The flagellar-specific transcription factor, sigma28, is the Type III secretion chaperone for the flagellar-specific anti-sigma28 factor FlgM

Abstract

The sigma(28) protein is a member of the bacterial sigma(70)-family of transcription factors that directs RNA polymerase to flagellar late (class 3) promoters. The sigma(28) protein is regulated in response to flagellar assembly by the anti-sigma(28) factor FlgM. FlgM inhibits sigma(28)-dependent transcription of genes whose products are needed late in assembly until the flagellar basal motor structure, the hook-basal body (HBB), is constructed. A second function for the sigma(28) transcription factor has been discovered: sigma(28) facilitates the secretion of FlgM through the HBB, acting as the FlgM Type III secretion chaperone. Transcription-specific mutants in sigma(28) were isolated that remained competent for FlgM-facilitated secretion separating the transcription and secretion-facilitation activities of sigma (28). Conversely, we also describe the isolation of mutants in sigma(28) that are specific for FlgM-facilitated secretion. The data demonstrate that sigma(28) is the Type III secretion chaperone for its own anti-sigma factor FlgM. Thus, a novel role for a sigma(70)-family transcription factor is described.

Figure 1.

Known flagellar T3S-chaperones do not facilitate FlgM secretion. A graph showing the relative amount of intracellular FlgM (FlgM_IN—unshaded bars) compared with extracellular FlgM levels (FlgM_OUT—black bars). Secretion assays were performed as described in Materials and Methods. FlgM_IN is defined as the level of FlgM present in whole cell lysates of mid-log phase cultures, while FlgM_OUT is defined as the amount of FlgM detected in the supernatant of the same culture. The secretion assays were performed in a wild-type background (HBB⁺ fliC ⁺) using null mutants of all three known T3S-chaperones compared with the wild-type strain LT2 (TH437): ΔflgN = TH5937; ΔfliS = TH5737; ΔfliT = TH5831; ΔfliS ΔfliT = TH5935; ΔfliS ΔfliT ΔflgN = TH5999. All data, including error bars, are shown relative to LT2 FlgM_IN levels. The data shown are an average of three independent repeats of the secretion assays.

Figure 2.

FlgM secretion is dependent upon the σ²⁸ protein. A graph showing the FlgM secretion profiles for different σ²⁸ mutant alleles compared with σ²⁸⁺. All secretion assays were performed in a strain that had a HBB⁺ FliC⁻ Hin⁻ phenotype (σ²⁸⁺) where fliC expression is “ON” (Bonifield and Hughes 2003). The data shown are an average of three independent repeats and are represented as relative to FlgM_IN levels for σ²⁸⁺. Strains used are highlighted in Table 3 with a full description of each geno-type.

Figure 3.

The stabilities of FlgM and σ²⁸ are interdependent. The stabilities of σ²⁸ (solid lines) and FlgM (dashed lines) were followed after growth was stopped at mid-log (OD₆₀₀ = 0.6–0.8) by the addition of spectinomycin to inhibit protein synthesis (Aldridge et al. 2003). Stability assays were performed for three independent repeats. Average protein levels were calculated as relative values of the T₀ time point for each sample. σ²⁸ was more stable over time in the flgM⁺ strain (solid-line diamonds [strain TPA368]) when compared with the ΔflgM strain (solid line-squares [strain TPA378]). In contrast, FlgM was much more stable in the absence of σ²⁸ (cf. σ²⁸⁺ background, dashed-line diamonds [strain TPA368], with the Δσ²⁸ mutant, dashed-line triangles [TPA376]). A degree of the change in FlgM stability in all these backgrounds is due to secretion rather than stability. Average OD₆₀₀ values for all strains used over the 60 min plus the calculations used for the protein half-lives can be found in the Supplemental Material.

Figure 4.

σ²⁸ mutants isolated or used during this study. Circled residues: σ²⁸ mutants isolated previously shown biochemically to have wild-type H14D (σ²⁸*[H14D]) or altered FlgM binding properties V33E and V213E (σ²⁸*[V213E]) (Chadsey and Hughes 2001). Squares: σ²⁸⁻ alleles used to show that the two activities of σ²⁸ are independent—σ²⁸⁻[R91C L207P] shaded in gray; σ²⁸⁻[Y190C S209L] shaded in black. Hexagons: characterized σ²⁸* mutants isolated during the FlgM-secretion mutant screen—σ²⁸*[E193V], black hexagon; σ²⁸*[S226R], white hexagon; σ²⁸*[E209G, V221A], gray hexagons.

Figure 5.

Putative σ^28† mutants exhibit weak σ²⁸* activity. β-Galactosidase activity of a fliC-lacZ transcriptional fusion (fliC5050::MudJ) in various flagellar mutant backgrounds. The activities of σ²⁸ mutants were compared with σ²⁸⁺, flgM ⁻, σ²⁸*[H14D] (increased-stability σ²⁸ with wild-type FlgM-binding), and σ²⁸*[V213E] (severe FlgM binding defective σ²⁸) in strains with either a HBB⁺ (white bars) or HBB⁻ (black bars) phenotype. Surprisingly, even though σ²⁸*[V33E], σ²⁸*[G77W L135S], σ²⁸*[E193V], σ²⁸*[S226R], and σ²⁸*[E209G V221A] all exhibited lower σ²⁸ transcriptional activity for HBB⁺ strains compared with HBB⁺ σ²⁸⁺, they all showed a weak σ²⁸* phenotype in the HBB⁻ strains. In contrast, significant increases in fliC-lacZ transcription for both HBB⁺ and HBB⁻ strains were observed for the σ²⁸*[H14D] and σ²⁸*[V213E] mutants. A complete description of the genotypes of all HBB⁺ strains used in this analysis is to be found in Table 3. HBB⁻ strains were constructed by transduction of the flgG574::Tn10 allele.