The E. coli K-12 strains, P. aeruginosa strains, and plasmids used in this study are listed in Supplementary Table S1. Luria–Bertani (LB) medium was used as the enrichment medium for growth of E. coli and P. aeruginosa. Chromobacterium violaceum CV026 and Agrobacterium tumefaciens NTL4 (pZLR4) were grown on LB medium at 28°C. Antibiotics were used at the following concentrations (in micrograms per milliliter): sodium ampicillin, 100; kanamycin sulfate, 30; chloramphenicol, 30; and gentamicin, 10 (for E. coli) or 100 (for P. aeruginosa). L-Arabinose was used at a final concentration of 0.01%. Isopropyl-β-D-thiogalactoside (IPTG) was used at a final concentration of 1 mM, and 5-bromo-4-chloro-3-indolyl-β-D-galactoside (X-Gal) was used at a concentration of 20 μg/mL.

To clone the P. aeruginosa acpP1, acpP2, and acpP3 genes, genomic DNA was extracted from P. aeruginosa strain PAO1 using the Takara DNA extraction kit. The PCR products, amplified from strain PAO1 genomic DNA using Pfu DNA polymerase and the primers listed in Supplementary Table S2, were digested by NdeI and HindIII and inserted into pBAD24m (Zhu et al., 2010) between the same sites to produce plasmids: pBAD24m-acpP1, pBAD24m-acpP2, and pBAD24m-acpP3. The acpPs genes were confirmed by sequencing by Shanghai Sangon, Inc. (Shanghai, China).

To produce plasmids pCD4 or pCD5, the PCR fragments that were amplified from plasmids pBAD24m-acpP1 or pBAD24m-acpP2, using the primers listed in Supplementary Table S2, were digested with XbaI and HindIII and cloned into pET-28 (b) at the same sites. The acpP3 fragment of pBAD24m-acpP3 digested by NdeI and HindIII was cloned between the NdeI and HindIII sites of pET-30 (a) to yield plasmid pCD6. To construct plasmids pCD8, pCD9, and pCD10, the plasmids pCD4, pCD5, and pCD6 were digested with XbaI and HindIII, respectively, and the DNA fragments were purified and inserted into the same sites of pBluescript SK(+), respectively. To yield plasmids pCD1, pCD2, and pCD3, the PCR fragments were amplified from plasmids pCD8, pCD9, or pCD10 using the primers listed in Supplementary Table S2, respectively, and then digested with double restriction enzymes, for which the sites had been designed into the primer sequences. After purification, the fragments were ligated with the vector pTac85 (Marsh, 1986), and they were digested with the same restriction enzymes, respectively.

The pET-28b-derived plasmids pCD4, pCD5, and pCD6 were introduced into E. coli strain BL21 (DE3), and the respective proteins, PaAcpP1, PaAcpP2, and PaAcpP3, were expressed at high levels and purified as described previously (Thomas and Cronan, 2005). The plasmids pYFJ84 and pCD15 were introduced into E. coli strain BL21 (DE3) for expression of Vibrio harveyi AasS and P. aeruginosa PcpS proteins. The AasS and PcpS proteins were also purified as described previously (Barekzi et al., 2004; Jiang et al., 2006).

The form of the ACPs was detected by measuring their migration pattern using conformationally sensitive gel electrophoresis. Briefly, the standard assay mixtures contained, in a volume of 40 μL, 0.1 M Tris-HCl (pH 8.8); 1 mM CoA, 25 mM MgCl₂, 2 mM DTT, 100 μM P. aeruginosa ACPs, and 2 μg P. aeruginosa PcpS. The mixture was pre-incubated at 37°C for 1 h. The reaction products were resolved by conformationally sensitive gel electrophoresis on 17.5% polyacrylamide gels containing 0.5 M urea optimized for separation. For purification of P. aeruginosa holo-ACPs, 2 mL of the above reaction mixture was used and incubated at 37°C for 4 h. The purification of holo-ACPs was carried out according to the methods described previously (Zhao et al., 2005). Briefly, two volumes of acetone were added to the reaction mixture, and the mixture was allowed to precipitate overnight at -20°C. The precipitates were pelleted at 20,000 × g for 20 min at 4°C and then washed twice with three volumes of acetone. The pellet was air-dried and re-suspended in 200 μL Tris-HCl (20 mM, pH 7.4), and the suspension was centrifuged at 20,000 × g for 30 min at 4°C. The supernatants were collected and stored at -80°C. Their molecular masses were determined by matrix-assisted laser desorption/ionization (MALDI)-time-of-flight (TOF)-mass spectrometry (MS) (Bruker Autoflex III; Freemont, CA, United States) according to the methods described previously (Mao et al., 2015).

