Biosynthetic metabolomes of cysteinyl-containing immunoresolvents

Abstract

Resolution of inflammation is an active process regulated by specialized proresolving mediators where we identified 3 new pathways producing allylic epoxide–derived mediators that stimulate regeneration [i.e., peptido-conjugates in tissue regeneration (CTRs)]. Here, using self-limited Escherichia coli peritonitis in mice, we identified endogenous maresin (MaR) CTR (MCTR), protectin (PD) CTR (PCTR), and resolvin CTR in infectious peritoneal exudates and distal spleens, as well as investigated enzymes involved in their biosynthesis. PCTRs were identified to be temporally regulated in peritoneal exudates and spleens. PCTR1 and MCTR1 were each produced by human recombinant leukotriene (LT) C₄ synthase (LTC₄S) and glutathione S-transferases (GSTs) [microsomal GST (mGST)2, mGST3, and GST-μ (GSTM)4] from their epoxide precursors [16S,17S-epoxy-PD (ePD) and 13S,14S-epoxy-MaR (eMaR)], with preference for GSTM4. Both eMaR and ePD inhibited LTB₄ production by LTA₄ hydrolase. LTC₄S, mGST2, mGST3, and GSTM4 were each expressed in human M1- and M2-like macrophages where LTC₄S inhibition increased CTRs. Finally, PCTR1 showed potent analgesic action. These results demonstrate CTR biosynthesis in mouse peritonitis, human spleens, and human macrophages, as well as identification of key enzymes in these pathways. Moreover, targeting LTC₄S increases CTR metabolomes, giving a new strategy to stimulate resolution and tissue regeneration.—Jouvene, C. C., Shay, A. E., Soens, M. A., Norris, P. C., Haeggström, J. Z., Serhan, C. N. Biosynthetic metabolomes of cysteinyl-containing immunoresolvents.

The acute inflammatory response is a protective mechanism initiated by the host following an injury and an infection. Ideally, this response should lead to tissue repair and regeneration as well as clearance of the invading organisms, thus allowing complete resolution and a return to homeostasis (1). It is now established that resolution of inflammation is an active process orchestrated by local mediators including specialized proresolving mediators (SPMs) (2–4). These SPMs are biosynthesized de novo from polyunsaturated fatty acids elucidated by this laboratory, which include arachidonic acid (AA)–derived lipoxins (LXs), eicosapentaenoic acid–derived E-series resolvins, and docosahexaenoic acid (DHA)–derived D-series resolvins (RvDs), protectins (PDs), and maresins (MaRs). Each of these structurally distinct families identified in human tissues (5–16) as well as in fish (17, 18) activates resolution programs (2, 19). SPMs actively promote resolution of inflammation by counterregulating proinflammatory mediators such as eicosanoids [e.g., prostaglandins (PGs)] and proinflammatory cytokines (e.g., IL-6), as well as limiting neutrophil infiltration (2, 19). Each also enhances the clearance of infections by activating bacterial phagocytosis and efferocytosis of apoptotic cell debris (19, 20).

Recently, 3 new branching pathways of peptide-containing mediators were uncovered that are produced during infectious inflammation, activate tissue regeneration, and promote resolution of inflammation and infection (21–26). These include MaR conjugates in tissue regeneration (CTRs) (MCTRs), PD CTRs (PCTRs), and resolvin CTRs (RCTRs) (19) (see abbreviations for complete chemical nomenclature in Supplemental Table S1). The complete stereochemistry and physiologic actions of each were confirmed by matching with synthetic materials prepared by total organic synthesis (22–28). The MCTR, PCTR, and RCTR pathways share biosynthetic components with the MaR, PD, and RvD pathways, respectively [mainly the same allylic epoxide precursor 13S,14S-epoxy-MaR (eMaR), 16S,17S-epoxy-PD (ePD), and 7S,8S-epoxy-resolvin] (19). These tissue regeneration metabolomes appear to follow similar sequential enzymatic steps (19, 22, 23, 25). Thus, the enzymes required for initiating the biosynthesis of MCTR1, PCTR1, and RCTR1 are of considerable interest.

The MCTR, PCTR, and RCTR pathways and the cysteinyl leukotriene (LT) (cysLT) pathway utilize some shared enzymes (22). CysLTs, peptide-containing mediators biosynthesized from AA, are potent bronchoconstrictors that play important functions in asthma and allergic reactions (29). AA is first converted into an allylic epoxide by lipoxygenation. This epoxide, LTA₄, is converted by LTC₄ synthase (LTC₄S) into LTC₄ that in turn is successively transformed by γ-glutamyl transferase and dipeptidase to produce LTD₄ and LTE₄ (30). Three other members of the glutathione S-transferase (GST) superfamily, microsomal GST (mGST) type 2 and 3 (mGST2 and mGST3), and soluble GST-μ (GSTM)4 can also catalyze conjugation of LTA₄ with glutathione (22, 30, 31). Because LTC₄S and GSTM4 were found to catalyze the biosynthesis of MCTR1 (22), we tested these 4 enzymes (LTC₄S, mGST2, mGST3, and GSTM4) in PCTR biosynthesis.

Because infections and their complications are important worldwide health issues, with ∼1.7 million adult cases of sepsis annually in the United States alone (32), we investigated infectious self-resolving peritoneal exudates to assess whether MCTRs, PCTRs, or RCTRs are produced in vivo, given their actions in tissue regeneration (23, 25, 26), and to address potential enzymes in their biosynthesis. Here, we report that CTRs are present and temporally produced in both peritonitis exudates and spleens during Escherichia coli infections. The enzymes LTC₄S, mGST2, mGST3, and GSTM4 were each present in both macrophages from infectious exudates and human macrophages, and they each proved to be capable of MCTR and PCTR biosynthesis.

MATERIALS AND METHODS

Materials

FVB male mice (6–8 wk old) were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Three spleen biopsies from healthy middle-aged white females and males were purchased from the Cooperative Human Tissue Network (National Cancer Institute, Philadelphia, PA, USA). Recombinant human LTA₄ hydrolase (LTA₄H), LTC₄S, mGST2, and mGST3 were prepared as previously described in refs. 30, 33, and 34, and TK05 (LTC₄S inhibitor) was earlier identified through high-throughput screening (35). Recombinant human GSTM4 was purchased from Creative Biomart (Shirley, NY, USA). Synthetic eMaR, ePD, [¹³C]₂¹⁵N-MCTR1, [¹³C]₂¹⁵N-MCTR2, and [¹³C]₂¹⁵N-MCTR3 were prepared by the lab of Dr. N. A. Petasis (University of Southern California, Los Angeles, CA, USA); their total organic synthesis will be reported separately. Each was further validated using liquid chromatography–tandem mass spectrometry (LC-MS/MS). Synthetic LTA₄ and deuterium-labeled d8-5-hydroxyeicosatetraenoic acid, d5-RvD2, d5-LXA₄, d4-LTB₄, d4-PGE₂, d5-LTC₄, and d5-LTD₄ to monitor recovery as internal standard were purchased from Cayman Chemicals (Ann Arbor, MI, USA).

