Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2011 Aug;163(7):1533-49.
doi: 10.1111/j.1476-5381.2011.01444.x.

Inhibition of COX-2 Expression by Endocannabinoid 2-arachidonoylglycerol Is Mediated via PPAR-γ

Affiliations
Free PMC article

Inhibition of COX-2 Expression by Endocannabinoid 2-arachidonoylglycerol Is Mediated via PPAR-γ

Huizhi Du et al. Br J Pharmacol. .
Free PMC article

Abstract

Background and purpose: Endocannabinoids have both anti-inflammatory and neuroprotective properties against harmful stimuli. We previously demonstrated that the endocannabinoid 2-arachidonoylglycerol (2-AG) protects hippocampal neurons by limiting the inflammatory response via a CB(1) receptor-dependent MAPK/NF-κB signalling pathway. The purpose of the present study was to determine whether PPARγ, an important nuclear receptor, mediates 2-AG-induced inhibition of NF-κB phosphorylation and COX-2 expression, and COX-2-enhanced miniature spontaneous excitatory postsynaptic currents (mEPSCs).

Experimental approach: By using a whole-cell patch clamp electrophysiological recording technique and immunoblot analysis, we determined mEPSCs, expression of COX-2 and PPARγ, and phosphorylation of NF-kB in mouse hippocampal neurons in culture.

Key results: Exogenous and endogenous 2-AG-produced suppressions of NF-κB-p65 phosphorylation, COX-2 expression and excitatory synaptic transmission in response to pro-inflammatory interleukin-1β (IL-1β) and LPS were inhibited by GW9662, a selective PPARγ antagonist, in hippocampal neurons in culture. PPARγ agonists 15-deoxy-Δ(12,14) -prostaglandin J(2) (15d-PGJ(2)) and rosiglitazone mimicked the effects of 2-AG on NF-κB-p65 phosphorylation, COX-2 expression and mEPSCs, and these effects were eliminated by antagonism of PPARγ. Moreover, exogenous application of 2-AG or elevation of endogenous 2-AG by inhibiting its hydrolysis with URB602 or JZL184, selective inhibitors of monoacylglycerol lipase (MAGL), prevented the IL-1β- and LPS-induced reduction of PPARγ expression. The 2-AG restoration of the reduced PPARγ expression was blocked or attenuated by pharmacological or genetic inhibition of the CB(1) receptor.

Conclusions and implications: Our results suggest that CB(1) receptor-dependent PPARγ expression is an important and novel signalling pathway in endocannabinoid 2-AG-produced resolution of neuroinflammation in response to pro-inflammatory insults.