For acylation of P. aeruginosa holo-ACPs, we used the protocols described by Jiang et al. (2006). The reaction mixture contained, in a volume of 40 μL, 10 mM ATP, 20 mM MgSO₄, 0.1 M Tris-HCl (pH 7.8), 1 mM dithiothreitol, 0.1 mg P. aeruginosa holo-ACPs, 0.5 mM octanoic acid and 4 μg V. harveyi AasS. The reaction mixtures were incubated at 37°C for 2 h and resolved by conformationally sensitive gel electrophoresis on 17.5% polyacrylamide gels containing 2.5 M urea optimized for the separation.

Cell-free extracts of P. aeruginosa were prepared according the processes described previously (Hoang and Schweizer, 1999). The exponentially growing cells (OD₆₀₀ of 0.8 to 1.0) in 100 mL LB medium were harvested by centrifugation and then suspended and washed in 2 mL 0.1 M sodium phosphate buffer (pH 6.5). Cell lysates were prepared by passing the cell suspensions three times through a French pressure cell (19,000 lb/in²). Cell debris was removed by ultracentrifugation for 1 h at 200,000 × g, and the supernatants were saved as cell extracts. The protein concentration of the cell extract was determined using the Quick Start Bradford dye reagent. The fatty acid synthetic reaction mixtures included, in a volume of 100 μL, 0.1 M Tris-HCl (pH 7.8), 1 mM dithiothreitol, 100 μM P. aeruginosa holo-ACPs (AcpP1, AcpP2, or AcpP3), 100 μM NADH, 100 μM NADPH, 100 μM malonyl-CoA, 100 μM acetyl-CoA and 5 mg P. aeruginosa cell-free extract. Reactions were incubated at 37°C for 2 h. Analysis of the ACP thioesters was performed by conformationally sensitive gel electrophoresis on 17.5% polyacrylamide gels containing 2.5 M urea optimized for the separation. Cerulenin was added to the reaction mixtures at a final concentration of 100 μM when needed.

To disrupt the P. aeruginosa acpPs genes, suicide plasmids were constructed as follows. The 1.3-kb DNA fragment containing acpP1 amplified from P. aeruginosa genomic DNA using the primers listed in Supplementary Table S2 was digested with BamHI and HindIII and cloned between the same sites of pBluescript SK (+) to produce pCD16. A gentamicin resistance cassette amplified from p34s-Gm (Dennis and Zylstra, 1998) using the primers listed in Supplementary Table S2 was digested by EcoRI and inserted into the same site of pCD16 to produce pCD17. Subsequently, the BamHI–HindIII fragment from pCD17 was inserted between the same sites of pK18mobsacB (Schafer et al., 1994) to yield the suicide plasmid pCD18. A similar method was used to construct the suicide plasmids pCD21 and pCD25. All these plasmids were introduced into P. aeruginosa PAO1 via conjugal transfer from E. coli S17-1 (Supplementary Figure S1A). Single-crossover integrants into the strain PAO1 chromosome were selected by chloramphenicol and gentamicin resistance. Cultures grown from the integrant colonies were plated onto LB medium containing 10% sucrose and gentamicin in order to select for the loss of the vector sequences from the PAO1 chromosome. The successful construction of the designed mutations was evaluated by PCR analysis using the primers listed in Supplementary Table S2. Two single mutant strains were obtained: PA-A2 (acpP2::Gm^r) and PA-A3 (acpP3::Gm^r) (Supplementary Figures S1B,C), but no acpP1 insertion mutant was obtained.

To replace P. aeruginosa acpP1 with E. coli acpP, the DNA fragments acpP1up, EcacpP and acpP1down were amplified using the primers listed in Supplementary Table S2. The above three DNA fragments were ligated by overlap PCR and cloned into the T-vector, pMD19, to produce pCD29. The BamHI–HindIII fragment from pCD29 was cloned between the same sites of pK18mobsacB to yield plasmid pCD30. The gentamicin resistance cassette of p34s-Gm, digested by BamHI, was inserted into the same site of pCD30 to produce plasmid pCD31. Using similar procedures, the replacement strain PA-A1 (acpP1::EcacpP) was obtained (Supplementary Figure S3).

