The phosphoenolpyruvate phosphotransferase system regulates Vibrio cholerae biofilm formation through multiple independent pathways

Abstract

The bacterial phosphoenolpyruvate phosphotransferase system (PTS) is a highly conserved phosphotransfer cascade that participates in the transport and phosphorylation of selected carbohydrates and modulates many cellular functions in response to carbohydrate availability. It plays a role in the virulence of many bacterial pathogens. Components of the carbohydrate-specific PTS include the general cytoplasmic components enzyme I (EI) and histidine protein (HPr), the sugar-specific cytoplasmic components enzymes IIA (EIIA) and IIB (EIIB), and the sugar-specific membrane-associated multisubunit components enzymes IIC (EIIC) and IID (EIID). Many bacterial genomes also encode a parallel PTS pathway that includes the EI homolog EI(Ntr), the HPr homolog NPr, and the EIIA homolog EIIA(Ntr). This pathway is thought to be nitrogen specific because of the proximity of the genes encoding this pathway to the genes encoding the nitrogen-specific sigma factor sigma(54). We previously reported that phosphorylation of HPr and FPr by EI represses Vibrio cholerae biofilm formation in minimal medium supplemented with glucose or pyruvate. Here we report two additional PTS-based biofilm regulatory pathways that are active in LB broth but not in minimal medium. These pathways involve the glucose-specific enzyme EIIA (EIIA(Glc)) and two nitrogen-specific EIIA homologs, EIIA(Ntr1) and EIIA(Ntr2). The presence of multiple, independent biofilm regulatory circuits in the PTS supports the hypothesis that the PTS and PTS-dependent substrates have a central role in sensing environments suitable for a surface-associated existence.

FIG. 1.

Schematic diagram of the 25 V. cholerae PTS components. Homologs of the EI, HPr, and EII proteins are indicated by the same colors. Furthermore, proteins belonging to the same EIIA superfamily are color coded (blue, glucose superfamily; violet, mannitol-fructose superfamily; yellow, lactose superfamily). Sugar specificities are indicated for PTS components if they are known (16). Parentheses indicate presumptive specificities. The VC1824 HPr-like domain is surrounded by a dashed line to indicate its low level of similarity to HPr. The sugar specificity of VC1281 is based on previously described data (3). NAG, N-acetylglucosamine.

FIG. 2.

PTS mutant displays distinct biofilm phenotypes in LB broth and minimal medium (MM) supplemented with glucose. Total growth and biofilm accumulation were determined for wild-type V. cholerae (WT), a ΔEI mutant, a ΔEIIA^Glc mutant, and a ΔPTS mutant in minimal medium supplemented with 0.5% (wt/vol) glucose (MM + glucose) and LB broth (LB). The values are the averages of four experimental replicates, and the error bars indicate standard deviations. There was a statistically significant difference (P = 0.0062)between biofilm formation by wild-type V. cholerae and biofilm formation by the ΔPTS mutant in minimal medium supplemented with glucose (indicated by an asterisk) but not in LB broth (P = 0.116).

FIG. 3.

EI-P represses biofilm formation and vps gene transcription in LB broth (biofilm regulatory pathway 1). (A) Total growth and biofilm accumulation in LB broth by wild-type V. cholerae (WT) or a ΔEI mutant rescued with either a control pBAD plasmid (pCTL), a pBAD plasmid expressing a wild-type EI allele (pEI), or a pBAD plasmid expressing an EI allele encoding an H-to-A mutation at position 189 (pEI^H189A). Protein expression was induced by addition of 0.04% l-arabinose. The values are the averages of four experimental replicates, and the error bars indicate standard deviations. Biofilm formation by the ΔEI mutant rescued with a wild-type EI allele was not statistically different from biofilm formation by wild-type V. cholerae (P = 0.290), while biofilm formation by a ΔEI mutant rescued with the EI^H189A allele was statistically different (P = 0.0007). (B) Western blot analysis demonstrating that the wild-type EI and EI^H189A alleles are well expressed in the ΔEI genetic background. (C) Quantification of total growth and biofilm formation over time for wild-type V. cholerae and a ΔEI mutant. The values are the means of two experimental replicates. The ΔEI mutant demonstrated more biofilm accumulation than wild-type V. cholerae throughout the course of the experiment.

FIG. 4.

