Differential modulation of Ca v 2.1 hannels y almodulin and Ca 2+ -inding protein 1 Amy Lee 1, Ruth E. Westenroek 1, Françoise Haeseleer 2, Krzysztof Palzewski 1 3, Todd Sheuer 1 and William A. Catterall 1 Departments of 1 Pharmaology, 2 Opthalmology and 3 Chemistry, University of Washington Shool of Mediine, Seattle, Washington 98195-7280, USA Correspondene should e addressed to W.A.C. (watt@u.washington.edu) Pulished online: 4 Feruary 2002, DOI: 10.1038/nn805 Ca v 2.1 hannels, whih mediate P/Q-type Ca 2+ urrents, undergo Ca 2+ /almodulin (CaM)- dependent inativation and failitation that an signifiantly alter synapti effiay. Here we report that the neuronal Ca 2+ -inding protein 1 (CaBP1) modulates Ca v 2.1 hannels in a manner that is markedly different from modulation y CaM. CaBP1 enhanes inativation, auses a depolarizing shift in the voltage dependene of ativation, and does not support Ca 2+ -dependent failitation of Ca v 2.1 hannels. These inhiitory effets of CaBP1 do not require Ca 2+, ut depend on the CaMinding domain in the α 1 suunit of Ca v 2.1 hannels (α 1 2.1). CaBP1 inds to the CaM-inding domain, o-immunopreipitates with α 1 2.1 from transfeted ells and rain extrats, and oloalizes with α 1 2.1 in disrete mirodomains of neurons in the hippoampus and ereellum. Our results identify an interation etween Ca 2+ hannels and CaBP1 that may regulate Ca 2+ -dependent forms of synapti plastiity y inhiiting Ca 2+ influx into neurons. Calium entry into ells through voltage-gated Ca 2+ hannels initiates a wide range of ellular proesses inluding protein phosphorylation, gene expression and neurotransmitter release 1. Neuronal Ca 2+ hannels onsist of a pore-forming α 1 suunit and auxiliary β, α 2 δ and sometimes γ suunits 2, and their funtion depends onsideraly on interations with additional regulatory fators. For example, the ativation of G-protein-oupled reeptors y neurotransmitters inhiits Ca v 2.1 and Ca v 2.2 hannels, whih mediate P/Q-type and N- type Ca 2+ urrents, respetively, through the inding of G-protein βγ suunits to distint sites on the Ca 2+ hannel α 1 suunit 3 5. These hannels are also inhiited y diret interations with synapti SNARE (solule NSF attahment protein reeptor proteins) a proess that may optimize oupling etween Ca 2+ entry and synapti vesile fusion 6 8. Charaterizing the funtional interations etween Ca 2+ hannels and other signaling moleules is therefore ruial to understanding how many Ca 2+ -dependent proesses in neurons are regulated. We have shown previously that the prominent Ca 2+ sensor CaM inds to a CaM-inding site (CBD) in the aroxy-terminal domain of the α 1 2.1 suunit and mediates the dual feedak regulation of Ca v 2.1 hannels y Ca 2+ ions 9,10. A seond site, loated amino-terminal to the CBD, is analogous to the IQ domain that mediates Ca 2+ /CaM-dependent inativation of Ca v 1 (L-type) hannels 11 13. The IQ domain of α 1 2.1 interats with CaM in vitro and also ontriutes to the regulation of Ca v 2.1 hannels y CaM 14,15. Ca 2+ /CaM mediates oth failitation and enhaned inativation of Ca v 2.1 hannels in transfeted ells during trains of depolarizations 10,15. Presynapti Ca v 2.1 hannels in the rain undergo similar forms of Ca 2+ - dependent modulation that an lead to oth synapti failitation and depression 16 18. Beause Ca v 2.1 hannels are essential to neurotransmitter release at most entral synapses 19,20, regulation y CaM may ontriute widely to mehanisms of ativity-dependent synapti plastiity. Calmodulin is the est haraterized memer of a superfamily of Ca 2+ -inding proteins that