Exocytotic catecholamine release is not associated with cation flux through channels in the vesicle membrane but Na+ influx through the fusion pore

Exocy o ic ca echolamine elease is no associa ed wi h ca ion lux h ough channels in he esicle memb ane bu Na+ in lux h ough he usion po e Liang-Wei Gong1,3, Guille mo Al a ez de Toledo2, and Man ed Lindau1,4 1School o Applied and Enginee ing Physics, Co nell Uni e si y, I haca, New Yo k 14850 2Depa men o Physiology and Biophysics, Facul y o Medicine, Uni e si y o Se ille, E-41009 Se ille, Spain Abs ac Release o cha ged neu o ansmi e molecules h ough a na ow usion po e equi es cha ge compensa ion by o he ions. I has been p oposed ha his may occu by ion low om he cy osol h ough channels in he esicle memb ane, which would gene a e a ne ou wa d cu en . We es ed his hypo hesis in ch oma in cells using cell-a ached pa ch ampe ome y measu ing simul aneously ca echolamine elease om single esicles and ionic cu en ac oss he pa ch memb ane. No de ec able cu en was associa ed wi h ca echolamine elease indica ing ha <2% (i any) o ca ions en e he esicle h ough i s memb ane. Ins ead we show ha lux o ca echolamines h ough he usion po e, measu ed as an ampe ome ic oo signal, dec eases when he ex acellula ca ion concen a ion is educed. The esul s e eal ha he a e o ansmi e elease h ough he usion po e is coupled o ne Na+ in lux h ough he usion po e as p edic ed by elec odi usion heo y applied o usion po e pe mea ion and sugges a p e usion a he han pos usion ole o esicula ca ion channels. Neu o ansmi e and ho mone elease occu s by exocy osis, which begins wi h he o ma ion o a na ow usion po e1,2. Fusion po es in mas cells and ch oma in cells ha e ypically an ini ial conduc ance o ~330 pS1,3 h ough which esicula se o onin and ca echolamines a e eleased, espec i ely4,5. The ini ial usion po e o a synap ic esicle is ypically >280 pS and elease o neu o ansmi e should occu wi h a ime cons an <500 μs due o he small esicle olume6. Many neu o ansmi e s and ho mones a e o ganic ions ha ca y cha ge wi h hem. Among hese ace ylcholine, se o onin and ca echolamines a e mos ly mono alen ca ions a physiological o mo e acidic esicula pH. E lux o cha ged molecules h ough a na ow usion po e would apidly cha ge he small capaci ance o he esicle and a mechanism o cha ge compensa ion is equi ed. Release o ca echolamines om a single esicle can be de ec ed elec ochemically7. The ini ial lux o ca echolamines h ough he ea ly na ow usion po e can be measu ed as an ampe ome ic oo signal p eceding he ampe ome ic spike3,5,8. The ampli ude o ampe ome ic oo cu en s is ypically ~5 pA. Since mos ca echolamine molecules will be in mono alen ca ionic o m bu wo elec ons a e ans e ed pe molecule in he oxida ion gi ing ise o he ampe ome ic cu en 9, he ca echolamines eleased du ing he oo signal ca y an ionic cu en o ~2.5 pA wi h hem. This would cha ge a ypical bo ine ch oma in g anule 4co espondence should be add essed o M.L. ([email p o ec ed]). 3P esen add ess: Depa men o Cell Biology, Howa d Hughes Medical Ins i u e, Yale Uni e si y School o Medicine, 295 Cong ess A enue, New Ha en, CT 06510 NIH Public Access Au ho Manusc ip Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. Published in inal edi ed o m as: Na Cell Biol. 2007 Augus ; 9(8): 915–922. doi:10.1038/ncb1617. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip wi h 2.5 F capaci ance by 1 V/ms. Synap ic esicles wi h ypical ~70 aF capaci ance6 con ain se e al housand ansmi e molecules pe esicle and elease o only 450 mono alen molecules would cha ge he esicle o >1 V. The e o e, a mechanism is necessa y o compensa e he cha ge mo emen associa ed wi h he lux o cha ged ansmi e o allow o apid and con inued ansmi e elease h ough he usion po e. E idence o he p esence o K+ pe meable channels in he memb ane o synap ic esicles and o he neu osec e o y esicles including ch oma in g anules10-15 led o he hypo hesis ha du ing elease en y o ca ions may occu h ough ca ion channels in he esicle memb ane16. I his we e he dominan mechanism media ing ion exchange, he