Monoacylglycerol Lipase Regulates Fever Response

Abstract

Cyclooxygenase inhibitors such as ibuprofen have been used for decades to control fever through reducing the levels of the pyrogenic lipid transmitter prostaglandin E2 (PGE2). Historically, phospholipases have been considered to be the primary generator of the arachidonic acid (AA) precursor pool for generating PGE2 and other eicosanoids. However, recent studies have demonstrated that monoacyglycerol lipase (MAGL), through hydrolysis of the endocannabinoid 2-arachidonoylglycerol, provides a major source of AA for PGE2 synthesis in the mammalian brain under basal and neuroinflammatory states. We show here that either genetic or pharmacological ablation of MAGL leads to significantly reduced fever responses in both centrally or peripherally-administered lipopolysaccharide or interleukin-1β-induced fever models in mice. We also show that a cannabinoid CB1 receptor antagonist does not attenuate these anti-pyrogenic effects of MAGL inhibitors. Thus, much like traditional nonsteroidal anti-inflammatory drugs, MAGL inhibitors can control fever, but appear to do so through restricted control over prostaglandin production in the nervous system.

Fig 1. Mgll ^-/- and Mgll ^+/+ mice have similar core body temperature profiles.

CBT profile of Mgll ^-/- and Mgll ^+/+ male mice over 24 hrs. No statistically significant differences were observed across genotypes. Data are shown as mean ± sem, n = 6 mice per group, p>0.05.

Fig 2. Genetic or pharmacological ablation of MAGL reduces fever response in peripheral LPS-induced fever model.

(A) CBT profile following i.p. injection of LPS (100 μg/kg) or vehicle (Saline) of Mgll ^+/+ mice treated i.p. with JZL184 (40 mg/kg). *p<0.05, Mgll ^+/++ Veh vs. Mgll ^+/+ + LPS; ^#p<0.05, Mgll ^+/++ LPS vs. Mgll ^-/- + LPS; ^&p<0.05, Mgll ^+/+ + Veh vs. Mgll ^-/- + LPS. (B) CBT profile of Mgll ^-/- and Mgll ^+/+ mice following i.p. injection of LPS (100 μg/kg) or vehicle (Saline) as indicated. Injection was performed at time 0. Data are shown as mean ± sem, n = 6 mice per group, *p<0.05, Mgll ^+/++ Veh + Veh vs. Mgll ^+/+ + Veh + LPS; # p<0.05, Mgll ^+/++ Veh+ LPS vs. Mgll ^+/++ JZL184 + LPS.

Fig 3. Genetic or pharmacological ablation of MAGL reduces fever response in central IL-1β-induced fever model.

(A) CBT profile following icv injection into POA of IL-1β (500 pg/0.5 μl) or aCSF of Mgll ^+/+ mice treated i.p. with JZL184 (40 mg/kg). *p<0.05, Mgll ^+/+ + Veh vs. Mgll ^+/+ + IL-1β; ^#p<0.05 Mgll ^+/ + IL-1β vs. Mgll ^-/-+ IL-1β. (B) CBT profile of Mgll ^-/- and Mgll ^+/+ mice following icv injection into POA of IL-1β (500 pg/0.5 μl) or aCSF as indicated. Injection was performed at time 0. Data are shown as mean ± sem, n = 6 mice per group. *p<0.05, Mgll ^+/++ Veh + Veh vs. Mgll ^+/++ Veh + IL-1β; ^#p<0.05 Mgll ^+/++ Veh+ IL-1β vs. Mgll ^+/++ JZL184 + IL-1β.

Fig 4. Anti-pyrogenic effects of MAGL inhibitors are independent of CB1 cannabinoid receptor activity.

CBT profile following i.p. injection of LPS (100 μg/kg) of Mgll ^+/+ mice receiving i.p. injection of rimonabant (1 mg/kg) and/ or JZL184 (40 mg/kg). Rimonabant did not affect the hypothermic effects of JZL184. Data are shown as mean ± sem, n = 6 mice per group, *p<0.05, Mgll ^+/++ Veh + Veh + Veh vs. Mgll ^+/++ Veh + Veh + LPS; ^#p<0.05, Mgll ^+/++ Veh + Veh + LPS vs. Mgll ^+/++ JZL184 + Veh + LPS; ^&p<0.05, Mgll ^+/++ Veh + Rim + LPS vs. Mgll ^+/++ Veh + Rim + Veh; ^$p<0.05, Mgll ^+/++ Veh + Rim + LPS vs. Mgll ^+/++ JZL184 + Rim + LPS. p>0.05 (NS) Mgll ^+/++ Veh + Veh + Veh vs Mgll ^+/++ Veh + Rim + Veh, Mgll ^+/++ JZL184 + Rim + LPS vs Mgll ^+/ + JZL184 + Veh + LPS, Mgll ^+/++ Veh + Veh + LPS vs Mgll ^+/++ Veh + Rim + LPS.