Delineation of the GPR15 receptor-mediated Gα protein signalling profile in recombinant mammalian cells.

The GPR15 receptor is a G protein-coupled receptor (GPCR), which is activated by an endogenous peptide GPR15L(25-81) and a C-terminal peptide fragment GPR15L(71-81). GPR15 signals through the Gi/o pathway to decrease intracellular cyclic adenosine 3',5'-monophosphate (cAMP). However, the activation profiles of the GPR15 receptor within Gi/o subtypes have not been examined. Moreover, whether the receptor can also couple to Gs , Gq/11 and G12/13 is unclear. Here, GPR15L(25-81) and GPR15L(71-81) are used as pharmacological tool compounds to delineate the GPR15 receptor-mediated Gα protein signalling using a G protein activation assay and second messenger assay conducted on living cells. The results show that the GPR15 receptor preferentially couples to Gi/o rather than other pathways in both assays. Within the Gi/o family, the GPR15 receptor activates all the subtypes (Gi1 , Gi2 , Gi3 , GoA , GoB and Gz ). The Emax and activation rates of Gi1, Gi2 , Gi3, GoA and GoB are similar, whilst the Emax of Gz is smaller and the activation rate is significantly slower. The potencies of both peptides toward each Gi/o subtype have been determined. Furthermore, the GPR15 receptor signals through Gi/o to inhibit cAMP accumulation, which could be blocked by the application of the Gi/o inhibitor pertussis toxin.

INTRODUCTION

The G protein‐coupled receptor 15 (GPR15) is a class A G protein‐coupled receptor (GPCR), which was cloned in 1996 , and shown to be a co‐receptor for human immunodeficiency virus and simian immunodeficiency virus infection. , , The GPR15 receptor is widely expressed in the human body, such as the colon, skin and peripheral blood. Published studies have shown that the GPR15 receptor plays an important role in immune disorders such as ulcerative colitis, dermatosis , and multiple sclerosis. In addition, the GPR15 receptor has been identified as a robust biomarker for tobacco smoking. , , These findings indicate that the GPR15 receptor is a potential therapeutic target for multiple diseases.

The endogenous ligand of the GPR15 receptor, a 57mer peptide termed GPR15L(25–81), has recently been identified by our collaborated laboratory and two other independent research groups. , , The GPR15L(25–81) peptide is encoded by chromosome 10 open reading frame 99 (C10orf99) in humans. It is a soluble basic amphiphilic peptide that contains 57‐amino acid residues (molecule weight 6.5 KD). It contains two intramolecular disulphide bridges (Cys40 to Cys63 and Cys41 to Cys60) and thus shows resemblance to the structure of chemokine peptides but not to their peptide sequence. , , The C‐terminal of the GPR15L(25–81) peptide is highly conserved among species and is essential for agonist activity. When deleting the C‐terminal region, the peptide is incapable of activating the receptor. However, when progressively deleting the N‐terminal amino acids, all truncated constructs (e.g. GPR15L(71–81)) are still capable of activating the GPR15 receptor. The potencies of those truncated variants were positively correlated to their length. , , In this study, we used the GPR15L(25–81) and its C‐terminal fragment GPR15L(71–81) as pharmacological tool compounds to probe the GPR15 receptor signalling.

There are 17 mammalian Gα proteins that together with Gβ and Gγ subunits form a functional trimeric G protein capable of coupling to GPCRs upon activation by agonists. They have been divided into four classes based on the evolutionary distance, which are Gi/o (Gi1, Gi2, Gi3, GoA, GoB, Gz, Gt1, Gt2, Gg), Gs (Gs, Golf), Gq/11 (Gq, G11, G14, G15/16) and G12/13 (G12, G13). The activation of each class of Gα protein leads to distinct downstream signalling pathways. Canonically, the Gi/o inhibits the adenylyl cyclase (AC) and decreases the intracellular cyclic adenosine 3′,5′‐monophosphate (cAMP) concentration. The Gs activates the adenylyl cyclase and increases the intracellular cAMP concentration. The Gq/11 activates the beta‐type phospholipase C and leads to the generation of inositol trisphosphate (IP3) and diacylglycerol. G12/13 activation regulates the small GTPases Rho and leads to c‐Jun N‐terminal kinase activation.

