Magnesium Promotes Flagellation of Vibrio fischeri

Abstract

The bacterium Vibrio fischeri requires bacterial motility to initiate colonization of the Hawaiian squid Euprymna scolopes. Once colonized, however, the bacterial population becomes largely unflagellated. To understand environmental influences on V. fischeri motility, we investigated migration of this organism in tryptone-based soft agar media supplemented with different salts. We found that optimal migration required divalent cations and, in particular, Mg²⁺. At concentrations naturally present in seawater, Mg²⁺ improved migration without altering the growth rate of the cells. Transmission electron microscopy and Western blot experiments suggested that Mg²⁺ addition enhanced flagellation, at least in part through an effect on the steady-state levels of flagellin protein.

The symbiosis between the marine bacterium Vibrio fischeri and the Hawaiian squid Euprymna scolopes provides a model for exploring the communication that occurs between a bacterium and its host in a natural setting (30, 39, 45, 55). Juveniles of E. scolopes hatch without V. fischeri cells present inside the symbiotic organ (the light organ), and rapidly acquire these bacteria from the surrounding seawater (31, 57). Colonization begins with aggregation of V. fischeri cells in mucus on the surface of the light organ (37, 38, 40), followed by movement of these bacteria through pores into ducts, apparently toxic passageways that limit nonspecific invaders (11, 40), and ultimately into crypts where the bacteria multiply (46).

Only motile cells of V. fischeri initiate symbiotic colonization of E. scolopes. Nonmotile mutants fail to colonize (16, 33, 59), presumably because they fail to migrate out of the bacterial aggregates formed on the surface of the light organ (40). Apparently, normal initiation also requires optimal motility, because several hypermotile mutants colonize with delayed kinetics (32).

Once V. fischeri cells initiate colonization, the majority of symbiotic bacteria within the E. scolopes light organ become nonflagellated (33, 46). However, within an hour of their release from the light organ into seawater, V. fischeri cells regrow their flagella (46). These observations suggest that environmental conditions inside the light organ inhibit flagellation, while those outside favor it (46).

Environmental influences on the motility of the enteric bacteria Escherichia coli and Salmonella enterica serovar Typhimurium have been well documented (for reviews, see references 3 and 29). These influences include nutrient availability, temperature, ionic composition, pH, and surface interactions (2, 21, 23, 27, 35, 48, 50). Most known environmental influences act at the level of transcription initiation or, to a lesser extent, message stability. These operate through at least one nucleoid protein (H-NS) and a host of transcription factors, including the cyclic AMP (cAMP) receptor protein, LrhA (CRP), and several two-component response regulators (1, 7, 14, 15, 21, 22, 26, 41, 49-51, 54). Control of message stability involves the small RNA-binding protein CsrA (56).

Environmental conditions also can affect flagellation posttranscriptionally. For example, flagellation of Rhizobium spp. depends on divalent cations to maintain cross-links between flagellin subunits. In the absence of those cations, the flagellin subunits dissociate, resulting in nonmotile cells (44).

During studies of V. fischeri chemotaxis (13), we noticed differences in migration through soft agar depending on the medium used. In this study, we determined that optimal migration of V. fischeri required divalent cations. In particular, we found that Mg²⁺ influenced motility by promoting flagellation of this organism.

MATERIALS AND METHODS

Strains and media.

Strains used in this study are described in Table 1. Cells were grown in TBS (1% tryptone, 342 mM NaCl) (13) or TB-SW (1% tryptone, 210 mM NaCl, 35 mM MgSO₄, 7 mM CaCl₂, and 7 mM KCl) (13). TBS was supplemented with MgSO₄, MgCl₂, CaCl₂, KCl, BaCl₂, or SrOH₂ at a variety of concentrations. Agar was added to a final concentration of 0.25% for soft agar and 1.5% for solid agar.

TABLE 1.

Strains used in this study

Species	Strain	Description	Source or reference
V. anguillarum	PKJ	Wild-type isolate	E. G. Ruby
V. fischeri	ES114	Isolate from E. scolopes	8
V. fischeri	ES235	Isolate from E. scolopes	9
V. fischeri	EM17	Isolate from E. morsei	8
V. fischeri	H905	Seawater isolate	24
V. fischeri	MJ1	Isolate from monocentrid fish	47
V. fischeri	MJ11	Isolate from monocentrid fish	25
V. harveyi	B392	Wild-type isolate	36, 43
V. orientalis	ATCC 33934	Seawater isolate	60
V. parahaemolyticus	KNH1	Isolate from Kaneohe Bay, Hawaii	40
V. splendidus	ATCC 33869	Wild-type isolate	43

Soft agar and motility assays.

