(A) The crossed-domain conformation of CviR:CL, with the two monomers colored orange and blue (Table S1). Importantly, although a portion of one of the interdomain linkers (residues 190-196; dashed line) is disordered in the crystal structure, the conformation of the other linker is supported by clear electron density. The CL molecules (Figure 1A) are shown using a space-filling representation. (B) Comparison of the x-ray structures of TraR bound to 3-oxo-C8-HSL and DNA (Zhang et al., 2002) and CviR bound to CL. (C) Surface representation of the LBD of CviR:CL. In the left panel, the region that interacts with the DBD of the other monomer is highlighted in blue. In the right panel, a subset of the LBD residues that contact the DBD are highlighted in red. (D) CviR:CL with the DNA recognition helices highlighted in red. See also Figure S2.

(A) Purified CviR or CviR M89S bound to the indicated AHL was included in DNA binding reactions with a fluorescein-labeled DNA containing the CviR′ binding site. Data are plotted as the change in fluorescence polarization resulting from CviR binding (ΔF) divided by fluorescence polarization in the absence of CviR (F_o). We note that using gel-shift assays, we previously (Swem et al., 2009) failed to detect the decreased DNA binding by the CviR:C10-HSL complex reported here. We attribute this discrepancy to our earlier use of much higher DNA concentrations, which had the effect of obscuring the relatively modest difference in affinity. (B) CviR-RNA polymerase interaction assessed by bacterial adenylate cyclase two-hybrid (BACTH) analysis. E. coli BTH101 was transformed with plasmids expressing the Bordetella pertussis adenylate cyclase T25 domain fused to CviR and the T18 domain fused to either CviR, the N-terminal domain of Chromobacterium violaceum RNA polymerase alpha subunit (αNTD), the C-terminal domain of C. violaceum RNA polymerase alpha subunit (αCTD), or region 4 of C. violaceum RNA polymerase sigma-70 (σ4). T25-zip and T18-zip are fusions to the constitutively interacting GCN4 leucine zipper. Strains were grown on LB agar with the addition of the indicated AHL at 10 μM final concentration. CviR is unstable and rapidly degraded in the absence of AHL, accounting for the absence of BACTH signal in plates lacking AHL. (C) Closeup view of the ligand-binding pockets of CviR:C6-HSL and CviR:C10-HSL. The ligand and Met 89 residues in the CviR:C6-HSL and CviR:C10-HSL complexes are colored cyan and gray, respectively. (D) Stereo view of the ligand-binding pockets of CviR bound to C6-HSL (cyan) and CL (gray), respectively. See also Figure S3.

(A) E. coli vioA-gfp-expressing CviR′ or CviR′ S89M/N77Y was grown in the presence of the indicated concentration of C6-HSL, C10-HSL, or CL. GFP fluorescence was measured and normalized to OD₆₀₀. Data are mean ± SD; n = 3. (B) E. coli vioA-gfp expressing CviR′ S89M/N77Y was grown in the indicated concentration of C10-HSL or CL (agonist), or in the presence of 1 μM C6-HSL plus the indicated concentration of C10-HSL or CL (antagonist). Samples were analyzed as in panel A. (C) The crossed-domain conformation of CviR′:C6-HSL (Table S1), with the two monomers colored green and yellow. The C6-HSL molecules are shown in magenta using a space-filling representation. (D) Comparison of the cross-domain conformations of CviR:CL (orange and blue) and CviR′:C6-HSL (green and yellow). The N-terminal LBDs were aligned in order to emphasize the different orientations of C-terminal DBDs. (E) The interactions between Met 253 and Met 257 side chains (cyan) and the cavity created when C6-HSL (magenta) binds in place of the native autoinducer is shown. The protein backbone is colored green and a surface representation for the LBD is gray. See also Figure S4.

(A) Agonist/antagonists studied in this work. (B) At high cell density, autoinducer (C6-HSL; circles) accumulates, binding to and stabilizing CviR. CviR:C6-HSL functions as a transcriptional activator of genes, including vioA, controlled by quorum sensing.

(A) Agonist activity was evaluated by growing an E. coli CviR reporter strain 7 h in the indicated AHL concentrations. Induction of vioA-gfp was measured by fluorimetry. Most error bars (SD, n = 3) are not visible because they are smaller than the symbols used. (B) Antagonist activity was evaluated by growing the reporter strain 7 h in 1 μM C6-HSL autoinducer plus the indicated concentrations of C8-HSL, C10-HSL, or CL. (C) Transcriptional activation by CviR was evaluated using in vitro transcription reactions containing RNA polymerase and α³²P-UTP. Radiolabeled transcripts corresponding to vioA or to the constitutively transcribed RNAI (control) were analyzed by urea-PAGE and phosphorimaging. (D) Data in panel C were quantified using ImageQuant and are displayed as mean ± SD; n = 4.

(A) E. coli vioA-gfp expressing CviR was grown in the indicated concentrations of C10-HSL or C10-HSL plus 1 μM C6-HSL. GFP fluorescence was measured and normalized to the optical density of the culture at 600 nm (OD₆₀₀). Data are mean ± SD; n = 3. (B) E. coli vioA-gfp expressing CviR M89F or CviR M89L was grown in the indicated concentrations of C10-HSL or C10-HSL plus 1 μM C6-HSL and analyzed as in panel A. (C) E. coli vioA-gfp expressing CviR M89A or CviR M89S was grown in the presence of the indicated concentrations of C6-HSL or C10-HSL and analyzed as in panel A. (D) CviR in vitro transcription. CviR or CviR M89S bound to C10-HSL were included in in vitro transcription reactions containing RNA polymerase and α³²P-UTP. Radiolabeled transcripts corresponding to vioA or the constitutively transcribed RNAI were analyzed by urea-PAGE and phosphorimaging. (E) Data in panel D were quantified using ImageQuant and are displayed as mean ± SD; n=4.