1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Biofilm development is conceived as a developmental process in which free swimming cells attach to a surface, first transiently and then permanently, as a single layer. This monolayer of immobilized cells gives rise to larger cell clusters that eventually develop into the biofilm, a three-dimensional structure consisting of large pillars of bacteria interspersed with water channels. Previous studies have shown that efficient development of the Vibrio cholerae biofilm requires a combination of pili, flagella and exopolysaccharide. Little is known, however, regarding the requirements for monolayer formation by wild-type V. cholerae. In this work, we have isolated the wild-type V. cholerae monolayer and demonstrated that the environmental signals, bacterial structures, and transcription profiles that induce and stabilize the monolayer state are unique. Cells in a monolayer are specialized to maintain their attachment to a surface. The surface itself activates mannose-sensitive haemagglutinin type IV pilus (MSHA)-mediated attachment, which is accompanied by repression of flagellar gene transcription. In contrast, cells in a biofilm are specialized to maintain intercellular contacts. Progression to this stage occurs when exopolysaccharide synthesis is induced by environmental monosaccharides. We propose a model for biofilm development in natural environments in which cells form a stable monolayer on a surface. As biotic surfaces are degraded with subsequent release of carbohydrates, the monolayer develops into a biofilm.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

In most natural environments, microbes are found attached to surfaces in communal structures known as biofilms. Associations between microbes and surfaces and among microbes in a biofilm may increase the availability of nutrients to certain members of the community and may also expand the community's repertoire of defences against environmental stresses (Hoyle et al., 1992; Bower and Daeschel, 1999; Yildiz and Schoolnik, 1999; Mah and O’Toole, 2001). Thus, biofilms may provide microbes with an adaptive advantage in the environment.

Genetic and microscopic studies of biofilm formation by Pseudomonas aeruginosa and other Gram-negative organisms have suggested that development of a mature, three-dimensional biofilm involves the following consecutive, discrete stages: (i) the planktonic stage (ii) the monolayer stage, and, finally (iii) the biofilm (Costerton, 1995; O’Toole et al., 2000; Watnick and Kolter, 2000; Danese et al., 2001). The planktonic stage consists of free-swimming bacteria. These bacteria carry appendages such as flagella or pili that may accelerate biofilm development (Pratt and Kolter, 1998; Kachlany et al., 2001; Sauer and Camper, 2001; Kang et al., 2002; Bechet and Blondeau, 2003; Di Martino et al., 2003). It should be noted, however, that many Gram-positive and Gram-negative bacteria are able to form biofilms in the absence of pili and flagella (McKenney et al., 1998; Watnick and Kolter, 1999; Toledo-Arana et al., 2001; Watnick et al., 2001; Klausen et al., 2003; Reisner et al., 2003). Thus, these appendages are not absolutely required for surface attachment.

The bacterial monolayer is formed by single cells that initially associate with the surface transiently and then become permanently immobilized. Very few genetic studies of the wild-type bacterial monolayer exist. This is because the wild-type bacterial monolayer is an intermediate stage in biofilm development and therefore difficult to isolate. The monolayer state has been isolated by mutagenesis, however. In P. fluorescens, a secreted protein has been identified that is required for the transition from transient to permanent attachment (Hinsa et al., 2003). Furthermore, in several Gram-negative organisms including V. cholerae and P. aeruginosa, mutations that arrest biofilm development in the monolayer state have been identified. In LB broth, mutation of genes involved in exopolysaccharide synthesis arrests V. cholerae cells in the monolayer stage of development (Watnick and Kolter, 1999). Mutation of genes involved in type IV pilus biogenesis arrest P. aeruginosa in the monolayer state (O’Toole and Kolter, 1998). Because of the paucity of monolayer-associated cells, quantification of gene transcription in this state is a challenge.

The final stage in biofilm development is defined by the formation of a highly organized, three-dimensional structure composed of bacterial pillars surrounded by water channels through which nutrients and waste products may be transported (Costerton et al., 1995; Kolter and Losick, 1998). Although the biofilm itself is heterogeneous, some general genetic and transcriptional features of the biofilm stage have been observed. Exopolysaccharide is required for the three-dimensional structure of most types of biofilms (Bonet et al., 1993; Allison et al., 1998; Watnick and Kolter, 1999; Danese et al., 2000). Furthermore, gene transcription in the biofilm is distinct from that in planktonic cells (Sauer and Camper, 2001; Whiteley et al., 2001; Tremoulet et al., 2002; Wen and Burne, 2002; Yoshida and Kuramitsu, 2002; Sauer, 2003; Schembri et al., 2003; Stanley et al., 2003). Genes required for the synthesis of flagella and pili, which are required for motility and surface attachment, are repressed in the biofilms of many Gram-negative bacteria (Sauer and Camper, 2001; Whiteley et al., 2001; Sauer et al., 2002; Schembri et al., 2003). These structures are only required during the initial stages of biofilm development and have the potential to destabilize the mature biofilm. In contrast, exopolysaccharide synthesis genes, which are crucial to the maintenance of the biofilm structure, show increased transcription in biofilm cells. Thus, progression to the biofilm state from the planktonic state requires a change in the gene transcription, the extracellular matrix composition, and the three-dimensional structure of the bacterial population.

