Ajulemic acid, a synthetic cannabinoid, increases formation of the endogenous proreso

Thread starter #1
Ajulemic acid (AjA), a synthetic nonpsychoactive cannabinoid, and lipoxin A4 (LXA4), an eicosanoid formed from sequential actions of 5- and 15-lipoxygenases (LOX), facilitate resolution of inflammation. The purpose of this study was to determine whether the ability of AjA to limit the progress of inflammation might relate to an increase in LXA4, a known anti-inflammatory and proresolving mediator. Addition of AjA (0—30 μM) in vitro to human blood and synovial cells increased production of LXA4 (ELISA) 2- to 5-fold. Administration of AjA to mice with peritonitis resulted in a 25—75% reduction of cells invading the peritoneum, and a 7-fold increase in LXA4 identified by mass spectrometry. Blockade of 12/15 LOX, which leads to LXA4 synthesis via 15-HETE production, reduced (>90%) the ability of AjA to enhance production of LXA4 in vitro. These results suggest that AjA and other agents that increase endogenous compounds that facilitate resolution of inflammation may be useful for conditions characterized by inflammation and tissue injury.–Zurier, R. B., Sun, Y.-P., George, K. L., Stebulis, J. A., Rossetti, R. G., Skulas, A., Judge, E., Serhan, C. N. Ajulemic acid, a synthetic cannabinoid, increases formation of the endogenous proresolving and anti-inflammatory eicosanoid, lipoxin A4.

THE CANNABIS PLANT HAS BEEN a source of medicinal preparations since the earliest written records on pharmacobotany (1)⇓ . A major obstacle to acceptance of the drug has been its potent psychoactivity. This problem has been studied in recent years in attempts to discover synthetic analogs that would retain medicinal properties without the psychotropic effects. A class of cannabinoid, the carboxyl tetrahydrocannabinols (2)⇓ , shows potential as therapy that is free from cannabimimetic central nervous system activity. These substances, which are metabolites of tetrahydrocannabinol (THC), the psychoactive principle of Cannabis, do not produce behavioral changes in humans at doses several times greater than THC doses given to the same volunteers (3)⇓ . The parent compound in this series, the THC metabolite THC-11-oic acid (Fig. 1⇓ ), is effective in animal models of inflammation and pain at oral doses of 20—40 mg/kg (4⇓ , 5)⇓ . However, more potent activity is needed for clinical use.

It has been known for some time that modifications of the pentyl side chain of THC increase its potency (4)⇓ . In particular, extending the chain length to 7 carbons and introducing branching close to the ring leads to compounds with potencies that are 50—100 times greater than that of THC. This strategy was employed in designing the structure of 1′,1′-dimethylheptyl-THC-11-oic-acid [trivial name ajulemic acid (AjA)] (5)⇓ . This synthetic analog of THC-11-oic acid is a potent anti-inflammatory and analgesic agent in several animal models (6⇓ , 7)⇓ . We have shown (8)⇓ that oral administration of AjA at a dose of 0.1 mg/kg 3×/wk reduces significantly the severity of adjuvant-induced olyarthritis in rats. Histomorphological evaluation of the joints suggested that although synovial inflammation occurred in AjA-treated animals, it did not progress to cartilage degradation, bone erosion, and distortion of joint architecture as observed in rats given placebo. Thus, AjA treatment appears to facilitate resolution of inflammation in this animal model of joint tissue injury.

Lipoxins are a new class of eicosanoids that arise from the sequential actions of lipoxygenases. Because these compounds are generated through an interaction between lipoxygenase pathways, the term lipoxins (lipoxygenase interaction products) was introduced (9)⇓ . In humans, lipoxins are formed in vivo during multicellular responses such as inflammation. They serve as stop signals in that they prevent leukocyte-mediated tissue injury and stimulate the uptake of apoptotic polymorphonuclear leukocytes (PMNs) at sites of inflammation, thereby facilitating resolution of inflammation. A major problem in joints of patients with rheumatoid arthritis is that inflammation–designed to be protective–often does not resolve. Stable analogs of lipoxin A4 (LXA4) block chemotaxis of PMNs and reduce inflammation in animal models (10)⇓ . As noted, AjA treatment of rats with adjuvant arthritis appears to promote resolution of synovial inflammation. Therefore, we designed studies to determine whether AjA–added to cells in vitro or administered to mice in vivo–is associated with an increase in LXA4. We present results here that indicate that AjA does stimulate production of LXA4, a process associated with counterregulation of an inflammatory response in a murine model of peritoneal inflammation.

