... Calmodulin bifurcates the local Ca 2+ signal that modulates P/Q-type Ca 2+ channels

Transcription

1 ships. This protool ws repete in Kres solution with 6.5 mm K +. To ientify the postsynpti leptin urrent, I±V reltionships were performe similrly with slow voltge rmps (5 mv s -1 from -1 to -2 mv) efore n 1 min fter ing leptin (1 nm). GABA-meite IPSs were reore using sl internl eletroe solution ((in mm): 14 sl, 1 HEPES, 5 Mgl 2, 1 BAPTA, 5 Mg-ATP,.3 -GTP). Both mini IPSs n lrge mplitue (presumly multisynpti) IPSs were oserve in the untrete slies. TTX (1 mm) olishe lrge IPSs. We quire t efore n fter rug ition t -5-mV holing potentil in 2-s sweeps every 4 s for the times inite in the gures. Mini-postsynpti urrents were nlyse using Axogrph 4 (Axon Instruments). IPSs n exittory postsynpti urrents (EPSs) were istinguishe on the sis of their ey onstnts; in ition, pirotoxin (1 mm) loke ll IPSs. POM neurons reeive low EPS tone, n the frequeny ws not moulte y ny of the tretments esrie here. Immunostining for light n eletron mirosopy We rrie out oule immunoytohemistry for PY n POM using ifferent olour iminoenziine (DAB) hromogens on xe mouse hypothlmi oring to pulishe protools 27. For eletron mirosopy, pre-emeing immunostining for - enorphin ws one with n AB Elite kit (Vetor Lortories) n DAB retion, followe y post-emeing lelling of GABA n PY using rit nti-gaba, 1:1 (v/v), n gol-onjugte (1-nm) got nti-rit IgG or sheep nti-py n golonjugte (25-nm) got nti-sheep IgG. Setions were ontrste with sturte urnyl ette (1 min) n le itrte (2±3 s), n exmine using Philips M-1 eletron mirosope. Reeive 19 Deemer 2; epte 12 Mrh Zhng, Y. et l. Positionl loning of the mouse oese gene n its humn homologue. ture 372, 425±432 (1994); errtum ii. 374, 479 (1995). 2. Frooqi, I. S. et l. Effets of reominnt leptin therpy in hil with ongenitl leptin e ieny.. Engl. J. Me. 341, 879±884 (1999). 3. mp el, L. A., Smith, F. J., Guisez, Y., Devos, R. & Burn, P. Reominnt mouse OB protein: eviene for peripherl signl linking iposity n entrl neurl networks. Siene 269, 546±549 (1995). 4. Kim, M. S. et l. The entrl melnoortin system ffets the hypothlmo-pituitry thyroi xis n my meite the effet of leptin. J. lin. Invest. 15, 15±111 (2). 5. Hynes, W. G., Morgn, D. A., Wlsh, S. A., Mrk, A. L. & Sivitz, W. I. Reeptor-meite regionl symptheti nerve tivtion y leptin. J. lin. Invest. 1, 27±278 (1997). 6. Hynes, W. G., Morgn, D. A., Djlli, A., Sivitz, W. I. & Mrk, A. L. Intertions etween the melnoortin system n leptin in ontrol of symptheti nerve trf. Hypertension 33, 542±547 (1999). 7. Hknsson, M. L., Brown, H., Ghilri,., Sko, R.. & Meister, B. Leptin reeptor immunoretivity in hemilly e ne trget neurons of the hypothlmus. J. eurosi. 18, 559±572 (1998). 8. Klr, S. P. et l. Interting ppetite-regulting pthwys in the hypothlmi regultion of oy weight. Enor. Rev. 2, 68±1 (1999). 9. one, R. D. The entrl melnoortin system n energy homeostsis. Trens Enorinol. Met. 1, 211±216 (1999). 1. Elis,. F. et l. Leptin ifferentilly regultes PY n POM neurons projeting to the lterl hypothlmi re. euron 23, 775±786 (1999). 11. Glum, S. R. et l. Leptin, the oese gene prout, rpily moultes synpti trnsmission in the hypothlmus. Mol. Phrmol. 5, 23±235 (1996). 12. Spnswik, D., Smith, M. A., Groppi, V. E., Logn, S. D. & Ashfor, M. L. Leptin inhiits hypothlmi neurons y tivtion of ATP-sensitive potssium hnnels. ture 39, 521±525 (1997). 13. Lee, K., Dixon, A. K., Rihrson, P. J. & Pinnok, R. D. Gluose-reeptive neurones in the rt ventromeil hypothlmus express KATP hnnels ompose of Kir6.1 n SUR1 suunits. J. Physiol. (Lon.) 515, 439±452 (1999). 14. Shirishi, T., Sski, K., iijim, A. & Oomur, Y. Leptin effets on feeing-relte hypothlmi n peripherl neuronl tivities in norml n oese rts. utrition 15, 576±579 (1999). 15. Bgnol,D.et l. Antomy of n enogenous ntgonist: reltionship etween Agouti-relte protein n proopiomelnoortin in rin. J. eurosi. [online] (ite 25 Aug. 99), org/gi/ontent/full/19/18/r26. (1999). 16. Butler, A. A. et l. A unique metoli synrome uses oesity in the melnoortin-3 reeptore ient mouse. Enorinology 141, 3518±3521 (2). 17. Young, J. I. et l. Authenti ell-spei n evelopmentlly regulte expression of pro-opiomelnoortin genomi frgments in hypothlmi n hinrin neuronsoftrnsgeni mie. J. eurosi. 18, 6631±664 (1998). 18. Frnklin, K. B. J. & Pxinos, G. The Mouse Brin in Stereotxi oorintes (Aemi, Sn Diego, 1997). 19. Kelly, M. J., Loose, M. D. & Ronnekleiv, O. K. Opiois hyperpolrize -enorphin neurons vi m-reeptor tivtion of potssium onutne. euroenorinology 52, 268±275 (199). 2. Slugg, R. M., Hywr, M. D., Ronnekleiv, O. K., Low, M. J. & Kelly, M. J. Effet of the m-opioi gonist DAMGO on meil sl hypothlmi neurons in -enorphin knok-out mie. euroenorinology 72, 28±217 (2). 