Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 83 (1), 267-82

N-Arachidonyl Glycine Does Not Activate G Protein-Coupled Receptor 18 Signaling via Canonical Pathways

Affiliations

N-Arachidonyl Glycine Does Not Activate G Protein-Coupled Receptor 18 Signaling via Canonical Pathways

Van B Lu et al. Mol Pharmacol.

Abstract

Recent studies propose that N-arachidonyl glycine (NAGly), a carboxylic analogue of anandamide, is an endogenous ligand of the Gα(i/o) protein-coupled receptor 18 (GPR18). However, a high-throughput β-arrestin-based screen failed to detect activation of GPR18 by NAGly (Yin et al., 2009; JBC, 18:12328). To address this inconsistency, this study investigated GPR18 coupling in a native neuronal system with endogenous signaling pathways and effectors. GPR18 was heterologously expressed in rat sympathetic neurons, and the modulation of N-type (Ca(v)2.2) calcium channels was examined. Proper expression and trafficking of receptor were confirmed by the "rim-like" fluorescence of fluorescently tagged receptor and the positive staining of external hemagglutinin-tagged GPR18-expressing cells. Application of NAGly on GPR18-expressing neurons did not inhibit calcium currents but instead potentiated currents in a voltage-dependent manner, similar to what has previously been reported (Guo et al., 2008; J Neurophysiol, 100:1147). Other proposed agonists of GPR18, including anandamide and abnormal cannabidiol, also failed to induce inhibition of calcium currents. Mutants of GPR18, designed to constitutively activate receptors, did not tonically inhibit calcium currents, indicating a lack of GPR18 activation or coupling to endogenous G proteins. Other downstream effectors of Gα(i/o)-coupled receptors, G protein-coupled inwardly rectifying potassium channels and adenylate cyclase, were not modulated by GPR18 signaling. Furthermore, GPR18 did not couple to other G proteins tested: Gα(s), Gα(z), and Gα(15). These results suggest NAGly is not an agonist for GPR18 or that GPR18 signaling involves noncanonical pathways not examined in these studies.