Materials and Methods
Reagents
AjA was obtained from Organix (Woburn, MA, USA). Its purity was monitored on high-pressure liquid chromatography by comparison with material synthesized previously (11)⇓ . The sample was 97% chemically pure and was >99% chiral pure in the R,R enantiomer. TNFα was from R&D Systems (Minneapolis, MN, USA). Zymosan A, media, and all other reagents were from Sigma Chemical Co. (St. Louis, MO, USA). AjA was dissolved in dimethyl sulfoxide (DMSO), then diluted with minimal essential medium (MEM) and 2% fetal bovine serum (FBS) to achieve appropriate concentrations. The concentration of DMSO was kept constant at 0.3%.

Methods
Human peripheral blood (60 ml) obtained by venepuncture (after informed consent approved by the University of Massachusetts Medical School Committee on Protection of Human Subjects), was mixed with 60 ml phosphate-buffered saline (PBS). Aliquots of 1 ml were used for experiments. Samples from separate donors were run in triplicate.

PMNs were obtained from the red blood cell pellet after density-gradient centrifugation of whole blood using histopaque (Sigma). After removing the plasma layer, the peripheral blood mononuclear cell (PBMC) layer, and the histopaque layer, red blood cells were lysed by adding 5 ml cold distilled water. PMN pellets were collected after centrifuging extracts for 10 min at 900 g.

Fibroblast-like synovial cells (FLSs) were from synovial fluid, as we have described (12)⇓ . Synovial fluid was aspirated from joints of patients with rheumatoid arthritis, inflammatory polyarthritis, or osteoarthritis. Fluid was collected in heparinized syringes, then centrifuged at 300 g for 15 min. The resulting pellets were suspended in 7 ml of MEM with 15% heat-inactivated FBS, 1% nonessential amino acids, and 1% penicillin/streptomycin solution, and were plated in 25-ml tissue culture flasks. Cultures were incubated at 37°C with 5% CO2 for 24—48 h, after which medium was aspirated and cultures were washed with PBS to remove nonadherent cells. Growth medium was replaced every 3 to 4 days. After 10 to 14 days, adherent cells were removed from flasks by trypsinization, washed, and transferred to 6-well tissue culture plates in fresh medium. FLSs were passaged (split 1:3) when they reached confluence, generally at 11 to 14 days. Passages 2 through 6 were used for experiments.

Zymosan A particles were suspended in PBS at 10 mg/ml. Serum (2%) was added, and the zymosan suspension was incubated for 1 h at 37°C, then washed once with PBS. Zymosan was then resuspended in serum-free RPMI at the desired concentration.

Cell viability
The integrity of cell membranes was assessed by Trypan blue exclusion. Viability was >95% in all experiments.

Murine peritonitis
Peritonitis was induced in FVB male mice, 6 to 8 wk of age (Charles River Laboratories, Durham, NC, USA), that were fed laboratory Rodent Diet 5001 (Purina Mills, St. Louis, MO, USA). After anesthesia with isoflurane, AjA (1.5 mg/kg), or vehicle was administered in 100 μl PBS intravenously through the tail vein. In other experiments, AjA in safflower oil, or safflower oil alone, was given by mouth for 3 days to male CD-1 mice (6—8 wk old) before Zymosan A (1 mg/ml PBS) was injected into the peritoneal space. In accordance with the Harvard Medical Area Standing Committee of Animals protocol No. 02570 and the University of Massachusetts Medical School Animal Research Review Group, mice were sacrificed after 2—3 h, and peritoneal lavages were collected in Dulbecco's PBS (minus magnesium and calcium). Aliquots of lavage were stained with Trypan blue, and cells were counted using light microscopy. For differential leukocyte counts, 300 μl of the lavage was added to 300 μl of 15% bovine serum albumin and centrifuged onto microscope slides at 2200 rpm for 4 min using a Cytofuge (Statspin, Norwood, MA). Slides were allowed to air dry, and cells were visualized using a modified Wright-Giemsa stain (Sigma).

