A Novel Role for Enzyme I of the Vibrio cholerae Phosphoenolpyruvate Phosphotransferase System in Regulation of Growth in a Biofilm

Abstract

Glucose is a universal energy source and a potent inducer of surface colonization for many microbial species. Highly efficient sugar assimilation pathways ensure successful competition for this preferred carbon source. One such pathway is the phosphoenolpyruvate phosphotransferase system (PTS), a multicomponent sugar transport system that phosphorylates the sugar as it enters the cell. Components required for transport of glucose through the PTS include enzyme I, histidine protein, enzyme IIA^Glc, and enzyme IIBC^Glc. In Escherichia coli, components of the PTS fulfill many regulatory roles, including regulation of nutrient scavenging and catabolism, chemotaxis, glycogen utilization, catabolite repression, and inducer exclusion. We previously observed that genes encoding the components of the Vibrio cholerae PTS were coregulated with the vps genes, which are required for synthesis of the biofilm matrix exopolysaccharide. In this work, we identify the PTS components required for transport of glucose and investigate the role of each of these components in regulation of biofilm formation. Our results establish a novel role for the phosphorylated form of enzyme I in specific regulation of biofilm-associated growth. As the PTS is highly conserved among bacteria, the enzyme I regulatory pathway may be relevant to a number of biofilm-based infections.

The ability of human beings to control their environment facilitates colonization of inhospitable habitats. In contrast, the limited ability of bacteria to regulate their environment dictates that they seek out and colonize optimal surroundings. For some bacteria, the response to an encounter with a hospitable environment is the formation of a biofilm, a surface-associated, three-dimensional, structured bacterial community enclosed in an extracellular matrix. Biofilm formation brings with it many benefits, including surface colonization, sharing of resources, protection against predators, and resistance to antimicrobial agents (5, 7, 17). Therefore, it is not surprising that bacterial species are predominantly found in association with surfaces.

Monosaccharides such as glucose are universal energy sources and the hallmark of a hospitable environment for many bacterial pathogens. Thus, bacterial surface colonization and biofilm formation are highly influenced by the type of carbohydrates available in the environment (3, 4, 13, 15, 20). Moreover, global regulators of carbon metabolism such as Csr and cyclic AMP (cAMP) receptor protein (CRP) of Escherichia coli, Crc of Pseudomonas aeruginosa, CcpA of Bacillus subtilis, and CRP of Klebsiella pneumoniae and Shewanella oneidensis have previously been identified as regulators of surface adhesion and biofilm formation (2, 4, 11, 22, 24, 25). These regulators themselves are under the influence of the environment via complex signaling pathways that are not completely understood.

Bacteria produce many different types of carbohydrate transporters. Among these, the phosphoenolpyruvate (PEP) transport system (PTS) stands out not only as the primary transporter of PTS-dependent sugars but also as a global regulator of the bacterial behaviors and metabolic processes that fine-tune the cell's physiology to the environment at hand. The PTS is highly conserved among bacteria and commonly transports monosaccharides such as glucose, mannose, and fructose and disaccharides such as sucrose and cellobiose. It is unique in that it consists of a multienzyme phosphotransfer cascade that ultimately activates the transported sugar by phosphorylation. The general components of this phosphotransfer cascade include the cytoplasmic proteins enzyme I (EI) and histidine protein (HPr). Bacteria possess multiple enzymes II (EII) that are sugar specific and comprise a cytoplasmic A subunit as well as B, C, and sometimes D subunits, which are located in the inner membrane. In the PTS phosphotransfer cascade, a phosphate group is transferred sequentially from PEP to EI, to HPr, to the relevant EII, and finally to the sugar as it is translocated across the membrane (Fig. 1A). The components of the PTS have many regulatory functions. For instance, in its unphosphorylated state, EI regulates the chemotactic response to PTS sugars. Unphosphorylated HPr stimulates utilization of glycogen. EIIA^Glc regulates transport and utilization of alternative carbon sources through inducer exclusion and catabolite repression. In Streptococcus mutans, EIIAB^Man has been demonstrated to activate biofilm formation (1). However, a role for EI in repression of biofilm-associated growth has not been described previously (6).

FIG. 1.

Schematic representation of the phosphotransfer cascade comprising the PTS (A) and the genomic organization of genes encoding the general and glucose-specific PTS components (B).

Vibrio cholerae is a native inhabitant of diverse aquatic environments, where it has been found in biofilms on the surfaces of plants, insects, plankton, and larger crustaceans. Due to its ability to adhere to the epithelium of the small bowel and elaborate cholera toxin, it is also the causative agent of the diarrheal disease cholera. V. cholerae biofilm development occurs in discrete stages. First, free-living cells encounter a surface and become permanently attached to it as a monolayer. Then, through the action of one or more environmental signals, VPS exopolysaccharide synthesis is activated, leading to the formation of a multicellular biofilm (8, 13, 27, 29).