Figures

Figure 1
Figure 1
Exogenous application of 2-AG suppresses NF-κB phosphorylation and COX-2 expression and elevates PPARγ expression in response to pro-inflammatory IL-1β and LPS insults. Hippocampal neurons in culture were treated with IL-1β (10 ng·mL−1) for 6 h or LPS (1 µg·mL−1) for 16 h in the absence and presence of 2-AG or GW9662 (5 µM). The different time points used for the treatments of IL-1β and LPS were based on our previous studies where we identified that IL-1β- or LPS-induced COX-2 expression and NF-kB phosphorylation reached the peaks at these time points (Zhang and Chen, 2008). 2-AG or GW9662 was added 30 min before IL-1β or LPS application. (A1–A2) Immunoblot analysis of 2-AG suppression of LPS-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (B1–B2) Immunoblot analysis of 2-AG suppression of IL-1β-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (C1–C2) 2-AG restores LPS-induced down-regulation of PPARγ (n = 3). (D1–D2) 2-AG restores IL-1β-induced down-regulation of PPARγ (n = 3). **P < 0.01, compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §P < 0.05, §§P < 0.01 compared with IL-1β+2-AG or LPS+2-AG.
Figure 2
Figure 2
2-AG inhibits COX-2 elevation-induced enhancement of mEPSCs. Hippocampal neurons in culture were treated with IL-1β (10 ng·mL−1) for 16 h or LPS (2 µg·mL−1) for 24 h in the absence and presence of 2-AG (1 µM) or GW9662 (5 µM). The different time points used for the treatment of IL-1β and LPS were based on our previous studies where we identified that IL-1β or LPS significantly enhanced synaptic activity at these time points (Sang et al., 2005; Zhang and Chen, 2008). (A1) Representative sweeps of mEPSCs recorded in vehicle control-, LPS-, LPS + 2-AG- and LPS + 2-AG+GW9662-treated neurons. Scale bar: 20 pA/2 s. (A2) Cumulative probability of mEPSCs frequency recorded in neurons with different treatments. (A3) Mean percentage changes in the frequency of mEPSCs in neurons with different treatments. (A4) Cumulative probability of mEPSCs amplitude. (A5) Mean percentage changes in the amplitude of mEPSCs. (B1) Representative sweeps of mEPSCs recorded in vehicle control-, IL-1β-, IL-1β+2-AG (1 µM)- and IL-1β+2-AG+GW9662 (5 µM)-treated neurons. (B2) Cumulative probability of mEPSCs frequency recorded in neurons with different treatments. (B3) Mean percentage changes in the frequency of mEPSCs. (B4) Cumulative probability of mEPSCs amplitude. (B5) Mean percentage changes in the amplitude of mEPSCs. **P < 0.01 compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §§P < 0.01 compared with IL-1β+2-AG or LPS+2-AG (n = 24–32).
Figure 3
Figure 3
Endogenous 2-AG suppresses NF-κB phosphorylation and COX-2 expression and prevents down-regulation of PPARγ expression induced by IL-1β and LPS. Hippocampal neurons in culture treated with IL-1β (10 ng·mL−1) or LPS (1 µg·mL−1) were the same as described in Figure 1. Selective MAGL inhibitors URB602 (URB, 10 µM) and JZL184 (1 µM) were added to the culture 30 min before IL-1β or LPS application in order to elevate endogenous 2-AG. (A1–A2) Immunoblot analysis of URB suppression of LPS-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (B1–B2) Immunoblot analysis of URB suppression of IL-1β-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (C1–C2) Immunoblot analysis of JZL184 suppression of LPS-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (D1–D2) Immunoblot analysis of JZL184 suppression of IL-1β-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (E1–H2). Elevation of endogenous 2-AG by inhibiting MAGL with URB602 or JZL184 restores LPS- or IL-1β-induced down-regulation of PPARγ (n = 3 per group). **P < 0.01, compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §P < 0.05, §§P < 0.01 compared with IL-1β+ URB or JZL184, LPS+URB or JZL184.
Figure 4
Figure 4
MAGL inhibitor URB602 attenuates COX-2 elevation-induced enhancement of mEPSCs. Hippocampal neurons in culture treated with IL-1β or LPS were the same as described in Figure 2. (A1) Representative sweeps of mEPSCs recorded in vehicle control-, LPS-, LPS+URB (10 µM)- and LPS+URB+GW9662 (5 µM)-treated neurons. Scale bar: 20 pA/2 s. (A2) Mean percentage changes in the frequency of mEPSCs in neurons with different treatments. (A3). Mean percentage changes in the amplitude of mEPSCs. (B1) Representative sweeps of mEPSCs recorded in vehicle control-, IL-1β-, IL-1β+ URB (10 µM)- and IL-1β+ URB+GW9662 (5 µM)-treated neurons. (B2)Mean percentage changes in the frequency of mEPSCs. (B3) Mean percentage changes in the amplitude of mEPSCs. **P < 0.01 compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §§P < 0.01 compared with IL-1β+ URB or LPS+URB (n = 20–34).
Figure 5
Figure 5
