Inhibition by Calcium of Mammalian Adenylyl Cyclases*

(Received for publication, June 7, 1999, and in revised form, September 2, 1999)

Jean-Louis Guillou‡§, Hiroko Nakata‡, and Dermot M. F. Cooper¶

From the Department of Pharmacology, University of Colorado Health Sciences Center, Denver, Colorado 80262

Ca21 regulates mammalian adenylyl cyclases in atype-specific manner. Stimulatory regulation is moder-ately well understood. By contrast, even the concentra-tion range over which Ca21 inhibits adenylyl cyclasesAC5 and AC6 is not unambiguously defined; even less sois the mechanism of inhibition. In the present study, wecompared the regulation of Ca21-stimulable and Ca21-inhibitable adenylyl cyclases expressed in Sf9 cells withtissues that predominantly express these activities inthe mouse brain. Soluble forms of AC5 containing eitherintact or truncated major cytosolic domains were alsoexamined. All adenylyl cyclases, except AC2 and the sol-uble forms of AC5, displayed biphasic Ca21 responses,suggesting the presence of two Ca21 sites of high (;0.2mM) and low affinity (;0.1 mM). With a high affinity, Ca21

(i) stimulated AC1 and cerebellar adenylyl cyclases, (ii)inhibited AC6 and striatal adenylyl cyclase, and (iii) waswithout effect on AC2. With a low affinity, Ca21 inhib-ited all adenylyl cyclases, including AC1, AC2, AC6, andboth soluble forms of AC5. The mechanism of both highand low affinity inhibition was revealed to be competi-tion for a stimulatory Mg21 site(s). A remarkable selec-tivity for Ca21 was displayed by the high affinity site,with a Ki value of ;0.2 mM, in the face of a 5000-foldexcess of Mg21. The present results show that high andlow affinity inhibition by Ca21 can be clearly distin-guished and that the inhibition occurs type-specificallyin discrete adenylyl cyclases. Distinction between thesesites is essential, or quite spurious inferences may bedrawn on the nature or location of high affinity bindingsites in the Ca21-inhibitable adenylyl cyclases.

Profound physiological significance derives from the regula-tion of adenylyl cyclase by Ca21, which provides a confluence oftwo major signaling pathways. For instance, Ca21 stimulationof adenylyl cyclase has been implicated in learning-memoryfunctions (1–3). Compelling evidence for this assertion comesfrom studies with Aplysia (4) and with Drosophila mutants (5)and more recently mice that have had adenylyl cyclase genesdeleted (6). On the other hand, Ca21 inhibition of adenylylcyclase has been proposed to contribute to oscillations and/orpacemaking in cardiac tissue (7) and the maintenance of endo-thelial cell permeability (8). A central issue in implicating theCa21 responsiveness of an adenylyl cyclase in a physiologicalprocess concerns the correspondence between the concentra-

tions of Ca21 achieved in response to physiological stimuliversus those required to regulate these adenylyl cyclases invitro.1 This issue seems moot with Ca21-stimulable adenylylcyclases, which are stimulated by Ca21 in vitro and are alsostimulated in response to various physiological means of ele-vating Ca21 in vivo (2, 7, 9). Such correlations are mutuallysupportive. The situation with Ca21-inhibitable adenylyl cycla-ses is more complex. The averaged cytosolic concentration ofCa21 achieved in intact cells upon triggering either capacita-tive- or voltage-gated Ca21 entry (;1 mM) corresponds to theconcentrations reported to be effective at inhibiting these ad-enylyl cyclases in some in vitro studies (10). However, a widearray of in vitro inhibitory sensitivities have been reportedfrom various sources (10–25). For instance, both monophasic(11–13, 21, 23) and biphasic (10, 14–19, 24) inhibition by Ca21

have been reported, spanning sub- to supramicromolar concen-tration ranges.

Some of the discrepancies may emanate from somewhat un-controlled assay conditions, e.g. with respect to pH or freeEGTA concentrations, which will confound estimates of freeCa21 (26). However, mixed populations of adenylyl cyclases inselected tissues may also confound analysis. The nine isoformsof mammalian adenylyl cyclases that have been cloned to date(7, 27) possess distinct characteristics including their sensitiv-ity to Ca21, which allows them to be classified as (i) Ca21-stimulated (AC1, AC8, and, possibly, AC3), (ii) Ca21-insensi-tive (AC2, AC4, and AC7), and (iii) Ca21-inhibited (AC5 andAC6; Refs. 7, 28, and 29). Heterogeneous preparations of suchadenylyl cyclases could give rise to a variety of Ca21 concen-tration responses. There is also the issue of low affinity Ca21

inhibition, which appears to be a property of all adenylyl cy-clases. This inhibition is believed to reflect competition by Ca21

for an allosteric regulatory site for Mg21 (16, 17, 20–22), al-though the precise mechanism is unknown. This low affinityinhibition by Ca21 can be confounded with high affinity inhi-bition, particularly if Ca21 concentrations are not rigorouslyestablished or controlled (26).

Ideally, well controlled, direct comparisons of expressed ad-enylyl cyclase isoforms should resolve these ambiguities. In-deed, in membranes from transfected HEK 293 cells, AC1 andAC8 were stimulated by submicromolar concentrations ofCa21, while they displayed inhibition in response to higherconcentrations, mirroring what happens in most brain tissues(9). However, again, the situation with the Ca21-inhibitableadenylyl cyclases is less clear. A monophasic inhibition by Ca21

spanning sub- to supramicromolar concentrations was associ-ated with the first description of canine AC5 (25). A secondreport on rabbit AC5 indicated an inhibition by very low sub-micromolar (suprananomolar) Ca21 concentrations (30). In thefirst description of AC6, its activity in transfected cell mem-

* This work was supported by National Institutes of Health GrantGM 32483 (to D. M. F. C.). The costs of publication of this article weredefrayed in part by the payment of page charges. This article musttherefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

‡ These authors contributed equally to this work.§ Supported by the Foundation FYSSEN (Paris).¶ To whom correspondence should be addressed: Dept. of Pharmacol-

ogy, Campus Box C-236, University of Colorado Health Sciences Cen-ter, 4200 E. Ninth Ave., Denver, CO 80262. Tel.: 303-315-89-64; Fax:303-315-70-97; E-mail: dermot.cooper@uchsc.edu.

