Expression of the primary carbohydrate component of the Bordetella bronchiseptica biofilm matrix is dependent on growth phase but independent of Bvg regulation

Abstract

We previously showed that the Bvg virulence control system regulates biofilm formation in Bordetella bronchiseptica (Y. Irie, S. Mattoo, and M. H. Yuk, J. Bacteriol. 186:5692-5698, 2004). Analyses of the extracellular components of B. bronchiseptica biofilm matrix revealed that the major sugar component in the matrix was xylose, and linkage analysis indicated a majority of it to be in a 4-linked polymeric form. The production of xylose was independent of Bvg regulation but instead was dependent on bacterial growth phase. In addition, N-acetyl-glucosamine in the matrix was found to be important for the initial development of the biofilm. These results suggest that B. bronchiseptica biofilm formation is growth phase dependent in addition to being regulated by the Bvg virulence system.

FIG. 1.

Deconvolution micrograph of Bordetella bronchiseptica biofilm grown on a glass chamber slide stained with Alexa Fluor 350-conjugated lectin wheat germ agglutinin (WGA). Recombinant B. bronchiseptica expressing YFP is shown in yellow, and WGA is shown in blue. The central picture shows a horizontal section while two flanking pictures show vertical sections. The blue line and green line indicate positions of the y-z section and x-z section, respectively. Red lines indicate positions of the horizontal section.

FIG. 2.

Quantitative analysis of Bordetella bronchiseptica biofilm matrix with WGA ELLA. Bacteria were grown in 96-well plates for 24 h, and ELLA was performed as described in Materials and Methods. The Bvgⁱ phase-locked strain (RB53i) and the ΔcyaA mutant form biofilms, while the wild-type bacteria grown in Bvg⁺ phase (RB50) and Bvg⁺ phase-locked (RB53) strains do not (15). The strong binding of WGA to the wells containing the biofilm-forming strains indicates the presence of surface-associated N-acetyl-glucosamine residues within the biofilm matrix. The absence of WGA binding to the samples containing RB50 and RB53 indicates the absence of matrix material. OD₄₀₅, optical density at 405 nm.

FIG. 3.

Crystal violet quantitative assay of 6-h biofilm grown in 96-well plates precoated with dispersin B (DspB). Wells were precoated with 50 μg/ml of DspB and then inoculated with Bvgⁱ phase-locked B. bronchiseptica for growth for 6 h. Biofilm formation was significantly lower in the presence of DspB than the controls without the enzyme and heat-inactivated DspB. OD₅₉₅, optical density at 595 nm.

FIG. 4.

Caulobacter crescentus xylose bioassay. (A) The Caulobacter xylose bioassay strain was added to serially diluted xylose, glucose, and N-acetyl-glucosamine solution, and the GFP fluorescence read-outs were measured by a fluorescence microplate reader. Only xylose induced GFP expression in a concentration-dependent manner. (B) Supernatants of overnight cultures of B. bronchiseptica strains RB50 (wild type), RB53 (Bvg⁺ phase locked), RB53i (Bvgⁱ phase locked), and RB54 (Bvg⁻ phase locked) were assayed. Xylose production appears to be Bvg independent, as all supernatant induced GFP expression of the Caulobacter bioassay strain to similar levels. (C) Culture supernatants of RB50 (grown in Bvg⁺ phase) and RB53i (Bvgⁱ phase locked) were collected at different growth stages and assayed for xylose. The table at the bottom shows relative fluorescence units that have been normalized with B. bronchiseptica CFU at the time of supernatant collection, and the reading from the late log phase was adjusted to 1 individually for Bvg⁺ readings and Bvgⁱ readings. The relative fluorescence unit/CFU values of early stationary phase, mid-stationary phase, and late stationary phase represent the fold increases of fluorescence units/B. bronchiseptica CFU compared to the late log values. Most of the xylose was produced in the stationary phase.