The current vaccine against tuberculosis, Mycobacterium bovis strain bacille Calmette-Guerin (BCG), offers potential advantages as a live, innately immunogenic vaccine vehicle for the expression and delivery of protective recombinant antigens (Stover, C.K., V.F. de la Cruz, T.R. Fuerst, J.E. Burlein, L.A. Benson, L.T. Bennett, G.P. Bansal, J.F. Young, M.H. Lee, G.F. Hatfull et al. 1991. Nature [Lond]. 351:456; Jacobs, W.R., Jr., S.B. Snapper, L. Lugosi and B.R. Bloom. 1990. Curr. Top. Microbiol. Immunol. 155:153; Jacobs, W.R., M. Tuckman, and B.R. Bloom. 1987. Nature [Lond.]. 327:532); but as an attenuated intracellular bacterium residing in macrophages, BCG would seem to be best suited for eliciting cellular responses and not humoral responses. Since bacterial lipoproteins are often among the most immunogenic of bacterial antigens, we tested whether BCG expression of a target antigen as a membrane-associated lipoprotein could enhance the potential for a recombinant BCG vaccine to elicit high-titered protective antibody responses to target antigens. Immunization of mice with recombinant BCG vaccines expressing the outer surface protein A (OspA) antigen of Borrelia burgdorferi as a membrane-associated lipoprotein resulted in protective antibody responses that were 100-1,000-fold higher than responses elicited by immunization with recombinant BCG expressing OspA cytoplasmically or as a secreted fusion protein. Furthermore, these improved antibody responses were observed in heterogeneous mouse strains that vary in their immune responsiveness to OspA and sensitivity to BCG growth. Thus, expression of protective antigens as chimeric membrane-associated lipoproteins on recombinant BCG may result in the generation of new candidate vaccines against Lyme borreliosis and other human or veterinary diseases where humoral immunity is the protective response.

SummaryWe used polyclonal antisera and monodonal antibodies (mAbs) to inhibit the growth of clonal populations of two strains of Borrelia burgdorferi, the Lyme disease agent, and thereby select for antibody-resistant mutants, mAbs were directed at the outer membrane proteins, OspA or OspB. Mutants resistant to the growth-inhibiting properties of the antibodies were present in the populations at frequencies ranging from 10 -s to 10 -2. The several escape variants that were examined were of four classes. Class I mutants were resistant to all mAbs; they lacked OspA and OspB and the linear plasmid that encodes them. Two other proteins were expressed in larger amounts in class I mutants; mAbs to these proteins inhibited the mutant but not the wild-type cells. Class II mutants were resistant to some but riot all mAbs; they had truncated OspA and/or OspB proteins. Class III mutants were resistant only to the selecting mAb; they had full-length Osp proteins that were not bound by the selecting antibody in Western blots. In two class III mutants resistant to different anti-OspA mAbs, missense mutations were demonstrated in the ospA genes. Class IV mutants were likewise resistant only to selecting antibody, but in this case the selecting antibody still bound in Western blots.

Function-oriented immunoassays, such as complement fixation and neutralization, are not commonly used in the study of the Lyme disease agent, Borrelia burgdorferi. To determine whether such assays provide information additional to matrix-based methods, such as ELISA, polyclonal antisera and monoclonal antibodies were examined for their abilities to agglutinate viable borreliae and inhibit their in vitro growth in microtiter plates. Different strains of B. burgdorferi in both high and low passage were examined, and the related spirochete Borrelia hermsii and antibodies to it served as controls. Agglutination and complement-independent inhibition of growth with polyclonal sera from rats, mice, and humans and with monoclonal antibodies to outer membrane proteins OspA and OspB was demonstrated. Growth inhibition was obtained with Fab fragments as well as with whole IgG molecules. In comparison with an ELISA using whole cells, the growth inhibition and agglutination assays were generally more specific.

Borrelia burgdorferi B31 cells lacking all linear plasmids or all but the 49-kb linear plasmid expressed the otherwise silent gene for the outer membrane protein OspC. In the first demonstration of a function for a circular plasmid of Borrelia spp., ospC was located on a 27-kb circular plasmid of B31.

A nonmotile mutant of Borrelia burgdorferi, the etiologic agent of Lyme disease, was isolated and characterized. The mutant was compared with the wild-type predecessor as well as with a motile back-revertant of the same genetic background. The mutant lacked, by morphologic, biochemical, and immunologic criteria, the major structural protein of flagella, flagellin. This mutation was not associated with major DNA rearrangements or with failure of transcription. An apparent consequence of a loss of flagella was reduced ability to penetrate human endothelial cell layers in vitro. In another assessment of functional significance, the flagella-less mutant was equal if not superior to flagella-bearing, isogenic isolates when examined in an enzyme-linked immunosorbent assay for anti-B. burgdorferi antibodies in the sera of Lyme disease patients. These studies of a mutant, the first among pathogenic Borrelia spp. to be characterized, indicate that the flagellum and motility it confers play a role in B. burgdorferi's invasion of human tissues. A flagellaless B. burgdorferi may be useful as the basis of a more specific immunoassay and a vaccine for protection against Lyme disease. (J. Clin. Invest. 1991. 88:82-92.)