|Home | About | Journals | Submit | Contact Us | Français|
Neisseria is a Gram-negative pathogen with phospholipids composed of straight chain saturated and monounsaturated fatty acids, the ability to incorporate exogenous fatty acids, and lipopolysaccharides that are not essential. The FabI inhibitor, AFN-1252, was deployed as a chemical biology tool to determine whether Neisseria can bypass the inhibition of fatty acid synthesis by incorporating exogenous fatty acids. Neisseria encodes a functional FabI that was potently inhibited by AFN-1252. AFN-1252 caused a dose-dependent inhibition of fatty acid synthesis in growing Neisseria, a delayed inhibition of growth phenotype, and minimal inhibition of DNA, RNA, and protein synthesis, showing that its mode of action is through inhibiting fatty acid synthesis. Isotopic fatty acid labeling experiments showed that Neisseria encodes the ability to incorporate exogenous fatty acids into its phospholipids by an acyl-acyl carrier protein-dependent pathway. However, AFN-1252 remained an effective antibacterial when Neisseria were supplemented with exogenous fatty acids. These results demonstrate that extracellular fatty acids are activated by an acyl-acyl carrier protein synthetase (AasN) and validate type II fatty acid synthesis (FabI) as a therapeutic target against Neisseria.
Neisseria is a genus of mucosal colonizing bacteria found in a variety of animals. Of the 11 species of Neisseria associated with humans, Neisseria gonorrhoeae and Neisseria meningitidis are obligate human pathogens. N. meningitidis is a common cause of meningitis and other meningococcal diseases in infants and children with 3,000 cases/year in the United States (1). N. gonorrhoeae is the causative agent of the sexually transmitted disease gonorrhoeae, with >100 million new cases/year worldwide (2). The incidence of multidrug-resistant N. gonorrhoeae has markedly increased in recent years, raising the prospect of untreatable gonorrhoeae and highlighting the need for developing new antibiotics and discovering new antibiotic targets in Neisseria (2). One emerging antibiotic drug target is bacterial type II fatty acid synthesis (FASII)3 (3,–6). All characterized bacteria have the ability to utilize exogenous fatty acids for phospholipid synthesis (7); thus, it is important to understand the pathways for fatty acid uptake and whether extracellular fatty acids can bypass FASII inhibition (7,–9).
Neisseria are Gram-negative bacteria with phospholipids containing saturated and unsaturated fatty acids (10) and LPS containing 3-hydroxy fatty acids (11). Escherichia coli utilizes an acyl-CoA synthetase (FadD) to activate exogenous fatty acids (12), and they are incorporated into phospholipids by the PlsB/PlsC acyltransferase system that utilizes either acyl-ACP or acyl-CoA (7, 13). E. coli cannot bypass FASII inhibitors because there is no mechanism for the generation of acyl-ACP from extracellular fatty acids, and FASII is the only source for the 3-hydroxy fatty acids required to synthesize the essential LPS (14, 15). The fatty acid composition of Neisseria resembles E. coli (7); however, LPS is not essential in Neisseria. N. meningitides with deficient LPS synthesis are viable under laboratory conditions (16, 17), and an LPS-deficient N. meningitidis strain has been isolated from a human patient (18). Neisseria belongs to the β-Proteobacteria class, and phosphatidic acid synthesis occurs by the more widely distributed PlsX/PlsY/PlsC acyltransferase system (19). In this pathway, the acyl-ACP end-products of FASII are converted to acyl-PO4 by PlsX, and the acyl-PO4 serves as the acyl donor for PlsY to acylate sn-glycerol-3-phosphate (19). Gram-positive bacteria, such as S. aureus, activate exogenous fatty acids using a fatty acid kinase to produce acyl-PO4 that is either used by PlsY or converted to acyl-ACP by PlsX (20, 21). Neisseria does not contain a fatty acid kinase homolog. Exogenous unsaturated fatty acids are incorporated into phospholipid (22), but the activation pathway is unknown. There is insufficient information available to determine whether Nesseria could potentially bypass FASII inhibitors by obtaining fatty acids from the host.
