Genetic analysis of flagellar-mediated surface sensing by Pseudomonas aeruginosa PA14

Abstract

Surface sensing is a key aspect of the early stage of biofilm formation. For Pseudomonas aeruginosa PA14, the type IV pili (T4P), the T4P alignment complex, and PilY1 were shown to play a key role in c-di-GMP signaling upon surface contact. The role of the flagellar machinery in surface sensing is less well understood for P. aeruginosa. Here, we show, consistent with findings from other groups, that a mutation in the gene encoding the flagellar hook protein (ΔflgK) or flagellin (ΔfliC) results in a strain that overproduces the Pel exopolysaccharide (EPS) with a concomitant increase in c-di-GMP levels. We use a candidate gene approach and genetic screens, combined with phenotypic assays, to identify key roles for the MotAB and MotCD stators and the FliG protein, a component of the flagellar switch complex, in stimulating the surface-dependent, increased c-di-GMP level noted for these flagellar mutants. These findings are consistent with previous studies showing a role for the stators in surface sensing. We also show that mutations in the genes coding for the DGCs SadC and RoeA, as well as SadB, a protein involved in early surface colonization, abrogate the increased c-d-GMP-related phenotypes of the ΔflgK mutant. Together, these data indicate that bacteria monitor the status of flagellar synthesis and function during surface sensing as a mechanism to trigger the biofilm program.

Importance: Understanding how the flagellum contributes to surface sensing for P. aeruginosa is key to elucidating the mechanisms of biofilm initiation by this important opportunistic pathogen. Here, we take advantage of the observation that mutations in the flagellar hook protein or flagellin enhance surface sensing. We exploit this phenotype to identify key players in this signaling pathway, a critical first step in understanding the mechanistic basis of flagellar-mediated surface sensing. Our findings establish a framework for the future study of flagellar-based surface sensing.

Keywords: Pseudomonas aeruginosa; biofilm; flagella; stators; surface sensing.

Fig 1

Mutations that eliminate stator production or impact stator occupancy suppress the ΔflgK mutant hyper-signaling phenotype. (A and D) Representative CR images of the indicated strains cultured on M8 plates solidified with 1% agar for 16 h at 37°C, followed by 4 days at room temperature. (B) Quantification of c-di-GMP levels in the indicated strains grown either on M8 swarm agar plates (left) or in M8 liquid (right) for extraction of nucleotides and measurement of c-di-GMP levels via mass spectrometry. (C and E) Quantification of c-di-GMP levels in the indicated strains grown on M8 agar swarm plates for 16 h prior to harvest for extraction of nucleotides and measurement of c-di-GMP levels via mass spectrometry. Experiments (B, C, and E) were performed in triplicate with three technical replicates per strain. Data in experiment in (B) were analyzed using an unpaired t-test. Data in experiments (C and E) were analyzed by ANOVA with Dunnett’s multiple comparisons test. Significant differences are shown for comparisons to the ΔflgK strain. ns, non-significant difference; *P < 0.05 and **P < 0.01.

Fig 2

Mutations that impact proton binding suppress hyper-signaling. (A) The top panel shows representative CR images of the indicated strains. The proton-binding aspartate residue MotB-D30, analogous to the E. coli MotB-D32, allele is mutated to alanine in the ΔflgK deletion strain. The middle panel shows the Western blot for the MotB-His₆ WT and D30A variant epitope-tagged proteins detected in lysate samples from surface-grown strains using an anti-His antibody (α-His). MotB-His₆ protein levels were quantified and normalized to a cross-reacting band (bottom panel, ctrl) detected in all samples and used as a loading control. Numbers below the middle panel show the mean and standard deviation (SD), in parentheses, from three independent experiments, normalized to the WT, which is set to 1.0. Statistical analysis was performed using ANOVA with Dunnett’s test for multiple comparisons. Significant differences are shown for comparisons to the ΔflgK motB⁺-His₆ strain, with *P < 0.05 and ****P < 0.0001. (B). The top panel shows representative CR images of the indicated strains. The proton-binding aspartate residue of MotD-D23 is mutated to alanine in the ΔflgK deletion strain. The middle panel shows the MotD WT and D23A His₆-epitope tagged proteins detected and quantified as described in panel A. Significant differences are shown for comparisons to the ΔflgK motD⁺-His₆ strain. *P < 0.05; ***P < 0.0005. (C). Quantification of c-di-GMP for the indicated strains grown on M8 swarm plates for 16 h. Experiments were performed in triplicate with three technical replicates per strain and analyzed by ANOVA with Tukey’s post-test comparison. Significant differences are shown for comparisons either to the ΔflgK motB⁺::His strain or to the ΔflgK motD⁺::His strain as indicated. ns, non-significant difference; significant differences noted as follows: ***P < 0.001; ****P < 0.0001.

Fig 3

The DGCs SadC and RoeA are required for hyper-signaling in the flgK mutant. (A) Representative CR images of the indicated strains. (B) Quantification of c-di-GMP extracted from swarm-grown strains. Experiments were performed in triplicate with three technical replicates per strain and analyzed by ANOVA with Dunnett’s post-test comparison. Significant differences are shown for comparisons to the ΔflgK mutant; *P < 0.05, **P < 0.005.

Fig 4

Testing candidate genes for their impact on the ΔflgK mutant hyper-signaling phenotype. (A). Representative CR plate images of (A) the ΔfliL mutation in the WT and the ΔflgK mutant background; (B) impact of stator mutations, and sadC and roeA DGC mutations on the ΔfliF mutant phenotypes; (C) mutations in the fliG and fliMN genes and impact of the ΔmotABCD mutation in these mutant backgrounds; (D) the ΔflhF mutant and impact of the ΔmotABCD mutation in this mutant; (E) mutations in the fimV and fimW genes in the WT and ΔflgK mutant backgrounds; and (F) mutation of sadB in the WT and ΔflgK mutant backgrounds. All CR assays were performed on M8 medium solidified with 1% agar and incubated for 16 h at 37°C followed by an additional 4 days at room temperature.

Fig 5

Proposed model for flagellar-mediated impacts on biofilm-relevant surface signaling. (1) Surface contact is sensed as increased load on the flagellum, which transmits this signal by modulating stator incorporation. We propose that mutations in fliC and flgK leading to loss of the flagellar filament mimic this signaling process. (2) PilY1 acts as a mechanosensor at the T4P tip to detect surface contact and relay this outside-in signal via the minor pilins (including PilW and PilX) and the alignment complex (PilMNOP) to SadC (45, 48, 49, 80). B/T/U refers to extension (PilB) and retraction ATPases (PilT/U). (3) Assembly of a complete flagellar basal body inhibits production of c-di-GMP in favor of motility. Disruption of basal body assembly elicits Pel production in a stator-independent pathway. (4) The DGCs, SadC and RoeA, contribute to pools of c-di-GMP that repress motility and activate production of Pel EPS to promote biofilm formation. (5) The FlgZ-c-di-GMP complex facilitates FlgZ-MotC interaction to remove the MotCD stator from the motor, disabling surface motility and allowing for interaction of MotC with SadC to further stimulate c-di-GMP production in a positive feedback loop (36, 37).