GPRC5B Activates Obesity-Associated Inflammatory Signaling in Adipocytes

Sci. Signal., 20 November 2012
Vol. 5, Issue 251, p. ra85
DOI: 10.1126/scisignal.2003149

GPRC5B Activates Obesity-Associated Inflammatory Signaling in Adipocytes

  1. Yeon-Jeong Kim,
  2. Takamitsu Sano,
  3. Takuji Nabetani,
  4. Yoshimi Asano, and
  5. Yoshio Hirabayashi*
  1. Laboratory for Molecular Membrane Neuroscience, RIKEN Brain Science Institute, Hirosawa 2-1, Wako-shi, Saitama 351-0198, Japan.
  1. *To whom correspondence should be addressed. E-mail: hirabaya{at}


A genome-wide association study identified a strong correlation between body mass index and the presence of a 21-kb copy number variation upstream of the human GPRC5B gene; however, the functional role of GPRC5B in obesity remains unknown. We report that GPRC5B-deficient mice were protected from diet-induced obesity and insulin resistance because of reduced inflammation in their white adipose tissue. GPRC5B is a lipid raft–associated transmembrane protein that contains multiple phosphorylated residues in its carboxyl terminus. Phosphorylation of GPRC5B by the tyrosine kinase Fyn and the subsequent direct interaction with Fyn through the Fyn Src homology 2 (SH2) domain were critical for the initiation and progression of inflammatory signaling in adipose tissue. We demonstrated that a GPRC5B mutant lacking the direct binding site for Fyn failed to activate a positive feedback loop of nuclear factor κB–inhibitor of κB kinase ε signaling. These findings suggest that GPRC5B may be a major node in adipose signaling systems linking diet-induced obesity to type 2 diabetes and may open new avenues for therapeutic approaches to diabetic progression.


Obesity increases the risk of disrupting nutrient homeostasis and developing insulin resistance, which leads to type 2 diabetes (1). During the development of type 2 diabetes, adipocyte dysfunction accompanied by chronic inflammatory signaling plays a crucial role in energy homeostasis and insulin resistance (2). Enlarged adipocytes produce large amounts of the chemokine monocyte chemoattractant protein-1 (MCP1), which promotes macrophage infiltration into adipose tissue in obese subjects (36). These macrophages secrete proinflammatory cytokines, including tumor necrosis factor–α (TNF-α), interleukin-6 (IL-6), and IL-1β, which ultimately lead to defects in insulin signaling and systemic insulin resistance (710).

Adipocytes are not only lipid storage depots, but they are also secretory cells that produce inflammatory cytokines and adipokines. Although how low-grade inflammation in adipose tissue is induced by overnutrition remains unknown, inhibitor of κB (IκB) kinase ε (IKKε) is required for high-fat diet (HFD)–induced obesity and adipose inflammation (11). IKKε deficiency protects against diet-induced adipose inflammation and insulin resistance (11). The gene encoding IKKε is transcriptionally induced by nuclear factor κB (NF-κB) downstream of inflammatory stimuli. In turn, IKKε activates NF-κB and phosphorylates IκB (12). Therefore, this positive feedback loop may contribute to the adipose inflammation caused by overnutrition.

GPRC5B is an orphan heterotrimeric guanine nucleotide–binding protein (G protein)–coupled receptor (GPCR) in the class C family, which includes metabotropic glutamate receptors, γ-aminobutyric acid B receptors, and taste receptors (13). GPRC5B was first cloned as a retinoic acid–induced gene. This protein potentially mediates the cellular effects of retinoic acid on G protein signaling. To evaluate its potential physiological roles, we generated GPRC5B-LacZ knock-in mice and found that GPRC5B is robustly expressed in both adipose tissues and the central nervous system (14).

Although the molecular function of GPRC5B is completely unknown to date, a genome-wide association study identified a strong correlation between body mass index (BMI) and the human GPRC5B gene (15). The 21-kb copy number variation (CNV) that lies 50 kb upstream of GPRC5B is associated with BMI. The identification of this CNV was based on the tagging of single-nucleotide polymorphism (SNP) rs12444979 in a large meta-analysis for population-based samples, whereas the deletion allele of the CNV is tagged by the non-risk allele of the SNP. This result suggests the potential involvement of GPRC5B in obesity. Here, we report that GPRC5B enhanced the tyrosine kinase activity of the Src family kinase (SFK) member Fyn through an interaction with the Src homology 2 (SH2) domain of Fyn. This interaction led to activation of the NF-κB signaling axis mediated by IKKε in adipose tissue during the progression of diet-induced obesity in mice.