The V. cholerae genome contains 24 open reading frames encoding putative components of the PTS. Homologs of the general components of the V. cholerae PTS are encoded at loci VC0964 (encoding EIIA^Glc), VC0965 (encoding EI), and VC0966 (encoding HPr). A homolog of EIIBC^Glc is encoded at locus VC2013 (Fig. 1B). In previous analyses of stage-specific gene transcription during biofilm development, we found that 11 of these open reading frames were coregulated with the vps genes (21). Based on these findings, we investigated the hypothesis that sugar transport by the PTS is essential to the biofilm mode of growth. Here, we present the surprising finding that the V. cholerae PTS selectively represses growth of biofilm-associated cells. Furthermore, we present evidence that the phosphorylated form of EI is responsible for this regulation. This work uncovers a novel role for the PTS EI in regulation of V. cholerae biofilm formation.

MATERIALS AND METHODS

Bacterial strains, plasmids, and media.

Relevant bacterial strains, plasmids, and the primer sets required to generate these strains and plasmids are listed in Table 1. All of the mutants were constructed as previously described (9). The ΔPTS ΔvpsA double mutant (PW877) was constructed using the previously generated suicide plasmid pAJH9, which carries a vpsA deletion fragment (13).

TABLE 1.

Strains, plasmids, and primers used in this study

Strain, plasmid, or primer	Genotype or sequence	Source or reference
Strains
E. coli
SM10λpir	thi thr leu tonA lacY supE recA::RP4-2-Tc::Mu λpirR6K Kmr	20
PW744	SM10λpir (pWM91ΔPTS)	This study
PW814	SM10λpir (pWM91ΔEIIAGlc)	This study
PW862	SM10λpir (pWM91ΔEI)	This study
PW863	SM10λpir (pWM91ΔHPr)	This study
PW831	SM10λpir (pWM91ΔEIIBCGlc)	This study
PW398	SM10λpir (pWM91ΔvpsA)	12
V. cholerae
PW357	MO10 lacZ::vpsLp→lacZ; Smr	8
PW751	MO10 ΔPTS; Smr	This study
PW836	MO10 ΔEIIAGlc; Smr	This study
PW864	MO10 ΔEI; Smr	This study
PW867	MO10 ΔHPr; Smr	This study
PW844	MO10 ΔEIIBCGlc; Smr	This study
PW396	MO10 ΔvpsA; Smr	12
PW877	MO10 ΔPTS ΔvpsA; Smr	This study
Plasmids
pWM91	oriR6K mobRP4 lacI ptac tnp mini-Tn10Km; Kmr Apr	18
pWM91Δpts	pWM91 carrying a fragment of PTS operon harboring an internal, unmarked deletion	This study
pWM91ΔEIIAGlc	pWM91 carrying a fragment of VC0964 harboring an internal, unmarked deletion	This study
pWM91ΔEI	pWM91 carrying a fragment of VC0965 harboring an internal, unmarked deletion	This study
pWM91ΔHPr	pWM91 carrying a fragment of VC0966 harboring an internal, unmarked deletion	This study
pWM91ΔEIIBCGlc	pWM91 carrying a fragment of VC2013 harboring an internal, unmarked deletion	This study
pBAD-TOPO-EIIAGlc	pBAD-TOPO carrying a fragment of VC0964	This study
pBAD-TOPO-EI	pBAD-TOPO carrying a fragment of VC0965	This study
pBAD-TOPO-HPr	pBAD-TOPO carrying a fragment of VC0966	This study
pBAD-TOPO-EI(H189A)	pBAD-TOPO carrying a fragment of VC0965 with mutation at position 189 (H to A)	This study
Primers
Construction of V. cholerae PTS deletion
P342	CGAACGTTTTAAATCGAACCA
P343	TAACGAGCGGCCGCACCTGCTGGTAACAAAATGGT
P344	TGCGGCCGCTCGTTATTTTGCTTCTTTAACAAACTGTGC
P345	CCCAGTCAGGGGCTTTTT
Construction of V. cholerae EI deletion
P439	CACTGATTGGCCTTCTTCTG
P440	TAACGAGCGGCCGCAGCAGAAATTGAAGCTTGCGT
P441	TGCGGCCGCTCGTTATGGAGATGCTAGAATGCCTG
P442	CAGAAAACGGCCTTCACACT
Construction of V. cholerae HPr deletion
P443	CAGTGGCTTGCTCTTCGATA
P444	TAACGAGCGGCCGCAGCTCTGATGGACCAACTTCA
P445	TGCGGCCGCTCGTTAGATTTCTACTTGCTTCTCGT
P446	CTGTGTTACCAAGTTCAGG
Construction of V. cholerae EIIAGlc deletion
P447	CACGGTGCTTTCCCATTAAC
P448	TAACGAGCGGCCGCACCGATCCTACGCGTGACCAAG
P449	TGCGGCCGCTCGTTAACCCATTGTGTCATGCTCCT
P450	CTCTCTCCAGCGGTTCTGAC
Construction of V. cholerae EIIBCGlc deletion
P451	AGCTGTTTGGTGGAACTGGT
P452	TAACGAGCGGCCGCAGAGTGGATCCGTAACAACGG
P453	TGCGGCCGCTCGTTAGCTGGGACTAACGGCATACC
P454	CACCTTATCTCGCCCCAGTA
Construction of V. cholerae EI complementation fragment
P455	CGCTTTCGCTGCGATGAATTTATCTACGC
P456	ATTTCAGGCATTCTAGCATCTCCAGG
Construction of V. cholerae HPr complementation fragment
P457	TGAAGTTGGTCCATCAGAGC
P458	TACGAGAAGCAAGTAGAAATCAC
Construction of V. cholerae EIIAGlc complementation fragment
P459	CTTGGTCACGCGTAGGATCGGTG
P460	GGTCTGTTTGACAAACTTAAGAAGC
Construction of V. cholerae EI (H189A) complementation fragment
P510	CATGATAGAAGTGGCAGAGGTACGG
P511	CCGTACCTCTGCCACTTCTATCATG

