Proresolving actions of a new resolvin D1 analog mimetic qualifies as an immunoresolvent

Abstract

Resolution of inflammation is an active process driven by several new families of endogenous lipid mediators collectively coined specialized proresolving mediators (SPM). Here, we report a synthetic analog of resolvin D1 (RvD1) and aspirin-triggered RvD1, benzo-diacetylenic-17R-RvD1-methyl ester (BDA-RvD1), which was prepared using fewer steps than required for total organic synthesis of natural SPM. BDA-RvD1 was resistant to further metabolism by human recombinant 15-prostaglandin dehydrogenase, a major inactivation pathway for RvD1. In ischemia-reperfusion-initiated second organ injury, BDA-RvD1 intravenously (1 μg) reduced neutrophil infiltration into the lungs by 58 ± 9% and was significantly more potent than native RvD1. BDA-RvD1 at 100 ng/mouse also shortened the resolution interval, R_i, of Escherichia coli peritonitis with a similar potency as RvD1, by ∼57%, from R_i 10.5 h to 4.5 h. With isolated human phagocytes, BDA-RvD1 at picomolar concentrations (10⁻¹² M) stimulated phagocytosis of zymosan A particles. BDA-RvD1 activated human recombinant G protein-coupled receptor 32/DRV1, an RvD1 receptor, in a dose-dependent manner. These results indicate that, both in vivo in mice and with isolated human cells, BDA-RvD1 shares defining proresolving actions of RvD1, including inhibiting leukocyte infiltration and stimulating phagocytosis. Moreover, they provide evidence for a new analog mimetic and example of an immunoresolvent, namely an agent that stimulates active resolution of inflammation, for a potential new therapeutic class.

inflammation is a critical protective response to acute injury and infection; however, overly robust and persistent responses can lead to tissue damage and chronic inflammation (26, 33). Specialized proresolving mediators (SPMs) are novel families of endogenous lipid mediators including resolvins, protectins, and maresins that signal the active resolution of inflammation. They were identified within self-resolving exudates using a systems approach and liquid chromatography-tandem mass spectrometry (LC-MS-MS)-based lipidomic profiling to limit further polymorphonuclear neutrophil (PMN) infiltration, enhance macrophage uptake of apoptotic PMN, and enhance microbial killing, recently reviewed (26). Total organic synthesis of each SPM facilitated further elucidation of their structure-function relationships (31), including deorphaning target G protein-coupled receptors (GPR) [e.g., GPR32/DRV1 for resolvin D1 (RvD1)] (17), identification of downstream intracellular signaling pathways (9), and discovery of novel in vivo resolving functions, including inhibition of leukocyte infiltration, augmentation of nonphlogistic phagocytosis of apoptotic cells, debris, and pathogens, as well as tissue repair (26). SPMs have delicate chemical structures that signal and are rapidly further metabolized in vivo (31); hence the development of stable synthetic analogs that retain the bioaction of natural mediators is of interest.

Synthetic analogs of SPMs resolvin E1 (RvE1) and lipoxin A₄ (LXA₄) were designed to increase resistance to further metabolic inactivation, improve chemical stability, and facilitate total organic synthesis (2, 32). RvD1 is converted locally to 17-oxo-RvD1, which reduces its activity (31). First-generation RvD1 (7S,8R,17S-trihydroxy-docosa-4Z,9E,11E,13Z,15E,19Z-hexaenoic acid) synthetic analogs including 17-(R/S)-methyl-RvD1 methyl ester and 7R-hydroxy-19-para-fluorophenoxy-RvD1 methyl ester were designed and characterized (12, 14, 34).

RvD1 is a potent SPM, produced endogenously in human cells by sequential reactions of 15-lipoxygenase and 5-lipoxygenase on docosahexaenoic acid (28, 31). There are biologically relevant levels of RvD1 in human tissues including, but not limited to, plasma (8, 21) and breast milk (38). Among its many reported actions (reviewed in Refs. 26 and 30), RvD1 reduces TNF-α signaling (18, 28), decreases pain via regulation of transient receptor potential channels (3, 13), is important in skeletal homeostasis (22), and is protective in experimental models of hyperoxia-induced lung injury (20), fibromyalgia (15), colitis (4), allergic airway, and salivary gland secretion related to Sjögren's syndrome (23, 25). Given the protective bioactions of RvD1, we sought to introduce a new RvD1 analog with increased chemical stability and ease of total organic synthesis that retains the potent biological activities.

