GST

D

Dte

AbbppnnGmGlaTrrpmUtorcncaeecmarol

K

Gmr

*EAFrtpuev

Neuroscience 137 (2006) 1347–1358

0d

ABAB HETERODIMERIC RECEPTORS PROMOTE Ca2� INFLUX VIA

TORE-OPERATED CHANNELS IN RAT CORTICAL NEURONS AND

RANSFECTED CHINESE HAMSTER OVARY CELLS

sGAt(senspan2tbt2ltt

G((dhpthcirarGtl(TeictWtac

. C. NEW, H. AN, N. Y. IP AND Y. H. WONG*

epartment of Biochemistry, the Molecular Neuroscience Center, andhe Biotechnology Research Institute, Hong Kong University of Sci-nce and Technology, Clearwater Bay, Hong Kong, China

bstract—The GABAB receptors are generally considered toe classical Gi-coupled receptors that lack the ability to mo-ilize intracellular Ca2� without the aid of promiscuous Groteins. Here, we report the ability of GABAB receptors toromote calcium influx into primary cultures of rat corticaleurons and transfected Chinese hamster ovary cells. Chi-ese hamster ovary cells were transfected with GABAB1(a) orABAB1(b) subunits along with GABAB2 subunits. In experi-ents using the fluorometric imaging plate reader platform,ABA and selective agonists promoted increases in intracel-

ular Ca2� levels in transfected Chinese hamster ovary cellsnd cortical neurons with the expected order of potency.hese effects were fully antagonized by selective GABABeceptor antagonists. To investigate the intracellular pathwaysesponsible for mediating these effects we employed severalharmacological inhibitors. Pertussis toxin abolished GABABediated Ca2� increases, as did the phospholipase C� inhibitor73122. Inhibitor 2-aminethoxydiphenyl borane acts as an an-

agonist at inositol 1,4,5-trisphosphate receptors and at store-perated channels. In all cell types, 2-aminethoxydiphenyl bo-ane prevented Ca2� mobilization. The selective store-operatedhannel inhibitor 1-[2-(4-methoxyphenyl)-2-[3-(4-methoxyphe-yl)propoxy]ethyl-1H-imidazole hydrochloride prevented in-reases in intracellular Ca2� levels as did performing thessays in Ca2� free buffers. In conclusion, GABAB receptorsxpressed in Chinese hamster ovary cells and endogenouslyxpressed in rat cortical neurons promote Ca2� entry into theell via the activation of store-operated channels, using aechanism that is dependent on Gi/o heterotrimeric proteins

nd phospholipase C�. These findings suggest that the neu-onal effects mediated by GABAB receptors may, in part, relyn the receptor’s ability to promote Ca2� influx. © 2005 Pub-

ished by Elsevier Ltd on behalf of IBRO.

ey words: GPCR, FLIPR, calcium, influx, SOC, GABA.

ABA is the main inhibitory neurotransmitter in the mam-alian CNS. Its effects are exerted through cell surface

eceptors, which are distributed throughout the nervous

Corresponding author. Tel: �2358-7328; fax: �2358-1552.-mail address: boyung@ust.hk (Y. H. Wong).bbreviations: CHO, Chinese hamster ovary; FBS, fetal bovine serum;LIPR, fluorometric imaging plate reader; GPCR, G protein-coupledeceptor; HBSS, Hanks’ balanced salt solution; IP3, inositol 1,4,5-risphosphate; PKA, protein kinase A; PKC, protein kinase C; PLC�,hospholipase C�; PTX, pertussis toxin; RFU, relative fluorescencenits; SKF96365, 1-[2-(4-methoxyphenyl)-2-[3-(4-methoxyphenyl)propoxy]

athyl-1H-imidazole hydrochloride; SOC, store-operated channels; VGCC,oltage-gated calcium channels; 2-APB, 2-aminethoxydiphenyl borane.

306-4522/06$30.00�0.00 © 2005 Published by Elsevier Ltd on behalf of IBRO.oi:10.1016/j.neuroscience.2005.10.033

1347

ystem and are either ionotropic (termed GABAA andABAC) or metabotropic (GABAB; Bowery et al., 1987).ttempts to clone GABAB receptors identified three pro-

eins with the topology of G protein-coupled receptorsGPCRs), GABAB1(a) and GABAB1(b) as well as a GABAB2ubunit (Jones et al., 1998; Kaupmann et al., 1998; Whitet al., 1998). Expression of any of these subunits alone didot generate a functional receptor. However, co-expres-ion of a GABAB1 subunit and GABAB2 resulted in theroduction of a functional GABAB receptor with affinitiesnd effector coupling efficiencies equivalent to endoge-ous GABAB receptors (Jones et al., 1998; Restituito et al.,005). A great deal of subsequent research has revealedhat the heterodimeric GABAB receptor is bound togethery a coiled-coil interaction between the carboxy terminalails of a GABAB1 and a GABAB2 subunit (Calver et al.,002). Only the GABAB1 subunit is able to bind extracel-

ular ligands and agonist activation transmits conforma-ional changes to the GABAB2 subunit, which is then ableo functionally interact with G proteins (Pin et al., 2004).

Functional GABAB receptors predominantly interact withproteins of the Gi/o family but not with Gs or Gq proteins

Odagaki and Koyama, 2001) and, indeed, pertussis toxinPTX) is effective at inhibiting many GABAB receptor-me-iated effects (Bowery et al., 2002). GABAB receptorsave been described to couple to several intracellularathways in neuronal cells endogenously expressinghese GPCRs. For example GABAB receptor agonists in-ibit the forskolin-stimulated accumulation of cAMP in ratortical brain slices (Knight and Bowery, 1996), while an

ncrease in cAMP production has also been reported in theat frontal cortex upon the application of GABAB receptorgonists, which is presumably mediated by G�� subuniteleased from Gi/o proteins (Onali and Olianas, 2001).ABAB receptors are also known to modulate changes in

he membrane K� flux of neuronal cells as well as regu-ating Ca2� flux through voltage-gated calcium channelsVGCCs; Parramon et al., 1995; Bowery et al., 2002).here are scattered reports of GABAB receptor-mediatedffects that are PTX insensitive (Bowery et al., 2002) but it

s not yet clear whether GABAB receptors are able toouple to Gz, a PTX insensitive member of the Gi/o familyhat is extensively expressed in neuronal cells (Ho and

ong, 2001), or to other unidentified G proteins and adap-or molecules (Calver et al., 2002). Furthermore, there isn isolated report that GABAB receptors promote the ac-umulation of inositol 1,4,5-trisphosphate (IP3) in bovine

drenal chromaffin cells (Parramon et al., 1995).

cerrtumtp2ifC(td

M

CCn(fvUDaNRUMaCwe

G

C3tUcGwicpZ3CnaGdrm1

P

Pc

tbsompsFbadir

W

CemsptGc

M

Tmcpdi1a

Md

Ca((mdtpFtitaiwU1acts(avaov

D. C. New et al. / Neuroscience 137 (2006) 1347–13581348

In an effort to broaden our understanding of the intra-ellular events activated by GABAB receptors, we havexamined their signal transduction mechanisms using bothecombinant and native cells. Unlike many Gi/o-coupledeceptors, GABAB receptors appear to be efficiently linkedo Ca2� mobilization. Intracellular Ca2� levels can be reg-lated by a variety of mechanisms (Verkhratsky, 2005),any of which are known to be activated by GPCRs, and

he mobilization of Ca2� is a key regulator of numeroushysiological and pathophysiological events (Verkhratsky,005). Here, we report the ability of GABAB heterodimers

n primary cultures of rat cortical neurons and in trans-ected Chinese hamster ovary (CHO) cells to promotea2� entry into the cell via store-operated channels

SOCs). GABAB receptors activate SOCs via a pathwayhat utilizes a PTX-sensitive and phospholipase C� (PLC�)-ependent mechanism.