The plasmids used to perform the rescue experiments were generated as follows. The genes at loci VC0964 (encoding EIIA^Glc), VC0965 (encoding EI), and VC0966 (encoding HPr) of the V. cholerae genome were amplified using the PCR. The gene encoding the EI (H189A) point mutant was generated by amplification of two gene fragments with overlapping internal primers P510 and P511 carrying a replacement of H189 for A189 (Table 1). These two fragments were ligated by splicing by overlap extension (9). Amplified products were cloned into a pBAD-TOPO expression vector (Invitrogen) to yield the rescue constructs pBAD-TOPO-EI, pBAD-TOPO-El (H189A), pBAD-TOPO-HPr, and pBAD-TOPO-EIIA^Glc. All insertions were confirmed by DNA sequence analysis.

All experiments were done in minimal medium (MM) alone or supplemented with 0.5% (wt/vol) of the specified sugar or carbon source (Sigma). The composition of the medium was adapted from Kapfhammer et al. (12) and is given in Table 2. Where noted, MM was supplemented with ampicillin (100 μg/ml) for plasmid maintenance and with 0.02% (wt/vol) l-arabinose to induce protein expression. A 0.1 M concentration of phosphate-buffered saline (PBS; pH 7.0) was used to rinse the cells.

TABLE 2.

Components of MM

Component	Concn
Salts
KCl	10 mM
CaCl2·2H2O	340 μM
MgSO4·7H2O	800 μM
FeSO4·7H2O	0.78 μM
KH2PO4	176 μM
K2HPO4	160 μM
(NH4)2SO4	60 μM
NaCl	100 mM
Amino acids
Alanine	0.3 g liter−1
Arginine HCl	0.2 g liter−1
Cysteine	0.2 g liter−1
Glycine	0.13 g liter−1
Histidine HCl·H2O	0.17 g liter−1
Isoleucine	0.3 g liter−1
Leucine	0.5 g liter−1
Lysine	0.5 g liter−1
Methionine	0.1 g liter−1
Phenylalanine	0.2 g liter−1
Serine	0.2 g liter−1
Threonine	0.2 g liter−1
Tyrosine	0.04 g liter−1
Valine	0.4 g liter−1
Sugars	0.5% (wt/vol)

Quantitative analysis of total growth and biofilm formation in MM.

Quantification of surface association was performed as described previously (13). Briefly, the strains to be tested were grown overnight on LB agar plates at 27°C. The following morning, bacterial colonies were resuspended in PBS and used to inoculate borosilicate tubes filled with 300 μl of the specified medium to yield an initial optical density at 655 nm (OD₆₅₅) of 0.05. After incubation for 24 h at 27°C, the planktonic cell suspension was removed, and the OD of this cell suspension was measured using a Benchmark Plus microplate spectrophotometer (Bio-Rad). The remaining surface-attached cells were dispersed in 300 μl PBS by vigorous vortexing in the presence of glass beads, and the OD of the resulting suspension was used to quantify surface attachment. Measurements of total growth were initially made by adding glass beads to a formed biofilm without removing planktonic cells, dispersing both planktonic and biofilm cells, and then recording an OD₆₅₅. We found, however, that this measurement was equivalent to the sum of separately made planktonic and biofilm OD₆₅₅ measurements. Thus, all reported measurements of total growth were made by the latter method. All measurements were performed in triplicate and repeated multiple times. Results are reported as a mean measurement. Error bars represent the standard deviation.

Flow cell experiments.

Biofilms were formed in flow cells consisting of a section of square borosilicate tubing measuring 70 mm (length) by 3 mm (height) by 3 mm (width) (Fiber Optic Center, Inc.) that was incorporated into a laminar flow circulation system driven by a Watson Marlow 205S peristaltic pump. The strains to be tested were incubated overnight in MM supplemented with glucose. After sterilization with sodium hypochlorite (Austin's), the flow cell was inoculated with 300 μl of the overnight culture diluted in MM supplemented with glucose to a final OD₆₅₅ of 0.1. After static incubation for 1 h at room temperature, flow was initiated at a constant rate of 3 ml h⁻¹. Biofilms were visualized after staining with the fluorescent dye FM 1-43 (Invitrogen) as follows. One hundred microliters of a 1-μg ml⁻¹ solution of FM 1-43 in sterile PBS was injected into the flow cell. After a 5-min incubation period, stained cells on the top surface of the flow chamber were visualized with an Eclipse TE-2000-E microscope (Nikon) equipped with a fluorescein isothiocyanate optical filter and an Orca digital charge-coupled device camera (Hamamatsu). A computer equipped with IP Lab (BD Biosciences) and AutoQuant X software (Media Cybernetics) was used for image acquisition and processing. In each experiment, strains were tested in duplicate, and all experiments were repeated multiple times.

