Borrelia burgdorferi B31 with and without outer membranes contained nearly identical amounts of outer surface proteins A and B. The majority of each immunogen also was localized intracelhularly by immunocryoultramicrotomy. These results are inconsistent with the widely held belief that outer surface proteins A and B are exclusively outer membrane proteins.Lyme disease and syphilis are infectious disorders caused by the spirochetal pathogens Borrelia burgdorferi and Treponema pallidum, respectively. These diseases share many clinical and microbiological features, including the abilities of their respective pathogens to persist for prolonged periods in individuals with high titers of specific antibodies (19,31). In the case of syphilis, efforts to elucidate this phenomenon revealed that the outer membrane of T. pallidum is a fragile phospholipid bilayer with a paucity of integral membrane proteins (23,25,35). The major membrane immunogens of T. pallidum, molecules formerly thought to be surface exposed (1,17,21), are now believed to be lipoproteins anchored by fatty acids to the periplasmic leaflet of the cytoplasmic membrane (8,22,23,28,33). The recognition that the outer membrane of B. burgdorferi is similarly labile (2) and therefore easily disrupted during experimental manipulation led us to consider the possibility that the major lipoprotein immunogens of B. burgdorferi, outer surface proteins (Osps) A and B (7), also have been incorrectly characterized as exclusively surface-exposed molecules (3,4,18). Interestingly, a reevaluation of published data provides some support for this contention. Periplasmic constitutents, namely, endoflagella (15) washing to remove this material could disrupt or even remove the outer membranes. We therefore focused our analyses on intact cells and protoplasmic cylinders, reasoning that obvious reductions of putative outer membrane proteins should be detectable in the cylinder fractions following selective removal of the outer membranes.Three preparations of B. burgdorferi B31 (high passage) were investigated. One consisted of viable organisms pelleted directly from BSK II medium at 13,700 x g for 10 min. In the other two, organisms were incubated on ice for 30 min with or without 0.1% Triton X-114; this incubation was followed by three successive centrifugations (each at 13,700 x g for 10 min at 4°C) and resuspensions (washes) in phosphate-buffered saline (PBS), pH 7.4. Portions of each preparation were removed for SDS-polyacrylamide gel electrophoresis (PAGE) or prepared for whole-mount and ultrathin-section electron microscopy.Nearly all of the organisms taken directly from BSK II medium were intact ( Fig. la and b). Repeated centrifugation and resuspension, manipulations comparable to those used in surface localization studies (3,4,18), disrupted many of the organisms (Fig. lc and ld); scattered membrane vesicles also were seen (data not shown). Outer membranes were uniformly absent from the detergent-incubated organisms, and membrane vesicles were not seen despite extensive ...