A Tandem Liquid Chromatography–Mass Spectrometry-based Approach for Metabolite Analysis of Staphylococcus aureus

Summary

Here we describe a protocol for the extraction of metabolites from Staphylococcus aureus and their subsequent analysis via liquid chromatography and mass spectrometry.

Abstract

In an effort to thwart bacterial pathogens, hosts often limit the availability of nutrients at the site of infection. This limitation can alter the abundances of key metabolites to which regulatory factors respond, adjusting cellular metabolism. In recent years, a number of proteins and RNA have emerged as important regulators of virulence gene expression. For example, the CodY protein responds to levels of branched-chain amino acids and GTP and is widely conserved in low G+C Gram-positive bacteria. As a global regulator in Staphylococcus aureus, CodY controls the expression of dozens of virulence and metabolic genes. We hypothesize that S. aureus uses CodY, in part, to alter its metabolic state in an effort to adapt to nutrient-limiting conditions potentially encountered in the host environment. This manuscript describes a method for extracting and analyzing metabolites from S. aureus using liquid chromatography coupled with mass spectrometry, a protocol that was developed to test this hypothesis. The method also highlights best practices that will ensure rigor and reproducibility, such as maintaining biological steady state and constant aeration without the use of continuous chemostat cultures. Relative to the USA200 methicillin-susceptible S. aureus isolate UAMS-1 parental strain, the isogenic codY mutant exhibited significant increases in amino acids derived from aspartate (e.g., threonine and isoleucine) and decreases in their precursors (e.g., aspartate and O-acetylhomoserine). These findings correlate well with transcriptional data obtained with RNA-seq analysis: genes in these pathways were up-regulated between 10- and 800-fold in the codY null mutant. Coupling global analyses of the transcriptome and the metabolome can reveal how bacteria alter their metabolism when faced with environmental or nutritional stress, providing potential insight into the physiological changes associated with nutrient depletion experienced during infection. Such discoveries may pave the way for the development of novel anti-infectives and therapeutics.

Introduction

Bacterial pathogens must contend with many challenges within the host environment. In addition to direct attack by immune cells, the host also sequesters nutrients essential for bacterial survival and replication, generating nutritional immunity^1,2. To survive these hostile environments, bacterial pathogens deploy virulence factors. Some of these factors allow the bacteria to evade the immune response; other factors include secreted digestive enzymes, such as hyaluronidase, thermonuclease, and lipase, which may enable the bacteria to replenish missing nutrients by consuming tissue-derived constituents^3,4,5. Indeed, bacteria have evolved regulatory systems that tie the physiological state of the cell to the production of virulence factors^6,7,8,9,10.

A growing body of evidence points to CodY as a critical regulator linking metabolism and virulence. Although first discovered in Bacillus subtilis as a repressor of the dipeptide permease (dpp) gene¹¹, CodY is now known to be produced by nearly all the low G+C Gram-positive bacteria^12,13 and regulates dozens of genes involved in carbon and nitrogen metabolism^{14,15,16,17,18,19}. In pathogenic species, CodY also controls the expression of some of the most important virulence genes^{20,21,22,23,24,25,26,27}. CodY is activated as a DNA-binding protein by two classes of ligands: branched-chain amino acids (BCAAs; isoleucine, leucine, and valine [ILV]) and GTP. When these nutrients are abundant, CodY represses (or in some cases, stimulates) transcription. As these nutrients become limited, CodY activity is progressively reduced, resulting in a graded transcriptional response that re-routes precursors through various metabolic pathways connected to central metabolism^28,29,30.
Tandem liquid chromatography coupled to mass spectrometry (LC-MS) is a powerful technique that can accurately identify and quantify small-molecule intracellular metabolites³¹. When paired with transcriptome analysis (e.g., RNA-Seq), this analytical workflow can provide insight into the physiological changes that occur in response to environmental or nutritional stress. Here, we present a method for metabolite extraction from Staphylococcus aureus cells and subsequent analysis via LC-MS. This approach has been used to demonstrate the pleiotropic effects of CodY on S. aureus physiology.

Protocol

1. Preparation of Buffer Solutions

1. Prepare phosphate-buffered saline (PBS; pH 7.4) by diluting a stock solution of 10x PBS to a final concentration of 1x with ultrapure (distilled and deionized) water.
2. Prepare quenching solution by combining 2 mL of acetonitrile, 2 mL of methanol, 1 mL of ultrapure H₂O, and 19 µL (0.1 mM final concentration) of formic acid.
3. Prepare LC-MS solvent A by adding formic acid (0.2% [v/v] final concentration) to ultrapure water.
4. Prepare LC-MS solvent B by adding formic acid (0.2% [v/v] final concentration) to acetonitrile.
   NOTE: All solutions should be prepared using the highest-purity reagents available (generally high-performance liquid chromatography grade). Solutions should be prepared fresh before each experiment and stored on ice prior to use.

2. Establishment of Steady-state S. aureus Growth

1. Streak S. aureus strains of interest for isolation on tryptic soy agar (TSA) from a frozen glycerol stock. Incubate at 37 °C for 16-24 h.
2. Inoculate 4 mL of tryptic soy broth (TSB) or another suitable medium in sterile glass incubation tubes with single colonies of each strain. Incubate inclined (~70° angle) with rotation at 60 rotations per min (rpm) at 37 °C for 16-20 h.
   NOTE: Overnight cultures are prone to oxygen gradients when using standard methods, including those described in step 2.2, which affect cellular physiology. Thus, we employ a multiple back-dilution strategy to ensure biological steady state (see steps 2.4-3.2, below).
3. Use a spectrophotometer to measure the optical density of the cultures from step 2.2 at 600 nm (OD₆₀₀). Use sterile medium as an optical reference (blank). Dilute these cells to an OD₆₀₀ of 0.05 in 50 mL of sterile TSB medium (pre-warmed to 37 °C) in separate, 250 mL DeLong flasks.
4. Incubate the cultures at 37 °C in a water bath with shaking at 280 rpm.
5. Every 30 min, take OD₆₀₀ measurements; as the optical densities increase, it may become necessary to dilute the cultures with TSB so that they remain within the linear absorbance range of the spectrophotometer.
6. When cultures from step 2.5 achieve an OD₆₀₀ of ~0.8-1.0, subculture them into 50 mL of 37 °C TSB to an OD₆₀₀ of 0.01-0.05 and repeat steps 2.4 and 2.5.

3. Sample Collection Setup

1. Prepare a bed of crushed dry ice in an appropriate vessel (e.g., glass dish, ice bucket, or cooler).
2. As the optical densities of the cultures approach the desired harvest point, add 1 mL of quenching solution to a 35 mm untreated Petri dish and pre-cool on dry ice for ≥5 min.
   NOTE: The "desired harvest point" will vary depending on experimental goals. For example, if one were to examine metabolites during aerobic growth, key indicators of this state include acetate excretion and re-assimilation of the acetate during the post-exponential growth phase^32,33. Generally, this point should be within a specific growth stage (e.g., exponential phase). The specific OD₆₀₀ values associated with this stage may vary between different bacterial strains and growth media.
3. Place a stainless steel filter frit (pre-chilled to -20 °C) in a rubber stopper and place it atop a vacuum flask attached to a house vacuum or vacuum pump.
4. Apply the vacuum and place a mixed cellulose ester membrane (0.22 µm pore size) on top.
   NOTE: It is critical to use a filter with a diameter equal to that of the frit and to properly center this filter to ensure that the sample is drawn through the filter rather than over the edge. Wetting the membrane with ice-cold, sterile H₂O may help with positioning the membrane.

4. Sample Harvest

1. At an OD₆₀₀ of ~0.4-0.5, use a serological pipette to remove 13 mL of culture from the flask and to apply the sample to the filter.
2. After the entire sample has been filtered, immediately wash the filter with ≥5 mL of ice-cold PBS to wash away medium-associated metabolites.
3. Disconnect the vacuum and use a pair of sterile tweezers to remove the filter from the frit. Invert the filter (cell-side down) into the pre-chilled quench solution.
   NOTE: It is important to perform the above steps quickly (i.e., within seconds) and as soon as the liquid has been removed to ensure the rapid quenching of the cells, arresting metabolic activity.
4. Incubate the filter in quench solution on dry ice for ≥20 min.
5. Using sterile tweezers, invert the filter (cell-side up) in the petri dish and use a micropipette to rinse the cells off of the membrane into the quench solution.
6. Re-suspended the cells in quench solution and then transfer the cell suspension to a sterile 2 mL impact-resistant tube containing ~100 µL of 0.1 mm silica beads. Store this on dry ice or at -80 °C.

5. Metabolite Extraction

1. Thaw samples on wet ice and disrupt the cells in a homogenizer with four 30 s bursts at 6,000 rpm, with 2 min cooling periods on dry ice between cycles.
2. Clarify the lysates for 15 min in a pre-chilled, refrigerated microcentrifuge at maximum speed (i.e., 18,213 x g at ≤4 °C).
3. Transfer the supernatant to a clean microcentrifuge tube.
4. Using a micropipette, transfer a small portion of the sample to a microcentrifuge tube for the quantification of residual peptide content in step 6; store the remainder at -80 °C.
   NOTE: The volume of reserved sample varies, depending on the BCA assay used in step 6.1. This sample should be stored on wet ice for immediate analysis or frozen at -80 °C.

6. Bicinchoninic Acid (BCA) Assay

1. Perform a BCA assay as recommended by the kit manufacturer, using samples from step 5.4 to determine the residual peptide concentration for each sample.

7. LC-MS

1. Mix 75 µL of S. aureus extract with 75 µL of LC-MS solvent B, prepared in step 1.4.
2. Vortex to mix and spin at 13,000 x g for 5 min.
3. Place 100 µL of supernatant into a liquid chromatography (LC) vial and cap it. Ensure that no air bubbles are trapped in the sample.
4. Load the LC vials onto the LC-MS autosampler and edit the running list in the software "Offline Worklist Editor."
   1. Fill out the "Sample Name" (e.g., wild-type-1), "Sample Position" (e.g., P1-A1), "Method" (e.g., Formic Acid-Negative Method), and "Data File" (e.g., wild-type-1) columns. Click the button "Save Worklist" button. Open the "Mass Spectrometry Data Acquisition Workstation" software and input the previously saved worklist. Click the "Start Worklist Run" button to start the continuous LC-MS measurement.
5. Separate the samples on a column, link the column to a time of flight (TOF) spectrometer, and couple the TOF spectrometer with the LC system. Use a mobile-phase gradient as follows: 0-2 min, 85% solvent B; 3-5 min, 80% solvent B; 6-7 min, 75% solvent B; 8-9 min, 70% solvent B; 10-11.1 min, 50% solvent B; 11.1-14 min, 20% solvent B; and 14.1-24 min, 5% solvent B; end with a 10 min re-equilibration period at 85% solvent B and a flow rate of 0.4 mL min^-1.
6. Using an isocratic pump, infuse a reference mass solution with the run to allow for simultaneous mass axis calibration.
   NOTE: This step is based on the standard TOF spectrometer manual.
   1. Use the mixture of acetic acid D4 and hexakis (1H,1H,3H-tetrafluoropropoxy) phosphazine as the reference mass solution to perform the real-time calibration. Use the isocratic pump with the flow rate of 2.5 mL min^-1 for the infusion.

8. Batch Correction of Ion Counts

1. Designate any sample to serve as a reference sample for batch correction (e.g., wild-type, replicate 1).
2. Calculate the sum of the ion counts for all metabolites within the reference sample. Repeat this calculation for all samples.
3. Divide the total ion count of each sample by the total ion count of the reference sample to generate a ratio.
4. Divide the ion count for each metabolite within a sample by the sample/reference ratio to obtain a batch-corrected ion count for each metabolite.

9. Peptide Normalization

1. Divide the batch-corrected ion count values for each sample obtained in step 8 by the peptide concentration determined with the BCA assay in step 6 to yield a normalized value for each metabolite.
   NOTE: The normalized, batch-batch corrected ion counts for each metabolite obtained in step 9.1 can be directly compared between strains and subjected to statistical analysis (e.g., a Mann-Whitney U-test). Alternatively, a metabolite known to be unchanged either by the treatment or genetic background may be used as a normalizer to detect changes due to metabolite decomposition. The inclusion of a known amount of L-norvaline or gluraric acid in the extraction buffer can be used to correct for loss during sample processing³⁴.

Results

We have analyzed intracellular metabolite pools in S. aureus during in vitro growth in a rich, complex medium. As proof of principle, we compared metabolite profiles between the methicillin-susceptible S. aureus osteomyelitis isolate UAMS-1 (wild-type [WT]) and an isogenic strain lacking the global transcriptional regulator CodY (ΔcodY)²⁶. Steady-state, exponential cultures of the WT and codY strains were establishe...

Discussion

All small-molecule metabolites are connected to one another through their common origins in central metabolic pathways. During exponential growth, bacterial cells are at biological and metabolic steady state, providing a snapshot of the physiological state under specific conditions. CodY monitors nutrient sufficiency by responding to ILV and GTP. As ILV and GTP pools drop, CodY activity is likely progressively reduced, adjusting the expression of its target genes to adapt to increasing nutrient depletion

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Material/Equipmenta
DeLong Culture Flask (250 mL)	Belco	2510-00250
Sidearm Flask, 500 mL	Pyrex	5340
3-hole Rubber Stopper, #7	Fisher	14-131E
Stainless Steel Filter holder/frit	VWR	89428-936
Petri Dish, 35 mm	Corning	430588	Not tissue culture treated
Mixed cellulose ester membrane, 0.22 μm pore size	Millipore	GSWP02500
Impact-resistant tubes, 2 mL	USA Scientific	1420-9600
Silica Beads, 0.1 mm	Biospec Products Inc	11079101Z
Precellys 24 homogenizer	Bertin Instruments	EQ03119-200-RD000.0
Micro BCA Protein Assay Kit	Pierce (Thermo Scientific)	23235
Cogent Diamond hydride type C column	Agilent	70000-15P-2
Accurate-Mass Time-of-Flight (TOF) LC-MS, 6200 Series	Agilent	G6230B
Quat Pump, 1290 Series	Agilent	G4204A
Bin Pump, 1290 Series	Agilent	G4220A
Valve Drive, 1290 Series	Agilent	G1107A
Isocratic Pump, 1290 Series	Agilent	G1310B
TCC, 1290 Series	Agilent	G1316C
Sampler, 1290 Series	Agilent	G4226A
Thermostat, 1290 Series	Agilent	G1330B
Chemical
Tryptic Soy Broth	Becton Dickinson	211825
Difco Agar, Granulated	Becton Dickinson	214530	Solid media contains 1.5% [w/v] agar
Phosphate-buffered saline (pH 7.4) 10x	Ambion	AM9624	Dilute fresh to 1x with ultra-pure water
Acetonitrile	Fisher Scientific	A955-500	Optima LC-MS
Methanol	Fisher Scientific	A456-500	Optima LC-MS; toxic
Formic Acid	Sigma Aldrich	94318	For mass spectrometry, 98%
Software
MassHunter	Agilent	G3337AA
Bacterial Strain	Species	Strain	Genotype
SRB 337	Staphylococcus aureus	USA200 MSSA UAMS-1	wild type
SRB 372	Staphylococcus aureus	USA200 MSSA UAMS-1	ΔcodY::erm
aChemicals and materials listed are specific to the method described and do not include standard laboratory chemicals or supplies.