Vibrio cholerae is both a natural inhabitant of aquatic environments and the intestinal pathogen that causes cholera (Colwell and Spira, 1992; Colwell and Huq, 1994; Colwell, 1996). It has been found in diverse marine, estuarine and fresh water environments in association with zooplankton, phytoplankton, crustaceans, insects and plants (Huq et al., 1983; 1986; Islam et al., 1989; Tamplin et al., 1990; Shukla et al., 1995; Castro-Rosas and Escartin, 2002). Adhesion to surfaces during infection of the human intestine and colonization of the aquatic environment may play an important role in the success of V. cholerae as a pathogen and an environmental organism. The toxin co-regulated pilus (TCP), a type IV bundle forming pilus, is primarily responsible for colonization of the mammalian intestine by V. cholerae (Taylor et al., 1987; Herrington et al., 1988; Thelin and Taylor, 1996; Tacket et al., 1998). In contrast, in LB broth, attachment of V. cholerae O1 El Tor to abiotic surfaces is accelerated by the mannose-sensitive haemagglutinin type IV pilus (MSHA) and the polar flagellum (Watnick et al., 1999; Chiavelli et al., 2001). Interestingly, the dependence of V. cholerae biofilm development on MSHA and the flagellum appears to be strain-dependent (Watnick and Kolter, 1999; Watnick et al., 2001). This suggests that V. cholerae produces a third mediator of initial attachment whose regulation varies among strains. In all V. cholerae strains studied, formation of a three-dimensional biofilm structure requires the vps genes, which encode enzymes involved in the synthesis of VPS, an exopolysaccharide found in the biofilm matrix (Wai et al., 1998; Mizunoe et al., 1999; Watnick and Kolter, 1999; Yildiz and Schoolnik, 1999). Regulation of VPS synthesis is complex. vps transcription, VPS synthesis, and biofilm development are activated by many environmental signals including (i) addition of monosaccharides to the growth medium and (ii) incorporation into a biofilm (Haugo and Watnick, 2002; Kierek and Watnick, 2003). Repressors of vps transcription, VPS synthesis and biofilm development have also been identified. High cell densities repress VPS synthesis and biofilm development through the action of HapR (Zhu et al., 2002; Hammer and Bassler, 2003; Vance et al., 2003; Zhu and Mekalanos, 2003). This suggests that quorum sensing acts to limit accumulation of cells on a surface. The flagellum is also likely to be a repressor of biofilm development as flagellar mutants display inappropriate VPS synthesis (Watnick et al., 2001). This appears to be a common theme as inverse regulation of exopolysaccharide and flagellum synthesis has been observed in other organisms (Burdman et al., 1998; Garrett et al., 1999). Lastly, CytR, a repressor of nucleoside uptake and catabolism, has been shown to repress vps transcription and decrease surface adhesion (Haugo and Watnick, 2002). VpsR, a σ54-dependent transcriptional activator, appears to be a central regulator of vps transcription through which many of the pathways controlling biofilm development pass (Yildiz et al., 2001). Strains harbouring engineered mutations in many of the pathways that repress VPS synthesis display a wrinkled or rugose colony morphology on LB-agar plates (Jobling and Holmes, 1997; Watnick et al., 2001; Haugo and Watnick, 2002; Hammer and Bassler, 2003; Zhu and Mekalanos, 2003). Increased VPS synthesis has been demonstrated in these colonies. Spontaneous rugose strains also arise (White, 1938; Morris et al., 1996; Wai et al., 1998; Mizunoe et al., 1999; Yildiz and Schoolnik, 1999; Ali et al., 2002). It has been suggested that mutations in the hapR gene give rise to these rugose strains (Hammer and Bassler, 2003).

Biofilm formation has been described as a developmental process in which the bacterial cell evolves from a planktonic cell to a monolayer cell to a terminally differentiated biofilm cell (Costerton, 1995; Palmer and White, 1997; O’Toole et al., 2000; Watnick and Kolter, 2000; Danese et al., 2001). Whereas studies of the planktonic and biofilm stages of biofilm development have begun to provide insights into the genetic requirements and transcription profiles of the biofilm-associated cell, the monolayer stage has not been studied using these techniques. Furthermore, the monolayer has not been proven to be a developmental stage which is distinct from the planktonic and biofilm stages both in its structural requirements and in its profile of gene transcription.

In this work, we have developed a growth medium that allows us to arrest wild-type V. cholerae cells in the monolayer stage and study this monolayer independently from the biofilm. Using this medium, we show that monolayer cells have genetic requirements and transcription profiles that are distinct from both planktonic and biofilm-associated cells.

Results and discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

A minimal medium lacking monosaccharides arrests V. cholerae O139 cells in the monolayer stage of biofilm development

Most studies that are relevant to V. cholerae biofilm development have been performed in LB broth (Wai et al., 1998; Mizunoe et al., 1999; Watnick et al., 1999; 2001; Watnick and Kolter, 1999; Yildiz and Schoolnik, 1999; Nesper et al., 2001; Yildiz et al., 2001; Haugo and Watnick, 2002; Zhu et al., 2002;Bomchil et al., 2003; Hammer and Bassler, 2003; Vance et al., 2003). In this medium, progression of cells from the planktonic state to association with a three-dimensional biofilm is not synchronized. However, at any point in biofilm development, the majority of cells are found either in the planktonic state or in association with a biofilm, and this feature of biofilm development in LB broth has allowed comparative studies of planktonic and biofilm-associated cells. Because the monolayer is a transient stage in biofilm development by wild-type V. cholerae, it has been more difficult to isolate and study. As a result, very little is known about the genetics of wild-type V. cholerae monolayer formation.

