PsrA, the Pseudomonas sigma regulator, controls regulators of epiphytic fitness, quorum-sensing signals, and plant interactions in Pseudomonas syringae pv. tomato strain DC3000

Abstract

Pseudomonas syringae pv. tomato strain DC3000, a pathogen of tomato and Arabidopsis, occurs as an epiphyte. It produces N-acyl homoserine lactones (AHLs) which apparently function as quorum-sensing signals. A Tn5 insertion mutant of DC3000, designated PsrA(-) (Psr is for Pseudomonas sigma regulator), overexpresses psyR (a LuxR-type regulator of psyI) and psyI (the gene for AHL synthase), and it produces a ca. 8-fold-higher level of AHL than does DC3000. The mutant is impaired in its ability to elicit the hypersensitive reaction and is attenuated in its virulence in tomato. These phenotypes correlate with reduced expression of hrpL, the gene for an alternate sigma factor, as well as several hrp and hop genes during early stages of incubation in a Hrp-inducing medium. PsrA also positively controls rpoS, the gene for an alternate sigma factor known to control various stress responses. By contrast, PsrA negatively regulates rsmA1, an RNA-binding protein gene known to function as negative regulator, and aefR, a tetR-like gene known to control AHL production and epiphytic fitness in P. syringae pv. syringae. Gel mobility shift assays and other lines of evidence demonstrate a direct interaction of PsrA protein with rpoS promoter DNA and aefR operator DNA. In addition, PsrA negatively autoregulates and binds the psrA operator. In an AefR(-) mutant, the expression of psyR and psyI and AHL production are lower than those in DC3000, the AefR(+) parent. In an RpoS(-) mutant, on the other hand, the levels of AHL and transcripts of psyR and psyI are much higher than those in the RpoS(+) parent, DC3000. We present evidence, albeit indirect, that the RpoS effect occurs via psyR. Thus, AefR positively regulates AHL production, whereas RpoS has a strong negative effect. We show that AefR and RpoS do not regulate PsrA and that the PsrA effect on AHL production is exerted via its cumulative, but independent, effects on both AefR and RpoS.

FIG. 1.

Southern blot analysis using the psrA⁺ DNA probe from P. syringae pv. tomato strain DC3000 under high-stringency conditions. Lane 1, P. syringae pv. tomato strain DC3000; lane 2, P. syringae pv. maculicola strain ES4326; lanes 3 to 6, P. syringae pv. syringae strains B3A, B456, B728a, and 301D, respectively; lane 7, P. syringae strain BR2R; lanes 8 to 10, P. syringae pv. phaseolicola strains 1448A, PDDC3019, and PM132, respectively; lane 11, P. corrugata strain 0782-6; lane 12, P. tabaci strain Rif5; lanes 13 and 14, P. syringae pv. glycinia strains Race 1 and Race 5, respectively; lanes 15 and 16, P. viridiflava strains SF312A and MI-4, respectively; lane 17, P. savastanoi strain 2009; lanes 18 and 19, P. aeruginosa strains PAO1 and PAO2006, respectively; and lane 20, P. fluorescens strain Pf7-14.

FIG. 2.

(A) Effect of disruption of PsrA on the elicitation of the HR by DC3000. Site 1, DC3000; site 2, PsrA⁻ mutant AC820. Leaf panels were infiltrated with 1 × 10⁶ CFU/ml of bacterial cells. (B) Northern blot analysis of hrpL, hrpZ, hrpA, hopP to hopK, and hopP to hopJ of DC3000 (lane 1) and AC820 (lane 2). (C) Disease symptoms caused by DC3000 (left) and AC820 (right) in tomato leaves. Leaves were dip inoculated in bacterial suspensions (2 × 10⁸ CFU/ml). The pictures were taken 7 days after inoculation. (D) Northern blot analysis of rsmA1 of DC3000 (lane 1) and AC820 (lane 2). Each lane contained 10 μg of total RNA.

FIG. 3.

Northern blot analysis of rpoS and aefR of DC3000 (lane 1) and its PsrA⁻ mutant AC820 (lane 2). Each lane contained 10 μg of total RNA.

FIG. 4.

(A) Primer extension analysis of rpoS, aefR, and psrA. Lane 1 represents the RNA sample from DC3000. For rpoS, 10 μg of total RNA was used, and for aefR and psrA, 40 μg of total RNA was used. The nucleotides on the left of each panel refer to the nucleotide sequence beyond the transcriptional start site. The asterisk indicates the residue at which transcription was initiated. (B) PsrA-binding sequences upstream of aefR, rpoS, and psrA based on sequence analysis. Numbers correspond to the nucleotide positions in relation to the transcriptional start sites.

FIG. 5.

Gel mobility shift assays for binding of purified MBP-PsrA protein to aefR, rpoS, and psrA DNAs. DNA fragments were end labeled with [α-³²P]dATP. Each reaction mixture contained 2 ng of labeled DNA probe. The amounts of protein and unlabeled DNA used in each reaction are indicated at the top.

FIG. 6.

(A) Northern blot analysis of psyI and psyR of DC3000 (lane 1) and its PsrA⁻ mutant AC820 (lane 2). Each lane contained 20 μg of total RNA. (B) Relative light units produced by spent cultures of DC3000 and AC820 in E. coli strain VJS533 harboring pHV200I. Bacterial cultures were grown in KB at 28°C to a Klett value of ca. 600 and used for assay. Error bars indicate standard deviations.

FIG. 7.

(A) Northern blot analysis of psyI and psyR of DC3000 (lane 1), its AefR⁻ mutant AC821 (lane 2), and RpoS⁻ mutant AC822 (lane 3). Each lane contained 20 μg of total RNA. (B). Relative light units produced by spent cultures of DC3000, AC821, and AC822 in E. coli strain VJS533 harboring pHV200I. Bacterial cultures were grown in KB at 28°C to a Klett value of ca. 600 and used for assay. Error bars indicate standard deviations.

FIG. 8.

Northern blot analysis of (A) rpoS of DC3000 (lane 1) and its AefR⁻ mutant AC821 (lane 2), (B) aefR of DC3000 (lane 1) and its RpoS⁻ mutant AC822 (lane 2), and (C) psrA of DC3000 (lane 1), its AefR⁻ mutant AC821 (lane 2), and RpoS⁻ mutant AC822 (lane 3).

FIG. 9.

A model depicting the regulatory effects of PsrA in P. syringae pv. tomato strain DC3000. PsrA positively controls transcription of rpoS and negatively regulates expression of aefR by binding to their promoter/operator. RpoS has a negative effect, whereas AefR has a positive effect, on AHL production and on transcript levels of psyI and psyR. RpoS and AefR do not regulate each other. Thus, the PsrA effect on AHL production is exerted via its cumulative, but independent, effects on both AefR and RpoS. PsrA effects on rpoS, aefR, psyI, and psyR transcript as well as AHL production are similar at different temperatures (i.e., 18°C and 28°C in KB medium) as well as in different media (i.e., in KB and MGY media). AHL production is autoregulated. The GacS/GacA system positively regulates the expression of rpoS and psyR as well as AHL production (8). The GacS/GacA system does not have a significant effect on psrA and aefR transcripts. PsrA negatively regulates expression of rsmA1, which, in turn, affects the transcript levels of hrpL (7). HrpL is required for expression of hrp, hop, and avr genes as well as virulence and elicitation of the HR (5, 13, 28, 50).