Herein, we report the structure and actions of a second-generation RvD1 analog, benzo-diacetylenic-17R-RvD1-methyl ester (BDA-RvD1). Results obtained with this analog indicate that, both in vivo in mice and with isolated human cells, BDA-RvD1 shares defining proresolving actions with RvD1 (i.e., resolving sterile and infective challenges and enhancing phagocytosis by human macrophages). Also, this synthetic analog activates human GPR32, one of the RvD1 receptors (17). Thus BDA-RvD1 is a lead synthetic immunoresolvent, providing a new research tool and a potential new therapeutic class.

MATERIALS AND METHODS

BDA-RvD1 preparation and UV, LC-MS-MS, and GC-MS.

The RvD1 analog, BDA-RvD1, was prepared by total organic synthesis via custom synthesis with Greg Keyes (Cayman Chemical, Ann Arbor, MI). BDA-RvD1 physical properties were determined by UV spectrometry, LC-MS-MS, and gas chromatography/mass spectrometry (GC-MS). An Agilent 8453 UV-Visible Spectrophotometer (Agilent Technologies, Santa Clara, CA) was used to determine the chromophore of BDA-RvD1 when suspended in methanol. LC-MS-MS was performed in the positive ion mode, using a MS QTrap 6500 (AB Sciex, Framingham, MA) that was equipped with a Shimadzu LC-20AD HPLC and a Shimadzu SIL-20AC autoinjector (Shimadzu, Kyoto, Japan). An Agilent Poroshell 120 EC-C18 column (100 mm × 4.6 mm × 2.7 μm) was kept at 50°C, and BDA-RvD1 was eluted with a mobile phase consisting of methanol-water-acetic acid of 50.00:50.00:0.01 (vol/vol/vol) that was ramped to 80.00:20.00:0.01 (vol/vol/vol) from 2 min to 11 min, maintained until 14.5 min and then rapidly ramped to 98.00:2.00:0.01 (vol/vol/vol) for the next 0.1 min. This was subsequently maintained at 98.00:2.00:0.01 (vol/vol/vol) for 5.4 min, and the flow rate was maintained at 0.5 ml/min. The QTrap 6500 was operated in positive ionization mode using multiple-reaction monitoring (MRM) coupled with information-dependent acquisition and an enhanced product ion scan (9). The MRM parameters were collision energy, 20 V; declustering potential, 40 V; entrance potential, 40 V; collision cell exit potential, 15 V; and temperature of ion source, 650°C. A second benzene-containing RvD1 analog, benzo-17R/S-RvD1, was also prepared via a custom synthesis and assessed for biological activity. Benzo-17R/S-RvD1 has the structure of BDA-RvD1 except that it has cis-double bonds in place of the acetylenic bonds of BDA-RvD1, is racemic at the 17-position alcohol, and was used as the carboxy-free acid. Structure of benzo-17R/S-RvD1 was confirmed by LC-MS-MS using conditions reported previously (8). Diagnostic and prominent ions were present including a parent ion with m/z 373 [M-H] and diagnostic fragmentation ions at m/z 355 [M-H-H₂O], m/z 337 [M-H-2H₂O], m/z 311 [M-H-CO₂-H₂O], m/z 288 [306-H₂O], m/z 244 [306-CO₂-H₂O], and m/z 213 [231-H₂O] (Supplemental Material S2; supplemental material for this article is available online at the American Journal of Physiology Lung Cellular and Molecular Physiology website).

Physical properties of BDA-RvD1 and RvD1 were assessed and validated just before each experiment, with RvD1 in accordance with published criteria (31). For GC-MS analysis, BDA-RvD1 was taken to dryness using a stream of N₂, suspended in MeOH (10 μl), and treated with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) [10 min at room temperature, obtained from Supelco (Bellefonte, PA)]. GC-MS was performed with a Hewlett-Packard 593 mass-selective quadrupole detector and HP6890 GC system (Agilent column HP-5MS, 30 mm × 0.25 mm × 0.25 μm). Samples were injected with hexane as the solvent, and the temperature program was initiated at 150°C and held for 2 min and reached 230°C at 10 min (10°C/min) and then 280°C at 20 min (5°C/min). Reference saturated fatty acid methyl ester carbons C₁₆-C₂₄ gave the following retention times (min): C₁₆, 9.2; C₁₈, 11.2; C₂₀, 13.3; C₂₂, 15.7; and C₂₄, 18.2; these were used to calculate the respective C value of synthetic Benzo-RvD1.

Conversion by EOR.