The goal of this work is to characterize the pathway for incorporation of extracellular fatty acids in Neisseria and determine whether exogenous fatty acids can overcome the inhibition of FASII. Labeling and complementation experiments show that Neisseria incorporates exogenous fatty acids following their activation by an acyl-ACP synthetase. FASII inhibitors with the most complete track record of development are inhibitors of enoyl-ACP reductase (FabI) (23,–26). FabI catalyzes the reduction of trans-2-enoyl-ACP, a rate-controlling step required to complete each round of elongation in fatty acid synthesis (27). AFN-1252 (23, 28,–33) is a FabI therapeutic that has completed Phase II clinical trials and proved efficacious against Staphylococcus aureus infections in humans. We selected AFN-1252 as a chemical probe to test whether exogenous fatty acids can bypass FabI (FASII) inhibition. Inhibition of Neisseria FabI by AFN-1252 cannot be bypassed by exogenous fatty acid supplementation. This work establishes an acyl-ACP synthetase pathway for fatty acid uptake in Neisseria and validates FabI inhibitors as a viable strategy for the development of Neisseria therapeutics.
Cell culture supplies were from BD Biosciences, and chemicals were from Sigma-Aldrich or Fisher unless otherwise indicated. Radiochemicals were from American Radiolabeled Chemicals.
Briefly, chocolate blood agar plates were made by suspending and autoclaving 7.2 g of Difco GC Medium base in 100 ml of water. To this solution was added 100 ml of sterile 2% hemoglobin solution (BBL hemoglobin, bovine, freeze-dried) and 2 ml of IsoVitaleX supplement.
The Shockley-Johnston-modified gonococcal (SJ-GC) mediumwas used for the planktonic growth of N. gonorrhoeae. The SJ-GC medium was made by combining three separate solutions. Solution 1 contained 15 g of proteose peptone 3, 4 g of K2HPO4, 1 g of KH2PO4, 5 g of NaCl, 5 g of casamino acids, and 1 g of soluble potato starch in 1 liter of distilled water. Solution 2 contained 22 g of dextrose and 3 g of MgSO4 in 50 ml of distilled water. Solution 3 contained 0.0001 g of thiamine pyrophosphate, 0.05 g of glutamine, and 0.2 g of cysteine hydrochloride. Solution 1 was autoclaved for 15 min at 120 °C, whereas solutions 2 and 3 were filter-sterilized. The three solutions were combined once they reached room temperature to make the SJ-GC medium. The SJ-GC medium contained potato starch, a known drug binder, so the chemically defined Graver-Wade (GW) medium was used for MIC determination for both N. gonorrhoeae and N. meningitidis. The GW medium was made as described (34). Bovine serum albumin (10 mg/ml) was added to the GW medium to increase the fatty acid binding capacity of the medium. N. gonorrhoeae strain NCTC 12700 and N. meningitidis strain ATCC 13077 were streaked from 10% glycerol stock onto chocolate blood agar plates and incubated under 10% CO2 and at 37 °C. Both species were grown in liquid culture by resuspending colonies from chocolate blood agar plates in the appropriate medium and grown at 37 °C with gentle shaking (125 rpm).
The MICs were determined by growing strains to an A600 of 0.4 before being diluted 200-fold into GW medium. A 100-μl aliquot of diluted cells was added to each well of a U-bottom 96-well plate containing 100 μl of GW medium with the indicated concentration of compound. The plate was incubated in a Star-Pac gas-permeable bag (Garner U.S. Enterprises) at 37 °C and 10% CO2 atmosphere for 24 h, and the A600 was determined. Data were derived from two biological replicates.
The fabI (NGO1666) and the two predicted acyl-CoA/ACP synthetase (NGO0530 and NGO1213) genes of N. gonorrhoeae strain FA 1090 (NCBI Microbial Genomes Database) were optimized for expression in E. coli through GeneArt Gene Synthesis Technology (Life Technologies, Inc.). A NdeI restriction site was engineered at the 5′-end of the gene with the start codon in the NdeI site, whereas a His6 tag, stop codon, and an EcoRI restriction site were sequentially engineered at the 3′-end of the gene. The genes were cloned into the pPJ131 plasmid (a modified version of the pBluescript plasmid with the multiple cloning site from pET28a) and the pET21a plasmid via the NdeI and EcoRI (New England Biolabs) restriction sites (24).
The pPJ131-NgfabI plasmid was transformed into the fabI temperature-sensitive E. coli strain JP1111 to determine complementation. Strain JP1111 is viable at 30 °C but nonviable at 42 °C without fabI gene complementation. The JP1111 cells were transformed with the pPJ131-NgfabI plasmid, the pPJ131 parent plasmid, and the pBluescript plasmids expressing fabI from E. coli and S. aureus (35) and then plated on Luria-Bertani (10 g of tryptone, 5 g of yeast extract, and 5 g of NaCl in 1 liter of distilled water for liquid media with an additional 7.5 g of agarose for plates) plates at 30 °C with 100 μg/ml carbenicillin. The transformed cells were restreaked onto Luria-Bertani plates with 100 μg/ml carbenicillin and grown at 30 or 42 °C to determine whether the N. gonorrhoeae fabI complements the E. coli fabI activity.
