Ralstonia solanacearum RSp0194 Encodes a Novel 3-Keto-Acyl Carrier Protein Synthase III

Fatty acid synthesis (FAS), a primary metabolic pathway, is essential for survival of bacteria. Ralstonia solanacearum, a β-proteobacteria member, causes a bacterial wilt affecting more than 200 plant species, including many economically important plants. However, thus far, the fatty acid biosynthesis pathway of R. solanacearum has not been well studied. In this study, we characterized two forms of 3-keto-ACP synthase III, RsFabH and RsFabW, in R. solanacearum. RsFabH, the homologue of Escherichia coli FabH, encoded by the chromosomal RSc1050 gene, catalyzes the condensation of acetyl-CoA with malonyl-ACP in the initiation steps of fatty acid biosynthesis in vitro. The RsfabH mutant lost de novo fatty acid synthetic ability, and grows in medium containing free fatty acids. RsFabW, a homologue of Pseudomonas aeruginosa PA3286, encoded by a megaplasmid gene, RSp0194, condenses acyl-CoA (C2-CoA to C10-CoA) with malonyl-ACP to produce 3-keto-acyl-ACP in vitro. Although the RsfabW mutant was viable, RsfabW was responsible for RsfabH mutant growth on medium containing free fatty acids. Our results also showed that RsFabW could condense acyl-ACP (C4-ACP to C8-ACP) with malonyl-ACP, to produce 3-keto-acyl-ACP in vitro, which implies that RsFabW plays a special role in fatty acid synthesis of R. solanacearum. All of these data confirm that R. solanacearum not only utilizes acetyl-CoA, but also, utilizes medium-chain acyl-CoAs or acyl-ACPs as primers to initiate fatty acid synthesis.


Introduction
Fatty acid synthesis (FAS) is a primary bacterial metabolic pathway. Fatty acids are not only bacterial cell components (membrane phospholipids, lipoproteins, and lipoglycans) [1,2,3,4], but also intermediates, used to synthesize other end products, such as cofactors (lipoate [5] and biotin [6,7]), quorum sensing [QS] signal molecules [8], and energy storage components [9]. In most bacteria, fatty acid synthase (FAS II) is composed of a series of small soluble proteins, and each enzyme, encoded by a separate gene, carries out a different catalytic step in the pathway [1,2,4]. In the fatty acid biosynthetic pathway of bacteria, 3-ketoacyl-ACP synthase (KAS) condenses acyl-acyl carrier protein (acyl-ACP) or acyl coenzyme A (acyl-CoA) with malonyl-ACP to produce 3-ketoacyl-ACP. Most bacteria contain at least two classes of KAS enzymes. KAS III class enzymes, which have a Cys-His-Asn active site triad, are responsible for the initiation of the fatty acid chains (Fig 1A), whereas the KAS I/II classes of condensing enzymes have a Cys-His-His catalytic triad, and catalyze synthesis of long acyl chains in the elongation steps [4,10,11]. In general, 3-ketoacyl-ACP synthase III (FabH) has two roles in bacterial fatty acid biosynthesis. First, as the initiator of elongation, this enzyme determines the amount of fatty acids that will be produced by the pathway, and is therefore subject to stringent feedback regulation by long-chain acyl-ACPs [12]. Second, the substrate specificity of this enzyme is a major determining factor for membrane fatty acid composition of bacteria [10,13]. In Escherichia coli, FabH prefers to use acetyl-CoA as a primer, leading to the only production of straightchain fatty acids in this organism [12,14]. In Listeria monocytogenes and many gram-positive bacteria, the FabH proteins are highly selective for branched-chain acyl-CoA as substrates, and primarily isoand anteiso-branched chain fatty acids are produced in these bacteria [15,16]. In Mycobacterium tuberculosis, the type II fatty acid biosynthesis pathway is employed to produce very long-chain mycolic acids, and accordingly the MtFabH enzyme prefers long-chain acyl-CoA substrates [10]. Recently, Yuan et al. identified two novel types of 3-ketoacyl-ACP synthase III in Pseudomonas aeruginosa, FabY and PA3286 [9,17]. FabY, defined as a novel class of 3-ketoacyl synthase KASI/II domain condensation enzymes, possesses a typical Cys-His-His catalytic triad and catalyzes the condensation of acetyl-CoA with malonyl-ACP in initiation steps of fatty acid biosynthesis of P. aeruginosa [17]. PA3286, which belongs to the 3-ketoacyl-acyl carrier protein synthase III family, condenses malonyl-ACP with octanoyl-  CoA shunting from the fatty acid β-oxidation degradation cycle, to produce 3-ketodecanoyl-ACP, which is a key intermediate common for synthesis of saturated fatty acids (SFA), unsaturated fatty acids (UFA) and lipopolysaccharides (LPS) [9].