Our group and others have recently shown that the GPR15 receptor signals via the Gi/o pathway in cAMP second messenger assays. , However, the coupling “fingerprint” within the Gi/o protein family is unclear. Moreover, whether the GPR15 receptor also couples to the Gs, Gq/11 and/or G12/13 signalling pathways remains to be determined. In this study, we examined the coupling selectivity of the GPR15 receptor toward 11 Gα proteins representing the four families (Gi1, Gi2, Gi3, GoA, GoB, Gz, Gs, Gq, G11, G15 and G13) by conducting the NanoLuc bioluminescence resonance energy transfer (BRET) based G protein activation assay. Wherein the masGRK3ct‐Nluc serves as BRET energy donor is anchored on the inside of the cell membrane, and the functional Venus tagged Gβγ dimmer serves as the acceptor is formed based on the bimolecular fluorescence complementation (BiFC) of Venus (155–239)‐Gβ1 and Venus (1–155)‐Gγ2. The BRET will happen when the Venus‐Gβγ is released from the heterotrimer and subsequently interacts with the masGRK3ct‐Nluc. We tested the downstream signalling of the GPR15 receptor among the Gi/o, Gs and Gq/11 pathways, by conducting homogenous time‐resolved FRET (HTRF)‐based cAMP and IP1 assays. Moreover, the potencies of the GPR15L(25–81) and GPR15L(71–81) peptides have been determined with both assay formats. The delineation of the GPR15 receptor signalling can contribute to future efforts of drug discovery.

MATERIALS AND METHODS

The study was conducted in accordance with the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies.

Reagents and materials

Dulbecco's Modified Eagle's Medium (DMEM, Cat. No: 12077549), opti‐MEM (Cat. No: 51985026), trypsin‐ethylenediaminetetraacetic acid (Cat. No: 11590626), penicillin/streptomycin (Cat. No: 11548876), dialyzed foetal bovine serum (FBS, Cat. No: 26400036), Hanks' balanced saline solution without Ca2+ and Mg2+ (Cat. No: 11540476), Dulbecco's phosphate‐buffered saline (Cat. No: 14190169), Lipofectamine™ LTX reagent with PLUS™ reagent (Cat. No: 15338100), Lipofectamine 2000 (Cat. No: 11668019), pertussis toxin (PTX, CAS#: 70323‐44‐3, Cat. No.: PHZ1174), pluronic F‐68 non‐ionic surfactant (CAS#: 9003‐11‐6, Cat. No: 24040032), 6‐well cell culture plate (Cat. No: 353046) and 10‐cm tissue culture dish (Cat. No.: 353003) were obtained from Thermo‐Fisher Scientific (Waltham, MA, USA). The 4‐(2‐hydroxyethyl) piperazine‐1‐ethanesulfonic acid (HEPES, Cat. No: H4034‐500G), antibiotic G418 (CAS#: 108321‐42‐2, Cat. No: G8168‐50ML), bovine serum albumin (BSA, Cat. No: A2153‐50G), cell dissociation solution (Cat. No: C5914‐100ML), adenylyl cyclase activator forskolin (FSK, CAS#: 66575‐29‐9, Cat. No: F6886‐10MG), broad‐spectrum phosphodiesterase (PDE) inhibitor 3‐isobutyl‐1‐methylxanthine (IBMX, CAS#: 28822‐58‐4, Cat. No: I5879‐250MG), carbamoylcholine chloride (Carbachol, CAS#: 51‐83‐2, Cat. No: C4382‐1G), isoproterenol bitartrate salt (Isoproterenol, CAS#: 54750‐10‐6, Cat. No: I2760‐500MG), glucagon (CAS#: 16941‐32‐5, Cat. No.: G2044‐5MG), lithium chloride (LiCl, CAS#: 7447‐41‐, Cat. No: 310468‐100G‐D) and dimethyl sulfoxide (DMSO, CAS#: 67‐68‐5, Cat. No.: D2650‐100ML) were obtained from Sigma Aldrich (St. Louis, MO, USA). The 384‐well white opaque microplate for cAMP assay (Cat. No: 784075) was obtained from Greiner. The 384‐well white opaque microplate for IP‐one assay (Cat. No.: 6007299) and 96‐well white opaque (Cat. No.: 6005688) for BRET assay were obtained from PerkinElmer. cAMP‐Gs dynamic kit (Cat. No: 62AM4PEC) and IP‐One‐Gq kit (Cat. No: 62IPAPEC) were obtained from Cisbio (Codolet, France). Human GPR15L(25–81) peptide sequence: KRRPAKAWSGRRTRLCCHRVPSPNSTNLKGHHVRLCKPCKLEPEPRLWVVPGALPQV, Cat. No.: 03489) and human GPR15L(71–81) peptide (sequence: LWVVPGALPQV, custom made) were obtained from Phoenix Pharmaceuticals. Nano‐Glo® Luciferase Assay System (Cat. No: N1110) was obtained from Promega. Envision multimode plate reader was obtained from PerkinElmer (Cat. No.: 2104). The multi‐mode plate reader POLARstar Omega was from BMG LABTECH.