EI and HPr are in the same biofilm regulatory pathway. (A) Total growth and biofilm accumulation by wild-type V. cholerae (WT), a ΔHPr mutant, a ΔFPr mutant, and a ΔHPr ΔFPr double mutant. Biofilm accumulation by a ΔHPr mutant is significantly different from that by wild-type V. cholerae (P = 0.002), while biofilm accumulation by a ΔHPr ΔFPr is not significantly different from that by ΔHPr (P = 0.758). (B) Total growth and biofilm accumulation by wild-type V. cholerae (WT), as well as ΔEI and ΔHPr ΔFPr mutants rescued either with a control pBAD plasmid (pCTL) or a pBAD plasmid expressing a wild-type EI allele (pEI). Protein expression was induced by addition of 0.04% l-arabinose. Deletion of HPr and FPr blocked the ability of EI to repress biofilm formation. (C) Measurement of vps gene transcription in wild-type V. cholerae and various PTS mutants harboring a chromosomal vpsL-lacZ fusion. vpsL transcription in the ΔEI (P = 0.0002), ΔHPr (P < 0.0001), and ΔHPr ΔFPr (P = 0.0001) mutants is significantly different from that in wild-type V. cholerae. The error bars indicate standard deviations. RLU, relative light units.

FIG. 5.

EIIA homologs EIIA^Glc, EIIA^Ntr1, and EIIA^Ntr2 regulate biofilm formation: quantification of total growth and biofilm accumulation by wild-type V. cholerae, as well as strains having mutations in each of the 10 genes encoding EIIA domains. The biofilm accumulation by the ΔEIIA^Glc (P < 0.0001), ΔEIIA^Ntr (P < 0.0001), and ΔVC1824 (P = 0.0004) mutants was significantly different from that by wild-type V. cholerae (WT). The error bars indicate standard deviations.

FIG. 6.

Differential transcription of PTS components is not solely responsible for the patterns of biofilm regulation observed in LB broth compared with those observed in minimal medium. The EI, HPr, FPr, EIIA^Glc, EIIA^Ntr1, and ΔEIIA^Ntr2 transcript levels in wild-type V. cholerae grown in LB broth were compared to the transcript levels in MM supplemented with pyruvate. The data were analyzed in triplicate using the ΔΔC_t method. clpX was used as a standard. The error bars indicate standard deviations.

FIG. 7.

Activation of biofilm formation by EIIA^Glc does not require phosphorylation and is independent of EI and HPr (biofilm regulatory pathway 2). (A) Quantification of total growth and biofilm accumulation by a ΔEIIA^Glc mutant rescued with a pBAD plasmid carrying a control sequence (pCTL) and pBAD plasmids with a wild-type EIIA^Glc allele (pEIIA^Glc) and a variety of mutant EIIA^Glc alleles. Protein expression was induced with 0.04% arabinose. Biofilm accumulation by the ΔEIIA^Glc mutant rescued with a wild-type EIIA^Glc glucose allele (P = 0.0024), an EIIA^Glc allele encoding a histidine-to-alanine substitution at position 191 (pEIIA^GlcH91A) (P < 0.0001), or an EIIA^Glc allele encoding a histidine-to-aspartate substitution at position 191 (pEIIA^GlcH91D) (P < 0.0001) was significantly increased compared with biofilm accumulation by the ΔEIIA^Glc mutant rescued with a control sequence. (B) Quantification of total growth and biofilm accumulation by a ΔPTS mutant (lacking EI, HPr, and EIIA^Glc) rescued with a pBAD plasmid carrying a control sequence (pCTL), a wild-type EIIA^Glc allele (pEIIA^Glc), an EIIA^Glc allele encoding a histidine-to-alanine substitution at position 191 (pEIIA^GlcH91A), or an EIIA^Glc allele encoding a histidine-to-aspartate substitution at position 191 (pEIIA^GlcH91D). Protein expression was induced with 0.04% arabinose. Biofilm accumulation by the ΔPTS mutant rescued with any of the EIIA^Glc alleles was significantly increased compared with biofilm accumulation by the ΔPTS mutant rescued with a control sequence (P ≤ 0.0002 for all comparisons). The error bars indicate standard deviations. (C) Western blot demonstrating that the wild-type and mutant EIIA^Glc alleles used in these experiments are well expressed and produce full-length proteins.