GPRC5B is a lipid raft–associated phosphoprotein in the plasma membrane

To examine the molecular function of GPRC5B, we purified GPRC5B-containing protein complexes from human embryonic kidney (HEK) 293 cells transfected with plasmid encoding GPRC5B by sequential-step purification with affinity chromatography and size-exclusion chromatography. We found that GPRC5B interacted mainly with caveolin-1 and rarely with flotillins (Fig. 1, A and B). These protein components are major scaffolding proteins in plasma membrane microdomains, which are considered to play a pivotal role in regulating various signaling pathways, including insulin signaling (1618). We demonstrated that GPRC5B colocalized with caveolin-1, flotillins, and Fyn in 3T3-L1 adipocytes and fibroblasts (Fig. 1C and fig. S1). We also found that GPRC5B was enriched in detergent-resistant membrane fractions (Fig. 1D). These observations suggest that the molecular function of GPRC5B is closely associated with signaling events that occur in lipid rafts.

Fig. 1

Biochemical features of GPRC5B. (A) The components of GPRC5B-containing complexes were confirmed by Western blotting analysis of fractions after size-exclusion chromatography. Data represent two independent experiments. 5B-Flag, Flag-tagged GPRC5B; Cav1, caveolin-1; Flot1, flotillin 1. (B) GPRC5B-containing protein complexes from GPRC5B-overexpressing HEK 293 cells were directly immunoprecipitated (IP) with Flag-agarose and analyzed by Western blotting (WB) with antibodies specific for the indicated proteins. Data represent three independent experiments. (C) Single-plane confocal images indicated that GPRC5B localized on membrane microdomains in 3T3-L1 adipocytes. 3T3-L1 adipocytes were transfected by electroporation with expression plasmids for the indicated proteins. After 24 hours, the cells were fixed and observed under a confocal laser microscope. Lipid droplets are labeled with lipidTOX neutral lipid stain. Images represent three independent experiments. (D) GPRC5B was fractionated into detergent-resistant membrane (DRM) fractions. The transferrin receptor was used as a marker of detergent-soluble membranes (DSM). The indicated fractions were analyzed by Western blotting with antibodies specific for the indicated proteins. Data represent three independent experiments. (E) An in vitro kinase assay revealed that Fyn phosphorylated GPRC5B. Data represent three independent experiments. (F) Fyn is the kinase responsible for the phosphorylation of GPRC5B in cells. HEK 293 cells were cotransfected with expression plasmids encoding GPRC5B-Flag and Fyn or its mutants. After 24 hours, the cells were stimulated with 1 mM H2O2. Fyn-KD, a kinase-defective mutant (Y528F); Fyn-CA, a constitutively active mutant (K299M). Data represent three independent experiments.

We next demonstrated that GPRC5B was a phosphoprotein that contained multiple phosphorylated sites. Mass spectrometric (MS) analysis showed that mouse GPRC5B contained three evolutionarily conserved phosphotyrosine residues at Tyr307, Tyr330, and Tyr383 (fig. S2, C to E), as well as eight phosphoserine and two phosphothreonine residues in its C-terminal cytoplasmic tail (fig. S2F). When cells were treated with pervanadate, a protein tyrosine phosphatase inhibitor, Western blotting analysis with polyclonal antibodies specific for phosphorylated tyrosines revealed an increase in the extent of the tyrosine phosphorylation of GPRC5B compared to that in untreated cells (figs. S2B and S3).

Because we had found trace amounts of Fyn in the GPRC5B-containing complexes, we next examined whether the tyrosine kinase Fyn directly phosphorylated GPRC5B. An in vitro kinase assay revealed that recombinant Fyn directly phosphorylated the C-terminal fragment of GPRC5B (Fig. 1E). Oxidative stress or hydrogen peroxide (H2O2) activates SFKs, including Fyn (19, 20). We demonstrated that H2O2-induced tyrosine phosphorylation of GPRC5B increased within 5 min of stimulation (fig. S4A). Pretreatment with PP2, a specific SFK inhibitor, blocked this phosphorylation in a dose-dependent manner (fig. S4B). Moreover, we detected the H2O2-induced enhanced phosphorylation of GPRC5B in cells transfected with a plasmid encoding a constitutively active mutant of Fyn (Fig. 1F). In contrast, in cells transfected with plasmid encoding a kinase-defective mutant of Fyn, tyrosine phosphorylation of GPRC5B was completely abolished as a consequence of the dominant-negative effect of the Fyn mutant (Fig. 1F). These findings suggest that Fyn is the kinase responsible for the tyrosine phosphorylation of GPRC5B.