exhiit four EF-hand motifs, one or more of whih may e nonfuntional in the oordination of Ca 2+ (ref. 21). Inluded in this superfamily are the neuronal Ca 2+ - inding proteins (NCBPs) that, unlike CaM, are loalized primarily in neurons 22. Some NCBPs an sustitute for CaM in vitro 23,24, whih suggests that NCBPs may regulate effetors that are typially thought to e modulated y CaM. Here we have studied the interation of CaBP1, an NCBP loated in the retina and rain 25, with Ca v 2.1 hannels. We show that CaBP1 inds to the CBD of the α 1 2.1 suunit, ut with properties and funtional onsequenes that are different from those of CaM. Our findings expand the repertoire of modulatory interations that take plae etween Ca 2+ hannels and Ca 2+ -inding proteins and indiate that NCBPs, in addition to CaM, may have a role in the ativitydependent regulation of neuronal Ca 2+ influx. RESULTS CaBP1 interats with the CBD of α 1 2.1 Although CaBP1 is a neuron-speifi Ca 2+ -inding protein that shares nearly 56% amino aid sequene identity with CaM 25, it differs in having a onsensus site for N-terminal myristoylation, an alternatively splied region, inativating amino aid sustitutions in the seond of the four EF-hand motifs, and an extra turn in the helial domain that links the N- and C-terminal loes 210 nature neurosiene volume 5 no 3 marh 2002
a (Fig. 1a). To determine whether CaBP1 an sustitute for CaM in interations with Ca v 2.1 hannels, we tested the aility of CaBP1 to interat with various intraellular domains of the α 1 2.1 suunit. In yeast two-hyrid assays, CaBP1 ativated transription of HIS3 and laz reporter genes only in yeast that had een otransformed with α 1 2.1 onstruts that inluded the CBD (Fig. 1 and ). CaBP1 did not interat with the IQ domain or with a ontrol plasmid that laked the CaBP1 oding region. These results indiated that CaBP1 may modulate Ca v 2.1 hannel funtion through interations with the CBD. To onfirm that CaBP1 assoiated with the CBD in the intat hannel, we tested whether CaBP1 o-immunopreipitated with Ca v 2.1 hannels from otransfeted tsa-201 ells. In this assay, CaM o-immunopreipitates with α 1 2.1 in a Ca 2+ -dependent manner only when the ells are exposed to Ca 2+ ionophore 9. Under these onditions CaBP1 also oimmunopreipitated with α 1 2.1. When Ca 2+ was uffered with 10 mm EGTA, however, the assoiation of CaBP1 with the hannel was not affeted (Fig. 2a). This o-immunopreipitation of CaBP1 was speifi eause CaBP1 was not immunopreipitated with ontrol IgG or with α 1 2.1-speifi antiodies Fig. 1. CaBP1 inds speifially to the CBD of α 1 2.1. (a) Diagram of CaBP1 and CaM. The four Ca 2+ -inding EF-hand motifs are shown as oxes, and key strutural differenes etween CaBP1 and CaM are indiated y arrows. () Diagram of the rat rain α 1 2.1 suunit (ra) showing the intraellular domains that were tested for interation with CaBP1 in yeast two-hyrid assays. The amino aid oundaries of the indiated onstruts are given in parentheses. () β-galatosidase assays of yeast otransformed with the α 1 2.1 onstruts shown in () and either CaBP1 or ontrol vetor (pact2). in ells transfeted with CaBP1 alone, and CaBP1 did not oimmunopreipitate with α 1 2.1 suunits that laked the CBD. Thus, despite its Ca 2+ independene, the interation etween CaBP1 and Ca v 2.1 hannels requires the same intraellular domain of Ca v 2.1 that inds CaM. CaBP1 assoiates with neuronal Ca v 2.1 hannels To determine whether CaBP1 assoiated with endogenous Ca v 2.1 Ca 2+ hannels, o-immunopreipitation experiments were done with extrats from rat ereellum, whih