Journal ofNeurochemistryLippincott—Raven Publishers, Philadelphia© 1996 International Society for Neurochemistry

Exocytosis Coupled to Mobilization of Intracellular Calciumby Muscarine and Caffeine in Rat Chromaffin Cells

Xi Guo, Dennis A. Przywara, Taruna D. Wakade, and Arun R. Wakade

Department of Phannacology, Wayne State University School of Medicine, Detroit, Michigan, U.S.A.

Abstract: We used cultured rat chromaffin cells to testthe hypothesis that Ca2~entry but not release from inter-nal stores is utilized for exocytosis. Two protocols wereused to identify internal versus external Ca2~sources: (a)Ca2~surrounding single cells was transiently displacedby applying agonist with orwithout Ca2~from an ejectionpipette. (b) Intracellular stores of Ca2~were depleted bysoaking cells in Ca2~-free plus 1 mM EGTA solution be-fore transient exposure to agonist plus Ca2~.Exocytosisfrom individual cells was measured by microelectrochem-ical detection, and the intracellular Ca2~concentration([Ca2~],) was measured by indo-1 fluorescence. KCI (35mM) and nicotine (10 tiM) caused an immediate increasein [Ca2~],and secretion in cells with or without internalCa2~stores, but only when applied with Ca2~in the ejec-tion pipette. Caffeine (10 mM) and muscarine (30

1zM)evoked exocytosis whether or not Ca

2~was included inthe pipette, but neither produced responses in cells de-pleted of internal Ca2~stores. Pretreatment with ryano-dine (0.1 ~tM)inhibited caffeine- but not muscarine-stim-ulated responses. Elevated [Ca2~]and exocytosis exhib-ited long latency to onset after stimulation by caffeine(2.9 ±0.38 s) ormuscarine (2.2 ±0.25 s). However, theduration of caffeine-evoked exocytosis (7.1 ±0.8 s) wassignificantly shorter than that evoked by muscarine (33.1±3.5 s). The duration of caffeine-evoked exocytosis wasnot affected by changing the application period between0.5 and 30 s. An ~-~20-srefractory period was found be-tween repeated caffeine-evoked exocytotic bursts eventhough [Ca2~]continued to be elevated. However, mus-carine or nicotine could evoke exocytosis during the caf-feine refractory period. We conclude that muscarine andcaffeine mobilize different internal Ca2~stores and thatboth are coupled to exocytosis in rat chromaffin cells.The nicotinic component of acetylcholine action dependsprimarily on influx of external Ca2~.These results andconclusions are consistent with our original observationsin the perfused adrenal gland. Key Words: Catechola-mine secretion —Chromaffin cells— Exocytosis—Mus-carine receptors—Calcium mobilization— Ryanodine.J. Neurochem. 67, 155—162 (1996).

mines (Douglas and Poisner, 1965; Kirpekar et al.,1982; Wakade and Wakade, 1983; Borges et al., 1987;Marley, 1987; Nakazato et al., 1988; Warashina et a!.,1989). Biochemical (Knight and Baker, 1986; Evingeret a!., 1992) and electrophysiological (Knight and Ma-conochie, 1987; Akaike et al., 1990a,b; Inoue and Ku-riyama, 1991; Neely and Lingle, 1992) properties ofchromaffin cells after stimulation with muscarinic ago-fists and antagonists have been examined. Althoughthis wealth of information has greatly advanced ourunderstanding of the actions of muscarinic receptors,it still remains unclear how these receptors affect Ca2~mobilization to stimulate exocytosis in several speciesof chromaffin cells.

In bovine adrenal chromaffin cells, muscarinic re-ceptor activation had been reported both to inhibit (De-rome et al., 1981) and to facilitate (Forsberg et al.,1986) nicotinic receptor-mediated secretion. Currentlyit is accepted that muscarinic receptors have only aminor role in catecholamine secretion in the bovinemodel (Fisher eta!., 1981; Cheek and Burgoyne, 1985;Cheek, 1989; Kim and Westhead, 1989). This conclu-sion has been supported by analysis of the distributionof agonist-evoked elevated intracellular Ca2~concen-tration (IiCa2~}

1).Muscarinic agonists or caffeine,which are poor secretagogues in bovine chromaffincells, cause a diffuse rise in [Ca

2~i1 across the cell but

do not elevate subplasma membrane [Ca2~]

1 as domore potent secretagogues, which act via Ca

2~entryfrom the external medium (Cheek et a!., 1989; O’Sulli-van et a!., 1989). These studies have led to the generalconclusion that entry of extracellular Ca2~via recep-tor- or voltage-operated channels is linked to exo-cytosis, whereas internally released Ca2~is not (Ba!-lesta et al., 1989; Bunn and Marley, 1989; Cheek,1989; Burgoyne, 1991; Yamagami et al., 1991).

In the rat adrenal gland, we demonstrated that omis-

Several laboratories have provided convincing evi-dence that stimulation of muscarinic receptors by neu-rally released or exogenously applied acetylcholine islinked to exocytotic secretion of adrenal catechola-

Received October 25, 1995; revised manuscript received February1, 1996; accepted February 12, 1996.

Address correspondence and reprint requests to Dr. D. A. Przy-wara at Department of Pharmacology, Wayne State UniversitySchool of Medicine, 540 East Canfield, Detroit, MI 48201, U.S.A.

Abbreviation used: [Ca2i,, intracellular Ca2~concentration.

