Periplasmic flagellar export apparatus protein, FliH, is involved in post-transcriptional regulation of FlaB, motility and virulence of the relapsing fever spirochete Borrelia hermsii

Abstract

Spirochetes are bacteria characterized in part by rotating periplasmic flagella that impart their helical or flat-wave morphology and motility. While most other bacteria rely on a transcriptional cascade to regulate the expression of motility genes, spirochetes employ post-transcriptional mechanism(s) that are only partially known. In the present study, we characterize a spontaneous non-motile mutant of the relapsing fever spirochete Borrelia hermsii that was straight, non-motile and deficient in periplasmic flagella. We used next generation DNA sequencing of the mutant's genome, which when compared to the wild-type genome identified a 142 bp deletion in the chromosomal gene encoding the flagellar export apparatus protein FliH. Immunoblot and transcription analyses showed that the mutant phenotype was linked to the posttranscriptional deficiency in the synthesis of the major periplasmic flagellar filament core protein FlaB. Despite the lack of FlaB, the amount of FlaA produced by the fliH mutant was similar to the wild-type level. The turnover of the residual pool of FlaB produced by the fliH mutant was comparable to the wild-type spirochete. The non-motile mutant was not infectious in mice and its inoculation did not induce an antibody response. Trans-complementation of the mutant with an intact fliH gene restored the synthesis of FlaB, a normal morphology, motility and infectivity in mice. Therefore, we propose that the flagellar export apparatus protein regulates motility of B. hermsii at the post-transcriptional level by influencing the synthesis of FlaB.

Figure 1. Protein and immunoblot analyses of flagella deficient B. hermsii compared to other isolates of B. hermsii and B. burgdorferi.

(A–B) SDS-PAGE, Coomassie blue stain and immunoblots with an anti-FlaB monoclonal antibody H9724. Whole-cell lysates include the B. hermsii non-motile mutant (Mutant clone), wild type clone (WT clone), the parental strain (DAH), two additional B. hermsii isolates (FRO and HS1), and B. burgdorferi B-31. The non-motile mutant is deficient in the amount of the 39-kDa protein FlaB (arrow) (A). The anti-FlaB monoclonal antibody confirms the paucity of FlaB in the lysate of the mutant (arrow) (B). Molecular mass standards are shown on the left in kDa.

Figure 2. Electron microscopic (EM) analyses of parental wild-type B. hermsii and non-motile mutant.

(A–F) The morphology of the mutant is strikingly different than the wild-type B. hermsii. By Scanning EM, the wild-type cells show the typical flat-wave morphology (A) compared to the straight shape of the non-motile mutant (B). By Cryo-EM, the wild-type spirochetes show the normal shape and numerous periplasmic flagella (C) compared to the mutant that is straight with few or no periplasmic flagella (D). Cross-sections viewed by transmission EM also show numerous flagella in the periplasmic space of the parental wild-type spirochetes (E) compared to none in the example of the non-motile mutant of B. hermsii shown here (F). All scale bars represent 0.2 µm.

Figure 3. Electron microscope analysis of intact B. hermsii cells for the presence of periplasmic flagella.

(A–D) 3-dimensional tomographic renderings of B. hermsii wild-type and mutant cells. The wild-type cell has numerous intact axial filaments each delineated by different colors (A), while the three mutant cells show either no (B), one intact (C), or several irregular and truncated filaments. (D). Ortho-slice views (A′–D′) taken from the same respective samples show the presence of basal bodies (black arrowheads) including the mutant cell lacking axial filaments (B & B′). The scale bar represents 0.2 µm.

Figure 4. ORF diagram of the chromosomal gene fliH (BH0289), anti-FlaB immunoblot analysis and quantification of fliH mRNA using realtime-RT PCR analysis of wild-type B. hermsii, fliH mutant, fliH mutant pBhSV2 and complemented mutant pBhSV2::pflgB-fliH.