We hypothesized that if cells in the monolayer were in rapid exchange with the planktonic state, monolayer cells would be most like planktonic cells. If the monolayer stage was merely a biofilm prior to significant accumulation of cells, monolayer cells would be indistinguishable from biofilm cells. Finally, the most interesting possibility was that cells in the monolayer would display characteristics that were distinct both from those of planktonic cells and those of biofilm-associated cells. To distinguish between these possibilities, we sought to isolate and study the V. cholerae monolayer. We had recently shown that V. cholerae forms a vps-dependent biofilm in amino acid-based media only if it is supplemented with monosaccharides (Kierek and Watnick, 2003). This suggested to us that it would be possible to isolate and study the V. cholerae monolayer using a defined medium that did not contain monosaccharides. To test this, we devised a minimal medium (MM) containing minerals and a mixture of purified amino acids but lacking monosaccharides (see Table 2). Wild-type V. cholerae cells cultured in this medium attached singly to the surface of the culture vessel and formed a confluent monolayer (Fig. 1A). Throughout this process, transiently attached cells were noted to wriggle about on the surface, detach, and either re-attach at another location or swim away. Progression to a multilayered structure was not seen over the course of 24 h. Supplementation of MM with mannose, however, led to the formation of microcolonies and eventually to a multilayered biofilm (see Fig. 3A). Thus, we identified conditions that would enable us to isolate the monolayer stage and then induce progression to the biofilm.

Table 2. . Medium composition.
Basal salts
KCl10 mM
CaCl2.2H2O340 µM
MgSO4.7H2O800 µM
FeSO4.7H2O0.6 µM
Phosphate ammonium mix
KH2PO488 µM
K2HPO4160 µM
(NH4)2SO460 µM
NaCl100 mM
Amino acids
Alanine0.3 g l−1
ArginineHCl0.2 g l−1
Cysteine0.2 g l−1
Glycine0.13 g l−1
HistidineHCl.H2O0.17 g l−1
Isoleucine0.3 g l−1
Leucine0.5 g l−1
Lysine0.5 g l−1
Methionine0.1 g l−1
Phenylalanine0.2 g l−1
Serine0.2 g l−1
Threonine0.2 g l−1
Tyrosine0.04 g l−1
Valine0.4 g l−1

Figure 1. Monolayer formation by wild-type V. cholerae (WT, MO10) and ΔvpsA (PW396), ΔflaA (PW412), and ΔmshA (PW361) mutants after incubation in MM alone at 27°C for 24 h.

A. Phase-contrast micrographs of monolayers. Bar = 10 µm.

B. Total surface area covered by surface-attached cells. Observations represent the mean of three separate experiments.


Figure 3. Biofilm formation by wild-type V. cholerae (WT, MO10) and ΔvpsA (PW396), ΔflaA (PW412), ΔmshA (PW361), ΔmshAΔvpsL (PW448), and ΔflaAΔvpsL (PW450) mutants in MM supplemented with mannose.

A. Phase-contrast micrographs of biofilm formation. Bar = 10 µm.

B. Comparison of mean cluster size of surface-associated cells in MM alone and supplemented with mannose as noted in the key. Observations represent the mean of three separate experiments.

MSHA but not the flagellum is critical for V. cholerae O139 monolayer formation in MM

Previous experiments suggested that, in LB broth, the flagellum was required for initial surface attachment by V. cholerae O1 El Tor and O139 (Watnick and Kolter, 1999; Watnick et al., 2001). In contrast, MSHA was required for initial surface attachment by V. cholerae O1 El Tor but not V. cholerae O139. Thus, we hypothesized that the flagellum and not MSHA would be required for monolayer formation by V. cholerae O139. To test this hypothesis, we documented and quantified wild-type and mutant monolayer formation as follows. Wild-type V. cholerae O139 (strain MO10) as well as ΔmshA,ΔflaA, and ΔvpsA mutants were incubated in polystyrene microtitre wells filled with MM, wells were rinsed, and the bottoms of the wells were visualized by phase-contrast microscopy (Fig. 1A). Wild-type cells as well as the ΔvpsA and the ΔflaA mutants formed a confluent monolayer on the bottom of the well. In contrast, the ΔmshA mutant was unable to attach and form a monolayer. Using image analysis software, we quantified the well surface area covered by wild-type V. cholerae and mutant cells. As shown in Fig. 1B, the surface area covered by wild-type V. cholerae,ΔvpsA mutant, and ΔflaA mutant cells was substantial. In contrast, the surface area covered by ΔmshaA cells was negligible. This suggested that MSHA but not the flagellum is required for monolayer formation.

The flagellum has previously been implicated in surface attachment by V. cholerae. Furthermore, the ΔflaA mutant naturally accumulates at the bottom of the well. To ensure that the monolayer formed by the ΔflaA mutant was truly attached to the surface, we evaluated attachment by a ΔflaAΔmshaA double mutant. Parameters for the monolayer formed by this mutant were similar to those for the monolayer formed by the ΔmshaA mutant (data not shown). These results suggest that the ΔflaA mutant forms a true monolayer and that, as is the case for wild-type V. cholerae, formation of the monolayer is dependent on MSHA.

