Dc

KD

a

ARRAA

KGSLC

1

mpcaadntmi

tlcwi

o

0d

Biosensors and Bioelectronics 25 (2010) 1675–1680

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

journa l homepage: www.e lsev ier .com/ locate /b ios

etection of G protein-coupled receptor-mediated cellular response involved inytoskeletal rearrangement using surface plasmon resonance

exin Chen, Hideru Obinata, Takashi Izumi ∗

epartment of Biochemistry, Gunma University Graduate School of Medicine, 3-39-22 Showa-machi, Maebashi, Gunma 371-8511, Japan

r t i c l e i n f o

rticle history:eceived 29 September 2009eceived in revised form 2 December 2009ccepted 4 December 2009vailable online 16 December 2009

eywords:protein-coupled receptor

a b s t r a c t

G protein-coupled receptors (GPCRs) form a superfamily of cell surface receptors that play fundamentalroles in physiology and pathophysiology. Although GPCRs have been the most successful targets fordrug discovery, there still remain many orphan GPCRs, which provides opportunities for developmentof novel drugs. Here, we introduce a new method for evaluation of GPCR activation utilizing a surfaceplasmon resonance (SPR) sensor. Cells expressing GPCRs were cultured directly on an SPR sensor chipand stimulated with GPCR ligands, resulting in SPR responses that were dependent on the type of G alphasubunits coupling with receptors. Namely Gi- and/or G12/13-coupled receptors evoked SPR responsesbut G - or G -coupled ones did not. Analyses on the intracellular signal pathways revealed that small

urface plasmon resonance

iving cellytoskeletal rearrangement

s q

G protein Rho/Rac-mediated actin rearrangement plays an important role in the signal transductionpathways leading to the SPR responses. An SPR response was also evoked by insulin-like growth factor-1,which stimulates Rac-dependent stress fiber formation via its receptor-tyrosine kinase. Thus, this methodprovides a unique opportunity for real-time monitoring of cellular responses involved in cytoskeletalrearrangements, and may be useful in ligand/drug discovery for certain types of receptor, such as Gi- andG12/13-coupled receptors.

© 2009 Elsevier B.V. All rights reserved.

. Introduction

G protein-coupled receptors (GPCRs), a superfamily of trans-embrane receptors, play fundamental roles in physiology and

athophysiology by mediating a wide variety of biological pro-esses in response to various agonists, including photons, odorants,mines, peptides, proteins, nucleotides and lipids (Fredriksson etl., 2003). GPCRs are potential targets for clinical therapeutics andrug discovery. At present, about 100 GPCRs, whose ligands haveot been identified, are called orphan GPCRs. De-orphaning ofhese GPCRs is expected to lead to the discovery of novel bioactive

olecules and better understandings of physiological or patholog-cal processes.

Heterotrimeric G proteins act as molecular switches thatransduce conformational changes in GPCRs to activate intracel-

ular effectors. Alpha subunits of heterotrimeric G protein arelassified into subfamilies based on intracellular signaling path-ays: Gs, Gi, Gq and G12/13. Gs activates adenylyl cyclase and

ncreases intracellular cAMP level, whereas Gi inhibits adenylyl

∗ Corresponding author. Tel.: +81 27 2207940; fax: +81 27 2207948.E-mail addresses: chenkexi@health.gunma-u.ac.jp (K. Chen),

binata@med.gunma-u.ac.jp (H. Obinata), takizumi@med.gunma-u.ac.jp (T. Izumi).

956-5663/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2009.12.006

cyclases. Gq activates phospholipase C and increases intracellu-lar calcium concentration. G12/13 activates Rho guanine nucleotideexchange factor (RhoGEF), and induces the formation of actinstress fibers (Offermanns, 2003). Ligand screening of orphanGPCRs has been performed based on changes of intracellu-lar cAMP or calcium concentration and receptor internalization.However, there still remain many orphan GPCRs, which encour-aged us to develop new screening methods based on differentprinciples.

