Global Metabolomic Analysis of a Mammalian Host Infected withBacillus anthracis

Chinh T. Q. Nguyen,a Vivekananda Shetty,b Anthony W. Maressoa

Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas, USAa; Alkek Center for Molecular Discovery, Baylor College of Medicine,Houston, Texas, USAb

Whereas DNA provides the information to design life and proteins provide the materials to construct it, the metabolome can beviewed as the physiology that powers it. As such, metabolomics, the field charged with the study of the dynamic small-moleculefluctuations that occur in response to changing biology, is now being used to study the basis of disease. Here, we describe a com-prehensive metabolomic analysis of a systemic bacterial infection using Bacillus anthracis, the etiological agent of anthrax dis-ease, as the model pathogen. An organ and blood analysis identified approximately 400 metabolites, including several key classesof lipids involved in inflammation, as being suppressed by B. anthracis. Metabolite changes were detected as early as 1 daypostinfection, well before the onset of disease or the spread of bacteria to organs, which testifies to the sensitivity of this method-ology. Functional studies using pharmacologic inhibition of host phospholipases support the idea of a role of these key enzymesand lipid mediators in host survival during anthrax disease. Finally, the results are integrated to provide a comprehensive pic-ture of how B. anthracis alters host physiology. Collectively, the results of this study provide a blueprint for using metabolomicsas a platform to identify and study novel host-pathogen interactions that shape the outcome of an infection.

Metabolomics, a quickly emerging “omics” field in systemsbiology, is the global analysis of small molecules in a biolog-

ical sample (1). Since the 1950s, the central dogma of biologicalinformation has been the transition from genes to transcript toprotein (2). Only recently has the use of high-throughput systemsbiology been used to study the products of protein activity, themetabolites. The metabolome represents all endogenous and ex-ogenous low-molecular-mass (�1-kDa) molecules present in abiological state, providing an instantaneous “snapshot” of thecell’s metabolic and physiological activity (3). Applications ofglobal metabolomic analysis fall into three broad categories: (i)disease diagnosis, (ii) biomarker and drug discovery, and (iii)study of metabolic pathways and their perturbations due to exter-nal factors (1). The analysis of altered metabolites has the potentialfor discovery of new biomarkers, thus providing the possibilityof earlier intervention and insights into the mechanisms ofdiseases (4).

The diagnosis of a bacterial infection encompasses an assess-ment of clinical symptoms, positive culture of an organism fromtissues or blood, and/or reliance on often expensive, outsourcedmolecular methods (5). Metabolomics provides a unique perspec-tive on bacterial infections as it is able to comprehensively char-acterize a vast number of metabolic changes in response to abiological perturbation within the host (2). In addition to thepotential for biomarker discovery, the metabolic profiles obtainedfrom this analysis can give insight into the identity and nature ofmolecules involved in the immune response, detect alterations inhost physiology, and identify novel pathways altered during infec-tion (6). The uses of metabolomics are quickly emerging in bothclinical and basic research settings to address fundamental ques-tions of bacterial pathogenesis. Efforts in clinical research havefocused on biomarker discovery and refining diagnostic methods,as these endeavors serve as the beginning steps toward personal-ized medicine (7, 8). In particular, for sepsis infections, metabo-lomics has been used to understand the dynamic physiologicalchanges in individual patient metabolic profiles as an alternative

to treating them as homogeneous populations (7–9). On the otherhand, basic research studies focused on identifying key metabolicpathways have given insight into physiological changes in specifictissues and/or host systems (10–12). However, a comprehensiveanalysis of major host organs and tissues has not yet been per-formed.

In this study, we utilized Bacillus anthracis as a model pathogento investigate the metabolic changes present during various stagesof infection. B. anthracis is a spore-forming bacterium that is theetiological agent of anthrax and a weapon of bioterrorism (13, 14).Upon exposure to the host, spores are engulfed by local macro-phages, where they germinate into vegetative cells and replicate asthe macrophages travel to lymph nodes (15–18). Vegetative bacillithen escape the cell and produce key virulence factors that con-tribute to the manifestation of disease. These include anthraxtoxin, a tripartite toxin system that consists of one receptor bind-ing component, protective antigen (PA), and two catalytic sub-units, lethal factor and edema factor (LF and EF) (19–21). LFexhibits metalloprotease activity, cleaving mitogen-activated pro-tein kinase kinases (MAPKKs) of the MAPK signaling pathwayand suppressing subsequent proinflammatory responses (20–23).EF acts as an adenylyl cyclase that converts ATP to cyclic AMP

Received 22 July 2015 Returned for modification 4 September 2015Accepted 23 September 2015

Accepted manuscript posted online 5 October 2015

Citation Nguyen CTQ, Shetty V, Maresso AW. 2015. Global metabolomic analysisof a mammalian host infected with Bacillus anthracis. Infect Immun 83:4811–4825.doi:10.1128/IAI.00947-15.

Editor: S. R. Blanke

Address correspondence to Anthony W. Maresso, maresso@bcm.edu.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00947-15.

Copyright © 2015, American Society for Microbiology. All Rights Reserved.

December 2015 Volume 83 Number 12 iai.asm.org 4811Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

(cAMP), thus increasing intracellular cAMP concentrations,which contribute to disrupting cytokine production and mediat-ing tissue destruction (24). Vegetative bacteria quickly spread intothe lymphatic and circulatory systems, initiating a systemic infec-tion (25). The propensity of this bacterium to transition from aninactive spore to a disseminated infection with multiorgan in-volvement that culminates in massive bacteremia and toxemiaoffers a unique opportunity to assess the small-molecule metabo-lome during a developing infection (26). Here, we report the firstwhole-organism metabolomic analysis of mice infected with B.anthracis, a bacterial pathogen that causes systemic disease, usingboth spore and vegetative cell infection. A global analysis of wholeblood and organs revealed bacillus-induced alterations of energyproduction and lipid mediators, the latter of which are largelysuppressed. Functional studies using pharmacological inhibitionof host phospholipase A2 (PLA2) enzymes further invoked a dis-ruption of lipid mediators during anthrax disease. This work dem-onstrates the usefulness of metabolomics in the identification andanalysis of small molecules involved in the innate host response toa bacterial infection.

MATERIALS AND METHODSMouse infections. A global metabolomic analysis of mice was used toassess the effect of B. anthracis infection on host metabolism. Female A/Jmice (6 to 8 weeks old) (n � 5 per experimental group) were purchasedfrom Jackson Laboratories. The Sterne 34F2 strain was used for eachmetabolomics experiment. This strain lacks the pOX2 plasmid and doesnot produce capsule (15). Spores were made from vegetative cells accord-ing to the protocol described by Kim and Goepfert (27). Vegetative cellswere generated from freezer stocks of strain 34F2. First, the 50% lethaldose (LD50) was determined by administering phosphate-buffered saline(PBS) suspensions of B. anthracis Sterne 34F2 vegetative cells or sporessubcutaneously into the left hind leg of mice. All subcutaneous inocula-tions in this study were performed as 50-�l injections into the subcutane-ous fatty layer of the ventral side of the right hind leg near the last mam-mary gland of female mice. Inoculum doses ranged from 1 � 103 to 1 �105 (vegetative cells) and from 2 � 101 to 2 � 104 (spores) bacilli perinoculum. Mice were monitored at 12-h intervals for 1 week. The LD50swere calculated by the Reed and Muench method and were determined tobe 1.5 � 103 for vegetative cells and 1 � 102 for spores (see Fig. S1A and Bin the supplemental material) (28). The median times to death (MTD)were calculated from the shortest survival time in which percent survivalwas less than or equal to 50%. Once the LD50 was determined, infections(performed using either vegetative cells or spores) were repeated usingthis dose, and either blood samples (both vegetative and spore infections)or organ samples (spores only) were harvested at 1 and 3 days postinfec-tion (dpi).

Ethics statement. All animal protocols for these studies were ap-proved by the Baylor College of Medicine Institutional Animal Use andCare Committee (BCM-IACUC; Animal Assurance no. AN-5177; proto-col no. AN-5177:1; IACUC protocol no. D1485 and D1491) and compliedwith the Public Health Service Policy on Humane Care and Use of Labo-ratory Animals and the regulations under the Animal Welfare Act (USDAregistration no. A3823-01).

