We have attempted in the preceding experiments the beginning of an analysis of bacterial anaphylaxis and its relation to the occurrences in the animal body during an infectious disease. We have shown that the sensitization of the tissues of guinea pigs, as indicated by the isolated uterus, required 3 to 5 days even when passive sensitization was employed, and that in these relations conditions with bacterial sensitization were entirely analogous to those revealed for serum anaphylaxis by Dale and Weil especially. It has become apparent that the sensitized uterus reacted not at all with whole bacteria or whole red cells, or, in other words, that before reaction with sensitized organs could occur an extraction or solution of the bacterial cell must take place. That bacteria yield some of their substance to the circulating blood during the course of infection was to be expected, but it has been definitely indicated, we think, by our complement fixations.
The mechanism of injury in the sensitized animal or in the human being so far along in typhoid fever that antibodies have begun to develop is in part one in which antigen, derived from the bacilli and brought into solution, or rather suspension, in the blood stream, reacts with antibodies which are from the beginning, or have subsequently become, integral parts of the cell protoplasm, the entire process taking place within the cell. This last point is indicated by the failure to sensitize by simply soaking the normal uterus in antiserum.
This, however, cannot be the entire story of injury. We know that typhoid antigen injected into normal animals in moderate amounts will render them gradually sick and eventually kill them. Also, a sufficient amount injected into a normal animal will occasionally produce acute symptoms, in every respect similar to the reaction produced in sensitized animals by smaller doses. We have shown that such acute symptoms in normal animals were not due in any degree to tissue sensitiveness, since even very large quantities of antigen will produce no response on the part of the normal uterus. It is reasonable to suppose, therefore, that the injury, gradual or acute, in the normal animal, is in no respect referable to tissue sensitiveness to the whole antigen, but rather must be referred to some series of phenomena which occur in the circulation. The acute shock of normal animals may possibly, therefore, be entirely due to an intravascular reaction. Whether this is one of antigen-splitting, or of antienzyme removal in the sense of Jobling is a point on which these experiments throw no light.
It is true that we have never succeeded in producing acute toxic symptoms either in the whole animal or in the isolated uterus with serum from animals acutely ill. This we eliminate as negative evidence inasmuch as we believe that the toxic substances need at no given time be present in the blood stream in sufficient concentration to render such an experiment successful. They are probably absorbed and do their injury almost as rapidly as formed, an assumption which is based on the speed with which symptoms develop.
It is possible, and not to be denied on the basis of any experiment that we can devise at present, that the gradual illness of the normal animal and the occasional acute shock of these animals may be based on entirely different mechanisms. In both cases, however, in normal animals, they seem to be intravascular. And since the symptoms of acute shock which can be produced in sensitized animals with moderate doses can also, though only occasionally, be produced in normal animals with larger doses, it is reasonable to suppose that the poisons produced intracellularly in the one may be similar to those produced intravascularly in the other.
It does not seem likely that the specific circulating antibodies are in any way sources of increased injury to an animal spontaneously infected with bacteria. If sufficiently powerful at the beginning they may even prevent tissue injury, first by increasing phagocytosis, then by producing intravascular agglutination, and finally, as indicated by our experiments, even by removing a part of the antigen from possible reaction with the cell, though in this last respect our experiments indicate that they functionate imperfectly. It is more probable that their chief protective action to the sensitized body lies in removing the whole bacteria from the possibility of intravascular disintegration, which, as we have shown, is prerequisite to anaphylactic injury of the tissues of the host.
We would tentatively summarize our opinion as to the occurrences in the typhoid-infected body as follows: Early injury is probably due to disintegration of part of the bacteria in the course of which albumose-like bodies are liberated, and, following which, intravascular reactions result in the formation of toxic substances, perhaps by some form of proteolysis.
Since the accumulation of bacteria during these stages is relatively slight, this form of injury probably plays little part in producing symptoms. Indeed, the experiment by which acute injury is produced in the normal guinea pig by the sudden injection of several times the lethal dose of partly dissolved bacteria, finds no analogy in the spontaneously diseased body.
At this time the tissues are not sensitive, but as antigen absorption progresses and the tissues are stimulated to react, sensitiveness develops, which renders them much more delicately amenable to injury by direct reaction with even small amounts of dissolved but otherwise unaltered antigen. This process is directly counteracted by circulating antibodies which tend to remove the bacteria from the possibility of yielding their antigen to solution by agglutinating them, aiding phagocytosis, and to a slight extent even neutralizing dissolved antigen.
It seems likely, therefore, that the symptoms which appear as the incubation time ends are largely those due to cellular sensitization which probably begins before any considerable amount of circulatory antibodies is present. The circulating antibodies would seem to have little or nothing to do with intravascular injury, the ferments responsible for this, however much it may occur, probably consisting of the non-specific proteases studied in this connection by Jobling.
Finally it appears that highly sensitized animals are more easily killed by typhoid antigen than are normal animals, provided they do not dispose over unusually large amounts of circulating antibodies.
Cure would consist of a gradual checking of growth and final destruction of the bacteria, and the consequent cessation of antigen liberation, but delicate hypersusceptibility would probably persist for some time after cure and immunity have been established. Just what the relation between tissue hypersusceptibility and immunity is remains a problem for further study.