Moss - What genes cant do - 2003 (522929), страница 36
Текст из файла (страница 36)
The concept of tumor suppresser genes originated from twoindependent lines of work, cytogenetic studies on hybrid cells derivedfrom somatic cell fusion experiments and epidemiological studies onchildhood cancers. In both cases evidence accumulated in support of theidea that the genetic lesions associated with the occurrence of cancerwere recessive in nature. Cancer was the result of the loss of both allelesat a certain locus, and this locus was in some manner associated withconstraining cells from entering into unregulated growth.The idea of the existence of tumor suppressor genes arose, in part,from experiments that involved the fusion of two distinct cells and thusfollowed from a technical breakthrough.
The first evidence of the possibility of viable cellular fusion involving multinucleate cells was made inrelation to the observation of pathological tissue by Johannes Müller(Faber 1893). However, the first experimental demonstration of viablecell fusion was not achieved until 1965 (Harris & Watkins 1965). In analyzing the implications of their achievment, the following observationswere made (Harris 1970).Dialectics of Disorder: Normalization and Pathology as Process1431. That an inactivated virus could be used to provide a general methodfor fusing animal cells together under controlled conditions.2.
That fusion could be induced between cells from widely differentspecies.3. That the fused cells were viable.One of the many intriguing avenues of inquiry which this experimentalbreakthrough enabled was the possibility of fusing normal with malignant cells or transformed cells with different degrees of malignant potential, and then observing the results.
In early studies in which cells ofgreater and lesser malignant potential were fused together and the malignant potential of the resulting hybrids were evaluated, a retention of themore highly malignant phenotype was observed. These results could havelent credence to an interpretation of the cancerous phenotype as anexpression of genetic dominance. The results however were deemed tobe inconclusive.In none of these cases, however, were the chromosomes of the tumors analyzedin detail, so that no assessment can be made of the extent to which the resultsmight have been complicated by loss of chromosomes or by selection of atypical variants in vivo (Harris, Miller et al.
1969).What soon became apparent was that the tetraploid (fours sets of chromosomes) fusion products of cells derived from different species, andwith less frequency those derived from animals of the same species havea marked tendency to lose certain chromosomes. Once apprised of thisfactor the question could be experimentally posed again.
The results weresurprising. Fusing a variety of tumorigenic mouse cells with cells of anontumorigenic mouse cell line, Harris and coworkers found that themalignant phenotype was suppressed in all cases. In addition theyobserved “that hybrids resulting from such fusions produce segregantsin which a loss of chromosomes is associated with a reversion to malignancy” (Harris, Miller et al. 1969). The implication of these experimentsappeared to be clear. The malignant phenotype is enabled only when allof the chromosomes which carry some factor or factors are lost.The ability to suppress a malignant phenotype proved to be a widespread phenomenon of somatic cell fusion experiments using both rodentand human fusion partners.
The proclivity of tetraploid hybrids to drop144Chapter 4chromosomes provided the opportunity to identify the particular chromosomes required for suppressing a malignant phenotype. Where thespontaneous loss of a certain chromosome correlated with the expression of the malignant phenotype, it was surmised that that chromosome,when present, suppressed that malignant phenotype. Mouse chromosome 4, for example, was found to be responsible for suppressing a widerange of tumor cells including those expressing retroviral oncogenes(Harris 1988).
Human chromosome 1, which was shown to have significant homology with mouse chromosome 4, was likewise seen to becapable of suppressing a wide range of human tumor cell phenotypes.Studies on human chromosome 11 found it to be associated with thesuppression of malignancy in fusion experiments consisting of humanuterine carcinoma and a normal human fibroblast (Harris 1988). Furthertechnical advances enabled researchers to demonstrate directly thatsingle copies of chromosome 11, when delivered by the new method ofmicrocell transfer, were able to suppress the malignant phenotype ofeither uterine carcinoma or of Wilms tumor cells (Harris 1988).
