BDSF Is a Degradation-Prone Quorum-Sensing Signal Detected by the Histidine Kinase RpfC of Xanthomonas campestris pv. campestris

Abstract

Diffusible signal factors (DSFs) are medium-chain fatty acids that induce bacterial quorum sensing. Among these compounds, BDSF is a structural analog of DSF that is commonly detected in bacterial species, and it is the predominant in planta quorum-sensing signal in Xanthomonas campestris. How BDSF is sensed in Xanthomonas spp. and the functional diversity between BDSF and DSF remain unclear. In this study, we generated genetic and biochemical evidence that BDSF is a low-active regulator of X. campestris pv. campestris quorum sensing, whereas trans-BDSF does not seem to be a signaling compound. BDSF is detected by the sensor histidine kinase RpfC. Although BDSF has relatively low physiological activities, it binds to the RpfC sensor with a high affinity and activates RpfC autophosphorylation to a level that is similar to that induced by DSF in vitro. The inconsistency in the physiological and biochemical activities of BDSF is not due to RpfC-RpfG phosphorylation or RpfG hydrolase. Neither BDSF nor DSF controls the phosphotransferase and phosphatase activities of RpfC or the ability of RpfG hydrolase activity to degrade the bacterial second messenger cyclic di-GMP. We demonstrated that BDSF is prone to degradation by RpfB, a critical fatty acyl coenzyme A ligase involved in the turnover of DSF-family signals. rpfB mutations lead to substantial increases in BDSF-induced quorum sensing. Although DSF and BDSF are similarly detected by RpfC, our data suggest that their differential degradation in cells is the major factor responsible for the diversity in their physiological effects. IMPORTANCE The diffusible signal factor (DSF) family consists of quorum-sensing signals employed by Gram-negative bacteria. These signals are a group of cis-2-unsaturated fatty acids, such as DSF, BDSF, IDSF, CDSF, and SDSF. However, the functional divergence of various DSF signals remains unclear. The present study demonstrates that though BDSF is a low active quorum-sensing signal, it binds histidine kinase RpfC with a higher affinity and activates RpfC autophosphorylation to the similar level as DSF. Rather than regulation of enzymatic activities of RpfC and its cognate response regulator RpfG encoding a c-di-GMP hydrolase, BDSF is prone to degradation in bacterial cells by RpfB, which effectively avoided the inhibition of bacterial growth by accumulating high concentrations of BDSF. Therefore, our study sheds new light on the functional differences of quorum-sensing signals and shows that bacteria balance quorum sensing and growth by fine-tuning concentrations of signaling chemicals.

Keywords: BDSF; DSF; Xanthomonas campestris; functional divergence; histidine kinase; quorum-sensing.

FIG 1

The quorum-sensing activity of BDSF is lower than that of DSF. (A) Chemical structures of BDSF and DSF. (B) Addition of BDSF and DSF increased the extracellular protease production of the rpfF mutant. In each panel, rpfF mutant bacterial colonies on NYG medium are presented, with the effects of 4 μL of BDSF or DSF added nearby the rpfF mutant to trigger quorum sensing presented on the right. The protease activity was determined by measuring the diameters (in cm) of the protein degradation zones after a 36-h incubation (n = 4). (C and D) Addition of DSF (C) and BDSF (D) increased the biofilm formation of the rpfF mutant. A crystal violet staining method was used to quantify the biofilm formation (n = 4). (E) engXCA transcription in response to BDSF or DSF. Each experiment was repeated four times. The vertical bars represent the standard deviations (n = 4). The asterisks indicate a significant difference compared to the unstimulated control.

FIG 2

trans-BDSF is a low-activity inducer of quorum sensing. (A) Extracellular protease production assay. In each panel, rpfF mutant bacterial colonies on NYG medium are presented, with the effects of 4 μL of trans-BDSF added nearby the rpfF mutant to trigger quorum sensing presented on the right. The protease activity was determined by measuring the diameters (in cm) of the protein degradation zones after a 36-h incubation (n = 4). (B) Biofilm production. A crystal violet staining method was used to quantify the biofilm formation of the rpfF mutant (n = 4). (C) PengXCA reporter activity assay in response to BDSF or trans-BDSF. Each experiment was repeated four times. The vertical bars represent the standard deviations (n = 4). The asterisks indicate a significant difference compared to the unstimulated control.

FIG 3

RpfC sensor and transmembrane regions are critical for sensing BDSF. (A) Schematic view of the full-length RpfC embedded in the membrane and RpfG. Secondary structures of RpfC-RpfG were predicted using Pfam. Transmembrane regions were predicted using TMpred. (B) Extracellular protease production assay. In each panel, rpfF rpfC double-mutant (deletions in various regions) bacterial colonies on NYG medium are presented, with the effects of 4 μL of BDSF added nearby the rpfF rpfC mutant to trigger quorum sensing presented on the right. (C) The protease activity was determined by measuring the diameters (in cm) of the protein degradation zones after a 36-h incubation (in panels A and B, n = 4). (D) Biofilm production. A crystal violet staining method was used to quantify the biofilm formation of various rpfF rpfC mutants (n = 4). BDSF (50 μM) was added to the bacterial culture. (E) PengXCA reporter activity assay to evaluate the effect of 50 μM BDSF. Each experiment was repeated four times. The vertical bars represent the standard deviations (n = 4). The asterisks indicate a significant difference compared to the unstimulated control (rpfF mutant).

