The RpoS Gatekeeper in Borrelia burgdorferi: An Invariant Regulatory Scheme That Promotes Spirochete Persistence in Reservoir Hosts and Niche Diversity

Abstract

Maintenance of Borrelia burgdorferi within its enzootic cycle requires a complex regulatory pathway involving the alternative σ factors RpoN and RpoS and two ancillary trans-acting factors, BosR and Rrp2. Activation of this pathway occurs within ticks during the nymphal blood meal when RpoS, the effector σ factor, transcribes genes required for tick transmission and mammalian infection. RpoS also exerts a 'gatekeeper' function by repressing σ⁷⁰-dependent tick phase genes (e.g., ospA, lp6.6). Herein, we undertook a broad examination of RpoS functionality throughout the enzootic cycle, beginning with modeling to confirm that this alternative σ factor is a 'genuine' RpoS homolog. Using a novel dual color reporter system, we established at the single spirochete level that ospA is expressed in nymphal midguts throughout transmission and is not downregulated until spirochetes have been transmitted to a naïve host. Although it is well established that rpoS/RpoS is expressed throughout infection, its requirement for persistent infection has not been demonstrated. Plasmid retention studies using a trans-complemented ΔrpoS mutant demonstrated that (i) RpoS is required for maximal fitness throughout the mammalian phase and (ii) RpoS represses tick phase genes until spirochetes are acquired by a naïve vector. By transposon mutant screening, we established that bba34/oppA5, the only OppA oligopeptide-binding protein controlled by RpoS, is a bona fide persistence gene. Lastly, comparison of the strain 297 and B31 RpoS DMC regulons identified two cohorts of RpoS-regulated genes. The first consists of highly conserved syntenic genes that are similarly regulated by RpoS in both strains and likely required for maintenance of B. burgdorferi sensu stricto strains in the wild. The second includes RpoS-regulated plasmid-encoded variable surface lipoproteins ospC, dbpA and members of the ospE/ospF/elp, mlp, revA, and Pfam54 paralogous gene families, all of which have evolved via inter- and intra-strain recombination. Thus, while the RpoN/RpoS pathway regulates a 'core' group of orthologous genes, diversity within RpoS regulons of different strains could be an important determinant of reservoir host range as well as spirochete virulence.

Keywords: Borrelia burgdorferi; Lyme disease; gene regulation; host adaptation; persistence; rpoS; sigma factor; vector borne disease.

FIGURE 1

RpoS_Bb (BB0771) is a divergent RpoS that retains residues and structural features important for RpoS function and promoter selectivity. (A) Overlay of RpoS_Ec structure (PDB: 5IPL-F) (Liu et al., 2016) and RpoS_Bb homology model. Indicated are conserved isoleucine (I128 and I79 in RpoS_Ec and RpoS_Bb, respectively) (Iwase et al., 2017) and positively charged residues (K173 in RpoS_Ec and R121 and K122 in RpoS_Bb) (Barne et al., 1997) known or predicted to be important for promoter selectivity by RpoS. The C-terminal domain (CTD, residues 315–330 in RpoS_Ec) conserved in RpoS homologs from γ-proteobacteria is indicated (Liu et al., 2016). (B) Functional subdomains in the RpoS_Bb homology model were defined according to the corresponding regions in RpoS_Ec (Lonetto et al., 1992) using the aligned structures in panel A. (C–F) Electrostatic distribution in RpoS_Ec (C), RpoS_Bb homology model (D), RpoD_Ec (PDB: 6CY9) (E), and RpoD_Bb (BB0712) homology models (F). The negatively charged glutamic acid residue (E458 in RpoD_Ec and E474 in RpoD_Bb) in σ region 3.1 used to distinguish between RpoD (σ⁷⁰) housekeeping and RpoS σ factors (Barne et al., 1997) is indicated.

FIGURE 2

ospA is expressed throughout the entire tick phase of the enzootic cycle. (A) Cartoon depicting the tandem insertion of a PospA-gfp/PflgB tdTomato “dual color” cassette into the cp26 bbb21-22 intergenic region (Kawabata et al., 2004); the resulting B. burgdorferi strain was designated BbP1981. The upstream region used to monitor expression of ospA (PospA) includes sequences required for RpoS-mediated repression of the ospA in mammals (Grove et al., 2017). (B) Representative quad-plots for temperature-shifted in vitro PospA-gfp +/tdTomato+ double-positive spirochetes. Numbers indicate the percentage of positive events in each quadrant. (C) Representative epifluorescence images of the PospA-gfp/PflgB-tdTomato dual color strain (BbP1981) following temperature-shift at 37°C in vitro; cultivation in DMCs; in I. scapularis larvae fed to repletion on BbP1981-infected C3H/HeJ mice; in unfed nymphs; in BbP1981-infected nymphs fed for 72–96 h on a naïve C3H/HeJ mice (Transmission Nymphs); and in naïve nymphs fed for ∼36–48 h on C3H/HeJ mice 4 weeks post-infection (Acquisition Nymphs). Scale bar, 20 μm.