21. Powis, J. E., Bins, J. S. & Ferguson, A. V. Leptin epolrizes rt hypothlmi prventriulr nuleus neurons. Am. J. Physiol. 274, R1468±R1472 (1998). 22. Horvth, T. L., Behmnn, I., ftolin, F., Klr, S. P. & Lernth,. Heterogeneity in the neuropeptie Y-ontining neurons of the rt rute nuleus: GABAergi n non-gabaergi supopultions. Brin Res. 756, 283±286 (1997). 23. Broerger,., Lnry, M., Wong, H., Wlsh, J.. & Hokfelt, T. Sutypes Y1 n Y2 of the neuropeptie Y reeptor re respetively expresse in pro-opiomelnoortin- n neuropeptie-yontining neurons of the rt hypothlmi rute nuleus. euroenorinology 66, 393±48 (1997). 24. King, P. J., Wiowson, P. S., Doos, H.. & Willims, G. Regultion of neuropeptie Y relese y neuropeptie Y reeptor ligns n lium hnnel ntgonists in hypothlmi slies. J. eurohem. 73, 641±646 (1999). 25. Grieo, P., Blse, P. M., Weinerg, D., Meil, T. & Hruy, V. J. D-Amino i sn of gmmmelnoyte-stimulting hormone: importne of Trp(8) on humn M3 reeptor seletivity. J. Me. hem. 43, 4998±52 (2). 26. siffry, A., Gors, T. J. & Plkovits, M. europeptie Y innervtion of ATH-immunoretive neurons in the rute nuleus of rts: orrelte light n eletron mirosopi oule immunoleling stuy. Brin Res. 56, 215±222 (199). 27. Horvth, T. L., ftolin, F. & Lernth,. GABAergi n teholminergi innervtion of meiosl hypothlmi et-enorphin ells projeting to the meil preopti re. eurosiene 51, 391±399 (1992). 28. owley, M. A. et l. Integrtion of PY, AGRP, n melnoortin signls in the hypothlmi prventriulr nuleus: eviene of ellulr sis for the ipostt. euron 24, 155±163 (1999). 29. Hls, J. L. et l. Physiologil response to long-term peripherl n entrl leptin infusion in len n oese mie. Pro. tl A. Si. USA 94, 8878±8883 (1997). Aknowlegements We wish to thnk V. J. Hruy for the D-Trp 8 -gmsh, O. K. Ronnekliev, R. G. Allen n M. R. Brown for ntiser n J. T. Willims n J. M. Brunege for vie. This work ws supporte y the IH, Fogrty Interntionl Reserh ollortive Awr, the Interntionl Sholr Progrm of the Howr Hughes Meil Institute, n Ageni ionl e PromoioÂn ienti y TehnoloÂgi. orresponene n requests for mterils shoul e resse to R.D.. (e-mil: one@ohsu.eu) or M.J.L. (e-mil: low@ohsu.eu).... lmoulin ifurtes the lol 2+ signl tht moultes P/Q-type 2+ hnnels rl D. DeMri*, Tuk Wh Soong², Br A. Alseikhn*, Ree S. Alvni* & Dvi T. Yue* * The Johns Hopkins University Shool of Meiine, Deprtments of Biomeil Engineering n eurosiene, Progrm in Moleulr n ellulr Systems Physiology, 72 Rutln Avenue, Bltimore, Mryln 2125, USA ² tionl eurosiene Institute, 11 Jln Tn Tuk Seng, Singpore 38433, n Deprtment of Physiology, tionl University of Singpore... Aute moultion of P/Q-type ( 1A ) lium hnnels y neuronl tivity-epenent hnges in intrellulr 2+ onentrtion my ontriute to short-term synpti plstiity 1±3, potentilly enrihing the neuroomputtionl pilities of the rin 4,5. An unonventionl mehnism for suh hnnel moultion hs een propose 6,7 in whih lmoulin (M) my exert two opposing effets on iniviul hnnels, initilly promoting (`filittion') n then inhiiting (`intivtion') hnnel opening. Here we report tht suh ul regultion rises from surprising 2+ -trnsution pilities of M. First, lthough filittion n intivtion re two ompeting proesses, oth require 2+ -M ining to single `-like' omin on the roxy til of 8 1A ; previously ienti e `' M-ining site 6,7 hs no etetle role. Seon, expression of M mutnt with impirment of ll four of its 2+ -ining sites (M 1234 ) elimintes oth forms of moultion. This result on rms tht M is the 2+ sensor for hnnel regultion, n inites tht M my ssoite with the hnnel even efore lol 2+ onentrtion rises. Finlly, the ifuntionl pility of M rises from ifurtion of 2+ signlling y the loes of M: 2+ ining to the mino-terminl loe seletively initites hnnel intivtion, wheres 2+ sensing y the roxyterminl loe inues filittion. Suh loe-spei etetion provies ompt mens to eoe lol 2+ signls in two wys, n to seprtely initite istint tions on single moleulr omplex. To simplify the issetion of the moleulr mehnisms, we stuie reominnt P/Q-type ( 1A / 2 / 2 ) hnnels expresse in mmmlin HEK293 ells. Figure 1 shows tht 2+ -epenent Mmilln Mgzines Lt ATURE VOL MAY 21

