Exogenous Polyunsaturated Fatty Acids Impact Membrane Remodeling and Affect Virulence Phenotypes among Pathogenic Vibrio Species

ABSTRACT

The pathogenic Vibrio species (V. cholerae, V. parahaemolyticus, and V. vulnificus) represent a constant threat to human health, causing foodborne and skin wound infections as a result of ingestion of or exposure to contaminated water and seafood. Recent studies have highlighted Vibrio's ability to acquire fatty acids from environmental sources and assimilate them into cell membranes. The possession and conservation of such machinery provokes consideration of fatty acids as important factors in the pathogenic lifestyle of Vibrio species. The findings here link exogenous fatty acid exposure to changes in bacterial membrane phospholipid structure, permeability, phenotypes associated with virulence, and consequent stress responses that may impact survival and persistence of pathogenic Vibrio species. Polyunsaturated fatty acids (PUFAs) (ranging in carbon length and unsaturation) supplied in growth medium were assimilated into bacterial phospholipids, as determined by thin-layer chromatography and liquid chromatography-mass spectrometry. The incorporation of fatty acids variably affected membrane permeability, as judged by uptake of the hydrophobic compound crystal violet. For each species, certain fatty acids were identified as affecting resistance to antimicrobial peptide treatment. Significant fluctuations were observed with regard to both motility and biofilm formation following growth in the presence of individual PUFAs. Our results illustrate the important and complex roles of exogenous fatty acids in the membrane physiology and virulence of a bacterial genus that inhabits aquatic and host environments containing an abundance of diverse fatty acids.

IMPORTANCE Bacterial responses to fatty acids include, but are not limited to, degradation for metabolic gain, modification of membrane lipids, alteration of protein function, and regulation of gene expression. Vibrio species exhibit significant diversity with regard to the machinery known to participate in the uptake and incorporation of fatty acids into their membranes. Both aquatic and host niches occupied by Vibrio are rife with various free fatty acids and fatty acid-containing lipids. The roles of fatty acids in the environmental survival and pathogenesis of bacteria have begun to emerge and are expected to expand significantly. The current study demonstrates the responsiveness of V. cholerae, V. parahaemolyticus, and V. vulnificus to exogenous PUFAs. In addition to phospholipid remodeling, PUFA assimilation impacts membrane permeability, motility, biofilm formation, and resistance to polymyxin B.

INTRODUCTION

Vibrio cholerae, Vibrio parahaemolyticus, and Vibrio vulnificus comprise the major human-pathogenic members of the Vibrio genus. While contaminated water or waterborne organisms represent their mode of transmission to humans, each species relies on a unique subset of virulence factors to cause disease. V. cholerae, V. parahaemolyticus, and V. vulnificus are responsible for gastrointestinal infections and diarrheal disease, while V. vulnificus also accesses skin wounds to establish a flesh-eating disease and/or septicemia with limb- and life-threatening consequences (1). Environmental factors that distinguish the aquatic reservoir and the human host have been experimentally linked to Vibrio survival and physiology. Examples of affected virulence pathways and phenotypic adaptations include the ToxR regulon in V. cholerae, the thermostable direct hemolysin in V. parahaemolyticus, and the capsular polysaccharide of V. vulnificus (2–4).

As inhabitants of both aquatic and host environments, Vibrio species have evolved well-tuned systems of survival and persistence based upon nutrient availability. Better-characterized pathways include iron, phosphate, and monosaccharide acquisition (5–8). Fatty acids represent environmental nutrients that were long considered solely as energy-yielding carbon sources but are garnering increased attention as signaling molecules and building blocks for membrane remodeling. In V. cholerae, transcriptional regulation by ToxT involves participation of palmitoleic acid, and the uptake and assimilation of environmentally relevant fatty acids have been demonstrated (9, 10). Unsaturated fatty acids repressed ctxAB and tcpA expression and enhanced motility in V. cholerae (11, 12). Other studies have linked fatty acids to influencing virulence cascades, motility, and cholera toxin binding to host intestinal epithelial cells (13, 14).

Gram-negative bacterial acquisition and utilization of long-chain fatty acids, as elucidated in Escherichia coli, begins with recognition and uptake by a dedicated outer membrane transporter, FadL (15). Passage of fatty acids across the periplasm is not well understood, and neither is the mechanism of transit through the inner membrane. Upon reaching the cytosolic face of the inner membrane, the acyl-coenzyme A (CoA) synthetase FadD generates acyl-CoA thioesters primed for entrance into the β-oxidation pathway or for reentry into the membrane for phospholipid construction (16, 17). A third destination for fatty acids involves their recognition as signaling molecules, either as free fatty acids in the periplasm or as acyl-CoA thioesters in the cytosol (18–20). It is increasingly evident that some bacteria have acquired or evolved extra homologs involved in fatty acid catabolism (21, 22).