To obtain double or triple P. aeruginosa acpPs mutant strains, we also constructed plasmid pCD22, in which the gentamicin resistance cassette in plasmid pCD 21 was replaced with a tetracycline resistance cassette (Dennis and Zylstra, 1998). Using similar procedures to those described above, on a background of PA-A1, we obtained the double mutant strains PA-A12 (acpP1::EcacpP acpP2::Gm^r) (Supplementary Figure S4A) and PA-A13 (acpP1::EcacpP acpP3::Gm^r) (Supplementary Figure S4B). On a background of PA-A3, we obtained the double mutant strain PA-A23 (acpP2::Tc^r acpP3::Gm^r) (Supplementary Figure S4C), and on a background of PA-A23, we produced a triple mutant strain, PA-A123 (acpP1::EcacpP acpP2::Tc^r acpP3::Gm^r) (Supplementary Figure S4D).

To confirm the mutant strains, all acpP allelic genes were sequenced by Shanghai Sangon, Inc. (Shanghai, China).

The cultures were grown aerobically at 37°C in LB medium overnight. Cells were harvested and washed three times with sterile water. Fatty acid methyl esters were synthesized and extracted as described previously (Stead, 1989). Briefly, cellular lipids were saponified by the addition of 2 mL of sodium hydroxide/methanol solution at 100°C for 40 min with shaking (800 rpm). The fatty acids were then methylated by the addition of 4 mL of hydrochloric acid/methanol solution at 80°C for 30 min, and immediately cooled to below 20°C. Fatty acid methyl esters were obtained by three extractions, each with 1 mL of petroleum ether. The solvent was removed under a stream of nitrogen, and the residue was dissolved in 100 μL of hexane. The crude extract was passed through a 0.22 μm Mini-star filter unit, and 2 μL of the extract was analyzed by gas chromatography–mass spectrometry (GC-MS).

The motility assay was carried out as described previously (Huang et al., 2016). The swarming motility of P. aeruginosa was investigated using the following media: 0.45% tryptone, 0.13% yeast extract, 0.5% agarose, 0.22% NaCl, and 0.5% agar. The plates were air-dried for 5–10 min before use. The bacterial cells (OD₆₀₀ to 0.8) were gently inoculated using a toothpick at the center of the media surface, and the plates were incubated at 30°C for 24 h. The twitching motility assay was performed using the following media: 1% tryptone, 0.5% yeast extract, 0.5% NaCl, and 1% agarose. The cells were stabbed into the bottom of a Petri dish containing each of the above media using a toothpick, and incubated at 37°C for 20 h. The movement of the colony on the interface between the medium and the dish was observed.

Rhamnolipids were extracted as previously described (Huang et al., 2016). Briefly, cells were removed from PPGAS (NH₄Cl 20 mM, KCl 20 mM, Tris-HCl pH 7.2 120 mM, MgSO₄ 1.6 mM, glucose 0.5%, and peptone 1.0%) cultures grown at 30°C for 24 h by centrifugation (10 min at 6,000 × g), and the supernatant was acidified to pH 2 with concentrated HCl. Equal volumes (1 mL) of acidified supernatant and chloroform:methanol (2:1) were mixed and vortexed for 1 min, and the lower organic phase was collected after centrifugation (10 min at 10,000 × g). The extraction was repeated, and the pooled organic phases were evaporated to dryness, re-suspended in 1 mL methanol, filtered through a 0.45-μm membrane, evaporated to dryness, and re-suspended in 20 μL methanol. The samples were analyzed by thin-layer chromatography (TLC) using silica 60 F₂₅₄ plates (Merck, Darmstadt, Germany) with a mobile phase consisting of chloroform:methanol:acetic acid (65:15:2, v/v/v). Orcinol reagent dissolved in 15% H₂SO₄ at a final concentration of 2% was used to visualize rhamnolipid spots at 100°C for 2–5 min. 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. To identify the fatty acid component of the rhamnolipid, rhamnolipids were extracted from 5 mL supernatant using the same procedures as described above. Fatty acid methyl esters were synthesized using rhamnolipid and extracted as described previously (Zhu et al., 2010). Briefly, the rhamnolipids were dissolved in 1.2 mL dry methanol, and 0.2 mL of 25% (vol/vol) sodium methoxide was added. After the solution was allowed to stand for 15 min at room temperature, 1.2 mL of 2 M HCl was added, and the fatty acid methyl esters were obtained by three extractions each with 1.2 mL of petroleum ether. The solvent was removed under a stream of nitrogen, and the residue was analyzed by measuring the formation of hydroxylacyl-ACP. The standard mixture contained 0.1 M sodium phosphate (pH 7.0), 10 mM ATP, 20 mM MgSO₄, 1 mM DTT, 100 μM E. coli ACP, 1 μg/μL His-tagged V. harveyi AasS purified from E. coli and fatty acid methyl esters obtained from rhamnolipid. The mixture was incubated at 37°C for 1 h and resolved by conformationally sensitive gel electrophoresis on 17.5% polyacrylamide gels containing 2.5 M urea optimized for the separation.

