Moss - What genes cant do - 2003 (522929), страница 37
Текст из файла (страница 37)
The experimental systemthey used for this work consisted of an aneuploid2 mouse fibroblast cellline called “NIH 3T3 cells” and a methodology known as transfectionfor introducing fragments of foreign DNA into these cells. Morphological transformation of the 3T3 cells (which generally meant a changefrom a flattened to a more spherical shape), loss of contact inhibition(that is, the ability to continue to divide and replicate even after lateralcontact with other cells has been made), and an ability to grow in softagar (which normal fibroblasts are not capable of doing) became thestandard criteria for identifying a neoplastic phenotype.With the recombinant technology for altering DNA sequence and anassay for in vitro transformation it was possible to explore experimentally the meaning of activation.Much of the experimental work carried out during the 1980s in thequickly expanding field of oncogene research consisted of attempts toclarify the exact nature of activation.
Oncogenes and protooncogenesfrom viral and cellular sources were isolated and subject to manipula-Dialectics of Disorder: Normalization and Pathology as Process147tion. Various sequences were altered and the effects of the alterationwere assayed for carcinogenic capacity through transfection into the NIH3T3 cell line. The dominantly acting oncogene theory of cancer causation turned on these highly publicized studies. In a scathing 1988 review,Henry Harris took the Bishop and Varmus theory severely to task. Hiscriticisms were enumerated as follows (Harris 1988):1. It has been shown that the introduction of an oncogene into NIH 3T3cells or other untransformed cell lines of this type, all of which are aneuploid, produces multiple stable changes in the genome of the recipientcell.
Against this complex background of genetic changes, no conclusionconcerning the dominance or recessivity of the mode of action of theinterpolated oncogene is possible, even when the parameter being studiedis no more than morphological transformation in vitro.2. The great majority of morphologically transformed cells are notmalignant in the sense that they are capable of progressive growth invivo. When the morphologically transformed cells are injected into anappropriate host, it can easily be shown by karyological analysis that thecells capable of progressive growth in vivo (those that generate thetumor) are a highly selected subpopulation.3.
Once malignant cells have been so selected, continued expression ofthe interpolated oncogene is not required for maintenance of the malignant phenotype.4. In the few cases where the question has been specifically examined ingenuine tumors, it has been found that mutated oncogenes are frequentlypresent in the hemizygous condition.5. When malignant cells containing known oncogenes that are activelyexpressed are fused with diploid fibroblasts, malignancy is suppressedwhether or not the oncogene remains active in the hybrid cell.The idea that the NIH 3T3 cell-oncogene transfection system demonstrates that activated proto-oncogenes orchestrate neoplastic transformation in a genetically dominant fashion was attacked by Harris onseveral levels. The cells that appear to be transformed in culture generally do not prove to be tumorigenic in animals.
That only a subsetare tumorigenic in animals suggests that something beyond transfectionwith the oncogene is required. With respect to such cells that are clearly148Chapter 4malignant, continued expression of the oncogene does not appear to berequired, suggesting that the oncogene is not orchestrating the phenotype.
In addition, the aneuploidy of the cell line and the tendency for anytransfection to result in numerous genetic changes makes it meaninglessto speak of genetic dominance because the transformed phenotypecannot be attributed to a single locus, nor can the presence of a wildtype allele be assumed. Finally, even if the genetic characterization of thecells were sufficient to make attributions of genetic dominance meaningful, such an attribution would still be factually inaccurate inasmuchas the transformed phenotype is suppressed in the cell fusionexperiments.An obvious way to have gotten around the genetic ambiguities of theaneuploid cell line would have been to transform ostensibly normal cells.Failure to do so was not based on a lack of effort.
As reported by EricStanbridge:Despite intensive efforts to transform normal human fibroblasts or epithelial cellswith varying combinations of activated cellular oncogenes, the results have beenuniformly negative (Stanbridge 1990).Perhaps the most convincing evidence against the dominant oncogenethesis came from the then surprising results of high-tech transgenicmouse experiments. Mice were “constructed” such as to contain eitherthe myc or ras oncogenes in every cell of their body.
