SEROLOGICAL activity of various fractions of culture filtrates of the tubercle bacillus.

As culture filtrates of Mycobacterium tuberculosis have been shown by numerous workers to be antigenically heterogeneous, a series of experiments has been started to identify some of the different antigens likely to be useful in laboratory tests for measuring the antibodies in human blood. The work involves the fractionation of concentrated culture filtrates by various chemical procedures, the chemical analysis of the fractions, and their subsequent testing and cross-testing with various serological procedures.A number of fractions made from filtrates of heated and unheated cultures have been tested for activity in the Middlebrook-Dubos haemagglutination test. As might be expected, the fractions were found to vary in potency, some showing at least a thousand times more erythrocyte-sensitizing activity than others. Cross-inhibition tests made with the most active preparations indicated the existence of at least two different substances responsible for sensitizing erythrocytes to agglutination, one being demonstrable only in certain unheated fractions.Antibodies detected with the tannic acid haemagglutination test have a different serological specificity from those detected by the Middlebrook-Dubos test. Different heated fractions gave strong cross-reactions with each other in the tannic acid haemagglutination procedure, and little difference could be seen between them; in contrast, four active fractions from unheated filtrates were distinct from each other, showing very little cross-reaction.Results obtained with another haemagglutination procedure, consisting of agglutination of normal erythrocytes by antiserum-antigen mixtures, showed that this method detected another antigen, not measured in the tannic acid or Middlebrook-Dubos techniques under the conditions of the tests.The present experiments have shown that the culture filtrates and fractions thereof contain a series of distinct antigens or at least characteristic combinations of antigens that differ in serological specificity and in their ability to react in the different tests. Further experiments are necessary to determine whether the serological differences are of any special significance in relation to tuberculous disease or bacterial type specificity.