The findings described here serve to demonstrate two outstanding biologic characteristics of the primitive cells of malignant attributes occurring in virus-induced myeloblastosis: the potentiality (1) to proliferate at exponential growth rates in vitro and (2) to liberate the myeloblastosis virus at a constant rate in a given culture series. Attainment of this status of the myeloblast culture system has been dependent on the increasing recognition of the requirements of the cells with respect both to their own growth processes and, as well, to the activities concerned with the synthesis of virus.

With respect to cell growth, the myeloblasts possess fully those potentials for proliferation which enable them to multiply at high constant rates in vitro. The results were the same with myeloblasts freshly obtained from the blood of diseased chickens; with those derived by treatment of normal bone marrow with virus in vitro; or those held for long periods in culture under highly varied conditions. This property of the cell could not be predicted in advance and has been appreciated, finally, only by the evolution of a medium capable of supplying consistently the nutrient and growth factors required for the stabilization of the growth processes involved. In its capacity to proliferate in culture, the myeloblast associated with the avian myeloblastosis virus differs profoundly from the myeloblast of the normal bone marrow, which does not survive in culture under the conditions thus far investigated (5). Thus, the proliferative capacity of the virus-associated cell in culture constitutes an outstanding biologic criterion of the influence of the agent.

Myeloblastosis in the intact bird is a highly malignant disease, in which the only manifestation of neoplasia is the tremendous overgrowth of myeloblasts. Other elements, such as reticular cells of the bone marrow and spleen, macrophages, and endothelial cells, are obviously susceptible to infection, but only the myeloblasts display neoplastic traits. These characteristics are not lost in culture, and, furthermore, they are evident in cells emanating from normal bone marrow exposed in culture to the myeloblastosis virus. Studies have not yet been made to determine whether every virus-exposed cell is malignant or is actually infected with the virus. It is not too remote to assume, however, that for practical purposes all the myeloblasts in the blood of the diseased bird and those in culture are virus-infected. This is based on the previously mentioned limited capacity of the normal myeloblast to proliferate, either in the bird or in culture. Thus, cells not under the influence of the virus would be expected to mature to granulocytes in the bird or to succumb in culture. Furthermore, there can be seen in thin sections, under proper conditions (8, 5), evidence of virus in essentially every cell.

From these considerations it may be concluded that the myeloblasts in vitro constitute essentially pure cultures of virus-induced malignant cells, which can be maintained indefinitely in a state of exponential growth. The cells occur as individuals in suspension in the blood of the host from which they can be procured as wished and quantitated by the simple process of enumeration in a hemocytometer. In addition, the cells grow best in suspension in shaking cultures which provide the same advantages for quantitation. With these assets, the cultures comprise a practically ideal material for the study of the factors governing the growth of a virus-induced tumor cell.

The factors influencing growth of the myeloblasts have not yet been fully defined. The usefulness of the culture system for extensive and controlled studies has depended on the stabilization of growth behavior. Aside from the supplementary nutritive and growth factors supplied by medium 199, the essential constituents consist of chicken serum in 50 percent concentration and the very high concentration of folic acid. As already seen, exponential cell growth and constant rates of virus liberation have been observed (4) in a medium of whole serum and, in other studies (5), in a medium of 50 percent chicken serum in medium 199 without additional folic acid supplement. This was a rare experience, however, because of great variation in the efficacy of different batches of serum (1). With folic acid, the behavior of the cultures has remained uniform. Some information has been gained of the amino acid requirements, and studies have been made of the influence of B vitamins other than folic acid. The basis of the need for high concentrations of serum is not known, but it is clear that no substitute for it has been found. Fortunately, serum collected easily and economically in large volumes at a chicken abattoir yields entirely satisfactory results with the addition of the high concentrations of folic acid. The results of studies on the evolution of the present medium will be discussed in detail in a later report.

The phenomena of synthesis and liberation of the myeloblastosis virus by myeloblasts in culture have been the primary stimulus to the continued efforts to grow the cells under controllable conditions. Some of the most important fundamental questions, concerned with the virus tumor problem in current work, are those pertaining to the factors governing the onset and maintenance of cell infection. For the study of such virus-host cell relationships the myeloblast culture system seems thus far without equal. The characteristics of cell growth have already been described. The agent can be titrated rapidly and accurately by virtue of the enzyme activity to dephosphorylate adenosine triphosphate. Furthermore, this assay measures, not infectious units, but absolute number of virus particles. In this respect, it is a fortunate circumstance that the enzyme activity is independent of the infectious moiety of the virus (18), and is highly stable under the conditions of tissue culture. The results of the assays are interpretable with complete assurance as representative of the output of virus in terms of actual particles, as can be fully corroborated by direct counts with the electron microscope. Of equal significance is the occurrence, under proper conditions, of ultra-structures specific to the virus-infected cell. These, obviously associated with virus synthesis and liberation, are of a size and other characteristics suitable not only for study by electron microscopy, but by the methods of cytochemistry and other analyses with the phase and light microscope (19).