1 Earlier biochemical studies had established that the in vitro effect ofCa21 was direct, readily reversible, and not secondary to phosphoryla-tion or a variety of potentially Ca21-dependent processes (18, 29,60, 61).

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 274, No. 50, Issue of December 10, pp. 35539–35545, 1999© 1999 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 35539

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

branes was inhibited by submicromolar concentrations of Ca21

(31), although higher concentrations were not explored. Re-cently, a chimeric, soluble form of AC5, comprising its fusedcytosolic (C1 and C2) domains was reported to display highaffinity inhibition by Ca21, which was lost upon truncation ofthe C1 region (32). Furthermore, AC2, which is classified as aCa21-insensitive adenylyl cyclase, displays inhibition in re-sponse to high concentrations of Ca21. It therefore seems es-sential to clearly define the effective concentration ranges forCa21 inhibition of individual adenylyl cyclases.

Consequently, in the present study, we compared the effectsof a comprehensive range of Ca21 concentrations on adenylylcyclase activity from a variety of sources. Adenylyl cyclaseactivity in membranes of Sf9 cells expressing AC1, AC2, andAC6 was compared with adenylyl cyclase activity of two se-lected brain tissues that express respectively AC5 (the stria-tum; Refs. 33 and 34) and AC1 (the cerebellum; Ref. 35). Also,the full-length and truncated forms of the AC5 C1/C2 chimerawere assessed. Unequivocal evidence is provided demonstrat-ing that, with a high affinity, Ca21 inhibits AC6 (and striatalAC5), stimulates AC1 (and cerebellar AC1), and does not mod-ulate AC2. In the supramicromolar concentration range, alladenylyl cyclases tested showed the same low affinity inhibi-tion by Ca21. Further analyses reveal that both the selectivehigh affinity inhibition occurring in AC6 (and AC5) and the lowaffinity inhibition occurring in all adenylyl cyclase isoformsinvolve competitive mechanisms with Mg21 activation of theenzyme. As for the two soluble forms of AC5, our results showthat both chimeras display only low affinity inhibition and thatthe site of high (and low) affinity inhibition must be sought byother experimental approaches.

EXPERIMENTAL PROCEDURES

Materials—Forskolin was from Calbiochem. [a-32P]ATP and[3H]cAMP were from Amersham Pharmacia Biotech, and other re-agents were from Sigma.

Expression of AC1, AC2, and AC6 in Sf9 Cells—Recombinant bacu-loviruses encoding AC1 and AC2 were generously provided by Drs. A. G.Gilman (University of Texas Southwestern Medical Center, Dallas) andR. Iyengar (Mt. Sinai Medical School, New York), respectively (36, 37).For construction of recombinant baculovirus encoding AC6, the entirecoding region of mouse AC6 was engineered by polymerase chain reac-tion into the baculovirus vector, pBlueBacHis2, between KpnI and SalIrestriction sites. The culture of Sf9 cells and the production, cloning,and amplification of recombinant baculovirus were performed accordingto the method of Summers and Smith (38). Sf9 cells were usuallyinfected with 1 plaque-forming unit/cell of baculovirus and were har-vested 48 h after infection.

Preparation of Membranes from Sf9 Cells—Membranes were pre-pared from Sf9 cells expressing individual isoforms of adenylyl cyclase,as described previously (39). The cell suspensions were centrifuged,washed with Phillip’s buffer containing protease inhibitors (20 mg/mlsoybean trypsin inhibitor, 4 mg/ml leupeptin, 12 units/ml kallikreininactivator, 4 mg/ml antipain, 52.4 mg/ml benzamidine, 52.3 mg/mlphenylmethylsulfonyl fluoride, and 2 mg/ml pepstatin A). Followingcentrifugation and subsequent lysis of cells in hypo-osmotic buffer con-taining protease inhibitors, samples were sheared mechanically byhomogenization with a Wheaton Dounce homogenizer with 25 strokesand passage through a 22-gauge needle 10 times. The lysates werecentrifuged at 270 3 g for 10 min. The crude membranes were preparedby pelleting this supernatant at 20,000 3 g. Purified membranes wereprepared by fractionation of the lysates using a continuous gradient of5–50% sucrose in lysis buffer. The plasma membranes banding between30 and 42% sucrose were removed and washed. The membranes wereresuspended in buffer containing 40 mM Tris, pH 7.4, 800 mM EGTA,and 0.25% bovine serum albumin, at a final protein concentration of0.5–2.5 mg/ml (as determined by the method of Lowry et al. (40) usingbovine serum albumin as a standard). The preparations were stored inliquid nitrogen.

Preparation of Soluble Forms of AC5—The plasmids encoding thesoluble forms of AC5 comprising either the C1 or C1a domain linked tothe C2 domain (VC1C2 and VC1aC2, respectively) were kindly providedby Dr. T. B. Patel (University of Kentucky, Lexington; Ref. 32). The

proteins were expressed in the Escherichia coli strain TP2000, which isincapable of producing cAMP (42). The expression of VC1C2 and VC1aC2,cell lysis and assays were performed as described previously (32).

Preparation of Mouse Cerebellum and Striatum Membranes—Malemice of the C57bl/6 strain were decapitated. The brain was removedand placed on dry ice. The cerebellum was detached and homogenizedimmediately. A transverse section of the forebrain was made at thelevel of the optic chiasma, and the dorsal striatum were dissected basedon the method described by Glowinski and Iversen (43). The tissueswere homogenized and centrifuged at 1000 3 g for 10 min in 20 volumesof a cold Tris-sucrose (50 mM, 10%) buffer, pH 7.4, containing 0.8 mM

EGTA and a mixture of protease inhibitors as described above. Thesupernatant was centrifuged at 15,000 3 g for 10 min, and the resultingpellet was washed and centrifuged three times before final suspensionin a volume of Tris buffer (50 mM) sufficient to give a concentration ofproteins in the range of 1 mg/ml as determined by the method of Lowry(40). Aliquots were immediately frozen in liquid nitrogen until use.Before use for the adenylyl cyclase assay, the samples were furtherwashed and centrifuged two times in 2 ml of a Hepes buffer (50 mM), pH7.4, containing 240 mM EGTA and 0.25% bovine serum albumin.

Adenylyl Cyclase Activity Measurements—The adenylyl cyclase ac-tivity of infected Sf9 cells or mouse brain tissues (striatum and cere-bellum) was measured in the presence of the following components: 12mM phosphocreatine, 2.5 units of creatine phosphokinase, 0.1 mM

cAMP, 1 mM MgCl2, 0.1 mM ATP, 0.04 mM GTP, 0.5 mM isobutylmeth-ylxanthine, 70 mM HEPES buffer, pH 7.4, and 1 mCi of [a-32P]ATP.(MgCl2 and ATP concentrations were varied in some experiments, asindicated). Because we were inhibiting adenylyl cyclase activity to lowlevels (at high [Ca21], particularly at low [Mg21]), maximizing thesevalues was desirable to generate robust kinetic parameters; conse-quently, basal activity was enhanced with forskolin for all sourcesexcept cerebellum and striatum, which displayed elevated basal activ-ities. Forskolin stimulation was employed in some experiments fromthese latter sources and yielded unchanged parameters (not shown).Calmodulin (1 mM) was included for the assay of cells expressing AC1and cerebellar membranes. Free Ca21 concentrations were establishedfrom a series of CaCl2 solutions buffered with 60 mM EGTA in the assayand were calculated as described previously in detail (44). The reactionmixture (final volume, 100 ml) was incubated at 30 °C for 30 min.Reactions were terminated with sodium lauryl sulfate (0.5%) contain-ing 1.5 mM cAMP and 22 mM ATP. [3H]cAMP (;10,000 cpm) was addedas a recovery marker, and the [32P]cAMP formed was quantified asdescribed previously (45). Data points are presented as mean activitiesof triplicate determinations (error bars are indicated when they arelarger than the data point symbols in Figs. 1–5). For each assay, at leasttwo adenylyl cyclases were compared, and a cerebellum Ca21 dose-response curve was determined as an internal control to monitor theaccuracy of the Ca21 buffering system.

Assays for Effect of Ca21 on Mg21 or MgATP22 Requirement—Theseassays were conducted as described above except that concentrations ofMg21 or ATP varied. Free Mg21 concentrations were established from aseries of MgCl2 solutions buffered with 60 mM EGTA and were calcu-lated as described previously (44). In order to investigate the respectiveeffects of occupancy of high and low affinity regulatory sites for Ca21,two selected concentrations of Ca21 (2.75 and 86 mM) were included inthese assays. After calculation, the concentrations of free Ca21 actuallyranged from 2.59 to 3.43 mM and from 81.6 to 90.1 mM, respectively, inthe presence of the lowest and the highest concentration of Mg21 orATP.

Statistics—Nonlinear regression was used to fit a competition curvewith either one or two components to the data (Graphpad, Inplot4).Goodness of fit was quantified by the least-squares method (F-test) andwas deemed better suited to one model when p was ,0.05.

Paired Student’s t tests (two-tailed) were performed to compare themeans of Km or Ka values determined with different concentrations ofCa21.

RESULTS

Species-specific Effects of Ca21 on Adenylyl Cyclase Activity—The effects of a broad range of Ca21 concentrations on adenylylcyclase activity were compared in membranes from Sf9 cellsexpressing AC6 and striatum (which predominantly expressesAC5; Ref. 33). Clearly, Ca21 elicits a biphasic decline in activitywith increasing concentration (Fig. 1). In both cases, curve-fitting analyses revealed that the inhibition curves alwaysfitted significantly better to a two-component model (p , 0.05),

Inhibition by Calcium of Mammalian Adenylyl Cyclases35540

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

for which the effective concentrations giving 50% of the re-sponse (EC50) were in the ranges of submicromolar concentra-tions and submillimolar concentrations, respectively (Table I).By contrast, the effects of Ca21 on the adenylyl cyclase activityin Sf9 cells expressing AC1 and in cerebellum membranes werebidirectional. Stimulation occurred within the range of submi-cromolar concentrations of Ca21, whereas inhibition occurredat submillimolar concentrations (Fig. 1). Curve fitting of thesedata indicated that stimulation occurred with a Ka in thesubmicromolar range and inhibition in the supramicromolarrange (Table I). In Sf9 cells expressing AC2, Ca21 also inhib-ited adenylyl cyclase activity (Fig. 1). However, curve-fittinganalyses of these data revealed only one component, for whichthe EC50 was in the low affinity (submillimolar) concentrationrange (Table I).