155

156 X. GUO ET AL.

sion of CaC!2 from the perfusion medium abolishedcatecholamine secretion evoked by nicotine and KC1but not by muscarine (Malhotra et al., 1988a). Thisobservation suggested that muscarine utilizes internalCa

2~mobilized via inosito! 1 ,4,5-tnsphosphate (Ma!-hotra et al., 1988b). Addition of EGTA to Ca2~-freemedium abolished the secretory response to all ago-nists (Maihotra et a!., 1988a). More recently we haveused cultured chromaffin cells of the rat to extend ourstudies and have found that, as in the intact adrenalgland, both nicotinic and muscarinic receptors partici-pate in secretion of catecholamines (Chowdhury et a!.,1994). However, when cultured chromaffin cells werein Ca2~-freemedium for 15—20 mm, the secretoryresponse to muscarine was completely absent, as wasthe case for nicotine and excess KC1. Failure of musca-rifle to induce catecholamine secretion in cultured ratchromaffin cells under these conditions could be dueto lack of external Ca2~or depletion of internal storesduring the period of exposure to Ca2~-free solution.

We have developed two protocols to resolve theissue of extrace!lular versus intracellular sources ofCa2~in the secretory response of rat chromaffin cellsto muscarinic stimulation. In one design, we main-tained the chromaffin cells in a Ca2~-containing me-dium and transiently changed the environment sur-rounding single cells by brief application of agonistwith or without Ca2~from an ejection pipette. In an-other protocol, intracellular Ca2~stores were depletedby soaking chromaffin cells in Ca2~-free plus ! mMEGTA solution, and then the cells were transientlyexposed to agonistplus Ca2~.Exocytosis from individ-ual cells was measured by microelectrochemica! detec-tion (Leszczyszyn et al., 1990; Wightman et al., 1991;Chow et a!., 1992; Chowdhury et a!., 1994). Parallelstudies were conducted on the same batch of culturedchromaffin cells to measure [Ca2~]~ using the indo-1ratiometric technique (Grynkiewicz et a!., 1985) andour previously established methods (Przywara et a!.,1991, 1993). With this approach we demonstrate thatinternally mobilized Ca2~by either caffeine or musca-rine is coupled to exocytosis in rat chromaffin cells.

MATERIALS AND METHODS

Primary cultures of rat chromaffin cellsChromaffin cells were isolated from 19—31-day-old

Sprague—Dawley rat pups by enzymatic digestion using acombination of previously described protocols (Unsicker eta!., 1978; Tischler et a!., 1982; Lillien and Claude, 1985).Animals were killed by exposure to C0

2, adrenal glandswere removed, and medulla was dissected in ice-cold M-199 culture medium (GIBCO) andbisected. Medullary frag-ments were pooled andincubated (37°C,45 mm) with gentlerocking in 1.5 mg/mi of type I collagenase and 0.5 mg/ml of DNase (Worthington) in phosphate-buffered saline.Tissue fragments were transferred to trypsin (1.25 mg/m!in phosphate-buffered saline) and DNase (0.5 mg/ml) andincubated for 30 mm, and cells were dissociated by tritura-tion. Cells were washed twice by centrifugation (600 g, 5

mm) in M- 199 with 15% fetal bovine serum (GIBCO; heat-inactivated, charcoal-treated), resuspended in fresh mediumcontaining 0.! p~Mdexamethasone, and plated on collagen-coated glass coverslips in 35-mm-diameter culture dishes.Cultures were maintained at 37°Cin a water-saturated atmo-sphere of 95% air/5% CO2. and medium was changed every2 days. Cells were used after 2—7 days in culture. For moni-toring of exocytosis or [Ca

2~], (see below), cells were trans-ferred to HEPES-buffered solution containing 119 mMNaC1, 4.7 mM KC1, 1.2 mM MgSO

4, 2.5 mM CaC12, 10mM HEPES buffer, and 10 mM glucose, pH 7.4 with NaOH.Ca

2~-free solution had no added CaCl2 and was supple-

mented with 1 mM EGTA.

Electrochemical detection of exocytosisIndividual secretory events from single rat chromaffin

cells were detected using established microelectrochemicaltechniques (Leszczyszyn eta!., 1990; Wightman eta!., 199!;Chow eta!., 1992; Chowdhury eta!., 1994). Electrodeswereconstructed as described by Chow et a!. (1992) using an 8-,~tmcarbon fibercannulated and sealed into a pulled polyeth-ylene tube (final tip diameter, <9 tim) glued into a glasscapillary. The capillary was back filled with 3 M KC!, andan Ag/AgC1 wire was used for electrical connection to anEl 400 potentiostat (Ensman Instrumentation, Bloomington,IN, U.S.A.) coupled to apersonal computer for dataacquisi-tion and analysis (Data Translations A/D-D/A board andGlobal Lab software). A significant improvement in re-cording of exocytotic events in this study over our previouswork (Chowdhury eta!., 1994)resulted from better electrodesealing within the polyethylene by reheating the tip aftercutting the pulled electrode. The exposed tip of the carbonfiber electrode was positioned within 1 ~.tmof a chromaffincell, and thepotential washeld at 650 mV to recordcurrentsresulting from oxidation of catecholamines released duringindividual exocytotic events (Leszczyszyn et a!., 1990;Wightman et al., 1991).

Secretagogues in control or Ca2~-freebath solutions were

applied by pressure ejection (PLI-!00 pico-injector; MedicalSystems) from pulled glass capillaries positioned ~20 ~imfrom the cell. Tip diameters ranged from 4 to 10 /.tm, andejection pressure was adjusted between 0.8 and 1.8 psi toachieve flow rates of ~-~20nl/s. In most experiments eachcell was used as its own control. Traces shown in Resultsare representative of five to 29 experiments repeated in twoor three different batches of cultured chromaffin cells.