Mouse peritonitis

Animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Brigham and Women’s Hospital (2016N000145) and complied with institutional guidelines, the U.S. Department of Agriculture Animal Welfare Act, Public Health Service Policy on Humane Care and Use of Laboratory Animals, and the Institute for Laboratory Animal Research Guide for the Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD, USA). Mice were fed ad libitum with PicoLab Rodent Diet 20-5053 (LabDiet, St. Louis, MO, USA). Mice were injected intraperitoneally with E. coli [10⁵ colony-forming units (CFUs), serotype O6:K2:H1] in sterile saline as previously described in Chiang et al. (36). Peritoneal exudates were harvested by lavage using 5 ml Dulbecco’s PBS (DPBS) without calcium or magnesium at 0 (naive), 4, 12, 24, 48, and 72 h. Spleens were also harvested from these mice at the same time intervals. The samples were then analyzed by LC-MS/MS and flow cytometry.

Lipid mediator SPMs and CTR metabololipidomics

Ice-cold methanol containing deuterium-labeled internal standards including d8-5-hydroxyeicosatetraenoic acid, d5-RvD2, d5-LXA₄, d4-LTB₄, d4-PGE₂, d5-LTC₄, d5-LTD₄, [¹³C] ₂¹⁵N-MCTR1, [¹³C] ₂¹⁵N-MCTR2, and [¹³C] ₂¹⁵N-MCTR3 (500 pg each) were added to each sample before lipid extraction for quantification and recovery of the lipid mediators (LMs). Spleens were gently pressed in the methanol solution using a Tissue Grinder Kit (Kimble Chase Life Science, Vineland, NJ, USA). After centrifugation at 1000 g for 10 min at 4°C, supernatants were collected and LM were extracted per optimized methods using an Extrahera automated extractor (Biotage, Uppsala, Sweden) as described in refs. 23 and 37). Briefly, samples were acidified to an apparent pH 3.5, loaded onto 3-ml SPE Isolute C18 100-mg cartridges (Biotage), and rapidly neutralized with double-distilled water. Two different systems were used for the LM-SPMs and the CTR mediator metabolipidomics (Supplemental Table S2). Methyl formate was used to elute LM-SPMs [system I (37)], and methanol was used to elute MCTR, PCTR, RCTR, and cysLT [system II (23)]. Both fractions were brought to dryness under a gentle stream of nitrogen gas using an automated evaporation system (TurboVap LV; Biotage) and immediately resuspended in a methanol-water mixture (50:50, v/v) for LC-MS/MS injections. The LC-MS/MS consisted of a QTrap 5500 (AB Sciex, Framingham, MA, USA) equipped with an LC-20AD HPLC (Shimadzu, Tokyo, Japan). A Poroshell 120 EC-C18 column (100 mm × 4.6 mm × 2.7 µm; Agilent Technologies, Santa Clara, CA, USA) was kept in a column cover regulated at 50°C. LM-SPMs were eluted from this column according to system I (37 and Supplemental Table S2) with a gradient of methanol-water–acetic acid from 40:50:0.01 (v/v/v) to 100:0:0.01 (v/v/v) at a flow rate of 0.5 ml/min. To monitor and quantify LM-SPM amounts, targeted multiple-reaction monitoring (MRM) and enhanced product ion in a negative mode (ion spray voltage: −4000 V) were used. MCTR, PCTR, RCTR, and cysLT were eluted from the liquid chromatography column according to system II (23 and Supplemental Table S2) consisting of a gradient of methanol-water–acetic acid from 55:45:0.1 (v/v/v) to 100:0:0.1 (v/v/v) at a flow rate of 0.6 ml/min. To monitor and quantify MCTR, PCTR, RCTR, and cysLT amounts, targeted MRM (Supplemental Table S3) and enhanced product ion in a positive mode (ion spray voltage: 5200 V) were used. Their limits of detection were ∼0.1 pg. For each compound, linear calibration curves were obtained using synthetic authentic LMs with r² values of 0.98–0.99. Reported criteria were used for identification of each molecule, including tandem mass spectrometry (MS/MS) matching to at least 6 diagnostic ion fragments per molecule and a matching retention time to authentic synthetic standards, where the synthetic standards were also qualified by NMR. Documentation for each of the products can be found at http://serhanlab.bwh.harvard.edu/wp-content/uploads/2019/05/UPDATED-2019-Spectra-Book.pdf. LC-MS/MS data are presented using Prism software v.7 (GraphPad Software, La Jolla, CA, USA) and Cytoscape software v.3.7.1 (https://cytoscape.org) (38).

Flow cytometry

Human peripheral blood mononuclear cells (PBMCs) and human PBMC–derived macrophages were enumerated and assessed for viability by Trypan blue followed by fixation with a 4% paraformaldehyde solution. FcR-mediated, nonspecific antibody binding was blocked with Human TruStain FcX solution (422302; BioLegend, San Diego, CA, USA). Human cells were extracellularly stained with anti-human allophycocyanin (APC)-conjugated CD54 (clone HA58; 559771; BD Biosciences, San Jose, CA, USA) and anti-human phycoerythrin (PE)-conjugated CD80 (clone 2D10; 305208; BioLegend) for M1-like macrophages, or anti-human APC-conjugated CD206 (clone 19.2; 550889; BD Biosciences), and anti-human peridinin chlorophyll protein–cyanine (Cy) 5.5–conjugated CD163 (clone RM3/1; 326512; BioLegend) for M2-like macrophages (16). For intracellular staining, cells were permeabilized with 1× FoxP3 Perm Buffer (BioLegend) per the manufacturer’s instructions and blocked with 10% human serum. Primary antibody staining was first optimized and occurred individually in 1× FoxP3 Perm Buffer at the following dilutions: 1:500 LTC₄S (polyclonal; ab91507; Abcam, Cambridge, MA, USA), 1:500 mGST2 (polyclonal; ab196503; Abcam), 1:500 mGST3 (polyclonal; ab74749; Abcam), 1:1000 GSTM4 (polyclonal; PA5-77214; Thermo Fisher Scientific, Waltham, MA, USA), or rabbit IgG isotype control (6102; Thermo Fisher Scientific). Secondary antibody staining occurred in 1× FoxP3 Perm Buffer with 1:250 F(ab′)2-donkey anti-rabbit PE-conjugated IgG H+L (polyclonal; 12-4739-81; Thermo Fisher Scientific). Samples were analyzed with a FACS Canto II Flow Cytometer (BD Biosciences) and FlowJo 10 software (BD Biosciences). Debris [forward scatter (FSC)-area (A) vs. side scatter (SSC)-A] and doublets (FSC-height (H) vs. FSC-A → FSC-A vs. FSC-width (W) → SSC-H vs. SSC-A) were excluded before further gating on populations of interest. Unstained controls were used for differentiating cell vs. antibody fluorescence as well as single-stain controls for calculating fluorochrome wavelength overlap compensation. Data are represented as either mean fluorescence intensity (MFI) or total cell number.