GPRC5B stimulates the kinase activity of Fyn

Similar to other SFKs, the catalytic activity of Fyn is regulated by inter- and intramolecular interactions through SH2 or SH3 domains (21). Therefore, we next examined whether GPRC5B regulated the kinase activity of Fyn. Because Fyn is an essential kinase for the phosphorylation of caveolin-1 (19), we measured the kinase activity of Fyn by assessing the extent of tyrosine phosphorylation of caveolin-1 (pTyr14) in cells stimulated with H2O2. We found that caveolin-1 phosphorylation in GPRC5B−/− mouse embryonic fibroblasts (MEFs) was reduced compared with that in wild-type cells (Fig. 2A). We next examined which domain of Fyn interacted with GPRC5B protein in glutathione S-transferase (GST) pulldown assays. We found that GPRC5B specifically interacted with the SH2 domain of Fyn in a phosphorylation-dependent manner (Fig. 2B and fig. S5A). In contrast, the SH3 domain of Fyn failed to interact with GPRC5B (Fig. 2B and fig. S5B), even though the C terminus of GPRC5B contains two putative proline-rich motifs (fig. S5C). Synthetic peptides corresponding to the C terminus of GPRC5B interacted with GST-SH2 peptide, suggesting that Fyn and GPRC5B directly interacted through this domain (fig. S5).

Fig. 2

GPRC5B stimulates the kinase activity of Fyn. (A) Time-course analysis of H2O2-stimulated phosphorylation of caveolin-1 (pY14). Caveolin-1 phosphorylation was decreased in GPRC5B−/− MEFs compared to that in wild-type (WT) MEFs [n = 4 experiments; analysis of variance (ANOVA): *P < 0.05; ***P < 0.001]. (B) GST pulldown (PD) assay. In pervanadate (PV)–treated cells, pGPRC5B specifically interacted with the SH2 domain of Fyn. Data represent three independent experiments. (C) GST pulldown assay. Phosphorylation of pTyr383 was critical for the binding of GPRC5B to the Fyn-SH2 domain (n = 6 Western blots; ANOVA: *P < 0.05; **P < 0.01; ***P < 0.001). (D) Fyn-dependent phosphorylation of caveolin-1 (pY14) was rescued in GPRC5B−/− MEFs stably expressing WT GPRC5B (5B-WT) but not in MEFs expressing mutant GPRC5B (5B-Y383F). n = 4 Western blots; **P < 0.01; ***P < 0.001.

We next determined which phosphorylated sites were associated with the interaction between the Fyn-SH2 domain and GPRC5B. We observed an ~75% reduction in the interaction between GST-SH2 and the Y383F mutant form of GPRC5B, as well as an ~25% reduction in the interaction between GST-SH2 and the Y307F and Y330F mutants of GPRC5B (Fig. 2C). This finding indicates that pTyr383 is a critical residue in the interaction between GPRC5B and the SH2 domain of Fyn. In addition, H2O2-induced phosphorylation of caveolin-1 was increased in cells transfected with plasmid encoding wild-type GPRC5B but not in cells transfected with plasmid encoding the Y383F mutant GPRC5B (Fig. 2D and fig. S6). These findings suggest that the molecular interaction between GPRC5B and Fyn was critical in enhancing the kinase activity of Fyn. Thus, GPRC5B stimulated Fyn through the interaction between its phosphorylated C terminus and the SH2 domain of Fyn.

GPRC5B-deficient mice exhibit metabolic phenotypes

We previously found that the Drosophila ortholog of GPRC5B, BOSS, is closely associated with energy homeostasis and insulin signaling (22). To address the biological role of GPRC5B in vivo, we characterized the metabolic phenotypes of GPRC5B-deficient mice. When fed an HFD, GPRC5B−/− mice weighed 18% less than did GPRC5B+/+ mice at 16 weeks of age (Fig. 3, A and B). Moreover, the adipose tissue and liver of GPRC5B+/+ mice were substantially heavier than those of GPRC5B−/− mice under HFD conditions (fig. S7). We observed enlarged adipocytes in GPRC5B+/+ mice but not in GPRC5B−/− mice (Fig. 3, C to E). We also observed apparent hepatic steatosis (fatty liver) in GPRC5B+/+ mice fed an HFD (fig. S8); however, GPRC5B mRNA and protein in liver were below the limit of detection (fig. S9).