Measurement of [Ca2~]~[Ca2~]

1was measured as previously described (Przywaraet al., 1991). Chromaffin cells cultured on glass coverslipswere incubated with 0.5— 1.0 jjM indo- !-acetoxymethylesterat room temperature for 1 h and washed to remove excessdye. The coverslip was secured in a Leiden chamber to thestage of an ACAS 570 confocal laser photometer (MeridianInstruments, Lansing, MI, U.S.A.). Cells were illuminatedby laser light (353—361 nm), and indo-1 fluorescence wasrecorded at 405 (Ca

2-bound) and 485 nm (Ca2~-free)wavelengths. Ratioing (405 nm1485 nm) was used to elimi-nate artifacts due to variations in thickness or indo- 1 distribu-tion in the cells (Grynkiewicz et a!., 1985). Calibration offluorescence and estimation of [Ca2~I, were as previouslydescribed (Przywara eta!., 1993)using in vitro solutions thatmimic intracellular ionic conditions. The calibration buffercontained 100 mM KC1, 1 mM EGTA, 50 mM HEPES, and1 ,aM indo-! salt, pH 7.2 at 25°C.The Ca2~concentration

J. Neurochem., Vol. 67, No. 1, 1996

EXOCYTOSIS BY INTERNAL Ca2~ 157

FIG. 1. Control of external Ca2~and exocytosis during KCI de-polarization. Amperometric detection of individual exocytoticevents (a and b) and [Ca2~]

1determined by indo-1 fluorescenceratio (C and d) were recorded from single rat chromaffin cells ina 2.5 mM Ca

2~-containing bath solution. KCl (35 mM, with orwithout Ca2~as indicated) was applied for 500 ms (arrows) or10 s (horizontal bars) by a pressure ejection pipette positioned—20 ~m from the cell. The three records in (a) are from thesame cell and are representative of 10 experiments repeated indifferent batches of cultured cells. The three records in (b) arefrom the same cell and are representative of 16 and 14 observa-tions with or without, respectively, Ca2~in the ejection pipette.[Ca2~]

1traces are from different cells and are representative ofsix observations under each experimental condition.

was adjusted using CaC12 standard solution. The free Ca2~

concentration was calculated using the ACAS 570 softwareto solve a double quadratic equation that accounts for Ca2~binding to indo-1 and EGTA. The K

0 values used for Ca2~-

EGTA and Ca2~-indo-!were 0.151 and 0.250 tiM, respec-tively (Grynkiewicz et al., 1985). [Ca2~]~was monitored inan area of cytosol using the point analysis mode of thephotometerwith continuous sampling (20-ms delay betweendata points) before, during, and after application of secreta-gogues from an ejection pipette. [Ca2~ I, traces shown inResults are representative of four to 10 observations.

RESULTS

Before examining receptor-mediated responses, weconfirmed and extended our protocols to identify inter-nal and external sources of Ca2~used in exocytosis.Depolarization by excess KC1 was used to identifyresponses due to Ca2~entry from the externa! medium,and caffeine was used to mobilize internal Ca2~stores.

Exocytosis by Ca2~influx during KC1depolarization

Chromaffin cells in a 2.5 mM Ca2~bath solutionwere stimulated by a brief (500-ms) pulse of 35 mMKCI, with or without Ca2~,from a pressure ejectionpipette aimed at the cell. KC! plus Ca2~caused animmediate burst of exocytotic events (Fig. la, toptrace). However, when KC1 was administered without

Ca2~in the pipette, there was no response (Fig. la,middle trace). This was not due to an inability of thece!! to respond, because reapplication of KC1 plus Ca2~could still stimulate exocytosis from the same cell (Fig.1 a, bottom trace). In parallel tests using chromaffincells loaded with the Ca2~fluorescent dye indo-!, aprominent elevation of [Ca2~j

1 was stimulated by a500-ms pulse of KC1 plus Ca

2~,but no change in[Ca2~j~ was observed when Ca2~was omitted fromthe ejection pipette (Fig. ic). Elevated [Ca2~], andexocytotic events stimulated by KC1 plus Ca2~weremaintained for several seconds after the 500-ms appli-cation. To determine if the environment presented bythe ejection pipette could be maintained for longertimes, the experiments were repeated using a 10-s ap-plication period (Fig. lb). Depolarization by 35 mMKC! plus Ca2~caused robust exocytosis during the 10-s application period, and exocytotic events continuedfor at least 10 s afterward (Fig. ib, top trace). How-ever, no exocytotic events were detected when KC!was applied without Ca2~(Fig. lb, midd!e trace). Thecell again exhibited robust exocytosis when retestedwith KC1 plus Ca2~(Fig. ib, bottom trace). The pro-longed exocytosis during and after the stimulation pe-riod is consistent with the prolonged elevation of[Ca2~]

1 observed in parallel experiments in indo-l-loaded cells (Fig. id). The inability of 35 mM KC1to evoke responses when administered without Ca

2~in the pipette, even though Ca2+ was present in thebath solution, indicates that the cell was exposed (ex-clusively) to the contents of the ejection pipette andthat Ca2~present in the surrounding medium was tran-siently displaced.

