angela r. howe and d. james surmeier- muscarinic receptors modulate n-, p-, and l-type ca^2+...

8/3/2019 Angela R. Howe and D. James Surmeier- Muscarinic Receptors Modulate N-, P-, and L-type Ca^2+ Currents in Rat

1/12

The Journal of Neuroscience, January 1995, 15(l): 456469

Muscarinic Receptors Modulate N-, P-, and L-type Ca2+ Currents inRat Striatal Neurons through Parallel PathwaysAngela R. Howe and D. James SurmeierDepartment of Anatomy and Neurobiology, College of Medicine, University of Tennessee, Memphis, Tennessee 38163

Muscarinic modulation of calcium currents was studied inacutely isolated striatal neurons from the adult rat using thewhole-cell configuration of the patch-clamp technique. Mus-carinic agonists reduced calcium currents through two dis-tinct signaling pathways. One pathway depended upon PTX-sensitive G-proteins and targeted N- and P-type currents.The other pathway depended upon PTX-insensitive G-pro-teins and was rendered inactive by high intracellular con-centrations of BAPTA and targeted L-type currents. Themodulation of N- and P-type currents was relieved by strongdepolarizing prepulses, whereas the modulation of L-typecurrents was not. These findings support the proposition thatparallel signaling pathways exist between muscarinic re-ceptors and calcium channels.[Key words: muscarine, acetylcholine, neuromodulation,dihydropyridine, w-agatoxin, w-conotoxin, patch clamp, acutedissociation]Manipulation of intrastriatal cholinergic signaling has been along-standingstrategy or treating basalgangliadisorders Calne,1980). The intrastriatal cholinergic network arisesexclusivelyfrom large interneurons with widely branching axonal arbors(Bolam et al., 1984; Phelpset al., 1985; DiFiglia, 1987; Wilsonet al., 1990). These nterneurons richly innervate neurons hatproject from the striatum, resulting in levels of acetylcholineand its synthetic enzymes that are among the highest of anybrain region (Vilaro et al., 1991). In spite of its functional im-portance, the impact of acetylcholine release n the excitabilityof striatal neurons s not well understood.Within the striatum, the postsynaptic effectsof acetylcholineare mediated primarily by muscarinic receptors. Five musca-rinic receptorshave been dentified through molecular cloningand n situ hybridization (Kubo et al., 1986; Bonner et al., 1987,1988; Bonner, 1989; Laio et al., 1989). Thesesubtypescan begrouped nto two broad categoriesbasedupon their preferentialcoupling o second-messengerystems.The m2 and m4 receptorsubtypes nhibit adenylate cyclase for review, seeHulme et al.,1990; Nathenson, 1990). On the other hand, the ml, m3, and

Received Mar. 10, 1994; revised May 24, 1994; accep ted J une 23, 1994.We gratefully acknowled ge the gift of the PTX -a protomer from Dr. SusanSenogles as well as her helpful advice on its use and handling. We thank Dr. LorinaDudkin for her excellent technical assistance as well as Drs. Bob Foehring, ReeseScroggs, and Charles Wilso n for critical reading of the manuscript. This work wassuuoorted by USPH S Grants NS 28889 and NS 26473 to D.J.S. and MH-10400to-A.R.H. -Correspondence should he addressed to D. James Surmeier, Ph.D., D epartmentof Anatomy and Neurobiology College of Medicine, University of Tennessee, 855Monroe Avenue, Memphis, TN 38 163.

Copyright 0 1995 Society for Neuro science 0270-6474/95/150458-12$05 .00/O

m5 receptor subtypes activate phospholipaseC (PLC) whichreleasesnositol 1,3,5 riphosphate IP,) and diacylglycerol DAG)from membrane phospholipid (Hokin and Hokin, 1954; Ber-ridge and Irvine, 1989).The principal neuron of the striatum-the medium-spiny neuron-expresses high levelsof two of thesefive receptors-the ml and m4 subtypes (Andersen and MC-Kinney, 1988; Bernard et al., 1992), suggestinghat acetylcho-line may be able to act through at least two signalingpathwaysto affect excitability.One peripheral neuron expressingboth ml and m4 receptorsubtypes that has been extensively studied is the rat superiorcervical ganglion (SCG) neuron (Beechet al., 1991, 1992; Ber-nheim et al., 1991, 1992; Mathie et al., 1992). In thesecells,muscarinic agonists inhibit N- and L-type calcium currentthrough fast and slow signalingpathways. The fast pathway ismediatedprimarily by the m4 receptor, is membranedelimited,and dependent upon the activation of a pertussis oxin (PTX)sensitiveG-protein. This pathway targetsonly N-type currents.The slow pathway is mediated primarily by the m 1 receptor, isdependentupon the activation of a PTX insensitive G-protein,and involves a BAPTA-sensitive, diffusible secondmessenger.This pathway targets both N- and L-type currents.We have attempted to test whether this model of muscarinicmodulation can be generalized to brain neurons and striatalneurons, n particular. To do so, we have usedwhole-cell volt-age-clamp ecording techniqueswith acutely isolatedadult neu-rons. As shown below, activation of muscarinic receptors instriatal neurons triggers two parallel signaling cascades hatmodulate N-, P-, and L-type currents. Although the two sig-naling cascades re similar in many respects o thosedescribedin SCG neurons, the PTX-insensitive pathway in striatal neu-rons only targets L-type currents. This difference suggestshatmuscarinic signaling mechanisms, particularly those of theBAPTA-sensitive pathway, may not be identical in peripheraland central neurons.Materials and MethodsAcute-dissociation. Neostriatalneuronsrom adult (>25 d postnatal)male atswereacutely solated singproceduresimilar o thosepre-viously eported by our group (Surmeier et al., 1991, 1992). In brief,ratswere ecapitatedndermetofanenesthesia.rainswerehenquicklyremoved, ced, andblockedprior to slicing.Slices 400pm) werecutusing a Vibr oslic e (Campden Instr., London, England ) and transferredto a ow calcium 1 b0&), HEPES-bufferedaline olution in mM:140Na isethionate, KCl, 4 MgCl,, 0.1 CaCl,,23glucose, 5HEPES, H7.4, 300-305mOsm/liter).Sliceswere hen ncubatedor l-6 hr in aNaHCO,-buffered alinebubbled ith 95% 0,, 5% CO, (in mM: 12 6NaCl. 2.5 KCl. 2 M&l,. 1 CaCl,. 1.25NaH,PO,. 26 NaHCO,. 10glucoie,1 pyrubic acyd,PH = 7.4:300-305&O&liter). Slices%erethen emovednto the owcalcium uffer,andwith theaidofa dissectingmicroscope,egions f the dorsalprecommissuraltriatumwere e-