Similarmethods localized putative tumor suppresser capabilities to nine different chromosomes (Levine 1993). A next step in the logic of this researchprogram was to locate and identify the specific tumor suppressor genessituated on these chromosomes.Somatic cell hybridization studies were one source of the tumor suppressor gene model; epidemiological studies on the childhood cancerretinoblastoma was another. Retinoblastoma arises from cells of theembryonal neural retina and occurs only in young children. In mostcases, retinoblastoma arises sporadically (not along family lines) withan incidence of approximately 1 : 20,000, yet in about one-third ofcases overall the tumor did appear to follow some heritable pattern(Stanbridge 1990). A. J. Knudson in 1971 proposed a “two-hit” modelin order to account for this split between heritable and sporadic formsof the cancer. Consistent with the idea of a recessive pattern, Knudsonsuggested that the same locus was involved in both sporadic and familial forms of the disease but that in the familial form one of the alleleswas already mutated in the germ line.
A single somatic mutation involving the homologous locus in the unaffected chromosome would then beenough to generate the tumor in the predisposed individual, whereas insporadic retinoblastoma two mutations must occur somatically in theDialectics of Disorder: Normalization and Pathology as Process145same retinal precursor cell. Cytogenetic studies corroborated this theoryand located the site of the retinoblastoma lesion to chromosome 13, bandq14 (Stanbridge 1990).A third line of research which led in the direction of associating carcinogenesis with the loss of both alleles at some locus was that whichsought to correlate tumorigenesis with the loss of chromosomal heterozygosity. In the inherited forms of retinoblastoma, the somatic eventthat “knocks out” the other allele, thus uncovering the previouslyrecessive germ line mutation, is usually a chromosomal aberration—achromosome loss, deletion, mitotic recombination, or gene conversion(Marshall 1991).
Such events result in the absence of genetic materialinherited from one of the parents and can be detected by the loss of heterozygosity for chromosomal markers flanking the locus. Since similarevents uncovering recessive somatic mutations occur in the sporadicforms of retinoblastoma, the consistent loss of heterozygosity in tumorscan be used as an indication of the presence of a tumor suppresser gene.A putative tumor suppresser gene called p53 was detected on the basisof the correlation of the loss of heterozygosity on the short arm of humanchromosome 17 with the occurrence of breast cancer, small cell lungcancer, astrocytomas, and colon cancer (Marshall 1991). This gene hasthus far emerged by some margin as the most frequent site at whichgenetic alterations are associated with human oncogenesis, far surpassing the frequency of correlations found with any of the retrovirally determined proto-oncogenes.
In addition to P53 and the retinoblastomarelated gene (located on chromosome 13q14), loss of heterozygositystudies indicated the presence of tumor suppressor genes as follows: theWilm’s tumor (WT-1) gene located on chromosome 11p13, the adenomatous polyposis (APC) gene located on chromosome 5q21, and thedeleted in colorectal carcinoma (DCC) gene located on chromosome18q21 (Levine 1993).Cancer Genes: Dominant, Recessive, or None of the Above?Bishop and Varmus had theorized that the retroviral findings pointedthe way to a set of normal growth-related genes, the c-oncs or protooncogenes, which become the cause of sporadic cancer when subjectto some form of somatic mutation, which became referred to as146Chapter 4“activated.” A single such event—the activation of any of these protooncogenes—would then determine the cancer phenotype and therebycount as a dominant allele.
By such reasoning, when a proto-oncogenebecomes activated, it acquires a new function by means of which itorchestrates the transformation to a malignant phenotype. In this theactivated proto-oncogene resembles the virogene-oncogene of theHuebner and Todaro hypothesis except that it is deemed to derive itspower from its location within an enzymatic control circuit which regulates cell growth. Cell fusion experiments performed by Harris et al.,however, suggested that tumorigenesis can be stopped by the presence ofa normal chromosome, implying that in order for carcinogenesis toensure, it is necessary (regardless what mutational activations may be promoting cancer) that certain (tumor suppressor) genes be entirely lacking.The retinoblastoma and p53 studies appeared to indicate that the lack ofcertain alleles alone may be sufficient to result in certain kinds of cancer.In order to support their contention that certain single genes, theproto-oncogenes, could become dominantly acting causes of cancer uponactivation, Bishop and Varmus endeavored to demonstrate that protooncogenes could be experimentally altered (activated) and made to becapable of inducing neoplastic transformation.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