Figures

Fig. 1.
Fig. 1.
Expression and localization of heterologously expressed GPR18-EGFP. Confocal images of (A) GPR18-EGFP, (B) EGFP, and (C) EGFP-KRas tail constructs expressed in SCG neurons (top panels) or HeLa cells (bottom panels). Insets of top panels: line plots of fluorescence intensity from dashed line across injected SCG neuron. Note the “rim-like” fluorescence of the GPR18-EGFP construct, similar to the membrane-bound EGFP-KRas tail-expressing cells and different from cytosolic EGFP-expressing cells. Scale bar is 20 µm. (D) Sample Western blot of EGFP-tagged constructs. Primary anti-GFP antibody (1:2000, NeuroMab) and secondary anti-mouse HRP-conjugated antibody (1:1000, Thermo) were used. Band in GPR18-EGFP lane at approximately 80 kDa, band in EGFP lane at 27 kDa, and band in EGFP-KRas tail lane at 29 kDa. (E) Stripped and reprobed blot for loading controls α-tubulin (1:2000, Cell Signaling) and cyclophilin B (1:5000, Abcam) corresponding to bands at 51 and 19 kDa, respectively. MW, molecular weight.
Fig. 2.
Fig. 2.
Live-cell staining and imaging of hemagglutinin-tagged GPR18. Confocal images of stained HeLa cells transfected with (A) 3xHA-GPR18 or external HA-tagged GPR18, (B and C) GPR18-3xHA or internal HA-tagged GPR18. EGFP was cotransfected to label transfected cells. (A and B) Live-cell staining with biotin-labeled anti-HA antibody (1:200; Covance) and streptavidin conjugated Qdot655 secondary antibody (1:200; Molecular Probes). (C) HA staining followed fixation and permeabilization of HeLa cells. First panels are images obtained from the GFP channel (500- to 550-nm emission band-pass filter), second panels are fluorescent images from the far-red channel (650- to 710-nm emission band-pass filter), and the last panels are merged images pseudo-colored green and purple. Scale bar is 20 µm.
Fig. 3.
Fig. 3.
NAGly-mediated potentiation of N-type (Cav2.2) Ca2+ channel currents (ICa) in rat SCG neurons. Data are from whole cell patch-clamp recordings of rat sympathetic neurons obtained at room temperature (20–24°C). (Ai) Sample superimposed ICa traces evoked from an uninjected SCG neuron using the double-pulse ICa protocol, shown as inset. Two 25-millisecond test pulses to 10 mV from a holding potential of −80 mV, separated by a 50-millisecond conditioning pulse to 80 mV. For both sample traces displayed in the figure, solid black trace is the baseline ICa, solid gray trace is the ICa during application of 10 μM NAGly, y-axis scale bar is 0.5 nA, and x-axis scale bar is 10 milliseconds. (Aii) Time course of ICa amplitude in an uninjected neuron during exposure to 1 and 10 μM NAGly (solid gray line) and 10 μM norepinephrine (NE, dashed black line). ○ represents the prepulse ICa, ● represents the postpulse ICa. (Aiii) Time course of the FR of the same sample cell in Aii during exposure to 1 and 10 μM NAGly and 10 μM NE. (B) Ca2+ current-voltage relationship of uninjected neuron before (○) and during 10 μM NAGly (formula image) application. Note the voltage-dependent enhancement of Ca2+ currents and the hyperpolarizing shift in the IV curve peak. (Ci) Sample superimposed ICa traces evoked from an untagged GPR18-injected SCG neuron using the double-pulse protocol. (Cii) Time course of ICa amplitude in an untagged GPR18-injected neuron during exposure to 1 and 10 μM NAGly (solid gray line) and 10 μM norepinephrine (NE, dashed black line). (Ciii) Time course of the FR of the same sample cell in Cii during exposure to 1 and 10 μM NAGly and 10 μM NE. (D) Changes in ICa amplitude produced by NAGly (1 or 10 μM) and NE (10 μM) from each cell are represented as individual points in the dot plot graph. Drug responses were normalized to baseline ICa using the equation Idrug/Ibaseline × 100, where Idrug and Ibaseline are ICa amplitudes during and before drug application, respectively. Mean ± S.E.M. drug responses are represented as lines on graph. The n values for each group are indicated on graph in parentheses. Means values between uninjected and untagged GPR18-injected neurons were not significantly different (unpaired t test, P > 0.05).
Fig. 4.
Fig. 4.