ontains high onentrations of α 1 2.1 and CaBP1 mrna 25,26. Immunolots of CaBP1 showed two proteins (28 and 36 kda) that speifially oimmunopreipitated with α 1 2.1 (Fig. 2, left) ut not with ontrol IgG (Fig. 2, middle). The 36-kDa speies might represent aldendrin, a larger isoform of CaBP1 that is produed from alternative spliing 25,27. The 28-kDa speies was onsistent in size with the predited moleular mass of the long CaBP1 isoform that we used in transfeted ells (Fig. 2, right), in support of a physiologial interation etween neuronal Ca v 2.1 hannels and CaBP1. To identify the potential ellular sites of interation etween CaBP1 and α 1 2.1, we immunostained rat rain setions with antiodies speifi for oth proteins. Compared with the immunostaining of α 1 2.1, the immunostaining for CaBP1 was generally far more restrited within the rain and more ommonly assoiated with somatodendriti regions than with nerve a Fig. 2. CaBP1 assoiates with the α 1 2.1 suunit in tsa-201 ells and rat rain. (a) Lysates from ells transfeted with Ca v 2.1 plus CaBP, CaBP1 alone or Ca v 2.1 CBD plus CaBP1 were sujeted to immunopreipitation (i.p.) with affinity-purified α 1 2.1-speifi antiodies or ontrol IgG as indiated. Experiments were done with 10 mm EGTA (lanes 1 and 2) or 2 mm Ca 2+ (lanes 3 6). Blots were proed with α 1 2.1- (top) or CaBP1- speifi antiodies (ottom). () Rat ereellar proteins immunopreipitated with α 1 2.1-speifi antiodies (CNA5) or ontrol IgG were immunolotted with α 1 2.1- (top) or CaBP1-speifi antiodies (ottom). Lysate from tsa-201 ells transfeted with CaBP1 was used as a ontrol. nature neurosiene volume 5 no 3 marh 2002 211
a d e f terminals. However, CaBP1 and α 1 2.1 showed similar patterns of puntate staining in the CA1 region of the hippoampus and in the moleular layer of the ereellum (Fig. 3a f). As a large proportion of puntate laeling of α 1 2.1 in the ereellum oloalizes with that of syntaxin 28, it is likely that CaBP1 and Ca v 2.1 hannels oexist in at least some presynapti nerve terminals. Immunostaining for CaBP1 and for α 1 2.1 also overlapped in lusters along the dendrites of ereellar Purkinje neurons and in strutures that resemled dendriti spines (data not shown), however, whih suggests that CaBP1 may assoiate with Ca v 2.1 hannels in the post- as well as in the presynapti memrane. CaBP1 enhanes inativation of Ca v 2.1 hannels To eluidate the funtional onsequenes of the interation etween CaBP1 and Ca v 2.1 hannels, we determined the effet of CaBP1 on Ca 2+ urrents (I Ca ) in whole-ell path-lamp reordings of transfeted tsa-201 ells. We first ompared the effets of transfeted CaBP1 and endogenous CaM on inativation of I Ca. We a have shown previously that Ca 2+ /CaM enhanes the inativation of I Ca during step depolarizations when intraellular reording solutions ontain 0.5 mm EGTA 9,10. Here, Fig. 3. CaBP1 oloalizes with α 1 2.1 in rat rain setions. Rat rain setions were doule-laeled with antiodies speifi for CaBP1 and α 1 2.1. Laeling for CaBP1 is shown in green (a, d) and for α 1 2.1 in red (, e). In the merged images (, f), doulelaeled strutures appear yellow. Representative examples are shown from the moleular layer of the ereellum (a ) and the CA1 region of the hippoampus (d f). Sale ars, 5 µm (a ) and 50 µm (d f). inativation of I Ca aused y Ca 2+ /CaM proeeded with a single exponential time ourse (τ = 852.3 ± 63.7 ms at +20 mv, n = 18) that was relatively insensitive to the test voltage (Fig. 4a and ). By ontrast, CaBP1 aused I Ca to deay signifiantly faster than in ells that were transfeted with only Ca v 2.1. In almost all of the ells that were otransfeted with CaBP1, the deay of I Ca evoked y +20- and +30-mV pulses was est fit y a doule exponential funtion, with a slow omponent similar to ontrol and a fast omponent omprising 30 40% of the peak urrent (Fig. 4). With a +10- mv test pulse, however, iphasi inativation was deteted in only 11 out of 20 ells that had een otransfeted with CaBP1. At this test voltage, whih eliits the peak inward I Ca, Ca 2+ /CaMdependent inativation is maximal 10. Therefore, the asene of a fast phase of inativation in some ells otransfeted with CaBP1 might have resulted from more effetive ompetition y Ca 2+ /CaM. Competition etween CaM and CaBP1 for Ca v 2.1 hannels was supported further y the oservation of a marked redution in Ca 2+ -dependent inativation in ells otransfeted with CaBP1 (Fig. 5a and ). Enhaned inativation aused y CaM results in a signifiant redution in the residual urrent at the end of a 1-seond depolarizing test pulse normalized to the peak urrent (I res /I pk ) for I Ca as ompared with I Ba (refs. 9, 10). By ontrast, in ells otransfeted with CaBP1, I res /I pk was already redued when Ba 2+ was the harge arrier and was not signifiantly different for I Ca and I Ba (Fig. 5a and ). Ca 2+ -dependent inativation was not affeted in the same way y otransfetion Fig. 4. CaBP1 enhanes the inativation of I Ca in tsa- 201 ells transfeted with Ca v 2.1 hannels. (a) Representative traes of I Ca from ells transfeted with Ca v 2.1 either alone (ottom) or with CaBP1 (top). Currents were evoked y 1-s pulses to the indiated voltages from a holding potential of 80 mv and were saled for omparison. () Time onstants for the inativation of Ca v 2.1 hannel urrents in the asene of CaBP1. Test urrents were evoked y pulses to the indiated voltages as desried in (a) and fit with a single exponential funtion. Data were averaged from 6 18 ells. () Time onstants for inativation of I Ca in ells otransfeted with CaBP1. Test urrents were evoked y the same voltages as in (a) and (), ut urrent traes were fit with a doule exponential funtion. Fast (τ fast, filled ars) and slow time onstants (τ slow, open ars) were averaged from 7 20 ells. 212 nature neurosiene volume 5 no 3 marh 2002
Fig. 5. Fast, Ca 2+ -independent inativation of Ca v 2.1 hannels y CaBP1 differs from the a modulation of Ca v 2.1 hannels y CaM. (a) Ca v 2.1 hannel urrents reorded with Ca 2+ or Ba 2+ as the permeant ion. Test pulses were applied from a holding voltage of 80 mv to +10 mv (Ca 2+ ) or 0 mv (Ba 2+ ) for Ca v 2.1 either alone or otransfeted with CaM, or to +20 mv (Ca 2+ ) or +10 mv (Ba 2+ ) for ells otransfeted with CaBP1, to aount for the positive shift in voltage-dependent ativation aused y CaBP1. The intraellular solution ontained 0.5 mm EGTA. () The residual urrent amplitude at the end of a test pulse (I res, indiated in a) was normalized to the peak urrent (I pk ) for ells transfeted with Ca v 2.1 either alone or with CaBP1 or CaM. () Representative urrents evoked y a test pulse to +30 mv (+20 mv for I Ba ) in ells transfeted with wild-type or mutant Ca v 2.1 laking the CBD (Ca v 2.1 CBD ) either alone or otransfeted with CaBP1. Intraellular solutions ontained 0.5 mm EGTA exept where 10 mm BAPTA is indiated and extraellular solutions ontained 10 mm Ca 2+ exept where Ba 2+ is indiated. (d) Current amplitudes at 200 ms (I 200, indiated in ) were normalized to the peak urrent (I pk ) and plotted for the different onditions. Reordings were d from tsa-201 ells transfeted with Ca v 2.1 or Ca v 2.1 CBD either alone (open