Ralstonia solanacearum, a soil-borne, destructive plant pathogen, causes a bacterial wilt that affects many economically important plants such as tomato (Lycopersicon esculentum), potato (Solanum tuberosum), tobacco (Nicotiana tabacum), and banana (Musa acuminata) [18,19]. R. solanacearum is known to utilize a complex regulatory pathway to activate production of the major virulence factors. Two signal transition pathways using hrp and phc, have been reported [20]. In the phc pathway, a LysR-type transcriptional regulator, PhcA, plays a central role in response to cell density by using a quorum-sensing mechanism that involves the specific signal molecule, 3-hydroxypalmitic acid methyl ester (3-OH-PAME). It has been demonstrated that 3-OH-PAME is synthesized by PhcB, which is thought to convert 3-hydroxypalmitoyl-ACP to 3-hydroxypalmitoyl methyl ester [21]. The 3-hydroxypalmitoyl-ACP comes from the bacterial fatty acid synthesis pathway. This indicates that novel antimicrobials targeting the enzymes in the R. solanacearum fatty acid synthesis pathway would interfere with quorum sensing of R. solanacearum and this could be helpful to control the bacterial wilt caused by R. solanacearum.
Although fatty acid analyses of total membranes of R. solanacearum demonstrated typical fatty acids found in many proteobacteria, primarily palmitic (C 16:0 ), palmitoleic (C 16:1 ), and cis-vaccenic (C 18:1 ) acids [21,22], little is known about the R. solanacearum fatty acid synthesis pathway. Cheng et al. have characterized five R. solanacearum putative 3-ketoacyl-ACP synthase homologues (RSc1054, RSp0358, RSp0357, RSp0361, and RSc0427) in R. solanacearum, and found that only RSc1054 possesses long-chain 3-ketoacyl-ACP synthase activity and is required for fatty acid synthesis [23]. Recently, our laboratory has studied two 3-ketoacyl-ACP reductase homologues (RSc1052 and RSp0359) in R. solanacearum and the data suggested both enzymes are active in fatty acid synthesis but function differently in determining the profile of cellular fatty acids (unpublished data).
In this study, we identified two forms of 3-ketoacyl-ACP synthase III, FabH and RSp0194, in R. solanacearum. Results have shown that RsFabH, similar to E. coli FabH, catalyzes the condensation of acetyl-CoA with malonyl-ACP in the initiation steps of fatty acid biosynthesis, and RSp0194, like P. aeruginosa PA3286, prefers to condense medium-chain acyl-CoAs (such as octanoyl-CoA or decanoyl-CoA) substrates with malonyl-ACP. However, it is also able to condense short-chain acyl-ACPs (such as butanoyl-ACP, hexanoyl-ACP, and octanoyl-ACP) with malonyl-ACP in elongation steps to produce decanoyl-ACP. This implies RSp0194 plays a unique role in fatty acid synthesis of R. solanacearum.