2/12

The Journal of Neuroscience, January 1995, 15(l) 459

moved and placed in a Cell Stir (Wheaton, Millville, NJ) containingprotease (type XIV, 1.5 mg/ml, Sigma Chemical Co., St. Louis, MO)in HEPES buffered Hanks balanced salt solution (HBSS, Sigma Chem-ical Co.) at 35C. Other enzymes, such as trypsin and papain, were alsoused with comparable results; however, protease treatment providedthe highest yield of viable cells. After 20-30 min of the enzyme treat-ment, the tissue was rinsed several times in the low calcium buffer andmechanically dissociated using a series of fire-polished Pasteur pipettes.The cell suspension was then plated into a 35 mm Lux petri dish mount-ed on the stage of an inverted microscope containing 1 ml of HEPES-buffered HBSS saline. After allowing the cells to settle (about 5 min),the solution bathing the cells was changed to our normal external re-cording solution.Whole-cellecordings.hole-cell recordings employed standard tech-niques (Hamill et al., 198 1; Surmeier et al., 1992). Electrodes were pulledfrom Coming 7052 or LA1 6 (Dagan Instr ., Inc., Minneapolis, MN) glassand fire polished prior to use. The internal recording solution consistedof (in mM) 180 N-methyl-D-glucamine (NMG), 40 HEPES, 4 MgCl,,O-10 EGTA, or 0.1-20 BAPTA, 12 phosphocreatine, 2 Na,ATP, 0.2Na,GTP, 0.1 leupeptin, pH = 7.2 with H,PO,, 270-275 mOsm/liter .The external recording solution consisted of (in mM): 135 NaCl, 20CsCl, 1 MgCl,, 10 HEPES, 0.001 TTX, 5 BaCl,, pH = 7.5 with NaOH,300-305 mOsm/liter. The nicotinic receptor antagonist, mecamylamine(10 PM , RBI, Inc., Natick, MA) was added to all solutions to blocknonmuscarinic receptors. Nifedipine and (-)Bay K 8644 (RBI, Inc.,Natick, MA) were made up as concentrated stocks in 95% ethanol andthen diluted immediately prior to use. These solutions were protectedfrom ambient light. Final ethanol concentrations never exceeded 0.05%.Although previous experiments had shown that this concentration hasno effecton currents, ethanol controls were run for every cell. w-ConotoxinGVIA (w-CgTx, Peninsula Lab., Belmont, CA; RBI, Inc.) and w-agatoxinIVA (w-AgTx, Peptides International, Louisville, KY) were prepared asconcentrated stocks in water, then frozen in aliquots that were thawedand diluted immediately prior to use. When using o- AgTx, cytochromec (Calbiochem, San Diego, CA) was added to all solutions (0.01%) toprevent nonspecific binding to plastics (see Mintz et al., 1992). Car-bachol, muscarine, and atropine (RBI Inc.) were prepared as fresh con-centrated stocks in water immediately prior to use. GTPrS or GDP-&S (Calbiochem) were substituted for GTP in some experiments at aconcentration of 600 PM and l-5 mM, respectively. To determine thePTX sens itivi ty of the signaling pathway, we utilized the catalytic do-main of the toxin (PTX a protomer, List Laboratories, Campbell, CA)at a concentration of 10 &ml. The protomer was added to the solutionused to backfill the recording pipette. Dialysis o f the protomer into thecell appeared to be complete within 10 min (Pusch and Neher, 1988).Electrode resistances were typ icall y 3-6 MQ in the bath. Recordingswere obtained using a Dagan 3900 Integrating patch clamp equippedwith the 39llA Whole-cell Expander module (Dagan Instr. , Minne-apolis, MN) or an Axon Instruments 200 patch clamp. Recordings weremonitored and controlled with a PC 486 clone using ~CLAMP (v. 5.0-5.5) with a 125 kHz interface (Axon Instr. , Inc. , Foster Cit y, CA). Afterseal rupture, series resistance compensation (70-80%) was routinelyemployed; given the amplitudes of our currents, series resistance errorsshould never have exceeded 2-3 mV. Recordings were made only frommedium-sized neurons (somal diameter 5-10 pm) that had at most oneor two short processes. The adequacy of voltage control was assessedafter compensation by examining the tail currents generated by strongdepolarizations. Cells in which tail currents did not decay rapidly andsmoothly at subthreshold potentials were eliminated from the study.Potentials were not corrected for the liquid junction potential with theNMG internal solution, which was estimated to be about 7 mV.Drugs were applied using a grav ity-fed sewer pipe system. The arrayof application capillaries (- 150 rrn i .d.) was positioned a few hundredmicrometers away from the cell under study. Solution changes weremade by altering the position of the array with a DC drive systemcontrolled by a microprocessor-based controller (Newport-Klinger, Inc.,Irvine, CA). Complete solution changes were achieved within less than1 set, as udged by changes in reversal potential. As a further check, thetime course of channel block by Cd2+ (200 PM) was also determined.Cd2+ block of whole-cell Ca2+ current developed with a time constantless than 500 msec in all cells tested (n = 5).Data were gathered using a combination of voltage step and rampprotocols. In most cases, those currents not blocked by Cd*+ (200 PM )were subtracted from control records (in nearly all cells, there was littleor no evidence for Cd2+-insensitive currents a t membrane potentials

below 0 mV, above this potential, small tail currents that were sensitiveto Cl- replacement were observed). Peak current-voltage plots derivedfrom Ba*+ currents evoked with short (10-30 msec) step protocols andfrom slow voltage ramps (0.25-0.5 mV/msec) were very similar in mostcells in which they were compared (n = 10/l 1). Faster ramps yieldedcurrent-voltage curves that were shif ted toward more depolarized po-tentials (see Bargas et al., in press).Data analysis. ose-response curves were fit with an Langmuir iso-therm of the form: Swat = I,,, ,,,, ;( 1 + [M]/EC,,))n + I,,,,,,,,, where I,,,was the maximum ramp current, I,,,,,,, was the blockable current, [M]was the agonist concentration, EC,, was the concentration of agonistproducing 50% block, Icc,,,a.,as the portion of the current resistant tomodulation, and n was usually set to one (varying n between 0.5 and1.5 did not signif icantly enhance the fits for the data described here).Fitting was done using a least squares criterion with a commercial lyavailable stat istical package (SYSTAT, Inc., Evanston, IL) running ona Macintosh computer. Statistical analyses also used this program. Sta-tistics on large samples (n > 20) are presented as means (w) + SEM.For smaller samples where means do not necessarily give a good measureof central tendency, medians and ranges are given. When the samplesizes allowed, box plots o f the data were drawn. In the box plot, themedian is represented as the central bar of the box, the edges of the boxare the interquartiles (technically fourths) , the whiskers are lines drawnto the most extreme points in the sample that are not outliers (definedas points beyond interquartile + 1.5.interquartile range); outliers areshown as circles or asterisks (Tukey, 1977).ResultsMuscarinic agonists eversibly inhibit Cal+ current through aG-protein signalingpathwayThe muscarinic agonists, muscarine and carbachol, reversiblyinhibited whole-cell Ca2+current in rat striatal neurons Fig. 1).Muscarine dose-response lots were well fit with a single so-therm having an EC,, in the low nanomolar range. Shown inFigure 1C is a representative dose-response lot. The EC,, inthis neuron was 1.0 nM; similar resultswere obtained in threeother cells (EC,, = 6.5 + 1.4 nM). Saturating concentrationsofmuscat-me1 PM) reduced the peak whole-cell Ca2+current anaverageof 46.0 f 6.4% (n = 11). But, as shown n the box plotin Figure 3C, the relative reduction in current amplitude atsaturating concentrations varied somewhat from cell to cell.Carbachol dose-response lots were well fit with a single so-therm having EC,,s in the low micromolar range. A represen-tative dose-response lot from a singleneuron is shown n Fig-ure 1D. The EC,, in this neuron was2.7PM; similar resultswereseen n 11 other neurons EC,, = 3.25 + 0.89 PM). On average,saturating concentrations of carbachol (10 PM) were not as ef-fective in reducing peak currents asmuscarine mean eduction= 32.3 * 1.2%, n = 106). However, as shown n the box plotshown n Figure 1D, there was variability in the magnitudeofthe modulation from cell to cell. As shown in Figure 2A, themodulation by muscarine was blocked by the muscarinic re-ceptor antagonist atropine (1 FM) (n = 5); similar resultswereobtained with carbachol n = lo), arguing hat the effectsof bothagonistswere mediated by muscarinic receptors.To test for the involvement of a GTP-binding protein (G-protein) in the modulation, we utilized the nonhydrolyzableGTP analog, GTPyS. Once bound, GTPrS cannot be metab-olized, resulting in the persistent activation of the G-proteinalpha subunit (for review, see Breitwieser and Szabo, 1988).Dialysis for 4-5 min with GTPyS (600 PM, 0 GTP) preventedthe complete reversal of the reduction in currents produced bycarbachol (Fig. 2B)-consistent with the involvement of aG-protein in the signalingpathway. Shown n the inset of Figure2B is a box plot summarizing the percent recovery of the car-bachol modulation 1O-l 5 min after seal upture in control and


3/12

460 Howe and Surme ier * Muscarinic Modulation of Ca*+ Currents

1.0 1.5 210 2:5 3.0TIME0 (minutes)

n=ll

0 100 200TIME (ms)

0.8 -

0.6.

BO- _g;; 60-s:40-sex20-

0 n=1060.6 A-14 -12 -10 -8 -6 -4LOG (muscarine) YM -2LOG (carbachol)Figure 1. Muscarinic agonists reduce Ca*+ currents in a reversible, dose-dependent fashion. Whole-cell calcium current was elicited by a rampvoltage protocol (0.5 mV/sec). Data points represent peak I, evoked by the ramp. A, Plot ofpeak current as a function oftime during the applicationof carbachol (10 PM). The timing of carbachol application is coded by the bar above the plot and the shading of the symbols . B, Representativeramp currents taken f rom points labeled 1 and 2 in A. C, A representative dose-response curve for muscarine was generated using increasingconcentrations of agonist. The data points were well fi t with a single isotherm having an ECSo of 1.0 nM. The inset is a box plot of the percentreduction in the presence of muscarine in 11 cells. D, A representative dose-response curve for carbachol was generated as in C. The data pointswere fi t with a single isotherm having an EC,, of 2.7 PM. The inset is a box plot of the percent reduction in the presence of carbachol in 106 cells.Recordings were made in the absence of extrinsic calcium chelators.