Plasmid DNA constructs

Plasmid encoding N‐terminal 3xHA tagged GPR15‐pcDNA3.1 + (Cat. No: GPR015TN00, human) was obtained from cDNA.org. Plasmids encoding Venus(155–239)‐Gβ1 (human); Venus(1–155)‐Gγ2 (human); mas GRK3ct‐Nluc‐HA (human); Gi1 (rat); Gi2 (rat); Gi3 (rat); GoA (human); GoB (human), Gz (human); Gq (human); G11 (human); G15 (human, also known as human G16); Gs (human); G13 (human); flag‐ric‐8A (human); PTX‐S1 (human); pcDNA3.1 + (human) were kindly provided by Dr. Kirill Martemyanov, Scripps Research, FL, USA.

Cell lines and culturing

Cells (passage < 30) were cultured in a 37°C incubator in a humidified 5% CO2 atmosphere. The culture media for HEK293A cells were DMEM supplemented with 10% dFBS and 100 U/ml penicillin‐streptomycin. The hGPR15‐HEK293A (HEK293A cells stably expressing human GPR15 receptor) recombinant cell line was established by selection with G418. The culture media for the stable cell line was DMEM supplemented with 10% dFBS and 500 μg/ml G418. As for the inhibition of Gi/o, 100 ng/ml of PTX was administered to the cells 24 h before the experiment.

Transient transfection for BRET G protein activation assay

At 20–24 h before the BRET assay, 2 × 105 per well of HEK293A cells were seeded into a 6‐well plate. The cells were transfected with 1.26 μg GPR15 receptor DNA, 0.21 μg Venus (155–239)‐Gβ1 DNA, 0.21 μg Venus (1–155)‐Gγ2 DNA, 0.21 μg masGRK3ct‐Nluc DNA and an optimized amount of Gα protein DNA (supporting information Table S1). For the successful expression of G15, co‐transfection with the chaperone Flag‐Ric‐8A (0.21 μg) is necessary. To avoid the positive BRET signal induced by endogenous Gi/o, the G proteins from the non‐Gi/o family (i.e. Gq, G11, G15, Gs and G13) were co‐transfected with 0.21 μg PTX‐S1 DNA. The transfection reagents used were Lipofectamine LTX reagent with PLUS reagent and/or Lipofectamine 2000.

BRET G protein activation assay

The assay step has previously been described by Masuho et al. in detail and is briefly summarized here. Kinetic assay on LUMIstar plate reader: 25‐μl/well cell suspension (containing 50 000 to 100 000 cells) was added to a white opaque flat‐bottom 96‐well culture plate. Whereafter 25 μl/well furimazine (1:250 dilution based on the manufacturer's instruction) was added automatically with one of the built‐in injectors. The plate was shaken and incubated for 3 min. BRET was then measured by simultaneous measurement of emission at 475 nm (donor) and 535 nm (acceptor). After 5 s of the basal BRET signal measurement, 50 μl/well of the agonist (2 μM GPR15L(25–81) or 20 μM of GPR15L(71–81)) or BRET buffer was added to the wells automatically with another built‐in injectors. BRET signals were then measured for 13 s with 0.06 s (Gz) or 0.02 s (remaining G proteins) intervals. Endpoint assay on Envision plate reader: The same procedure as for the LUMIstar was followed except that the agonist was added manually with the multiple channel pipette whereafter the plate was read immediately (practically around 10 s after agonist addition).