Mouse peritoneal cells were enumerated and assessed for viability by Trypan blue followed by fixation with a 4% paraformaldehyde solution. FcR-mediated, nonspecific antibody binding was blocked with anti-mouse CD16/CD32 (clone 93; 14-0161-85; Thermo Fisher Scientific). Peritoneal cells were extracellularly stained with anti-mouse FITC-conjugated Ly6C (clone HK1.4; 128006; BioLegend), anti-mouse PE-conjugated Ly6G (clone 1A8; 127608; BioLegend), anti-mouse PE-Cy7–conjugated CD11b (clone M1/70; 25-0112-81; Thermo Fisher Scientific), anti-mouse peridinin chlorophyll protein–Cy5.5–conjugated CD45 (clone 30-F11; 103132; BioLegend), and anti-mouse APC-conjugated F4/80 (clone BM8; 17-4801-82; Thermo Fisher Scientific) (23). Intracellular staining and analysis were performed as previously described.

Human spleen MCTR, PCTR, and RCTR

Studies were performed in accordance with the Partners Human Research Committee (1999P001279). Human spleen sections (150 mg) were used for conversion time-course studies. Each human spleen section was lightly pressed and dissociated in 2 ml DPBS with calcium and magnesium and incubated with 200 ng of either PCTR1, MCTR1, or RCTR1 for 2 h at 37°C. Aliquots (200 µl) were taken at 0, 5, 30, 60, 90, and 120 min, and methanol was added to stop the reaction. Samples were taken to LC-MS/MS and endogenous levels of each CTR were measured and subtracted from each time point to show the exogenous conversion only.

Conversion of epoxides to LTC₄, MCTR1, and PCTR1 by LTC₄S and GSTs

Human recombinant LTC₄S, mGST2, mGST3, or GSTM4 (1 µg/ml) was incubated with LTA₄ [obtained by hydrolysis of LTA₄ methyl ester with lithium hydroxide (39)], eMaR, or ePD (25 µM) in 25 mM Tris-HCl (pH 8.0), 0.05% Triton X-100, and 5 mM reduced glutathione for 15 min at 37°C as previously described in Dalli et al. (22).

LTA₄, eMaR, and ePD inhibit LTB₄ production

Human recombinant LTA₄H (100 ng/ml) was incubated with synthetic LTA₄ (5 µM) in 10 mM Tris-HCl (pH 8.0) for 30 min at 37°C to monitor the conversion of LTA₄ to LTB₄. For suicide inhibition experiments as previously described in Dalli et al. (40), LTA₄H (100 ng/ml) was incubated first with LTA₄ (5 µM), eMaR (5 µM), or ePD (5 µM) for 15 min and then with LTA₄ (5 µM) for 15 min at 37°C.

Human PBMC-derived macrophage cell culture

Human PBMCs from deidentified healthy human volunteers (Boston Children’s Hospital, Boston, MA, USA) were purchased. PBMCs were isolated by density gradient with a Ficoll-Histopaque (MilliporeSigma, Burlington, MA, USA) as previously described in Dalli et al. 22. Next, cells were incubated in tissue culture–treated 6-well plates (Corning, Corning, NY, USA) for 1 h at 37°C in DPBS^+/+ to promote cell adhesion. Supernatants were decanted, and then adherent cells were cultured for 6 d in RPMI 1640 containing 10% fetal bovine serum (VWR, West Chester, PA, USA), 5 mM l-glutamine (Lonza, Basel, Switzerland), 5% penicillin-streptomycin (Lonza), and either granulocyte M-CSF (20 ng/ml; Peprotech, Rocky Hill, NJ, USA) for M1-like macrophages or M-CSF (20 ng/ml; Peprotech) for M2-like macrophages. After 6 d of differentiation, macrophages were polarized for 48 h toward M1-like macrophages with IFN-γ (20 ng/ml; Peprotech) and LPS (100 ng/ml; Peprotech) or M2-like macrophages with IL-4 (20 ng/ml; Peprotech) added to the RPMI-1640 with 10% fetal bovine serum, 5 mM l-glutamine, and 5% penicillin-streptomycin. Macrophages were incubated with either 0.01% DMSO (vehicle) or 5 μM LTC₄S inhibitor (TK05) for 1 h at 37°C prior to an additional 2 h incubation at 37°C with E. coli (serotype O6:K2:H1) with a 1:50 ratio as previously described in Werz et al. (16).

Mouse visceral and abdominal pain

FVB male mice (8–9 wk old) were habituated daily to a large translucent plexiglass cage. To launch inflammatory visceral and abdominal pain, mice were intraperitoneally injected with 1 mg/ml zymosan A (MilliporeSigma) and 2 μg/ml carbacyclin (a stable prostacyclin analog; Cayman Chemicals) (41, 42). In parallel, 750 ng/ml PCTR1 or vehicle (sterile saline solution) was intraperitoneally injected into these mice. Then, they were visually monitored for a 30-min period for number of writhing responses (wave of constrictions and elongations caudally along the abdominal wall accompanied by twisting of the trunk and extension of the limbs) by a blinded observer.

Statistics

Statistical analyses were performed with Prism software v.7 using 1- and 2-way ANOVA with Tukey’s multiple comparisons test and unpaired 1- and 2-tailed Student’s t tests. Values of P < 0.05 were considered statistically significant.