Lipoxin A4
Samples were diluted 1:2 with methanol, then acidified to pH 3.5 with 1 N hydrochloric acid, then centrifuged (8000 g for 15 min). Supernatants were then applied to Supelco C18 minicolumns (Supelco, Bellfonte, PA, USA); columns were washed with water, then with 1 ml hexane, and lipoxin was eluted with methyl formate. The column was filled with 99% methyl formate, capped, placed in a 15-ml centrifuge tube, and centrifuged 30 min at 1500 g. Extracted samples were dried under nitrogen, resuspended in new extraction buffer, and stored at −80°C until assay. LXA4 was monitored by ELISA (Neogen Corporation, Lansing, MI, USA). The sensitivity of the assay is ∼10 pg/ml.

Liquid chromatography tandem mass spectrometry (LC/MS/MS) identification of LXA4
Peritoneal lavages were collected, and 2 vol of methanol was added. Samples were then extracted with solid-phase C18 cartridges. The resulting methyl formate eluants were taken to dryness with a stream of nitrogen and resuspended in methanol (100 μl) in preparation for LC/MS/MS analysis. Samples were analyzed using liquid chromatography-photodiode array detector (PDA)-tandem mass spectrometry (LC_PDA_MS/MS) (ThermoFinnigan, San Jose, CA, USA). Reverse-phase LC was conducted with a LUNA C18—2 (100×2×5 mm) column. The column was kept in a column heater (30°C). The LC system used a P-4000 quaternary LC pump (ThermoFinnigan). The column was eluted at a flow rate of 0.2 ml/min with methanol:water:acetic acid (65:35:0.01, v/v/v) from 0 to 10 min.

Statistical analysis
Changes in LXA4 production and in cell counts were analyzed for significance by a one-tailed Student's t test for groups with equal variance. For experiments in which it was necessary to control for variability among groups, significance was assessed by an analysis of covariance (ANCOVA) (13)⇓ . In both cases, a value of P < 0.05 was considered significant.

RESULTS
AjA increased LXA4 production by TNFα stimulated PMNs in a dose-dependent manner (Fig. 2A⇓ ). In a series of 4 experiments, 30 μM AjA increased LXA4 release from TNFα-stimulated PMNs 2.60 ± 0.35-fold (mean±SD); P = 0.04 vs. untreated control cells.

Optimal generation of LXA4 requires transcellular biosynthesis. The addition of AjA to whole blood in vitro also enhanced production of LXA4 in a dose-dependent manner (Fig. 2B⇓ ). In cells in which LXA4 was increased maximally by zymosan, LXA4 was not increased substantially more by AjA (not shown). The observations indicate that AjA is by itself an agonist for LXA4. Although the ratio of AjA-induced LXA4 to baseline LXA4 was similar across experiments, the absolute amounts of LXA4 induced by stimulation of cells by zymosan and/or by AjA varied widely from experiment to experiment. Therefore, results from 3 experiments presented in Fig. 2B⇓ were assessed by ANCOVA.

Although joint tissue injury in patients with rheumatoid arthritis is likely due to a multicellular assault on cartilage and bone, studies in animals and humans suggest that joint damage can proceed with participation of synovial cells alone (14)⇓ . Therefore, we examined the influence of AjA on LXA4 production by 4 different cultures of human FLSs. We have not observed differences in responses of FLSs to AjA whether cells were derived from patients with inflammatory polyarthritis (rheumatoid, psoriatic) or osteoarthritis. Addition of AjA to FLSs increased LXA4 release from unstimulated and TNFα (1 ng/ml)-stimulated cells in a concentration-dependent manner (Fig. 3⇓ ). In these experiments, TNFα did not stimulate LXA4 maximally (not shown). Thus, AjA did stimulate further LXA4 production in TNFα-stimulated cells.

We next assessed the actions of AjA in an in vivo murine model of acute inflammation. Results from these experiments in which 6 CD-1 mice were administered 10 mg/kg/day AjA by mouth for 3 days (total daily dose ∼0.3 mg) before intraperitoneal injection of zymosan are shown in Fig. 4⇓ . AjA treatment reduced the total number of cells invading the peritoneum by 69%.

