Controlling synaptotagmin activity by electrostatic screening

Transcription

1 Controlling synptotgmin tivity y eletrostti sreening Yongsoo Prk, Jvier M Hernnez, Geert vn en Bogrt, Shee Ahme, Mtthew Holt, Dietmr Rieel & Reinhr Jhn Exoytosis of neuroseretory vesiles is meite y the SNARE (solule N-ethylmleimie sensitive ftor tthment protein reeptor) proteins syntxin-, synptorevin n SNAP-5, with synptotgmin funtioning s the mjor C + sensor for triggering memrne fusion. Here we show tht ovine hromffin grnules reily fuse with lrge unilmellr liposomes in SNARE-epenent mnner. Fusion is enhne y C +, ut only when the trget liposomes ontin phosphtiylinositol-,5- isphosphte n when polyphosphte nions, suh s nuleoties or pyrophosphte, re present. C + -epenent enhnement is meite y enogenous synptotgmin-. Polyphosphtes operte y n eletrostti mehnism tht reverses n intivting is ssoition of synptotgmin- with its own memrne without ffeting trns ining. Hene, the lning of trns- n is-memrne intertions of synptotgmin- oul e ruil element in the pthwy of C + -epenent exoytosis. Neurons ommunite through exoytoti relese of neurotrnsmitters. In the resting stte, neurotrnsmitters re store in synpti vesiles or seretory grnules. Upon epolriztion, voltge-gte lium hnnels open, resulting in n influx of C + tht triggers fusion of the storge-vesile memrne with the plsm memrne, thus relesing the neurotrnsmitter into the extrellulr spe. Mny neurons ontin, in ition to synpti vesiles, nother lss of seretory vesiles, whih hve n eletron-ense ore n lrger imeter n re referre to s lrge ense-ore vesiles (LDCVs). LDCVs orrespon to the seretory vesiles of neuroenorine ells suh s the hromffin ells of the renl meull. LDCVs store proteins n pepties in ition to lssil neurotrnsmitters,. In neurons, LDCVs lso unergo exoytosis in response to C +, ut their relese is regulte ifferently from tht of synpti vesiles, requiring ursts of tion potentils. However, the relese mhinery, s fr s it is known, utilizes the sme proteins s synpti vesile exoytosis, inluing SNAREs n synptotgmins. Owing to the ese of ess to hromffin ells with miroeletroes, exoytosis of LDCVs in these ells hs serve s moel for stuying the mehnism of SNARE-meite exoytosis with eletrophysiologil pprohes 5. Vesile oking, tivtion of the fusion mhinery (priming), C + -epenent triggering n susequent memrne merger re rrie out y evolutionrily onserve protein omplexes funtioning in ll neurons n neuroenorine ells. Of these, SNARE proteins re, t present, onsiere the tlysts of the fusion retion 9. SNAREs meiting synpti exoytosis inlue syntxin- n SNAP-5 (t the plsm memrne) n synptorevin (t the vesile memrne),,. SNAREs re smll, memrne-nhore proteins tht ontin one or two onserve strethes of 7 mino is rrnge in hept repets, terme SNARE motifs. Relese is triggere y n influx of C + from the extrellulr spe in response to epolriztion. Upon triggering, the SNARE motifs form helil omplex tht riges the memrnes (the trns omplex). Complex ssemly proees towr the memrne nhors in the N-to-C-terminl iretion, thus pulling the memrnes together n inititing fusion s the SNARE omplex relxes into the is onfigurtion,. Triggering is meite y synptotgmins, trnsmemrne proteins of synpti vesiles n hromffin grnules (CGs) ontining two C + -ining C omins CA n CB tht in three or two C + ions, respetively,5. Synptotgmins intert with SNAREs n ii lipis. Both intertions re wiely onsiere essentil for their funtion. However, how extly synptotgmins elerte fusion is unler. We hve reently shown tht intertions with vesiulr ii lipis seem to intivte synptotgmins. Inorportion of synptotgmin- into liposomes ontining % phosphtiylserine (PS) results in ining of synptotgmin- to its own memrne (is ining), preventing trns ining to trget memrnes ontining ii phospholipis n/or SNARE proteins,7. Ntive vesiles ontin ~5% ii phospholipis, inluing PS n phosphtiylinositol (PI). It is not known how the intivting is ssoition of synptotgmins is prevente in vivo,9,. Here, we hve investigte the role of is n trns ining of synptotgmin- on C + -epenent fusion in vitro. To prevent prolems tht oul rise from the ifferenes etween rtifiil n iologil memrnes, we use purifie ovine CGs s one of the fusion prtners. Our t show tht CGs fuse with liposomes rrying syntxin-a n SNAP-5A in SNARE-epenent mnner. C + inrese Deprtment of Neuroiology, Mx Plnk Institute for Biophysil Chemistry, Göttingen, Germny. Fility for Eletron Mirosopy, Mx Plnk Institute for Biophysil Chemistry, Göttingen, Germny. Corresponene shoul e resse to R.J. (rjhn@gwg.e). Reeive Mrh; epte 7 July; pulishe online Septemer ; oi:./nsm.75 nture struturl & moleulr iology VOLUME 9 NUMBER OCTOBER 99

2 the rte of fusion severlfol, ut only when PI,5-isphosphte (PI(,5)P ) ws present in the trget memrne n, notly, when ws inlue in the uffer. No hyrolysis ws require, n the stimultory effet of ws ttriute to n eletrostti effet tht prevents the intivting is ining of synptotgmin C omins to the vesile memrne, therey llowing enogenous synptotgmins to intert in trns with the lipis n/or SNAREs in the trget memrne. RESULTS Chrteriztion of purifie hromffin grnules To reonstitute CG fusion in vitro, we purifie CGs from ovine renl meull using moifie protool in whih entrifugtion in ontinuous surose-ensity grient is the finl purifition step (Supplementry Fig. ). We onfirme the purity of the CGs y western lots, whih showe tht the finl frtion ws eplete of mrkers of other orgnelles (Fig. ). VAMP-, SNARE foun only on immture seretory grnules, ws lso remove uring purifition (Supplementry Fig. ), initing tht our protool yiels mture CGs t high purity. Cryo-EM of purifie CGs revele heterogeneous size istriution (Fig. ), with n verge imeter of 7.7 ±. nm (s.e.m.). Size heterogeneity of CGs in hromffin ells hs een oserve previously, with imeters of 5 nm, ut