MAGL inhibitor JZL184 attenuates COX-2 elevation-induced enhancement of mEPSCs. Hippocampal neurons in culture treated with IL-1β or LPS were the same as described in Figure 2. (A1) Representative sweeps of mEPSCs recorded in vehicle control-, LPS-, LPS+JZL184 (1 µM)- and LPS+JZL184+GW9662 (5 µM)-treated neurons. Scale bar: 20 pA/2 s. (A2) Mean percentage changes in the frequency of mEPSCs in neurons with different treatments. (A3). Mean percentage changes in the amplitude of mEPSCs. (B1) Representative sweeps of mEPSCs recorded in vehicle control-, IL-1β-, IL-1β+ JZL184 (1 µM)- and IL-1β+ JZL184+GW9662 (5 µM)-treated neurons. (B2) Mean percentage changes in the frequency of mEPSCs. (B3) Mean percentage changes in the amplitude of mEPSCs. **P < 0.01 compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §§P < 0.01 compared with IL-1β+ JZL184 or LPS+JZL184 (n = 23–30).
Figure 6
Figure 6
PPARγ agonists inhibit phosphorylation of NF-κB and expression of COX-2 induced by IL-1β-and LPS. Hippocampal neurons in culture-treated with IL-1β (10 ng·mL−1) or LPS (1 µg·mL−1) were the same as described in Figure 1. PPARγ agonists 15d-PGJ2 (2 µM) or rosiglitazone (Ros, 1 µM) were added to the culture 30 min before IL-1β or LPS. (A1–A2) Immunoblot analysis of 15d-PGJ2 suppression of LPS-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (B1-–B) Immunoblot analysis of 15d-PGJ2 inhibition of IL-1β-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (C1–C2) Immunoblot analysis of Ros suppression of LPS-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW9662 (n = 3). (D1–D2) Immunoblot analysis of Ros suppression of IL-1β-induced NF-κB-p65 phosphorylation and COX-2 expression in the absence and presence of GW (n = 3). **P < 0.01, compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §P < 0.05, §§P < 0.01 compared with IL-1β+15d-PGJ2 or Ros, LPS+15d-PGJ2 or Ros.
Figure 7
Figure 7
PPARγ agonist 15d-PGJ2 reduces COX-2 elevation-induced enhancement of mEPSCs. Hippocampal neurons in culture-treated with IL-1β or LPS were the same as described in Figure 2. (A1) Representative sweeps of mEPSCs recorded in vehicle control-, LPS-, LPS+15d-PGJ2 (2 µM)- and LPS+15d-PGJ2+GW9662 (5 µM)-treated neurons. Scale bar: 20 pA/2 s. (A2) Mean percentage changes in the frequency of mEPSCs in neurons with different treatments. (A3). Mean percentage changes in the amplitude of mEPSCs. (B1) Representative sweeps of mEPSCs recorded in vehicle control-, IL-1β-, IL-1β+15d-PGJ2 (2 µM)- and IL-1β+15d-PGJ2+GW9662 (5 µM)-treated neurons. (B2) Mean percentage changes in the frequency of mEPSCs. (B3) Mean percentage changes in the amplitude of mEPSCs. **P < 0.01 compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §§P < 0.01 compared with IL-1β+15d-PGJ2 or LPS+15d-PGJ2 (n = 23–30).
Figure 8
Figure 8
PPARγ activator rosiglitazone reduces COX-2 elevation-induced enhancement of mEPSCs. Hippocampal neurons in culture treated with IL-1β or LPS were the same as described in Figure 2. (A1) Representative sweeps of mEPSCs recorded in vehicle control-, LPS-, LPS + Ros (1 µM)- and LPS + Ros + GW9662 (5 µM)-treated neurons. Scale bar: 20 pA/2 s. (A2) Mean percentage changes in the frequency of mEPSCs in neurons with different treatments. (A3) Mean percentage changes in the amplitude of mEPSCs. (B1) Representative sweeps of mEPSCs recorded in vehicle control-, IL-1β-, IL-1β+ Ros (1 µM)- and IL-1β+ Ros+GW9662 (5 µM)-treated neurons. (B2) Mean percentage changes in the frequency of mEPSCs. (B3) Mean percentage changes in the amplitude of mEPSCs. **P < 0.01 compared with the vehicle control; ##P < 0.01 compared with IL-1β or LPS; §§P < 0.01 compared with IL-1β+ Ros or LPS + Ros (n = 23–30).
Figure 9
Figure 9
Exogenous and endogenous 2-AG prevents IL-1β-and LPS-induced decrease in PPARγ expression via a CB1 receptor-dependent mechanism. Hippocampal neurons in culture were treated with LPS (1 µg·mL−1) for 6 h. 2-AG (A1–A2), URB602 (B1–B2), JZL184 (C1–C2) or rimonabant (RIM, 1 µM) were administered 30 min before LPS. Expression of PPARγ was detected using the immunoblot analysis. **P < 0.01 compared with the vehicle control; #P < 0.05, ##P < 0.01 compared with LPS; §P < 0.05 compared with LPS+2-AG, + URB or +JZL184 (n = 3 per group). Hippocampal neurons in culture from mice deficient in the CB1 receptor were treated with LPS (1 µg·mL−1) for 6 h in the absence and presence of 2-AG, URB602 or JZL184 (D1–D2). **P < 0.01 compared with the vehicle control. There are no statistically significant differences between LPS and LPS+2-AG, LPS+URB602 or LPS+JZl184 (n = 3 per group).

Similar articles

See all similar articles

Cited by 35 articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

LinkOut - more resources

Feedback