Having shown an induction of LXA4 by AjA in vitro, we next examined whether LXA4 generation was also increased in vivo. To this end, we analyzed LXA4 in peritoneal lavages from 4 AjA-treated animals by LC/MS/MS and compared these with lavages from 5 mice that received vehicle. Complete MS/MS analysis of the fraction at 7.2 min resulted in a fragmentation pattern consistent with that for LXA (Fig. 5A⇓ ). The prominent MS/MS product ions that are diagnostic for LXA4 were noted at m/z 351 ([M-H]) (Fig. 5A, a⇓ ), m/z 333 ([M-H]—H2O), m/z 315 ([M-H]—2H2O), m/z 307 ([M-H]—CO2 (Fig. 5A, b⇓ ), m/z 289 ([M-H]—H2O—CO2), m/z 271 ([M-H]—2H2O—CO2), m/z 251 ([M-H]—CHO(CH2)4CH3) (Fig. 5A, d⇓ ), m/z 233 ([M-H]—CHO(CH2)4CH3—H2O), m/z 207 ([M-H]—CHO(CH2)4CH3—CO2), m/z 189 ([M-H]—CHO(CH2)4CH3—H2O—CO2), m/z 235 ([M-H]—CHO(CH2)3COOH) (Fig. 5A, c⇓ ), and m/z 115 (CHO(CH2)3COOH) (Fig. 5A, c′⇓ ). Having confirmed the identity of the material eluted at 7.2 min as LXA4 via MS/MS identification of diagnostic ions, it was then determined that this material represented ∼0.5 ng of LXA4 (Fig. 5A⇓ ). When compared to baseline levels of 0.06 ng/mouse obtained in the mice who received vehicle only, this represented a 7.33-fold increase in LXA4 generation (Fig. 5B⇓ ). AjA treatment reduced total cell counts in peritoneal lavages but did not reduce numbers of mononuclear cells significantly (not shown). The reduced total cell count was due to a reduction in the PMN counts in peritoneal lavages from AjA-treated mice (Fig. 5C⇓ ). Taken together, these results demonstrate that LXA4 generation is associated with reduction of PMN infiltration by AjA in vivo.

In an effort to determine which biosynthetic pathway might be influenced by AjA, we treated human whole blood with inhibitors of cytoplasmic phospholipase A2 (cPLA2), and 12/15 lipoxygenase (12/15 LOX). Results suggest that most of the increase in LXA4 initiated by AjA resulted from the conversion of endogenous arachidonic acid to LXA4 by the 15 LOX-initiated route. The addition to cells of the 12/15 LOX inhibitor baicalein blocked the AjA-induced increase in LXA4 substantially (Fig. 6⇓ ). It is likely that AjA stimulated the release of arachidonic acid from phospholipid precursors in both murine exudate cells and human whole blood. However, the addition to mouse peritoneal exudate cells or human whole blood of the cPLA2 inhibitor methylarachidonylfluorophosphate did not alter the AjA-induced increase in LXA4 (data not shown), thus precluding our assessment of specific phospholipases in the process.

DISCUSSION
Results of experiments presented in this report indicate that addition of AjA in vitro to human PMNs, whole blood, or FLSs increases production of LXA4 by these cells. AjA itself is an agonist for LXA4 generation and can stimulate further LXA4 production in activated cells. In addition, administration of AjA in vivo increases LXA4 in and reduces total cellular and PMN infiltration into peritoneal fluid in murine peritonitis, a time-honored model for acute inflammation (15)⇓ . The anti-inflammatory actions of LXA4 are well documented (16)⇓ . It blocks neutrophil activation (17)⇓ and antagonizes peptido-leukotrienes (18)⇓ . In addition, in a murine model of inflammatory ear edema, LXA4 analogs are more potent anti-inflammatory agents than equimolar concentrations of dexamethasone (19)⇓ . Thus, maintaining plasma or tissue levels of AjA should provide a way to limit tissue injury during states of mild chronic inflammation. For example, in a model of chronic inflammation (adjuvant-induced arthritis in rats), oral administration of 0.1 mg/kg AjA 3×/wk for 1 mo prevented joint tissue injury (8)⇓ . In the short-term studies of acute inflammation presented in this article, mice were treated with an oral dose of 10 mg/kg AjA for 3 days or 1.5 mg/kg i.v. once. Use of higher doses of AjA used in these experiments does not preclude the potential clinical use of AjA. Although dose translation from rodent to human is tenuous at best, using rodent body surface area as described by Reagan-Shaw et al. (20)⇓ for converting the 10 mg/kg/day dose of AjA in mice, yields an equivalent dose of 56 mg/day for a 70-kg human. In the only published clinical trial of AjA (21)⇓ , administration of 40 and 80 mg/day AjA for 7 days reduced neuropathic pain without induction of cannabimimietic effects.