Exocytosis by internal Ca2~release induced bycaffeine

Caffeine (10 mM), which is known to mobilizeinternal Ca2~without a dependence on external Ca2~,was tested using 500-ms and 10-s application with andwithout Ca2~in the ejection pipette (Fig. 2). Unlikethe immediate response to KC1, caffeine-induced re-sponses followed the application with a latency of hun-dreds of milliseconds to several seconds. Furthermore,exocytosis and elevated [Ca2~ I, were unaffected whenCa2~was omitted from the ejection pipette, supportingthe idea that caffeine responses were due to mobiliza-tion of intracellular Ca2~stores. Elevated [Ca2~], fol-lowing caffeine stimulation also behaved differentlythan that produced by excess KC!. [Ca2~ I, remainedelevated near its peak value for >10 s after the 500-msapplication of caffeine plus Ca2~(Fig. 2c), whereasexocytosis terminated within ‘—5 s (Fig. 2a). Pro-longed elevation of [Ca2~ I, also occurred after a 10-sapplication of caffeine plus Ca2~(Fig. 2d), althoughexocytosis did not continue beyond the application pe-riod (Fig. 2b). Caffeine also produced fluctuations in[Ca2~], when applied for short or long durations as isapparent in Fig. 2d (upper trace) following cessationof caffeine application.

J. Neurochem., Vol. 67, No. 1, 1996

158 X. GUO ET AL.

FIG. 2. Caffeine-evoked responses are independent of extracel-lular Ca2~.Exocytotic events (a and b) and [Ca2]

1 (c and d)were recorded from single rat chromaffin cells following 500 ms(arrows) or lOs (horizontal bars) application of 10 mMcaffeineeither with or without Ca

2~in the ejection pipette as indicated.Records are from different cells and are representative of sevento 10 observations under each experimental condition. Otherdetails are as in Fig. 1.

We have previously shown that caffeine-sensitiveCa2~stores in sympathetic neurons are depleted after‘—10 mm in Ca2~-free medium (Wakade eta!., 1990).To verify the involvement of intracellular Ca2~mobili-zation in the caffeine responses, internal Ca2~storeswere depletedby keeping the chromaffin cells in Ca2~-free medium (plus 1 mM EGTA) for 15 mm beforetesting. Under these conditions, caffeine administeredwith Ca2~for 500 ms or 10 s was unable to evokeexocytosis, whereas 35 mM KC1 produced a typicalburst of exocytotic events (Fig. 3a and b). The effectsof caffeine and KC1 on [Ca2~ I, were consistent withtheir effects on exocytosis, with KC1 but not caffeinecausing a stimulated increase in [Ca2~I

1 in Ca2~-de-

pleted cells (Fig. 3c and d). The [Ca2~j~signal evokedby KC1 declined more rapidly in Ca2~-freemediumthan in Ca2~-containing medium (compare Figs. icand 3c). The possibility that KC1 responses were dueto nonspecific Ca2~influx through a “leaky” plasmamembrane following prolonged exposure to Ca2~-freeEGTA medium is ruled out by the failure of 2.5 mMCa2~(present with caffeine in the ejection pipette) tocause exocytotic or Ca2~responses. In control experi-ments, ejection pipettes containing 2.5 mM Ca2~HEPES without any drug were used to monitor stimu-!atory effects of Ca2~on cells maintained in a Ca2~-

free environment. In five experiments, there were fewexocytotic events produced by Ca2~alone (data notshown). These results indicate that chromaffin cellsdepleted of their internal Ca2~stores by treatment withCa2+ -free EGTA solution can be effectively used todetermine whether a test agent utilizes external Ca2~

or is totally dependent on the liberation of Ca 2+ frominternal stores.

Effects of muscarine on exocytosis and Ca2~mobilization

The above experiments provided suitable back-ground todetermine the source of Ca2~used by musca-rinic receptors to stimulate secretion in cultured ratchromaffin cells. Muscarine (30 [tM), like caffeine,exhibited a latency between application and the startof exocytosis (Fig. 4a and b) and the rise in [Ca2~]

1(Fig. 4c and d). Also like caffeine, muscarine waseffective when applied either with or without Ca

2~inthe ejection pipette. However, unlike caffeine, musca-rime-evoked responses were maintained far beyond theexposure period, whether it was for 500 ms (Fig. 4a)or 10 s (Fig. 4b and see Kinetics of caffeine- andmuscarine-evoked responses below). Oscillations of[Ca2~], were often apparent following exposure tomuscarine (Fig. 4c, upper trace, for example). In cellsdepleted of internal Ca2~,500-ms application of mus-came plus Ca2~was unable to evoke exocytosis orelevate ~Ca2~]~(Fig. 5a and b, lower traces). Thesefindings support the idea that mobilization of internalCa2~ is involved in muscarine-evoked exocytosis.However, a significant difference between muscarine-and caffeine-induced responses was apparent during10-s stimulation of cells depleted of their internal Ca2~stores (Fig. Sa and b, upper traces). Application ofmuscarine plus Ca2~produced exocytotic events (Fig.Sa) and elevated [Ca2~j~ (Fig. Sb), which began afteran initial delay and lasted for several seconds. Applica-tion of caffeine plus Ca2~was without effect under thesame conditions (Fig. 3b), suggesting that muscarine-sensitive Ca2~stores may be replenished during the10-s exposure period or that muscarine may activate

FIG. 3. Ca2~influx and exocytosis in cells depleted of internalCa2~.Exocytotic events (a and b) and [Ca2i~ (C and d) wererecorded from rat chromaffin cells maintained in Ca2~-free me-dium (plus 1 mM EGTA) and stimulated by 500 ms (arrows) or10 s (horizontal bars) application of 10 mM caffeine or 35 mMKCI as indicated. Records are from different cells and are repre-sentative of six observations under each experimental condition.

