The EAL-domain protein FcsR regulates flagella, chemotaxis and type III secretion system in Pseudomonas aeruginosa by a phosphodiesterase independent mechanism

Abstract

The second messenger c-di-GMP regulates the switch between motile and sessile bacterial lifestyles. A general feature of c-di-GMP metabolism is the presence of a surprisingly large number of genes coding for diguanylate cyclases and phosphodiesterases, the enzymes responsible for its synthesis and degradation respectively. However, the physiological relevance of this apparent redundancy is not clear, emphasizing the need for investigating the functions of each of these enzymes. Here we focused on the phosphodiesterase PA2133 from Pseudomonas aeruginosa, an important opportunistic pathogen. We phenotypically characterized P. aeruginosa strain K overexpressing PA2133 or its inactive mutant. We showed that biofilm formation and motility are severely impaired by overexpression of PA2133. Our quantitative proteomic approach applied to the membrane and exoprotein fractions revealed that proteins involved in three processes were mostly affected: flagellar motility, type III secretion system and chemotaxis. While inhibition of biofilm formation can be ascribed to the phosphodiesterase activity of PA2133, down-regulation of flagellar, chemotaxis, and type III secretion system proteins is independent of this enzymatic activity. Based on these unexpected effects of PA2133, we propose to rename this gene product FcsR, for Flagellar, chemotaxis and type III secretion system Regulator.

Figure 1

PAK/pJN-FcsR is impaired for biofilm formation. (a) Pellicle formation assays. The figure shows the pellicles associated to the glass tubes stained with crystal violet after removal of the liquid culture. This assay was performed for PAK, PAK/pJN-FcsR and PAK/pJN-FcsRE₆₀A. The concentrations of L-arabinose used as inducer are indicated. (b) Biofilm formation assays. Biofilm formation on solid surface was evaluated using microtiter dish binding assay for PAK, PAK/pJN-FcsR and PAK/pJN-FcsRE₆₀A. L-arabinose concentrations used are indicated. Crystal violet staining retained on the biofilms was measured spectrophotometrically at 570 nm for quantification. *Indicates statistically significant difference determined by ANOVA and Tukey’s post hoc test (p < 0.05).

Figure 2

Quantitative analysis of proteins identified in membrane fractions of PAK and PAK/pJN-FcsR using shotgun approach. The figure shows the Volcano plot generated using the PatternLab for Proteomics TFold module. Each dot in the plot represents a protein identified at least in 5 replicates of all conditions, plotted according its p-value (log2 (p)) and fold change (log 2 (fold change)). Red dots represent proteins that do not satisfy neither fold change nor statistic criteria, and thus are considered unchanged between strains. Green dots satisfy fold change criterion but not statistical one. Orange dots correspond to low abundant proteins satisfying both fold change and q value criteria, but due to the low number of spectra they deserved further validation. Finally, blue dots correspond to proteins satisfying all statistical filters and represent the differentially expressed proteins between strains. The identity of each blue dot is shown in the figure (see also Table 2).

Figure 3

Comparative analysis of exoproteomes of PAK and PAK/pJN-FcsR by 2D DIGE. (a) Multiple images for one DIGE gel showing spot profiles for PAK (Cy3 labeled, green), PAK/pJN-FcsR (Cy5 labeled, red) and internal standard (Cy2 labeled, blue). (b) Overlay image of PAK (green) and PAK/pJN-FcsR (red) channels. Differential spots between strains (considering all four DIGE gels, p-values ≤ 0.05 and fold change ≥ 25%) are indicated and were further identified by MS analysis. Protein identities, fold changes and p values are listed in Supplementary Table S4. (c) 3D view of selected spots with significant normalized abundance fold changes: spot 1 CdrA (fold change 4.1) spot 4 flagellar hook-associated protein (fold change 11.2) and spot 7 flagellin (fold change 15.2).

Figure 4

Flagellar motility inhibition is independent of FcsR phosphodiesterase activity. (a) Swimming motility assays. Assays were performed in triplicates for PAK, PAK/pJN-FcsR and PAK/pJN-FcsRE₆₀A strains. The swimming areas were measured and the relative swimming area (% of wild type PAK) was plotted. The concentrations of L-arabinose used as inducer are indicated. *Indicates statistically significant difference determined by ANOVA and Tukey’s post hoc test, p < 0.05. (b) Effect of FcsR and FcsRE₆₀A on flagella assembly. Flagella isolated from PAK, PAK/pJN-FcsR (0.2% arabinose), PAK/pJN-FcsRE₆₀A (0.2% arabinose) and PAK-FliC^- strains were analyzed by SDS-PAGE followed by MALDI-TOF/TOF mass spectrometry. The major band of approximately 45 kDa detected in PAK (but absent in all other strains) was unambiguously identified as type A flagellin by mass spectrometry. (c) MALDI mass spectrum of tryptic peptides isolated from the 45 kDa gel band in (b). The m/z values observed can be assigned to FliC sequences (m/z 1053.4: sequence 126–135; m/z 1413.7: sequence 303–316; m/z 1615.7: sequence 2–16; m/z 2555.2: sequence 222–247; m/z 2613.3: sequence 67–91; m/z 2874.6: sequence 368–394).

Figure 5

Transmission electron microscopy of PAK and PAK/pJN-FcsR. Electron microscope images showing the presence of flagellum and type IV pili in PAK strain. Only pili were observed in PAK/pJN-FcsR. Flagellum is indicated by a black arrow. Type IV pili are indicated by dashed arrows. Scale bar: 500 nm.

Figure 6

TTSS regulation is independent of FcsR phosphodiesterase activity. Secreted extracts of PAK, PAK/pJN-FcsR and PAK/pJN-FcsRE₆₀A grown in Ca²⁺ depleted media were separated on SDS-PAGE and the presence of specific TTSS proteins was detected by Western blot. ExoS and PcrV were sequentially detected in the same blotting membrane. Original images are displayed in Supplementary Figure 7. (a) Cropped view of immunoblotting using anti-ExoS antibody. ExoS is only detected in PAK strain (b) Cropped view of immunoblotting using anti-PcrV antibody. PcrV is only detected in PAK strain.