Modulation of Dihydropyridine-SensitiveCalcium Channels in Drosophila by acAMP-Mediated Pathway

Anindya Bhattacharya, Gang-Guo Gu, Satpal Singh

Department of Biochemical Pharmacology, 308 Hochstetter Hall, State University of New York atBuffalo, Buffalo, New York 14260-1200

Received 6 October 1998; accepted 13 January 1999

ABSTRACT: Drosophila has proved to be a valu-able system for studying the structure and function ofion channels. However, relatively little is known aboutthe regulation of ion channels, particularly that of Ca21

channels, inDrosophila.Physiological and pharmacolog-ical differences between invertebrate and mammalianL-type Ca21 channels raise questions on the extent ofconservation of Ca21 channel modulatory pathways. Wehave examined the role of cyclic adenosine monophos-phate (cAMP) cascade in modulating the dihydropyri-dine (DHP)-sensitive Ca21 channels in the larval mus-cles of Drosophila, using mutations and drugs thatdisrupt specific steps in this pathway. The L-type (DHP-sensitive) Ca21 channel current was increased in thedunce mutants, which have high cAMP concentrationowing to cAMP-specific phosphodiesterase (PDE) dis-ruption. The current was decreased in the rutabagamutants, where adenylyl cyclase (AC) activity is altered

thereby decreasing the cAMP concentration. Thedunceeffect was mimicked by 8-Br-cAMP, a cAMP analog,and IBMX, a PDE inhibitor. The rutabaga effect wasrescued by forskolin, an AC activator. H-89, an inhibitorof protein kinase-A (PKA), reduced the current andinhibited the effect of 8-Br-cAMP. The data suggestmodulation of L-type Ca21 channels ofDrosophilavia acAMP-PKA mediated pathway. While there are differ-ences in L-type channels, as well as in components ofcAMP cascade, betweenDrosophila and vertebrates,main features of the modulatory pathway have beenconserved. The data also raise questions on the likelyrole of DHP-sensitive Ca21 channel modulation in syn-aptic plasticity, and learning and memory, processesdisrupted by the dnc and the rut mutations. © 1999 John

Wiley & Sons, Inc. J Neurobiol 39: 491–500, 1999

Keywords: Drosophila;larval muscle; L-type Ca21 cur-rent; dnc; rut; cAMP

Voltage-dependent calcium channels provide a majorpathway for calcium entry into the cell (Hille, 1992;McDonald et al., 1994; Varadi et al., 1995). Sincecalcium influx through these channels plays a criticalrole in several physiological processes such as exci-tation-contraction coupling, secretion, gene expres-sion, learning and memory, and activation of calcium-

dependent ion channels (Rampe and Kane, 1994),they are subject to tight regulation by intracellularcomponents. For example, phosphorylation plays animportant role in modulating Ca21 channels. Proteinkinases and phosphatases make functionally impor-tant modifications to these channels. Of the kinases,the protein kinase A (PKA) pathway has been foundto play a crucial role in modulating vertebrate L-typecalcium channels (McDonald et al., 1994). Interest-ingly, channels in different cell types are modulatedby the PKA pathway in different ways (Sperelakis etal., 1994; Xiong et al., 1994). For example, activationof the PKA pathway increases L-type current in theskeletal muscles of bullfrog (Kokate et al., 1993),

Correspondence to:S. SinghContract grant sponsor: NIH; contract grant number: GM-

50779Contract grant sponsor: NSF; contract grant numbers: IBN-

9011427, MCB-9604457Contract grant sponsor: Sigma Xi

© 1999 John Wiley & Sons, Inc. CCC 0022-3034/99/040491-10

491

whereas it inhibits the L-type current in the vascularsmooth muscles of rabbit portal vein (Xiong et al.,1994).

Mechanisms that modulate Ca21 channels in in-vertebrates, particularly inDrosophila,are not as wellstudied as in vertebrates. Almost nothing is knownabout the role of signaling pathways and the extent oftheir involvement, in Ca21 channel modulation inDrosophila. L-type Ca21 channels inDrosophilashow pharmacological differences from vertebrate L-type channels. For example, the most common type ofBa21-conducting channels seen inDrosophila brainmembrane preparations is sensitive to phenylalkyl-amines but resistant to DHPs (Greenberg et al., 1989;Pelzer et al., 1989), while the DHP-sensitive channelsin the larval muscles are resistant to phenylalkyl-amines (Gielow et al., 1995). This is in contrast tovertebrates, in which the L-type channels are sensitiveto all the three categories of blockers, viz. dihydro-pyridines, phenylalkylamines, and benzothiazepines.Other types of Ca21 channels inDrosophila alsoshow significant pharmacological differences fromvertebrate Ca21 channels. These differences raisequestions as to whether and to what extent the channelmodulatory pathways are conserved between insectsand vertebrates, and highlight the importance of ex-amining the pathways that modulate these channels inDrosophila.

Studies on modulation of channels can be helpedgreatly by mutations and drugs that disrupt specificcomponents of the signal transduction pathways. Anextraordinary repertoire of available mutations inDrosophila thus provides a useful tool for such stud-ies. The larval muscle preparation ofDrosophila isparticularly suited for these studies because the L-typeCa21 current in these muscles can be studied in iso-lation from other ionic currents (Gielow et al., 1995).We have studied the modulation of L-type Ca21 chan-nels in larval muscles via cyclic adenosine monophos-phate (cAMP)-mediated pathway by combining mu-tations and drugs that disrupt this pathway.