Glucose consumption assays.

Strains were incubated for 24 h without agitation in tubes filled with MM supplemented with glucose. The concentration of glucose remaining in the culture medium at the end of the incubation was measured with the Sigma glucose (HK) assay kit. Briefly, cultures were centrifuged to pellet the cells, and 2 to 10 μl of supernatant was added to 200 μl of glucose assay reagent. After incubation for 5 min at room temperature, an A₃₄₀ was measured and the glucose concentration (mg/ml) was determined by the following relationship: [glucose] = (A₃₄₀ × total assay volume × dilution factor × 0.029)/sample volume.

Western blot analysis.

V. cholerae strains harboring a pBAD-TOPO expression vector encoding either wild-type EI or the mutant protein EI (H189A) fused to a C-terminal His tag were grown overnight at 37°C with shaking in MM supplemented with 0.5% glucose and 0.02% arabinose. A 1.5-ml volume of culture was centrifuged, and cells were resuspended in 1 ml MM supplemented with 0.5% glucose and 0.02% arabinose. An OD₆₅₅ was measured, and cultures were diluted as necessary to yield equivalent cell densities. Two hundred microliters of the adjusted culture was mixed with 200 μl of sample buffer (Bio-Rad) and then disrupted by boiling for 15 min and sonicating for 2 s. Proteins in the resulting suspension were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 12% polyacrylamide gel (NuSep). The separated proteins were transferred to a nitrocellulose membrane using a Mini Trans-Blot cell (Bio-Rad). The membrane was incubated first with a mouse anti-His (C terminal) primary antibody (Invitrogen; dilution 1:5,000) and then with an anti-mouse peroxidase-conjugated secondary antibody (Jackson ImmunoResearch; dilution 1:10,000). Reacting bands were visualized using the ECL enhanced chemiluminescence detection system (Amersham) according to the method specified by the manufacturer.

β-Galactosidase measurements of vpsL transcription.

For quantification of vpsL transcription, strains carrying a chromosomal vpsL-lacZ fusion were grown to the stationary phase in LB broth (pH 7.4) at 27°C with shaking. Twenty microliters of these cultures was used to inoculate tubes filled with 2 ml of MM supplemented with 0.05% glucose. These cultures were incubated overnight at 27°C. In the morning, the final OD₆₅₅ of each culture was measured and used for normalization of β-galactosidase activity as described below. An additional 1.5 ml of each culture was removed, centrifuged for 5 min at 3,000 rpm to pellet cells, washed with 1 ml of Z buffer, and then resuspended in 100 μl of Z buffer (19). After three freeze-thaw cycles, 17 μl of a 4-mg/ml solution of ONPG (o-nitrophenyl-β-d-galactopyranoside; Sigma) was added to the lysed cells, and this preparation was incubated at 37°C for 23 h. Cell debris was then removed by centrifugation, and the OD₄₁₅ of the resulting supernatant was measured. All β-galactosidase measurements are reported as the OD₄₁₅ of the supernatant divided by the final OD₆₅₅ of the respective culture. β-Galactosidase measurements were performed in triplicate and reported as a mean measurement. Error bars represent the standard deviation.

RESULTS

The PTS represses glucose-activated growth and biofilm formation.

We previously reported that mannose activates vps gene transcription, biofilm formation, and transcription of a subset of the V. cholerae genes encoding components of the PTS (13, 20, 21). Based on these observations, we hypothesized that the PTS might play an important role in biofilm formation in environments rich in PTS sugars. To test this, we first constructed a V. cholerae mutant (ΔPTS; Fig. 1B) containing an in-frame deletion spanning the genes encoding EIIA^Glc (VC0964), HPr (VC0965), and EI (VC0966). Based on our knowledge of the E. coli PTS, we predicted that this mutant would display reduced uptake of PTS-dependent sugars leading to a defect in total growth and biofilm formation. To determine whether our prediction was correct, wild-type V. cholerae and the ΔPTS mutant were incubated for 24 h in borosilicate tubes filled with MM either alone or supplemented with mannose, glucose, or galactose. After this incubation, we quantified the total density of cells in the final culture as well as the density of cells associated with the surface. As shown in Fig. 2A, both wild-type V. cholerae and the ΔPTS mutant formed a sparse biofilm in the absence of monosaccharides. When MM was supplemented with mannose, a large increase in total growth and biofilm formation was observed for wild-type V. cholerae. In contrast, when cultured in MM supplemented with mannose, total growth and biofilm formation by the ΔPTS mutant were similar to those of wild-type V. cholerae cultured in the absence of mannose. This suggested to us that V. cholerae uses the PTS as its primary mode of mannose transport. Growth of both wild-type V. cholerae and the ΔPTS mutant in MM supplemented with galactose led to similar increases in total growth but no increases in biofilm-associated growth, suggesting that galactose is not transported by the PTS and that stimulation of surface-associated growth may be unique to PTS transport. Because the PTS is the main transport system for glucose in E. coli, we predicted that growth of wild-type V. cholerae and the ΔPTS mutant in MM supplemented with glucose would yield results similar to those observed for mannose supplementation. To our surprise, however, in the presence of glucose, deletion of the PTS genes led to a dramatic increase in both total growth and biofilm formation. These results lead us to hypothesize that, when V. cholerae is grown in the presence of glucose, the V. cholerae PTS is able to repress growth and biofilm formation.

