Abstract
GPR40 is a G protein-coupled receptor (GPCR) whose endogenous ligands have recently been identified as medium- and long-chain free fatty acids (FFAs), and it is thought to play an important role in insulin release. Despite recent research efforts, much still remains unclear in our understanding of its pharmacology, mainly because the receptor-ligand interaction has not been analyzed directly. To study the pharmacology of GPR40 in a more direct fashion, we developed a flow cytometry-based binding assay. FLAG-tagged GPR40 protein was expressed in Sf9 cells, solubilized, immobilized on immunomagnetic beads, and labeled with the fluorescent probe C1-BODIPY-C12. Flow cytometry analysis showed that C1-BODIPY-C12 specifically labels a single class of binding site in a saturable and reversible manner with an apparent dissociation constant of ∼3 μM. The FFAs that activate GPR40 competed with C1-BODIPY-C12 binding; thus, medium- to long-chain FFAs could compete, whereas short-chain FFAs and methyl linoleate had no inhibitory effect. Furthermore, ligands that are known to activate GPR40 competed for binding in a concentration-dependent manner. All the ligands that inhibited the binding promoted phosphorylation of extracellular signal-regulated kinase (ERK)-1/2 in human embryonic kidney (HEK) 293 cells that expressed GPR40 and [Ca2+]i responses in mouse insulinoma (MIN6) cells that natively express GPR40; however, pioglitazone, a thiazolidinedione that failed to compete for the binding, did not activate ERK or [Ca2+]i response. This study showed that a flow cytometry-based binding assay can successfully identify direct interactions between GPR40 and its ligands. This approach would be of value in studying the pharmacology of GPCRs.
Free fatty acids (FFAs) have been demonstrated as ligands for orphan GPCRs (GPR40, GPR41, GPR43, GPR84, and GPR120) and have been proposed to play important roles in a various physiological responses (Briscoe et al., 2003; Brown et al., 2003; Itoh et al., 2003; Kotarsky et al., 2003; Le Poul et al., 2003; Nilsson et al., 2003; Hirasawa et al., 2005; Wang et al., 2006). They are activated by FFAs of different chain lengths with varying degrees of specificity, such that GPR40 (known as free fatty acid receptor 1; FFAR1), GPR84, and GPR120 prefer medium- to long-chain fatty acids (Itoh et al., 2003; Hirasawa et al., 2005; Wang et al., 2006), whereas GPR41 and GPR43, which are now called FFAR3 and FFAR2, respectively, prefer chain lengths of C2 to C5 (Le Poul et al., 2003; Nilsson et al., 2003). GPR40 is coupled to Gq, which results in the activation of phospholipase C (Hardy et al., 2005) and subsequent increases in the intracellular calcium concentration ([Ca2+]i) (Itoh et al., 2003). In addition, GPR40 has been reported to promote the phosphorylation of ERK-1/2 (Yonezawa et al., 2008). GPR40 mRNA is expressed primarily in the pancreas, brain, and monocytes (Briscoe et al., 2003; Itoh et al., 2003). A number of in vitro and in vivo studies have now demonstrated that FFAs promote glucose-stimulated insulin secretion in pancreatic β-cells via GPR40 (Briscoe et al., 2003; Itoh et al., 2003; Poitout, 2003; Steneberg et al., 2005; Feng et al., 2006); therefore, research into GPR40 has the potential to lead to the development of antidiabetes drugs.
Despite intensive research efforts, the pharmacology of GPR40 is not yet fully understood because most of the data that have been collected were obtained by monitoring the GPR40-mediated second messengers, and direct analysis of the receptor-ligand relationship has not yet been performed. The development of a direct binding assay for GPR40 (and other FFARs) has been hindered mainly by the lack of specific labeled probes. Here, we show that some fluorescent-labeled FFAs can specifically label GPR40. Furthermore, we show that a flow cytometry-based binding assay that uses one of these ligands, C1-BODIPY-C12, as a specific probe can be used to monitor successfully the interaction of GPR40 with its ligands.
