The Pseudomonas aeruginosa diguanylate cyclase GcbA, a homolog of P. fluorescens GcbA, promotes initial attachment to surfaces, but not biofilm formation, via regulation of motility

Abstract

Cyclic di-GMP is a conserved signaling molecule regulating the transitions between motile and sessile modes of growth in a variety of bacterial species. Recent evidence suggests that Pseudomonas species harbor separate intracellular pools of c-di-GMP to control different phenotypic outputs associated with motility, attachment, and biofilm formation, with multiple diguanylate cyclases (DGCs) playing distinct roles in these processes, yet little is known about the potential conservation of functional DGCs across Pseudomonas species. In the present study, we demonstrate that the P. aeruginosa homolog of the P. fluorescens DGC GcbA involved in promoting biofilm formation via regulation of swimming motility likewise synthesizes c-di-GMP to regulate surface attachment via modulation of motility, however, without affecting subsequent biofilm formation. P. aeruginosa GcbA was found to regulate flagellum-driven motility by suppressing flagellar reversal rates in a manner independent of viscosity, surface hardness, and polysaccharide production. P. fluorescens GcbA was found to be functional in P. aeruginosa and was capable of restoring phenotypes associated with inactivation of gcbA in P. aeruginosa to wild-type levels. Motility and attachment of a gcbA mutant strain could be restored to wild-type levels via overexpression of the small regulatory RNA RsmZ. Furthermore, epistasis analysis revealed that while both contribute to the regulation of initial surface attachment and flagellum-driven motility, GcbA and the phosphodiesterase DipA act within different signaling networks to regulate these processes. Our findings expand the complexity of c-di-GMP signaling in the regulation of the motile-sessile switch by providing yet another potential link to the Gac/Rsm network and suggesting that distinct c-di-GMP-modulating signaling pathways can regulate a single phenotypic output.

FIG 1

P. aeruginosa DGC impacts initial attachment to surfaces but not biofilm formation. (A) GcbA, a homolog of the previously characterized P. fluorescens DGC GcbA, is a PleD-like response regulator harboring conserved receiver (Rec) and DGC (GGDEF) domains. Numbers indicate amino acids corresponding to the start and end of the proteins and respective domains. (B) Confocal laser scanning microscopy (CLSM) images of PAO1 wild-type and ΔgcbA and ΔctpL mutant biofilm cells grown for 6 days under flowing conditions and stained with the LIVE/DEAD BacLight viability stain. Size bars, 100 μm. The CLSM was followed by COMSTAT analysis to evaluate biofilm biomass accumulation (C) and average and maximum thickness (D). (E) Evaluation of attachment to a polystyrene surface as determined by crystal violet (CV) staining following 6 or 24 h of growth under shaking conditions. Attachment is expressed as a percentage relative to that of the wild-type PAO1 control. (F) Growth curve of wild-type PAO1 and the isogenic ΔgcbA mutant strain. All assays were repeated at least three times, with at least 6 images acquired per CSLM analysis and at least 8 replicates used per attachment or growth curve assay. *, significantly different from PAO1 (P < 0.01).

FIG 2

gcbA is cotranscribed with ctpL, a chemotaxis transducer affecting initial surface attachment. (A) Genomic organization of gcbA and ctpL and RT-PCR results indicating cotranscription of gcbA and ctpL. P₁ and P₂ indicate the positions of primers PA4843delF2 and PA4844delR1, respectively (see Table S2 in the supplemental material), used to test cotranscription of gcbA and ctpL. cDNA, complementary DNA after reverse transcription; gDNA, genomic DNA used as a positive control. The absence (−) or presence (+) of reverse transcriptase is indicated. (B) Evaluation of attachment to a polystyrene surface by crystal violet (CV) staining following 6 h of growth under shaking conditions. Attachment is expressed as percentage relative to that of the wild-type PAO1 control. All assays were repeated at least three times. *, significantly different from PAO1 (P < 0.01).

FIG 3

GcbA suppresses swimming and swarming motilities in a manner independent of surface hardness via regulation of flagellar reversal rates. (A to C) The swarming and swimming motilities of the indicated strains were analyzed using 0.5% and 0.3% agar, respectively, following 24 h of growth. Additionally, the surface hardness dependence of GcbA effects on swarming was assessed by testing the motilities of the PAO1 wild-type and ΔgcbA mutant strains on 0.45, 0.5, and 0.55% agar. Motility assays were repeated at least three times, with a minimum of four plates used per strain-agar concentration combination. (D) Flagellar reversal rates in M63 medium with 3 or 15% Ficoll, representing swimming and swarming conditions, respectively, were measured as changes in direction of movement of cells. Rates are expressed as reversals per cell per minute. The assays were repeated at least three times, with five videos acquired per sample and 10 cells analyzed per video, for a total of 150 cells analyzed per strain-viscosity combination. *, significantly different from PAO1 or the respective vector control PAO1/pJN105 (P < 0.01).