Fig. 3

GPRC5B deficiency confers mice with an obesity-resistant phenotype. (A) GPRC5B+/+ and GPRC5B−/− mice at 16 weeks of age after being fed an HFD. (B) Changes in body weight under the feeding conditions. Under HFD conditions, GPRC5B−/− mice weighed less than WT mice (n = 8 mice of each genotype; ANOVA: P < 0.05). (C) Representative images of hematoxylin and eosin–stained epididymal WAT from mice kept under HFD conditions. Scale bars, 100 μm. Images represent five mice of each genotype. (D and E) Frequency distribution (D) and box-and-whisker plot (E) of adipocyte diameters measured in about 300 adipocytes from two mice selected independently (Student’s t test: ***P < 0.001). (F) Food intake per mouse measured over 1 week under HFD conditions (GPRC5B+/+, n = 11 mice; GPRC5B−/−, n = 8 mice; Student’s t test). (G) Volume of O2 consumption (VO2) measured over 24 hours (n = 5 mice of each genotype; ANOVA: *P < 0.05). (H) Calculated respiratory quotient (RQ) values were similar across the genotypes (n = 5 mice of each genotype; Student’s t test).

GPRC5B mRNA and protein are most abundant in the brain (11, 23). The expression of GPRC5B was shown in the granular cell layer of the olfactory bulb, cerebral cortex, dentate gyrus, Purkinje cell layer, and hypothalamus (14). In open-field tests, GPRC5B−/− mice show a decrease in the total distance traveled and an increase in total center time compared with wild-type mice. These results indicate that GPRC5B−/− mice have an altered spontaneous activity pattern and a decreased response in a new or unfamiliar environment (14). However, these spontaneous activities do not explain why GPRC5B−/− mice have an obesity-resistant phenotype. In addition, daily food intake was not affected by GPRC5B deficiency.

We next examined the metabolic rate of the mice. Oxygen consumption in GPRC5B−/− mice was 15% greater than that in GPRC5B+/+ mice (Fig. 3G). In both strains of mice, the respiratory quotient, a measure of fuel-partitioning patterns, fluctuated ~0.8 under HFD conditions (Fig. 3H), indicating that fuel selection of carbohydrates and lipids did not differ between the mice of either genotype. We observed an increased abundance of uncoupling protein 1 (UCP1) in the brown adipose tissues of GPRC5B−/− mice compared to that of wild-type mice, which indicated that enhanced thermogenesis occurred in GPRC5B−/− mice (fig. S10). Together, these findings suggested that GPRC5B deficiency conferred resistance to diet-induced obesity caused by a high metabolic rate.

GPRC5B promotes adipose inflammation and insulin resistance in a Fyn-dependent manner

The positive feedback loop of the NF-κB–IKKε signaling axis in adipocytes was demonstrated as a key player linking obesity and adipose inflammation (11). Therefore, we focused on HFD-induced chronic inflammation in white adipose tissue (WAT), which might explain the molecular function of GPRC5B in defined knockout phenotypes. We observed reduced IKKε and TANK-binding kinase 1 (TBK1) abundance in GPRC5B−/− mice compared to that in GPRC5B+/+ mice (Fig. 4A). The production of inflammatory cytokines, including MCP1 and TNF-α, accompanied by macrophage infiltration in WAT is important for triggering systemic insulin resistance (310). The amounts of MCP1 and TNF-α mRNAs in WAT were consistently decreased in GPRC5B−/− mice compared to those in wild-type mice (Fig. 4B). The abundance of mRNA for F4/80, a macrophage marker, was also reduced in GPRC5B−/− mice compared to that in wild-type mice (Fig. 4B), as was the abundance of the mRNA of inducible nitric oxide synthase (iNOS) (Fig. 4B). These data suggest that GPRC5B deficiency reduces the extent of macrophage infiltration and inflammatory signaling in WAT during obesity progression.

Fig. 4

GPRC5B deficiency reduces inflammation in adipose tissue. (A) IKKε and TBK1 protein abundances in epididymal WAT of 16-week-old mice under HFD conditions (n = 7 mice; ANOVA: **P < 0.01; ***P < 0.001). (B) Quantitative reverse transcription polymerase chain reaction (RT-PCR) analysis showed that MCP1, TNF-α, F4/80, and iNOS mRNA abundances in WAT were reduced in GPRC5B−/− mice compared to those in WT mice (n = 7 mice of each genotype; Student’s t test: **P < 0.01; ***P < 0.001). (C) Tyrosine phosphorylation of GPRC5B in WAT from mice under different feeding conditions. Increased phosphorylation of GPRC5B was observed in WAT under HFD conditions (HFD: n = 6 mice; chow: n = 8 mice; Student’s t test: **P < 0.01). (D) IKKε protein abundance in SVF-derived adipocytes. TNF-α–induced IKKε protein was decreased in GPRC5B−/− and Fyn−/− adipocytes compared to that in WT adipocytes (n = 4 Western blots; ANOVA: ***P < 0.001). (E) Circulating glucose concentrations during the GTT (left) and ITT (right) in mice (n = 9 mice of each genotype; ANOVA: *P < 0.05; **P < 0.01).