Sample collection and preparation for metabolomics studies. Globalmetabolomic analyses were conducted using whole blood and organs ofA/J mice infected with B. anthracis. Specifically, the vegetative cell infec-tion analysis utilized whole blood collected via cardiac puncture (with 4.5mM EDTA added to prevent clotting) from control and infected mice at 1and 3 dpi (n � 8 per experimental group). One of the mice in each groupwas analyzed by the investigators in a blind manner to determine whetherthe metabolomic changes that were identified were predictive of the typeand relative stage of infection. The spore infection utilized whole bloodand organ tissues from lungs, liver, kidneys, and spleen harvested from

animals at 3 dpi. A total of 12 mice, 6 infected and 6 control, were used. Toreduce the effect of the presence of metabolites originating from blood inconfounding metabolic profiles of organ tissues, transcardiac perfusionswere performed on anesthetized mice to flush out blood from the vascu-lature, as described in reference 29. Briefly, animals were first anesthetizedby intraperitoneal (i.p.) administration of a lethal dose of a 1.2% solutionof 2,2,2-tribromoethanol–2-methyl-2-butanol– 0.9% NaCl at a final con-centration of 0.2 ml/0.01 kg body weight and transcardiac perfusions wereperformed. To ensure that animals were properly sedated before perfu-sion procedures were performed, each mouse was tested for physical re-sponsiveness by pedal reflex determination. Sedated animals were placedon operating platforms located in a biosafety cabinet. A horizontal inci-sion was made below the ribcage, and two vertical incisions were madeinto the skin, each above the ends of the ribcage, leaving the thoracic cavityexposed. Using a peristaltic pump, a butterfly needle with a Luer-lok at-tachment was secured to the efflux end of the pump. The influx end wasplaced in a reservoir of cold sterile PBS with 5 mM EDTA. A needle wasplaced into the left ventricle of the heart, a small, 3-mm-long incisionwas made at the right atrium, and the pump was allowed to run for 5 minper animal to flush out circulating blood from tissues. In addition, theheart was excised to ensure exsanguination. Perfused organs were har-vested after perfusions were completed. Perfusion efficiencies were ana-lyzed as a proxy for the amount of whole blood still present in tissuevasculature after perfusions by determining the average optical density(OD) at 403 nm per gram of perfused tissues in comparison to nonper-fused tissues, a value that is directly proportional to the amount of hemo-globin present (see Fig. S3 in the supplemental material) (30).

Metabolomic profiling. A global metabolomics approach was per-formed using a pipeline developed by Metabolon Inc. Samples for analy-sis, either from blood or organs, were prepared using a solvent extractionmethod (MicroLab Star system from the Hamilton Company) as previ-ously described (31). Briefly, the resulting extract was split into equal partsand applied to gas chromatography/mass spectrometry (GC/MS) and liq-uid chromatography tandem MS (LC/MS/MS) platforms. The LC/MSportion of the platform is based on a Waters Acquity ultraperformance LC(UPLC) and Thermo-Finnigan linear trap quadrupole (LTQ) mass spec-trometer, consisting of an electrospray ionization source and linear iontrap mass analyzer. For GC/MS, samples were derivatized and separatedusing a 5% phenyl column with a temperature ramp from 40° to 300°C ina 16-min period. Samples were then analyzed on a Thermo-FinniganTrace DSQ fast-scanning single-quadrupole mass spectrometer usingelectron impact ionization. All compounds were identified by their reten-tion time and mass after comparison to library entries of purified stan-dards in Metabolon’s database of �5,000 molecules.

ELISA for the detection of B. anthracis lethal factor. An indirectenzyme-linked immunosorbent assay (ELISA) for the detection andquantification of lethal factor was adapted from past studies (32–34).Serum samples from control and infected A/J mice were normalized to 3mg/ml total protein of bovine serum albumin (BSA) by a Bradford assay.Briefly, a standard curve was generated with known concentrations ofBSA and used to determine the total concentration of samples. In doingso, a consistent serum concentration normalized to 3 mg/ml was main-tained across all samples. Samples were added in duplicate to Immulon(Thermo-Scientific) plates and incubated overnight at 4°C. LF seriallydiluted in carbonate coating buffer (0.06 M Na2CO3, 0.14 M NaHCO3,100 ml distilled H2O [pH 9.6]) was used to generate a standard curve.Wells were blocked with 2% nonfat milk and 100 �g/ml BSA, and thereaction mixture was washed three times with 0.05% Tween 20 –PBS(wash buffer) and incubated overnight with anti-B. anthracis LF (bD-17;Santa Cruz) at a 1:100 dilution in PBS overnight at 4°C. Wells were treatedwith horseradish peroxidase-conjugated donkey anti-goat IgG (SantaCruz) and washed, and a colorimetric reaction was allowed to developusing 1-Step Ultra TMB (3,3=,5=5=-tetramethylbenzidine)-ELISA sub-strate solution (Thermo Scientific). Average absorbance values were mea-sured at 450 nm and normalized by subtracting these values from the

Nguyen et al.

4812 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

average absorbance of blank PBS wells. The negative control was 3 mg/mlBSA–PBS, and the positive controls were supernatants of B. anthracisovernight cultures grown in Ristroph media, a toxin-inducing media (35).One-way analysis of variance (ANOVA) followed by Dunnett’s test formultiple comparisons was used to compare negative-control groups tothe respective positive-control and infected groups.

Data normalization, statistical analysis, and bioinformatics. Theanalysis of the metabolomic data was performed in the Maresso labora-tory. Raw spectral counts were log transformed and normalized by me-dian centering. Missing values were inputted using the lowest value de-tected for a particular metabolite. Statistical analysis of fold changes bytwo-way ANOVA and Welch’s two-sample t test were used to determinemetabolites with significant results (P � 0.05) for samples obtained fromthe vegetative cell and spore infections, respectively. ANOVA and Welch’stest were used to compare the average spectral counts from the mock-infected negative-control group with those from the experimental groupfor each metabolite. Bioinformatics analyses such as partial least-squaresdiscriminant analysis (PLS-DA), random forest (RF) analysis, and hierar-chical clustering by heat map analysis were performed using the Metabo-analyst web portal (www.metaboanalyst.ca). PLS-DA was used to identifyinitial trends and clusters in data sets. Principal components denote thepercentage of the contribution that can explain the original data set afterlinear transformation. Component 1 is defined as representing the highestpossible variability in the data set after transformation is applied. Com-ponent 2 represents the second highest variability after transformation(36). The RF algorithm was used to classify metabolite importance basedon the computed mean decrease accuracy (or prediction accuracy), whichis determined by random permutation of variables and running observedvalues through decision trees. For each run, the prediction accuracy isreassessed for random noise or decreases in value if the variable is impor-tant to the classification. Hierarchical clustering by heat map analysis wasused to demonstrate fold changes of metabolites in comparison to controlgroup results.

Pharmacological inhibition of PLA2 during infection. Female A/Jmice (6 to 8 weeks old) were given intraperitoneal (i.p.) injections in theright hind leg with 200 �l of 80 mg of quinacrine (QC)–PBS (broad-spectrum PLA2 inhibitor) or a 100 �M solution of PACOCF3 (3% di-methyl sulfoxide [DMSO]–saline solution) (cytosolic PLA2 [cPLA2]-spe-cific inhibitor) as a single or repeated infection, followed by subcutaneousinjection of the left hind leg with B. anthracis spores (n � 5 per group).Several doses of QC administered to mice were assessed to determine anappropriate concentration that did not yield negative physiological sideeffects (data not shown). Mice given multiple doses of inhibitor wereadministered the drug 24 h after the last injection. Disease sequelae andmortality were monitored at 12-h intervals. The effects of inhibitor andinfection on disease severity were evaluated using Kaplan-Meir survivalcurves and a log rank test (P � 0.05) and GraphPad Prism 6.0 software(GraphPad Software, Inc.). A disease index was developed to evaluatevisible disease symptoms and physiological health conditions using thefollowing five criteria: ruffled fur, lethargy, hunched posture, edema, andeschar formation (scoring range, 0.00 to 1.00, with 1.00 representing thegreatest degree of visible disease progression for a particular criterion).Kruskal-Wallis nonparametric ANOVA was used to compare the areaunder the curve (AUC) data for the average daily health score values foreach animal (P � 0.05).

RESULTSEstablishing a vegetative cell infection model for a metabolomicanalysis. We used a well-characterized murine model of anthraxdisease to examine metabolomic changes during bacterial infec-tion. This model was chosen because (i) B. anthracis can initiate aprogressive infection from a low dose from either a spore or veg-etative cell form, thereby allowing one to monitor early-stage al-terations in the host during a developing infection, (ii) the exper-iments can be performed under biosafety level 2 (BSL2)

conditions, thereby enabling the timely completion and experi-mental manipulation of the conditions, and (iii) the course of thedisease models anthrax disease in humans (37, 38). A/J mice wereinfected (by subcutaneous injection) with B. anthracis vegetativecells or spores to first determine the LD50 for these forms of bacilli.The LD50 for the vegetative cell infection was determined to be1.5 � 103 cells, while that of the spore infection was determined tobe 1.0 � 102 spores (see Fig. S1A and B in the supplemental ma-terial). For the first metabolomic analysis, the vegetative form waschosen and mice were mock infected (1% saline solution) or in-fected with one LD50 of B. anthracis. Following infection, bloodwas collected at 1 and 3 dpi (experimental design shown in Fig.1A); indeed, very few or, in most cases, no bacilli were detected inthe organs, as determined by a CFU estimation (data not shown).

We chose 1 and 3 dpi for sample collection because these weretime points considered before the onset of robust anthrax disease.In doing so, we wanted to avoid assessing nonspecific changes inmetabolites that result from the initiation of death. Thus, by sam-pling animals well before death, we could observe significantchanges during early stages of infection before the onset of sub-stantial bacterial dissemination (data not shown). Samples wereharvested on day 1 because it represents an early and convenienttime point— exactly 24 h after infection. Day 3 was chosen be-cause it is 1 or 2 days before the mice develop full-blown anthraxdisease in this model.