NAGly-mediated inhibition of LVA-ICa heterologously expressed in rat SCG neurons. (A) Sample ICa trace from an uninjected SCG neuron, elicited by the low-voltage ICa protocol illustrated as inset. Two 25-millisecond test pulses, the first to −40 mV and the second to 10 mV, separated by a 60-millisecond pulse to −60 mV to inactivate LVA-ICa. (B) Sample superimposed ICa traces from a Cav3.1-injected SCG neuron, elicited by the low-voltage ICa protocol. For all sample traces displayed in the figure, solid black trace is the baseline ICa, solid gray trace is the ICa during application of 10 μM NAGly, y-axis scale bar is 0.5 nA, and x-axis scale bar is 10 milliseconds. Note the inhibition of LVA-ICa elicited during the first test pulse and the potentiation of HVA-ICa elicited during the second test pulse by application of NAGly. (C) Sample superimposed ICa traces from an SCG neuron coinjected with Cav3.1 and GPR18, elicited by the low-voltage ICa protocol. In this set of experiments, an untagged version of GPR18 was used. (D) Changes in ICa amplitude, both LVA- and HVA-ICa, produced by NAGly (10 μM) from each cell are represented as individual points in the dot plot graph. NAGly responses were normalized to baseline ICa. Mean ± S.E.M. NAGly responses are represented as lines on graph. The n values for each group are indicated on graph in parentheses. Mean values between Cav3.1 alone and Cav3.1 with GPR18-injected groups were not significantly different (unpaired t test, P > 0.05).
Fig. 5.
Fig. 5.
Effectiveness of various proposed agonists of GPR18 to inhibit ICa in GPR18- or CB1R-expressing SCG neurons. In this set of experiments, an untagged version of GPR18 was used. N-arachidonoyl-l-serine, AEA, and NE were applied to uninjected, GPR18- and CB1R-injected neurons. Because of the lack of a confirmed endogenous receptor, Abn-Cbd and O-1602 were applied to uninjected and GPR18-injected neurons only. Changes in ICa amplitude in response to various agonists from each cell are represented as individual points in the dot plot graph, normalized to baseline ICa. Mean ± S.E.M. drug responses are represented as lines on graph. The n values for each group are indicated on graph in parentheses. For N-arachidonoyl-l-serine, AEA, and NE treatment groups, a one-way ANOVA followed by Newman-Keuls post-test was used to compare groups. For Abn-Cbd and O-1602, an unpaired t test was used. *P < 0.05, ***P < 0.001.
Fig. 6.
Fig. 6.
Mutant class A GPCRs designed to induce or alter tonic activity of receptors. (A) Left: Confocal image of GPR18 D118T-EGFP construct expressed in HeLa cells. Note the fluorescence primarily inside of the cell in the internal membrane network, reminiscent of the endoplasmic reticulum. Scale bar is 20 µm. Right: Sample superimposed ICa traces from an untagged GPR18 D118T-injected SCG neuron, elicited by the double-pulse ICa protocol. For all sample traces displayed in the figure, solid black trace is the baseline ICa, dashed black trace is the ICa during application of 10 μM NE, solid gray trace is the ICa during application of 10 μM NAGly, y-axis scale bar is 0.5 nA, and x-axis scale bar is 10 milliseconds. Horizontal dashed gray line from the peak of the postpulse ICa highlights the relief of tonic Ca2+ channel inhibition by the conditioning pulse. (B) Sample superimposed ICa traces from an ADRA2A D130T-injected SCG neuron, elicited by the double-pulse ICa protocol. Note the kinetic slowing and inhibition of prepulse ICa and the large relief of tonic Ca2+ channel inhibition by the conditioning pulse in the baseline ICa trace. (C) Responses of a proposed nonconstitutively active mutant of GPR18, GPR18 A108N (Qin et al., 2011), to NAGly and other agonists of GPR18. Left: Confocal image of GPR18 A108N-EGFP construct expressed in HeLa cells. Note the “rim-like” fluorescence in transfected cells. Scale bar is 20 µm. Right: Sample superimposed ICa traces from an untagged GPR18 A108N-injected SCG neuron, elicited by the double-pulse ICa protocol. Below: Time course of ICa amplitude in an untagged GPR18 A108N-expressing neuron during exposure to 10 μM NAGly, 10 μM AEA, and 10 μM Abn-Cbd. ○ represents the prepulse ICa, ● represents the postpulse ICa. Note the potentiation of pre- and postpulse ICa after application of NAGly and lack of ICa response to AEA and Abn-Cbd application. (D) Changes in ICa amplitude, normalized to baseline ICa, from each cell, are represented as individual points in the dot plot graph. Mean ± S.E.M. drug responses are represented as lines on graph. The n values for each group are indicated on graph in parentheses. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was used. *P < 0.05, ***P < 0.001.
Fig. 7.
Fig. 7.