Incubations were conducted using conditions similar to those reported previously (19). Briefly, 0.5 μg of recombinant eicosanoid oxidoreductase (EOR) (Cayman Chemical) and BDA-RvD1, benzo-RvD1, or RvD1 (100 ng) was incubated in buffer containing Tris·HCl (0.1 M, pH 9.0) and NAD+ (1 mM). At indicated intervals, aliquots were taken and subjected to LC-MS/MS (see above).

Ischemia-reperfusion-induced second-organ lung injury.

All experiments were performed in accordance with approval by the Harvard Medical Area Standing Committee on Animals (protocol no. 02570). Mice were anesthetized by intraperitoneal injection of a mixture of xylazine (80 mg/kg) and ketamine (10 mg/kg). Hindlimb ischemia was initiated using tourniquets consisting of a rubber band placed on each hindlimb, as conducted previously (14). Mice were subjected to hindlimb ischemia for 1 h, after which the tourniquets were removed to initiate reperfusion. RvD1 and BDA-RvD1 were each administered for direct comparison at 1 μg/mouse in vehicle (0.1% ethanol in sterile saline) and compared with vehicle alone. They were administered by tail vein 5 min before reperfusion. At the end of a 2-h reperfusion period, mice were euthanized with an overdose of isoflurane, and the lungs were quickly harvested. Lungs were frozen in liquid nitrogen and stored at −80°C or fixed in 10% buffered formalin and processed for histological assessment by the Histology Core of Boston Children's Hospital. Right lungs were homogenized from individual mice and centrifuged, and tissue myeloperoxidase (MPO) levels in the resulting supernatants were determined using a mouse MPO ELISA normalized by tissue weight (R&D Systems, Minneapolis, MN).

E. coli peritoneal infections.

Escherichia coli (E. coli) (serotype O6:K2:H1) were cultured in lysogeny broth, harvested at mid-log phase [OD_600nm≈0.5; 0.5×10⁹ colony-forming units(CFU)/ml], and washed in sterile saline before inoculation into mouse peritoneum. Experiments were carried out with male FVB mice (6–8 wk; Charles River, Wilmington, MA). Mice were administered RvD1 (100 ng/mouse), BDA-RvD1 (100 ng/mouse), benzo-17R/S-RvD1 (100 ng/mouse), or vehicle (saline containing 0.1% ethanol) injections into the peritoneal cavity together with live E. coli (10⁵ CFU), as done previously (5). Mice were euthanized at designated time points by isoflurane overdose, and peritoneal exudates were collected by lavage with 5 ml PBS. Leukocyte numbers and differential counts were assessed using Turk's solution, light microscopy, and flow cytometry analysis. For flow cytometry, aliquots of exudate cells were incubated with anti-mouse CD16/32 blocking antibody (15 min) and then incubated with FITC-conjugated anti-mouse Ly6G (clone 1A8; BD Bioscience, San Jose, CA) and CD11b (clone M1/70; eBioscience, San Diego, CA). PMN were determined as Ly6G- and CD11b-positive events (FACS Canto II). Resolution indices were calculated as before (5), where T_max is the time interval when PMN reaches maximum, T₅₀ is the time interval corresponding to 50% PMN reduction, and the resolution interval (R_i) is the interval between T_max and T₅₀.

Human macrophage phagocytosis.

Human peripheral blood monocytes obtained from deidentified healthy human volunteers were purchased from the Children's Hospital Boston blood bank and isolated by density-gradient Ficoll-Histopaque isolation, as reported previously (17). Peripheral blood mononuclear cells were differentiated to macrophages by 7-day incubation with 10 ng/ml granulocyte-macrophage colony-stimulating factor in RPMI 1640 (10% fetal bovine serum) (17). Macrophages were then incubated with vehicle (0.1% ethanol in Dulbecco's PBS containing Ca²⁺ and Mg²⁺), RvD1 (0.1 pM–10.0 nM), BDA-RvD1 (0.1 pM–10.0 nM), or benzo-17R/S-RvD1 (0.1 pM–100.0 nM) for 15 min at 37°C. FITC-labeled zymosan was then added, and cells were incubated 45 min at 37°C. Phagocytosis was assessed using an M3 SpectraMax plate reader (5, 17).

Ligand-GPR32 interactions using ECIS.

Ligand-receptor interactions were monitored by measuring impedance across cultured Chinese hamster ovary (CHO)-hGPR32 cell monolayers using an electric cell-substrate impedance-sensing system (ECIS; Applied Biophysics, Troy, NY), as before (16). Surface expression of hGPR32 was confirmed by flow cytometry. CHO-hGPR32 cells and wild-type CHO cells were stained with rabbit anti-human GPR32 antibody (GeneTex, Irvine, CA) or nonimmunized rabbit IgG isotype control (30 min) followed by incubation with phycoerythrin-conjugated F(ab′)2 donkey anti-rabbit IgG (20 min). Staining was assessed using FACSDiva CantoII (BD Biosciences) and analyzed using FlowJo (Tree Star, Ashland, OR).