Surface plasmon resonance (SPR) is an optical sensing techniquethat is based on the phenomenon of attenuated total reflection andhas been developed for real-time analysis of molecule–moleculeinteractions. It has been reported that SPR is also able to detectcellular responses in antigen-stimulated mast cells (Hide et al.,2002) and in EGF-stimulated keratinocytes (Yanase et al., 2007).These SPR responses were assumed to reflect changes in overallcytoskeletal rearrangements, though the underlying mechanismsare unknown.

In this study, we determined whether an SPR sensor could be

utilized to detect cellular responses mediated by GPCRs in livingcells directly cultured on the sensor chip. We also analyzed intracel-lular signaling pathways leading to the SPR responses. Our resultsindicate that the SPR responses reflect actin cytoskeletal rearrange-ments downstream of Gi and/or G12/13.

1 Bioele

2

2

gKScPS22aW

2

atrv

2

HtSieeuwa

2

swta

Fig. S1 (in the Supporting Information). Clear SPR responses were

TS

Sap

676 K. Chen et al. / Biosensors and

. Materials and methods

.1. Materials

Nutrient mixture F-12 HAM, fetal bovine serum, Trizma base,lucagon, lysophosphatidic acid (LPA), cytochalasin D (cyto D),i16425 and Alexa Fluor 546 phalloidin were purchased fromigma–Aldrich Japan. Sphingosine-1-phophate (S1P) was pur-hased from Biomol; insulin-like growth factor-1 (IGF-1) fromeproTech; JTE013 from Tocris; bovine serum albumin (BSA) fromerologicals; cAMP-Screen system from Applied Biosystems; Fura-AM from Dojindo; F127 from Molecular Probes; Lipofectamine000, opti-MEM I medium and neomycin from Invitrogen; andnti-G13 antibody from Santa Cruz. All other chemicals were fromako.

.2. Plasmid DNA for expression of GPCR

Coding sequences of muscarinic acetylcholine receptor (M1, M2nd M3), dopamine receptor (D1 and D2), leukotrienes B4 recep-or (BLT1), platelet-activating factor receptor (PAFR) and glucagoneceptor (GCGR) were amplified by PCR and ligated with pcDNA3ector (Invitrogen).

.3. Cell preparation

Chinese hamster ovary (CHO)-K1 cells were maintained inam’s F-12 medium containing 10% fetal bovine serum. Cells were

ransfected with plasmid DNA using Lipofectamine 2000 reagent.table CHO cell clones that express GPCRs were established by lim-ting dilution and selection with 1 mg/ml G418 (Invitrogen). Thexpression of GPCRs was confirmed by RT-PCR, and by ligand-voked changes in intracellular calcium and/or cAMP concentrationsing established protocols (Yokomizo et al., 1997). Activated Racas detected using an EZ-DetectTM Rac1 Activation Kit (PIERCE)

ccording to the manufacturer’s instructions.

.4. SPR assay

SPR assays were performed using a dual-channel SPR670-MACSystem (Moritex) in which a semiconductor laser acts as a sourceith a fixed wavelength of 670 nm in conjunction with a pho-

odiode sensor to measure reflection light. This system providessensitivity of 0.001◦/s. SPR sensor chips (Moritex, gold-layered

able 1ummary of SPR responses in CHO cells.

Receptor Ligand tested G protein

M1 Carbachol Gq

M2 Carbachol Gi

M3 Carbachol Gq

D1 Dopamine Gs

D2 Dopamine Gi

BLT1 Leukotriene B4 Gi, Gq

PAFR PAF Gi, Gq

GCCR Glucagon Gs

LPA1 LPA Gi, Gq, G12/13

S1P2 S1P Gi, Gq, G12/13

IGF-1R IGF-1 Not applicable

PR responses in CHO cells are summarized with the results of calcium and cAMP assanalyzed after the application with each ligand for 2 min. The responses of endogenous rlatelet-activating factor; LPA, lysophosphatidic acid; S1P, sphingosine-1-phophate; IGF-a Mean ± S.D. of EC50 value from triplicate measurements.b No response up to 300 nM.c Inhibitory effect on the 10 �M forskolin-induced cAMP accumulation.d No response up to 100 nM.e Mean ± S.D. of EC50 value from quadruplicate measurements.