Functional coupling of endogenous α2-adrenergic receptors to Gαz but not untagged GPR18. (Ai) Sample superimposed ICa traces evoked from a Gαz-injected SCG neuron using the double-pulse ICa protocol. For both sample traces displayed in Figure, solid black trace is the baseline ICa, dashed black trace is the ICa during application of 10 μM NE, solid gray trace is the ICa during application of 10 μM NAGly, y-axis scale bar is 0.5 nA, and x-axis scale bar is 10 milliseconds. (Aii) Time course of ICa amplitude in a Gαz-injected neuron during exposure to 10 μM NAGly (solid gray line) and 10 μM NE (dashed black line). ○ represents the prepulse ICa, ● represents the postpulse ICa. Note the slower onset and off-rate of NE compared with uninjected SCG neurons (Fig. 3Aii). (Aiii) Time course of the facilitation ratio (FR) of the same sample cell in Aii during exposure to 10 μM NAGly and 10 μM NE. (Bi) Sample superimposed ICa traces evoked from an SCG neuron coexpressing Gαz and untagged GPR18, using the double-pulse ICa protocol. (Bii) Time course of ICa amplitude in a Gαz + untagged GPR18-injected neuron during exposure to 10 μM NAGly (solid gray line) and 10 μM NE (dashed black line). (Biii) Time course of FR of the same sample cell as in Bii. (C) Changes in ICa amplitude produced by NAGly (10 μM) and NE (10 μM) from each cell are represented as individual points in the dot plot graph. Mean ± S.E.M. drug responses are represented as lines on graph. NE-mediated responses were elicited from SCG neurons after overnight PTX treatment. The n values for each group are indicated on graph in parentheses. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was performed. ***P < 0.001. (D) Testing possible coupling of constitutively active GPR18 with Gαz in SCG neurons. (Di) The basal FR, a sensitive indicator of tonic receptor activity, was determined from the ratio of postpulse to prepulse ICa of the first recording obtained from each cell. Basal FR values from each cell are represented as individual points in the dot plot graph. Mean ± S.E.M. basal FR are represented as lines on graph. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was performed. ***P < 0.001. (Dii) Changes in ICa amplitude in response to NE, normalized to baseline ICa, from each cell, are represented as individual points in the dot plot graph. Mean ± S.E.M. NE response are represented as lines on graph. No significant difference was observed between groups, as determined by one-way ANOVA followed by Newman-Keuls post-test (P > 0.05).
Fig. 8.
Fig. 8.
Functional coupling of heterologously expressed mGluR2 receptors to Gα15 but not untagged GPR18. (A) Demonstration of functional coupling of mGluR2 to Gα15. For controls of functional coupling to Gα15, see Supplemental Fig. 3. (Ai) Sample superimposed ICa traces evoked from an SCG neuron coexpressing Gα15β1γ2 and mGluR2 using the double-pulse ICa protocol, following overnight PTX treatment. For both sample traces displayed in the figure, solid black trace is the baseline ICa, dashed black trace is the ICa during application of 10 μM NE, dashed gray trace is the ICa during application of 100 μM glutamate, y-axis scale bar is 0.5 nA, and x-axis scale bar is 10 milliseconds. (Aii) Time course of ICa amplitude in a Gα15β1γ2 + mGluR2-injected neuron during exposure to 10 μM NE (dashed black line) and 100 μM glutamate (dashed gray line). ○ represents the prepulse ICa, ● represents the postpulse ICa. (Aiii) Time course of the FR of the same sample cell in Aii during exposure to 10 μM NE and 100 μM glutamate. (Bi) Sample superimposed ICa traces evoked from an SCG neuron coexpressing Gα15β1γ2 and untagged GPR18, using the double-pulse ICa protocol, following overnight PTX treatment. (Bii) Time course of ICa amplitude in an SCG neuron coexpressing Gα15β1γ2 and untagged GPR18 during exposure to 10 μM NAGly (solid gray line) and 10 μM NE (dashed black line). (Biii) Time course of the FR of the same sample cell in Bii. (C) Changes in ICa amplitude produced by NAGly (10 μM) and NE (10 μM) from each cell are represented as individual points in the dot plot graph. Mean ± S.E.M. drug response are represented as lines on graph. Drug responses were tested in SCG neurons after overnight PTX treatment. The n values for each group are indicated on graph in parentheses. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was performed, but no significant difference was observed (P > 0.05). (D) To test possible coupling of constitutively active GPR18 with Gα15 in SCG neurons, basal ICa density was measured. Basal ICa density was determined from the postpulse ICa divided by the capacitance of the cell, calculated from integrating the area under the current trace obtained from a 10-mV step applied before cell capacitance compensation. ICa density values from each cell are represented as individual points in the dot plot graph. Mean ± S.E.M. ICa density is represented as lines on graph. ICa density was measured in SCG neurons after overnight PTX treatment. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was performed. *P < 0.05.
Fig. 9.
Fig. 9.