Statistics.

Data are presented as means ± SE. In ischemia-reperfusion, the criterion for statistical significance was P < 0.05 using one-way ANOVA followed by a post hoc Tukey's multiple-comparisons test using GraphPad Prism 6. In E. coli peritonitis and human macrophage incubations, significance was P < 0.05 using Student's t-test when comparing treatment group to vehicle.

RESULTS

Physical characterization of BDA-RvD1.

Given that RvD1 is more prone to enzymatic inactivation and that its total organic synthesis requires many steps, we designed and characterized the physical properties of a more stable synthetic analog BDA-RvD1 (structure shown in Fig. 1A) using UV spectrometry, LC-MS-MS, and GC-MS. The UV spectrum of BDA-RvD1 gave a distinct UV chromophore, with a λ_max in methanol at 248 nm, corresponding to a bathochromic shift of a benzene ring conjugated to a double bond (Fig. 2A). Operated in positive-ion mode, LC-MS-MS of the carboxy-methyl ester BDA-RvD1 gave a peak that eluted at 10.0 min (Fig. 2B). The MS-MS spectrum for the product under the peak gave a parent ion with m/z of 385 [M+H]. Within the same peak and MS-MS spectrum, further evidence for the structure of BDA-RvD1 was provided by the presence of ions with m/z 367 [M+H-H₂O], m/z 349 [M+H-2H₂O], m/z 331 [M+H-3H₂O], m/z 309 [M+H-COOCH₃-H₂O], m/z 291 [M+H-COOCH₃-2H₂O], m/z 273 [M+H-COOCH₃-3H₂O], m/z 241 [259-H₂O], m/z 193 [287-COOCH₃-2H₂O], m/z 181 [199-H₂O], and m/z 152 [185-CH₃-H₂O] (Fig. 2C).

Fig. 1.

Resolvin D1 (RvD1) and benzo-diacetylenic-17R-RvD1-methyl ester (BDA-RvD1), conversion by eicosanoid oxidoreductase (EOR). A: structures of RvD1 and the RvD1 analog BDA-RvD1. B: time course of conversion was monitored by quantifying the formation of 17-oxo-RvD1, a further rapid metabolite of RvD1 by liquid chromatography-tandem mass spectrometry (LC-MS-MS) (see materials and methods). Substrates (∼100 ng) were incubated with recombinant EOR (∼0.5 μg) for the indicated intervals. Results are expressed in nanograms of product; means ± SE; n = 3. C: representative MS-MS spectrum of 17-oxo-RvD1.

Fig. 2.

Physical properties of BDA-RvD1. A: UV absorption spectrum. B: multiple-reaction monitoring (m/z Q1/Q3, 385/309). C: MS-MS fragmentation of BDA-RvD1.

To obtain further evidence in support of the structure of BDA-RvD1, GC-MS of its methyl ester trimethylsilyloxy derivative (OTMS) was conducted. GC-MS spectrum ions of the OTMS-derivatized product (m/z 600) were consistent with the proposed structure of BDA-RvD1, with assigned ions m/z 503 [534-OCH₃], m/z 479 [M-OTMS-OCH₃], m/z 413 [534-OTMS-OCH₃], m/z 361 [M-2OTMS-COOCH₃], m/z 341 [431-OTMS], m/z 323 [355-OCH₃-H], m/z 299 [329-OCH₃+H], m/z 287 [M-3OTMS-CH₃-CH₂-CH₃+H], m/z 197 [228-OCH₃], m/z 191 [431-2OTMS-COOCH₃-H], m/z 181 [271-OTMS], m/z 155 [245-OTMS], and m/z 116 [355-2OTMS-COOCH₃] (see Supplemental Material S1).

BDA-RvD1 is resistant to rapid further metabolism by EOR.

We first questioned whether BDA-RvD1, whose structure was designed to impart stability, is resistant to further metabolism compared with RvD1 (see structures in Fig. 1A). There are several functional groups in endogenous RvD1 that could render it susceptible to local enzymatic inactivation, particularly by EOR, formerly designated 15-prostaglandin dehydrogenase, before its activity on several other families of lipid mediators was known (7). As expected, RvD1 incubated with human recombinant EOR was converted as determined by LC-MS-MS to 17-oxo-RvD1 in a time-dependent manner (Fig. 1B). BDA-RvD1 was not converted to 17-oxo-RvD1 within this rapid time frame (Fig. 1), indicating that this analog is resistant to conversion and inactivation by recombinant EOR.

