Linoleic acid as a GPR40 agonist (C₁₈H₃₂O₂; mol. wt. 280.45; purity 99%) and G418 (an aminoglycoside antibiotic, similar in structure to gentamicin B1) were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA). Reagents were dissolved in dimethyl sulfoxide (DMSO; Duchefa-Farma, Haarlem, Netherlands) and diluted in DMSO to the indicated final concentrations. RPMI 1640, fetal bovine serum (FBS), trypsin-EDTA, and penicillin/streptomycin were purchased from Gibco BRL (Grand Island, NY, USA). All other reagents were of analytical grade or complied with the standards for cell culture.

The rat immortalized pancreatic β-cell line RIN-5f (CRL-2058) was purchased from ATCC and maintained in RPMI 1640 medium containing 2.05 mM L-glutamine, 17.86 mM sodium bicarbonate, 25 mM glucose, 10 mM HEPES, 1 mM sodium pyruvate, and 10% heat-inactivated FBS. Cells were incubated in a humidified atmosphere at 37℃ with 5% CO₂.

For construction of the pCMV6-Neo-GPR40 expression plasmid, the vector pEZ-M02-GPR40 was purchased from GeneCopoeia (MD, USA) and inserted into the pCMV6-Neo vector (ORIGENE, MD, USA). E. coli DH-5α cells harboring the selected plasmid DNA were cultured in LB broth media. Plasmid DNA was extracted using a plasmid mini kit (Qiagen, Hilden, Germany). The extracted plasmid DNA was treated with the XmnI and XhoI restriction enzymes. The digested GPR40 gene was separated in a 1.0% agarose gel, eluted using a MEGA-spin Agarose Gel Extraction Kit (Intron, Seongnam, Korea), and ligated into the pCMV6-Neo vector plasmid. The pCMV6-Neo-GPR40 was verified by XmnI/XhoI double digestion and sequencing (Bioneer, Daejeon, Korea). The confirmed vector was prepared using a QIAFilter Plasmid Midi Kit (Hilden, Germany) and used to transfect the cells.

To construct a stable cell line for human GPR40 (NM_005303), RIN-5F cells were transfected with pCMV6-Neo-GPR40 in the transfection reagent Nucleofector Solution V using an Amaxa Nucleofector II device (Amaxa Biosystems, Cologne, Germany). The transfected cells were then selected and maintained in culture medium containing 300 µg/ml G418; this stably GPR40-overexpressing cell line was designated "RIN-40." RIN-40 cells were maintained in G418 selective medium and confirmed by assaying GPR40 mRNA and protein levels.

The cells were plated in 96-well plates (2×10⁵ per well) for 2 days (to 70~80% confluence). On the day of the experiment, the cell culture medium was aspirated and the plate was washed twice with Krebs-Ringer bicarbonate HEPES (KRBH) buffer. Cells were rested at 37℃ for 30 min in KRBH containing 25 mM glucose. Compounds were dissolved in DMSO and added to cells for 2 min. After the reaction, cells were fixed with 10% formalin for 1 h. The fixed cells were treated with 2 µM Fura-2AM in 1 mM EGTA buffer for 30 min. Intracellular Ca²⁺ was measured using a total internal reflection fluorescence (TIRF) microscope.

RIN-5F and RIN-40 cells were seeded in 24-well plates at a density of 2×10⁵/well and cultured for 2 days. The cells were washed twice with glucose-free KRBH buffer containing 103.45 mM NaCl, 5.33 mM KCl, 5.63 mM Na₂HPO₄, 0.407 mM MgSO₄, 1.28 mM CaCl₂, 10 mM HEPES and 17.86 mM NaHCO₃ and then incubated in KRBH buffer with 0.05% bovine serum albumin (BSA) and 2.5 mM glucose for 30 min at 37℃. Then, cells were washed once more with glucose-free KRBH buffer and treated with 500-µL KRBH buffer containing 25 mM glucose and with or without the indicated concentrations of reagents in DMSO (final DMSO concentration of 0.1%). After 2 h, supernatant from each plate was collected and the insulin level determined using an Insulin (Rat) High-range ELISA (ALPCO, Windham, NH). The data were expressed as fold changes compared to the amount of insulin secreted (ng/ml insulin vs. mg/ml total protein).