RESULTS

Identification of PCTRs and cysLTs in peritoneal exudates and spleens during self-limited E. coli peritonitis in mice

We used self-limited peritonitis, namely, an inflammation that resolves on its own (1), to assess whether MCTRs, PCTRs, or RCTRs are produced in vivo and to determine their relationship to agonist-induced cysLTs production. Intraperitoneal inoculation with E. coli (10⁵ CFUs/mouse) induced a self-resolving host response with a maximal polymorphonuclear neutrophil (F4/80⁻Ly6G⁺) infiltration at 12 h followed by a rapid decline (Fig. 1A and Supplemental Fig. S1). Monocyte (F4/80⁻Ly6C⁺) infiltration also peaked at 12 h, whereas the number of macrophages (F4/80⁺CD11b⁺) in the exudates was at its minimum at 12 h. An accumulation of macrophages was observed at 48 h (Fig. 1A and Supplemental Fig. S1). These leukocyte dynamics and exudate numbers were consistent with self-limited peritoneal infection (36).

Figure 1.

Temporal distribution of PCTRs and cysLT during self-limited E. coli peritonitis in mice. Male 6–8-wk-old FVB mice were injected intraperitoneally with a self-limiting dose of E. coli (10⁵ CFUs; serotype O6:K2:H1). Peritoneal exudates (2.5 ml) and spleens (∼100 mg) were harvested at 0 (naive), 4, 12, 24, 48, and 72 h postinfection (2 separate experiments with 3 mice/time point in each experiment). LC-MS/MS metabololipidomics was performed on exudates. A) Left, total number of polymorphonuclear neutrophils (PMNs), monocytes, and macrophages in E. coli–infected peritoneal exudates over a 72-h time course. Right, debris and doublets were excluded before further gating on CD45⁺ cells. Representative flow cytometry gating strategy for identifying macrophages (F4/80⁺CD11b⁺), monocytes (F4/80⁻Ly6C⁺), and neutrophils (F4/80⁻Ly6G⁺) in mouse peritoneal lavage before E. coli inoculation and 12 and 72 h postinfection. *P < 0.05, ***P < 0.001, ****P < 0.0001 (unpaired, 2-tailed Student’s t test vs. 0 h) B) Left, time courses of cysLTs and PCTRs identified in peritoneal exudates during self-limited E. coli peritonitis. Results are expressed as means ± sem. *P < 0.05 (unpaired, 2-tailed Student’s t test vs. 0 h). Right, representative MS/MS spectra for LTC₄ (magnification, ×2; m/z [100–610]) and PCTR1 (magnification, ×10; m/z [180–540]). See Materials and Methods and Supplemental Tables S2 and S3 for LC-MS/MS settings. C) Left, temporal distribution of the PCTRs identified in the spleen during self-limited E. coli peritonitis. Results are expressed as means ± sem. *P < 0.05, **P < 0.01 (unpaired, 2-tailed Student’s t test vs. 0 h). Right, representative MS/MS spectra for PCTR3 (magnification, ×10; m/z [100–440]). See Materials and Methods and Supplemental Tables S2 and S3 for LC-MS/MS settings.

Using targeted LC-MS/MS, PCTR1, PCTR3, MCTR1, MCTR3, RCTR3, LTC₄, LTD₄, and LTE₄ were identified by MS/MS matching in peritoneal exudates during self-limited E. coli peritonitis (Fig. 1B and Supplemental Fig. S2A, B). Each was identified using published criteria obtained from the structure elucidation of MCTRs, PCTRs, and RCTRs, which included specific chromatographic retention times (Supplemental Fig. S2A) and identification of at least 6 characteristic diagnostic ions present in their MS/MS spectra that were matched with synthetic standards (Fig. 1B) (22, 23). Figure 1B shows representative MS/MS spectra for LTC₄ and PCTR1 obtained in the inflammatory exudates. LTC₄ gave the following diagnostic ions of mass-charge ratios (m/z): m/z 626 (= M + H), m/z 608 (= M + H − H₂O), m/z 497, m/z 479 (= 497 − H₂O), m/z 319, m/z 308, m/z 301 (= 319 − H₂O), m/z 189, m/z 171 (= 189 − H₂O), and m/z 131. PCTR1 gave diagnostic ions of m/z 650 (= M + H), m/z 632 (= M + H − H₂O), m/z 614 (= M + H − 2H₂O), m/z 521, m/z 503 (= 521 − H₂O), m/z 418, m/z 400 (= 418 − H₂O), m/z 343, m/z 325 (= 343 − H₂O), m/z 308, m/z 257 (= 275 − H₂O), m/z 245, m/z 231, m/z 227 (= 245 − H₂O), and m/z 179 (Fig. 1B). These matched ions from authentic synthetic PCTR1 (24). PCTR1, PCTR3, MCTR1, MCTR3, RCTR3, LTC₄, LTD₄, and LTE₄ were identified and present in the naive mice before E. coli inoculation (Fig. 1B and Supplemental Fig. S2A, B). The identification and quantification of CTRs was determined using isotopic-labeled internal standards ([¹³C] ₂ ¹⁵N-MCTR1, [¹³C] ₂ ¹⁵N-MCTR2, and [¹³C] ₂ ¹⁵N-MCTR3]. In the exudates, cysLTs were each identified and rapidly formed during the infection as anticipated (Fig. 1B) (25). The terminal cysLT, LTE₄, was statistically significantly increased in the infectious peritoneal exudates 4 h postinfection compared with naive exudates (4.6 ± 0.6 pg/exudates at 0 h vs. 27.2 ± 9.5 pg/exudates at 4 h; Fig. 1B) and decreased 12 h postinfection (2.5 ± 0.5 pg/exudates at 12 h vs. 4.6 ± 0.6 pg/exudates at 0 h; Fig. 1B). The production of PCTR3 also increased during self-limited infectious peritonitis with statistically significantly more PCTR3 at 12 h postinfection (1.5 ± 0.4 pg/exudates at 12 h vs. 0.4 ± 0.2 pg/exudates at 0 h; Fig. 1B). Furthermore, the expression of the enzymes (LTC₄S, mGST2, mGST3, and GSTM4) involved in the biosynthesis of cysLT and potentially in the biosynthesis of each MCTR, PCTR, and RCTR pathway was quantified in infectious exudates. In peritoneal macrophages (CD45⁺F4/80⁺ macrophages), mGST3 and LTC₄S peaked at 12 h, whereas mGST2 and GSTM4 were the most abundant at 48 h (Supplemental Fig. S3).

