Identification of intracellular carriers for the endocannabinoid anandamide

Abstract

The endocannabinoid anandamide (arachidonoyl ethanolamide, AEA) is an uncharged neuromodulatory lipid that, similar to many neurotransmitters, is inactivated through its cellular uptake and subsequent catabolism. AEA is hydrolyzed by fatty acid amide hydrolase (FAAH), an enzyme localized on the endoplasmic reticulum. In contrast to most neuromodulators, the hydrophilic cytosol poses a diffusional barrier for the efficient delivery of AEA to its site of catabolism. Therefore, AEA likely traverses the cytosol with the assistance of an intracellular carrier that increases its solubility and rate of diffusion. To study this process, AEA uptake and hydrolysis were examined in COS-7 cells expressing FAAH restricted to the endoplasmic reticulum, mitochondria, or the Golgi apparatus. AEA hydrolysis was detectable at the earliest measurable time point (3 seconds), suggesting that COS-7 cells, normally devoid of an endocannabinoid system, possess an efficient cytosolic trafficking mechanism for AEA. Three fatty acid binding proteins (FABPs) known to be expressed in brain were examined as possible intracellular AEA carriers. AEA uptake and hydrolysis were significantly potentiated in N18TG2 neuroblastoma cells after overexpression of FABP5 or FABP7, but not FABP3. Similar results were observed in COS-7 cells stably expressing FAAH. Consistent with the roles of FABP as AEA carriers, administration of the competitive FABP ligand oleic acid or the selective non-lipid FABP inhibitor BMS309403 attenuated AEA uptake and hydrolysis by approximately 50% in N18TG2 and COS-7 cells. Taken together, FABPs represent the first proteins known to transport AEA from the plasma membrane to FAAH for inactivation and may therefore be novel pharmacological targets.

Fig. 1.

Generation of FAAH variants with distinct subcellular localizations. (A) Constructs used in this study. FAAH-eGFP is shown with its N-terminal transmembrane helix (residues 1–29) in red, residues 30–579 in black, and eGFP in green. In TOM-ΔTMFAAH-eGFP, the N-terminal helix was replaced with the N terminus of mouse TOM20 (blue, residues 1–33). In GRASP-ΔTMFAAH-eGFP, the Golgi resident protein grasp65 (gray) was fused to the N terminus of ΔTMFAAH-eGFP. (B) Western blot indicates similar protein levels in homogenates of COS-7 cells stably expressing FAAH-eGFP, TOM-ΔTMFAAH-eGFP, or GRASP-ΔTMFAAH-eGFP. Blots were probed with eGFP antibodies with β-actin serving as a loading control. (C) Similar rates of [¹⁴C]AEA hydrolysis (100 μM) were observed in COS7-FAAH-eGFP, TOM-ΔTMFAAH-eGFP, and GRASP-ΔTMFAAH-eGFP homogenates (n = 3). (D) FAAH fusion protein localization is restricted to specific organelles. Top panel shows FAAH-eGFP fusion proteins (green); middle panel depicts the marker of interest (red); and bottom panel is the merged image (yellow). FAAH-eGFP localized to the ER and mostly overlapped with the ER marker, calreticulin. TOM-ΔTMFAAH-eGFP co-localized with the mitochondrial dye Mitotracker CM-H2Xros and was excluded from the ER. As confirmed by double labeling with GM130, GRASP-ΔTMFAAH-eGFP localized to the Golgi apparatus in COS-7 cells. Note that overexpression of GRASP-ΔTMFAAH-eGFP resulted in an enlargement of the Golgi apparatus. (E) Proteinase K protection analysis of FAAH-eGFP, TOM-ΔTMFAAH-eGFP and GRASP-ΔTMFAAH-eGFP proteins in membrane fractions of COS-7 cells. Membranes were either left untreated or were incubated with 500 μg/ml proteinase K in the presence or absence of 1% Triton X-100. Samples were subsequently resolved by SDS/PAGE and probed with eGFP and calreticulin antibodies.

Fig. 2.

Rapid inactivation of AEA in cells expressing spatially restricted FAAH variants. [¹⁴C]AEA (100 nM) uptake (black bars) and hydrolysis (gray bars) were similar (P > 0.05) in COS7-FAAH-eGFP, TOM-ΔTMFAAH-eGFP, or GRASP-ΔTMFAAH-eGFP cells after (A) 5-minute or (B) 10-second incubations (n = 3). (C) Similar levels of intracellular [¹⁴C]AEA hydrolysis after 3-second incubation in COS7-FAAH-eGFP, TOM-ΔTMFAAH-eGFP, or GRASP-ΔTMFAAH-eGFP cells (P > 0.05). Hydrolysis of [¹⁴C]AEA in untransfected COS-7 cells was significantly lower than in FAAH-eGFP expressing controls. **, P < 0.01 (n = 3).

Fig. 3.

Effect of FABP overexpression upon AEA uptake and hydrolysis. (A) Following a 10-minute incubation, [¹⁴C]AEA uptake (black bars) and hydrolysis (gray bars) were elevated in COS-7-FAAH-eGFP cells after FABP5 or FABP7, but not FABP3, transfection. **, P < 0.01 compared with vector-transfected controls (n = 3–5). (B) Western blot confirms the overexpression of FABP3, FABP5, or FABP7 in COS-7-FAAH-eGFP- and N18TG2-transfected cells. (C) Overexpression of FABP3, FABP5, or FABP7 had no effect upon [¹⁴C]AEA hydrolysis by COS-7-FAAH-eGFP homogenates (P > 0.05). (D) [¹⁴C]AEA uptake and hydrolysis by N18TG2 cells are enhanced after transfection with FABP5 and FABP7, but not FABP3. *, P < 0.05 and **, P < 0.01 compared with vector transfected controls (n = 3–5). (E) RT-PCR analysis confirmed the endogenous expression of FABP3, FABP5, and FABP7 in brain and the expression of FABP3 and FABP5 in N18TG2 cells. β-Actin serves as a control.

Fig. 4.

Effect of FABP inhibition upon AEA internalization and hydrolysis by FAAH. (A) Treatment of COS-7-FAAH-eGFP and N18TG2 cells for 5 minutes with 100 μM oleic acid, a ligand for FABPs, significantly reduced [¹⁴C]AEA uptake (black bars) and hydrolysis (gray bars). *, P < 0.05 compared with vehicle-treated controls (n = 3–5). (B, C) Treatment with 20–100 μM BMS309403, a selective competitive FABP inhibitor, significantly decreased [¹⁴C]AEA uptake and metabolism by COS-7-FAAH-eGFP (B) or N18TG2 (c) cells at 5 min. *, P < 0.05 and **, P < 0.01 compared with vehicle controls (n = 3–5). (D, E) Increasing concentrations (20–100 μM) of BMS309403 reduced the hydrolysis of [¹⁴C]AEA after cellular uptake by COS7-FAAH-eGFP (D) or N18TG2 (E) cells at 3 sec. *, P < 0.05 and **, P < 0.01 compared with vehicle-treated controls (n = 3–5).