α-Methylmannoside (AMM) interferes with V. cholerae monolayer attachments

Alpha-methylmannoside is a non-metabolizable form of mannose that is used to interrupt attachments mediated by interaction with a mannose moiety. Alpha-methylmannoside has previously been used to study mannose-sensitive interactions of E. coli pili with plastic (Pratt and Kolter, 1998). Because MSHA attachments are mannose-sensitive and MSHA was required for monolayer formation by wild-type V. cholerae and the ΔflaA mutant, we hypothesized that monolayer formation by these two strains would be sensitive to AMM. To test this, we incubated wild-type V. cholerae with a surface bathed in MM supplemented with AMM. In this case, very few cells attached to the surface (data not shown). We then questioned whether addition of AMM after monolayer formation could disrupt V. cholerae attachments to the surface. To address this, we added AMM to the medium after monolayer formation by wild-type V. cholerae and a ΔflaA mutant had taken place. Representative micrographs illustrating this experiment are shown in Fig. 2A, and quantification of the surface area covered by cells before and after addition of AMM is shown in Fig. 2B. When AMM was added after monolayer formation, it removed a subset (approximately 40%) of the attached population from the surface. Interestingly, the monolayer formed by the ΔflaA mutant was resistant to AMM. We hypothesize that the flagellum is required to break the association of MSHA with the surface. When this happens, the mannose binding site on MSHA is exposed. Alpha-methylmannoside binds to this site, and MSHA is unable to mediate further surface attachments. Because the ΔflaA mutant has no flagellum, it is unable to break contact with the surface and the MSHA mannose binding site is never exposed. Thus, once surface attachment is formed, it is resistant to inhibition by AMM.


Figure 2. Effect of α-methyl mannoside (AMM) on surface association by wild-type V. cholerae (WT, MO10) and a ΔflaA mutant (PW412) in MM alone.

A. Phase-contrast micrographs of monolayers before (MM) and after (MM + AMM) treatment with AMM. Bar = 10 µm.

B. Per cent total surface area covered by monolayer cells before and after treatment with AMM as noted in the key. Observations represent the mean of three separate experiments.

In this study, two populations of monolayer-associated wild-type V. cholerae cells were observed. One was sensitive to the action of AMM, whereas the other was resistant. By analogy to ΔflaA mutants, which are resistant to AMM, we hypothesize that AMM-resistant monolayer-associated cells no longer have an active flagellum. Thus, like the ΔflaA mutant cells, these AMM-resistant cells are unable to break their attachment to the surface. This suggests the possibility that the V. cholerae transition from transient to permanent attachment may be associated with loss of flagellum-based motility.

Transcription of flaA is decreased in the V. cholerae monolayer

Because cells in the wild-type V. cholerae monolayer were immobilized, we hypothesized that transcription of genes encoding the polar flagellum would be altered in the monolayer. In order to test this hypothesis, we measured transcription of flaA in the planktonic state and monolayer. Given the relative paucity of cells in the monolayer, measurement of transcription in this state required the use of quantitative PCR techniques. Figure 5A shows transcription of flaA by wild-type V. cholerae, ΔvpsA mutant and ΔmshA mutant cells in the planktonic and monolayer states (–Mannose). Compared to transcription in the planktonic state, we observed that transcription of flaA was repressed by approximately sevenfold or greater in the V. cholerae monolayer (= 0.04).


Figure 5. Normalized levels of mRNA transcripts in planktonic (grey bars) or surface-associated (black bars) wild-type V. cholerae (WT, MO10) and ΔvpsA (PW396), ΔflaA (PW412) and ΔmshA (PW361) mutant cells. Cells were incubated in MM alone (–Mannose) or supplemented with mannose (+Mannose).

A. Quantification of flaA mRNA.

B. Quantification of vpsL mRNA. Observations represent the mean of at least three separate experiments.

The mannose-sensitive haemagglutinin type IV pilus is the only structure we have identified that is required for surface attachment in the V. cholerae monolayer. Interestingly, flaA transcription is repressed in planktonic ΔmshA mutant cells, suggesting that flaA transcription is activated by MSHA. Because MSHA both mediates surface attachment and activates flaA transcription, we hypothesize that a change in the state of MSHA upon binding to a surface could result in repression of flaA transcription in response to a surface, ultimately resulting in the formation of a stable monolayer.

Surface attachment is not sufficient to induce vpsL transcription

Numerous previous studies of gene transcription have demonstrated increased transcription of exopolysaccharide genes in biofilms. Thus, it has been suggested that exopolysaccharide synthesis is induced upon contact with a surface. Because V. cholerae formed only a monolayer upon contacting a surface bathed in MM, it seemed unlikely that VPS was being synthesized. However, this might have been due to the absence of monosaccharide substrates in the environment. Thus, we hypothesized that vps gene transcription might still be induced upon contact with the surface. We measured vpsL transcription in planktonic and monolayer-associated wild-type V. cholerae, ΔvpsA mutant, ΔmshA mutant, and ΔflaA mutant cells (see Fig. 5B, – Mannose). No induction of vpsL transcription was seen upon surface attachment by either wild-type V. cholerae or any of the mutants studied. Whereas this observation does not rule out a requirement for surface attachment in induction of vps gene transcription, it suggests that surface attachment is not sufficient for induction of vps gene transcription. Thus, in V. cholerae, the regulatory pathways governing transcription of flagellar and exopolysaccharide synthesis genes in response to a surface are divergent. Furthermore, the monolayer state appears to be defined by repression of flagellar genes and loss of motility.