FIG 4

GcbA mediates the transition from polar to longitudinal attachment. (A) Attachment of the PAO1 wild-type and ΔgcbA mutant strains to a glass surface following 6 h of growth under flowing conditions and of the ΔgcbA/pJN-gcbA strain following 1 or 6 h of growth, was assessed via bright-field microscopy, with images captured by using ProgRes CapturePro software. An example of reversibly or polarly attached cells is indicated with a dashed arrow, while that of irreversibly attached cells (adhered longitudinally along their entire length) is indicated with a solid arrow. (B) Total numbers of attached cells per 400 μm² were quantified using ImageJ software with the cell counter tool. (C) Numbers of polarly, reversibly attached cells and irreversibly attached cells were quantified using ImageJ software with the cell counter tool. The rate of polar attachment is reported as a percentage of total attached cells. The assays were repeated at least in triplicate, with a minimum of 10 images acquired per strain per replicate. *, significantly different from PAO1 (P < 0.01).

FIG 5

P. fluorescens gcbA rescues the P. aeruginosa gcbA mutant phenotypes. In order to test whether the P. fluorescens GcbA (GcbA-Pfl) is functional in P. aeruginosa and will rescue the P. aeruginosa ΔgcbA (ΔgcbA-Paer) phenotypes, gcbA-Pfl under the control of the arabinose-inducible P_BAD promoter within the pMQ72 vector (pGcbA-Pfl) or the pMQ72 empty vector control was introduced into the P. aeruginosa PAO1 and the gcbA-Paer mutant strains. The respective strains were tested for swarming (A) and swimming (B) motilities, attachment to glass under flowing conditions (C), and rates of polar attachment (D). All assays were repeated at least in triplicate. *, significantly different from PAO1 (P < 0.01).

FIG 6

GcbA-mediated suppression of motility is independent of EPS production. Effects of gcbA overexpression in the wild-type PAO1 and Δpel, Δpsl, and Δpel Δpsl EPS mutant strains on swarming (A and C) and swimming (B and D) motilities were analyzed using 0.5% and 0.3% agar, respectively, following 24 h of growth. All assays were repeated at least in triplicate. *, significantly different from the respective strain not overexpressing gcbA (P < 0.01).

FIG 7

GcbA is an active diguanylate cyclase. (A) Elution profiles of enzymatically produced c-di-GMP. Diguanylate cyclase assays were performed using 25 μM GTP and 1 μg of the total cell protein of indicated E. coli strains harboring either gcbA under the control of the arabinose-inducible P_BAD promoter in the pJN105 vector or the empty pN105 vector control. E. coli expressing the DGC PleD was used as a positive control. The reaction mixtures were analyzed by HPLC for the presence of c-di-GMP 0, 60, and 120 min postinitiation of DGC assays. Representative peaks corresponding to c-di-GMP are shown. (B) Wild-type PAO1 exhibits a hyperaggregative phenotype in liquid culture upon overexpression of gcbA. c-di-GMP levels in extracts obtained from cells of indicated P. aeruginosa strains grown planktonically (C) or as biofilms (D) were determined using HPLC analysis. c-di-GMP levels are shown as pmol per mg of total cell protein. The wild-type PAO1 strain harboring the empty vector pJN105 was used as a vector control. Error bars denote standard deviations. All assays were repeated at least in triplicate. *, significantly different from PAO1 biofilm cells and the vector control PAO1/pJN105 (P < 0.01).

FIG 8

GcbA and DipA regulate attachment and motility by distinct yet converging pathways. (A) Swarming (0.45, 0.50, and 0.55% agar) and swimming (0.3% agar) motilities of the wild-type PAO1 and ΔgcbA, ΔdipA, and ΔdipA ΔgcbA mutant strains were assessed following 24 h of growth. (B) Total numbers of attached cells per 400 μm² were quantified using ImageJ software with the cell counter tool. (C) Numbers of polarly attached cells were quantified using ImageJ software, and the rate of polar attachment is reported as the percentage of total attached cells. (D) Flagellar reversal rates in M63 medium with 3% or 15% Ficoll, representing swimming and swarming conditions, respectively, were measured as changes in direction of movement of cells. Rates are expressed as reversals per cell per minute. (E) c-di-GMP levels extracts obtained from cells of indicated P. aeruginosa strains grown planktonically were determined by HPLC analysis. c-di-GMP levels are shown as pmol per mg total cell protein. All assays were repeated at least in triplicate. *, significantly different from PAO1 (P < 0.01).

FIG 9

Multicopy expression of rsmZ restores attachment and motility of ΔgcbA cells to wild-type levels. (A) Evaluation of attachment to a polystyrene surface by crystal violet staining following 6 h of growth under shaking conditions. Attachment is expressed as a percentage relative to the wild-type PAO1 control. (B) Total numbers of attached cells per 400 μm² were quantified using ImageJ software with the cell counter tool. (C) Numbers of reversibly attached cells were quantified using ImageJ software, and the rate of polar attachment is reported as a percentage of total attached cells. (D) The swarming and swimming motilities of the indicated strains were analyzed using 0.5% and 0.3% agar, respectively, following 24 h of growth. All assays were repeated at least in triplicate. *, significantly different from PAO1 (P < 0.01).