BigR, a transcriptional repressor from plant-associated bacteria, regulates an operon implicated in biofilm growth

Abstract

Xylella fastidiosa is a plant pathogen that colonizes the xylem vessels, causing vascular occlusion due to bacterial biofilm growth. However, little is known about the molecular mechanisms driving biofilm formation in Xylella-plant interactions. Here we show that BigR (for "biofilm growth-associated repressor") is a novel helix-turn-helix repressor that controls the transcription of an operon implicated in biofilm growth. This operon, which encodes BigR, membrane proteins, and an unusual beta-lactamase-like hydrolase (BLH), is restricted to a few plant-associated bacteria, and thus, we sought to understand its regulation and function in X. fastidiosa and Agrobacterium tumefaciens. BigR binds to a palindromic AT-rich element (the BigR box) in the Xylella and Agrobacterium blh promoters and strongly represses the transcription of the operon in these cells. The BigR box overlaps with two alternative -10 regions identified in the blh promoters, and mutations in this box significantly affected transcription, indicating that BigR competes with the RNA polymerase for the same promoter site. Although BigR is similar to members of the ArsR/SmtB family of regulators, our data suggest that, in contrast to the initial prediction, it does not act as a metal sensor. Increased activity of the BigR operon was observed in both Xylella and Agrobacterium biofilms. In addition, an A. tumefaciens bigR mutant showed constitutive expression of operon genes and increased biofilm formation on glass surfaces and tobacco roots, indicating that the operon may play a role in cell adherence or biofilm development.

FIG. 1.

BigR is homologous to prokaryotic HTH transcriptional factors and binds to upstream sequences (US) of its own operon. (A) Schematic view of the Xylella XF0768-XF0764 operon, including genes encoding BigR, putative membrane proteins 1 through 3, and a hydrolase (blh). (B) Sequence alignment between BigR and putative members of the ArsR/SmtB family, generated by ClustalW (http://www.ebi.ac.uk/clustalW). BA, Brucella abortus (NCBI no. AAX76167); ML, Mesorhizobium loti (NP_103569); SM, Sinorhizobium meliloti (NP_435817); AT, Agrobacterium tumefaciens C58 (AAK89928); RR, Rhodospirillum rubrum (ZP_00268696); VC, Vibrio cholerae (NP_233031); XV, Xanthomonas campestris pv. vesicatoria (YP_364177); RE, Ralstonia eutropha (ZP_00171423); BV, Burkholderia vietnamiensis (ZP_00422748); YI, Yersinia intermedia (ZP_00831742); CV, Chromobacterium violaceum (NP_899754); BJ, Bradyrhizobium japonicum (NP_772413); SP, Silicibacter pomeroyi (YP_165254). ArsR is from Escherichia coli (P15905), and SmtB is from Synechococcus elongatus (P30340). Invariable residues are boldfaced; asterisked residues are restricted to the HTH domain. Amino acids that are conserved only in the uncharacterized members of the ArsR/SmtB family (all except ArsR and SmtB) are shaded with white letters. Acidic residues in the N termini are boldfaced and underlined, and the residues responsible for metal coordination in ArsR and SmtB are shaded with boldface letters. The arrow points to the initial methionine of ΔBigR. (C) EMSA showing that BigR (lanes 1 and 2) and ΔBigR (lanes 3 and 4), with (lanes 1 and 3) or without (lanes 2 and 4) the His₆ tag, recognize the US of the blh gene. Arrow indicates shifted bands. FP, free probe.

FIG. 2.

BigR binds to an AT-rich element spanning the −10 region of the Xylella blh promoter. (A) Nucleotide sequence of the promoter showing the −35 and −10 elements, the putative ribosome-binding site (RBS), and the initial ATG codon (underlined). The transcription start site determined by RACE is indicated by the arrow in the DNA sequencing electropherogram. The −10 region overlaps with the dyad symmetry sequence (bold italics) of an imperfect palindrome (opposing arrows), the BigR box. Two additional palindromes, P1 and P2, are indicated by opposing arrows. P2 overlaps with a sequence (italics) that is similar to regulatory elements bound by MarA and ArsR (23, 29). (B) EMSA experiments with BigR6his (+) and the different DNA probes, represented by bars corresponding to the promoter elements shown in panel A (black, P1; light gray, P2; dark gray, −35 element; white, −10 element plus RBS), showing that the protein binds to the promoter fragments harboring the AT-rich sequence. Shifted bands (arrows) and the free probe (−) are indicated.

FIG. 3.

BigR binds to the palindromic AT-rich sequence, and this alters its secondary structure. (A) Footprint analysis of the Xylella blh promoter in the presence of decreasing amounts of BigR relative to that with a no-protein control (−). The footprinted area is read from the sequencing reaction on the left. (B) Gel shift using the double-stranded (ds) TATA probe (the BigR box sequence). A 10-fold excess (+) of the unlabeled ds-TATA probe was able to displace the labeled probe shifted by both BigR and ΔBigR (arrows). FP, free probe. (C) CD analysis of BigR in the presence of the ds-TATA probe. The resulting spectrum of the mixed protein and DNA (filled squares) differs from the theoretical sum (open circles) of the spectra measured separately, indicating changes in the secondary structure of the protein upon interaction with the DNA. The two minimum points at 208 and 222 nm also indicate that BigR has a high content of α-helices.

