There has been much effort in recent years, by what might be called summit meetings on the tumor-virus problem, to find a short cut to the solution of the numerous questions in this field by a rather indiscriminate exchange of ideas by workers in widely separated areas. While it cannot be denied that such exchanges may be highly stimulative and entertaining, it is certain that the problem cannot be talked to a solution. This can be attained only by the careful and progressive accumulation of information by investigators like you who are intimately concerned with direct experimentation.

Comparative Micromorphology of Normal, Transforming, and Malignant Cells

The purpose of micromorphologic comparisons of normal and malignant cells has changed rapidly within the past several years, due to improvements in microscopy, increased identification of morphologically recognizable cellular components, and increased availability of controlled experimental material. It is no longer sufficient, at the micromorphologic level, to describe merely size and shape of nuclear and cytoplasmic organelles. It now becomes our obligation to investigate structures of molecular dimensions with the particular view of determining (a) sites and modes of exchange of genetic information within and between cells, and (b) the relation of this exchange to normal and abnormal synthetic processes. This is particularly necessary in the study of virus-tumor cells in view of the close relationship, or exchange, of viral and cellular genetic information.

Detailed structural studies, in normal cells, of such processes as spermatogenesis, nucleo-cytoplasmic exchange, and nucleolar-chromosomal interrelations have already suggested the possibility of gaining an insight into normal genetic control of synthesis. Further, the visualization of cell border activity, ribonucleoprotein synthesis, and endomembrane continuities has opened the door for the visualization of pathways of exchange between environment and controlled cellular systems. Numerous investigators have afforded us a steadily increasing body of information on the ultrastructure of tumor cells, virus-infected cells, and viruses themselves. More pertinent to the study to be presented in the following pages is the recent work on the fine internal structure of virus particles and tumor cell inclusions. Specifically pertinent is the visualization of the coiled thread structure of polio virus (1), polyoma (2-5), adenovirus (6), and pox virus (7). Related to this is the recent demonstration of inclusion-body fine structure in myeloblastosis (8, 9), stomatitis (10), swine pox (11), and parotid tumor cells (S).

The survey to be presented is less a comparison of normal and malignant cells, but rather, more a comparison of virus-infected cultured cells, transforming cultured cells, and \rirus-induced tumor cells. It is hoped to find certain common denominators that can serve as clues to the determination of sites and modes of virus formation, and of the relation of these to normal and abnormal cell processes.


The nucleus normally comprises peripherally arranged chromatin, one or more nucleoli and nucleolar fragments, and a nucleoplasm containing various granules and fine filaments. The nucleus is bounded by an envelope continuous with membranes of the endoplasmic reticulum, and facultatively continuous, thus, with the plasma membrane of the cell. At various intervals the nuclear envelope is interrupted by the nuclear pores, affording continuity of nucleoplasm and cytoplasm. In normal cells the nuclear organelles pass through morphologic cycles in association with growth and metabolic events. Some nuclear changes observed in transforming and tumor cells are similarly related, i.e., increase in nuclear size, nucleolar size, vacuolization and fragmentation, and increase in numbers and width of nuclear pores with the increased synthetic activity of the cell (12, IS).

Viral particles, or viral precursory material, have been observed in the nuclei of numerous infected cells, e.g., herpes (14, 15), polyoma (2-5), Friend leukemia agent (16,17), and adenoviruses (18, 19). The question immediately arises as to how such material is transmitted to the cytoplasm. The mode of transmission apparently varies with virus type. Herpes virus, for example, may pick up a membrane from the nuclear envelope and probably be transmitted across the envelope by reverse pinocytosis. Cytopathic viruses, e.g., polio virus and hemadsorption virus (20), may cause breakdown of the nuclear envelope or enlargement of the nuclear pore to the point where the particles may simply diffuse directly into the cytoplasm. Noncytopathic viruses probably use the normal pathway of nucleo-cytoplasmic exchange, i.e., the channel of the nuclear pore. Evidence has been accumulating for the normal nuclear control of cytoplasmic processes by the production and migration of ribonucleic acid (RNA) within the nucleolus and the subsequent passage of nucleolar material through the channel to the cytoplasm. It is interesting to note that in the case of poliovirus (21), nucleolar material seems to pass to the cytoplasm, re-form a well-defined nucleolus, and that the filaments of the cytoplasmic nucleolar structure appear to be in direct association with material involved in poliovirus crystal formation. In the parotid tumor cells of polyoma-infected mice the intranuclear polyoma particles (figs. 1, 2, 3, and 4) appear to be closely associated with nucleolar filaments in the formative stages (fig. 1). In the leukemic thymus we have observed filamentous material in the nucleoplasm near the pores, within the pore channels, and continuous filaments from pore channel to cell border, where definitive virus particles are forming (fig. 12). The nucleolar material could thus be directly involved in virus formation. However, as in normal synthetic processes in the cell, the nucleolar material probably provides the transport mechanism and it is nucleolar-associated material that passes through the pore from nucleus to cytoplasm. Further, it is apparent that, in some virus tumor cells, particles per se do not pass from nucleus to cytoplasm, but rather a viral material in the form of fine filaments passes through the nuclear pore channel.