the verge imeter (5 nm) is lrger thn tht foun in our stuy. This is proly euse of the inlusion, in the previous stuies, of immture CGs, whih re known to e lrger. CGs eome smller n more onense through the removl of wter n the sheing of memrne uring the mturtion proess,5. Our t show tht our protool yiels minly mture CGs t high purity. SNARE-epenent fusion of hromffin grnules We investigte whether purifie CGs re le to fuse with lrge unilmellr liposomes (LUVs) ontining SNAP-5A n syntxin-a. Proteoliposomes ontining N-(7-nitro-,,-enzoxizol--yl) (NBD)- n rhomine-lele phospholipis were reonstitute with Synptorevin Synptophysin VMAT- DBH SDHA EEA- Clnexin LAMP- Rpt- Ctlse N + /K + -se H P SP CG Frequeny (%) 5 5 Dimeter (nm) Totl fluoresene (%) stilize eptor omplex known s the N omplex: preforme omplex of syntxin-a (lking the N-terminl H omin) n SNAP-5A (ontining C-terminl frgment of synptorevin resiues 9 9 (Sy 9 9 )) (ref. ). The lipi omposition of liposomes ws 5% phosphtiylholine, 5% phosphtiylethnolmine, % PS, 5% holesterol, % PI n % PI(,5)P. Fusion ws monitore y lipi-mixing ssy in whih fluoresene resonne energy trnsfer (FRET) etween the two fluorophore-lele lipis is reue s result of fusion with unlele lipis, leing to lipi ilution n equenhing of the onor fluorophore 7. We oserve roust fusion when we omine CGs with proteoliposomes ontining the SNARE eptor omplex (Fig. n Supplementry Fig. ). Memrne fusion ws SNARE speifi, s shown y ompetitive inhiition y solule frgment of synptorevin (Sy 9 ) or solule omplex of H omin of syntxin A (SyxH) n SNAP-5A n y inution with the light hin of tetnus neurotoxin (TeNT), protese seletively leving synptorevin (Fig. n Supplementry Fig.,). Furthermore, enogenous synptorevin in the CG memrne ws ple of forming SNARE omplexes, s shown y SDS-PAGE, in whih synptorevin ssemle in the ternry SNARE omplex ws foun to e resistnt to levge y TeNT light hin (Fig. ). Lysophosphtiylholine, whih estilizes the negtive urvture of stlk-type fusion intermeites y inuing positive urvture, inhiite CG fusion in ose-epenent mnner (Fig. e n Supplementry Fig. ). We otine similr results when we monitore ontent mixing (using FRET-se ssy) inste of lipi mixing (Supplementry Fig.,e), initing tht SNARE-epenent fusion of CGs is omplete n not rreste t hemifusion. C + enhnes fusion in the presene of Exoytosis of CGs is triggere, like tht of synpti vesiles, y n inrese in intrellulr C +. We therefore investigte whether ition of C + influene fusion etween purifie CGs n SNARE-ontining proteoliposomes. We oserve slight ut sustntil inhiitory effet t C + onentrtions > µm (Fig. No ition TeNT mutnt SyxH/SN5 Sy 9 TeNT Figure SNARE-epenent fusion of CGs with LUVs. () Western lot nlysis of mrker proteins speifi to CGs (synptorevin, synptophysin, vesiulr monomine trnsporter- (VMAT-) n opmine-et-hyroxylse (DBH)); mitohonri (suinte ehyrogense omplex suunit A (SDHA)); enoplsmi retiulum (lnexin); erly enosomes (erly enosome ssoite protein (EEA-)); lysosomes (lysosoml-ssoite memrne protein (LAMP-)); protesomes (S protesome AAA-se suunit RPT (Rpt-)); peroxisomes (tlse) n plsm memrne (N + /K + -se). See Online Methos for etils; see Supplementry Figure for frtiontion sheme n omplete lots. P, nulei n ell eris; S, superntnt; P, rue SyxH/SN5 SyxH/SN5 + + TeNT TeNT + + e Totl fluoresene (%) Non-hete Ternry omplex Synptorevin Hete No ition LPC µm LPC µm LPC 5 µm CG frtion. () Size istriution of CGs s etermine y ryo-em (n = 7). Inset, imge of typil CGs; sle r, nm. () Fusion of purifie CGs with LUVs ontining stilize SNARE eptor omplex s mesure y fluoresene equenhing ssy. Preinution of LUVs with Sy 9 n of CGs with SyxH SNAP-5A (SyxH/SN5) ompletely loke lipi mixing. Lipi mixing ws lso inhiite when CGs were preinute with TeNT light hin ut not when they were inute with n intive TeNT light-hin mutnt. () Immunolot etetion of enogenous synptorevin, issoite or in ternry SNARE omplexes, fter inution of CGs with exess SyxH SN5 (or without (two left lnes), with (+) or without ( ) TeNT, hete or non-hete. (e) Fluoresene vlues normlize s perentge of mximum onor fluoresene fter inution with.% Triton X- etergent. Lysophosphtiylholine (LPC) ws e efore the fusion retion; no ition, sl fusion without LPC. 99 VOLUME 9 NUMBER OCTOBER nture struturl & moleulr iology

3 , CCI Totl fluoresene (%) No ition CCI Sy 9,, CCI Totl fluoresene (%), CCI No ition CCI Sy 9,, CCI Fluoresene nisotropy Sy 9 TeNT,, CCI No ition CCI, CCI, CCI GTP, CCI -γ-s, CCI -γ-s GTP Totl fluoresene (%) e, CCI ADP, CCI AMP, CCI ADP AMP Totl fluoresene (%) f No CCI CCI Normlize fusion (%) -γ-s GTP ADP AMP Pyrophosphte Figure Role of polyphosphtes in C + -epenent vesile fusion s mesure y lipi-mixing ssy. () C + (CCl ) enhnes CG fusion with LUVs ontining the stilize eptor omplex only in the presene of. Preinution with Sy 9 olishe C + -inue CG fusion. No ition, sl fusion without or C +. () Fusion of purifie synpti vesiles from rt rin ws lso enhne y C +. Lels s in. () TeNT leves synptorevin, n o-tretment with n C + loks SNARE ssemly. At the en of eh retion, we e unlele Sy 9 to isple ll remining Sy 9 9. No ition, sl fusion without or C +. () hyrolysis is not require for C + -epenent enhnement of CG fusion. ws reple with -γ-s or GTP. (e) C + -epenent enhnement of CG fusion ws lso oserve in the presene of ADP n AMP, ut with lower effiy. (f) Eletrostti effet of polyphosphtes on CG fusion in the presene n sene of C +. Dt were tken t -min time points n normlize s the perentge of fusion in the presene of. Free C + ( µm) ws use in the presene of 5 mm unless otherwise inite. All quntittive t re men ± s.. from three or more inepenent experiments. n Supplementry Fig.,). While sreening for metolites tht oul ffet fusion, we note tht, in the presene of, C + i not inhiit fusion ut rther enhne it sustntilly (Fig. n