GTP$S dialyzed cells. Dialysis with l-5 mM GDP-P-S did noteliminate the modulation produced by carbachol.Lowering theATP concentration in the internal solution (1 mM) did not sig-nificantly enhance he effects of GDP-P-S. Similar negative re-sultswith GDP-P-S have been eported for modulations n otherneuronswith known G-protein involvement (Ikeda and Scho-field, 1989).N-, L-, and P-type components f the whole cell calciumcurrent are modulatedWhole-cell calcium currents in adult, medium-spiny striatalneurons are of the high-voltage-activated (HVA) type. Fourclasses f Ca*+current contribute to whole-cell currents n thesecells: N-, L-, and P-type currents and a fourth, pharmacologi-cally uncharacterizedcurrent (Bargas t al., in press).We utilizedthe calcium channel blockers w-conotoxin GVIA (o-CgTx),w-agatoxin IVA (w-AgTx) and nifedipine to determine whichcomponentsof the whole-cell calcium current are modulatedby the muscarinic agonist carbachol.

To simplify the task of determining whether P-type currentswere modulated, N- and L-type currents were first eliminatedwith saturatingconcentrationsof the N-channelblocker w-CgTx(2 PM) and the L-channel blocker nifedipine (5 PM). In the pres-enceof theseantagonists,carbachol nhibited a significant com-ponent of the remaining current. A plot of the peak currentevoked by a voltage ramp asa function of time and drug treat-ment in a typical neuron is shown n Figure 3A. In this cell, thesubsequent ddition of the P-channelblocker w-AgTx (100 nM)reduced he effectsof carbachol,demonstrating hat P-type cur-rents were modulated (Figure 34. Washing led to a partialrecovery that was argely attributable to reversal of the nifedi-pine block of L-type currents (Hoehn et al., 1993; Bargaset al.,in press).A statistical summary s shown n the inset; ascan beseen rom the box plot, the addition of w-AgTx (in the presenceof N- and L-type blockers) reduced the modulation producedby carbachol by more than 80% in most neurons n = 7). It isunclear whether the small residualmodulation seenn the pres-enceof all three channel antagonists epresentsmodulation of


4/12


750J--;;r;;------- ~--_--_----- ----"- atrdne wash 1

-13004~-2.5 3.5 4.0 4.5 510TIME (minutes)

---/5.5

308

256

204$00152 z B& 60L

100 E8 2010 11 12 iiiIl..? ,.:- r--\ a In, L-l

I IIVIE (manures) control GTPySFigure 2. A muscariniceceptor ndGTP-binding rotein G-protein)are nvolved in the signaling athway.A, The modulationwaselimi-nated n the presencef the muscarinic ntagonist tropine 1 PM) n-dicating he nvolvementof the muscariniceceptorn the modulatorypathway. , waselicitedby a 20 msec tep o 0 mV from a holdingpotentialof -90 mV. B, Substitution f GTP in the internalsolutionwith the nonhydrolyzable TP analog,GTPrS (600 M) resultedn anirreversible uppressionf I, in responseo receptor ctivationby 10WM carbachol. he insetshowshe averageecovery rom muscarinicsuppressionor cellswith GTP versusGTPyS in the nternalsolution.

unblocked L-, N-, or P-type currents or modulation of the re-sistantcurrent. Nevertheless, he impact of w-AgTx on the mod-ulation clearly demonstrates hat P-type currents were modu-lated.A similar paradigm was used o determine whether N-typecurrents were modulated by activating muscarinic receptors. nthe presenceof L- and P-type channel antagonists,carbacholreduced a significant component of the remaining current. Aplot of the peak current evoked by a voltage ramp asa functionof time and drug application in an exemplary neuron is shownin Figure 3B. The subsequent ddition of the N-channel blockerw-CgTx reduced the modulation, demonstrating that N-typecurrents weremodulated. Washing ed to a partial recovery thatwas argelyattributable to reversalof the nifedipine block. Shownin the inset s a box plot summary of the percentage eductionin the carbachol modulation produced by w-CgTx in the pres-ence of nifedipine and w-AgTx (n = 4).Although its modulation hasbeen eported infrequently (Ber-nheim et al., 1992; Mathie et al., 1992; Sayer et al., 1992; for

0 Y, I I I I I I50 6.1 7.2 63 9.4 10.5 116 12.7 13.6II TIME (minutes)u 500

Fa 4000m5 300:5 20005h 100015 00 5.75 6.50 7.25 8 00TIME (mtnutes)

Figure 3. N- and P-typecurrents re modulated y muscarinic go-nists.A, The modulation f P-typecurrentwasstudiedn the presenceof the L-channel lockernifedipine 5 NM)and the N-channel lockerw-CgTx 1PM). Carbachol(l0 M) inhibiteda significant omponent fthe remaining urrent.Subsequentdditionof the P-channel lockerw-AgTx (100 nM) eliminated he modulationdemonstratinghe in-volvementof P-typecurrents. he nsetshows statistical ummary fthe percentage odulation liminated y the additionof o-AgTx. B,The modulation f N-type currentwasstudiedn the presencef theL-channel lockernifedipine 5 PM)and heP-channellockerw-AgTx(100nM). Carbachol 10PM ) inhibiteda significant omponent f theremainingurrent.Subsequentddition f theN-channel locker -CgTx(1 PM ) eliminated he modulation emonstratinghe involvementofN-typecurrents. he nset hows statistical ummary fthe percentagemodulation liminated y the additionof w-CgTx.review, seeTsien et al., 1988) L-type current wasalso modu-lated by carbachol in striatal neurons. An approach similar tothat described or N- and P-type currents consistently revealedan occlusionof the carbacholmodulation by treatment with theL-type channelantagonistnifedipine (5-10 PM) (n = 4, data notshown). Another way in which the involvement of L-type chan-nels can be examined is with the use of the dihydropyridineagonist (-)Bay K 8644. Bay K 8644 selectively enhances hecurrent through L-type channels Kokubun and Reuter, 1984;Plummer, 1989; Jonesand Jacobs, 1990) shifting the peak ofthe I-Vcurve to more negative potentials (Fig. 4A,B, Fox et al.,1987; Tsien et al., 1988; Tsien, 1993) and slowing the deacti-vation tail currents (Fig. 4B; Fox et al., 1987;Tsein et al., 1988;Tsein, 1993). The slow component of the deactivation tail inthe presenceof Bay K 8644 is carried exclusively throughL-channels (Plummer et al., 1989; Jones and Jacobs, 1990).Alterations in this slow tail current were usedas a measureofmuscarineseffect on L-type currents. As shown n Figure 4C,the application of carbacholin the presenceof Bay K 8644reducedpeak currents. In addition, carbachol reduced he slowcomponentof the tail current as udgedby the differencecurrentsshown in Figure 40. This modulation readily reversed withwashing.The statistical summary of the carbachol-induced e-


5/12

462 Howe and Surmeier * Muscarinic Modulation of Ca2+ Currents

C

50 B 50-55 -55

+BaiK 6644

0 BAPTA

0 200TIME (ms)

6 -800

l measurement point for D400 260 300TIME (ms) 340

0 BAPTA

control iBayK 8644)

