Serine Hydroxymethyltransferase ShrA (PA2444) Controls Rugose Small-Colony Variant Formation in Pseudomonas aeruginosa

Abstract

Pseudomonas aeruginosa causes many biofilm infections, and the rugose small-colony variants (RSCVs) of this bacterium are important for infection. We found here that inactivation of PA2444, which we determined to be a serine hydroxymethyltransferase (SHMT), leads to the RSCV phenotype of P. aeruginosa PA14. In addition, loss of PA2444 increases biofilm formation by two orders of magnitude, increases exopolysaccharide by 45-fold, and abolishes swarming. The RSCV phenotype is related to higher cyclic diguanylate concentrations due to increased activity of the Wsp chemosensory system, including diguanylate cyclase WspR. By characterizing the PA2444 enzyme in vitro, we determined the physiological function of PA2444 protein by relating it to S-adenosylmethionine (SAM) concentrations and methylation of a membrane bound methyl-accepting chemotaxis protein WspA. A whole transcriptome analysis also revealed PA2444 is related to the redox state of the cells, and the altered redox state was demonstrated by an increase in the intracellular NADH/NAD⁺ ratio. Hence, we provide a mechanism for how an enzyme of central metabolism controls the community behavior of the bacterium, and suggest the PA2444 protein should be named ShrA for serine hydroxymethyltransferase related to rugose colony formation.

Keywords: Pseudomonas aeruginosa; biofilm formation; rugose; serine hydroxymethyltransferase; small colony variants.

Figure 1

Mechanism of how ShrA controls c-di-GMP concentrations. (A) Proposed mechanism of how ShrA controls rugose morphology. The red box highlights the reaction scheme for SHMT ShrA. (B) Organization of the genes encoding ShrA and the glycine cleavage system. (C) Scheme for the Wsp chemosensory system. For the metabolites, THF is tetrahydrofolate, SAM is S-adenosylmethionine, SAH is S-adenosylhomocysteine and c-di-GMP is cyclic diguanylate; for the proteins, Gcv is the glycine cleavage system, MetF is a 5,10-methylenetetrahydrofolate reductase, MetH is a methionine synthase, MetK is a methionine methyltransferase, WspA is a membrane bound methyl-accepting chemotaxis protein, WspC is a methyltransferase, WspF is a methylesterase, WspD is a scaffold protein, WspE is a histidine kinase, and WspR is a diguanylate cyclase.

Figure 2

Inactivation of shrA leads to RSCV morphology. (A) Colony morphology on Congo-red plates after 2 days at 37°C showing the smaller size of the shrA mutant. (B) Colony morphology after 6 days at 25°C, after 3 days at 37°C, and after 3 days at 37°C with 30 mM KNO₃. (C) Morphology of gcvP2, gcvT2 and sdaA mutants in the shrA operon.

Figure 3

Inactivation of shrA increases biofilm formation in shake flasks and in microtitre plates. (A) The shrA mutant formed biofilms at the glass surface as indicated by the yellow arrow after overnight incubation in LB at 37°C with shaking at 250 rpm. Total biofilm formation (at the liquid/solid and air-liquid interfaces) (B), and bottom biofilm formation on the polystyrene plates (C) by PA14 and the shrA mutant after incubation in LB at 37°C without shaking. Six wells were used for each culture. Error bars indicate standard deviations from three independent cultures.

Figure 4

Inactivation of shrA increases biofilm formation in flow cells. Biofilms of P. aeruginosa PA14 and the shrA mutant were formed in flow cell chambers in 5% LB at 37°C. After 72 h of incubation, biofilms were stained with SYTO9 for 20 min in the dark. Random biofilm images were obtained using a confocal microscope, and the representative images shown were produced by IMARIS. Scale bar indicates 10 μm.

Figure 5

Inactivation of shrA increases EPS production and reduces motility. (A) EPS production by the Congo-red assay after incubation in LB for 24 h at 25°C or 16 h at 37°C. Error bars represent standard deviations from two independent cultures. (B) Swarming motility and (C) swimming motility of P. aeruginosa PA14 and the shrA mutant at 37°C after 24 h. Three plates were used for each culture and two independent cultures were used for each strain.

Figure 6

Inactivation of shrA increases c-di-GMP concentrations. Chromatography traces of (A) 10 μM synthetic c-di-GMP and nucleotide extracts from (B) P. aeruginosa PA14, (C) the shrA mutant, and (D) the shrA mutant spiked with 7.5 μM c-di-GMP. (E–H) spectra of the corresponding c-di-GMP peak of (A–D), respectively.

Figure 7

ShrA reduces biofilm formation. (A) Total biofilm formation and (B) bottom biofilm formation on polystyrene plates for production of ShrA in the shrA mutant via pMQ70-shrA, and (C) total biofilm formation on polystyrene plates for production of ShrA in P. aeruginosa PA14 via pMQ70-shrA. LB containing 0.05% arabinose at 37°C was used. (D) Total biofilm formation of P. aeruginosa PA14 on polystyrene plates for production of ShrA via pMJT1-shrA in LB containing 0.2% arabinose at 37°C for 24 h. Six wells were used for each culture and three independent cultures were used for each strain. Error bars represent the standard deviations.

Figure 8

ShrA produces primarily Ser not Gly via the HPLC-fluorometric assay. (A) HPLC standards for 10 μM of each amino acid. (B) Ser is the product of the forward reaction of Gly (20 to 30 mM) + (6R,S)-5,10-methylenetetrahydrofolate (2 to 3 mM). (C) Negative control (buffer, no ShrA) for the forward reaction: Gly + (6R,S)-5,10-methylenetetrahydrofolate was not converted to Ser. (D) Gly is the product of the backward reaction of Ser (4–6 mM) and (6R,S)-tetrahydrofolate (2–3 mM): Ser was converted to Gly. (E) Negative control (buffer, no ShrA) for the backward reaction: Ser + (6R,S)-tetrahydrofolate was not converted to Gly.