Medium supplementation with mannose allows progression from the monolayer stage to the biofilm stage

We hypothesized that progression from the monolayer to the biofilm stage would be induced by addition of monosaccharides to the growth medium. To test this, we formed wild-type and mutant monolayers in MM. After monolayer formation, MM was replaced with fresh MM supplemented with mannose. This two-step approach was required for efficient surface adhesion because mannose, like AMM, interferes with MSHA-dependent surface attachment. The effect of mannose on both growth and biofilm development was measured. Interestingly, addition of mannose did not significantly improve the growth of V. cholerae (data not shown). However, after addition of mannose, wild-type V. cholerae and ΔflaA mutants formed intercellular contacts both on the surface (Fig. 3A) and in the planktonic phase (data not shown). The former observation was confirmed by quantifications demonstrating that the cell cluster size of surface-attached wild-type V. cholerae and the ΔflaA mutant increased by approximately a factor of four upon addition of mannose (Fig. 3B). As expected, surface-attached ΔvpsA mutant cells did not demonstrate a similar increase in cell cluster size upon addition of mannose. A similar result was obtained for the ΔflaAΔvpsL double mutant. Thus, we conclude that intercellular interactions are mediated by the VPS exopolysaccharide and that these interactions do not require a surface.

The ΔmshA mutant showed the most striking surface attachment phenotype. Although the ΔmshA mutant was completely unable to form a monolayer, medium supplementation with mannose induced formation of intercellular contacts in the planktonic phase. In biofilm development by the ΔmshA mutant, we observed that clusters of cells as well as single cells dropped to the surface to join the nascent biofilm. Thus, we conclude that the ΔmshA mutant bypasses the monolayer stage in biofilm development.

To determine whether attachment of the ΔmshA mutant to the surface was mediated by VPS, we studied surface attachment by a ΔmshAΔvpsL double mutant. In fact, a ΔmshAΔvpsL double mutant was completely unable to associate with the surface. This suggests that, although monolayer formation is a prerequisite for efficient biofilm formation, VPS synthesis allows the bacterium to bypass the monolayer stage.

These observations may explain the puzzling observation that a V. cholerae O139 ΔmshA mutant has no attachment defect in LB broth (Watnick et al., 2001). Yeast extract, a component of LB broth, contains large amounts of mannose due to the presence of mannan in the yeast cell wall. Mannose interferes with MSHA-mediated surface attachment but enhances VPS synthesis and, thus, VPS-mediated surface attachment. Thus, the components of LB broth may block MSHA-mediated surface attachment and stimulate VPS synthesis and VPS-mediated attachment. As a result, when V. cholerae O139 is grown in LB broth, MSHA may play a minor role in surface attachment.

The biofilm is resistant to AMM

We previously observed that monolayer attachment was dependent on MSHA and that, for a subset of monolayer-associated cells, attachment was sensitive to AMM. Because VPS is able to mediate attachments to the surface, we predicted that surface attachment in the biofilm would not be solely dependent on MSHA and would be resistant to the action of AMM. In fact, as shown in Fig. 4A and quantified in Fig. 4B, both the wild-type V. cholerae and ΔflaA mutant biofilms were completely resistant to the action of AMM. These experiments demonstrate that the MSHA pilus is required for formation of the monolayer, whereas VPS is required for formation and maintenance of the biofilm.


Figure 4. Effect of α-methyl mannoside (AMM) on surface association by wild-type V. cholerae (WT, MO10) and a ΔflaA mutant (PW412) in MM supplemented with mannose.

A. Phase-contrast micrographs before (MM + M) and after (MM + M + AMM) treatment with AMM. Bar = 10 µm.

B. Percentage of the total surface area covered by biofilm-associated cells before and after treatment with AMM as noted in the key. Observations represent the mean of three separate experiments.

Supplementation with mannose represses transcription of flaA both in planktonic and biofilm-associated cells

We hypothesized that supplementation of the medium with mannose would leave levels of flaA transcription unaltered in planktonic cells and would increase transcription of vps genes in biofilm-associated cells. To test this, we used quantitative PCR to measure levels of flaA transcription in planktonic and biofilm-associated cells grown in medium supplemented with mannose (Fig. 5A, + Mannose). To our surprise, we found that supplementation of the medium with mannose decreased planktonic flaA transcription to the levels found in the biofilm. It is possible that interaction of MSHA with mannose is similar to interaction of MSHA with a surface in its effect on flaA transcription. Although repression of flaA transcription was also observed for ΔmshA mutant biofilm and planktonic cells, our parallel measurements of flaA transcription in MM alone suggest that mutation of mshA causes a repression of flaA transcription that is independent of medium supplementation with mannose.

Although flaA is repressed in planktonic cells grown in medium supplemented with mannose, we still observe single motile cells in the medium. Thus, although it seems probable that repression of flaA is necessary for cessation of motility, the temporal correlation between repression of flaA and the disappearance of flagellum-based motility is not yet clear.

Supplementation with mannose increases transcription of vpsL in planktonic and biofilm-associated cells but not in the monolayer

Induction of exopolysaccharide synthesis gene transcription upon association with a surface is a well-documented phenomenon in bacterial biofilm development (Davies et al., 1993; Prigent-Combaret et al., 1999; Whiteley et al., 2001; Haugo and Watnick, 2002; Sauer et al., 2002). Thus, we hypothesized that vpsL transcription would be greater in biofilm-associated cells than in planktonic cells. Again, we used quantitative PCR to measure vpsL transcription in planktonic and biofilm-associated wild-type and mutant cells cultured in medium supplemented with mannose (Fig. 5B, + Mannose). Interestingly, we found that supplementation with mannose increased vpsL transcription in planktonic V. cholerae by approximately fourfold or greater.