The dunce (dnc) mutation specifically disruptscAMP-specific phosphodiesterase (PDE) (Byers et al.,1981; Davis and Kiger, 1981; Chen et al., 1986;Nighorn et al., 1994), while therutabaga(rut) muta-tion affects calcium/calmodulin-sensitive adenylyl cy-clase (Livingstone et al., 1984; Levin et al., 1992).There is an increase of cAMP levels in thednc mu-tants (Byers et al., 1981; Davis and Kiger, 1981),whereasrut mutations decrease cAMP concentration(Livingstone et al., 1984; Feany, 1990). These muta-tions, originally identified for affecting learning andmemory (Dudai et al., 1976; Aceves-Pina et al., 1983;Tully, 1991), have been used to study the importance

of cAMP in learning and memory (Dudai et al., 1976;Drain et al., 1991; Davis et al., 1995), pacemakerfunction (Levine et al., 1994), synaptic arborization(Zhong et al., 1992), synaptic plasticity (Zhong andWu, 1991; Delgado et al., 1992), habituation (Engeland Wu, 1996), growth cone motility (Kim and Wu,1996), and modulation of K1 currents (Delgado et al.,1991; Zhong and Wu, 1993; Zhong, 1995; Alshuaiband Mathew, 1998). To examine the modulation ofthe L-type Ca21 channels inDrosophila,we used thednc and rut mutations along with pharmacologicalagents that affect cAMP-PKA pathway. These com-pounds included 8-Br-cAMP, an analog of cAMP;IBMX, an inhibitor of phosphodiesterase; forskolin,an activator of adenylyl cyclase; and H-89, an inhib-itor of PKA. Data presented here suggest that theL-type Ca21 channels are modulated by cAMP via thePKA pathway inDrosophila.

MATERIALS AND METHODS

Drosophila Strains

The Canton-S (CS) strain was used as control in all exper-iments. Thednc2 andrut2 strains were obtained from C. F.Wu (Department of Biology, University of Iowa, Iowa City,Iowa) anddnc1 andrut1 from R. Davis (Department of CellBiology and Neurology, Baylor College of Medicine, Hous-ton, Texas).

Drugs and Solutions

8-Bromoadenosine 39,59-cyclic monophosphate (8-Br-cAMP), 3-isobutyl-1-methylxanthine (IBMX), and forsko-lin were purchased from Sigma Chemical Company (St.Louis, MO). N-[2-((p-Bromocinnamyl)amino)ethyl]-5-iso-quinolinesulfonamide (H-89) was obtained from Calbio-chem-Novabiochem International (La Jolla, CA). All finalconcentrations of drugs were made fresh every day. Drugswere applied at the final concentration to the bath solution.The dissection solution contained (in mM): NaCl (77.5),KCl (5), MgCl2 (4), NaHCO3 (2.5), trehalose (5), sucrose(115), and Hepes (5). Recording saline contained (in mM):NaCl (77.5), KCl (5), MgCl2 (4), NaHCO3 (2.5), trehalose(5), sucrose (115), Hepes (5), triethylammonium (TEA)(20), 4-aminopyridine (4-AP) (1), quinidine (0.1), andBaCl2 (10) (Stewart et al., 1994; Gielow et al., 1995; Gu andSingh, 1997). The pH was adjusted to 7.1. Current recordedin all experiments was barium current through Ca21 chan-nels.

Preparation

Calcium channel currents of the larval-body wall muscles(Jan and Jan, 1976; Wu and Haugland, 1985; Gielow et al.,

492 Bhattacharya et al.

1995) were recorded using the two-microelectrode voltage-clamp technique. Wandering third-instar larvae were usedthroughout the study (Singh and Wu, 1989). Flies weregrown on a standard cornmeal medium at 21°C (Chopra andSingh, 1994). Larvae were pinned dorsal side up on adissection dish. The cuticle along the dorsal midline was cutand pinned back. All internal organs were removed. Record-ings were made from muscle fiber 12 (Crossley, 1978) andwere completed within 30 min from the start of dissection(Gielow et al., 1995).

Electrophysiology

Electrodes were pulled from thin-walled, 1.0-mm borosili-cate glass capillaries (World Precision Instruments, Sara-sota, FL) using a David Kopf Instruments puller (Model750). Voltage electrode was filled with 2.5M KCl and thecurrent electrode with a 3:1 mixture of 2.5M KCl:2 Mpotassium citrate (Wu and Haugland, 1985; Singh and Wu,1989). Resistances of both electrodes were in the range of10–20 MV. Currents were elicited by 500-ms voltage stepsfrom a holding potential of240 mV to potentials between260 and140 mV, in 10-mV increments. Interpulse dura-tion was 10 s. Potassium currents were blocked by TEA,quinidine, and 4-AP (Wu and Haugland, 1985; Gho andMallart, 1986; Singh and Wu, 1989; Gielow et al., 1995).Control currents were recorded separately for each set ofexperiments. Recording temperature was maintained at21°C with a peltier junction and measured for each larvawith a thermocouple probe after the experiment (Gu andSingh, 1997).