FIG. 2.

Characterization of the growth of a ΔPTS mutant in static culture and in a flow cell. (A) Quantification of total (TOT) growth and biofilm (BF)-associated cell growth by wild-type (WT) V. cholerae and a ΔPTS mutant in MM alone or supplemented with sugars as specified (0.5% [wt/vol]). (B) Accumulation of total cells, planktonic (PK) cells, and biofilm-associated cells over time for wild-type V. cholerae and a ΔPTS mutant (PTS). (C) Comparison of total growth and biofilm growth by wild-type V. cholerae as well as ΔPTS, ΔvpsA, and ΔPTS ΔvpsA mutants. (D) Wild-type V. cholerae and ΔPTS mutant biofilms formed in a flow cell chamber over 24 h. Biofilms were imaged by fluorescence microscopy after staining with the fluorescent lipid-soluble dye FM-43.

The PTS specifically regulates growth of V. cholerae in a biofilm.

Transport of glucose and mannose by the PTS activates biofilm accumulation, but biofilm accumulation in glucose-rich medium is increased further by deletion of the PTS. Because of these seemingly contradictory effects of the PTS on biofilm growth, we hypothesized that the impact of the PTS on biofilm accumulation might vary with growth phase. To test this, we quantified total growth, planktonic growth, and biofilm-associated cell growth over time for both wild-type V. cholerae and a ΔPTS mutant cultured in MM supplemented with glucose. As shown in Fig. 2B, accumulation of both planktonic and surface-attached wild-type V. cholerae reached a plateau after 19 h of incubation in this medium. Interestingly, while accumulation of ΔPTS mutant planktonic cells also reached a plateau after 19 h, surface-associated cells continued to increase. This resulted in greater total growth and biofilm accumulation by the ΔPTS mutant compared to wild-type V. cholerae. These results suggested to us that the PTS might play a unique role in regulation of biofilm-associated growth.

The PTS does not regulate growth of a biofilm-defective mutant.

We hypothesized that if the PTS was a surface-specific regulator of growth, it would have no effect on the growth of a V. cholerae mutant that was unable to form a biofilm. To test our hypothesis, we compared total growth and surface-associated growth of wild-type V. cholerae, a ΔPTS mutant, a biofilm-defective mutant carrying a deletion of the vpsA operon, and a biofilm-defective mutant carrying a deletion of both the PTS genes and the vpsA operon. As shown in Fig. 2C, while deletion of the PTS genes in a wild-type background resulted in greater total growth and surface-associated growth as compared with wild-type V. cholerae, deletion of these genes in a ΔvpsA mutant background had no effect on total growth and surface-associated growth. These results further support our hypothesis that regulation of V. cholerae growth by the PTS is limited to surface-associated cells.

Flow cell studies replicate the PTS biofilm phenotype.

Because of the potential for active recruitment of planktonic cells, surface-specific modulation of growth cannot be adequately evaluated in static culture. Therefore, we also compared surface-associated growth of wild-type V. cholerae and the ΔPTS mutant in a flow cell. As shown in Fig. 2D, after 24 h of growth in a flow cell, greater surface accumulation was observed for the ΔPTS mutant than for wild-type V. cholerae. This observation further supports the hypothesis that the PTS regulates the growth of biofilm-associated V. cholerae.

Identification of V. cholerae PTS components that are involved in glucose transport.

After establishing that a V. cholerae ΔPTS mutant displayed increased surface-associated growth in the presence of glucose, our goal was to establish the role of these putative PTS components in glucose transport and to identify additional glucose-specific PTS components. For this purpose, mutants were constructed carrying single in-frame deletions of each component of the putative glucose-specific PTS transport pathway, namely the genes encoding EI, HPr, EIIA^Glc, and EIIBC^Glc. The corresponding mutants were first tested for their ability to consume glucose present in the culture medium. As shown in Fig. 3A, at the end of a 24-h incubation, wild-type V. cholerae had consumed all of the glucose in the medium. Although the amount of glucose transported by each PTS mutant was variable, all consumed less glucose than wild-type V. cholerae. These data suggest we have identified the components of the V. cholerae glucose PTS. They also suggest that V. cholerae is able to consume environmental glucose in the absence of a functional PTS.

FIG. 3.

Characterization of the ΔPTS mutant as well as ΔEI, ΔHPr, ΔEIIA^Glc, and ΔEIIBC^Glc mutants. (A) Total glucose consumed. (B) Total growth and biofilm formation in MM supplemented with 5 mg/ml glucose alone or also with 500 μM cAMP. WT, wild type.