Supplementry Fig. ). Similrly, fusion of synpti vesiles purifie from rt rin ws lso inrese y C + in the presene of ut inhiite slightly in the sene of (Fig. ), in greement with erlier experiments,9,,9. We sw no suh enhnement when we use MgCl inste of CCl (Supplementry Fig.,). Fusion ws ompletely loke when we e Sy 9 s ompetitive inhiitor, onfirming tht C + -enhne fusion is meite y SNARE proteins. Enhnement ourre whether n C + were e simultneously or sequentilly (Supplementry Fig. ). Furthermore, we oserve the sme rtes of C + -enhne fusion uner ll onitions esrie when we reple potssium glutmte in the uffer with potssium gluonte, KCl or NCl (t not shown). We otine similr results when we use liposomes ontining : full-length syntxin-a SNAP-5A inry eptor omplex, lthough the rtes were lower, s expete (Supplementry Fig. ). To onfirm tht SNARE zippering ours uring fusion, we took vntge of the ft tht Sy 9 9, the C-terminl synptorevin frgment tht stilizes the N eptor omplex, is isple y ining of full-length synptorevin from vesile memrnes, llowing SNARE omplex ssemly to e monitore y fluoresene nisotropy. When we inute CGs with liposomes ontining the N omplex with Alex Fluor lele Sy 9 9, we oserve erese of fluoresene nisotropy, initing SNARE-omplex zippering. Inution with n C + inrese peptie isplement, wheres or C + lone h no effet (Fig. n Supplementry Fig. e), initing tht C + -epenent enhnement of vesile fusion is ssoite with n inrese in numer of ssemle SNARE omplexes. We next investigte whether hyrolysis is require for the stimultory effet of C + -enhne fusion, whih woul suggest the involvement of n se in the CG memrne. However, this turne out not to e the se, s we lso oserve synpti vesile fusion when we reple with the non-hyrolyzle nlog -γ-s (Fig. f n Supplementry Fig. 5). Furthermore, we oserve no effet on fusion when we reple with GTP. The eletrostti effet of seems to epen primrily on numer of negtive hrges, s ADP n AMP lso enhne C + -epenent fusion, leit with reue effiy (Fig. e,f). Finlly, we teste pyrophosphte, whih ws s effiient s ADP (Fig. f), suggesting tht the pyrophosphte moiety of the nuleoties, rther thn the ses, is ruil for C + -epenent vesile fusion. C + enhnement epens on synptotgmins n PI(,5)P In hromffin ells, memers of the synptotgmin fmily meite fst exoytosis in response to C +. Inee, our results suggest tht C + -epenent enhnement of fusion is rought out y enogenous synptotgmins in the CG memrne, whih in to negtively hrge phospholipis y eletrostti intertions, prtiulrly in the presene of phosphoinosities. We use two pprohes to test this. In the first, we e monolonl ntioy speifi to the ytoplsmi omin of synptotgmin- (ref. ) to the fusion retion, whih resulte in prtil inhiition of C + -epenent enhnement ut h no effet on sl fusion rte (Fig. ). A monolonl ntioy speifi for the ytoplsmi omin of synptotgmin- reue C + -epenent enhnement of CG fusion, ompre to n ntioy speifi for the intrvesiulr omin of synptotgmin- (Fig. ). In the seon pproh, we e inresing onentrtions of solule ytoplsmi frgment of synptotgmin- (the CAB omin) s ompetitive inhiitor, whih resulte in the progressive inhiition of C + -epenent enhnement, gin with no ffet on sl fusion nture struturl & moleulr iology VOLUME 9 NUMBER OCTOBER 99

4 Totl fluoresene (%) A. A. 5 CCI No CCI Totl fluoresene (%) No CAB CAB. µm CAB µm CAB µm 5 CCI No CCI 5 rtes (Fig. ). Together, these finings suggest tht the C + effet on fusion is meite y enogenous synptotgmin-, lthough we nnot exlue the possiility tht other C + sensors (suh s synptotgmin-7) tht might ontriute to the effet my e present. To further verify the involvement of synptotgmin-, we exmine whether elertion of fusion is epenent on the presene of PI(,5)P in the trget memrne. In hromffin ells, plsm-memrne levels of PI(,5)P regulte priming y ontrolling the size of the relesle CG pool. PI(,5)P is lso importnt for C + -epenent vesile fusion owing to its enhnement of C + ffinity of synptotgmin- (ref. 5), n it umultes hevily t sites of oke vesiles. Notly, in the sene of PI(,5)P, the sl fusion rte ws lower, n no C + -epenent enhnement ws oserve (Fig. ). To gin more insight into the epenene of the C + effet on PI(,5)P, we i titrtions t ifferent PI(,5)P onentrtions in the trget memrne (Fig. ). As PI(,5)P onentrtions inrese, lower C + onentrtions were require to hieve the sme response. These t Totl fluoresene (%) No CCI CCI % PI(,5)P No PI(,5)P Figure Roles of synptotgmin- n PI(,5)P in C + -epenent enhnement of vesile fusion. () A monolonl ntioy speifi for the ytoplsmi omin of synptotgmin- (A.) reue C + -epenent enhnement of CG fusion, wheres no effet ws oserve on the sl fusion rte (no CCl ). An ntioy speifi for the intrvesiulr omin of synptotgmin- (A.) ws use s ontrol. () Aition of reominnt CAB omin of synptotgmin- loke C + -epenent enhnement of CG fusion in ose-epenent mnner. No effet on sl fusion ws oserve. () PI(,5)P is require for C + -epenent enhnement of CG fusion. Omission of PI(,5)P from the LUV memrne reue sl fusion n olishe C + -epenent enhnement of fusion. () Higher PI(,5)P onentrtions inrese C + sensitivity for CG fusion y lowering EC 5. Enhnement of vesile fusion y Fusion (normlize response) % PI(,5)P % PI(,5)P 5% PI(,5)P CCI (µm) f Totl fluoresene (%) No CCI CCI.5% PI(,5)P No PI(,5)P.5% PI(,5)P Sy 9, CCI suggest tht higher lol onentrtions of PI(,5)P enhne C + sensitivity y lowering hlf-mximum effetive onentrtion (EC 5 ), in greement with stuies rrie out with purifie synptotgmin- (ref. 5). Next, we exhnge PI(,5)P for PI -phosphte (PIP), phosphoinositie speies speifilly ssoite with enosoml n utophgosoml memrnes 7,. C + -epenent enhnement still ourre ut ws less roust in the presene of PIP thn in the presene of PI(,5)P (Fig. e n Supplementry Fig. ). We lso oserve enhnement in the sene of PI(,5)P when we