ExpR, a LuxR homolog of Erwinia carotovora subsp. carotovora, activates transcription of rsmA, which specifies a global regulatory RNA-binding protein

Abstract

N-acyl homoserine lactone (AHL) is required by Erwinia carotovora subspecies for the expression of various traits, including extracellular enzyme and protein production and pathogenicity. Previous studies with E. carotovora subsp. carotovora have shown that AHL deficiency causes the production of high levels of RsmA, an RNA binding protein that functions as a global negative regulator of extracellular enzymes and proteins and secondary metabolites (Rsm, regulator of secondary metabolites). We document here that ExpR, a putative AHL receptor belonging to the LuxR family of regulators, activates RsmA production. In the absence of AHL, an ExpR(+) E. carotovora subsp. carotovora strain compared to its ExpR(-) mutant, produces higher levels of rsmA RNA and better expresses an rsmA-lacZ transcriptional fusion. Moreover, the expression of the rsmA-lacZ fusion in Escherichia coli is much higher in the presence of expR(71) (the expR gene of E. carotovora subsp. carotovora strain Ecc71) than in its absence. We also show that purified preparation of MBP-ExpR(71) binds (MBP, maltose binding protein) rsmA DNA. By contrast, MBP-ExpR(71) does not bind ahlI (gene for AHL synthase), pel-1 (gene for pectate lyase), or rsmB (gene for regulatory RNA that binds RsmA), nor does ExpR(71) activate expression of these genes. These observations strongly suggest transcriptional activation of rsmA resulting from a direct and specific interaction between ExpR(71) and the rsmA promoter. Several lines of evidence establish that N-3-oxohexanoyl-L-homoserine lactone (3-oxo-C6-HL), the major AHL analog produced by E. carotovora subsp. carotovora strain Ecc71, inhibits ExpR(71)-mediated activation of rsmA expression. These findings for the first time establish that the expR effect in E. carotovora subsp. carotovora is channeled via RsmA, a posttranscriptional regulator of E. carotovora subspecies, and AHL neutralizes this ExpR effect.

FIG. 1.

(A) RLUs produced by HPLC fractions of crude AHL extract from Ecc71 in E. coli strain VJS533 harboring pHV200I. (B) TLC analysis of AHLs. Lane 1, mixture of synthetic 3-oxo-C6-HL and 3-oxo-C8-HL; lane 2, crude AHL extract of Ecc71; and lane 3, HPLC purified 71AHL (ca. 7 nmol). (C) LCMS/MS scan for HPLC-fractionated 71AHL. (D) Agarose plate assays for Pel, Peh, Prt, and Cel activities of AC5094 (ExpR⁺ AhlI⁻). Thirty microliters of culture supernatant was applied in each well. (E and F) Northern blot analysis and Western blot analysis of rsmA or RsmA of AC5094. Each lane contained 10 μg of total RNA for Northern blot analysis, and 10 μg of total protein for Western blot analysis. (D, E, and F) Lane 1, in the absence of AHL (i.e., same volume of water was added); lane 2, in the presence of 71AHL (300 μl of 1 mM 71AHL added to a 6-ml culture to yield a final concentration of 50 μM). Bacteria were inoculated in minimal salts plus sucrose (0.5%, wt/vol) medium supplemented with or without 71AHL. Total RNAs and proteins were extracted after 5 h incubation at 28°C, and after 8 h incubation culture supernatants were collected by centrifugation (10,000 rpm, 10 min) for exoenzyme assays.

FIG. 2.

Overexpression and purification of MBP-ExpR₇₁ protein. Lanes 1 and 2, lysates of E. coli DH5α carrying the vector (pMAL-c2g) without or with 1 mM IPTG; lanes 3 and 4, lysates of DH5α carrying pAKC1220 (expR₇₁ in pMAL-c2g) without or with 1 mM IPTG; and lane 5, purified MBP-ExpR₇₁. Lanes 1 to 4 contained 5 μg of total bacterial proteins, and lane 5 contained 2 μg of purified MBP-ExpR₇₁. Samples were analyzed by SDS-PAGE in a 10% (wt/vol) polyacrylamide gel.