Extraction of quorum-sensing signal molecules was performed as described previously (Yuan et al., 2012). P. aeruginosa cells were grown in LB to the early stationary phase, and cells were removed from 10 mL growth medium by centrifugation at 12,000 × g for 15 min at 4°C. Quorum-sensing signal molecules were twice extracted from 10 mL of supernatant for each sample, using either an equal volume of ethyl acetate for homoserine lactones (HSL) or an equal volume of acidified ethyl acetate for 2-heptyl-3-hydroxy-4(1H)-quinolone (PQS). The organic extracts were concentrated to dryness using a nitrogen bubbler, and the residue was resuspended in 100 μL of ethyl acetate (HSL) or methanol (PQS). For determination of AHL-like molecules with short acyl chains, the biosensor C. violaceum CV026 was utilized (McClean et al., 1997). A. tumefaciens NTL4 (pZLR4) was utilized to detect AHLs with long acyl chains (Cha et al., 1998). HPLC analysis was carried out with an HPLC series (Agilent Technologies, Palo Alto, CA, United States) equipped with an Elite C18 column (4.6 mm × 250 mm, 5 μm particle size) that was maintained at 45°C, with a UV detector set at 250 nm. The mobile phase comprised 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B) (Ortori et al., 2007). The flow rate was 1 mL/min. The elution conditions were as follows: 0 min 35% B, linear gradient to 60% B for 10 min and then a linear gradient from 60 to 95% B over 5 min, then 5 min 95% B and then ramped back to the starting conditions in 9 min. The column was re-equilibrated for a total of 5 min. A 10-μL volume was injected onto the column. PQS content was assessed by normal-phase thin-layer chromatography (TLC) on activated silica 60 F254 plates (Merck) according the method described previously (Yuan et al., 2012). Authentic acyl- and 3-oxo-acyl-HSL, and PQS (Sigma) were used as standards.

Pseudomonas aeruginosa cells were grown in LB at 37°C for 12 h, and cells were removed from 10 mL growth medium by centrifugation at 5,000 × g for 10 min at 4°C. LasA protease activity was determined by measuring the ability of P. aeruginosa culture supernatants to lyse boiled Staphylococcus aureus RN4200 cells, as described previously (Huang et al., 2016). A 30-mL overnight culture of S. aureus grown in tryptic soy broth was placed in a boiling water bath for 10 min and then centrifuged for 10 min at 10,000 × g. The resulting pellet was re-suspended in 10 mM Na₂HPO₄ (pH 7.5) and adjusted to an OD₆₀₀ of 0.9. A 100-μL aliquot of bacterial supernatant was then added to 900 μL of S. aureus suspension, and the OD₆₀₀ was determined after 5, 10, 15, 20, 25, 30, 35, 40, 45, 60, 75, 90, and 105 min. LasB protease activity in P. aeruginosa culture supernatants was determined using the elastin Congo red assay (Yuan et al., 2012). After a 3-h digestion period at 37°C (100 mM Tris, 1 mM CaCl [pH 7.5] with 20 mg of elastin Congo red substrate [Sigma] plus 100 μL of spent supernatant), suspensions were clarified by centrifugation (16,000 × g, 5 min, RT), and the absorbance at 495 nm was measured to determine the amount of liberated dye.

The pyocyanin assay is based on the absorbance of pyocyanin at 520 nm in acidic solution (Essar et al., 1990). A 5-mL sample of culture grown in LB was extracted with 3 mL of chloroform and then re-extracted into 1 mL of 0.2 N HCl to give a pink to deep red solution. The absorbance of this solution was measured at 520 nm. Concentrations, expressed as μg of pyocyanin produced per mL of culture supernatant, were determined by multiplying the optical density at 520 nm (OD₅₂₀) by 17.072.