Given the important role of the spleen as a lymphoid organ in mice (43) and in immune function and production of SPMs (19), we profiled spleens during self-limited infectious peritonitis to determine CTR biosynthesis in the spleen during peritonitis. In the spleen of these mice, PCTR1, PCTR2, PCTR3, MCTR1, MCTR3, RCTR2, RCTR3, LTC₄, LTD₄, and LTE₄ were identified (Fig. 1C and Supplemental Fig. S2C). There were 2 observed waves of biosynthesis, one during the initial phase with a peak at 4 h and another during the resolution phase that was initiated at ∼24 h with a maximum at 48 h (Fig. 1C and Supplemental Fig. S2C). PCTR3 showed statistically significant increases in the spleen during the infection at 4 h postinfection (0.2 ± 0.1 pg/exudates at 0 h vs. 6.0 ± 2.5 pg/exudates at 4 h; Fig. 1C) and 48 h postinfection (0.2 ± 0.1 pg/exudates at 0 h vs. 12.9 ± 5.8 pg/exudates at 48 h; Fig. 1C). Each CTR was identified using published criteria, which includes specific chromatographic retention times (Supplemental Fig. S2A) and identification of at least 6 characteristic diagnostic ions present in their MS/MS spectra (Fig. 1C) (22, 23). PCTR3 gave the following diagnostic ions: m/z 464 (= M + H), m/z 446 (= M + H − H₂O), m/z 428 (= M + H − 2H₂O), m/z 275, m/z 257 (= 275 − H₂O), m/z 231, m/z 201 (= 245 − CO₂), m/z 187 (= 231 − CO₂), and m/z 121. These results indicate that, during self-limited E. coli peritonitis in mice, the spleen, a second distal organ, contributes to the endogenous biosynthesis of PCTRs.

Human spleen converts PCTR1, MCTR1, and RCTR1 to further pathway products

The human spleen is enriched in MCTRs, PCTRs, and especially in RCTRs (23). Hence, we questioned whether MCTR1, PCTR1, or RCTR1 are further metabolized by human spleen. To address this, human spleen tissues were incubated with either PCTR1, MCTR1, or RCTR1, extracted, and samples were taken to LC-MS/MS. Each mediator was again identified using specific chromatographic retention times and at least 6 characteristic diagnostic ions in the MS/MS spectra (Fig. 2A). PCTR3 gave diagnostic ions of (m/z) 464 (= M + H), m/z 446 (= M + H − H₂O), m/z 428 (= M + H − 2H₂O), m/z 343, m/z 325 (= 343 − H₂O), m/z 257 (= 275 − H₂O), m/z 245, m/z 239 (= 275 − 2H₂O), m/z 231, m/z 213 (= 231 − H₂O), and m/z 121 (Fig. 2A). MCTR3 gave diagnostic ions of m/z 464 (= M + H), m/z 446 (= M + H − H₂O), m/z 343, m/z 325 (= 343 − H₂O), m/z 191, m/z 187 (= 205 − H₂O), m/z 173 (= 191 − H₂O), m/z 161 (= 205 − CO₂), m/z 121, and m/z 109 (Fig. 2A). RCTR3 gave diagnostic ions of m/z 480 (= M + H), m/z 462 (= M + H − H₂O), m/z 444 (= M + H − 2H₂O), m/z 341 (= 359 − H₂O), m/z 323 (= 359 − 2H₂O), m/z 247, m/z 217, m/z 203, m/z 199 (= 217 − H₂O), m/z 185 (= 203 − H₂O), and m/z 121 (Fig. 2A). These ions matched those obtained from authentic and synthetic standards (23, 24, 26). PCTR1, MCTR1, and RCTR1 were each rapidly transformed into additional bioactive products with different kinetics (Fig. 2B). MCTR1 conversion was faster than PCTR1 and RCTR1 with MCTR3, representing ∼85% of all MCTRs after 120 min of incubation compared with ∼80% for PCTR3 and ∼45% for RCTR3. Moreover, the conversions of MCTR1 into MCTR2 and RCTR1 into RCTR2 were more rapid than the conversion of PCTR1 into PCTR2. MCTR2 and RCTR2 each increased to ∼50% within the first 5 min of the incubations, whereas PCTR2 increased to only ∼15% compared with time zero. PCTR2, MCTR2, and RCTR2 decreased, whereas PCTR3, MCTR3, and RCTR3 increased (Fig. 2B). These results demonstrate that the human spleen can biosynthesize each of the PCTR3, MCTR3, and RCTR3 from their respective precursors.

Figure 2.

Human spleens convert PCTR1, MCTR1, and RCTR1. Human spleen tissue (150 mg) (n = 3) was incubated with either 200 ng PCTR1, MCTR1, or RCTR1 in DPBS^+/+ at 37°C for 2 h. Aliquots were taken at 5 min and then every 30 min and subjected to LC-MS/MS. A) Representative LC-MS/MS spectra for PCTR3, MCTR3 (magnification, ×6; m/z [100–450]), and RCTR3. See the Materials and Methods section and Supplemental Tables S2 and S3 for LC-MS/MS settings. B) Time course of PCTR1, MCTR1, and RCTR1 conversion by human spleens. Values are represented as means ± sem (n = 3 different human spleens). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (1-way ANOVA with Tukey’s multiple comparison vs. 0 h).

PCTR1 and MCTR1 formation by human LTC₄S and 3 GSTs

Because GST catalyzes the conjugation of glutathione to the epoxide LTA₄ to produce potent bioactive molecules such as LTC₄ (30), we assessed the role of both GST and LTC₄S in the production of CTRs. In cysLT biosynthesis, following the delocalization of the charge over the conjugated triene system of LTA₄ by action of LTC₄S, the allylic epoxide opens and results in the formation of glutathionyl-LTA₄ (Fig. 3A). LTC₄S, mGST2, mGST3, and GSTM4 catalyze the conversion of LTA₄ into LTC₄ (22, 30, 31) and, additionally, LTC₄S and GSTM4 can convert the epoxide eMaR into MCTR1 (22). We determined the ability of these 4 enzymes to convert the epoxides eMaR and ePD into MCTR1 and PCTR1. Each enzyme converted eMaR into MCTR1 and ePD into PCTR1 and was directly compared with the conversion of LTA₄ into LTC₄ (Fig. 3B, C). The identification by LC-MS/MS of each product was confirmed by retention times matched to authentic synthetic standards (Fig. 3B) and the identification of at least 6 diagnostic ions in their MS/MS spectra (Supplemental Fig. S4). In these in vitro conditions, LTA₄ was the preferred substrate for all 4 enzymes. Indeed, LTC₄ represented >75% conversion from epoxide precursor formed by these enzymes. GSTM4 gave the highest conversion of eMaR and ePD (Fig. 3C), suggesting that GSTM4 is the predominant enzyme in the biosynthesis of CTRs. These results show the ability of LTC₄S, mGST2, mGST3, and GSTM4 to convert eMaR and ePD into MCTR1 and PCTR1.