ctronics 25 (2010) 1675–1680

chip, 13 mm × 20 mm × 0.7 mm) were sequentially washed withacetone, 70% ethanol and phosphate-buffered saline (PBS) at roomtemperature under sterilized conditions. Cells were directly seededonto the SPR sensor chips without any chemical treatment onthe surface of the chips, then placed in 35 mm culture dishes ata density of 1 × 106 cells/dish in Ham’s F-12 medium containing10% fetal calf serum and cultured overnight. After incubation inHEPES–Tyrode’s–BSA buffer (HTB buffer, 25 mM HEPES–NaOH, pH7.4, 140 mM NaCl, 2.7 mM KCl, 1.0 mM CaCl2, 12 mM NaHCO3,5.6 mM d-glucose, 0.37 mM NaH2PO4, 0.49 mM MgCl2 and 0.1%fatty acid-free BSA) for 30 min, the cells were mounted on theSPR system and further washed with HTB buffer at a flow rateof 30 �l/min through the flow channels until the baseline signalbecame stable. SPR measurements were performed using two flowchannels; ligands for GPCRs dissolved in HTB buffer were addedto one channel (the sensing channel) and the corresponding vehi-cle to the other (the reference channel). SPR responses are shownas maximum changes in resonance angle in the sensing channelminus those in the reference channel unless otherwise noted. Usu-ally, the cells on the SPR sensor chip were able to be stimulatedseveral times at intervals of approximately 30 min.

2.5. Measurement of intracellular calcium concentration

CHO cells were loaded with 5 �M Fura 2-AM in HTB buffer con-taining 1.25 mM probenecid and 0.02% Pluronic F127 for 1 h at37 ◦C, and washed with HTB buffer. Changes in intracellular cal-cium concentrations upon ligand stimulation were monitored witha FLEX-station scanning fluorometer system (Molecular Devices).

3. Results

3.1. SPR responses in CHO-K1 cells expressing exogenous GPCRs

We examined ligand-induced SPR responses in CHO cells sta-bly expressing each GPCR. The results are summarized in Table 1.The typical SPR responses in these CHO cells are shown in

observed in M2 and D2 receptors (Gi-coupled), and PAFR and BLT1(Gq- and Gi-coupled), whereas no SPR response was observed inother Gq- or Gs-coupled GPCRs. These results suggest that SPRresponses can be evoked by signals mediated via Gi-coupled recep-tors.

Ca2+ cAMP SPR

300 ± 20 nMa Not done No responseb

Not done 3 ± 2 �Ma,c 5 ± 2 �Ma

100 ± 20 nMa Not done No responseb

Not done 3 ± 2 nMa No responsed

20 ± 10 nMa 30 ± 10 nMa,c 30 ± 10 nMa

5 ± 1 nMe Not done 0.3 ± 0.2 nMe

0.04 ± 0.02 nMe Not done 0.1 ± 0.1 nMe

Not done 10 ± 5 nMe No responsed

No responsed No responsec,d 5 ± 2 nMa

No responsed No responsec,d 4 ± 2 nMa

Not done Not done 7 ± 3 nMa

ys. The responses of CHO cells stably expressing exogenous human GPCRs wereeceptors (LPA1, S1P2 and IGF-1R) expressing in CHO cells were also analyzed. PAF,1, insulin-like growth factor-1.

K. Chen et al. / Biosensors and Bioele

Fig. 1. LTB4-induced responses in CHO-BLT1 cells. (A) CHO-BLT1 cells were stim-ulated with increasing concentrations of LTB4 for 2 min, as indicated by solid bar.Typical SPR charts of the sensing channel are shown. (B) SPR responses to variousconcentrations of LTB4. The EC50 value for the SPR response was about 0.3 nM. (C)Lcm

3

mc(tTwcto

TB4-induced intracellular calcium mobilization was measured. The EC50 value foralcium response was about 5 nM. Data are the mean ± S.D. of triplicate measure-ents.