The dynamics of GPRC5B phosphorylation may be important for elucidating the signaling network as well as for understanding the functional link between GPRC5B and obesity-associated insulin resistance. We found that tyrosine-phosphorylated GPRC5B protein in WAT was increased in abundance under HFD conditions compared to that in mice fed normal chow (Fig. 4C). These data suggest that the tyrosine phosphorylation of GPRC5B is closely linked with dysfunction in adipocytes. We also observed increased IKKε protein abundance in GPRC5B−/− MEFs as well as in 3T3-L1 adipocytes transfected with plasmid encoding GPRC5B (fig. S11, A and C). The increased abundance of GPRC5B in 3T3-L1 adipocytes led to increased MCP1 in the culture medium (fig. S11B). Stable transfection of cells with plasmid encoding the Y383F mutant GPRC5B did not enhance the abundance of IKKε (fig. S11C). Inhibition of the kinase activity of Fyn with a dominant-negative Fyn mutant markedly reduced the abundance of IKKε in 3T3-L1 adipocytes (fig. S11D). Reciprocal activation of NF-κB was also observed in parallel experiments with an NF-κB reporter assay (fig. S11E). We confirmed that GPRC5B and Fyn affected the abundance of IKKε in TNF-α–treated adipocytes derived from stromal vascular cells (Fig. 4D). These data suggest that amplification of the kinase activity of Fyn by GPRC5B is critical for the positive feedback regulation of the IKKε–NF-κB signaling axis, which leads to cytokine production in adipocytes.

Glucose tolerance tests (GTTs) and insulin tolerance tests (ITTs) consistently revealed that GPRC5B−/− mice had a faster glucose clearance rate than did GPRC5B+/+ mice (Fig. 4E). Circulating insulin and glucose concentrations in GPRC5B−/− mice were statistically significantly lower than those in GPRC5B+/+ mice under HFD conditions (table S1). The amounts of adiponectin in the serum were not substantially different between the two mouse strains. Leptin concentrations, however, were reduced in GPRC5B−/− mice compared to those in wild-type mice (table S1). This was probably a result of the smaller size of adipocytes and increased leptin sensitivity in GPRC5B−/− mice. These data suggested that GPRC5B deficiency ameliorated diet-induced insulin resistance by reducing inflammatory signaling in WAT through the kinase activity of Fyn.


Here, we demonstrated that the interaction of GPRC5B and the Fyn-SH2 domain drove the local amplification of Fyn kinase activity. This interaction controlled the positive feedback loop of the IKKε–NF-κB signaling axis, leading to cytokine production, inflammation in adipose tissue, and insulin resistance during the progression of diet-induced obesity (fig. S12). GPRC5B protein is mainly found in the brain and adipose tissues; it is almost undetectable in skeletal muscle and liver (fig. S9B). Our previous study showed that GPRC5B in the brain is mostly found in Purkinje neurons and dentate gyrus granule cells (14). Our present analyses failed to reveal any histological abnormalities in the brains of GPRC5B−/− mice. Moreover, feeding behaviors and physical activity related to obesity in GPRC5B−/− mice were not altered compared to those in wild-type mice (Fig. 3F). These findings suggest that the brain does not play a major role in insulin resistance. Although GPRC5B was absent from the liver, we observed substantial lipid accumulation in the liver in wild-type mice under HFD conditions (fig. S8). A study of mice with an adipose-specific conditional knockout of c-Jun N-terminal kinase 1 (JNK1) suggests that an inflammatory response and stress in adipocytes triggers hepatic steatosis (24). Thus, GPRC5B activation in adipose tissue may cause lipid accumulation in the liver under HFD conditions.

Macrophage infiltration may be an essential step for dysfunction and inflammatory signaling in adipocytes during obesity progression (6). In addition to obesity, increased production of MCP1 in adipocytes may promote macrophage infiltration. Similarly, macrophage infiltration is reduced in the adipose tissue of IKKε-knockout mice compared to that in wild-type mice (3). Here, we found that overexpression of GPRC5B in 3T3-L1 adipocytes increased the amount of MCP1 secreted (fig. S11B). On the other hand, macrophage infiltration was decreased in the adipose tissue of GPRC5B−/− mice (Fig. 4B). These results correlated with the GPRC5B-induced increase in IKKε abundance and NF-κB activation in adipocytes.