The CAS-LB plates were prepared according to a modification of the published protocol (Yuan et al., 2012). A 10× blue dye solution consisting of 1 mM chrome azurol S (CAS) (Acros), 2 mM cetyltrimethylammonium bromide, and 500 μM FeCl₃ 6H₂O was sterilized by autoclaving, and then 10 mL was added to 100 mL of molten LB Miller agar (1.5%). Plates were allowed to solidify and dried at room temperature for 1 h before streaking cultures containing each strain to be tested.

An analysis of variance for the experimental datasets was performed using JMP software version 5.0 (SAS Institute Inc., Cary, NC, United States). Significant effects of treatment were determined by the F-value (P = 0.05). When a significant F-test was obtained, a separation of means was accomplished by Fisher’s protected LSD (least significant difference) at P = 0.05.

In order to test whether P. aeruginosa acpP1, acpP2, and acpP3 function in fatty acid biosynthesis, these genes were inserted into the IPTG-inducible vector pTac85 (Marsh, 1986) to yield plasmids pCD1 (carrying Pa acpP1), pCD2 (carrying Pa acpP2), and pCD13 (carrying Pa acpP3), respectively. The plasmids were then used to transform E. coli strain CY1877, a strain in which the chromosomal acpP gene had been deleted and replaced with a chloramphenicol resistance cassette. The deletion replacement event was done in the presence of an ampicillin-resistant pBAD24-derived plasmid expressing the E. coli acpP gene to allow survival of the acpP deletion strain. Growth of strain CY1877 was dependent on the addition of arabinose because transcription of the acpP gene was from a vector araBAD promoter. Only E. coli CY1877 carrying the plasmid encoding acpP1 (pCD1) grew on LB medium in the presence of IPTG and the absence of arabinose, whereas derivatives of strain CY1877 carrying empty vector, or plasmids encoding acpP2 or acpP3 failed to grow under this condition (Figure 1A). P. aeruginosa acpP1 complemented the E. coli acpP mutation, showing that P. aeruginosa AcpP1 can function as an ACP and replace functional E. coli ACP. To examine further whether P. aeruginosa AcpP2 and AcpP3 are involved in fatty acid synthesis, these ACPs were overexpressed and purified as described in the section “Materials and Methods.” First, phosphopantetheinylatation of these three ACPs was tested. As expected, after incubation with P. aeruginosa phosphopantetheinyl transferase (PcpS) and CoA, the migration of AcpP1 did not change significantly, suggesting that P. aeruginosa AcpP1 may be activated from apo-ACP to holo-ACP by E. coli holo-ACP synthase (AcpS) (Figure 1B), which is consistent with the finding that acpP1 complemented the E. coli acpP mutation. To confirm the above observation, we analyzed AcpP1, which was purified from E. coli cells directly or from phosphopantetheinylatation with P. aeruginosa PcpS, by MALDI-TOF-MS (Supplementary Figure S5). The data showed that AcpP1 purified from E. coli cells includes apo- and holo- forms of ACPs, while after phosphopantetheinylatation with P. aeruginosa PcpS, the AcpP1 all became holo-AcpP1 (Supplementary Figures S5A,B). However, both AcpP2 and AcpP3 showed different migration rates after incubation with PcpS and CoA (Figure 1B), indicating that the purified AcpP2 and AcpP3 proteins from E. coli were apo-ACPs, and that AcpP2 and AcpP3 could not be activated properly in E. coli from apo-ACP to holo-ACP by E. coli AcpS, which were confirmed by analysis with MALDI-TOF-MS (Supplementary Figures S5C–F). We also tested whether these ACPs can be acylated by V. harveyi acyl-ACP synthetase (AasS). After incubation of these holo-ACPs with octanoic acid, ATP and AasS, all reaction mixtures produced new bands on native gels, showing that all these holo-ACPs could be acylated by AasS (Figure 1C). Finally, whether these ACPs can be used as substrates during fatty acid synthesis by P. aeruginosa in vitro was examined. As expected, P. aeruginosa AcpP1 was a good substrate for fatty acid synthesis. The reaction mixture containing holo-AcpP1, malonyl-CoA, acetyl-CoA, NADH, NADPH and P. aeruginosa cell-free extract produced short-chain acyl-ACPs in the presence of cerulenin, an antibiotic that inhibits the activity of long chain 3-keto-acyl-ACP synthases. In the absence of cerulenin, the reaction mixture produced various long-chain acyl-ACPs (Figure 1D). However, P. aeruginosa AcpP2 and AcpP3 were poor substrates for fatty acid synthesis in vitro: only the reaction mixture containing AcpP2 produced a trace amount of short-chain acyl-ACPs in the absence of cerulenin (Figure 1D), but the reaction mixture containing AcpP3 did not produce any acyl-ACPs. All these results imply that P. aeruginosa AcpP2 and AcpP3 are not involved in fatty acid synthesis in P. aeruginosa.

To examine further the physiological functions of the three P. aeruginosa ACP homologs, we attempted to inactivate each of the putative acpP genes by allelic replacement. Three pK18mobsacB-borne suicide plasmids used to insert a gentamicin resistance cassette into acpP1, acpP2, or acpP3 genes were constructed and yielded pCD18, pCD21, and pCD25, respectively (Supplementary Figure S1). These plasmids were introduced into P. aeruginosa PAO1 via conjugal transfer from E. coli S17-1 (Supplementary Figure S1). After negative selection, mediated by Bacillus subtilis sacB, and identification by colony PCR using the primers listed in Supplementary Table S2, two single mutant strains were obtained: PA-A2 (acpP2::Gm^r) and PA-A3 (acpP3::Gm^r) (Supplementary Figures S1A,B), but no acpP1 insertion mutant was obtained, which indicates that P. aeruginosa acpP1 seems to be an essential gene.

To confirm this finding, two new suicide plasmids, pCD28 (Supplementary Figure S2) and pCD31 (Supplementary Figure S3A), were constructed. The plasmid pCD28 was used to delete acpP1, while pCD31 was used to replace P. aeruginosa acpP1 with E. coli acpP. Both plasmids were introduced into P. aeruginosa PAO1 by conjugation. Cultures grown from the single-integrant colonies were plated onto LB medium containing sucrose. Colonies that were sensitive to gentamicin were analyzed by PCR using the primers listed in Supplementary Table S2. Unfortunately, no acpP1 deletion strain was obtained (Supplementary Figure S2). However, we obtained a replacement mutation strain, PA-A1 (acp1::EcacpP), in which the P. aeruginosa acpP1 gene was replaced with the E. coli acpP gene (Supplementary Figures S3B,C). These results indicated that acpP1 is essential for the growth of P. aeruginosa.

Using similar procedures, we also constructed three double mutants, PA-12 (acpP1::EcacpP, acpP2::Gm^r) (Supplementary Figure S4A), PA-13 (acpP1::EcacpP, acpP3::Gm^r) (Supplementary Figure S4B), and PA-23 (acpP2::Tc^r, acpP3::Gm^r) (Supplementary Figure S4C), and one triple mutant, PA-123 (acpP1::EcacpP, acpP2::Tc^r, acpP3::Gm^r) (Supplementary Figure S4D). Next, the growth of these mutant strains was tested. All mutant strains were viable and grew as well as wild type strain PAO1 (data not shown).

To investigate further the functions of the three ACP homologs in fatty acid synthesis in P. aeruginosa, the fatty acid composition of the acpPs mutant strains grown in LB medium was first determined by gas chromatography–mass spectrometry (GC-MS). The fatty acid profiles of all the mutant strains were almost the same and were not distinct from that of wild type strain PAO1 (Table 1 and Supplementary Table S3). This indicated that E. coli ACP protein may replace functions of P. aeruginosa AcpP1, and that AcpP2 and AcpP3 may not play roles in fatty acid synthesis.

In P. aeruginosa, fatty acids are not only the compulsory components of membrane phospholipids, but are also utilized to produce exo-products, including three acylated quorum-sensing signal molecules, rhamnolipids and siderophore pyoverdine (Hoang et al., 2002; Zhu and Rock, 2008; Yuan et al., 2012). To confirm the above hypothesis and to determine whether AcpP2 or AcpP3 is specified to synthesize these exo-products, the production of these exo-products in P. aeruginosa acpPs mutant strains was examined.

