Compartmentalization of endocannabinoids into lipid rafts in a dorsal root ganglion cell line

Abstract

Background and purpose: N-arachidonoyl ethanolamine (AEA) and 2-arachidonoyl glycerol (2-AG) are endogenous cannabinoids binding to the cannabinoid receptors CB1 and CB2 to modulate neuronal excitability and synaptic transmission in primary afferent neurons. To investigate the compartmentalization of the machinery for AEA and 2-AG signalling, we studied their partitioning into lipid raft fractions isolated from a dorsal root ganglion X neuroblastoma cell line (F-11).

Experimental approach: F-11 cells were homogenized and fractionated using a detergent-free OptiPrep density gradient. All lipids were partially purified from methanolic extracts of the fractions on solid phase cartridges and quantified using liquid chromatography tandem mass spectrometry (LC/MS/MS). Protein distribution was determined by Western blotting.

Key results: Under basal conditions, the endogenous cannabinoid AEA was present in both lipid raft and specific non-lipid raft fractions as was one of its biosynthetic enzymes, NAPE-PLD. The 2-AG precursor 1-stearoyl-2-arachidonoyl-sn-glycerol (DAG), diacylglycerol lipase alpha (DAGLalpha), which cleaves DAG to form 2-AG, and 2-AG were all co-localized with lipid raft markers. CB1 receptors, previously reported to partition into lipid raft fractions, were not detected in F-11 membranes, but CB2 receptors were detected at high levels and partitioned into non-lipid raft fractions.

Conclusions and implications: The biochemical machinery for the production of 2-AG via the putative diacylglycerol pathway is localized within lipid rafts, suggesting that 2-AG synthesis via DAG occurs within these microdomains. The observed co-localization of AEA, 2-AG, and their synthetic enzymes with the reported localization of CB1 raises the possibility of intrinsic-autocrine signalling within lipid raft domains and/or retrograde-paracrine signalling.

Figure 1

Distribution of protein lipid raft markers across F-11 fractions. The lipid raft markers caveolin-1 and flotillin-1 were concentrated in lipid raft fractions 10–11. Neuronal NO synthase was spread across lipid raft and non-lipid raft fractions. N-acyl phosphatidylethanolamine phospholipase D was highest in fraction 6 and elevated in fractions 11–12. Cannabinoid CB₂ receptors were localized to non-lipid raft fractions 1–5. The enzyme diacylglycerol lipase α was localized to lipid raft fractions 10–12. Blots are representative of three blots per protein.

Figure 2

(a) Distribution of lipid molecule lipid raft markers. Lipids are presented as a percent of the highest fraction in each fractionation experiment. Cholesterol was significantly higher in fractions 10–13, P<0.05, n=7, one-way ANOVA with Fisher's LSD post hoc. The glycosphingolipid ganglioside M₃ (GM₃) was significantly higher in fractions 10–12 compared with all other fractions P<0.001, (n=7). The levels in these fractions were also significantly different from each other, P<0.005, one-way ANOVA with Fisher's LSD post hoc. Sphingomyelin (18:0) was highest in fractions 11–12, P<0.05, n=4, one-way ANOVA with Fisher's LSD post hoc. (b) Distribution of total proteins in fractions. Total protein levels were highest in lipid raft fraction 11 and non-lipid raft fraction 4. Protein levels were low in fractions 7–9 and were below our detection limit in fractions 13–16. Error bars represent s.e.mean.

Figure 3

(a) Distribution of free arachidonic acid and prostaglandins. Fraction 6 contained the highest levels of arachidonic acid (n=6; ⁺P<0.005, one-way ANOVA, Fisher's LSD post hoc). Its prostaglandin metabolites, PGE₂ and PGF_2α were highest in fractions 1–6 where they were significantly elevated compared to lipid raft fractions 10–13 (n=6–7; PGE₂, ^#P<0.001; PGF_2α ^*P<0.01: one-way ANOVA, Fisher's LSD post hoc). Data are presented as the total quantity in pmol recovered from 600 μl of each fraction. Error bars represent s.e. mean. (b) Distribution of acyl ethanolamines. AEA was significantly higher in fraction 6 compared with all fractions except for lipid raft fraction 10, n=12, ^#P<0.05, one-way ANOVA, Fisher's LSD post hoc. Palmitoyl ethanolamine (PEA) showed a trend similar to AEA. However, no significant differences were observed in the distribution of PEA between fractions. Data are presented as the quantity in fmol recovered from 600 μl of each fraction. (c) Distribution of exogenously administered [²H₈]-AEA and its metabolite [²H₈]-arachidonic acid. [²H₈]-AEA was significantly higher in fractions 10 (P<0.01, n=3) and 5 (^*P<0.05, n=3) compared with all other fractions, one-way ANOVA, Fisher's LSD post hoc. Levels of its metabolite [²H₈]-arachidonic acid were highest in fraction 5 (^#P<0.05, excluding fraction 6, n=3). Data are presented as the total quantity in pmol recovered from 600 μl of each fraction. Error bars represent s.e. mean.

Figure 4

(a) Distribution of 2-AG and its putative DAG precursor. 2-AG (left y axis, squares) was significantly higher in lipid raft fractions 10–11 (^#P<0.001; n=8, one-way ANOVA, Fisher's LSD post hoc). Arachidonoyl-containing DAG (1-stearoyl-2-arachidonoyl sn-glycerol) was most abundant in fractions 11–12 (^*P<0.0001, n=7, one-way ANOVA, Fisher's LSD post hoc). Data are presented as the total quantity in pmol recovered from 600 μl of each fraction. (b) The distribution of a DAG lacking arachidonoyl chains. There were no significant differences between levels of DAG (1-palmitoyl-2-oleoyl sn-glycerol) in the cell fractions. Error bars represent s.e. mean.

Figure 5

(a) Distribution of arachidonoyl-containing lyso-phosphatidyl inositol (lyso-PI) and PI. The arachidonoyl-containing lyso-PI was highest in fractions 2–6 where the levels were significantly different from fractions 7–16 (^#P<0.005, n=6, one-way ANOVA, Fisher's LSD post hoc). The highest levels of the arachidonoyl-containing phosphatidyl inositol (18:0; 20:4) were found in fractions 3 and 5–6, which significantly differed from fractions 7–16 in PI levels (^*P<0.05, n=6, one-way ANOVA, Fisher's LSD post hoc). Data for PI are presented as the total quantity in pmol recovered from 600 μl of each fraction. Due to the absence of a synthetic standard for lyso-PI (20:4), data are presented as integrated area under the curve in counts. (b) Distribution of arachidonoyl-containing phosphatidic acids. The levels of arachidonoyl-containing phosphatidic acid 1-palmitoyl-2-arachidonoyl sn-glycerol-3-phosphate were highest in fractions 1–3 and 5–6 and were significantly different from fractions 7–16 (^#P<0.05, n=3–5, one-way ANOVA, Fisher's LSD post hoc). The highest levels of the arachidonoyl-containing phosphatidic acid 1-stearoyl-2-arachidonoyl-sn-glycerol-3-phosphate were found in fractions 1–2 and 5–6 and were significantly different from fractions 7–16 (^*P<0.05, n=3–5, one-way ANOVA, Fisher's LSD post hoc).