Moss - What genes cant do - 2003 (522929), страница 38
Текст из файла (страница 38)
But why speakof the association of the activity, possibly elevated, of some gene productwith carcinogenesis in terms of function? Certainly no manner ofetiological explanation could justify the use of the term function here.Neither the normal proto-oncogene nor its activated state (whatever thatmight be) is present because of the function it has served causing cancerin previous generations.What is really at stake in ascribing function to the activity of the oncogene is an attempt to retain a gene-centered interpretation of carcinogenesis.
The depiction of cancer derived from the experience of the tumorsuppresser research program is a different one. Tumor suppressor genescannot be the motor-force of cancer causation. The positing of their existence does not constitute a genetically directed story of cancer progression. Where a tumor suppressor function or any other function is lostthe execution of carcinogenesis has not thereby been assimilated into agenetic model.
Genetic lesions that result either in the loss of biochemical activity or in an increase of biochemical activity could effect a lossof function. As we will see, it is the loss of functions that pertain to terminal differentiation that appears to lead the way to malignancy. Againthe reflections of Henry Harris will be useful in articulating thisDialectics of Disorder: Normalization and Pathology as Process151alternative framework in which the dominant-recessive distinction is nolonger meaningful.My present ideas on this subject have their origins in a histological observationmade by Stanbridge and Ceredig (1981).
They found that when cells of a linederived from a human carcinoma were fused with normal human fibroblasts,hybrids in which malignancy was suppressed acquired a different morphology invivo from segregants in which the malignant phenotype had reappeared as a consequence of chromosome loss.
The malignant segregants grew progressively asundifferentiated epithelial tumors; but the hybrid cells in which malignancy wassuppressed assumed an increasingly elongated, shape and gradually ceased tomultiply. Now this is just what normal fibroblasts do in, for example, a healingwound. The cells are at first induced to multiply, but as they secrete and organize their characteristic collagenous extracellular matrix, they gradually elongateand stop multiplying. . . .
The observations of Stanbridge and Ceredig thus suggested that the hybrids in which malignancy was suppressed were executing, atleast in part, the differentiation programme of a normal fibroblast, whereas themalignant segregants were not.Essentially the same result was obtained when tumour cells were fused withnormal keratinocytes. Terminal differentiation in the keratinocytes involves thesynthesis of the protein involucrin, which is cross-linked by a keratincoytespecific transglutaminase to form an insoluable envelope.
. . . When hybridsbetween normal human keratinocytes and cells of a line derived from a humancarcinoma were examined, it was found that those in which malignancy was suppressed continued to synthesize involucrin, but malignant segregants did not.When injected into the animal the non-malignant hybrids showed the characteristic histological features of terminally differentiated keratinocytes and ceasedto multiply, whereas the malignant segregants continued to multiply as undifferentiated epithelial tumors. It appears that once again the suppression of themalignancy in these cases involves the imposition on the hybrid cells of the terminal differentiation programme of the normal cell with which the tumour cellis fused (Harris 1990).In the explanatory model suggested by these findings, cells stop dividing when they enter into a pathway of terminal differentiation.
In thisview the ability to continue to divide, as is the case in tumorigenesis,does not require the acquisition of new functions but rather is on theorder of a default condition of the cell—of any cell—when its progression along a path of differentiation has been impeded. Understood thisway, the significance of either oncogene lesions resulting in an increaseof some biochemical activity or tumor suppresser gene lesions resultingin the loss of some biochemical activity is that in either case it is a disruption of terminal differentiation, that is, a loss of function, which is152Chapter 4associated with the onset of tumorigenesis. Along with abandoning theidea that oncogenes are characterized by a gain in function goes thespecifically gene-centered notion that certain dominantly acting allelesdetermine the neoplastic phenotype.Following the idea that some impediment to the pattern of differentiation which is specific to the tissue and stage of differentiation of a cellis the common feature of both oncogene and tumor suppressor genelesions in the causation of cancer, Harris “invented” a genetic lesion totest his hypothesis.
The extracellular matrix molecule fibronectin, whichis secreted by fibroblasts, has been shown to provide a mediating linkbetween fibroblasts and other components of the extracellular environment within tissue. Proper attachment of fibroblasts to the extracellularmatrix is believed to be a critical step in fibroblast differentiation.
Harristransfected nontumorigenic hybrid cells derived from tumor and normalfibroblast fusions with sections of the fibronectin gene in an “anti-sense”4configuration. The anticipated result of such a procedure is that ofinhibiting the synthesis of fibronectin. The anti-sense fibronectin RNAtranscript, which is complementary to the normal transcript, pairs withit to form an inactive RNA duplex, and in so doing it blocks translationand synthesis of the fibronectin protein.
Of five clones with fibronectinproduction abolished or greatly reduced, “in four of these, the malignantphenotype had reappeared and the hybrid cells again produced rapidlyprogressive tumours when injected into appropriate animals” (Harris1990).If one can imagine the fibronectin anti-sense gene arising naturally asa mutation in a cell one can see the fallacy of the Bishop dominantrecessive gain in function–loss of function distinction. The pattern ofinvolvement in transformation of this gene would follow that of theoncogene and not that of the tumor suppressor gene.
Its tumorigeniceffects would be associated with its transcriptional activity; the moreactivity, the stronger the tumorigenic effects. By the Bishop distinctionthe anti-sense gene would constitute a dominantly acting oncogene witha gain of function. Yet the mode of action of this mutation would be toblock the synthesis of fibronectin and thereby impede progress towardterminal fibroblast differentiation. The production of fibronectin antisense gene activity in fact represents a loss of function.Dialectics of Disorder: Normalization and Pathology as Process153The example of the anti-sense fibronectin gene provides a perspicuousoccasion for revisiting the Gene-P/Gene-D distinction in the context ofcarcinogenesis. If we imagine that a fibronectin anti-sense allele5 cameabout somehow spontaneously, it would be quite plausible to imaginethat its expression could correlate with some elevated incidence of tissuespecific, (fibroblastic) tumor formation at certain developmental stages.Our fibronectin anti-sense gene would then be a proper Gene-P for acertain sarcoma (fibroblastic tumor).
If it were in the germ-line, it wouldindeed follow a Mendelian pattern of inheritance. As a Gene-P (whichcould be identified with a molecular probe), it could do what a Gene-Pcan do. It could pragmatically serve as a predictor of an elevated cancerrisk at certain times and under certain conditions. Where conflactionaryconfusion arose (and this is exactly what Bishop’s analysis entails), itwould be treated as a specific molecule (Gene-D) which orchestrates(programs, dictates, determines, controls, effects, and so forth) thetumorigenic phenotype. The capacity and proclivity for fibroblasts toundergo abnormal growth in the absence of secreted fibrnonectin—which would be restricted to certain tissue-developmental and microenvironmental contexts—would not be explained, let alone programmed(dictated, executed), by the absence of the fibronectin.
An unconflatedGene-P (that did not confuse prediction with function) would notmasquerade as a substitute for understanding the actual biology of neoplastic growth.Despite extensive efforts, no mechanistic pathway has been identifiedto account for the means by which activated proto-oncogenes productsfunction to transform cells. Most likely, as in the case of the anti-sensefibronectin gene (but with less specificity), the activated proto-oncogeneproducts act to disturb processes of differentiation in certain cells atcertain times, resulting in a loss of the ability of these cells to obtain orretain a state of terminal differentiation.Further examination of the nature of cancer-related genes lendssupport to Harris’s theory.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