Figure 3.

Selective conversion of ePD and eMaR to PCTR1 and MCTR1 by specific human GST. Each human recombinant enzyme (LTC₄S, mGST2, mGST3, and GSTM4; 1 µg/ml) was incubated with each of the 3 synthetic epoxides (25 µM) in 25 mM Tris-HCl (pH 8.0), 0.05% Triton X-100, and 5 mM reduced glutathione for 15 min at 37°C. LTC₄, MCTR1, and PCTR1 were each identified and quantified by LC-MS/MS. A) Mechanism of the S_N2 nucleophilic substitution reaction on the epoxide leading to the conversion to LTC₄, PCTR1, and MCTR1 by these enzymes. B) Representative MRM chromatograms of LTC₄, MCTR1, and PCTR1 produced by these enzymes. C) Percentage conversion of the epoxides by the 4 enzymes. Data are expressed as the percentages of each conjugate compound out of the total produced by 1 enzyme (n = 3). D) LTA₄, eMaR, and ePD inhibit human LTA₄H. LTA₄H (100 ng/ml) was first incubated with LTA₄ (5 µM), eMaR (5 µM), or ePD (5 µM) for 15 min and then with LTA₄ (5 µM) for 15 min at 37°C for inhibition experiments. Top, quantification of LTB₄ formation following the incubation of LTA₄H and LTA₄. Bottom, percentage inhibition of LTB₄ formation by the 3 epoxides. Values are expressed as means ± sem, n = 3. *P < 0.05, **P < 0.01 (1-tailed Student’s t test vs. LTA₄H + LTA₄ condition).

Because LTA₄ and eMaR each inhibit LTA₄H (40, 44), the enzyme responsible for the conversion of LTA₄ to LTB₄, we tested the ability of ePD to also inhibit LTA₄H. To this end, human recombinant LTA₄H was incubated with each of the epoxides and then LTA₄ was added. Treatment of LTA₄H with LTA₄ led to ∼30% “suicide” inhibition of the enzymes, reducing LTB₄ formation (Fig. 3D). eMaR and ePD were each more potent than LTA₄ at equal molar concentrations, with ∼60% inhibition of LTB₄ production by LTA₄H (Fig. 3D). Therefore, eMaR and ePD are not only the precursors of potent proresolving mediators, but they each prevent on their own the biosynthesis of the potent chemoattractant LTB₄ (29, 30); in this regard, both eMaR and ePD were equally potent.

M1- and M2-like macrophages express LTC₄S, mGST2, mGST3, and GSTM4

Because macrophages play a central role in resolution of inflammation and infection (45), we next assessed the expression of the enzymes involved in CTR biosynthesis. Given the different phenotypes and functions of M1- and M2-like macrophages, it was deemed critical to consider both phenotypes. Both M1- and M2-like macrophages were prepared, and their respective phenotypes were confirmed using cell surface markers identified by flow cytometry (Supplemental Fig. S5A).

Both human M1- and M2-like macrophages expressed a similar distribution of LTC₄S, mGST2, mGST3, and GSTM4 involved in PCTR, MCTR, RCTR, and cysLT biosynthesis (Fig. 4 and Supplemental Fig. S5B, C). GSTM4 was the most abundant enzyme in these macrophages, whereas mGST2 was the least abundant (Fig. 4A). These enzymes were not expressed in human PBMCs (d 0) and their expression increased during differentiation (from d 0 to d 6) and polarization (d 6 to 8) for both M1- and M2-like macrophage phenotypes (Fig. 4B and Supplemental Fig. S5C).

Figure 4.

Human M1- and M2-like macrophages express the enzymes involved in the biosynthesis of the CTRs. Human PBMCs were isolated from the whole blood of 3 individual donors and differentiated into M1-like and M2-like phenotypes. A) MFI of LTC₄S, mGST2, mGST3, and GSTM4 in human PBMCs and M1- and M2-like macrophages. The gray dotted line represents the MFI for the isotype control. Values are represented as means ± sem. Two-way ANOVA with Tukey’s multiple comparisons test. M1- and M2-like compared with PBMC for each enzyme. *P < 0.05, ****P < 0.0001. B) Representative histograms of LTC₄S, mGST2, mGST3, and GSTM4 expression in human M2-like macrophages as identified by flow cytometry at d 0 (undifferentiated PBMCs), d 4 and 6 (during differentiation), and d 8 (postpolarization).

Small-molecule inhibitor of LTC₄S increases CTR production

Inhibition of LT biosynthesis is a target of clinical development, and inhibitors of LTA₄H and 5-lipoxygenase (LOX) were developed that reduce LTB₄ production (46). Recently, a new LTC₄S inhibitor, TK05 (Fig. 5A), was identified by high-throughput screening that targets LTC₄S (35). We tested TK05 on PCTR, MCTR, and RCTR production in M2-like macrophages after E. coli challenge. Challenging TK05-pretreated M2-like macrophages with E. coli inhibited LTC₄, LTD₄, and LTE₄ biosynthesis and increased MCTR2, MCTR3, RCTR3, PCTR1, and PCTR2 (Fig. 5B). These results demonstrate that inhibition of LTC₄S by TK05 enhanced the proresolving CTRs.

Figure 5.

Inhibition of LTC₄S enhances the production of PCTR, MCTR, and RCTR. Human M2-like macrophages were incubated for 1 h with the LTC₄S inhibitor TK05 (5 µM) or 0.01% DMSO (vehicle) at 37°C and then with E. coli for 2 additional hours (O6:K2:H1; ratio, 1:50). A) Schema depicting TK05 structure and mode of action. B) Network pathway visualization of PCTR, MCTR, RCTR, and cysLT amounts in M2-like macrophages. Quantification was obtained by LC-MS/MS analyses. Node size represents the amount of CTRs and cysLTs (pg/10⁷ cells) after TK05 treatment, and node color represents the fold change of each CTR and cysLT after TK05 treatment relative to vehicle (n = 3). See Fig. 7 for structures. DPEP, dipeptidase; GGT, γ-glutamyl transferase.

PCTR1 diminishes inflammatory visceral and abdominal pain

Because resolvins and PDs each reduce pain (47), we investigated whether PCTR1 impacts visceral and abdominal pain. To address visceral and abdominal pain, an abdominal constriction and pain-writhing response was induced in mice (41, 42). To evoke an observational pain response, both zymosan A (1 mg/ml) and a prostacyclin-stable analog (carbacyclin, 2 µg/ml) were administered together and gave substantial writhing responses (Fig. 6). PCTR1 is produced in the picomolar range (in peritoneal exudates during peritonitis) and its administration at the micromolar range reduced the number of writhing responses [12.5 ± 1.3 writhes/30 min for inflammatory pain condition vs. 7.5 ± 1.7 writhes/30 min for inflammatory pain plus PCTR1 condition; P = 0.03 (Fig. 6)]. Hence, PCTR1 acts in the magnitude of concentration required to evoke a visceral pain response by strong inflammatory stimuli. These results indicate an analgesic action of PCTR1 in reducing an inflammatory visceral pain.

Figure 6.

PCTR1 attenuates visceral and abdominal pain in mice. Male FVB age-matched mice were injected intraperitoneally with either 750 ng/ml PCTR1, 1 mg/ml zymosan A, 1 mg/ml zymosan A + 2 µg/ml carbacyclin (inflammatory pain condition), or 1 mg/ml zymosan A + 2 µg/ml carbacyclin + 750 ng/ml PCTR1 (inflammatory pain + PCTR1 condition) and observed over a 30-min period for number of writhing responses. Values are means ± sem (n = 4–13 mice/group). Groups with different letters from each other represent P < 0.05 (unpaired, 2-tailed Student’s t test).

DISCUSSION

During an infection, such as self-limited E. coli peritonitis, SPMs are temporally regulated and accelerate the resolution of infection. Notably, each of the SPMs enhance phagocytosis and killing of E. coli (19, 20, 36). The resolution phase also biosynthesizes mediators that control tissue regeneration (21). Along these lines, we have elucidated structures of a new family of mediators that stimulate tissue regeneration and also possess proresolving functions, namely, MCTRs, PCTRs, and RCTRs (21–26). The structure and stereochemistry of each bioactive member of these families was determined by matching and coelution with synthetic materials prepared by total organic synthesis. Their biosynthesis pathways were assessed using LC-MS/MS–based tracking of deuterium from precursors, and the structures of the bioactive products were deduced from physical properties of methyl esters and Rainey Nickel desulphurization (21–26). Here, we identified MCTR, PCTR, RCTR (Fig. 7), and cysLT in mouse inflammatory exudates and spleens during self-limited peritonitis using their diagnostic ions in MS/MS spectra and chromatographic retention times matching those of authentic and synthetic standards (Fig. 1B, C and Supplemental Fig. S2A). CTRs and cysLTs were each identified in peritoneal exudates, the primary site of infection (Fig. 1B and Supplemental Fig. S2A, B), as well as in a secondary site, which, in this case, is the spleen (Fig. 1C and Supplemental Fig. S2C). Endogenous production of PCTRs and cysLTs was statistically significantly regulated temporally, whereas MCTR and RCTR production was not statistically significant over the time course (Fig. 1B, C and Supplemental Fig. S2).

Figure 7.

MCTR, PCTR, and RCTR biosynthetic pathways and structures. MCTR, PCTR, and RCTR biosynthesis from DHA were established (see main text for details). The complete stereochemistry of MCTRs 1–3, PCTRs 1–3, and RCTRs 1–3 was each established (22–26). In humans, 12-LOX carries the 14-lipoxygenation, and 15-LOX carries the 17-lipoxygenation. By definition, each of the CTRs’ bioactions include proresolving, anti-inflammatory, and tissue regeneration. 14S-HpDHA, 14S-hydroperoxy-DHA.

MCTR, PCTR, and RCTR were each identified earlier in human spleens (23, 25), suggesting biosynthesis of each of the 3 CTR families in humans. Human spleen actively biosynthesizes CTRs and converts each of the precursors (MCTR1, PCTR1, and RCTR1) sequentially to MCTR2, PCTR2, and RCTR2, as well as the respective CTR3s in each biosynthetic pathway (Fig. 2). These results demonstrate for the first time the ability of the human spleen to actively biosynthesize CTRs and are consistent with those obtained earlier for CTR biosynthesis in human macrophages incubated with the intermediate 17-hydroperoxy-DHA, the product of DHA with 15-LOX (25). In human spleen, conversion of MCTR1 into MCTR2, and MCTR3 was more rapid than conversion of either RCTR1 or PCTR1 (Fig. 2B). This robust formation of MCTR also proved to be the case with isolated human recombinant enzymes directly incubated with either eMaR or ePD allylic epoxides (Fig. 3C). Again, MCTR1 was preferentially produced in quantities greater than PCTR1 in side-by-side incubations. These differences suggest that biosynthesis along the 3 separate CTR pathways likely reflects the enzyme affinities for each substrate, given the unique structures of the allylic epoxide intermediates eMaR and ePD.

Similar enzymes appear to be used in the cysLT and CTR pathways, namely, the GSTs. This enzyme family was initially described for their roles in the detoxification of xenobiotic substrates (48) and later recognized for their specific roles in cysLT biosynthesis (30). LTA₄ is converted into LTC₄ by LTC₄S and mGST2, mGST3, and GSTM4 (22, 30), and eMaR is transformed into MCTR1 by LTC₄S and GSTM4 (22). In the present report, we identified LTC₄S and GSTM4, as well as mGST2 and mGST3 for the first time, 2 other members of the membrane-associated proteins in eicosanoids and glutathione metabolism family (30, 31) in peritoneal CD45⁺F4/80⁺ macrophages during self-limited peritonitis in mice (Supplemental Fig. S3). The expression of these enzymes was temporally regulated in the peritoneal infectious exudates. Both LTC₄S and mGST3 peaked at 12 h (Supplemental Fig. S3), which coincided with the peak of inflammation (Fig. 1A). Of interest, both mGST2 and GSTM4 peaked at 48 h (Supplemental Fig. S3), coinciding with the known resolution phase (36) during E. coli peritonitis (Fig. 1A). The temporal separation of these GST enzymes in inflammatory exudates likely reflects their function in the resolution of inflammation and tissue regeneration by their contributions to CTR biosynthesis (2).

We report here for the first time the conversion of eMaR by human recombinant mGST2 and mGST3, as well as the production of PCTR1 independently by these 4 human recombinant enzymes (Fig. 3C). Although LTC₄S, mGST2, mGST3, and GSTM4 can convert the allylic epoxide precursors and share high structural similarity, the production proved different between each combination of enzyme and respective substrate (Fig. 3A–C). Residues in the glutathione binding site are highly conserved in each of these enzymes, whereas they present several changes in the residues in the hydrophobic substrate-binding site that could presumably interact with eMaR and ePD (49, 50). These 4 enzymes gave higher conversion with LTA₄, followed by eMaR and ePD (LTA₄ > eMaR > ePD) (Fig. 3C), which may be caused by the configuration of the different structures of these epoxides in the catalytic site (Fig. 3A). For example, shielding of the ω-end of the allylic epoxides from the membrane by a key residue in the catalytic site, tryptophan 116, leads to a suboptimal orientation and position of the epoxide group for conjugation with glutathione (51). GSTM4 had the highest production of MCTR1 and PCTR1 compared with mGST2 and mGST3 (Fig. 3C). Moreover, GSTM4 was the most abundant of these 4 enzymes in human macrophages (Fig. 4A). These 2 lines of evidence suggest that GSTM4 has a pivotal role in the formation of the first members of the bioactive CTR families, namely, MCTR1, PCTR1, or RCTR1. Together, these results demonstrate a potential new function of these GSTs that can include tissue regeneration and resolution of inflammation via their roles in the biosynthesis of CTRs (23).

The enzymes responsible for LT biosynthesis are targets for drug development (46, 52–54), given the functions of LTs and their exacerbated production in many diseases. Selective inhibitors of LTA₄H have been shown to decrease LTB₄ production and increase LX biosynthesis (46). Along these lines, we found that eMaR inhibits LTB₄ production by LTA₄H (40), and this inhibitory action was also a property of ePD (Fig. 3D). In earlier studies, we found that inhibition of cysLT biosynthesis was associated with an up-regulation in the amount of both resolvins and PDs produced by human macrophages (22). In the present manuscript, using a specific LTC₄S inhibitor, namely, TK05 (35), the biosynthesis in M2-like macrophages of cysLT decreased, and MCTR2, RCTR3, PCTR1, PCTR2, and MCTR3 production increased (Fig. 5B). Thus, targeting LTC₄S with TK05 or related compound is a promising potential therapeutic for reducing inflammation while increasing resolution.

The PD metabolome exerts specific functions in the resolution of inflammation (2) because both PD1 (55) and PCTR1 (24) enhance phagocytosis and macrophage recruitment as well as limit neutrophil infiltration. Moreover, the PCTR family apparently also plays important functions during mouse infectious peritonitis with their different time course between the peritoneal exudates and the spleen (Fig. 1). PCTR1 was temporally regulated during mouse peritonitis (Fig. 1) and was significantly increased by M2-like macrophages when LTC₄S was inhibited (Fig. 5B). In the present experiments, we uncovered that PCTR1 reduced the visceral and abdominal pain responses evoked by zymosan A plus a prostacyclin-stable analog in doses within the same order of magnitude (Fig. 6). Additionally, neuro-PD1 protects against neuropathic pain via activating the G protein-coupled receptor 37 (GPR37) (56, 57), which adds a new member to the known proresolving receptors (58) that evoke resolution programs in multiple organs. For example, RvDs regulate macrophage autophagy (59) and are protective in lung injury (60, 61), ocular disease (62), postmyocardial infarct depression (63), human B-cell IgE production (64), and nociception (65, 66). In general, the actions of CTRs parallel those of resolvins and other proresolving mediators by limiting neutrophil infiltration, stimulating phagocytosis, killing and clearance of microbes (23), wound healing (67), and the return to tissue function via activation of tissue regeneration programs (23, 24), functions uncovered for the CTRs.

In summary, we report on the CTRs and their relationship to cysLT biosynthesis in self-limited peritonitis in mice and in human macrophages. MCTRs, PCTRs, RCTRs, and cysLTs pathways share specific enzymes, namely, LTC₄S, mGST2, mGST3, and GSTM4, with a preference for GSTM4 in the biosynthesis of both MCTR1 and PCTR1 (Fig. 7). The intermediates eMaR and ePD each blocked LTB₄ production by LTA₄H, and specific inhibition of LTC₄S in human macrophages increased PCTR1, PCTR2, MCTR2, MCTR3, and RCTR3. In addition to the proresolving and tissue regeneration actions of CTRs (24, 25), PCTR1 reduced inflammatory visceral and abdominal pain. This study also provides evidence for potentially novel strategies to increase endogenous CTRs to evoke their protective, tissue regeneration, and proresolving functions via selective inhibition of LTC₄S while also reducing cysLTs and their potent contributions to disease pathology.

Glossary

AA

arachidonic acid

APC

allophycocyanin

CFU

colony-forming unit

CTR

conjugate in tissue regeneration

Cy

cyanine

cysLT

cysteinyl LT

DHA

docosahexaenoic acid

DPBS

Dulbecco’s PBS

eMaR

13S,14S-epoxy-maresin

ePD

16S,17S-epoxy-PD

FSC

forward scatter

GST

glutathione S-transferase

GSTM

GST-μ

LC-MS/MS

liquid chromatography–MS/MS

LM

lipid mediator

LOX

lipoxygenase

LT

leukotriene

LTA₄H

LTA₄ hydrolase

LTC₄S

LTC₄ synthase

LX

lipoxin

MaR

maresin

MCTR

MaR CTR

MFI

mean fluorescence intensity

mGST

microsomal GST

MRM

multiple-reaction monitoring

MS/MS

tandem mass spectrometry

PBMC

peripheral blood mononuclear cell

PCTR

PD CTR

PD

protectin

PE

phycoerythrin

PG

prostaglandin

RCTR

resolvin CTR

RvD

D-series resolvin

SPM

specialized proresolving mediator

SSC

side scatter

AUTHOR CONTRIBUTIONS

C. C. Jouvene, A. E. Shay, and C. N. Serhan designed experiments and wrote the paper; C. C. Jouvene, A. E. Shay, M. A. Soens, and P. C. Norris performed experiments and analyzed data; C. N. Serhan conceived and supervised research; J. Z. Haeggström provided reagents and contributed to manuscript preparation; and all authors provided critical review of the manuscript.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.