The biological effects of long-chain polyunsaturated fatty acids (PUFAs) include, but are not limited to, molecular contributions in membrane-associated processes, critical membrane structural roles, and gene regulation (23–25). The contributions of PUFAs to membrane structure, protein function, and gene regulation draw attention to organisms that possess an excess of fatty acid handling genes, such as Vibrio species (17). Expanding the repertoire of accessible fatty acids may serve to broaden survival and pathogenesis capabilities.

Endogenous mechanisms for membrane lipid homeostasis have been addressed in many studies. These adaptations enhance bacterial survival under a variety of environmental conditions (26–29). However, the ramifications of exogenously acquired modifications to membrane lipids and other bacterial functions are not well understood. The effects of exogenous fatty acids on Gram-positive and Gram-negative bacterial growth and virulence have been mentioned in the literature. Unsaturated fatty acids have been implicated in promoting motility in Proteus mirabilis and virulence gene expression in Staphylococcus aureus, Listeria monocytogenes, and Pseudomonas aeruginosa (19, 30–32). Conversely, inhibitory growth effects of polyunsaturated fatty acids have been reported in P. aeruginosa, Helicobacter pylori, and Neisseria gonorrhoeae (33–35).

The first published indication of the enhanced fatty acid scavenging ability of Vibrio was by Wier et al., who observed a dramatically altered fatty acid content of V. fischeri following incubation within its symbiotic partner, the bobtail squid (36). A year later, the fatty acids in bile were found responsible for phospholipid alterations in V. cholerae, a study that demonstrated the assimilation of various exogenous fatty acids into major phospholipids (10). This study also discussed the additional molecular machinery possessed by Vibrio cholerae, including multiple homologs to FadL, FadD, and acyltransferases (10). In addition to scavenging free fatty acids, Vibrio species express a variety of lipases that can generate fatty acids through catalysis of larger lipid molecules (3, 37–39). In particular, a recent discovery described the surface-exposed lysophospholipase VolA, a lipoprotein conserved among Vibrio species that is transcriptionally and perhaps physically coupled to a fatty acid transporter (40). The expanded fatty acid handling capabilities of Vibrio warrant further investigation, especially with regard to possible roles as environmental cues and structural modifiers of membrane phospholipids. In the present study, we expand and further define the exogenous long-chain PUFA handling capabilities of Vibrio species and test the hypothesis that the resultant phospholipid remodeling affects virulence phenotypes and confers homeoviscous advantage in the context of physiologically relevant stresses. We report here that the three pathogenic Vibrio species utilize a plethora of long-chain polyunsaturated fatty acids, and their incorporation into membrane phospholipids affects membrane permeability, motility, biofilm formation, and subsequent exposure to antimicrobial stress.

RESULTS

PUFA exposure results in altered phospholipid profiles of Vibrio species.

To test the effect of various PUFAs on Vibrio phospholipids, bacteria were grown in the presence and absence of micromolar concentrations of each fatty acid. Cultures were grown in minimal medium to exclude any fatty acid contributions of complex media, as are present in Luria and tryptic soy broths. Bacterial phospholipids were extracted and examined by thin-layer chromatography (TLC). The major phospholipids produced by Vibrio include phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and cardiolipin (CL) (Fig. 1). Each of the three Vibrio species exhibited migrational shifts upward as carbon number and unsaturation of the exogenously supplied fatty acid increased, indicative of production of phospholipid species with more variable hydrophobicity compared to the control. In some cases, two distinct species of the same phospholipid can be discerned, particularly as carbon length and unsaturation is increased (see arrows). In addition, preference for the more hydrophobic species is evident for the PG of V. parahaemolyticus (Fig. 1).

FIG 1.

Thin-layer chromatography of phospholipids extracted from Vibrio species grown in the presence of individual polyunsaturated fatty acids. Bacteria were grown to the exponential phase of growth (OD₆₀₀ of ∼0.8) in G56 minimal medium (pH 7.4; 3% NaCl for V. parahaemolyticus only) at 37°C with or without 300 μM the indicated fatty acids (linoleic acid [18:2], alpha-linolenic acid [18:3α], gamma-linolenic acid [18:3γ], dihomo-gamma-linolenic acid [20:3], arachidonic acid [20:4], eicosapentaenoic acid [20:5], and docosahexaenoic acid [22:6]) prior to extraction of phospholipids by the method of Bligh and Dyer and thin-layer chromatography in the solvent system chloroform-methanol-acetic acid (65:25:10, vol/vol/vol). The plates were charred and scanned to produce the final image. Arrows depict examples of two distinct phospholipid species as the result of exogenous fatty acid incorporation.

UPLC–ESI-MS analyses indicate assimilation of exogenous PUFAs into Vibrio phospholipids.