BL21AI cells harboring the pET21a-NgfabI were grown in Luria-Bertani medium with 100 μg/ml carbenicillin at 37 °C with 225-rpm shaking until A600 reached 0.6–0.8. The culture was then induced with 1 mm isopropyl β-d-1-thiogalactopyranoside and 0.2% arabinose (w/v) and grown at 17 °C with 225-rpm shaking overnight. Cells were pelleted and washed twice with 20 mm Tris, pH 8.2, and finally resuspended in 20 mm Tris, pH 8.2 (20 ml/liter of culture). Cells were lysed via a cell disruptor and the C-terminal His6-tagged NgFabI was purified via standard nickel chelation chromatography (24).
The NgFabI enzymatic activity was determined by measuring the conversion of NAD(P)H to NAD(P)+ at 340 nm. The enzyme reactions were 100 μl in volume and performed and monitored in Costar UV half-area 96-well plates with a SpectraMax 340 instrument taking 340-nm readings at 10-s intervals at 37 °C. The substrate crotonyl-ACP was synthesized as described previously using E. coli ACP (23). The velocity of the NgFabI enzyme (200 nm) was measured by adding 50 μm crotonyl-ACP and either 200 μm NADPH or NADH in 20 mm Tris, pH 8.0. Upon finding that NADH, but not NADPH, catalyzed the NgFabI reaction, NADH was used for future measurements. The apparent Km of crotonyl-ACP was determined by adding 200 nm NgFabI to 200 μm NADH and 3, 6, 12, 25, 50, 100, or 150 μm crotonyl-ACP. The apparent Km of NADH was determined by adding 200 nm NgFabI to 50 μm crotonyl-ACP and 5, 10, 20, 40, 80, 160, and 320 μm NADH. The reaction was mixed for 10 s by the mix function on the plate reader, and data were acquired at 10-s intervals for 5 min. Initial velocity was calculated from the linear phase of the progress curve and fit using a standard Michaelis-Menten equation to determine the apparent Km. The IC50 of AFN-1252 against NgFabI was measured as above at saturating substrate concentrations (50 μm crotonyl-ACP and 200 μm NADH) against different concentrations of AFN-1252 (0, 7.8, 15.6, 31.2, 62.5, 125, 250, and 500 nm). All kinetic experiments were run in duplicates. Under standard steady state conditions, inhibitors are kept at concentrations 10-fold or more above the concentration of the enzyme, so that the formation of the enzyme inhibitor complex does not alter the concentration of free inhibitors. However, AFN-1252 had affinity values in the nanomolar range for the NgFabI, and nanomolar concentrations of NgFabI are necessary to generate a detectable signal in our kinetic assay. Thus, the data were fit to the Morrison quadratic equation for fitting tight binding inhibitors (23, 36), which allows for the determination of affinity in terms of free and bound concentrations of enzyme and inhibitor, accounting for the impact of enzyme inhibitor binding on the free concentration of inhibitor. The resulting Kiapp was the apparent dissociation constant of AFN-1252 to the NgFabI.
N. gonorrhoeae was grown in SJ-GC medium until A600 reached 0.5, and then aliquoted to 10-ml cultures. Increasing concentrations of AFN-1252 (DMSO only, 1, 2, 4, 8, and 16 μm) were added to each culture and incubated for 15 min at 37 °C with 125-rpm shaking. Next, 10 μCi of [1-14C]acetate was added to each culture and incubated for 45 min at 37 °C with 125-rpm shaking. The cells were harvested via centrifugation (4,000 × g) and washed twice with phosphate-buffered saline. The lipid fraction of the cells was extracted via the Bligh and Dyer method (37), and the radioactive incorporation was measured via liquid scintillation counting (Tri-Carb 2910 TR, PerkinElmer Life Sciences). Data were plotted as fractional incorporation relative to the DMSO-treated cells and fitted via a standard IC50 equation. Plotted data were derived from two biological replicates.
