1. Our experiments show that proteotoxin (anaphylatoxin) produced with typhoid bacilli and active guinea pig serum, by the general method of Friedberger, possesses aggressin-like action in that, when injected together with sublethal doses of bacteria, it renders them lethal.
2. This action is not specific, since even when typhoid bacilli are used in its production, the proteotoxin possesses this property not only for typhoid bacilli, but for staphylococci and prodigiosus bacilli, the only other organisms with which we worked.
3. This action is manifest in the experiment only when quantities of proteotoxin are used which can produce a certain degree of systemic poisoning.
4. It is not due to anti-opsonic action since the phagocytosis of staphylococci in the presence of active serum takes place readily in spite of the presence of relatively considerable amounts of the poison.
5. The aggressin-like action of the proteotoxin is probably due to the leucopenia that it causes. It is not likely that this leucopenia is dependent upon a negative chemotactic action, in the ordinary sense of this term, but it is probably due to the general intoxication of the animal, resulting among other things in the poisoning of those tissue elements which ordinarily react to infection with the mobilization of leucocytes in the circulation.
It has seemed to us that the results of our experiments were of interest also in throwing some light on the general problem of virulence.
The thought naturally suggests itself that Bail's so called aggressins were nothing more or less than proteotoxins, or anaphylatoxin. It has been shown by Friedberger and Nathan (3) that poisons of this description may be formed in the peritoneal cavities of animals, and if we examine the method by which Bail (4) prepared his aggressive exudates, it seems likely that they contained proteotoxins. Moreover, although Bail's aggressins are described as non-toxic in themselves, this is a discrepancy to which, it seems to us, no great weight need be attached in this connection, since the poisonous properties of the proteotoxins are clearly apparent only when considerable amounts, two to three cubic centimeters, are intravenously injected into young guinea pigs, and below this quantity a threshold is rapidly reached beyond which no visible reaction whatever is achieved. We have not yet, it is true, shown that animals may be immunized against bacterial infection by treatment with the proteotoxins, a result obtained by Bail with his aggressins. But we feel that the mere fact that tolerance to proteotoxin may be established, a phenomenon which has been the subject of an earlier paper, is a step in this direction. Further than this the analogy cannot be carried since we do not know as yet whether the proteotoxin tolerance is specific,—a fact which would be necessary in order to identify completely these substances with the substances studied by Bail. In view, however, of the many objections that have been voiced against this phase of Bail's work, and of the fact that a rapid accumulation of leucocytes is described as characteristic of aggressin immunity, it seems not unlikely that even in this aspect of the problem aggressin and proteotoxin may eventually be identified.
It is probable, also, if we reason logically from the premises which our experiments supply, that the power of invasion of many microorganisms is intimately related to their ability to react with the active blood constituents of the infected animal. This has already been suggested in the work of Embleton and Thiele (5). If, indeed, the condition of bacterial anaphylaxis or hypersusceptibility depends upon the establishment of a balance between the injected bacterial antigen and the blood constituents, such that proteotoxins are formed, a supposition rendered probable by much recent work, then our experiments would tend to show that bacterial anaphylaxis implies, at the same time, a condition of lessened resistance to invasion by the bacteria (6). It may even be that generalized invasion, after localization, must await the striking of such a balance, an interval which, in the case of diseases like typhoid fever, may form a part of the incubation time. At the same time, in such diseases, the action of the proteotoxin would explain the leucopenia that accompanies the bacterial invasion.
We do not, of course, expect to explain virulence in general by these phenomena. In the case of organisms like the pyogenic cocci, the pneumococcus, and some other bacteria, there is apparently another important element contributed in the resistance of the invader to direct reaction with the active serum constituents. In many of these organisms the slight extent or absence of serum lysis or even bactericidal action would point in this direction, and here the appearance of proteotoxin would be secondary in importance or possibly absent entirely until late in severe infections.
The non-specificity of the aggressive action of proteotoxin goes far toward explaining the increased susceptibility to other, secondary infections apparent during the course of many cases of typhoid fever and in some other diseases. It should also have an important bearing on our understanding of mixed infections.