Étude microcinématographique du mécanisme d'invasion cancéreuse en cultures de tissu normal associé aux cellules malignes

Abstract

The phenomenon of malignant invasion was studied with the aid of phasecontrast time-lapse microcinematography on associated cultures of normal and malignant tissues. The mixed cultures were established on a very thin plasma coat covering standard size coverslips adapted to special perfusion culture chambers. The normal tissue was represented by the outgrowth from fragments of C3H mouse embryo heart tissue. Fibroblast-like cells as well as cells of a more polygonal shape forming pavement-like sheets, probably of endothelial origin, were the main cell components of this outgrowth. The malignant cells explanted as compact colonies, close to the normal tissue fragments, were represented either by the highly malignant N1strain cells of typically fibroblast-like appearance or by the “hybrid” clone M6, of a different morphology but similarly malignant. The results of the malignant versus normal tissue confrontation were examined by frame after frame analysis of microcinematographic recordings taken on mixed cultures of various ages up to 22 days. Normal cells characterized by moderate activity of ectoplasmic membranes had a good mutual coherence and single cells detached from the colony were observed only exceptionally. The junction of two normal cell outgrowths coming into contact leads to a perfectly coherent junction without overlapping. Quite differently, the malignant N1 and M6 cells showed a very strong and abrupt activity of ectoplasmic membranes and pseudopodia. Their outgrowth in dense cultures formed a disordered non-coherent and intermingled pattern. Single “sentinel” cells frequently detached from the whole colony, especially in the case of the N1 cells. They actively and rapidly migrated at a relatively long range. When the normal and the malignant outgrowth met each other in the mixed cultures, the malignant cells had a tendency to infiltrate the normal cell colony. This penetration was essentially due to progression of the single “sentinel” cells, and occurred particularly in areas where the normal outgrowth showed gaps or lacunar arrangements. On the contrary, the progression of the malignant cells was restrained and even stopped when opposed by coherent and compact sheets of normal cells, especially the pavementlike endothelial structures. Malignant cells, when arrested in their progression, often shifted along the demarcation line changing their previous orientation. At this point, our conclusions do not entirely fit with Abercrombie's conception of complete lack of “contact inhibition” regarding malignant cells. They are in agreement, however, with other observations especially those of Leighton concerning the intracapillary blocage of malignant cell emboli observed in sponge matrix cultures of mouse embryo heart tissue.