e the ii phospholipi PS, whih hs net hrge of, t onentrtion of % (Supplementry Fig. ). We otine similr results for the epenene on PI(,5)P when we use synpti vesiles inste of CGs (Fig. f n Supplementry Fig. 5). prevents is ining of synptotgmins y hrge shieling Wht is the mehnism y whih polyphosphte nions enhne C + - n synptotgmin-meite fusion of CGs n synpti e % ontrol No CCI CCI No PIP % PIP % PIP C + t PI(,5)P onentrtions of % (. ± µm), % ( ± µm) n 5% (7. ± µm) in LUV memrnes (pprent EC 5 vlues for C + in prentheses). C + -epenent fusion ws normlize to present the mximum fusion s (see Online Methos for etils). All quntittive t re men ± s.. from three or more inepenent experiments. (e) C + -epenent enhnement of vesile fusion ws quntifie s perentge of sl fusion in the presene of PI(,5)P (PIP ) n PIP. (f) Purifie synpti vesiles from rt rin lso require PI(,5)P in the trget LUVs for C + -epenent enhnement of fusion. Free C + ( µm) ws use in the presene of 5 mm unless otherwise inite. All quntittive t re men ± s.. Fluoresene nisotropy.9..7 CG C + AMP. Anisotropy (net hnges, ).5..5 No CCI CCI + CCI GTP + CCI AMP + CCI Anisotropy (net hnges, ) CG, no CG,, CCI (µm) SV, no SV,, CCI (µm) Figure Effet of on ssoition of synptotgmin- to the vesile memrne. (,) CAB ining to CG memrnes ws monitore y fluoresene nisotropy. () CGs were inute with purifie frgment of synptotgmin- enompssing oth C omins (Syt 97, Alex Fluor lele t SC). Bining resulte in n inrese of nisotropy ue to reue moility of the fluorophores 5. Aition of (5 mm) reues free C + onentrtion to µm ue to heltion; see Online Methos for etils. Net hnges of fluoresene nisotropy intensity y C + re shown in. (,) C + -epenene of CAB ining to purifie CGs () n synpti vesiles (SVs; ) in the presene n sene of 5 mm. (e,f) Fluoresene nisotropy over time (e) n net hnges uner inresing C + onentrtions (f) of memrne ining using liposomes (Lip) Anisotropy (net hnges, ) f e Fluoresene nisotropy Anisotropy (net hnges, ) Liposome (% PI(,5)P ) C + Lip, no Lip,, CCI (µm) ontining % PI(,5)P. All quntittive t re men ± s.. from three or more inepenent experiments. All C + -onentrtions were orrete to ount for -epenent heltion. 99 VOLUME 9 NUMBER OCTOBER nture struturl & moleulr iology

5 Figure 5 Effet of ii phospholipi onentrtion in the liposome memrne on C + -epenent ining of synptotgmin-. ( ) C + -epenent ining of the CAB omin ws monitore y fluoresene nisotropy s in Figure, using LUVs ontining % (), 5% () n % () phosphtiylserine (PS). Left pnels show exemplry fluoresene nisotropy tres; mile pnels show net hnges of nisotropy in epenene on free C + onentrtion; right pnels show hnges in epenene on C + n of memrne fusion. Donor liposomes were reonstitute with full-length synptotgmin- (Syt-) n synptorevin- (Sy), wheres eptor LUVs ontine stilize SNARE eptor omplex (Syx/SN) n % PI(,5)P (see igrm ove pnels). Dt were normlize to the fusion tivity mesure in the sene of C + in the sme experiment. All quntittive t re men ± s.. from three or more inepenent experiments. vesiles? Consiering the t ove, we hypothesize tht polyphosphtes suh s Liposome. n relte ompouns might tivte enogenous synptotgmins in the memrnes of CGs n synpti vesiles. Previous. stuies hve suggeste tht the C omins of memrne-nhore synptotgmins might intert with their nhor memrnes (is ining), whih oul interfere with or even prevent ining to the trget memrne,7,9. Trns ining of synptotgmins, however, is elieve to e require for synptotgmins to rive C + -epenent exoytosis. Beuse we were unle to iretly mesure ssoition of enogenous synptotgmins, we use fluoresene nisotropy to monitor ining of n exogenously e CAB frgment of synptotgmin-, lele with Alex Fluor, to CG n synpti vesile memrnes (Fig. ). Aition of C + inrese fluoresene nisotropy, initing tht the CAB omins of synptotgmin- in to CG memrnes in C + -epenent mnner (Fig. ). Aition of or GTP fter C + erese the nisotropy signl, suggesting issoition of the CAB frgment from the memrne. Aition of AMP h smller effet on nisotropy, thus mirroring the effets of AMP on fusion enhnement. We otine very similr results when using purifie synpti vesiles inste of CGs (Fig. ). Inrese of the free C + onentrtion to > mm i not ffet the ility of to prevent ining (Fig.,; see Online Methos for etils). Tken together, these results suggest tht synptotgmins in to their own memrnes (is ining) in response to C +, whih olishes trns intertions, in greement with our previous oservtions using rtifiil vesiles 7. To nlyze in more etil the effet of on the C + -epenent ining of synptotgmins to ii memrnes, we rrie out ining experiments using PI(,5)P -ontining liposomes with the sme omposition s those use in the fusion ssys. Aition of C + resulte in inrese ining of the CAB omin, n this ining ws not reverse y (Fig. e,f). Cis ining of synptotgmins is prevente y polyphosphte nions t physiologil onentrtions, wheres trns intertions with PI(,5)P -ontining trget memrnes o not seem to e inhiite. C + -epenent ining of the C omins to memrnes is strongly enhne y PI(,5)P, ut it lso epens on the onentrtion of ii memrne lipis suh s PS or PI. We therefore Fluoresene nisotropy Fluoresene nisotropy Fluoresene nisotropy.. % PS Liposome... 5% PS % PS Liposome C + mm C +.. C + 5 mm Anisotropy (net hnges, ) Anisotropy (net hnges, ) Anisotropy (net hnges, ) CAB No CCl (µm) No CCl (µm) No CCl (µm),,, speulte tht the inhiitory effet of n its prlogs might e effetive only t moerte onentrtions of suh ii lipis. To test this, we mesure C + -epenent ining of synptotgmin- to liposomes t inresing onentrtions of PS, in the presene n sene of (Fig. 5). As esrie previously, the lium sensitivity of C + -epenent ining ws inrese with inresing PS onentrtions. Notly, reverse ining t PS onentrtions of % n 5% ut not % (for referene, synpti vesile memrnes re ~5% ii phospholipis). Inee, the C + ose-response urves for CAB ining to CGs (Fig. ) n synpti vesiles (Fig. ) resemle tht of liposomes ontining 5% PS (Fig. 5), initing tht onentrtion of 5% PS mimis the eletrostti environment of ntive vesile memrnes. Finlly, we teste whether we oul reproue is intivtion, its prevention y n trns elertion of fusion using liposomes reonstitute with synptorevin n synptotgmin- s onor vesiles inste of purifie CGs n synpti vesiles. Speifilly, we ske whether synptotgmin- is ple of enhning SNARE-meite fusion in C + -epenent mnner, n we foun tht it ws (Fig. 5). C + -epenent elertion of fusion orrelte with the prevention of is ining y (Fig. 5, right), ut we i not oserve the enhnement when the memrnes of the vesiles inorporting synptotgmin- n synptorevin ontine more thn % PS (Fig. 5, right). These oservtions gree with reent report showing tht C + -epenent enhnement of fusion epens on the mount of PS in synptotgmin- ontining liposomes 9. Notly, C + -epenent fusion of liposomes ws less effiient thn fusion of ntive vesiles. DISCUSSION Exoytosis of neuroseretory vesiles is meite y SNAREs n triggere y C + -oun synptotgmins. However, the mehnism y whih synptotgmins enhne fusion is still unler. In vitro Fusion (% inrese) Fusion (% inrese) Fusion (% inrese) Syt- Sy No Syx/SN, CCl (µm) No CCl (µm) No CCl (µm),, nture struturl & moleulr iology VOLUME 9 NUMBER OCTOBER 995

6 C + -ining loop of synptotgmin- Asprtte Aenosine Polysi pth in CB Lysine Aenosine Synptorevin No C + Cis intivtion Synptotgmin- Trns intivtion 5% PS (vesile memrne) 5% PS (vesile memrne) % Pl(,5)P (plsm memrne) C + Pl(,5)P Figure Effet of polyphosphtes on C + -epenent ining of synptotgmins to memrnes ontining ii phospholipis. () In the sene of, synptotgmin- ins to the vesile memrne in C + -epenent mnner (EC 5 = µm C + ). Asprtte resiues of the C omins (lk lls) provie prtil oorintion for C + ions, with more omplete oorintion sphere ontriute y ii memrne lipis. (), whih heltes C + with K = µm (ref. ), ompetes with C + -epenent memrne ining of synptotgmin- y shieling the oorintion site of ii phospholipis. () In the presene of PI(,5)P, C + ining etween the memrne n the C omins is enhne, with n EC 5 = 5 µm C + (Fig. f), proly euse of polysi pth within the CB omin tht ins PI(,5)P on its own. The high C + ffinity of CAB ining to PI(,5)P -ontining trget memrnes nnot e ompete for y. () In the sene of, C + intivtes synptotgmin- through is ssoition (ining to vesile memrne). seletively olishes is ining of synptotgmin-, leving the CAB omins tive to intert with the plsm memrne ontining PI(,5)P, s require for C + triggering of exoytosis. reonstitution of SNARE-epenent memrne fusion in the presene of synptotgmins hs yiele onfliting results, with the effets of C + rnging from slight inhiition to enhnement uner wiely vrying onitions,9,9. We hve previously shown tht memrnenhore synptotgmin- ins to its own memrne (is ining) when ii phospholipis re present, preventing trns intertions with the trget memrne,7. Here we hve shown tht is ining of synptotgmin- ours in ntive CGs n synpti vesiles, ut tht suh is ining is prevente y polyphosphte nions, inluing, t physiologil onentrtions. n nlogous ompouns proly operte y hrge sreening tht is, y ompeting iretly with ii memrne lipis to helte C + n isrupt is ining of synptotgmins (Fig. ). A ntive vesile memrne tht ontins ~5% ii phospholipis hs C + ffinity, with EC 5 = ± 9 µm, for ining of the CAB omin (Fig. ), n on its own heltes C + with similr ffinity (EC 5 = µm, t not shown; see lso n ref. ), therey inhiiting CAB ining to the vesile memrne. Notly, sreening is effetive only when the onentrtion of ii phospholipis oes not exee 5% n when no PI(,5)P is present in the memrne. PI(,5)P -ontining trget memrnes hve muh higher C + ffinity thn vesile memrnes for CAB ining (EC 5 = 5 ± 9 µm), n synptotgmins n then t in trns, resulting in mjor C + -epenent enhnement of SNARE-epenent fusion. Our t she light on the mehnisms y whih synptotgmin C + sensors might operte etween the vesile n the plsm memrne. It is well estlishe tht oth C omins show highly oopertive C + -epenent ining to memrnes ontining ii phospholipis 5,, with higher onentrtions of ii phospholipis in the memrne resulting in higher ffinities. Furthermore, synptotgmins ontin si pth in the CB omin tht ins to PI(,5)P in n t lest prtilly C + -inepenent mnner, n enhnes the C + sensitivity of exoytosis 5. Our t suggest tht there my e elite lne etween is n trns ining of synptotgmins: wheres the trget memrne ontining high onentrtions of PI(,5)P llows for strong C + -epenent (n prtilly C + -inepenent) ining of synptotgmins, the onentrtion of ii phospholipis in the vesile memrne seems to e juste to regulte is ining of synptotgmins. We o not yet know whether polyphosphte-epenent sreening of intivting is ining hs role in the regultion of C + -epenent exoytosis uner physiologil onitions. In permeilize neuroenorine ells, triggering of exoytosis y C + epens on the presene of. Moreover, we hve shown previously tht in ell-free preprtion ompose of inverte lwns of plsm memrne ontining oke seretory vesiles, ition of C + eliits exoytosis, ut only when is present 7. epenene might simply reflet the involvement of -using enzymes suh s N-ethylmleimie sensitive fusion protein or PI -kinse. However, it is oneivle tht the strit epenene on the presene of even uring the C + -triggering phse oul lso e ttriutle to the prevention of intivting is ining of synptotgmins to the vesile memrne. How o these t ontriute to the unerstning of the stillete mehnism y whih C + -ining to synptotgmins elertes the rte of exoytosis y more thn five orers of mgnitue? Mny moels hve een suggeste for the tion of synptotgmins in the fusion pthwy, suh s the unloking or tivtion of rreste trns SNARE omplexes, genertion of