The qualitative changes observed by TLC were further investigated using ultraperformance liquid chromatography-electrospray ionization mass spectrometry (UPLC–ESI-MS). Extracted phospholipids from each species following growth with (and without) individual PUFAs were subjected to chromatographic separation using gradient elution. Analysis of 300-ppm (total lipid extract) samples yielded chromatograms reflecting structural changes to the phospholipid profile compared to the control (Fig. 2). Figure 2 shows mass filtered chromatograms for each species grown in the presence of select fatty acids. Although all 3 Vibrio species were grown in the presence of all fatty acids, for the sake of clarity Fig. 2 only shows chromatograms comparing two fatty acid-exposed samples to their respective controls. Comparison of all other combinations can be found in the supplemental material. Alterations to the phospholipid profile are apparent due to the appearance of new chromatographic peaks not present in the control sample.

FIG 2.

Ultraperformance liquid chromatography-mass spectrometry of isolated phospholipids from Vibrio species grown in the presence and absence of fatty acids. Cultures of V. cholerae (A), V. parahaemolyticus (B), and V. vulnificus (C) were grown with or without 300 μM the indicated fatty acid at 30°C in G56 (pH 7.4) to the exponential phase of growth. Lipids were extracted by the method of Bligh and Dyer, and an extra wash step was included to increase the purity of the isolated lipids. Phospholipids were dissolved in a diluent (50:50 [30:70 25 mM ammonium acetate-methanol]-methanol) and injected into a Waters UPLC for gradient elution using a reversed-phase C₈ column. Detection was by quadrupole mass spectrometry in full scan mode (m/z 650 to 850) following electrospray ionization in the negative mode. Alterations of each chromatogram, compared to the control, indicate changes to the phospholipid profile resulting from exposure to individual fatty acids.

Like the TLC results, the LC-MS results show a prevalence of PE and PG species, and that these phospholipids display a variety of fatty acyl chains dependent on which fatty acid was supplemented. Analysis of the [M-H]⁻ ions shows that the new peaks correspond to phospholipid species comprised of at least one acyl chain matching the exogenously supplied fatty acid. This is most clearly shown in the extracted ion chromatograms (XIC) in Fig. 3. For example, chromatograms of V. cholerae phospholipids from fatty acid-supplemented cultures had peaks with [M-H]⁻ ions of m/z 767.5, 769.5, and 793.5, which are absent from the control (see Fig. S1 in the supplemental material). These ions correspond to PG species with fatty acyl compositions of 16:0/20:5, 16:0/20:4, and 16:0/22:6, respectively (Fig. 3A). Likewise, the same fatty acyl combinations were identified for PE species (m/z 736.5, 738.5, and 762.5; Fig. 3A; also see the supplemental material) from fatty acid-supplemented cultures. For V. parahaemolyticus (Fig. 3B), incorporation of 20:3, 20:5, and 22:6 fatty acids was indicated by the appearance of [M-H]⁻ ions with m/z 767.5, 793.5, and 771.5 (for modified PG species) and 736.5, 740.5, and 762.5 (for modified PE species), all of which are absent from the control. These ions correspond to PG(16:0/20:5), PG(16:0/22:6), PG(16:0/20:3), PE(16:0/20:5), PE(16:0/22:6), and PE(16:0/20:3), respectively. Likewise, the extracted ion chromatograms of V. vulnificus (Fig. 3C) are composed of ions corresponding to PE and PG species consisting of the fatty acyl groups 18:3, 20:3, and 20:4, which are not seen in the control.

FIG 3.

Extracted ion chromatograms of isolated phospholipids from Vibrio species grown in the presence and absence of fatty acids. V. cholerae (A), V. parahaemolyticus (B), and V. vulnificus (C). Representative structures of phosphatidylglycerol and phosphatidylethanolamine species, predicted using the parent ion m/z (see the supplemental material) and the Lipid Maps Database (http://www.lipidmaps.org/), are indicated. The structures displayed are speculative, since the acyl chains could be at either sn position.

All of the fatty acids tested yielded altered chromatographic patterns for each Vibrio species tested (Fig. S1). Mass spectral analysis reveals singly charged parent ions with masses that correspond to a variety of lipid species that are consistent with incorporation of the supplemented fatty acid into the membrane phospholipids. For each Vibrio species, it is observed that both PG and PE phospholipids show incorporation of all fatty acids tested (Fig. S2 to S4).

LC-MS was performed for Vibrio species grown at both 30°C and 37°C, and there were no differences observed between chromatograms. Furthermore, to assess whether PUFA incorporation is an active process, heat-killed V. cholerae was exposed to a subset of fatty acids (18:2, 20:4, and 22:6) prior to lipid extraction and LC-MS analyses. The chromatograms from fatty acid-exposed samples were identical to those of the unexposed control (data not shown), suggesting a concerted bacterial effort to facilitate uptake and incorporation of exogenous fatty acids into membrane phospholipids.

Exogenous fatty acids affect hydrophobic compound uptake in Vibrio species.

Having observed modifications to bacterial membrane phospholipid structures following fatty acid exposure, we assessed the resultant membrane permeability under the same conditions. By measuring bacterial uptake of the hydrophobic compound crystal violet, it was evident that the fatty acids variably, yet significantly, altered permeability among each Vibrio species (Fig. 4). In general, all three species experienced decreases in membrane permeability upon exposure to all fatty acids. V. cholerae displayed overall lower permeability than V. parahaemolyticus and V. vulnificus. Of note, arachidonic acid (20:4) did not change permeability in V. cholerae but decreased permeability by 20% in V. parahaemolyticus and by almost 30% in V. vulnificus. The omega-3 fish fatty acid, docosahexaenoic acid (22:6), minimally affected permeability in V. parahaemolyticus yet induced the greatest effect on V. cholerae. Clearly, exposure to fatty acids impacts permeability in Vibrio species, and there are some intrinsic differences in membrane permeability between V. cholerae and its pathogenic relatives.

FIG 4.

Effect of exogenous fatty acids on hydrophobic compound uptake in Vibrio species. Bacteria were grown at 37°C in G56 (pH 7.4) with and without 300 μM the indicated fatty acids to the exponential phase of growth (OD₆₀₀ of ∼0.8). Cultures were gently pelleted, washed with PBS, and resuspended in an equal volume of PBS (OD₆₀₀ of 0.4). The amount of crystal violet (CV) in the supernatant following centrifugation was measured (OD₅₉₀) at regular intervals and expressed graphically as a percentage of CV uptake. Results represent means from three independent determinations of CV uptake. All standard deviations were less than 5% (not graphed for visual clarity), and asterisks indicate significant difference (*, P < 0.002) compared to the control.

Exogenous fatty acids alter polymyxin B resistance in Vibrio species.

It was hypothesized that the assimilation of exogenous fatty acids would affect bacterial response to some antimicrobials, particularly antimicrobial peptides that rely on membrane insertion to exert their effect. We investigated the impact of fatty acids on polymyxin B resistance using microtiter plate MIC assays (Fig. 5). The experiments were performed two ways to compare the effect of fatty acid availability during the assay incubation. For all experiments, bacteria were grown in the presence of fatty acids to logarithmic phase prior to preparing them as an inoculum for the assay. Assays were performed with fatty acids absent (Fig. 5A, C, and E) and present (Fig. 5B, D, and F) during incubation with 2-fold concentrations of polymyxin B. Statistically significant (P < 0.002) shifts in MIC, from 100 μg/ml to 50 μg/ml, were observed when 18:2 and 20:4 were available during the assay (Fig. 5A and B). For V. parahaemolyticus, exposure to 18:3α and 20:4 in the assay shifted the MICs from 200 μg/ml to 50 μg/ml and 100 μg/ml, respectively (Fig. 5C and D). For V. vulnificus, access to 20:3, 20:4, and 22:6 during incubation with polymyxin B conferred protection, shifting MICs from 100 μg/ml to 200 μg/ml (Fig. 5E).

FIG 5.

Effect of exogenous fatty acids on polymyxin B resistance in Vibrio species. V. cholerae (A and B), V. parahaemolyticus (C and D), and V. vulnificus (E and F) were grown at 37°C in G56 (pH 7.4) with and without 300 μM the indicated fatty acids to the exponential phase of growth (OD of ∼0.8) Cultures were pelleted, washed with G56, and resuspended in G56 to an OD₆₀₀ of 0.1. The assay was performed two ways. For both sets of assays, bacteria were grown to logarithmic phase in the presence and absence of each fatty acid. The bacteria were pelleted and prepared as inoculum for the polymyxin B assays. In one set of assays (A, C, and E) there were no fatty acids present during incubation with polymyxin B. The other set of assays (B, D, and F) contained fatty acids at a final concentration of 300 μM during incubation with polymyxin B. The bacterial suspension was distributed into microtiter plates, and 2-fold concentrations of polymyxin B were added. After 20 h of incubation at 37°C, the optical density (OD₆₀₀) was read using a BioTek Synergy microplate reader. Values represent the means (±SD) from two independent experiments, each performed in triplicate. All standard deviations (<0.02) are masked by the marker symbols. Symbols circled by a dotted line indicate significant differences (P < 0.002) compared to the control (no fatty acid) at the particular polymyxin B concentration.

Exogenous fatty acids affect swimming motility in Vibrio species.

Another important phenotype directly linked to virulence in Vibrio species is motility. Small molecules, such as fatty acids, often serve as signals for chemotaxis and quorum sensing (41, 42). Furthermore, the effects of fatty acids on bacterial motility have been documented (30, 43). To investigate the impact of individual polyunsaturated fatty acids on Vibrio motility, bacteria were inoculated into motility plates prepared with either LB or marine broth supplemented with 0.35% agar. In general, bacterial motility was less affected by fatty acid supplementation in LB than marine broth (Fig. 6). Statistical significance was determined using the Student t test with a P value threshold of 0.002. V. cholerae demonstrated a significant decrease in motility when exposed to 18:3α and 22:6 in LB, whereas exposure to all fatty acids significantly lowered motility in marine broth. V. parahaemolyticus displayed decreased swimming motility in the presence of all fatty acids tested regardless of media; statistical significance (*, P < 0.002) was prevalent, with one fatty acid (20:3) causing significant decreases in marine broth but not in LB. Significant decreases in V. vulnificus motility were observed in LB supplemented with five fatty acids (18:2, 18:3γ, 20:3, 20:4, and 20:5), whereas three fatty acids (18:2, 18:3α, and 20:5) elicited significant increases in motility in marine broth. In summary, V. cholerae and V. parahaemolyticus responded similarly to fatty acids regardless of media, whereas V. vulnificus demonstrated an inverse response between media. Growth curves in marine broth did not explain the differences in motility, although some fatty acids alter lag phase depending upon media and Vibrio species (see the supplemental material).