BDA-RvD1 protects against second-organ lung injury.

RvD1 reduces PMN-mediated tissue injury as demonstrated in ischemia-reperfusion-induced second-organ lung injury (14). Therefore, we next sought to determine whether the RvD1 analog BDA-RvD1 shares RvD1-protective actions in this acute lung injury model. Ischemia was induced by applying tourniquets bilaterally to the hindlimbs of mice for 1 h, followed by tourniquet removal for a 2-h reperfusion period. Five minutes before tourniquet removal, BDA-RvD1 (1 μg/mouse) or vehicle (sterile saline containing 0.1% ethanol) was administered via tail vein. For the purpose of direct comparison, one group of mice was administered RvD1 (1 μg/mouse). After 2 h of reperfusion, lungs were collected for histological and MPO-based assessment of tissue leukocyte infiltration. BDA-RvD1 significantly decreased PMN infiltration to the lungs by 57.8 ± 8.7% compared with vehicle (P < 0.01; Fig. 3A). BDA-RvD1 was significantly more potent than RvD1 (∼3.5×) in protecting lungs after ischemia-reperfusion (P < 0.05; Fig. 3A). Histological assessment confirmed MPO measures of increased PMN infiltration in lungs following ischemia-reperfusion (Fig. 3B, left, middle). Lungs of mice administered BDA-RvD1 before reperfusion displayed fewer infiltrated PMN and a preservation of bronchiolar structure (Fig. 3B, right).

Fig. 3.

BDA-RvD1 protects against second-organ lung injury. Tourniquets were applied bilaterally to the hindlimbs of 6–8-wk-old male FVB mice to initiate ischemia for 1 h. 5 min before reperfusion (tourniquet removal), 1 μg of RvD1, BDA-RvD1, or vehicle (sterile saline containing 0.1% ethanol) was administered via tail vein injection. Lungs were collected following 2 h of reperfusion. A: lung polymorphonuclear leukocyte (PMN) infiltration as assessed by myeloperoxidase (MPO). Results are means ± SE, n = 5 mice per group, **P < 0.01 vs. vehicle; #P < 0.05 between groups. B: tissue histology. Hematoxylin and eosin (H and E) staining of lungs (scale bar = 10 μm; magnification, ×20, inset magnification, ×40). Arrows denote infiltrated PMN. Representative H and E from n = 5 mice. I/R, ischemia-reperfusion.

BDA-RvD1 accelerates resolution of infection.

RvD1 potently accelerates the resolution of E. coli peritonitis, shortening R_i and increasing bacterial clearance (5). We therefore investigated whether BDA-RvD1 also shortens the resolution of infection, comparing its actions directly to RvD1. In these experiments, mice were inoculated with 10⁵ CFU live E. coli (serotype O6:K2:H1) to initiate self-resolving infections (5), along with 100 ng RvD1, BDA-RvD1, or vehicle (sterile saline containing 0.1% ethanol). At 12, 24, and 48 h after E. coli inoculation, peritoneal exudates were collected and leukocytes enumerated. Vehicle-treated E. coli peritonitis gave an R_i of 10.5 h, which is the time for peritoneal exudate PMN number to reduce by half (T₅₀) from their maximum influx values (T_max; ∼10.8 ± 2.2 × 10⁶ cells/exudate, Fig. 4A). Administration of RvD1 accelerated R_i by 52%, shortening resolution from ∼10.5 h to ∼5.0 h (Fig. 4A). BDA-RvD1 was essentially as potent as RvD1, with an R_i of ∼4.5 h, ∼57% shortened R_i compared with vehicle (Fig. 4B). Thus BDA-RvD1 potently accelerates R_i in self-resolving E. coli peritonitis, sharing proresolving bioactions with RvD1. Given the equipotence of BDA-RvD1 compared with RvD1 in stimulating resolution of inflammation during infection, we questioned whether a second benzene ring containing RvD1 analog, benzo-17R/S-RvD1, also has resolving actions during E. coli infection. Benzo-17R/S-RvD1 has essentially a similar structure to BDA-RvD1 (see materials and methods), save the cis-double carbon-carbon bonds in place of the acetylenic carbon-carbon bonds (see Fig. 1), is racemic at the 17-position alcohol, and was used as the carboxy-free acid (Supplemental Material S2). We assessed leukocyte infiltration 24 h postinfection because BDA-RvD1 treatment significantly reduced PMN infiltration at this time point compared with vehicle treatment (Fig. 3B). Self-resolving E. coli peritoneal infections were initiated as described above, with mice administered benzo-17R/S-RvD1 (100 ng/mouse ip) or vehicle at the initiation of infection. Benzo-17R/S-RvD1 decreased infiltrated PMN numbers at 24 h, to 1.5 ± 0.3 × 10⁶ compared with the 6.7 ± 1.7 × 10⁶ in mice given vehicle (P < 0.05, n = 5–6/group) (Supplemental Material S3). Both RvD1 analogs appeared to be equally potent because there was no significant difference in the magnitude of the infiltrated PMN 24 h postinoculation between BDA-RvD1 and benzo-17R/S-RvD1. Thus two benzene ring-containing analogs of RvD1 presented herein decrease activation of leukocyte responses and promote resolution during infection.

