Differential spectrum of mutations that activate the Escherichia coli bgl operon in an rpoS genetic background

Abstract

The bgl promoter is silent in wild-type Escherichia coli under standard laboratory conditions, and as a result, cells exhibit a beta-glucoside-negative (Bgl-) phenotype. Silencing is brought about by negative elements that flank the promoter and include DNA structural elements and sequences that interact with the nucleoid-associated protein H-NS. Mutations that confer a Bgl+ phenotype arise spontaneously at a detectable frequency. Transposition of DNA insertion elements within the regulatory locus, bglR, constitutes the major class of activating mutations identified in laboratory cultures. The rpoS-encoded sigmaS, the stationary-phase sigma factor, is involved in both physiological as well as genetic changes that occur in the cell under stationary-state conditions. In an attempt to see if the rpoS status of the cell influences the nature of the mutations that activate the bgl promoter, we analyzed spontaneously arising Bgl+ mutants in rpoS+ and rpoS genetic backgrounds. We show that the spectrum of activating mutations in rpoS cells is different from that in rpoS+ cells. Unlike rpoS+ cells, where insertions in bglR are the predominant activating mutations, mutations in hns make up the majority in rpoS cells. The physiological significance of these differences is discussed in the context of survival of natural populations of E. coli.

FIG. 1.

The bgl operon of E. coli. The region upstream of the structural genes is termed bglR. Activation of the operon occurs predominantly by insertions in bglR. Negative regulatory elements, such as the inverted repeat that can extrude into a cruciform and the H-NS binding region, ensure the silencing of the operon in wild-type cells. The catabolite gene activator protein-cyclic AMP (CAP-cAMP) binding site, present upstream of the promoter, overlaps with the H-NS binding site. BglG, which functions as an antiterminator at the two Rho-independent terminators, brings about salicin-inducible transcription of the bgl genes upon activation. PEP, phosphoenolpyruvate.

FIG. 2.

(A) Schematic representation of the molecular analysis of the bgl operon of wild-type cells and Bgl⁺ mutants. (a to d) Expected bands of genomic DNA digested with EcoRV. (d) The 1.2-kb fragment obtained by SspI digestion of the plasmid carrying the wild-type operon used as a probe. This detects two bands, a 2.3-kb downstream fragment (a) and an upstream fragment (b), which is 1.9 kb in the wild-type and noninsertionally activated operons. (c) This increases in size to 2.6, 3.1, or 3.2 kb upon activation of the operon by the insertion of IS1, IS5, or IS10, respectively. (e and f) Expected bands obtained in PCR analysis. Primers SM1 and SM2 amplify an ∼560-bp region in wild-type and noninsertionally activated operons (e). (f) Insertion of IS1, IS5, or IS10 results in a larger product, 1.3, 1.8, or 1.9 kb, respectively. (B) Representative Southern analysis of RV (lanes 1 to 8) and SM2 (lanes 9 to 15) papillae. Except for RVp3 (bglR::IS5) (lane 3) and RVp5 (wild type) (lane 5), all strains show an increase in size in the 1.9-kb band suggestive of IS1 insertion. All seven SM2 papillae show bands similar to that of the wild-type strain. The wild type, RV, and RV⁺ (bglR::IS1) are controls. PCR analysis of representative papillae of SM2 (C) and RV (D). SM2p1.18 is activated by IS5 (lane 2) and SM2p1.19 is activated by IS1 (lane 3), whereas SM2p1.17, 1.21, 1.22, 1.23, and 1.24 show products with sizes similar to that of the wild type (lanes 1, 4, 5, 6, and 7, respectively). RVp1.12 is activated by IS5 (lane 5) and RVp1.13 is activated by IS1 (lane 6), whereas RVp1.3, 1.6, 1.8, and 1.9 show products with sizes similar to that of the wild type (lanes 1, 2, 3, and 4, respectively). λ/D and λ/P are size markers. The wild type, RV, RV⁺ (bglR::IS1), and RVp3 (bglR::IS5) were used as controls in the PCR analysis.

FIG. 3.

Detection of bgl transcript levels from representative SM2 papillae, SM2p1.17, SM2p1.24, and SM2p1.29 (Bgl⁺) with the S1 nuclease protection assay as described previously (35). Wild-type MM1 (bglR⁰) and activated AE328 (bglR::IS1) are the controls. No transcript can be detected in MM1, but the SM2 papillae show transcript levels comparable to those of the insertionally activated strain, indicating that the Sal⁺ phenotype of the mutants is due to enhanced bgl transcription.

FIG. 4.

Southern analysis of the bglJ region of representative SM2 mutants showing an insertion of ∼1.4 kb. Genomic DNA of the mutants was digested with BamHI and probed with linearized pJL3. RV (Bgl⁻), which has wild-type bglJ, shows a band of ∼6 kb while SM2p1.17, 1.25, 1.26, 1.29, and 1.31 (Bgl⁺) show bands of ∼7.5 kb, indicating an insertion.