Whole cells were lysed in Pro-prep protein extraction solution (Intron Biotechnology, Seoul, Korea), and the protein concentration in the lysates was measured using a Bio-Rad protein assay kit (Bio-Rad, Hercules, CA, USA). Equal amounts of proteins (10 µg) were run on a 10% SDS-PAGE gel and transferred onto a PVDF membrane by electroblotting. The membranes were then washed with TBST containing 5% BSA at room temperature and incubated for 2 h with the following antibodies: 1:1000 anti-ERK1/2 and anti-phospho-ERK1/2 (Cell Signaling, Beverly, MA, USA), 1:500 anti-GPR40 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and 1:5000 anti-β-actin (AbFrontier, Seoul, Korea), which was used as an internal control. The membranes were washed in TBST and incubated for 1 h with horseradish peroxidase-conjugated anti-mouse and anti-rabbit immunoglobulin antibodies (1:1000) under the same conditions. After washing with TBST, signals were detected using an enhanced chemiluminescence detection reagent (AbFrontier, Seoul, Korea), visualized, and measured by densitometry using a ChemiDoc XRS digital imaging system and the Quantity One software (ver. 4.4.1; Bio-Rad Laboratories, Hercules, CA, USA).

The RIN-5f and RIN-40 cells were seeded (10⁶/well) in six-well plates for 48 h and were made quiescent in serum-free medium for 24 h, followed by treating with 30 µM linoleic acid to RIN-40 for an additional 2 h. Total RNA was isolated using a HiYield Total RNA Mini Kit according to the manufacturer's protocol (RBCBioscience, New Taipei, Taiwan). The quality of total RNA was confirmed by running samples on an Agilent 2100 Bioanalyzer (Agilent Technology, CA, USA). RIN (RNA Integrity Number, cut-off threshold: 7.0) values were 9.6 for RIN-5f and RIN-40, and 9.7 for RIN-40 treated linoleic acid. Agilent's Rat Oligo Microarray (44K) analysis was carried out at eBiogen Inc. (Seoul, Korea).

The sample quality control was based on the Pearson correlation of a sample with other samples in the whole experiment. Hybridized images were scanned using an Agilent DNA microarray scanner and quantified with Feature Extraction Software (Agilent Technology, Palo Alto, CA, USA). Probe features were divided with 'present,' 'marginal,' and 'absent' flags, and the present and marginal flags were retained for further analyses. The genes flagged as 'present' and 'marginal' were selected as reliable for further analysis. All data normalization and selection of fold-changed genes were performed using GeneSpringGX 7.3 (Agilent Technology, USA). To assess functional relationships between genes, DAVID (http://david.abcc.ncifcrf.gov/) as a web-based singular enrichment analysis (SEA), was used, providing mainly annotation and gene ontology (GO) term enrichment analysis to highlight the most relevant GO terms associated with a given gene list. The enrichment p value was calculated for each term from the list of genes of interest with a single-linkage method. Then, enriched terms were listed in a simple linear text format. In the DAVID annotation system, the EASE score and a modified Fisher's exact test between 'in pathway' and 'not in pathway' were used to measure gene enrichment in terms of annotation. Due to the redundant nature of annotations, a Functional Annotation Chart was used to present similar/relevant annotations repeatedly. The grouping algorithm is based on the hypothesis that similar annotations should have similar gene members. Functional Annotation Clustering integrates the same techniques as kappa statistics to measure the degree of common genes between two annotations, and fuzzy heuristic clustering (as used in the Gene Functional Classification Tool) to classify the groups of similar annotations according to kappa values. In this sense, the greater the number of common genes shared by annotations, the greater the likelihood they will be grouped together. The Group Enrichment Score, the geometric mean of members' p-values in a corresponding annotation cluster, is used to rank their biological significance. Thus, the members of the top-ranked annotation groups most likely have consistently lower p-values. To determine whether certain functional categories were over-represented in the gene lists determined in the microarray experiments, reliable genes were first filtered using the 'present' and 'marginal' flags and re-filtered according to being 1.5-, 2-, and 3-fold up- and downregulated. The filtered gene lists were applied to DAVID and annotated functional pathways in GO and KEGG, which can identify the locations of genes in related pathways. Annotated pathways were displayed with statistical significance values in terms of enrichment scores, nominal p value, and FDR (p<0.05). For discrete genes in certain pathway we identified biological functions in the KEGG pathways and verified their functions using GenBank and literature searches using PubMed (http://www.ncbi.nlm.nih.gov/).