.2. The sensitivity of the SPR assay

We compared the EC50 in the SPR assay with those in the calciumobilization and cAMP accumulation assays. In BLT1-expressing

ells, SPR responses were evoked in a dose-dependent mannerFig. 1A). The EC50 for the SPR assay was about 10-fold lower thanhat in the calcium mobilization assay for BLT1 (Fig. 1B and C,

able 1). Also, for M2, D2 and PAFR, SPR responses were observedith ligand concentrations at around the EC50 in the calcium or

AMP assays (Table 1). These results indicate that the sensitivity ofhe SPR assay is comparable to or even higher than that of calciumr cAMP assays.

ctronics 25 (2010) 1675–1680 1677

3.3. Involvement of actin cytoskeletal rearrangement inGPCR-mediated SPR response

We analyzed intracellular signaling pathways leading to SPRresponses via BLT1. First, effects of pertussis toxin (PTX), a spe-cific inhibitor of Gi, were investigated. As shown in Fig. 2A,LTB4-induced SPR response was completely blocked by PTX pre-treatment, indicating the importance of Gi-mediated signaling forthe SPR response. The pre-treatment with PTX did not affect cellu-lar machinery of calcium mobilization, judging from ATP-inducedintracellular calcium increase (Fig. 2A, inset).

We next examined effects of cyto D, an inhibitor of actin poly-merization, on LTB4-induced SPR responses. Cyto D pre-treatmentmarkedly inhibited the SPR response (Fig. 2C), but did not affectcalcium mobilization (data not shown). When CHO-K1 cells onthe sensor chip were treated with cyto D via the flow channel,a decrease in resonance angle was observed (Fig. 2D). The reso-nance angle returned to the basal level when cyto D was removed.These results indicate that actin cytoskeletal rearrangement playsan important role in the SPR responses.

We confirmed LTB4-induced stress fiber formation by confocalmicroscopy. LTB4 induced a rapid formation of stress fibers within3 min that was inhibited by PTX or cyto D pre-treatment (Fig. S2A).We attempted to confirm LTB4-induced Rac activation in our assaysystem because it has been reported that Rac plays an importantrole in stress fiber formation (Hall, 1998) and that LTB4 inducesRac activation in Rat-2 fibroblasts cells (Woo et al., 2002). Rac wasrapidly activated by LTB4 in BLT1-CHO cells, and the activation wasinhibited by PTX (Fig. 2B). Taken together, these results suggestthat the LTB4-induced actin rearrangement involving Rac activationparticipated in the SPR response.

3.4. Contribution of G12/13-mediated signals to SPR response

To examine whether G12/13-coupled receptor can also evoke SPRresponses, we investigated the SPR responses evoked by LPA andSIP. Among several receptor subtypes, LPA receptor 1 (LPA1) andS1P receptor 2 (S1P2) were mainly expressed in CHO-K1 cells, asrevealed by RT-PCR analysis (data not shown). Both LPA1 and S1P2have been reported to couple with Gq, Gi and G12/13 (Moolenaar,1999; Sugimoto et al., 2003).

When CHO-K1 cells were stimulated with LPA or S1P, clear SPRresponses were obtained with an EC50 in the nanomolar range(Table 1). The SPR responses were significantly reduced in the pres-ence of receptor-specific antagonists (Fig. 3A). LPA- or S1P-evokedresponses were not detected in the calcium mobilization or thecAMP accumulation assays (Table 1). These results indicate thatthe sensitivity of SPR assay is high enough to detect signals fromendogenously expressed receptors that are difficult to detect byconventional assay systems.

