This method provides a framework for studying incorporation of exogenous fatty acids from complex host sources into bacterial membranes, particularly Staphylococcus aureus. To achieve this, protocols for the enrichment of lipoprotein particles from chicken egg yolk and subsequent fatty acid profiling of bacterial phospholipids utilizing mass spectrometry are described.
Staphylococcus aureus and other Gram-positive pathogens incorporate fatty acids from the environment into membrane phospholipids. During infection, the majority of exogenous fatty acids are present within host lipoprotein particles. Uncertainty remains as to the reservoirs of host fatty acids and the mechanisms by which bacteria extract fatty acids from the lipoprotein particles. In this work, we describe protocols for enrichment of low-density lipoprotein (LDL) particles from chicken egg yolk and determining whether LDLs serve as fatty acid reservoirs for S. aureus. This method exploits unbiased lipidomic analysis and chicken LDLs, an effective and economical model for the exploration of interactions between LDLs and bacteria. The analysis of S. aureus integration of exogenous fatty acids from LDLs is performed using high-resolution/accurate mass spectrometry and tandem mass spectrometry, enabling the characterization of the fatty acid composition of the bacterial membrane and unbiased identification of novel combinations of fatty acids that arise in bacterial membrane lipids upon exposure to LDLs. These advanced mass spectrometry techniques offer an unparalleled perspective of fatty acid incorporation by revealing the specific exogenous fatty acids incorporated into the phospholipids. The methods outlined here are adaptable to the study of other bacterial pathogens and alternative sources of complex fatty acids.
Methicillin-resistant S. aureus (MRSA) is the leading cause of healthcare-associated infection and the associated antibiotic resistance is a considerable clinical challenge1,2,3. Therefore, the development of novel therapeutic strategies is a high priority. A promising treatment strategy for Gram-positive pathogens is inhibiting fatty acid synthesis, a requirement for membrane phospholipid production that, in S. aureus, includes phosphatidylglycerol (PG), lysyl-PG, and cardiolipin4. In bacteria, fatty acid production occurs via the fatty acid synthesis II pathway (FASII)5, which is considerably different from the eukaryotic counterpart, making FASII an attractive target for antibiotic development5,6. FASII inhibitors primarily target FabI, an enzyme required for fatty acid carbon chain elongation7. The FabI inhibitor triclosan is broadly used in consumer and medical goods8,9. Additional FabI inhibitors are being developed by several pharmaceutical companies for the treatment of S. aureus infection10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26. However, many Gram-positive pathogens, including S. aureus, are capable of scavenging exogenous fatty acids for phospholipid synthesis, bypassing FASII inhibition27,28,29. Thus, the clinical potential of FASII inhibitors is debated due to considerable gaps in our knowledge of the sources of host fatty acids and the mechanisms by which pathogens extract fatty acids from the host27,28. To address these gaps, we developed an unbiased lipidomic analysis method to monitor incorporation of exogenous fatty acid from lipoprotein particles into membrane phospholipids of S. aureus.
During sepsis, host lipoprotein particles represent a potential source of host-derived fatty acids within the vasculature, as a majority of host fatty acids are associated with the particles30. Lipoproteins consist of a hydrophilic shell composed of phospholipids and proteins that enclose a hydrophobic core of triglycerides and cholesterol esters31. Four major classes of lipoproteins–chylomicron, very low-density lipoprotein, high-density lipoprotein, and low-density lipoprotein (LDL)—are produced by the host and function as lipid transport vehicles, delivering fatty acids and cholesterol to and from host cells via the vasculature. LDLs are abundant in esterified fatty acid including triglycerides and cholesterol esters31. We have previously demonstrated that highly purified human LDLs are a viable source of exogenous fatty acids for PG synthesis, thus providing a mechanism for FASII inhibitor bypass32. Purifying human LDLs can be technically challenging and time consuming while commercial sources of purified human LDLs are prohibitively expensive to use on a routine basis or to perform large-scale bacterial screens. To address these limitations, we have modified a procedure for the enrichment of LDLs from chicken egg yolk, a rich source of lipoprotein particles33. We have successfully used untargeted, high-resolution/accurate mass spectrometry and tandem mass spectrometry to monitor incorporation of human LDL-derived fatty acids into the membrane of S. aureus32. Unlike previously reported methods, this approach can quantify individual fatty acid isomers for each of the three major staphylococcal phospholipid types. Oleic acid (18:1) is an unsaturated fatty acid present within all host lipoprotein particles that is readily incorporated into S. aureus phospholipids29,30,32. S. aureus is not capable of oleic acid synthesis29; therefore, the quantity of phospholipid-incorporated oleic acid establishes the presence of host lipoprotein-derived fatty acids within the staphylococcal membrane29. These phospholipid species can be identified by the state-of-the-art mass spectrometry method described here, offering unprecedented resolution of the membrane composition of S. aureus cultured in the presence of a fatty acid source it likely encounters during infection.
NOTE: The following protocol for enrichment of LDL particles from chicken egg yolk is derived from Moussa et al. 200233.
1. Preparation of chicken egg yolk for enrichment of LDL particles
2. Fractionation of LDL-containing plasma from chicken egg yolk
3. Isolation of LDL particles from plasma
4. Assessment of chicken LDLs as a source of fatty acids
5. Incubation of S. aureus with LDLs for membrane lipid analysis.
6. Extraction of S. aureus membrane lipids
7. Analysis of S. aureus lipid profiles using high resolution/accurate mass spectrometry
8. Database searching to identify endogenous S. aureus and exogenous LDL-derived lipids
The protocol for the enrichment of LDL from chicken egg yolk is illustrated in Figure 1. This process begins by diluting whole egg yolk with saline and separating the egg yolk solids referred to as granules from the soluble or plasma fraction containing the LDLs (Figure 1)33. The LDL content of the plasma fraction is further enriched by precipitation of the ~ 30-40 kDa β-livetins (Figure 2)33. The presence of protein bands at 140, 80, 65, 60 and 15 kDa correlate with the apoproteins of LDLs (Figure 2)33,39. Treatment with triclosan inhibits growth of S. aureus in fatty acid-free media32. We have previously demonstrated that supplementing cultures with egg yolk plasma or purified human LDLs as exogenous fatty acid sources overcomes triclosan-induced growth inhibition (Figure 3)32. Similarly, supplementation of triclosan-treated cultures with enriched egg yolk LDL restores growth (Figure 3). Further, addition of egg yolk LDLs support the growth of a previously characterized S. aureus fatty acid auxotroph (Figure 4)32. For the most accurate mass spectrometry-based profiling of S. aureus incorporation of exogenous fatty acids, it is important to limit the presence of free fatty acids in the growth medium. The free fatty acid composition of 1% tryptone broth and chicken egg yolk LDLs diluted in tryptone broth was determined by employing flow injection high-resolution/accurate mass spectrometry and found minimal quantities of free fatty acid (Figure 5). The same untargeted mass spectrometry analysis was performed to determine the fatty acid composition of S. aureus phospholipids after exposure to chicken egg yolk LDLs. Orthogonal partial least-squares discriminant analysis (OPLS-DA)40 of abundant S. aureus membrane phospholipids demonstrated clear class separation of untreated and chicken egg yolk LDL-treated conditions, as shown in the OPLS-DA scores plot (Figure 6A). The OPLS-DA loadings plot indicated numerous phosphatidylglycerol species as important variables in the PLS-DA model. Notably, phospholipids containing unsaturated fatty acids, a molecular marker of exogenous fatty acid incorporation, are enriched in the LDL supplemented cultures compared to cells incubated in the absence of LDLs (Figure 6B). Previous studies have found that chicken egg yolks are a rich source of unsaturated fatty acids with oleic acid (18:1) being the most abundant41,42. In agreement with these observations, we found oleic acid to be the most common unsaturated fatty acid utilized for phospholipid synthesis when S. aureus cultures were supplemented with chicken egg yolk LDLs (Figure 6C). Table 1 further illustrates that the fatty acid profiles of membrane phospholipids are altered when S. aureus is grown in the presence of egg yolk LDL.
Figure 1: An illustration of LDL enrichment from chicken egg yolk utilizing centrifugation and ammonia sulfate precipitation. (A) The reagents necessary for the enrichment of LDL from chicken egg yolk. (B) The flow chart depicts the significant steps of the LDL enrichment process. Please click here to view a larger version of this figure.
Figure 2: Protein profile of chicken egg yolk prior to and after enrichment for LDL. Protein lysates were prepared using RIPA buffer. Protein lysate (15 µg) was loaded into an 8% acrylamide SDS-PAGE gel. Gels were stained overnight with Bio Rad Protein reagent. The molecular weights in kDa of LDL associated proteins are denoted along the right side of the image. M: protein marker, Y: chicken egg yolk, and LDL: chicken egg yolk LDL enrichment Please click here to view a larger version of this figure.
Figure 3: Egg yolk-derived LDLs protect S. aureus from triclosan-induced FASII inhibition. The growth of S. aureus was monitored over time via measurement of OD600 in 1% tryptone broth in the following conditions: 1% tryptone broth (TB), 1 µM triclosan (TCS), 1 µM triclosan with 1% egg yolk plasma (TCS + EYP), 5% egg yolk LDL (LDL), or 1 µM triclosan with 5% egg yolk LDL (TCS + LDL). The mean from three independent experiments is shown. Error bars represent the standard deviation of the mean. Please click here to view a larger version of this figure.
Figure 4: Growth of a S. aureus fatty acid auxotroph is supported by egg yolk-derived LDL. The growth of a fatty acid auxotroph in 1% tryptone broth (TB) with or without 5% egg yolk LDL (LDL) supplementation was monitored over time via measurement of OD600. The mean from three independent experiments is shown. Error bars represent the standard deviation of the mean. Please click here to view a larger version of this figure.
Figure 5: Free fatty acid content measured in 1% tryptone broth or chicken egg yolk LDL. Free fatty acids were detected by flow injection high-resolution/accurate mass spectrometry and tandem mass spectrometry. Normalized numbers of ions per mg of protein was determined for 1% tryptone broth and 1% tryptone broth supplemented with 5% chicken egg yolk LDLs. Please click here to view a larger version of this figure.
Figure 6: Chicken egg yolk low-density lipoproteins are a reservoir of exogenous fatty acids for synthesis of S. aureus phosphatidylglycerol. (A) Scores plot of orthogonal partial least-squares discriminant analysis of chicken egg yolk LDL-treated and untreated S. aureus membrane phospholipids identified using high resolution/accurate mass spectrometry. (B) Percentage of unsaturated phosphatidylglycerol (UPG) compared to total membrane PG of S. aureus grown in the absence (WT) or presence (WT + LDL) of chicken egg yolk LDLs. (C) Unsaturated fatty acid (UFA) profile of membrane PG of S. aureus grown without (WT) or with (WT + LDL) chicken egg yolk LDLs graphed as a percentage of the amount of total PG fatty acids. Please click here to view a larger version of this figure.
WT cultured in tryptone broth | WT cultured in tryptone broth supplemented with LDLs | |||||
Phosphatidyl glycerol (TC:TDB)a | Normalized ion abundance/mg of protein | SD | Fatty acidsc | Normalized ion abundance/mg of protein | SD | Fatty Acidsc |
24:0 | 0 | 0 | NDb | 0.052031116 | 0.02677 | ND |
26:0 | 0 | 0 | ND | 0.009539117 | 0.00362 | ND |
28:0 | 0.127937113 | 0.04528 | 15:0_13:0 | 0.167643281 | 0.02392 | 15:0_13:0 |
28:1 | 0.006765427 | 0.00157 | ND | 0.002776821 | 0.00372 | 15:0_13:1 |
30:0 | 8.680180809 | 2.68375 | 15:0_15:0 | 14.04873592 | 2.4531 | 15:0_15:0 |
30:1 | 0 | 0 | ND | 0.010152161 | 0.00449 | 15:1_15:0, 13:1_17:0 |
31:0 | 4.150511117 | 1.31658 | 16:0_15:0, 14:0_17:0 | 10.17590926 | 1.88431 | 16:0_15:0, 14:0_17:0, 18:0_13:0 |
31:1 | 0.016156004 | 0.01216 | 13:1_15:0, 12:1_19:0 | 0.473478683 | 0.09063 | 13:1_15:0, 18:1_13:0, 12:1_19:0 |
32:0 | 29.29259262 | 8.82993 | 15:0_17:0 | 48.24342037 | 8.95664 | 15:0_17:0, 16:0_16:0 |
32:1 | 0.02074815 | 0.00941 | ND | 0.307044942 | 0.07305 | 18:1_14:0, 16:1_16:0 |
33:0 | 9.000460122 | 2.78194 | 18:0_15:0, 16:0_17:0 | 15.4531776 | 2.98171 | 18:0_15:0, 16:0_17:0 |
33:1 | 0.162934812 | 0.04796 | ND | 2.921832928 | 0.30851 | 18:1_15:0 |
33:2 | 0 | 0 | ND | 0.167492702 | 0.03211 | 18:1_15:1, 18:2_15:0 |
34:0 | 12.3064043 | 3.70242 | 19:0_15:0, 17:0_17:0 | 18.40129157 | 3.21385 | 19:0_15:0, 17:0_17:0 |
34:1 | 0 | 0 | ND | 1.423605186 | 0.20066 | 18:1_16:0 |
34:2 | 0.000470922 | 0.00082 | ND | 0.156133734 | 0.03929 | 18:2_16:0 |
35:0 | 5.727462455 | 1.74583 | 20:0_15:0, 18:0_17:0 | 7.771538992 | 1.28515 | 20:0_15:0, 16:0_19:0, 18:0_17:0 |
35:1 | 0.17337586 | 0.05727 | 20:1_15:0 | 0.772202525 | 0.08526 | 20:1_15:0, 18:1_17:0 |
35:2 | 0 | 0 | ND | 0.038758757 | 0.01481 | 18:2_17:0, 18:1_17:1 |
36:0 | 0.671004303 | 0.2116 | 21:0_15:0, 19:0_17:0 | 0.967295024 | 0.2572 | 21:0_15:0, 20:0_16:0, 19:0_17:0, 22:0_14:0 |
36:2 | 0 | 0 | ND | 0.495485065 | 0.04473 | 18:1_18:1, 18:2_18:0 |
36:3 | 0 | 0 | ND | 0.059268233 | 0.02291 | 18:2_18:1, 20:3_16:0, 20:2_16:1 |
37:0 | 0.060466411 | 0.01961 | 22:0_15:0, 20:0_17:0 | 0.114526894 | 0.01852 | 22:0_15:0, 20:0_17:0, 18:0_19:0 |
38:2 | 0 | 0 | ND | 0.079469521 | 0.02872 | 18:2_20:0, 16:1_20:1 |
aDetected as [M-H]– ions. TC, total chain length; TDB, total number of double bonds. | ||||||
bND, not determined | ||||||
cFatty acids are listed in order of isomer abundance. An underscore between fatty acid designations indicates that each fatty acid may be present in either the SN1 or SN2 position, as tandem mass spectrometry alone cannot rule out the possibility that lipid species exist as a mixture of positional isomers. |
Table 1: Fatty acid profile of S. aureus cultured in the presence of chicken egg yolk LDLs. We used an unbiased lipidomic analysis utilizing high-resolution/accurate MS and MS/MS to determine the fatty acid profile of S. aureus PG. S. aureus was incubated in the presence or absence of chicken egg yolk LDLs, and the PG profile of these cells was compared to that of cells cultured in 1% tryptone broth.
S. aureus incorporates exogenous fatty acids into its membrane phospholipids27,32,43. Phospholipid synthesis using exogenous fatty acids bypasses FASII inhibition but also alters the biophysical properties of the membrane27,32,44. While incorporation of exogenous fatty acids into phospholipids of Gram-positive pathogens is well documented, gaps remain in the identity of host fatty acid reservoirs and the structural alterations to each of the three major staphylococcal phospholipid types that result from exogenous fatty acid incorporation. Here, we describe protocols which can be employed to: i) enrich LDL particles from chicken egg yolk, a source of fatty acids, ii) determine the effects of exogenous fatty acids on the growth of S. aureus, and iii) utilize an unbiased lipidomic analysis for monitoring exogenous fatty acid incorporation into membrane phospholipids of S. aureus. The advanced mass spectrometry method provided in this study offers an extraordinary perspective of the membrane composition of S. aureus grown in the presence of exogenous fatty acids.
Several Gram-positive pathogens utilize exogenous fatty acids for membrane synthesis and, as with S. aureus, the possible sources of exogenous fatty acids during infection are poorly understood27,43. The growth analysis described here can be modified to evaluate the proliferation of other Gram-positive pathogens in the presence of lipoproteins if the growth kinetics of each pathogen are considered. Additionally, other complex host sources of exogenous fatty acids could be tested using this protocol if the potential effects of the fatty acid source on background optical density are controlled. Moreover, the described mass spectrometry method for analysis of bacterial lipids is sufficiently flexible to enable lipidome evaluation from virtually any bacterial species. As lipid accurate mass data is collected in ‘full scan’ MS mode, little a priori knowledge of the lipid content of the bacterial species of interest is required, unlike targeted analytical methods based on known fragmentation patterns of specific lipids32,45,46. In the ‘untargeted’ analytical workflow we describe, the downstream data analysis, and particularly the searching of accurate mass peak lists against a lipid database, are key steps that are highly adaptable and may support a broad range of bacterial species and experimental treatments. When constructing or choosing a searchable database to enable identification of lipid species, researchers must consider a wide range of hypothetical endogenous lipid species, while also allowing for detection of novel or unforeseen exogenous lipids derived from the experimental treatment.
In the present study, a high resolution/accurate mass spectrometer (Table of materials) was employed due to its ultra-high resolution/accurate mass capabilities. Alternatively, numerous other high resolution/accurate mass spectrometry platforms could be successfully implemented to perform untargeted lipid analysis. Similarly, a wide range of sample introduction methods including direct infusion, desorption electrospray ionization, or matrix-assisted laser desorption ionization, that enable direct analysis of lipid extracts, could be utilized to rapidly collect untargeted lipidomic data. The inclusion of liquid chromatography prior to sample introduction, when used in combination with high resolution/accurate mass spectrometry, may permit the resolution of some isobaric lipid species during full-scan MS data collection. However, the inclusion of chromatography necessitates ensuring that the chromatographic method of choice is versatile enough to enable separation and detection of unanticipated or novel lipid species that may be present following experimental treatments. Database searching to identify lipids present in the dataset may be performed using any publicly available searchable database. While LIMSA software enables facile development of user-defined databases of tens of thousands of hypothetical lipid species, numerous other options exist for identifying lipids from high resolution/accurate mass MS peak lists. The LIPID MAPS consortium (www.lipidmaps.org) provides tools for searching computational and experimental databases of hypothetical lipids using high resolution/accurate MS-generated peak lists within a user-defined mass tolerance, and many software vendors offer their own solutions for analyzing lipidomics data.
Successful growth curve and exogenous fatty acid analysis is dependent on several factors including LDL purity and limiting background fatty acid levels. Proper identification of the LDL-containing fraction is crucial. The above protocol and Figure 1 illustrate the correct fraction to retain for each step of the enrichment process. We have had success with the use of 40% ammonium sulfate (purity ≥99.5%) for the precipitation and subsequent removal of β-livetins. However, others have reported that the purity and concentration of ammonia sulfate added to the egg yolk plasma can significantly impact this step47. Limiting ammonium sulfate contamination in the LDL enrichment is important for the LDL preparation to be used in bacterial assays, as high concentrations of ammonium sulfate can restrict growth48. During dialysis, ample free space in the dialysis tubing must be provided to allow for the diffusion and removal of the ammonium sulfate. Further optimization of the dialysis process may include additional water changes during the overnight incubation. We have found overnight dialysis to provide the best results, although Moussa et al. report dialysis of 6 h is sufficient. It is critical that the starting concentration of cells in the growth curves are kept consistent between trials. For S. aureus, diluting cells to an initial OD600 of 0.05 has provided the most consistent results. Additionally, high concentrations of FASII inhibitors can result in non-specific effects on bacterial cells. For example, triclosan concentrations above 7 µM induce cytoplasmic membrane damage in S. aureus, therefore it is necessary that the concentration of this compound remain below this level49. We have found a final triclosan concentration of 1 µM results in reproducible growth assays. When evaluating a potential source of exogenous fatty acids for bacterial phospholipid synthesis, it is important to minimize the fatty acid contribution of the culture medium. In the above assays, the culture medium of 1% tryptone supports adequate growth of S. aureus and has minimal fatty acid contamination29,32.
Limiting background fatty acid levels is particularly important for downstream mass spectrometry-based fatty acid profiling. Others have reported the quantity of free fatty acids in chicken egg yolk is naturally low41 and our analysis supports this conclusion (Figure 5). Using tryptone broth and thoroughly washing cells with PBS after incubation are essential. Additionally, it is important to consider the growth phase of the cells. We chose mid-log phase cells to ensure ample bacterial phospholipid synthesis. Other potential contaminating sources of exogenous fatty acids may be introduced after bacterial growth, such as during the lipid extraction steps or subsequent sample preparation prior to mass spectrometry analysis50. Exogenous fatty acids introduced at any step of sample preparation could be detected as free fatty acids during mass spectrometry analysis. Baking laboratory glassware in a high temperature oven (at least 180 °C) overnight can remove exogenous fatty acids from test tubes used for lipid extraction and lipid storage. Additionally, laboratory supplies including plastics may be rinsed with methanol to reduce fatty acid background50. Residual fatty acids from previous analyses may also contaminate internal surfaces of the mass spectrometer itself. Inclusion of analytical blanks during mass spectrometry analysis for determination of background levels of free fatty acids is therefore strongly advised.
The authors have nothing to disclose.
We thank members of the Hammer laboratory for their critical evaluation of the manuscript and support of this work. Dr. Alex Horswill of the University of Colorado School of Medicine kindly provided AH1263. Dr. Chris Waters laboratory at Michigan State University provided reagents. This work was supported by American Heart Association grant 16SDG30170026 and start-up funds provide by Michigan State University.
Ammonium sulfate | Fisher | BP212R-1 | ≥99.5% pure |
Cell culture incubator | Thermo | MaxQ 6000 | |
Centrafuge | Thermo | 75-217-420 | Sorvall Legen XTR, rotor F14-6×250 LE |
Costar assay plate | Corning | 3788 | 96 well |
Filter paper | Schleicher & Schuell | 597 | |
Large chicken egg | N/A | N/A | Common store bought egg |
Microplate spectrophotometer | BioTek | Epoch 2 | |
NaCl | Sigma | S9625 | |
S. aureus strain AH1263 | N/A | N/A | Provided by Alex Horswill of the University of Colorado |
Dialysis tubing | Pierce | 68700 | 7,000 MWCO |
Tryptone | Becton, Dickison and Company | 211705 | |
0.5 mm zirconium oxide beads | Next Advance | ZROB05 | |
Bullet Blender | Next Advance | BBX24B | |
Methanol (LC-MS grade) | Fisher | A4561 | |
Chloroform (reagent grade) | Fisher | MCX10559 | |
Isopropanol (LC-MS grade) | Fisher | A4611 | |
Dimyristoyl phosphatidylcholine | Avanti Polar Lipids | 850345C-25mg | |
Ammonium bicarbonate | Sigma | 9830 | ≥99.5% pure |
Ammonium formate | Sigma | 70221-25G-F | |
Xcalibur software | Thermo Scientific | OPTON-30801 | |
LTQ-Orbitrap Velos mass spectrometer | Thermo Scientific | high resolution/accurate mass MS | |
Agilent 1260 capillary HPLC | Agilent | ||
SpeedVac Vacuum Concentrators | Thermo Scientific |