2 5 mm st spike 1 ms 1,4 ms (test pulse) 5 ms +1 mv B 2+ urrents evoke y APW trin (1 Hz) 1 n +2 (prepulse) r 8 n = 6 f =.28 ±.3 +5 (test pulse) 9 mv Intivtion 5 Test pulse (mv) f B filittion n intivtion in suh reominnt hnnels repitulte moultory ehviour omptile with tht of presynpti hnnels 1,2. To mimi physiologil responses, we tivte reominnt hnnels using trins of tion-potentil wveforms. The resulting 2+ urrents filitte with repetitive spikes, n then intivte over longer timesle. Suh ehviour ws sent in orresponing B 2+ urrents (see Supplementry Informtion), tting with the high seletivity of M 9 for 2+ over B 2+. To quntify the intivtion, we use prolonge squre-pulse epolriztions (Fig. 1, left), in whih 2+ urrent eye lmost ompletely wheres the orresponing B 2+ urrent remine sustntil. On verge (Fig. 1, right), the frtion of pek 2+ urrent present fter epolrizing for 8 ms (r 8 ) ore eep, U-shpe epenene upon test-pulse voltge, onsistent with genuine 2+ -epenent intivtion 8. The orresponing B 2+ reltion eline only moestly, re eting slower voltge-epenent mehnism. Hene, the ifferene etween 2+ n B 2+ r 8 (f) provie roust inex of pure 2+ -epenent intivtion. Using shorter pulses with negligile intivtion (Fig. 1, left top), filittion ws reily resolve s slower phse of 2+ urrent inrese (), following n initil rpi tivtion (). The phenomenon oul e unerstoo s fst tivtion in norml moe of gting (Fig. 1, ottom, pthwy ), followe y slower, 2+ -riven onversion to filitte gting moe with enhne open proility (pthwy ). Suh senrio ws on rme in three wys. First, voltge prepulse shoul `prefilitte' the orml P o P/Q-type (α 1A ) L-type (α 1 ) ± prepulse Filitte P o.75 Filitte O O orml I II III IV α 1A( ) 9 I +2 (prepulse) Filittion +5 mv (test pulse) 5 ms I II III IV I Intivtion +1 (test pulse) 2+ Q = normlize hrge ifferene ± prepulse 5 ms B 2+ e.25 Reltive filittion 2+ g 5 5 Prepulse (mv) g =.21 ±.2 n = 9 B 2+ Figure 1 Filittion n intivtion in P/Q-type 2+ hnnels., 2+ urrents show filittion (irle) n intivtion (squre) uring trins of tion potentil wveforms (APW)., Left, intivtion of 2+ (grey) n B 2+ (lk) urrents uring step epolriztion. 2+ tres for intivtion mpli e out twie to mth B 2+ tres, n til urrents lippe t orers. B 2+ sle, 1 na. Right, r 8, verge from n ells. f ws tken t 1 mv., 2+ -epenent filittion in prepulse protools. Left, test pulse without prepulse. Right, test pulse following prepulse, with rpi tivtion to lrger filitte level. 2+ sle, 1 na. P O, stey-stte open proility. Sme ell s in ws use. Bottom, interprettion of phses ±., Initil rpi tivtion;, slower phse of inrese;, the hnnels lrey in the filitte moe rpily tivte., Qunti tion of filittion. Top, test-pulse 2+ urrents otine without (grey) n with (lk) prepulse, from fter normliztion to unity t test-pulse termintion. DQ, integrl of ifferene etween normlize tres. Reltive filittion RF ˆ DQ=t ˆ :33 for upper tres, where t is the filittion time onstnt (,1 ms) (see Methos). Bottom, testpulse B 2+ urrents, sme ell (RF ˆ :11). Sle, 2 na. B 2+ tres sle to re et lower open proility of norml moe. e, Averge reltive filittion. g tken t 2 mv. ±e, See Supplementry Informtion. 1 IV α 1A(IM/EE) IV α 1A(IM/AA) IMEYYRQSKAK EE xxxrgxxxr IV IMEYYRQSKAK AA xxxrgxxxr + prepulse 2+ prepulse g=.23 ±.4 (n = 6) +2 (test pulse) g=.2 ±.2 (n = 5) +5 (test pulse) g=.1 ±.1 (n = 7) 2+ B 2+ f =.29 ±.2 (n = 5) 3 ms +3 (test pulse) f =.1 ±.1 (n = 5) 2+ B (test pulse) f =.3 ±.5 (n = 5) Figure 2 Struturl eterminnts of filittion n intivtion., P/Q-type n L-type hnnel pore-forming 1 -suunits. hnnel omins (I±IV). Proximl thir of -til (I region) ontins onsensus -hn (), motif () n omin. ±, Formt essentilly s in Fig. 1, exept. Sles, 1 na, exept prepulse in ( na)., onsensus 8 shown for referene. Higher hrge rrier (Methos) neessitte ifferent test-pulse voltges. Fster intivtion require shorter intivtion test pulses, n use of r 1 (frtion of pek urrent fter 1-ms epolriztion) to quntify intivtion. f, ifferene in B 2+ n 2+ r 1 t 3 mv. ±, See Supplementry Informtion. ATURE VOL MAY Mmilln Mgzines Lt 485

3 hnnels, suh tht susequent test pulse woul rpily tivte the hnnels lrey in the filitte gting moe (Fig. 1, right n ottom, pthwy ). The overly of normlize test-pulse 2+ urrents (otine 6 prepulse) unersores suh prefilittion (Fig. 1, top). Seon, test-pulse B 2+ urrents shoul tivte y single, rpi omponent regrless of prepulse epolriztion (Fig. 1, ottom), euse of hnnel `trpping' in the norml moe. Finlly, if 2+ entry rives filittion, then prefilittion shoul er ell-shpe epenene upon prepulse potentil (Fig. 1e). This ws on rme y the verge reltion etween reltive filittion n prepulse voltge. Here, the reltive filittion is proportionl to the frtion of hnnels filitte y prepulse, s reltive filittion is erive from the ifferene in hrge (DQ; Fig. 1) rrie y normlize test-pulse urrents (eliite 6 prepulse; Fig. 1). The orresponing B 2+ reltive filittion (Fig. 1e) showe only smll monotoni inrese, re eting kgroun G-protein moultion 1. Hene, the ifferene etween 2+ n B 2+ reltive filittions (g; Fig. 1e) gve mesure of pure 2+ -epenent filittion. How oes M initite suh kinetilly isprte proessesð filittion n intivtionðon single trget moleule? One