Although AjA does increase release of arachidonic acid from cell membranes (22)⇓ , the cPLA2 inhibitor methylarachidonylfluorophosphate did not block AjA-induced increases in LXA4. Thus, either the inhibitor did not block cPLA2 adequately, or sufficient arachidonate was available by virtue of cell activation and/or the action of other phospholipases (23)⇓ . For example, THC-induced release of arachidonic acid from mouse peritoneal cells is mediated by phospholipase D activity (24)⇓ . Blockade of 12/15 LOX, which converts endogenous arachidonic acid to LXA4 by the 15 LOX route (25)⇓ , does interfere with the ability of AjA to enhance production of LXA4. We have also observed (26)⇓ that AjA increases expression of cyclooxygenase 2, leading to an increase in prostaglandin J2 (PGJ2), presumably by an increase in PGD synthase. Both PGD2 and PGJ2, like LXA4, facilitate resolution of inflammation (27⇓ , 28)⇓ .

Efforts to limit tissue injury by facilitating resolution of inflammation in, for example, a disease like rheumatoid arthritis, must include induction of naturally occurring proresolving agonists such as LXA4 by stromal cells, such as fibroblasts, at the site of inflammation. We show here that AjA increases LXA4 secretion from human FLSs derived from patients with inflammatory arthritis. The amounts of LXA4 generated by the FLSs in culture are small in these experiments. However, human synovial fibroblasts exhibit functional LXA4 receptors, and nanomolar concentrations of LXA4 block IL-1β-induced production by these cells of the inflammatory cytokines IL-6 and IL-8, and of matrix metalloproteinases (29)⇓ . Therefore, it is of interest that AjA also reduces production of IL-1β from human peripheral blood and synovial fluid monocytes (30)⇓ , of IL-6 from human monocyte-derived macrophages (31)⇓ , and of matrix metalloproteinases from human synovial cells (32)⇓ .

A major defect in patients with diseases characterized by chronic inflammation is a lack of physiological resolution of inflammation. Given that inflammation is a primitive protective response and given the hard-earned knowledge that suppression of inflammation can be freighted with adverse events (33)⇓ , it makes sense to facilitate the next step in this physiological process: resolution (36)⇓ . As noted, both AjA and LXA4 facilitate resolution of inflammation. Our observation (Fig. 6⇓ ) that inhibition of 12/15 LOX blocks the AjA-induced increase in LXA4 suggests that most of the increase results from the conversion of endogenous arachidonic acid to LXA4 by the 15-lipoxygenase initiated route. Whatever the precise mechanisms whereby AjA increases LXA4, it is clear that reduction by AjA of acute inflammation in a murine model of peritonitis is associated with a concomitant increase in endogenous LXA4.

It has long been known that not all products of the cyclooxygenase pathway are mediators of inflammation (34)⇓ . Indeed, several of the eicosanoids modulate inflammation, and prostaglandin E1 (PGE1) and prostaglandin J2 (PGJ2), for example, function as anti-inflammatory agents (27⇓ , 35)⇓ . Similarly, by virtue of limiting PMN infiltration to the inflamed site, and enhancing clearance of apoptotic PMNs by macrophages (9⇓ , 10)⇓ , LXA4 hastens resolution of inflammation. LXA4, generated in rat kidneys during experimental immune complex-mediated glomerulonephritis, antagonizes the tissue-injuring actions of leukotrienes in this animal model (37⇓ , 38)⇓ . In addition, an LXA4 analog reduces expression of interferon gamma-induced genes associated with nephritis (39)⇓ . Thus, development of analogs of LXA4 (28)⇓ and of compounds that increase endogenous LXA4 production may prove useful as therapeutic agents for diseases characterized by chronic inflammation and related tissue injury.

Source: fasebj.org