RIN-5f and RIN-40 cells were seeded (10⁶/well) in six-well plates for 48 h, and exposed to various reagents. Total RNA was isolated using a HiYield Total RNA Mini Kit according to the manufacturer's protocol (RBCBioscience, New Taipei, Taiwan). First-strand complementary DNA (cDNA) was synthesized from 1 µg of total RNA in a 20-µL reaction volume using the AccuPower CycleScript RT PreMix, as recommended by the manufacturer (Bioneer, Daejeon, Korea). The cDNA synthesis thermal cycling program included the following three steps: 37℃ for 1 min, 47℃ for 3 min, 55℃ for 1 min, then 95℃ for 5 min. Using the PCR mixture, initial DNA polymerase activation was carried out at 95℃ for 5 min, followed by 35 cycles of 95℃ for 10 s, 60℃ for 10 s, and 72℃ for 10 min. A final extension was carried out at 72℃ for 5 min. The PCR products were visualized by ethidium bromide staining after separation by electrophoresis in a 2% agarose gel in Tris borate ethylenediamine tetra-acetic acid (TBE) buffer (pH 8.3). Images were captured using a Fluor-S Max MultiImaging system and band densities were assessed with the Quantity One 1-D software (ver. 4.6.0). The reaction mixture for real-time quantitative reverse transcription PCR (qRT-PCR) contained cDNA, 1× SYBR green Taq polymerase mixture (Toyobo, Osaka, Japan), and primers, and was performed using an Exicylcler 96 instrument (Bioneer, Daejeon, Korea). Relative gene expression levels were compared using an invariant endogenous control (β-actin). The ΔΔCT method was used for relative quantification according to the manufacturer's guidelines. The sequences of the primers used for PCR amplification are shown in Table 1.

The DAVID web-based tool was used to extract the major biological features in the large gene lists. Statistical significance was expressed as nominal p values and the corrected p value though FDR. For the insulin secretion test, statistical analyses were conducted using a Kruskal-Wallis test, followed by Dunn's post hoc test or Mann-Whitney U-test using the Prism software (GraphPad Software, San Diego, CA, USA). p values<0.05 were considered to indicate statistical significance.

To evaluate the effects of overexpressing GPR40, we established RIN-40 from the original rat pancreatic β-cell line RIN-5f. Microscopic observation of RIN-40 cells revealed that they were morphologically similar in overall appearance to RIN-5f (data not shown). By RT-PCR, the RIN-40 cell line showed abundant expression of GPR40 mRNA (human GPR40), but little endogenous Gpr40 mRNA (rat GPR40; Fig. 1A). Neither GPR40 nor Gpr40 mRNA was detected in the RIN-5f cell line. As expected, the GPR40 protein level in RIN-40 was considerably greater than that in RIN-5f (Fig. 1B). To determine whether insulin secretion was mediated by RIN-40 activation, the dose-response relationship between linoleic acid as a GPR40 agonist and insulin secretion was measured. The basal level of glucose-stimulated insulin secretion in RIN-40 was about 50-fold greater than that in RIN-5f (Fig. 1C). As the concentration of linoleic acid increased, glucose-stimulated insulin secretion increased significantly, in a dose-dependent manner. The maximal efficacy was at 30 µM linoleic acid with a ca. threefold increase (from 50.00±7.46 to 135.10±9.20 ng/ml/protein mg), compared with the basal level in RIN-40 (p<0.001). To confirm that the linoleic acid-induced action was mediated through Gα q-coupled GPR40 signaling, intracellular Ca²⁺ mobilization was measured in linoleic-acid-treated RIN-5f and RIN-40 cells. GPR40 activation induced by 30 µM linoleic acid caused an increase in intracellular Ca²⁺ mobilization in RIN-40 cells, to a level ca. fivefold greater than that in RIN-5f cells (Fig. 1D). On the basis of these observations, RIN-40, a stable GPR40-overexpressing cell line, was established to be suitable for investigation of the function of GPR40, with the exception of the effects of endogenous Gpr40, and was used for further experiments.