J. Neurochem., Vol. 67, No. 1, 1996

EXOCYTOSIS BY INTERNAL Ca2~ 159

FIG. 4. Exocytosis and elevated [Ca2~]~stimulated by muscarine(Mus). Exocytotic events (a and b) and [Ca2~l

1(c and d) wererecorded from rat chromaffin cells in 2.5 mM Ca

2~-containingmedium. Muscarine (30 ~M) was applied for 500 ms (arrows)or lOs (horizontal bars) eitherwith or without Ca2~in the ejectionpipette as indicated. Records are from different cells and arerepresentative of 12—21 observations under each experimentalcondition.

Ca2~influx as well as Ca2~mobilization. The —‘3-slatency between the start of muscarine application andthe first response in Fig. 5 argues against nonspecificCa2~entry through leaky plasma membranes, whichwould occur immediately on exposure to muscarmneplus Ca2~.

Effect of ryanodine on caffeine- and muscarine-evoked responses

Although caffeine and muscarine both mobilized in-tracellular Ca2~to evoke exocytosis, differences in theduration of responses evoked by each agonist and theability of muscarine to evoke responses in Ca2~-de-pleted cells following a 10-s exposure suggest thatmuscarine and caffeine act via different pathways ormobilize different Ca2~stores. Ryanodine (0.1—10

FIG. 5. Muscarine (Mus)-evokedresponse in cells depleted of in-ternal Ca2~. Exocytotic events(a) and [Ca2~],(b)were recordedfrom rat chromaffin cells main-tained in Ca2~-free medium (plus1 mM EGTA) and stimulated by10-s application (horizontal bars)of 30 ~iMmuscarine. Results arefrom different cells and are repre-sentative of 10 observations.

FIG. 6. Effects of ryanodine on caffeine- and muscarine (Mus)-evoked responses. Exocytotic events (a and b) and [Ca2~]~(cand d) were recorded from chromaffin cells in 2.5 mM Ca2~-containing medium and stimulated by 30 ~iMmuscarine or 10mM caffeine for 10 s (horizontal bars). Ryanodine was addedtothe bath (final concentration, 10 ~.tM),and 20 mm was allowedbefore reapplication of agonists as indicated. Records are fromdifferent cells and are representative of five or six observations.

1iM) was used to examine this issue further. Pretreat-ment with 0.1 ~M ryanodine for 20 mm had no effecton muscarine-evoked exocytosis but blocked caffeine-evoked exocytosis (data not shown). Increasing theryanodine concentration to 10 ~sMalso failed to inhibitmuscarmne-evoked exocytosis (Fig. 6a) and elevated[Ca

2~j~ (Fig. 6c) but completely blocked caffeine-evoked exocytosis in a reversible manner (Fig. 6b).Parallel to effects on exocytosis, ryanodine alsoblocked the caffeine-evoked increase in [Ca2~ 1~(Fig. 6d).

Kinetics of caffeine- and muscarine-evokedresponses

It was apparent that latency to onset and durationof action were substantially different for each of thesecretagogues tested. Following the start of KC1 appli-cation the latency to the first exocytotic eventwas 0.36±0.06 s (n = 22). For muscarine and caffeine thelatency averaged 2.2 ±0.25 s (n = 15) and 2.9 ±0.38s (n = 12), respectively. The duration of responseswas unique to each secretagogue. Increasing the dura-tion of exposure to KC! from 0.5 to 10 s caused anaccompanying increase in the duration of exocytosis(see Fig. 1, for example), whereas caffeine producedan equally short burst of exocytosis (7.6 ±1.1, 7.1±0.8, and 7.3 ±0.9 s) during exposure for 0.5, 5,and 10 s, respectively. Muscarine, on the other hand,typically produced exocytosis that lasted beyond the20-s recording period (see Fig. 4, for example). In aseparate set of experiments, muscarmne (30 ~sMfor 5

J. Neurochem., Vol. 67, No. 1, 1996

160 X. GUO ET AL.

FIG. 7. Temporal properties of caffeine-evoked exocytosis. Exo-cytotic events (a and c—f) and [Ca2~] (b) were recorded fromchromaffin cells in 2.5 mM Ca2~-containing medium. Caffeine(10 ~tM)was applied for 30 s (a and b) or repeatedly for 5 s (cand d). Records in (c) and (d) show 15- and 20-s periods,respectively, after the end of the first application. Results arefrom different cells and are representative of 18—25 observa-tions. Muscarine (30 ~M; Mus; e; six cells) or nicotine (10 ~M;Nic; f; six cells) was applied 5 s after caffeine using a secondejection pipette.

s) produced exocytotic events that lasted for an aver-age of 31.1 ±3.5 s (n = 12).

To determineif the exocytotic pattern following caf-feine could be prolonged, exocytosis was monitoredduring maintained caffeine exposure and during re-peated, brief stimulation by caffeine. Cells exposed tocaffeine for up to 30 s still exhibited an exocytoticresponse of —~7s in duration with typical latency afterthe onset of stimulation (Fig. 7a). The short durationof exocytosis was not due to reduction of [Ca2~I

1,which remained e!evated (Fig. 7b; also see Fig. 2d),or to depletion of releasable vesicles because applica-tion of muscarine (Fig. 7e), nicotine (Fig. 7f), or KC!(data not shown) was able to evoke exocytosis imme-diately after the caffeine-evoked response. However,repeated application of caffeine was not able to pro-duce a second round of exocytosis before an —‘20-srecovery period (Fig. 7c and d).