Whole blood was then subjected to an untargeted metabolo-mic analysis using a combination of gas and liquid chromatogra-phy and mass spectrometry, an approach that identified 271 totalmetabolites in whole blood. There were clear differences betweeninfected and uninfected mice at both 1 and 3 dpi, as revealed bypartial least-squares regression (PLS-DA), with a greater separa-tion observed later in infection, indicating considerable variationin the metabolite composition as the infection progressed (Fig.1B). Analysis of statistically significant metabolites identified bytwo-way ANOVA (P � 0.05) revealed that more metabolites haddecreased in abundance than had increased in abundance at bothtime points compared to the control group results (Fig. 1C). Ofthe 60 significant metabolites whose abundance had changed inthe 1 dpi groups, 57 metabolites decreased in abundance, whereasonly 3 metabolites exhibited increased levels. Similar trends wereobserved 3 dpi; 58 metabolites showed significant decreases inlevels whereas 16 metabolites were increased in abundance in re-sponse to infection (Fig. 1C). All altered metabolites were nextsegmented by metabolic pathways, which revealed major changesin lipid, amino acid, and carbohydrate pathways (Fig. 1D and E).Of these, the changes in lipid levels, the majority of which haddecreased by 3 dpi, were the most dramatic. These findings indi-cate that during the early stages of anthrax disease, well before anyobservable signs of disease or detectable bacilli in major organsystems, the host demonstrates dramatic alterations in lipid levels.It should be noted that anthrax toxin was detected in blood ofthese mice, which indicates that the mice were indeed infected (seeFig. S4 in the supplemental material).

Effect of B. anthracis vegetative infection on the levels ofpolyunsaturated fatty acids (PUFAs) and lysophospholipids (ly-solipids). We wondered if there were metabolites that were pre-dictive of anthrax disease. A comparison of (1 dpi) infected anduninfected groups by random forest (RF) analysis yielded a pre-dictive accuracy of 64.3% (Fig. 2A), which is only slightly betterthan what is expected by chance (50% accuracy). However, dis-

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4813Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

FIG 1 Global metabolomic analysis of blood from mice infected with B. anthracis vegetative cells. (A) Infection timeline. Mice were infected with vegetative cells(v.c.) via a subcutaneous route (see Materials and Methods), and whole blood from infected and control A/J mice that were separated into 5 groups was collectedat 1 dpi (control, C1d; infected, I1d) and 3 dpi (control; C3d; infected, I3d) (n � 7 per group). The fifth group consisted of one mouse from each of the four groupsdescribed above; the mice in this group were analyzed in a blind manner (see Fig. S4 in the supplemental material). (B) PLS-DA analysis of each group,demonstrating a clear separation of controls and infected samples from both time points. (C) Identification of statistically significant metabolites by two-wayANOVA t test (P � 0.05) of normalized metabolomics data. (D and E) Categorization of metabolites involved in glucose metabolism, including gluconeogenesis,glycolysis, oxidative phosphorylation, and the Krebs cycle, at 1 dpi (D) and 3 dpi (E). Note the decrease in levels of lipids relative to the other metabolites.

Nguyen et al.

4814 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

tinct trends were present even after 1 dpi. Among the top 15 mostimportant metabolites found by RF analysis, 7 were associatedwith lipid pathways, the majority belonging to carnitine-derivedmolecules. Carnitine and its fatty acid derivative, palmitoylcarni-tine, have been shown to modulate the activity of caspases in-volved in apoptosis, suggesting that early stages of B. anthracisinfection can alter the apoptotic/survival status of host cells (39,40). In addition, sphinganine, a precursor in sphingosine biosyn-thesis, showed reduced levels at 1 dpi compared to control groupresults, which may indicate altered ceramide signaling in apop-totic events in response to infection (Fig. 2B) (41). Carnitine-conjugated fatty acids such as decanoylcarnitine and succinylcar-nitine showed lower levels in the infected group than in thecontrol group (Fig. 2B).

An RF analysis of the 3 dpi group yielded a different story,showing 92.9% accuracy of infection prediction compared to thecontrol and indicating that the changes in metabolic profiles be-tween the two groups are substantial (Fig. 2C). PUFAs, a class oflipid molecules important in inflammation signaling (Fig. 3A),showed alterations at both time points in response to infection.PUFAs of the �-3 family, which includes eicosapentaenoate (EPA;20:5 �-3) and docosahexaenoate (DHA; 22:6 �-3), identified byRF analysis were shown to be correlated with the decreased levelsof essential fatty acids �-linolenate (18:3 �-3) and linoleate (18:2�-6) (Fig. 2B and 3B). Other members of the �-3 and �-6 path-ways such as dihomo--linolenate (18:3 �-6) and arachidonicacid (AA) also showed significant decreases in abundance at 3 dpi(Fig. 3). AA is the key intermediate of the arachidonic acid cascade

FIG 2 Random forest (RF) analysis of metabolite changes in blood of animals infected with B. anthracis vegetative cells. Mice were infected with vegetative cellsvia a subcutaneous route (see Materials and Methods). Data represent mean decrease accuracy and hierarchical clustering by heat map analysis of control andinfected blood for the top 15 metabolites at 1 dpi (A and B) and 3 dpi (C and D) determined using Pearson rank correlation distance and complete linkage. Heatmap colors reflect relative abundances of metabolites identified by random forest analysis. Red or blue is used to show that the abundance of the indicatedmetabolite was higher or lower than the mean metabolite abundance, respectively. GPC, glycerophosphocholine; GPE, glycerophosphoethanolamine.

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4815Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

which stems into multiple cell signaling pathways with diversefunctions, including production of eicosanoids, considered to bekey mediators of inflammation (Fig. 3A). This notion is reflectedby the decreased levels of 12-hydroxyeicosatetraenoic acid (12-HETE), an eicosanoid derived from arachidonic acid in response

to the vegetative cell infection at both time points (Fig. 3B). Asecond source for AA operates through the breakdown of phos-pholipid membranes (Fig. 3A). Products of membrane digestioncalled lysophospholipids (lysolipids) also showed reduced levelsin infected blood at both time points (Fig. 3B). Eicosanoids and

FIG 3 Lipid pathways affected by B. anthracis vegetative cell infection. (A) Schematic model of the two inflammatory lipid pathways that were altered byinfection: the polyunsaturated fatty acid (PUFA) biochemical pathway and the pathway involved in membrane breakdown by host PLA2 enzymes (cytosolic andsecreted PLA2 classes cPLA2 and sPLA2 are shown). These sources merge in the production of arachidonic acid (AA), which feeds into the eicosanoid biosynthesispathways catalyzed by cyclooxygenase (COX), which yields prostaglandins (PGs) and thromboxanes (TXs), and by lipoxygenase (LOX) enzymes, which yieldleukotrienes (LTs) and hydroxyeicosatetraenoic acids (HETEs). Eicosanoid production and lysolipid production can both promote inflammation. Lipidmetabolites whose levels were lower following infection are denoted in red with red arrows. (B) Representative lipids from the PUFA and phospholipidbreakdown pathways, comparing levels at 1 and 3 dpi. Box-and-whisker plots show metabolite minimum, lower quartile, median, upper quartile, median(middle line), mean (), and extreme (open circle) values and sample maximums. Metabolites involved in PUFA pathways are in gray, eicosanoids are in red,and lysolipids are in yellow.

Nguyen et al.

4816 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

lysolipids are both bioactive lipids produced by AA metabolismthat contribute to the recruitment of immune cells and initiationof inflammation. In particular, lysophosphatidylcholines(lysoPCs) such as 1-arachidonoyl– glycerophosphocholine (1-arachidonoyl–GPC) are molecules that can accumulate in bloodand are considered neutrophil-activating factors, and their levelswere found to be reduced in the infected groups at both timepoints (Fig. 2D) (7, 8). Taken together, these data suggest thatvegetative cell infection by B. anthracis generates an overall reduc-tion of levels of circulating lipids, several of which are reportedlyinvolved in the innate immune response during infection.

Whole-organ metabolomics during B. anthracis spore infec-tion. During systemic anthrax, bacilli spread to the lungs, liver,kidneys, and spleen. Having established that there are globalchanges in lipids in blood as an outcome of B. anthracis infection,we next assessed which metabolites were altered in major organsystems during infection. Spores were used for this arm of ouruntargeted metabolomic analysis because they represent the infec-tious form observed in nature. This also allowed us to compare themetabolomes of spore versus vegetative forms of the infection.Mice were mock infected with 1% saline solution or infected with1.0 � 102 spores (determined to be the LD50 of B. anthracis) (seeFig. S1B in the supplemental material), and samples of lungs, liver,kidneys, spleen, and whole blood were collected at 3 dpi fromcontrol and infected mice (n � 6 per experimental group) (Fig.4A). Of note, to eliminate any contamination of these tissues withblood, each organ was perfused (see Materials and Methods), aprocess that resulted in a substantial reduction in the levels ofhemoglobin in these organs, signifying that the blood had beenremoved (see Fig. S3). The examination of blood in a differentcontext also provides a control by which to determine how con-sistent these findings are between two very different modes ofinfection. Following gas/liquid chromatography and mass spec-trometry of organ samples, metabolomic profiling identified a to-tal of 339 metabolites in the lung, 387 metabolites in the liver, 384metabolites in the kidneys, 378 metabolites in the spleen, and 319metabolites in the blood. Analysis by Welch’s two-sample t test(P � 0.05) revealed 39 statistically significant metabolites betweeninfected and uninfected mice from liver tissues, 32 metabolitesfrom the kidneys, 14 metabolites from the spleen, and 43 metab-olites from blood (Fig. 4B). Similarly to the first metabolomicanalysis, the abundances of the vast majority of altered metabo-lites (40 metabolites) in the blood were decreased, which providedconfidence in the consistency of our infections and this analysis(Fig. 4B). Furthermore, large changes were again associated withlipids, with more than 17 decreased in abundance (Fig. 4C). Only7 metabolites in the lungs (all of which had increased in abun-dance) had changed compared to the control, the lowest numberof changes among all of the organ comparisons. The liver dem-onstrated �20 metabolites whose abundance increased or de-creased following spore infection (Fig. 4B). In contrast, thekidneys displayed a net increase in levels of metaboliteswhereas the spleen showed more metabolites whose abun-dances were reduced in response to infection (9 metabolites).Taken together, these results indicate that the host undergoessubstantial changes in metabolite levels in major organ systemsin response to B. anthracis infection. In addition, many of thesame changes, especially decreases in levels of lipids, were againobserved in the blood, regardless of whether the spore or veg-etative form was analyzed.

