Poly-N-acetylglucosamine mediates biofilm formation and antibiotic resistance in Actinobacillus pleuropneumoniae

Abstract

Most field isolates of the swine pathogen Actinobacillus pleuropneumoniae form tenacious biofilms on abiotic surfaces in vitro. We purified matrix polysaccharides from biofilms produced by A. pleuropneumoniae field isolates IA1 and IA5 (serotypes 1 and 5, respectively), and determined their chemical structures by using NMR spectroscopy. Both strains produced matrix polysaccharides consisting of linear chains of N-acetyl-D-glucosamine (GlcNAc) residues in beta(1,6) linkage (poly-beta-1,6-GlcNAc or PGA). A small percentage of the GlcNAc residues in each polysaccharide were N-deacetylated. These structures were nearly identical to those of biofilm matrix polysaccharides produced by Escherichia coli, Staphylococcus aureus and Staphylococcus epidermidis. PCR analyses indicated that a gene encoding the PGA-specific glycoside transferase enzyme PgaC was present on the chromosome of 15 out of 15 A. pleuropneumoniae reference strains (serotypes 1-12) and 76 out of 77 A. pleuropneumoniae field isolates (serotypes 1, 5 and 7). A pgaC mutant of strain IA5 failed to form biofilms in vitro, as did wild-type strains IA1 and IA5 when grown in broth supplemented with the PGA-hydrolyzing enzyme dispersin B. Treatment of IA5 biofilms with dispersin B rendered them more sensitive to killing by ampicillin. Our findings suggest that PGA functions as a major biofilm adhesin in A. pleuropneumoniae. Biofilm formation may have relevance to the colonization and pathogenesis of A. pleuropneumoniae in pigs.

Fig. 1

Elution profile of a crude extracellular extract of A. pleuropneumoniae IA1 biofilm cells on a Sephacryl S-300 column irrigated with water. Aliquots (200 μl) of each 5-ml fraction were assayed for neutral sugars (●, A₄₈₅) and aminosugars (○, A₅₃₀). Void (V_o) and total (V_t) volumes of the column are indicated with arrows.

Fig. 2

¹H NMR spectra of PGA from of A. pleuropneumoniae strains IA5 (a), A. pleuropneumoniae strain IA1 (b), and PGA from S. epidermidis RP62A (c). The spectrum in panel (c) is from [8]. Peak assignments are according to Mack et al. [9].

Fig. 3

Partial 500-MHz heteronuclear ¹H-¹³C chemical shift correlation (HSQC) spectrum of A. pleuropneumoniae IA1 PGA in D₂O at 25°C. Correlation peaks are labeled with residue identification of β-GlcNH₂ (A) and β-GlcNAc (B), as in [9].

Fig. 4

Growth and detachment of A. pleuropneumoniae biofilms in 96-well microtiter plates. (A) Growth of biofilm-forming strain IA5N and isogenic pgaC mutants EI1001 and EI1005 in MHB. (B) Growth of wild-type strains IA1 and IA5 in MHB containing 0 or 20 μg/ml of dispersin B. Biofilms in panels A and B were stained with crystal violet. (C) Detachment of IA5 biofilm colonies by dispersin B. A well containing a small number of biofilm colonies (dark spots) was aspirated and filled with 100 μL of dispersin B (20 μg/mL in PBS) and photographed at the indicated times.

Fig. 5

Killing of A. pleuropneumoniae biofilms by ampicillin. Biofilms were grown for 18 h in MHB and then for 4 additional h in MHB with 0, 10 or 100 μg/ml of ampicillin and 0 or 20 μg/ml of dispersin B. CFUs were determined by dilution plating. Values indicate the mean number of CFUs/tube. Error bars indicate range.