To create an unmarked deletion mutant of RsfabH, the 500 base pair (bp) DNA fragments flanking up or down the fabH gene were amplified with Pfu DNA polymerase using R. solanacearum genomic DNA as the template, and either RsfabH Up EcoRI and RsfabH Up XbaI (for Up fabH), or RsfabH Dn XbaI and RsfabH Dn HindIII (for Down fabH) as primers (S2 Table), respectively. The products of these PCR reactions were purified and overlapping PCR was carried out using RsfabH Up EcoRI and RsfabH Dn HindIII as the primers. The resulting 1,000 bp DNA fragment was digested with EcoRI and HindIII and inserted between the same sites of pK18mobscaB [25] to yield pYH10. Using the same methods, an RsfabW deletion plasmid pYH11 was also constructed.
Deletion of the RsfabH and RsfabW genes E. coli strain S17-1 carrying plasmid pYH10 or pYH11 was conjugated with R. solanacearum GMI1000 on BG plates for 24 h at 30°C. After appropriate dilutions, the cultures were spread onto BG plates containing chloramphenicol plus kanamycin to select for integration of the plasmid into the genome of GMI1000. Several colonies were inoculated into BG medium (for selection of the ΔfabW strain) or BG medium containing 0.01% octanoic acid (for selection of the ΔfabH strain), at 30°C for 48 h. After appropriate dilutions, the cultures were spread onto BG plates containing 10% sucrose and octanoic acid. Colony PCR utilizing the primers listed in S2 Table was carried out to screen colonies sensitive to kanamycin. We successfully obtained RsfabH deletion strain RsmH and RsfabW deletion strain RsmW. To obtain the ΔfabH ΔfabW double deletion strain, the Vibrio harveyi AasS expression plasmid pYH5 (pSRK-aasS) was first introduced into strain RsmH (ΔfabH). Then, using the same method as described above, the double deletion mutant ΔfabH ΔfabW/ pSRK-aasS strain, RsmD, was obtained.

Assays of de novo fatty acid synthesis
To examine the function of fabH in R. solanacearum fatty acid synthesis, de novo fatty acid synthesis of RsmH (ΔfabH) strain was measured by sodium [1-14 C] acetate incorporation into membrane phospholipids. The RsmH (ΔfabH) strain was grown at 30°C in BG supplemented with 0.01% octanoic acid for 36 h. The cells were harvested and washed three times with fresh BG to remove unincorporated fatty acids. The cells were transferred into 10 mL BG medium, and then 5 μL of sodium [1-14 C] acetate was added. The culture was incubated for an additional 4 h. Labeled fatty acids were extracted, analyzed by thin layer chromatography (TLC), and quantitated by phosphorimaging [26].

Cell-free extract preparation
Cell extracts of R. solanacearum strains were prepared from exponentially growing cells. Cell cultures were harvested by centrifugation and then re-suspended in lysis buffer (0.1 M sodium phosphate, pH 7.5, 5 mM β-mercaptoethanol, and 1 mM EDTA). Cell lysates were prepared by passing the cultures three times through a French pressure cell, then they were centrifuged for 1 h at 260,000 × g, and the supernatants were dialyzed against lysis buffer for 24 h and saved as cell extracts.

R. solanacearum genes encode two different 3-ketoacyl-acyl carrier protein synthase III homologues
To identify putative 3-ketoacyl-ACP synthase III encoding genes in R. solanacearum, the sequences of FabH from E. coli, and FabY and PA3286 [9,17] from P. aeruginosa were used as query sequences for a BLAST analysis of the R. solanacearum GMI1000 genome [19]. The chromosomal RSc1050 gene (named RsfabH in this study) is annotated as encoding an E. coli FabH homologue. The RSp0194 gene (named RsfabW in this study), located on a megaplasmid, encodes a strong homologue of P. aeruginosa PA3286. Alignment of R. solanacearum FabH and E. coli FabH showed that RsFabH is 53.4% identical to E. coli FabH (Fig 1C). Moreover, RsfabH is located within a fatty acid synthesis gene cluster (plsX, fabH, fabD, fabG1, acpP, and fabF) (Fig 1B) and it has been shown that the gene cluster encoded FabF proteins have roles in R. solanacearum fatty acid biosynthesis [23]. RsFabW is 60.9% identical to PA3286, and more importantly, contains the Cys-His-Asn catalytic active triad like the 3-ketoacyl-ACP synthase III family of proteins ( Fig 1D). These alignments suggest that the R. solanacearum genome encodes at least two different 3-ketoacyl-ACP synthase III homologues.