Effect of B. anthracis spore infection on lipid metabolism.Lipids and amino acids identified in whole blood of mice infectedwith B. anthracis spores were the most abundant metabolites thatshowed decreased levels, reflecting trends similar to those observedwith the vegetative cell infection (Fig. 4C). RF analysis revealed that 10of 20 of the most important metabolites were lipids (Fig. 4D). Simi-larly to the vegetative cell infection analysis, the RF analysis identifiedthat the levels of lipids involved in inflammation were decreased afterspore infection. Consistent with the vegetative cell analysis, the levelsof lipids involved in PUFA metabolism and lysolipid production weredecreased, indicating they may play an important role during bothstages of B. anthracis infection. Among these lipids were PUFAs suchas linoleate (18:2 �-6) and dihomo--linolenate, both of which werefound to be decreased in abundance compared to the control (Fig.4D). This notion is reflected in decreased levels of downstream prod-ucts in lysolipids such as 1-linoleoyl-glycerophosphoethanolamine(1-linoleoyl-GPE) (18:2) and 1-linolenoyl-GPC (18:3 �-3) (Fig. 4D).The changes seen in whole blood from the spore infection at 3 dpiwere less significant than those found with the vegetative cell infection(Fig. 4B). No new group or class of metabolites with distinct relation-ships, aside from similar (albeit smaller) changes in lipids with in-flammatory properties found in the vegetative cell infection, wasidentified by bioinformatics analysis of the spore infection. Takentogether, these data suggest that B. anthracis infection suppressessmall-lipid mediators prior to the onset of anthrax disease.

Effect of B. anthracis spore infection on the energy status inthe spleen and liver. Aside from their roles in the immune re-sponse and signaling, fatty acids can be used as an energy source(42–45). For example, it is known that fatty acid oxidation candrive ATP production in rat heart tissues (46). Up to 50% of theATP produced in myocardial tissues can be a result of fatty acidcontributions, suggesting that energy demands cannot be metsolely by glucose supplies (46). It is possible that lipids producedin organ systems can be released into the vasculature and affect thecomposition of metabolites found in whole blood. In particular,fatty acid production can occur in the liver and the fatty acids canbe released into circulation (44). Liver tissues showed significantdecreases in the levels of lipid metabolites associated with fattyacid biosynthesis (Fig. 5A). Specifically, the levels of coenzyme A(CoA) and 4=phosphopantetheine, a metabolite in the CoA bio-synthesis pathway, were decreased in infected liver tissues relativeto the control, suggesting alterations in normal hepatocellular en-ergy metabolism (Fig. 6C). In addition, carbohydrates participat-ing in energy metabolism were also affected (Fig. 5A). Levels ofglycolysis intermediates in both the spleen and liver were de-creased, suggesting additional decreases in host energy consump-tion due to infection (Fig. 5C and 6B and C). In the liver, levels ofmaltooligosaccharides, carbohydrates involved in glycogen pro-duction that can be converted into glucose, were also decreasedduring infection, suggesting that the stored glycogen that had beendepleted had not been replenished (Fig. 6D). These results suggestthat both arms of energy production, carbohydrates and lipids, innormal liver function are disrupted in response to spore infection.

Effect of B. anthracis spore infection on glucose metabolismin the kidney. Further separation of significant metabolites intoaffected pathways revealed the increases in the levels of lipid andcarbohydrate-derived metabolites to be the most abundant (Fig.5B). Prominent increases in levels of lipid metabolites associatedwith increased oxidative stress were of particular interest. Specif-ically, a higher abundance of cholesterol and cholestanol, a deriv-

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4817Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

FIG 4 Metabolomics analysis of blood and organs of animals infected with B. anthracis spores. (A) Infection timeline. Mice were infected with spores via a subcutaneousroute (see Materials and Methods). Whole blood (WB) and organ (lung, liver, kidney, and spleen) tissues from infected and control A/J mice were collected at 3 dpi. (B)Identification of statistically significant metabolites by Welch’s two-sample t test (P � 0.05). (C) Classification of altered metabolites into categories of affected pathways.(D) Hierarchical clustering by heat map analysis (Person rank correlation distance and complete linkage) of the top 20 metabolites identified by RF analysis of wholeblood. Red or blue is used to show that the abundance of the indicated metabolite was higher or lower than the mean metabolite abundance, respectively. GSSG,glutathione disulfide.

4818 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

ative of cholesterol metabolism, was observed in response to in-fection (Fig. 6F). In addition, plant sterol-derived campesterol, amarker of cholesterol absorption, also displayed increased levels rel-ative to the uninfected kidney samples (Fig. 6F) (47). The levels ofseveral metabolites involved in glucose metabolism were revealed tobe significantly increased in response to infection (Fig. 6E). Interme-diates participating in gluconeogenesis, including glucose-6-phos-phate (G6P), phosphoenolpyruvate (PEP), and 3-phosphoglycerate(3PG) (Fig. 6E), were identified. G6P and PEP levels increased almost3-fold in infected kidneys relative to the control.

The role of PLA2 in the systemic response to B. anthracisinfection and anthrax disease. Lysolipids are produced by the ac-tivities of phospholipase A enzymes, which catalyze the cleavage ofmembrane phospholipids. There was no obvious preference for thesn-1 position or the sn-2 position among the lysolipids altered in thevegetative cell or spore infection. This may indicate both PLA1 and/orPLA2 may be inhibited or somehow downmodulated as the infectiondevelops. However, AA production and subsequent lysolipid release

are due only to the action of PLA2 breakdown of membranes. Thelatter finding invokes a more prominent role for PLA2 than for themembers of other classes of phospholipases in this response (48).Products of PLA2 activity can act as direct mediators of inflammationand can also initiate multiple downstream inflammatory signalingcascades in response to an infection. The levels of several of theselipids were found to be decreased in blood in both the vegetative celland spore infections, perhaps as a mechanism directed by B. anthra-cis to dampen the innate immune response.

Along these lines, we tested the hypothesis that PLA2 was im-portant in the host response to B. anthracis infection by assessingthe progression to anthrax disease in mice administered PLA2 in-hibitors. Mice were injected intraperitoneally (i.p.) with quina-crine (QC) or PACOCF3, each of which is an inhibitor of PLA2,and then immediately inoculated with B. anthracis spores, fol-lowed by another dose of inhibitor every 24 h. Single-blind exper-iments were conducted to assess disease severity and mortality byhealth index scores, monitored at 12-h intervals (Fig. 7 and Table 1).

FIG 5 Summary of the altered pathways in organs following infection with B. anthracis spores. Metabolites were identified by Welch’s two-sample t test (P �0.05), comparing control and infected samples from the liver (A), kidneys (B), and spleen (C) following infection of A/J mice with B. anthracis strain Sterne 34F2(as outlined in Materials and Methods).

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4819Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

Briefly, before scores were given, research personnel removedidentifying information from each experimental group and ran-domized the placement order from the previous day on the shelvesin which the cages were housed. Different research personnel then

recorded the scores without having knowledge of which experi-mental group was being assessed. The same researcher personnelwere then used to access health scores for all time points to elim-inate technical variation. Infected mice given quinacrine, a broad-

FIG 6 The analysis of organ tissues as it pertains to energy metabolism following infection with B. anthracis spores. (A) Changes in metabolites in energyproduction, consumption, and storage pathways that were significantly altered following infection of A/J mice with B. anthracis strain Sterne 34F2 (spores). (Bto F) Representative metabolite data, comparing control and infected spleens (B), livers (C and D), and kidneys (E and F).

Nguyen et al.

4820 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

spectrum PLA2 inhibitor, displayed a more rapid and severe ap-pearance of visible disease symptoms than animals given QC orspores alone (Table 1). Disease severity was correlated to increasedmortality and resulted in decreased MTD in infected mice receiv-ing the inhibitor (Fig. 7A and B) (Table 1). No lethal effects wereobserved in mice given QC alone. As expected, increasing thespore inoculum from 5 � 102 to 2 � 103 spores augmented diseaseseverity and decreased the MTD even further compared to lowerspore doses. Similar results were observed with mice givenPACOCF3, an inhibitor that specifically targets the cytosolic formof PLA2 (cPLA2) (Fig. 7C) (Table 1). Since lipid inflammatorycascades can be amplified by the activities of PLA2 and since theinhibition of PLA2 under three different conditions worsens theoutcome of B. anthracis infection, the findings from these func-tional studies suggest that B. anthracis may inhibit inflammatorylipid production (as directly determined by the metabolomics

studies) as a means to dampen the host immune response. Theseresults also suggest that inhibition of phospholipases and theirlipid products is important for host survival to B. anthracis.