Figure4. Bay K 8644-enhanced tail currents are modulated by carbachol in the absence of intracellular calcium chelators. A, Addition of theL-channel agonist Bay K 8644 (1 FM ) selectively enhances current through L-type channels, shifting the peak of the I-V curve to more negativepotentials and slowing the deactivation tail current. B, An expanded view of the tail currents shown in A. C, In the absence of intracellular calciumchelators, 10 PM carbachol reduced peak currents and the slow tail current. D,An expanded view of the tail currents in control and in the presenceof carbachol. Also plotted is the difference current showing that the carbachol-induced reduction was of the slow component of the tail current.duction in tail amplitudes (measured12 msec nto the tail, seethe arrow shown n Fig. 4B) for a subsetof neurons n = 15) isshown in the inset of Figure 40. In this sample, the medianreduction in tail amplitude wasapproximately 30%. This resultconfirms the inference that L-type currents were reduced bycarbachol.The modulation of L-type current appears o involve apathway distinct from that of the other types of Ca2+current instriatal neuronsIn SCG cells, he modulation of L-type currents s abolishedbyhigh ntracellular concentrationsofthe calcium chelator BAPTA(Beechet al., 1991; Bernheim et al., 1991), whereas he mod-ulation of N-type currents is not. In striatal neurons, the mod-ulation of N- and P-currents appeared o be unaffectedby highconcentrationsof intracellular BAPTA (20 mM), indicating thatthe signalingpathway targeting thesechannels s not dependentupon intracellular calcium (n = 5, data not shown). n contrast,modulation of L-type currents was blocked by dialysis with 20mM BAPTA, as udged from failure of carbachol to reduce BayK 8644-enhanced ail currents (Fig. 5A,B). Shown in the insetof Figure 5B is a summary of the change n the Bay K 8644-

enhanced ail current produced by carbachol n recordingsusinginternal solutionscontaining 20 mM BAPTA and no addedCa2+(n = 10). The free concentration of Ca2+ n the dialysate, ratherthan BAPTA per se, appeared o be the crucial factor in theseexperiments. In most cells (92%, n = 39), raising he estimatedfree Ca2+concentration to 150 nM in the presenceof BAPTA(20 mM) restored the ability of muscarinic agonists o reduceBay K 8644-enhanced ail currents. An example of the effect ofraising intracellular Ca*+ is shown in Figure 5,C and D. Boththe peakand slowcomponent of the tail current were modulatedby carbachol in this cell. A statistical comparisonof the reduc-tion in the slow tail measuredwith a low BAPTA internal (O-0.1 mM BAPTA) to that seenusing an internal solution con-taining 20 mM BAPTA with 10 mM Ca2+added s shown n theinset. Although the reductions in tail amplitude in these wosituations were both significantly greater than those producedin the presenceof 20 mM BAPTA without added Ca2+ lowBAPTA: p < 0.00 1, Kruskal-Wallis ANOVA; high BAPTA: p< 0.00 1, Kruskal-Wallis ANOVA), they were not significantlydifferent from each other (p > 0.05, Kruskal-Wallis ANOVA).A second eature distinguishing the muscarinic pathways nSCG neurons s the PTX sensitivity of the G-protein involved


6/12


A 50 B 50-55 -55-80I I

20 mM BAPTA + 0 Ca2+-800 , 1 -800 L0 200 400 260

TIME (ms)360 340

TIME (ms)

-1200 20mMBAPTA+lOmMC&0 260TIME (ms)

-1600400 260

BAPTA LowlOO- +Ca*+ B APTA

300 340TIME (ms)

Figure 5. Chelation of intracellular Cal+ eliminates +&e reduction of the Bay K 8644-enhanced tail current but restoration of intracellular freeCa2+ to near 150 nM reenables the modulation. A, The application of carbachol (10 PM) following dialysis of the cell with a solution containing 20rnM BAPTA did not abolish the reduction in peak current but did abolish the reduction in the slow tail current. B, An expanded view o f the tailcurrent showing that the slow component was unaffected by carbachol. The difference current shows that only a rapidly deactivating current wasreduced. Shown in the inset is a summary of the change in the Bay K 8644-enhanced tail current produced by carbachol under these conditions.C, When carbachol was applied to cells containing the same concentration of BAPTA but with 10 mM Caz+ (to bring the free Ca2+ concentrationto near 150 nM), the modulation of the peak current was enhanced and the reduction of the slow tail current was restored. D, An expanded viewof the tail currents in control solution and after application of carbachol; also shown are the difference currents. Shown in the inset is a statisticalcomparison of the reduction in the slow tail measured with a low BAPTA internal to that seen using an internal solution containing 20 mM BAPTAwith 10 mM added Ca*+.

(Beech et al., 1992). In these cells, the rapid modulation ofN-type current is sensitive to PTX; the slow modulation of N-andL-type currents is not. The PTX-sensitivity ofthe G-proteininvolved in muscarinic modulation of calcium currents in stri-atal neuronswasdetermined by dialyzing cellswith the catalyticdomain of PTX, the PTX a-protomer (Haydon et al., 1991;Elmslie, 1992). As shown n the plot of peak currents in Figure6, carbachol reduced whole-cell currents. To determine car-bacholseffect on N- and P-type currents, carbacholwasappliedin the presence f the L-type current antagonist nifedipine. Asshown at the first downward arrow, carbachol was nitially ca-pable of modulating thesecurrents. Washing out nifedipine re-stored the peak currents to their previous values. After severalminutes of PTX, a protomer dialysis, nifedipine was reappliedand carbacholseffectsmeasured.Now, carbacholhad no effecton the whole-cell currents (secondarrow), suggesting hat thesignalingpathway affecting N- and P-type currents wasblocked.

After washingout the nifedipine, carbachol was reapplied, re-vealing that the modulation of the L-type currents was stillintact, and little changed rom its initial value (as udged by thedifference in the first and secondmodulations). Similar resultswere observed n four cells.The loss of carbachols effect on N- and P-type currents intheseexperiments wasnot likely to have beena consequence frundown, as he modulation of thesecurrents wasnormally veryrobust. In a sampleof 10 cellswhere the modulation of N- andP-type currents was solated, he medianmodulation 10-l 5 minafter seal upture wasclose o 90% of that measured min afterinitiation of whole-cell recording (seebox plot inset of Fig. 6).Desensitization alsowasnot a factor, as epeatedapplication ofcarbachol in control cells led to very little attenuation in theresponsemagnitude (see nset Fig. 6). In six cells tested withrepeatedapplications (like those of the inset), the median mod-ulation after 10min was89%ofthe initial value. Taken together,


7/12

464 Howe and Surme ier * Muscarinic Modulation of Ca*+ Currents

725

675

625

575s2 520m

b475

$ 425

325

275

4:o 610 6'0 Id.0 li.0TIME (minutes)

Figure 6. The modulation of N- and P-type currents s blockedbydialysiswith the a-protomer f PTX. Peak ampcurrentsare plottedasa functionof time ollowingseal upture.The nternalsolutioncon-tained the PTX a-protomer (10 @ml). Carbachol(10 PM ) was appliedperiodically (shown by the bars at the top and the shading of the symbols).Initially, the L-type currentantagonist ifedipine id not occlude hecarbachol reduction. However, a fter a few minutes ofdialysis, nifedipinewas able to completelyocclude he modulation,suggestinghat themuscariniceceptors ereno longerable o affectN- and P-typecur-rents. The time course was corrected for a 20% rundown in the currentsby subtraction f an exponentialit to the currents uringcontrolpe-riods. The box plot is a statistica l summary of the percentage changein the magnitude f the modulation ensitiveo carbachol t 5 and 10min into the recording showing that the observed eff ect is not due torundownof the modulation.Nor is it a resultof desensitization,srepeated pplicationsf carbacholed to very little attenuation f theresponsesshownn the nset.these esultssuggesthat a PTX-sensitive G-protein is nvolvedin the modulation of N- and P-type currents but not of L-typecurrents.A third featuresdistinguishingmuscarinic pathways in SCGcells s the voltage dependence f the modulation (Beechet al.,1992).The voltage dependence f the muscarinic effects n stri-atal neuronswas studied by comparing the modulation of cur-rents evoked by a step to -20 mV with and without strong(+ 100 mV) depolarizing conditioning steps 30 msec). In allneurons, strong depolarizing prepulsesenhanced the kineticsand amplitude ofcurrents evoked by the test step Fig. 7A, inset),asdescribed or a number of other neurons Bean, 1989;Bolandand Bean, 1993). When intracellular Ca2+was not extrinsicallybuffered, the current reductions produced by carbachol wereattenuated in most cells (315)by depolarizing prepulses.How-ever, the magnitude of the attenuation, if present, was highlyvariable (1 -60%). One possibleexplanation for the variationis that when Caz+ s unbuffered t may be ow or high, effectivelyenabling or disenabling the Ca*+-dependent pathway that isthought to be voltage independent seebelow). To test this pos-sibility, intracellular free Ca2+was buffered into the low nano-molar range by dialyzing with a solution containing 20 mMBAPTA and no Ca2+. As shown above, this maneuver limitsthe muscarinic modulation to N- and P-type currents. Underthesecircumstances,he reductions producedby carbacholwere

more consistently attenuated by depolarizing prepulses Fig.7A,B) (median modulation after the prepulsewas 63% of con-trol, see nset Fig. 7B).This result suggestshat some of the variation in voltagedependenceseenwithout Ca2+ buffers could have been a con-sequence f a voltage-independent modulation of L-type cur-rents (Beech et al., 1992; Mathie et al., 1992).To test whetherthe modulation of L-type currentswasvoltagedependent, -)BayK 8644enhanced tail currents were examined n the samepro-tocol with free Ca*+ buffered to near 150 nM. Under thesecon-ditions (seeabove), the slow tail is reduced by carbachol (Fig.7C). Strong depolarizing prepulsesdid not attenuate the car-bachol-induced reduction in theseslow tail currents (Fig. 70).In fact, the median modulation was arger following a strongconditioning depolarization (median modulation = 119%, seeinset Fig. 70). Although the samplesweresmall, statisticalanal-ysis of the change n modulation in these wo conditions sug-gested he differenceswere significant p < 0.05, Kruskal-WallisANOVA).A final feature distinguishing he muscarinicpathways n SCGcells s onsetkinetics (Bemheim et al., 1991). One pathway thattargets N-type channels s fast (T < 1 set), whereas he secondpathway that targets N- and L-type currents is slower (T - 3-35 set, depending on agonist concentration). To determinewhether fast and slow componentscould be resolved n striatalneurons, est steps o 0 mV were delivered oncea secondduringdrug application. Rapidly moving the perfusion barrelseffectedan agonistconcentration jump that was complete n less han asecond see nset, Fig. 8A). At saturating concentrations of ag-onist (10 PM carbachol), plots of peak current as a function oftime were well fit with a singleexponential in all casesn = 10).An example s shown n Figure 8A; in this neuron, the onsetofthe modulation developed as a singleexponential with a timeconstant of 3.86 set; other cellshad similar kinetics (median =3.01 set, n = 10; see nset, Fig. 8B). The kinetics of the reductionin the slow, Bay K 8644-enhanced ail current subsequent ocarbachol application were similar (2.5 < T < 6.5 set, n = 13,data not shown). To determine whether a fast component waspresentbut merely maskedby large L-type current modulation,intracellular Ca2+waschelated to low levels with 20 mM BAP-TA. Under these conditions, the onset of the modulation wasnot significantly faster.An example s shown n Figure8B. Here,the change n current amplitude waswell fit, with an exponentialhaving a single time constant of 3.25 set; other neurons weresimilar (median = 2.8 set, n = 9, see nset Figure 8B). Thedifferences n the time constants n the two buffering conditionswere not significant (p > 0.05, Kruskal-Wallis ANOVA).DiscussionMuscarinic receptors modulate HVA Ca currentsIn rat striatal neurons, activation of muscarinic receptors rig-gers wo parallel signalingcascadeshat result n the modulationof N-, P-, and L-type HVA Ca+ currents. The muscarinic re-ductions in Ca*+ currents were dosedependentand reversible.This finding is consistentwith previous reports in striatal neu-rons (Misgeld et al., 1986)and in other cell types (Adams et al.,1982; Brown and Selyanko, 1985; Heschleret al., 1986, 1987;Gahwiler and Brown, 1987; Wanke et al., 1987; Brown et al.,1989; Toselli and Lux, 1989;Fisher and Johnston, 1990;Beechet al., 1991; Bemheim et al., 1992; Mathie et al., 1992)showingmuscarinic reductions n Ca+ currents. As in rat SCG neurons(Bemheim et al., 1992), it is highly likely that these striatal


8/12


A -20-80400,.20 mM BAP TA + 0 Ca2+ +carbachola o-s5$ -4oo-

-800 -

-1200n 30 60 0TIMI! jmsec)C

400 -0

z-4!sii!L$ -10000

-2000

nu-10

-80 400+carbacholyxf OYontrol $5 -1000I 0'0

5 mM BAPTA+ 2.8 mM Can+[Caz+j - 150 nM30 I -2000,60 0

I +carbachol

I30TIME (msec) 60+lOO

-101 -80

II

tBay K 6644r Tagll

$a25- Cd Ca2+

+carbachol

-, 60TIME (msec) TIME (msec)Figure 7. The modulation of N- and P-type currents was attenuated by depolarizing prepulses, but the reduction of L-type currents was not. A,Whole-cell calcium currents were elicited by a step to -20 mV from a holding potential of -80 mV in a cell dialyzed with an internal solutioncontaining 20 rnM BAPTA, 0 Ca+. The application of carbachol (10 PM) reduced the currents evoked by the step. Conditioning with a step to+ 100 mV (30 msec) enhanced the currents evoked by the step to -20 mV, as shown in the inset. B, Conditioning steps to + 100 mV (30 msec)attenuated the modulation. The bar shows the amplitude of the modulation without the conditioning step in A. The inset is a statistical summaryshowing the percentage control modulation remaining after the conditioning step. C, When the internal Ca*+ was elevated to near 150 nM (5 mMBAPTA, 2.8 mM Ca*+ ), carbachol(5 PM ) was able to modulate the slow Bay K 8644 tail current (as shown above). D,Conditioning steps to + 100mV in these recording conditions did not attenuate the reduction in the slow tail currents produced by carbachol. The inset is a statistical summaryshowing the percentage control modulation remaining after the conditioning step under conditions of high and low intracellular Ca2+.pathways depend upon ml and m4 class eceptors in striatalneurons. The principal neuron of the striatum-the medium-spiny projection neuron-is known to colocalize ml and m4receptors Andersonand McKinney, 1988;Bernard et al., 1992).As theseneuronsconstitute more than 95% of all striatal neu-rons, the vas t majority of our recordings were undoubtedly fromthis cell type.N- and P-type currents are modulated through a PTX-sensitiveG-proteinIn spite of our inability to definitively isolatem 1 and m4 path-ways at the receptor level, two signalingpathways with featuresknown to be associatedwith each receptor were identified bymanipulation of G-proteins and intracellular Ca*+. In a varietyof cell types, m4 receptors couple with PTX-sensitive G,/ G,classproteins, whereasm 1 receptors do not (Nathenson, 1987;Peralta et al., 1988;Hulme et al., 1990; Bemheim, et al., 1992).Dialysis with the a-protomer of PTX eliminated the muscarinicmodulation of N- and P-type currents but did not noticeablyaffect modulation of L-type currents. This result is consistent

with studies n peripheral neurons showing that activation ofthe PTX-sensitive pathway reducesN-type currents (Marchettiet al., 1986; Bemheim et al., 1991; Beechet al., 1992).Recentwork with central neurons Mintz and Bean, 1993)has evealedreductions in P-currents subsequent o activation of receptors(GABA,) that have been shown to couple to PTX-sensitiveG-proteins (Alford and Grillner, 1991; Knott et al., 1993). Inthese reports, the modulation of N- and P-currents remainedwhen intracellular CaZ+ was buffered to nominally low nano-molar levels and was elieved by strongly depolarizing prepulses(i.e., voltage dependent).The modulation of N- and P-type cur-rents by muscarinic agonists n striatal neuronspossessedothof these eatures, ndicating that the striatal PTX-sensitive sig-naling pathway is similar to that found in other central andperipheral neurons.Only the onset kinetics of the modulation of N- and P-typecurrents differed from previous descriptions. In SCG neurons,the rapid (T < 1set), PTX-sensitive reduction in N-type currentsis associatedwith a membranedelimited, G-protein-mediatedmodulation. Rather than having a time constant of less han a


9/12

466 Howe and Surme ier - Muscarinic Modulation of Cap+ Currents

250

50 10 20 30 40 50 60 70TIME (set)

time

70 I I I I I I 1IO 20 30 40 50 60 70 ETIME (set)

Figure 8. The kinet ics o f the muscarinic reduction were similar with and without Ca2+ chelators. A, Calcium current was elicited by a 10 msecstep to 0 mV f rom a holding potential of -80 mV. The test step was repeated once every second. The onset of the modulation was well f it with asingle exponential curve (I = Z,,exp( -(t - Q/T) + constant, where Z,. was the reducible part o f the current, to was the time o f drug application,and the constant was the nonreducible part o f the current). 7 in this cell was 3.86 sec. The inset is the single exponential fi t of Cd*+ blockade ofthe current having a T of 400 msec, indicating the speed with which complete solution changes are achieved with our perfusion system. E, Thesame experiment was performed in the presence of a high concentration of intracellular calcium chelator (20 mM BAPTA). A single exponentialcurve having a 7 of 3.25 set was fi t to the onset of the muscarinic suppression. The inset is a box plot showing the distribution of time constantsin neurons dialyzed without buffer and those dialyzed with 20 mM BAPTA. The distributions were not statistically different (p > 0.05, Kolmoogorov-Smimov, df = 8; p > 0.05, Kruskal-Wallis analysis of variance).second, the onset of the striatal modulation of N- and P-type G-protein effects on Ca*+ channels (Jones, 1991; Boland andcurrents at carbachol concentrations 3-4 times the IC, , usually Bean, 1993). Hence, kinet ics alone do not appear to be a reliablehad time constants in the 2-3 set range (see Fig. 8B). It is means of distinguishing direct and indirect signaling pathways.diff icul t, however, to assign much significance to this difference. One reason for this experimental variabil ity may be that in smallIn other cell types, onset kinetics very similar to ours have been cells (e.g., striatal neurons) patching (or dissociation) disruptsreported for modulations thought to be mediated by direct the cytoskeleton in such a way that the spatial constraints upon


10/12


signaling elements (receptors, G-proteins, and targets) are lost.Constraints upon these signaling elements appear to be crucialto the rapid kinetics of membrane del imi ted pathways (Brown,1993). The loss of these constraints lead to kinetic predictionsvery similar to kinetics we observed (Brown, 1993).L-Type currents are modulated through a BAPTA-sensitivepathway involving a PTX-insensitive G-proteinIn the absence of extrinsic Ca *+ chelator, muscarinic agonistsreduced L-type currents in striatal neurons. The involvementof L-type currents was deduced not only from the occlusiveeffects of the dihydropyridine antagonist n ifedip ine but alsofrom the abi lity of muscarinic agonists to reduce the slow com-ponent of Bay K 8644-enhanced tail currents (Plummer et al.,1989; Jones and Jacobs, 1990). Although init ially L-type cur-rents in neurons were not thought to be neuromodulatory tar-gets, recent work has revealed several types of receptor nega-tively coupled to this class of channel (Bley and Tsien, 1990;Beech et al., 199 1, 1992; Bemheim et al., 199 1, 1992; Mathieet al., 1992; Sayer et al., 1993).

The insensitivity of the L-type current modulat ion to dialysiswith the PTX catalytic subunit (a-protomer) is consistent withthe hypothesis that this modula tion is mediated by ml classreceptors (Wanke et al., 1987). A similar linkage has been es-tablished in rat SCG neurons, where m l receptors modulateL-type currents through a PTX-insensitive G-protein and a sol-uble second messenger (Bemheim et al., 1992). In these neurons,as in striatal neurons, the modulat ion is relatively slow (T > 2set), is not attenuated by depolarizing condition ing steps, butis eliminated by high concentrations of Ca*+ chelator (Beech etal., 1992). In spite of the BAPTA sensitivity, the SCG signalingpathway does not appear to involve Ca2+ other than in a per-missive way. There are reasons to believe that fluctuations inintracellular Ca*+ may play a more direct role in the modulat ionof L-type currents in striatal neurons. In these neurons, m lreceptors are coupled to phospholipase C and the productionof diacylg lycerol (DAG) and inosi tol triphosphate (IP,) (Peraltaet al., 1988). In a variety of cell types, IP, triggers the releaseof Ca*+ from intracellular stores. Preliminary photometry ex-periments (Surmeier, Harrington, and Hille, unpublished ob-servations) have confirmed this inference by showing that mus-carinic agonists are capable of inducing transient elevations incytosolic Ca2+ levels in striatal neurons, in contrast to SCGneurons (Bemheim et al., 199 1). Furthermore, caffeine, a potentliberator of Ca2+ from ryanodine-sensitive Ca2+ stores (Endo,1977; Leitjen and Van Breeman, 1984) also mimics the abilityof muscarinic agonists to reduce L-type currents in striatal neu-rons (Howe and Surmeier, 1993). Given these facts, it is plau-sible to assume that Ca2+ per se is causally linked to the reduc-tion in striatal L-type currents by muscarinic agonists. Aschematic depicting our working hypotheses about this signalingpathway and the pathway target ing N- and P-type channels isshown in Figure 9.What is less clear is why elevating free Ca*+ in the presenceof 20 mM BAPTA also enables modulat ion of L-current. In thiscircumstance, Ca*+ should still be well buffered. There are twopossible explanations. One possibility is that elevation of cy-tosolic Ca*+ above 100 nM by muscarinic agonists is necessarybut not sufficient to bring about the modulat ion of L-currents.Although a number of cellular signaling molecules are Ca2+dependent, the one directly linked to m 1 receptors (protein ki-nase C) has been reported to enhance L-type currents (Yang and

Muscarinic receptorsml. m4

PTX-sensitive PTX-insensitiveG-protein G-protein

Ca2+-independentpathway Ca2+-dependentpathway

N, P-type Ca2+ channels L-type Ca2+ channelsFigure 9. A schematic summarizing the elements involved in the mus-carinic modulation of HVA currents in neostriatal neurons. Muscarinicreceptorsoupleo PTX-sensitivend-insensitive-proteins. hePTX-sensitive pathway reduces N- and P-type currents through a mechanismthat is not affected by chelating Ca2+ to low nanomolar levels. The PTX-insensitive pathway reduces L-type currents through signaling elementsthat are blocked by Ca2+ chelation.Tsien, 1993). Another possibility is that dialyzed BAPTA can-not penetrate all cellular compartments equally well and, as aconsequence, oesnot buffer Ca2+ fficiently throughout the cell.Dialysis with high millimolar concentrationsof BAPTA in theabsence f added Ca2+may block the modulation by depletingreleasableCaZ+ stores-a consequence ot brought about bydialysiswith a CaZ+supplemented olution. We are esting hesepossibilities by combining perfused pipette whole-cell patchclamp measurementswith coincident photometric measure-ments of free Ca*+In spite of these caveats, the hypothesis that Ca2+per se sthe muscarinicsignal s consistentwith the well-describedCa2+-dependent inactivation of L-type currents (Eckert and Chad,1984; Kalman et al., 1988; Armstrong, 1989). In a variety ofcell types, elevation of intracellular Caz+ eads o a reduction inL-type current. There are a number of enzymes hat are depen-dent upon calmodulin and intracellular Ca2+ hat could be in-volved in this process Stryer, 1988; Hille, 1992). One likelycandidate is the calcium/calmodulin-dependent phosphatase,calcineurin (Armstrong, 1989). In musclecells, calcineurin de-phosphorylates protein kinase A (PKA) sites on dihydropyri-dine-sensitive, L-type channelsand, in so doing, reduces rans-membrane urrent. If a similarmechanisms mportant in striatalneurons,PKA activation should enhanceL-type currents. Thisappears to be the case n a subset of medium-sized striatalneurons Surmeier et al., unpublishedobservations).Assumingthat there is constitutive activity of PKA (or a kinasewith sim-ilar substratespecificity), dephosphorylation of L-type channelsby calcineurin could produce the pattern of effects we haveobserved. This interaction-between Dl dopamine receptor-mediatedactivation of PKA and muscarinic eceptor activationof calcineurin - could be a cellular counterpart of the ong-stand-ing clinical observation that dopaminergic and cholinergic ag-


11/12

466 Howe and Surme ier - Muscarinic Modulation of Ca2+ Currents

onists oppose one another in the control of basal gangl ia-de-pendent movements (Barbeau, 1962).Functional implicationsBy reducing the entry of Ca >+ through voltage-dependent chan-nels, acetylcholine will influence a variety of cellular functions.The integrative properties of striatal neurons, for example, shouldbe altered in at least two ways. First, by minimizing the con-tribution of Ca*+ -dependent regenerative events in the dendrites(Lev-Ram et al., 1992; Amitai et al., 1993) acetylcholine shouldreduce the conduction of postsynaptic charge to the somaiini tialsegment. In agreement with this inference, activation of striatalmuscarinic receptors has been shown to reduce the amplitudeof synaptic potentials evoked in response to cortical stimulation(Calabresi et al., 1991). Second, the reduction of HVA Ca2+entry during spiking should attenuate Ca*+-dependent potas-sium currents that slow discharge in a variety of neurons (Lan-caster et al., 1990; McCobb and Beam, 199 l), including striatalneurons (Pineda et al., 1992). By acting on a portion of thedendritic tree and the soma ofa neuron, cholinergic interneuronscould functional ly shrink targeted dendrites and enhance thesomatic response to sustained excitatory inputs arising else-where.Muscarinic receptors may also affect presynaptic transmitterrelease and junctional plasticity through Ca*+ channels. Ace-&choline is known to reduce depolarization-induced GABArelease (Hashimoto et al., 1986; Augustine et al., 1987; Marchiet al., 1990; Raiteri et al., 1990). Muscarinic reductions in N-and P-type currents, which control transmitter release in thestriatum (Turner et al., 1993), provides a cellular mechanismfor this inhibition. Alterations in Ca2+ entry into the soma anddendrites can also be expected to modulate second-messengerpathways that are Ca*+/ca lmodul in dependent. Changes of thissort are of importance in synaptic plasticity (see Kotsyuk, 1992,for review) and structural changes contingent upon immediateearly gene induction (Kley et al., 1987; Murphy et al., 199 1).

