MECHANISMS OF PERSISTENT AND MASKED INFECTIONS IN TISSUE CULTURE

I t has been repeatedly observed that certain infected tissues that initially cannot be made to yield virus upon isolation will do so after the tissues have been subcultured i.n aitro (Rowe et al., 1953, and Hull, Minner, and Smith, 1956). This cultivation in vitro subjects the cells to new environmental influences. The cells are transformed from a resting state to one of rapid proliferation; furthermore, they are removed from the humoral influences of the host. Once isolated, the viruses are apparently cytopathogenic to the progeny of the cells that formerly harbored them. If one disregards the possibility that the cultures were repeatedly contaminated after isolation, one must assume that the virus was present in the tissue in some nontransmissible or masked form. The factors involving the length of time the virus had persisted in the tissue and whether it had been undergoing some restricted multiplication present questions of considerable interest. Several years ago we made the observation that poliovirus could be repeatedly isolated from subcultures of HeLa cells in which prolonged and extensive cellular multiplication was occurring (Ackermann and Kurtz, 1955). In this experimental system there was strong evidence that the virus not only persisted, but to some degree replicated. More recently, this type of observation in vitro has been extended to the adenoviruses (Ginsberg and Boyer, 1956). The aim of this presentation is to consider what relation these in vitro systems bear to the phenomenon described by Rowe and Hull and their associates, on the one hand, and to the problem of recurring infections and persisting immunity, on the other. The first efforts a t explanation should be in terms of the known framework and principles of the single infectious sequence, and one should be cautious in postulating new principles and laws. From this viewpoint it may be worth while to note that there is a nontransmissible phase in the normal infectious sequence of animal viruses. Further, if the normal sequence of development is interrupted by the use of metabolic antagonists after infection has been initiated, as has been done with influenza virus (Ackermann and Maassab, 1955), or by the use of deficient medium as applied to psittacosis (Morgan, 1956), the infected cell can remain in a state of virostasis for long periods, even for days. The period of virostasis may be terminated upon change in the environment, and the infectious sequence will then continue and will yield virus. In the case in which virostasis was induced with an inhibitor of protein synthesis, viral development still continued in this noninfectious stage. The fact that virus in the noninfectious state is quite durable, perhaps more so than when in the extracellular infectious form, is further illustrated by recombination experiments in which irradiated influenza virus was found to react