ca echolamine molecules eleased h ough he usion po e would be eplaced by ca ions om he cy osol, gi ing ise o a ne ou wa d cu en . To es his hypo hesis we pe o med pa ch ampe ome y expe imen s on ch oma in cells o simul aneous de e mina ion o ca echolamine lux, usion po e conduc ance and ne memb ane cu en . RESULTS Fusion po e conduc ance de e mines lux o ca echolamines du ing oo signal Fig. 1 shows a cell a ached pa ch ampe ome y eco ding o an exocy o ic e en wi h an unusually long ampe ome ic oo signal (Fig. 1a). The ca echolamine concen a ion o his esicle was ~0.8 M and ~40% o he esicle con en s we e eleased du ing he oo signal. The pa ch admi ance measu emen shows luc ua ing alues in he eal pa (Fig. 1b) and he imagina y pa (Fig. 1c). F om hese aces he ime cou se o usion po e conduc ance was calcula ed (Fig. 1d, solid line). The luc ua ions in usion po e conduc ance obse ed in his ace a e accompanied by co esponding luc ua ions in he ampe ome ic cu en (Fig. 1d, do ed line) and hus in he lux o ca echolamine molecules. The co ela ions a e well esol ed because in his expe imen he ca bon ib e elec ode (CFE) was e y close o he plasma memb ane pa ch minimizing he ime o di usion om he elease si e o he CFE. E en ually he usion po e expanded (a ow), which was accompanied by he onse o he ampe ome ic spike indica ing apid elease o he emaining ca echolamine molecules om he esicle. Fig. 1e shows he a io o ca echolamine lux/ usion po e conduc ance ( lux/conduc ance a io) du ing he oo signal, gi ing a a he cons an alue o ~3×107 molecules pe second pe nS o usion po e conduc ance (molecules s−1nS−1) o he pa indica ed by he ho izon al line, in ag eemen wi h p e ious es ima es5. A la e imes he lux/conduc ance a io dec eases somewha , p esumably due o deple ion o ee ca echolamine in he esicle. Fig. 1 shows he da a poin s o he pa indica ed by he ho izon al line in Fig. 1e plo ed as lux s. usion po e conduc ance and e eals a p opo ional ela ionship. I shows ha he na ow usion po e is he di usion ba ie and ha i s size luc ua ions limi he lux o ca echolamine du ing he ampe ome ic oo signal. Fig. 1g shows he usion po e opening and oo signal o a esicle wi h simila ca echolamine concen a ion. The lux o ca echolamine again shows a ime cou se ha pa allels ha o he usion po e conduc ance. The lux/conduc ance a io (Fig. 1h) is simila (3.1×107 molecules s−1nS−1) and is a he cons an o he du a ion o he oo signal. Due o he sho e oo du a ion his esicle eleases only ~25% o i s con en s du ing he oo signal. Fig. 1i shows he da a o a esicle wi h a e y low ca echolamine concen a ion o ~0.26 M. This esicle had an a e age usion po e conduc ance o ~270 pS wi h a li e ime o 550 ms. Du ing he i s 150 ms a e ini ial opening o he usion po e he lux o ca echolamine again pa allels he ime cou se o usion po e conduc ance. The lux/conduc ance a io (Fig. 1j) du ing his ime was lowe (1.4×107 molecules s−1nS−1), consis en wi h he lowe esicula ca echolamine concen a ion. F om his esicle o e 70% o he con en s we e eleased du ing he oo signal and i can be seen ha a dec ease in lux due o ca echolamine deple ion al eady se s in a e 150 ms. The ac ha such deple ion is much less e iden in esicles wi h highe ca echolamine Gong e al. Page 2 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip concen a ion indica es ha he ee ca echolamine concen a ion is bu e ed, p esumably by binding si es on he g anula ma ix. The mean lux/conduc ance a io o all e en s ha we e eco ded wi h a CFE close enough o he pa ch memb ane ha he lux was accu a ely e lec ed in he ampe ome ic oo signal was 2.2±1.2×107 molecules s−1nS−1 (SD, n=13). Since he ionic ca echolamine cu en is hal he ampe ome ic cu en , an ampe ome ic cu en o ~1 pA, as in Fig. 1d, co esponds o an ionic ca echolamine cu en o ~0.5 pA h ough he usion po e, which would cha ge he 3 F esicle a a a e o ~170 mV/ms, and which should apidly educe he lux o molecules. Howe e , he lux is main ained wi h a a he cons an lux/conduc ance a io o hund eds o milliseconds indica ing ha a mechanism o cha ge compensa ion is p esen . Ca