Serological Relationship between Clostridium Bifermentans and Clostridium Sordellii Based upon Soluble Antigens.

The problem of whether Clostridium sordellii should be assigned species status or be retained as a pathogenic variant of C. bifermentans is a perplexing one. Despite marked similarities in morphology, colony formation, and biochemical reactions (Clark and Hall, J. Bacteriol. 33:23, 1937; Stewart, J. Bacteriol. 35:13, 1938), Brooks and Epps (J. Gen. Microbiol. 21:144, 1958) were able to distinguish between strains of C. bifermentans which were urease-negative and nonpathogenic and strains of C. sordellii which were urease-positive and varied in their pathogenicity. Walker (J. Pathol. Bacteriol. 85:41, 1963), using spore agglutinogens from these two organisms, found cross-agglutination to occur. The present report describes serological similarities and differences in the soluble antigens of the C. bifermentans-C. sordellii complex as studied by agargel diffusion. Four strains of C. bifermentans and nine strains of C. sordellii, of which four were pathogenic (P) and five nonpathogenic (NP), were employed. The technique was essentially that described by Ellner and Bohan (J. Bacteriol. 83:284, 1962), except that the medium was modified so as to be nonantigenic in this system. Among the four strains of C. bifermentans, six distinct antigens could be distinguished, of which at least five were common to all strains (Fig. 1). Among the nine strains of C. sordellii, seven distinct antigens were observed, of which at least five were common to all strains (Fig. 2 and 3). Although pathogenicity tests clearly distinguished between P and NP strains of C. sordellii, no antigenic differences could be detected. WVhen antigen pools representing either C. sordellii or C. bifermentans were tested against both homologous and heterologous sera, it became apparent that four antigens were common to both species. In addition to these four common antigens, the C. sordellii antigen pool produced three unique lines when tested with homologous serum. On the other hand, the C. bifermentans antigen pool demonstrated two unique lines with either homologous or heterologous serum (Fig. 4). After absorption of C. bifermentans serum with the pooled C. sordellii antigen, however, the heterologous reaction was eliminated, and two lines were obtained with the homologous antigen (Fig. 5). In like manner, after absorption of C. sordellii serum with pooled C. bifermentans antigen, the heterologous reaction was eliminated and three lines appeared with the homologous antigen (Fig. 6). The above studies clearly indicate that, although the majority of antigens are common to both "species," each possesses unique speciesspecific antigens. Since our own studies currently in progress show many antigens to be common to a wide variety of clostridia, it is believed that the existence of these unique species-specific antigens constitutes an additional criterion for individual species status.