A recent report suggested that the C1b region of AC5 mightmediate high affinity inhibition by Ca21 (32). Consequently,the Ca21 sensitivity of the full-length and truncated solubleforms of AC5 (VC1C2 and VC1aC2, respectively) were exam-ined. In both cases, Ca21 produced an inhibition similar to thatmeasured in the cells expressing AC2 (Fig. 1). Curve fittinganalyses showed that, as in material expressing AC2, the in-hibition was monophasic and, as with AC2, the EC50 was in thelow affinity concentration range.

Effect of Ca21 on Mg21 Activation of Adenylyl Cyclase—For a

considerable period, it has been believed that low affinity inhi-bition of adenylyl cyclase activity by Ca21 reflected competitionfor a Mg21 activation site (11, 16–18, 20–22). However, thisissue has not been addressed in the light of current informationon adenylyl cyclase diversity. Consequently, the effects of oc-cupancy of high affinity and low affinity Ca21 inhibition siteson the activation by Mg21 of various adenylyl cyclases wereinvestigated . The effects of two concentrations of free Ca21,corresponding to the plateau region of the first inhibitory phase(2.75 mM; cf. Fig. 1) and within the second phase (86 mM; cf. Fig.1) of the Ca21 inhibition curves, on the activation by Mg21 ofAC1, AC2, AC6, and striatal adenylyl cyclase were determined.Concentration-activity curves were generated using varyingconcentrations of Mg21 (Fig. 2). The results showed that Mg21

increased adenylyl cyclase activity concentration-dependently,as expected. The two selected concentrations of Ca21 inhibitedthe Mg21 stimulation in membranes expressing AC6. Similareffects were observed in striatal membranes. In contrast, thestimulation by Mg21 was inhibited only by the high Ca21

concentration (86 mM) in membranes expressing AC2. Similarassays were also performed with AC1 membranes. Only theeffect of the high concentration of Ca21 was investigated (in theabsence of calmodulin), since at low concentrations AC1 wouldbe stimulated by Ca21. Indeed, in this condition, Ca21 alsoreduced the activation produced by Mg21 in AC1. Reciprocalplots of velocity versus Mg21 concentration are shown in Fig. 3.The linear regression analyses performed on these data indi-cate that, in AC6 and striatal membranes, both 2.75 and 86 mM

Ca21 inhibited Mg21 activation competitively; i.e. the intercept(1/Vmax) was unchanged by Ca21, whereas the slope (Ka/Vmax)increased (Fig. 3, Table II). The inhibition detected in AC2 wasalso competitive at the high concentration of Ca21 (Fig. 3, TableII). Finally, Ca21 also demonstrated competitive inhibition at86 mM Ca21 in AC1 membranes. The mean Ka for Mg21 wasabout 1 mM in all of the adenylyl cyclases examined in theabsence of Ca21 (Table II). However, in the presence of 2.75 mM

Ca21, this value was significantly increased (by about 2-fold),with Ca21-inhibitable adenylyl cyclases. This increase reached6–8-fold and occurred in all adenylyl cyclase isoforms when 86mM Ca21 was included in the assays (Table II).

Effects of Ca21 on the MgATP22 Requirement in AC6, AC2,AC1, or Striatum—Since Mg21 not only activates adenylylcyclase, but also participates in the formation of the cyclasesubstrate, MgATP22, we investigated the effects of inhibitionby Ca21 (3.41 and 89 mM) on the relationship between activityand substrate (MgATP22) concentration. In these experiments,the concentration of free Mg21 was held constant (10 mM) sothat it could be determined whether the antagonistic effect ofCa21 on Mg21 activation reported in Fig. 2 (and Table I) mightreflect some effect on the participation of Mg21 in the substrate

FIG. 1. Effects of Ca21 on adenylyl cyclase activity in mem-branes of cells expressing AC6, AC2, AC1, or soluble forms ofAC5 and striatal or cerebellum membranes. Membranes express-ing various adenylyl cyclases were assayed in the presence of theindicated free Ca21 concentrations, as described under “ExperimentalProcedures.” Forskolin was included in the assay as follows: 100 mM forAC6 and soluble forms of AC5 and 10 mM for AC2 and AC1. No forskolinwas included in assays of either striatal or cerebellar membranes. Eachpanel shows a representative experiment that was repeated at leastthree times with similar results. Data are mean adenylyl cyclase activ-ity (6 S.E.) of triplicate determinations.

TABLE IHigh and low affinity effects of calcium

Values are mean 6 S.E. of effective concentrations of calcium giving50% of the response for the high affinity and the low affinity sites,respectively, as determined from curve-fitting analyses of experimentsanalogous to those depicted in Fig. 1. The number of determinationsperformed on each adenylyl cyclase is indicated (n).

High affinity Low affinity

mM mM

AC6 in Sf9 cells (n 5 6) 0.148 6 0.039 0.060 6 0.005AC2 in Sf9 cells (n 5 9) None 0.066 6 0.013AC1 in Sf9 cells (n 5 7) 0.122 6 0.021 0.032 6 0.004

Striatum (n 5 3) 0.079 6 0.030 0.058 6 0.002Cerebellum (n 5 20) 0.094 6 0.012 0.055 6 0.010

VC1C2 (n 5 7) None 0.081 6 0.017VC1aC2 (n 5 6) None 0.097 6 0.017

Inhibition by Calcium of Mammalian Adenylyl Cyclases 35541

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

formation. The concentration-response curves for MgATP22

obtained for each adenylyl cyclase are shown in Fig. 4. Theresults indicate that the two concentrations of Ca21 reducedthe response to substrate MgATP22 dose-dependently in bothmembranes expressing AC6 and striatum. As expected, onlythe high concentration (89 mM) altered the [MgATP22] depend-ence of AC2. Finally, in membranes expressing AC1, a reduc-tion of the activity in response to increasing concentrations ofMgATP22 also occurred at the high concentration of Ca21.Reciprocal plots of activity versus MgATP22 concentrations are

shown in Fig. 5. Linear regression analyses of these data indi-cated that the low concentration of Ca21 produced a noncom-petitive inhibition on the substrate MgATP22 activation forboth AC6 and striatum. These plots were largely parallel, al-though there was also a slight indication of a Km effect. Finally,noncompetitive patterns of inhibition were revealed in all of theadenylyl cyclase preparations examined when the high concen-tration of Ca21 was included in the assays. As indicated inTable II, in all cases, the mean Km value was significantlydecreased (p , 0.05) with respect to controls in the presence of

FIG. 2. Effects of Ca21 on the stimu-lation of Mg21 of adenylyl cyclase ac-tivity. Adenylyl cyclase activity was de-termined in membranes expressing AC6,AC2, and AC1, respectively, and in stria-tum membranes in the absence of Ca21

(E) and in the presence of 2.75 mM Ca21

(M) or 86 mM Ca21 (‚). Forskolin was in-cluded in the assay as follows: 25 mM forAC6 and 10 mM for AC2 and AC1. Noforskolin was added in striatal assays.Data are means 6 S.E. of triplicate deter-minations from an experiment that wasrepeated twice (striatum) or three times(AC6, AC2, and AC1) with similar results.

FIG. 3. Double reciprocal plots of ef-fects of Ca21 on Mg21 stimulation ofadenylyl cyclase activity. Plots weremade from the data presented in Fig. 2and analyzed by linear regression analy-sis. Free Ca21 concentrations in the assaywere as follows: 0 (E), 2.75 mM (M), and 86mM (‚).

TABLE IIKinetic parameters for Mg21 and MgATP as a function of Ca21 concentrations

Values are means 6 S.E. of Ka and Km values extracted from linear regression analyses performed on the double reciprocal plots of Mg21 andMgATP22 stimulation of adenylyl cyclase activity. The number of assays performed for each adenylyl cyclase is indicated (n).

Mg21 (Ka) MgATP (Km)

0 mM Ca21 2.75 mM Ca21 86 mM Ca21 0 mM Ca21 3.41 mM Ca21 89 mM Ca21

mM mM

AC6 (n 5 3) 1.02 6 0.12 1.77 6 0.36a 8.08 6 1.23a 68.30 6 3.35 69.09 6 4.59 46.27 6 10.3a

Striatum (n 5 2) 0.91 6 0.21 2.09 6 0.61a 8.11 6 4.70a 46.16 6 2.72 43.92 6 1.07 34.73 6 0.25a

AC2 (n 5 3) 1.02 6 0.25 1.07 6 0.28 6.92 6 1.52a 77.26 6 2.83 76.37 6 1.02 48.11 6 6.45a

AC1 (n 5 3) 1.32 6 0.06 ND 6.10 6 0.76a 383.7 6 26.3 ND 241.0 6 20.1a

a A significant difference (p , 0.05) with respect to control (no added Ca21) as determined by a paired Student’s t test (two tails).

Inhibition by Calcium of Mammalian Adenylyl Cyclases35542

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

89 mM Ca21, although the effect on Vmax was most prominent.Overall, these analyses are consistent with the predominant

effect of Ca21 being to antagonize stimulation by Mg21, alongwith a minor noncompetitive effect on substrate utilization.

DISCUSSION

The stimulation of adenylyl cyclases AC1 and AC8 by Ca21 isunderstood to occur in the submicromolar concentration rangeand to be mediated by calmodulin (9, 46–48). In addition, thecalmodulin binding sites are known for both AC1 and AC8 (49,50). By contrast, even the concentration ranges over whichCa21 inhibits adenylyl cyclases are not unambiguously under-stood; even less so is the mechanism of inhibition. There havebeen a variety of studies reporting monophasic inhibition (11–13, 21, 23) as well as some showing evidence for two distinctbinding sites of low and high affinity for Ca21, giving rise tobiphasic inhibition (14–19). When AC5 and AC6 were identi-fied, they were characterized as being Ca21-inhibited adenylylcyclase isoforms (25, 31). When expressed in intact cells, thesecyclases are inhibited by capacitative Ca21 entry (51). ThemRNAs of these species are also expressed in tissues, such asthe heart, in which the adenylyl cyclase activity is inhibited byCa21 (52, 53). To resolve the confusion that attends Ca21

inhibition of adenylyl cyclase, examination of material express-ing a single isoform would seem to be a rational procedure tounderstand the Ca21 regulation of adenylyl cyclase. Conse-quently, in the present study we first compared the regulationof a presumably Ca21-stimulable and a Ca21-inhibitable ad-enylyl cyclase expressed in Sf9 cells with tissues that predom-inantly express these activities in the mouse brain. It wasgratifying to note that the response to Ca21 of Sf9 cell mem-branes expressing AC1 and AC6 was virtually identical to theresponse of cerebellar and striatal membranes, respectively,which are a major source of the analogous adenylyl cyclasemRNAs. Such findings indicate that no co-factor or post-trans-lational modification is lacking between moths and mice toalter adenylyl cyclase responsiveness. Consequently, it seemsreasonable to characterize further the Ca21-regulation in mem-branes from either source.

The present studies clearly indicate that Ca21 inhibits AC6(and AC5) over precisely the same concentration range as itstimulates AC1. Specifically, our results demonstrate clearlythat, in both adenylyl cyclases, the effects of Ca21 progress ina biphasic manner, strongly suggesting interactions with twodistinct regulatory binding sites of low and high affinity. Inter-

FIG. 4. Effects of Ca21 on theMgATP dependence of adenylyl cy-clase activity. Adenylyl cyclase activitywas determined in the membranes ex-pressing AC6, AC2, and AC1, respec-tively, and in striatum membranes in theabsence of Ca21 (E) and in the presence of3.41 mM Ca21 (M) or 89 mM Ca21 (‚). For-skolin concentration included in the assaywas 25 mM for AC6 and 10 mM for AC2 andAC1. No forskolin was added in striatalassays. Data are means 6 S.E. of tripli-cate determinations from an experimentthat was repeated twice (striatum) orthree times (AC6, AC2, and AC1) withsimilar results.

FIG. 5. Double reciprocal plots foreffects of Ca21 on MgATP depend-ence of adenylyl cyclase activity.Plots were made from the data presentedin Fig. 4 and analyzed by linear regres-sion analysis. Free Ca21 concentrations inthe assay were as follows: 0 (E), 3.41 mM

(f), and 89 mM (‚).

Inhibition by Calcium of Mammalian Adenylyl Cyclases 35543

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

action with sites of high affinity for Ca21 (;0.15 mM) conferopposite regulation in AC6 (inhibition) and AC1 (stimulation),whereas interaction with sites of low affinity (;0.06 mM) in-hibits both isoforms. The effects of Ca21 on striatal adenylylcyclase activity were identical to those detected in AC6. Previ-ous reports had found that the basal adenylyl cyclase activityin the striatum was higher but less stimulated by Ca21, orinsensitive to Ca21, as compared with other brain areas, in thepresence of calmodulin (54). More recently, the adenylyl cyclaseactivity of the rat striatum was shown to exhibit a daily oscil-lation with a peak occurring around 10:00–12:00 a.m., duringwhich Ca21 inhibition was selectively manifest (55). Accord-ingly, we dissected striatum during this critical phase, and inagreement with the latter findings, our data showed that Ca21

potently inhibits the striatal adenylyl cyclase activity. Finally,comparisons with AC2 demonstrated that, in sharp contrast,this adenylyl cyclase was insensitive to low concentrations ofCa21. Nevertheless, like the other adenylyl cyclases examinedhere, AC2 was inhibited when Ca21 reached submillimolarlevels. Together, these findings demonstrate that the variousisoforms of adenylyl cyclases behave differently when Ca21

interacts at high affinity binding sites, whereas they respondsimilarly when Ca21 binds low affinity sites.

Next, we compared the mechanisms that yielded high affin-ity inhibition in AC6 and striatal adenylyl cyclase, comparedwith the low affinity inhibition that is common to all adenylylcyclases. Previous studies of adenylyl cyclase activity in vari-ous tissues had proposed that high concentrations of Ca21

reduced activity depending on the substrate concentration in anoncompetitive manner (11, 12, 16, 17, 21, 56). When compe-tition between Ca21 and Mg21 was considered, more contradic-tions were encountered, even between studies carried out onthe same tissues. Specifically, some studies showed that Ca21

inhibition of adenylyl cyclase occurred noncompetitively withMg21 (11, 19, 23), while others found that high concentrationsof Ca21 compete with Mg21, whereas lower concentrations donot (56). In other studies, it was proposed that Ca21 inhibitsadenylyl cyclase by competing with Mg21 and suggested that acommon metal site bound both ions (16, 17, 20–22). Here, weshow that high affinity inhibition by Ca21 of either AC6 orstriatal adenylyl cyclase does not involve competition with theMgATP22 substrate. Instead, high affinity inhibition reflectscompetition by low concentrations of Ca21 for the Mg21 acti-vation in both AC6 and striatal adenylyl cyclase, so that theaffinity for Mg21 is reduced selectively in these isoforms. Onthe other hand, low affinity inhibition produced by Ca21 inAC6, AC1, and AC2, as well as in the striatal adenylyl cyclase,also stems largely from a competitive action on the Mg21 acti-vation of the enzyme, with a minor effect on substrateutilization.

Recent deletion analyses have suggested the existence of aMg21 binding site essential for catalysis and distinct from theATP-bound cation (57). Along with insight gained from crystal-lographic studies (58), the latter findings led these authors toconclude that one conserved aspartate residue (correspondingto Asp-396 in the C1a region of canine AC5) plays a critical rolein coordinating catalytic Mg21 ions. Specifically, this Mg21 waspredicted to facilitate the nucleophilic attack of the 39-hydroxylgroup of ATP on the a-phosphate. It is conceivable that Ca21

competes with Mg21 at this site, which is largely conservedamong all adenylyl cyclases, to yield low affinity inhibition. Bycontrast, the site of high affinity inhibition by Ca21 of AC5 andAC6 would be expected to be mediated by a region unique toAC5 and AC6.

In the latter context, Scholich et al. (32) reported recentlythat Ca21 inhibited a soluble form of AC5, composed of the C1

and C2 regions of the enzyme, but not a shorter form lackingthe unconserved C1b domain. These findings were taken tosuggest that the 112-amino acid C1b region in AC5 mediatedhigh affinity Ca21 inhibition of the enzyme activity. In agree-ment with this study, we also found that Ca21 would inhibit theC1C2 protein. However, since we were conducting our experi-ments in parallel with Ca21-stimulated cerebellar adenylylcyclase, as an internal control of estimated Ca21 concentra-tions, it was clear that the inhibition occurred with submilli-molar concentrations of Ca21, not with submicromolar concen-trations. Furthermore, in our experiments, the truncated formof C1C2, namely the C1aC2 construct, displayed the same lowaffinity inhibition by Ca21. Based on these results, it seemsunlikely that the Ca21 site responsible for high affinity inhibi-tion resides within the cytosolic domains of AC5. However, it ispossible that the soluble chimera expressed in Escherichia colidoes not adopt the structural configuration of the native ACV.Alternatively, the absence of other domains might explain theabsence of high affinity inhibition in the VC1C2 protein. In anycase, it is reasonable to conclude that further studies areneeded to clarify the structural features conferring Ca21 inhi-bition on adenylyl cyclases.

The present study has clearly separated the high and lowaffinity effects of Ca21. With a high affinity, Ca21 stimulatesAC1 and cerebellar membrane adenylyl cyclase in a mannerthat requires dissociable calmodulin. Again, with a high affin-ity, Ca21 inhibits AC6 expressed in Sf9 cells and striatal ad-enylyl cyclase independently of added calmodulin. These lowconcentrations of Ca21 do not affect AC2 expressed in Sf9 cells.With a low affinity, Ca21 inhibits all adenylyl cyclases, includ-ing AC1, AC2, AC6, striatal AC5, and soluble forms of AC5,intact or with the C1b region deleted. Kinetically, the mecha-nism of both high and low inhibition by Ca21 appears similarand appears to involve competitive inhibition for a stimulatoryMg21 site(s). The inhibition does not involve competitive mech-anisms for the substrate MgATP22. A Ki value of ;0.2 mM forthe high affinity inhibitory site, in the face of a 5000-fold excessof Mg21, suggests a remarkable selectivity for Ca21. This se-lectivity for Ca21 over Mg21 is analogous with the specificity ofE-F-hand-containing proteins for Ca21 (59). However, the na-ture or location of this site remains elusive. No E-F hand orother obvious Ca21-binding motif is present in AC5 or AC6 (31).In addition, earlier biochemical studies appeared to eliminatethe participation of calmodulin in the inhibition; for instance,neither calmodulin antagonists nor EGTA washing modulatedthe effect (60), although this does not exclude the possibilitythat a tightly bound calmodulin is involved (as with phospho-rylase kinase). On the other hand, the low affinity site appearsto represent a more conventional competition between Ca21

and Mg21, which, as with a number of metabolic enzymes,shows an approximately 10-fold difference in effective concen-trations of the two cations (62, 63). Given that averaged, cyto-solic concentrations of Ca21 rarely reach 50 mM, except tran-siently around the mouths of ion channels, it is difficult toimagine that this inhibition is of any physiological utility. Onthe other hand, it is important to take pains to distinguish thisinhibition from high affinity inhibition, or quite spurious infer-ences may be drawn on the nature or location of high affinitybinding sites.

Acknowledgments—We thank Drs. A. G. Gilman, R. Iyengar, and T.Patel for constructs used in this study. We also thank Dr. Jeff Karpenfor useful comments and the University of Colorado Health SciencesCenter Cancer Center for propagating baculovirus-infected cells.

REFERENCES

1. Kandel, E. R., Klein, M., Castellucci, V. F., Schacher, S., and Goelet, P. (1986)Neurosci. Res. 36, 498–520

2. Chetkovich, D. M., Gray, R., Johnston, D., and Sweatt, J. D. (1991) Proc. Natl.

Inhibition by Calcium of Mammalian Adenylyl Cyclases35544

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from

Acad. Sci. U. S. A. 88, 6467–64713. Mons, N., Guillou, J.-L., and Jaffard, R. (1999) Cell. Mol. Life Sci. 55, 523–5354. Yovell, Y., Kandel, E. R., Dudai, Y., and Abrams, T. W. (1987) Proc. Natl. Acad.

Sci. U. S. A. 84, 9285–92895. Dudai, Y., and Zvi, S. (1984) Neurosci. Lett. 47, 119–1246. Storm, D. R., Hansel, C., Hacker, B., Parent, A., and Linden, D. J. (1998)

Neuron 20, 1199–12107. Cooper, D. M. F., Karpen, J. W., and Mons, N. (1995) Nature 374, 421–4248. Stevens, T., Nakahashi, Y., Cornfield, D., McMurtry, I. F., Cooper, D. M. F.,

and Rodman, D. M. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 2696–27009. Fagan, K. A., Mahey, R., and Cooper, D. M. F. (1996) J. Biol. Chem. 271,

12438–1244410. Fagan K. A., Mons, N., and Cooper, D. M. F. (1998) J. Biol. Chem. 273,

9297–930511. Steer, M. L., and Levitzki, A. (1975) J. Biol. Chem. 250, 2080–208412. Hanski, E., Sebilla, N., and Levitzki, A. (1977) Eur. J. Biochem. 76, 513–52013. Demirel, E. Ugur, O., and Ongun Onaran, H. (1998) Pharmacology 57,

222–22814. Potter, J. E., Piascik, M. T., Wisler, P. L., Robertson, S. P., and Johnson, C. L.

(1980) Ann. N. Y. Acad. Sci. 356, 220–23115. Oldham, S. B., Molloy, C. T., and Lipson, L. G. (1984) Endocrinology 114,

207–21416. Oldham, S. B., Rude, R. K., Molloy, C. T., and Lipson, L. G. (1984) Endocri-

nology 115, 1883–189017. Giannattasio, G., Bianchi, R., Spada, A., and Vallar, L. (1987) Endocrinology

120, 2611–261918. Boyajian, C. L., and Cooper, D. M. F. (1990) Cell Calcium 11, 299–30719. Colvin, R. A., Oibo, J. A., and Allen, R. A. (1991) Cell Calcium 12, 19–2720. Drummond, G. I., and Duncan, L. (1970) J. Biol. Chem. 245, 976–98221. Lasker, R. D., Downs, R. W., Jr., and Aurbach, G. D. (1982) Arch. Biochem.

Biophys. 216, 345–35522. Piascik, M. T., Addison, P., and Babich, M. (1985) Arch. Biochem. Biophys. 241,

28–3523. Matsuzaki, S., and Dumont, J. (1972) Biochim. Biophys. Acta 284, 227–23424. Nakahashi, Y., Nelson, E., Fagan, K. A., Mons, N., Gonzalez, E., Guillou, J. L.,

and Cooper, D. M. F. (1997) J. Biol. Chem. 272, 18093–1809725. Ishikawa Y., Katsushika, S., Chen L., Halmon, N. J., Kawabe J., and Homcy,

C. J. (1992) J. Biol. Chem. 267, 13553–1355726. Cooper, D. M. F. (1994) Methods Enzymol. 238, 71–8127. Smit, M. J, and Iyengar, R. (1998) Adv. Second Messenger Phosphoprotein Res.

32, 1–2128. Sunahara, R. K., Dessauer, C. W., and Gilman, A. G. (1996) Annu. Rev.

Pharmacol. Toxicol. 36, 461–48029. Cooper, D. M. F., Karpen, J. W., Fagan, K. A., and Mons, N. (1998) Adv. Second

Messenger Phosphoprotein Res. 32, 23–5130. Wallach, J., Droste, M., Kluxen, F. W., Pfeuffer, T., and Frank, R. (1994) FEBS

Lett. 338, 257–26331. Yoshimura, M., and Cooper, D. M. F. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,

6716–672032. Scholich, K., Barbier, A. J., Mullenix, J. B., and Patel, T. B. (1997) Proc. Natl.

Acad. Sci. U. S. A. 94, 2915–292033. Mons, N., and Cooper, D. M. F (1994) Mol. Brain Res. 22, 236–24434. Glatt, C. E., and Snyder, S. H. (1993) Nature 361, 536–53835. Mons, N., Yoshimura, M., and Cooper, D. M. F. (1993) Synapse 14, 51–5936. Taussig, R., Quarmby, L. M., and Gilman, A. G. (1993) J. Biol. Chem. 268,

9–1237. Jacobowitz, O., Iyengar, R.(1994) Proc. Natl. Acad. Sci. U. S. A. 91,

10630–1063438. Summers, M. D., and Smith, G. E. (1987) Tex. Agric. Exp. Stn. Bull. 155, 1–5639. Caldwell, K. K., Newell, M. K., Cambier, J. C., Prasad, K. N., Masserano, J. M.,

Schlegel, W., and Cooper, D. M. F. (1988) Anal. Biochem. 175, 177–19040. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol.

Chem. 193, 265–27541. Roy, A., and Danchin, A. (1982) Mol. Gen. Genet. 188, 465–47142. Beauve, A., Boesten, B., Crasnier, M., Danchin, A., and O’Gara, F. (1990) J.

Bacteriol. 172, 2614–262143. Glowinski, J., and Iversen, L. L. (1966) J. Neurochem. 13, 655–66944. Ahlijanian, M. K., and Cooper, D. M. F., (1987) J. Pharmacol. Exp. Ther. 241,

407–41445. Salomon, Y., Londos, C., and Rodbell, M. (1974) Anal. Biochem. 58, 541–54846. Cali, J. J. Zwaagstra, J. C., Mons, N., Cooper, D. M. F., and Krupinski, J.

(1994) J. Biol. Chem. 269, 12190–1219547. Tang, W. J., Krupinski, J., and Gilman, A. G. (1991) J. Biol. Chem. 266,

8595–860348. Levin, L. R., and Reed, R. R. (1995) J. Biol. Chem. 270, 7573–757949. Wu, Z., Wong, S. T., and Storms, D. R. (1993) J. Biol. Chem. 268, 23766–2376850. Gu, C., and Cooper, D. M. F. (1999) J. Biol. Chem. 274, 8012–802151. Cooper, D. M. F., Yoshimura, M., Zhang, Y., Chiono, M., and Mahey, R. (1994)

Biochem. J. 297, 437–44052. Yu, H. J., Ma, H., and Green, R. D. (1993) Mol. Pharmacol. 44, 689–69353. Cooper, D. M. F., and Brooker, G. (1993) Trends Pharmacol. Sci. 14, 34–3654. Dokas, L. A, and Ting, S. M. (1993) Neurobiol. Aging 14, 65–7255. Chern, Y., Lee, E. H. Y., Lai, H. L., Wang, H. L., Lee, Y. C., and Ching, Y. H.

(1996) FEBS Lett. 385, 205–20856. Bellorin-font, E., Martin, K. J., Freitag, J. J., Anderson, C., Sicard, G.,

Slatopolsky, E., and Klahr, S. (1981) J. Clin. Endocrinol. Metab. 52,499–507

57. Zimmermann, G., Zhou, D., and Taussig, R. (1998) J. Biol. Chem. 273,19650–19655

58. Tesmer, J. J. G. Sunahara, R. K., Johnson, R. A., Gosselin, G., Gilman A. G.,and Sprang, S. R. (1999) Science 285, 756–760

59. Drake, S. K., Lee, K. L., and Falke, J. J. (1996) Biochemistry 35, 6697–670560. Boyajian, C. L., Garritsen, A., and Cooper, D. M. F. (1991) J. Biol. Chem. 266,

4995–500361. Garritsen, A., Zhang, Y., Firestone, J. A., Browning, M. D., and Cooper,

D. M. F. J. Neurochem. (1992) 59, 1630–163962. Cerdan, S., Lusty, C. J., Davis, K. N., Jacobson, J. A., and Williamson, J. R.

(1984) J. Biol. Chem. 259, 323–33163. Fraichard, A., Trossat, C., Perotti, E., Pugin, A. (1996) Biochimie (Paris) 78,

259–266

Inhibition by Calcium of Mammalian Adenylyl Cyclases 35545

at CN

RS

on October 6, 2006

ww

w.jbc.org

Dow

nloaded from