DISCUSSION

The broad conclusion from the above results is thatCa

2~release from internal Ca2~stores by caffeine ormuscarine is coupled to exocytosis in rat chromaffincells. Caffeine is well known to mobilize internal Ca2~stores in various cell types. Unlike agents that usedexternal Ca2~,caffeine-induced responses were simi-!ar whether or not Ca2~was included in the ejection

pipette and exhibited a latency consistent with timerequired to enter the cells and mobilize internal Ca2~.The occurrence of [Ca2~j

1 fluctuations during caffeinetreatment is also consistent with earlier reports of fluc-tuations in IICa

2~1~following mobilization of internalCa2~by caffeine (Ma!garoli et a!., 1990). However,Ca2~entry due to action potentials may also contributeto the fluctuations. It is important that cells maintainedin Ca2~-freeEGTA solution showed no response tocaffeine even when presented along with externalCa2~,supporting the idea that internal stores were de-pleted by this treatment and that the external Ca2~wasunable to produce any noticeable response. The abilityof internal Ca2~mobilized by caffeine to stimulateexocytosis in rat chromaffin cells is in agreement withrecent findings that caffeine effectively increases[Ca2~]~ and causes exocytosis in bovine chromaffincells (von Rtiden and Neher, 1993).

The use of agents that affect internal Ca2~mobiliza-tion may or may not affect Ca2~stores utilized byreceptor-mediated pathways. In this regard there isstrong evidence that caffeine- and ryanodine-sensitiveCa2~stores are not identical to those coupled to cellsurface receptors through production of inosito! 1,4,5-trisphosphate. Our findings show that muscarine uti-lizes internal Ca2~to evoke secretion and has effectssimilar, but not identical, to those of caffeine. Likecaffeine, muscarine was effective either with or with-out Ca2~in the ejection pipette, exhibited latency, andcaused [Ca2~j, fluctuations and was noteffective whenapplied for 500 ms in cells depleted of internal Ca2~stores. A!! these findings indicate that muscarine actsthrough mobilization of internal Ca2~similar to caf-feine. However, there were distinct differences be-tween the two agonists. In ce!!s maintained in Ca2~-containing medium, caffeine applied for 500 ms to 30 scaused only a brief ( —‘7 s) burst of exocytosis, whereasmuscarine-evoked exocytosis persisted for —30 s. It isinteresting that the elevated [Ca2~], stimulated by thetwo agonists in the presence of Ca2~was of similarmagnitude and duration. Finally, only caffeine-evokedresponses were sensitive to block by ryanodine. Thus,the difference in exocytosis by the two agonists islikely due to participation of different internal Ca2~stores, as we!! as the fact that muscarine acts via recep-tor-mediated generation of inosito! 1 ,4,5-trisphosphateand diacy!g!ycero!, which may facilitate the releaseprocess (Ma!hotra et a!., 1988b, 1989).

An intriguing difference between muscarine and caf-feine was found when the agonists were applied for10 s to cells depleted of internal Ca2* stores. Afterapplication of muscarine plus Ca2~for several secondsthere was a rise in [Ca2~]~that was accompanied byexocytosis (Fig. 5). This cannot simply be the resultof Ca2~entry into Ca2~-depleted cells because underidentical conditions Ca2~plus caffeine was withouteffect. Furthermore, as demonstrated by the effects ofKC! plus Ca2~applied to Ca2~-depletedcells, Ca2~entry from the external medium occurs immediately.

J. Neurochem., Vol. 67, No. 1, 1996

EXOCYTOSIS BY INTERNAL Ca2~ 161

With muscarine, there was always a substantial latencybefore [Ca2~1~ was elevated and exocytosis com-menced, consistent with the involvement of receptor-mediated pathways and intracellular Ca2~mobiliza-tion. A possible explanation for the response to musca-rine plus Ca2~in Ca2~-depleted cells is that a partialreplenishing of Ca2~stores occurs during the applica-tion of muscarine plus Ca2~,followed by mobilizationof these stores by second messengers generated duringthe latent period. This interpretation is consistent withthe hypothesis of capacitative coupling of Ca2~in-flux with the emptying of second messenger (inosito!1 ,4,5-trisphosphate) -sensitive Ca2~ stores (Putney,1990a,b). An alternate possibility is that muscarinicreceptor-coupled pathways directly activate a Ca2~current. A muscarinic receptor-activated Ca2~currenthas been reported in chromaffin cells from the adrenalglands of chicken (Knight and Maconochie, 1987),guinea pig (Inoue and Kuriyama, 1991), and pig(Forsberg and Xu, 1993). In either case, it appearsthat muscarine may be having two effects: one depen-dent on internal Ca2~stores and the other on externalCa2~.This conclusion is supported by the ability ofryanodine to delay the muscarine-stimu!ated rise inCa2~concentration (Fig. 6c). One interpretation ofthese ~Ca2~]

1data is that ryanodine inhibits the musca-rime-induced early release of internal Ca

2~stores butnot the later [Ca2~], elevation that does not depend oncytosolic stores.

The present study allows us to speculate on the phys-iologic significance of having two types of cholinergicreceptor to regulate catecho!amine secretion in a singlechromaffin cell (Chowdhury et al., 1994). Activationof nicotimic receptors by neurally released acety!cho-line causes immediate catecho!amine secretion via de-polarization-induced Ca2~influx. An alternate sourceof Ca2~to maintain the secretory response if nicotinicreceptors or the external Ca2~source become compro-mised is available through activation of muscarinicreceptors and mobilization of Ca2~from internalstores, thus assuring a maintained source of catechola-mines in the bloodstream during long-term activationof the adrenal gland.

