Membrane disruption by antimicrobial fatty acids releases low-molecular-weight proteins from Staphylococcus aureus

Abstract

The skin represents an important barrier for pathogens and is known to produce fatty acids that are toxic toward gram-positive bacteria. A screen of fatty acids as growth inhibitors of Staphylococcus aureus revealed structure-specific antibacterial activity. Fatty acids like oleate (18:1Δ9) were nontoxic, whereas palmitoleate (16:1Δ9) was a potent growth inhibitor. Cells treated with 16:1Δ9 exhibited rapid membrane depolarization, the disruption of all major branches of macromolecular synthesis, and the release of solutes and low-molecular-weight proteins into the medium. Other cytotoxic lipids, such as glycerol ethers, sphingosine, and acyl-amines blocked growth by the same mechanisms. Nontoxic 18:1Δ9 was used for phospholipid synthesis, whereas toxic 16:1Δ9 was not and required elongation to 18:1Δ11 prior to incorporation. However, blocking fatty acid metabolism using inhibitors to prevent acyl-acyl carrier protein formation or glycerol-phosphate acyltransferase activity did not increase the toxicity of 18:1Δ9, indicating that inefficient metabolism did not play a determinant role in fatty acid toxicity. Nontoxic 18:1Δ9 was as toxic as 16:1Δ9 in a strain lacking wall teichoic acids and led to growth arrest and enhanced release of intracellular contents. Thus, wall teichoic acids contribute to the structure-specific antimicrobial effects of unsaturated fatty acids. The ability of poorly metabolized 16:1 isomers to penetrate the cell wall defenses is a weakness that has been exploited by the innate immune system to combat S. aureus.

Fig 1

Structure-specific fatty acid toxicity in S. aureus. The MICs of a series of fatty acid structures were determined using the broth microdilution assay described in Materials and Methods. Data shown are for the laboratory S. aureus strain RN4220. MIC data for four virulent S. aureus strains, B. subtilis, and S. pneumoniae are provided in Fig. S1 in the supplemental material. Fatty acid structures are described according to the following scheme: number of carbons:number of double bonds:Δposition of double bond. All double bonds were cis except where specifically indicated as trans.

Fig 2

Rapid S. aureus growth arrest and membrane depolarization by 16:1Δ9. (A) S. aureus strain RN4220 was grown to mid-log phase in LB–0.1% Brij 58 and treated with either 4 mM 18:1Δ9 (○) or 100 μM 16:1Δ9 (●) fatty acids at the time indicated by the arrow. Cell growth was monitored and compared to that of the untreated control (■) cell population. The slight increase in turbidity at the arrow is due to the added fatty acids. An experiment with a shorter time course showed that growth arrest occurred <5 min after the addition of 100 μM 16:1Δ9 (inset). (B) The kinetics of growth inhibition were compared under the following conditions: with AFN-1252, a fatty acid synthesis inhibitor (○, 60 ng/ml); with mupirocin, a protein synthesis inhibitor (●, 120 ng/ml); with CCCP, a protonophore (□, 5 μM); and in untreated cells (■). (C) Metabolic labeling with [¹⁴C]acetate for lipids, [³H]uracil for RNA, [³H]thymidine for DNA, and ³H-labeled amino acids for protein synthesis was performed as described in Materials and Methods. Strain RN4220 was grown to mid-log phase and labeled for 20 min without additions to obtain the 100% incorporation in the absence of lipid in tryptone broth (dashed line). The amount incorporated was compared to that in cells treated with either 18:1Δ9 (100 μM) or 16:1Δ9 (100 μM) and then labeled with the indicated tracer for 20 min. Incorporation of label was normalized to cell number. (D) Strain RN4220 grown in tryptone broth was treated with either 1% DMSO (control) or CCCP (5 μM) to collapse the proton gradient and subjected to dual-color flow cytometry analysis within 120 s. (E) Strain RN4220 grown in tryptone broth was treated with either 100 μM 18:1Δ9 or 16:1Δ9 and subjected to dual-color flow cytometry analysis within 120 s as described in Materials and Methods. (F) C₁₂-resazurin fluorescence indicates the reducing environment within the cell. Strain RN4220 was treated with either 1% DMSO (control), 16:1Δ9 (100 μM), 18:1Δ9 (100 μM), or CCCP (5 μM) for 15 min, and the C₁₂-resazurin dye was added. Fluorescence was recorded as described in Materials and Methods. Triplicate measurements were recorded. Error bars in panels C and D are standard deviations from triplicate measurements. A representative result from multiple experiments are shown in panels A, B, D, and E.

Fig 3

Release of solutes and ACP from S. aureus by toxic fatty acids. (A and B) Two assays were used to determine the permeability properties of cells treated with toxic fatty acids. Controls contained 1% DMSO; fatty acids were added to a final concentration of 100 μM, nisin was 25 μg/ml, and CCCP was 5 μM. (A) The TO-PRO-3 iodide dye fluoresces when bound to DNA but cannot penetrate intact cytoplasmic membranes. The increase in TO-PRO-3 iodide fluorescence was measured 20 min after the treatment of triplicate samples of strain RN4220 with the indicated compounds. (B) Strain RN4220 was treated for 20 min with the indicated compounds, and the cells and medium were separated by centrifugation. Triplicate samples of cells and medium were independently analyzed for their ATP content as described in Materials and Methods. (C) Release of ACP from S. aureus. Strain RN4220 was treated with a 200 μM concentration of the indicated compounds for 30 min, and the cells were separated from the medium by centrifugation. Proteins in the medium were isolated and concentrated by anion exchange chromatography, and samples of the cells and medium were separated by urea gel electrophoresis. ACP was detected using Western blotting and an anti-ACP antibody. (D) The release of ACP following the treatment of strain RN4220 with 200 μM 16:1Δ9 was determined as a function of time by collecting the cells by centrifugation and analyzing the ACP content by immunoblotting with anti-ACP antibody following urea gel electrophoresis. The same number of extracted cells was loaded into each lane. (E) Strain RN4220 was untreated or treated with the indicated lipids for 30 min. The cells were harvested, and an equal number of cells was loaded into each lane. Lipids used were 16:1Δ9 (200 μM), 18:1Δ9 (200 μM), laurylamine (12-NH₂, 200 μM), alkylglycerol (12-GE, 200 μM), and sphingosine (Sph, 35 μM). Experiments were performed in triplicate, and error bars represent standard deviations. Images were representative of multiple experiments.

Fig 4

16:1Δ9 is a poor substrate for phospholipid synthesis. Following labeling of strain RN4220 with exogenous fatty acids, the lipids were extracted and separated by thin-layer chromatography to identify the neutral lipids present. PL, phospholipid; FA, free fatty acid; DAG, diacylglycerol. (A) Strain RN4220 metabolically labeled with 500 μM [¹⁴C]18:1Δ9 (specific activity, 2.2 mCi/mole) for 30 min in TB–0.1% Brij 58. (B) Strain RN4220 metabolically labeled with 20 μM [¹⁴C]16:1Δ9 for 30 min. (C) Strain RN4220 metabolically labeled with 500 μM [¹⁴C]16:1Δ9 (specific activity, 2.2 mCi/mole) for 30 min. (D to F) Cells were grown to mid-log phase with the indicated fatty acid, the lipids were extracted, and the molecular species of PtdGro were determined by mass spectroscopy as described in Materials and Methods. The fatty acid compositions of the samples are provided in Fig. S1 in the supplemental material. (D) PtdGro molecular species in strain RN4220 grown in tryptone broth without a fatty acid supplement. (E) PtdGro molecular species in strain RN4220 grown in the presence of 25 μM 18:1Δ9. (F) PtdGro molecular species in strain RN4220 grown in the presence of 25 μM 16:1Δ9. Data shown are representative results from multiple experiments.

Fig 5

Role of metabolism and wall teichoic acids in fatty acid intoxication. (A) Strain RN4220 or PDJ28 (ΔgpsA) was grown to mid-log phase and either treated with AFN-1252 to arrest fatty acid synthesis (RN4220) or deprived of glycerol for 1 h (PDJ28) to arrest phospholipid synthesis. The cells were then treated with 100 μM 16:1Δ9 or 500 μM 18:1Δ11 for 30 min, and the amounts of ATP in the supernatant and cell pellet were determined as described in Materials and Methods. (B) MIC determination for 16:1Δ9 (circles) and 18:1Δ9 (squares) in strain EBII53 (ΔtarO) harboring a plasmid expressing tarO under the control of IPTG inducer. Cells were grown in TB with (open symbols) or without (filled symbols) 1 mM IPTG. (C) Strain RN4220 (control) or strain EBII53 in the presence or absence of 1 mM IPTG was prelabeled with [¹⁴C]acetate. The cells were harvested, washed, and treated with 100 μM 16:1Δ9 for 2 h. Cells and supernatant were collected by centrifugation, and the lipids were extracted and quantified by liquid scintillation counting. (D) Strains RN4220 and EBII53 were grown and treated as described in panel C, except in the absence of labeled acetate. Cell supernatants were collected, and the protein released into the medium was measured as described in Materials and Methods. Experiments were performed in triplicate, and error bars indicate standard deviations.