Properties of R. solanacearum ΔfabH and ΔfabW deletion mutant strains
To identify the physiological functions of the two KASIIIs in R. solanacearum fatty acid biosynthesis, knockout strains for deletion of the RsfabH or RsfabW genes were made by allelic replacement. Two suicide plasmids, pYH10 and pYH11, were constructed, and then introduced into strain GMI1000 by conjugation (see Materials and Methods). The conjugants were selected on BG medium containing kanamycin. Cultures from conjugants were then plated onto BG medium containing sucrose in order to select for the loss of the suicide plasmid sequences from the GMI1000 genome (for selection of the RsfabH deletion strain, BG medium also contained octanoic acid). The colonies that were sensitive to kanamycin and resistant to sucrose were assayed by PCR analysis using the primers listed in S2 Table (data not shown). Two deletion mutant strains were obtained: RsmH (ΔfabH) and RsmW (ΔfabW).
We first tested the growth of the RsmH and RsmW strains on BG medium. Strain RsmH failed to grow on BG medium except in the presence of octanoic acid, whereas strain RsmW grew on BG medium in the absence of octanoic acid, as did the wild type strain GMI1000 (Fig  2A). When either E. coli fabH or R. solanacearum fabH was expressed from plasmid pSRK-Gm in strain RsmH, growth on BG medium in the absence of octanoic acid was restored (Fig 2A). EcfabH genes encode 3-ketoacyl-ACP synthase III, the enzyme required for initiation of fatty acid biosynthesis in bacteria [10]. In order to confirm the function of fabH in R. solanacearum fatty acid synthesis, the ability of fatty acid synthesis in strain RsmH was determined by measuring sodium [1-14 C] acetate incorporation into membrane phospholipids as described in the Materials and Methods section. The RsmH strain synthesized only traces of fatty acids when grown in medium lacking octanoic acid (Fig 2B, lane 2). Complementation via episomal expression of E. coli fabH or R. solanacearum fabH restored fatty acid biosynthesis in RsmH strain (Fig 2B, lanes 3 and 4). These data confirmed that deletion of fabH caused octanoic acid auxotrophy in R. solanacearum, and the RsfabH-encoded protein had 3-ketoacyl-ACP synthase III activity.
We also tested if other fatty acid species restored RsmH growth on BG medium. Data showed that short-chain fatty acids, such as butanoic (C 4 ) or hexanoic (C 6 ) acid, failed to restore growth of RsmH, while the long-chain fatty acids, tetradecanoic (C 14 ), hexadecanoic (C 16 ), or octadecanoic (C 18 ) acids supported RsmH growth, but it was much weaker than the growth on BG medium supplied with octanoic (C 8 ), decanoic (C 10 ), or dodecanoic (C 12 ) acids (Fig 3A and 3B). These results suggested that R. solanacearum possesses another enzyme(s) that allows exogenous fatty acids to enter the fatty acid synthesis pathway, and has a strong preference for fatty acids of medium chain length.