N. gonorrhoeae was grown in SJ-GC medium until A600 reached 0.5, and then aliquoted to 10-ml cultures. AFN-1252 (4 μm) or DMSO (untreated) was added to each culture and incubated for 15 min at 37 °C with 125-rpm shaking. Next, the cell cultures were labeled for 45 min at 37 °C with 125-rpm shaking with 10 μCi of a 3H-labeled amino acid mixture, [3H]thymidine, or [3H]uracil to measure protein, DNA, and RNA synthesis, respectively. The cells were collected via filtration and washed with 3 ml of phosphate buffered saline three times. The radioactive incorporation into the cells was measured via liquid scintillation counting of the filter. Data are plotted as fractional incorporation relative to the untreated cells normalized to the final A600 of the cultures. Plotted data were derived from two biological replicates.
N. gonorrhoeae was grown in SJ-GC medium until A600 reached 0.25 and then aliquoted to 10-ml cultures. Increasing concentrations of AFN-1252 (DMSO only, 1, 2, 4, 8, and 16 μm) were added to each culture and incubated for 15 min at 37 °C with 125-rpm shaking. Next, 5 μCi of [14C]16:0 was added to each culture and incubated for 45 min at 37 °C with 125-rpm shaking. The cells were harvested via centrifugation (4,000 × g) and washed twice with phosphate-buffered saline. The lipid fraction of the cells was extracted via the Bligh and Dyer method (37), and the total radioactive incorporation was measured via liquid scintillation counting. The distribution of radioactivity into the phospholipid fraction was determined by thin-layer chromatography on Analtech Silica Gel G plates developed with chloroform/methanol/acetic acid (98/2/1, v/v/v). The percentage of incorporation into phospholipids was multiplied by the total radioactive incorporation to determine the total radioactive incorporation into phospholipids. Data were plotted as fractional incorporation relative to the DMSO-treated cells and fitted via a standard IC50 equation. Plotted data were derived from two biological replicates.
N. gonorrhoeae was grown in SJ-GC medium until A600 reached 0.25 and then aliquoted to 10-ml cultures. DMSO control, 100 μm [7,7,8,8-D4]palmitic acid (D4-16:0, Cambridge Isotope Laboratory), or 100 μm [methyl-D3]lauric acid (D3-12:0, Cambridge Isotope Laboratory) was added to the culture and incubated for 2 h. The cells were harvested and washed twice in phosphate-buffered saline, extracted via the Bligh and Dyer method (37), and analyzed via molecule species profiling.
Phospholipid molecular species profiling for PE was performed as described previously (38). Phospholipid molecular species fingerprints were determined using direct infusion electrospray ionization-mass spectrometry technology (39, 40). Mass spectrometry analysis was performed using a FinniganTM TSQ® Quantum (Thermo Electron, San Jose, CA) triple quadruple mass spectrometer. The instrument was operated in positive ion mode for PE analysis. Acyl chain lengths were assigned from the mass based on product scans of the particular mass peak or predications from LipidMaps for previously identified peaks. Minor isobaric species may be present at each mass, but the major molecular species represented at each major mass peak is labeled on the figures. Two biological replicates were analyzed for each condition, and representative spectra are shown in the figures.
The pPJ131-NGO0530 and pPJ131-NGO1213 plasmids were transformed into strain LCH30 (aas-1 fadD88) along with the parent pPJ131 plasmid (empty vector), pPJ131 plasmid expressing the Chlamydia trachomatis acyl-ACP synthetase (aasC), or pBluescript plasmid expressing E. coli acyl-CoA synthetase (fadD) as controls. Cultures expressing the respective constructs were grown to A600 =0.25 in 10 ml of Luria Bertani medium (41) with 100 μg/ml carbenicillin at 37 °C, and then 100 μm D3-14:0 was added to the medium. The cultures were grown for an additional 1.5 h and then harvested. The cells were washed twice, the lipids were extracted, and the incorporation of D3-14:0 into PE was analyzed by mass spectrometry.
Enoyl-ACP reductase catalyzes the reduction of trans-2-enoyl-ACP into acyl-ACP and is a rate-determining step in FASII (42, 43). The predicted FabI sequences from the different species of the genus Neisseria have >98% amino acid sequence identity and are homologous to the E. coli FabI (60% identity, e-value = 1e−90). Whether the predicted Neisseria FabI was a bona fide enoyl-ACP reductase was assessed by determining whether expression of the Neisseria FabI complemented the growth of an E. coli fabI(Ts) mutant. The E. coli codon-optimized fabI gene from N. gonorrhoeae strain FA 1090 was cloned into the pPJ131 expression vector and then transformed into E. coli strain JP1111 (fabI(Ts)) (44). Strain JP1111 was able to grow without fabI gene complementation at the permissive temperature (Fig. 1A). However, strain JP1111 was not able to grow without fabI gene complementation at the non-permissive temperature (Fig. 1B). The E. coli, S. aureus, and N. gonorrhoeae fabI genes all complemented growth at the non-permissive temperature (Fig. 1B), demonstrating that the N. gonorrhoeae fabI functioned as an enoyl-ACP reductase in the heterologous E. coli FASII system.