echolamine elease is no associa ed wi h ne ou wa d cu en I a signi ican ac ion o ca ions en e s he esicle ia channels in he esicle memb ane (Fig. 2a), hen he lux o ca echolamine h ough he usion po e would be associa ed wi h a ne ou wa d cu en . Fig. 2b shows an exocy o ic e en measu ed by pa ch ampe ome y. I he ionic cha ges o he eleased ca echolamines we e o be compensa ed by ca ions en e ing h ough ion channels in he esicle memb ane, he ne ou wa d cu en h ough he pa ch memb ane would be expec ed o be hal o he ampe ome ic cu en . In con as o his expec a ion, no pa ch ou wa d cu en associa ed wi h he ampe ome ic spike o ampe ome ic oo signal is de ec able in Fig. 2b (bo om ace). No e ha he ampe ome ic and pa ch cu en s a e scaled such ha he expec ed pa ch cu en would appea on he g aphs wi h he same size as he ampe ome ic cu en . The e en o Fig. 2c shows a simila eco ding o an expanding usion po e bu he oo cu en was a he small. To imp o e he esolu ion we a e aged 11 e en s wi h mean oo ampli ude >1 pA and oo du a ions >20 ms. All e en s we e aligned a he ime o ull po e expansion (onse o he spike). E en his a e aged eco d (Fig. 2d) showed no de ec able pa ch cu en associa ed wi h oo cu en . The pa ch cu en ace ha would be expec ed i compensa o y ca ion lux h ough he esicula memb ane occu ed, would ac ually be equal o he scaled ampe ome ic cu en (Fig.2d, do ed signal). Howe e , since he ampe ome ic oo and spike ampli udes a e educed due o di usional b oadening, he a e age pa ch cu en should be e en la ge han he do ed cu en signal indica es. I ion lux h ough he memb ane would only compensa e he cha ge while he usion po e is small, hen he pa ch cu en du ing he ampe ome ic oo would be equal o g ea e han he oo cu en and e mina e wi h he onse o he spike (a ow, g ey do ed line). To es ima e an uppe limi o ca ion lux h ough he esicle memb ane du ing he oo signal we de e mined he mean pa ch cu en du ing he pla eau phase o he a e aged ampe ome ic oo (be ween e ical dashed lines in Fig. 2c) a e i ing he a e aged pa ch cu en p eceding he onse o he oo wi h a s aigh line and sub ac ing his i as he baseline. The mean pa ch cu en du ing his ime was 0.004±0.013 pA (s.e.m.) and hus <0.03 pA (wi hin 2 s.e.m.). The mean pla eau oo cu en was 2.9 pA, which co esponds o a ca ionic ca echolamine cu en o 1.45 pA. The ne ou wa d cu en is hus <2% o he ca ionic ca echolamine cu en indica ing ha <2% o he ca echolamine cha ge (i any) is eplaced by ca ions en e ing he esicle h ough channels in he esicle memb ane. Fusion po e conduc ance depends on ex acellula ion concen a ion Ha ing excluded he en y o ca ions h ough he esicle memb ane, cha ge compensa ion mus occu ia he usion po e. Ei he ca echolamine elease du ing he oo signal is accompanied by co- elease o anions o by en y o ca ions om he ex acellula space h ough he usion po e. To dis inguish be ween hese wo possibili ies we examined he e ec s o ex acellula Gong e al. Page 3 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip ca ion concen a ion on elease om single esicles. Fi s we analyzed he in luence o changing ex acellula ions on usion po e p ope ies by pa ch ampe ome y eco dings a ying he pipe e solu ion (Fig.3). The mean capaci ance s ep size and mean ampe ome ically measu ed cha ge (quan al size) we e no signi ican ly di e en in he di e en solu ions. Indi idual usion po e openings we e quan i ied by de e mining ini ial and mean usion po e conduc ance and usion po e du a ion3 (Fig. 3a,b). The mean usion po e conduc ance du ing he oo signals was 740±68 pS in solu ion A and was educed o 515±32 pS in solu ion B and o 326±20 pS in solu ion C (Fig. 3c) (see Me hods o composi ions o solu ions A, B, C). A co esponding educ ion was also ob ained o he ini ial usion po e conduc ance (solu ion A: 383±27 pS, solu ion B: 298±29 pS, solu ion C: 214±16 pS) (Fig. 3d). The usion po e du a ion om he ini ial opening o he usion po e o a apid inc ease o a conduc ance >1 nS was ~26ms, in ag eemen wi h p e ious esul s3, and was no signi ican ly di e en in he di e en solu ions (Fig. 3e). Fo