EXPERIMENTAL PROCEDURES

aterials

HO-K1 cells were obtained from the American Type Cultureollection (ATCC CCL-61; Rockville, MD, USA). Cell and neuro-al culture reagents were purchased from Life TechnologiesGaithersburg, MD, USA). Ninety-six-well plates were obtainedrom Corning, Inc. (Corning, NY, USA). The bicistronic cloningector pBudCE4.1 was purchased from Invitrogen (Carlsbad, CA,SA). The GABAB1 and GABAB2 cDNAs were kindly provided byr. Fiona Marshall (GlaxoSmithKline, Stevenage, UK). GABAB1(b)nd GABAB2 subunits were subcloned into pBudCE4.1 usingotI/KpnI and HindIII/XbaI restriction enzyme sites, respectively.eceptor ligands and inhibitors were obtained from Tocris (Bristol,K) or Sigma-RBI (St. Louis, MO, USA). Fluo-4 AM was fromolecular Probes, Inc. (Eugene, OR, USA). Antibodies directedgainst GABAB1 and GABAB2 subunits were obtained fromhemicon International (Temecula, CA, USA). Other chemicalsere of analytical grade and purchased from commercial suppli-rs.

eneration of stable cell lines

HO cells were seeded into 25 cm2 culture flasks at a density of�105 cells per flask. On the following day, 2 h before transfec-ion, fresh F12 medium with 10% (v/v) fetal bovine serum (FBS), 50/ml penicillin and 50 �g/ml streptomycin was added to the cells. Toonstruct the GABAB1(b)/2/CHO cell line, 10 �g of GABAB1(b) andABAB2 cDNAs subcloned into the bicistronic pBUDCE4.1 vectoras introduced into the cells using the calcium phosphate precip-

tation method (Sambrook et al., 1989). For the GABAB1(a)/2/CHOell line, 10 �g of each of the GABAB1(a) cDNA subcloned intocDNA3.1 (�) and the GABAB2 cDNA subcloned into pcDNA3.1/eo (�) were used. Transfected cells were cultured for 16 h at7 °C in a water-saturated atmosphere of 95% air and 5% CO2.ells were washed with phosphate-buffered saline and cultured inormal growth medium for a further 24 h before selection by theddition of 400 �g/ml zeocin (GABAB1(b)/2/CHO) or 800 �g/ml418 and 400 �g/ml zeocin (GABAB1(a)/2/CHO). Following theeath of all mock-transfected cells, zeocin resistant cells weree-plated at a low density. Individual colonies were isolated andaintained in growth medium containing 200 �g/ml G418 and/or00 �g/ml zeocin, as appropriate.

reparation of cortical neuronal cultures

rimary cultures of cortical neurons were initiated from the corti-

es of 17–18 day old rat (Sprague–Dawley) embryos. Cortical f

issue was washed with HEPES buffered salt solution and incu-ated with 0.25% trypsin at 37 °C for 15 min. The neurons wereupplemented with 5% (v/v) horse serum followed by two roundsf centrifugation (1250 r.p.m., 5 min) and resuspension in DMEMedium. Cells were seeded in 96-well, black-walled microtiterlates at a density of 20,000 cells per well in DMEM mediumupplemented with 10% (v/v) horse serum and 1 mM L-glutamine.our hours after seeding the medium was replaced with neuro-asal medium supplemented with 2% (v/v) B27, 50 U/ml penicillinnd 50 �g/ml streptomycin. The medium was replaced after 3ays and cultured neuronal cells were assayed by fluorometric

maging plate reader (FLIPR) assay 9–11 days following prepa-ation.

estern blotting analysis

rude membrane proteins from stably transfected cell lines werextracted as previously described (Liu and Wong, 2005). Fiftyicrograms of membrane protein was resolved on a 12.5% (v/v)

odium dodecyl sulfate–polyacrylamide gel and transferred toolyvinylidene fluoride membrane. Protein expression was de-ected using antiserum recognizing the C-terminal tail regions ofABAB1 or GABAB2 subunits and an enhanced chemilumines-ence (ECL) kit.

easurement of cAMP levels

ransfected cells were labeled with [3H]adenine (1 �Ci/ml) in F12edium containing 1% (v/v) FBS for 20–24 h. Labeled cells were

hallenged with appropriate drugs for 30 min at 37 °C, in theresence of 50 �M forskolin, and assayed for cAMP levels, asescribed previously (Ho et al., 2002). When required, cells were

ncubated in the presence of 100 ng/ml of PTX for approximately6 h before forskolin and agonist challenge. EC50 values are givens the mean�S.D. from four independent experiments.

easurement of intracellular Ca2� using a FLIPRevice

HO cells were seeded into 96-well, black-walled microtiter platest a density of 4�104 cells per well in F12 medium containing 10%v/v) FBS. Where indicated, cells were incubated with PTX100 ng/ml) for 16 h prior to the assay. The following day, theedium was removed and replaced with 200 �l of labeling me-ium consisting of 1:1 (v/v) Opti-MEM:Hanks’ balanced salt solu-ion (HBSS), 2.5% (v/v) FBS, 20 mM HEPES, pH 7.4, 2.5 mMrobenecid and 2 �M Fluo-4 AM and assayed using an optimizedLIPR protocol (New and Wong, 2004). Pharmacological inhibi-

ors were added at appropriate time points before the assay wherendicated. Agonists and antagonists were prepared as a 5� solu-ion in HBSS, 20 mM HEPES, pH 7.4 and 2.5 mM probenecid andliquotted into polypropylene 96-well plates. Following a 60 min

ncubation of the cells in the labeling medium, cell and drug platesere placed in a FLIPR (Molecular Devices, Sunnyvale, CA,SA). Changes in fluorescence were monitored over a period of20 s following excitation at a wavelength of 488 nm and detectiont 510–560 nm. Fifty microliters of drug solution were added to theell medium at time�10 s. When required, inhibitors were addedo the labeling medium at the indicated times before the drugolutions. Data were collected as relative fluorescence unitsRFU), which denotes the fluorescent signal obtained over anrbitrarily set baseline. For each treatment, the response to theehicle was subtracted from the drug-induced response. Gener-lly, agonist-induced responses were several thousand RFUsver basal. Data were analyzed using Excel and GraphPad Prism,ersion 3.02. EC and IC values are given as the mean�S.D.

50 50rom no less than three independent determinations.

Ge

IrettrswGcw(ssGp

wsatpb

vbfluiFoifstwai

Go

PtGChtataC

FctsflaGcbakswa

FGrnppp

D. C. New et al. / Neuroscience 137 (2006) 1347–1358 1349

RESULTS

eneration and characterization of CHO cells stablyxpressing GABAB1(b)/2 subunits

n order to examine the signaling properties of the GABABeceptor, we have established a CHO cell line stably co-xpressing the GABAB1(b) and GABAB2 receptors using

he bicistronic expression vector pBudCE4.1. Zeocin resis-ant clonal cells were isolated and tested for their ability toespond to GABAB receptor agonists by inhibiting the for-kolin-stimulated intracellular cAMP levels. Several clonesere identified that responded to 500 �M of the selectiveABAB agonist (RS)-baclofen (to be referred to as ba-

lofen) by inhibiting cAMP accumulation up to 50–60%,hereas parental CHO cells did not respond to baclofen

data not shown). Membrane preparations of the best re-ponding clone were subjected to Western blotting analy-is using antisera directed against C-terminal regions ofABA or GABA subunits. Antisera against GABA

CHO

CHO

GABA

B1(b

)/2/C

HO

GABA

B1(b

)/2/C

HO

Antisera: GABAB1 GABAB2

100 kDa120 kDa

A

B

EC50 = 3.7 2.3 M(n=4)

No PTX

With PTX

025

50

75

100

-9 -8 -7 -6 -5 -4 -3

% F

ors

kolin

Res

po

nse

Log [Baclofen] (M)

ig. 1. Characterization of GABAB subunits stably expressed in CHOells. (A) GABAB1(b)/2/CHO cells were harvested and membrane pro-eins were prepared. Fifty micrograms of the membrane proteins wereeparated on a 12% SDS-PAGE gel and transferred to polyvinylideneuoride membrane by electroblotting. Membranes were probed withntibodies raised against the C-terminal region of either GABAB1 orABAB2 subunits. Fluorographs were visualized with an enhanced

hemiluminescence detection kit. (B) GABAB1(b)/2/CHO cells were la-eled with [3H]adenine in the absence (�) or presence (�) of PTXnd assayed for cAMP accumulation in the presence of 50 �M fors-olin and increasing concentrations of baclofen. [3H]cAMP was mea-ured as a fraction of the total tritiated adenine nucleotides per assayell. The normalized data are presented as the means�S.D. of dataveraged across four independent experiments performed in triplicate.

B1 B2 B1

roteins identified a band of approximately 100 kDa thatst

as not expressed in parental CHO cells (Fig. 1A). Anti-era against GABAB2 subunits labeled a diffuse band atpproximately 128 kDa, whose expression was not de-ected in CHO cells (Fig. 1A). This is in agreement withrevious data identifying GABAB subunit expression in ratrain and transfected HEK-293 cells (4).

The ability of these GABAB1(b)/2/CHO cells to acti-ate intracellular Gi/o-coupled pathways was assessedy the ability of the GABAB agonist baclofen to inhibit theorskolin-stimulated production of cAMP. When chal-enged with baclofen, the cAMP levels in forskolin-stim-lated GABAB1(b)/2/CHO cells were reduced by approx-

mately 65% with an EC50 value of 3.7�2.3 �M (n�4;ig. 1B). This is in good agreement with an EC50 valuef 3.7 �M reported for GABAB receptor-mediated cAMP

nhibition in HEK-293 cells and an EC50 value of 59 �Mor incorporation of GTP�S into rat brain membranestimulated by GABA (White et al., 1998). Furthermore,he inhibition of cAMP production was entirely abolishedhen cells were preincubated with PTX prior to theddition of 1 mM baclofen (Fig. 1B), suggesting the

nvolvement of Gi/o proteins.

ABAB receptors mediate the mobilizationf intracellular Ca2�

revious reports have demonstrated that GABAB recep-ors coexpressed with chimeric G proteins allows these

i/o-coupled receptors to promote intracellular increases ina2� levels (Wood et al., 2000; Liu et al., 2003). As weave previously observed that several Gi/o-coupled recep-ors can promote Ca2� mobilization in the absence ofrtificial chimeras (New and Wong, 2004), we examinedhe ability of GABAB receptors expressed in CHO cells toctivate pathways that trigger increases in intracellulara2� levels. GABAB1(b)/2/CHO cells were challenged with

00

20

40

60

80

100

120 Baclofen

-9 -8 -7 -6 -5 -4 -3

GABASKF97541CHO

Log [Drug] (M)

% M

axim

alR

esp

on

se

ig. 2. Ca2� mobilization in GABAB1(b)/2/CHO cells challenged withABAB receptor agonists. FLIPR assays were used to construct dose-

esponse curves of GABAB1(b)/2/CHO cell activation by three ago-ists, baclofen, GABA and SKF97541. The normalized data areresented as the means�S.D. of data averaged across three inde-endent experiments performed in triplicate. The response of thearental CHO cell line to baclofen was also examined and re-

ponses normalized to the maximal baclofen-induced response inhe stable cell line.

bGsCCddTrc

nsa3tf(r1bt13aC(so

IC

HdePblrb

ampI4csh

Fooaopparwct

Fbbr

D. C. New et al. / Neuroscience 137 (2006) 1347–13581350

aclofen and the Ca2� levels monitored in FLIPR assays.ABAB1(b)/2/CHO cells, but not parental CHO cells, re-

ponded to baclofen with large increases in intracellulara2� with EC50 values in the micromolar range (Fig. 2).a2� mobilization was apparent within 2–3 s of drug ad-ition, reaching a maximum value approximately 20 s afterrug addition before returning to basal levels within 50 s.o confirm that the Ca2� mobilization was mediated by GABAB

eceptors, we tested the response of the GABA /CHO

A

B

C

-6 -5 -4 -3 -2-1

0

1

2

3

4

Log [CGP 46381] (M)

Lo

g [

dr-

1]

0

0

20

40

60

80

100

CGP 55845

CGP 46381

-11 -10 -9 -8 -7 -6 -5 -4 -3

Log [Drug] (M)

% M

axim

al R

esp

on

se

00

20

40

60

80

100

1200 M

1 M

10 M

-9 -8 -7 -6 -5 -4 -3

100 M1 mM

[CGP 46381]

Log [Baclofen] (M)

% M

axim

al R

esp

on

se

ig. 3. Selective GABAB receptor antagonists inhibit agonist inductionf Ca2� mobilization in GABAB1(b)/2/CHO cells. (A) The ability of 10 �Mf baclofen to mobilize Ca2� in GABAB1(b)/2/CHO cells using FLIPRssays was determined in the presence of increasing concentrationsf antagonists CGP55845 and CGP46381. The normalized data areresented as the means�S.D.s of data averaged across three inde-endent experiments performed in duplicate or triplicate. (B) Thebility of increasing concentrations of CGP46381 to shift the dose-esponse curve of GABAB1(b)/2/CHO cells to baclofen (1 �M to 1 mM)as determined (in triplicate and normalized to the response of theells to 1 mM baclofen in the absence of antagonist) and C, wasransformed into a Schild plot to determine the KB of CGP46381.

B(1b)/2

ells to different GABAB receptor agonists and antago-Ta

ists. Dose-response curves constructed using FLIPR as-ays showed that the agonist GABA had a similar potencynd efficacy to baclofen (Fig. 2), with EC50 values of.5�3.72 �M (n�3) versus 4.43�1.8 �M (n�3), respec-ively. However, agonist SKF97541 was approximately 10-old more potent with an EC50 value of 0.37�0.05 �Mn�3; Fig. 2). The potency order of these three agonistseplicates that seen in previous studies (Seabrook et al.,990; Wood et al., 2000). The baclofen-induced Ca2� mo-ilization was completely abolished by co-administration with

he antagonists CGP55845 or CGP46381 with IC50 values of.5�2.1 �M (n�3) and 22.2�34 �M (n�3), respectively (Fig.A). To determine the KB value of CGP46381, a Schildnalysis was performed. Increasing concentrations ofGP46381 shifted the baclofen dose-response to the right

Fig. 3B) with a KB value of 0.82 �M and a slope of 0.94,uggesting potent, competitive inhibition (Fig. 3C) as previ-usly reported (Olpe et al., 1993; Wood et al., 2000).

dentification of intracellular pathways mediatinga2� release

aving determined that GABAB receptors are able to me-iate Ca2� mobilization in CHO cells, we proceeded toxamine the intracellular pathways mediating this effect.TX treatment of GABAB1(b)/2/CHO cells for 16 h prior toaclofen stimulation completely abolished the agonist-stimu-

ated increases in Ca2� levels (Fig. 4). This indicates that theesponses determined in FLIPR assays are entirely mediatedy PTX-sensitive Gi/o family heterotrimeric proteins.

G�� subunits released from Gi/o proteins are able toctivate PLC� (Rebecchi and Pentyala, 2000), potentiallyodulating Ca2� release and we reasoned that such aathway may be operational in GABAB1(b)/2/CHO cells.ndeed, the PLC� inhibitor U73122 at a concentration of

�M reduced Ca2� mobilization induced by a saturatingoncentration of baclofen by 86.3�7.8% (n�3; data nothown). Increasing the concentration of U73122-10 �M in-ibited responses to agonists by 99.1�4.0% (n�3; Fig. 5A).

00

20

40

60

80

100

No PTXPTX Treatment

-9 -8 -7 -6 -5 -4 -3

Log [Baclofen] (M)

% M

axim

al R

esp

on

se

ig. 4. Effect of PTX on the response of GABAB1(b)/2/CHO cells toaclofen. GABAB1(b)/2/CHO cells were seeded in 96-well plates 6 hefore the addition of PTX (100 ng/ml) or vehicle to the cells. Dose-esponse curves were constructed using baclofen in the FLIPR platform.

he normalized data are presented as the means�S.D. of data averagedcross three independent experiments performed in triplicate.

IUisdi

Cc

inb7rras

FeSbp ndent exp

D. C. New et al. / Neuroscience 137 (2006) 1347–1358 1351

n contrast, 4 �M or 10 �M of the inactive homologue of73122, U73343, was unable to suppress the baclofen-

nduced Ca2� mobilization. U73343 (10 �M) inhibited re-ponses by �3.6�9.2% (n�3; Fig. 5A). These resultsemonstrate that GABAB receptors induce increases in

ntracellular Ca2� levels via a PLC�-mediated pathway.To identify the downstream effectors and the source of

a2�, GABA /CHO cells were challenged with ba-

-20

0

20

40

60

80

100

120

% B

aclo

fen

-In

du

ced

Res

po

nse

No A

dditi

on

10

M U

7312

2

10

M 2

-APB

10

M U

7334

3

A

0

20

40

60

80

100

120

140

% B

aclo

fen

-In

du

ced

Res

po

nse

B

20 M

Ruthenium

red

Da

No

Addition

ig. 5. Effect of pharmacological inhibitors on the response of GABABffect of the indicated inhibitors. U73122, U73343, 2-APB, calphostiKF96365 and all inhibitors indicated in panel B were added 10 min inasal value obtained in the presence of the inhibitor. None of the inhresented as the means�S.D. of data averaged across three indepe

B1(b)/2

lofen following treatment with various pharmacological C

nhibitors. The IP3 receptor antagonist 2-aminethoxydiphe-yl borane (2-APB) dose-dependently inhibited Ca2� mo-ilization with 10 �M inhibiting the response by at least5% and 100 �M completely abolishing baclofen-inducedesponses (102.1�4.1%, n�3; Fig. 5A). The inhibition ofesponses by 2-APB implicates intracellular Ca2� storess the likely source of the Ca2� measured in FLIPR as-ays, but 2-APB is also known to inhibit store-operated

0 M

SKF

9636

5PB

100

M S

KF96

365

200

nM C

alph

ostin

C

10

M H

89

10 M

Nifendipene

1 M

-conotoxin

GVIA

1 nM

SNX-482

O cells to 100 �M baclofen. FLIPR assays were used to examine theH89 were added to the cell labeling medium 20 min before assay.. For each treatment the data were measured as an increase over theed had a significant effect on basal values. The normalized data areeriments performed in triplicate.

110

0 M

2-A

10 M

ntrolene

1(b)/2/CHn C andadvance

ibitors us

a2� entry over the same concentration range that antag-

op11chSnwmucC

pnicleCmerwd

Ea

Waih

2cbe(nr

GC

TGcGCl92br(cS

Cc

Tmipla

F(ea ependen

D. C. New et al. / Neuroscience 137 (2006) 1347–13581352

nizes IP3 receptors (Bootman et al., 2002). Therefore, wereincubated the cells with the selective SOC inhibitor-[2-(4-methoxyphenyl)-2-[3-(4-methoxyphenyl)propoxy]ethyl-H-imidazole hydrochloride (SKF96365) for 10 min beforehallenge with baclofen. SKF96365 (10 �M) significantly in-ibited FLIPR responses to baclofen while 100 �MKF96365 completely abolished the response (95�14.3%,�3; Fig. 5A), suggesting that the majority of Ca2� measuredithin the cell was the result of influx from the extracellularedium. This was confirmed when the assays were repeatedsing Ca2� free buffers. Under these assay conditions, ba-lofen was ineffective in promoting increases in intracellulara2� levels even at a concentration of 1 mM (Fig. 6).

As the baclofen-induced Ca2� fluxes were PLC� de-endent, we tested the ability of the selective protein ki-ase C (PKC) inhibitor calphostin C to inhibit increases in

ntracellular Ca2� levels. At a concentration of 200 nM,alphostin C was unable to significantly affect Ca2� mobi-

ization, indicating that this kinase and its downstreamffectors do not play a role in GABAB receptor-induceda2� fluxes. As GABAB receptors have been shown toediate cAMP production in some cell types (10), we alsoxamined the role of protein kinase A (PKA) in GABABeceptor-mediated Ca2� flux. The selective inhibitor H89as completely ineffective in inhibiting the baclofen-in-uced responses (Fig. 5A).

xamining the role of ryanodine receptorsnd Ca2� channels

e also examined the contribution of ryanodine receptorsnd VGCC-mediated Ca2� fluxes to the GABAB receptor-

nduced intracellular Ca2� levels. Ryanodine receptor in-

-20

0

20

40

60

80

100

% M

axim

al R

esp

on

se

With Calcium

Vehicle 10 MU73122

50SKF

ig. 6. Effect of extracellular Ca2� on the GABAB receptor-induced100 �M) in FLIPR assays was determined in the presence (with calciuffect of U73122 and SKF96365 on baclofen-induced basal Ca2� levere presented as the means�S.D. of data averaged across three ind

ibitors ruthenium red and dantrolene at concentrations of e

0 �M and 10 �M, respectively, were not able to signifi-antly decrease the response of GABAB1(b)/2/CHO cells toaclofen (Fig. 5B). Similarly, the VGCC blockers nifendip-ne (L-type channel blocker; 10 �M), -conotoxin GVIAN-type channel blocker; 1 �M) or SNX-482 (R-type chan-el blocker; 30 nM) were unable to attenuate the GABABeceptor-mediated responses (Fig. 5B).

ABAB1(a)/2 receptors also promote store-operateda2� entry

o determine the ability of GABAB1(a) containing heterodimericABAB receptors to activate Ca

2� influx via SOCs, weonstructed and examined the signaling properties ofABAB1(a)/2/CHO cells. As observed with GABAB1(b)/2/HO cells, baclofen promoted increases in intracellular Ca2�

evels, which were antagonized by CGP55845 (IC50�.8�10.5 �M (n�3); Fig. 7A). Furthermore, PTX, U73122,-APB and SKF96365 were all able to completely inhibitaclofen-induced responses over similar concentrationanges as were effective in GABAB1(b)/2 expressing cellsFig. 7B). This indicates that GABAB1(a)- and GABAB1(b)-ontaining receptors are both able to promote Ca2� influx viaOCs using a PTX-sensitive, PLC�-dependent pathway.

a2� influx through SOCs in primary cultures of ratortical neurons

o determine whether GABAB receptors are able to pro-ote Ca2� influx in neuronal cells endogenously express-

ng GABAB receptors, a similar set of experiments waserformed using primary cultures of cortical neurons iso-

ated from rats and cultured in 96-well plates. Baclofen wasble to dose-dependently increase intracellular Ca2� lev-

Calcium-free

Vehicle 10 MU73122

50 MSKF96365

lar Ca2� levels. The response of GABAB1(b)/2/CHO cells to baclofenence (calcium-free) of Ca2� in the extracellular labeling medium. Thetypes of labeling medium was also determined. The normalized data

t experiments performed in duplicate or triplicate.

M96365

intracellum) or absls in both

ls with an EC50 value of 3.7�4.4 �M (n�4; Fig. 8A). The

rCFaP(cf(sodcS

ioaGToeap

G

Floaotm perform

D. C. New et al. / Neuroscience 137 (2006) 1347–1358 1353

esponse induced by 1 mM baclofen was antagonized byGP55845 with an IC50 value of 9.95�6.5 �M (n�3;ig. 8B), confirming that GABAB receptors were medi-ting the response. Pretreatment of cells for 16 h withTX rendered 1 mM baclofen completely ineffective

Fig. 9), indicating that, as in CHO cells, Gi/o proteins inortical neurons also transduce the Ca2� mobilizing ef-ects of GABAB receptors. The PLC� inhibitor U7312210 �M) completely inhibited baclofen-induced re-ponses, whereas U73343 was ineffective (Fig. 9). Asbserved in GABAB1(b)/2/CHO cells, 2-APB dose-depen-ently inhibited agonist-induced responses with 100 �Mompletely blocking increases in Ca2� levels (Fig. 9).

A

B

00

20

40

60

80

100

120

-8

Log [C

% M

axim

al R

esp

on

se

No

Addition

10 M

U73122

PTX

0

20

40

60

80

100

120

% B

aclo

fen

-In

du

ced

Res

po

nse

ig. 7. Effect of GABAB antagonists and pharmacological inhibitors oine stably expressing GABAB1(a) and GABAB2 receptor subunits was en 100 �M baclofen-induced Ca2� levels was characterized using FLIPcross three independent experiments performed in duplicate or triplican baclofen-induced Ca2� levels were also examined. For each treatm

he presence of the inhibitor. None of the inhibitors used had a sigeans�S.D. of data averaged across three independent experiments

OC inhibitor SKF96365 completely abolished agonist- c

nduced responses at 10 and 50 �M (Fig. 9). The ryan-dine receptor inhibitors ruthenium red and dantrolene,s well as the VGCC blockers nifendipene, -conotoxinVIA and SNX-482 were ineffective (data not shown).he accumulated data confirm that the PLC�-mediatedpening of SOCs by the stimulation of GABAB receptorsndogenously expressed in cortical neurons operates inmanner similar to GABAB receptors exogenously ex-

ressed in CHO cells.

DISCUSSION

ABA receptors are typical G -coupled receptors, which

-6 -5 -4

5845] (M)

100 M

43

10 M 10 M100 M

2-APB SKF96365

ponse of GABAB1(a)/2/CHO cells to 100 �M baclofen. (A) A CHO celld. The effect of the selective GABAB receptor antagonist CGP55845ormalized data are presented as the means�S.D.s of data averagede effects of PTX (100 ng/ml), U73122, U73343, 2-APB and SKF96365data were measured as an increase over the basal value obtained ineffect on basal values. The normalized data are presented as theed in triplicate.

-7

GP 5

10 M

U733

n the resstablisheR. The n

te. (B) Thent thenificant

B i

an efficiently inhibit adenylyl cyclase (Bowery et al.,

2GCGClefmtsseialacGiCoac

srfiamGpsss

SIGtbaapsdamfpPathc2rm2

tataci2i(c7radsawt(ms

FcCponotm

D. C. New et al. / Neuroscience 137 (2006) 1347–13581354

002), but generally thought to require the expression of

q/i chimeras to activate PLC� and mobilize intracellulara2� (Wood et al., 2000; Liu et al., 2003). Although most

i-coupled receptors are unable to mobilize intracellulara2� in the absence of Gq/i chimeras, a number of Gi-

inked receptors appear to possess an inherent ability tolevate intracellular Ca2� levels. Muscarinic M2 and M4,

ormyl peptide-receptor-like-1 (New and Wong, 2004), so-atostatin SST2 (Nunn et al., 2004) and �-opioid recep-

ors (Smart et al., 1997) are all able to mediate PTX-ensitive elevation of intracellular Ca2� levels. The presenttudy suggests that activating GABAB receptors containingither the B1(a) or B1(b) subunit can also lead to an

ncrease in intracellular Ca2� levels, primarily through thectivation of SOCs. This notion is supported by several

ines of evidence. Firstly, the pharmacological profiles forgonist-induced Ca2� responses in GABAB1(b)/2/CHOells are in agreement with known selectivity of theABAB receptor (Fig. 2); GABAB receptor antagonists

nhibited the agonist-induced responses in GABAB1(a)/2/HO and GABAB1(b)/2/CHO cells (Figs. 3 and 7A). Sec-ndly, a selective inhibitor of SOCs, SKF96365 (Merritt etl., 1990), abolished the GABA receptor-mediated

A

B

00

20

40

60

80

100

-9 -8 -7 -6 -5 -4

Log [CGP55845] (M)

% M

axim

al R

esp

on

se

00

20

40

60

80

100

-6 -5 -4 -3

Log [Baclofen] (M)

% M

axim

al R

esp

on

se

ig. 8. Ca2� mobilization in primary cultures of rat cortical neuronshallenged with a GABAB receptor agonist and antagonized byGP55845. (A) FLIPR assays were used to determine the response ofrimary rat cortical neurons challenged with increasing concentrationsf baclofen. (B) Primary cultures of cortical neurons were simulta-eously challenged with 1 mM baclofen and increasing concentrationsf CGP55845. For both panels, the normalized data are presented ashe means�S.D. of data averaged across three independent experi-ents performed in duplicate or triplicate.

B

hanges in intracellular Ca2� (Figs. 5A and 7B). The ab- r

ence of any response when Ca2� free buffers were useduled out the possibility that SKF96365 was exerting ef-ects other than on SOCs (Fig. 6). Lastly, the baclofen-nduced Ca2� fluxes and their sensitivity to antagonistsnd pharmacological inhibitors were also observed in pri-ary cultures of rat cortical neurons (Figs. 8 and 9), whereABAB1 and GABAB2 subunits are endogenously co-ex-ressed (Kaupmann et al., 1998). These results stronglyuggest that activation of GABAB receptors in the CHOtable cell lines and rat cortical neurons can lead to thetimulation of SOCs.

The mechanism by which GABAB receptors activateOCs appears to involve multiple signaling intermediaries.

nhibition of Ca2� influx by PTX treatment of cells identified

i/o heterotrimeric proteins as mediators of GABAB recep-or-generated signals (Figs. 4, 7B and 9). Complete inhi-ition of Ca2� influx by U73122 (but not by its inactivenalog) implicates the involvement of PLC� (Figs. 5A, 7Bnd 9). To date there are no known instances of Gi/oroteins directly activating PLC� and it has been demon-trated that constitutively active mutants of Gi/o subunitso not promote PLC� activity (Tsu et al., 1995). However,ctivation of PLC� by G�� dimers is an increasingly com-on observation (Rebecchi and Pentyala, 2000). It there-

ore seems likely that GABAB receptor activation of Gi/oroteins leads to the release of G�� dimers that activateLC�, which in turn triggers downstream events leading ton increase in intracellular levels of Ca2�. GABAB recep-or-mediated release of G�� subunits from Gi/o proteinsas previously been shown to result in increased adenylylyclase activity in the rat frontal cortex (Onali and Olianas,001). Furthermore, Xenopus spinal growth cones areepelled by a gradient of baclofen, which is apparentlyediated by Gi protein activation of PLC� (Xiang et al.,002).

Activation of PLC� leads to the generation of IP3 andhe subsequent release of Ca2� from intracellular stores,s well as the generation of diacylglycerol and the activa-ion of PKC (Berridge, 1993). 2-APB is an IP3 receptorntagonist that has been widely used as a pharmacologi-al tool to characterize IP3-mediated Ca

2� release fromntracellular stores (Bootman et al., 2002). However,-APB has also been shown to be equally effective at

nhibiting the Ca2� influx from the extracellular mediumBootman et al., 2002) and, therefore, its inhibition of ba-lofen-mediated Ca2� flux in our experiments (Figs. 5A,B and 9) does not allow us to conclude a role for IP3eceptors in the activation of SOCs. However, the lack ofn effect on baclofen-induced Ca2� influx by calphostin Coes allow us to conclude that PKC and its effectors do notignificantly contribute to the GABAB receptor-mediatedctivation of SOCs (Fig. 5A). This observation is consistentith previous findings that thapsigargin- and orexin recep-

or-induced SOC activation are also unaffected by PKCVenkatachalam et al., 2003; Larsson et al., 2005). A sum-ary of the proposed mechanism of Ca2� entry is pre-

ented schematically in Fig. 10.These accumulated data suggested to us that GABA

B

eceptor activation leads to a Gi/o-dependent, PLC�-medi-

atcpeqS

odcnSkf

Ftiim perform

FiUrPV

D. C. New et al. / Neuroscience 137 (2006) 1347–1358 1355

ted rapid influx of Ca2� into the cell through SOCs. Ex-ensive studies on the process of Ca2� entry have indi-ated that several mechanisms may mediate PLC�-de-endent Ca2� influx (Putney et al., 2001; Venkatachalamt al., 2002), with contradictory data reported on the re-uirement of IP3 generation and IP3 receptor activation.everal studies have shown that IP3 is required for store-

10 M

U73122

No

Addition

PTX-20

0

20

40

60

80

100

120

% B

aclo

fen

-In

du

ced

Res

po

nse

ig. 9. Effect of PTX and pharmacological inhibitors on the response oreated for 16 h before assay using the FLIPR platform. The other inhibncubation times are as specified in the legend to Fig. 6). For each tren the presence of the inhibitor. None of the inhibitors used had a s

eans�S.D. of data averaged across three independent experiments

IP

PKA

ATP

cAMP

GABAVGCCs

ββββGααααi/o

AC

ER

2-APB

PTX

ig. 10. Proposed mechanism for the activation of Ca2� entry mediatndicate that agonist activation of GABAB receptors leads to an influx of73122, 2-APB and SKF96365 suggests that a mechanism of capacita

ole of intermediaries between PLC�, IP3, intracellular stores and SOC

arekh and Putney, 2005 for full discussions). Our experimental evidence showGCCs for SOC entry.

perated Ca2� entry (Zubov et al., 1999) but others haveemonstrated that in lacrimal and rat basophilic leukemiaells treated with the PLC� inhibitor U73122, Ca2� influx isot restored by the application of exogenous IP3 and thatOC activating pathways are operational in IP3 receptornockout DT40 B-lymphocytes (Broad et al., 2001). There-ore, there appear to be IP3 dependent and independent

10 M 10 100 M M 50 M

2-APB SKF96365

rat cortical neurons to 1 mM baclofen. For PTX (100 ng/ml), cells wereapplied at the indicated concentrations in advance of the assays (the

e data were measured as an increase over the basal value obtainedeffect on basal values. The normalized data are presented as the

ed in duplicate or triplicate.

PKC

PLCββββ

Ca2+

?Ca2+

U73122

SKF96365

2-APB

SOC

BAB receptors. The data presented are illustrated schematically andthe extracellular medium. The inhibition of this phenomenon by PTX,

um entry operates to allow Ca2� influx through SOCs. The nature andt yet fully elucidated (the reader is directed to Putney et al., 2001 and

10 M

U73343

f primaryitors wereatment thignificant

3

B

γγγγ

ed by GACa2� fromtive calcis are no

s that GABAB receptors do not require the activation of PKA, PKC or

mPtbum

tracvdvutwPSG

cowo(nol2sca

ps(oi5pmrCisCs

ifitotrtia

itlatuG

timppnGa2rtParqGsctLntbVrii

Sol(p1ggAGtpTana2

bLas

D. C. New et al. / Neuroscience 137 (2006) 1347–13581356

echanisms of SOC activation. Our data confirm theLC� dependence of GABAB receptor-mediated capacita-

ive calcium entry, although the lack of selective, mem-rane permeable IP3 receptor antagonists has preventeds from demonstrating IP3 receptor involvement in thisechanism.

Several other GPCRs have been demonstrated to ac-ivate SOCs, including metabotropic glutamate subtype 1eceptors in dopamine neurons in rat brain slices (Tozzi etl., 2003), muscarinic receptors expressed in lymphaticell lines (Broad et al., 2001) and endothelin receptors inascular smooth muscle cells (Kawanabe et al., 2002). Toate, no systematic study on the ability of GPCRs to acti-ate SOCs in neuronal and non-neuronal cells has beenndertaken. It is, therefore, unclear whether SOC activa-ion is a property common to all Gq-coupled receptors asell as those Gi/o-coupled GPCRs that are able to activateLC�, or whether there is an extra degree of control ofOC opening that can only be overcome by a subset ofPCRs.

It has recently been reported that the orexin GPCRsan promote Ca2� influx into CHO cells via the activationf nonstore-operated Ca2� entry (Larsson et al., 2005). Asith SOCs, these channels are believed to be composedf subtypes of the transient receptor potential channelTRPC) family (Minke and Cook, 2002; Parekh and Put-ey, 2005). A distinguishing characteristic of nonstore-perated entry, compared with Ca2� entry via SOCs, is its

ack of sensitivity to 2-APB and SKF96365 (Larsson et al.,005). The sensitivity to these two blockers of the re-ponses that we observed in transfected CHO cells andortical neurons indicates that GABAB receptors do notctivate nonstore-operated Ca2� channels.

GABAB receptors have previously been shown to sup-ress (Bowery et al., 2002) or facilitate Ca2� influx viaeveral sub-types of VGCCs, notably L-type channelsParramon et al., 1995; Carter and Mynlieff, 2004). Webserved no signs of GABAB receptor-mediated VGCC

nflux in transfected CHO cells or rat cortical neurons (Fig.B). Ryanodine receptors are activated by some Gi/o-cou-led GPCRs (Maghazachi, 2000) in a PLC�-independentanner (Putney et al., 2001), but we conclude that GABAB

eceptors do not activate this system when expressed inHO cells or rat cortical neurons (Fig. 5B). However, our

nvestigation measured Ca2� levels on a whole cellcale, and we do not rule out the possibility of localizeda2� gradients controlled by VGCCs or ryanodine-sen-itive stores.

It was apparent from the experiments in which Ca2�

nflux was inhibited by SKF96365 or by the use of Ca2�

ree buffers that we could not detect Ca2� release fromntracellular stores, even though depletion of Ca2� fromhe ER is thought by some to be required before SOCspen (Putney et al., 2001). This would seem to suggesthat in our experimental system the IP3 receptor-inducedelease of Ca2� from intracellular stores following activa-ion of GABAB receptors is on a small scale. Alternatively,t is possible that Ca2� release from intracellular stores is not

prerequisite for SOC activation. Consistent with these find- (

ngs are observations that in skeletal muscle cells, IP3 recep-ors mediate SOC opening without necessarily triggering re-ease from internal Ca2� stores (Launikonis et al., 2003). Were currently investigating the relative contributions of ex-racellular Ca2� influx and intracellular Ca2� mobilizationpon activation of PLC�-dependent pathways by other

i/o- and Gq-coupled receptors.GABAB receptor involvement has been demonstrated

o regulate a variety of neuronal functions both directly andndirectly, by modulating the activity of other neurotrans-

itter systems. GABAB receptors generate inhibitoryostsynaptic potentials that are associated with large hy-erpolarizations, which have an inhibitory effect on manyeurons (Mott and Lewis, 1994). However, postsynapticABAB receptors in the thalamus are excitatory, with suchctivity contributing to absence epilepsy (Calver et al.,002). Postsynaptically, GABAB receptor activity can indi-ectly affect neuronal function by, for example, promotinghe phosphorylation and inhibition of GABAA receptors in aLC�-dependent manner (Hahner et al., 1991). Postsyn-ptic activity is especially effective at inhibiting NMDAeceptor-mediated responses using a mechanism that re-uires G proteins (Morrisett et al., 1991) and, therefore,ABAB receptors are likely to play an inhibitory role inynaptic plasticity and long-term potentiation. Presynapti-ally, GABAB receptors have a depressant effect on exci-atory responses in numerous brain regions (Mott andewis, 1994), leading to the modulation of the release ofumerous neurotransmitters (Calver et al., 2002). Many of

hese post- and presynaptic effects are, in part, regulatedy the GABAB receptor regulation of K

� channels andGCCs. However, in light of our findings that GABAB

eceptor activation in neuronal cells leads to increases inntracellular Ca2� levels, we suggest that the role of SOCsn neurophysiology be considered.

It is currently unclear as to whether Ca2� influx throughOCs is simply required for the replenishment of ER storesr whether it serves some independent function. Neverthe-

ess, SOC activation has been implicated in neuroplasticityBaba et al., 2003), DNA synthesis and ordered cell cyclerogression in retinal neuroepithelial cells (Sugioka et al.,999), while reduced SOC activity is associated with theeneration of the A�42 protein, a key component in theeneration of plaques associated with the development oflzheimer’s disease (Yoo et al., 2000). We anticipate thatABAB receptor-mediated activation of SOCs contributes

o a complex interplay of Ca2� releasing and sequesteringathways that regulate the intracellular levels of Ca2�.his may be required for the regulation of electric chargend the effective control of neuronal functions, includingeurotransmitter release, excitability, synaptic plasticitynd gene expression (Mott and Lewis, 1994; Verkhratsky,005).

The broad expression profile of GABAB receptors atoth pre- and post-synaptic sites in the CNS (Mott andewis, 1994) has led to their use as therapeutic targets,nd treatments with anti-convulsant, anti-ulcer, anti-amne-ic and many other efficacies are being actively pursued

Bolser et al., 1995; Kerr and Ong, 1995; Nava et al., 2001;

BrtwdssrdtaIviS

Atrst2KBn

B

B

B

B

B

B

B

B

C

C

H

H

H

J

K

K

KK

L

L

L

L

M

M

M

M

M

N

N

N

O

O

D. C. New et al. / Neuroscience 137 (2006) 1347–1358 1357

rown et al., 2003). We have previously examined theesponses of GABAB receptors in FLIPR assays using aransient transfection system (Liu et al., 2003). We foundeak but consistent Ca2� increases in response to highoses of baclofen that were potentiated by the chimeric G

ubunits 16z25 and 16z44. Our generation of CHO cellstably expressing GABAB receptors provides a much moreobust screening platform that replicates the signal trans-uction pathways seen in neuronal cells. We believe thathis will enable more efficient drug screening for agonists,ntagonists and allosteric modulators of GABAB receptors.n conclusion, we have identified a GABAB receptor-acti-ated signal transduction pathway in cortical neurons thats mediated by Gi/o proteins and leads to the activation ofOCs.

cknowledgments—The authors are grateful to Agnes Chan, Es-ella Tong, Fred Luk and Fanny Ip for assistance with the prepa-ation of the cortical neurons and the subcloning. This work wasupported in part by the Hong Kong Jockey Club and grants fromhe Innovation and Technology Commission of Hong Kong (ITS/26/01 and ITS/113/03), the Research Grants Council of Hongong (HKUST 3/03C), and the University Grants Committee (AoE/-15/01). N.Y.I. and Y.H.W. were recipients of the Croucher Se-ior Research Fellowship.

REFERENCES

aba A, Yasui T, Fujisawa S, Yamada RX, Yamada MK, Nishiyama N,Matsuki N, Ikegaya Y (2003) Activity-evoked capacitative Ca2�

entry: implications in synaptic plasticity. J Neurosci 23:7737–7741.erridge MJ (1993) Inositol trisphosphate and calcium signalling. Na-

ture 361:315–325.olser DC, Blythin DJ, Chapman RW, Egan RW, Hey JA, Rizzo C, Kuo

SC, Kreutner W (1995) The pharmacology of SCH 50911: a novel,orally-active GABA-B receptor antagonist. J Pharmacol Exp Ther274:1393–1398.

ootman MD, Collins TJ, Mackenzie L, Roderick HL, Berridge MJ,Peppiatt CM (2002) 2-Aminoethoxydiphenyl borate (2-APB) is areliable blocker of store-operated Ca2� entry but an inconsistentinhibitor of InsP3-induced Ca

2� release. FASEB J 16:1145–1150.owery NG, Hudson AL, Price GW (1987) GABAA and GABAB recep-

tor site distribution in the rat central nervous system. Neuroscience20:365–383.

owery NG, Bettler B, Froestl W, Gallagher JP, Marshall F, Raiteri M,Bonner TI, Enna SJ (2002) International Union of Pharmacology.XXXIII. Mammalian gamma-aminobutyric acidB receptors: struc-ture and function. Pharmacol Rev 54:247–264.

road LM, Braun FJ, Lievremont JP, Bird GS, Kurosaki T, Putney JWJr (2001) Role of the phospholipase C-inositol 1,4,5-trisphosphatepathway in calcium release-activated calcium current and capaci-tative calcium entry. J Biol Chem 276:15945–15952.

rown JT, Gill CH, Farmer CE, Lanneau C, Randall AD, Pangalos MN,Collingridge GL, Davies CH (2003) Mechanisms contributing to theexacerbated epileptiform activity in hippocampal slices of GABAB1receptor subunit knockout mice. Epilepsy Res 57:121–136.

alver AR, Davies CH, Pangalos M (2002) GABAB receptors: frommonogamy to promiscuity. Neurosignals 11:299–314.

arter TJ, Mynlieff M (2004) �-Aminobutyric acid type B receptorsfacilitate L-type and attenuate N-type Ca2� currents in isolatedhippocampal neurons. J Neurosci Res 76:323–333.

ahner L, McQuilkin S, Harris RA (1991) Cerebellar GABAB receptorsmodulate function of GABAA receptors. FASEB J 5:2466–2472.

o MK, Wong YH (2001) G signaling: emerging divergence from G

z isignaling. Oncogene 20:1615–1625.

o MKC, New DC, Wong YH (2002) Co-expressions of different opioidreceptor types differentially modulate their signaling via G16. Neu-rosignals 11:115–122.

ones KA, Borowsky B, Tamm JA, Craig DA, Durkin MM, Dai M, YaoWJ, Johnson M, Gunwaldsen C, Huang LY, Tang C, Shen Q,Salon JA, Morse K, Laz T, Smith KE, Nagarathnam D, Noble SA,Branchek TA, Gerald C (1998) GABAB receptors function as aheteromeric assembly of the subunits GABABR1 and GABABR2.Nature 396:674–679.

aupmann K, Malitschek B, Schuler V, Heid J, Froestl W, Beck P,Mosbacher J, Bischoff S, Kulik A, Shigemoto R, Karschin A, BettlerB (1998) GABAB-receptor subtypes assemble into functional het-eromeric complexes. Nature 396:683–687.

awanabe Y, Hashimoto N, Masaki T (2002) Ca2� channels involvedin endothelin-induced mitogenic response in carotid artery vascu-lar smooth muscle cells. Am J Physiol 282:C330–C337.

err DI, Ong J (1995) GABAB receptors. Pharmacol Ther 67:187–246.night AR, Bowery NG (1996) The pharmacology of adenylyl cyclase

modulation by GABAB receptors in rat brain slices. Neuropharma-cology 35:703–712.

arsson KP, Peltonen HM, Bart G, Louhiuori LM, Penttonen A, Anti-kainen M, Kukkonen JP, Akerman KEO (2005) Orexin-A-inducedCa2� entry: evidence for involvement of TRPC channels and pro-tein kinase C regulation. J Biol Chem 280:1771–1781.

aunikonis BS, Barnes M, Stephenson DG (2003) Identification of thecoupling between skeletal muscle store-operated Ca2� entry andthe inositol trisphosphate receptor. Proc Natl Acad Sci USA 100:2941–2944.

iu AMF, Ho MKC, Wong CSS, Chan JHP, Pau AHM, Wong YH (2003)G16/z chimeras efficiently link a wide range of G protein-coupledreceptors to calcium mobilization. J Biomol Screen 8:39–49.

iu AMF, Wong YH (2005) �-Opioid receptor-mediated phosphoryla-tion of I�B kinase in human neuroblastoma SH-SY5Y cells. Neu-rosignals 14:136–142.

aghazachi AA (2000) Intracellular signalling events at the leadingedge of migrating cells. Int J Biochem Cell Biol 32:931–943.

erritt JE, Armstrong WP, Benham CD, Hallam TJ, Jacob R, Jaxa-Chamiec A, Leigh BK, McCarthy SA, Moores KE, Rink TJ (1990)SKandF 96365, a novel inhibitor of receptor-mediated calciumentry. Biochem J 271:515–522.

inke B, Cook B (2002) TRP channel proteins and signal transduction.Physiol Rev 82:429–472.

orrisett RA, Mott DD, Lewis DV, Swartzwelder HS, Wilson WA (1991)GABAB-receptor-mediated inhibition of the N-methyl-D-aspartatecomponent of synaptic transmission in the rat hippocampus.J Neurosci 11:203–209.

ott DD, Lewis DV (1994) The pharmacology and function of centralGABAB receptors. Int Rev Neurobiol 36:97–223.

ava F, Carta G, Bortolato M, Gessa GL (2001) �-Hydroxybutyric acidand baclofen decrease extracellular acetylcholine levels in thehippocampus via GABAB receptors. Eur J Pharmacol 430:261–263.

ew DC, Wong YH (2004) Characterization of CHO cell lines stablyexpressing a G16/z chimera for high throughput screening ofGPCRs. Assay Drug Dev Tech 2:269–280.

unn C, Cervia D, Langenegger D, Tenaillon L, Bouhelal R, Hoyer D(2004) Comparison of functional profiles at human recombinantsomatostatin sst2 receptor: simultaneous determination of intracel-lular Ca2� and luciferase expression in CHO-K1 cells. Br J Phar-macol 142:150–160.

dagaki Y, Koyama T (2001) Identification of G subtype(s) involvedin �-aminobutyric acidB receptor-mediated high-affinity guanosinetriphosphatase activity in rat cerebral cortical membranes. Neuro-sci Lett 297:137–141.

lpe HR, Steinmann MW, Ferrat T, Pozza MF, Greiner K, Brugger F,Froestl W, Mickel SJ, Bittiger H (1993) The actions of orally activeGABA receptor antagonists on GABAergic transmission in vivo

Band in vitro. Eur J Pharmacol 233:179–186.

O

P

P

P

P

R

R

S

S

S

S

T

T

V

V

V

W

W

X

Y

Z

D. C. New et al. / Neuroscience 137 (2006) 1347–13581358

nali P, Olianas MC (2001) ��-Mediated enhancement of corticotrop-in-releasing hormone-stimulated adenylyl cyclase activity by acti-vation of gamma-aminobutyric acidB receptors in membranes of ratfrontal cortex. Biochem Pharmacol 62:183–190.

arekh AB, Putney JW Jr (2005) Store-operated calcium channels.Physiol Rev 85:757–810.

arramon M, Gonzalez MP, Herrero MT, Oset-Gasque MJ (1995)GABAB receptors increase intracellular calcium concentrations inchromaffin cells through two different pathways: their role in cate-cholamine secretion. J Neurosci Res 41:65–72.

in JP, Kniazeff J, Binet V, Liu J, Maurel D, Galvez T, Duthey B,Havlickova M, Blahos J, Prezeau L, Rondard P (2004) Activationmechanism of the heterodimeric GABAB receptor. Biochem Phar-macol 68:1565–1572.

utney JW Jr, Broad LM, Braun FJ, Lievremont JP, Bird GS (2001)Mechanisms of capacitative calcium entry. J Cell Sci 114:2223–2229.

ebecchi MJ, Pentyala SN (2000) Structure, function, and control ofphosphoinositide-specific phospholipase C. Physiol Rev 80:1291–1335.

estituito S, Couve A, Bawagan H, Jourdain S, Pangalos MN, CalverAR, Freeman KB, Moss SJ (2005) Multiple motifs regulate thetrafficking of GABAB receptors at distinct checkpoints within thesecretory pathway. Mol Cell Neurosci 28:747–756.

ambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a labo-ratory manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Har-bor Laboratory.

eabrook GR, Howson W, Lacey MG (1990) Electrophysiologicalcharacterization of potent agonists and antagonists at pre- andpostsynaptic GABAB receptors on neurones in rat brain slices. Br JPharmacol 101:949–957.

mart D, Hirst RA, Hirota K, Grandy DK, Lambert DG (1997) Theeffects of recombinant rat �-opioid receptor activation in CHO cellson phospholipase C, [Ca2�]i and adenylyl cyclase. Br J Pharmacol120:1165–1171.

ugioka M, Zhou WL, Hofmann HD, Yamashita M (1999) Ca2� mobi-lization and capacitative Ca2� entry regulate DNA synthesis incultured chick retinal neuroepithelial cells. Int J Dev Neurosci

17:163–172.

ozzi A, Bengtson CP, Longone P, Carignani C, Fusco FR, BernardiG, Mercuri NB (2003) Involvement of transient receptor potential-like channels in responses to mGluR-I activation in midbrain do-pamine neurons. Eur J Neurosci 18:2133–2145.

su RC, Lai HW, Allen RA, Wong YH (1995) Differential coupling of theformyl peptide receptor to adenylate cyclase and phospholipase Cby the pertussis toxin-insensitive Gz protein. Biochem J 309:331–339.

enkatachalam K, van Rossum DB, Patterson RL, Ma HT, Gill DL(2002) The cellular and molecular basis of store-operated calciumentry. Nat Cell Biol 4:E263–E272.

enkatachalam K, Zheng F, Gill DL (2003) Regulation of canonicaltransient receptor potential (TRPC) channel function by diacylglyc-erol and protein kinase C. J Biol Chem 278:29031–29040.

erkhratsky A (2005) Physiology and pathophysiology of the calciumstore in the endoplasmic reticulum of neurons. Physiol Rev85:201–279.

hite JH, Wise A, Main MJ, Green A, Fraser NJ, Disney GH, BarnesAA, Emson P, Foord SM, Marshall FH (1998) Heterodimerization isrequired for the formation of a functional GABAB receptor. Nature396:679–682.

ood MD, Murkitt KL, Rice SQ, Testa T, Punia PK, Stammers M,Jenkins O, Elshourbagy NA, Shabon U, Taylor SJ, Gager TL,Minton J, Hirst WD, Price GW, Pangalos M (2000) The humanGABAB1b and GABAB2 heterodimeric recombinant receptor showslow sensitivity to phaclofen and saclofen. Br J Pharmacol 131:1050–1054.

iang Y, Li Y, Zhang Z, Cui K, Wang S, Yuan XB, Wu CP, Poo MM,Duan S (2002) Nerve growth cone guidance mediated by G pro-tein-coupled receptors. Nat Neurosci 5:843–848.

oo AS, Cheng I, Chung S, Grenfell TZ, Lee H, Pack-Chung E,Handler M, Shen J, Xia W, Tesco G, Saunders AJ, Ding K, FroschMP, Tanzi RE, Kim TW (2000) Presenilin-mediated modulation ofcapacitative calcium entry. Neuron 27:561–572.

ubov AI, Kaznacheeva EV, Nikolaev AV, Alexeenko VA, Kiselyov K,Muallem S, Mozhayeva GN (1999) Regulation of the miniatureplasma membrane Ca2� channel I by inositol 1,4,5-trisphos-

minphate receptors. J Biol Chem 274:25983–25985.

(Accepted 18 October 2005)(Available online 15 December 2005)

GABAB HETERODIMERIC RECEPTORS PROMOTE Ca 2 INFLUX VIA STORE-OPERATED CHANNELS IN RAT CORTICAL NEURONS AND...EXPERIMENTAL PROCEDURESMaterialsGeneration of stable cell linesPreparation of cortical neuronal culturesWestern blotting analysisMeasurement of cAMP levelsMeasurement of intracellular Ca 2 using a FLIPR device

RESULTSGeneration and characterization of CHO cells stably expressing GABAB1(b)/2 subunitsGABAB receptors mediate the mobilization of intracellular Ca 2 Identification of intracellular pathways mediating Ca 2 releaseExamining the role of ryanodine receptors and Ca 2 channelsGABAB1(a)/2 receptors also promote store-operated Ca 2 entryCa 2 influx through SOCs in primary cultures of rat cortical neurons

DISCUSSIONAcknowledgmentsREFERENCES