This paper has three main objectives: (1) to examine the conditions under which a virus, able to produce malignant tumor in the animal, can transform in vitro normal cells into characteristically altered cells, different from the normal cells in morphological and physiological properties; (2) to discuss the nature of the virus-cell interaction leading to this transformation; and (3) to investigate the relationship of the virus-transformed cells to cancer cells as they are found in the animal. For these purposes a single virus-cell system will be used, that constituted by the Rous sarcoma virus and by cells derived from the chicken embryo, a system extensively used in our laboratory.

We shall briefly consider the characteristics of this model system. The virus is constituted by particles about 80 ran in diameter, containing ribose nucleic acid, protein, and lipides. Virus stocks can be easily obtained by extracting the fully developed Rous sarcoma in chickens, or by collecting the supernatant of tissue cultures of chicken-embryo cells infected by the virus. The cell-free filtrate thus obtained has the ability to infect young chicks, and to reproduce in them the characteristic sarcoma. The infectious quality of the virus is considerably heat-labile, being lost at 37° C. in the usual media with a half life of 2 1/2 hours.

The virus is propagated in vitro in cells derived from chicken embryos; turkey-embryo cells would also be suitable. One important problem arises in the choice of a suitable source of embryos. It is known that of the many available breeds of chickens some are considerably resistant to the Rous virus; this resistance appears to be of a genetic nature, and is manifest also in the cell cultures derived from embryos of these breeds. A genetically sensitive breed must therefore be chosen; this can be accomplished only in an approximate way, since there are no completely inbred strains of chicken in this country. A number of poultry farms maintain relatively inbred stocks, some of which are satisfactory for work with the Rous virus. These stocks, however, are not entirely uniform in genetic constitution and contain a fraction of relatively resistant animals. Since cell cultures are usually produced in batches from randomly chosen populations of embryos, they usually contain a proportion of genetically resistant cells. The proportion of genetically resistant cells is generally small and does not cause important difficulties. If needed, cultures entirely constituted by genetically sensitive cells can be obtained, by making cultures from single embryos.

The cell cultures sometimes show, in addition to this type of resistance, another type of resistance to the Rous virus, of a physiological nature. Physiological resistance may be caused by some conditions of cultivation of the cells, such as age of the cultures, composition of the medium, and density of the cultures; the relative importance of the various conditions and their significance in causing physiological resistance are unknown.2

The cultures generally used with the Rous virus are monolayer cultures, in which the cells sparsely cover the glass bottom of a petri dish. A standard type of monolayer culture is obtained from decapitated 10-day chicken embryos dispersed in trypsin; these cultures contain mostly elongated cells, described as fibroblasts, deriving in majority from the body wall. For special purposes, cultures can be prepared either from trypsin-dispersed hearts of 10- to 16-day embryos-referred to as heart fibroblasts-or from lungs of 16-day embryos, also dispersed by trypsin. The lung cultures contain, in addition to fibroblast-like cells, many epithelial cells, which can be enriched by proper techniques. For best results sparse, young cultures, usually secondary 1-day-old, are used.

We want now to examine the effects of the exposure of suitable cultures to the Rous virus. Several effects are observed; they will be examined in succession.

The most noticeable effect is a morphological one. It was observed by Manaker and Group6 (1) that suitable cell cultures exposed to a virus preparation containing a small number of infectious units develop foci of characteristic round cells. This phenomenon by which an elongated cell, upon infection with the virus, becomes spherical will be defined as "conversion."

Subsequent study by Temin showed that conversion of elongated cells to round morphology is only a special case of conversion. Many other types of morphological conversion caused by the Rous virus were observed. In general we can say that the morphological cell type deriving from the interaction of chicken cells in vitro with the Rous virus depends on two factors at least: the type of virus and the type of cell used.

We shall examine the types of conversion observed as a function of the two main variables, the virus and the cell type.

We shall first consider the case in which identical cells are infected by different virus types. Populations of identical cells can be obtained as progeny of a single cell: They are called clonal populations. Different virus types can be obtained by the selection of viral variants in standard virus stocks; the variants are selected because they convert cells to morphological types different from the standard ones. Several variants of this type have been found thus far. When clonal populations of fibro-blast-type cells are exposed to different variant types of virus, different types of conversion are found. The two extreme cases are the conversion to a round, highly refractile cell type, which is the one produced by the standard virus, and the conversion to an extremely elongated cell type. Other virus variants convert the cells to intermediate types.

2 Footnote added in proof: It has been recently shown by Rubin (personal communication) that an important (actor causing cellular resistance to the Rous virus is a latent Infection of the cells with a related virus, probably of the avian lymphomatosis group.

It can be shown that these virus variants derive from the standard type as a consequence of a change occurring spontaneously, randomly, and independently of the cells in which the virus is propagated; the variant characteristic is hereditary. These changes are therefore mutations. Their frequency is of the order of 10~3 per virus particle, per generation.

That virus mutants can convert identical cells to different morphological types appears to have a considerable significance. On the one hand, in cells maintained in vitro, in the absence of virus infection, the morphological type tends to remain fairly constant over the generations, provided external conditions are uniform; on the other hand, it has been shown in one system that cell morphology changes with changes of the genie balance of the cells (S). Therefore the morphological type is an expression of the genie setup of the cell. The hereditary modifications of the morphology of the cells caused by the entry of the virus in them are thus likely to reflect a modification of the genie setup of the cell, due to the activity of the viral genome within the cell.