an aqueous usion po e, i s conduc ance depends on he conduc i i y o he solu ion ha ills he po e. The conduc i i ies o solu ions A, B, and C we e measu ed o be 18, 8.7, and 4.2 mScm−1, espec i ely. While hese conduc i i ies a y app oxima ely as 4:2:1, he ini ial and mean usion po e conduc ances a y app oxima ely as 4:3:2. The usion po e conduc ances hus a y less han he conduc i i ies o he ex acellula solu ions, which is expec ed because he ionic composi ion in he po e will be a mix u e o ex acellula and in a esicula ions and only he ex a esicula concen a ions we e a ied. Assuming ha he usion po e con ains a 1:1 mix u e o ex acellula and in a esicula solu ion, he esul would be consis en wi h a conduc i i y o ~11-12 mScm−1 o he in a esicula solu ion and unchanged usion po e dimensions. The usion po e will hus p esen a simila di usion ba ie o ca echolamines and o he ions in he di e en ex acellula solu ions. Flux o ca echolamines h ough he usion po e depends on ex acellula ca ions I he cha ge compensa ion occu s by co- elease o anions (such as ATP), he lux o ca echolamines h ough he usion po e should be simila in he di e en solu ions. I , howe e , he cha ge compensa ion occu s by en y o ca ions om he ex acellula solu ion h ough he usion po e, he lux o ca echolamines should be educed a lowe ex acellula ca ion concen a ions. We he e o e de e mined how he ampe ome ic oo signals a e a ec ed by changes in ex acellula ca ion concen a ion. To minimize di usional dis o ion we measu ed he oo cu en s using con en ional ca bon ib e ampe ome y (Fig. 4a) whe e he memb ane-CFE dis ance is gene ally e y small. We used ou di e en solu ions con aining (in mM) 140, 90, 40, and 0 NaCl. All solu ions con ained in addi ion 5 KCl, 5 CaCl2, 1 MgCl2 and 10 HEPES/ NaOH. Osmolali y was main ained nea physiological alues by adding app op ia e amoun s o glucose. The mean quan al size measu ed as cha ge o in eg a ed ampe ome ic spikes, he mean ampe ome ic spike ampli ude and he mean hal -wid h we e no signi ican ly a ec ed by changing he solu ions (da a no shown). Wi h 140 and 90 mM NaCl oo signals we e de ec able in ~72% o ampe ome ic e en s, dec easing o 66% in 40 mM NaCl and 44% in 0 NaCl (Fig. 4b). The mean cha ge o oo signals, i.e. he a ea unde he oo , was educed p og essi ely wi h educing he NaCl concen a ion ([NaCl]) in he ba h solu ion (Fig. 4c). The mean oo du a ion was no signi ican ly di e en in he di e en solu ions (Fig. 4d) and he educed oo cha ge e lec s p ima ily a dec ease in oo cu en ampli ude (Fig. 4e). I an ampe ome ic spike has no de ec able oo signal, i is no a p io i clea i he eason is oo sho du a ion o oo small ampli ude. The e o e, unde ec ed oo signals we e no included in he s a is ical analysis o mean oo signal p ope ies (Fig. 4c-e). Fo he mean ampe ome ic oo cu en , his would be co ec i he unde ec ed oo signals we e o oo sho du a ion. Since he mean usion po e du a ion is mo e han 10 imes longe han ou de ec ion limi o 3 Gong e al. Page 4 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip ms, i is, howe e , likely ha he unde ec ed oo signals had oo small ampli ude. The de ec ion limi s o oo signal was an ampli ude ≥1 pA and unde ec ed oo cu en s should be be ween 0 and 1 pA. We hus es ima ed he mean oo cu en s o he di e en [NaCl] assuming on a e age 0.5 pA o unde ec ed oo signals (Fig. 4 ). This analysis shows a u he educ ion a low [NaCl]. As p e iously epo ed, oo only e en s, also e med s and-alone oo signals17, we e also obse ed in ou eco dings. The pe cen age o such e en s ha did no lead o ull usion was 11.6% o 140 NaCl, 8.9% o 90 NaCl, 7% o 40 NaCl and 6.8% o 0 NaCl. These equencies a e simila o hose p e iously epo ed17. The mean ampe ome ic cu en ampli udes o hese e en s showed a simila dec ease wi h dec easing [NaCl] om 4.9 +/− 0.4 pA (n=33) in 140 NaCl solu ion o 2.1+/−0.2 pA (n=31) in 0 NaCl solu ion. The s and-alone oo signals had a smalle mean ampli ude han he p e-spike oo signals sugges ing a smalle usion po e. Howe e , we did no include hese e en s in he analysis because he de e mina ion o usion po e conduc ance o such e en s is o en unce ain because he lack o he inal capaci ance s ep lea es some unce ain y ega ding he phase se ing o he lock-in admi ance measu emen . E lux o ca echolamines h ough he usion po e occu s by elec odi usion The dec ease in ca echolamine e lux h ough he na ow usion po e when he ion concen a ion is educed indica es ha he ca ion en y a e ia he usion po e limi s he e lux o ca echolamine elease du ing he oo signal. Fo a quan i a i e es ima e we applied elec odi usion heo y using he Ne ns -Planck equa ion assuming a cons an ield ac oss he usion po e as in he Goldman-Hodgkin-Ka z cons an ield heo y (see Me hods). We assumed ha o pe mean ions, he pe meabili ies o he usion po e a e p opo ional o he di usion coe icien s o he espec i e ions in aqueous solu ion (see supplemen a y ig.1 and supplemen a y able 1). To es ima e he dependence o ca echolamine lux h ough he usion po e on he ex acellula [NaCl], we es ed ou di e en models: 1. Fusion po e selec i e o mono alen ca ions 2. Ca ion selec i e usion po e (allowing pe mea ion o mono- and di alen ca ions) 3. Non-selec i e usion po e, negligible in a esicula ee anion concen a ion 4. Non-selec i e usion po e, in a esicula ee anion concen a ion equi alen o 50 mM Cl−. The in a esicula ee ca echolamine concen a ion was he only ee pa ame e and was adjus ed o ep oduce he mean ampe ome ic oo cu en ampli ude measu ed in he p esence o 140 mM NaCl (see legend o Fig. 4 e, and supplemen a y ig. 1). Depending on he ex acellula [NaCl] he e e sal po en ial o he usion po e changes, hus changing he lux o ca echolamines h ough he usion po e (see supplemen a y ig. 1). Al hough an ini ial po en ial may be p esen ac oss he esicle memb ane be o e usion1, he e e sal po en ial o he usion po e will be a ained wi h he ime cons an gi en by ch oma in g anule capaci ance and usion po e conduc ance and is es ima ed o be abou 3 F/300 pS=10 μs. The expec ed changes in oo cu en wi h changing [NaCl] a e compa ed wi h he expe imen al da a o he ou di e en usion po e models in Figs. 4 e, . The mos ob ious esul is ha model 4, allowing signi ican co- elease o anions, shows a e y weak dependence o he oo signal on he ex acellula [NaCl] and is no consis en wi h he measu emen s. I can hus be concluded ha cha ge compensa ion occu s by en y o Na+ ions h ough he usion po e (Fig. 4g). Ca ion selec i e usion po e p ope ies (models 1 and 2) appea o i he da a somewha be e han he non-selec i e po e wi h negligible ee anion concen a ion in he esicle bu he di e ences a e ela i ely small. Gong e al. Page 5 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip DISCUSSION Release o cha ged ansmi e molecules h ough exocy o ic usion po es equi es a mechanism ha compensa es o he cha ge mo emen associa ed wi h he lux o hese molecules h ough he usion po e. We ha e shown he e ha en y o ca ions in o he esicle ia ca ion channels in he esicle memb ane is no a majo mechanism o ion exchange du ing elease. Ins ead, ca echolamine lux h ough he usion po e is associa ed wi h Na+ in lux h ough he usion po e. A mechanism ha in ol es cha ge compensa ion by co- elease o anions is inconsis en wi h he expe imen al esul s. In con as o he ampe ome ic oo cu en , he ampe ome ic spike kine ics a e no signi ican ly a ec ed by changing he ex acellula [NaCl], indica ing ha elease du ing he ampe ome ic spike is no limi ed by he usion po e as a di usion ba ie and no limi ed by he a e o ion exchange. In mas cells he usion po e is also no limi ing he a e o elease du ing his phase18. The elease kine ics o se o onin om mas cell g anules ollowing elec opo a ion depends on he ype o ca ion in he solu ion bu no on i s concen a ion o e a wide ange (1-100 mM), consis en wi h slow di usion wi hin he g anula ma ix and an appa en in a-ma ix di usion coe icien ha depends on he ca ion species bu no i s concen a ion19. I is possible ha ion exchange media ing dissocia ion om he in a esicula ma ix in ch oma in g anules may in ol e Ca2+ ions, which could no be omi ed in ou expe imen s because hey a e essen ial o s imula e