FIG. 8.

Evidence for involvement of Mlc and EIIBC^Glc in pathway 2: quantification of total growth and biofilm accumulation by wild-type V. cholerae and various pathway 2 mutants. Biofilm accumulation by the ΔEI (P = 0.0008) or ΔEIIBC^Glc (P = 0.0001) mutant was significantly greater than that by wild-type V. cholerae (WT). Biofilm accumulation by the ΔEIIA^Glc or ΔMlc mutants was significantly less than that by wild-type V. cholerae (P < 0.0001). Biofilm accumulation by the ΔEIIA^Glc ΔEIIBC^Glc mutant (P = 0.04) or the ΔEIIA^Glc ΔEIIBC^Glc ΔMlc mutant (P = 0.001) was significantly greater than that by the ΔEIIA^Glc mutant. Biofilm accumulation by the ΔEIIBC^Glc ΔMlc mutant (P = 0.0006) or the ΔEIIA^Glc ΔEIIBC^Glc ΔMlc mutant (P = 0.0005) was significantly greater than that by the ΔMlc mutant. The error bars indicate standard deviations.

FIG. 9.

Pathway 2 regulates biofilm accumulation at the transcriptional level. (A) Schematic diagram of pathways 2a and 2b. In this model, Mlc activates vps gene transcription through both EIIA^Glc-dependent and EIIA^Glc-independent pathways. EIIBC^Glc interferes with the action of Mlc. The numbers in circles indicate the predicted relative effect of each component in the regulatory cascade on vps gene transcription. (B) All the strains tested have vpsL-lacZ promoter fusions in the V. cholerae lacZ locus. The measurements of β-galactosidase activity reflect vpsL transcription. The measurements for the ΔEIIA^Glc (P < 0.0001), ΔEIIBC^Glc (P = 0.04), and ΔMlc (P = 0.0154) mutants were significantly different from that for wild-type V. cholerae. The vpsL transcription in the ΔEIIBC^Glc ΔMlc mutant was significantly different from that in the ΔEIIBC^Glc mutant (P = 0.0003) but not from that in the ΔMlc mutant (P = 0.766). The vps transcription in the ΔEIIA^Glc ΔEIIBC^Glc mutant was significantly different from that in the ΔEIIA^Glc mutant (P = 0.032) or the ΔEIIBC^Glc mutant (P < 0.0001). The vps transcription in the ΔEIIA^Glc ΔEIIBC^Glc ΔMlc mutant was significantly different both from that in the ΔEIIA^Glc ΔEIIBC^Glc mutant (P = 0.0086) or the ΔEIIBC^Glc ΔMlc mutant (P < 0.0001) but not from that in the ΔEIIA^Glc mutant (P = 0.415). The models on the right show the predicted impact of each genetic background on vps gene transcription. The relative contributions of the regulatory components to vps gene transcription as shown in panel A were added to obtain the predicted impact of the genetic background on vps gene transcription. WT, wild type; RLU, relative light units.

FIG. 10.

Activation of biofilm accumulation by Mlc is partially dependent on EIIA^Glc: quantification of biofilm accumulation and total growth by wild-type V. cholerae (WT) and ΔMlc, ΔEIIBC^Glc ΔMlc, and ΔEIIA^Glc ΔEIIBC^Glc ΔMlc mutants rescued with either a control vector (pCTL) or a plasmid carrying a wild-type mlc allele (pMlc). Biofilm formation by ΔMlc (P < 0.0001), ΔEIIBC^Glc ΔMlc (P < 0.0001), and ΔEIIA^Glc ΔEIIBC^Glc ΔMlc (P = 0.002) mutants rescued with the plasmid carrying the wild-type mlc allele was significantly greater than that by mutants carrying a control vector. The error bars indicate standard deviations.

FIG. 11.