NgFabI was expressed, and the protein was purified (Fig. 2A). The enzymatic activity of NgFabI was determined by measuring conversion of NADH to NAD+ at 340 nm arising from reduction of crotonyl-ACP to butyryl-ACP. NgFabI exhibited an apparent Km of 116.2 ± 18.3 μm for NADH (Fig. 2B) and an apparent Km for crotonyl-ACP of 7.2 ± 1.1 μm (Fig. 2C). NADPH did not support the enzymatic reaction (data not shown), consistent with FabI enzymes from other Gram-negative bacteria (38, 42). The kcat of NgFabI was 25.2 ± 2.8 min−1, similar to the velocities of other characterized FabI enzymes (38, 45). The complementation and biochemical characterization experiments showed that the predicted NgFabI functioned as an enoyl-ACP reductase.
AFN-1252 was a tight binding inhibitor of the NgFabI, meaning that the dissociation constant for AFN-1252 was lower than the concentration of NgFabI enzyme necessary to get a reliable signal in the spectrophotometric assay. Therefore, the Morrison quadratic equation for tight binding inhibitors was used to calculate the Kiapp for AFN-1252 (36). This treatment of the data takes into account the change in the concentration of the inhibitor due to its binding to the enzyme. The Kiapp of AFN-1252 for NgFabI at saturating concentrations of substrates was 5 ± 2 nm (Fig. 3A).
The effect of increasing concentrations of AFN-1252 on the growth of N. gonorrhoeae was determined. Bacteria usually continue to grow for one generation following the inhibition of fatty acid synthesis, causing a characteristic delayed growth arrest phenotype when fatty acid synthesis is inhibited (46). This growth phenotype was observed in the growth curves of N. gonorrhoeae treated with increasing concentrations of AFN-1252 (Fig. 3B). N. gonorrhoeae was able to grow with minimal decrease in growth rate for 45 min (corresponding to another generation) after AFN-1252 was added to the culture, regardless of the concentration of AFN-1252 added. Increasing concentrations of AFN-1252 exerted increasing inhibitory effects after this initial growth phase, with 4 and 8 μm AFN-1252 causing a complete cessation of growth. These data showed that AFN-1252 was able to block Neisseria growth with a phenotype that was consistent with AFN-1252 acting through inhibiting fatty acid synthesis.
The effect of increasing concentrations of AFN-1252 on the rate of fatty acid biosynthesis in N. gonorrhoeae was determined. N. gonorrhoeae cultures were treated with increasing concentrations of AFN-1252 for 15 min and then labeled with [14C]acetate for 45 min. Incorporation of [14C]acetate into the phospholipid fraction was determined. AFN-1252 caused a dose-dependent decrease in the incorporation of [14C]acetate, with greater than 95% inhibition achieved at 16 μm AFN-1252 (Fig. 3C). Metabolic labeling in the presence and absence of 4 μm AFN-1252 was used to assess its effect on the four major pathways of macromolecular biosynthesis (Fig. 3D). AFN-1252 caused a 75% inhibition of lipid biosynthesis, with less than 5% inhibition of protein, DNA, or RNA biosynthetic pathways in N. gonorrhoeae. This experiment showed that AFN-1252 selectively inhibited lipid biosynthesis. Together, the growth inhibition, pathway labeling, and biochemical analysis of FabI inhibition demonstrated that AFN-1252 blocked the growth of N. gonorrhoeae through its on-target inhibition of the FabI component of FASII.