Because supplementation of planktonic cells with mannose decreased flaA transcription and increased vpsL transcription, we suggest that mannose induces a biofilm-like state in the absence of a surface, perhaps by interacting with MSHA. In the biofilm, as expected, an approximately threefold induction of vpsL transcription was observed over that measured for planktonic cells grown in medium supplemented with mannose.

We hypothesized that, in the presence of mannose, vpsL transcription in ΔvpsA mutant monolayer cells would be similar to that of biofilm-associated wild-type V. cholerae cells. To our surprise, we found that vpsL was not induced in ΔvpsA mutant monolayer cells. In fact, vpsL transcription was approximately 10-fold less than that observed for the wild-type V. cholerae biofilm (P = 0.02). Furthermore, levels of flaA and vpsL transcription in the wild-type V. cholerae monolayer formed in medium without mannose were quite similar to those in the ΔvpsA mutant monolayer formed in medium with mannose. Thus, even when formed under these very different growth conditions by bacteria with different genetic backgrounds, the transcriptional footprint of the V. cholerae monolayer was similar. Although the genetic underpinning of this result has not yet been explored, we hypothesize that, in the presence of mannose, the induction of vpsL transcription that we observed upon association of V. cholerae with a surface is not a direct effect of surface association. Rather, basal levels of VPS synthesis are required for induction of vpsL transcription upon contact with a surface. This suggests to us that the environmental signal that effects induction of vps gene transcription is either VPS itself, increased cell density which arises as a function of VPS synthesis, or a property conferred by the architecture of the nascent biofilm structure. Thus, we must conclude that mannose and not the surface is the environmental signal that induces formation of the three-dimensional biofilm structure. Interestingly, VPS synthesis does not appear to be required for induction of vps genes in the planktonic state (Fig. 5B). Thus, these data suggest that the environmental cues that induce vpsL transcription in response to mannose are distinct in the planktonic and biofilm states.

We have previously noted that mutants in all stages of flagellar assembly demonstrate a rugose colony morphology (Watnick et al., 2001), suggesting that VPS synthesis is deregulated. Thus, we hypothesized that vpsL transcription would be elevated in all stages of ΔflaA mutant biofilm development. In fact, we found that vpsL transcription was elevated and equal in planktonic and biofilm-associated ΔflaA mutant cells. Furthermore, we noted that in all of our transcription measurements, elevation of vpsL transcription was accompanied by a repression of flaA transcription. This is consistent with the hypothesis that repression of flagellar synthesis is necessary but not sufficient for vpsL induction. However, these experiments do not address this hypothesis directly.

It is important to note that, whereas ΔflaA mutants are rugose on LB-agar plates, they demonstrate very low vpsL transcription in minimal medium without mannose. Thus, we point out that, whereas a rugose colony morphology suggests abnormal regulation of VPS synthesis, it does not imply constitutive synthesis of VPS. Thus, in a mannose-free environment, surface attachment by certain rugose and non-rugose variants may be indistinguishable.

The monolayer represents a distinct stage in biofilm development

In eukaryotic or prokaryotic developmental processes, consecutive stages are induced by distinct environmental signals that alter the global profile of gene transcription. The resulting change in cellular or multicellular structure heralds passage into the next developmental stage. We have demonstrated that V. cholerae biofilm development includes at least three distinct developmental stages. These are the planktonic stage, the monolayer stage, and the biofilm stage. As illustrated in Fig. 6, environmental signals catalyse transitions between these developmental stages through changes in gene transcription. The planktonic stage, characterized by single cells that exhibit flagellum-mediated motility, results when monosaccharides and a surface are absent from the environment. Planktonic V. cholerae exhibit high levels of flagellar gene transcription and repression of exopolysaccharide synthesis genes. The monolayer stage, characterized by single, non-motile cells attached to a surface, results when a surface is presented to the bacterium in the absence of monosaccharides. Under these conditions, the MSHA type IV pilus mediates attachment to the surface leading to repression of flagellar gene transcription. In this stage, our results suggest that modulation of flagellar gene transcription is the direct result of surface sensing. Progression to the biofilm stage requires only the presence of monosaccharides in the environment. Our experiments suggest that a surface is not absolutely necessary for creation of the biofilm state. In fact, when a surface is absent, planktonic cell clusters are observed. Biofilm-associated V. cholerae exhibit repression of flagellar gene transcription along with induction of exopolysaccharide synthesis gene transcription. Induction of the vps genes enables intercellular interactions that result in the characteristic three-dimensional biofilm structure.


Figure 6. A model for V. cholerae biofilm development in natural environments. Planktonic V. cholerae enter the monolayer stage upon encounter with a surface through the action of MSHA. This is accompanied by the loss of the flagellum. Degradation and transport of oligosaccharides and monosaccharides from the surface provide the signal for progression to the biofilm stage.

Download figure to PowerPoint

Whereas previous experiments have defined the characteristics of the planktonic stage and the biofilm stage of V. cholerae biofilm development, our experiments define the specific genetic requirements and transcription profile of the monolayer stage. These studies support a model of biofilm development in which the monolayer is a distinct developmental stage with specific genetic requirements and a transcription profile that differ both from those of the planktonic and biofilm stages.