Of the 30,367 probe sets on the Agilent Rat Oligo Microarray (44K), the 'reliable' signals were selected to characterize the gene expression profiles of RIN-5f and RIN-40 cells. A microarray analysis of the RIN-40 cell line using GeneSpringGX 7.3, showed that 1004 genes in GO biological pathways were expressed differentially by at least twofold compared with in RIN-5f cells (Table 2). Of the differentially expressed genes identified, 475 were upregulated, and 529 were downregulated. Genes related to signal transduction and cell differentiation represented >30% of the total. Upregulated and downregulated genes showed similar distributions in each biological pathway. Genes associated with signal transduction pathways whose expression changed by at least threefold are presented in Table 2. To identify the signaling pathways in which genes were involved, the genes applied on DAVID with at least a 1.5fold change were matched to specific KEGG pathways and classified by cellular function with statistical values, based on frequencies of genes in each pathway (Table 3). In the various biological pathways affected by overexpression of GPR40, many genes were related to chemokines (p_corrected=0.028), mitogen-activated protein (MAP) kinase (p_corrected=0.030), maturity onset diabetes of the young (p_corrected=0.031), mTOR (p_corrected=0.033), and the dilated cardiomyopathy signaling pathway (p_corrected= 0.049).

After checking signal intensity quality with the microarray using the RIN-40 response to 30 µM linoleic acid, 93 genes were found to be differentially expressed by more than twofold. Of these, 15 (22.7%) upregulated and 7 (25.9%) downregulated genes were involved in signal transduction (Table 2). The total number of genes whose expression was altered by linoleic acid (93) was fewer than that in the comparison between RIN-5f and RIN-40 cells. Whereas ~59% of genes related to cell differentiation, cell proliferation, signal transduction, and transcription were upregulated, ~63% of those related to the cell cycle, cell differentiation, cell proliferation, and signal transduction were downregulated by linoleic acid treatment. Table 4 shows representative genes related to signal transduction whose expression changed by at least twofold. In an analysis of involvement by pathway, the gene sets annotated included olfactory transduction (p=8.9×10^-23), neuroactive ligand-receptor interaction (p=0.002), MAP kinase signaling pathway (p=0.014), cytokine-cytokine receptor interactions (p=0.024), and regulation of the actin cytoskeleton (p=0.036). To confirm the microarray data, three upregulated genes [arrestin domain containing 3 (Arrdc3), early growth response 1 (Egr1), and serum/glucocorticoid-regulated kinase 1 (Sgk1)], and two downregulated genes [cAMP-specific phosphodiesterase 4B (Pde4b) and thrombospondin 1 (Thbs1)] were selected and qRT-PCR was performed using RNA obtained independently of that used in the microarray. The mRNA levels of all selected genes in the qRT-PCR experiment showed patterns similar to those obtained by microarray analysis (Table 5).

To assess the involvement of ERK signaling in linoleic-acid-induced insulin secretion, the phosphorylation of ERK according to duration of incubation in the presence of linoleic acid and insulin secretion under GPR40 activation induced by linoleic acid treatment were measured. The data in Fig. 2A show that ERK phosphorylation was increased rapidly by treatment with 30 µM linoleic acid for 3 min, followed by a decrease, in RIN-40 cells. ERK activation coincided with the increase in insulin secretion in RIN-40 cells, which did not occur in cells transfected with siRNA for GPR40 (siGPR40; Fig. 2B). Thus, these results indicate that the increase in insulin secretion induced by linoleic acid may be associated with ERK activation.