possiility is tht 2+ -M ins to two hnnel sites, one for eh moultory effet. Although the M-ining omin () site on the istl roxyl til (-til) of 1A reportely meite oth filittion n intivtion 6,7 (Fig. 2, left), M lso ins to n upstrem -like motif 8. Though the funtion of the 1A site ws unknown, M ining to the homologous L-type hnnel site (Fig. 2, right) is ruil to 2+ regultion 8,11±13, n the enompssing I (lium intivtion) region my trnsue 2+ -M ining to L-type hnnel moultion 14. To test for funtionl role of sites outsie the, we trunte the 1A -til so s to entirely remove the region ( 1A(D ; Fig. 2). Unexpetely, filittion n intivtion were not ppreily hnge ompre with the wil-type 1A, impliting other M-ining sites in hnnel regultion. To test the importne of the 1A site, we mutte the ritil isoleuine n methionine resiues of the -like motif 15 to negtively hrge glutmtes ( 1A(IM/EE) ; Fig. 2), in n ttempt to isrupt 2+ -M intertion. These muttions entirely olishe filittion n intivtion, suggesting tht oth regultory effets re initite through 2+ -M intertion with the site. The resiul voltge-epenent intivtion, ientil with either 2+ or B 2+, ws elerte out 4-fol. This t with the ul roles of trnsution n M ining serve y mny M-intertion omins 16 (see Supplementry Informtion). We lso exmine more moest muttions to neutrlly hrge lnines ( 1A(IM/AA) ; Fig. 2), in n ttempt to preserve prtil 2+ -M intertion. Suh muttions eliminte filittion, ut spre 2+ -epenent intivtion. The presume effets of hnnel muttions on 2+ -M intertion were veri e y gel-moility shift ssys, in whih mixtures of M n pepties spnning the 1A region were eletrophorese on nonenturing gels (Fig. 3±). Without 2+, M rn s single, peptie-free n, regrless of the onentrtion of the wil-type peptie WT (Fig. 3, left). By ontrst, with 2+, inresing peptie onentrtions inue gre interhnge towrs slower moility n (Fig. 3, right), representing peptie± 2+ -M omplex. As we oserve previously 8, the moility shift sturte t low peptie:m rtios, initing sustntil 2+ -M f nity for the peptie. By ontrst, the IM/ EE peptie (Fig. 3) showe no inition of peptie±m intertion, wheres the IM/AA peptie (Fig. 3) showe prtil moility shift, initing moest intertion with 2+ -M. Fluoresene experiments with nsyl-m 17 (Fig. 3) quntittively on rme this rnk orer of peptie± 2+ -M intertion. In sturting 2+, WT oun M with high f nity (K ˆ 66 nm), IM/AA oun with intermeite f nity (K ˆ 1:3 mm) n IM/EE showe no ppreile intertion. These iohemil stuies (Fig. 3) on rme the presumptions of funtionl muttionl nlysis in Fig. 2 (tht IM/EE woul eliminte, n IM/AA woul weken 2+ -M intertion), leing us to rgue tht 2+ -M ining to the 1A site lone meites oth filittion n intivtion. The initil question out how the intertion of M with single site proues two moultory effets ws then reple y similr puzzle involving the site. The pproh for solving this prox me y rst onsiering the fst kinetis of filittion (t < 1 ms, Fig. 1). Suh rpiity seems inomptile with the iffuse ytoplsmi loliztion typil of M. Might pom ( 2+ -free M) e oun to the hnnel omplex even efore 2+ entry (`pressoition'), thus elerting ( 2+ ) WT :M 4. ( 2+ ) IM/EE :M 4. ( 2+ ) IM/AA :M 4. Dnsyl-M frtion oun α 1A - WT WT : KIYAAMMIMEYYRQSKAKKLQ (xxxrgxxxr) ( ) α 1A IM/EE α 1A IM/AA (+ 2+ ) WT :M : KIYAAMMEEEYYRQSKAKKLQ EYYRQSKAKKLQ (xxxrgxxxr) ( ) (+ 2+ ) IM/EE :M : KIYAAMMAAEYYRQSKAKKLQ AAEYYRQSKAKKLQ (xxxrgxxxr) ( ) WT (+ 2+ ) IM/AA :M Peptie IM/AA IM/EE Figure 3 M ining to 1A site. ±, Gel-moility shift ssy of M ining to pepties (insets) spnning the 1A site. Without 2+ (±, left), lk of M moility shift. With 2+ (±, right), inresing WT :M rtios () use exhnge towrs slower-moility (peptie- 2+ -M) n (tringle); IM/EE inue no shift in 2+ (), n IM/AA proue prtil shift in 2+ ()., Dnsyl-M uoresene hrteriztion of 2+ -M ining to pepties (see Methos). Smooth urves, single-ining isotherm ts Mmilln Mgzines Lt ATURE VOL MAY 21

4 susequent 2+ -M intertion with the site? To exmine this question we oexpresse P/Q-type hnnels with M 1234 (Fig. 4, top), 2+ -insensitive M mutnt with sprtte-to-lnine muttions t the `x' position of ll four hns 8,18. M 1234 woul eliminte hnnel regultion if the M responsile for moultion ws preoun to the 1A omplex with preferentil ess to the site; otherwise, funtionl enogenous M shoul preserve norml hnnel regultion 8. The oserve elimintion of filittion n intivtion (Fig. 4) inites pressoition of pom, n unequivolly on rms tht M is the 2+ sensor for regultion. The ominnt-negtive tion of M 1234 llowe us to return to the question of how 2+ -M ining to single site initites ul moultory proesses. Mutnt Ms with seletive impirment of 2+ ining to either the -terminl (M 12 )orterminl