The strains and plasmids used in this study are listed in Supplementary Table S1. The E. coli K-12 strain DH5α was used in this study for gene cloning. The P. aeruginosa strains used in this study were the wild-type strain PAO1, fabV deletion strain PAO170 and fabI deletion strain PAO272. Luria-Bertani (LB) medium was used as the enriched medium for E. coli and P. aeruginosa growth. Antibiotics were used at the following concentrations (in micrograms per milliliter): sodium ampicillin, 100 (for E. coli); kanamycin sulfate, 30 (for E. coli) or 100 (for P. aeruginosa); gentamicin, 10 (for E. coli) or 100 (for P. aeruginosa); and triclosan, 3 (for P. aeruginosa). Isopropyl-β-D-thiogalactoside (IPTG) was used at a final concentration of 1 mmol/L.

Pseudomonas aeruginosa fabI (PafabI) and fabV (PafabV), V. cholerae fabV (VcfabV), E. faecalis fabK (EnfabK) and B. subtilis fabL (BsfabL) were amplified from genomic DNA obtained from wild-type strains of P. aeruginosa, V. cholerae, E. faecalis and B. subtilis, respectively. The primers are listed in Supplementary Table S2. The PCR was performed with Pfu DNA polymerase and the products were inserted into T-vector pMD19, to give plasmids pYH1 (PafabI), pYH2 (PafabV), pYH3 (VcfabV), pYH4 (EnfabK) and pYH5 (BsfabL), respectively. All the genes in the T-vectors were confirmed by sequencing, performed by Shanghai Sangon, Inc. The fab genes in T-vector were digested with NdeI and HindIII, and the appropriate fragments were isolated and ligated into pSRK-Km (Khan et al., 2008) to create plasmids pSRK-PI, pSRK-PV, pSRK-VV, pSRK-EK and pSRK-BL, respectively. All these plasmids were introduced into the P. aeruginosa PAO170 strain via electroporation to give strains: PI170 (PAO170/pSRK-PI), PV170 (PAO170/ pSRK-PV), VV170 (PAO170/pSRK-VV), EK170 (PAO170/pSRK-EK) and BL170 (PAO170/ pSRK-BL), respectively.

The cellular lipid assay was adapted from that of Stead (1989). Briefly, cultures of the P. aeruginosa strains were grown at 37°C in LB medium overnight. Cells were then harvested from 10 ml aliquots of the cultures and washed with fresh LB medium at room temperature. Cellular lipids were saponified by the addition of 1 ml NaOH in methanol solution (NaOH 45 g; methanol 150 ml; water 150 ml). The samples were placed in a boiling water bath for 30 min. The tubes were vortexed before and once during boiling. FAs were methylated by the addition of 2 ml of 6 mol/L HCl in methanol (325 ml 11.6 mol/L HCl, plus 275 ml methanol). The samples were heated at 80°C for 10 min and immediately cooled to below 20°C. The fatty acid methyl esters were extracted three times with 1.25 ml petroleum ether. The samples were dried under a stream of nitrogen in a fume hood. The esters were analyzed by gas chromatography-mass spectrometry (GC-MS) as described previously (Feng et al., 2015; Mao et al., 2016). The data are presented as percentages of the total FAs and represent the mean ± standard error for three independent determinations.

Cell-free extracts of P. aeruginosa were prepared from early stationary phase growing cultures (optical density at 600 nm of 0.8 to 1.0). Cells grown in LB medium at 37°C were harvested by centrifugation and were then suspended in lysis buffer (0.1 mol/L sodium phosphate, pH 7.5, 5 mmol/L β-mercaptoethanol, 1 mmol/L EDTA). The cell lysates were prepared by passing the cell suspensions three times through a French pressure cell. Cell debris was removed by ultracentrifugation for 1 h at 260,000 × g, and the supernatants were dialyzed against lysis buffer for 24 h and saved as cell extracts. The ENR activity of cell-free extracts was determined by using trans-2-decenoyl-ACP as the substrate, which was done by monitoring spectrophotometrically the decrease in the absorbance at 340 nm using an NADH extinction coefficient of 6,220 mol/L^-1. The trans-2-decenoyl-ACP was prepared by using a previously published procedure (Zhu et al., 2010). The reaction mixtures for activity assays contained 150 μmol/L NADH, 10 ng of the cell-free extracts, 100 μmol/L trans-2-decenoyl-ACP, and 0.1 mol/L LiCl in 0.1 mol/L sodium phosphate buffer (pH 7.0).

