JournalofPhysiology (1993), 469,pp. 153-178 153 With 11figures Printed inGreatBritain ON THE MECHANISM OF M-CURRENT INHIBITION BY MUSCARINIC ml RECEPTORS IN DNA-TRANSFECTED RODENT NEUROBLASTOMAxGLIOMA CELLS BY J. ROBBINS, S. J. MARSH AND D. A. BROWVN From the Department ofPharmacology, University College London, Gower Street, London WClE 6BT (Received 5 October 1992) SUMMARY 1. Acetylcholine (ACh) produces two membrane current changes when applied to NG108-15 mouse neuroblastomaxrat glioma hybrid cells transformed (by DNA transfection) to express ml muscarinic receptors: it activates a Ca2"-dependent K+ conductance, producing an outward current, andit inhibits a voltage-dependent K+ conductance (the M conductance), thus diminishing the M-type voltage-dependent K+ current (IK(M)) and producing an inward current. The present experiments were undertaken to find outhowfarinhibition ofIK(M) might be secondary to stimulation ofphospholipase C, by recording membrane currents and intracellular Ca2+ changes with indo-1 using whole-cell patch-clamp methods. 2. Bath application of 100 /M ACh reversibly inhibited IK(M) by 47-3+ 32% (n = 23). Followingpressure-application oft mmACh,themeanlatencytoinhibition was 420 ms at 35 °C and 1-79 s at 23 'C. Latencies to inhibition by Ba2+ ions were 148 ms at 35 'C and 92 ms at 23 'C. 3. TheinvolvementofaG-proteinwastestedbyadding05 mMGTP-y-Sor 10 mM potassium fluoride to the pipette solution. These slowly reduced IK(M), with half- times ofabout 30 and 20 min respectively, and rendered the effect ofsuperimposed ACh irreversible. Effects ofACh were not significantly changed after pretreatment for 24 h with 500ng ml-' pertussis toxin or on adding up to 10 mm GDP-,8-S to the pipette solution. 4. The role of phospholipase C and its products was tested using neomycin (to inhibit phospholipase C), inositol 1,4,5-trisphosphate (InsP3) and inositol 1,3,4,5- tetrakisphosphate (InsP4),heparin, andphorboldibutyrate (PDBu)andstaurosporin (to activate and inhibit protein kinase C respectively). Both neomycin (1 mM external) andInsP3 (100,UMintrapipette) inhibitedtheACh-inducedoutwardcurrent and/orintracellularCa2+transientbutdidnotblockACh-inducedinhibitionofIK(M). Intrapipette heparin (1 mM) blocked activation ofIK(Ca) and reduced Ach-induced inhibitions ofIK(M), but also reduced inhibition ofICa via endogenous m4 receptors. PDBu (with or without intrapipette ATP) and staurosporin had no significant effects. 5. ACh induced a transient rise in intracellular [Ca2+] but this did not appear to be responsible forinhibition ofIK(M) since (a) the latterprecededthe rise in [Ca2+] by MS1820 154 J. ROBBINS, S. J. MARSH AND D. A. BROWN 3-6 s, and (b) ACh still inhibited IK(M) when the rise in [Ca2+] was suppressed by (i) repetitive ACh applications, (ii) addition of100/M InsP3 tothepipette solution, and (iii) buffering with 20 mm BAPTA. All three procedures inhibited the ACh-induced outward current. 6. Inhibitors of phospholipase A2, lipoxygenase, cyclo-oxygenase or nitric oxide synthase had no significant effect on ACh-induced inhibition ofIK(M). 7. The presence ofGTP (2 mm), ATP (2 mM), dibutyryl-cAMP (1 mM), ATP-y-S, (500J#M), adenylyl-imidodiphosphate (AMP-PNP, 2 mM), calmodulin (10,M), calmodulinantipeptide (10 /tM) orthegrowth-associatedprotein, GAP-43 (2-5 mM) in the pipette did not affect M-current or modify responses to acetylcholine. 