ReferencesAdams PR, Brown DA, Constanti A (1982) Pharmacological inhi-bition of the M-current. J Physiol (Lond) 332:223-262.Alford S, Grillner S (199 1) The involvement of GABA, receptors andcoupled G-proteins in spinal GABAergic presynaptic inhibition. JNeurosci 11:3718-3726.-Amitai Y, Friedman A, Connors BW, Gutnick MJ (1993) Regener-ative activit y in apical dendrites of pyramidal cells in neocortex. CerebCortex 3:26-38.Anderson DJ . McKinnev M (1988) Muscarinic M2 receptor mediatedCAMP reduction in mechanically dissociated rat cortex. Brain Res

475~28-34.Armstrong DL (1989) Calcium channel regulation by calcineurin, acalcium activated phosphatase in mammalian brain. Trends Neurosci12:117-122.Augustine GJ, Charlton MP, Smith ST (1987) Calcium action in syn -aptic transmitter release. Annu Rev Neurosci 10:633-693.Barbeau A (1962) The pathogenesis o f Parkinsons disease: a newhypothesis . Can Med Assoc J 87:802-807.Bargas J, Howe A, Eberwine J, Cao Y, Surmeier DJ (1994) Cellularand molecular characterization of Ca*+ currents in acutely-isolated,adult rat neostriatal neurons. J Neurosci, in press.Bean BP (1989) Neurotransmitter inhibition of neuronal calcium cur-rents by changes in channel voltage dependence. Nature 340: 153-156.Beech DJ, Bemheim L, Mathie A, Hille B (199 1) Intracellular Cabuffers disrupt muscarinic suppression of Ca current and M currentin rat sympathetic neurons. Proc Nat1 Acad Sci USA 88:652-656.Beech DJ, Bemheim L, Hille B (1992) Pertussis toxin and voltage

dependence distinguish multiple pathways modulating calcium chan-nels of rat sympathetic neurons. Neuron 8:97-106.Bernard V, Normand E, Bloch B (1992) Phenotypical characterizationof the rat striatal neurons expressing muscarinic receptor genes. JNeurosci 12:3591-3600.Bemheim L, Beech DJ, Hille B (199 1) A diffusible second messengermediates one of the pathways coupling receptors to calcium channelsin rat sympathetic neurons.~Neurbn 6859-867.Bemheim L. Mathie A. Hille B (1992) Characterization of muscarinic,receptor subtypes inhibiting Ca current and M current in rat sym-pathetic neurons. Proc Nat1 Acad Sci USA 89:9544-9548.Bley KR, Tsien RW (1990) Inhibition of calcium and potassium chan-nels in sympathetic neurons by neuropeptides and other ganglionictransmitters. Neuron 3:379-39 1.Bolam JP, Wainer BH, Smith AD (1984) Characterization of cholin-ergic neurons in the rat neostriatum. A combination of choine ace-tyltransferase immunocytochemistry, Golgi-impregnation and elec-tron microscopy. Neuroscience 12:71 l-7 18. -Boland LM, Bean BP (1993) Modulation of N-type calcium channelsin bullfrog sympathetic neurons by luteinizing hormone-releasing hor-mone: kinetics and voltage dependence. J Neurosci 13:5 16-533.Bonner TI (1989) The molecular basis of muscarinic receptor diver-sit y. Trends Neurosci 12: 148-15 1.Bonner TI, Buckley NJ, Young AC, Brann MR (1987) Identificationof a fam ily of muscarinic acetylcholine receptor genes. Science 237:527-531.Bonner TI, Young AC, Brann MR, Buckley NJ (1988) Cloning andexpression ofthe human and rat m5 muscarinic acetylcholine receptorgenes. Neuron 1:4031110.Breitwieser GE, Szabo G (1988) Mechanism of muscarinic receptor-induced potassium channel activation as revealed by hydrolysis-re-sistant GTP analogues. J Gen Phys iol 9 1:469-493.Brown AM (1993) Membrane-delimited cell signaling complexes: di-rect ion channel regulation by G proteins. J Membr Biol 13 1:93-104.Brown DA, Selyanko AA (1985) Two components of muscarine-sen-sitive membrane current in rat sympathetic neurones. J Physiol (Lond)358:335-363.Brown DA, Marrion NV, Smart TG (1989) On the transduction mech-anism for muscarine-induced inhibition of M-current in cultured ratsympathetic neurones. J Physiol (Lond) 413:469488.Calabresi P. Mercuri NB. Bemardi G (199 1) Muscarinic modulationof synaptic and intrinsic membrane properties of striatal neurons. EurJ Neurosci S4: 116.Calne DB (1980) Therapeutics in neurology. London: Blackwells.DiFiglia M (1987) Synaptic organization of cholinergic neurons in themonkey neostriatum. J Corn; Neural 225:245-258.Eckert R. Chad JE (1984) Inactivation of Ca channels. Proc bioohvsMol Biol 44:2 15-267. . ,Elmslie KS (1992) Calcium current modulation in frog sympatheticneurones: multiple neurotransmitters and G proteins. J Physiol (Lond)45 1~229-246.Endo M (1977) Calcium release from the sarcoplasmic reticulum.Physiol Rev 57:71-108.Fisher R, Johnston D (1990) Differential modulation of single voltage-gated calcium channels by cholinergic and adrenergic agonists in adulthippocampal neurons. J Neurophysio164: 129 l-l 302.Fox AR, Nowycky MC, Tsien RW (1987) Kinetic and pharmacolog-ical properties distinguishing three types of calcium currents in chicksensory neurons. J Physiol (Lond) 394: 149-172.Gahwiler BH, Brown DA (1987) Muscarine effectscalcium current inhippocampal pyramidal cells in vitro.Neurosci Lett 78:301-306.Hamill OP, Marty A, Neher E, Sakman B, Sigworth FC (1981) Im-proved path-clamp techniques for high-resolution current recordingfrom cells and cell-free membrane patches. Pfluegers Arch 391:85-100.Hashimoto S, Tanaka C, Taniyama K (1986) Presynaptic muscarinicand a-adrenoceptor-mediated regulation o f GABA release frommyenteric neurons of the guinea-pig small intestine. Br J Pharmacol89:787-792.Hayden PG, Man-Son-Hing H, Doyle RT, Zoran M (199 1) FMRFam-ide modulation of secretory machinery underlying presynaptic inhi-bition of synapt ic transmission requires a pertussis toxin-sensitiveG-protein. J Neurosci 11:3851-3860.Heschler J, Kameyama M, Trautwein W (1986) On the mechanismof muscarinic inhibition of the cardiac Ca current. Pfluegers Arch407:182-189.


12/12


Heschler J, Rosenthal W, Trautwein W, Schultz G (1987) The GTPbinding protein, Go, regulates neuronal calcium channels. Nature 325:445-447.Hille B (1992) Ionic channels of excitable membranes. Sunderland,MA: Sinauer.Hoehn K, Watson TWJ, MacVicar BA (1993) Multiple types of cal-cium channels in acutely isolated rat neostriatal neurons. J Neurosci13:1244-1257.Hokin MR, Hokin LE (1954) Effects of acetylcholine on phospholipidsin the pancreas. J Biol Chem 209:549-558.Howe AR, Surmeier DJ (1993) Calcium-dependent modulation ofL-type calcium currents in acutely isolated rat neostriatal neurons.Sot Neurosci Abstr 19:978.Hulme EC, Birdsall NJM, Buckley NJ (1990) Muscarinic receptorsubtypes. Annu Rev Pharmacol Toxic01 30633-673.Ikeda SR, Schofield GG (1989) Somatostatin blocks a calcium currentin rat sympathetic ganglion neurons. J Physiol (Lond) 409:221-240.Jones SW (199 1) Time course of receptor-channel coupling in frogsympathetic neurons. Biophys J 60:502-507.Jones SW, Jacobs LS (1990) Dihydropyridine actions on calcium cur-rents of frog sympathetic neurons. J Neurosci 10:2261-2267.Kalman D, OLague PH, Erxleben C, Armstrong DL (1988) Calcium-dependent inactivation of the dihydropyridine-sensitive calciumchannels in GH3 cells. J Gen Physiol 92:531-547.Kley N, Loeffler JP, Pittius CW, Holt V (1987) Involvement of ionchannels in the induction of proenkephalin A gene expression bynicotine and CAMP in bovine chromaffin cells. J Biol Chem 262:40834089.Knott C, Maguire JJ, Moratalla R, Bowery NG (1993) Regional effectsof nertussis toxin in vivo and in vitro on GABA,, receptor binding inrat brain. Neuroscience 52:73-8 1.Kokubun S, Reuter H (1984) Dihydropyridine derivat ives prolong theonen state of Ca channels in cultured cardiac cells. Proc Nat1 AcadSci USA 8 1:48284827.Kostyuk PG (1992) Calcium ions in nerve cell function. London:Oxford UP.Kubo T, Fukada K, Mikami A, Maeda A, Takahashi H, Mishina M,Haga T, Haga K, Ichiyama A, Kangawa K, Kojima M, Matsuo H,Hirose T, Numa S (1986) Cloning, sequencing and expression ofcomnlementarv DNA encoding the muscarinic acetylcholine receptor.Nature 323:411416. -Lancaster B, Nicoll RA, Perkel DJ (1990) Calcium activated two typesof potassium channels in rat hippocampal neurons in culture. J Neu-rosci 10:23-30.Leitjen PAA, Van Breeman C (1984) The effec ts of caffe ine on thenoradrenaline-sensitive calcium store in rabbit aorta. J Physiol (Land)3571327-339.Lev-Ram V, Miyakawa H, Lasser-Ross N, Ross WN (1992) Calciumtransients in cerebellar Purkinie neurons evoked by intracellular stim-ulation. J Neurophysiol 68: 1 67-1177.Liao CF, Axe1 PN, Themmen RJ, Claude B, Bimbaumer M, BimbaumerL (1989) Molecular cloning and expression of a fi fth muscarinicacetylcholine receptor. J Biol Chem 264:7328-7337.Marchetti C, Carbone E, Lux HD (1986) Effect s of dopamine andnoradrenaline on Ca channels of cultured sensory and sympatheticneurons of chick. Pfluegers Arch 406: 104-l 11.Marchi M, Sanguineti P, Raiteri M (1990) Muscarinic receptors me-diate direct inhibition of GABA release from rat striatal nerve ter-minals. Neurosci Lett 116:347-35 1.Mathie A, Bemheim L, Hille B (1992) Inhibition of N- and L-typecalcium channels by muscarinic receptor activation in rat sympatheticneurons. Neuron 8:907-9 14.McCobb DR, Beam KG (199 1) Action potential waveform voltage-clamp commands reveal striking differences in calcium entry via lowand high voltage-activated calcium channels. Neuron 7:119-127.Mintz IM, Bean BP (1993) GABA, receptor inhibition of P-type cal-cium channels in central neurons. Neuron 10:889-898.Mintz IM, Adams ME, Bean BP (1992) P-type calcium channels inrat central and peripheral neurons. Neuron 9:85-95.Misgeld U, Calabrisi P, Dodt HU (1986) Muscarinic modulation of

calcium dependent plateau potentials in rat striatal neurons. PfluegersArch 407:482-487.Murphy TH, Worley PF, Baraban JM (199 1) L-type voltage-sensitivecalcium channels mediate synapt ic activation of immediate early genes.Neuron 7~625-635.Nathenson NM (1987) Molecular properties of the muscarinic ace-tylcholine receptor. Annu Rev Neurosci 10: 195-236.Peralta EG, Ashkenazi A, Winslow JW, Ramachandran J, Capon DJ(1988) Differential regulation of PI hydrolysis and adenyl cyclase bymuscarinic receptor subtypes. Nature 334:434-437.Phelps PE, Houser CR, Vaughn JE (1985) Immunocytochemical lo-calization of choline acetyltransferase within the rat neostriatum: acorrelated light and electron microscopic study of cholinergic neuronsand synapses. J Comp Neural 238:286-307.Pineda JC, Galarraga E, Bargas J, Cristancho M, Aceves J (I 992) Cha-rybdotoxin and apamin sens itiv ity o f the calcium-dependent repo-larization and the afterhyperpolarization in neostriatal neurons. JNeurophysiol 68:287-294.Plummer MR, Logothetis DE, Hess P (1989) Elementary propertiesand pharmacological sensitivit ies of calcium channels in mammalianperipheral neurons. Neuron 2: 1453-1463.Pusch M, Neher E (1988) Rates of diffusional exchange between smallcells and a measuring pipette. Pfluegers Arch 4 11:204-2 Il.Raiteri M, Marchi M, Paudice P, Pittaluga A (1990) Muscarinic re-ceptors mediating inhibition of g-aminobutyric acid release in ratcorpus striatum and their pharmacological characterization. J Phar-macol Exp Ther 254:496-50 1.Sayer RJ, Schwindt PC, Grill WE (1992) Metabotropic glutamatereceptor-mediated suppression of L-type calcium current in acutelyisolated neocortical neurons. J Neurophysiol 68:833-842.Stryer L (1988) Biochemistry. New York: Freeman.Surmeier DJ, Stefani A, Foehring R, Kitai ST (199 1) Developmentalexpression of a slowly-inactivating voltage-dependent potassium cur-rent in rat striatal neurons. Neurosci Lett 122:4 l-46.Surmeier DJ, Eberwine J, Wilson CJ, Stefani A, Kitai ST (1992) Do-pamine receptor subtypes co-localize in rat striatonigral neurons. ProcNat1 Acad Sci USA 89:10178-10182.Surmeier DJ, Bargas J, Howe AR (1993) Dopamine Dl receptor ac-tivation reduces calcium currents in acutely-dissociated rat neostriatalneurons. Sot Neurosci Abstr 19:978.Toselli M. Lux HD (19890 GTP-bindina nroteins mediate acetv lcho-line inhibition of voltage-dependent Cal&m channels in hippocampalneurons. Pfluegers Arch 4 13:3 19-32 1.Tsien RW (1993) Calcium channels in excitable cell membranes. AnnuRev Physiol45:341-358.Tsien RW, Lipscombe D, Madison DV, Bley KR, Fox AP (1988)Multiple types of neuronal calcium channels and their selective mod-ulation. Trends Neurosci 11:43 1438.Tukey JW (1977) Exploratory data analysis . Menlo Park, CA: Addi-son-Wesley.Turner TJ, Adams ME, Dunlap K (1993) Multiple calcium channelscoexist to regulate synaptosomal neurotransmitter release. Proc Nat1Acad Sci USA 90:95 18-9522.Usowicz MM, Augimori M, Cherksey B, Llinas R (1992) P-type cal-cium channels in the somata and dendrites of adult cerebellar purkinjecells. Neuron 9: 1185-I 199.Vilaro MT, Wiederhold KH, Palacios JM, Mengod G (1991) Mus-carinic cholinergic receptors in the rat caudate-putamen and olfactorytubercle belong predominantly to the m4 class: in situ hybridizationand receptor autoradiography evidence. Neuroscience 40: 159-l 67.Wanke E, Ferroni A, Malgaroli A, Ambrosini A, Pozzan T, MeldolesiJ (1987) Activation of a muscarinic receptor selectively inhibits arapidly inactivated Ca current in rat sympathetic neurons. Proc Nat1Acad Sci USA 84:43 1343 17.Wilson CJ, Chang HT, Kitai ST (1990) Firing patterns and synapticpotentials of identified giant aspiny intemeurons in the rat neostri-atum. J Neurosci 10:508-5 19.Yang J, Tsien RW (1993) Enhancement of N- and L-type calciumchannel currents by protein kinase C in frog sympathetic neurons.Neuron 10:127-136.

angela r. howe and d. james surmeier- muscarinic receptors modulate n-, p-, and l-type ca^2+...

Documents