Bacteria were grown at 28°C to mid-exponential phase (A₆₀₀ of ca. 0.3 to 0.7) in TBS or TB-SW. For soft agar motility studies, 10-ml aliquots of approximately equal numbers of bacterial cells were inoculated on the surface of soft agar motility plates and incubated at 28°C for 4 to 7 h. Soft agar motility plates were handled as described previously (58). Images were taken with a Kodak DX3600 Zoom digital camera (Eastman Kodak Company, Rochester, N.Y.).

Visualization of flagella by electron microscopy.

Wild-type V. fischeri cells (ES114) were grown to exponential phase in either TBS or TBS containing 35 mM MgSO₄. Samples were negatively stained with 2% phosphotungstic acid (pH 7.0) and observed with a JEM-1200EXII electron microscope (JEOL, Tokyo, Japan). Micrographs were taken at an accelerating voltage of 80 kV.

Western blot analysis.

ES114 cells were grown in the indicated medium, concentrated by microcentrifugation (2 min), resuspended in 10 mM Tris (pH 7.5), and then lysed by sonication. Equal amounts of proteins (10 μg), as determined by Lowry assay (28), were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12.5% acrylamide) and transferred to a nylon membrane. After blocking with dry milk in TBS-T (10 mM Tris, 150 mM sodium chloride, and 0.05% Tween 20), the membrane was treated with rabbit anti-Vibrio parahaemolyticus flagellin antibody (29a), followed by anti-rabbit immunoglobulin G secondary antibody conjugated to horseradish peroxidase (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pa.). To visualize the flagellin proteins, the membrane was incubated with Western Lightning Western blot chemiluminescence reagents (Perkin-Elmer, Torrance, Calif.) and exposed to film.

Measurements of motile cells.

Overnight cultures of V. fischeri strain ES114 were inoculated into two tubes of TBS broth (5 ml) or, as a control, TBS containing 35 mM MgSO₄ fresh medium. After incubation (3 h) at 28°C (A₆₀₀ of ca. 0.3), 180 μl of 1 M MgSO₄ was added to one of the TBS cultures. At 10-min intervals, 75-μl samples were removed. Cell motility was observed with an Olympus BX41 microscope with dark-field background and video recorded through an Olympus charge-coupled device color camera for 5 s. The percentage of motile cells was calculated by comparing the position of cells after 5 s relative to their position after 1 s.

RESULTS

Effect of medium and pregrowth conditions on migration.

During the course of experiments designed to investigate chemotaxis of V. fischeri cells in a tryptone-based soft agar medium (13), we observed enhanced migration in the presence of salts naturally present in seawater. For example, the cells migrated farther and formed rings that appeared denser when inoculated onto TB-SW (a tryptone-based medium that contains 210 mM NaCl, 35 mM MgSO₄, 7 mM CaCl₂, and 7 mM KCl) than when inoculated onto TBS (a tryptone-based medium with 342 mM NaCl) (compare Fig. 1A with B and 1C with D). Furthermore, the growth condition of the cells prior to inoculation onto soft agar (pregrowth) affected their subsequent migration. When inoculated onto TBS soft agar, cells pregrown in TBS broth migrated poorly relative to those pregrown in TB-SW broth (Fig. 1B and D). Although less dramatic, the same was true when cells were inoculated onto TB-SW soft agar (Fig. 1A and C).

FIG. 1.

Migration of V. fischeri on tryptone-based soft agar with various salt compositions. (A to D) Exponential-phase cells of V. fischeri, pregrown in TB-SW or TBS as indicted, were inoculated near the center of soft agar plates containing either TB-SW (A and C) or TBS (B and D). (E to H) Exponential-phase cells of V. fischeri grown in TBS were inoculated near the center of soft agar plates containing TBS containing 35 mM MgSO₄ (E and H), TBS containing 7 mM CaCl₂ (F), or TBS containing 7 mM KCl (G) and incubated at 28°C for 5 h. At 4 h, 5 μl of 2 M serine was spotted directly onto a TBS plate containing 35 mM MgSO₄ outside the migrating bands (arrow, H).