HTRF Gi cAMP assay

The assays were performed in a 384‐well microplate, suspension format as previously described in detail. GPR15L(25–81) and GPR15L(71–81) were threefold diluted in ligand buffer (Hanks' balanced saline solution supplemented with 20 mM HEPES, 1 mM CaCl , 1 mM MgCl , 3 μM forskolin, and 0.01% freshly added pluronic acid, pH adjusted to 7.4 with NaOH). The highest concentration (2 × final concentration) of the GPR15L(25–81) and GPR15L(71–81) peptides were 2 μM and 60 μM, respectively. The 5 μl/well of peptide solution was transferred to a 384‐well microplate and put aside for later use. HEK293A‐GPR15 cells with 80–90% growth confluence were harvested with the non‐enzymatic cell dissociation solution. The cells were suspended in 37°C preheated cell suspension buffer (Hanks' balanced saline solution supplemented with 20 mM HEPES, 1 mM CaCl , 1 mM MgCl , and 100 μM of freshly added IBMX, pH adjusted to 7.4 with NaOH) with a concentration of 8 × 105 cells/ml. Whereafter 5 μl/well (i.e. 4000 cells/well) of cell suspension was added to the 384‐well plate. The plate was centrifuged for 10 s at 500 rpm and incubated at room temperature for 30 min. Afterwards, 10 μl/well of freshly made detection solution (Lysis buffer supplemented with 2.5% cAMP‐cryptate and 2.5% anti‐cAMP‐d2) was added to the plate in a dim light environment. The plate was incubated for 1 h in the dark at room temperature. Whereafter the plate was read in an Envision plate reader to detect emission at 665 nm and 615 nm simultaneously.

HTRF Gs cAMP assay

The assay protocol was performed as described previously. The Gs cAMP assay procedure was very similar to the Gi cAMP assay procedure with the only difference being that the ligand buffer used in the Gs cAMP assay was without forskolin.

HTRF Gq IP1 assay

The assay protocol was performed as described previously. The 5 μl/well of agonists (GPR15L(25–81), GPR15L(71–81), isoproterenol and carbachol) prepared in ligand buffer (Hanks' balanced saline solution supplemented with 20mM HEPES, 1mM CaCl2, 1mM MgCl2, and 0.01% freshly added pluronic acid, pH adjusted to 7.4 with NaOH) with 2 × final concentration was added to the 384‐well microplate. The 5 μl of cell suspensions supplemented with 30 mM LiCl and 20 000 HEK293A‐GPR15 recombinant cells was added to the agonist‐containing plate. The plate was incubated for 30 min at 37°C. Whereafter 10 μl of IP1 detection solution (IP1‐cryptate: anti‐IP1‐d2: lysis buffer = 1: 1: 38) was added and followed by incubating the plate in the dark for 1 h at room temperature. The emissions at 665 nm and 615 nm were then recorded simultaneously with the Envision plate reader.

Data analysis

All statistical analysis was performed with Prism 8 (GraphPad Software, San Diego, CA, USA). Data are presented as the standard error of the mean of at least three independent experiments. Concentration‐response curves were fitted with a log (agonist) vs. response‐variable slope (four parameters) model to determine the EC50 value. The kinetic parameter comparisons between the Gz and other Gi/o subunits were performed with a one‐way analysis of variance test followed by Dunnett's multiple comparison test. The kinetic parameter comparisons between the GPR15L(25–81) and the GPR15L(71–81) were performed with Student's t test. The p values were indicated by asterisk(s): *, p < 0.05; **, p < 0.01; ***, p < 0.001. The HTRF ratio = emission of the d2 at 665 nm/emission of the Eu3+ cryptate labelled anti‐cAMP/IP1 antibody at 615 nm. cAMP or IP1 concentrations were calculated from cAMP or IP1 standard curves. The BRET ratio = emission of Venus at 535 nm/emission of NanoLuc at 475 nm with a 30‐nm band path. The BRET ratio amplitude = BRET ratio after agonist or buffer application − BRET ratio before agonist or buffer application. The E max and the activation rate 1/τ were determined by fitting the kinetic readouts with the one‐phase exponential decay equation.

RESULTS

GPR15 receptor preferentially couples to Gi/o family

We examined the coupling profiles of GPR15 toward 11 representative Gα proteins spanning the four Gα protein families with the NanoLuc BRET G protein activation assay. At 20–24 h before the experiment, we co‐transfected the GPR15 receptor, a specific Gα protein and the BRET sensors (Venus(155–239)‐Gβ1, Venus(1–155)‐Gγ2 and masGRK3ct‐Nluc‐HA) into wild‐type HEK293A cells with optimized DNA amounts (supporting information Table S1). We found that both the GPR15L(25–81) (Figure 1A,B) and the GPR15L(71–81) (supporting information Figure S1A, B) peptides led to effective activation of all the members of the Gi/o family (Gi1, Gi2, Gi3, GoA, GoB and Gz). No response to either peptide was detected when the cells were treated with the BRET buffer (supporting information Figure S2), or when cells were only transfected with the Gα protein and the BRET sensors (i.e. no GPR15 receptor present) (supporting information Figure S3A,B). The negative control results demonstrate that the BRET signals in Figure 1A,B, and supporting information Figure 1A, 1B were specifically mediated by the GPR15 receptor and its cognate GPR15L peptides. No response was detected when testing the peptides on cells transfected with the GPR15 receptor and Gα proteins (Gs, Gq, G11, G15 and G13) representing the Gs, Gq/11 and G12/13 families, indicating that the GPR15 receptor does not interact with the Gs, Gq/11 and G12/13 proteins (Figure 1C,D,E; supporting information Figure S1C, D, E). To ensure assay validity, we included the GCGR that promiscuously couples to all four Gα protein families as a positive control. Our results show that when GCGR was stimulated with 10 μM glucagon, GCGR indeed coupled to the Gs, Gq, G11, G15 and G13 (supporting information Figure S4). Taken together, our results show that among the four Gα protein families (Gi/o, Gs, Gq/11 and G12/13), the GPR15 receptor preferentially couples to the Gi/o signalling pathway and activates all the members of this family effectively.

FIGURE 1

Real‐time measurement of G protein coupling profiles of the GPR15 receptor activated by GPR15L(25–81) peptide. (A, B) GPR15, Gα, Venus 155–239 Gβ1, Venus 1–155 Gγ2 and masGRK3ct‐Nluc‐HA transfected HEK293A cells treated with 1 μM of GPR15L(25–81) peptide and activated heterotrimeric Gi/o proteins (Gi1, Gi2, Gi3, GoA, GoB and Gz) to release the Gβγ subunits leading to BRET signal increase. Note the difference in scale of Y axis. (C–E) No response was detected when the cells were treated with 1 μM of GPR15L(25–81) peptide toward G proteins from Gs, Gq/11 and G12/13 families. The arrow indicates the administration of the agonist. Data plotted as mean ± SEM (error bars) of 3–5 grouped independent experiments performed in duplicates. BRET: Bioluminescence resonance energy transfer; SEM: standard error of the mean

The kinetic parameters of the Gi/o activation mediated by the GPR15 receptor

Based on data obtained from the kinetic assay, we quantified the E max and activation rate kinetic parameters. Our results showed that there were no significant (p > 0.05) kinetic parameter differences between the GPR15L(25–81) and the GPR15L(71–81) toward all Gi/o proteins (i.e. Gi1, Gi2, Gi3, GoA, GoB and Gz) (Figure 2A,B). The activation kinetics among the Gi1, Gi2, Gi3, GoA and GoB were quite similar. However, regarding the Gz activation, when stimulated with the GPR15L(25–81), Gz produced a much lower E max and slower activation rate for both versions of peptides (Figure 2A,B). Taken together, the GPR15 receptor showed similar selectivity preference among the Gi1, Gi2, Gi3, GoA and GoB with much lower preference for Gz. Moreover, the GPR15L(25–81) and the GPR15L(71–81) induced similar G protein coupling profiles.

FIGURE 2

The E max and activation rate of the Gi/o activation induced by GPR15L(25–81) and GPR15L(71–81). (A) the E max and (B) the activation rates of Gi1, Gi2, Gi3, GoA, GoB and Gz induced by 1 μM GPR15L(25–81) and 10 μM GPR15L(71–81). The E max of Gz is significantly (p < 0.01) smaller than Gi1, Gi2, Gi3, GoA and GoB. The activation rate of Gz is significantly slower than Gi1, Gi2, Gi3, GoA and GoB (p < 0.001). No significant difference was detected between the GPR15L(25–81) and the GPR15L(71–81) triggered response. The comparisons between the Gz and other Gi/o subunits were performed with Dunnett's multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001). The comparisons between the GPR15L(25–81) and the GPR15L(71–81) were performed with the students t test. Data plot as mean ± SEM (error bars) of 3–5 grouped independent experiments performed in two replicates. BRET: Bioluminescence resonance energy transfer; SEM: standard error of the mean

GPR15L peptides activate each Gi/o subtype in a concentration‐dependent manner

To gain further insight, we determined the potency of the GPR15L(25–81) and its C‐terminal fragment GPR15L(71–81) toward activation of the Gi1, Gi2, Gi3, GoA and GoB proteins with the BRET G protein activation assay (Table 1). HEK293A cells transiently expressing the GPR15 receptor, the Gi/o protein and the BRET pair were stimulated with increasing concentrations of the GPR15L peptides. The E max of each concentration was recorded to determine the concentration‐response curve. Our results showed that both GPR15L(25–81) (Figure 3) and GPR15L(71–81) (supporting information Figure S5) activated each Gi/o protein in a concentration‐dependent manner. The GPR15L(25–81) was more potent than the GPR15L(71–81) on each Gi/o protein. The potency ratio (i.e. EC50 of GPR15L(71–81)/EC50 of GPR15L(25–81)) was 45, 23, 45, 58 and 89 toward Gi1, Gi2, Gi3, GoA and GoB, respectively (Table 1).

TABLE 1

pEC50, EC50 and potency ratio summary of the GPR15L(25–81) and GPR15L(71–81) peptide on Gi/o proteins obtained from BRET G protein activation assay and the Gi cAMP assay. GPR15L(25–81) is more potent than the C‐terminal truncated peptide GPR15L(71–81). Potency ratio calculated as EC50 of GPR15L(71–81)/EC50 of GPR15L(25–81). Where a ratio >1 indicates that GPR15L(25–81) is more potent at the individual G protein than GPR15L(71–81), and a larger potency ratio correlates to higher potency differences between the two versions of peptides. Data shown as mean ± SEM of 3–5 grouped independent experiments. BRET: Bioluminescence resonance energy transfer; SEM: standard error of the mean

	BRET G protein activation assay					Gi cAMP assay
	Gi1	Gi2	Gi3	GoA	GoB	Gi/o
GPR15L(71–81) pEC50	6.14 ± 0.05	5.77 ± 0.04	6.04 ± 0.12	5.52 ± 0.09	5.60 ± 0.10	6.39 ± 0.07
GPR15L(71–81) EC50 (nM)	724	1700	912	3020	2510	407
GPR15L(25–81) pEC50	7.79 ± 0.06	7.14 ± 0.04	7.69 ± 0.10	7.28 ± 0.03	7.54 ± 0.02	7.96 ± 0.17
GPR15L(25–81) EC50 (nM)	16	72	20	52	28	10
Potency ratio	45	23	45	58	89	40

FIGURE 3

Concentration‐response curves of GPR15L(25–81) peptide toward distinct G protein from Gi/o family. Concentration‐response curves of the GPR15L(25–81) peptide on Gi1, Gi2, Gi3, GoA and GoB proteins were determined by conducting the BRET G protein activation assay on HEK293A cells transiently expressing the GPR15 receptor, the Gα protein and the BRET pair. The E max of each concentration was buffer corrected and normalized to the max dose‐response (100%). Data plotted as mean ± SEM (error bars) of 3–5 grouped independent experiments performed in two to three replicates. BRET: Bioluminescence resonance energy transfer; SEM; standard error of the mean.

GPR15 receptor signals through Gi/o rather than Gs and Gq/11 in the downstream signalling pathway

The results described above were obtained from the BRET G protein activation assay, which reflects the proximity of the GPCR and G protein, upstream of the signal cascade. Next, we also investigated the downstream signalling pathway by measuring second messenger molecules cAMP and IP1. We measured cAMP accumulation in Gi/o and Gs mode by including/excluding the adenylate cyclase activator forskolin, respectively. The pan‐phosphodiesterase inhibitor IBMX was applied to prevent the cAMP degradation in the cAMP assay. The Gq pathway was assessed by measurement of IP1 accumulation in the presence of LiCl to prevent further breakdown to inositol monophosphate. All three pathways were endpoint assays based on HTRF technology. ,

In the Gi cAMP assay, both GPR15L peptides led to decreased intracellular cAMP concentrations in a concentration‐dependent manner (Figure 4A). The GPR15L(25–81) peptide was 40‐fold more potent than its C‐terminal fragment GPR15L(71–81) (Table 1). Moreover, the inhibition could be eliminated by the Gi/o inhibitor PTX (Figure 4A) demonstrating that the negative regulation of cAMP production was mediated by Gi/o proteins.