REFERENCES

Akaike A., Mine Y., Sasa M., and Takaori S. (1990a) Voltage andcurrent clamp studies of muscarinic and nicotinic excitation ofthe rat adrenal chromaffin cells. J. Pharmacol. Exp. Ther. 255,333—339.

Akaike A., Mine Y., Sasa M., and Takaori S. (199Db) A patchclamp study of muscarinic excitation of the rat adrenal chromaf-fin cells. J. Pharmacol. Exp. Ther. 255, 340—345.

Ballesta J. J., Borges R., Garcia A. G., and Hidalgo M. J. (1989)Secretory and radioligand binding studies on muscarinic recep-tors in bovine and feline chromaffin cells. J. Physiol. (Lond.)418, 411—426.

Borges R., Ballesta J. J., and Garcia A. G. (1987) M2 muscarinic-associated ionophore at the cat adrenal. Biochem. Biophys. Res.Commun. 144, 965—972.

Bunn S. J. and Marley P. D. (1989) Effects of angiotensin II on

cultured bovine adrenal medullary cells. Neuropeptides 13,121— 132.

Burgoyne R. D. (1991) Control of exocytosis in adrenal chromaffincells. Biochim. Biophys. Acta 1071, 174—202.

Cheek T. R. (1989) Spatial aspects of calcium signalling. J. CellSci. 93, 211—216.

Cheek T. R. and Burgoyne R. D. (1985) Effect of activation ofmuscarinic receptors on intracellular free calcium and secretionin bovine adrenal chromaffin cells. Biochim. Biophys. Ada 846,167—173.

Cheek T. R., Jackson T. R., O’Sullivan A. J., Moreton R. B., Ber-ridge M. J., and Burgoyne R. D. (1989) Simultaneous measure-ments ofcytosolic calcium and secretion in single bovine adre-nal chromaffin cells by fluorescent imaging of fura-2 in cocul-tured cells. J. Cell Biol. 109, 1219—1227.

Chow R. H., von Rtiden L., and Neher E. (1992) Delay in vesiclefusion revealed by electrochemical monitoring of single secre-tory events in adrenal chromaffin cells. Nature 356, 60—63.

Chowdhury P. S., Guo X., Wakade T. D., Przywara D. A., andWakade A. R. (1994) Exocytosis from a single rat chromaffincell by cholinergic and peptidergic neurotransmitters. Neurosci-ence 59, 1—5.

Derome G., Tseng R., Mercier P., Lemaire 1., and Lemaire S. (1981)Possible muscarinic regulation of catecholamine secretion me-diated by cyclic GMP in isolated bovine adrenal chromaffincells. Biochem. Pharmacol. 30, 855—860.

Douglas W. W. and Poisner A. M. (1965) Preferential release ofadrenaline from the adrenal medulla by muscarine and pilocar-pine. Nature 208, 1102—1103.

Evinger M. J., Hemmick L. M., Requnathan S., Rels D. J., and RossM. E. (1992) Nicotine and muscarine stimulate expression ofPNMT promoter constructs transfected into primary chromaffincells. Soc. Neurosci. Abslr. 18, 588.

Fisher S. K., Holz R. W., and Agranoff B. W. (1981) Muscarinicreceptors in chromaffin cell cultures mediate enhanced phos-pholipid labelling but not catecholamine secretion. J. Neuro-chem. 37, 491—497.

Forsberg E. J. and Xu Y. (1993) Cation-selective channels activatedby muscarinic agonists in porcine adrenal chromaffin cells. Soc.Neurosci. Abstr. 19, 280.

Forsberg E. J., Rojas E., and Pollard H. B. (1986) Muscarinic recep-tor enhancement of nicotinic-induced catecholamine secretionmay be mediated by phosphoinositide metabolism in bovineadrenal chromaffin cells. J. Biol. Chem. 261, 4915—4920.

Grynkiewicz G., Poenie M., and Tsien R. Y. (1985) A new genera-tion of Ca2~indicators with greatly improved fluorescence prop-erties. .1. Biol. Chem. 260, 3440—3450.

lnoue M. and Kuriyama H. (1991) Muscarinic receptor is coupledwith a cation channel through a GTP-binding protein in guinea-pig chromaffin cells. J. Physiol. (Lond.) 436, 511—529.

Kim K. T. and Westhead E. W. (1989) Cellular responses to Ca2~from extracellular and intracellular sources are different asshown by simultaneous measurements of cytosolic Ca2~andsecretion from bovine chromaffin cells. Proc. Nati. Acad. Sci.USA 86, 9881—9885.

Kirpekar S. M., Prat J. C., and Schiavone M. T. (1982) Effect ofmuscarine on release of catecholamines from the perfused adre-nal gland of the cat. Br. J. Pharmacol. 77, 455—460.

Knight D. E. and Baker P. F. (1986) Observations on the muscarinicactivation of catecholamine secretion in the chicken adrenal.Neuroscience 19, 357—366.

Knight D. E. and Maconochie D. J. (1987) Muscarine induces aninward current in chicken chromaffin cells. J. Physiol. (Lond.)394, 147P.