The RsmW strain grew well on BG medium in the absence of octanoic acid (Fig 2A). To determine if RsfabW is responsible for growth of RsmH on BG medium in the presence of fatty acids, double deletion of the RsfabH and RsfabW strains was constructed. Because strains lacking both RsFabH and RsFabW activities might be nonviable, we first introduced plasmid pYH5 that encodes Vibrio harveyi acyl-ACP synthetase (AasS) from plasmid pSRK-Gm into RsmH and then deleted fabW by allelic replacement. Double deletion mutant RsmHW [ΔfabH ΔfabW/pYH5 (pSRK-aasS)] was obtained by selection on BG medium containing octanoic acid and IPTG. The double deletion strain RsmD was unable to grow on BG medium in the absence of octanoic acid (Fig 2A). Without IPTG induction, RsmHW was unable to grow on BG medium containing any species of fatty acids (Fig 3A and 3B). However, under the induction by IPTG, RsmD grew on BG medium containing C 4 , C 6 , C 8 , or C 10 fatty acid, but failed when the supplement was long chain fatty acids, such as C 12 , C 14 , C 16 , or C 18 fatty acids ( Fig  3A and 3B). The lack of growth on BG medium containing fatty acids in the absence of IPTG induction indicated that double mutant RsmHW lost both RsFabH and RsFabW activities and failed to synthesize fatty acids, whereas growth on BG medium containing short or medium chain fatty acid under IPTG induction indicated that the presence of acyl-ACP synthetase (AasS) activity, and RsmHW allows exogenous fatty acids to enter R. solanacearum and restore fatty acid synthesis. These results confirmed that fabW is responsible for growth of strain RsmH on BG medium in the presence of fatty acids, and confirmed that RsfabW encodes a 3-ketoacyl-ACP synthase III similar to P. aeruginosa PA3286 that is able to condense malonyl-ACP with medium-chain acyl-CoAs to initiate fatty acid biosynthesis.

Expression and purification of the R. solanacearum RsFabH and RsFabW
In order to allow direct in vitro assays of RsFabH and RsFabW KAS activities, both proteins were expressed in E. coli and the N-terminally His 6 -tagged versions of these two proteins were successfully purified by nickel-chelate chromatography (S1A Fig). The purified RsFabH and RsFabW proteins had monomeric molecular masses of 36 kDa and 42 kDa on SDS-PAGE, respectively, which were in agreement with the values calculated from the sequences of the tagged proteins (36.6 and 42.4 kDa, respectively). We also purified the E. coli fatty acid biosynthetic proteins FabD, FabG, FabZ, FabI, and holo-ACP, plus the V. harveyi acyl-ACP synthetase (AasS) (see Materials and Methods).

In vitro enzymatic activities of RsFabH and RsFabW
To determine the function of RsFabH, the initiation reaction of the fatty acid synthesis were reconstituted by the addition of EcFabD, 3-ketoacyl-ACP synthase III (EcFabH or RsFabH), EcFabG, EcFabA, and EcFabI, and the products was analyzed by conformationally sensitive gel electrophoresis. In the absence of KAS, only holo-ACP was observed during electrophoresis. Addition of EcFabH or RsFabH to the reaction mixture led to formation of butyryl-ACP (S1B Fig, lane 1 and lane 2). Upon addition of the long chain E. coli 3-ketoacyl-ACP synthase I, FabB, to the reactions, all reactions produced long chain acyl-ACP species (S1B Fig, lane 3 and  lane 4). These data clearly showed that, like E. coli FabH, RsFabH could complete the initial cycle of fatty acid synthesis to produce butyryl-ACP.