lolize memrne urvture in the plsm memrne or perturtion of the hyrophilihyrophoi ounry of the memrnes t the ontt site. Although we nnot yet exlue ny of these mehnisms, it hs reently een suggeste tht synptotgmins my t s istne regultors tht pull the vesile n the plsm memrne it loser upon C + triggering, thus triggering SNARE ssemly n fusion. It is possile tht synptotgmins first in in trns to the trget memrne n tht this trns ining is then followe y is ining involving the C omins to the vesile memrne, shortening the istne etween the memrnes. Suh two-step mehnism oul llow for regultion of is ining, for exmple, y fine-tuning the onentrtion of ii phospholipis n/or phosphorylte vrints of PI in the vesile memrne t the ontt site. Methos Methos n ny ssoite referenes re ville in the online version of the pper. Note: Supplementry informtion is ville in the online version of the pper. 99 VOLUME 9 NUMBER OCTOBER nture struturl & moleulr iology

7 Aknowlegments We re inete to G. Mieskes (Deprtment of Neuroiology, Mx Plnk Institute for Biophysil Chemistry) for the rrngement of renl glns n logistil ssistne. This work ws supporte y grnts from the Alexner von Humolt Fountion (to Y.P.) n the US Ntionl Institutes of Helth ( P GM79- A to R.J.). AUTHOR CONTRIBUTIONS J.M.H. ssiste in the genertion of SNARE-ontining lrge unilmellr liposomes n performe the light-sttering experiments. G.v..B. provie lele proteins n ssiste in the fluoresene nisotropy experiments. S.A. n M.H. provie purifie synpti vesiles. D.R. performe EM. Y.P. n R.J. esigne the stuy n wrote the pper. Experiments were onute minly y Y.P. All uthors isusse the results n ommente on the mnusript. COMPETING FINANCIAL INTERESTS The uthors elre no ompeting finnil interests. Pulishe online t Reprints n permissions informtion is ville online t reprints/inex.html.. Augustine, G.J. How oes lium trigger neurotrnsmitter relese? Curr. Opin. Neuroiol., ().. De Cmilli, P. & Jhn, R. Pthwys to regulte exoytosis in neurons. Annu. Rev. Physiol. 5, 5 5 (99).. Prk, Y. & Kim, K.T. Short-term plstiity of smll synpti vesile (SSV) n lrge lrge ense-ore vesile (LDCV) exoytosis. Cell. Signl., 5 7 (9).. Neher, E. Vesile pools n C + miroomins: new tools for unerstning their roles in neurotrnsmitter relese. Neuron, 9 99 (99). 5. Neher, E. A omprison etween exoyti ontrol mehnisms in renl hromffin ells n glutmtergi synpse. Pflugers Arh. 5, ().. Jhn, R. & Sheller, R.H. SNAREs engines for memrne fusion. Nt. Rev. Mol. Cell Biol. 7, (). 7. Mrtens, S. & MMhon, H.T. Mehnisms of memrne fusion: isprte plyers n ommon priniples. Nt. Rev. Mol. Cell Biol. 9, 5 55 ().. Mrtin, T.F. The moleulr mhinery for fst n slow neuroseretion. Curr. Opin. Neuroiol., (99). 9. Rizo, J. & Rosenmun, C. Synpti vesile fusion. Nt. Strut. Mol. Biol. 5, 5 7 ().. Mlsm, J., Kreye, S. & Sollner, T.H. Memrne fusion: SNAREs n regultion. Cell. Mol. Life Si. 5, ().. Sühof, T.C. & Rothmn, J.E. Memrne fusion: grppling with SNARE n SM proteins. Siene, 7 77 (9).. Stein, A., Weer, G., Whl, M.C. & Jhn, R. Helil extension of the neuronl SNARE omplex into the memrne. Nture, 55 5 (9).. Wieerhol, K. & Fsshuer, D. Is ssemly of the SNARE omplex enough to fuel memrne fusion? J. Biol. Chem., 5 (9).. Fernnez, I. et l. Three-imensionl struture of the synptotgmin CBomin: synptotgmin- s phospholipi ining mhine. Neuron, 57 9 (). 5. Chpmn, E.R. How oes synptotgmin trigger neurotrnsmitter relese? Annu. Rev. Biohem. 77, 5 ().. Stein, A., Rhkrishnn, A., Rieel, D., Fsshuer, D. & Jhn, R. Synptotgmin tivtes memrne fusion through C + -epenent trns intertion with phospholipis. Nt. Strut. Mol. Biol., 9 9 (7). 7. Vennekte, W. et l. Cis- n trns-memrne intertions of synptotgmin-. Pro. Ntl. A. Si. USA 9, 7 ().. Tkmori, S. et l. Moleulr ntomy of trffiking orgnelle. Cell 7, (). 9. Holt, M., Rieel, D., Stein, A., Shuette, C. & Jhn, R. Synpti vesiles re onstitutively tive fusion mhines tht funtion inepenently of C +. Curr. Biol., 75 7 ().. Mhl, L.K., Sequeir, S.M., Guresko, J.M. & Sollner, T.H. Clium-inepenent stimultion of memrne fusion n SNAREpin formtion y synptotgmin. J. Cell Biol. 5, 7 ().. Eton, B.A., Hugwitz, M., Lu, D. & Moore, H.P. Biogenesis of regulte exoytoti rriers in neuroenorine ells. J. Neurosi., 7 7 ().. Grner, C.P., Prie, S.D., Lyskowski, A. & Fox, A.P. Mouse hromffin ells hve two popultions of ense-ore vesiles. J. Neurophysiol. 9, 9 (5).. Plttner, H., Artlejo, A.R. & Neher, E. Ultrstruturl orgniztion of ovine hromffin ell ortex nlysis y ryofixtion n morphometry of spets pertinent to exoytosis. J. Cell Biol. 9, (997).. Kim, T., Gonre-Lewis, M.C., Arnoutov, I. & Loh, Y.P. Dense-ore seretory grnule iogenesis. Physiology (Bethes), (). 5. Melolesi, J., Chieregtti, E. & Luis Mlosio, M. Requirements for the ientifition of ense-ore grnules. Trens Cell Biol., 9 ().. Poti, A.V., Stein, A. & Fsshuer, D. N- to C-terminl SNARE omplex ssemly promotes rpi memrne fusion. Siene, 7 7 (). 7. Struk, D.K., Hoekstr, D. & Pgno, R.E. Use of resonne energy trnsfer to monitor memrne fusion. Biohemistry, 9 99 (9).. Chernomorik, L.V. et l. Lysolipis reversily inhiit C + -, GTP- n ph-epenent fusion of iologil memrnes. FEBS Lett., 7 7 (99). 9. Lee, H.K. et l. Dynmi C + -epenent stimultion of vesile fusion y memrnenhore synptotgmin-. Siene, 7 7 ().. Dvletov, B.A. & Suhof, T.C. A single C omin from synptotgmin- is suffiient for high-ffinity C + -phospholipi ining. J. Biol. Chem., 9 (99).. Zhng, X., Rizo, J. & Suhof, T.C. Mehnism of phospholipi ining y the CAomin of synptotgmin-. Biohemistry 7, 95 (99).. Chpmn, E.R. & Jhn, R. Clium-epenent intertion of the ytoplsmi region of synptotgmin with memrnes. Autonomous funtion of single C-homologous omin. J. Biol. Chem. 9, (99).. Di Polo, G. & De Cmilli, P. Phosphoinosities in ell regultion n memrne ynmis. Nture, 5 57 ().. Milosevi, I. et l. Plsmlemml phosphtiylinositol-,5-isphosphte level regultes the relesle vesile pool size in hromffin ells. J. Neurosi. 