Research was conducted to investigate the biosynthesis of pertussis toxin (PT) as well as the structure and mechanism of action of the toxin. We have focused much or our work on the analysis of the secretion of PT from B. pertussis. Previously, we identified nine ptl genes that are essential for the secretion of the toxin. These genes include proteins that belong to the Type IV family of transporters. In order to gain further insight into the mechanism of secretion, we studied the ability of individual components of the toxin to be secreted. In order to do this, we constructed mutants of B. pertussis that expressed only the S1 subunit of the toxin or only the B oligomer of the toxin. We found that the B. oligomer was not secreted in the absence of the S1 subunit. Likewise, in the absence of the B oligomer, the S1 subunit was not secreted by a Ptl-dependent mechanism. These data indicate that it is the holotoxin form of the protein that is released from the organism by the Ptl transport system. In order to further understand the biogenesis of PT, we initiated studies to localize the subunits of the toxin prior to secretion. We first localized the S1 subunit of the toxin in strains that produce only S1, S1+B oligomer, and S1 + B oligomer + Ptl proteins (wild-type). We found that the S1 subunit localized to the outer membrane of the bacterium. Thus, the outer membrane may serve as the site of assembly of the toxin. After the toxin assembles, the Ptl proteins may act using a piston-like action to push the toxin through the outer membrane and into the extracellular space. Such a model would be consistent with the pilus structures that are often seen with Type IV transporters. We have also examined the stoichiometry of the subunits of pertussis toxin and the Ptl proteins by fusing a phoA gene to the ATG start codon of various ptx and ptl genes. The ptx and ptl genes are known to be part of the same operon. We found that the transcription/translation of genes decreased as the distance from the promoter increased. We found that the ratio of the ptxS1-phoA fusion to the ptlF-phoA fusion was 9:1. These results indicate that the ptl proteins may be synthesised at lower levels than the pertussis toxin subunits. This finding led to the hypothesis that the Ptl proteins may be rate limiting in secretion and, in fact, this appears to be the case since duplication of the ptx gene region results in higher levels of secretion of the toxin.