Equipment and Software

A Macintosh IISi computer provided the voltage-clampcommand pulses through a 12-bit digital to analog converterusing a MacADIOS II/16 board from GW Instruments(Somerville, MA). A TEC 01C/02/03 amplifier (NPI Elec-tronic GmbH, Haeldenstrasse, Germany) was used for re-cordings. Data were acquired after 16-bit analog-to-digitalconversion. Further analysis was performed by a programwritten in Think-C (Symantec, Cupertino, CA).

Data Handling

Currents were digitally sampled every 500ms, except dur-ing examination of capacitative transients (Wu andHaugland, 1985), which were sampled every 100ms. Cur-rents were digitally corrected for linear leakage with respectto currents from260 mV. Current densities, expressed asnanoamperes per nanofarad (nA/nF), were calculated bydividing the absolute currents by cell capacitance to avoiddifference due to fiber size. All traces depict average datafrom a number of fibers as mentioned. Current amplitude foreach pulse was measured at the peak value to generate acurrent–voltage (I/V) relationship. Currents recorded underdifferent conditions were compared for their peak values at

210 mV. All data are expressed as mean values6 standarderror of means. Statistical analyses were performed usingStudentt test.

RESULTS

DHP-Sensitive Current Is Altered in dncand rut Mutants

The DHP-sensitive calcium channel current was re-corded from the wild-type (CS),dnc,andrut larvae tostudy the effect of disruption of cAMP metabolism onthe current (Fig. 1). A holding potential of240 mVwas used, which inactivates the amiloride-sensitivecalcium current, yielding only the L-type (DHP-sen-sitive) current in response to depolarizing pulses(Gielow et al., 1995; Gu and Singh, 1997). Averagecurrent traces from the wild-type (CS),dnc2, and therut2 larvae are shown in Figure 1(A–C), respectively.Current–voltage (I/V) relationships are shown in Fig-ure 1(D). Peak current in the wild-type larval muscleswas 48.56 1.0 nA/nF. It was 126.66 5.2% and 78.66 7.5% of the wild-type current indnc2 and rut2,respectively. The increase in the current indnc2 andits decrease inrut2 were statistically significant (dnc2:p , .001; rut2: p , .01). Since mutations in thedncand therut genes resulted in an increase and decreasein the cAMP level, respectively, the data suggest thatan increase in the cAMP concentration correlates withan increase in the DHP-sensitive Ca21 current,whereas a decrease in the cAMP level correlates witha decrease in the current. Figure 1(D) also shows I/Vrelationships for currents recorded from additionaldnc and rut alleles, dnc1 and rut1. There were nosignificant differences in input resistances and capac-itances between CS,dnc,and rut muscles.

Amplitude of the currents showed slight variabilityfrom experiment to experiment. The wild-type (CS)current varied from 46.26 1.5 nA/nF to 48.56 1.5nA/nF during different experiments discussed here.Control recordings were performed independently foreach set of experiments, and percent effects on cur-rents under different recording conditions were calcu-lated with respect to controls performed specificallyfor those recording conditions. Currents from thednc2

fibers varied between 119.3% and 126.8% of thewild-type current, while currents from therut2 fibersvaried between 63.9% and 78.6% of the wild-typecurrent.

Effect of cAMP Analog on Currents

These experiments with thedncand therut mutationsindicate that the DHP-sensitive current may be influ-

Calcium Channel Modulation inDrosophila 493

enced by cAMP concentrations. If an increase in thecurrent is indeed due to higher cAMP levels, it may bepossible to increase the wild-type current by increas-ing cAMP concentration pharmacologically. This wasattempted with 8-Br-cAMP, a cell-permeable analogof cAMP. The current increased in the wild-typemuscles with as low as 10 nM 8-Br-cAMP in therecording solution. Average traces for the wild-typecurrent without or with 10 nM 8-Br-cAMP are shownin Figure 2(A,B), respectively. 8-Br-cAMP increasedthe wild-type current to 118.46 2.9% (p , .001). I/V

Figure 2 Effect of 8-Br-cAMP on the current. (A,B) Av-erage traces from CS without and with 10 nM 8-Br-cAMP,respectively. I/V curves for CS (L5 15; F 5 36), CStreated with 10 nM 8-Br-cAMP (L 5 11; F5 14) anddnc2

(L 5 5; F5 14) are shown in (C). 8-Br-cAMP increased thewild-type current to 118.46 2.9% (p , .001), close to thednc2 value.

Figure 1 Effect of thednc and therut mutations on theDHP-sensitive current. Currents shown here and in allsubsequent figures were elicited with 500-ms voltagepulses (see Materials and Methods). (A–C) Average cur-rent traces from wild-type (CS),dnc2, and rut2 respec-tively. Current voltage (I/V) relationships for thesestrains and fordnc1 and rut1 are shown in (D). Thecurrent increased in thednc mutants, whereas it de-creased in therut mutants. Indnc2 [number of larvae (L)5 5; number of fibers (F)5 6), the peak current at210mV was 126.66 5.2% (p , .001) of the control (L5 3;F 5 7). Current inrut2 (L 5 5; F 5 6) was 78.66 7.5%(p , .01) of the control. Fordnc1: L 5 6; F 5 9; Forrut1: L 5 4; F 5 6.

