Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 284 (43), 29817-27

Atypical Responsiveness of the Orphan Receptor GPR55 to Cannabinoid Ligands

Affiliations

Atypical Responsiveness of the Orphan Receptor GPR55 to Cannabinoid Ligands

Ankur Kapur et al. J Biol Chem.

Abstract

The cannabinoid receptor 1 (CB(1)) and CB(2) cannabinoid receptors, associated with drugs of abuse, may provide a means to treat pain, mood, and addiction disorders affecting widespread segments of society. Whether the orphan G-protein coupled receptor GPR55 is also a cannabinoid receptor remains unclear as a result of conflicting pharmacological studies. GPR55 has been reported to be activated by exogenous and endogenous cannabinoid compounds but surprisingly also by the endogenous non-cannabinoid mediator lysophosphatidylinositol (LPI). We examined the effects of a representative panel of cannabinoid ligands and LPI on GPR55 using a beta-arrestin-green fluorescent protein biosensor as a direct readout of agonist-mediated receptor activation. Our data demonstrate that AM251 and SR141716A (rimonabant), which are cannabinoid antagonists, and the lipid LPI, which is not a cannabinoid receptor ligand, are GPR55 agonists. They possess comparable efficacy in inducing beta-arrestin trafficking and, moreover, activate the G-protein-dependent signaling of protein kinase CbetaII. Conversely, the potent synthetic cannabinoid agonist CP55,940 acts as a GPR55 antagonist/partial agonist. CP55,940 blocks GPR55 internalization, the formation of beta-arrestin GPR55 complexes, and the phosphorylation of ERK1/2; CP55,940 produces only a slight amount of protein kinase CbetaII membrane recruitment but does not stimulate membrane remodeling like LPI, AM251, or rimonabant. Our studies provide a paradigm for measuring the responsiveness of GPR55 to a variety of ligand scaffolds comprising cannabinoid and novel compounds and suggest that at best GPR55 is an atypical cannabinoid responder. The activation of GPR55 by rimonabant may be responsible for some of the off-target effects that led to its removal as a potential obesity therapy.