GPRC5B−/− mice were leaner than wild-type mice (Fig. 3, A and B). This decreased adiposity may have affected insulin sensitivity in vivo; however, we demonstrated that GPRC5B directly activated IKKε–NF-κB signaling, which was dependent on the kinase activity of Fyn in vitro. This suggests that GPRC5B is a crucial factor for inflammation in adipose tissue. Links between metabolic regulation and Fyn kinase activity were suggested by experiments with Fyn-knockout mice that showed a lean phenotype and improved glucose tolerance resulting from insulin sensitivity in peripheral tissues (25, 26). Fyn-deficient mice have decreased body weight even under normal chow-feeding conditions. In contrast, the body weights of GPRC5B−/− mice and wild-type mice differed only when the mice were fed an HFD (Fig. 3B). This indicates that GPRC5B−/− mice have a modest phenotype compared to that of Fyn-deficient mice. Given that the function of GPRC5B is to regulate Fyn kinase activity, the differences in the metabolic phenotypes between GPRC5B- and Fyn-deficient mice may be understandable.

It is not apparent whether the phenotypes of GPRC5B−/− mice are related to a deficit in specific agonist-induced G protein–dependent signaling. In particular, GPRC5B has an unusually short extracellular N terminus—corresponding to the agonist-binding domain—compared to other GPCRs in the class C family. Even in comparison with the Drosophila ortholog BOSS (22, 27), the N terminus of GPRC5B is quite short (fig. S13). This indicates that the ligand-binding function of GPRC5B may have been deleted during evolution. However, signaling events through the molecular interactions of GPRC5B are similar to those of transmembrane adaptor proteins, including linker for activation of T cells (LAT), phosphoprotein associated with glycosphingolipid-enriched microdomains (PAG), non–T cell activation linker (NTAL, also known as LAT2), and Lck-interacting membrane protein (LIME), which contribute to the regulation of SFKs in lipid rafts during immune cell development (28). We demonstrated that GPRC5B played a pivotal role in the local amplification of Fyn-mediated tyrosine phosphorylation on membrane rafts. Fyn is usually activated by growth factors, including insulin, and cellular stress, which is crucially important for obesity and the inflammatory process in adipose tissue. Diet-induced obesity and insulin resistance progressively developed and were accompanied by insulin-induced negative feedback regulation, adipose inflammation, and cellular stress in adipocytes. During this process, GPRC5B may contribute to the enhanced activation of localized Fyn, leading to the promotion of inflammatory signaling mediated by IKKε.

Materials and Methods


All mouse experiments were performed according to procedures approved by the Animal Experiments Committee at RIKEN. GPRC5B was inactivated by targeted disruption in exon 2, as described previously (14). Mice were housed in an animal facility with 12-hour light:12-hour dark cycles and at a constant temperature of 23 ± 2°C. The animals were fed standard chow (CRF-1, Oriental Yeast) or an HFD (High Fat Diet 32, CLEA Japan Inc.) consisting of 25.5% (w/w) protein, 29.4% carbohydrates, and 32% fat. The mice were fed an HFD from the age of 4 weeks and had access to water ad libitum. Oxygen consumption was measured for 24 hours with an O2-CO2 metabolism measuring system (MK-5000RQ, Muromachi). GTT was performed by an intraperitoneal injection of glucose (2 g/kg body weight) after an overnight fast. ITT was performed by an intraperitoneal injection of insulin (0.75 U/kg body weight; Eli Lilly) after a 4-hour fast. Blood glucose was measured in tail-vein blood samples at the designated times with an ACCU-CHEK system (Roche). Hormone concentrations in serum were measured by enzyme-linked immunosorbent assay according to the manufacturer’s instructions. Cholesterol, triglyceride, and free fatty acid concentrations were measured with appropriate quantification kits.


Polyclonal antisera against GPRC5B were raised in New Zealand rabbits with a synthetic peptide corresponding to the mouse GPRC5B C-terminal sequence (amino acid residues 394 to 410). The antibody was affinity-purified with the antigen peptide. For the phosphosite-specific antibodies, polyclonal antisera were raised against the following phosphorylated peptides: pY307 (ENPPNpYFDTSQ), pY330 (HLPRApYMENK), and pY383 (FRSNVpYQPTEM). Nonspecific antibodies were removed, and antibodies specific for phosphorylated residues were affinity-purified with the corresponding phosphopeptides. The following commercial antibodies were used: anti–caveolin-1, anti–p-caveolin-1 (pY14), anti-actin, anti-IKKε, and anti-TBK1 (Cell Signaling Technology); anti-Fyn and anti-pFyn (pY528) (BD Biosciences); anti-pSrc (pY418) and anti–transferrin receptor (Invitrogen); anti–Flag M2 (Sigma-Aldrich); anti-flotillin (Santa Cruz Biotechnology); and anti-GST (Novagen). Tyrosine phosphorylation was detected with anti-phosphotyrosine antibodies (PY20 or 4G10, Millipore).