The ability of N. gonorrhoeae to utilize exogenous fatty acids was determined by examining the incorporation of isotopic labeled exogenous fatty acids into the N. gonorrhoeae phospholipids. The major phospholipid class in N. gonorrhoeae was PE, comprising greater than 70% of the total phospholipids (47). Molecular species profiling was used to determine the acyl chain composition of the N. gonorrhoeae PE. The major PE molecular species of N. gonorrhoeae had a molecular mass of 690.52, corresponding to PE containing a 16:0 fatty acid in the 1-position and a 16:1 fatty acid in the 2-position (Fig. 4A). The 16:1 fatty acid is known to be palmitoleic acid (16:1Δ9), which has a cis double bond between the 9- and 10-carbons (48). Smaller amounts of 16:0/18:1, 16:1/18:1, 16:1/16:1, 14:0/16:1, 16:0/14:0, 14:0/14:0, and 18:1/18:1 PE molecular species were also detected. A growing culture of N. gonorrhoeae was labeled with 100 μm D4-16:0. Two new prominent molecular peaks at m/z = 694.62, corresponding to D4-16:0/16:1, and m/z = 700.64, corresponding to D4-16:0/D4-16:0, appeared (Fig. 4B). The D4-16:0 was also incorporated into other, minor molecular species where 16:0 was normally found. This experiment showed that N. gonorrhoeae was able to incorporate exogenous D4-16:0 into both acyl chain positions. Next, cells were labeled with D3-12:0 to determine whether exogenous fatty acids were elongated by Neisseria. Cells grown in the presence of D3-12:0 had new PE molecular species corresponding to the elongation and incorporation of D3-12:0 (Fig. 4C). The most prominent peak occurred at m/z = 693.57, corresponding to D3-16:0/16:1, and a number of smaller peaks containing D3-12:0, D3-14:0, and D3-16:0 were observed. These experiments showed that N. gonorrhoeae was able to elongate D3-12:0 prior to incorporating the fatty acid into phospholipid. This result means that Neisseria encoded an enzyme or pathway that produced acyl-ACP from exogenous fatty acids.
N. gonorrhoeae have two genes (NGO0530 and NGO1213) that encode members of the superfamily of proteins containing the AFD_class_I domain, which corresponds to the binding site for the acyl-adenylate intermediate in acyl-ACP and acyl-CoA synthetases (12, 49). The functions of the two potential acyl-CoA/ACP synthetase genes were determined by their expression in E. coli and isotopic fatty acid labeling. E. coli strain LCH30 (aas-1 fadD88) lacks both acyl-ACP synthetase (aas) and acyl-CoA synthetase (fadD) and was unable to incorporate exogenous fatty acids into phospholipid (50). Thus, when strain LCH30 harboring the empty pPJ131 expression vector was grown in the presence of D3-14:0, there was no detectable incorporation of D3-14:0 into the phospholipids (Fig. 5A). Strain LCH30 expressing the E. coli acyl-CoA synthetase (FadD) did incorporate D3-14:0 into phospholipid, but there was no elongation of the fatty acid (Fig. 5B). Acyl-CoA can be used for phospholipid synthesis in E. coli but cannot be converted to acyl-ACP and elongated. In contrast, strain LCH30 expressing the C. trachomatis acyl-ACP synthetase (aasC) converted D3-14:0 to D3-16:0 prior to incorporation into phospholipid (Fig. 5C), consistent with the formation of acyl-ACP that can either be used for phospholipid synthesis or enter the FASII elongation cycle (49). Strain LCH30 expressing NGO0530 incorporated D3-14:0 both as D3-14:0 and elongated D3-14:0 to D3-16:0, demonstrating that NGO0530 functioned as an acyl-ACP synthetase (Fig. 5D). Thus, we have named the NGO0530 gene aasN for Neisseria acyl-ACP synthetase. Strain LCH30 expressing NGO1213 incorporated D3-14:0 only as D3-14:0, demonstrating that NGO1213 functioned as an acyl-CoA synthetase (Fig. 5E). The exogenous fatty acid labeling experiments conducted in N. gonorrhoeae showed that D3-12:0 was incorporated into the phospholipids overwhelmingly as the elongated D3-14:0 and D3-16:0 fatty acids (Fig. 4C). The existence of AasN accounts for the activation and elongation of exogenous fatty acids into the N. gonorrhoeae phospholipids. The acyl-CoA synthetase, NGO1213, did not appear to have a role in the incorporation of exogenous fatty acids into the N. gonorrhoeae phospholipids.