FIGURE 4

Concentration‐response curves of the GPR15L peptides in three HTRF based assays of Gi, Go and Gq downstream pathways. (A) Concentration‐response curves of full‐length GPR15L(25–81) and C‐terminal peptide GPR15L(71–81) with a Gi cAMP assay conducted on the HEK293A‐GPR15 recombinant cells. The activation effect of the GPR15L peptides could be blocked by the Gi/o inhibitor PTX. (B) Concentration‐response curves of the GPR15L peptides in the Gs cAMP assay. The endogenously expressed ß2‐adrenergic receptor and its agonist isoproterenol was used as positive control. (C) Concentration‐response curves of the GPR15L peptides in the Gq IP1 assay. The endogenously expressed muscarinic acetylcholine receptor (mAChR1, mAChR3 and mAChR5) and its agonist carbachol was used as positive control. Data were buffer corrected and then normalized to the maximal response of FSK (forskolin), isoproterenol and carbachol, respectively. Data plotted as mean ± S.E.M. (error bars) of 3–5 grouped independent experiments performed in three replicates. HTRF: homogenous time‐resolved Förster resonance energy transfer; PTX: pertussis toxin; FSK: forskolin; Iso: isoproterenol; carb.: carbachol; SEM: standard error of the mean

In the Gs cAMP assay, no cAMP response was detected by either GPR15L peptide. To ensure assay validity, the Gs coupled β2‐adrenergic receptor, which is endogenously expressed in the parental HEK293A cells, was used as a positive control. We detected a robust cAMP level increase when stimulating with the cells with the β2‐adrenergic receptor agonist isoproterenol (Figure 4B).

In the Gq IP1 assay, no IP1 response was detected by either GPR15L peptide. To ensure assay validity, we used endogenously expressed Gq coupled muscarinic acetylcholine receptors as a positive control. We detected a robust IP1 level increase when stimulating the HEK293A cells with the muscarinic receptor agonist carbachol (Figure 4C).

DISCUSSION

In this study, by combining the upstream G protein activation assay and the canonical downstream assays, we demonstrated that, among the four G protein families (Gi/o, Gs/olf, Gq/11 and G12/13), the GPR15 receptor preferentially couples to the Gi/o family. Moreover, the GPR15 receptor couples to all members of the Gi/o protein subtypes (Gi1, Gi2, Gi3, GoA, GoB and Gz) and with the least preference to the Gz subtype. The endogenous peptide ligand GPR15L(25–81) is more potent than its C‐terminal peptide GPR15L(71–81). Except for this, both versions of peptides display the same overall signalling profiles.

We thus validate former studies which have shown that the GPR15 receptor signals through Gi/o. , , However, those results were all obtained from the distal part of the signal cascades, where accuracy may be affected due to potential signalling crosstalk. To gain further insight, we conducted a BRET‐based G protein activation assay, which enables the direct examination of the GPR15 receptor‐mediated Gα protein activation. Our results show that the GPR15 receptor coupling profiles are highly consistent between the upstream and downstream assays. Recently, it has been shown that some GPCRs paradoxically recruit/activate G proteins without activating the downstream signalling pathways. , Our results clearly show that this behaviour is not observed for GPR15.

Suply et al. observed calcium signalling when the GPR15 receptor co‐expressed with G16 in CHO‐K1 cells was stimulated by GPR15L peptides, which suggests that GPR15 can activate the promiscuous G16 protein and putatively also other members of the Gq/11 family. However, we did not detect activation of Gq, G11 or G15, or IP1 accumulation in HEK293A cells. This discrepancy could either be caused by differences in cell background, in assay sensitivity, or in assay formats.

Our concentration‐response experiments show that the endogenous peptide ligand GPR15L(25–81) (EC50 = 10 nM) is 40‐fold more potent than its C‐terminal peptide GPR15L(71–81) (EC50 = 407 nM) in the Gi cAMP assay. The GPR15L(25–81) is also more potent (23‐ to 89‐fold) than GPR15L(71–81) in the BRET Gi/o protein activation assay. This potency difference agrees well with our previous findings. Wherein the C‐terminal fragments of GPR15L(25–81), peptides were tested with the same Gi cAMP assay as we did here, but using T‐REx 293 cells which need doxycycline to induce GPR15 expression, and where we found that the ability of GPR15L peptides to activate GPR15 are positively correlating to their lengths. We were incapable of determining concentration‐response curves for Gz in the BRET G protein activation assay due to the very small Emax of Gz. To boost the signal, co‐transfection with regulators of G protein signalling proteins may be necessary.

In conclusion, our results demonstrate that the GPR15 receptor shows a high preference for Gi/o signalling when expressed in HEK293A. So far, no such studies have been performed on cells with native GPR15 expressions such as the colon colorectal adenocarcinoma cell lines SW48 and HT29, or some lymphoblast cells such as PM1, Hut78 and NC37, , , which will be important in future studies to determine if the signalling profile in recombinant systems is translatable to the ex vivo situation. To this end, it will also be important to develop better tool ligands such as the first antagonists and small molecule ligands as the current pharmacological toolbox is very limited and not suitable for in vivo studies.

CONFLICT OF INTEREST

The authors declared that there are no conflicts of interest in this study.

Table S1 Optimized DNA transfection ratio of the GPR15 receptor, Gα protein, Ric‐8A, PTX‐S1 and BRET sensors. Ratio 1 equals 0.21 μg DNA. The Ric‐8A is a chaperone for the successful expression of the G15 protein. The PTX‐S1 is a Gi/o inhibitor to avoid potential basal BRET signal caused by the endogenous Gi/o protein of HEK293A cells. The total DNA amount was normalized to 5 μg with the pcDNA3.1+. N.A. represents no transfection. PTX: pertussis toxin

Table S2 DNA transfection ratio of glucagon receptor (GCGR), Gα protein, Ric‐8A, PTX‐S1, and BRET sensors. Ratio 1 equals 0.21 μg DNA. The Ric‐8A is a chaperone for the successful expression of the G15 protein, and the PTX‐S1 is a Gi/o inhibitor to avoid potential basal BRET signal caused by the endogenous Gi/o protein of HEK293A cells. The total DNA amount was normalized to 5 μg with the pcDNA3.1+. N.A. represents no transfection. PTX: pertussis toxin, BRET: bioluminescence resonance energy transfer

Figure S1 Real‐time measurement of G protein activation profiles of the C‐terminal peptide GPR15L(71–81) on the GPR15 receptor. (A, B) GPR15 receptor expressing HEK293A cells treated with 10 μM of GPR15L(71–81) peptide activated the heterotrimeric G protein of Gi/o (Gi1, Gi2, Gi3, GoA, GoB, Gz) leading to the release of the Gβγ subunits and rapid increase in BRET signal. (C‐E) No response was detected from 10 μM GPR15L(71–81) peptide in HEK293A cells co‐expressing GPR15 and G proteins from the Gs, Gq/11 and G12/13 families. The arrow indicates the administration of the agonist. Data plotted as mean ± S.E.M. (error bars) of 3–5 grouped independent experiments performed in duplicates. BRET: bioluminescence resonance energy transfer.

Figure S2 No buffer response detected in BRET G protein activation assay. HEK293A cells transfected with Gi/o (Gi1, Gi2, Gi3, GoA, GoB Gz), GPR15 receptor and the BRET sensors did not respond to buffer. Data plotted as mean ± S.E.M. (error bars) of one representative experiment performed in duplicates. BRET: bioluminescence resonance energy transfer.

Figure S3 Emax of Gi/o, Gs, Gq/11 and G12/13 without GPR15 receptor present. No BRET signal was detected from (A) 1 μM GPR15L(25–81) or (B) 10 μM GPR15L(71–81) in HEK293A cells transfected with Gα protein (Gi1, Gi2, Gi3, GoA, GoB, Gz, Gs, Gq, G11, G15, G13) and BRET sensors in the absence of the GPR15 receptor. BRET: bioluminescence resonance energy transfer

Figure S4 Glucagon receptor (GCGR)‐mediated Gα protein activation. The wild‐type HEK293A cells were transfected with Gα protein, GCGR and BRET sensors as indicated in Table 2(A). When stimulated with 10 μM of glucagon peptide, the GCGR couples to Gs, Gq, G11, G15 and G13. (B) No response was detected when the transfected cells were treated with buffer. Data plotted as mean ± S.E.M. (error bars) of three grouped independent experiments performed in duplicates. BRET: bioluminescence resonance energy transfer.

Figure S5 Concentration‐response curves of the C‐terminal GPR15L(71–81) peptide toward distinct G proteins from the Gi/o family. Concentration‐response curves of GPR15L(71–81) on HEK293A cells transiently transfected with indicated Gi1, Gi2, Gi3, GoA, or GoB proteins, GPR15 and BRET sensors. Data were buffer corrected and normalized to the max dose‐response (100%). Data plotted as mean ± S.E.M. (error bars) of 3–5 grouped independent experiments performed in two to three replicates.

Click here for additional data file.