In a characteristictransgenic study an activated myc oncogene had been fused with anMMTV promoter to ensure its expression in the pancreas, lung, brain,salivary gland, and breast of the mouse. Tumors arose only in the murinebreast, only on a clonal basis, and only after a latency period (Weinberg1989). Such results indicated that tumorigenesis had occurred in at mosta few out of millions of cells expressing the oncogene, that it was dependent upon additional lesions occurring and that even this effect waslimited to cells of a specific tissue and differentiation state. Oncogeneresearcher Robert Weinberg felt compelled to conclude thatthe lesson from this is dramatic and clear: A single oncogene like ras or myc isunable to malignantly transform the great majority of cells in which it isexpressed.
Although millions of cells in a tissue remain quite normal, only a fewwill go on to generate a tumor mass. The expression of these single oncogenesmay be necessary for tumorigenesis, but is hardly sufficient (Weinberg 1989).Dialectics of Disorder: Normalization and Pathology as Process149The realization that the expression of a single activated proto-oncogeneis not a sufficient basis for cancer should not have been such a surprisewhen viewed in the light of 50 years of research in experimental carcinogenesis, which had long since ruled against any single-hit model ofcancer.By the end of the 1980s a consensus among molecular oncologistsemerged which reendorsed a multistep model of carcinogenesis, postulating a progressive accumulation of lesions to both oncogenes andtumor suppresser genes, with tumor suppresser genes the more apparantly pervasive.
The use of Mendelian language, however, did not disappear. Attempting to advance a more genetically ecumenical theory ofcarcinogenesis, albeit one in which the distinction between oncogenesand tumor suppresser genes was still held to be of more than just historical interest, Bishop redeployed the terms dominant and recessive withan altered specification:The genetic damage found in cancer cells is of two sorts: dominant, with targetsknown colloquially as proto-oncogenes; and recessive, with targets known variously as tumor suppresser genes, growth suppresser genes, recessive oncogenes,or anti-oncogenes. The dominant damage typically results in a gain of function,whereas the recessive lesions cause loss of function (Bishop 1991).Bishop suggested that the distinction between dominant oncogenes andrecessive tumor suppressor genes pertained to function.
What can thismean? Is this distinction tenable? Where the distinction between the twosets of cancer-implicated genes is least problematic would appear topertain to biochemical activity. Those genes that have been referred toas oncogenes or activated proto-oncogenes have been implicated in carcinogenesis in relation to the biochemical activity of their gene products.Those genes that have been referred to as tumor suppresser genes havebeen implicated in carcinogenesis with respect to the loss of expressionand/or the absence of the biochemical activity of their gene products.
Sowhy not just distinguish these putative cancer-related classes in terms ofbiochemical activity? The ascription of function is different and strongerthan that of biochemical activity. Bishop means to make a stronger claimabout the relationship of genes to carcinogenesis than only a distinctionin terms of biochemical activity would imply. In analyzing the warrantof his stronger claim we can begin to see the research program in150Chapter 4molecular oncology, by its own dynamics, pushing the gene-centeredconcept to its limits.The problem with attempting to distinguish oncogenes from tumorsuppresser genes in terms of the more empirically tractable criterion ofbiochemical activity is that the normal c-oncs, or unactivated protooncogenes, already have biochemical activity.
Where loss of activitycould distinguish normal from abnormal states of the tumor suppressergenes, this distinction does not work for the oncogenes. Nor has a changein activity, that is, between the normal activity of the proto-oncogenesand the abnormal activity of the activated proto-oncogenes, ever beenestablished. Indeed, Bishop himself in 1982, and many investigatorssince, had suggested that there is no qualitative change of activity, onlya quantitative increase of the same activity associated with transformation (Bishop 1982). So in the view of Bishop what the proto-oncogenegains in the course of becoming an activated oncogene is not a newbiochemical activity but the function of causing cancer.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