loe (M 34 ) (Fig. 4,, left) provie the key insight. Expression of M 12, in whih the sprtte-to-lnine muttions esrie for M 1234 were restrite to the -terminl loe, strikingly suppresse 2+ -epenent intivtion (Fig. 4), wheres filittion ws inistinguishle from the ontrol (Fig. 1e). Expression of M 34 (Fig. 4), in whih the nlogous muttions were restrite to the -terminl loe, yiele the opposite results: 2+ -epenent filittion ws essentilly sent, wheres intivtion remine unhnge ompre with the ontrol (Fig. 1). These unexpete nings were on rme y itionl popultion t (Fig. 4) tht gve the inies for the strength of intivtion (f ) n filittion (g); these were poole from even more ells thn were use for the full voltge protools. We therefore onlue tht 2+ ining to the -terminl loe of M seletively initites intivtion, wheres 2+ ining to the -terminl loe seletively triggers filittion. Our experiments highlight signlling pilities of M tht my generlize to numerous iologil systems. First, M regultion through the -like site of 1A s to n emerging theme, s homologous site in L-type ( 1 ) 2+ hnnels unerlies their regultion y M 8,11±14, n nlogous -like motifs in R-type ( 1E ) 2+ hnnels 8 n hnnels 19 intert with M. Seon, pressoition of pom with moleulr omplexes my e prevlent mehnism to spee their moultion y 2+ -M. Both L-type 2+ hnnels 8,11 n smll-onutne K hnnels 2,21 my exploit suh sheme. Moreover, FRET ( uoresene resonne energy trnsfer) experiments in our lortory inite 9 M (prepulse) +5 mv (test pulse) Reltive filittion 5 ms g =.4 ±.1 n = 8 +1 (test pulse) B r 8 Intivtion n = 6 f =.1 ± Prepulse (mv) 5 5 Test pulse (mv) M g =.24 ±.2 n = 9 B r 8 n = 7 f =.2 ± M g =.5 ±.1 n = r 8 n = 5 f =.28 ± e Filitte open Filitte intivte M 2+ o 2+ oun M 1234 M 12 (8) (8) M 34 (7) Filittion (g) Intivtion (f) (7) Resting orml open orml intivte Figure 4 Seletive initition of filittion n intivtion y ifferent loes of M. ±, Formt s in Fig. 1. Prepulse n B 2+ intivtion sle, 1 na., o-expression of M 1234 with hnnels., o-expression of M 12 (efetive -terminl loe) with hnnels., o-expression of M 34 (efetive -terminl loe) with hnnels., Summrize effets of vrious Ms. M inites enogenous M without reominnt expression. For eh M, the sme set of ells ws use for filittion n intivtion. e, Propose mehnism of P/Q-type hnnel regultion. ±e, See Supplementry Informtion. ATURE VOL MAY Mmilln Mgzines Lt 487

5 pressoition of M to multiple types of 2+ hnnels in living ells 22, n pom ining sites hve een ienti e in hnnels 19 n rynoine reeptors 23. Thir, our results show tht M, ting through single ining region, n use loe-spei 2+ ining to seletively ontrol istint proesses, perhps s propose for P/Q-type hnnels (Fig. 4e). ApoM pressoites with resting hnnels t sturle site tht provies preferentil ess to the region. Depolriztion opens hnnels, n the ensuing 2+ ining to the -terminl loe of M enles one form of intertion with the site, whih inues filitte hnnel onformtions. 2+ ining to the -terminl loe enles seon form of intertion tht fvours intivte onformtions. Seletive elimintion of intivtion or filittion y M 12 n M 34 (Fig. 4) shows tht the proesses re istint. Suh mehnism oul provie intriguing esign vntges, espeilly euse it seems likely tht the loes of M seletively etet fetures of the lol 2+ signl tht rise from lol n istnt soures of 2+ (Fig. 5). The -terminl loe my respon preferentilly to the spike-like omponent of lol 2+ onentrtion tht is ttriutle to lol hnnel tivity (-loe M reout). The -terminl loe my etet the slow omponent of the signls whih result from ggregte ellulr 2+ signlling (loe M reout) (see Supplementry Informtion). There re suggestive iohemil preeents for loe-spei M signlling 16,24, n funtionl preeents where one loe of M preferentilly regultes Prmeium 25 n mmmlin 8,2 hnnels, s well s yest physiology 26. However, the P/Q-type hnnel my provie the rst exmple of M loe-spei initition of two regultory effets on single moleulr trget omplex. If wiespre, suh ifurtion of lol 2+ signls y M my e ritil rtionle for the two-loe esign of lmoulin. M Lol hnnel i (t) Lol [ 2+ ] Lol [ 2+ ] + EGTA -loe M reout -loe M reout s Figure 5 Deoing lol 2+ onentrtions ([ 2+ ]). Dotte lines, zero. Lol hnnel i(t ), shemtize urrents of one 2+ hnnel. Theoretil lol [ 2+ ] ner hnnel 3, re eting tivity of ssoite hnnel (spikes) n istnt 2+ soures (slow phse). Lol [ 2+ ] + EGTA, postulte lol 2+ with severl mm ytoplsmi EGTA, feturing ttenution of slow phse with spring of spikes 3. -loe M reout, spike omponent. Deution from spring of filittion with mm EGTA 7, filittion epenene on M -terminl loe, n thir tre. -loe M reout, slow phse. Deution from intivtion lunting y mm EGTA 7, intivtion epenene on M - terminl loe, n thir tre. Approximte timese. See Supplementry Informtion. Methos Moleulr iology Muttions of the 1A -like region 8 (Fig. 2, IMEYYRQSKAK) were me y PR, with humn 1A s templte 27.For 1A(IM/EE) n 1A(IM/AA) mutnts (Fig. 2, ), forwr mutgeni oligonuleoties spnne oth the region n BglII site, 2 p upstrem of the referene isoleuine (ol). The reverse oligo h n XI site n sequene omplementry to the 1A stop oon. The resulting 899-se-pir (p) PR prouts were sulone into humn 1A /pda3 (Invitrogen) y BglII±XI. For 1A(D) (Fig. 2), the forwr oligo nnele 44 p upstrem of n XhoI site preeing IVS6. The reverse oligo ws omplementry to the mino-i streth LDP, 51 mino is ownstrem from the referene isoleuine. This ws followe y premture stop oon n XI site. The resulting 779-p PR frgment ws sulone into humn 1A /pda3 y XhoI±XI. All PR prouts were entirely sequene. Ms were lone into pda3 (ref. 8). Gel-moility-shift ssy M gel-shift ssys were performe with 15% non-enturing polyrylmie gels mostly s esrie 8.In 2+ -free experiments, 5 mm EGTA ws e to retion uffers n 2 mm EGTA ws inlue in the running uffer, ut no heltor or exogenous 2+ ws e uring gel sting. In experiments with 2+, 1±2 mm 2+ ws present in the retion uffer n.1 mm 2+ ws inlue in oth the running uffer n gel (uring sting). Dnsyl-M stuies Puri e reominnt M protein ws erivtize 28 with nsyl (5-imethylminonphthlene-1-sulphonyl) hlorie (Moleulr Proes). The nsyl-m ws then ilyse ginst 2 mm MOPS (ph 7.2). During ining ssys, 1 nm nsyl-m ws mixe with vrious onentrtions of 1A - WT peptie (Fig. 3), in room-temperture uffer 17 ontining 2 mm l 2, 15 mm l n 5 mm Tris-l (ph 7.5). omprle results were otine with 15 mm l 2 (not shown). Fluoresene emission (4± 65 nm, 2 nm nwith) ws monitore y spetro uorometer (SPF-5, SLM Instruments), using 34 nm exittion (2.5 nm nwith). Bkgroun-sutrte uoresene emission t 49 nm (F 49 ) spei e the pprent frtion of M oun to peptie (B pp ), with the reltion B pp ˆ F 49 2 F 49 no peptieš = F 49 1 mm peptieš -F 49 no peptieš. This lgorithm exploits peptie-epenent lue shift n enhnement of nsyl emission spetr 17. The reltions etween B pp n [ 1A - WT peptie] were lest-squres t with the reltion B pp ˆ B pp;mx = K 1A - WT peptieš, n the tul frtion of M oun to peptie etermine s B ˆ B pp =B pp;mx (Fig. 3). We pplie n nlogous proess to the reltion etween B pp n [ IA - IMIAA peptie]. For the 1A - IM/EE peptie, the eqution for B pp use 1-mM peptie mesurement (inste of 1-mM). The pepties were from HHMI Biopolymer Lortory, Johns Hopkins University Shool of Meiine. Eletrophysiology The methos here were mostly s esrie 8. DA for wil-type (or mutnt) humn 1A (ref. 27) ws trnsiently otrnsfete with 2 n 2 (ref. 8) (n vrious Ms 8 s require) in HEK293 ells. Two to three ys lter, whole-ell reorings were otine t room temperture. 2 minimize voltge intivtion 29, enhning resolution of 2+ - epenent regultion. Bth solution ontine (in mm): TEA-MeSO 3, 14; HEPES (ph 7.3), 1; n l 2 or Bl 2, 5; t 3 mosm, juste with gluose. Internl solution (in mm): s-meso 3, 135; sl 2, 5; EGTA, ; Mgl 2, 1; Mg±ATP, 4; n HEPES (ph 7.3), 1; t 29 mosm, juste with gluose. To enhne the resolution of 1A(IM/EE) (Fig. 2), th solution ontine 2 mm Bl 2 or l 2 n 119 mm TEA-MeSO 3. Lrger surfepotentil shift use y 2 mm hrge rrier require inrese test-pulse voltges (Fig. 2). The protools were otherwise the sme s for the experiments with 5 mm hrge rrier. For tion potentil wveforms (APWs) (Fig. 1), voltges were preorrete for -11 mv juntion potentil 1, with sels me in B 2+ th solution; otherwise, reporte voltges were unorrete, n true voltge my e otine y sutrting 11 mv from reporte vlues. APWs sle uniformly from those reore in lyx of Hel 29, here with 2-ms hlf with n voltge rnge from -8 to 34 mv. urrents were ltere t 2 khz n smple t 1 khz, exept in APW experiments (5 khz lowpss, 25 khz smpling). Series resistne ws typilly 1±2 MQ fter more thn 7% ompenstion. Leks n pitive trnsients were sutrte y P/8 protool. Test-pulse epolriztions were elivere every 6 s (filittion protool) or 1 s (intivtion protool). 2+ -epenent filittion ws etermine using the normlize hrge ifferene DQ, otine y integrting the ifferene etween normlize tres 6 prepulse (Fig. 1, top). The frtion of hnnels filitte y prepulse (F filitte ) is iretly proportionl to DQ ivie y the time onstnt (t) of filittion, yieling reltive filittion (RF ˆ DQ=t). This follows y ssuming tht ll hnnels re initilly in norml moe t test-pulse onset, n tht susequent shifts to filitte moe our mono-exponentilly with time onstnt t. Then, RF ˆ F filitte 3 P o; filitte 2 P o; norml Š=P filitte, where P o, filitte n P o, norml re stey-stte open proilities in filitte n norml moes, respetively. t ws expliitly etermine from 2+ tres in eh ell efore lultion of RF. B 2+ RF lulte y using t vlues etermine from 2+ tres in the sme ell. For knokouts of filittion (Figs 2,, 4, ), t ws set to 1 ms (t out the verge for ontrol filittion, Fig. 1) in RF lultions. All verge t were presente s men 6 s.e.m., fter nlysis y ustom-written softwre in MATLAB (MthWorks). Smooth-urve ts to t were one y eye, exept in Fig. 3. In Fig. 4, n importnt feture ws to inlue only ells in whih oth g n f inies for filittion n intivtion were etermine. This onstrint exlue the Mmilln Mgzines Lt ATURE VOL MAY 21