FIG. 3.

Gel mobility shift assays for binding of purified MBP-ExpR₇₁ protein to (A) rsmA and (B) pel-1, ahlI, and rsmB DNAs. DNA fragments were end-labeled with [α-³²P]dATP. Two nanograms of probe DNAs was added in each reaction. The amounts of proteins, unlabeled DNA, and 71AHL or synthetic 3-oxo-C6-HL used in each reaction are indicated on the top of each figure.

FIG. 4.

(A) Northern blot analysis and (B) Western blot analysis of E. coli strain DH5α carrying pCL1920 or pAKC936 (expR₇₁) with rsmA plasmids from SCC3193 (pAKC1204), Eca12 (pAKC1205), EC153 (pAKC1206), and Ecc71 (pAKC875). Total RNAs and proteins were extracted from bacteria grown at 28°C in LB medium supplemented with spectinomycin and tetracycline to a Klett value of ca. 150. For Northern blot analysis, each lane contained 10 μg of total RNA, and for Western blot analysis, each lane contained 5 μg of total bacterial protein. Equal loading of RNA was checked by hybridization of the blot with a probe corresponding to 16S rRNA (rDNA). Lane 1, DH5α carrying pLARF5 plus pCL1920 (cloning vectors); 2, DH5α carrying pCL1920 and rsmA⁺ plasmids; and 3, DH5α carrying expR₇₁⁺ (pAKC936) and rsmA⁺ plasmids.

FIG. 5.

(A) Northern blot analysis of rsmA and (B) Western blot analysis of RsmA of E. carotovora subsp. carotovora strains. Lane 1, AC5006 (AhlI⁺ ExpR⁺); 2, AC5098 (AhlI⁺ ExpR⁻); 3, AC5091 (AhlI⁻ ExpR⁺); and 4, AC5099 (AhlI⁻ ExpR⁻). Total RNAs and proteins were extracted from bacteria grown at 28°C in minimal salts plus sucrose (0.5%, wt/vol) and celery extract medium to a Klett value of ca. 200. Each lane contained 10 μg of total RNA for Northern blot analysis and 10 μg of total protein for Western blot analysis. Equal loading of RNA was checked by hybridization of the blot with a probe corresponding to 16S rRNA (rDNA). (C) β-Galactosidase assays of AC5006, AC5098, AC5091, and AC5099 carrying pMP220 (vector) or pAKC1100 (rsmA-lacZ fusion). Bacterial constructs were grown at 28°C in minimal salts medium plus sucrose and tetracycline to a Klett value of ca. 200 and harvested for the assays. Bars represent standard errors.

FIG. 6.

(A) Agarose plate assays for Pel, Peh, Prt, and Cel activities of AC5091 (AhlI⁻ ExpR⁺, lane 1) and AC5099 (AhlI⁻ ExpR⁻, lane 2). Fifty microliters of culture supernatant was applied in each well. (B) Northern blot analysis of pel-1, peh-1, and celV of AC5091 (1) and AC5099 (2). Each lane contained 10 μg of total RNA. Equal loading of RNA was checked by hybridization of the blot with a probe corresponding to 16S rRNA (rDNA). Bacteria were grown in minimal salts plus sucrose (0.5%, wt/vol) and celery extract medium at 28°C to a Klett value of ca. 200 for total RNAs extraction. Culture supernatants were used for exoenzyme assays.

FIG. 7.

A speculative model depicting the regulatory events in the expression of rsmA in E. carotovora subsp. carotovora. The basal level of rsmA expression (upper pathway) is controlled by several factors, including RpoS and RsmC. Increased pool of free ExpR but not ExpR-AHL complex (lower pathway) causes overproduction of RsmA, which in turn suppresses exoprotein and antibiotic production and pathogenicity by controlling mRNA stability.