exocy osis. Expe imen s wi h elec ode a ays e ealed a educed di usion coe icien o ca echolamines nea he cell su ace20 sugges ing ha he ampe ome ic spike kine ics may also be limi ed by slow di usion nea he cell memb ane. Release o ca echolamines h ough he usion po e is no associa ed wi h cha ge compensa ion by co- elease o anions. This implies ha he usion po e is ei he ca ion selec i e o he ee concen a ion o he majo g anula anion ATP is e y low. I he usion po e o ch oma in g anules we e ca ion selec i e one would ha e o pos ula e ha he usion po e o glu ama e gic synap ic esicles should ha e e y di e en p ope ies since a ca ion selec i e usion po e would no allow e ec i e elease o anionic glu ama e. Since he same se o SNARE p o eins media es exocy osis in neu ons and ch oma in cells21, a ca ionic usion po e in ch oma in cells would be inconsis en wi h a common usion po e lined by SNARE p o ein domains as ecen ly p oposed22. Howe e , ou analysis assumes ha he pe meabili ies o di e en ions h ough he usion po e a e p opo ional o hei di usion coe icien s, which is a ough es ima e. In pa icula wi h complex ionic mix u es pe meabili ies may a y depending on ionic composi ion. The e o e, a p ecise ag eemen should no necessa ily be expec ed and a usion po e pe meable o anions and negligible ee anion concen a ion inside he ch oma in g anule canno be excluded. ATP in ch oma in g anules has a s uc u e simila o c ys alline ATP, no o ATP in solu ion23, sugges ing a low ee ATP concen a ion. The usion po e models indica e in a esicula ee ca echolamine concen a ions o 130-370 mM, much lowe han he o al esicula ca echolamine concen a ion24, which is consis en wi h binding o he g anula ma ix25,26 o al e na i ely ca echolamines and ATP o ming a highly non-ideal solu ion27. Expe imen s employing solu ions whe e Cl− ions a e subs i u ed by la ge anions, wi h much lowe di usion coe icien s o e en oo la ge o pe mea e he usion po e should allow a be e es ima e o he usion po e pe meabili y o anions. Ca echolamine elease h ough na ow usion po es is signi ican . Ampe ome ic eco dings indica e ha ~10% o ca echolamine elease occu s h ough na ow usion po es and unde basal s imula ion ca echolamine elease may p edominan ly occu h ough na ow usion po es e aining he p o ein con en s o he esicle28. Unde hese condi ions he a e o elease om he esicle is he e o e de e mined by elec odi usion. Fas ansmi e elease om small synap ic esicles is belie ed o occu en i ely h ough a na ow usion po e. I has been epo ed Gong e al. Page 6 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip ha in To pedo synap ic esicles mos o he ace ylcholine (ACh) o ATP con en is adso bed o an in a esicula ma ix co e29 analogous o he si ua ion in ch oma in cells. The hypo hesis ha cha ge compensa ion in hese esicles occu s ia ca ion channels in he esicle memb ane16 has been ein o ced by he ecen inding ha unc ional TRPM7 channels in choline gic esicles play a c i ical ole in ansmi e elease15,30. The esul s p esen ed he e o ca echolamine elease indica e ha pos usion ca ion cu en s h ough he esicula memb ane a e no de ec able and a e no equi ed o sus ain lux o ansmi e h ough he usion po e. I his is also he case o choline gic esicles i would sugges ha he TRPM7 modula ion o quan al size may be a p e usion mechanism a ec ing loading o he esicle15, 30 a he han a pos usion ole in media ing ion exchange. METHODS Cells and Solu ions Bo ine ch oma in cells we e p epa ed and cul u ed as desc ibed31. Reco dings we e made on days 1–7 in cul u e. Fo pa ch ampe ome y32, he ba h solu ion con ained (in mM) 140 NaCl, 5 KCl, 5 CaCl2, 1 MgCl2, 10 HEPES/NaOH, 10 glucose, and pH was adjus ed o 7.3 wi h NaOH and he osmolali y was a ound 310 mmol/kg. Pipe e solu ion A con ained (in mM) 50 NaCl, 100 TEACl, 5 KCl, 10 CaCl2, 1 MgCl2, 10 HEPES/NaOH, he pH was adjus ed o 7.3 and he osmolali y was adjus ed o ~290 mmol/kg. In pa ch ampe ome y expe imen s o s udy usion po e conduc ance wi h low ca ion concen a ion in he pipe e, pipe e solu ions con ained (in mM): (solu ion B) 0 NaCl, 50 TEACl, 5 