5, (5). 5. Rhkrishnn, A., Stein, A., Jhn, R. & Fsshuer, D. The C + ffinity of synptotgmin- is mrkely inrese y speifi intertion of its C B omin with phosphtiylinositol-,5-isphosphte. J. Biol. Chem., (9).. vn en Bogrt, G. et l. Memrne-protein sequestering y ioni protein-lipi intertions. Nture 79, (). 7. Botelho, R.J., Sott, C.C. & Grinstein, S. Phosphoinositie involvement in phgoytosis n phgosome mturtion. Curr. Top. Miroiol. Immunol., ().. Gillooly, D.J. et l. Loliztion of phosphtiylinositol--phosphte in yest n mmmlin ells. EMBO J. 9, (). 9. Li, Y. & Shin, Y.K. The importne of n symmetri istriution of ii lipis for synptotgmin- funtion s C + sensor. Biohem. J., 9 ().. Bhll, A., Chik, M.C., Tuker, W.C. & Chpmn, E.R. C + -synptotgmin iretly regultes t-snare funtion uring reonstitute memrne fusion. Nt. Strut. Mol. Biol., ().. Kyoung, M. et l. In vitro system ple of ifferentiting fst C + -triggere ontent mixing from lipi exhnge for mehnisti stuies of neurotrnsmitter relese. Pro. Ntl. A. Si. USA, E E ().. Wilson, J.E. & Chin, A. Cheltion of ivlent tions y, stuie y titrtion lorimetry. Anl. Biohem. 9, 9 (99).. Kuo, W., Herrik, D.Z., Ellen, J.F. & Cfiso, D.S. The lium-epenent n lium-inepenent memrne ining of synptotgmin-: two moes of CB ining. J. Mol. Biol. 7, 9 (9).. Vrlji, M. et l. Post-trnsltionl moifitions n lipi-ining profile of inset ell expresse full-length mmmlin synptotgmin-. Biohemistry 5, 999 (). 5. Li, L. et l. Phosphtiylinositol phosphtes s o-tivtors of C + ining to C omins of synptotgmin-. J. Biol. Chem., ().. Bker, P.F. & Knight, D.E. Clium-epenent exoytosis in ovine renl meullry ells with leky plsm memrnes. Nture 7, (97). 7. Brszzewski, M. et l. A novel site of tion for α-snap in the SNARE onformtionl yle ontrolling memrne fusion. Mol. Biol. Cell 9, 77 7 ().. vn en Bogrt, G. et l. Synptotgmin- my e istne regultor ting upstrem of SNARE nuletion. Nt. Strut. Mol. Biol., 5 (). nture struturl & moleulr iology VOLUME 9 NUMBER OCTOBER 997

8 ONLINE METHODS Mterils. Clein,, ADP, AMP, GTP, -γ-s n pyrophosphte were purhse from Sigm (St. Louis, MO). l-α-lysophosphtiylholine (LPC) n other lipis were from Avnti (Alster, AL). Antioies to synptorevin (lone numer 9.), synptophysin (lone numer 7.), synptotgmin- (monolonl ntioies. n.) n VAMP- (tlog no. ) were from Synpti Systems (Göttingen, Germny). Antioies to SDHA (tlog no. 75), EEA- (tlog no. 9), lnexin (tlog no. ), LAMP- (tlog no. 7), Rpt- (tlog no. 9), tlse (tlog no. 77) n N + /K + -se (tlog no. 77) were from Am (Cmrige, MA). All ntioies were ilute to :, for use. Purifition of hromffin grnules n synpti vesiles. CGs were purifie s previously esrie 9, with severl moifitions. Fresh ovine renl glns were otine from lol slughterhouse. After the ortex n ft were trimme wy, the meulle were mine with sissor in -mm surose uffer ( mm surose, mm HEPES (ph 7.) juste with KOH) n then homogenize using oole glss Teflon homogenizer t, r.p.m. PMSF ( µm) ws e to prevent protein egrtion. All susequent steps were rrie out t C. The smple ws entrifuge t,g for 5 min t C, fter whih the pellet ontining nulei n ell eris (P) ws isre. The superntnt (S) ws further entrifuge (,g, 5 min, C) n then sujete to n itionl yle of resuspension n entrifugtion for wshing. The resulting pellet (P, rue CG frtion) ws resuspene in -mm surose uffer n loe on top of ontinuous surose grient (from mm to. M) to remove other ontminnts, inluing mitohonri. CGs were ollete from the pellet fter entrifugtion t 7, r.p.m. for min in Bekmn SW Ti rotor n resuspene with the uffer ( mm potssium glutmte, mm potssium ette, mm HEPES n KOH (ph 7.)). The frtion iretly on top of the pellet ws remove, n the pellet lone ws resuspene only to purify mture CGs. Synpti vesiles from rt rin were purifie s previously esrie. Briefly, rt rins were homogenize in homogeniztion uffer supplemente with protese inhiitors, using glss Teflon homogenizer, with strokes t 9 r.p.m. The homogente ws entrifuge for min t,g, n the resulting superntnt ws further entrifuge for 5 min t 5,g. The superntnt S ws store on ie for lter use. The synptosome pellet ws lyse y ition of ie-ol wter, n three strokes t, r.p.m. were pplie. Protese inhiitors n HEPES were e to the lyste immeitely. The lyste ws entrifuge for 5 min t 7,g, n the superntnt LS ws omine with the S. The mixture of LS n S ws entrifuge for 5 min t,g. The resulting superntnt CS ws overli onto.7-m surose ushion n entrifuge for h t,g. The pellet ws resuspene in olumn uffer ( mm Tris- HCl, mm KCl (ph 7.)) n loe onto Sephryl S- size-exlusion hromtogrphy olumn ( m). Protein purifition. All SNARE onstruts were se on rt sequenes n were lone in the pet vetor. The TeNT light hin (oth wil-type n the intive EA mutnt 5 ), n SNARE proteins inluing the solule form of synptorevin lking the trnsmemrne omin (Sy 9 ) n CAB omin of synptotgmin- (resiues 97 ), were expresse in Esherihi oli n purifie y Ni + -NTA ffinity hromtogrphy followe y ion-exhnge hromtogrphy with Mono S olumn on n Äkt system (GE Helthre, Pistwy, NJ). The stilize Q-SNARE omplex, referre to s the N omplex n ontining syntxin-a ( ), SNAP-5A (with ll ysteines reple y lnines) n the C-terminl synptorevin frgment (9 9), ws purifie s esrie previously. The : inry Q-SNARE omplex ontining syntxin- A ( ) n SNAP-5A (no ysteine) ws expresse using otrnsformtion 5. The N omplex, the syntxin-a SNAP-5A : inry omplex, SNAP-5A (no ysteine) n syntxin-a (, n (SyxH)) were purifie y Ni + -NTA ffinity hromtogrphy followe y ion-exhnge hromtogrphy on Mono Q olumn (GE Helthre, Pistwy, NJ) in the presene of 5 mm n-otyl-β--gluosie. For nisotropy mesurements, point-mutte CAB (SC) 5 n Sy 9 9 (T79C) in the N omplex were lele with Alex Fluor C5 mleimie. Preprtion of proteoliposomes. Unless