Materials and Methods
Materials. Fluorescent-labeled FFA analogs were purchased from Invitrogen (Carlsbad, CA) (Table 1). They included: 4,4-difluoro-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene (BODIPY), 4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoic acid (C1-BODIPY-C12), 5-butyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene-3-nonanoic acid (C4-BODIPY-C9), 4,4-difluoro-5-octyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoic acid (C8-BODIPY-C5), and 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoic acid (BODIPY-C5). Troglitazone was a gift from Daiichi Sankyo Co., Ltd. (Tokyo, Japan). Pioglitazone and rosiglitazone were purchased from Alexis Biochemicals (San Diego, CA). Ciglitazone and MEDICA16 were purchased from Sigma (St. Louis, MO). GW9508 was synthesized according to the previously reported procedure (Garrido et al., 2006) and purchased from Namiki Shoji Co., Ltd. (Tokyo, Japan). The FLAG-human GPR40 (hGPR40)/pcDNA5/FRT/TO plasmid was prepared as described previously (Hirasawa et al., 2008). In brief, hGPR40 cDNA was obtained by polymerase chain reaction using genomic DNA as a template and ligated into the multicloning site of the mammalian expression vector pcDNA5/FRT/TO (Invitrogen) together with an N-terminal FLAG-tag. All other materials were from standard sources and of the highest purity commercially available.
Cell Culture. The Flp-In T-REx 293 cells (Invitrogen) were cultured in Dulbecco's modified Eagle's medium (DMEM; Sigma) that had been supplemented with 10% fetal bovine serum (FBS), 10 μg/ml blasticidin S (Funakoshi, Tokyo, Japan) and 100 μg/ml Zeocin (Invitrogen). In this study, we used FBS, which we had confirmed to have no inducible effect for T-REx system. MIN6 cells were cultured in DMEM containing 25 mM glucose supplemented with 10% fetal calf serum, 60 μM β-mercaptoethanol, and 0.5% streptomycin and penicillin sulfate. All mammalian cells were grown at 37°C in a humidified atmosphere of 5% CO2/95% air. The Spodoptera frugiperda ovarian (Sf9) cells were cultured in Grace's medium (Invitrogen) that had been supplemented with 10% FBS and grown in monolayer culture at 27°C in 200-ml flasks.
T-REx Expression System. We used Flp-In T-REx 293 cells to develop stable cell lines (T-Rex hGPR40) in which the expression of GPR40 could be induced with doxycycline (Dox). Cells were transfected with FLAG-hGPR40/pcDNA5/FRT/TO using Lipofectamine reagent (Invitrogen), and selected with DMEM that had been supplemented with 10% FBS, 10 μg/ml blasticidin S, and 100 μg/ml hygromycin B (Sigma). GPR40 protein expression was induced with 10 μg/ml Dox for 24 h.
ERK Assay. Activation of ERK-1/2 in T-REx hGPR40 cells was assayed by Western blotting as described previously (Hirasawa et al., 2008). In brief, cells were serum-starved for 2 h and treated with each compound that was being tested at a concentration of 100 μM. After 10 min of incubation with each compound, total cell extracts were prepared and subjected to Western blotting using anti-phospho- and anti-total-kinase antibodies.
Baculovirus Expression System. To construct the pBAC-FLAG-hGPR40 plasmid, the FLAG-hGPR40 cDNA fragment was prepared from the FLAG-hGPR40/pcDNA5/FRT/TO plasmid and inserted into the pBAC transfer plasmid (Takara Bio, Shiga, Japan). To generate recombinant baculovirus, Sf9 cells were cotransfected with the pBAC-FLAG-hGPR40 plasmid and BacVector-1000 Triple Cut Virus DNA (Novagen, Madison, WI). The recombinant baculovirus was then purified by plaque assay. For expression, the Sf9 cells were grown to a density of 60 to 70% and then infected with recombinant baculovirus at a multiplicity of infection of 10 for 72 h.