KCl, 10 CaCl2, 1 MgCl2, 10 HEPES/ NaOH pH 7.3, and (solu ion C) 0 NaCl, 0.5 TEACl, 5 KCl, 10 CaCl2, 1 MgCl2, 10 HEPES/ NaOH pH 7.3, osmola i ies we e adjus ed o ~310 mmol/kg wi h glucose. Fo ampe ome y, he ba h solu ions con ained (in mM): (140 NaCl solu ion) 140 NaCl, 5 KCl, 5 CaCl2, 1 MgCl2, 10 HEPES/NaOH, 10 glucose; (90 NaCl solu ion) 90 NaCl, 5 KCl, 5 CaCl2, 1 MgCl2, 10 HEPES/NaOH; (40 NaCl solu ion) 40 NaCl, 5 KCl, 5 CaCl2, 1 MgCl2, 10 HEPES/NaOH; o (0 NaCl solu ion) 5 KCl, 5 CaCl2, 1 MgCl2, 10 HEPES/NaOH; he pH was adjus ed o 7.3 and he osmola i ies we e adjus ed o ~290 mmol/kg wi h glucose. All he expe imen s we e pe o med a oom empe a u e. Pa ch ampe ome y Changes in memb ane capaci ance and ca echolamine elease we e eco ded simul aneously by he cell-a ached pa ch ampe ome y32. B ie ly, cell-a ached pa ch clamp was pe o med wi h a CFE in oduced in o he pa ch pipe e. The CFE was posi ioned a a dis ance 1-5 μm om he ip opening unde he mic oscope. The CFE was con inuously held a +700 mV. Ampe ome ic cu en s we e il e ed wi h an 8-pole Bessel il e se a 3 kHz. Pa ch pipe es we e pulled in ou s ages wi h a p og ammable pulle (P-97; Su e Ins umen s) and coa ed wi h a s icky wax (Ke ). Pipe es we e i e-polished and had a ypical esis ance in he ba h o a ound 2 MΩ. Pipe e esis ance ypically inc eased up o 3-4 MΩ when he CFE app oached he ip opening. Fo capaci ance measu emen s, we used a pa ch clamp ampli ie (EPC-7; HEKA-Elek onik). Command ol age was applied o he ba h. Changes o pa ch admi ance we e measu ed as desc ibed p e iously33,34 wi h a lock-in ampli ie (SR830; S an o d Resea ch Sys ems) using a sine wa e ampli ude o 50 mV ( oo mean squa e) a a equency o 20 kHz. The ou pu il e was se o a 1ms ime cons an , 24 db. The pa ch cu en is domina ed by he capaci i e cu en s ha a ise om he 20 kHz sine wa e ol age applied o measu e changes in memb ane capaci ance. Howe e , he 20 kHz cu en was il e ed ou by a 3 kHz low-pass il e such ha pa ch cu en s below his equency can be well esol ed du ing he capaci ance measu emen . Fusion po e openings we e analyzed as desc ibed p e iously3,32,34. Po e conduc ance (GP) Gong e al. Page 7 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip and esicle capaci ance we e calcula ed om he eal (Re) and imagina y (Im) pa s o he admi ance a e baseline sub ac ion: GP = (Re2 + Im2)/Re and CV = [(Re2 + Im2)/Im]/ω. The ini ial ise o GP e lec s he s ep esponse o he low pass il e se ing o he lock-in ampli ie (1 ms, 24 dB). A his se ing 90% o he inal alue is eached wi hin 7 ms. The ini ial GP alue was hus aken as he GP alue a 7-10 ms ( ime o ini ial GP) a e he onse o he inc ease in he GP ace. The usion po e du a ion was de ined as he ime in e al be ween ime o ini ial GP and he ime whe e GP inc eased beyond 1 nS. The mean GP was aken as he a e age GP alue du ing his ime in e al. To de e mine esicula ca echolamine concen a ions he numbe o molecules eleased om a gi en esicle was calcula ed om he in eg a ed ampe ome ic spike cha ge assuming a ans e o 2 elec ons pe molecule and he esicle olume was es ima ed om he capaci ance s ep size assuming sphe ical geome y and a speci ic memb ane capaci ance o 9 F cm−1 as desc ibed5. Ampe ome y Con en ional ampe ome y o ca echolamine de ec ion used 5 μm ca bon ib es (ALA Scien i ic Ins umen s). The ip o he elec ode was posi ioned closely agains he cell su ace o minimize he di usion dis ance om elease si es. The ampe ome ic cu en , gene a ed by oxida ion o ca echolamines a he exposed ip o he CFE, was measu ed using a homemade ampe ome ic ampli ie , ope a ed in he ol age-clamp mode a a holding po en ial o +700 mV. Ampe ome ic signals we e low-pass il e ed a 1 kHz and digi ized a 4 kHz. The da a we e collec ed, and hen analyzed by compu e using IGOR so wa e (Wa eMe ics, Lake Oswego, OR). Sec e ion was induced by p essu e ejec ion o 20 μM ionomycin (Sigma, S . Louis) dilu ed in he co esponding ba h solu ion om a ~ 5 μm ipped mic opipe e loca ed ~ 30 μm away. Ampe ome ic eco dings we e analyzed wi h a p og am w i en in his labo a o y o ex ac oo