8. It is concluded that ACh-induced inhibition ofIK(M) is unlikely to be mediated byproducts ofphospholipase C stimulation, eitherindividually orincombination. It is suggested that activation of phospholipase C and inhibition of IK(M) represent parallel pathway responses to ACh, and that IK(M) might be inhibited by a direct effect of the activated G-protein(s) or through another (unidentified) enzymatic product. INTRODUCTION InNG108-15 mouseneuroblastoma xratgliomahybridcellstransformedbyDNA transfection to express ml (or m3) muscarinic acetylcholine receptors, application of acetylcholineproducestwomembranecurrentchanges: itactivatesaCa2+-dependent K+ current, IK(Ca), and inhibits an M-type voltage-dependent K+ current, IK(M) (Fukuda et al. 1988; Neher, Marty, Fukuda, Kudo & Numa, 1988; Robbins, Caulfield, Higashida & Brown, 1991). Activation of IK(ca) probably results from the formation of inositol 1,4,5- trisphosphate(InsP3)andsubsequentreleaseofintracellularCa2+,since(i)bothml and m3 receptor activation stimulated inositol phosphate production (Fukuda et al. 1988), (ii) outward current activation coincides with a rise in intracellular [Ca2+] (Neher et al. 1988), and (iii) a comparable effect is induced by the intracellular iontophoretic injection of either InsP3 or Ca2+ ions (Higashida & Brown, 1986; Robbins, Cloues & Brown, 1992a). The mechanism responsible for IK(M) inhibition is less clear. Experiments on the analogous effect ofbradykinin have led to the suggestion that it might result from theparallelformation ofdiacylglycerols (DAG) andsubsequent activation ofprotein kinase C (PKC), since the effect ofbradykinin could be partly replicated by phorbol esters (Higashida & Brown, 1986; Schiifer, Behe & Meves, 1991) and partly blocked by inhibitors ofprotein kinase C (Schiifer et al. 1991). However, in other cells where IK(M) is inhibited by muscarinic receptor agonists, this mechanism has been discountedbecause agonist-inducedinhibitionwasnotpreventedbyproteinkinaseC inhibitors (e.g. frog ganglion cells, Bosma & Hille, 1989). Alternative suggestions include (i) a response to the rise inintracellular [Ca2+] (Kirkwood, Simmons, Mather & Lisman, 1991; but see Pfaffinger, Leibowitz, Subers, Nathanson, Almers & Hille, 1988; Beech, Bernheim, Mathie & Hille, 1991; Marrion, Zucker, Marsh & Adams, 1991), (ii) a Ca2+-independent effect ofInsP3 (Dutar & Nicoll, 1988), or (iii) a direct effect ofthe activated G-protein (Lopez, 1992; Chen & Smith, 1992). In the present experiments, therefore, we have attempted to find out more about M-CURRENT TRANSDUCTION 155 the mechanism responsible for IK(M) inhibition by acetylcholine in ml-transformed NG108-15 cells, using whole-cell patch-electrode recording in combination with intracellular [Ca2+] measurements with indo-1. These transformed cells have several advantages for such a study. First, they have a homogeneous population of expressed ml receptors, obviating some of the complexities of primary neurones. Second, a great deal is already known about their biochemical responses to muscarinic receptor stimulation. Third, the activation of IK(Ca) provides a useful electrophysiological measure of InsP3-induced Ca2+ release against which to judge the effectiveness of procedures designed to suppress this (or an antecedent) component ofresponse. We wouAd hope that the results obtained may be relevant to mechanisms for IK(M) inhibitien in other cells. METHODS Tissue culture. Neuroblastomaxglioma (NG108-15) cells transfected with the porcine brain acetylcholine muscarinic ml receptor (subclone PM1-8, Fukuda et al. 1988) were continuously grownat37 0CinDulbecco'sminimalessential medium (DMEM, highglucose) containing5%fetal calf serum (FCS), hypoxanthine (30,uM), aminopterine (12,CM), thymidine (4-8/,M) and L- glutamine (2 mM) and in the presence of 10% C02. Cells were grown to 70-80% confluence then passaged 1:3 every 3-4 days. For electrophysiological recording, cells were transferred to 35mm Petri dishes (density 2000-5000 cells per dish) which had been precoated with polyornithine; for microfluorimetry, they were plated onto polyornithine-coated glass coverslips (22 mmx22 mm). The cellswere differentiated by changing the medium 24 hlater to one inwhich the FCS hadbeen reduced to 1%, aminopterine omitted, and 10,UM prostaglandin E1 and 50,UM 3-isobutylmethyl- xanthine (IBMX) added. Electrophysiology. Thewhole-cellvariantofthepatch-clamptechniquewasusedindiscontinuous voltage-clamp mode (Axoclamp-2, Axon Instruments, Foster City, CA, USA) as detailed in Robbins, Trouslard, Marsh& Brown (1992b). CellsweresuperfusedwithamodifiedKrebssolution at35°C, ofcomposition (mM): NaCl, 120; KCI, 3; glucose, 11 1; NaHCO2, 22-6; MgCl2, 1-2; Hepes, 5; CaCl2, 2-5; tetrodotoxin, 00005; pH was 7-36 when gassed with 95% 02-5% CO2. Flow rates were between 5 and 10mlmin-. Electrodes (3-5MQ) were normally filled with a solution containing (mM); KOOCCH3, 90; KCI, 20; Hepes, 40; MgCl2, 3; EGTA, 3; CaCl2, 1. The calculated free calcium concentration was 40nim (programme REACT 2-01; G. L. Smith, Physiology Department, Glasgow University); measured resting levels of calcium were around 45 nm (see below).Accessresistancewasbetween4and9MQl.Cellswerevoltageclampedatbetween -20and -30mV andM-current deactivation tailswereevoked by hyperpolarizing steps for 1 sto -50 or -60 mV(seeRobbinsetal. 1992b). Current-voltagerelationshipswereobtainedusingincremental voltage stepsof10mVbetween -110and -10 mV. Calcium currentswererecordedunder similar conditions except that the electrode solution was CsCl based and the external medium contained tetraethylammonium chloride instead of NaCl. Currents were evoked by stepping to 0mV for 500ms from a holding potential of -90mV (see Caulfield, Robbins & Brown, 1992, for details). Pressure ejection. A separate micropipette was used to pressure-apply BaCl2 (100mM), CdC12 (1 mm) or acetylcholine (1-10 mM) dissolved in the external superfusate. The pressures routinely used were 138kPafor 50-300ms. The pipette was placed upstream ofthe cell and as close to the cell as possible without inducing pressure artifacts. Intracellular iontophoresis. In some experiments cells were also impaled with a second electrode filled with either InsP3 (100-500/SM) or CaCl2 (100200 mM), and InsP3 or Ca2+ injected by iontophoresis using current pulses of +10-40nA for 0 1-1 s (see Robbins et al. 1992a). Microfluorimetry. The method used for measuring intracellular [Ca2+] has been previously described in detail (Robbins et al. 1992b). In brief, cells loaded with either indo-1 AM (the acetoxymethyl ester of indo-1) or indo-1 (via the patch-pipette) were excited with UV light (360nm) and the emission at 408 and 488nm measured simultaneously to produce a ratio (R) 408/488. This was converted to free calcium by the expression: [Ca2+] = [(R-Rmin)/ (Rm-R)]xKd(FO/FS),whereRm. = 40,Rmin = 0-38andKd((FI,/F) = 1400 nM. Electrode solutions 156 J. ROBBINS, S. J. MARSH AND D. A. BROWN were as detailed above but with the inclusion ofindo-1 tetrapotassium salt (01 mM) plus 0.1 mm BAPTA, with no added calcium or EGTA. Some experiments were done on cells preloaded for 30-40minwith 5/uM indo-1-AM ester, to make calcium measurements during the transition from cell-attached to whole-cell recording mode. Drugs and chemicals. The following were used (with sources): acetylcholine chloride (ACh), apamin,bradykinin (BK), tetrodotoxin (TTX),ethyleneglycolbis-(fl-aminoethylether)N,N,N',N'- tetraacetic acid (EGTA), 1,2-bis(2-aminophenoxy)ethane -N,N,N',N'-tetraacetic acid (BAPTA), atropine sulphate, adenosine triphosphate (ATP), dibutyryl-cAMP, adenosine 5'-O-(3-thiotri- phosphate) (ATP-y-S), guanosine triphosphate (GTP), guanosine 5'-O-(3-thiotriphosphate) (GTP- y-S), guanosine 5'-O-(2-thiodiphosphate) (GDP-,8-S), mastoparan, arachidonic acid (AA), 4- bromophenacyl bromide (BPB), indomethacin, neomycin sulphate, potassium fluoride, doxo- rubicin (Doxo), phorbol dibutyrate (PDBu), oleic acid, linoleic acid, elaidic acid, heparin low molecular weight, nordihydroguaiaretic acid (NDGA),NV-nitro-L-arginine (L-NOARG), 8-bromo- cyclic guanosine monophosphate (8-bromo-cGMP), sodium nitroprusside and 2,3-butanedione monoxime (BDM) (all from Sigma, Dorset, UK); 1,4,5-inositol trisphosphate (InsP3), 1,3,4,5- inositol tetrakisphosphate (InsP4), BAPTA AM, indo-1, indo-1 AM and calmodulin (CAM) (from Calbiochem, CA, USA); staurosporine and adenylyl-imidodiphosphate (AMP-PNP) (Boehringer Mannheim, Germany); Bordatella pertussis (PTX) toxin (Porton Products, Dorset, UK); charybdotoxin (ChTX; Latoxan, Rosans, France); DL-muscarine chloride (RBI, MA, USA); eicosatetraynoicacid(ETYA;CaymanChemical,MI,USA);andokadaicacid(ScientificMarketing Associates, Herts, UK). Drugs that were not water soluble were dissolved in ethanol or dimethyl sulphoxide(DMSO), anddilutedat > 1:1000inKrebssolutionforuse. (ControldilutionsofDMSO or ethanol were without effect.) GAP-43 C-terminal decapeptide was a gift from Dr G. Milligan (Department ofBiochemistry, GlasgowUniversity). Calmodulin bindingpeptide (CBP), atwenty- nine amino acid polypeptide (Kelly, Weinberger & Waxman, 1988) was synthesized by Dr S. Bansal(DepartmentofPharmaceuticalChemistry,KingsCollege,London).Purifiedproteinkinase C (PKC)wasagiftfromDrP. J. Parker(ICRF, London). Thecyclophilin-cyclosporinA (Cp-CsA) was kindly supplied by Dr R. J. Docherty (Sandoz Institute, London). RESULTS Controlperfusion ofajust-submaximal ACh concentration (100/tM) to asample of twenty-three ml-transformed cells produced an inward current at the standard holding potential ofabout -30 mV, and inhibitedIK(M) deactivation relaxations at -60 mV by47-3+3-2% (mean+S.E.M.; n = 23; see Fig. 1 and Table 1). This agrees withprevious observations thattheIK(M) deactivation relaxationsinthese cellswere inhibited maximally by about 50% at > 1 mm acetylcholine (Robbins et al. 1991). Current-voltage curves (Fig. 1C) showed that the inward current resulted from a reduction in the normal outward rectification observed at potentials positive to -60 mV, with little or no change in the linear component ofthe curve negative to -60 mV. This accords with the supposition that the inward current results from inhibition ofIK(M) (Robbins et al. 1992b). In twelve outofthe twenty-three cells, the inward currentwas preceded by atransient outward current, aspreviously reported by Fukuda et al. (1988) and Neher et al. (1988) (Fig. 1). In a separate sample of eight cells impaled with potassium citrate-filled microelectrodes (cf. Fukuda et al. 1988), 100,UM ACh inhibitedIK(M) by 60+11 2%; an initial outward current was observed in six of these. Hence, whole-cell patch clamping, and consequent exchange of the normal intracellular solution with the electrode solution, did not materially alter the effect ofACh. Bradykinin (BK) also activates IK(Ca) and inhibits IK(M) in NG108-15 cells, by activating endogenous BKreceptors (Higashida & Brown, 1986; Schiifer etal. 