First, the rhl QS signal molecule, N-butanoyl-L-homoserine lactone (C₄-HSL), produced by P. aeruginosa acpPs mutant strains was tested using the C. violaceum reporter strain CV026, which produces a purple halo in response to acyl-HSLs. The purple halo around mutant strains PA-A2, PA-A3, and PA-A23 was as large as that around wild type strain PAO1 (Figure 2A). The amount of C₄-HSL produced by PA-A2, PA-A3, or PA-A23 was also determined by high performance liquid chromatography (HPLC), which showed that all these mutant strains produced almost the same amount of C₄-HSL (about 4.1 μM), which was similar to the amount produced by the PAO1 strain (Figure 2B). However, the purple halo around mutant strain PA-A1 was weak and small in comparison with that of wild type strain PAO1 (Figure 2C). The HPLC analysis showed that the amount of C₄-HSL produced by PA-A1 (about 2.4 μM) was significantly lower than that produced by the PAO1 strain (about 4.1 μM) (P < 0.01) (Figure 2C). In contrast, the double mutant strains, PA-A12 and PA-A13, and the triple mutant, PA-123, produced a similar amount of C₄-HSL to the PA-A1 strain (Figure 2B). However, expression of the wild-type P. aeruginosa acpP1 gene in mutant strain PA-A1 increased the production of C₄-HSL, and under IPTG induction the amount of C₄-HSL was restored to the level of the wild type strain (Figure 2C).

The las QS signal molecule, N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C₁₂-HSL), produced by these mutant strains was also investigated using the reporter strain A. tumefaciens NL4(pZLR4) and HPLC. The data showed that replacement of acpP1 with E. coli acpP reduced the ability of P. aeruginosa to produce 3-oxo-C₁₂-HSL (Figures 3A,B), but mutation of acpP2 or acpP3 did not affect the production of the las QS signal molecule (Figures 3B,C).

We analyzed the levels of PQS in supernatant extracts of the mutant strains by thin-layer chromatography (TLC). The level of PQS was reduced 40% in the supernatant extract of mutant PA-A1 in comparison with that produced by wild type strain PAO1. However, disruption of acpP2 or acpP3 did not cause reduction of the amount of PQS produced by P. aeruginosa strains (data not shown).

Rhamnolipids are secreted surfactant glycolipids formed by acylation of L-rhamnose using 3-hydroxydecanoyl-ACP (Zhu and Rock, 2008). Therefore, the levels of rhamnolipids produced by P. aeruginosa mutant strains were examined. First, TLC was used to analyze the rhamnolipids produced by P. aeruginosa strains. The amount of rhamnolipids in acpP1 mutant strain PA-A1 was decreased, but complementation with plasmid pSRK-Pa, in which the wild-type acpP1 gene was cloned into pSRK-Km, restored the ability to produce rhamnolipids to the level of that in wild type strain PAO1 (Figure 4A). Quantification of rhamnolipids by colorimetric detection of rhamnose showed a twofold decrease in the acpP1 mutant PA-A1, while the amount of rhamnolipids produced by complemented strain PK-Pa (PA-A1/pSRKPa) was up to 80% of that of the wild type strain PAO1 (Supplementary Figure S6). We also analyzed the 3-hydroxyacids composition of the rhamnolipids produced by acpP1 mutant strains. The data showed that, although the same fatty acid species were present in PA-A1 as in the wild type PAO1, the 3-hydroxyacids were quantitatively reduced among the rhamnolipids produced by PA-A1 (Figure 4B). Next, we investigated the amount of rhamnolipids produced by P. aeruginosa acpP2 or acpP3 mutant strains. The level of rhamnolipids produced by PA-A2, PA-A3, or PA-A23 was almost the same as that in strain PAO1. In strain PA-A1, disruption of a single acpP2 or single acpP3, or disruption of both acpP2 and acpP3 genes, did not cause the reduction in rhamnolipids seen following replacement of P. aeruginosa acpP1 with the E. coli acpP gene (data not shown).

Pyoverdine is the dominant siderophore of P. aeruginosa and is assembled from tetradecanoyl-ACP (Quadri et al., 1999). Secretion of pyoverdine by P. aeruginosa acpPs mutant strains was examined by culture on LB–chrome azurol S (CAS) indicator plates. Zones around the cultures that change from blue to yellow–orange indicate where the siderophores have sequestered Fe³⁺ away from the blue CAS-Fe³⁺ complex. Both wild type strain PAO1 and mutant strain PA-A1 generated similar prominent clear zones, which suggested that replacement of acpP1 with E. coli acpP did not affect siderophore secretion (Figure 4C). Moreover, the yellow–orange halo around mutant strains PA-A2, PA-A3, and PA-A23 was as large as that around strain PAO1. In addition, on a background of strain PA-A1, mutation of acpP2 or acpP3 did not reduce siderophore secretion by P. aeruginosa (data not shown).