6 possiility tht seletion is might hve le to n pprent ifferentil elimintion of regultory mehnisms. Suh n rteft oul hve ourre if mutnt Ms vrily inhiite filittion, intivtion, or othðepening on the prtiulr ell uner oservtion. Approprite hnges (or lk thereof) in f n g vlues, s etermine from the sme ells, exlue suh senrio (Fig. 4; see Supplementry Informtion). Reeive 14 Deemer 2; epte 3 Mrh Borst, J. G. & Skmnn, B. Filittion of presynpti lium urrents in the rt rinstem. J. Physiol. (Lon.) 513, 149±155 (1998). 2. uttle, M. F., Tsujimoto, T., Forsythe, I. D. & Tkhshi, T. Filittion of the presynpti lium urrent t n uitory synpse in rt rinstem. J. Physiol. (Lon.) 512, 723±729 (1998). 3. Forsythe, I. D., Tsujimoto, T., Brnes-Dvies, M., uttle, M. & Tkhshi, T. Intivtion of presynpti lium urrent ontriutes to synpti epression t fst entrl synpse. euron 2, 797±87 (1998). 4. Aott, L. F., Vrel, J. A., Sen, K. & elson, S. B. Synpti epression n ortil gin ontrol. Siene 175, 22±224 (1997). 5. Tsoyks, M. V. & Mrkrm, H. The neurl oe etween neoortil pyrmil neurons epens on neurotrnsmitter relese proility. Pro. tl A. Si. USA 94, 719±723 (1997). 6. Lee, A. et l. 2+ /lmoulin ins to n moultes P/Q-type lium hnnels. ture 399, 155± 159 (1999). 7. Lee, A., Sheuer, T. & tterll, W. A. 2+ /lmoulin-epenent filittion n intivtion of P/ Q-type 2+ hnnels. J. eurosi. 2, 683±6838 (2). 8. Peterson, B. Z., DeMri,. D., Aelmn, J. P. & Yue, D. T. lmoulin is the 2+ sensor for 2+ - epenent intivtion of L-type lium hnnels. euron 22, 549±558 (1999). 9. ho, S. H., Suzuki, Y., Zysk, J. R. & heung, W. Y. Ativtion of lmoulin y vrious metl tions s funtion of ioni rius. Mol. Phrmol. 26, 75±82 (1984). 1. olerft, H. M., Ptil, P. G. & Yue, D. T. Differentil ourrene of relutnt openings in G-proteininhiite - n P/Q-type lium hnnels. J. Gen. Physiol. 115, 175±192 (2). 11. Zuhlke, R. D., Pitt, G. S., Deisseroth, K., Tsien, R. W. & Reuter, H. lmoulin supports oth intivtion n filittion of L-type lium hnnels. ture 399, 159±162 (1999). 12. Zuhlke, R. D., Pitt, G. S., Tsien, R. W. & Reuter, H. 2+ -sensitive intivtion n filittion of L-type 2+ hnnels oth epen on spei mino i resiues in onsensus lmoulin-ining motif in the 1 suunit. J. Biol. hem. 275, 21121±21129 (2). 13. Qin,., Olese, R., Brnsy, M., Lin, T. & Birnumer, L. 2+ -inue inhiition of the ri 2+ hnnel epens on lmoulin. Pro. tl A. Si. USA 96, 2435±2438 (1999). 14. Peterson, B. Z. et l. ritil eterminnts of 2+ -epenent intivtion within n -hn motif of L-type 2+ hnnels. Biophys. J. 78, 196±192 (2). 15. Houusse, A. & ohen,. Trget sequene reognition y the lmoulin superfmily: implitions from light hin ining to the regultory omin of sllop myosin. Pro. tl A. Si. USA 92, 1644±1647 (1995). 16. Elshorst, B. et l. MR solution struture of omplex of lmoulin with ining peptie of the 2+ pump. Biohemistry 38, 1232±12332 (1999). 17. Ehlers, M. D., Zhng, S., Bernhrt, J. P. & Hugnir, R. L. Intivtion of MDA reeptors y iret intertion of lmoulin with the R1 suunit. ell 84, 745±755 (1996). 18. Putkey, J. A., Sweeney, H. L. & mpell, S. T. Site-irete muttion of the trigger lium-ining sites in ri troponin. J. Biol. hem. 264, 1237±12378 (1989). 19. Mori, M. et l. ovel intertion of the voltge-epenent soium hnnel (VDS) with lmoulin: oes VDS quire lmoulin-meite 2+ -sensitivity? Biohemistry 39, 1316±1323 (2). 2.Keen,J.E.et l. Domins responsile for onstitutive n 2+ -epenent intertions etween lmoulin n smll onutne 2+ -tivte potssium hnnels. J. eurosi. 19, 883±8838 (1999). 21. Fnger,. M. et l. lmoulin meites lium-epenent tivtion of the intermeite onutne K hnnel, IKl. J. Biol. hem. 274, 5746±5754 (1999). 22. Erikson, M. G. & Yue, D. T. FRET revels tethering of lmoulin to lium hnnel omplex in single living ells. Biophys. J. 8, 196 (21). 23. Roney, G. G. et l. lium ining to lmoulin les to n -terminl shift in its ining site on the rynoine reeptor. J. Biol. hem. 276, 269±274 (21). 24. Brth, A., Mrtin, S. R. & Byley, P. M. Spei ity n symmetry in the intertion of lmoulin omins with the skeletl musle myosin light hin kinse trget sequene. J. Biol. hem. 273, 2174± 2183 (1998). 