Deletion of EI reproduces the phenotype of the ΔPTS mutant.

To determine which of the components of the PTS regulates surface-associated growth, we quantified total growth and biofilm-associated growth by wild-type V. cholerae as well as ΔEI, ΔHPr, ΔEIIA^Glc, and ΔEIIBC^Glc mutants (Fig. 3B). Only the ΔEI mutant demonstrated an increase in total growth and biofilm-associated growth that was similar to that of the ΔPTS mutant. Interestingly, deletion of the genes encoding the other three components of the PTS resulted in decreased biofilm formation compared with wild-type V. cholerae.

EI rescues the phenotype of the ΔPTS mutant.

We reasoned that if a single component of the PTS were responsible for regulation of V. cholerae surface attachment, a plasmid encoding this component should rescue the phenotype of the ΔPTS mutant. To test this, we constructed the expression vectors pBAD-TOPO-EI, pBAD-TOPO-HPr, pBAD-TOPO-EIIA^Glc carrying the genes encoding EI, HPr, or EIIA^Glc, respectively. To confirm that these plasmids were functional, we first tested their ability to rescue the phenotypes of the corresponding single PTS mutants. Total growth and biofilm formation were compared for wild-type V. cholerae rescued with a control plasmid, the relevant deletion mutant rescued with a control plasmid, and the same deletion mutant rescued with a plasmid encoding the wild-type protein. As shown in Fig. 4A, B, and C, we were able to document complete rescue of the ΔEI, ΔHPr, and ΔEIIA^Glc mutant phenotypes by pBAD-TOPO-EI, pBAD-TOPO-HPr, and pBAD-TOPO-EIIA^Glc, respectively. Of note, the plasmid encoding HPr yielded reproducible rescue of the ΔHPr mutant phenotype only when it was tested immediately after electroporation into the ΔHPr mutant. Thus, mutants were newly transformed with pBAD-TOPO-HPr prior to each experiment.

FIG. 4.

Rescue of the various PTS mutant phenotypes with single components of the PTS. Quantification of total (TOT) growth and biofilm (BF) growth is presented for a ΔEI mutant harboring a pBAD expression vector encoding either β-galactosidase (lacZ), wild-type EI (EI), or an unphosphorylatable mutant of EI [EI (H189A)] (A); a ΔHPr mutant harboring a pBAD expression vector encoding either β-galactosidase (lacZ) or wild-type HPr (HPr) (B); a ΔEIlA^Glc mutant harboring a pBAD expression vector encoding either β-galactosidase (lacZ) or wild-type EIIA^Glc (EIIA) (C); and a ΔPTS mutant harboring a pBAD expression vector encoding either β-galactosidase (lacZ), wild-type EI (EI), an unphosphorylatable mutant of EI [EI (H189A)], wild-type HPr (HPr), or wild-type EIIA^Glc (EIIA^Glc) (D). In each case, wild-type V. cholerae (WT) rescued with the control vector pBAD-lacZ is shown as a control. (E) Western blot demonstrating expression of pBAD-encoded wild-type EI (pBAD-EI) and EI (H189A) proteins in the ΔEI and ΔPTS mutant backgrounds. In all experiments, the medium was supplemented with 0.02% arabinose to induce protein expression.

To determine if any single component of the PTS was able to rescue the phenotype of the ΔPTS mutant, we transformed each of these plasmids into the ΔPTS mutant and measured total growth and biofilm formation. As shown in Fig. 4D, transformation of the ΔPTS mutant with the rescue construct encoding EI restored levels of total growth and biofilm formation to those measured for wild-type V. cholerae. Rescue constructs encoding HPr and EIIA^Glc, however, had no effect on total growth and biofilm formation by the ΔPTS mutant. These data support the conclusion that EI is responsible for repression of biofilm-associated growth by the PTS.

Evidence that EI must be phosphorylated to repress surface-associated growth.

We have demonstrated that biofilm-associated growth is activated by deletion of EI and that deletion of HPr or EIIA^Glc alone leads to repression of biofilm-associated growth. Because a deletion of either HPr or EIIA^Glc would be predicted to block phosphotransfer through the PTS leading to an increase in the phosphorylated form of EI, we formed the hypothesis that repression of surface-associated growth requires the phosphorylated form of EI. An alternative method for blocking the PTS phosphotransfer cascade that does not rely on mutagenesis is the cultivation of wild-type V. cholerae in the absence of PTS carbon sources. We hypothesized that if the phosphorylated form of EI (EI-P) were responsible for repression of surface-associated growth, surface-associated growth should be maximally repressed when wild-type V. cholerae is cultured in the presence of non-PTS carbon sources. To test this, we compared total growth and biofilm formation by wild-type V. cholerae and a ΔEI mutant in the presence of a number of non-PTS carbon sources. As shown in Fig. 5, wild-type V. cholerae grew well in the presence of these non-PTS carbon sources. However, wild-type V. cholerae surface association was negligible. Deletion of EI resulted in large increases in surface association in the presence of these non-PTS carbon sources, suggesting that, in the absence of EI, non-PTS carbon sources are, indeed, able to support biofilm-associated growth. These data support a role for EI-P in repression of surface-associated growth.

FIG. 5.

Quantification of total (TOT) cell growth and biofilm (BF)-associated cell growth after incubation of wild-type (WT) V. cholerae or the ΔEI mutant (EI) for 24 h in MM alone or supplemented with the indicated non-PTS carbon sources.

An expression vector encoding an EI mutant that is locked in the unphosphorylated state does not rescue the ΔEI and ΔPTS mutant phenotypes.

The observations described above suggested that the phosphorylated form of EI was responsible for repression of biofilm-associated growth. To test this more directly, an expression vector encoding a form of EI that is locked in the unphosphorylated state was constructed by replacement of the conserved histidine residue in charge of phosphorylation transfer with alanine (H189A). Both the ΔEI and ΔPTS mutants were transformed with this plasmid. We first confirmed that expression of the plasmid-encoded EI (H189A) was comparable to that of the plasmid-encoded wild-type EI (Fig. 4E). In spite of excellent expression of the mutant protein, EI (H189A) was unable to rescue either the EI mutant (Fig. 4A) or the PTS mutant (Fig. 4D). These results further support the hypothesis that the phosphorylated form of EI is required for repression of biofilm-associated growth.

EI-P represses vps gene transcription.

Because biofilm formation is related to exopolysaccharide synthesis, we questioned whether EI-P was acting at the transcriptional level. To address this, β-galactosidase activity was measured in wild-type V. cholerae and ΔEI, ΔHPr, ΔEIIA^Glc, and ΔEIIBC^Glc mutants carrying chromosomal vpsL-lacZ reporter fusions (Fig. 6). Indeed, the level of β-galactosidase activity measured for a ΔEI mutant cultured in MM supplemented with glucose was higher than that measured for wild-type V. cholerae. In contrast to the ΔEI mutant, ΔHPr, ΔEIIA^Glc, and ΔEIIBC^Glc mutants displayed decreased β-galactosidase activity. These mutants are predicted to harbor higher levels of EI-P due to a block in the PTS phosphotransfer cascade. These data suggest that the phosphorylated form of EI also acts to repress transcription of the vps genes. Therefore, while EI-P may act at other levels as well, EI-P acts at the transcriptional level to repress surface association.

FIG. 6.

Quantification of vpsL transcription in wild-type (WT) V. cholerae or ΔEI, ΔHPr, ΔEIIA^Glc, and ΔEIIBC^Glc mutants harboring a chromosomal fusion of the vpsL promoter to the lacZ gene.

The β-galactosidase activity measured for the ΔHPr mutant was intermediate between that measured for the ΔEI mutant and that measured for the ΔEIIA^Glc and ΔEIIBC^Glc mutants. We hypothesize that this intermediate phenotype is due to the presence of another functional HPr homolog in the V. cholerae genome. In fact, deletion of the HPr homolog encoded at VC0966 significantly decreases but does not completely block sugar transport through the PTS (data not shown).

Catabolite repression effected by EIIA^Glc does not contribute to activation of surface association by the V. cholerae PTS.

In E. coli, cAMP synthesized by adenylyl cyclase serves as a second messenger that activates utilization of non-PTS carbon sources in the cell. Adenylyl cyclase, in turn, is activated by EIIA^Glc in its phosphorylated state. Thus, when PTS sugars are scarce, EIIA^Glc is maintained in its phosphorylated state, adenylyl cyclase is activated, and increased intracellular levels of cAMP enable utilization of non-PTS substrates (6).

It has previously been demonstrated that cAMP activates biofilm formation by E. coli (10). We hypothesized that if V. cholerae EIIA^Glc also activated adenylyl cyclase, the low levels of cAMP in a ΔEIIA^Glc mutant might contribute to the observed decrease in surface association for this mutant. If this were the case, we reasoned that addition of cAMP to the growth medium should at least partially rescue the biofilm formation defect of the ΔEIIA^Glc mutant. However, we observed that at concentrations as low as 500 μM, cAMP actually repressed total growth and biofilm accumulation by wild-type V. cholerae and the ΔPTS and ΔEI mutants (Fig. 3B). Supplementation of the growth medium with a variety of concentrations of cAMP from 0 to 2 mM had no effect on total growth and biofilm accumulation by the ΔEIIA^Glc mutant (Fig. 3B and data not shown). This suggests that, in contrast to what is observed for E. coli, catabolite repression effected by EIIA^Glc does not activate surface association by V. cholerae.

Supplementation of the medium with concentrations as low as 500 μM cAMP reduced environmental glucose consumption by the ΔPTS and ΔEI mutants but not wild-type V. cholerae (Fig. 3A), suggesting that cAMP also represses glucose consumption by V. cholerae in the absence of EI.

DISCUSSION

This study arose from our observation that PTS sugars such as mannose and glucose activate biofilm formation by V. cholerae, while non-PTS sugars such as galactose do not. Because very few studies of the V. cholerae PTS exist, we first set out to identify components of the PTS that were responsible for glucose transport. Our measurements suggest that glucose transport in V. cholerae is accomplished through both PTS and non-PTS pathways.