1991). In contrast to ACh, the effect of BK wanes rapidly during continued exposure, as shown in Fig. 2A. However, in the continued presence ofBK, ACh could still inhibit M-CURRENT TRANSDUCTION 157 3minafter A Control ACh (100 Wash |1 nA 10s B Control ACh (100/ctM) Wash i bo rfH 11 nA 1 s C 70- 5.0- a co 3.0, 0 1*0 -110-100 -90 -80 -70 -60 -50 -40 -30 -20 -10 Membrane potential (mV) Fig. 1. Effects ofAChonpotassium currents inml transformed NG108-15 cells. The cell inAwasclampedat -28mV(VH)andsteppedto -58mVfor1 sbefore,duringandafter superfusion with 100/M ACh (at bar). During these hyperpolarizing steps (lower trace) the recorded current shows a slow inward relaxation, due to deactivation ofIK(M) (see Robbinsetal. 1992b);thetransientinwardcurrentattheendofthestepisaCa2+current. Note thatACh induced a transient outward current (at arrow), then an inward current; duringthelattertheamplitudeoftheIK(M) deactivationrelaxationwasreduced. Records inB show current responses ofanother cell to a series ofvoltage steps from VH-26mV to command potentials between -16 and -106mV in 10mV increments, recorded before, duringandaftersuperfusionwith 100/MACh. ThegraphsinCshowtheabsolute currentlevelattainedattheendofeachvoltagestepinB (ordinates,nA) plottedagainst thecommandpotential (abscissae, mV)before (0), during(@)andafter(A) superfusion with ACh. 158 J. ROBBINS, S. J. MARSH AND D. A. BROWN TABLE 1. Effects oftransduction pathway modifiers on M-current inhibition evoked by acetylcholine (100/M) Concentration M-currentt M-current§ Number Compound (mM) Routet amplitude inhibition ofcells Control - 330+34 473+3-2 23 G-proteins GTP-y-S 05 Int 118+24* 53-3+6-9 6 GDP-,f-S 10.0 Int 477+173 56-6_6-2 5 Mastoparan 041 Int 400+106 47-8+14 4 PTX (500ngml-') Ext 578+107 41P5±50 8 Calcium Calcium free Ext 650+216 450+49 5 BAPTA 20-0 Int 317+73 42-8+4_5 6 BAPTA AM 0*1 Ext 519+62 398+52 4 Phospholipase C pathway Heparin 1.0 Int 546+131 130±3.8** 7 Neomycin 1.0 Ext 311+37 408+47 6 Neomycin 1*0 Int 413+99 495+45 4 Doxorubicin 0*1 Int 300+67 490+2-5 4 InsP3 041 Int 268+86 746+67** 9 InsP4 0*1 Int 371+56 42-0+5-9 6 PDBu 01001 Ext 424+71 33-2+3.7* 11 PDBu (+5mm ATP) 0.001 Ext 819+180 425+54 4 PDBu (sharp) 0.001 Ext 700+253 42-6+3-6 5 Staurosporine 0-002 Int 236+42 400+±1*2 5 Phospholipase A2 BPB 0.01 Ext 488+102 450+6-9 4 ETYA 0.01 Ext 320+96 49-8+49 5 NDGA 005 Ext 338+99 60-8+5-3 5 Indomethacin 005 Ext 260±80 55-8+6-8 5 Arachidonic acid 005 Ext 288+65 416+7-5 5 NO synthase pathway L-NOARG 041 Ext 600+195 41P7+5-5 6 8-Bromo-cGMP 1.0 Ext 510+195 5 Sodium nitroprusside 1-0 Ext 490+177 5 Nucleotides ATP 50 Int 725+130 45-8+641 4 GTP 2-0 Int 225+38 540+40 3 AMP-PNP 2-0 Int 300+63 61P8+7-4 5 ATP-y-S 05 Int 320+42 458+69 5 (+1 mm GTP) Phosphorylation-dephosphorylation BDM 20-0 Ext 590+176 666+61 5 Okadaic acid 0005 Ext 1250+164 52-6+7-3 8 Cp-CsA 0100002-0-00005 Int 588+226 538+52 4 Calmodulin CAM 0.01 Int 413+99 440+8-3 4 CBP 0-01 Int 808+249 463+96 3 Significant difference from control: *P <005, **P <010005. t Int, internal; Ext, external. I Mean±+S.E.M. amplitude (pA), measured at -50 to -60mV. § Mean+S.E.M. percentage inhibition. M-CURRENT TRANSDUCTION 159 IK(M). Conversely, superimposition ofBK on the partial inhibition produced by ACh induced further inhibition (Fig. 2B). Thus, these two agonists seem to inhibit IK(M) in a non-interactive manner, and do not cross-desensitize. Response latency To find out how rapidly ACh could inhibit IK(M), we tested the effect of focal pressure-application ofhigh concentrations (1-10 mM) ofACh (Fig. 3; see Methods). A BK(10#M) 700 - ACh(100#M)El 600 0* no 500 0 - 'E 400 / E 2300 0 200 100 l l l l l 0 0 1 2 3 4 5 6 7 8 9 10 11 12 B ACh(000lM) 500 BK(10tM) t0. 400 300 / E 0 0w 0~~~~~~~ o~~~~~~~~~~ 0 1 2 3 4 5 6 7 8 9 10 11 12 Time (min) Fig. 2. Effects ofACh and bradykinin onIK(M) in two ml-transformed NG108-15 cells. GraphsshowM-currentdeactivationrelaxation amplitudes atthecommandpotential of -60mVevokedfromaholdingpotentialof-30mVfor 1 severy30s. Bradykinin (BK, 10/LM)andACh(100,UM)wereappliedbybathperfusionforthedurationofthefilledand openbarsrespectively. InA, BKwasappliedfirstandproducedaninwardcurrentwhich thenwanedin the continued presence ofBK. Superimposed AChthen induced a second inward current. InB, ACh produced a steady inward current, which was increased on subsequent application ofBK. In these experiments, the initial outward current (due to activation ofIK(Ca) see Fukuda et al. 1988, and Neher et al. 1988) was suppressed by adding 100 nm apamin and 10 nm charybdotoxin to the bathing fluid and pressure-pipette solution. The inward current produced by these ACh applications was fully blocked by 1 ,UM atropine, hence resulted exclusively from muscarinic receptor activation. To control 160 J. ROBBINS, S. J. MARSH AND D. A. BROWN for the time taken for ejected ACh to reach the cell, we also pressure-applied Ba2+ ions: this probably blocks the M-channels directly (Robbins et al. 1992b). As shown in Fig. 3, the effect of Ba21 was clearly more rapid than that of ACh. The mean latencyto observable inward current development afterthe onset ofapressure-pulse A 100pA B 100pA is Fig. 3.TimecourseofIK(M)inhibitionbybarium (Ba2+)andAChrecordedat35°C(A)and 23°C (B) (two cells). Micropipettes containing 100mm Ba2+ or 1 mm ACh were placed closetothecellsurfaceandejectedusing 100ms, 138kPapressurepulses.Cellswereheld at -20 mV and bathed in a solution containing apamin (100nM) and charybdotoxin (10nM) in order to block contaminating IK(Ca) of Ba21 was 147-7+14-1 ms (n = 12) at 35°C (minimum latency, 81 ms) and 91P7+8-3 ms (n = 3) at 23 'C. In contrast, the mean latencies for the inward current induced by 1 mm ACh were 419-5+63-7 ms (n = 10) at 35 'C (minimum latency, 202 ms) and 1793+288 ms at 23 'C (n = 8). Increasing the concentration ofACh to 10 mM did not shorten the latency (412-5+51P9 ms at 35 °C; n = 4). Thus, assuming ACh andBa2+reach the cell membrane atthe same time, there is an additionaldelay of about 270 ms at 35 'C between activation ofthe ml receptors and inhibition of M-CURRENT TRANSDUCTION 161 A 120- 1T T~~~~~~~~~~ 80 ~60 10 20 30 40 50 E) 80 iX oA~Co oA~ ~TT~~~~ a-= @ 020 T~~~~~~~ z TA ±~~~ 0 10 20 30 40 50 60 Time (min) 401 40-n2 B Control LACh, 1t00 nm Wash (5 I0 IM' i as [0.25 nA GTP-y-S 500zM--S Wash (500Itm)~~~~~~~~~~~~~~~~~II 6mpin wa min Fig. 4. Effect of irreversible G-protein activation on I,(.). The graph in A shows the change in M-current (I,(.)) amplitude with time following breakthrough with pipettes containingthenormalintracellularsolution (0,seeMethods), orwiththenormalsolution supplementedwith500/LmGTP-y-S (0)or 10mmpotassiumfluoride (A). Cellswereheld at -30mVandsteppedto -60mVfor1 severy30 s.M-currentamplitudewasmeasured astheamplitudeofthedeactivationrelaxationsinducedbythehyperpolarizingsteps (see Methods and Fig. 1), and expressed as a fraction of that recorded immediately after breakthrough (time 0). Eachpoint is the mean fromfive cells; bars show S.E.M. Records in B show responses oftwo cells to a briefbath application ofACh, recorded with the normal pipette solution (upper trace) and with 500#M GTP-y-S in the pipette (lower trace). Note that IK(M) recovers within 6min of removing ACh in the absence of GTP-y-S, butremains depressed 60minafterremovingAChinthepresence ofGTP-y-S. 162 J. ROBBINS, S. J. MARSH AND D. A. BROWN IK(M). This lengthens to about 1P7 s at 23 °C, giving a temperature coefficient, Q10, over the range 22-35 °C, for the transduction process ofabout 3-6. NG108-15 cells possess endogenous m4 muscarinic acetylcholine receptors, activation of which inhibits an wo-conotoxin-sensitive component of Ca2+ current (Caulfield et al. 1992). There is no evidence to suggest that this involves any intracellular transduction process; instead, by analogy with equivalent effects in sympathetic neurones (Mathie, Bernheim & Hille, 1992), it seems more likely that this effect involves a local ('membrane-delimited') response. It therefore seemed interesting to find out how rapidly ACh could inhibit the Ca2+ current. To test this, we applied long (2 s) depolarizing steps to 0 mV from -90 mV using Cs+-filled electrodes (see Methods), and then pressure-applied either 1 mm ACh or 1 mm Cd2+ (which blocks the Ca2+ channels directly) at a predetermined time (usually 50-100 ms) aftertheonsetofthevoltagestep. Meanlatenciestodetectablereduction inICa (at 35 °C) were: Cd2+, 69-3+3-5 ms (n = 4); ACh, 68-1+3-3 ms (n = 6). These were indistinguishable, and probably within the limits of the pressure-ejection system. Notwithstanding, theeffectofAChonICawas clearly much more rapidthan its effect on IK(M) G-proteins To test whether a G-protein was involved in the transduction pathway between muscarinic receptor activation and M-current inhibition, we added GTP (2 mM), GTP-y-S (0-5 mM), GDP-,f-S (up to 10 mM) or potassium fluoride (10 mM) to the pipette solution. Addition ofGTP did not modify the response to 100fM ACh (see Robbins et al. 1991). Addition ofGTP-y-S or potassium fluoride caused a slow run- down of the current, with half-times of about 20 min (for potassium fluoride) and 30 min (for GTP-y-S) (Fig. 4A). When ACh (100,tM) was applied in the presence of 05 mm intracellular GTP-y-S, maximum inhibition of residual IK(M) was not significantly increased (Table 1), but there was no longer any recovery following wash-out ofACh (Fig. 4B). These effects agree qualitatively with those previously reported in sympathetic ganglion cells (Pfaffinger, 1988; Brown, Marrion & Smart, 1989; Lopez & Adams, 1989). However, and unlike previous observations on sympathetic ganglion cells where apartial block ofthe agonist-evoked response was seen (see above, and Simmons & Mather, 1991), GDP-fl-S, even at 10 mm, had no significant inhibitory effect on the action of ACh (Table 1). (GDP-fl-S was also ineffective in preventing inhibition of ICa by noradrenaline; I. McFadzean, Department of Pharmacology, Kings' College London, personal communication. Resistance to GDP-,/-S might result from a high rate of GTP production in these cells, as evidenced by the lack of any requirement for added GTP in the pipette solution.) In agreement with previous observations on sympathetic neurones (Pfaffinger, 1988), pretreatment for 24 h with 500 ng ml-' pertussis toxin did not significantly reduce the effect ofACh (Table 1). In parallel experiments on these cells, the same pretreatment regimen suppressed the inhibition ofICa by acetylcholine (Higashida, Hashii, Fukuda, Caulfield, Numa & Brown, 1990). We also tested the effects of adding to the pipette solution 100#M mastoparan (which directly activates certain G-proteins; Higashijima, Uzu, Nakajima & Ross, 1988) or 2-5 mm ofthe C-terminal decapeptide ofGAP-43 (which has been reported