Pseudomonas aeruginosa uses a cyclic-di-GMP-regulated adhesin to reinforce the biofilm extracellular matrix

Abstract

Pseudomonas aeruginosa, the principal pathogen of cystic fibrosis patients, forms antibiotic-resistant biofilms promoting chronic colonization of the airways. The extracellular (EPS) matrix is a crucial component of biofilms that provides the community multiple benefits. Recent work suggests that the secondary messenger, cyclic-di-GMP, promotes biofilm formation. An analysis of factors specifically expressed in P. aeruginosa under conditions of elevated c-di-GMP, revealed functions involved in the production and maintenance of the biofilm extracellular matrix. We have characterized one of these components, encoded by the PA4625 gene, as a putative adhesin and designated it cdrA. CdrA shares structural similarities to extracellular adhesins that belong to two-partner secretion systems. The cdrA gene is in a two gene operon that also encodes a putative outer membrane transporter, CdrB. The cdrA gene encodes a 220 KDa protein that is predicted to be rod-shaped protein harbouring a beta-helix structural motif. Western analysis indicates that the CdrA is produced as a 220 kDa proprotein and processed to 150 kDa before secretion into the extracellular medium. We demonstrated that cdrAB expression is minimal in liquid culture, but is elevated in biofilm cultures. CdrAB expression was found to promote biofilm formation and auto-aggregation in liquid culture. Aggregation mediated by CdrA is dependent on the Psl polysaccharide and can be disrupted by adding mannose, a key structural component of Psl. Immunoprecipitation of Psl present in culture supernatants resulted in co-immunoprecipitation of CdrA, providing additional evidence that CdrA directly binds to Psl. A mutation in cdrA caused a decrease in biofilm biomass and resulted in the formation of biofilms exhibiting decreased structural integrity. Psl-specific lectin staining suggests that CdrA either cross-links Psl polysaccharide polymers and/or tethers Psl to the cells, resulting in increased biofilm structural stability. Thus, this study identifies a key protein structural component of the P. aeruginosa EPS matrix.

Fig. 1

A diagram of a tertiary structure model of CdrA. The model was constructed by the de novo modelling of six separate segments and reattaching these segments by overlapping regions of structures. A. The global model of CdrA predicts a β-helix dominated structure with several exposed loops containing α-helices. The arrow indicates a predicted integrin binding motif. B. The same model was colour-coded to represent a trinary confidence measure based on secondary structure. Green indicates high confidence, yellow indicates intermediate confidence and red indicates low confidence. C. The N-terminal domain of CdrA contains a predicted signal peptide (blue), a haemagglutination site (orange) and a putative sugar-binding domain (purple). D. An exposed loop of the CdrA model contains a predicted integrin binding motif (red). Amino acids of the integrin binding motif are highlighted in red.

Fig. 2

RT-PCR data of cdrA expression under conditions of high and low c-di-GMP. cdrA is highly expressed under conditions where c-di-GMP is elevated. Relative transcript levels for cdrA are shown for wild-type P. aeruginosa (PAO1), a wspF mutant (elevated c-di-GMP), PAO1 expressing a diguanylate cyclase, PA1120 (elevated c-di-GMP), a wspF mutant expressing a c-di-GMP degrading phosphodiesterase PA2133 (low c-di-GMP), a wspFpelApslBCD mutant harbouring a vector control (elevated c-di-GMP) and a wspFpelApslBCD mutant expressing a c-di-GMP degrading phosphodiesterase (low c-di-GMP). Data are shown for four biological replicates.

Fig. 3

Western blot analysis of CdrA. Polyclonal antisera was used to monitor CdrA expression. In the top panel, CdrA was detected in the cellular fractions and culture supernatants in strains engineered to overexpress CdrA. The detection of CdrA in supernatants was dependent upon CdrB. In the lower two panels, CdrA expression was only detected in planktonic cultures of a ΔwspF background, not PAO1.

Fig. 4

The effect of cdrAB expression on biofilm formation in a static microtitre dish assay. A. Biofilm formation as measured by crystal violet staining by a cdrAB overexpression strain. Open bars indicate control treatments without arabinose induction and grey bars indicate 0.2% arabinose-induced treatment. B. cdrA contributes to biofilm formation. Biofilm formation by wspF, wspFcdrA, wspFpelApslBCD and wspFpelApslBCDcdrA mutant strains were quantified by measuring crystal violet staining after 20 h of static growth in a microtitre dish assay.

Fig. 5

CdrA contributes to biofilm development under conditions of liquid flow. A. Top view and side view of PAO1 and PAO1ΔcdrA biofilms at 20× magnification. B. Top view and side view of PAO1 ΔwspF and PAO1 ΔwspFcdrA biofilms at 40× magnification. C. Relative transcript levels of cdrA for PAO1 cells grown planktonically for 24 h as compared with cells grown in a biofilm. D. Western blot analysis using CdrA antisera of PAO1 grown planktonically and as a biofilm for 24 h.

Fig. 6

Overexpression of cdrAB results in aggregation in liquid culture and small colonies on solid medium. A. Phase contrast micrographs of P. aeruginosa PAO1 harbouring the expression vector pBADcdrAB and pMJT-1. Bacteria were cultured in LB medium and 0.5% arabinose. B. Liquid cultures of P. aeruginosa PAO1 harbouring the expression vectors pBADcdrA, pBADcdrB, pBADcdrAB, and pMJT-1 cultured in LB medium and 1.0% arabinose. C. Colonies of PAO1 harbouring the expression vector pBADcdrAB cultivated on VBMM congo red agar plates supplemented with increasing concentrations of arabinose.

Fig. 7

CdrA-mediated aggregation is dependent on Psl and inhibited by the addition of mannose. A. Aggregation of wild-type PAO1 and mutant strains that no longer produce the Pel and/or Psl exopolysaccharides were evaluated after induction of cdrAB with 1% arabinose and 3 h of growth. B. Structure of the Psl polysaccharide. C. Inhibition of aggregation by the addition of sugars. Analysis of the sugar-binding specificity of cdrA was monitored indirectly by measuring the % relative aggregation of PAO1 pBADcdrAB strain in which cdrAB expression has been induced by addition of 1% arabinose.

Fig. 8

Psl and CdrA Co-immunoprecipitation analysis. This figure depicts a western analysis using CdrA antisera. The Co-IP eluants show what eluted off of Psl antisera coated beads after incubation with supernatants of selected strains. The input pool shows the amount of CdrA present in the supernatants that were incubated with the beads.

Fig. 9

Lectin staining reveals CdrA is required for Psl association with a developing biofilm. Biofilms produced by Pseudomonas aeruginosa PAO1 ΔwspF (A, side view; C, bottom-up view) and ΔwspFcdrA (B, side view; D, bottom-up view) were stained with lectin. Bacterial cells were stained with Syto9 (green) and Psl stained with HHA-Tritc labelled lectin (red).