DISCUSSION

Metabolomics is becoming increasingly important for the analysisof host-pathogen interactions, particularly in the discovery of bio-markers of disease (11). Recent studies have shown that metabo-lomics can be used to identify metabolites whose levels are alteredin response to an infection (10, 49–51). In particular, with respectto bacterial infections, the rapid emergence of multidrug-resistantGram-positive and Gram-negative bacteria fuels the need fornovel strategies in diagnosis and treatment (11). Diagnosis pri-marily relies on nonmolecular tests that are time- and labor-in-tensive. Here, we report the first global metabolomic analysis ofmultiple organ systems following a systemic bacterial infection.Our results reveal (i) that there are clear differences between in-fected and uninfected animals in the small-molecule metabolomeof blood and organs, (ii) that these differences highlight majorchanges in amino acid, energy, and lipid metabolism as early as 1dpi, (iii) that major alterations in two different arms of inflamma-tion-related lipid metabolism suggest that B. anthracis infectionleads to decreased levels of signaling molecules that function in theinnate immune response to pathogens, and, finally, (vi) that thetargeted inhibition of phospholipases accelerated anthrax disease,suggesting a novel role for these enzymes in host survival during B.anthracis infection. In addition to the points discussed above,since similar results were observed for the spore and vegetativeforms of the infection, for different mouse groups, and, generally,for blood and organs, the findings presented in this report suggestthat B. anthracis infection induces a pronounced reduction in thesteady-state levels of small-lipid mediators, some of which areimportant in activating an innate immune response, during theonset of anthrax disease.

In this study, we chose to use B. anthracis as a model organism toidentify metabolomics changes during blood-borne bacterial infec-tions. We choose to use a murine infection model and the Sternestrain to facilitate these studies. Some limitations to using this strain(versus a fully virulent pXO1- and pXO2-containing strain) includethe fact that anthrax toxin is primarily thought to be the main factordriving overall disease severity. Therefore, based on our LD50 deter-mination, we were careful in selecting time points at 1 and 3 dpi,which are likely before the onset of significant toxemia.

Due to the sensitivity of the metabolomics platform, selectionof an animal model can result in various outcomes depending onthe species and strain used for infection. We were aware that thereis no ideal animal model capable of perfectly recapitulating hu-man anthrax disease. Past studies have used guinea, rabbits, non-human primates, and mice to study anthrax in mammals (52).While the guinea pig, rabbit, and nonhuman primate models arerelevant for inhalational studies, the mouse models are generallyaccepted for studying early stages of infection (52). Since this wasthe primary focus of our study, we chose to use the A/J mousemodel due to the feasibility of handling a larger sample size in anuntargeted metabolomics platform.

The LD50 was chosen as the infectious dose so that the metabo-lomics analysis would be performed on mice that were undergo-ing a developing infection. By doing so, we are able to determine ifour metabolomics platform was sensitive enough to detect metab-olites during an early state in an infection that was not yet robust

FIG 7 Effects of PLA2 inhibition on the lethality of B. anthracis spore infec-tion. (A to C) Kaplan-Meier survival curves of A/J mice given i.p. injections ofa broad-range PLA2 inhibitor, quinacrine (QC), or a cPLA2-specific inhibitor,PACOCF3 (C), followed by a subsequent subcutaneous infection with B. an-thracis spores. Animals were inoculated with inhibitor only, with inhibitorimmediately followed by spores, with inhibitor at multiple doses (M) and thenspores, or with spores only at 5 � 102 spores/dose (A) or 2 � 103 spores/dose(B and C). All differences in groups were statistically significant by the log-ranktest (P � 0.0001).

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4821Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

in nature. Even if the infection did not progress, it was expectedthat the bacterial inoculation would produce a host response andcorresponding changes in host metabolites. As such, this ap-proach was meant to also avoid nonspecific changes in physiologythat might be associated with the process of dying. No mortalitieswere observed at either day in the mice used for this study, whichis consistent with the idea that we were measuring early changes inthe development of the infection. Attempts by the host defense tomitigate early stages of B. anthracis infection rely heavily on theactivity of innate immune cells, which can induce activation ofhost-defense mechanisms that initiate inflammatory pathways(21). Expression of secreted and cytosolic families of the PLA2

enzymes is one common host defense strategy employed by mac-rophages to hydrolyze membrane lipids (48, 53). These enzymescan act as a bactericidal component and/or simultaneously pro-duce lysolipids which can activate immune cells (54–57). B. an-thracis lethal and edema toxins decrease the expression of the se-creted phospholipase sPLA2-IIA in alveolar macrophages (56, 58,59). However, it is unclear whether this downregulation can affectlysis of B. anthracis membranes and, more importantly, whether itcan impair lipid mediator production or subsequent inflamma-tory status. Some preliminary data determined using a targetedmetabolomic approach with toxin-deficient strains indicate thatthe levels of PUFAs are still reduced following infection (data notshown). This may indicate that other virulence factors play a moreprominent role in this response. In addition, on the host side, it isalso unknown whether other subclasses of PLA2 enzymes are ac-tively participating in the response during early infections. Animalstudies using a broad-spectrum PLA2 inhibitor, QC, in conjunc-tion with spore infection resulted in increased disease severity andmortality. Interestingly, previous studies showed that addition ofQC (in a culture setting) inhibits lethal toxin cytotoxicity inmouse peritoneal macrophages (60). These findings are somewhatat odds with the in vivo results presented here, which show thatadministration of QC increases disease severity and mortality in B.anthracis-infected mice. However, these two distinctions together

may suggest that PLA2 enzymes play an important role during B.anthracis infection. PLA2 enzymes could have a protective effect inguarding against systemic infection but may be deleterious to thehost in local tissue environments. More-specific forms of PLA2

such as cPLA2 are known to target membranes that contain AA,suggesting a more prominent role for cPLA2 than for sPLA2 inpromoting inflammatory pathways (61). The finding that inhibi-tion of cPLA2 by PACOCF3 also leads to increased mortality sug-gests that although cPLA2 does not have a direct role in producingdeleterious effects on B. anthracis, it is still important in the re-sponse to this type of infection.

Our analysis also revealed vast differences in metabolic net-works, illustrating the complexity and interconnection of biolog-ical systems due to infection. Hence, no single metabolome can becompletely identical to the next, and each represents a uniqueprofile exclusive to each organism (3). Despite this variation, wewere able to successfully assign blind blood samples to their re-spective infection groups based on the results of PLS-DA cluster-ing analysis. This finding supports the idea of the potential use ofmetabolomics in the diagnosis of infections before the onset ofsymptoms (see Fig. S4 in the supplemental material). Addition-ally, the use of different animals in our study instead of the sameone for each time point resulted in sufficient resolution to detectminor but distinct changes in entirely different subjects. In thisanalysis, the use of metabolomics analysis also led to the successfulidentification of new biochemical pathways altered by B. anthracisthat had not been previously reported. Studies of innate immuneresponses to infection by bacilli have focused on cytokines andchemokines, proteins that can evoke inflammation (21, 24, 62).However, the metabolomic approach described here has identi-fied lipids and associated pathways involved in inflammation, im-plicating a new class of molecules that could be equally importantto host defenses during B. anthracis infection.

The profiles from infected whole blood from B. anthracis veg-etative cell and spore infections provided us with a connection toinfection status and the life cycle of B. anthracis. Comparisons of

TABLE 1 Average heath index scores and median times to death following PLA2 inhibition in B. anthracis-infected micea

Treatment

Avg heath index score

MTDDay 1 Day 2 Day 3 Day 4 Day 5

QC � 5 � 102 spores*QC only 0.25 � 0.00 0 0 0 0 NA5 � 102 spores only 0 0.20 � 0.27 1.62 � 0.85 3.08 � 2.45 0.50 4.0QC 5 � 102 spores 0.70 � 0.21 0.99 � 0.36 2.32 � 0.29 4.38 � 0.88 3.8QC(M) 5 � 102 spores 1.20 � 0.21 1.59 � 0.29 3.12 � 0.22 4.75 3.6

QC � 2 � 103 spores*

QC only 0 0 0 0 0 NA2 � 103 spores only 0 1.80 � 0.89 3.50 � 0.90 3.25 5 3.2QC 2 � 103 spores 0.70 � 0.27 2.63 � 1.30 3.25 2.2QC(M) 2 � 103 spores 0.75 0.35 2.81 � 1.60 2.3

PA � 2 � 103 spores*PA only 0 0.20 � 0.00 0.20 � 0.00 0 0 NA2 � 103 spores only 0 0.37 � 0.16 2.59 � 0.19 3.5PA 2 � 103 spores 0 0.86 � 0.12 2.57 � 0.25 4.0PA(M) 2 � 103 spores 0 1.04 � 0.55 2.71 � 0.14 3.5

a The table shows average daily health index scores for surviving mice of each group � standard deviations (SD). *, P � 0.05 by Kruskal-Wallis test comparing the area under thecurve (AUC) of daily average health score values for each animal. (M), multiple doses; MTD, median time to death; NA, not applicable.

Nguyen et al.

