Contributions of multiple binding sites and effector-independent binding to CodY-mediated regulation in Bacillus subtilis

Abstract

CodY is a branched-chain amino acid-responsive transcriptional regulator that controls, directly or indirectly, the expression of more than 100 genes and operons in Bacillus subtilis. Using DNase I footprinting and gel-shift experiments, we identified two CodY-binding regions upstream of a B. subtilis gene (bcaP, previously known as yhdG) that encodes a transporter of branched-chain amino acids. Mutational analysis revealed that both CodY-binding regions contribute to repression in vivo and do so independently of each other. Thus, a single CodY-binding site is apparently sufficient for substantial CodY-dependent regulation. By analyzing affinities of wild-type and mutant CodY-binding sites for CodY and their regulation by wild-type CodY and forms of CodY with various levels of activation by branched-chain amino acids, we concluded that unliganded CodY cannot repress transcription in vivo and that the level of endogenously produced effectors is sufficient for CodY-mediated regulation of promoters with stronger sites. Because the sites with higher affinity apparently respond to lower concentrations of CodY effectors and saturate faster as the concentrations of effectors increase, having two sites of binding with different affinities for CodY permits a promoter to respond to a wider range of intracellular concentrations of effectors.

FIG. 1.

Plasmid maps and sequence of the bcaP regulatory region. (A) Schematic maps of the bcaP inserts used to construct lacZ fusions. The location of the transcription start point is indicated by the bent arrow. CodY-binding motifs are shown as rectangles. The coordinates indicate the boundaries of different fusions with respect to the transcription start point. The repression ratio is the ratio of expression values for the corresponding fusions in the codY null mutant and wild-type strain in the medium containing 16 amino acids. (B) The sequence of the coding (nontemplate) strand of the bcaP regulatory region. The likely initiation codon, −10 and −35 promoter regions, transcription start site, and CodY-binding motifs are bold. The directions of transcription and translation are indicated by arrows. Sequences protected by CodY in DNase I footprinting experiments on the template strand of DNA are underlined. The boundaries of DNA fragments used to construct various lacZ fusions are indicated by vertical arrows.

FIG. 2.

Determination of the bcaP transcription start point and CodY-binding regions. (A) Primer extension analysis of the bcaP mRNA. Primer oBB102 annealing to the lacZ gene of the bcaP-lacZ fusion was extended with reverse transcriptase using as the template total RNA from fusion-containing strains BB2505 (codY⁺) (lane 1) and BB2548 (codY) (lane 2) grown in the medium containing 16 amino acids. The sequence of the template strand of pBB1419 determined from reaction mixtures primed with oBB102 is shown to the left. The apparent transcription start site of the bcaP gene is bold and marked by the +1 notation. A bent arrow indicates the direction of transcription. Additional bands observed in the primer extension lanes reflect unspecific binding of oBB102 to DNA and were present even when total RNA used for reverse transcription was isolated from a strain that did not contain the bcaP-lacZ fusion (data not shown). (B) DNase I footprinting analysis of CodY binding to the bcaP regulatory region. The bcaP283p⁺ DNA fragment labeled on the template strand was incubated with increasing amounts of purified CodY in the presence of ILV and GTP and then with DNase I. The sequence of the bcaP region was determined by using pBB1419 as the template and oBB102 as the primer and is shown to the right. The apparent transcription start site and direction of bcaP transcription are shown by the bent arrow. The protected areas are indicated by the vertical lines. CodY concentrations used (nM of monomer) are indicated above each lane. (C) Same as described for panel B. The gel was run longer to resolve the upstream CodY-binding sites. Site Ia is indicated by an arrow.

FIG. 3.

Independent binding of CodY to bcaP sites I and II. The bcaP283p⁺ (A and B), bcaP167p⁺ (A), and bcaP235p⁺ (B) DNA fragments were analyzed by DNase I footprinting as described for Fig. 2. The CodY concentrations used (nM of monomer) are indicated below each lane.

FIG. 4.

Gel-shift assay of CodY affinity for bcaP DNA fragments. Labeled DNA fragments were incubated with increasing amounts of purified CodY in the presence of ILV and GTP. CodY concentrations used (nM of monomer) are indicated below each lane. The arrow indicates a likely complex in which CodY is bound only to one binding site as opposed to two sites simultaneously.

FIG. 5.

Expression of bcaP283-lacZ fusions in codY mutants defective in interaction with ILV. Cells were grown in the medium containing 16 amino acids. Activity of each fusion was expressed as the percentage of activity of the same fusion in the codY null mutant. All strains are derivatives of strains BB2833 to BB2840.

FIG. 6.

Differential regulation of bcaP sites I and II in codY mutants defective in interaction with ILV. Cells were grown in the medium containing 16 amino acids. Activity of each fusion was expressed as the percentage of activity of the same fusion in the codY null mutant. All strains are derivatives of strains BB2833 to BB2840.

FIG. 7.

Interaction of wild-type CodY and CodY(F71R) with the bcaPp⁺ and bcaPp₈ versions of site I. Labeled bcaP235p⁺ and bcaP235p₈ DNA fragments were incubated with increasing amounts of purified wild-type CodY in the presence (A) or absence (B) of ILV and GTP. CodY concentrations used (nM of monomer) are indicated below each lane. (C and D) As described above but with CodY(F71R).