ON THE MECHANISM OF THE SERUM SENSITIZATION OF ACID-FAST BACTERIA.

Serum sensitization of the acid-fast bacteria causes two definite and directly observable changes in the bacterial surface: 1. A change from a surface readily wet by oil to a surface more readily wet by aqueous salt solution than by oil. This change is observed by microscopic examination of the bacteria in a saline-oil interface; thus detected, the surface alteration is said to constitute a "positive interface reaction." 2. An increased cohesiveness of the sensitized bacteria. This may be detected either by centrifuging the bacteria and then shaking up the sediment (resuspension reaction), or by observation of the clumps in the saline-oil interface. The interface reaction is serologically specific and confirms the existence of qualitative differences among acid-fast bacteria. The interface reaction parallels the binding of agglutinins as detected by the resuspension reaction, but not agglutination as ordinarily tested for. The interface reaction is less sensitive,-i.e., gives lower titers-than the resuspension reaction in about the average ratio of 1:3. The interface reaction in most instances runs approximately parallel to the complement fixation reaction; under at least one set of conditions, however, the interface reaction is correlated with the binding of agglutinin but not with the complement fixation reaction. How much of the bacterial surface must be covered with agglutinin in order to produce agglutination varies greatly with the bacterial strain used. The bacterial surfaces are modified by treatment with fresh normal sera in a manner quantitatively less but qualitatively not observably different from the effects of immune sera. Heating normal human, sheep, goat, or rabbit sera for 30 minutes at 56 degrees C. has usually diminished but not abolished their effect on the bacterial surface. Similar inactivation of guinea pig sera left them without detectable effect on the bacterial surface. The agglutination prezone is shown to be due to interference by excess colloidal material with the collisions of the bacteria prerequisite to clumping. The prezone maybe abolished by centrifugation and resuspension of the sediment. Antibodies may be partially dissociated from the sensitized bacteria by alkali, with return of the bacterial surface toward its normal, unsensitized condition. A carbohydrate yielding on hydrolysis a positive pentose test has been detected in the specific alcohol extracts of acid-fast bacteria studied by Furth and Aronson.(27) The tentative suggestion is made that the alcohol-soluble antigens of acid-fast microorganisms may be conjugated lipins owing their specificity to carbohydrate haptenes. Protective antipneumococcus globulins after heat denaturation have shown behavior in the saline-tricaprylin interface indistinguishable from that of maximally sensitized acid-fast bacteria. This strengthens the evidence suggesting that sensitized bacteria are coated with denatured globulin.