FIG 6.

Effect of exogenous polyunsaturated fatty acids on motility of Vibrio species. Luria and marine broth (0.35% agar) motility plates were prepared and supplemented with 300 μM the appropriate fatty acid (after cooling to 55°C) or left unsupplemented. One microliter of inoculum (OD₆₀₀ of 0.1) was pipetted onto motility plates and observed after incubation at 30°C (Luria) or 37°C (marine). Values represent the means (±SD) from two independent experiments, each including four biological replicates. Significant differences in motility compared to the control were determined by Student's t test for each species (*, P < 0.002).

Fatty acids impact biofilm formation in Vibrio species.

The ability to form biofilms is an important phenotype for bacterial survival and persistence. To test the effect of fatty acids on biofilm formation, each Vibrio species was grown in the presence of fatty acid for 24 h and biofilms were measured using a crystal violet assay. Addition of all tested PUFAs significantly increased (P < 0.002) biofilm formation in V. cholerae (Fig. 7). Moreover, addition of the same PUFAs increased biofilm formation in V. parahaemolyticus except for 18:3γ and 22:6. Three fatty acids (18:3α, 20:3, and 20:4) increased biofilm formation in V. vulnificus, whereas exposure to 18:2, 18:3γ, and 22:6 caused significant decreases in biofilm formation. As observed for motility, growth curves could not account for differences in biofilm formation.

FIG 7.

Incubation with exogenous fatty acids alters biofilm formation in Vibrio species. Overnight cultures were pelleted, washed, resuspended in appropriate media, and inoculated onto microtiter plates (starting OD₆₀₀ of 0.1) in octuplet. Each culture was grown in the presence of 300 μM the indicated fatty acids. After 24 h of incubation, the biofilm assay by O'Toole was performed (66). The absorbance (OD₅₉₀) was measured using a BioTek Synergy microplate reader. Values represents the means (±SD) from two independent experiments performed in octuplet. Significant differences in biofilm formation were observed as determined by Student's t test and are indicated by asterisks (*, P < 0.002).

DISCUSSION

The current study demonstrates a broad range of effects of polyunsaturated fatty acids on the membrane structure and virulence behavior of pathogenic Vibrio species. Significantly expanding on a previous study (10), the results here broaden the fatty acid assimilation capability by identifying additional, relevant fatty acids that are predicted to be incorporated into phospholipids. Furthermore, this study extends exogenous fatty acid-mediated phospholipid remodeling to other pathogenic members of the genus. Beyond structural alterations and consequences to membrane permeability, fatty acids also impacted antimicrobial peptide resistance and caused striking changes in motility and biofilm formation. Accordingly, it is now recognized that fatty acids can play an important role in the environmental and host niches inhabited by Vibrio species.

Pertaining to the structural modification of phospholipids with exogenously acquired fatty acids, our results using thin-layer chromatography and UPLC-MS provide a convincing argument for the environmental relevance of this phenomenon. First, the concentration of fatty acids used in this study (300 μM) is within the expected range of estimated physiologically and environmentally relevant values (44–46). Second, the combination of altered migration patterns in TLC and divergent UPLC chromatogram profiles strongly suggest wholesale changes in the membrane phospholipid composition. Third, the identification of exogenous fatty acid-containing PG and PE species using the LIPID MAPS Structure Database (67) strengthens the likelihood that the predicted phospholipid structures are indeed correct.

Interestingly, fatty acids exert species-specific effects with regard to membrane permeability. There seems to be no correlation to (i) degree of saturation, (ii) carbon length, or (iii) stereochemical conformation (omega-3 versus omega-6), and membrane permeability does not predict antimicrobial peptide resistance, motility, or biofilm formation. However, fluctuations in membrane permeability trigger several pathways involved in maintaining membrane homeostasis, adaptations that affect bacterial survival depending upon the type and level of stress (68–70). It is also important to consider the effects of membrane lipid disorganization on protein structure and function, such as during stress response, transport, and protein folding, a phenomenon prevalent in both prokaryotes and eukaryotes (71). The assimilation of highly unsaturated fatty acids into the membrane would be expected to disrupt bilayer dynamics. It is important to note that the human-pathogenic Vibrio species encode a putative fatty acid cis/trans isomerase (VCA0552; VPA0677; VV2_0127) that may play a role in alleviating membrane perturbations resulting from incorporation of cis bond-containing isomers. In fact, a search for homologues to VCA0552 using the NCBI Taxonomy Browser reveals that the fatty acid cis/trans isomerase is well conserved among all Vibrio species. Depending upon which fatty acid is integrated, Vibrio may have evolved to respond accordingly by altering both membrane homeostasis and cellular processes. A recent study by Mizan et al. correlated V. parahaemolyticus biofilm formation with cell surface hydrophobicity, a finding that was corroborated in the present study with the exception of two fatty acids (18:3γ and 22:6) (53).