A model for V. cholerae biofilm development in natural environments

Based on the data gathered in this work, we propose the following model for environmental biofilm development. In natural aquatic environments, high concentrations of monosaccharides in the liquid phase are not common. Thus, planktonic V. cholerae must establish stable associations with surfaces by entering the monolayer state. Many aquatic surfaces are composed of polysaccharides. These include the exoskeletons of aquatic insects and zooplankton, the cell walls of plants and fungi, and the carbohydrate capsules of aquatic bacteria. Such a surface may be degraded by the V. cholerae monolayer, resulting in release of oligosaccharides and monosaccharides into the local environment. These surface-derived monosaccharides may provide the signal to progress to the biofilm stage.

Because V. cholerae are stably attached to a surface in the absence of VPS synthesis, one must question why the bacterium expends the energy to create a three-dimensional structure. One possibility is that biofilm development maximizes the number of cells fed by a surface. Another is that restricted diffusion within the biofilm structure allows biofilm-associated cells to retain monosaccharides released from surfaces. A third rationale is that formation of a multicellular structure in which all bacteria are surrounded by an exopolysaccharide matrix allows local storage of VPS, a carbohydrate polymer that may be degraded and consumed when nutrients become scarce. We hypothesize that when ample substrates are available for synthesis of VPS, all of these factors make biofilm development a favourable process in natural environments.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Bacterial strains, plasmids and media

The bacterial strains and plasmids used in this study are listed in Table 1. All experiments were done in minimal medium (MM) alone or supplemented with mannose. The composition of the medium is given in Table 2. No carbon source other than the amino acids is required for growth of V. cholerae in this medium. When used, mannose and α-methyl mannoside were added to a final concentration of 0.2%. Phosphate buffered-saline (pH 7.0, 0.1 M) was used to rinse the cells when required.

Table 1. . The bacterial strains and plasmids used in this study.
Strain/plasmidGenotypeSource or reference
E. coli
SM10λpir thi thr leu tonA lacY supE recA::RP4-2-Tc::MuλpirR6K;Kmr Miller and Mekalanos (1988)
PW326SM10λpir (pAJH5) Haugo and Watnick (2002)
PW402SM10λpir (pSM1)This study
V. cholerae
PW328MO10 ΔvpsL, Smr Haugo and Watnick (2002)
PW357MO10lacZ::vpsLp[RIGHTWARDS ARROW]lacZ, Smr Haugo and Watnick (2002)
PW361MO10 ΔmshA, Smr 
PW396MO10 ΔvpsA, Smr Kierek and Watnick (2003)
PW412MO10 ΔflaA, SmrThis study
PW447MO10 ΔflaAΔmshA, SmrThis study
PW448MO10 ΔvpsLΔmshA, SmrThis study
PW450MO10 ΔflaAΔvpsL, SmrThis study
pWM91 oriR6KmobRP4 lacI pTac tnp miniTn10Km; Kmr, Apr Metcalf et al. (1996)
pAJH5pWM91 carrying a fragment of vps operon harbouring an internal, unmarked deletion Haugo and Watnick (2002)
pSM1pWM91 carrying a fragment of flaA harbouring an internal, unmarked deletionThis study

Construction of the deletion mutants

The ΔflaA mutant was constructed in the following manner. A 494 bp fragment located just upstream of the start codon of flaA and a 449 bp fragment downstream of the stop codon were amplified by PCR using primer pairs P108, P109 and P110, P111 respectively (see Table 3 for sequences). Internal primers included a complementary 15 bp sequence at their 3′ and 5′ ends, respectively, as previously described (Haugo and Watnick, 2002). These two fragments were joined using the gene splicing by overlap extension technique (Horton et al., 1990; Lefebvre et al., 1995). This resulted in the construction of a fragment with a 1077 bp deletion in flaA. The fragment containing the deletion was purified and ligated into pCR2.1TOPO. This fragment was then removed from pCR2.1TOPO by digestion with SpeI and XhoI and ligated into pWM91 to create the suicide plasmid pSM1. This plasmid was used to create flaA deletions in the relevant strains by double homologous recombination and sucrose selection as previously described (Haugo and Watnick, 2002).

Table 3. . The sequence of the primers used.
PrimerDescriptionSequenceProduct size
Construction of V. cholerae flaA deletion
P108forward primer for downstream fragment5′-TTTGCCCTAAACCCTCTGAA-3′ 
P109reverse primer for downstream fragment5′-TTACGAGCGGCCGCAGCCAAACTCTGCAATCTCGT-3′449
P110forward primer for upstream fragment5′-TGCGGCCGCTCGTAAGCCGACACGTTGGTATTTACG-3′ 
P111reverse primer for upstream fragment5′-ATCGCAGAAAGCCTTTGGTA-3′494
cDNA synthesis and real time RT-PCR
P192forward primer for flaA5′-TCCTTTATTGCCGAACAACC-3′ 
P193reverse primer for flaA5′-ATCCGTTTGACCGTTGATGT-3′171
P194forward primer for vpsL5′-ATCGCACCATAGTGAATCGCT-3′ 
P195reverse primer for vpsL5′-TCTGTGCCCATCCAGTAATGC-3′ 69
P198forward primer for rDNA5′-GAGCGGCAGCACAGAGGA-3′ 
P199reverse primer for rDNA5′-TTTCCCAGGCATTACTCACCC-3′ 67

The ΔflaAΔvpsL and ΔflaAΔmshA double mutants were constructed in a similar manner by introducing the suicide plasmid pSM1 into strains PW328 (ΔvpsL) and PW361 (ΔmshA) respectively. The ΔmshAΔvpsL double mutant was constructed by introducing the suicide plasmid pAJH5, carrying the vpsL deletion fragment, into PW361 (ΔmshA) strain.