AFN-1252 blockade at the FabI step in S. aureus results in cellular ACP being converted to short-chain enoyl-ACP and prevents the incorporation of exogenous fatty acids by tying up free ACP needed to ligate the incoming fatty acids. However, N. gonorrhoeae also encoded an acyl-CoA synthetase, which raised the possibility that exogenous fatty acids were converted into acyl-CoA that would bypass FabI inhibition by acting as the acyl donor for the acyltransferases. Therefore, the effect of AFN-1252 on the incorporation of exogenous fatty acid in N. gonorrhoeae was characterized via fatty acid labeling experiments to determine whether exogenous fatty acid incorporation was affected by FabI inhibition. A growing culture of N. gonorrhoeae was treated with increasing concentrations of AFN-1252 for 15 min and then labeled with [14C]16:0 for 45 min. AFN-1252 caused a concentration-dependent decrease in the incorporation of [14C]16:0 into the phospholipids of N. gonorrhoeae, with >90% inhibition achieved at 16 μm AFN-1252 (Fig. 5F). This result showed that the N. gonorrhoeae exogenous fatty acid incorporation mechanism was linked to the inhibition of endogenous fatty acid synthesis.
The SJ-GC medium contains potato starch, which is a known drug binder, so the potency of AFN-1252 against N. gonorrhoeae was tested in the chemically defined GW medium. Long-chain fatty acids have potent inhibitory effects against N. gonorrhoeae (51), so conditions to deliver the 16:0 and 16:1Δ9 fatty acids, the major components of N. gonorrhoeae phospholipids, without inhibiting N. gonorrhoeae growth were examined. N. gonorrhoeae was refractory to 100 μm 16:0 in SJ-GC medium, whereas the unsaturated 16:1Δ9 fatty acid had an MIC of 12.5 μm. N. gonorrhoeae was refractory to a 100 μm concentration of either 16:0 or 16:1Δ9 in SJ-GC medium containing 10 mg/ml BSA, showing that SJ-GC medium containing BSA was an acceptable method to deliver the fatty acids (Table 1).
AFN-1252 demonstrated a minimal inhibitory concentration of 1.25 μm against N. gonorrhoeae in SJ-GC medium (Table 2). Adding 10 mg/ml BSA to the medium shifted the MIC to 5 μm. Supplementing the medium with 50 μm 16:0 and 50 μm 16:1Δ9 did not change the AFN-1252 MIC. The cofactor lipoic acid is an important product of FASII, and adding 1 μm lipoic acid in addition to the fatty acids did not change the MIC, showing that exogenous FASII products cannot complement FabI inhibition by AFN-1252 in N. gonorrhoeae. Similar MIC experiments were performed on N. meningitidis. Unlike N. gonorrhoeae, neither 16:0 nor 16:1Δ9 inhibited N. meningitidis growth (Table 1). AFN-1252 demonstrated a minimal inhibitory concentration of 1.25 μm against N. meningitidis in SJ-GC medium that shifted to 5 μm when 10 mg/ml BSA was present in the medium (Table 2). However, fatty acid and lipoic acid supplementation did not increase the MIC. Lipoic acid was provided because it is an essential cofactor for ketoacid dehydrogenases and is derived from octanoyl-ACP arising from FASII (52,–54). S. aureus lacking FASII require exogenous lipoate in addition to fatty acids to support maximum growth (35, 55). These experiments showed that exogenous fatty acids cannot bypass the inhibition of FabI by AFN-1252 in both N. gonorrhoeae and N. meningitidis.
One product of FASII in Gram-negative bacteria is 3-hydroxyacyl-ACP, which is used in the synthesis of LPS. LPS synthesis is essential for E. coli (15) but is nonessential for N. meningitides (16,–18). The LpxC inhibitor CHIR-090, which has nanomolar affinity against the Neisseria LpxC (56), was tested against N. gonorrhoeae to determine whether LPS was also nonessential in N. gonorrhoeae and whether the effects of FabI inhibition were through inhibiting 3-hydroxy fatty acid synthesis. CHIR-090 demonstrated potent activity (MIC = 78 nm) against E. coli, but N. gonorrhoeae was refractory to up to 20 μm CHIR-090 (Fig. 6). This result was consistent with LPS being nonessential in Neisseria, and therefore the growth-arresting effects of AFN-1252 were through inhibiting fatty acid synthesis required for phospholipids and not LPS.