4822 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

blood metabolites from the two analyses revealed similar trends toperturbed metabolic pathways, providing a link to infection pro-gression. The spore infection is more representative of naturalinfections, as it is the form of B. anthracis that the host encounters,whereas the vegetative cell form represents a later stage of infec-tion. These changes to the pathogen’s life cycle are correlated tothe differences seen in our results. There were fewer significantchanges found in the spore infection, and the changes were not aspronounced as those found in the vegetative cell infection. Asidefrom the inflammatory lipids, no other class of metabolites dis-played significant trends corresponding to spore infection alone,suggesting a steady accumulation of metabolic perturbations in-duced by the transition from the bacillus spore to the vegetativecell. Similar trends in decreased levels of PUFAs and lysolipidswere found in the two analyses; however, fewer of these lipids wereidentified in the spore infection whereas the decrease was mostpronounced during the vegetative cell infection at 3 dpi. This sug-gests that the vegetative cell is the more disruptive form of B.anthracis or, perhaps more intriguing, that the infection by thespore form is stealthier.

Our results showed that the decreases in the levels of lipidbiomolecules with links to inflammation originated from two ma-jor pathways that converge in the production of AA. The eico-sanoid biosynthesis pathway centers on the metabolism of PUFAprecursors and is thus important in modulating host inflamma-tory status. We choose to further explore the link between theseinflammatory lipids and their relationship to infection. The sec-ond pathway involves the PLA2 breakdown of phospholipidmembranes, which can also produce AA and, in addition, re-lease lysolipids. Lysolipids are not related to eicosanoids butcan serve functions similar to those of inflammatory media-tors. The metabolomics findings presented here support the ideaof the importance of AA, since decreases in levels remained con-stant for both time points of vegetative cell infection, suggesting acontribution of both pathways to AA production. Decreases ob-served in levels of �-6 PUFAs and lysolipids from infected bloodduring 3 dpi suggests that both arms of AA production were dis-rupted. These results are consistent with other metabolomicsstudies of infection whose results have indicated the importance ofeicosanoids and lysolipids during infection (45, 50, 63). However,it appears that decreases in levels of eicosanoids and lysolipids arenot universal traits during bacterial infections. Serum profiles ofpatients infected with Mycobacterium leprae revealed increasedlevels of anti-inflammatory PUFAs and accumulation of phos-pholipids (12, 45). Feces and liver profiles from mice subjected toinfection with Salmonella enterica serovar Typhimurium showedincreases in levels of eicosanoid biosynthesis products (50). Not alltypes of infections evoke the activity of PLA2 enzymes as an im-mune defense mechanism against invading pathogens. It has beenshown that PLA2 is a crucial factor in hepatitis C virus (HCV)virion core protein production and envelopment, as well as for thesecretion of particles from host cells (64). Ultimately, PLA2 is animportant prerequisite host component needed for improving theinfectivity of HCV. This suggests that downmodulation of eico-sanoids and lysolipids is specific to B. anthracis and thus may beused as a biomarker to differentiate B. anthracis infections fromother types of infection. While we chose to focus on possiblemechanisms of B. anthracis targeting host lipid production, anequally plausible alternative hypothesis is that decreases seen inlevels of lipids may be due to lipid consumption by the host and/or

bacteria. Perhaps the host needs to consume more lipids to com-pensate for the energy losses observed in other areas of metabo-lism. Additional experimentation would be needed to addressthese issues.

While changes associated with lipids found in whole bloodwere the primary focus of this study, there were also many bio-chemicals from whole organs that were substantially changed inresponse to infection. We chose to analyze the lung, liver, kidneys,and spleen for three main reasons. First, we wanted to understandhow infection altered the host’s metabolism on a systematic level.Therefore, we chose to analyze these major organs due to theirimportant physiological functions. Second, our metabolomics ex-perimental design was intended to be untargeted so as to cast awide net. Surveying other tissues would have been beyond whatwas feasible to perform in this current study. That said, it will beinteresting to determine what metabolites are altered in other tis-sues and cells relevant to anthrax disease, including regionallymph nodes and macrophages, respectively.

We discovered that organ metabolites involved in energymetabolism showed differential levels among the experimentalgroups, which could correspond to changes to the inflammatorystatus observed in blood. Global analysis of A/J mice infected withB. anthracis spores showed that the metabolic profiles of the kid-neys, liver, and, to a lesser extent, spleen revealed complex net-works involved in energy metabolism. Lung tissues from infectedmice were considered relatively unchanged compared to the con-trol and were excluded from any further analysis. Given that sub-cutaneous sublethal infections were used, this lack of change sug-gests that host responses were not yet active in the lungs. Renalglucose metabolism plays a significant role in the energy produc-tion and storage involved in maintaining homeostatic functionsand responding to pathological alterations (65, 66). Increased lev-els of metabolites involved in gluconeogenesis in infected kidneyscould reflect the host’s growing energy consumption needs in re-sponse to the invading pathogen. Although the liver is consideredthe more prominent site of gluconeogenesis, renal glucose releasecan contribute to �20% of the total glucose released into thecirculation (65).

Energy production, consumption, and storage are split be-tween the use of carbohydrates and the use of lipids. CoA, animportant coenzyme present in both arms of energy production,was found to be decreased in abundance in infected liver tissuesrelative to the control. CoA plays a role in �-oxidation and inaddition feeds into the first step of the tricarboxylic acid (TCA)cycle (Fig. 6) (42). Alterations in its levels are consistent with de-creased free fatty acid levels in blood later during infection, sup-porting the notion of compromised liver function and inter-connection between major metabolic pathways and tissues.Additionally, decreases in levels of starch byproducts (maltooli-gosaccharides) indicate alterations in glycogen metabolism, sug-gesting that the depleted liver glycogen stores were not replen-ished, as levels of glucose and gluconeogenesis intermediates werealso decreased during infection. Our results are in agreement withthose of other studies that show that systemic infection can lead todecreased liver gluconeogenic function, suggesting that B. anthra-cis acts similarly during the course of spore infection (67–69).

Although we chose to focus our interpretation on energy me-tabolism in these organ systems, there were many other pathwaysfound in our analysis that exceed the scope of this paper. It ispossible that the complex metabolic networks that span organs

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4823Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

and whole blood are interconnected. To our knowledge, this con-cept has not been investigated extensively during a B. anthracisinfection. Organ-specific effects, such as in microenvironments incertain organ tissues, and global effects may be occurring simul-taneously, but it is possible that these organ-specific effects cantranslate to global changes. Functional studies would be requiredto confirm which of these scenarios plays the more prominent roleduring infection.

In summary, we provide the first systemic, multiorgan meta-bolomic analysis of an infected mammal, using B. anthracis as amodel bacterial pathogen. Downmodulation of lipid mediators ofthe host immune response was observed and was functionallyconfirmed to result in enhanced disease when the enzymes thatgenerate these lipids were inhibited. We propose the use ofmetabolomics to identify new pathogenic processes and biomark-ers for infection.

ACKNOWLEDGMENTS

This work was supported by grants from the National Institutes of Healthto A.W.M. (AI097167 and AI096314) and seed funds from Baylor Collegeof Medicine’s Alkek Center for Metagenomics and Microbiome Research.

Animal studies were performed with the assistance of Sabrina Green.B. anthracis Sterne 7702 and toxin-deficient mutant strains were gener-ously given by Shauna McGillivray of Texas Christian University.

REFERENCES1. Hollywood K, Brison DR, Goodacre R. 2006. Metabolomics: current

technologies and future trends. Proteomics 6:4716 – 4723. http://dx.doi.org/10.1002/pmic.200600106.

2. Fontana JM, Alexander E, Salvatore M. 2012. Translational research ininfectious disease: current paradigms and challenges ahead. Transl Res159:430 – 453. http://dx.doi.org/10.1016/j.trsl.2011.12.009.

3. Zhang A, Sun H, Wang P, Han Y, Wang X. 2012. Recent and potentialdevelopments of biofluid analyses in metabolomics. J Proteomics 75:1079 –1088. http://dx.doi.org/10.1016/j.jprot.2011.10.027.

4. Jansen JJ, Szymanska E, Hoefsloot HCJ, Smilde AK. 2012. Individualdifferences in metabolomics: individualised responses and between-metabolite relationships. Metabolomics 8:94 –104. http://dx.doi.org/10.1007/s11306-012-0414-8.

5. Cockerill FR, III, Smith TF. 2004. Response of the clinical microbiologylaboratory to emerging (new) and reemerging infectious diseases. J ClinMicrobiol 42:2359 –2365. http://dx.doi.org/10.1128/JCM.42.6.2359-2365.2004.

6. Maertzdorf J, Weiner J, Kaufmann SHE. 2012. Enabling biomarkers fortuberculosis control. Int J Tuberc Lung Dis 16:1140 –1148. http://dx.doi.org/10.5588/ijtld.12.0246.

7. Christaki E, Giamarellos-Bourboulis EJ. 2014. The beginning of person-alized medicine in sepsis: small steps to a bright future. Clin Genet 86:56 –61. http://dx.doi.org/10.1111/cge.12368.

8. Schmerler D, Neugebauer S, Ludewig K, Bremer-Streck S, BrunkhorstFM, Kiehntopf M. 2012. Targeted metabolomics for discrimination ofsystemic inflammatory disorders in critically ill patients. J Lipid Res 53:1369 –1375. http://dx.doi.org/10.1194/jlr.P023309.