inite otherwise, the lipi omposition of proteoliposomes (molr rtios) onsists of 5% PC (l-α-phosphtiylholine), 5% PE (l-α-phosphtiylethnolmine), % PS (l-α-phosphtiylserine), 5% holesterol n 5% PI (l-α-phosphtiylinositol). PI(,5)P or PIP, t the inite onentrtions, reple PI. For FRET-se equenhing ssys,.5%,-ioleoyl-sn-glyero--phosphoethnolmine-n- (7-nitroenz--ox-,-izol--yl) (NBD-DOPE) n.5%,-ioleoyl-snglyero--phosphoethnolmine-n-lissmine rhomine B sulfonyl mmonium slt (Rhomine-DOPE) were use s onor n n eptor ye, respetively. Synptorevin- n synptotgmin- ontining liposomes onsist of 5% PC, % PE, % PS n % holesterol. When 5% PS or % PS ws use, PC ontents were juste oringly. As esrie 5, liposomes were extrue using polyronte memrnes of pore size nm (Avnti Polr lipis) to give uniformly istriute lrge unilmellr vesiles (LUVs) in the imeter rnge of nm s onfirme y fielflow frtiontion ouple with multingle lser light sttering (FFF-MALLS, Wytt Tehnology Corportion, Snt Brr, CA, USA). Inorportion of the proteins into liposomes ws hieve y n-otyl-β-gluosie (OG)-meite reonstitution. Proteoliposomes ontining the stilize eptor omplex ( N omplex) or the syntxin-a SNAP-5A inry omplex in : stoihiometry (: omplex) were prepre y etergent-ssiste insertion of proteins s esrie previously,5. N omplex in 5 mm OG ws mixe with LUVs (lipi-to-protein rtio of 5: (n/n)). In se of : omplex inorportion with LUVs, lipi-to-protein rtio ws :. For ontent-mixing ssys, 5 mm lein (95/55 nm) ws enpsulte in proteoliposomes s esrie previously 5. Lipis were issolve in iethyl ether (.5 ml) n resuspene with.5 ml of 5 mm lein (N + -lein ) in -mm HEPES-KOH (ph 7.), 75 mm KCl n mm DTT. Content mixing ws speifi to SNARE proteins inorporte in liposomes n lekge, etermine y quenhing lein leke into the meium y ition of Co +, ws only 5% of totl lein (for etils see ref. ). Fusion retion. CG fusion retions were one t 7 C. For eh retion, 5 µg of CGs n µl of proteoliposomes were mixe in ml of uffer ontining mm potssium glutmte, mm potssium ette, mm HEPES-KOH (ph 7.) n 5 mm MgCl. Unless inite otherwise, eptor liposomes ontine the stilize Q-SNARE omplex, terme N omplex. The : (syntxin-a SNAP-5A) Q-SNARE omplex ws lso teste for SNARE- n C + -epenent fusion (Supplementry Fig. ). For C + -epenent fusion, 5 mm N + - ws e. shoul e me freshly for experiments euse is esily estroye y freezing n thwing. Fluoresene equenhing signl ws mesure y FluoroLog n FluoroMx (HORIBA Join Yvon), with wvelengths of nm (slit with of nm) for exittion n 5 nm (slit with of nm) for emission. Fluoresene vlues were normlize s the perentge vlue of the mximum onor fluoresene inue y.% Triton X- etergent tretment t the en of eh experiment. No ition represents sl fusion without ny tretment or C +. Quntifition of vesile-fusion t of lipi-mixing n ontent-mixing ssy is presente s perentge y normlizing sl fusion fter min of retion time. C + -epenent fusion with the ifferent onentrtions of PI(,5)P (Fig. ) ws normlize oring to Fusion nor = (T T ) / (T mx T ), where T is the perentge of totl fluoresene inue y vesile fusion, T inites sl fusion without C + n T mx inites fusion t µm C + (mximum level). Cryo-eletron mirosopy. Smples were oun in Vitroot Mrk IV (FEI Compny) to glow-ishrge ron foil overe gri. The suspension ws lotte for s t lot fore = n vitrifie t C n 97% humiity. The smples were evlute with CM trnsmission eletron mirosope, n pitures were tken with TemCm A slow sn CCD mer (TVIPS, Guting, Germny). Fluoresene-nisotropy mesurements. Anisotropy mesurements were rrie out in FluoroLog spetrometer in T-onfigurtion equippe for polriztion (Moel FL, Join Yvon). All experiments were one t 7 C in ml of uffer ontining mm potssium glutmte, mm potssium ette, mm HEPES-KOH (ph 7.) n 5 mm MgCl. N + - n CCl were trete s inite. Alex Fluor lele proteins were exite t nm (slit with of nm), n their emission ws mesure t 5 nm (slit with of nm). For monitoring of SNARE ssemly, µg CGs were inute with nture struturl & moleulr iology oi:./nsm.75

9 % PI(,5)P -ontining liposomes tht inorporte the N omplex (Sy 9 9 lele with Alex Fluor t T79C). For monitoring of the ining of the CAB omin, nm CAB (SC, Alex Fluor lele) ws inute with µg CGs or protein-free liposomes ontining %, 5% or % PS. The G ftor ws lulte oring to G = I HV / I HH, where I is the fluoresene intensity, the first susript letter inites the iretion of the exiting light, n the seon susript letter the iretion of emitte light. The intensity of the vertilly (V) n horizontlly (H) polrize emission light fter exittion y vertilly polrize light ws mesure. The nisotropy (r) ws etermine oring to r = (I VV G I VH ) / (I VV + G I VH ). C + lirtion. ontins negtively hrge oxygen toms tht in to Mg +, C + or Sr +, therey helting ivlent tions. C + onentrtions were lirte with Fluo-5N, low-ffinity C + initor with K = 9 µm, n experiment t were orrelte with simultion tht lultes the free C + onentrtions ( Sttistil nlysis. All quntittive t re men ± s.. from three or more inepenent experiments. Dose-response urves were fit using four-prmeter logisti equtions (PL) to lulte EC 5 (SigmPlot). 9. Smith, A.D. & Winkler, H. A simple metho for the isoltion of renl hromffin grnules on lrge sle. Biohem. J., (97). 5. Li, Y. et l. A single muttion in the reominnt light hin of tetnus toxin olishes its proteolyti tivity n removes the toxiity seen fter reonstitution with ntive hevy hin. Biohemistry, 7 7 (99). 5. Weer, T. et l. SNAREpins: miniml mhinery for memrne fusion. Cell 9, (99). 5. Cypionk, A. et l. Disrimintion etween oking n fusion of liposomes reonstitute with neuronl SNARE proteins using FCS. Pro. Ntl. A. Si. USA, (9). 5. Kenll, D.A. & MDonl, R.C. Chrteriztion of fluoresene ssy to monitor hnges in the queous volume of lipi vesiles. Anl. Biohem., (9). oi:./nsm.75 nture struturl & moleulr iology