Preparation of the GPR40-Bead Complex. GPR40 was complexed to magnetic beads according to the following procedure. Cells that had been infected with recombinant baculovirus were collected and solubilized with 1% digitonin/PBS that contained 1% protease inhibitor cocktail (Nacalai Tesque, Kyoto, Japan). The lysates were centrifuged in a microcentrifuge for 5 min at 15,000 rpm to sediment the insoluble material. The soluble fraction was removed and incubated with 4 μg of anti-FLAG M2 antibody (Sigma) for 1 h on a rotator at 4°C. After 1 h of incubation with 15 μl of Magnabind protein G beads (Pierce, Rockford, IL), the GPR40-bead complex was collected using an external magnetic field and transferred into 170 μl of 1% digitonin/PBS that contained 1% protease inhibitor cocktail. For some Western blotting experiments, the GPR40-bead complex that had been collected was heated with SDS-polyacrylamide gel electrophoresis sample buffer for 5 min at 70°C, the beads were removed using an external magnetic field, and the supernatant was analyzed.
Flow Cytometry Analysis. The fluorescence emissions were analyzed using a FACSCalibur (BD Biosciences, San Jose, CA). For the binding analysis, samples of the GPR40-bead complex that had been resuspended in 1% digitonin/PBS that contained 1% protease inhibitor cocktail were used. Thirty thousand events were analyzed per sample, using forward scatter versus side scatter dot-plot gating to resolve the primary population of bead particles. The data were collected from FL1 (BODIPY fluorescence) in log mode. The data were normalized by using fluorescence calibration beads (BD Biosciences, San Jose, CA).
Flow Cytometry-Based Binding Assay. A 10-μl aliquot of the bead complex (∼8.8 × 104 counts/μl) was incubated for 5 min at room temperature in a 3.5-ml U-bottomed plastic tube with test compounds that had been diluted with 0.1% digitonin/PBS containing 1% protease inhibitor cocktail. Each BODIPY-labeled FFA was diluted in 0.1% digitonin/PBS containing 1% protease inhibitor cocktail and added to the bead-complex. After the samples had been allowed to equilibrate for 5 min, their fluorescence was measured using the FACSCalibur.
The reversibility of binding was monitored in following procedure. In brief, GPR40-bead complex was preincubated for 5 min with C1-BODIPY-C12 until equilibrium was obtained. After the addition of 100 μM troglitazone at the 0-s time point, the fluorescence intensity was monitored. Data were best fitted using a two-phase exponential decay function.
[Ca2+]i Measurement. [Ca2+]i were recorded by conventional Ca2+ imaging method using an image processor (Argus 50; Hamamatsu Photonics, Hamamatsu, Japan). For [Ca2+]i measurement, MIN6 cells were loaded with fura-2 acetoxymethyl ester (fura-2 AM; Dojindo, Tokyo, Japan) by incubation in 2 μM fura-2 acetoxymethyl ester for 30 min at 37°C. [Ca2+]i measurement was performed at 30°C in Tyrode's solution. Fluorescence of fura-2 was measured by applying UV light at 340 and 380 nm alternatively and by leading emission light through a 505-nm dichroic mirror (DCLP; Omega Optical, Brattleboro, VT). Fluorescence was detected by an SPD-CCD camera (MC681APD-R0B0; Texas Instruments, Dallas, TX). Ca2+ images were acquired at intervals of 20 s and processed to calculate F340/F380 later using NIH Image (http://rsbweb.nih.gov/nih-image/).
Data Analysis. Curve fitting and parameter estimations were performed by using the Igor Pro 3.1.4 software (WaveMetrics, Lake Oswego, OR). The level of significance for the difference between sets of data was assessed using an unpaired Student's t test. Data were expressed as means ± S.E. p < 0.05 was considered statistically significant.
Results
Analysis of the GPR40-Bead Complex. The first step in the development of the binding assay was to immobilize solubilized FLAG-tagged GPR40 protein on magnetic protein G beads via an anti-FLAG antibody, as shown in Fig. 1. To determine whether the GPR40 protein was enriched and immobilized on the beads, Western blotting was performed. The first three lanes of Fig. 2, left, show the Western blotting of whole-cell lysates. The anti-FLAG antibody detected a protein of approximately 30 kDa and a smear of proteins from 75 to 250 kDa in the lysate from cells that had been infected with the recombinant baculovirus, which expressed FLAG-tagged GPR40. These proteins were not detected in the lysates from cells that were uninfected or that had been infected with nonrecombinant baculovirus.
The Western blotting performed on the supernatant collected after removing beads from the GPR40-bead complex was shown in Fig. 2, right. For the samples that were derived from either uninfected cells or cells that had been infected with nonrecombinant baculovirus, Western blotting with the anti-FLAG antibody showed that there were several bands in common between these two samples, which presumably correspond to proteins that are bound nonspecifically by the antibody. In contrast, the anti-FLAG antibody produced very strong signals at approximately 30 kDa and in a smear from 75 to 250 kDa in the sample from the cells that had been infected with the recombinant baculovirus. The results confirmed that GPR40 protein that had been prepared from cells that had been infected with recombinant baculovirus was enriched by using immunomagnetic beads.
Fluorescent Ligands for GPR40. To develop a binding assay with the GPR40-bead complex, we examined a variety of radiolabeled ligands, which included [3H]rosiglitazone, but it was difficult to detect specific binding (data not shown). In an attempt to find a fluorescence-labeled probe that allowed the detection of GPR40, we then examined the effects of BODIPY-FFAs on ERK-1/2 activation in T-Rex hGPR40 cells, in which GPR40 expression can be induced by Dox. As shown in Fig. 3A, C4-BODIPY-C9 and C1-BODIPY-C12, but not BODIPY, BODIPY-C5, or C8-BODIPY-C5, resulted in potent ERK-1/2 activation after Dox induction. These two compounds activated ERK-1/2 to a degree similar to that of linolenic acid (LA), which is the natural ligand for GPR40. Furthermore, we examined whether these BODIPY compounds could specifically label GPR40 protein that had been immobilized on the immunomagnetic beads. We performed flow cytometry analysis with beads that had been prepared with lysates from either uninfected Sf9 cells, cells that had been infected with nonrecombinant baculovirus, or cells that had been infected with recombinant baculovirus. As shown in Fig. 3B, C8-BODIPY-C5, C4-BODIPY-C9, and C1-BODIPY-C12, but not BODIPY or BODIPY-C5, showed strong fluorescence intensity with beads to which GPR40 protein was bound but not with beads that had been incubated with lysate from either the uninfected cells or cells that had been infected with the nonrecombinant baculovirus. Similar fluorescence intensity was observed with C8-BODIPY-C5, C4-BODIPY-C9, and C1-BODIPY-C12, but only C4-BODIPY-C9 and C1-BODIPY-C12, not C8-BODIPY-C5, activated ERK-1/2. Hence, we further examined whether C8-BODIPY-C5 acts as an antagonist for GPR40. As shown in Fig. 3C, C8-BODIPY-C5 inhibited LA-induced ERK-1/2 phosphorylation in a concentration-dependent manner, indicating that C8-BODIPY-C5 may act as an antagonist for GPR40. These results indicated that C8-BODIPY-C5, C4-BODIPY-C9, and C1-BODIPY-C12 could specifically label solubilized GPR40 protein that had been immobilized on immunomagnetic beads; however, C8-BODIPY-C5 has an antagonistic effect, whereas C4-BODIPY-C9 and C1-BODIPY-C12 have agonistic effects, respectively.
Flow Cytometry-Based Binding Assay. Using these BODIPY compounds, we performed saturation experiments. The relative fluorescence intensity of the BODIPY compound that was bound to the GPR40-bead complex was determined at five different concentrations ranging from 1 to 10 μM. Nonspecific binding was determined in the presence of 100 μM troglitazone, which is known to activate GPR40 (Stoddart et al., 2007). Because the three BODIPY compounds showed similar specific binding (in a typical study, specific binding was approximately 20% of total binding at 7 μM BODIPY compound), we used C1-BODIPY-C12 as a specific probe in the following experiments. As shown in Fig. 4A, C1-BODIPY-C12 binding was saturable, with an apparent equilibrium dissociation constant (Kd) of 2.7 ± 0.7 μM (n = 3). In addition, the reversibility of specific C1-BODIPY-C12 binding was assessed. As shown in Fig. 4B, addition of 100 μM troglitazone at equilibrium produced a rapid dissociation (apparent half-time of dissociation was ∼15 s) of the bound C1-BODIPY-C12, demonstrating the reversibility of C1-BODIPY-C12 binding. After approximately 60 s, almost 25% of the C1-BODIPY-C12 was dissociated.
The ability of FFAs to compete for C1-BODIPY-C12 binding sites was examined. FFAs competed for the C1-BODIPY-C12 binding sites in a concentration-dependent manner, and apparent inhibitory binding activities were detected in saturated FFAs of C12 and C16 length and in C18 length unsaturated FFAs. However, methyl-linoleate did not show any inhibitory activity, which suggests that the carboxyl group is indispensable for this interaction. Some eicosanoids also showed inhibitory activity at levels that were comparable with those of long-chain FFAs. The apparent Ki values that were obtained are summarized in Table 2.
We then examined whether the binding to GPR40 of the chemical compounds MEDICA16, GW9508, rosiglitazone, and troglitazone, which were previously shown to activate GPR40 (Kotarsky et al., 2003; Stoddart et al., 2007; Sum et al., 2007), could be monitored by this flow cytometry-based binding assay. As shown in Fig. 4C, these compounds, together with an additional thiazolidinedione, ciglitazone, inhibited the C1-BODIPY-C12 binding in a concentration-dependent manner; however, another thiazolidinedione that was examined, pioglitazone, had no effect on the C1-BODIPY-C12 binding. We further examined whether the binding profile of these compounds correlated with GPR40-mediated ERK activation in T-REx hGPR40 cells as well as with the [Ca2+]i response in MIN6 cells, which endogenously express GPR40. MEDICA16, GW9508, and all the thiazolidinediones that had been shown to inhibit the binding of C1-BODIPY-C12 were found to activate ERK-1/2 in cells expressing GPR40 (Fig. 5), and they also increased [Ca2+]i in MIN6 cells (Fig. 6A and B). On the other hand, pioglitazone, which had been shown to have no direct effect on C1-BODIPY-C12 binding, did not activate the GPR40-mediated ERK-1/2 (Fig. 5) or [Ca2+]i response (Fig. 6, A and B). Thus, the binding profile of the chemical compounds obtained by C1-BODIPY-C12 binding assay seemed to correlate well with their biological effect via GPR40.
Discussion
This study demonstrated that specific C1-BODIPY-C12 binding sites that were identified by a flow cytometry-based assay had the characteristics that were expected for binding sites on the FFAR GPR40. The binding of C1-BODIPY-C12 was of apparently high affinity and was saturable. FFAs competed for the C1-BODIPY-C12 binding sites in a concentration-dependent manner. In addition, as summarized in Table 2, there was good agreement between the apparent Ki values (obtained using the flow cytometry-based binding assay) and apparent EC50 values (obtained by [Ca2+]i monitoring) that were estimated for the FFAs. Thus, the binding properties of the C1-BODIPY-C12 binding sites closely resembled those of GPR40 as delineated by pharmacological procedures ([Ca2+]i) in CHO cells that stably express human GPR40 (Itoh et al., 2003), although both parameters were apparent. Furthermore, the binding profile of not only the FFAs but also the synthetic chemical compounds correlated well with their ability to stimulate ERK-1/2 activity; thus, all the synthetic chemical compounds that inhibited the binding of C1-BODIPY-C12 were shown to activate ERK-1/2 in cells that expressed GPR40, whereas the thiazolidinedione that had no direct effect on the C1-BODIPY-C12 binding (pioglitazone) did not activate the GPR40-mediated ERK response. Moreover, these ligand binding profiles were also found to correlate well with their effect on [Ca2+]i response in MIN6 cells.