To further investigate the influence of pregrowth conditions, we monitored the rate of migration. We harvested mid-exponential-phase cells grown in either medium, gently washed them with TBS to reduce salt carryover while minimizing flagellum breakage, and then inoculated soft agar medium. Cells pregrown in TB-SW, and inoculated onto the same medium, migrated rapidly (Fig. 2). After a 1-h lag, those pregrown in TBS, and inoculated onto TB-SW, migrated at about the same rate. Thus, the difference in ring diameter seen in Fig. 1A and C resulted from a lag, not a differential rate of migration. Similarly, pregrowth in TB-SW allowed cells to begin migrating immediately on TBS, albeit slowly (Fig. 2). This accounts for the difference in ring diameter seen in Fig. 1B and D. On the basis of these experiments, we hypothesized that the observed lags in and rates of migration resulted from the differential salt compositions of the pregrowth medium and the inoculated soft agar medium.

FIG. 2.

Effect of pregrowth conditions on migration of V. fischeri. Cells of V. fischeri strain ES114 were pregrown to mid-exponential phase in TBS (triangles) or TB-SW (squares), inoculated onto TBS (open symbols) or TB-SW (solid symbols) soft agar plates, and incubated at 28°C. Migration of the cells was determined hourly by measuring the diameter of the outer migrating rings. Error bars represent the standard deviations of a representative experiment performed in triplicate.

Effect of Mg²⁺ and Ca²⁺ on migration of V. fischeri in soft agar.

To identify the salt component that caused the observed differences in lag and migration, we examined the behavior of V. fischeri on TBS medium supplemented with MgSO₄, CaCl₂, KCl, or additional NaCl. Addition of either MgSO₄ (35 mM) or CaCl₂ (7 mM) significantly enhanced migration (Fig. 1E and F and 3A), while KCl (7 mM) exerted, at most, a small effect (Fig. 1G and 3A). In contrast, addition of NaCl to TBS (to final concentrations between 220 and 380 mM) failed to alter migration (data not shown), suggesting that the observed effects did not result from a change in osmolarity. MgSO₄ also affected the appearance of the chemotactic rings. In particular, a gap occurred between the inner and outer rings of cells grown on TBS-Mg²⁺ (Fig. 1E) but not of cells grown on TB-SW (Fig. 1C). The reason for these differences remains unknown; however, like the rings formed by cells grown on TB-SW (13), the inner ring formed on TBS-Mg²⁺ responds to serine (arrow in Fig. 1H) and the outer ring responds to thymidine (data not shown). This observation argues that the salt composition does not substantially alter chemotaxis.

FIG. 3.

Mean migration of V. fischeri on tryptone-based soft agar plates with various salt compositions. (A) Exponential-phase cells of V. fischeri, grown in TBS, were inoculated near the center of TBS soft agar plates or TBS plates containing 35 mM MgSO₄, 7 mM CaCl₂, or 7 mM KCl. (B) Exponential-phase cells of V. fischeri, grown in TBS, were inoculated near the center of TBS soft agar plates containing the specified concentration of MgSO₄. (C) Exponential-phase cells of V. fischeri, grown in TBS, were inoculated near the center of TBS soft agar plates containing the specified concentration of CaCl₂. All plates were incubated at 28°C for 5 h, after which time the diameters of the outer migrating bands were measured. Error bars represent the standard deviations of a representative experiment performed in triplicate.

We then asked whether differences in growth could account for the observed differences in migration. The cells reached a higher optical density in TB-SW than in TBS. We attributed this growth difference to the presence of KCl, because its addition to TBS resulted in growth similar to that of TB-SW-grown cells (data not shown). Thus, the slight effect that KCl exerts on migration cannot be distinguished from growth, and we have not pursued its further investigation. In contrast, the addition of MgSO₄ or CaCl₂ exerted no effect upon growth (data not shown). Therefore, these salts must play specific roles in enhancing migration of V. fischeri cells.

Effect of divalent cations on migration of V. fischeri.

To further explore the importance of salts in promoting migration of V. fischeri, we first determined the range of MgSO₄ concentrations that promote migration. Cells migrated farther with the addition of as little as 0.2 mM and as much as 200 mM MgSO₄; optimal migration occurred between 20 and 40 mM (Fig. 3B and data not shown). Because the addition of MgCl₂ (35 mM) enhanced migration to a similar extent as did MgSO₄ (35 mM) (data not shown), we concluded that the Mg²⁺ cation promotes migration of V. fischeri.