FIGURE 3

Repression regulation of ospA in mice is not complete until > 72 h after syringe-inoculation and continues throughout the mammalian phase. (A) Representative composite two-photon microscopy images of ear tissues from Myd88^–/– mice 24 and 72 h after needle-inoculation with either PospA-gfp (BbP1981) or PflaB-gfp (BbP1286). Arrowheads indicate intact GFP+ spirochetes. Three-dimensional z-stack images were rendered using Volocity software from images of sequential x, y planes taken at different levels. Hair follicles and dermal collagen fibers fluoresce yellow-orange and blue, respectively, due to second-harmonic generation. A minimum of 20 fields per tissues were examined. Representative images for day 8 and 13 are shown in Supplementary Figure 4A. (B,C) Representative epifluorescence images of 7 μm cryosections of ear and patella collected from Myd88^–/– mice 36 days post-infection (p.i.) by needle with BbP1981 to detect PospA-gfp and PflgB-tdTomato (B) or BbP1286 to detect PflaB-gfp (C). Scale bar, 20 μm.

FIGURE 4

Reciprocal regulation of OspC and OspA by RpoS occurs only within mammals and requires an intact RpoN/RpoS pathway. (A) Whole-cell lysates from wild-type strain B31 5A4 (WT; BbP1781), isogenic ΔrpoS mutant (BbP1752) and complemented mutant (comp; BbP1754) strains following temperature-shift in vitro and cultivation in DMCs separated by SDS-PAGE and stained with silver. (B) Whole cell lysates of wild-type strain 297 and isogenic ΔrpoS, ΔrpoN, and ΔbosR mutants following temperature-shift from 23 to 37°C in vitro and cultivation in DMCs. Molecular weight markers (kDa) are shown at the left of each gel.

FIGURE 5

RpoS is required for persistence in mice. Retention of empty vector or rpoS-complementing plasmid by BbP1974 (WT+empty vector) and BbP1754 (rpoScomp) strains in tissues collected from infected mice 4, 8, 12, 16, and 20 weeks post-infection. Retention was assessed by PCR of individual colonies obtained by semi-solid phase plating of (A) murine ear and skin minimally cultured in BSK-II without antibiotic and (B) larvae fed to repletion on infected C3H/HeJ mice at 4 week intervals beginning 4 weeks post-infection. Columns represent the average of at least 10 colonies per cultured tissue per time point. Numbers on x-axis indicate the number of weeks post-infection. p-Values for pairwise comparisons (WT+empty vector and rpoScomp at the same time point) were determined using a two-tailed t-test; ^∗p ≤ 0.05.

FIGURE 6

Loss of the rpoS-complementing plasmid by ΔrpoS B. burgdorferi is associated with production of antibodies against OspA and BBA62 in mice. ELISA assay performed using sera collected from individual C3H/HeJ mice and purified recombinant proteins for tick-phase gene products (OspA and BBA62) repressed by RpoS in mammals (A) and genes products upregulated (OspC and DbpA) or unaffected (FlaB) by RpoS (B). Sera was collected at the designated time points (in weeks) following syringe-inoculation with either WT+empty vector (BbP1974) or rpoScomp (BbP1754). Columns represent the average and standard error of the mean from three mice per strain, per time point (4- to 20-weeks). p-Values for pairwise comparisons (WT vs. rpoScomp) were determined using a two-tailed t-test. ^∗p ≤ 0.05; ^∗∗p ≤ 0.001; ^∗∗∗p ≤ 0.0001.

FIGURE 7

Characterization of bba34 transposon mutant and complemented strains in vitro. (A) Cartoon schematic of wild-type (WT), bba34 transposon mutant (Δa34tn) and allelic-exchange strategy used to generate bba34 complement (bba34comp). Arrows shown above the WT loci indicate location of primers used to amplify upstream and downstream fragments in bba34comp allelic replacement construct. qRT-PCR (B) and immunoblot (C) analyses of WT, Δa34tn and a34comp strains following temperature-shift in vitro. Transcript levels for bba34 were normalized using a TaqMan-assay for flaB. Error bars indicate standard errors of the mean (3 biological replicates per strain, assayed in quadruplicate). p-Values for pairwise comparisons were determined using a two-tailed t-test. Whole cell lysates were separated on a 12.5% SDS polyacrylamide gel and then either stained with Coomassie blue or transferred to nitrocellulose for immunoblotting with polyclonal antisera against strain B31 OspC (generated as part of these studies), RpoS (Hyde et al., 2007) and FlaB (Caimano et al., 2005). Molecular markers (kDa) are shown on the left.