Testing modulation of IGIRK channels by untagged GPR18. (Ai) Sample superimposed IGIRK traces evoked from a GIRK4 S143T-injected SCG neuron using the voltage ramp, shown as inset. A 200-millisecond voltage ramp from −140 to −40 mV, from a holding potential of −60 mV, was used to elicit IGIRK. For both sample traces displayed in the figure, solid black trace is the baseline IGIRK, dashed black trace is the IGIRK during application of 10 μM NE, solid gray trace is the IGIRK during application of 10 μM NAGly, and y-axis scale bar is 1 nA. (Aii) Time course of IGIRK amplitude in a GIRK4 S143T-injected neuron during exposure to 10 μM NAGly (solid gray line) and 10 μM NE (dashed black line). (Bi) Sample superimposed IGIRK traces evoked from an SCG neuron coinjected with GIRK4 S143T and untagged GPR18 using the IGIRK voltage ramp protocol. (Bii) Time course of IGIRK amplitude in a GIRK4 S143T + GPR18-injected neuron during exposure to 10 μM NAGly (solid gray line) and 10 μM NE (dashed black line). (C) Changes in IGIRK amplitude produced by NAGly (10 μM) and NE (10 μM) from each cell are represented as individual points in the dot plot graph. Drug responses were normalized to baseline IGIRK. Mean ± S.E.M. drug responses represented as lines on graph. The n values for each group are indicated on graph in parentheses. To compare groups, a one-way ANOVA followed by Newman-Keuls post-test was performed. *P < 0.05, ***P < 0.001.
Fig. 10.
Fig. 10.
Monitoring modulation of adenylate cyclase using the BRET-based cAMP sensor CAMYEL. Data are obtained from live HEK cells loaded in a microplate luminometer. HEK cells were transfected with empty vector or selected G protein–coupled receptor cDNA, CAMYEL cDNA, and PEI. Approximately 16 hours after transfection, cells were transferred to a black 96-well microplate and loaded into a luminometer for recording. Net BRET was calculated as A/Dd, where A is the light intensity measured from the acceptor channel (542/27 nm) for 1 second, D is the light intensity measured from the donor channel (460/60 nm) for 1 second, and d is the background or spectral overlap, calculated previously as the A/D value for Rluc8 alone. Net BRET values are inversely related to cAMP levels (high net BRET = low intracellular cAMP levels, low net BRET = high intracellular cAMP levels). (A and B) Test of Gαi/o-mediated inhibition of forskolin stimulated cAMP production. (Ai) Time course of net BRET readings from empty vector and untagged GPR18-expressing HEK cells. F, forskolin (1 μM); NG, NAGly (10 μM). (Aii) Net BRET values 4 minutes after injection of NAGly from each sample well are represented as individual points in the dot plot graph. Mean ± S.E.M. net BRET values are represented as lines on graph. The n values for each group are indicated on graph in parentheses. No significant difference was observed between empty vector and GPR18-transfected cells (unpaired t test, P > 0.05). (Bi) Time course of net BRET readings from empty vector and mGluR2-expressing HEK cells. F, forskolin (1 μM); G, glutamate (100 μM). PTX was applied to HEK cells during overnight incubation. (Bii) Net BRET values 4 minutes after injection of glutamate from each sample well are represented as individual points in the dot plot graph. Mean ± S.E.M. net BRET values are represented as lines on graph. A significant increase in mean net BRET value after glutamate application was observed in the mGluR2-transfected group (one-way ANOVA followed by Newman-Keuls post-test, ***P < 0.001). (C and D) Test of Gαs-mediated stimulation of cAMP production. (Ci) Time course of net BRET readings from empty vector and untagged GPR18-expressing HEK cells. F, forskolin (10 μM). (Cii) Net BRET values 4 minutes after injection of NAGly from each sample well are represented as individual points in the dot plot graph. Mean ± S.E.M. net BRET values are represented as lines on graph. No significant difference was observed between empty vector and GPR18-transfected cells (unpaired t test, P > 0.05). Di) Time course of net BRET readings from empty vector and D1R-expressing HEK cells. D, dopamine hydrochloride (10 μM); F, forskolin (10 μM). CTX was applied to HEK cells during overnight incubation. (Dii) Net BRET values 4 minutes after injection of dopamine from each sample well are represented as individual points in the dot plot graph. Mean ± S.E.M. net BRET values are represented as lines on graph. A significant decrease in mean net BRET value after dopamine application was observed in the D1R-transfected group (one-way ANOVA followed by Newman-Keuls post-test, ***P < 0.001).

Similar articles

See all similar articles

Cited by 23 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback