g protein-dependent ca2+signaling complexes in polarized cells

Invited review

G protein-dependent Ca2+ signalingcomplexes in polarized cells

S. Muallem, 1 T. M. Wilkie 2

1Department of Physiology, University of Texas southwestern Medical Center, Dallas, USA 2Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, USA

Summary Polarized cells signal in a polarized manner. This is exemplified in the patterns of [Ca2+]i waves and [Ca2+]ioscillations evoked by stimulation of G protein -coupled receptors in these cells. Organization of Ca2+-signaling com-plexes in cellular microdomains, with the aid of scaffolding proteins, is likely to have a major role in shaping G protein-coupled [Ca2+]i signal pathways. In epithelial cells, these domains coincide with sites of [Ca2+]i-wave initiation and local[Ca2+]i oscillations. Cellular microdomains enriched with Ca2+-signaling proteins have been found in several cell types.Microdomains organize communication between Ca2+-signaling proteins in the plasma membrane and internal Ca2+

stores in the endoplasmic reticulum through the interaction between the IP3 receptors in the endoplasmic reticulum andCa2+-influx channels in the plasma membrane. Ca2+ signaling appears to be controlled within the receptor complex bythe regulators of G protein-signaling (RGS) proteins. Three domains in RGS4 and related RGS proteins contributeimportant regulatory features. The RGS domain accelerates GTP hydrolysis on the Gα subunit to uncouple receptorstimulation from IP3 production; the C-terminus may mediate interaction with accessory proteins in the complex; andthe N-terminus acts in a receptor-selective manner to confer regulatory specificity. Hence, RGS proteins have both cat-alytic and scaffolding function in Ca2+ signaling. Organization of Ca2+-signaling proteins into complexes withinmicrodomains is likely to play a prominent role in the localized control of [Ca2+]i and in [Ca2+]i oscillations. © HarcourtPublishers Ltd

Cell Calcium (1999) 26 (5), 173–180© Harcourt Publishers Ltd 1999Article No. ceca.1999.0077

INTRODUCTION

Polarized cells, such as epithelial cells and neurons, per-formed their dedicated functions in a polarized manner.The cellular functions include electrolyte and fluid trans-port [1], exocytosis [2] and neurotransmission [3]. Thesevectorial functions are regulated by changes in cytoplas-mic Ca2+ concentrations ([Ca2+]i). G protein-coupledreceptors (GPCR) play a central role in regulating [Ca2+]i

in all mammalian cells, including polarized cells. At phys-iological concentrations, extracellular signals stimulateoscillatory charges in [Ca2+]i [4]. In polarized cells, [Ca2+]i

oscillations occur in localized microdomains [5,6] orpropagate as a [Ca2+]i wave that always initiates in thesame cellular pole, independent of the stimulated GPCR[5–7]. [Ca2+]i oscillations then promote rhythmic stimula-tion of cellular activity. The polarized nature of the cellu-lar activities regulated by GPCR requires polarized

Received 15 September 1999Accepted 15 September 1999

Correspondence to: Shmuel Muallem, Department of Physiology, Universityof Texas Southwestern Medical Center, Dallas, TX 75235, USA.

organization and functioning of [Ca2+]i signaling in cellu-lar microdomains. A central question in cell signaling ishow signaling proteins are organized into complexeswithin microdomains and how specificity of signaling isachieved. In the present review, we will focus our discus-sion on the organization of GPCR signaling complexes inepithelial cells.

GENERAL FEATURES OF Ca2+ SIGNALING

[Ca2+]i is regulated by two types of Ca2+ pumps and twotypes of Ca2+ channels, which reside in the plasma mem-brane (PM) and endoplasmic reticulum (ER). In restingcells, [Ca2+]i is determined by pump-leak turnover acrossthe PM, whereas Ca2+ transport by the ER serves as arapidly-acting, high-capacity and high-affinity Ca2+

buffering system. The well-established biochemicalsequence of events leading to a single [Ca2+]i transientbegins with binding of agonists to their receptors andactivation of phospholipase C (PLC) by G proteins (Gq orGi) [8]. Activation of PLC causes the enzymatic break-down of PIP2, generating 1,4,5–IP3 (IP3) and diacylglyc-

173

174 S Muallem, T M Wilkie

Fig. 1 [Ca2+]i is regulated by multiprotein complexes withinmicrodomains of polarized epithelial cells. Gq-mediated signalinginitiates at the plasma membrane producing IP3, which binds to theIP3R to release Ca2+ from internal stores. The IP3R binds to andregulates ICRAC activity in an IP3-dependent manner. IntracellularCa2+ is pumped into the ER or out of the cell by SERCA and PMCApumps respectively. Scaffolding proteins (gray bi-sphericalsymbols) are postulated to help assemble microdomains of theseCa2+ signaling proteins. PM, plasma membrane; ER endoplasmicreticulum; GPCR, G protein-coupled receptor; RGS, regulator of Gprotein signaling (a Gαq GAP); PLCβ, phospholipase Cβ; [Ca2+]i,intracellular Ca2+ concentration; IP3R, inositol -1,4,5-trisphosphatereceptor; SERCA, sarcoplasmic/endoplasmic Ca2+ ATPase; PMCA,plasma membrane Ca2+ ATPase; ICRAC, Ca2+ release activated Ca2+

current; agonist binding is depicted (star); palmitoylation andisoprenylation of Gα and Gγ are noted by insertions into the plasmamembrane.

erol [4]. IP3 releases Ca2+ from a compartmentalized ERCa2+ pool [9,10]. Ca2+ release is followed by activation of acapacitative Ca2+ influx pathway in the PM [11], one formof which is ICRAC [12]. Activation of the ER and PM chan-nels causes a rapid increase in [Ca2+]i which is the firstphase of the Ca2+ signal. In the second phase, Ca2+ isremoved from the cytosol by the action of the Ca2+

pumps [4,9]. The relative contribution of the two pumpsto the removal of [Ca2+]i from the cytoplasm depends onstimulus intensity [13]. During intense stimulation, whenthe activity of the IP3 channel is persistently high, the PMCa2+ pump extrudes most of the Ca2+. During weak stim-ulation, the IP3 channels appear to open only periodi-cally, and most of the released Ca2+ is pumped back intothe ER [14–16]. This sequence of events results in a tran-sient change in [Ca2+]i. Periodic repetition of Ca2+ releaseand re-sequestration during weak agonist stimulationgives rise to [Ca2+]i oscillations. A possible relationshipsbetween the proteins mediating the [Ca2+]i transient isillustrated by the model in Figure 1.

In recent years, it has become clear that the details of[Ca2+]i signaling are considerably more complex than thegeneral features outlined above would suggest. A majorcontributor to this complexity is the expression of multi-ple isoforms of Ca2+ signaling proteins in the same cell.For example, pancreatic acinar cells express three of thefour Gq class α subunits [17]. The activity of Gq proteinsis regulated by a recently discovered family of proteins,termed regulators of G proteins signaling (RGS) proteins[18,19]. There are at least 22 different RGS proteins inmammals [18]. Polarized cells, like pancreatic acinar andsalivary gland cells, probably express several RGS pro-teins that may recognize different signaling complexes.Four isoforms and 16 splice variants of the PM Ca2+

pump (PMCA) are known [20,21]. Two isoforms, PMCA1aand 4b, are expressed in most cells [21] including pancre-atic acinar cells (our unpublished observation). Similarly,four isoforms of the SERCA pumps [22,23] that areexpressed in a cell-specific manner [22–24] have beenidentified. Epithelial cells can express different combina-tions of SERCA pumps [24]. At least three isoforms of theIP3R [25] and three isoforms of the Ca2+-induced Ca2+

release channel (or ryanodine receptor (RyR)) [26] areknown. Many epithelial cells express all three IP3R iso-forms [25,27]. These cells express at least one RyR iso-form, since they respond to cyclic ADP ribose [27,28], aligand that activates Ca2+-induced Ca2+ release [29].

Ca2+ SIGNALING COMPLEXES

Faced with such complexity, how is specificity in Ca2+ sig-naling achieved? Rapidly developing evidence indicatesthat cells group signaling proteins into complexes thatare localized in specialized microdomains with the aid of

Cell Calcium (1999) 26(5), 173–180

scaffolding proteins (see reviews in this issue). A well-known example for G protein-dependent signaling is theorganization of Ca2+ signaling proteins by the scaffoldingprotein INAD in Drosophila photoreceptors (discussed byTsunoda and Zuker in this issue). Although several mam-malian homologues of INAD have been identified[30,31], their role in organization of signaling complexesis not known. Similarly, the scaffolding protein(s) respon-sible for organization of Ca2+ signaling complexes inepithelial cells have not been identified, although theirexistence is strongly suggested by functional and struc-tural evidence.

The quantal properties of Ca2+ release from internalstores (IS) [10,32] is one type of functional evidence forcompartmentalization of Ca2+ signaling components andtheir organized into complexes. This phenomenon hasbeen observed in many cell types [33] and it cannot beexplained by a pump-leak turnover across the storemembrane [34]. Thus, unidirectional 45Ca2+ flux mea-surement showed that cell stimulation at sub-maximalintensity could access only part of the agonist-mobiliz-able Ca2+ pool [34,35]. This behavior may contribute tolocalized changes in Ca2+ in the form of puffs [36] andsparks [37], which are high [Ca2+]i microdomainsobserved in many cell types [38,39].

© Harcourt Publishers Ltd 1999

G protein-dependent Ca2+ signaling complexes in polarized cells 175

Clear evidence for the existence of specialized cellularregions of [Ca2+]i signaling in epithelial cells was providedby the original observation of Kasai and Augustine [40].These authors reported that stimulation of pancreaticacinar cells with acetylcholine generated a [Ca2+]i wavethat was initiated in the luminal pole and propagated tothe basal pole. This observation has since been con-firmed by several groups [6,7,41] and found in otherepithelial cells [27,42]. Interestingly, at low stimulusintensity the Ca2+ signals can be initiated and remainconfined to the secretory pole [5,6]. The [Ca2+]i wavesevoked by repetitive stimulation with the same agonistwere the same in all aspects, including the site of Ca2+

wave initiation and propagation pattern, whereas differ-ent agonists acting on the same cell-evoked agonist-spe-cific [Ca2+]i waves [7]. The consistency in whichstimulation of GPCR initiate the Ca2+ waves in the lumi-nal pole reflects higher sensitivity of this region to IP3

[5,6,43].The structural basis for the quantal behavior of Ca2+

release and the polarized, agonist-specific [Ca2+]i wave arelikely to be facilitated by polarized expression and recruit-ment of Ca2+ signaling proteins. Immunolocalization stud-ies showed the polarized expression of both Ca2+ pumpsand Ca2+ channels in epithelial cells [24,27]. Polarizedexpression of Ca2+ channels and Ca2+ pumps appearsimportant for initiation of [Ca2+]i wave in the luminal poleand the rate of their propagation to the basal pole [24,27].In a model system, cell stimulation caused clustering ofGFP-tagged β2 adrenergic receptors in microdomains ofthe plasma membrane [44–46]. Furthermore, the stimu-lated receptors recruited adapter (scaffolding), effectorand modulatory proteins to form a G protein-mediatedsignaling complex [46].

Many Ca2+-signaling proteins contain protein-bindingmotifs that are likely to mediate their polarized and local-ized expression. The β2 adrenergic receptor binds to thePDZ domain of the Na+–H+ exchanger regulatory factor[47,48]. This interaction was required for inhibition of thetype 3 Na+–H+ exchanger by stimulation of the β2 adren-ergic receptor [47]. The metabotropic glutaminergicreceptor binds to the EVH domain of the scaffold proteinHomer through its PPXXFR biding motif [49,50]. Notably,the IP3 and ryanodine receptors contain the sameHomer-binding motif [50]. Ca2+ signaling may depend onprotein interactions with the scaffold, because disruptionof Homer coiled–coiled domain, which mediates Homermultimerization, inhibited IP3-dependent Ca2+ releasefrom internal stores [50]. Finally, isoform 4b of PMCA,which is expressed in epithelial cells and neurons, bindsstrongly to PDZ domains of the scaffold proteins PSD95and Drosophila discs-large [51]. These interactions sug-gest that scaffolding proteins play a central role in theorganization and regulation of Ca2+ signaling complexes.


COMMUNICATION BETWEEN THEENDOPLASMIC RETICULUM AND THE PLASMAMEMBRANE

Another aspect of the organization of Ca2+-signalingcomplexes is the communication between Ca2+ releasechannels in the endoplasmic reticulum (ER) and Ca2+

influx channels in the plasma membrane (PM). Ca2+ sig-naling commences upon Ca2+ release from the ER andleads to the subsequent activation of Ca2+ influx acrossthe PM. Gating of the Ca2+-entry pathway by the Ca2+

content in the ER is termed capacitative Ca2+ entry [11].The nature of the Ca2+-entry pathway and how it sensesCa2+ content in the stores are the least understoodaspects of Ca2+ signaling.

Ca2+ entry is mediated by a Ca2+ channel called ICRAC

[11,12], which may be one or a combination of severalisoforms of the store-operated human homologues of theDrosophila TRP channel (hTRP) [11,52,53]. The way inwhich ICRAC senses Ca2+ content of the stores is currentlythe subject of intense investigation. Recent work sug-gests that Ca2+ influx is regulated by an exocytosis-likemechanism [54,55]. Hence, modulators of regulated exo-cytosis, such as C3 transferase, BoNT A toxin andSNAP25, inhibited activation of ICRAC in Xenopus oocytes[54] and solidification of the actin cytoskeleton inhibitedICRAC activation in smooth muscle cell lines [55].However, in both studies complete disassembly of theaction cytoskelaton, which is known to modulate regu-lated exocytosis in many cell types, had no effect on acti-vation of Ca2+ influx by store depletion [54,55]. Hence, itis possible that proteins which regulate exocytosis deter-mine the availability of ICRAC channels in the PM and/orthe proximity of the ER to the PM over relatively longperiods of time, rather than acutely regulate ICRAC inser-tion in the plasma membrane as a result of store deple-tion.

Another mechanism for gating of ICRAC, which is notmutually exclusive with regulation by exocytosis, is gat-ing by conformational coupling. The inability to identifya soluble messenger that regulates ICRAC and the exis-tence of subplasmalemal domains expressing high levelsof IP3Rs lead Irvine [56] and Berridge [57] to postulateregulation of ICRAC by IP3R. This mode of gating is analo-gous to EC coupling in skeletal muscle [58], and becameknown as the conformational coupling hypothesis.Recent work provided experimental support for thismode of gating. We showed that IP3Rs interact with andgate the store-operated hTRP3 channels [59]. The N-ter-minal domain of the IP3R was sufficient for activation ofhTRP3, but full gating (activation upon depletion of Ca2+

stores and inhibition when the Ca2+ stores are full)requires full-length IP3R [60]. Regulation by conforma-tional coupling is probably due to direct interaction

Cell Calcium (1999) 26(5), 173–180


between the IP3R and hTRP. The two channels can be co-immunoprecipitated [60] and domains in IP3R and hTRPthat interact with each other were found to regulate Ca2+

influx [61]. Regulation by conformational coupling canalso be demonstrated with the native Ca2+ selective chan-nel Imin present in patches excised from the PM of A431cells [62]. All single channel properties of Imin that wereexamined were found to be similar to those reported forICRAC [63], suggesting that native ICRAC channels are regu-lated by conformational coupling. The significance ofconformational coupling is that the interaction of theIP3Rs with ICRAC may help to assemble Ca2+ signalingcomplexes. IP3Rs may be anchored to the cytoskeletonthrough interaction with a scaffolding protein likeHomer [50] or only through their binding to ICRAC. If ICRAC

is related to the TRP channels it may bind to thecytoskeleton through ankirin-binding domains, whichare found in the N-terminal domain of the TRP channels[11,52]. In any case, the interaction between IP3Rs andICRAC recruits the ER component of Ca2+ signaling pro-teins to the signaling complex at the plasma membrane.

RGS PROTEINS AND G PROTEIN-MEDIATED Ca2+

SIGNALING

Agonist binding to GPCR at the plasma membrane initi-ates Ca2+ signaling transduced by heterotrimeric G pro-teins. As with all mammalian cell-types, polarizedepithelial cells express two classes of G proteins thatmediate Ca2+ signaling via activation of PLCβ to produceIP3. All isoforms of PLCβ are activated by Gq class α sub-units [64–66]. PLCβ2 and PLCβ3 are also activated byGβγ subunits, primarily released from Gi class G proteins[67], which is the basis for pertussis toxin inhibition ofPLCβ activity and Ca2+ signaling evoked by Gi-coupledreceptors [68]. In addition, recent work showed that thespecificity, intensity and duration of Gq and Gi signalingis controlled by a family of GTPase activating proteins(GAPs), the so-called RGS proteins [69–71]. Specificity inthe interactions between these signaling components is acritical issue for appropriate cellular responses to a multi-tude of stimuli.

GPCR AND RGS PROTEINS REGULATE GTP-BINDING AND HYDROLYSIS ON Gα

G proteins act like a molecular switch; in the inactive orbasal state, the α subunit binds GDP, and in the activestate it binds GTP. Agonist-bound GPCRs activate G pro-teins by promoting the binding of GTP. GTP bindinginduces conformational changes in the Gα subunit thatdrives dissociation of Gβγ and receptor and the associa-tion of Gα–GTP and/or Gβγ with effector proteins [8].

Cell Calcium (1999) 26(5), 173–180

Thus, information is transduced from extracellular ligandbinding to effector proteins which regulate [Ca2+]i andother second messengers inside the cell.

RGS proteins appear to have multiple roles in control-ling Ca2+ signaling. Genetic and biochemical studies ini-tially suggested that RGS proteins could attenuate Gprotein signaling by accelerating GTP hydrolysis on Gαsubunits [69,72,73]. Mammals express at least 25 RGSproteins, most of which are Gq and/or Gi–GAPs[18,72,74]. Several studies demonstrated that RGS pro-teins could inhibit Ca2+ signaling evoked by either Gi- orGq-coupled receptors [75–78]. However, RGS proteinsappear to encompass more than the simple role ofsignaling attenuators because they can also promote the assembly and operation of G protein signalingcomplexes.

The scaffolding functions of RGS proteins were firstappreciated by Doupnik et al. who found that RGS4expressed in Xenopus oocytes accelerated both the inacti-vation of the Gi-coupled G protein-regulated inwardrectifying K+ (GIRK) channel and, surprisingly, agonist-dependent GIRK activation [79]. Their interpretation wasthat RGS4 brought Gβγ within proximity of its effectortarget, the GIRK channel, to accelerate the kinetics ofboth signaling activation and inactivation. Similar resultswere observed with the related protein, RGS8 [80].Another interpretation was that RGS proteins mightenhance the rate of Gβγ dissociation from Gα–GTP [81],thus accelerating the kinetics of GIRK activation.

Recombinant RGS4 and closely related RGS proteinsalso exhibit both scaffolding and GAP activity whenintroduced into mammalian cells through a patch-pipette. Different domains of the RGS protein convey sep-arate activities. All RGS proteins have a commonstructural feature of about 125 amino acids, referred toas the RGS box [73]. Amino-acid identity within the RGSbox of the Gi/Gq GAPs ranges between 30 and 90%. TheRSG box is flanked by N- and C-termini that appear to beconserved within subfamilies of RGS proteins [73,82].Comparisons of GAP activity from several RGS proteinsshowed that the RGS box and full-length proteins simi-larly accelerated GTP hydrolysis on Gi and Gqα subunitsin a single turnover assay [83]. However, in intact cells,the RGS box was a 10 000-fold less potent inhibitor ofGq-coupled signaling then was full length RGS4 [75]. TheN- and C-terminal sequences that flank the RGS box arerequired for full activity in cells and probably determineRGS specificity within the signaling complex.

The N-termini of RGS4 and RGS16 have two functions,membrane targeting [84,85] and interaction with thereceptor-complex [75]. Full-length RGS proteins regu-lated Gq-signaling in a receptor-selective manner [76],whereas the RGS box alone did not discriminate betweenGPCR-types [75]. Interestingly, an N-terminal peptide of



33 amino acids from RGS4 or RGS16 conveyed receptor-selective inhibition of Gq-signaling [75]. Furthermore,when added together into cells, the N-terminal peptideand RGS box acted synergistically to partially restore theactivity of full-length RGS protein. Whereas the N-termi-nus affects both the mechanism and potency of RGS4action, the C-terminus appears to increase the local con-centration of RGS4 in the vicinity of the receptor-com-plex [75]. Thus, RGS4 and RGS16 regulate Gq classsignaling by the combined action of the RGS box, whichaccelerates GTP hydrolysis on Gαq, and the N- and C-ter-mini, which convey high-affinity and receptor-selectiveinhibition of Gq signaling (modeled in Fig. 1).

The simplest mechanism for receptor-selective regula-tion is protein–protein interaction between receptors andthe N-termini of the RGS4 and RGS16. However, directreceptor binding is not sufficient to explain all RGS4activities that have been detected in the signaling com-plex, which may include receptor, Gqαβγ, effector, lipids,RGS protein and perhaps unknown component(s) (Fig. 1).It appears that selectivity of RGS action within the signal-ing complex provides a mechanism for intracellular regu-lation of signaling specificity.

REGULATION OF RGS PROTEINS CAN INITIATECa2+ OSCILLATIONS

Episodic RGS activity, dependent on feedback regulationwithin the Gq-receptor complex, may provide a biochemi-cal pathway for the generation of Ca2+ oscillations. Ca2+

oscillations are considered the physiologically relevantform of all Ca2+ signaling in cells controlling basic anddiverse functions such as progression through the cellcycle, cell differentiation, release of neurotransmitters, andactivation of transcription factors [11]. Ca2+ oscillations areassumed to be a purely biophysical phenomenon, basedon the discovery of the bell-shape dependence of IP3Rchannel activity on cytoplasmic Ca2+ [86] and the observa-tion that non-hydrolyzable analogs of IP3 at constant con-centration stimulated Ca2+ oscillations [87]. However,Hirose et al. recently observed that a Gq-coupled agonistevoked oscillations in [IP3]i synchronous with Ca2+ oscilla-tions [88]. In another study, Ito et al. compared the rate ofCa2+ release and of Ca2+ wave propagation, evoked byeither maximal agonist stimulation or uncaging high con-centrations of IP3 [43]. These studies showed that the rateof wave propagation was much slower than the rate of IP3-mediated Ca2+ release. This suggests that biochemicalevents at the receptor-complex determine the rate of IP3

production to control Ca2+ wave propagation [43].What is the biochemical mechanism that initiates oscil-

lations in IP3? PLCβ is a plausible regulatory targetbecause the product of its enzymatic activity is IP3.However, the regulatory mechanism need not act directly


on PLCβ. We propose that feedback regulation of RGSproteins within the Gq–receptor complex may producepulses of IP3 to initiate Ca2+ oscillations in cells [75,76].Preliminary observations from our laboratories usingwild-type and mutant forms of RGS4 support a role forRGS proteins in regulating Ca2+ oscillations.

An oscillator requires an on-switch and an off-switchsensitive to feedback regulation. In most models of [Ca2+]i

oscillations both switches are placed in the regulation ofthe IP3R by IP3 and low [Ca2+]i (on switch) and high [Ca2+]i

(off switch). However, the results of Ito et al.[43] and ofHirose et al. [88] suggest that biochemical events in thereceptor complex generate oscillations in IP3 to govern[Ca2+]i oscillations. RGS regulation of G proteins, and thusIP3 levels, can provide an alternative mechanism to initi-ate [Ca2+]i oscillations. In this mechanism, hormone bind-ing provides the on-switch by promoting GTP binding toGαq. The traditional view is that the G protein activitycycle proceeds from Gα–GTP binding to subunit dissoci-ation, effector activation and GTP hydrolysis, ready totransit another round of the cycle with persistent agoniststimulation. We propose that RGS proteins could uncou-ple receptor catalyzed loading of Gqα–GTP from PLCβactivation, even in the presence of persistent agonist[75,77]. In one scenario, the RGS proteins tonically inhibitGq in the receptor complex. Activation of the receptor bylow agonist concentration results in periodic inhibition ofRGS activity to generate pulsatile changes in [IP3]. Withincreasing agonist concentrations, the activity of the RGSproteins is further inhibited to increase the frequency ofIP3 production and, thus, [Ca2+]i oscillations. At saturatingagonist concentration the activity of the RGS proteins iscompletely inhibited, resulting in a sustained [Ca2+]i

response. In another model, the RGS proteins in the Gq-receptor complex are initially inactive, which allowsPLCβ activation upon agonist binding to receptor,Periodic activation of RGS protein GAP activity, perhapsby feedback regulation, could periodically terminate pro-duction of IP3 and Ca2+ release, resulting in oscillatorychanges in IP3 and Ca2+ concentrations over time. At highstimulus intensity, activated Gqα may overwhelm theability of RGS proteins to inhibit signaling. The role ofRGS proteins in controlling [Ca2+]i oscillations willundoubtedly be explored very vigorously in the future.In summary, proximity of the Gq-receptor complex, IP3

regulated intracellular Ca2+ stores, and ICRAC, perhaps co-localized by direct intermolecular binding and/or via aseries of scaffold proteins, may precisely regulate cellularresponses to local changes in [Ca2+]i.

REFERENCES

1. Petersen OH. Stimulus-secretion coupling: cytoplasmic calciumsignals and the control of ion channels in exocrine acinar cells.J Physiol 1992; 448: 1–51.

Cell Calcium (1999) 26(5), 173–180


2. Arvan P, Castle D. Sorting and storage during secretory granulebiogenesis: looking backward and looking forward. Biochem J1998; 332: 593–610.

3. Geppert M, Sudhof TC. Rab3 and synaptotagmin: the yin andyang of synaptic membrane fusion. Annu Rev Neurosci 1998;21: 75–95.

4. Berridge MJ. Inositol trisphosphate and calcium signaling.Nature 1993; 361: 315–325.

5. Kasai H, Li YX, Miyashita Y. Subcellular distribution of Ca2+

release channels underlying Ca2+ waves and oscillations inexocrine pancreas. Cell 1993; 74: 669–677.

6. Thorn P, Lawrie AM, Smith PM, Gallacher DV, Petersen OH.Local and global cytosolic Ca2+ oscillations in exocrine cellsevoked by agonists and inositol trisphosphate. Cell 1993; 74:661–668.

7. Xu X, Zeng W, Diaz J, Muallem S. Spacial compartmentalizationof Ca2+ signaling complexes in pancreatic acini. J Biol Chem1996; 271: 24684–24690.

8. Gudermann T, Schoneberg T, Schultz G. Functional andstructural complexity of signal transduction via G-protein-coupled receptors. Annu Rev Neurosci 1997; 20: 399–427.

9. Muallem S. The ins and outs of Ca2+ in exocrine cells. In:Putney JW Jr, (Ed.) Advances in Second Messenger andPhosphoprotein Research: Inositol Phosphates and CalciumSignaling. Raven Press, New York 1992; 26: 349–366.

10. Muallem S, Pandol SJ, Beeker TG. Hormone-evoked calciumrelease from intracellular stores is a quantal process. J BiolChem 1989; 264: 205–212.

11. Putney JW Jr, McKay RR. Capacitative calcium entry channels.Bioessays 1999; 21: 38–46.

12. Parekh AB, Penner R. Store depletion and calcium influx.Physiol Rev 1997; 77: 901–930.

13. Zhang B-X, Zhao H, Loessberg PA, Muallem S. Activation of theplasma membrane Ca2+-pump during agonist stimulation ofpancreatic acini. J Biol Chem 1992; 267: 15419–15425.

14. Zhang B-X, Muallem S. Feedback inhibition of Ca2+ release byCa2+ is the underlying mechanism of agonist-evokedintracellular Ca2+ oscillations in pancreatic acinar cells. J BiolChem 1992; 267: 24387–24393.

15. Hajnoczky G, Thomas AP. Minimal requirements for Ca2+

oscillations driven by the IP3 receptor. EMBO J 1997; 16:3533–3543.

16. Hofer AM, Landolfi B, Debellis L, Pozzan T, Curci S. Free [Ca2+]dynamics measured in agonist-sensitive stores of single livingintact cells: a new look at the refilling process. EMBO J 1998;17: 1986–1995.

17. Wilkie TM, Gilbert DJ, Olsen AS et al evolution of themammalian G Protein α subunit multigene family. Nature Genet1992; 1: 85–91.

18. Berman DM, Gilman AG. Mammalian RGS proteins: barbariansat the gate. J Biol Chem 1998; 273: 1269–1272.

19. De Vries L, Gist Farquhar M. RGS proteins: more than just GAPsfor heterotrimeric G proteins. Trends Cell Biol 1999; 9:138–144.

20. Carafoli E, Garcia-Martin E, Guerini D. The plasma membranecalcium pump: recent developments and future perspectives.Experientia 1996; 52: 1091–1100.

21. Penniston JT, Enyedi A, Verma AK, Adamo HP, Filoteo AG.Plasma membrane Ca2+ pumps. Ann NY Acad Sci 1997; 834:56–64.

22. MacLennan DH, Rice WJ, Green NM. The mechanism of Ca2+

transport by sarco(endo)plasmic reticulum Ca2+-ATPases. J BiolChem 1997; 272: 28815–28818.

23. Guerini D. The Ca2+ pumps and the Na+/Ca2+ exchangers.Biometals 1998; 11: 319–330.

Cell Calcium (1999) 26(5), 173–180

24. Lee MG, Xu X, Zeng W et al. Polarized expression of Ca2+ pumps inpancreatic and salivary gland cells. Role in initiation andpropagation of [Ca2+]i waves. J Biol Chem 1997; 272: 15771–15776.

25. Patel S, Joseph SK, Thomas AP. Molecular properties of inositol1,4,5-trisphosphate receptors. Cell Calcium 1999; 25: 247–264.

26. Marks AR. Intracellular calcium-release channels: regulators ofcell life and death. Am J Physiol 1997; 272: H597–H605.

27. Lee MG, Xu X, Zeng W et al. Polarized expression of Ca2+

channels in pancreatic and salivary gland cells. Correlationwith initiation and propagation of [Ca2+]i waves. J Biol Chem1997; 272: 15765–15770.

28. Thorn P, Gerasimenko O, Petersen OH. Cyclic ADP-riboseregulation of ryanodine receptors involved in agonist evokedcytosolic Ca2+ oscillations in pancreatic acinar cells. EMBO J1994; 13: 2038–2043.

29. Lee HC. Mechanisms of Ca2+ signaling by cyclic ADP-ribose andNAADP. Physiol Rev 1997; 77: 1133–1164.

30. Philipp S, Flockerzi V. Molecular characterization of a novelhuman PDZ domain protein with homology to INAD fromDrosophila melanogaster. FEBS Lett 1997; 413: 243–248.

31. Ullmer C, Schmuck K, Figge A, Lubbert H. Cloning andcharacterization of MUPP1, a novel PDZ domain protein. FEBSLett 1998; 424: 63–68.

32. Totorici G, Zhang B-X, Xu X, Muallem S. Compartmentalizationof Ca2+ signaling and Ca2+ pools in pancreatic acini: Implicationfor the quantal behavior of Ca2+ release. J Biol Chem 1994; 269:29621–26928.

33. Parys JB, Missiaen L, Smedt HD, Sienaert I, Casteels R.Michanisms responsible for quantal Ca2+ release from inositoltrisphosphate-sensitive calcium stores. Pflugers Arch 1996; 432:359–367.

34. Muallem S, Schoeffield MS, Fimmel CJ, Pandol SJ. Agonist-sensitivecalcium pool in the pancreatic acinar cell. II. Characterization ofreloading. Am J Physiol 1988; 255: G229–G235.

35. Muallem S, Pandol SJ, Beeker TG. Calcium mobilizinghormones activate the plasma membrane Ca2+ pump ofpancreatic acinar cells. J Membr Biol 1988; 106: 57–69.

36. Parker I, Choi J, Yao Y. Elementary events of the InsP3-inducedCa2+ liberation in Xenopus oocytes: hot spots, puffs and blips.Cell Calcium 1996; 20: 105–121.

37. Parker I, Zang WJ, Wier WG. Ca2+ sparks involving multiple Ca2+

release sites along Z-lines in rat heart cells. J Physiol (Lond)1996; 497: 31–38.

38. Neher E. Vesicle pools and Ca2+ microdomains: new tools forunderstanding their roles in neurotransmitter release. Neuron1998; 20: 389–399.

39. Muallem S, Lee MG. High [Ca2+]i domains, secretory granulesand exocytosis. Cell Calcium 1997; 22: 1–4.

40. Kasai H, Augustine GJ. Cytosolic Ca2+ gradients triggeringundirectional fluid secretion from exocrine pancreas. Nature1990; 348: 735–738.

41. Nathanson MH, Fallon MB, Padfield PJ, Maranto AR.Localization of the type 3 inositol 1,4,5-trisphosphate receptorin the Ca2+ wave trigger zone of pancreatic acinar cells. J BiolChem 1994; 269: 4693–4696.

42. Takemura H, Yamashina S, Segawa A. Millisecond analyses ofCa2+ initiation sites evoked by muscarinic receptor stimulationin exocrine acinar cells. Biochem Biophys Res Commun 1999;259: 656–660.

43. Ito K, Miyashita Y, Kasai H. Kinetic control of multiple forms ofCa2+ spikes by inositol triphosphate in pancreatic acinar cells. JCell Biol 1999; 146: 405–413.

44. Barak LS, Ferguson SS, Zhang J, Caron MG. A β-arrestin/greenfluorescent protein biosensor for detecting G protein-coupledreceptor activation. J Biol Chem 1997; 272: 27497–27500.



45. Stoffel RH, Pitcher JA, Lefkowitz RJ. Targeting G protein-coupledreceptor kinases to their receptor substrates. J Membr Biol 1997;157: 1–8.

46. Luttrell LM, Ferguson SS, Daaka Y et al. β-arrestin-dependentformation of β2 adrenergic receptor-Src protein kinasecomplexes. Science 1999; 283: 655–661.

47. Hall RA, Premont RT, Chow CW, et al. The β2-adrenergicreceptor interacts with the Na+/H+-exchanger regulatory factorto control Na+/H+ exchange. Nature 1998; 392: 626–630.

48. Hall RA, Ostedgaard LS, Premont RT et al. A C-terminal motiffound in the β2-adrenergic receptor, P2Y1 receptor and cysticfibrosis transmembrane conductance regulator determinesbinding to the Na+/H+ exchanger regulatory factor family ofPDZ proteins. Proc Natl Acad Sci USA 1998; 95: 8496–8501.

49. Xiao B, Tu JC, Petralia RS et al. Homer regulates the associationof group 1 metabotropic glutamate receptors with multivalentcomplexes of homer-related, synaptic proteins. Neuron 1998;21: 707–716.

50. Tu JC, Xiao B, Yuan JP et al. Homer binds a novel proline-richmotif and links group 1 metabotropic glutamate receptors withIP3 receptors. Neuron 1998; 21: 717–726.

51. Kim E, DeMarco SJ, Marfatia SM, Chishti AH, Sheng M, StrehlerEE. Plasma membrane Ca2+ ATPase isoform 4b binds tomembrane-associated guanylate kinase (MAGUK) proteins viatheir PDZ (PSD-95/Dlg/ZO-1) domains. J Biol Chem 1998; 273:1591–1595.

52. Birnbaumer L, Zhu X, Jiang M et al. On the molecular basis andregulation of cellular capacitative Ca2+ entry: roles for Trpproteins. Proc Natl Acad Sci USA 1996; 93: 15195–15202.

53. Kiselyov K, Muallem S. Fatty acids, diacylglycerol, Ins(1,4,5)P3

receptors and Ca2+ influx. Trends Neurosci 1999 22: 334–337.54. Yao Y, Ferrer-Montiel AV, Montal M, Tsien RY. Activation of

store-operated Ca2+ current in xenopus oocytes requires SNAP-25 but not a diffusible messenger. Cell 1999; 98: 475–485.

55. Patterson RL, van Rossum DB, Gill DL. Store-operated Ca2+

entry: evidence for a secretion-like coupling model. Cell 1999;98: 487–499.

56. Irvine RF. ‘Quantal’ Ca2+ release and the control of Ca2+ entryby inositol phosphates – a possible mechanism. FEBS Lett 1990;263: 5–9.

57. Berridge MJ. Capacitative calcium entry. Biochem J 1995; 312:1–11.

58. Franzini-Armstrong C, Protasi F. Ryanodine receptors of striatedmuscles: a complex channel capable of multiple interactions.Physiol Rev 1997; 77: 699–729.

59. Kiselyov K, Xu X, Mozhayeva G et al. Functional Interactionbetween InsP3 receptors and store-operated h Trp3 channels.Nature 1998; 396: 478–482.

60. Kiselyov K, Mignery GA, Zhu MX, Muallem S. The N-terminaldomain of the IP3 receptor gates store-operated hTrp3 channels.Molecular Cell 1999; 4: (in press).

61. Vannier B, Peyton M, Boulay G et al. Mouse trp2, thehomologue of the human trpc2 pseudogene, encodes mTrp2, astore depletion-activated capacitative Ca2+ entry channel. ProcNatl Acad Sci USA 1999; 96: 2060–2064.

62. Zubov AI, Kaznacheeva EV, Nikolaev AV et al. Regulation of theminiature plasma membrane Ca2+ channel Imin by inositol1,4,5-trisphosphate receptors. J Biol Chem 1999; 274:25983–25985.

63. Kerschbaum HH, Cahalan MD. Single-channel recording of astore-operated Ca2+ channel in Jurkat T lymphocytes. Science1999; 283: 836–839.

64. Taylor SJ, Chae HZ, Rhee SG, Exton JH. Activation of the β-1isozyme of phospholipase C by α subunits of the Gq class of Gproteins. Nature 1991; 350: 516–518.


65. Smrcka AV, Hepler JR, Brown KO, Sternweis PC. Regulation ofpolyphosphoinositide-specific phospholipase C activity bypurified Gq. Science 1991; 251: 804–807.

66. Wu DQ, Lee CH, Rhee SG, Simon MI. Activation ofphospholipase C by the α subunits of the Gq and G11 proteinsin transfected Cos-7 cells. J Biol Chem 1992; 267: 1811–1817.

67. Sternweis, PC. The active role of Gβγ in signal transduction.Curr Opin Cell Biol 1994; 6: 198–203.

68. Simon MI, Strathmann MP, Gautam N. Diversity of G proteinsin signal transduction. Science 1991; 252: 802–808.

69. Berman DM, Wilkie TM, Gilman AG. GAIP and RGS4 areGTPase-activating proteins for the Gi subfamily of G protein αsubunits. Cell 1996; 86: 445–452.

70. Watson N, Linder ME, Druey KM, Kehrl JH, Blumer KJ. RGSfamily members: GTPase-activating proteins for heterotrimericG-protein α-subunits. Nature 1996; 383: 172–175.

71. Hunt TW, Fields TA, Casey PJ, Peralta EG. RGS10 is a selectiveactivator of G α i GTPase activity. Nature 1996; 383: 175–177.

72. Dohlman HG, Thorner J. RGS proteins and signaling byheterotrimeric G proteins. J Biol Chem 1997; 272: 3871–3874.

73. Koelle MR, Horvitz HR. EGL-10 regulates G protein signaling inthe C. elegans nervous system and shares a conserved domainwith many mammalian proteins. Cell 1996; 84: 115–125.

74. Koelle MR. A new family of G-protein regulators – the RGSproteins. Curr Opin Cell Biol 1997; 9: 143–147.

75. Zeng W, Xu X, Popov S, et al. The N-terminal domain of RGS4confers receptor-specific inhibition of G-protein signaling. J BiolChem 1998; 273: 34687–34690.

76. Xu X, Zeng W, Popov S et al. RGS4 determines signalingspecificity of Gq-coupled receptors. J Biol Chem 1999; 274:3549–3556.

77. Luo X, Zeng W, Xu X et al. Alternate coupling of receptors to Gsand Gi in pancreatic and submandibular gland cells. J BiolChem 1999; 274: 17684–17690.

78. Saugstad JA, Marino MJ, Folk JA, Hepler JR, Conn PJ. RGS4inhibits signaling by group I metabotropic glutamate receptors.J Neurosci 1998; 18: 905–913.

79. Doupnik CA, Davidson N, Lester HA, Kofuji P. RGS proteinsreconstitute the rapid gating kinetics of Gβγ- activated inwardlyrectifying K+ channels. Proc Natl Acad Sci USA 1997; 94:10461–10466.

80. Saitoh O, Kubo Y, Miyatani Y, Asano T, Nakata H. RGS8accelerates G-protein-mediated modulation of K+ currents.Nature 1997; 390: 525–529.

81. Chuang H, Yu M, Jan YN, Jan LY. Evidence that the nucleotideexchange and hydrolysis cycle of G proteins causes acutedesensitization of G-protein gated inward rectifier K+ channels.Proc Natl Acad Sci USA 1998; 95: 11727–11732.

82. Zheng B, De Vries L, Farquhar MG. Divergence of RGS proteins:evidence for the existence of six mammalian RGS subfamilies.Trends Bio Sci 1999; (in press).

83. Popov S, Yu K, Kozasa T, Wilkie TM. The regulators of G proteinsignaling (RGS) domains of RGS4, RGS10, and GAIP retainGTPase activating protein activity in vitro. Proc Natl Acad SciUSA 1997; 94: 7216–7220.

84. Srinivasa SP, Bernstein LS, Blumer KJ, Linder ME. Plasmamembrane localization is required for RGS4 function inSaccharomyces cerevisiae. Proc Natl Acad Sci USA 1998; 95:5584–5589.

85. Chen C, Seow KT, Guo K, Yaw LP, Lin SC. The membraneassociation domain of RGS16 contains unique amphipathicfeatures that are conserved in RGS4 and RGS5. J Biol Chem1999; 274: 19799–19806.

86. Bezprozvanny I, Watras J, Ehrlich BE, Bell-shaped calcium-response curves of Ins(1,4,5)P3- and calcium-gated channels

Cell Calcium (1999) 26(5), 173–180


from endoplasmic reticulum of cerebellum. Nature 1991; 351:751–754.

87. Wakui M, Potter BV, Petersen OH. Pulsatile intracellular calciumrelease does not depend on fluctuations in inositoltrisphosphate concentration. Nature 1989; 339: 317–320.

Cell Calcium (1999) 26(5), 173–180

88. Hirose K, Kadowaki S, Tanabe M, Takeshima H, Iino M.Spatiotemporal dynamics of inositol 1,4,5-trisphosphate thatunderlies complex Ca2+ mobilization patterns. Science 1999;284: 1527–1530.


g protein-dependent ca2+signaling complexes in polarized cells

Documents