We next varied the CaCl₂ concentration. Addition of low concentrations of CaCl₂ (2 to 20 mM) to TBS enhanced migration (Fig. 3C). In contrast, higher concentrations of CaCl₂ (40 to 66 mM) inhibited migration of the cells (Fig. 3C and data not shown). However, this decreased migration likely stems from an effect on growth, as these amounts also decreased the growth rate and peak optical density in liquid culture (data not shown).

Because both Mg²⁺ and Ca²⁺ could enhance migration, albeit not to the same extent, we asked whether migration of V. fischeri depended upon divalent cations, in general. We found that low concentrations of BaCl₂ (2 to 8 mM) enhanced migration to nearly the same degree as did 8 mM MgSO₄ (data not shown), while higher concentrations inhibited migration. Similarly, low concentrations of SrOH₂ (2 to 5 mM) also enhanced migration (data not shown). We could not test the effects of higher concentrations, as this salt was insoluble at higher concentrations. These results support a general divalent cation effect. They also suggest that optimal V. fischeri motility occurs under the conditions of high concentrations of Mg²⁺ that naturally exist in seawater (about 50 mM) (12).

Role of Mg²⁺ in migration of other bacterial species.

Because Mg²⁺-dependent migration has not been reported previously, we asked whether the ability to migrate farther in the presence of Mg²⁺ represents a common trait of marine Vibrio spp. We therefore examined migration of other marine isolates in TBS and TBS containing 35 mM Mg²⁺ (Fig. 4; Table 1; also data not shown). Mg²⁺ significantly enhanced migration of other V. fischeri isolates, including strains isolated from the fish Monocentris japonica (MJ1 and MJ11), the squids E. scolopes (ES235) and Euprymna morsei (EM17), and seawater (H905). Mg²⁺ also significantly enhanced migration of Vibrio orientalis (ATCC 33934) (data not shown). Migration of Vibrio splendidus (ATCC 33869), V. parahaemolyticus (KNH1), and Vibrio anguillarum (PKJ) was also enhanced; however, the effect of this salt on migration was relatively minor (Fig. 4 and data not shown). Finally, Mg²⁺ did not significantly influence migration of Vibrio harveyi B392 (Fig. 4). Furthermore, addition of low concentrations of MgSO₄ (2.5 to 10 mM) to TBS did not affect the migration of E. coli (data not shown), while high concentrations (20 to 80 mM) decreased migration, as reported previously (27).

FIG. 4.

Migration of various Vibrio strains on TBS soft agar supplemented with Mg²⁺. Exponential-phase cells of V. fischeri MJ1, V. splendidus ATCC 33869, V. parahaemolyticus KNH1, and V. harveyi B392, grown in TBS, were inoculated near the center of TBS soft agar plates or TBS supplemented with 35 mM MgSO₄. Pictures were taken after 7 h of incubation at 28°C.

Effect of Mg²⁺ on flagellation of V. fischeri.

We hypothesized that Mg²⁺ and other divalent cations could affect migration of V. fischeri by altering the kinetics or direction of flagellar rotation, altering the length of the filaments, increasing the number of flagella per cell, and/or increasing the proportion of flagellated bacteria. To distinguish among these possibilities, we first examined cells by phase-contrast microscopy. Those cells grown in TBS-Mg²⁺ were highly motile, while those grown in TBS exhibited poor to no motility (data not shown). We then determined by Western analysis that the Mg²⁺-grown cells contained higher amounts of flagellin protein than did those grown without Mg²⁺ (Fig. 5A). We next sheared and collected surface-associated flagellin and found that Mg²⁺-exposed bacteria possessed larger amounts (data not shown). Finally, we analyzed V. fischeri cells by using transmission electron microscopy (TEM). The majority of cells grown in TBS were nonflagellated (Fig. 5B and F). Furthermore, we observed no obvious basal-body-like structures at the poles of nonflagellated cells (Fig. 5C). In contrast, the majority of cells grown in TBS-Mg²⁺ expressed between one and three flagella (Fig. 5D to F and data not shown). Thus, Mg²⁺ enhances migration of V. fischeri by increasing both the number of flagella per cell and the proportion of cells that have flagella.

FIG. 5.