The swarming, swimming, and twitching motilities of P. aeruginosa were investigated using the following media: (1) swim plate [1% tryptone, 0.5% NaCl, 0.3% agar (Difco Bacto Agar)], (2) swarm plate (0.45% tryptone, 0.13% yeast extract, 0.22% NaCl, 0.5% glucose, 0.5% agar), and (3) twitch plate (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 1% agar). The agar media were air-dried for 5–10 min before use. For the swimming and swarming assays, plates were point-inoculated with bacteria from an overnight culture using a sterile toothpick on the surface, and the plates were incubated at 30°C for 24–48 h. For the twitching motility assay, the cells were stabbed into the bottom of twitching plates using a toothpick and incubated at 37°C for 24 h. The motility was assessed by examining the circle around the inoculation site formed by the growing bacteria.

For the rhamnolipid assay, P. aeruginosa strains were cultured to stationary phase in PPGAS (120 mmol/L Tris-HCl, pH 7.2, 20 mmol/L NH₄Cl, 20 mmol/L KCl, 1.6 mmol/L MgSO₄, 1% tryptone and 0.5% glucose) at 30°C. The supernatants were first adjusted to pH 2.0 with 1 mol/L HCl, and 1 ml of these supernatant samples was extracted twice with an equal volume of diethyl ether, followed by vacuum drying. The concentration of rhamnolipid was determined by measuring the concentration of rhamnose with the sulfuric acid-anthrone reagent (0.2% anthrone, 85% sulfuric acid) method, using rhamnose sugar as the standard, at 620 nm.

Chrome azurol S (CAS)-LB agar plates were used for the siderophore secretion assay: 10 ml sterilized 10X CAS solution (1 mmol/L chrome azurol S, 2 mmol/L cetyltrimethylammonium bromide, 500 μmol/L FeCl₃⋅6H₂O) was mixed with 100 ml LB agar to prepare CAS-LB agar plates. The plates were dried for 1 h at room temperature prior to inoculation, following which 10 μl of overnight P. aeruginosa culture was drop inoculated onto a large CAS-LB agar plate (ø 200 mm). Transparent circles were observed after 48 h of incubation at 37°C.

Cultures were grown in LB medium for 24 h at 37°C with shaking (250 rpm) to stationary phase. Supernatants were harvested by centrifugation (12,000 × g, 5 min at room temperature), and 5 ml of supernatant for each sample was extracted using an equal volume of ethyl acetate. The organic phase was concentrated to dryness. For the N-butyryl-L-homoserine lactone (C₄-HSL) assay, 10 μl of each sample resolved in 50 μl ethyl acetate was absorbed by a ø 0.5 cm filter paper and then tiled on an LB agar plate, which was newly covered with Chromobacterium violaceum CV026. Purple circles were observed after overnight incubation at 30°C. To detect N-3-oxo-dodecanoyl-L-homoserine lactone (3-oxo-C₁₂-HSL) production, the dry extracts were resolved in 50 μl acetonitrile and 10 μl of each sample was analyzed with HPLC according to the previously published procedure (Ortori et al., 2007). In addition, 10 mmol/L pure 3-oxo-C₁₂-HSL (Sigema) acetonitrile solution was used as a reference.

The pyocyanin assay is based on the absorbance of pyocyanin at 520 nm in acidic solution. A 5-ml sample of P. aeruginosa culture grown in LB was extracted with 3 ml of chloroform and then re-extracted into 1 ml of 0.2 mmol/L HCl to give a pink to deep red solution. The absorbance of this solution was measured at 520 nm. Concentrations, expressed as micrograms of pyocyanin produced per milliliter of culture supernatant, were determined by multiplying the optical density at 520 nm by 17.072.