To confirm the functions of RsFabW, which condenses malonyl-ACP with medium-chain acyl-CoAs to produce 3-keto-acyl-ACPs, we tested condensation of malonyl-ACP with octanoyl-CoA. First, malonyl-ACP was synthesized from holo-ACP and malonyl-CoA using E. coli FabD. Then, the reactions were reconstituted by the sequential addition of purified proteins and co-factors as described in the Materials and Methods section. After incubation of E. coli FabG, FabA, and FabI with malonyl-ACP, octanoyl-CoA, NADPH, and NADH, the reaction mixture did not form new products, and incubation of RsFabW with only malonyl-ACP, octanoyl-CoA, NADPH, and NADH, also failed to produce any new species. However, when RsFabW was added to the mixture of E. coli FabG, FabA, and FabI with malonyl-ACP, octanoyl-CoA, NADPH, and NADH, then decanoyl-ACP was formed (Fig 4A). Using the same approach, decanoyl-CoA was tested and the result showed that, like octanoyl-CoA, decanoyl-CoA was also one of the substrates necessary for RsFabW to convert dodecanoyl-ACP ( Fig  4A). We also tested butyryl-CoA and hexanoyl-CoA. Although both butyryl-CoA and hexanoyl-CoA were substrates of RsFabW, the product of the reaction was neither hexanoyl-ACP nor octanoyl-ACP. It was surprising that butyryl-CoA and hexanoyl-CoA were converted to decanoyl-ACP by RsFabW (Fig 4A). This suggested that RsFabW was able to condense acyl-ACPs and malonyl-ACP to produce longer acyl-ACPs. To confirm this possibility, we replaced octanoyl-CoA with octanoyl-ACP to perform the same reaction. Indeed, the combination of RsFabW, E. coli FabG, FabA, and FabI, in the reaction mixture produced decanoyl-ACP ( Fig  4B), and hexanoyl-ACP gave the same product, decanoyl-ACP (Fig 4B). However, when decanoyl-ACP was tested, no new products appeared (Fig 4B). Elongation of butyryl-ACP by RsFabW was also tested. Because AasS does not use butyric acid, butyryl-ACP was produced using the initial reaction of the fatty acid synthesis reconstituted by the sequential addition of E. coli FabD, FabH, FabA, FabG, and FabI, and malonyl-CoA, holo-ACP, NADPH, and NADH ( Fig 4C, lane 2). Upon addition of RsFabW, the reaction mixture produced hexanoyl-ACP and decanoyl-ACP (Fig 4C, lane 5). Thus, these data indicate that RsFabW is able to use butyryl-ACP, hexanoyl-ACP, and octanoyl-ACP as substrates and to elongate them to decanoyl-ACP.
The above ACP thioester intermediates in the reconstitution assays were also analyzed by MALDI-TOF-MS to confirm the identities of the products. We used hexanoyl-CoA as substrate and purified the products of the reaction mixture before measuring their mass by MAL-DI-TOF-MS. As a control (Fig 5A), the mass peak for the reaction not containing RsFabW occurred at 8980, corresponding to the holo-ACP molecules. For RsFabW-containing reaction mixtures, two new peaks appeared. The minor mass peak was at 9107, corresponding to octanoyl-ACP (holo-ACP + 128), and the major peak was at 9135, corresponding to decanoyl-ACP (holo-ACP + 155) (Fig 5B). These data confirm the identities determined by gel electrophoresis as indicated in Fig 4A and 4B, and substantiate the conclusion that RsFabW was able to condense either medium-chain acyl-CoAs or medium-chain acyl-ACPs with malonyl-ACP.
We next determined if RsFabW could condense acetyl-CoA with malonyl-ACP to initiate fatty acid synthesis. The reaction mixture contained E. coli FabD, FabA, FabG, and FabI, malonyl-CoA, acetyl-CoA, holo-ACP, NADPH, and NADH. When adding RsFabW to the mixture and incubating for 1 hour, the product was assayed by conformationally sensitive gel electrophoresis, showing that butyryl-ACP, hexanoyl-ACP, and decanoyl-ACP were formed (Fig 4C,  lane 3). However, when adding R. solanacearum FabH to the mixture, only butyryl-ACP was produced (Fig 4C, lane 2). In conclusion, in vitro, RsFabW is able to use acetyl-CoA to initiate fatty acid synthesis and then produce decanoyl-ACP.