Analysis of flagella from V. fischeri ES114 grown in the presence and absence of Mg²⁺. (A) Ten micrograms of protein extracted from cells grown to mid-exponential phase in TBS (lane 1) or TBS supplemented with 35 mM MgSO₄ (lane 2) was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to a membrane, and probed with antiflagellin antibody. The observed bands were located between the 36.2- and 49.9-kDa size standards, similar to that observed previously (34). (B and C) TEM of ES114 cells grown in TBS. (D and E) TEM of ES114 grown in TBS containing 35 mM MgSO₄. Scale bars, 200 μm (B and D) and 200 nm (C and E). (F) Histogram indicating the number of ES114 cells containing the indicated number of flagella per cell. ES114 was grown in TBS (shaded bars) or TBS containing 35 mM MgSO₄ (black bars). Arrows point to basal body structures.

Effect of Mg²⁺ on the kinetics of flagellin biosynthesis and motility.

To understand the migration kinetics shown in Fig. 2, we determined the rate of disappearance or appearance of flagellin protein following a Mg²⁺ shift-down (Fig. 6A) or shift-up (Fig. 6B). We observed a steady decline in cell-associated flagellin following the shift-down. However, even after 2 h of growth in TBS, some cell-associated flagellin protein remained. This result explains why cells pregrown in media containing Mg²⁺ consistently exhibited more rapid migration over the first few hours of incubation than did those pregrown without added Mg²⁺ (Fig. 2).

FIG. 6.

Response of V. fischeri flagellation and motility to changes in MgSO₄. (A) ES114 cells were grown in TBS containing 35 mM MgSO₄ to mid-exponential phase, collected, and resuspended in TBS lacking MgSO₄. Samples were collected over time and analyzed by a Western blot procedure as described for Fig. 5. Lanes 1 to 6 contain samples collected at the indicated times (min): 1, t = 0; 2, t = 15; 3, t = 30; 4, t = 45; 5, t = 60; 6, t = 120. Lanes 7 and 8 contain samples prepared from ES114 cells treated similarly to those in lanes 1 to 6 but grown for the duration of the experiment (overnight and t = 120 min) in either TBS (lane 7) or TBS containing 35 mM MgSO₄ (lane 8). (B) ES114 cells were grown in TBS to mid-exponential phase and then subcultured in TBS containing 35 mM MgSO₄. Samples were collected over time and analyzed by a Western blot procedure as described for Fig. 5. Lanes 1 to 7 contain samples collected at the indicated times (min): 1, t = 0; 2, t = 5; 3, t = 15; 4, t = 30; 5, t = 45; 6, t = 60; 7, t = 120. Lanes 8 and 9 contain samples prepared from ES114 cells treated similarly to those in lanes 1 to 7 but grown for the duration of the experiment (overnight and t = 120 min) in either TBS (lane 8) or TBS containing 35 mM MgSO₄ (lane 9). (C) ES114 cells grown overnight in TBS (white and black bars) or TBS containing 35 mM MgSO₄ (striped bars) were subcultured into the same medium to an A₆₀₀ of 0.3. Then, 35 mM MgSO₄ was added to one of the TBS cultures (white bars). Samples were examined for motility as described in Materials and Methods.

Shift-up experiments (Fig. 6B) showed a progressive increase in V. fischeri flagellin, with a large increase by 40 min and a maximum level reached between 1 and 2 h. This amount of time is similar to that required for E. coli and S. enterica serovar Typhimurium to become fully flagellated and motile (about 40 min) (4, 19). To determine whether this timing was similar for V. fischeri, we examined motility of the cells during this time frame. Within 40 min of MgSO₄ addition to TBS, we observed about a threefold increase in occurrence of motile cells (Fig. 6C). However, the percentage of motile cells in cultures exposed for 60 min to Mg²⁺ remained substantially lower than that for cells grown under conditions that included Mg²⁺ for the duration of the experiment. This suggests that additional time is required for the population to achieve full flagellation and motility following Mg²⁺ exposure under these conditions.

DISCUSSION

In this study, we demonstrated that divalent cations enhance the migration of V. fischeri cells. Mg²⁺ exerted the largest effect across a wide range of concentrations, including that present in seawater. Although Mg²⁺ enhanced the migration of some other marine Vibrio isolates, it exerted the most substantial influence on V. fischeri. This effect occurred at the level of flagellation and represents a novel mechanism for the control of motility.