Next, we examined whether G12/13-mediated cytoskeletal rear-rangement was involved in the SPR responses. When CHO-K1 cellswere transfected with the C3 exoenzyme that selectively ADP-ribosylates the Rho subfamily, the LPA-evoked SPR signal wasdecreased by nearly half (Fig. 3B, left). The effect of PTX pre-treatment was also partial and co-treatment with C3 exoenzymeand PTX almost completely inhibited the response (Fig. 3B, left).These results indicate that the LPA-mediated SPR response wascomposed of both Gi- and G12/13-derived signals. On the otherhand, C3 exoenzyme significantly reduced the S1P2-mediated SPRresponse but PTX showed little effect (Fig. 3B, right), suggesting

that G12/13-derived signals mainly contributed to inducing the SPRresponse to S1P2.

The expression of p115RhoGEF-RGS, a blocker of G12/13 sig-naling (Grabocka and Wedegaertner, 2005), partially suppressedLPA-induced SPR response (Fig. 3C). C3 exoenzyme markedly

1678 K. Chen et al. / Biosensors and Bioelectronics 25 (2010) 1675–1680

Fig. 2. SPR response caused by actin rearrangement through the Gi- and Rac- pathways in CHO-BLT1 cells. (A) CHO-BLT1 cells were pre-treated with 100 ng/ml PTX for 3 h,and the maximum SPR responses with 1 nM LTB4 were measured. Results of intracellular calcium mobilization with 10 �M ATP are shown in the inset. (B) CHO-BLT1 cellsw rated ut ree ind1 an + Sc t’s t te

itR

3

mtltfeiair((mIt(e

ere stimulated with 1 nM LTB4 for 1 min. GTP-Rac in the total cell lysates was sepao Western blot analysis using the anti-Rac antibody. Data are means + S.D. from thh, and the maximum SPR responses with 1 nM LTB4 were measured. Data are meyto D, indicated by the solid bar, was monitored in CHO-K1 cells. *P < 0.05 (Studen

nhibited LPA-induced stress fiber formation (Fig. S2B). Takenogether, G12/13 also contributes to the SPR response by inducingho activation and subsequent actin rearrangement.

.5. IGF-1-mediated SPR response

Because the SPR sensor was able to sensitively detect GPCR-ediated cytoskeletal rearrangement in living cells, we assumed

hat the SPR sensor might be utilized to detect various intracel-ular signals that affect cytoskeletal rearrangement. To examinehis idea, we measured the SPR response via insulin-like growthactor-1 receptor (IGF-1R), a receptor-tyrosine kinase, which is alsondogenously expressed in CHO-K1 cells and is reported to potentlynduce Rac activation and lamellipodium formation (Sugimoto etl., 2003). IGF-1 evoked a clear SPR response with EC50 at 7 nMn CHO-BLT1 cells (Table 1). In contrast to LTB4, IGF-1-inducedesponse was PTX-insensitive. Because phosphoinositide 3-kinasePI3K) is involved in lamellipodium formation upstream of RacWeiss-Haljiti et al., 2004), we next examined the effects of wort-

annin, a potent PI3K inhibitor, on IGF-1-induced SPR response.GF-1-induced SPR response was totally eliminated by wortmanninreatment, whereas LTB4-induced SPR response was not affecteddata not shown). IGF-1-induced stress fiber formation was alsoliminated by wortmannin but not by PTX (Fig. S2C). These results

sing an EZ-DetectTM Rac1 Activation Kit. Total Rac and the GTP-Rac were subjectedependent experiments. (C) CHO-BLT1 cells were pre-treated with 2 �M cyto D for

.D. from four independent experiments. (D) Change in resonance angle with 2 �Mst).

indicate that SPR could also detect signals from receptor-tyrosinekinases in cases where the activation of the receptor leads to actinrearrangement.