25. Kink,J. A.et l. Muttions in prmeium lmoulin inite funtionl ifferenes etween the - terminl n -terminl loes in vivo. ell 62, 165±174 (199). 26. Ohy, Y. & Botstein, D. Diverse essentil funtions revele y omplementing yest lmoulin mutnts. Siene 263, 963±966 (1994). 27. Sutton, K. G., MRory, J. E., Guthrie, H., Murphy, T. H. & Snuth, T. P. P/Q-type lium hnnels meite the tivity epenent feek of syntxin-1a. ture 41, 8±84 (1999). 28. Kini, R. L., Billingsley, M. L. & Vughn, M. Preprtion of uoresent, ross-linking, n iotinylte lmoulin erivtives n their use in stuies of lmoulin-tivte phosphoiesterse n protein phosphtse. Methos Enzymol. 159, 65±626 (1988). 29. Ptil, P. G., Broy, D. L. & Yue, D. T. Preferentil lose-stte intivtion of neuronl lium hnnels. euron 2, 127±138 (1998). 3. Song, L. S., Shm, J. S., Stern, M. D., Lktt, E. G. & hng, H. Diret mesurement of SR relese ux y trking ` 2+ spikes' in rt ri myoytes. J. Physiol. (Lon.) 512, 677±691 (1998). Supplementry informtion is ville on ture's Worl-Wie We site ( or s pper opy from the Lonon eitoril of e of ture. Aknowlegements We thnk W. Agnew,. hen, H. olerft, M. Erikson, S. Tkhshi, H. Agler n E. Soie for isussion; B. Peterson for initil ttempts to etet P/Q-type hnnel filittion; n T. Snuth for the gift of humn 1A lone. This work ws supporte y n IH RSA fellowship (.D.D.) n grnts from the IH (D.T.Y.) n I (T.W.S.). orresponene n requests for mterils shoul e resse to D.T.Y. (e-mil: yue@me.jhu.eu).... B ells quire ntigen from trget ells fter synpse formtion Funo D. Btist, Dgmr Ier & Mihel S. euerger Meil Reserh Lortory of Moleulr Biology, Hills Ro, mrige, B2 2QH, UK... Solule ntigen ins to the B-ell ntigen reeptor n is internlize for susequent proessing n the presenttion of ntigen-erive pepties to T ells 1. Mny ntigens re not solule, however, ut re integrl omponents of memrne; furthermore, solule ntigens will usully e enountere in vivo in memrne-nhore form, tethere y F or omplement reeptors 2±4. Here we show tht B-ell intertion with ntigens tht re immoilize on the surfe of trget ell les to the formtion of synpse n the quisition, even, of memrne-integrl ntigens from the trget. B-ell ntigen reeptor umultes t the synpse, segregte from the D45 oreeptor whih is exlue from the synpse, n there is orresponing polriztion of ytoplsmi effetors in the B ell. B-ell ntigen reeptor meites the gthering of ntigen into the synpse n its susequent quisition, therey potentiting ntigen proessing n presenttion to T ells with high ef y. Synpse formtion n ntigen quisition will proly enhne the tivtion of B ells t low ntigen onentrtion, llow ontext-epenent ntigen reognition n enhne the linking of B- n T-ell epitopes. To investigte the B-ell response to ntigen enountere s prt of n immune omplex tethere to ell surfe, immune omplexes omprising hen-egg lysozyme (HEL) ggregte with spei immunogloulin-g (IgG) monolonl ntioies were loe on to the surfe of n Fg reeptor (FgR)-expressing myeloi ell line. The immune omplexes h pthy istriution over the myeloi ell surfe (Fig. 1, ); however, on inution with ntigenspei B ells, ell ggregtes were forme in whih the immune omplexes on the myeloi ell n the B-ell ntigen reeptor (BR) on the B ell were gthere together into region of synpsis (Fig. 1, ). Similr results were otine using immune omplexes loe onto FgRI-expressing L-ell trnsfetnts. Immunoytohemistry suggeste tht there is gthering of tethere immune omplexes, meite y the BR (possily ie y the oligomeri nture of the BR 5 ) n ompnie y n pprent reorgniztion of the B-ell surfe, s juge y segregtion of the BR from D45 in the region of synpsis (Fig. 1e). This onentrtion of BR is reminisent of the pping of surfe IgM tht results from inution of B ells with polyvlent nti-igm ntiser 6,7. We therefore etermine whether the reorgniztion of the B ell epene on the polyvlent nture of the immune omplex. We generte trnsfetnts isplying HEL ntigen s presumptively monovlent integrl memrne ntigen n use one of these trnsfetnts (J[mHEL]6; Fig. 2) in exess s trget for HEL-spei B ells. onfol mirosopy showe tht, fter 1 min, most B ells were in onjuntion with trget; BR ws onentrte in the region of synpsis ut with ler exlusion of D45 (Fig. 2,, f; Supplementry Informtion movie 1). Reorgniztion of omponents of the B-ell memrne ws lso evient from epletion of D22 from the entre of most synpses (lthough often onentrte t the eges), where there ws onentrtion of gngliosie GM1 (whih is ssoite lssilly with mny Sr-fmily tyrosine kinses). There ws lso polriztion of ytoplsmi omponents, s juge y epletion of the signlinhiitory phosphtse SHP1 in the region of the synpse (Fig. 2±f), ut onentrtion of phosphotyrosine-ontining proteins s well ATURE VOL MAY Mmilln Mgzines Lt 489