494 Bhattacharya et al.

relationships are depicted in Figure 2(C). The I/Vcurve fordnc2 is shown as dotted line for comparison.These data further support the possibility that an in-crease in cAMP level may lead to an increase in thecurrent.

Effect of Phosphodiesterase Inhibitoron Currents

The dnc mutations raise the cellular cAMP levels bydisrupting PDE. This effect can be mimicked in thewild-type by inhibiting PDE pharmacologically.IBMX can be used as a potent, cell-permeable inhib-itor of PDE; and wild-type flies fed with IBMX showmemory decay like thednc flies (Asztalos et al.,1991). Application of 50mM IBMX increased thewild-type current amplitude (Fig. 3). Average tracesfor the wild-type current without and with 50mMIBMX are shown in Figure 3(A,B), respectively.IBMX increased the wild-type current to 133.36 4.4% of the control current (p , .001) and mim-icked thednc2 effect on the current [Fig. 3(C)]. Forcomparison, thednc2 I/V curve is included in Figure3(C) as a dotted line. This further strengthens the ideathat a rise in cAMP level by either mutations or drugsmay increase the DHP-sensitive Ca21 current. Wewere unable to record the current fromdnc larvae inthe presence of IBMX owing to increased musclecontraction.

In addition to increasing the amplitude of the cur-rent, thednc mutation as well as IBMX and 8-Br-cAMP seemed to make the kinetics of inactivation ofthe current slightly faster (not shown). However, al-though it was seen consistently with several repeti-tions of the experiment, the effect was not pronouncedenough to be clearly resolved by the available data.

Effect of Adenylyl Cyclase Activatoron Currents

As shown in Figure 1, a decrease in cAMP level inrutcorrelated with a decrease in the DHP-sensitive cur-rent. It was instructive to examine whether therutcurrent could be rescued by forskolin, a cell-perme-able activator of adenylyl cyclase. The effect of 1mMforskolin on therut current is shown in Figure 4.Average current traces from larvae with therut2 mu-tation (a hypomorphic allele) without and with 1mMforskolin are depicted in Figure 4(A,B). The corre-sponding I/V curves including that for the wild type(dotted line) are shown in Figure 4(C). Compared tothe current recorded fromrut2 alone (74.46 4.3%),application of 1mM forskolin increased the current to88.26 3.7% of the wild type. The effect was statis-

tically significant (p , .01). Thus, the effect ofrutmutation could be partially rescued pharmacologi-cally by a drug that activated adenylyl cyclase. Thesedata further strengthen the possibility of cAMP-de-pendent modulation of the L-type current. Forskolinfailed to elicit a further increase in current amplitudein rut1 larvae (data not shown). This is consistent withthe observation that therut1 adenylyl cyclase is de-fective in its responsiveness to forskolin (Dudai et al.,1985). We were unable to record stable currents fromwild-type larvae in the presence of forskolin owing toincreased muscle contraction.

Figure 3 Effect of IBMX on the current. Currents fromCS (L 5 7; F 5 13) and CS with 50mM IBMX (L 5 6; F5 10) are shown in (A,B), respectively. IBMX increased thewild-type current to 133.76 4.4% (p , .001). The I/Vrelationships for CS, CS with 50mM IBMX, and dnc2 (L5 5; F 5 10) are shown in (C).

Calcium Channel Modulation inDrosophila 495

Effect of PKA Inhibitor on Currents

Since a number of physiological actions of cAMP aremediated through PKA, and since phosphorylation isbelieved to be an important factor in modulating ionchannels, there was a likelihood that the current mightbe modulated by PKA. To test this possibility, weused a potent cell-permeable inhibitor of PKA, H-89.Figure 5 shows averaged data for Ca21 currents fromCS,dnc2, andrut2, either alone or in the presence of10 mM H-89 in the bath solution. Figure 5(G) repre-sents the corresponding bar graphs of the peak currentat 210 mV. Compared to control, 10mM H-89 in-hibited the current in wild type anddnc2 to values of76.6 6 5.4% and 80.56 3.3%, respectively. Inhibi-tion of the CS anddnc2 current was statistically sig-

nificant (p , .001). On the other hand, 10mM H-89did not further decrease therut2 current. These resultssuggest the possibility of a PKA-dependent effect onthe current. Thus, the effect of cAMP on the DHP-sensitive current was likely mediated via PKA.

8-Br-cAMP Effect Is Mediated via PKA

If increase in the DHP-sensitive current by highcAMP levels were mediated via PKA, inhibition ofPKA by H-89 would be expected to override the

Figure 5 Effect of H-89 on the current. Average currenttraces from CS,dnc2, and rut2 alone (A–C) or in thepresence of 10mM H-89 in the recording saline (D–F). Bargraph (G) represents the average of peak currents at the testpotential of210 mV for CS (L5 8; F 5 19),dnc2 (L 5 5;F 5 16), andrut2 (L 5 3; F 5 10) with and without 10mMH-89. H-89 inhibited the wild-type and thednc2 current to76.6 6 5.4% (p , .001) and 80.56 3.3% (p , .001),respectively. The current recorded fromrut2 larvae wasunaffected in the presence of 10mM H-89.

Figure 4 Effect of forskolin on the current. (A,B) Averagetraces of currents recorded fromrut2 alone (L5 5; F 5 16)and in the presence of 1mM forskolin in the recording saline(L 5 6; F 5 15), respectively. The corresponding I/Vcurves are shown in (C) along with that for the wild-type(dotted line). Forskolin increased therut2 current from 74.46 4.3% to 88.26 3.7% (p , .01) of the CS level.

496 Bhattacharya et al.

effect of high cAMP concentration. To examine this,the current was recorded in the presence of 10 nM8-Br-cAMP and 10mM H-89. As seen in Figure 6, anincrease in the current observed in the presence of8-Br-cAMP alone was inhibited when H-89 was also

present in the recording solution. Average peak cur-rent traces (210 mV) from CS, CS with 10 nM8-Br-cAMP, and CS in the presence of both 8-Br-cAMP and 10mM H-89 are shown in Figure 6(A). Inthe presence of 8-Br-cAMP alone, the current was118.56 2.9% of the CS value. With the addition of10 mM H-89 in the continued presence of 8-Br-cAMP, the current decreased to 85.86 3.8% of thewild type [Fig. 6(C)]. The I/V relationships, alongwith that fordnc2, are shown in Figure 6(B). The levelof the current observed with both 8-Br-cAMP andH-89 in the solution was close to the level of currentobserved with H-89 alone [compare Figs. 5(G) and6(C)]. Thus, our data suggest that the 8-Br-cAMPeffect on the Ca21 current is likely to be mediatedby PKA.

DISCUSSION

An extensive repertoire of mutations that disrupt stepsin signaling pathways provides a powerful tool toanalyze mechanisms of ion channel modulation inDrosophila. Our experiments with mutations anddrugs that affect the cAMP-PKA pathway stronglysuggest that the DHP-sensitive Ca21 channels inDro-sophilalarval muscles are modulated via this pathway(Fig. 7). Evidence for this comes from several angles.Increase in the current indnc mutants, which haveincreased cAMP levels, and its decrease inrut mu-tants, which have decreased cAMP, point to the likely

Figure 7 Schematic representation of the pathway modu-lating the DHP-sensitive channels inDrosophilalarval mus-cles. Activation of adenylyl cyclase (AC) (rut) leads tocAMP production. This leads to the activation of proteinkinase A (PKA). cAMP is degraded by phosphodiesterase(PDE) (dnc). PKA modulated the DHP-sensitive calciumchannel leading to increased channel current. Pharmacolog-ical agents used in this study to manipulate different steps inthe signaling pathway are shown as (1) and (2) for acti-vators and inhibitors, respectively. The same channel wasalso modulated by the phospholipase C (PLC)-mediatedprotein kinase C (PKC) pathway (see Discussion).

Figure 6 Effect of H-89 and 8-Br-cAMP on the current.(A) Average currents elicited under different recording con-ditions by a test pulse to210 mV (peak current) from aholding potential of 240 mV. (B) Corresponding I/Vcurves. For comparison,dnc2 (L 5 5; F 5 14) is shown asa dotted line. (C) Average peak currents for test pulses to210 mV for CS (L 5 15; F 5 36), CS with 10 nM8-Br-cAMP (L 5 11; F 5 14), and CS with both 10 nM8-Br-cAMP and 10mM H-89 (L 5 7; F 5 11). 8-Br-cAMPincreased the wild-type current to 118.46 2.9%, while thesimultaneous addition of 10 nM 8-Br-cAMP and 10mMH-89 produced only 85.56 3.8% (p , .001) of the wild-type current.

Calcium Channel Modulation inDrosophila 497

effect of cAMP on the current. Involvement of cAMPand PKA is further suggested by the ability of 8-Br-cAMP and IBMX to mimic thednc effect in thewild-type muscles, by the rescue of therut currentclose to the wild-type level by forskolin, and by thereduction of the wild type and thednccurrent by H-89close to the level of therut current. Further experi-ments are needed to determine the nature of the mod-ifications responsible for the modulatory effect ofcAMP on the channels. These modifications may in-volve phosphorylation or some other alteration of thechannel protein, or of an accessory molecule or someother modulatory component.

Intracellular messenger systems play a very impor-tant role in the modulation of ion channels in vertebratecells. Less is known about factors involved in the mod-ulation of Ca21 channels in invertebrates in general, andin Drosophila in particular. The cAMP cascade modu-lates Ca21 currents in the skeletal muscles, cardiac mus-cles, and smooth muscles of vertebrates (McDonald etal., 1994). L-type Ca21 channels inDrosophilaare phar-macologically somewhat different from the vertebrateL-type channels (see Introduction). In general, inverte-brate Ca21 channels have distinct differences in theirpharmacology compared to the higher vertebrate species(Skeer et al., 1996). Despite these differences that havearisen during evolution, and despite pharmacologicaldifferences between components of cAMP cascade be-tween Drosophila and mammals (Henkel-Tigges andDavis, 1989), the DHP-sensitive channels from larvalmuscles ofDrosophilashare modulatory pathways withvertebrate L-type channels. This implies the importanceof the cAMP-PKA pathway which seems to have beenconserved during evolution. It will be interesting toexamine whether the phenylalkylamine-sensitive, DHP-resistant channels (Pelzer et al., 1989) from the brainmembrane preparations ofDrosophilaare modulated bythe same pathway. There are indications that cAMP mayalso play a role in modulating a different type of Ca21

channels in cleavage-arrested embryonic neurons ofDrosophila,as the addition of theophylline and dibutyrylcAMP increases the current through these channels morethan it increases the background inward current(Alshuaib and Byerly, 1996).

Unlike in vertebrate muscle fibers, calcium formuscle contraction in insects enters the cell via volt-age-gated Ca21 channels (Hille, 1992). Regulation ofCa21 channels thus plays a critical role in regulatingmuscle excitability. Modulation ofDrosophila Ca21

channels via cAMP cascade assumes additional func-tional significance owing to the role of this pathway inregulating the contraction of insect muscles.

Both thedncand therut mutations were identified fora defect in learning and memory and were subsequently

found to have alterations in cAMP metabolism. ThecAMP-PKA pathway has been shown to play an impor-tant role in long-term facilitation and sensitization inother invertebrates such as crayfish andAplysia(Dixonand Atwood, 1989; Kandel and Schwartz, 1982). InDrosophila,cAMP plays an important role in learningand memory (Davis et al., 1995), synaptic plasticity(Zhong and Wu, 1991), habituation (Engel and Wu,1996), growth cone motility (Kim and Wu, 1996), andethanol intoxication (Moore et al., 1998). Shotwell(1983) correlated reduced PDE activity indnc mutantsto their learning deficiency in third-instar larvae. There isalso evidence of the involvement of ion channels, in-cluding Ca21 channels, in learning and memory (Collinand Alkon, 1992; Disterhoft et al., 1993). Mushroombody neurons in insects are involved in olfactory learn-ing (Davis, 1993). Thedncand therut gene products arepreferentially expressed in these neurons (Nighorn et al.,1991; Han et al., 1992). It will be instructive to examinewhether the cAMP cascade is involved in the modula-tion of Ca21 channels in the mushroom body neurons ofDrosophila,if these channels are affected in thedncandthe rut mutations, and if such a modulation has a rela-tionship with the learning and memory deficit in thedncand therut mutants. Recently, Davis et al. (1998) dem-onstrated the regulation of neurotransmitter release via aPKA-dependent retrograde pathway from larval musclesto motor neurons. It will be interesting to examinewhether our data on the modulation of Ca21 channels inDrosophila larval muscles via a cAMP pathway hasimplications for such a regulation of neurotransmitterrelease and if PKA-mediated neurotransmitter releaseplays a role in learning and memory.

A mutation such asdnc or rut can influence acurrent either as an acute effect, i.e., altered cAMPconcentration at a particular time may effect the chan-nels at that time, or as a chronic effect, i.e., a pro-longed alteration in the cAMP concentration mayhave indirect consequences such as developmentalregulation of the channels. Acute application of phar-macological agents in our experiments was able tomimic the mutational effect in the wild-type current aswell as rescue the mutant current close to the wild-type level. This strongly argues in favor of thedncand therut effect on the current being an acute ratherthan a chronic effect.

Experiments with mutations and drugs that disruptthe phosphoinositide cascade show that the L-typechannels studied in this report are also modulated bythe phospholipase Cb (PLC)–diacylglycerol (DAG)–protein kinase C (PKC) pathway (Gu and Singh,1997). It will be interesting to examine whether thetwo pathways (represented by cAMP-PKA and thePLC-DAG-PKC steps) leading to the modulation of

498 Bhattacharya et al.

the channels act independently of each other, if theyoverlap or interact, or if they converge and result inthe same eventual molecular modification in the twocases.

Further studies are needed to analyze the pathwaysupstream of adenylyl cyclase that may play a role inmodulating theDrosophila L-type channelsin vivo.Neurotransmitters such as octopamine and pituitaryadenylyl cyclase activating peptide, present in theneuromuscular synaptic cleft inDrosophila larvae,may activate G protein–coupled receptors, therebyactivating adenylyl cyclase and/or phospholipase C.These and/or other signal transduction pathways mayplay a significant role in larval physiology by modu-lating ion channel function and thereby muscle excit-ability. In this context, both the cAMP-PKA and thePLC-PKC signal transduction machineries becomeimportant pathways for further modulatory studies.There are several additional mutant as well as trans-genic strains ofDrosophilaavailable that affect stepsin the PLC and in the cAMP-mediated pathways.Combination of these strains with pharmacologicalagents can serve as a useful tool to understand themechanisms of modulation of the L-type channels.

This work was supported by grants from NIH (GM-50779) and NSF (IBN-9011427, MCB-9604457) to SS, anda grant-in-aid from Sigma Xi, The Scientific Research So-ciety, to AB.

REFERENCES

Aceves-Pina EO, Brooker R, Duerr JS, Livingstone MS,Quinn WG, Smith RF, Szibe PP, Tempel BL, Tully TP.1983. Learning and memory inDrosophila,studied withmutants. Cold Spring Harb Symp Quant Biol 48:831–840.

Alshuaib WB, Byerly L. 1996. Modulation of membranecurrents by cyclic AMP in cleavage-arrestedDrosophilaneurons. J Exp Biol 199:537–548.

Alshuaib WB, Mathew MV. 1998. Potassium current inDrosophilaneurons is increased by eitherduncemutationor cyclic AMP. J Neurosci Res 52:521–529.

Asztalos Z, Lossos M, Friedrich P. 1991. On the pharma-cological phenocopying of memory mutations inDro-sophila:alkylxanthines accelerate memory decay. BehavGenet 21:495–511.

Byers D, Davis RL, Kiger JA. 1981. Defect in cyclic AMPphosphodiesterase due to theduncemutation of learningin Drosophila melanogaster.Nature 289:79–81.

Chen CN, Denome S, Davis RL. 1986. Molecular analysisof cDNA clones and the corresponding genomic codingsequences of theDrosophila dunce1 gene, the structuralgene for cAMP phosphodiesterase. Proc Natl Acad SciUSA 83:9313–9317.

Chopra M, Singh S. 1994. Developmental temperature se-lectively regulates a voltage-activated potassium currentin Drosophila.J Neurobiol 25:119–126.

Collin C, Alkon DL. 1992. Neural correlates of memorystorage; the role of ion channels. Ion Channels 3:159–175.

Crossley AC. 1978. The morphology and development ofthe Drosophila muscular system. In: Ashburner M,Wrights TRF, editors. The genetics and biology ofDro-sophila, Vol. 2B. New York: Academic Press. p. 499–560.

Davis GW, DiAntonio A, Petersen SA, Goodman CS. 1998.Postsynaptic PKA controls quantal size and reveals aretrograde signal that regulates presynaptic transmitterrelease inDrosophila.Neuron 20:305–315.

Davis RL. 1993. Mushroom bodies andDrosophila learn-ing. Neuron 11:1–14.

Davis RL, Cherry J, Dauwalder B, Han PL, Skoulakis E.1995. The cyclic AMP system andDrosophila learning.Mol Cell Biochem 149/150:271–278.

Davis RL, Kiger JA. 1981.duncemutants ofDrosophilamelanogaster:mutants defective in the cyclic AMP phos-phodiesterase enzyme system. J Cell Biol 90:101–107.

Delgado R, Hidalgo P, Diaz F, Latorre R, Labarca P. 1991.A cyclic AMP-activated K1 channel inDrosophilalarvalmuscle is persistently activated indunce.Proc Natl AcadSci USA 88:557–560.

Delgado R, Latorre R, Labarca P. 1992. K1-channel block-ers restore synaptic plasticity in the neuromuscular junc-tion of dunce,aDrosophilalearning and memory mutant.Proc R Soc Lond B 250:181–185.

Disterhoft JF, Moyer JR, Thompson LT, Kowalska M.1993. Functional aspects of calcium-channel modulation.Clin Neuropharmacol 16(Suppl 1):S12–S24.

Dixon D, Atwood HL. 1989. Adenylate cyclase system isessential for long-term facilitation at the crayfish neuro-muscular junction. J Neurosci 9:4246–4252.

Drain P, Folkers E, Quinn WG. 1991. cAMP-dependentprotein kinase and the disruption of learning in transgenicflies. Neuron 6:71–82.

Dudai Y, Jan Y-N, Byers D, Quinn WG, Benzer S. 1976.dunce,a mutant ofDrosophiladeficient in learning. ProcNatl Acad Sci USA 73:1684–1688.

Dudai Y, Sher B, Segal D, Yovell Y. 1985. Defectiveresponsiveness of adenylate cyclase to forskolin in theDrosophila memory mutantrutabaga. J Neurogen 2:365–380.

Engel JE, Wu CF. 1996. Altered habituation of an identifiedescape circuit inDrosophilamemory mutants. J Neurosci16:3486–3499.

Feany MB. 1990. Rescue of the learning defect indunce,aDrosophila learning mutant, by an allele ofrutabaga,asecond learning mutant. Proc Natl Acad Sci USA 87:2795–2799.

Gho M, Mallart A. 1986. Two distinct calcium activated K1

currents in larval muscle fibers ofDrosophila.PfluegersArch 407:526–533.

Gielow ML, Gu GG, Singh S. 1995. Resolution and phar-

Calcium Channel Modulation inDrosophila 499

macological analysis of the voltage-dependent calciumchannels ofDrosophila larval muscles. J Neurosci 15:6085–6093.

Greenberg RM, Striessnig J, Koza A, Devay P, GlossmannH, Hall LM. 1989. Native and detergent-solubilizedmembrane extracts fromDrosophilaheads contain bind-ing sites for phenylalkylamine calcium channel blockers.Insect Biochem 19:309–322.

Gu GG, Singh S. 1997. Modulation of the dihydropyridine-sensitive calcium channels inDrosophilaby a phospho-lipase C-mediated pathway. J Neurobiol 33:265–275.

Han PL, Levin LR, Reed RR, Davis RL. 1992. Preferentialexpression of theDrosophila rutabagagene in mush-room bodies, neural centers for learning in insects. Neu-ron 9:619–627.

Henkel-Tigges J, Davis RL. 1989. Rat homologs of theDrosophila duncegene code for cyclic AMP phosphodi-esterases sensitive to rolipram and RO 20-1724. MolPharmacol 37:7–10.