Cell culture and transfection

HEK 293 cells, 3T3-L1 preadipocytes, and MEFs were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. All culture incubations were performed in a humidified incubator at 37°C and 5% CO2. MEFs were isolated from embryos on embryonic day 13.5. For immortalization, MEFs were passaged in serial cultures to generate cell lines, according to the 3T3 protocol. Adipocyte differentiation in 3T3-L1 or stromal vascular cells from epididymal fat pads was induced by treating the cells for 48 hours with 1 μM dexamethasone, 0.5 mM isobutylmethylxanthine, 5 μM troglitazone (Calbiochem), and insulin (10 μg/ml). The cells were then maintained in medium containing insulin (1 μg/ml). HEK 293 cells were transfected with the FuGENE HD transfection reagent (Roche) according to the manufacturer’s instructions. MEFs or 3T3-L1 adipocytes were transfected by electroporation under a preoptimized square pulse condition (1400 V, 20 ms, 1 pulse) with the NEON transfection system (Invitrogen). Stable transfectants were polyclonally selected with hygromycin (100 μg/ml) with the Flp-In expression system (Invitrogen).

Confocal microscopy

Colocalization analysis was performed on differentiated 3T3-L1 or MEFs transfected with plasmid encoding fluorescently tagged proteins. After 24 to 48 hours of transfection, cells were fixed in 3% paraformaldehyde for 15 min and then were washed with phosphate-buffered saline. Lipid droplets were labeled with lipidTOX neutral lipid stain (Invitrogen) according to the manufacturer’s protocol. Cells were mounted in Immu-Mount reagent (Thermo Scientific) and then observed under a confocal laser-scanning microscope (FV1000, Olympus).

Purification of GPRC5B-containing protein complexes

Protein complexes containing GPRC5B were purified from HEK 293 cells stably expressing Flag-tagged GPRC5B with Flag-affinity chromatography followed by size-exclusion chromatography. The protein extract prepared in lysis buffer [20 mM tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 10% glycerol, 1% NP-40] supplemented with a protease inhibitor cocktail (Roche) was subjected to affinity chromatography with Flag M2–agarose (Sigma-Aldrich). The protein complexes were eluted with 3× Flag peptide (Sigma-Aldrich) and subsequently purified by size-exclusion chromatography. The protein components in each size-exclusion chromatography fraction were identified by tandem MS.

MS analysis

Samples were reduced with dithiothreitol (DTT), alkylated with iodoacetamide, and subsequently digested in-solution with trypsin (Promega) or endoproteinase Glu-C (Roche). Liquid chromatography–tandem MS was performed on an LTQ mass spectrometer (Thermo Fisher Scientific) connected to a Paradigm MS2 high-performance liquid chromatography system (Michrom Bioresources). Solvent A was 0.1% formic acid in 2% acetonitrile, and solvent B was 0.1% formic acid in 98% acetonitrile. The components of the GPRC5B complexes were determined with Mascot v2.2.1 software. To identify phosphorylated sites in GPRC5B, we directly subjected the samples to phosphorylated peptide (phosphopeptide) enrichment methods, as described previously (29). The phosphopeptides were injected into an analytical column (L-Column Micro C18, 0.2 × 50 mm, CERI) at 2 μl/min and eluted with a linear gradient from 2 to 50% solvent B for 58 min.

Protein-protein interaction assay

Immunoprecipitations, GST pulldowns, and peptide-binding assays were performed to determine protein-protein interactions. Cell lysates were prepared in lysis buffer [20 mM tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% NP-40, 60 mM octylglucoside] supplemented with protease inhibitor and phosphatase inhibitor cocktails (Roche) with brief sonication. For immunoprecipitation, the lysates were incubated at 4°C for 2 hours with Flag M2–agarose or anti-GPRC5B antibody–coupled protein G magnetic beads (Dynabeads protein G, Invitrogen). For the GST pulldown assay, 1 μg of GST fusion protein–coupled glutathione-Sepharose beads (GE Healthcare) was incubated with 200 μg of cell lysates at 4°C for 1 hour. For the peptide binding assay, 50 pmol of biotinylated peptide coupled to Dynabeads M-280 streptavidin was incubated at 4°C for 1 hour with 2 μg of purified GST-SH2 protein. After incubation, the mixtures were extensively washed with lysis buffer. The proteins were then eluted with Laemmli sample buffer, resolved by SDS–polyacrylamide gel electrophoresis (SDS-PAGE), and detected by Western blotting or Coomassie staining.