We also compared the 3-ketoacyl-ACP synthase III activities of cell-free extracts of wild type strain GMI1000 and the ΔfabH (RsmH) or ΔfabW (RsmW) deletion mutants (S2A Fig). The cell-free extracts of the wild type strain GMI1000 could use acyl-CoAs (C 2 -CoA to C 10 -CoA) as substrates to synthesize long-chain acyl-ACP. In the ΔfabW mutant extract, only acetyl-CoA was elongated to long-chain acyl-ACPs, whereas the ΔfabH mutant extract elongated acyl-CoAs (C 4 -CoA to C 10 -CoA) to long-chain acyl-ACP species. The double mutant RsmD extract failed to elongate any of the acyl-CoAs. The substrate specificity of RsFabW In the above in vitro experiments, it has been demonstrated that RsFabW is able to use acyl-CoAs (C 2 to C 10 ) as primers to initiate fatty acid synthesis. In order to probe the substrate specificity of RsFabW, the decrease in absorbance at 340 nm of NADPH was monitored spectrophotometrically in reaction mixtures containing holo-ACP, malonyl-CoA, NADPH, EcFabD, EcFabG, RsFabW, and various straight-chain saturated acyl-CoAs (C 2 to C 10 ) ( Table 1). The data showed that although the conformationally sensitive gel assay indicated that RsFabW could elongate acetyl-CoA to decanoyl-CoA, RsFabW very weakly accepted acetyl-CoA as the primer; the activity for acetyl-CoA was 16.93 ± 0.83 μmol/m × μg. The best substrate was clearly octanoyl-CoA, and the activity of RsFabW for octanoyl-CoA was 195.77 ± 14.55 μmol/s × μg. For butyryl-CoA, hexanoyl-CoA, and decanoyl-CoA, the activities of RsFabW were 19.05 ± 1.61, 77.38 ± 6.06, and 97.88 ± 5.4 μmol/s×μg, respectively, and lower than that for C 8 -CoA, confirming the observations that the shorter exogenous fatty acids, such as butyric and hexanoic acid, did not rescue growth in ΔfabH strain.
To confirm that RsFabW uses butyryl-CoA and hexanoyl-CoA as substrates in vivo, we introduced plasmid pYH6 (pSRK-RsfabW) into the RsmH strain to construct the RsfabW overexpression strain RsmH/pYH6 (pSRK-RsfabW), and tested the growth of this strain on BG plates in the presence of acetic, butanoic, or hexanoic acid. Under IPTG induction strain

Discussion
In general, the 3-ketoacyl-ACP synthase III class of enzymes condenses acyl-CoAs with malonyl-ACP to initiate fatty acid biosynthesis in the dissociated, type II fatty acid synthase systems typified by bacteria [1,3,10,11]. According to the substrate specificity of enzymes, the 3-ketoacyl-ACP synthase III class of enzymes can be divided into three groups. The first group of these enzymes that is highly selective for acetyl-CoA as primers includes E. coli FabH [33] and P. aeruginosa FabY [17]. The second group of these enzymes includes many gram-positive bacteria FabHs, such as L. monocytogenes [15] and B. subtilis [34], which prefer branched-chain acyl-CoAs from branched-chain amino acid metabolism as substrates. The third group contains M. tuberculosis FabH [10], and P. aeruginosa PA3286 [9], which uses medium-chain acyl-CoAs as substrates and shunt intermediates from the type I fatty acid synthesis pathway (type I FAS) or the fatty acid β-oxidation degradation cycle into the type II fatty acid synthesis pathway. The R. solanacearum genome encodes two forms of 3-ketoacyl-ACP synthase III [19]. RsFabH catalyzes the condensation of acetyl-CoA with malonyl-ACP in initiation steps of fatty acid biosynthesis, and the RsmH mutant strain has lost the ability for de novo fatty acid synthesis. RsFabW can condense either acyl-CoAs (C 2 to C 10 -CoAs) or acyl-ACPs (C 4 to C 8 -ACPs) with malonyl-ACP to produce 3-keto-acyl-ACP in vitro. Moreover, although the RsmW mutant was viable, RsfabW was responsible for RsfabH mutant growth on medium containing medium-chain fatty acids, such