The effect of fatty acids on polymyxin B sensitivity are particularly intriguing. Alterations to the MIC of polymyxin were observed despite the bacteria not having access to fatty acid during the assay. This suggests that the preadaptation of the initial culture was sufficient to elicit the changes observed. As expected, the inclusion of fatty acids during the assay augmented the effects. The substantial decreases in MIC for V. cholerae and modest decreases in MIC for V. parahaemolyticus represent potential for fatty acids in synergistic treatment regimens. Conversely, as observed in V. vulnificus, there may be protective effects gained depending upon the fatty acid. Previous studies have documented the antioxidative properties of omega-3 fatty acids when incorporated into E. coli (54), and membrane incorporation of oleic acid and linoleic acid by Enterococcus faecalis heightens resistance to the antibiotic daptomycin and sodium dodecyl sulfate (SDS) (55).

The observed motility and biofilm effects may provide hints as to the role of exogenous fatty acids in the biphasic lifestyle of Vibrio species. For example, in its aquatic habitat Vibrio has been associated with fish, shellfish, and zooplankton (56, 57). These organisms possess high levels of PUFAs and, in the case of copepods, experience temperature-dependent and seasonal fluctuations in PUFA content (58–60). It is tempting to speculate that the fatty acid reservoir provided by plankton contributes to the established seasonality of Vibrio (56). Entering the human host exposes Vibrio to several stresses, including temperature increase, exposure to innate immune effectors, and extreme acidity. An adapted membrane fatty acid composition, acquired from its aquatic niche, may contribute to survival during the early stages of infection. Furthermore, the fatty acids encountered within the host may be instrumental for intestinal colonization and/or preparation for dispersal into the environment. Fatty acid responsiveness by V. cholerae may be a factor in the adoption of the hypervirulent state during exit from the host (61).

Several fatty acids (18:2, 18:3α, 20:3, 20:4, and 20:5) stimulated biofilm formation in all three Vibrio species, while V. cholerae and V. parahaemolyticus exhibited similar overall profiles (Fig. 6). Gamma-linolenic acid (18:3g) similarly affected biofilm formation in V. parahaemolyticus and V. vulnificus, a decrease of more than 50% (Fig. 6). This omega-6 fatty acid may act akin to another fatty acid messenger, cis-2-decenoic acid, identified in Pseudomonas aeruginosa (62). Indeed, there has emerged a quorum of studies highlighting the impact of exogenous fatty acids on bacterial behavior, both phenotypic and mechanistic (47–49, 63, 64).

The importance of fatty acids in the lifestyle of Vibrio species is underscored by the prevalence of genes predicted to be involved in fatty acid catabolism (Table 1). Multiple proteins with homology to FadL, FadD, and acyltransferases are identified throughout Vibrio genomes, supporting the idea that possession of extra molecular machinery allows for uptake and utilization of a wider repertoire of fatty acids than other bacteria. Another important gene recognized for its wide substrate spectrum is the acyl carrier protein synthetase characterized in V. harveyi. This gene product lacks significant homology among Vibrio species and is hypothesized to participate in the recycling of fatty acid products generated by bacterial luciferase to facilitate bioluminescence (50, 51). Table 1 also includes Vibrio proteins sharing homology to V. harveyi 1DA3 AasS (VME_37260). Although no strong hits were identified, the acyl-CoA synthetase motifs in V. harveyi AasS contribute to the redundancy of proteins with homology to FadD (52). Perhaps the acquisition and maintenance of these genes originated from the aquatic reservoir, where carbon scavenging is at a premium. Alternatively, the environmental associations with aquatic organisms (oysters, fish, etc.) may have guided the evolution of proteins targeting the abundant supply of PUFAs. Bacterial utilization of fatty acids, or other lipid mediators, is an intriguing mechanism that may impact virulence and/or environmental persistence. A better understanding of microbial recognition and response to niche-specific lipid signatures may reveal strategies for prevention, control, and treatment of several pathogens. The results of this study contribute to our current understanding of how and why bacteria have evolved to utilize a variety of exogenous lipid molecules, thus providing insight into their survival both outside and within the host, as well as uncovering pathways vulnerable to antimicrobial attack.

TABLE 1.

