Since the discovery of Borrelia recurrentis as a cause of relapsing fever (1), biologists have been intrigued by the ability of the organism to cause repeated attacks of disease in man. Early investigators attempted to elucidate the relapse phenomenon by comparing the serologic responses of recovered animals with original and relapse populations of spirochetes. Most found that peak antibody titers to relapse spirochetes subsequently appeared in the order in which the spirochete populations were isolated (2). This finding suggested that the surface antigens of the organism had been changed and that growth of a modified population was responsible for each relapse. However, this concept was not uniformly accepted. Some investigators could not detect an immune response with any serologic test then in use. As explained in a review by Schuhardt (3), there were many reasons for controversy about early theories on the mechanisms of relapse. Many strains studied had been grown by repeated passage in animals without a complete record of factors that affected the relapse phenomenon. A variety of tests were used to measure serologic responses, and there was serious disagreement about which tests gave valid results. In addition, the phenomenon was different in the many animal hosts used to study the process.In recent years, little progress has been made in elucidating the true nature of the phenomenon. According to one popular concept (3, 4), new serotypes arise from borreliae that escape the specific antibody response and retreat into organs where new antigens are unmasked. These changed organisms penetrate the blood stream and cause a relapse that, in turn, is terminated when the host produces a specific serologic response to that relapse population. This process is repeated until death intervenes or a complete immunity is established.The development of a medium for cultivating borreliae (5, 6) and the demonstration by Coffey and Eveland (7,8) that individual organisms in a relapse population can be identified by specific fluoresceinated antiserum could have enabled new approaches to a study of the phenomenon. Investigating the relapse phenomenon in rats, they identified four serotypes of B. hermsii with rabbit and rat antisera conjugated with fluorescein-isothiocyanate (FITC) 1 according to a modification of the method of Gordon et al. (9). Initially, we attempted to repeat their work with antisera prepared in rabbits against the original and three relapse populations of B. hermsii obtained from a single mouse. Borreliae in smears of mouse blood were stained with these antisera by indirect immunofluorescence, but organisms from one population could not be clearly distinguished from those of the other three. Furthermore, the variability *

The vector-borne bacterium Borrelia hermsii, a relapsing fever agent, switches gene expression of a surface protein between different antigenic variants, thereby causing sequential waves of immune escape within hosts and increasing the likelihood of transmission. Analogous programmed systems of antigenic variation occur in African trypanosomes and Plasmodium falciparum. In these examples, switch rates to individual variants differ over a wide range. We studied how B. hermsii determines switch rates in two experimental infections: one where variants were identified by specific antisera and one based on identification by DNA sequence. Unexpressed loci of variant antigens copy into a single expression site at rates determined by extragenic features of silent loci rather than similarity between coding sequences of variants at silent sites and the single expression site. Two elements, in particular, determine switch rates. One set of elements overlaps the 5 ends of the expressed gene and the silent loci; greater sequence identity between elements was associated with a higher switch rate. The second set of elements flanks the expression site on the 3 side and occurs at variable distances downstream from silent loci; the nearer an element to a silent locus, the greater the switch rate of that locus into the expression site. In combination, these two features of the genome provide a simple mechanism to modulate switch rate whereby silent loci form a hierarchy of switch rates into the expression site. Although the switching hierarchy causes changes in individual cells that are stochastic, ordering of variants within hosts is semipredictable.antibody ͉ Borrelia ͉ recombination ͉ relapsing fever ͉ vector-borne

The reappearance of borreliae in a patient's blood during a second, third, or fourth febrile crisis of relapsing fever has suggested to students of this disease that these spirochetes undergo antigenic variation (1-4). Meleney summarized the state of knowledge of this phenemenon in 1928 (4): "At the time of the crisis which terminates the attack of fever, there is rapid agglutination and destruction of the spirochetes with the subsequent formation of immune bodies in the blood. These substances are specific for the strain of spirochetes which was present during the preceding attack, but have no influence on the spirochetes of the succeeding relapse. The spirochetes of the relapse give rise, in turn, to immune substances which are specific for them but not for the spirochetes of the first attack." Schuhardt and Wilkerson (5) showed in 1951 that these antigenically distinct relapse strains appeared after inoculations of rats with single organisms of Borrelia turicatae. Coffey and Eveland (6) subsequently identified by immunofluorescence three novel serotypes of Borrelia hermsii in the blood of rats experiencing relapses. One of us (H. G. Stoenner) has further defined this phenomenon of antigenic variation by studying B. hermsii, an agent of relapsing fever in North America, in mice. The use of serotype-specific antisera permitted identification of 24 different serotypes among the progeny of a single organism inoculated into a mouse (7).We are attempting to identify and characterize the variable antigens of B. hermsii. Although there has been considerable interest among biologists in similar phenomena shown by the salivarian trypanosomes (8-12), relatively little is known of borrelial antigens, their locations in the cells, and the mechanism of antigenic variation. Using both polyclonal antisera and monoclonal antibodies, we have identified in whole cell lysates of B. hermsii an abundant protein that is serotype specific. Materials and MethodsOrganisms and Culture Conditions. The origin of the B. hermsii strain used in these studies has been described (13,14). It was designated strain HS1. Details of the strategy and methods used for identification and isolation of relapse serotypes is described elsewhere (7). In summary, HS1 was cloned by limiting dilution in Swiss mice of the Rocky Mountain Laboratories stock (RML). 1 Spirochetes were enumerated by dark-field microscopy (14). Preliminary experiments had shown that one borrelia inoculated intraperitoneally was sufficient to produce spiroche-1 Abbreviations used in this paper: BSA, bovine serum albumin; C, culture; FA, direct immunofluorescence assay; FCS, fetal calf serum; IFA, indirect immunofluorescence assay; NCP, nitrocellulose; PBS, phosphatebuffered saline, pH 7.4; PBS/Mg, phosphate-buffered saline with 5 mM MgCI2; RML, Rocky Mountain Laboratories; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Tris, tris-(hydroxymethyl)aminomethane; TSGAN, 50 mM Tris, pH 7.4, 150 mM NaCI, 5 mM EDTA, 0.25% gelatin, 0.05% sodium azide, and 0.05% Nonidet ...

No abstract