ars) or with CaBP1 (filled ars). Results represent averages of 5 13 ells. Asterisks indiate statistially signifiant differenes etween the paired groups (p 0.05). a with CaM instead of CaBP1, whih indiated that the faster, Ca 2+ -independent inativation was a speifi onsequene of the modulation of Ca v 2.1 hannels y CaBP1. To larify the effets of CaBP1 on fast inativation of Ca v 2.1 hannels, we measured the amplitude of I Ca at the 200-ms time point during a 1-s test pulse and normalized this to the peak urrent (I 200 /I pk, Fig. 5 and d). We used more positive test voltages to limit Ca 2+ entry, thus minimizing the ontriution of endogenous CaM in these experiments. With 0.5 mm EGTA, faster inativation of I Ca in ells with CaBP1 aused a signifiant derease in I 200 /I pk (0.45 ± 0.06 for CaBP1 versus 0.79 ± 0.03 for ontrol, p < 0.01). CaBP1 signifiantly enhaned fast inativation of I Ba (I 200 /I pk of 0.43 ± 0.09 for CaBP1 versus 0.74 ± 0.07 for ontrol, p < 0.05) and also of I Ca with 10 mm of the intraellular alium helator BAPTA (I 200 /I pk of 0.56 ± 0.08 for CaBP1 versus 0.85 ± 0.03 for ontrol, p < 0.02). The CBD was essential for these effets on inativation, eause CaBP1 had no effet on hannels in whih this domain had een deleted (Fig. 5 and d, Ca v 2.1 CBD ; p > 0.3). Together with iohemial analyses, these results support a Ca 2+ -independent assoiation of CaBP1 with the CBD, whih mediates a strong aeleration of Ca v 2.1 hannel inativation that does not require Ca 2+ influx or intraellular aumulation of Ca 2+. d Fig. 6. CaBP1 alters the voltage dependene of Ca v 2.1 ativation. Tail urrent voltage urves from tsa-201 ells transfeted with Ca v 2.1 (a ) or Ca v 2.1 CBD (d) either alone (open irles) or with CaBP1 (filled irles). Test pulses (10 ms) to the indiated voltages were applied from a holding voltage of 80 mv and peak tail urrents were measured upon the repolarization of ells to 40 mv, normalized to the largest tail urrent in the series, and plotted against test voltage. Test pulses were held for 10 ms, as ativation of urrents was omplete ut inativation was minimal during this time. Bath solutions ontained 10 mm Ca 2+ (a,, d) or Ba 2+ (), and intraellular solutions ontained 0.5 mm EGTA (a,, d) or 10 mm BAPTA (). Eah point represents the mean of 7 20 ells. nature neurosiene volume 5 no 3 marh 2002 213
Ca 2+ -dependent failitation is not supported y CaBP1 Ativity-dependent inreases in intraellular Ca 2+ ause an initial failitation of I Ca owing to the interation of Ca 2+ /CaM with Ca v 2.1 hannels (refs 10, 15). This Ca 2+ -dependent failitation was evident with 0.5 mm intraellular EGTA in paired-pulse protools, in whih Ca 2+ influx during a short prepulse indued a signifiant inrease in the tail urrent eliited y a susequent test pulse (Fig. 7a). With the same voltage protool, no failitation of I Ca was oserved in ells otransfeted with CaBP1 (Fig. 7). Beause of the strong voltagedependent enhanement of I Ca inativation aused y CaBP1, it was possile that paired-pulse failitation in ells otransfeted with CaBP1 might have een osured y the onset of inativation during the onditioning prepulse. Alternatively, CaBP1, unlike CaM, might not support Ca 2+ -dependent failitation of I Ca. To distinguish etween these possiilities, we analyzed the properties of I Ca during trains of short (5-ms) repetitive depoa CaBP1 shifts voltage dependene of Ca v 2.1 ativation In ells otransfeted with CaBP1 and Ca v 2.1, the normalized tail urrent voltage urve was shifted positively and was shallower than in ells transfeted with Ca v 2.1 alone (Fig. 6a). CaBP1 aused signifiant inreases in the half-ativation voltage, V 1/2 (12.8 ± 1.3 mv for CaBP1 versus 4.5 ± 0.9 mv for ontrol, p < 0.01), and slope fator of the tail urrent voltage urve ( 8. 