Putative fatty acid uptake and assimilation homologs in Vibrio species^a

Species	Gene(s) predicted forb:
	FadL	FadD	PlsB; PlsC	PlsX; PlsY	AasSVh
V. cholerae N16961	VCA0862, VC1042, VC1043	VC1985, VC2341, VCA1110	VC0093; VC2513	VC2024; VC0523	VC1985, VC2341, VC2484
V. cholerae O395	VCO395_A0560, VCO395_A0561, VCO395_0374	VCO395_A1570, VCO395_A1920, VCO395_0135	VCO395_A2422; VCO395_A2095	VCO395_A1610; VCO395_A0051	VCO395_A1570, VCO395 _A1920, VCO395_A2060
V. cholerae O139 (strain FC1105)	CBG35_04240, CBG35_04245, CBG35_11350	CBG35_00935, CBG35_06560	CBG35_14780; CBG35_08310	CBG35_01130; CBG35_15535	CBG35_00935, CBG35_03360
V. parahaemolyticus RIMD 2210633	VPA0222, VPA0860, VP2212, VP2213	VP0870, VP1964, VP2425, VP0351, VPA1152, VP1730	VP2947; VP2657	VP2057; VP0410	VP1964, VP0870, VP2425, VP1730, VP0351
V. vulnificus CMCP6	VV1_1971, VV1_1972, VV2_0249	VV1_0136, VV1_1742, VV1_0649	VV1_1165; VV1_0678 VV1_2699	VV1_3012; VV1_0626	VV1_0136, VV1_0649
V. alginolyticus ATCC17749	N646_1316, N646_4253, N646_3657, N646_1315	N646_3023, N646_1045, N646_1520, N646_2540, N646_4025	N646_2043; N646_1749	N646_1133; N646_2566	N646_1045, N646_3023, N646_1520
V. harveyi 1DA3	VME_36620, VME_04810, VME_24030, VME_24020	VME_14840, VME_38380, VME_13100, VME_37260	VME_22490; VME_11370	VME_27810; VME_09520	VME_14840, VME_38380, VME_13100, VME_27510, VME_16890
V. fischeri MJ11	VFMJ11_1637, VFMJ11_1945, VFMJ11_1946	VFMJ11_1834, VFMJ11_2368, VFMJ11_0535, VFMJ11_2376	VFMJ11_2567; VFMJ11_0392	VFMJ11_1870; VFMJ11_2359	VFMJ_2368, VFMJ_1834, VFMJ_0535

^aVibrio species possess an expanded repertoire of genes involved in fatty acid uptake. Putative fatty acid uptake and assimilation homologs in Vibrio species is shown. Representative Vibrio genomes were bioinformatically assessed (see Materials and Methods) for homologs to a long-chain fatty acid transporter (FadL), long chain fatty acyl-CoA synthetase (FadD), acyltransferases (PlsB/C/X/Y), and V. harveyi acyl-acyl carrier protein synthetase (AasS_Vh), involved in acquisition and assimilation of exogenous fatty acids. The identified homologues illustrate the potential for expanded utilization of exogenous fatty acid in Vibrio species.

^bHits receiving a maximum score greater than 200 are underlined.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Vibrio cholerae El Tor C6706, Vibrio parahaemolyticus (clinical isolate; East Tennessee State University Clinical Laboratory), and Vibrio vulnificus CAP-D-08 (clinical isolate; East Tennessee State University Clinical Laboratory) were used in this study. G56 minimal media (0.4% glucose, 0.4% Casamino Acids [Fisher BioReagents], supplemented with 300 mM NaCl) was used for growth of bacteria in each experiment unless described otherwise. Fatty acids used in this study were purchased from Cayman Chemicals and administered individually at a concentration of 300 μM for each experiment.

Bacterial lipid extraction and thin-layer chromatography.

Lipids were extracted from 14 ml of bacterial culture by the method of Bligh and Dyer (65) and spotted onto silica gel 60 TLC plates. Lipids were separated using a solvent system consisting of chloroform, methanol, and acetic acid (65:25:10, vol/vol/vol). Once dry, the plates were sprayed with a solution of 10% sulfuric acid in 100% ethanol and exposed to 150°C for approximately 1 min. The plates were scanned using a Canon CanoScan 9000F.

UPLC–ESI-MS.

Lipids were extracted from 20 ml of bacterial culture by the method of Bligh and Dyer (65). Prior to analysis, dried lipid extracts were massed and then brought up in sufficient diluent to produce a 300-ppm (total lipid) sample. Diluent consisted of a 50:50 mixture of solvents A and B, where A was 30:70 25 mM, pH 6.7, ammonium acetate-methanol and B was methanol. All reagents were Optima grade (Fisher Scientific), and 18.2 MΩ · cm water was produced from a Direct Q3 Milli-Q system (Millipore, Bedford, MA). Samples were prepared in LC-MS certified glass autosampler vials, and 5 μl was injected for analysis. Chromatographic separation was achieved using gradient elution on an Acquity UPLC system (Waters, Milford, MA) equipped with a BEH C₈ column (2.1 by 100 mm; 1.7-μm particles). Detection was by quadrupole mass spectrometry following electrospray ionization in the negative mode. The Quattro Micro mass spectrometer was operated with a capillary voltage of 1.5 kV and a cone voltage of 50 V. Desolvation gas was N₂ heated to 350°C and a flow rate of 750 liters/h. The quadrupoles were scanned over the m/z range 650 to 850 amu using a scan time of 0.5 seconds. MassLynx v. 4.1 software was used for LC-MS data collection and analysis.

