Implication of the anti-inflammatory bioactive lipid prostaglandin D2-glycerol ester in the control of macrophage activation and inflammation by ABHD6

Abstract

Proinflammatory macrophages are key mediators in several pathologies; thus, controlling their activation is necessary. The endocannabinoid system is implicated in various inflammatory processes. Here we show that in macrophages, the newly characterized enzyme α/β-hydrolase domain 6 (ABHD6) controls 2-arachidonoylglycerol (2-AG) levels and thus its pharmacological effects. Furthermore, we characterize a unique pathway mediating the effects of 2-AG through its oxygenation by cyclooxygenase-2 to give rise to the anti-inflammatory prostaglandin D2-glycerol ester (PGD2-G). Pharmacological blockade of cyclooxygenase-2 or of prostaglandin D synthase prevented the effects of increasing 2-AG levels by ABHD6 inhibition in vitro, as well as the 2-AG-induced increase in PGD2-G levels. Together, our data demonstrate the physiological relevance of the interaction between the endocannabinoid and prostanoid systems. Moreover, we show that ABHD6 inhibition in vivo allows for fine-tuning of 2-AG levels in mice, therefore reducing lipopolysaccharide-induced inflammation, without the characteristic central side effects of strong increases in 2-AG levels obtained following monoacylglycerol lipase inhibition. In addition, administration of PGD2-G reduces lipopolysaccharide-induced inflammation in mice, thus confirming the biological relevance of this 2-AG metabolite. This points to ABHD6 as an interesting therapeutic target that should be relevant in treating inflammation-related conditions, and proposes PGD2-G as a bioactive lipid with potential anti-inflammatory properties in vivo.

Keywords: COX-2; FAAH; anandamide; glyceryl prostaglandin; prostaglandin synthase.

Fig. 1.

ABHD6 but not MAGL inhibition increases 2-AG levels and reduces macrophage activation. Macrophages were activated by incubation with LPS (100 ng/mL) for 8 h. (A) The ABHD6 inhibitor WWL70 (WWL) dose-dependently increases 2-AG levels. (B) 2-AG and WWL70 dose-dependently decrease LPS-induced IL-1β mRNA expression. (C) 2-AG (10 µM) and WWL70 (10 µM), added 2 h post LPS, reduce IL-1β mRNA expression. (D) ABHD6 inhibition (WWL, 10 µM) decreases LPS-induced PG production in J774 cells. (E) Incubation with 2-AG (10 µM) or WWL (10 µM) decreases LPS-induced NO production by J774 cells. (F) The MAGL inhibitor JZL184 (JZL, 10 µM) does not reduce LPS-induced expression of IL-1β in J774 cells. (G) Following MAGL inhibition, 2-AG levels are not increased in J774 cells (JZL, 10 µM) but are increased in B16 melanoma cells (JZL, 1 µM). (H–J) ABHD6 inhibition (WWL, 10 µM) (H) increases 2-AG levels in thioglycolate-elicited peritoneal macrophages (TGEM) and BV2 microglial-like cells and (I and J) decreases IL-1β mRNA expression in LPS-stimulated (I) TGEM and (J) BV2 cells. Experiments were performed at least three times in triplicate. Compounds were added 1 h before LPS unless otherwise specified. Except for D, G, and H, the effect of LPS on DMSO-treated cells (Veh.) is set at 100%. Values are mean ± SEM. #, P < 0.05; ##, P < 0.01; ###, P < 0.001 for treatments vs. Veh. in LPS-untreated control cells. ***, P < 0.001 for treatments vs. Veh. in the presence of LPS.

Fig. 2.

