The experiments described above were begun because it appeared in earlier work with vaccine virus, in which death was being taken as the criterion of infection, that some mice sickened but did not die. This appeared to happen more frequently after inoculation with weak strains which had a low titer in the mouse as compared with other hosts and which gave a more irregular titration-curve in the mouse than with strains of high virulence for this animal. The experiments were extended when it became apparent that information having an important bearing on the mechanisms of infection might be obtained. The criterion of infection which was chosen for the present experiments was death following the initial inoculation of virus, or survival when the mice remaining from the initial titration were reinoculated with an amount of virus which consistently led to death of normal mice. This test, while more sensitive than the one previously used, was nevertheless subject to certain vagaries of its own as will appear. It was postulated that vaccinal immunity when it did appear was the result of actual infection, and was not consequent upon the introduction of enough virus to produce immunity in the absence of infection. This seemed reasonable since in most cases the resistance to a challenge inoculation was high and because vaccinal immunity is not readily induced by inoculation with other than relatively enormous quantities of inactive virus (4). Furthermore, the the immunity which is induced with inactive agents is difficult of demonstration. It early became evident, however, that vaccinal immunity is not absolute and under certain conditions a feeble resistance might be masked by the use of a rigorous test.

The question which led to the initiation of the experiments, whether death is a reliable index of vaccinal infection with all strains of virus, was soon answered. It is not. With IHD strain, virulent for rabbits and highly virulent for mice, there was little difference between the “lethal” and the “infectious” titers. The small discrepancy which was found did not seem to contravent the assumption that the minimal quantity of virus capable of causing infection was the same for rabbit and mouse. The CL strain, on the other hand, highly virulent for the rabbit and relatively avirulent for the mouse when mouse fatalities alone were considered, was shown to be only slightly less efficient than IHD strain in causing infection. With the less virulent CV II and CV I strains, there was a further loss of ability to infect.

With change of route of inoculation from intracerebral to subcutaneous, a sharp drop in apparent virulence of all strains of virus was shown. The “infectious unit” of IHD strain contained some hundred-fold more virus when the insertion was subcutaneous than when intracerebral insertion was used, the average increase being 102.2 times. The infectious unit of CL strain was 102.5 times larger on subcutaneous than on intracerebral insertion while with CV II strain the difference was 103.3, and with CV I was 102.4. These data were taken to indicate that ability to initiate infection by extracerebral routes was considerably less than ability to initiate an infection after intracerebral insertion.

That the immunity induced was not always absolute was indicated by several experiments. It was found that of a group of mice all inoculated (vaccinated) with similar quantities of active virus those reinoculated with a very small dose of virus were more apt to survive than those inoculated with a very large one. Both amounts of virus led to death of all of the normal mice inoculated. This variation in degree of immunity was more apparent among mice injected initially with CV II than IHD strain, and much more prominent among mice injected subcutaneously than those injected intracerebrally.

The results of previously published investigations in this and other laboratories (2, 5) have been interpreted as indicating that infection in vaccinia may be presumed to depend upon the entrance of an active viral particle into a susceptible host cell. On this basis, the varying resistance and susceptibility of different hosts would depend essentially upon the proportion of susceptible to resistant cells which they offered. It may be profitable to examine the data above in the light of this hypothesis to determine whether the results are consistent with it. If it be assumed that vaccinal infection can only follow the arrival of active virus in a susceptible cell, and that not all cells of a host are susceptible, one of several fates may overtake any single vaccinal particle following its inoculation. Thus it might arrive in a cell unsuited to it, a resistant cell, and failing to multiply therein become inactive. No special antiviral mechanism need be postulated but simply the spontaneous deterioration of the viral activity. No local lesion would appear and general immunity would not be expected to follow the insertion of even several particles under these conditions. Such an event appears frequently to follow intracerebral insertion of the weaker vaccinal strains, and more frequently to follow subcutaneous insertion of virus of all of the strains.

On the other hand, the particle may arrive in a susceptible cell and multiply therein. Escaping from this, it may enter other cells and so give rise to progressive disease. This viral multiplication, if the disease is nonfatal, is eventually controlled by the forces of acquired immunity. In regard to these it might be inferred that they would act more directly on the cells originally suitable for viral growth than on those originally resistant. With virus able to parasitize a wide range of host cells, it is to be expected that even with small inocula an extensive infection would result with a consequent solid immunity of surviving animals. Such a situation may be presumed to exist with IHD strain. The high fatality after intracerebral inoculation is probably due as much to vital function of the cells affected as to the virulence of the infection, but any mouse which survives the infection is solidly immune.

A third possibility is that virus may enter a susceptible cell, but in a tissue most of the cells of which are not suitable for viral growth. Here, even if multiplication of virus does occur, it is possible that no evident disease will be produced since the number of cells which can be infected is small. The stimulus to the forces of immunity would likewise be small, and it is not unreasonable to assume that not much resistance would need to be offered to check the growth of virus entirely so that the small response would nevertheless be quite adequate to control the infection. Such a situation might exist with subcutaneous inoculation of CV II strain, and in diminishing degree with CL and IHD strains. Here the initial improbability of finding a susceptible cell would be presumed to continue, and furthermore the immunity actually produced would be expected to vary with the size of the inoculum. With a large inoculum infection would begin simultaneously in many foci, giving a more adequate vaccinating stimulus even with a “weak” strain, whereas the infection which has its origin in only a few cells might give little or no measurable immunity.

In the discussion above, figures have been used indicating the titer of virus as measured in various ways and the suggestion has been made that the differences found may have their origin in part in the different number of susceptible cells present in different locations. As has been pointed out in previous discussions, it must be remembered that these figures indicate ratios and probabilities, and do not necessarily indicate the absolute number of cells available to the virus, or even in an exact sense the ratio of susceptible to nonsusceptible cells. It is not known, for example, what gradation of susceptibility may exist among a collection of cells nor what the fate of virus might be in a “partially susceptible” cell.

This content is only available via PDF.