LasA protease activity was determined by measuring the ability of P. aeruginosa culture supernatants to lyse boiled Staphylococcus aureus cells. A 30 ml overnight culture of S. aureus was placed in a boiling water bath for 10 min and then centrifuged for 10 min at 10,000 × g. The resulting pellet was resuspended in 10 mmol/L Na₂HPO₄ (pH 7.5) and adjusted to an OD₆₀₀ of 0.9. A 100 μL aliquot of P. aeruginosa supernatant was then added to 900 μL S. aureus suspension, and the OD₆₀₀ was determined after 5, 10, 15, 20, 25, 30, 35, 40, 45, 60, 75, 90, and 105 min.

Analysis of variance for experimental datasets was performed using JMP soft ware, version 5.0 (SAS Institute Inc., Cary, NC, USA). Significant effects of treatment were determined by the F value. When a significant F test was obtained, separation of means was accomplished by Fisher’s protected LSD (least significant difference) at P ≤ 0.05.

In order to confirm the role of FabV in P. aeruginosa resistance to triclosan, we tested the sensitivity to triclosan of PAO170 (ΔfabV) carried pSRK-PI (carries P. aeruginosa fabI), pSRK-PV (carries P. aeruginosa fabV) and pSRK-VV (carries V. cholerae fabV), respectively. As expected, the derivative of strain PAO170 carrying the vector plasmid (pSRK-Km) was very sensitive to triclosan (the triclosan MIC was about 1 μg/ml; Figure 2 and Table 1). In contrast, expression of PaFabI, PaFabV, and VcFabV proteins from pSRK-Km-derived plasmids increased triclosan resistance. However, the MIC of strain PAO170 (ΔfabV) carrying plasmid pSRK-PI (5 μg/ml) was much lower than that of strain PAO170 (ΔfabV) carrying plasmid pSRK-PV or pSRK-VV (above 2,000 μg/ml), which was the same as that of the wild-type strain PAO1 (Figure 2 and Table 1). These findings suggest that the sensitivity of mutant strain PAO170 (ΔfabV) to triclosan was due to its lack of triclosan-resistant ENR. To investigate this point further, we also introduced pSRK-EK (carries E. faecalis fabK) and pSRK-BL into P. aeruginosa fabV mutant strain PAO170, respectively. The test of sensitivity to triclosan showed that strain PAO170 (ΔfabV) carrying both plasmids was highly resistant to triclosan; the triclosan MIC was above 2,000 μg/ml (Supplementary Figure S1 and Table 1). These data confirmed that FabV, a triclosan-resistant ENR, leads to resistance of P. aeruginosa to triclosan.

Motility is strongly associated with the pathogenesis of P. aeruginosa. The bacterium exhibits movement on surfaces by utilizing three types of motility: swimming, swarming, and twitching (Inoue et al., 2008; Overhage et al., 2008). Figure 3 shows the colony spreading patterns of these types of motility in P. aeruginosa mutant strains. The twitching pattern of mutant strain PAO272 (ΔfabI) was almost the same as that of wild-type strain PAO1, whereas strain PAO170 (ΔfabV) formed a much smaller twitching pattern than wild-type strain PAO1 (Figure 3A), which suggests that deletion of the fabI gene had no effect on twitching motility of P. aeruginosa, but deletion of fabV inhibited P. aeruginosa twitching motility.