(A) Schematic alignment and genetic organization of the chromosomal regions surrounding fliH in B. hermsii. (B) B. hermsii wild-type (WT), fliH mutant, fliH mutant transformed with the empty vector pBhSV2, and fliH mutant complemented with pBhSV2::pflgB-fliH analyzed by immunoblot with anti-FlaB antibody. Coomassie blue staining and anti-GlpQ immunoblot are shown to demonstrate equal loading of the cell lysates. (C) Relative amounts of fliH mRNA normalized to glpQ mRNA in the WT, fliH mutant, fliH mutant pBhSV2 and fliH mutant pBhSV2::pflgB-fliH as determined by qRT-PCR analyses. *, ANOVA test P-value <0.05 versus WT.

Figure 5. Complementation of the fliH mutant with pBhSV2::pflgB-fliH restores the wild-type morphology and motility.

(A) Darkfield microscopy analysis of B. hermsii wild-type (WT) (i), fliH mutant (ii), fliH mutant transformed with empty vector pBhSV2 (iii) and fliH mutant complemented with pBhSV2::pflgB-fliH (iv). (B) Representative images of quantitative swimming plate assays show that the wild-type B. hermsii and fliH complemented mutant pBhSV2::pflgB-fliH swam equally whereas the fliH mutant swam significantly less (B & C). Results are the average of 3 independent biological experiments. The asterisk indicates that the swim diameter of the mutant was significantly less than those of the WT and complemented spirochetes (p<.05). Scale bars represent 10 µm.

Figure 6. Comparison of flaB gene transcription and synthesis of flagellar proteins in B. hermsii wild-type (WT) and fliH mutant.

(A) Fold change of flaB transcript normalized to glpQ transcript was obtained from 3 individual cultures of WT and fliH mutant spirochetes. (B) Relative amounts of FlaB protein were normalized to GlpQ protein measured by semi-quantitative immunoblot using 3 lysates of the WT and mutant B. hermsii. (C) Immunoblot analysis of purified periplasmic flagella from wild-type B. hermsii (lanes a), lysates of wild-type B. hermsii (lanes b) and the fliH mutant (lanes c) probed with anti-FlaA antibody (left panel) and anti- FlaB (H9724) antibody (right panel). Molecular mass standards are shown on the left in kDa. (D) Comparative anti-FlgE and anti-FliI immunoblot analyses of WT, fliH mutant, fliH mutant pBhSV2 and fliH mutant pBhSV2:: pflgB-fliH spirochetes.

Figure 7. Stability of FlaB in wild-type (WT) B. hermsii and the fliH mutant.

Lysates from spirochetes treated with 100 µg/ml of spectinomycin were analyzed by immunoblot using anti-FlaB and anti-GlpQ antibodies at the different time points shown.

Figure 8. FliH is required for B. hermsii infectivity in mice.

Kinetics of spirochetemia with wild-type B. hermsii (A), B. hermsii fliH mutant (B), B. hermsii fliH mutant complemented with pBhSV2::pflgB-fliH (C). Each strain was inoculated intraperitoneally into groups of 4 mice and spirochetemia was monitored daily for 14 days. Each graph represents the spirochetemia determined for one mouse. The wild-type and complemented B. hermsii produced primary spirochetemias and a relapse (A & C) while the FliH mutant spirochetes produced no detectable spirochetemia (B).

Figure 9. Mice inoculated with fliH mutant do not seroconvert to B. hermsii.

Serum samples were examined by immunoblot for antibodies to wild-type B. hermsii in mice 8 weeks after inoculation with wild-type (WT) B. hermsii (Mice #1 and #2), the fliH mutant (Mice #3 and #4), and the fliH mutant complemented with pBhSV2::pflgB-fliH (Mice #5 and #6). Representative results are shown for two of the four mice in each group and demonstrate the complete lack of antibody detectable in the mutant-infected mice. Molecular mass standards are shown on the left in kDa.