FIG 4

Identification of the amino acid residues in the RpfC sensor that are essential for sensing BDSF. rpfC mutants with point mutations generated by alanine-scanning mutagenesis in the rpfF background were tested. (A) Extracellular protease production by the mutants. In each panel, rpfF rpfC double mutant (various substitution sites) bacterial colonies on NYG medium are presented, with the effects of 4 μL of BDSF added nearby the rpfF rpfC mutant to trigger quorum sensing presented on the right. (B) The protease activity was determined by measuring the diameters (in cm) of the protein degradation zones after a 36-h incubation (in both panels A and B, n = 4). (C) Biofilm production by the mutants. A crystal violet staining method was used to quantify the biofilm formation of various mutants (n = 4). BDSF (50 μM) was added to the bacterial culture. (D) PengXCA reporter activity assay to evaluate the effect of 50 μM BDSF. Each experiment was repeated four times. The vertical bars represent the standard deviations (n = 4). The asterisks indicate a significant difference compared to the unstimulated control (rpfF mutant).

FIG 5

BDSF binds to the RpfC sensor to activate the autokinase. (A) BDSF and DSF similarly activate the RpfC autokinase. Lower panels show RpfC stained with Coomassie Brilliant Blue, which served as loading controls. (B and C) BDSF activates the RpfC autokinase in a dose-dependent manner. (B) Various concentrations of BDSF (0.05 to 5 μM) were added, along with [γ-³²P]ATP, to the reaction mixture for the autophosphorylation assay. (C) Quantification of the isotopic signal for the RpfC autophosphorylation presented in panel B. (D) Substitution of essential amino acids decreased the RpfC autophosphorylation level. (E and F) A microscale thermophoresis assay revealed that BDSF can bind to the full-length RpfC liposome (E) and RpfC sensor (F). The DSF concentrations ranged from 25 to 2,000 μM. The solid curve presents the fit of the data points to the standard KD-Fit function. Each binding assay was repeated three times, and black bars represent the standard deviations. K_d, dissociation constant.

FIG 6

Neither BDSF nor DSF regulated the RpfC phosphotransferase and phosphatase activities or the degradation of c-di-GMP by RpfG hydrolase. (A) RpfC phosphorylated RpfG at its Asp⁸⁰ residue. The RpfC liposome was autophosphorylated by [γ-³²P]ATP, and recombinant RpfG or RpfG^D80V was added to the reaction mixtures. The phosphorylation signal was detected by autoradiography. The lower panels present proteins stained with Coomassie Brilliant Blue. (B) The addition of BDSF and DSF did not affect the RpfC-RpfG phosphotransfer. (C) Full-length RpfC can function as a phosphatase to dephosphorylate RpfG. After the RpfC-RpfG phosphotransfer, RpfC liposomes were removed by centrifugation, and free ATP was removed by desalination. Fresh unphosphorylated RpfC liposomes were then added to the reaction mixture, which was incubated for various periods. The phosphorylation signal was detected by autoradiography. (D) The addition of BDSF and DSF did not affect the RpfC phosphatase activity toward RpfG. (E) The addition of BDSF and DSF did not affect the degradation of c-di-GMP by RpfG hydrolase. The ³²P-labeled c-di-GMP was synthesized using a recombinant tDGC cyclase and then coincubated with recombinant RpfG. If necessary, BDSF and DSF were added to the reaction mixtures. The products of the enzymatic reaction were separated by thin-layer chromatography, and the isotopic signal was recorded by autoradiography. For all panels, each experiment was repeated three times, and representative results are presented.

FIG 7

BDSF is prone to degradation in vivo and in vitro. (A) Quantification of biofilm production in the rpfF mutant and the rpfF rpfB double mutant. We added 10 μM BDSF and DSF to the bacterial cultures. (B) PengXCA reporter activity assay of the rpfF mutant and the rpfF rpfB double mutant. Labels are the same as in panel A. We added 10 μM BDSF and DSF to the bacterial cultures to induce engXCA transcription. (C and D) Fatty acyl-CoA ligase activity of RpfB to activate DSF and BDSF. Decreasing concentrations of CoA thiol groups resulting from fatty acyl-CoA synthesis were measured using DSF and BDSF as the substrates. For panels C and D, Mg²⁺ and Ca²⁺ were used as cofactors, respectively. (E and F) Bacterial growth curves. Various strains were grown in the type III secretion system-inducing medium (i.e., XCM2). If necessary, DSF or BDSF was added. For all panels, each experiment was repeated four times. The vertical bars represent the standard deviations (n = 4). The asterisks indicate a significant difference compared to the unstimulated control.