and spike in o ma ion acco ding o he c i e ia o Chow and Von Ruden35. The beginning o he cu en spike was loca ed whe e he leading edge o he ansien (which includes he “ oo ” signal when p esen ) exceeded he baseline cu en by wo imes he SD o he baseline noise le el. Foo and spikes we e selec ed o analysis only when he ampli ude o he spike was > 40 pA. De ec able oo signals had an ampli ude ≥1pA and a du a ion ≥3ms. Elec odi usion Calcula ions The cu en ca ied by ion S h ough a cylind ical usion po e o leng h l and c oss-sec ional a ea a is gi en by (1) whe e DS is he di usion coe icien o ion S, zS i s alency, [SV] i s in a esicula ee concen a ion and [SE] i s ex acellula ee concen a ion. V is he ol age p esen ac oss he usion po e and F,R,T ha e hei usual meaning. Fo solu ion A he conduc i i y o he solu ion illing he usion po e is app oxima ely σ=15 mScm−1. The usion po e conduc ance GP can be w i en as (2) and we can subs i u e he po e geome y in eq. 1 gi ing Gong e al. Page 8 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip (3), which is alid also o a bi a y po e geome ies36. We calcula e he quan i y GP/σ om solu ion A o be ~5×10−8 cm. The in a esicula concen a ions o ee Na+ and K+ a e bo h ~20 mM and ee Ca2+ and Mg2+ a e e y low37. The e ec i e ee concen a ion o ca echolamines and soluble anions a e no p ecisely known. S a is ical Analysis Da a is exp essed as Mean±SEM i no o he wise indica ed. Newman o one-way ANOVA es s was used o s a is ical analysis. Supplemen a y Ma e ial Re e o Web e sion on PubMed Cen al o supplemen a y ma e ial. Acknowledgmen s This wo k was suppo ed by Na ional Ins i u es o Heal h G an R01-NS38200 and he Nanobio echnology Cen e (a Na ional Science Founda ion Science and Technology Cen e , ag eemen No. ECS-9876771). We hank Joan Lenz o he cell p epa a ion and excellen echnical assis ance. REFERENCES 1. B ecken idge LJ, Alme s W. Cu en s h ough he usion po e ha o ms du ing exocy osis o a sec e o y esicle. Na u e 1987;328:814–817. [PubMed: 2442614] 2. Lindau M, Al a ez de Toledo G. The usion po e. Biochim. Biophys. Ac a 2003;1641:167–173. [PubMed: 12914957] 3. De nick G, Al a ez de Toledo G, Lindau M. Exocy osis o single ch oma in g anules in cell- ee inside-ou memb ane pa ches. Na . Cell Biol 2003;5:358–362. [PubMed: 12652310] 4. Al a ez de Toledo G, Fe nández-Chacón R, Fe nandez JM. Release o sec e o y p oduc s du ing ansien esicle usion. Na u e 1993;363:554–558. [PubMed: 8505984] 5. Albillos A, e al. The exocy o ic e en in ch oma in cells e ealed by pa ch ampe ome y. Na u e 1997;389:509–512. [PubMed: 9333242] 6. He L, Wu XS, Mohan R, Wu LG. Two modes o usion po e opening e ealed by cell-a ached eco dings a a synapse. Na u e 2006;444:102–105. [PubMed: 17065984] 7. Wigh man RM, e al. Tempo ally esol ed ca echolamine spikes co espond o single esicle elease om indi idual ch oma in cells. P oc. Na l. Acad. Sci. U. S. A 1991;88:10754–10758. [PubMed: 1961743] 8. Chow RH, Rüden L. . Nehe E. Delay in esicle usion e ealed by elec ochemical moni o ing o single sec e o y e en s in ad enal ch oma in cells. Na u e 1992;356:60–63. [PubMed: 1538782] 9. Bau JE, K is ensen EW, May LJ, Wiedemann DJ, Wigh man RJ. Fas -scan ol amme y o biogenic amines. Anal. Chem 1988;60:1268–1272. [PubMed: 3213946] 10. Rahamimo R, DeRieme SA, Sakmann B, S adle H, Yaki N. Ion channels in synap ic esicles om To pedo elec ic o gan. P oc. Na l. Acad. Sci. U. S. A 1988;85:5310–5314. [PubMed: 2455900] 11. Yaki N, Rahamimo R. The non-speci ic ion channel in To pedo ocella a used synap ic esicles. J Physiol 1995;485(P 3):683–697. [PubMed: 7562610] 12. Yin Y, Dayani hi G, Lemos JR. Ca(2+)- egula ed, neu osec e o y g anule channel in ol ed in elease om neu ohypophysial e minals. J Physiol 2002;539:409–418. [PubMed: 11882674] 13. A ispe N, Polla d HB, Rojas E. Calcium-independen K(+)-selec i e channel om ch oma in g anule memb anes. J. Memb . Biol 1992;130:191–202. [PubMed: 1283986] Gong e al. Page 9 Na Cell Biol. Au ho manusc ip ; a ailable in PMC 2010 May 17. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip