The sensor kinase DctS forms a tripartite sensor unit with DctB and DctA for sensing C4-dicarboxylates in Bacillus subtilis

Abstract

The DctSR two-component system of Bacillus subtilis controls the expression of the aerobic C4-dicarboxylate transporter DctA. Deletion of DctA leads to an increased dctA expression. The inactivation of DctB, an extracellular binding protein, is known to inhibit the expression of dctA. Here, interaction between the sensor kinase DctS and the transporter DctA as well as the binding protein DctB was demonstrated in vivo using streptavidin (Strep) or His protein interaction experiments (mSPINE or mHPINE), and the data suggest that DctA and DctB act as cosensors for DctS. The interaction between DctS and DctB was also confirmed by the bacterial two-hybrid system (BACTH). In contrast, no indication was obtained for a direct interaction between the transporter DctA and the binding protein DctB. Activity levels of uptake of [(14)C]succinate by bacteria that expressed DctA from a plasmid were similar in the absence and the presence of DctB, demonstrating that the binding protein DctB is not required for transport. Thus, DctB is involved not in transport but in cosensing with DctS, highlighting DctB as the first example of a TRAP-type binding protein that acts as a cosensor. The simultaneous presence of DctS/DctB and DctS/DctA sensor pairs and the lack of direct interaction between the cosensors DctA and DctB indicate the formation of a tripartite complex via DctS. It is suggested that the DctS/DctA/DctB complex forms the functional unit for C4-dicarboxylate sensing in B. subtilis.

FIG 1

In vivo binding of DctB-His₆ (DctB-H) to DctS-Strep (DctS-S) and copurification using a Strep-Tactin column (mSPINE) (A) and a Ni-NTA column (mHPINE) (C). (A) The two proteins were coexpressed in E. coli C43(DE3). Formaldehyde (FA) was added at an OD₅₇₈ of 1.3 to 1.6 where indicated. The membrane fraction was isolated and solubilized with 2% (wt/vol) Empigen. DctS-S was isolated from the membrane fraction using a Strep-Tactin column, and 20 μl of the eluate was subjected to SDS gel electrophoresis and Western blotting. As a control, DctB-H was expressed without DctS-S. (B) For testing the specificity of the mSPINE assay, the two noninteracting proteins DctS-S and His-DctP from Aromatoleum aromaticum were coexpressed in E. coli C43(DE3). At an OD₅₇₈ of 1.3 to 1.6, the cells were incubated with formaldehyde. The membrane fraction was solubilized as described above, and DctS-S was isolated using a Strep-Tactin column. (C) For the mHPINE assay, DctB-H was isolated from the solubilized membrane fraction using a Ni-NTA column. As a control, DctS-S was expressed without DctB-H.

FIG 2

Interaction of DctS with DctB in the bacterial two-hybrid system (BACTH). E. coli BTH101 was cotransformed with plasmid pairs encoding the T25 fusions and T18 fusions of the proteins as indicated: T25-DctS (pMW1697), T25-DctB (pMW1705), T18-DctS (pMW1696), DctS-T18 (pMW1695), DctR-T18 (pMW1699), and T18-DctB (pMW1708). The β-galactosidase activity was determined after growth to an OD₅₇₈ of 0.5 to 0.7. The strain expressing the pair of leucine zipper proteins T25-Zip and T18-Zip was used as a positive control for two interacting cytosolic proteins. The activity of the noninteracting proteins T18-DctS plus T25-Zip (80 Miller units, dashed line) represents the background. The data are from four or more independent biological replicates. Error bars represent the standard deviations.

FIG 3

Lack of interaction of DctA with the binding protein DctB: studies using the mSPINE assay followed by Western blotting (A) and the BACTH assay (B). For the mSPINE assay, DctA-Strep (DctA-S) and DctB-His (DctB-H) were coexpressed in E. coli C43(DE3) and the membranes were solubilized using 2% (wt/vol) Empigen BB detergent. DctA-S was isolated from the solubilized membrane fraction using a Strep-Tactin column. For the BACTH assays, E. coli BTH101 was cotransformed with plasmid pairs. In each pair, one plasmid encoded a T25 fusion and the other encoded a T18 fusion: T25-DctS (pMW1697) and T25-DctB (pMW1705) were the T25 fusions, and DctR-T18 (pMW1699), T18-DctA (pMW1704), and T18-DctS (pMW1696) were the T18 fusions. The β-galactosidase activity was determined after growth to an OD₅₇₈ of 0.5 to 0.7. The strain expressing the pair of leucine zipper proteins was used as a positive control for two interacting cytosolic proteins. The background activity obtained for noninteracting proteins like T18-DctS plus T25-Zip was below 80 Miller units (dashed line). All results are the averages of the measurements of at least four biological replicates. Error bars represent the standard deviations.