Fig. 4.

BDA-RvD1 shortens the resolution interval of Escherichia coli (E. coli) infection. 6–8-wk-old male FVB mice were inoculated intraperitoneally with E. coli (10⁵ colony-forming units). RvD1 (A, dashed line), BDA-RvD1 (B, dashed line), or vehicle (sterile saline containing 0.1% ethanol, A and B solid line) was delivered along with E. coli at the initiation of infection. Peritoneal lavages were collected 12, 24, and 48 h later, and total cell counts were enumerated by light microscopy. Flow cytometry was used to identify PMN. C: resolution indices were calculated. Results are means ± SE, *P < 0.05 vs. vehicle, n = 7–8 mice/group from 3 individual experiments. T_max, time interval when PMN reaches maximum; T₅₀, time interval corresponding to 50% PMN reduction; R_i, resolution interval.

BDA-RvD1 stimulates human macrophage phagocytosis.

Phagocytic uptake of pathogens, cellular debris, and apoptotic cells by leukocytes is a critical process in the resolution of infection and inflammation (26). Therefore, we next questioned whether BDA-RvD1 exerts proresolving actions on human leukocytes challenged with opsonized zymosan, which triggers phagocytosis via highly conserved Fc-mediated signaling pathways. Human macrophages were incubated with BDA-RvD1 before the addition of opsonized zymosan (1:10). BDA-RvD1 dose dependently increased phagocytosis ∼24% at 10 pM (Fig. 5; P < 0.05 vs. vehicle), giving a similar potency as RvD1 at equimolar concentration (10 pM; ∼30%, P < 0.05 vs. vehicle). With similar potencies to RvD1 and BDA-RvD1, benzo-17R/S-RvD1 increased phagocytosis in a dose-dependent manner (10 pM; ∼27%, P < 0.05 vs. vehicle, n = 5 human donors) (Supplemental Material S4). BDA-RvD1 and benzo-17R/S-RvD1 each significantly enhanced human macrophage phagocytic activity with a similar potency as RvD1.

Fig. 5.

BDA-RvD1 increases human macrophage phagocytosis. Monocyte-derived human macrophages were incubated with RvD1 (solid line) or BDA-RvD1 (dashed line) (0.1 pM-10.0 nM, 15 min) followed by addition of opsonized zymosan for 45 min (37°C). Results are means ± SE, n = 3 healthy subjects, *P < 0.05 vs. vehicle.

BDA-RvD1 is an agonist of human GPR32.

Given the proresolving actions of BDA-RvD1 on human cells, we next sought to examine the ligand-receptor interactions of BDA-RvD1 with the recently identified RvD1 receptor GPR32 (17). We used ECIS (24) to monitor receptor-mediated impedance changes across monolayers of human GPR32-overexpressing CHO cells (Fig. 6A). BDA-RvD1 elicited rapid and dose-dependent (10-1,000 nM) impedance changes (Fig. 6, B and C), with a similar potency to that obtained with RvD1 at 100 nM (Fig. 6B). The rate of GPR32 activation was associated with the concentration of BDA-RvD1, giving an R² = 0.825 when plotting the slopes of the first-order derivative of each dose (0–1,000 nM) between 0–2 min (data not shown). Thus, like RvD1, BDA-RvD1 activates human recombinant GPR32.

Fig. 6.