9. Seymour CW, Yende S, Scott MJ, Pribis J, Mohney RP, Bell LN, ChenYF, Zuckerbraun BS, Bigbee WL, Yealy DM, Weissfeld L, Kellum JA,Angus DC. 2013. Metabolomics in pneumonia and sepsis: an analysis ofthe GenIMS cohort study. Intensive Care Med 39:1423–1434. http://dx.doi.org/10.1007/s00134-013-2935-7.

10. Müller C, Dietz I, Tziotis D, Moritz F, Rupp J, Schmitt-Kopplin P.2013. Molecular cartography in acute Chlamydia pneumoniae infec-tions—a non-targeted metabolomics approach. Anal Bioanal Chem 405:5119 –5131. http://dx.doi.org/10.1007/s00216-013-6732-5.

11. Hoerr V, Zbytnuik L, Leger C, Tam PPC, Kubes P, Vogel HJ. 2012.Gram-negative and gram-positive bacterial infections give rise to a differ-ent metabolic response in a mouse model. J Proteome Res 11:3231–3245.http://dx.doi.org/10.1021/pr201274r.

12. Al-Mubarak R, Vander Heiden J, Broeckling CD, Balagon M, BrennanPJ, Vissa VD. 2011. Serum metabolomics reveals higher levels of polyun-

saturated fatty acids in lepromatous leprosy: potential markers for suscep-tibility and pathogenesis. PLoS Negl Trop Dis 5:e1303. http://dx.doi.org/10.1371/journal.pntd.0001303.

13. Guidi-Rontani C. 2002. The alveolar macrophage: the Trojan horse ofBacillus anthracis. Trends Microbiol 10:405– 409. http://dx.doi.org/10.1016/S0966-842X(02)02422-8.

14. Koehler TM. 2009. Bacillus anthracis physiology and genetics. Mol As-pects Med 30:386 –396. http://dx.doi.org/10.1016/j.mam.2009.07.004.

15. Kolstø A-B, Tourasse NJ, Økstad OA. 2009. What sets Bacillus anthracisapart from other Bacillus species? Annu Rev Microbiol 63:451– 476. http://dx.doi.org/10.1146/annurev.micro.091208.073255.

16. Duong S, Chiaraviglio L, Kirby JE. 2006. Histopathology in a murinemodel of anthrax. Int J Exp Pathol 87:131–137. http://dx.doi.org/10.1111/j.0959-9673.2006.00473.x.

17. Guidi-Rontani C, Weber-Levy M, Labruyère E, Mock M. 1999. Germi-nation of Bacillus anthracis spores within alveolar macrophages. Mol Mi-crobiol 31:9 –17. http://dx.doi.org/10.1046/j.1365-2958.1999.01137.x.

18. Schwartz M. 2009. Dr. Jekyll and Mr. Hyde: a short history of anthrax.Mol Aspects Med 30:347–355.

19. Guichard A, Nizet V, Bier E. 2012. New insights into the biological effectsof anthrax toxins: linking cellular to organismal responses. Microbes In-fect 14:97–118. http://dx.doi.org/10.1016/j.micinf.2011.08.016.

20. Tournier J-N, Rossi Paccani S, Quesnel-Hellmann A, Baldari CT. 2009.Anthrax toxins: a weapon to systematically dismantle the host immunedefenses. Mol Aspects Med 30:456 – 466. http://dx.doi.org/10.1016/j.mam.2009.06.002.

21. Ribot WJ, Panchal RG, Brittingham KC, Ruthel G, Kenny TA, Lane D,Curry B, Hoover TA, Friedlander AM, Bavari S. 2006. Anthrax lethaltoxin impairs innate immune functions of alveolar macrophages and fa-cilitates Bacillus anthracis survival. Infect Immun 74:5029 –5034. http://dx.doi.org/10.1128/IAI.00275-06.

22. Vitale G, Pellizzari R, Recchi C, Napolitani G, Mock M, Montecucco C.1999. Anthrax lethal factor cleaves the N-terminus of MAPKKS and in-duces tyrosine/threonine phosphorylation of MAPKS in cultured macro-phages. J Appl Microbiol 87:288. http://dx.doi.org/10.1046/j.1365-2672.1999.00893.x.

23. Tonello F, Montecucco C. 2009. The anthrax lethal factor and its MAPKkinase-specific metalloprotease activity. Mol Aspects Med 30:431– 438.http://dx.doi.org/10.1016/j.mam.2009.07.006.

24. Firoved AM, Miller GF, Moayeri M, Kakkar R, Shen Y, Wiggins JF,McNally EM, Tang W-J, Leppla SH. 2005. Bacillus anthracis edema toxincauses extensive tissue lesions and rapid lethality in mice. Am J Pathol167:1309 –1320. http://dx.doi.org/10.1016/S0002-9440(10)61218-7.

25. Liu S, Moayeri M, Leppla SH. 2014. Anthrax lethal and edema toxins inanthrax pathogenesis. Trends Microbiol 22:317–325. http://dx.doi.org/10.1016/j.tim.2014.02.012.

26. Guidi-Rontani C, Levy M, Ohayon H, Mock M. 2001. Fate of germi-nated Bacillus anthracis spores in primary murine macrophages. Mol Mi-crobiol 42:931–938. http://dx.doi.org/10.1046/j.1365-2958.2001.02695.x.

27. Kim HU, Goepfert JM. 1974. A sporulation medium for Bacillus anthra-cis. J Appl Bacteriol 37:265–267. http://dx.doi.org/10.1111/j.1365-2672.1974.tb00438.x.

28. Reed L, Muench H. 1938. A simple method of estimating fifty per centendpoints. Am J Epidemiol 27:493– 497.

29. Gage GJ, Kipke DR, Shain W. 2012. Whole animal perfusion fixation forrodents. J Vis Exp 2012:3564. http://dx.doi.org/10.3791/3564.

30. Balderas MA, Nobles CL, Honsa ES, Alicki ER, Maresso AW. 2012. Halis a Bacillus anthracis heme acquisition protein. J Bacteriol 194:5513–5521.http://dx.doi.org/10.1128/JB.00685-12.

31. Evans AM, DeHaven CD, Barrett T, Mitchell M, Milgram E. 2009.Integrated, nontargeted ultrahigh performance liquid chromatography/electrospray ionization tandem mass spectrometry platform for the iden-tification and relative quantification of the small-molecule complement ofbiological systems. Anal Chem 81:6656 – 6667. http://dx.doi.org/10.1021/ac901536h.

32. Watson LE, Kuo S-R, Katki K, Dang T, Park SK, Dostal DE, Tang W-J,Leppla SH, Frankel AE. 2007. Anthrax toxins induce shock in rats bydepressed cardiac ventricular function. PLoS One 2:e466. http://dx.doi.org/10.1371/journal.pone.0000466.

33. Turnbull PC, Broster MG, Carman JA, Manchee RJ, Melling J. 1986.Development of antibodies to protective antigen and lethal factor compo-nents of anthrax toxin in humans and guinea pigs and their relevance toprotective immunity. Infect Immun 52:356 –363.

Nguyen et al.

4824 iai.asm.org December 2015 Volume 83 Number 12Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

34. Grunow R, Porsch-Ozcürümez M, Splettstoesser W, Buckendahl A,Hahn U, Beyer W, Böhm R, Huber M, vd Esche U, Bessler W, Fr-angoulidis D, Finke EJ. 2007. Monitoring of ELISA-reactive antibodiesagainst anthrax protective antigen (PA), lethal factor (LF), and toxin-neutralising antibodies in serum of individuals vaccinated against anthraxwith the PA-based UK anthrax vaccine. Vaccine 25:3679 –3683. http://dx.doi.org/10.1016/j.vaccine.2007.01.056.

35. Ristroph JD, Ivins BE. 1983. Elaboration of Bacillus anthracis antigens ina new, defined culture medium. Infect Immun 39:483– 486.

36. Wold S, Esbensen K, Geladi P. 1987. Principal component analysis.Chemometrics Intell Lab Syst 2:37–52. http://dx.doi.org/10.1016/0169-7439(87)80084-9.

37. Chand HS, Drysdale M, Lovchik J, Koehler TM, Lipscomb MF, LyonsCR. 2009. Discriminating virulence mechanisms among Bacillus anthracisstrains by using a murine subcutaneous infection model. Infect Immun77:429 – 435. http://dx.doi.org/10.1128/IAI.00647-08.

38. Cote CK, Welkos SL, Bozue J. 2011. Key aspects of the molecular andcellular basis of inhalational anthrax. Microbes Infect 13:1146 –1155. http://dx.doi.org/10.1016/j.micinf.2011.07.005.

39. Vescovo G, Ravara B, Gobbo V, Sandri M, Angelini A, Della Barbera M,Dona M, Peluso G, Calvani M, Mosconi L, Dalla Libera L. 2002.L-Carnitine: a potential treatment for blocking apoptosis and preventingskeletal muscle myopathy in heart failure. Am J Physiol Cell Physiol 283:C802–C810. http://dx.doi.org/10.1152/ajpcell.00046.2002.

40. Mutomba MC, Yuan H, Konyavko M, Adachi S, Yokoyama CB, EsserV, McGarry JD, Babior BM, Gottlieb RA. 2000. Regulation of the activityof caspases by L-carnitine and palmitoylcarnitine. FEBS Lett 478:19 –25.http://dx.doi.org/10.1016/S0014-5793(00)01817-2.

41. Haimovitz-Friedman A, Kolesnick RN, Fuks Z. 1997. Ceramide signal-ing in apoptosis. Br Med Bull 53:539 –553. http://dx.doi.org/10.1093/oxfordjournals.bmb.a011629.

42. Stansbie D, Brownsey RW, Crettaz M, Denton RM. 1976. Acute effectsin vivo of anti-insulin serum on rates of fatty acid synthesis and activitiesof acetyl-coenzyme A carboxylase and pyruvate dehydrogenase in liverand epididymal adipose tissue of fed rats. Biochem J 160:413– 416. http://dx.doi.org/10.1042/bj1600413.

43. Fritsche K. 2006. Fatty acids as modulators of the immune response.Annu Rev Nutr 26:45–73. http://dx.doi.org/10.1146/annurev.nutr.25.050304.092610.

44. Frayn KN, Arner P, Yki-Järvinen H. 2006. Fatty acid metabolism inadipose tissue, muscle and liver in health and disease. Essays Biochem42:89 –103. http://dx.doi.org/10.1042/bse0420089.

45. Amaral JJ, Antunes LCM, de Macedo CS, Mattos KA, Han J, Pan J,Candéa ALP, Henriques MDGMO, Ribeiro-Alves M, Borchers CH,Sarno EN, Bozza PT, Finlay BB, Pessolani MCV. 2013. Metabonomicsreveals drastic changes in anti-inflammatory/pro-resolving polyunsatu-rated fatty acids-derived lipid mediators in leprosy disease. PLoS NeglTrop Dis 7:e2381. http://dx.doi.org/10.1371/journal.pntd.0002381.

46. Saddik M, Lopaschuk GD. 1991. Myocardial triglyceride turnover andcontribution to energy substrate utilization in isolated working rat hearts.J Biol Chem 266:8162– 8170.

47. Matthan NR, Resteghini N, Robertson M, Ford I, Shepherd J, PackardC, Buckley BM, Jukema JW, Lichtenstein AH, Schaefer EJ. 2010. Cho-lesterol absorption and synthesis markers in individuals with and withouta CHD event during pravastatin therapy: insights from the PROSPER trial.J Lipid Res 51:202–209. http://dx.doi.org/10.1194/jlr.M900032-JLR200.

48. Hurley BP, McCormick BA. 2008. Multiple roles of phospholipase A2during lung infection and inflammation. Infect Immun 76:2259 –2272.http://dx.doi.org/10.1128/IAI.00059-08.

49. Demirev PA, Fenselau C. 2008. Mass spectrometry for rapid character-ization of microorganisms. Annu Rev Anal Chem (Palo Alto Calif) 1:71–93. http://dx.doi.org/10.1146/annurev.anchem.1.031207.112838.

50. Antunes LCM, Arena ET, Menendez A, Han J, Ferreira RBR, BucknerMMC, Lolic P, Madilao LL, Bohlmann J, Borchers CH, Finlay BB. 2011.Impact of salmonella infection on host hormone metabolism revealed bymetabolomics. Infect Immun 79:1759 –1769. http://dx.doi.org/10.1128/IAI.01373-10.

51. Shin JH, Yang JY, Jeon BY, Yoon YJ, Cho SN, Kang YH, Ryu DH,Hwang GS. 2011. 1H NMR-based metabolomic profiling in mice infectedwith Mycobacterium tuberculosis. J Proteome Res 10:2238 –2247. http://dx.doi.org/10.1021/pr101054m.

52. Twenhafel NA. 2010. Pathology of inhalational anthrax animal models.Vet Pathol 47:819 – 830. http://dx.doi.org/10.1177/0300985810378112.

53. Ghosh M, Tucker DE, Burchett SA, Leslie CC. 2006. Properties of thegroup IV phospholipase A 2 family. Prog Lipid Res 45:487–510. http://dx.doi.org/10.1016/j.plipres.2006.05.003.

54. van der Meer-Janssen YPM, van Galen J, Batenburg JJ, Helms JB. 2010.Lipids in host-pathogen interactions: pathogens exploit the complexity ofthe host cell lipidome. Prog Lipid Res 49:1–26. http://dx.doi.org/10.1016/j.plipres.2009.07.003.

55. Harrison LM, Balan KV, Babu US. 2013. Dietary fatty acids and immuneresponse to food-borne bacterial infections. Nutrients 5:1801–1822. http://dx.doi.org/10.3390/nu5051801.

56. Gimenez AP, Wu Y, Paya M, Delclaux C, Touqui L, Goossens PL. 2004.High bactericidal efficiency of type iia phospholipase A2 against Bacillusanthracis and inhibition of its secretion by the lethal toxin. J Immunol173:521–530. http://dx.doi.org/10.4049/jimmunol.173.1.521.

57. Kelley DS. 2001. Modulation of human immune and inflammatory re-sponses by dietary fatty acids. Nutrition 17:669 – 673. http://dx.doi.org/10.1016/S0899-9007(01)00576-7.

58. Wu Y, Raymond B, Goossens PL, Njamkepo E, Guiso N, Paya M,Touqui L. 2010. Type-IIA secreted phospholipase A 2 is an endogenousantibiotic-like protein of the host. Biochimie 92:583–587. http://dx.doi.org/10.1016/j.biochi.2010.01.024.

59. Leslie CC. 1997. Properties and regulation of cytosolic phospholipase A2.J Biol Chem 272:16709 –16712. http://dx.doi.org/10.1074/jbc.272.27.16709.

60. Shin S, Kim YB, Hur GH. 1999. Involvement of phospholipase A2activation in anthrax lethal toxin-induced cytotoxicity. Cell Biol Toxicol15:19 –29. http://dx.doi.org/10.1023/A:1007546505528.

61. Murakami M, Taketomi Y, Sato H, Yamamoto K. 2011. Secreted phos-pholipase A2 revisited. J Biochem 150:233–255. http://dx.doi.org/10.1093/jb/mvr088.

62. Cui X, Li Y, Li X, Haley M, Moayeri M, Fitz Y, Leppla SH, EichackerPQ. 2006. Sublethal doses of Bacillus anthracis lethal toxin inhibit inflam-mation with lipopolysaccharide and Escherichia coli challenge but haveopposite effects on survival. J Infect Dis 193:829 – 840. http://dx.doi.org/10.1086/500468.

63. Cassol E, Misra V, Holman A, Kamat A, Morgello S, Gabuzda D. 2013.Plasma metabolomics identifies lipid abnormalities linked to markers ofinflammation, microbial translocation, and hepatic function in HIV pa-tients receiving protease inhibitors. BMC Infect Dis 13:203. http://dx.doi.org/10.1186/1471-2334-13-203.

64. Menzel N, Fischl W, Hueging K, Bankwitz D, Frentzen A, Haid S,Gentzsch J, Kaderali L, Bartenschlager R, Pietschmann T. 2012. MAP-kinase regulated cytosolic phospholipase A2 activity is essential for pro-duction of infectious Hepatitis C virus particles. PLoS Pathog 8:e1002829.http://dx.doi.org/10.1371/journal.ppat.1002829.

65. Gerich J, Meyer C, Woerle H, Stumvoll M. 2001. Renal gluconeogenesis:its importance in human glucose homeostasis. Diabetes Care 24:382–391.http://dx.doi.org/10.2337/diacare.24.2.382.

66. Ekberg K, Landau BR, Wajngot A, Chandramouli V, Efendic S, Brunen-graber H, Wahren J. 1999. Contributions by kidney and liver to glucoseproduction in the postabsorptive state and after 60 h of fasting. Diabetes48:292–298. http://dx.doi.org/10.2337/diabetes.48.2.292.

67. Guillem JG, Clemens MG, Chaudry IH, McDermott PH, Baue AE.1982. Hepatic gluconeogenic capability in sepsis is depressed beforechanges in oxidative capability. J Trauma 22:723–729. http://dx.doi.org/10.1097/00005373-198209000-00001.

68. Jones CG, Titheradge MA. 1993. The effect of treatment of the rat withbacterial endotoxin on gluconeogenesis and pyruvate metabolism in sub-sequently isolated hepatocytes. Biochem J 289:169 –172. http://dx.doi.org/10.1042/bj2890169.

69. Kushner I. 1982. The phenomenon of the acute phase response. Ann NY Acad Sci 389:39 – 48. http://dx.doi.org/10.1111/j.1749-6632.1982.tb22124.x.

70. Maslanka K, Smolenska-Sym G, Michur H, Wróbel A, Lachert E, BrojerE. 2012. Lysophosphatidylcholines: bioactive lipids generated during stor-age of blood components. Arch Immunol Ther Exp (Warsz) 60:55– 60.http://dx.doi.org/10.1007/s00005-011-0154-x.

71. Drobnik W, Liebisch G, Audebert F-X, Frohlich D, Gluck T, Vogel P,Rothe G, Schmitz G. 2003. Plasma ceramide and lysophosphatidylcho-line inversely correlate with mortality in sepsis patients. J Lipid Res 44:754 –761. http://dx.doi.org/10.1194/jlr.M200401-JLR200.

Metabolomics of B. anthracis Infection

December 2015 Volume 83 Number 12 iai.asm.org 4825Infection and Immunity

on March 17, 2021 by guest

http://iai.asm.org/

Dow

nloaded from