Immunization with chemically detoxified pertussis toxin can prevent severe whooping cough with an efficacy similar to that of the cellular pertussis vaccine, which normally gives unwanted side effects. To avoid the reversion to toxicity and the loss of immunogenicity that may follow chemical treatment of pertussis toxin, inactive toxins were constructed by genetic manipulation. A number of genetically engineered alleles of the pertussis toxin genes, constructed by replacing either one or two key amino acids within the enzymatically active S1 subunit, were introduced into the chromosome of strains of Bordetella pertussis, B. parapertussis, and B. bronchiseptica. These strains produce mutant pertussis toxin molecules that are nontoxic and immunogenic and that protect mice from the intracerebral challenge with virulent Bordetella pertussis. Such molecules are ideal for the development of new and safer vaccines against whooping cough.

No abstract

The introduction of two amino acid substitutions within the enzymatically active subunit Si of pertussis toxin (PT) abolishes its ADP-ribosyltransferase activity and toxicity on CHO cells (Pizza et al., Science 246:497-500, 1989). These genetically inactivated molecules are also devoid of other in vivo adverse reactions typical of PT, such as induction of leukocytosis, potentiation of anaphylaxis, stimulation of insulin secretion, and histamine sensitivity. However, the mutant PT molecules are indistinguishable from wild-type PT in sodium dodecyl sulfate-polyacrylamide gel electrophoresis and maintain all the physical and chemical properties of PT, including affinity for toxin-neutralizing poly-and monoclonal antibodies. Either alone or stabilized with formaldehyde, PT mutants are able to induce high levels of neutralizing antibodies and to protect mice in a dose-dependent fashion against intracerebral challenge with virulent B. pertussis. These results clearly show that these genetically inactivated PT molecules are nontoxic but still immunogenic and justify their development as a component of a new, safer aceliular vaccine against whooping cough.

SummaryPT9K/129G, a nontoxic mutant of pertussis toxin (PT) obtained by genetic manipulation, has been shown in animal models to be a promising candidate for new vaccines against whooping cough. To assess the safety and the immunogenicity of PT-9K/129G in humans, a pilot study has been performed in adult volunteers. The protein was found to be safe, capable of inducing high titers of toxin-neutralizing antibodies, and capable of generating immunological memory. In fact, vaccination caused an increase of cell-mediated response to PT, PT9K/129G, S1 subunit, and B oligomer, indicating that memory T cells are induced by the vaccine. Since PT-9K/129G is mitogenic for T lymphocytes in vitro, it was investigated whether this activity is also present in vivo. No variation was observed in the proportion of T cells (CD3'), T helper cells (CD4+), and cytotoxic T cells (CD8+), as well as in that of other lymphoid populations, by FACS analysis. Interestingly, no thorough correlation was found between humoral and cellular responses. In one case, a very high cellular response was present in absence of detectable antibodies, suggesting that the antibody response, which is the only parameter measured in most clinical trials, may not give a complete picture of the response induced by a vaccine.

Secretory antibodies of the IgA class (sIgA) are thought to have an important role in the defence against bacteria at mucosal surfaces--the level at which the infectious agents first come into contact with the host. However, the mechanism by which sIgA exert their antibacterial activity is still a matter of debate. After the recent discovery of receptors for the Fc portion of IgA (RFc alpha) on lymphocytes, monocytes and granulocytes of human, rabbit, guinea pig and mouse origin, it has been hypothesized that IgA also mediate antibody-dependent cellular cytotoxicity (ADCC). Indeed, ADCC mediated by human leukocytes against bacteria has been demonstrated in the presence of human circulating IgA. As RFc alpha have also been shown to bind sIgA, we decided to investigate whether sIgA could mediate antibacterial ADCC when bound to lymphocytes from the murine gut-associated lymphoid tissues (GALT) which first interact with the invading bacteria. By using Shigella X16 (a hybrid strain between the enteric pathogen Shigella flexneri and Escherichia coli) as target in an in vitro assay that measures cell-mediated antibacterial responses, we found that murine lymphocytes from GALT but not from other tissues are able to exert natural antibacterial activity against Shigella X16, and that sIgA significantly and specifically increase the natural antibacterial activity of GALT lymphocytes from mice and induce antibacterial activity in cells from the spleen, but not from the thymus or popliteal lymph nodes. Thus, we now propose a new role for sIgA in protecting the host against infectious agents at the mucosal level.