Heat inactivation of Vibrio cholerae.

Heat-inactivated bacteria (20 ml) were prepared by incubating bacterial suspensions, 1 × 10⁹ Vibrio organisms ml⁻¹, in a shaking water bath at 56°C for 1 h. The suspensions were incubated at room temperature for 24 h and then at 4°C for another 24 h. Complete inactivation was confirmed by testing growth of suspensions on LB and blood agar plates at 37°C for 3 days. The Vibrio organisms were incubated with select fatty acids (18:2, 20:4, and 22:6) for 3 h at 37°C. Lipids were extracted and analyzed by UPLC–ESI-MS as described above.

Crystal violet uptake assay.

Each Vibrio species was grown in 7 ml of G56 minimal medium in the presence or absence of 300 μM each individual PUFA to logarithmic phase (all cultures were captured at an optical density at 600 nm [OD₆₀₀] of 0.75 to 0.85). V. parahaemolyticus was supplemented with 3% NaCl. The cultures were gently pelleted, washed with phosphate-buffered saline (PBS), and resuspended in 5 ml PBS at an OD of 0.4. Crystal violet (5 μg/ml) was added to the cells and cultures were gently agitated (50 rpm). Every 5 min, 1 ml was removed and pelleted, and the supernatant was assessed spectrophotometrically at 590 nm. Inclusion of a control (containing CV but no bacteria) allowed normalization of the data. The amount of CV measured (representing dye not taken up) was converted to percentage of uptake using Excel. Three independent biological replicates were performed, and all standard deviations (SD) were calculated to be less than 5% (not graphed for visual clarity).

Swimming motility assay.

Soft-agar assays were prepared using Luria and marine broths (0.35% agar), supplemented with 300 μM the appropriate fatty acid. Marine agar was injected with 1 μl of inoculum (OD₆₀₀ of 0.1), and the diameters were measured after 10 h of incubation at 37°C for V. cholerae and V. parahaemolyticus. For V. vulnificus, diameters were measured after 24 h of incubation. Luria agar was also injected with 1 μl of inoculum (OD₆₀₀ of 0.1) and the diameters were measured at 30°C for V. cholerae (12 h), V. parahaemolyticus (6 h), and V. vulnificus (24 h). Two independent experiments were performed in quadruplicate, and P values were determined using the Student t test (2-tailed, paired).

Biofilm assay.

Assessment of biofilm formation was performed using the protocol by O'Toole (66). Briefly, Vibrio species were grown overnight in Luria broth, and cultures were prepared in 96-well microtiter plates containing G56 minimal media supplemented with each individual fatty acid or left unsupplemented. Following incubation for 20 h at 37°C, planktonic cells were removed and the plates were gently washed with distilled water (dH₂O). To each well, 125 μl of 5 μg/ml crystal violet solution was added and plates were incubated at room temperature for 15 min. Plates were washed 3 times with dH₂O and allowed to dry for at least 3 h. One hundred twenty-five microliters of 30% acetic acid was added and allowed to incubate for 15 min. The solubilized crystal violet was transferred to a fresh microtiter plate, and absorbance was read at 590 nm using a BioTek Synergy microplate reader. Two independent experiments were performed in octuplet, and P values were determined using the Student t test (2-tailed, paired).

Polymyxin B susceptibility assay.

Vibrio species were grown to exponential phase of growth in G56 minimal medium in the presence and absence of individual fatty acids (300 μM). Cultures were centrifuged, washed with G56, and resuspended at an OD₆₀₀ of 0.13. Bacterial inoculum (170 μl) was added to 2-fold concentrations of polymyxin B (30 μl) for a total of 200 μl per well (bacterial starting OD₆₀₀ of 0.1). One set of microtiter plate assays was performed without fatty acid supplementation, while another set contained fatty acids at a concentration of 300 μM during incubation. Plates were incubated for 20 h at 37°C, and absorbance was read at 600 nm using a BioTek Synergy microplate reader. Two independent experiments were performed in triplicate, and P values were determined using the Student t test (2-tailed, paired).

Bioinformatics.

Table 1 was generated by consulting National Center of Biotechnology Information's Basic Local Alignment Search Tool (BLAST) (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Briefly, the protein sequences of E. coli MG1655 FadL (b2344), FadD (b1805), phospholipid acyltransferases (b4041, b3018, b1090, and b3059), and V. harveyi AasS (VME_37260) constituted input for homolog searches in the representative sequenced Vibrio species. The expect threshold was set to 1e−10. Homologs with a maximum score greater than 200 are underlined. No Vibrio species analyzed produced a maximum score above 175 when using V. harveyi AasS as the query sequence.