Suspensions of yolk sac infected with rickettsiae of murine or of epidemic typhus, and indeed suspensions of liver or peritoneal washings from cotton rats infected with these organisms, when the suspensions contain a sufficient concentration of living rickettsiae, possess the capacity to hemolyze, in vitro, red cells of rabbit or sheep origin. Under the same conditions such material did not cause the hemolysis of cells from mice, cotton rats, or guinea pigs. Yolk sac and tissue suspensions of R. orientalis (Karp, Raub, and Gilliam strains) failed to hemolyze rabbit cells.
The hemolytic principle was destroyed by formol (0.5 per cent) and by heating to 56°C. for 1 hour. The close association of the "hemolysin" with the rickettsiae was indicated by their parallel movements on centrifugation and by the close parallel demonstrated between the effects produced on hemolytic activity and on such manifestations of the presence of living rickettsiae as toxicity for mice and lethality or total infectivity for cotton rats when infected yolk sac suspensions were exposed to elevated temperature or were concentrated by high-speed centrifugation.
The mechanism of hemolysis remains ill defined. It progresses slowly, but never to completion (i.e. hemolysis of all the cells in the substrate) even when the number of such cells is small. This may be because hemolysis progresses best at temperatures (±35°C.) which are unfavorable to the hemolytic principle and because erythrocytes vary in their susceptibility to the injurious agent. The concentration of red cells also may be a determining factor, either by affecting the rate at which the reaction proceeds or by aiding in the preservation of hemolytic activity. Although it was impossible to separate the "hemolysin" from the rickettsiae, examination of smear preparations of hemolyzing mixtures showed no consistent contact relation between cells and rickettsiae. Adsorption experiments revealed that, while removal of the cells from the hemolytic system did not reduce the hemolyzing capacity of the system, the cells carried away a significant tendency to hemolyze. It is suggested that the rate at which injury is inflicted on the red cells is dependent on the concentration of "hemolysin" present and that the process may proceed in the relative absence of rickettsiae once sufficient damage has been inflicted.
The toxicity of richly infected yolk sac suspensions for intravenously inoculated mice was found to be related, not to in vivo hemolysis, but apparently to a serious alteration in vascular permeability which caused a marked reduction in blood volume. This was evidenced by an increased concentration of red cells but not of plasma proteins. It is possible, but has not been demonstrated, that the factor responsible for altering vascular permeability of mice may be identical with that which causes lysis of rabbit and sheep cells.
Two applications of the phenomenon of in vitro hemolysis are indicated. Following the technique described in the text for the optimum demonstration of hemolysis, the determination of the degree of hemolytic activity may serve as a rapid, roughly quantitative in vitro method for assessing the infectivity of a newly made rickettsial preparation. The phenomenon also is the basis for an additional, and potentially useful in vitro serologic technique which differs from other techniques in that it depends on a property of living rickettsiae. It has been shown that homologous antisera are capable of inhibiting hemolysis, and optimum conditions for demonstrating this inhibitory effect are described.