Repression of biofilm formation by EIIA^Ntr1 does not require phosphorylation of the conserved histidine at residue 66 and is independent of pathway 1 (biofilm regulatory pathway 3). (A) Quantification of total growth and biofilm accumulation by wild-type V. cholerae or a ΔEIIA^Ntr1 mutant rescued with a control plasmid (pCTL), a plasmid carrying a wild-type EIIA^Ntr1 allele (pEIIA^Ntr1), a plasmid carrying an EIIA^Ntr1 allele resulting in an alanine substitution for the histidine at position 66 (pEIIA^Ntr1H66A), a plasmid carrying an EIIA^Ntr1 allele resulting in an aspartate substitution for the histidine at position 66 (pEIIA^Ntr1H66D), and a plasmid carrying a wild-type EIIA^Ntr2 allele (pEIIA^Ntr2). Only biofilm accumulation by the ΔEIIA^Ntr1 (pCTL) (P = 0.0003) and ΔEIIA^Ntr1 (pEIIA^Ntr2) (P = 0.0004) mutants was significantly different from that by wild-type V. cholerae (WT). The error bars indicate standard deviations. (B) Western blot demonstrating that the wild-type and mutant EIIA^Ntr1 alleles used in these experiments are well expressed and produce full-length proteins. (C) Quantification of total growth and biofilm accumulation by wild-type V. cholerae and strains having mutations in the other two components of the PTS^Ntr, EI^Ntr, and NPr. Biofilm accumulation by the two mutants was not significantly different from that by wild-type V. cholerae. (D) Quantification of total growth and biofilm formation by a ΔEI mutant or a ΔEI ΔEIIA^Ntr1 mutant rescued with a pBAD plasmid carrying a control sequence (pCTL), a wild-type EI allele (pEI), or a wild-type EIIA^Ntr1 allele (pEIIA^Ntr1). Biofilm accumulation by the ΔEI ΔEIIA^Ntr1 mutant rescued with a wild-type EI allele was significantly different from that by the ΔEI ΔEIIA^Ntr1 mutant carrying a control sequence (P = 0.0102).

FIG. 12.

Repression of biofilm formation by EIIA^Ntr2 does not require phosphorylation at histidine 172 and is rescued by EIIA^Ntr1 (biofilm regulatory pathway 3) (A) Quantification of total growth and biofilm accumulation by wild-type V. cholerae or a ΔEIIA^Ntr2 mutant rescued with a pBAD plasmid carrying a control sequence (pCTL), a wild-type ΔEIIA^Ntr2 allele (pEIIA^Ntr2), a ΔEIIA^Ntr2 allele encoding an H-to-A point mutation at conserved residue 172 (pEIIA^Ntr2H172A), a ΔEIIA^Ntr2 allele encoding an H-to-D point mutation at conserved residue 172 (pEIIA^Ntr2H172D), or a wild-type EIIA^Ntr1 allele (pEIIA^Ntr1). Protein expression was induced with 0.04% arabinose. Biofilm accumulation by a ΔEIIA^Ntr2 mutant rescued with wild-type EIIA^Ntr2 (P = 0.263), EIIA^Ntr2H172A (P = 0.776), EIIA^Ntr2H172D (P = 0.126), or EIIA^Ntr1 (P = 0.514) was not significantly different from that by wild-type V. cholerae (WT). The error bars indicate standard deviations. (B) Western blot demonstrating that the wild-type and mutant EIIA^Ntr2 alleles used in these experiments are well expressed and produce full-length proteins.

FIG. 13.

EIIA^Ntr1 and EIIA^Ntr2 operate at the level of transcription: measurements of vps gene transcription in wild-type V. cholerae, a ΔEIIA^Ntr1 mutant, and a ΔEIIA^Ntr2 mutant harboring a chromosomal vpsL-lacZ fusion. The vpsL transcription in the ΔEIIA^Ntr1 and ΔEIIA^Ntr2 mutants was significantly different from that in wild-type V. cholerae (WT) (P < 0.0001). The error bars indicate standard deviations. RLU, relative light units.

FIG. 14.

Schematic diagram of three pathways in the PTS network that regulate biofilm formation. Pathway 1 requires phosphorylation of the PTS components EI and HPr or FPr to repress transcription of the vps genes. Pathway 2a is comprised of the components EIIBC^Glc, Mlc, and EIIA^Glc. EIIA^Glc activates vps transcription in a phosphorylation-independent manner. Mlc potentiates the action of EIIA^Glc but may also activate vps gene transcription through an EIIA^Glc-independent route (pathway 2b). EIIBC^Glc interferes with the action of Mlc. Pathway 3 includes EIIA^Ntr1and EIIA^Ntr2, whose functions display some redundancy in repression of vps transcription.