FIG. 4.

The BigR box is conserved in promoters of related operons found in plant-associated bacteria. (A) Schematic view of BigR operons from X. fastidiosa (Xf), A. tumefaciens (At), S. meliloti (Sm), M. loti (Ml), Chromobacterium violaceum (Cv), and Nitrosomonas europaea (Ne). Joined arrows represent ORFs clustered into operons, and arrows with the same shading indicate orthologous ORFs. Open arrows fused to hatched rectangles (DUF442) represent conserved blh genes, whereas blh without DUF442 is shown by open arrows only. Striped arrows, ORFs unique to the N. europaea operon, which also harbors a DUF442 hydrolase (stippled arrow). Numbers above ORFs are the percentages of identity between the Xylella BLH/BigR proteins and their orthologs. (B) Nucleotide sequence alignment of upstream sequences of the blh genes showing the BigR box consensus. (C) Nucleotide sequence of the A. tumefaciens blh promoter showing two transcription start sites (arrows) and other promoter elements, including the BigR box (boxed) and its AT-rich palindrome (opposing arrows). The frequencies of the two transcription start sites are given in the sequencing run diagrams.(D) EMSA performed with BigR (+) and the X. fastidiosa or A. tumefaciens blh promoter probe, showing the BigR shifted bands (arrow).

FIG. 5.

BigR functions as a transcriptional repressor. (A) EGFP fluorescence in cell extracts of E. coli carrying the Xylella pGem-pxf115GFP (bars 1 and 2) or the Agrobacterium pGem-pat145GFP (bars 3 and 4) reporter plasmid. Cells carrying the reporters and the pET28-BigR plasmid (bars 2 and 4) showed a reduction in EGFP fluorescence relative to that for nontransformed cells (bar 5). (B) EGFP fluorescence in E. coli cells carrying the Xylella pSP-pxf115GFP (bar 1) or the Agrobacterium pBI-pat145GFP (bar 4) reporter plasmid relative to fluorescence levels driven by the same reporters in Xylella (bar 2) and Agrobacterium (bar 5) cells, respectively. The background fluorescence of nontransformed Xylella (bar 3) and Agrobacterium (bar 6) cells is shown. Results are means from three independent samples.

FIG. 6.

Mutations in the BigR box affected repressor binding and transcription from blh promoters. (A) Mutations within the BigR boxes of the Xylella and Agrobacterium blh promoters represented by m1, m2, and m3 (negative controls) relative to the wild-type (wt) sequences. In the m1 mutant, the first half of the palindrome was replaced by an unrelated sequence (white letters), but one of the −10 elements and the spacing between the −10 and −35 regions were maintained, whereas in the m2 and m3 mutants, deletions were made in the BigR box. (B) EGFP fluorescence of cell extracts of E. coli transformed with the X. fastidiosa (Xf) or A. tumefaciens (At) reporter plasmid carrying a wt or mutated promoter, relative to the background fluorescence of untransformed cells (control [C]). Results are means from four independent samples. (C) EMSA using the wt probe or a mutant probe from the X. fastidiosa or A. tumefaciens promoter in the presence (+) or absence (−) of BigR protein (P), showing loss of DNA binding activity with the mutated sequences. Arrows point to shifted bands.

FIG. 7.

EGFP fluorescence of bacterial reporter cells grown as biofilms compared to that of cells in suspension. X. fastidiosa (A and B) and A. tumefaciens (C and D) reporter cells were grown on the surfaces of glass slides (A and C) and washed in sterile water before visualization under the fluorescence microscope at ×1,000 magnification. Planktonic cells (B and D) were pelleted and resuspended in a small volume of water prior to visualization to obtain a density of bacterial cells comparable to that of the biofilms. A. tumefaciens reporter cells attached to tobacco roots (E) showed increased EGFP fluorescence relative to that of nonattached cells (F) at ×100 magnification.

FIG. 8.

Effects of bigR and blh gene disruption on operon activity and bacterial growth. (A) EGFP fluorescence levels measured in A. tumefaciens wild-type (wt), blh mutant (mut), and bigR mutant cells carrying the reporter plasmid. (Inset) Expression of blh in wt and bigR mutant cells, measured by qPCR. (B) Bacterial growth curves of wt, blh mutant, and bigR mutant cells grown in suspension, monitored by measuring the OD₆₀₀. (C) Comparison of biofilm formation by the blh or bigR mutant with that by wt cells, measured by the ratio of the A₆₀₀ of the stained biofilm to the turbidity (OD₆₀₀) of the planktonic culture (Abs/OD). Results are means from five independent samples. (D) Average number of A. tumefaciens cells (expressed in CFU) recovered from tobacco roots after bacterial coincubation. Results are means from three independent replicates.