as octanoic, decanoic, and dodecanoic acid. These results confirmed that R. solanacearum not only utilizes acetyl-CoA, but also medium-chain acyl-CoAs or acyl-ACPs as primers to initiate fatty acid synthesis. The growth of RsfabH mutants on medium containing long-chain fatty acids (C 14 to C 18 ) was weaker than on medium in the presence of medium-chain fatty acids (C 8 to C 12 ) (Fig 3). This was in agreement with the substrate specificity of RsFabW for C 8 -CoA and C 10 -CoA in vitro (Table 1). This indicated that exogenous long-chain fatty acids (C 14 to C 18 ) should be degraded to C 8 -CoA by β-oxidation, and then RsFabW would shunt C 8 -CoA into the fatty acid synthesis pathway to make LPS, UFA, and SFA required for bacteria growth. However, although FabW possessed strong sequence similarity to P. aeruginosa PA3286 (Fig 1D), FabW was different from P. aeruginosa PA3286 [9]. First, FabW not only uses acyl-CoAs (such as octanoyl-CoA or decanoyl-CoA) as primers, but also condenses short-chain acyl-ACPs (such as butanoyl-ACP, hexanoyl-ACP, and octanoyl-ACP) with malonyl-ACP. Second, FabW seems to have a unique role in fatty acid synthesis in R. solanacearum, but does not only shunt intermediates from β-oxidation degradation into fatty acid biosynthesis. R. solanacearum mainly invades xylem vessels of host plants [35], where this is a lack of sufficient phospholipids or free fatty acids to support this bacterial growth. RsFabW is also distinct from long-chain 3-ketoacyl-ACP synthase I/II, which can catalyze the condensation of long-chain acyl-ACP with malonyl-ACP in the elongation cycle of bacterial fatty acid synthesis [10]. RsFabW did not utilize decanoyl-ACP or longer acyl-ACPs as substrates. It has been demonstrated that R. solanacearum only has one long chain 3-ketoacyl-acyl carrier protein synthase, RsFabF1, which possesses both the activity of 3-ketoacyl-ACP synthase II and I [23]. Therefore, it is possible that RsFabW helps RsFabF1 to function in elongation reactions of medium-chain fatty acid synthesis. On the basis of our results, we suggest a model for fatty acid biosynthesis in R. solanacearum in Fig 6. Although V. harveyi aasS was able to bypass both the E. coli fabH mutation [29] and R. solanacearum ΔfabH ΔfabW double mutation when medium was supplemented with exogenous short-chain fatty acids, such as hexanoic or octanoic acid, the R. solanacearum ΔfabH ΔfabW double mutant is different from the E. coli fabH mutant. Decanoic acid did not support E. coli fabH carried the V. harveyi aasS-encoded plasmid for growth [29], but actually did maintain the ΔfabH ΔfabW double mutant that harbored aasS encoded plasmid for growth, though the growth was weaker than that observed with supplementation with hexanoic or octanoic acid. The growth of the E. coli fabH mutant carrying aasS-encoding plasmid was not supported by decanoic acid and this was because decanoyl-ACP only entered the fatty acid synthesis pathway beyond the point where unsaturated fatty acid synthesis branches from the common (saturated) pathway [29]. However, to date it is still not clear how R. solanacearum synthesizes unsaturated fatty acids. One possibility is that R. solanacearum converted decanoyl-ACP to cis-3-decenoyl-ACP directly, which is the key intermediate for unsaturated fatty acids synthesis. BLAST analysis of the R. solanacearum GMI1000 genome showed that RSc 0114 is 55% identical to Neisseria gonorrhoeae UfaA, which has been demonstrated to function in anaerobic unsaturated fatty acid synthesis [36]. Although R. solanacearum is an aerobic bacterium, whether or not R. solanacearum RSc0114 could convert decanoyl-ACP to cis-3-decenoyl-ACP directly needs to be further addressed.