The effect of ABHD6 inhibition on macrophage activation is COX-2 and PGD synthase dependent. Macrophages were activated by incubation with LPS (100 ng/mL) for 8 h. (A and B) R-flurbiprofen (R-flurbi; 10–100 µM) blocks the effect of ABHD6 inhibition (WWL, 10 µM) on LPS-induced IL-1β mRNA expression in (A) J774 macrophages and (B) thioglycolate-elicited peritoneal macrophages. (C) R-flurbiprofen (100 µM) increases 2-AG levels. (D) Only arachidonic acid-derived monoacylglycerols affect LPS-induced IL-1β expression. PG, palmitoylglycerol; OG, oleoylglycerol. (E) The PGD synthase inhibitor HQL79 (HQL, 3–10 µM) dose-dependently blocks the effect of WWL (10 µM) and 2-AG (10 µM) on IL-1β expression. (F) PGD₂-G (10 µM) decreases LPS-induced IL-1β, whereas PGE₂-G and PGF_2α-G (10 µM) have the opposite effect in J774 cells. (G) Dose-dependent effect of PGD₂-G on LPS-induced IL-1β expression in J774 cells. (H) The effect of PGD₂-G on LPS-induced IL-1β mRNA expression in macrophages is not blocked by antagonists of the PGD₂ receptors: DP1 (BWA686c, 1 µM) and DP2 (CAY10471, 1 µM). (I and J) In J774 cells, LPS increases (I) PGD₂-G and (J) PGE₂-G levels (measured by HPLC-MS) compared with control cells. The PGD synthase inhibitor HQL79 (10 µM) (I) blocks PGD₂-G production and (J) has no effect on PGE₂-G production. (K) PGD₂-G production is increased further when 2-AG (0.1 µM) is added to the cells and decreased by coincubation with R-flurbiprofen (100 µM) or HQL79 (10 µM). Experiments were performed at least three times in triplicate. Compounds were added 1 h before LPS unless otherwise specified. Except for I–K, the effect of LPS on DMSO-treated cells (Veh.) is set at 100%. Values are mean ± SEM. ##, P < 0.01; ###, P < 0.001 for treatments vs. Veh. in LPS-untreated control cells. **, P < 0.01; ***, P < 0.001 for treatments vs. Veh. in the presence of LPS; $$, P < 0.01; $$$, P < 0.001 for antagonists/inhibitors vs. treatment (WWL or 2-AG).

Fig. 3.

ABHD6 inhibition in vivo reduces LPS-induced inflammation. (A–D) Inflammation was induced in C57BL/6 mice (seven mice per group) by i.p. administration of LPS (300 µg/kg) to mice treated with vehicle (Veh.) or WWL70 (WWL, 20 mg/kg) 2 h before LPS. Control mice (Veh.) received i.p. injections of the corresponding vehicles. (A and C) ABHD6 inhibition (WWL) reduced the LPS-induced increase in proinflammatory cytokines in the (A) cerebellum and (C) lung. (B and D) In the same experiment, ABHD6 inhibition increased 2-AG levels in (D) the lung but not in (B) the cerebellum. (E and F) C57BL/6 mice (10 mice per group) received i.p. injections of vehicle (Veh.), JZL184 (JZL, 20 mg/kg), or WWL70 (WWL, 20 mg/kg) once a day for 4 d. Locomotor activity was assessed 4 h after the injection on day 1 and on day 4. (E) MAGL inhibition (JZL) led to a decrease in locomotion compared with the control group (Veh.) on both days 1 and 4, whereas ABHD6 inhibition (WWL) had no effect. (F) 2-AG levels on day 4 were increased in the brain following MAGL inhibition (JZL), but not ABHD6 inhibition (WWL). Values are mean ± SEM. #, P < 0.05; ##, P < 0.01; ###, P < 0.001 vs. control mice receiving only vehicle (Veh.) and no LPS. *, P < 0.05; **, P < 0.01 for treatments vs. Veh., both in the presence of LPS.

Fig. 4.

PGD₂-G reduces LPS-induced inflammation in vivo. Inflammation was induced in C57BL/6 mice (seven mice per group) by i.p. administration of LPS (300 µg/kg). WWL70 (WWL, 20 mg/kg), the CB₁ antagonist SR1 (3 mg/kg), and the CB₂ antagonist AM630 (10 mg/kg) were administered 2 h before LPS. The substrate-selective COX-2 inhibitor R-flurbiprofen (R-flu, 5 mg/kg) and PGD₂-G (20 mg/kg) were administered 30 min before LPS. (A) Spleen weight is reduced by WWL administration to inflamed mice, and this effect is blocked by R-flurbiprofen but not CB₁ or CB₂ antagonism. (B–D) LPS-induced proinflammatory cytokine production in the (B) cerebellum and (C and D) liver is decreased by WWL. (B) R-flurbiprofen leads to a further increase in LPS-induced IL-1β expression in the cerebellum and blocks the effect of WWL. (C and D) COX-2 inhibition and CB₁ antagonism block, in part, the effects of WWL on LPS-induced (C) IL-1β and (D) IL-6 expression in the liver. (E–H) LPS administration leads to increased (E) spleen weight and proinflammatory cytokines expression in the (F) cerebellum and (G and H) liver, which are decreased by PGD₂-G. For A–D, the effect of LPS (Veh.) is set at 100%. Values are mean ± SEM. #, P < 0.05; ##, P < 0.01; ###, P < 0.001 for treatments vs. Veh. in LPS-untreated control cells. *, P < 0.05; **, P < 0.01; ***, P < 0.001 for treatments vs. Veh. in the presence of LPS. $, P < 0.05; $$, P < 0.01 for antagonists/inhibitors vs. WWL.