BDA-RvD1 activates human recombinant G protein-coupled receptor 32 (GPR32). Ligand-receptor-dependent impedance changes in human GPR32-overexpressing Chinese hamster ovary (CHO) cells. A: histograms showing GPR32 expression in hGPR32-transfected CHO cells. Control unstained cells and isotype Ab-stained hGPR32-transfected CHO cells are also shown. B and C: impedance was continuously recorded with real-time monitoring across cell monolayers using an electric cell-substrate impedance-sensing system. Results are mean tracings from incubations of CHO-hGPR32 cells with 100 nM RvD1 or 100 nM BDA-RvD1 (n = 3, d = 2) (B), and 10 nM, 100 nM, or 1,000 nM BDA-RvD1 (n = 3, d = 2) (C). PE, phycoerythrin.

DISCUSSION

In the present report, we provide structural and biological evidence for BDA-RvD1, a new RvD1 analog mimetic. This analog was designed to retain and stabilize the backbone orientation of RvD1 postulated to possess the active pharmacophore (Fig. 1) and to reduce rapid inactivation. We report the unique physical properties of BDA-RvD1 using UV spectrometry, LC-MS-MS, and GC-MS, as well as its resistance to further metabolism by recombinant EOR. Sharing bioactions with RvD1 in vivo, BDA-RvD1 protected against second-organ ischemia-reperfusion-induced lung injury. BDA-RvD1 also enhanced host responses against in vivo infection in mice by accelerating R_i of E. coli peritonitis. BDA-RvD1 also shares RvD1 bioactions on human cells, as evidenced by the fact that BDA-RvD1 dose dependently increases phagocytosis of opsonized zymosan. Furthermore, we found that BDA-RvD1 activated GPR32 in a recombinant human GPR32-transfected CHO cell system. This is a GPR for RvD1 (16). Together, these results provide evidence for the function of a novel immunoresolvent designed as a mimetic for RvD1-initiated proresolving programs.

SPM are endogenously biosynthesized molecules that actively regulate the resolution of inflammatory responses. Dysregulation in RvD1 production and signaling is associated with human disease states involving chronic inflammation, including dementia (37). Both RvD1 and aspirin-triggered RvD1 (AT-RvD1) administration are beneficial in in vivo models of obesity (6), allergic airway responses (25), cognitive decline (35), resolution of inflammation, and tissue healing (as recently reviewed in Ref. 26). Synthetic RvD1 and AT-RvD1 preparation by total organic synthesis requires the use of complex synthesis routes with many steps (30). Additionally, RvD1 is further metabolized enzymatically by EOR (31), and the cis-double bonds and conjugated double-bond systems render it chemically labile. Hence, herein we report on an RvD1 analog that was designed and prepared without several of these inactivation-prone or chemically labile moieties while maintaining bioactive properties (see Fig. 1).

BDA-RvD1 is a synthetic analog of RvD1, produced from readily available precursors through a convergent coupling strategy similar to that used earlier for synthesis of RvD1 and benzo-LXA₄ (31, 32). Replacement of the E,E,Z,E tetraene backbone with a benzene ring adds further chemical stability to the molecule while maintaining the proposed pharmacophore of RvD1. To impart resistance to further local cellular metabolism (i.e., dehydrogenation) by oxidoreductases, R chirality of the 17-carbon hydroxyl group was incorporated into BDA-RvD1, similar to the metabolic resistance of AT-RvD1 (31). Resistance to metabolism by EOR was confirmed by the lack of BDA-RvD1 conversion to 17-oxo-BDA-RvD1 (Fig. 1). A second RvD1 analog also containing a benzene ring, namely benzo-17R/S-RvD1, was also described and characterized herein that contains the stabilizing benzene ring in place of the E,E,Z,E portion of the RvD1 backbone and is racemic at the 17 carbon. These design strategies employ similar principles as those used in the development of RvE1 and LXA₄ analogs (2, 29, 32), which enabled the identification of novel functions of these SPM in a range of experimental systems (11, 27, 30, 36). Another modification strategy with SPM analogs introduced with BDA-RvD1 is the addition of two acetylenic bonds in place of labile double bonds. Physical properties of BDA-RvD1 were described herein, including a UV spectrum with a distinct chromophore and LC-MS-MS and GC-MS profiles with diagnostic-specific ions (Fig. 2).

Lungs are prone to severe damage following surgical procedures, where vessels serving the surgical site are clamped to avoid blood loss (10). The ensuing ischemia activates circulating leukocytes, particularly PMNs, which infiltrate tissues upon reperfusion via permeated endothelial cells and release substantial inflammatory factors (10). RvD1 protects against ischemia-reperfusion-induced lung injury (14). Herein, we described that BDA-RvD1 shares this organ-protective action. When administered before reperfusion, BDA-RvD1 reduced infiltrating PMN (Fig. 3). BDA-RvD1 was more potent than RvD1, potentially attributable to a resistance to further conversion. Lung tissue has high expression of EOR (1); thus it is possible that the higher potency of BDA-RvD1 is due to a resistance to local lung metabolism because RvD1 is converted to its oxo-metabolites by this enzyme (31). The bioactions of BDA-RvD1 in hindlimb second-organ injury of the lung provide an in vivo demonstration of bioactions shared with RvD1 in a sterile injury by leukocytes from within.

Given the increasing prevalence of antibiotic-resistant bacteria strains, there is a need for new treatment strategies. RvD1 is protective in E. coli infection, increasing phagocyte uptake of E. coli, decreasing bacterial load, accelerating R_i, and increasing survival (5). RvD1 enhances the antimicrobial activity of the common antibiotics ciprofloxacin and vancomycin by augmenting the host's own antimicrobial responses. In the present studies, BDA-RvD1 elicited similar in vivo actions as RvD1 in self-resolving E. coli infection. BDA-RvD1 was essentially equipotent as RvD1 in accelerating the resolution of infections, by 57% and 52%, respectively. This equipotence in the peritoneum contrasts with the increased potency of BDA-RvD1 compared with RvD1 found in the lung, which may be attributable to the high local expression of EOR in the lung as noted above. Benzo-17R/S-RvD1 also shared bioactions with RvD1 by giving a significant reduction in infiltrated PMN at 24 h postinoculation, with a similar potency to BDA-RvD1 (Fig. 4). The resolving actions during infection shared with RvD1 provide further evidence for in vivo proresolving programs by an analog mimetic, namely BDA-RvD1. Thus SPM and their analog mimetics may be useful therapeutic tools for microbial infections.

One hallmark mechanism of resolution for SPM, including RvD1, is increasing phagocytosis by human leukocytes. With a similar potency to RvD1, we found that BDA-RvD1 and benzo-17R/S-RvD1 dose dependently augmented phagocytosis of opsonized zymosan by differentiated primary human macrophages (Fig. 5). The bell-shaped dose response of each compound on phagocytic action of human cells is in line with our earlier findings with SPM (5) and could be the result of several physiological events, including the activation of yet to be defined receptors at higher concentrations, as has been shown with the proinflammatory lipid mediator leukotriene B₄ and its agonist actions on receptors BLT1 (high affinity) and BLT2 (low affinity) (39). SPM evoke proresolving actions via GPRs (27); RvD1 is a potent agonist for human GPR32 (16, 17). We found that BDA-RvD1 dose dependently activated human GPR32 (Fig. 6). The ability of BDA-RvD1 to activate GPR32 in vitro provides evidence for shared activation pathways with RvD1, in line with proresolving actions both in mice in vivo and in isolated human cells in vitro.

In summary, herein, we describe the structure and proresolving actions of a novel synthetic RvD1-stable analog that is a functional mimetic of RvD1. BDA-RvD1 is a second-generation RvD1 synthetic analog that stabilizes the backbone configuration of native RvD1 with the presence of a ring and rapid local metabolism resistance with the R-chiral hydroxy group at the 17-carbon position. Given the complex chemical routes required to synthesize native SPM, BDA-RvD1 is an example of a potential cost-effective RvD1 analog mimetic that will permit further studies on the resolution of inflammation and may also be useful in the design of therapeutics targeting resolution pathways.

GRANTS

This work was supported in part by NIH grant no. P01GM095467 (C. N. Serhan) and by an Institut Mérieux Research Grant. S. Orr was supported by a Canadian Institutes of Health Research Fellowship Award.

DISCLOSURES

C. N. Serhan is an inventor on patents (resolvins) assigned to BWH and licensed to Resolvyx Pharmaceuticals. C. N. Serhan is a scientific founder of Resolvyx Pharmaceuticals and holds equity in the company of no known value. C. N. Serhan's interests were reviewed and are managed by the Brigham and Women's Hospital and Partners HealthCare in accordance with their conflict of interest policies.

AUTHOR CONTRIBUTIONS

Author contributions: S.K.O. and C.N.S. conception and design of research; S.K.O. and R.A.C. performed experiments; S.K.O. and R.A.C. analyzed data; S.K.O., R.A.C., J.D., N.C., and C.N.S. interpreted results of experiments; S.K.O. and R.A.C. prepared figures; S.K.O. drafted manuscript; S.K.O., R.A.C., J.D., N.C., and C.N.S. edited and revised manuscript; S.K.O., R.A.C., J.D., N.C., and C.N.S. approved final version of manuscript.