Leszczyszyn D. J., Jankowski J. A., Viveros 0. 1-1., Diliberto E. 1.,Near D. J., and Wightman R. M. (1990) Nicotinic-receptormediated catecholamine secretion from individual chromaffincells: chemical evidence for exocytosis. J. Biol. Chem. 265,14736—14737.

Lillien L. F. and Claude p. (1985) Nerve growth factor is a mitogenfor cultured chromaffin cells. Nature 317, 623—634.

J. Neurochem., Vol. 67, No. 1, 1996

162 X. GUO ET AL.

Malgaroli A., Fesce R., and Meldolesi J. (1990) Spontaneous[Ca

2], fluctuations in rat chromaffin cells do not require inosi-tol 1,4,5-trisphosphate elevations but are generated by a caf-feine- and ryanodine-sensitive intracellular Ca2~store. J. Biol.Chem. 265, 3005—3008.

Malhotra R. K., Wakade T. D., and Wakade A. R. (1988a) Compari-son of secretion of catecholamines from the rat adrenal medulladuring continuous exposure to nicotine, muscarine or excess K.Neuroscience 26, 313—320.

Malhotra R. K., Wakade T. D., and Wakade A. R. (19886) Vaso-active intestinal polypeptide and muscarine mobilize intracellu-lar Ca2~through breakdown of phosphoinositides to inducecatecholamine secretion. .1. Biol. Chem. 263, 2123—2126.

Malhotra R. K., Wakade T. D., and Wakade A. R. (1989) Cross-communication between acetylcholine and VIP in controllingcatecholamine secretion by affecting cAMP, inositol triphos-phate, protein kinase C and calcium in rat adrenal medulla. J.Neurosci. 9, 4150—4157.

Marley P. D. (1987) New insights into the non-nicotinic regulationof adrenal medullary function. Trends Pharmacol. Sd. 8, 411—413.

Nakazato Y., Ohga A., Oleshansky M., Tomita U., and Yamada Y.(1988) Voltage-independent catecholamine release mediated bythe activation of muscarinic receptors in guinea-pig adrenalglands. Br. J. Pharmacol. 93, 101—109.

Neely A. and Lingle C. J. (1992) Effects of muscarine on single ratadrenal chromaffin cells. J. Physiol. (Land.) 453, 133—166.

O’Sullivan A. J., Cheek T. R., Moreton R. B., Berridge M. J., andBurgoyne R. D. (1989) Localization and heterogeneity of ago-nist-induced changes in cytosolic calcium concentration in sin-gle bovine adrenal chromaffin cells from video imaging of fura-2. EMBO J. 8,401—411.

Przywara D. A., Bhave S. V., Bhave A., Wakade T. D., and WakadeA. R. (1991) Dissociation between intracellular Ca2~and mod-ulation of [3H]noradrenaline release in chick sympathetic neu-rons. J. Physiol. (Land.) 437, 201—220.

Przywara D. A., Chowdhury P. S., Bhave S. V., Wakade T. D., andWakade A. R. (1993) Barium-induced exocytosis is due to

internal calcium release and block ofcalcium efflux. Proc. Nail.Acad. Sci. USA 90, 557—561.

Putney J. W. J. (1990a) Receptor-regulated calcium entry. Pharma-col. Ther. 48, 427—434.

Putney J. W. J. (l990b) Capacitative calcium entry revisited. CellCalcium 11, 611—624.

Tischler A. S., Perlman R. L., Nunnemacher G., Morse G. M.,DeLellis R. A., Wolfe H. J., and Sheard B. E. (1982) Long-termeffects of dexamethasone and nerve growth factor on adrenalmedullary cells cultured from young adult rats. Cell Tissue Res.255, 525—542.

Unsicker K., Krisch B., Otten U., and Thoenen H. (1978) Nervegrowth factor-induced fiber outgrowth from isolated rat adrenalchromaffin cells: impairment by glucocorticoids. Proc. NatI.Acad. Sd. USA 75, 3498—3502.

von RDden L. and Neher E. (1993) A Ca-dependent early step inthe release of catecholamines from adrenal chromaffin cells.Science 262, 1061—1065.

Wakade A. R. and Wakade T. D. (1983) Contribution of nicotinicand muscarinic receptors in the secretion of catecholaminesevoked by endogenous and exogenous acetylcholine. Neurosci-

ence 10, 973—978.Wakade T. D., Bhave S. V., Bhave A., Przywara D. A., and Wakade

A. R. (1990) Ca2 mobilized by caffeine from the inositol1,4,5-trisphosphate-insensitive pooi of Ca2~in somatic regionsof sympathetic neurons does not evoke [3H]norepinephrine re-lease. J. Neurochem. 55, 1806—1809.

Warashina A., Fujiwara N., and Shimoji K. (1989) Characteristics ofnicotinic and muscarinic secretory responses in the rat adrenalmedulla studied by real-time monitoring of catecholamine re-lease. Biomed. Res. 10, 157—164.

Wightman R. M., Jankowski J. A., Kennedy R. T., Kawagoe K. T.,Schroeder T. J., Leszczyszyn D. J., Near J. A., Diliberto E. J.,and Viveros 0. H. (1991) Temporally resolved catecholaminespikes correspond to single vesicle releasefrom individual chro-maffin cells. Proc. Nati. Acad. Sci. USA 88, 10754—10758.

Yamagami K., Nishimura S., and Sorimachi M. (1991) InternalCa2~mobilization by muscarinic stimulation increases secretionfrom adrenal chromaffin cells only in the presence of Ca2~influx. J. Neurochem. 57, 1681—1689.

J. Neurochem., Vol. 67, No. 1, 1996