4. Discussion

SPR is one of the most powerful tools for label-free analy-sis of molecular interactions. In the case of GPCR investigations,SPR provides the possibility for monitoring ligand–receptor andreceptor–effector interactions. Direct analysis of a GPCR-agonistinteraction by SPR has been successfully reported using detergent-solubilized recombinant GPCR (neurotensin receptor-1) as ananalyte in a configuration where biotinylated agonist (neurotensin)was immobilized on the sensor chip (Harding et al., 2007). However,solubilization of GPCR is generally difficult and time-consuming.Moreover, SPR is unsuitable for analyses in which the molecularweight of the analyte is extremely small compared to that of themolecules immobilized on the sensor chip, as in the case of GPCRand its ligand. These difficulties have hampered the application of

SPR for ligand or agonist/antagonist screening of GPCRs.

In this study, we succeeded in monitoring GPCR-mediated cel-lular responses in CHO-K1 cells. We found that CHO-K1 cells canadhere to and proliferate on a gold-coated sensor chip withoutchemical modification of chip surface and there was no significant

K. Chen et al. / Biosensors and Bioelectronics 25 (2010) 1675–1680 1679

F esponc for SP PR asst

dTiv

ctiwm

tGcBvisGct

ig. 3. SPR responses induced by LPA or S1P through intrinsic GPCRs. (A–C) SPR ro-stimulated with specific antagonists (100 �M Ki16425 for LPA1 or 10 �M JTE013TX for 3 h before SPR assay. (C) Cells were transfected with p115RhoGEF before Sest).

etachment of the cells during SPR monitoring for several hours.hus, the system enabled us to treat cells as needed before and dur-ng measurement, including transfection, pre-/co-treatment witharious reagents and multiple applications of analytes.

The GPCR-induced SPR responses were markedly reduced byyto D (Fig. 2). Small G protein Rac and/or Rho were involved inhe responses (Figs. 2 and 3). Furthermore, the time course andntensity of ligand-induced stress fiber formation coincided well

ith those of SPR responses. These results suggest that the GPCR-ediated actin rearrangement participates in the SPR responses.Observed GPCR-mediated SPR responses were dependent on

he type of G alpha subunit coupling with the receptor; Gi- and/or12/13-coupled receptors evoked SPR responses, while Gs- or Gq-oupled ones did not (Table 1 and Fig. S1). Among these GPCRs, onlyLT1 showed higher sensitivity of the SPR assay than those of con-entional assays. Both Gi-and Gq-coupled with BLT1 are involved

n LTB4-induced calcium mobilization (Yokomizo et al., 1997). Wepeculate that BLT1 is more efficiently coupled with Gi than withq in CHO-BLT1 cells. The fact that LTB4-induced SPR response wasompletely inhibited by PTX-treatment (Fig. 2A) is compatible withhis idea. LTB4 is a potent chemoattractant in leukocytes, and also

ses were measured in CHO-K1 cells with 10 nM LPA or 10 nM S1P. (A) Cells were1P2). (B) Cells were transfected with C3 exoenzyme and/or treated with 100 ng/mlay. Data are mean + S.D. from four independent experiments. *P < 0.05 (Student’s t

induces chemotaxis of CHO-BLT1 cells in a PTX-sensitive manner(Yokomizo et al., 1997). Chemotaxis requires a number of distinctcytoskeletal rearrangements such as localized actin polymerizationand disassembly of the cortical myosin thick filament network. PAF,LPA and S1P are also known as chemoattractants, and all of theminduced SPR responses. Thus, the SPR could be useful for detectingcytoskeletal rearrangement induced by chemoattractants.

The range over which SPR sensor can detect density changesis, in principle, determined by the evanescent field (≤670 nm)but practically, SPR sensors are thought to detect density changeswithin around 300 nm of the sensor chip surface (Sjölander andUrbaniczky, 1991). On the other hand, the plasma membrane hasa thickness of several nanometers and is structurally supported bycytoskeleton. The cellular cortex is an actin-rich layer just beneaththe plasma membrane, responsible for maintaining cell shape andmotility. It can be rapidly rearranged in response to various stimuli

via the polymerization–depolymerization cycle of actin (Pesen andHoh, 2005), which leads to density changes in the cellular cortex.We assume that GPCR-mediated SPR responses result from den-sity changes of the cellular cortex by actin polymerization. Theschematic diagram of the signaling cascades through cell surface

