrett cancer will be briefly reviewed.
Growth factors and growth factor-receptors
Cells displaying appropriate receptors normally divide only when they are stimulated by growth factors. These positive signals act by overriding intracellular negative control systems that otherwise restrain from growth and block the cell-cycle. Most growth factors originate from cells in the neighbourhood of the affected cell and act as local mediators. Neighbouring cells compete for minute quantities of growth factors, and therefore the amount of growth factors available prevents cells from proliferating beyond a certain population density . Epidermal growth factor (EGF) and transforming growth factor alpha (TGFa) are two factors which have been investigated extensively and are structurally and functionally related. Both EGF and TGFa have been demonstrated in Barrett's mucosa, indicating the possibility of auto-/paracrine growth regulation [4, 5]. Moreover, a stepwise increase was found in TGFa from fundic-type Barrett's mucosa through intestinal-type Barrett's mucosa to Barrett's cancer . However, a proportion of poorly differentiated tumors failed to show any TGFa immunoreactivity .
Figure 1. Simplified diagram of the transduction of external growth stimuli to the nucleus, leading to increased cell proliferation.
EGF is a ligand that binds to a specific membrane receptor, the so-called epidermal growth factor receptor (EGFR). Both amplification of the EGFR-gene and overexpression of the receptor itself have been demonstrated in Barrett's mucosa and cancer [4, 6, 7]. However, association between the degree of EGFR-gene amplification and the intensity of EGFR overexpression could not be observed .
Cancer cells proliferate excessively mainly because, as a result of accumulated molecular genetic alterations, they are able to divide without stimulation from other cells and therefore are no longer subject to the normal "social" controls on cell proliferation. Any mutation that results in the production of an abnormally active protein, could promote cancer by encouraging the cell to proliferate in the absence of the appropriate extracellular signals. A mutation of a proto-oncogene into an oncogene, that causes its product to be overexpressed or hyperactive, results in excessive cell proliferation. The c-erb B-2 oncogene encodes a truncated form of EGFR that has a cytoplasmatic domain which is continuously active. Cells expressing this oncogene behave as though they were constantly being signaled to proliferate, even if no EGF is present . The prevalence of c-erb B-2 overexpression in Barrett's cancer varies between 11% and 60% [7-10]. Patients with c-erb B-2 positive tumors appeared to have a significantly poorer prognosis than patients with negative tumors [9, 10]. Hardwick et al. were able to show membranous c-erb B-2 overexpression in 8 out of 31 (26%) adenocarcinomas but in none of 27 dysplastic areas and in none of 20 non-dysplastic areas, suggesting that it is a late event in the dysplasia-carcinoma sequence .
Growth factor-receptors on the cell surface have an extracellular part, to which the ligand binds; a transmembranous part, which lies in the cell membrane; and an intracellular part, which transfers the signal into the cytoplasm. Activation of a receptor leads to a cascade-like phosphorylation of various cytoplasmatic mediators (kinases such as abl-, ras-, and src-proteins), which transduce the signal from the cell membrane into the nucleus.
Of the various cytoplasmatic mediators the ras-proteins have drawn much attention. Three different human ras-oncogenes have been detected, the K-, H-, and N-ras genes. Point mutations, which result in altered aminoacids in the codons 12, 13, or 61 transform the ras-protooncogenes into oncogenes. In the tumorigenesis of both colonic and pancreatic cancer the K-ras-oncogene appears to play an important role. In 90% of the pancreatic cancers a point mutation of codon 12 of this gene can be detected. However, in the tumorigenesis of Barrett's cancer, K-ras does not seem to play an important role. Analysis of 46 resection specimens with various grades of dysplasia and/or invasive cancer did not show any K-ras gene mutations . On the other hand, H-ras was consistently expressed in higher grades of dysplasia and carcinoma, but not in low grade dysplasia or nondysplastic Barrett's mucosa .
The src-protein is another cytoplasmatic mediator, which is known to play a role in the carcinogenesis of the colorectum. Src-activation was also studied in endoscopic tissue samples of Barrett's mucosa and cancer . Src-activity was 3-4 fold higher in nondysplastic Barrett's mucosa than in normal esophagus or duodenum, and 6- fold higher in esophageal adenocarcinoma than in adjacent nondysplastic mucosa. These data suggest, that activation of src may be an early event that occurs in Barrett's mucosa prior to the development of esophageal adenocarcinoma.
When the growth factor/receptor complex has activated the cytoplasmatic phosphorylation cascade, the expression of two classes of genes is induced in the nucleus: the early-response genes and the delayed-response genes. It seems that the delayed-response genes are induced by the products of the early-response genes. A well-studied early-response gene is the myc-protooncogene, which encodes a gene regulatory protein in the nucleus. When overexpressed by mutation, it prevents cells from entering the Go-resting phase and thereby can cause uncontrolled proliferation.
The role of myc-overexpression in the tumorigenesis of Barrett cancer was studied in
12 patients . Expression of myc could not be detected in non-dysplastic mucosa, but myc-overexpression of approximately equal intensity was consistently observed in all grades of dysplasia and in carcinoma.
Two or more specific oncogenes can collaborate synergistically to make cells cancerous. Oncogene collaboration has been described for the myc-gene and the bcl-2 gene. If myc alone is overexpressed, cells are driven round the division cycle inappropriately, but no cancer results because the progeny of such abnormally forced divisions are programmed to die. This programmed cell death (apoptosis) is characterized by typical morphological changes, in which the cells and their nuclei shrink and condense and are rapidly and efficiently phagocytosed. The bcl-2 gene encodes a protein which suppresses apoptosis. If the bcl-2 gene is overexpressed together with the myc-gene, the excess progeny cells, which are induced by the myc-protein, will now survive .
Unlike most oncogenes, rather than increasing the rate of cellular proliferation, the bcl-2 protein decreases the rate of cell death by preventing apoptosis. The bcl-2 protein has been found to be abnormally overexpressed as an early event in the dysplasia-carcinoma sequence of gastric cancer. However, in 36 esophageal resection specimens containing Barrett's mucosa, no immunoreactivity was seen in any of the cases with or without dysplasia or carcinoma. Apparently, in contrast to its role in gastric cancer, bcl-2 alteration is not an important event in the neoplastic progression of Barrett's mucosa .
The essential processes of DNA-replication and mitosis are governed by the so-called cell-cycle control system. Traditionally, the standard cell-cycle is subdivided in G1, S, G2, and M. The cell-cycle control system uses brakes, that can stop the cycle at specific checkpoints in G1 and G2. Mammalian cells have a safety mechanism to restrain them from entering the S-phase with damaged DNA. This mechanism depends on a protein called p53, which specifically binds to DNA. It accumulates in the cell in response to DNA damage and halts the cell-cycle control system at the G1 checkpoint to allow for DNA-repair . Alterations of the p53 tumor suppressor gene are the most common genetic lesion in human cancer. More than half of all humans dying of cancer do not do so unless their tumor evolves a p53 alteration . Mutations of the p53 gene result in the production of a mutant peptide, which is deprived of its normal regulatory function of suppression of cell turnover. It is characterized by a much longer half-life time and can therefore be visualized by immunohistochemical methods. Several studies have been performed, indicating an increasing positivity of the mutant p53 protein with increasing grade of dysplasia in Barrett's mucosa (Table I) [8, 17-22].
Cancer is characterized by a dysregulation of cell proliferation. An accumulation of multiple genetic alterations is needed to cause an invasive cancer. For the Barrett's metaplasia-dysplasia-carcinoma progression it has been shown, that activation of specific oncogenes (esp. EGFR, c-erb B-2, src, H-ras, and myc) and inhibition of specific tumor suppressor genes (esp. p53) play an important role. Various other oncogenes (esp. K-ras and bcl-2), while crucial in the tumorigenesis at different sites of the gastrointestinal tract, do not seem to underlie the malignant degeneration of Barrett's mucosa. The typical sequence of genetic alterations is, as yet, largely unknown. The presently available knowledge is too fragmentary to allow clinical application for early tumor detection.
5. Brito MJ, Filipe MI, Linehan J, Jankowski J. Association of transforming growth factor alfa (TGFa) and its precursors with malignant change in Barrett's epithelium: biological and clinical variables. Int J Cancer 1995;60:27-32.
10. Nakamura T, Nekarda H, Hölscher AH, Bollschweiller E, Horbec N, Becker K, Siewert JR. Prognostic value of DNA ploidy and c-erbB-2 oncoprotein overexpression in adenocarcinoma of Barrett's esophagus. Cancer 1994;73:1785-1794.
14. Kumble S, Cartwright CA, Omary MB, Triadafilopoulos G. Src activation in malignant and premalignant epithelia of Barrett's esophagus. Proceedings of American Gastroenterology Association, San Francisco 1996:A 868.
17. Ramel S, Reid BJ, Sanchez CA, Blount PL, Levine DS, Neshat K, Haggitt RC, Dean PJ, Thor K, Rabinovitch PS. Evaluation of p53 protein expression in Barrett's esophagus by two-parameter flow cytometry. Gastroenterology 1992;102:1220-1228.
20. Casson AG, Manolopoulos B, Troster M, Kerkvliet N, O'Malley F, Inculet R, Finley R, Roth JA. Clinical implications of p53 gene mutation in the progression of Barrett's epithelium to invasive esophageal cancer. Am J Surg 1994;167:52-57.
21. Hardwick RH, Shepherd NA, Moorghen M, Newcomb PV, Alderson D. Adenocarcinoma arising in Barrett's oesophagus: evidence for the participation of p53 dysfunction in the dysplasia/carcinoma sequence. Gut 1994;35:764-768.
22. Polkowski W, van Lanschot JJB, ten Kate FJW, Baak JPA, Tytgat GNJ, Obertop H, Voorn WJ, Offerhaus GJA. The value of p53 and Ki67 as markers for tumour progression in the Barrett's dysplasia-carcinoma sequence. Surg Oncol 1995;4:163-171.