The results which have been presented show that under the conditions of artificial cultivation at 37°C. definite differences exist between two smooth strains of Pneumococcus Type III both of which are highly virulent for mice by the intraperitoneal route, but which may be sharply distinguished in their virulence for rabbits. These differences consist in the size of the fully developed intact capsule and the interval of time required for its loss. The somewhat smaller capsule of the avirulent strain, well formed and easily demonstrable during the early period of growth, diminishes quickly, while the large capsule of the strain virulent for rabbits is retained for a considerably longer period. Closely correlated with the time at which this reduction of capsule occurs is the appearance of changes in the surface properties of the bacteria which are revealed by a shifting of the range of acid agglutination, susceptibility to clumping in anti-R serum and ingestion by normal adult human polymorphonuclear leucocytes and serum. Since it has been shown that these alterations as growth continues, result in a loss of characteristics which distinguish the strictly type specific, fully capsulated pneumococcus and ultimately lead to a state temporarily approximating that of the completely avirulent R form, and since under the experimental conditions they are inaugurated sooner, advance more rapidly and are more complete in the rabbit avirulent organism, we believe that they may partly account for difference in rabbit virulence of the two strains. In the following paper an attempt has therefore been made to correlate this behavior in vitro with the events attendant upon inoculation into the animal body.
The studies of Clark and Ruehl (16), Henrici (17), Bayne-Jones and Adolph (18) and others have demonstrated a marked increase in the size of the bacterial cell associated with the early phases of growth. These authors have dealt chiefly with noncapsulated rod forms and even Clark and Ruehl who included cultures of various cocci do not make reference to variations in capsule size. Recently Seastone (19) has called attention to the large volume occupied by young capsulated streptococci. Similarly we have found that increase and decrease of Pneumococcus Type III volume appears to be due largely to the formation of capsule in young cultures and its subsequent loss as the organisms age. Because of the relatively great proportion of capsule in comparison with soma, a greater disparity exists between the volume of young and old pneumococci than that found by those who have studied bacteria lacking this structure. Of interest in connection with our observations are those of Preisz (20) on the nature of the capsules of virulent anthrax bacilli and strains attenuated by cultivation at 42.5°C. The latter produced soft, rapidly dissolving capsules while such structures in the former were characteristically firm and were retained by the bacilli for longer periods. This worker also noted in confirmation of the earlier work of others, that the capsules of B. anthracis are lost during the course of growth in serum media and in the subcutaneous tissues of the susceptible mouse.
We have demonstrated that the R variants derived under the same conditions from the two smooth strains of Pneumococcus Type III reveal certain characteristics by which they may be distinguished from each other in respect to cell and colony morphology, growth in broth, as well as growth at 41°C. (cf. Paper I). By employing the method of Griffith, these two R variants have been induced to revert to the S form. Following the injection into mice of the various possible combinations of living R variant and the killed S organisms of either rabbit virulent or avirulent strain, as well as very large numbers of the R variant alone, S forms emerged which in their various attributes, notably that of virulence for rabbits, resembled the original smooth strain from which the particular R variant involved was dissociated. The function of the smooth killed organisms in the process of transformation appeared to be only that of a stimulus toward reversion to the S. They apparently play no rôle in determining the virulence or the growth properties of the resulting S form.
These observations indicate that the factors involved in virulence are conditioned by stable physiological properties peculiar to the individual strain and that although temporarily inactive during the R state, they are again resumed unaltered upon the transition to the S form. They serve also to reemphasize the fact, apparent from several studies but perhaps not sufficiently realized, that the R variants of the pneumococcus, even though obtained under the same conditions from the same type but from different strains, may vary definitely in their various attributes.
Finally, they strongly suggest that the degree of virulence of a given strain of a bacterial species may be determined not only by its ability to multiply in the environment of the host and to synthesize certain substances of definite chemical and antigenic properties, but also by the capacity to elaborate these in greater or lesser degree and under the conditions of parasitism within the animal body to maintain them in contact with the soma of the cell in such state that they afford an efficient barrier to the defensive mechanisms of the host.