RpoS is not central to the general stress response in Borrelia burgdorferi but does control expression of one or more essential virulence determinants

Abstract

Borrelia burgdorferi, the Lyme disease spirochete, undergoes dramatic changes in antigenic composition as it cycles between its arthropod and mammalian hosts. A growing body of evidence suggests that these changes reflect, at least in part, the need for spirochetes to adapt to the physiological stresses imposed by abrupt changes in environmental conditions and nutrient availability. In many microorganisms, global responses are mediated by master regulators such as alternative sigma factors, with Escherichia coli RpoS (sigmaS) serving as a prototype. The importance of this transcriptional activator in other bacteria, coupled with the report by Hubner et al. (A. Hubner, X. Yang, D. M. Nolen, T. G. Popova, F. C. Cabello, and M. V. Norgard, Proc. Natl. Acad. Sci. USA 98:12724-12729, 2001) demonstrating that the borrelial RpoS ortholog controls expression of OspC and decorin-binding protein A (DbpA), prompted us to examine more closely the roles of RpoS-dependent and -independent differential gene expression in physiological adaptation by the Lyme disease spirochete. We observed that B. burgdorferi rpoS (rpoSBb) was induced following temperature shift and transcript levels were further enhanced by reduced pH (pH 6.8). Using quantitative real-time reverse transcription-PCR (RT-PCR), we demonstrated that, in contrast to its ortholog (rpoSEc) in Escherichia coli, rpoSBb was expressed at significant levels in B. burgdorferi throughout all phases of growth following temperature shift. By comparing a B. burgdorferi strain 297 rpoSBb mutant to its wild-type counterpart, we determined that RpoSBb was not required for survival following exposure to a wide range of environmental stresses (i.e., temperature shift, serum starvation, increased osmolality, reactive oxygen intermediates, and increased or reduced oxygen tension), although the mutant was more sensitive to extremes of pH. While B. burgdorferi strains lacking RpoS were able to survive within intraperitoneal dialysis membrane chambers at a level equivalent to that of the wild type, they were avirulent in mice. Lastly, RT-PCR analysis of the ospE-ospF-elp paralogous lipoprotein families complements earlier findings that many temperature-inducible borrelial loci are controlled in an RpoSBb-independent manner. Together, these data point to fundamental differences between the role(s) of RpoS in B. burgdorferi and that in E. coli. Rather than functioning as a master regulator, RpoSBb appears to serve as a stress-responsive activator of a subset of virulence determinants that, together with the RpoS-independent, differentially expressed regulon, encompass the spirochete's genetic programs required for mammalian host adaptation.

FIG. 1.

Expression kinetics of rpoS_Bb and RpoS-dependent loci in B. burgdorferi. (A) Growth curve indicating time points at which samples were taken from wild-type strain 297, following temperature shift from 23 to 37°C in BSK-H medium at either pH 7.5 (solid squares) or 6.8 (open squares), for subsequent RNA and protein analyses. (B) qRT-PCR analysis of rpoS_Bb. RNAs were isolated from spirochetes cultivated at 23°C and after temperature shift to 37°C in BSK-H medium at either pH 7.5 or 6.8 and after cultivation within DMCs. (C) Whole-cell lysates of wild-type B. burgdorferi strain 297 (∼10⁷ per lane) cultivated at 23°C and following temperature shift to 37°C in BSK-H medium at pH 7.5 were separated by SDS-PAGE and then silver stained. Molecular mass markers (in kilodaltons) are indicated. Sample numbering in panels B and C corresponds to similarly numbered points in panel A. wt, wild type.

FIG. 2.

Induction kinetics of RpoS-dependent promoters differ in B. burgdorferi and E. coli. Analysis of RpoS-dependent B. burgdorferi (P_ospC) and E. coli (P_osmY) promoters with a GFP reporter system in E. coli. Samples from wild-type (ZK126; squares) and rpoS_Ec mutant (ZK1000; triangles) E. coli transformed with the indicated reporter constructs analyzed by flow cytometry. The top panel depicts a representative growth curve for both wild-type and rpoS_Ec mutant isolates. Mean fluorescence intensities for P_ospC-gfp (middle panel) and P_osmY-gfp (bottom panel) reporter constructs were plotted against time on the representative graphs. Each graph represents the average of three trials.

FIG. 3.

Growth phenotypes of wild-type and rpoS_Bb mutant bacteria at increased temperatures and reduced pH. Growth curve analyses of wild-type (WT; solid squares) and rpoS_Bb mutant (AH200; open circles) B. burgdorferi strain 297 cultivated in BSK-H medium following temperature shift to 37°C at pH 7.5 (A) and 6.8 (B). The graph depicting the growth of the wild-type isolates is the same as that shown in Fig. 1A.

FIG. 4.

RpoS is not required for survival following exposure to environmental stress. Survival of B. burgdorferi wild-type strain 297 and rpoS_Bb mutant AH200 following cultivation in RPMI with (+) or without (−) 6% normal rabbit serum (A), under various oxygen tensions (standard microaerophilic, increased oxygen [CO₂], and anaerobic) (B), or following exposure to high osmolarity (1 M NaCl) (C), exogenous peroxide (0.5 or 1 mM H₂O₂) (D), or acidic pH (pH 6.0) (E). Percent survival was calculated based on no-exposure controls for each sample. Percent survival values represent the averages of two independent trials. The log rank and P values for the results shown in panel E, determined by the Kaplan-Meier method, were χ² = 3.809 and P = 0.051. wt, wild type.

FIG. 5.

ospE-ospF-elp paralogous loci utilize predominantly RpoS-independent pathways. (A) Analysis of wild-type (WT) and rpoS_Bb mutant (AH200) B. burgdorferi strain 297 samples cultivated at 23°C and following temperature shift to 37°C in BSK-H medium at pH 7.5 or 6.8. Whole-cell lysates (∼10⁷ lysates per lane) were separated by SDS-PAGE and either silver stained (OspC) or immunoblotted with sera directed against DbpA, OspF, OspE, or FlaB. (B) RT-PCR of wild-type (wt) strain 297 and AH200 following temperature shift to 37°C with primers specific for individual ospE-ospF-elp loci and flaB with (+) or without (−) reverse transcriptase. DNA controls were also performed with 297 genomic DNA (not shown). Molecular weight markers in kilobases are indicated to the left.