In vitro kinase assay

An in vitro kinase assay was performed with the purified GST-fused C terminus of GPRC5B (amino acids 295 to 410, GST-5B-Ct) as the substrate and recombinant Fyn (Invitrogen) as the kinase. Aliquots of substrate protein (1 μg) were incubated at 30°C for 15 min with 0.1 μg of recombinant Fyn in kinase buffer [25 mM tris-HCl (pH 7.5), 5 mM β-glycerophosphate, 2 mM DTT, 0.1 mM Na3VO4, 10 mM MgCl2, 100 μM adenosine 5′-triphosphate]. The reaction mixture was mixed with Laemmli sample buffer and resolved by SDS-PAGE, and protein phosphorylation was detected by Western blotting.

Quantitative RT-PCR

Complementary DNA (cDNA) was synthesized with high-capacity cDNA reverse transcription kits (Applied Biosystems) from total RNA extracted with Trizol reagent (Invitrogen). Gene expression analysis of each tissue was performed with an ABI 7900HT system (Applied Biosystems). TaqMan PCR primers and probes were purchased from Applied Biosystems and used to determine the abundances of MCP1 (Mm00441242_m1), TNF-α (Mm00443258_m1), iNOS (Mm00440502_m1), and F4/80 (Mm00802530_m1) mRNAs. The relative amounts of these mRNAs were normalized to that of 36B4 mRNA (Mm00725448_s1).

Statistical analysis

The data sets were analyzed with ANOVA or Student’s t tests, where appropriate, with GraphPad Prism 5 software (GraphPad Software). All values are shown as means ± SEM.

Supplementary Materials

Fig. S1. Cellular localization of GPRC5B.

Fig. S2. GPRC5B contains multiple phosphorylated sites.

Fig. S3. Specificity of phosphorylation site–specific polyclonal antibodies for pGPRC5B.

Fig. S4. Tyrosine phosphorylation of GPRC5B in H2O2-stimulated cells.

Fig. S5. Interactions between the Fyn-SH2 domain and C-terminal phosphopeptides of GPRC5B.

Fig. S6. Increased GPRC5B abundance enhances the kinase activity of Fyn in HEK 293 cells.

Fig. S7. Tissue weights of mice fed an HFD.

Fig. S8. Representative images of liver sections from mice fed an HFD.

Fig. S9. Relative GPRC5B mRNA and protein abundance in mouse tissues.

Fig. S10. Enhanced UCP1 abundance in brown adipose tissue of GPRC5B−/− mice.

Fig. S11. Molecular interaction of GPRC5B and Fyn functionally leads to positive feedback regulation of the IKKε–NF-κB signaling axis.

Fig. S12. Functional role of GPRC5B in adipose inflammation.

Fig. S13. Sequence comparison of human GPRC5B and Drosophila BOSS.

Table S1. Serum parameters of mice in this study.

  • Received 18 April 2012.
  • Accepted 2 November 2012.

References and Notes

Acknowledgments: We thank E. Takahashi and K. Yamada for their critical comments on the animal behavioral analysis and T. Shimizu for support with animal care. We are grateful to the Support Unit for Bio-material Analysis, RIKEN BSI Research Resources Center, for help with the TaqMan PCR analyses, nucleotide sequencing, and chemical synthesis of peptides. We thank T. Yagi (Osaka University) for providing Fyn-knockout mice. We thank T. Kadowaki (The University of Tokyo) for critical reading of the manuscript. Funding: This work was supported by the RIKEN Brain Science Institute, the RIKEN President Discretionary Fund, the Core Research for Evolutional Science and Technology (CREST) of Japan Science and Technology Agency (JST) (2005–2010), and, in part, by the Naito Foundation Subsidy for Promotion of Specific Research Projects. Author contributions: Y.-J.K., T.S., and Y.H. designed the study and wrote the manuscript; T.S. was involved in generating the knockout mice and in animal care and performed the experiments determining the metabolic phenotypes of the mice; Y.-J.K. performed the experiments on the molecular mechanism and signaling pathways; T.N. performed the MS analyses; Y.A. contributed to the analysis of the specificity of antibodies and in the construction of the expression plasmids; and all authors discussed the results and commented on the manuscript. Competing interests: The authors declare that they have no competing interests.


Y.-J. Kim, T. Sano, T. Nabetani, Y. Asano, and Y. Hirabayashi, GPRC5B Activates Obesity-Associated Inflammatory Signaling in Adipocytes. Sci. Signal. 5, ra85 (2012).
Science Signaling. ISSN 1937-9145 (online), 1945-0877 (print). Pre-2008: Science's STKE. ISSN 1525-8882