In attempting to understand the effect of Mg²⁺ on motility of V. fischeri, we found that monitoring migration over a time course rather than examining a single arbitrarily chosen time point provided insight into the problem. In particular, we observed a lag in migration when cells were pregrown without Mg²⁺ relative to those grown in its presence. However, the lack of Mg²⁺ during pregrowth did not cause any defects in the rate of migration on Mg²⁺-containing soft agar—a phenotype not evident from examining a single time point. Migration is a complex behavior. It involves transport, metabolism, growth, and motility (6). Our results make it clear that it also depends upon the status of the cells in the inoculum.

How does Mg²⁺ influence flagellation of V. fischeri? In E. coli, S. enterica serovar Typhimurium, and other enteric bacteria, environmental factors influence flagellation predominantly by controlling transcription of the operon that encodes the master activator FlhDC (10). This may also be true for V. fischeri. Although relatively little is known about the regulatory genes that exert control over flagellar transcription in V. fischeri, it is apparent that its flagellar hierarchy is distinct from that of E. coli and likely similar to that of the closely related pathogen Vibrio cholerae (20, 33). Flagellation of both Vibrio species depends on σ⁵⁴ (20, 59) and FlrA, a σ⁵⁴-dependent master regulator of flagellar gene transcription (20, 33, 42). In V. cholerae, and likely in V. fischeri, distinct subsets of flagellar genes also are controlled by the two-component regulatory system FlrBC and by the flagellum-specific sigma factor σ²⁸ (20, 33). Although the flagellar hierarchy of V. fischeri is not fully understood, our preliminary experiments suggest that Mg²⁺ does not operate at the level of transcription (T. M. O'Shea and K. L. Visick, unpublished data). Rather, it appears to act at a posttranscriptional level.

Posttranscriptional control of flagellation has been documented (5, 18, 44, 53, 56). Notably, Mg²⁺ exerts an effect on the stability of flagella in Sinorhizobium meliloti. In this soil organism, divalent cations, present at relatively low concentrations (200 μM), provide cross-bridges between the flagellin subunits that form the flagellar filament (44). In contrast, 200 μM Mg²⁺ only slightly enhanced migration of V. fischeri cells, while substantial enhancement required about 1 to 2 mM (Fig. 3 and data not shown). Unlike those of S. meliloti, V. fischeri flagella are sheathed, encased within an extension of the outer membrane (32) (Fig. 5C). Furthermore, when grown without added Mg²⁺, the resulting nonflagellated V. fischeri cells do not possess basal body structures (Fig. 5E). One would predict the presence of such structures, if the lack of Mg²⁺ simply caused the disassembly of the flagellar filament. Together, these observations suggest that Mg²⁺ affects a different level, such as (i) the translation of all or a subset of flagellar transcripts, (ii) the stability of the resultant flagellar subunits, and/or (iii) the assembly of the basal body.

Mg²⁺ is a signal that the pathogen S. enterica serovar Typhimurium senses and uses to determine its location within a host (17). Does Mg²⁺ also serve as an environmental signal for V. fischeri? This organism primarily exists in seawater, where Mg²⁺ is abundant, as well as in the intestinal tracts of fishes and in the light organ of E. scolopes. The Mg²⁺ levels inside these host environments are unknown. However, recent evidence suggests that E. scolopes can control at least one environmental condition (osmolarity) within its symbiotic light organ (52).

Motility is absolutely required for V. fischeri cells to enter the light organ of E. scolopes (16, 32-34, 59). Following colonization, a switch occurs, resulting in a significant population of nonflagellated cells within the internal crypts of the light organ (33, 46). Furthermore, within about 45 to 60 min after their release from the light organ into seawater, V. fischeri cells regrow their flagella (46). This is similar to the time interval necessary for cells pregrown in TBS to produce flagellin and, at least for some, to become motile following exposure to Mg²⁺.

Whether, like S. enterica serovar Typhimurium (17), V. fischeri experiences a reduction in Mg²⁺ concentration in its natural environment remains an open question. Regardless, data presented in this paper support a novel mechanism for the control of flagellar biogenesis in V. fischeri. Determining the mechanism by which Mg²⁺ influences flagellar biogenesis in V. fischeri likely will provide insight into the control of flagellar assembly by bacteria and, potentially, an increased understanding of the environmental control of bacterium-host associations.