[CANCER RESEARCH 39, 2390-2399, July 19791 studies, decreased cytoplasmic spreading on substrata was characteristic of neoplastic rodent embryo fibroblast lines com pared with normal cell lines (5, 7, 10, 11). Transformation related changes in cell shape have also been identified by scanning electron microscopy in rodent liver cells (2, 15) and embryonic fibroblasts (4, 7, 17, 30). In chick embryo fibroblasts infected with a temperature-sensitive Rous sarcoma virus, changes in cell shape occurred rapidly after a shift to the permissive temperature (3, 16). In addition, scanning electron microscopic studies of normal and transformed fibroblasts have provided evidence for in creased surface activity in transformed cells, in the form of prominent microvilli (4, 17, 23) or ruffles (3). In comparisons of normal and oncogenically transformed rat liver cells, the latter also exhibited increased degrees of surface activity (2, 15). However, a recent comparison of tumorigenic and nontumori genic cell lines derived from a single clone of mouse embryo cells showed no consistent alterations in surface features, although greater variability of surface morphology was typical of the tumorigenic lines (30). In a similar study of tumorigenic and nontumorigenic rat fibroblasts in vitro, no universal pattern of surface morphology was found to be associated with tumor igenicity (6). In general, disturbances of cell shape appeared to be more characteristic of oncogenically transformed cells than were changes in surface architecture. It is important to determine whether morphological alterations are correlated with oncogenic transformation in cells from the lining epithelia, which constitute the major sites of cancer incidence in humans. In an experimental model of respiratory carcinogenesis, cell lines derived from carcinogen-exposed respiratory airway epithelium showed anchorage independ ence and tumorigenicity after varying lengths of time in culture (19). The characteristics of populations similar in origin but differing in tumorigenicity have been studied in cells from sequential passages of these lines. To accomplish the present objectives, we have compared cell shape and microvillar den sity in cells from populations derived before and after the time at which anchorage independence and tumorigenicity were expressed. Since cell shape and surface morphology vary considerably among cells of the lining epithelia in vivo (1, 14, 25), significant morphological variability was expected in the cells studied. Therefore, the approach taken was to assess the variability of shape and surface features in cells from small colonies, predominantly of clonal derivation. By this approach, a range of variability could be defined for preoncogenic popu lations; thus, the appearance of new characteristics or chang ing frequencies of expression of existing characteristics could be detected. MATERIALS AND METHODS Cell Culture. All of the cell lines used were derived from specific-pathogen-free, inbred female Fischer 344 rats. Two ABSTRACT The purpose of the present studies was to determine whether changes in cell shape and microvillar density accompanied oncogenic transformation of rat respiratory tract epithelial cells. Two cell lines which became oncogenic during in vitro culture (1000 W and 165 S) were studied. In relatively late passages, but not in early passages, the lines produced keratinizing squamous cell carcinomas when tested in syngeneic hosts. Small colonies, predominantly of clonal origin, were obtained at early and late times after initiation of the lines into in vitro culture. Scanning electron microscopic studies showed that preoncogenic and oncogenic populations differed with respect to the shapes of cells within colonies. Differences in cell shape were further analyzed by estimation of the height and the ratio of length to width for 20 cells sampled from each colony. Each cell was assigned to one of nine classes of cell shape. The frequency with which spindle-shaped cells were observed in colonies increased 3-fold with oncogenic transformation of the 1000 W and 165 S lines. The frequency did not increase during in vitro culture of a third highly oncogenic cell line, BP 3-0. The frequency of observation of spindle-shaped cells in the 1000 W line was not decreased by in vivo growth and rederivation. In fact, the tumor-derived subline, 1000 WT, had a 5-fold greater frequency of expression than did an early passage of the 1000 W line. The number of colonies in which this cell shape was observed also increased 5-fold and came to include nearly one-half of the colonies analyzed. Therefore, expression of spindle shape became prevalent in clonal subpopulations of the line. In early passages of the 1000 W and 165 S lines, most spindle-shaped cells were found at the edges of colonies. This observation suggested that the spindle shape was assumed in response to forces generated during colony expansion. In general, the 1000 W line, which was more oncogenic than the 165 S line, also showed more pronounced morphological al terations. The prevalence of ruffles was well correlated with oncogenicity in the 1000 W line. However, the cell lines differed with respect to the density of microvilli at the cell surface, and this feature did not seem well correlated with oncogenicity. The results suggested that cytoskeletal and/or adhesive mecha nisms implicated in shape maintenance were altered in parallel with oncogenic transformation of epithelial cells originating from the respiratory tract. INTRODUCTION Previous comparisons of normal and transformed cells in vitro have shown that certain morphological features are cor related with oncogenic transformation. In light microscopic , Research sponsored by the Division of Biomedical and Environmental Re search, United States Department of Energy, under contract w-7405-eng-26 with the Union Carbide Corporation. 2 To whom requests for reprints should be addressed. Received October 12. 1978; accepted March 21, 1979. CANCERRESEARCHVOL. 39 2390 Morphological Markers of Oncogenic Transformation in Respiratory Tract Epithelial Cells1 C. A. Heckman2 and A. C. Olson Biology Division. Oak Ridge National Laboratory. Oak Ridge. Tennessee37830 Research. on November 24, 2021. © 1979 American Association for Cancer cancerres.aacrjournals.org Downloaded from