On the basis of the above results, we conclude that neither acpP2 nor acpP3 is involved in fatty acid synthesis in P. aeruginosa and that they have no function in the production of various exo-products. Furthermore, although E. coli ACP replaces the function of P. aeruginosa AcpP1 in fatty acid synthesis, P. aeruginosa AcpP1 is required for the production of some exo-products, including three acylated quorum-sensing signal molecules and rhamnolipids.

The QS systems, including rhl, las, and PQS, not only control the production of many virulence factors by P. aeruginosa (such as LasA/LasB, rhamnolipid, procyanin, and others), but are also involved in the regulation of P. aeruginosa motility (Jimenez et al., 2012). Replacement of acpP1 by E. coli acpP caused P. aeruginosa to reduce the production of QS signal molecules. Therefore, the mutant strain PA-A1 was expected to have reduced production of some virulence factors and attenuated motility. To confirm this, we first tested the production of LasA, LasB, and procyanin. LasA protease activity was determined by measuring the ability of P. aeruginosa culture supernatants to lyse boiled S. aureus cells. The data showed that the activity of LasA in mutant strain PA-A1 was much lower than that in wild type strain PAO1, but complementation strain PK-Pa (PA-A1/pSRK-Pa) increased LasA to the level of that in wild type strain PAO1 (Figure 5A). The elastase proteolytic activity (LasB) was measured using the elastin Congo red assay. In comparison with that in wild type strain PAO1, the activity of LasB was decreased twofold in the PA-A1 strain, while the activity of LasB in complemented strain PK-Pa (PA-A1/pSRK-Pa) was up to 80% of that in wild type strain PAO1 (data not shown). However, procyanin production by mutant strain PA-A1 was significantly increased (P < 0.005) in comparison with wild type strain PAO1, and plasmid pSRK-Pa did not restore the ability of strain PA-A1 to produce procyanin to the level of that in the wild type strain, indicating that other factors may be involved in the regulation of the production of procyanin in mutant strain PA-A1 (Figure 5B).

We tested three types of motility: swimming, swarming, and twitching, in mutant strain PA-A1. The swimming pattern of mutant stain PA-A1 was almost the same as that of wild type strain PAO1, which suggested that replacement of acpP1 with E. coli acpP did not affect the swimming motility of P. aeruginosa (data not shown). However, replacement of acpP1 by E. coli acpP reduced the twitching ability of P. aeruginosa. The twitching pattern of mutant PA-A1 was much smaller than that of the PAO1 strain (Figure 6A). We also tested the effect of mutation of acpP2 or acpP3 on twitching motility. The data showed that neither single mutation of acpP2 or acpP3 nor double mutation of acpP2 and acpP3 affected P. aeruginosa twitching motility (Figures 6A,B).

Next, we investigated the swarming motility of mutant strain PA-A1. Replacement of acpP1 by E. coli acpP caused P. aeruginosa swarming to fail. On semisolid plates (containing 0.5% agar), wild type strain PAO1 formed a typical swarming pattern, while mutant strain PA-A1 only formed a small circular lawn (Figure 7A). However, when complemented with pSRK-Pa, the ability of PA-A1 to form a typical swarming pattern was restored (Figure 7B). We also noticed that mutant strains PA-A2, PA-A3, and PA-A23 could form a typical swarming pattern, indicating that mutation of acpP2 or acpP3 did not affect P. aeruginosa swarming motility (Figures 7A,B). Because replacement of acpP1 by E. coli acpP causes P. aeruginosa to reduce the production of QS signal molecules, we wondered if exogenous addition of AHLs signals could restore the ability of mutant strain PA-A1 to swarm on semisolid plates. To test this hypothesis, we added 2 μM C₄-HSL or 3-oxo-C₁₂-HSL to semisolid plates. After 24 h of incubation, on both types of semisolid plate, the swarming motility of mutant strain PA-A1 was restored (Figure 7C), even though the swarming patterns were still smaller than that of the wild type strain PAO1. This indicated that reduction of the AHLs signal was the main factor that caused P. aeruginosa to fail to swarm.