FIG 4

Interaction of DctA with DctS using the mSPINE (A) and mHPINE (B) methods followed by Western blotting. DctS-Strep (DctS-S) and DctA-His (DctA-H) were coexpressed in E. coli C43(DE3), and the membranes were solubilized using 2% Empigen BB detergent. For the mSPINE assay, DctS-S was isolated from the solubilized membrane fraction using a Strep-Tactin column. As a control, DctA-H was expressed without DctS-S and the membrane fraction was applied to a Strep-Tactin column. For the mHPINE assay, DctA-H was isolated from the solubilized membrane fraction using a Ni-NTA column.

FIG 5

Binding of DctB-His (DctB-H) and DctA-His (DctA-H) on DctS-Strep (DctS-S) and copurification by mSPINE. The three proteins were coexpressed in E. coli C43(DE3). DctS-S was isolated from the solubilized membrane fraction using a Strep-Tactin column. The proteins were detected in Western blots by antibodies directed against the His₆ or Strep tag, respectively. After blotting of the proteins, the nitrocellulose membrane was divided and the antibodies were used separately for detection of proteins. As a control, DctA-H and DctB-H were expressed without DctS-S and the membrane fraction was applied to a Strep-Tactin column.

FIG 6

Uptake of [¹⁴C]succinate by B. subtilis strains with various genetic backgrounds for dctB and dctA after aerobic growth of the strains on succinate. For transport measurements and growth experiments, the cells were grown in White minimal medium (24), which was enriched with 0.05% or 0.01% (wt/vol) yeast extract, respectively, and supplemented with 20 mM succinate. The transport rates and growth capabilities on succinate were determined for the B. subtilis 168 wild-type strain as a positive control, the wild-type strain plus plasmid-encoded DctA (pDctA, pMW1927) as a control for the addition of pDctA to B. subtilis cells, the DctB-deficient strain (DctB⁻), and the DctB⁻ strain supplemented with pMW1927. (A) Transport assays were performed at 37°C with 20 μM [¹⁴C]succinate. The transport rates are given in μmol × g [dry weight]⁻¹ × min⁻¹ and are the averages of at least three biological replicates. Error bars represent the standard deviations. (B) The bacterial growth is given as the final OD₆₀₀ after incubation in enriched White minimal medium at 37°C for 16 h.

FIG 7

Interaction of the DctB, DctA, and DctS proteins. The arrows between the different proteins are labeled to show whether the interaction experiments using the BACTH, mSPINE, and mHPINE experiments were positive (+) or negative (−). There was a strong interaction between DctB and DctS, and the mSPINE and mHPINE assays showed a strong interaction of DctA with the sensor kinase DctS. No interaction was detected between DctB and DctA, suggesting that the transport of C₄-dicarboxylates by DctA is independent of the presence of the binding protein DctB. In addition to the BACTH assays shown in Fig. 2 and Fig. 3B, the assay was used to analyze the interaction between DctS and DctA, but none of the T18 and T25 fusion pairs showed β-galactosidase activity above background levels (data not shown). In the BACTH assay, sterical hindrances of potential interaction partners can lead to a false-negative result (11), which is a common problem with membrane proteins which can move and rotate only in two dimensions.

FIG 8

Model for the functional tripartite DctS/DctA/DctB complex for sensing fumarate and succinate in B. subtilis. According to the model, DctB is the primary site for perception of fumarate and succinate, which are then transmitted to PAS_P of DctS. The putative sites for DctA/DctS interaction and the proposed routes for signal transfer are shown. Transport of fumarate and succinate modifies the interaction between DctA and DctS (dashed and kinked C-terminal helix in DctA), which affects activity or signal transmission to the kinase domain. Abbreviations: PAS_P, periplasmic PAS (Per-ARNT-Sim) domain; PAS_C, cytoplasmic PAS (Per-ARNT-Sim) domain; Kin, kinase domain; P, phosphoryl group; Succ, succinate.