Mutant strain PAO170 (ΔfabV) did not form a swimming pattern after 24 h of incubation on swimming assay plates at 30°C. However, when the incubation time was extended to 48 h, it was able to swim and form a typical swimming pattern, although the average swimming diameter of strain PAO170 (ΔfabV) was much smaller than that of the wild-type strain and the fabI mutant strain PAO272 (ΔfabI; Figure 3B). This result was not consistent with a previous study, in which Bai et al. (2007) reported that mini-Mu insertion in fabV caused the PA68 strain of P. aeruginosa to lose swimming motility. However, it is important to note that, when Bai et al. (2007) carried out the swimming assay for the fabV strain, they incubated the mutant strain for only 16 h on a swimming assay plate. Thus, we speculate that mutation of fabV did not inhibit P. aeruginosa swimming motility directly, but led to slow growth and therefore caused it to form the swimming pattern late. We tested the growth of strain PAO170 (ΔfabV) in swimming medium and confirmed that it grew more slowly than wild-type strain PAO1 (data not shown). Meanwhile, deletion of fabI had no effect on swimming motility of P. aeruginosa because, like wild-type strain PAO1, the mutant strain PAO272 (ΔfabI) was able to form a normal swimming pattern on the swimming assay plate after 24 h of incubation (Figure 3B).

The swarming motility of P. aeruginosa strains was also tested on semisolid plates (containing 0.5% agar) at 30°C (Figure 3C). Although both the fabI mutant strain PAO272 (ΔfabI) and the wild-type strain swarmed normally on semisolid plates after 24 h of incubation, the fabV mutant strain PAO170 failed to form a swarming pattern, even when the incubation time was extended to 48 h (Figure 3C). In swarming medium, the fabV mutant strain grew more slowly than wild-type strain PAO1 (data not shown). Therefore, to test whether the deficiency of swarming motility was due to the weak growth of the fabV mutant strain, we repeated the swarming assay on semisolid plates containing 0.4% agar. As expected, after 24 h of incubation the wild-type strain PAO1 formed obvious swarms on semisolid plates containing 0.4% agar, while the fabV mutant strain failed to form a swarming pattern (Figure 3D). Subsequently, using a toothpick, we inoculated wild-type strain PAO1 on a semisolid plate on which the fabV mutant strain had been cultured for 48 h. After a further 24 h of incubation, the PAO1 strain formed a swarming pattern, but the fabV strain still did not swarm (Figure 3D). These data indicate that it was disruption of fabV that caused P. aeruginosa to lose swarming motility. To test this hypothesis, we checked the swarming motility of complementary strain PAO170/pSRK-PV on a semisolid plate (containing 0.5% agar). Like wild-type strain PAO1, strain PAO170/pSRK-PV was able to form a typical swarming pattern after 24 h of incubation on the semisolid plate (Figure 3E). Moreover, it was interesting that not only P. aeruginosa fabV but also P. aeruginosa fabI could restore normal swarming motility to the fabV mutant (Figure 3E). We also tested whether V. cholerae fabV, E. faecalis fabK and B. subtilis fabL complemented the swarming motility of the fabV mutant. The data showed that all these ENR encoding genes were able to restore swarming motility to the fabV mutant (Figure 3E and Supplementary Figure S2). These results suggest that ENR is required for swarming motility in P. aeruginosa.