Hille B. 1992. Ionic channels of excitable membranes. Sun-derland, MA: Sinauer Associates.

Jan L, Jan Y. 1976. Properties of the larval neuromuscularjunction in Drosophila melanogaster.J Physiol (Lond)262:189–214.

Kandel ER, Schwartz JH. 1982. Molecular biology of learn-ing: modulation of transmitter release. Science 218:433–443.

Kim YT, Wu CF. 1996. Reduced growth cone motility incultured neurons fromDrosophilamemory mutants witha defective cAMP cascade. J Neurosci 16:5593–5602.

Kokate TG, Sperelakis N, Heiny J. 1993. Stimulation of theslow calcium current in bullfrog skeletal muscle fibers bycAMP and cGMP. Am J Physiol 265:C47–C53.

Levin LR, Han PL, Hwang PM, Feinstein PG, Davis RL,Reed RR. 1992. TheDrosophila learning and memorygene rutabaga encodes a Ca21/calmodulin-responsiveadenylyl cyclase. Cell 68:479–489.

Levine JD, Casey CI, Kalderon DD, Jackson FR. 1994.Altered circadian pacemaker functions and cyclic AMPrhythms in theDrosophila learning mutantdunce.Neu-ron 13:967–974.

Livingstone MS, Sziber PP, Quinn WG. 1984. Loss of calci-um/calmodulin responsiveness in adenylate cyclase ofruta-baga,a Drosophila learning mutant. Cell 37:205–215.

McDonald TF, Pelzer S, Trautwein W, Pelzer DJ. 1994.Regulation and modulation of calcium channels in car-diac, skeletal and smooth muscle cells. Physiol Rev 74:365–507.

Moore MS, DeZazzo J, Luk AY, Tully T, Singh CM,Heberlein U. 1998. Ethanol intoxication inDrosophila:genetic and pharmacological evidence for regulation bythe cAMP signaling pathway. Cell 93:997–1007.

Nighorn A, Healy MJ, Davis RL. 1991. The cyclic AMPphosphodiesterase encoded by theDrosophila gene isconcentrated in the mushroom body neuropil. Neuron6:455–467.

Nighorn A, Qiu Y, Davis RL. 1994. Progress in understand-ing the Drosophila dnclocus. Comp Biochem Physiol108B:1–9.

Pelzer S, Barhanin J, Pauron D, Trautwein W, Lazdunski M,Pelzer D. 1989. Diversity and novel pharmacologicalproperties of calcium channels inDrosophilabrain mem-branes. EMBO J 8:2365–2371.

Rampe D, Kane MJ. 1994. Activators of voltage-dependentL-type calcium channels. Drug Dev Res 33:344–363.

Shotwell SL. 1983. Cyclic adenosine 39:59-monophosphatephosphodiesterase and its role in learning inDrosophila.J Neurosci 3:739–747.

Singh S, Wu CF. 1989. Complete separation of four potas-sium currents inDrosophila.Neuron 2:1325–1329.

Skeer JM, Norman RI, Sattelle DB. 1996. Invertebratevoltage-dependent calcium channel subtypes. Biol Rev71:137–154.

Sperelakis N, Xiong Z, Haddad G, Masuda H. 1994. Reg-ulation of slow calcium channels of myocardial cells andvascular smooth muscle cells by cyclic nucleotides andphosphorylation. Mol Cell Biochem 140:103–117.

Stewart BA, Atwood HL, Renger JJ, Wang J, Wu CF. 1994.Improved stability ofDrosophila larval neuromuscularpreparations in haemolymph-like physiological solutions.J Comp Physiol 175:179–191.

Tully T. 1991. Genetic dissection of learning and memoryin Drosophila melanogaster.In: Madden J, editor. Neu-robiology of learning, emotion and affect. New York:Raven Press. p 29–66.

Varadi G, Mori Y, Mikala G, Schwartz A. 1995. Moleculardeterminants of Ca21 channel function and drug action.Trends Biochem Sci 16:43–49.

Wu CF, Haugland FN. 1985. Voltage clamp analysis ofmembrane currents in larval muscle fibers ofDrosophila:alteration of potassium currents inShakermutants. J Neu-rosci 5:2626–2640.

Xiong Z, Sperelakis N, Fenoglio-Preiser C. 1994. Regula-tion of L-type calcium channels by cyclic nucleotides andphosphorylation in smooth muscle cells from rabbit por-tal vein. J Vasc Res 31:271–279.

Zhong Y. 1995. Mediation of PACAP-like neuropeptidetransmission by coactivation of ras/raf and cAMP signaltransduction pathways inDrosophila. Nature 375:588–591.

Zhong Y, Budnik V, Wu CF. 1992. Synaptic plasticity inDrosophilamemory and hyperexcitable mutants: role ofcAMP cascade. J Neurosci 12:644–651.

Zhong Y, Wu CF. 1991. Altered synaptic plasticity inDrosophila memory mutants with a defective cyclicAMP cascade. Science 251:198–201.

Zhong Y, Wu CF. 1993. Differential modulation of potas-sium currents by cAMP and its long-term and short-termeffects: dunce and rutabaga mutants of Drosophila.J Neurogen 9:15–27.

500 Bhattacharya et al.