Phase-contrast microscopy and quantitative analysis of monolayer and biofilm structures

For phase-contrast microscopy, cells were grown in sterile 24-well polystyrene microtitre dishes. For monolayer formation, wells were filled with 300 µl of MM and inoculated with the relevant strain. The dishes were incubated at 27°C with gentle shaking for 24 h. For biofilm formation, wells were first filled with 300 µl of MM, inoculated with the relevant strain and incubated at 27°C. After 24 h, the planktonic cells were rinsed off with sterile PBS and then 300 µl of MM supplemented with mannose was added. The dishes were incubated for an additional 24 h to allow biofilms to develop. After the incubation, the planktonic cells were rinsed off with PBS, and biofilm or monolayer development at the bottom surface of the well was visualized with an Eclipse TE-200 microscope (Nikon) equipped with an Orca digital CCD camera (Hamamatsu). For treatment with α-methyl mannoside (AMM), biofilms and monolayers were formed as described above and visualized by phase-contrast microscopy. They were then rinsed and AMM was added to the PBS in the wells. The cells were incubated for three additional hours at 27°C, rinsed, and visualized again. A computer equipped with Metamorph Imaging software (Universal Imaging) was used for image acquisition, processing and quantification. This software quantifies the number of cells attached to the well surface, the surface area covered by cells, and the size of cell clusters.

RNA extraction

All the strains were grown in MM or MM supplemented with mannose in sterile 90 mm Petri dishes. Briefly, for preparation of planktonic cell RNA, 10 ml aliquots of the cultures were pelleted by centrifugation. Two millilitres of TRIzol were added to each pellet, and cells were lysed by repetitive pipetting. After removal of planktonic cells, the remaining surface-associated cells were rinsed with PBS three times and then lysed in situ with TRIzol. The lysates of both planktonic and surface-associated cells were incubated at room temperature for 10 min. Chloroform was then added to the samples at a ratio of 0.2 mls for every ml of TRIzol. Samples were shaken vigorously for 15 min and then incubated at room temperature for 3 min. The samples were centrifuged at 4500 g at 4°C for 10 min to separate the aqueous layer, which was then transferred to a microfuge tube. The RNA, along with 10 µg of carrier t-RNA, was incubated with isopropyl alcohol in a ratio of 0.5 ml for every ml of TRIzol to allow precipitation of RNA. Precipitated RNA was pelleted by centrifugation at 12000 g for 30 min at 4°C, washed with 75% ethanol, dried, and then dissolved in 100 µl of RNase-free water. To remove contaminating DNA, the total RNA was incubated with RNase-free DNAse I (Ambion) for 30 min at 37°C. Finally, the RNeasy Mini Kit (Qiagen) was used to clean up the RNA after DNase I digestion. The RNA samples were checked for presence of residual genomic DNA by standard PCR. A measurement of absorbance at 260 nm was used to quantify RNA concentration, and RNA was subsequently stored at − 80°C.

Primers for cDNA synthesis and real time RT-PCR

vpsL, flaA and rRNA primers were designed to amplify 100–150 bp fragments. The sequence of the primers and the size of products obtained using these primers are given in Table 3. Specificity of the primer pairs was confirmed by electrophoretic analysis of individual PCR reactions.

cDNA synthesis

Reverse transcription (RT) of RNA was performed with Superscript II RNase H-Reverse Transcriptase (Invitrogen) and nuclease-free water (Ambion). Two-hundred ng of total RNA was used for RT in a final volume of 30 µl. A pool of gene-specific reverse primers, each at a concentration of 5 µM, and a 10 mM dNTP mixture was added to the RNA before denaturation at 65°C for 5 min. To this mix was then added the first strand buffer [75 mM KCl, 50 mM Tris-HCl (pH 8.3), 3 mM MgCl2], 10 mM DTT, 60 units of RNase inhibitor and 200 units of Superscript II RT. Reactions without RT were performed to verify the absence of contaminating genomic DNA in each RNA sample. The reactions were incubated at 42°C for 50 min followed by 70°C for 15 min. The resulting cDNA was stored at − 20°C.

Real time RT-PCR

Quantitative RT-PCR was performed in 20 µl reactions using 1 µl of each cDNA sample, the SYBR Green RT-PCR Kit (Qiagen) and the specific primer pair. The PCR mixture was held at 50°C for 2 min and denatured at 95°C for 10 min. Forty amplification cycles were carried out at 94°C for 30 s followed by 54°C for 1 min and 72°C for 30 s. Dissociation curves were performed for each primer pair under the following conditions: 95°C for 15 s, 54°C for 20 s, and 95°C for 15 s. Template-free or reverse transcriptase-free reactions were included as controls. All PCR reactions were performed with the ABI Prism 7700 sequence detection system (PE Applied Biosystems). The data was analysed on a power Macintosh G4 (Apple Computer, Santa Clara, CA) linked directly to the ABI Prism 7700 sequence detection system by using the SDS (version 1.9) application software (PE Applied Biosystems) as described by the manufacturer. rRNA was used as an internal control for all the PCR reactions. Transcript amounts for all the strains and conditions were normalized against this gene. The PCR reaction for each independent experiment was repeated at least three times to check for consistency of the PCR reactions.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

We would like to thank Dr Anne Kane of the GRASP Center and her staff for their expert preparation of many reagents, which greatly accelerated the course of these experiments. This work was supported by an award from the Hood Foundation, NIH R01 AI50032, and the New England Medical Center GRASP Center NIH/NIDDK, P30 DK34928.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References