This work identifies an acyl-ACP synthetase (AasN) as a key participant in the activation of extracellular fatty acids in Neisseria (Fig. 7). Following their conversion to acyl-ACP, the fatty acid is either converted to acyl-PO4 by PlsX and esterified to the 1-position of sn-glycerol phosphate by PlsY, utilized directly by the PlsC acyltransferase, or enters the elongation cycle of FASII (Fig. 7). Thus, Neisseria differ from prototypical Gram-negative bacteria, which activate extracellular fatty acid to acyl-CoA for the PlsB/PlsC acyltransferase system. This flow of exogenous fatty acids into phospholipid is distinct from the fatty acid kinase pathway used in Gram-positive bacteria that use the PlsX/PlsY/PlsC acyltransferase system found in Neisseria. In the Gram-positive paradigm, extracellular fatty acids are phosphorylated by fatty acid kinase and are either used by PlsY or converted to acyl-ACP by PlsX (20, 21). Acyl-ACP synthetase is found in other Gram-negative bacteria, although none of these organisms use the PlsX/PlsY/PlsC acyltransferase system. The first acyl-ACP synthetase discovered is a subunit of 2-acyl-lysophospholipid acyltransferase of E. coli; however, this enzyme produces only enzyme-bound acyl-ACP that is not available to FASII or the other acyltransferases (57, 58). Vibrio harveyi expresses an acyl-ACP synthetase capable of releasing acyl-ACP that is used by FASII or the PlsB/PlsC acyltransferases (59, 60). C. trachomatis also expresses an acyl-ACP synthetase that produces acyl-ACP for FASII and the PlsE/PlsC acyltransferase system (49). The function(s) for an acyl-CoA synthetase in N. gonorrhoeae is not clear. In E. coli, acyl-CoAs are consumed by fatty acid β-oxidation, but Neisseria do not encode the genes necessary to break down acyl-CoA via this pathway. Furthermore, whereas both E. coli PlsB and PlsC acyltranferases are able to utilize acyl-CoA as acyl donors, the characterized PlsX/PlsY/PlsC acyltransferases cannot use acyl-CoA (19, 61). The observed elongation and incorporation of exogenous fatty acids in Neisseria is explained by AasN activation of exogenous fatty acids to acyl-ACP because exogenous fatty acids are elongated and used by the acyl-PO4/acyl-ACP-dependent PlsX/PlsY/PlsC acyltransferase system. The acyl-CoA synthetase does not appear to function in exogenous fatty acid incorporation in our experiments. The observations that exogenous fatty acids are elongated and that AFN-1252 blocks fatty acid incorporation into phospholipid support the conclusion that acyl-ACP is the key intermediate in Neisseria and not acyl-CoA. The presence of AasN is sufficient to account for the elongation and incorporation of exogenous fatty acids in Neisseria, and the role of acyl-CoA synthetase in Neisseria physiology remains to be determined.
The therapeutically relevant aspect of the study is that it establishes FASII inhibitors as effective therapeutics against Neisseria. All bacteria studied to date are capable of incorporating exogenous fatty acids into membrane phospholipids (7). Thus, it is important to experimentally determine whether extracellular fatty acid supplementation can overcome a blockade in FASII (7,–9). Neisseria is different from the prototypical Gram-negative E. coli model system in that it uses AasN to activate exogenous fatty acids and uses the acyl-PO4 pathway for phospholipid synthesis, and LPS is not essential for survival. These properties suggest that Neisseria may be able to escape FASII inhibitors by obtaining straight-chain saturated and unsaturated fatty acid from the host. However, our experiments establish that Neisseria cannot circumvent FabI inhibition by scavenging fatty acids from the environment. In S. aureus, FabI inhibition blocks the uptake of fatty acids due to the depletion of cellular ACP as the initiation module of FASII continues to feed acyl-ACP into the elongation cycle and the enoyl-ACP intermediates accumulate at the inhibited step (35). AFN-1252 also inhibits the incorporation of extracellular fatty acids into Neisseria consistent with the blockade at the FabI step triggering the same imbalance in fatty acid metabolism in both organisms. Although there remains more to understand about regulation of fatty acid metabolism in Neisseria, this work validates FASII, and specifically FabI, as a therapeutic target for the development of critically needed new Neisseria therapeutics.
J. Y., R. E. L., and C. O. R. designed and directed the study. J. Y., D. F. B., and M. W. F. performed the experiments. All authors contributed to and approved the manuscript.
We thank the Hartwell Center DNA sequencing shared resource for DNA sequencing, Deepak Bhasin and the Chemical Biology and Therapeutics Synthesis Core for the AFN-1252, and Chris Raetz for CHIR-090.
*This work was supported by National Institutes of Health Grant GM034496 (to C. O. R) and Cancer Center Support Grant CA21765 and by the American Lebanese Syrian Associated Charities. The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
3The abbreviations used are: