Host Fatty Acid Utilization by Staphylococcus aureus at the Infection Site

Abstract

Staphylococcus aureus utilizes the fatty acid (FA) kinase system to activate exogenous FAs for membrane synthesis. We developed a lipidomics workflow to determine the membrane phosphatidylglycerol (PG) molecular species synthesized by S. aureus at the thigh infection site. Wild-type S. aureus utilizes both host palmitate and oleate to acylate the 1 position of PG, and the 2 position is occupied by pentadecanoic acid arising from de novo biosynthesis. Inactivation of FakB2 eliminates the ability to assimilate oleate and inactivation of FakB1 reduces the content of saturated FAs and enhances oleate utilization. Elimination of FA activation in either ΔfakA or ΔfakB1 ΔfakB2 mutants does not impact growth. All S. aureus strains recovered from the thigh have significantly reduced branched-chain FAs and increased even-chain FAs compared to that with growth in rich laboratory medium. The molecular species pattern observed in the thigh was reproduced in the laboratory by growth in isoleucine-deficient medium containing exogenous FAs. S. aureus utilizes specific host FAs for membrane biosynthesis but also requires de novo FA biosynthesis initiated by isoleucine (or leucine) to produce pentadecanoic acid.IMPORTANCE The shortage of antibiotics against drug-resistant Staphylococcus aureus has led to the development of new drugs targeting the elongation cycle of fatty acid (FA) synthesis that are progressing toward the clinic. An objection to the use of FA synthesis inhibitors is that S. aureus can utilize exogenous FAs to construct its membrane, suggesting that the bacterium would bypass these therapeutics by utilizing host FAs instead. We developed a mass spectrometry workflow to determine the composition of the S. aureus membrane at the infection site to directly address how S. aureus uses host FAs. S. aureus strains that cannot acquire host FAs are as effective in establishing an infection as the wild type, but strains that require the utilization of host FAs for growth were attenuated in the mouse thigh infection model. We find that S. aureus does utilize host FAs to construct its membrane, but host FAs do not replace the requirement for pentadecanoic acid, a branched-chain FA derived from isoleucine (or leucine) that predominantly occupies the 2 position of S. aureus phospholipids. The membrane phospholipid structure of S. aureus mutants that cannot utilize host FAs indicates the isoleucine is a scarce resource at the infection site. This reliance on the de novo synthesis of predominantly pentadecanoic acid that cannot be obtained from the host is one reason why drugs that target fatty acid synthesis are effective in treating S. aureus infections.

Keywords: Staphylococcus aureus; fatty acid; fatty acid activation; fatty acid binding protein; fatty acid kinase; host-pathogen interactions; phospholipid; virulence.

FIG 1

Model for the utilization of host FAs for phosphatidylglycerol (PG) synthesis by S. aureus. (A) FA kinase is the only pathway for the activation of extracellular FAs in S. aureus. Host FAs bind to one of the two FA binding proteins. FakB1 specifically binds saturated FA (16:0), and FakB2 selectively binds monounsaturated FA (18:1). The FakB(FA) complexes are phosphorylated by FakA, converting host FAs to the key intermediate, acyl-PO₄. The resulting acyl-PO₄ has two fates: PlsX and PlsY. PlsX converts acyl-PO₄ to acyl-ACP that is elongated by FASII before being converted again to acyl-PO₄ by PlsX. FASII produces 15:0 ACP that is preferentially used by PlsC to acylate the 2 position. (B) Structure of PG. The major membrane phospholipid of S. aureus consists of a glycerol phosphate backbone (blue) that is first acylated in the 1 position by acyl-PO₄-dependent PlsY (green) followed by acylation of the 2 position by 15:0-ACP-selective PlsC (red). The glycerol headgroup (gold) is added by a series of three enzymes. There are four stereocenters (*) in the example show the abundant 17:0/15:0 PG molecular species (32:0 PG; m/z 721) of S. aureus.

FIG 2

Bacterial load and detection of S. aureus PG molecular species in the neutropenic thigh model. (A) Enumeration of the bacteria recovered from the infected thighs. The thighs were infected with the indicated S. aureus strains; 24 h later, the thighs were excised, and the number of colonies present was determined by plating serial dilutions. Groups were compared using the Student's t test using GraphPad software, and the P values are shown in red. The shaded bar represents the initial inoculum. (B) A reflection plot of the PG molecular species detected in the infected and uninfected thighs. Thigh lipids were extracted, and separated into phospholipid classes by liquid chromatography, and the total negative ion scan of the peak containing PG was determined by mass spectrometry. (Top) Representative example showing the data obtained from a thigh infected with S. aureus strain JLB2 (ΔfakA). (Bottom) Representative example of the total negative ion current in the PG peak detected in an uninfected thigh. The new PG molecular species arising in the S. aureus infected thigh are highlighted in red. (Insets) Examples of how mass spectrometry was used to determine the FA composition of the new peaks in the infected tissue. 31:0 PG consisted of predominately 16:0 and 15:0 PGs with significantly smaller amounts of 18:0 and 13:0 PGs. 32:0 PG contained 16:0 from host PG and was a mixture of 17:0 and 15:0 PGs and 18:0 and 14:0 PGs from S. aureus. 33:0 PG was 18:0 and 15:0 PGs, and 35:0 PG was 20:0 and 15:0 PGs from S. aureus.

FIG 3

Role of FA kinase in S. aureus PG molecular species composition in the infected thigh. The thigh lipid samples were separated by solid-phase extraction, and the PG molecular species containing 15:0 were detected using mass spectrometry by scanning for parent ions that lose m/z 241 corresponding to the loss of a 15:0 FA. The identities of the FAs in the 1 and 2 positions (blue) are shown in panels A and B for each molecular species (black). (A) Wild-type strain AH1263 grown in rich laboratory medium. (B) Strain AH1263 PG recovered from an infected thigh. (C) Strain JLB2 (ΔfakA) PG recovered from an infected thigh. (D) Strain JLB31 (ΔfakB1 ΔfakB2) PG recovered from an infected thigh. (E) Strain JLB27 (ΔfakB1) PG recovered from an infected thigh. (F) Strain JLB30 (ΔfakB2) PG recovered from an infected thigh. Representative spectra are shown, and the triplicate results are tabulated in Table S1 in the supplemental material. The peak labeled “X” is not a PG molecular species.

FIG 4

Impact of exogenous FAs on PG molecular species. Strains were grown in defined media (described in Materials and Methods) plus 0.1% Brij-58 and with or without 500 μM FAs. The identities of the FAs in the 1 and 2-positions (blue) are shown in panels A and B for each molecular species (black). (A) Strain AH1263 grown in defined medium. (B) PG molecular species of strain AH1263 grown in defined medium broth containing 500 μM 18:1. (C) Strain AH1263 grown in defined medium supplemented with 500 μM 16:0. (D) Strain JLB2 (ΔfakA) grown in defined medium containing 500 μM 18:1.

FIG 5

Impact of exogenous Ile on PG molecular species of wild-type strain AH1263. The identities of the FAs in the 1 and 2 positions (blue) are shown in panel A for each PG molecular species (black). (A) PG molecular species of strain AH1263 grown in defined medium as described in Materials and Methods containing 50 μg/ml Ile and Leu. (B) PG molecular species of strain AH1263 grown in defined medium lacking Ile. (C) PG molecular species of strain AH1263 grown in defined medium lacking Leu. (D) PG molecular species of strain AH1263 grown in defined medium lacking both Ile and Leu.