1680 K. Chen et al. / Biosensors and Bioele

Fig. 4. Schematic diagram of the signaling cascades through cell surface receptorsleading to SPR response. When cell surface receptors such as BLT1, S1P2 and IGF-1Rare stimulated by their ligands, intracellular signaling pathways are evoked via Gi

protein, G12/13 protein and PI3K, respectively. Then, small G proteins such as Rac andRdSc

ria

rsifui

ho are activated, followed by formation of F-actin in the cellular cortex, resulting inensity change in the evanescent field just above the surface of the SPR sensor chip.pecific inhibitors of signal molecules are also shown. PTX, pertussis toxin; Cyto D,ytochalasin D.

eceptors leading to SPR response is shown in Fig. 4. Our resultsndicate the possibility of utilizing cytoskeletal rearrangements asnovel indicator of cellular responses without any cell labelling.

In summary, we succeeded in monitoring GPCR-mediated SPResponses in living cells as a result of density changes on the sen-

or chip surface, probably due to actin rearrangement. Our systems therefore expected to comprise one of the most powerful toolsor screening of agonists or antagonists for GPCRs and is especiallyseful for G12/13-coupled receptors, for which an efficient screen-

ng method has not previously been developed. This system could

ctronics 25 (2010) 1675–1680

also be applicable for analysis of cellular responses involved incytoskeletal rearrangements, such as chemotaxis.

Acknowledgements

We would like to thank Dr. Shuh Narumiya for the gift of plas-mid DNA and Dr. Kohei Hosaka for useful suggestions. This researchwas funded by the G-COE program and Grants-in-Aid for scientificresearch from the Ministry of Education, Culture, Sports, Scienceand Technology of Japan (MEXT) and The Support Program forImproving Graduate School Education from MEXT to K.X C.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.bios.2009.12.006.

References

Fredriksson, R., Lagerstrom, M.C., Lundin, L.G., Schioth, H.B., 2003. Mol. Pharmacol.63, 1256–1272.

Grabocka, E., Wedegaertner, P.B., 2005. Oncogene 24, 2155–2165.Hall, A., 1998. Science 279, 509–514.Harding, P.J., Attrill, H., Ross, S., Koeppe, J.R., Kapanidis, A.N., Watts, A., 2007.

Biochem. Soc. Trans. 35, 760–763.Hide, M., Tsutsui, T., Sato, H., Nishimura, T., Morimoto, K., Yamamoto, S., Yoshizato,

K., 2002. Anal. Biochem. 302, 28–37.Moolenaar, W.H., 1999. Exp. Cell Res. 253, 230–238.Offermanns, S., 2003. Prog. Biophys. Mol. Biol. 83, 101–130.Pesen, D., Hoh, J.H., 2005. Biophys. J. 88 (1), 670–679.Sjölander, S., Urbaniczky, C., 1991. Anal. Chem. 63, 2338–2345.Sugimoto, N., Takuwa, N., Okamoto, H., Sakurada, S., Takuwa, Y., 2003. Mol. Cell Biol.

23, 1534–1545.Weiss-Haljiti, C., Pasquali, C., Ji, H., Gillieron, C., Chabert, C., Curchod, M.L., Hirsch, E.,

Ridley, A.J., van Huijsduijnen, R.H., Camps, M., Rommel, C., 2004. J. Biol. Chem.

279, 43273–43284.

Woo, C.H., You, H.J., Cho, Y.W., Eom, J.S., Chun, J.S., Yoo, Y.J., Kim, J.H., 2002. J. Biol.Chem. 277, 8572–8578.

Yanase, Y., Suzuki, H., Tsutsui, T., Hiragun, T., Kameyoshi, Y., Hide, M., 2007. Biosens.Bioelectron 22, 1081–1086.

Yokomizo, T., Izumi, T., Chang, K., Takuwa, Y., Shimizu, T., 1997. Nature 387, 620–624.