This study showed that the assay we developed can directly monitor the interaction between the FFAR GPR40 and its ligands. Besides GPR40, we have recently succeeded in monitoring the interaction between the FFAR GPR120 and its ligands by this technique (T. Hara, A. Mirasawa, Q. Sun, T. Koshimizu, T. Awaji, and G. Tsujimoto, manuscript in preparation). In addition to these FFARs, we had confirmed that this assay was applicable to other GPCRs (such as adrenergic receptors) in a preliminary series of experiments (data not shown). Hence, this technique seems to be applied for GPCRs in general.
This flow cytometry-based assay has advantages and disadvantages compared with other established methods. Similar to radioligand binding assay, this technique also requires specific probe to label the receptor, a fluorescent one in this case. Becasue this technique assesses the receptor-ligand interaction by monitoring the fluorescent emission, and the fluorescent measurement is less rigorous in quantifying the probe-labeled receptor compared with other methods (such as radioligand binding assay), the pharmacological parameters on receptor-ligand interactions obtained by this technique should be considered as “apparent” ones. Despite these problems, this technique has several advantages. As described in this study, the radioligand binding assay with [3H]rosiglitazone on the putative membrane preparation of the cloned GPR40 could not detect the receptor-ligand interactions, mainly because of its high nonspecific binding. The reasons for the successful monitoring by the flow cytometry-based binding assay would include that this technique uses the immunopurified receptor protein, which may minimize the nonspecific binding of probe. In addition, this assay measures fluorescent intensity continuously and does not need to separate the free and bound ligands. With these advantages and disadvantages, the availability of the direct flow cytometry-based binding assay will make it possible to perform further pharmacological characterization of GPCRs, especially those difficult to be monitored by other methods. In addition, the assay should prove useful for high-throughput screening of ligands for such GPCRs.
Footnotes
-
This work was supported in part by research grants from the Scientific Fund of the Ministry of Education, Science, and Culture of Japan (to G.T. and A.H.); the Program for Promotion of Fundamental Studies in Health Sciences of National Institute of Biomedical Innovation (NIBIO) (to G.T.); the Japan Health Science Foundation and the Ministry of Human Health and Welfare (to G.T.); in part by the Mitsubishi Foundation; the Shimadzu Science Foundation (to A.H.); and by the Society for Research on Umami Taste (to A.H.).
-
ABBREVIATIONS: FFA, free fatty acid; GPCR, G protein-coupled receptor; FFAR, free fatty acid receptor; ERK, extracellular regulated kinase; BODIPY, 4,4-difluoro-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene; C1-BODIPY-C12, 4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoic acid; C4-BODIPY-C9, 5-butyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene-3-nonanoic acid; C8-BODIPY-C5, 4,4-difluoro-5-octyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoic acid; BODIPY-C5, 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoic acid; GW9508, 3-[4-({[3-(phenyloxy)phenyl]methyl}amino)phenyl]propanoic acid; MEDICA16, β,β′-tetramethyl hexadecanedioic acid; hGPR40, human GPR40; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; Dox, doxycycline; PBS, phosphate-buffered saline; LA, linolenic acid; MIN6, mouse insulinoma cell line.
- Received September 23, 2008.
- Accepted October 16, 2008.
- The American Society for Pharmacology and Experimental Therapeutics