Figures

FIGURE 1.
FIGURE 1.
Confocal microscopy of co-transfected βarr2-GFP and GPR55. A, transiently transfected βarr2-GFP (in cytoplasm, see the arrow) and GPR55 in untreated HEK293 cells. B and C, 10 μm LPI induces redistribution of βarr2-GFP in HEK 293 cells expressing GPR55 (B) and HA-GPR55E (C). D and E, corresponding GFP and Alexa Fluor 568 fluorescence channels of a field of U2OS cells stably transfected with βarr2-GFP and HA-GPR55E. D, under basal conditions the βarr2-GFP is uniformly distributed in the cell cytoplasm. Panel E shows the plasma membrane staining of HA-GPR55E in these cells using anti-HA epitope mouse primary antibody and goat anti-mouse Alexa Fluor 568 secondary antibody. F depicts the overlay of panel D and E. G, cells as in E, in which treatment with primary anti-HA antibody is omitted. No membrane staining of receptor is observed in the Alexa Fluor 568 channel.
FIGURE 2.
FIGURE 2.
Agonist induced trafficking of βarr2-GFP in U2OS cells containing GPR55E. The redistribution of βarr2-GFP fluorescence was visualized after a 40-min treatment with 3 μm LPI (A), 30 μm SR141716A (C), and 30 μm AM251 (E). B, D, and F represent the concentration-response curves for LPI, SR141716A, and AM251, respectively. G shows the cytosolic distribution of βarr2-GFP in vehicle-treated cells. 4′,6-diamidino-2-phenylindole nuclear staining (arrows) shows nuclear exclusion of βarr2. H, the maximum number of β-arrestin aggregates (objects) was determined from the calculated plateaus of the concentration-response curves for each compound to determine relative efficacies. The data represent the mean ± S.E. from at least three independent experiments where triplicate images of fields containing multiple cells were analyzed. Representative images that were captured at 40× magnification are depicted.
FIGURE 3.
FIGURE 3.
Ligand induced trafficking of βarr2-GFP in U2OS cells containing CB1RE, a receptor with a phosphorylation site-enhanced C-tail. Shown are representative live cell confocal images of a line of U2OS cells permanently transfected with CB1E receptors and β2-arrestin GFP. Cells were treated for 40 min at 37 °C with media containing vehicle (A), 10 μm LPI (B), 10 μm SR141716A (C), and 3 μm CP55,940 (D).
FIGURE 4.
FIGURE 4.
CP55,940-mediated antagonism of βarr2-GFP in U2OS cells containing GPR55E. βarr2-GFP cytosolic fluorescence was visualized after a 40-min treatment of the GPR55E cells with the various combinations of ligands to assess the ability of CP55,940 to act as an antagonist. A, incubation with 3 μm CP55,940 alone. B, 100 nm CP55,940 was applied in combination with 3 μm LPI. C, 3 μm CP55,940 was applied along with 3 μm LPI. D, 100 nm CP55,940 was applied in combination with 30 μm AM251. E, 3 μm CP55,940 was applied with 30 μm AM251. F, 100 nm CP55,940 was applied along with 30 μm SR141716A. G, 3 μm CP55,940 was applied in combination with 30 μm SR141716A Images A-G were acquired 40× magnification. Panels H–I shows the dose-dependent CP55,940-mediated inhibition of β-arrestin redistribution in the absence (dotted line) or presence of 3 μm LPI (H), 30 μm SR141716A (I), and 30 μm AM251 (J). Each curve represents the analysis of at least three fields of cells at each set of ligand concentrations from three independent experiments ±S.E.
FIGURE 5.
FIGURE 5.
Agonist-mediated GPR55E internalization in U2OS Cells. Live U2OS cells expressing HA-GPR55E prelabeled with anti-HA mouse antibody and Alexa Fluor 568 secondary antibody were treated with various compounds and the membrane staining imaged by fluorescence microscopy at 40X magnification. Receptor internalization resulted in a loss of cell surface immunofluorescence. A, vehicle-treated cells show predominantly plasma membrane receptor staining. B, D, and F, internalization of membrane-bound GPR55 occurs upon treatment with 3 μm LPI, 30 μm SR141716A, and 30 μm AM251, respectively. C, E, and G, the co-application of 3 μm CP55,940 attenuated receptor internalization and plasma membrane receptor staining in the presence of 3 μm LPI (C), 30 μm SR141716A (E), and 30 μm AM251 (G). On-Cell Western analysis is shown in H and I of the concentration-dependent LPI-induced loss of plasma membrane receptor (H) and concentration-dependent CP55,940-mediated antagonism of membrane receptor loss in the presence of 10 μm LPI(I). *, p < 0.05 was deemed statistically significant.
FIGURE 6.
FIGURE 6.
Agonist-mediated GPR55E internalization in HEK293 cells transiently transfected with HA-GPR55E. Live HEK293 cells expressing HA-GPR55E prelabeled with anti-HA mouse antibody and Alexa Fluor 568 secondary antibody were treated with various compounds, and membrane staining was imaged by confocal microscopy at 63× magnification. Receptor internalization resulted in a loss of cell surface immunofluorescence. A, vehicle-treated cells show predominantly plasma membrane receptor staining. B–F, treatment with 30 μm 2-AG, 30 μm anandamide, 30 μm THC, 30 μm O-1602, and 10 μm CP55,940, respectively, resulted in no loss of receptor surface staining and resembled vehicle treated cells. G, I, and K, internalization of membrane-bound GPR55 occurs upon treatment with 3 μm LPI, 30 μm SR141716A, and 30 μm AM251, respectively. H, J, and L, the co-application of 3 μm CP55,940 in the presence of 3 μm LPI, 30 μm SR141716A, and 30 μm AM251, respectively, attenuated receptor internalization and restored plasma membrane receptor staining.
FIGURE 7.
FIGURE 7.
LPI mediated ERK1/2 phosphorylation in U2OS cells containing GPR55E. A, GPR55E cells treated for 10 min with 3 and 10 μm LPI demonstrated ERK1/2 activation significantly different from vehicle-treated cells (*, p < 0.05; **, p < 0.001). Co-application of 10 μm CP55,940 significantly inhibited 10 μm LPI-mediated ERK1/2 activation. In contrast, 10 μm CP55,940 (CP)-, 30 μm SR141716A (SR)-, and AM251 (AM)-mediated ERK activation was not different from vehicle treatment. The data represent the mean ± S.E. from at least three independent experiments performed in duplicate. B, 10 μm LPI fails to evoke ERK1/2 activation in untransfected U2OS cells, whereas activation of ERK occurs with 1 mm pervanadate treatment. C, in U2OS cells stably expressing the CB1RE, 10 μm LPI failed to activate ERK1/2, whereas treatment with 10 μm CB1 agonist, CP 55,940, resulted in a marked increase of ERK1/2 activation.
FIGURE 8.
FIGURE 8.
Agonist induced trafficking of PKCβII in HEK293 cells transiently transfected with PKCβII-GFP and GPR55. A, live HEK293 cells expressing PKCβII-GFP or PKCβII-GFP + GPR55 were treated with the indicated compounds, and membrane staining was imaged 30–45 min after drug application by confocal microscopy using a 40× oil objective. B, time series of live cell imaging of PKCβII-GFP recruitment after the addition of 10 μm AM251. Arrows indicate that within 60 s of the addition of AM251, membrane rearrangements begin to occur followed by protrusions and blebbing of the cell membranes.

Similar articles

See all similar articles

Cited by 82 PubMed Central articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Feedback