It has been reported that in defined minimal medium the fabV mutant strain PAO170 was viable, but its generation time (about 118 min) was longer than that of the wild-type strain PAO1 (about 40 min; Zhu et al., 2010). To determine whether nutrient conditions affect the growth of strain PAO170, its growth in LB medium was examined. The growth of stain PAO170 in LB medium was significantly slower than that of wild-type strain PAO1, confirming that nutrient conditions did not improve the growth of strain PAO170 (Figure 4A). Complementation of strain PAO170 with plasmid pSRK-PV increased the growth of PAO170 to a certain extent, but not to the level of the wild-type strain (Figure 4A). Plasmid pSRK-PI or pSRK-VV also allowed strain PAO170 to grow to the same level as pSRK-PV (Figure 4A). These results indicated that deletion of PafabV from the genome caused P. aeruginosa to grow slowly and that ENR encoding genes could restore growth of the fabV mutant to a certain extent. We also investigated the growth of PAO170 carrying E. faecalis fabK or B. subtilis fabL encoding plasmids and obtained similar results (Figure 4B). To investigate further possible reasons for the weak growth of strain PAO170, the fatty acid compositions of strain PAO170 were determined by GC-MS. Although the mutant strain PAO170 contained the same fatty acid species as the wild-type strain, PAO170 produced more 3-hydroxy fatty acids (3-HFAs) and fewer FAs (Table 2). The ratio of FAs to 3-HFAs in strain PAO170 was 2.58 and was lower than that of the wild-type strain (3.87), which showed significant difference (P < 0.05) between mutant strain PA170 and wild-type strain PAO1. Complementation of strain PAO170 with plasmid pSRK-PI or pSRK-VV restored the ability of strain PAO170 to produce the wild-type strain’s level of 3-HFAs and FAs. The ratio of FAs to 3-HFAs in strain PAO170 carrying pSRK-PI or pSRK-VV was 3.70 or 3.61, respectively. Moreover, strain PAO170 carrying pSRK-PV produced more FAs and fewer 3-HFAs than the wild-type strain. The ratio of FAs to 3-HFAs in strain PAO170 carrying pSRK-PV was 4.57 (Table 2), and there was highly significant difference between strain PV170 and PAO170 (P < 0.01). It is noteworthy that mutant strain PAO170 carrying pSRK-PV, pSRK-PI or pSRK-VV grew to the same level. This indicated that a change in fatty acid composition caused by deletion of fabV was not the main reason for weak growth in P. aeruginosa fabV mutant. Another possibility was that deletion of fabV led to decreased activity of ENR, which was a contributing factor in the lower growth rate of PAO170. We measured the activity of ENR in cell-free extract of PAO170 by using trans-2-decenoyl-ACP as a substrate and monitoring the decrease in NADH absorbance at 340 nm. The ENR activity (in μmol/min per mg extract protein) in cell-free extract of PAO170 was 0.175 ± 0.006, and was lower than that in the cell-free extract of wild-type strain PAO1 (0.478 ± 0.032). The ENR activity in strain PAO170 carrying plasmids pSRK-PI, pSRK-PV and pSRK-VV was 0.290 ± 0.001, 0.292 ± 0.004 and 0.300 ± 0.006, respectively. Although the ENR activity in strain PAO170 carrying plasmids with genes encoding foreign ENR was increased, it was still lower than that in wild-type strain PAO1. We also detected the ENR activity in cell-free extracts of PAO170 carrying pSRK-EK or pSRK-BL encoding plasmids. The ENR activity in both strains was increased, but did not reach the level of that in the wild-type strain either (data not shown). Therefore, we suggest that the level of activity of ENR, rather than the PafabV gene or protein product, is essential to the growth of P. aeruginosa.

Given that deletion of fabV decreased the activity of the ENR of P. aeruginosa significantly, decreased production of several exo-products, which are synthesized using the intermediates in the fatty acid synthetic pathway, would be expected in PAO170 mutant strains. We first looked at levels of rhamnolipids produced by P. aeruginosa strains. Rhamnolipids are secreted surfactant glycolipids assembled by rhamnosyltransferase using L-rhamnose and 3-hydroxydecanoyl-ACP from the FAS pathway (Deziel et al., 2003; Zhu and Rock, 2008). Rhamnolipid quantification by colorimetric detection of rhamnose showed a twofold decrease in the fabV mutant PAO170, while the amount of rhamnolipid produced by complemented strain PAO170/pSRK-PV was up to 80% of that of the wild-type strain PAO1 (Figure 5A).

Pyoverdine is the dominant siderophore of P. aeruginosa and is assembled from tetradecanoyl-ACP (Drake and Gulick, 2011; Hannauer et al., 2012). The siderophore produced by P. aeruginosa strains was examined on LB-CAS indicator plates. Yellow-orange halos around the cultures indicate where the siderophores have sequestered Fe³⁺ away from the blue CAS-Fe³⁺ complex. The halo produced by the PAO170 mutant strain was much smaller than that of the wild-type strain: its average diameter was only a quarter of that of the wild-type strain. After complementation with plasmid pSRK-PV, the average diameter of the yellow-orange halo of the PAO170 mutant strain increased to one half that of the wild-type strain (Figure 5B).