We have demonstrated that mutation of EI activates the growth of cells on a surface. This is not due to a decrease in the intracellular supply of glucose-6-phosphate. First, while deletion of the gene encoding EI has only a small impact on glucose transport, activation of surface-associated growth is considerable. Furthermore, supplementation of the medium with glucose-6-phosphate leads to an increase in growth of wild-type V. cholerae but does not rescue the phenotype of the ΔEI mutant (Fig. 5).

In comparison with wild-type V. cholerae, the ΔEI mutant displays the surprising phenotype of increased total growth in the face of decreased glucose transport. This suggests either that transport or catabolism of other carbon sources present in the growth medium such as amino acids is increased in the ΔEI mutant or that glucose utilization is altered in the ΔEI mutant. Specifically, as cells reach stationary phase, EI-P may coordinate cessation of growth with allocation of glucose to intracellular stores such as glycogen.

We have also shown that the phenotype of the ΔEI mutant does not arise from the effect of EI on the phosphorylation state of downstream PTS components. The ΔEI and ΔPTS mutants display similar increases in total growth and surface accumulation compared with wild-type V. cholerae, and the ΔPTS mutant can be rescued by delivery of the gene encoding the EI protein in trans, This suggests that EI is necessary and sufficient for repression of biofilm growth but HPr and EIIA^Glc are not.

Lastly, we have provided evidence that phosphorylation of EI leads to repression of surface-associated growth as depicted in Fig. 7. First, under two conditions where one would predict a high abundance of EI-P, namely in an EII mutant background (Fig. 7B) and during growth of wild-type V. cholerae in media containing non-PTS sugars (Fig. 7C), strong repression of surface-associated growth is observed. This repression is relieved by deletion of EI. Second, while EI provided in trans is able to rescue both the ΔEI and ΔPTS mutant phenotypes, a mutant EI protein locked in the unphosphorylated state is not able to rescue the phenotype of either of these mutants. While we have not yet studied the mechanism by which EI-P acts, we hypothesize that EI-P, a protein known to transfer phosphate, most likely transfers its phosphate to another regulatory molecule to effect repression of surface-associated growth. Experiments to address this hypothesis are under way in our laboratory.

FIG. 7.

Schematic depiction of vps gene regulation by the PTS based on our findings for wild-type (WT) V. cholerae grown in glucose-rich medium (A), for ΔEIIA and ΔEIIBC mutants grown in glucose-rich medium (B), and for wild-type V. cholerae grown in medium containing a non-PTS carbon source (C).

Very little is known about multiplication of cells enclosed in biofilm structures. Because of the high cell density and limited diffusion in the biofilm, mature biofilms have been likened to stationary-phase cultures, in which growth and metabolic activity decrease and the toxic end products of metabolism accumulate (14, 26). Our results demonstrate that the plateau in cell division within the biofilm is not simply the result of increased spatial constraints, decreased access to nutrients, or the accumulation of metabolites. Instead, the growth rate within the biofilm is an actively regulated process. Because accumulation of surface-associated cells is increased in the ΔPTS mutant even under conditions of flow, we suggest that the PTS, and, in particular, EI-P, is a biofilm-specific regulator of cell growth. Interestingly, EI is the first reported regulator to connect monosaccharide availability with cell division on surfaces and vps expression.

We propose that the novel regulatory pathway described here is relevant to survival of V. cholerae both in the environment and in the mammalian intestine. In the environment, the ability to respond to decreasing levels of glucose and other PTS sugars by arresting growth may enhance the survival fitness of the biofilm community. In contrast, in planktonic cells, the more important response to decreasing levels of glucose may be chemotaxis, a function in which EI has also been implicated (16). In the human small intestine, glucose is the principal metabolic sugar produced during intralumenal digestion of polysaccharides (23). Therefore, we hypothesize that glucose sensing and transport are important in colonization of and survival within the small bowel. In support of this, microarray analysis has demonstrated that transcription of the genes encoding EI and HPr is activated in vivo (28). Moreover, a signature-tagged mutagenesis screen has identified the V. cholerae EI as a potential colonization factor in the infant mouse intestine (18).

In enteric bacteria, phosphorylated EIIA^Glc is the key regulator of catabolic repression via the activation of adenylyl cyclase. Furthermore, glucose inhibits surface association and activation of adenylyl cyclase by EIIA^Glc-P is known to increase biofilm formation. Interestingly, in V. cholerae, a different behavior is observed. PTS sugars such as glucose activate biofilm development, cAMP inhibits total growth and surface association, and regulation of surface association by the PTS is independent of catabolite repression effected by EIIA^Glc. We suggest that this may reflect the different intestinal habitats of V. cholerae, which is thought to colonize the nutrient-rich epithelium of the small intestine, and enteric bacteria such as E. coli, which colonize the nutrient-depleted colon.

In both gram-negative and gram-positive bacteria, the PTS plays a central role in coordinating environmental sensing, transport, catabolism, and storage of monosaccharides with environmental supply. However, a role for the PTS in repression of biofilm-associated growth has not been described previously. In this article, we document a new role for the phosphorylated form of V. cholerae EI in repression of surface-associated cell growth and exopolysaccharide synthesis.