The rhl and las QS signal molecules, N-butanoyl-L-homoserine lactone (C₄-HSL) and N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C₁₂-HSL), are synthesized from S-adenosylmethionine and butanoyl-ACP or 3-oxododecanoyl-ACP, respectively (Parsek et al., 1999; Hoang et al., 2002). Both acyl-ACPs are intermediates in the FAS pathway. Therefore, we also measured the levels of both QS signal molecules produced by P. aeruginosa strains. The rhl and las QS signals were extracted from culture supernatants of P. aeruginosa strains. The acyl-HSLs were detected using an agar overlay of a C. violaceum reporter strain CV026, which produces a purple halo in response to acyl-HSLs. The purple halo around PAO170 mutant strain was weak and small in comparison to that of the wild-type strain (Figure 5C). Production of acyl-HSLs was increased to wild-type levels by complementation with strain PAO170/pSRK-PV. The 3-oxo-acyl-HSL was detected using HPLC. The level of 3-oxo-acyl-HSL showed a threefold decrease in the fabV mutant in comparison to that of the wild-type strain and there was highly significant difference between strain PV170 and PAO170 (P < 0.01). Complemented strain PAO170/pSRK-PV produced more 3-oxo-acyl-HSL than the fabV mutant, but not to the level of the wild-type strain (Figure 5D).

The PQS signal molecules also utilize 3-ketoacyl medium-chain fatty acid metabolites to synthesize 2-heptyl-3-hydroxy-4-quinolone (PQS; Girard and Bloemberg, 2008; Williams and Camara, 2009). We thus analyzed the levels of PQS in supernatant extract in P. aeruginosa strains. The level of PQS was reduced to a trace in the supernatant extract when fabV was deleted. In contrast, PQS production was markedly enhanced in the complemented strain (data not shown).

We also complemented the PAO170 mutant with other ENR encoding genes, such as P. aeruginosa fabI or V. cholerae fabV, and all complemented strains produced more rhamnolipids, pyoverdine, AHLs and PQS than the fabV mutant (Figures 5A–D).

The QS systems, including rhl, las and PQS, regulate the ability of P. aeruginosa to produce virulence factors (e.g., LasA/LasB and alkaline proteases, phospholipases, lipases, exotoxin A, rhamnolipid, pyocyanin, and others; Girard and Bloemberg, 2008; Williams and Camara, 2009). Thus, reduction of QS signal molecules should decrease production of these virulence factors. To confirm this, we tested the production of LasA and pyocyanin by P. aeruginosa strains. The data showed that the activity of LasA in PAO170 mutant was much lower than that in the wild-type strain, and the amount of pyocyanin produced by PAO170 strain was about 50% of that produced by the wild-type strain (Figures 5E,F). However, complementation of PAO170 mutant strain with ENR genes increased the activity of LasA and the amount of pyocyanin (Figures 5E,F).

Based on the above study, it would be expected that, on addition of QS signal molecules to cultures of P. aeruginosa strains, production of some virulence factors that are regulated by the QS system would be restored. First, we added C₄-HSL or 3-oxo-C₁₂-HSL to swarming plates and tested the swarming motility of the fabV mutant strain. Exogenous addition of both QS signals to the swarming plate restored the ability of the fabV mutant to form a typical swarming pattern after 24 h of incubation, although the pattern was still smaller than that of the wild type (Figure 6A). We also examined production of pyocyanin, rhamnolipid and LasA protease by PAO170 mutant strain after exogenous addition of QS signals. The data showed that, although not up to the levels of the wild-type strain, the production of all these virulence factors was increased to a certain extent (Figures 6B,C). These findings indicated that deletion of fabV reduced the activity of ENR in P. aeruginosa, subsequently depressed the production of QS signals and finally led to reduced virulence factors in P. aeruginosa.