7 ± 0.5 mv for CaBP1 versus 5.2 ± 0.6 mv for ontrol, p < 0.01). Similar to the other ations of CaBP1 on Ca v 2.1 hannels, these effets on I Ca ativation were essentially reprodued with intraellular BAPTA and extraellular Ba 2+ ut were not oserved with Ca v 2.1 CBD hannels (Fig. 6 d), whih indiates that the Ca 2+ -independent assoiation of CaBP1 with the CBD results in a newly identified, multifaeted regulation of Ca v 2.1 hannels. d Fig. 7. CaBP1 does not support Ca 2+ -dependent failitation of Ca v 2.1 hannels. (a, ) Voltage dependene of Ca v 2.1 Ca 2+ urrents evoked efore (P1, filled irles) and after (P2, open irles) a depolarizing prepulse. Tail urrents were measured y repolarizing ells to 40 mv for 5 ms after variale test voltages and normalized to the largest tail urrent evoked y P1. Inset, representative urrents evoked y a test pulse to +10 mv efore (filled irles) and after (open irles) the prepulse. Intraellular reording solution ontained 0.5 mm EGTA. Results were otained from ells transfeted with Ca v 2.1 either alone (a, n = 7) or with CaBP1 (, n = 10). (, d) Ca v 2.1 hannel urrents eliited y repetitive depolarizations. Test pulses (+20 mv () or +10 mv (d) to aount for voltage shifts ause y Ba 2+ sustitution) at a frequeny of 100 Hz were applied to ells transfeted with Ca v 2.1 either alone (open irles) or along with CaBP1 (filled irles). Peak urrent amplitudes were normalized to the first pulse in the series and plotted against time during the train. Every seond data point is shown. Intraellular reording solutions ontained 0.5 mm EGTA, and ath solutions ontained 10 mm Ca 2+ () or Ba 2+ (d). In (), n = 9 for open irles; n = 13 for losed irles. In (d), n = 5 for open irles; n = 11 for losed irles. larizations, whih should initially minimize the impat of voltage-dependent inativation and reveal failitation of I Ca early in the train. With 0.5 mm intraellular EGTA, Ca v 2.1 Ca 2+ urrents undergo a sustained failitation and gradually inativate elow initial urrent amplitudes after 800 ms of repetitive pulses (Fig. 7), an effet that depends on Ca 2+ /CaM 10. In ells otransfeted with CaBP1, failitation of I Ca was redued markedly, with urrent amplitudes rapidly inativating elow initial values only 200 ms into the train (Fig. 7). The maximum failitated I Ca amplitude at 50 ms in ells otransfeted with CaBP1 (1.04 ± 0.02, n = 13) was not signifiantly different from the Ca 2+ -independent failitation of Ba 2+ urrents in ells transfeted with Ca v 2.1 alone (1.04 ± 0.03, n =5, p = 0.80) or otransfeted with CaBP1 (1.02 ± 0.02, n = 11, p = 0.52; Fig. 7d), whih indiated that CaBP1 does not support Ca 2+ -dependent failitation of Ca v 2.1 hannels. Together with the enhaned inativation and positive shifts in ativation aused y CaBP1, the asene of Ca 2+ -dependent failitation would strongly limit voltage-dependent Ca 2+ entry through Ca v 2.1 hannels. These results highlight further the different modulation of these Ca 2+ hannels y CaBP1 and CaM. DISCUSSION We have shown that the neuronal Ca 2+ -inding protein CaBP1 interats with and modulates Ca v 2.1 hannels in a manner that is markedly different from that of CaM. CaBP1 ound to the CBD of α 1 2.1 ut aused signifiantly faster inativation of Ca v 2.1 hannel urrents than that aused y CaM. CaBP1 also positively shifted tail urrent-ativation urves and did not support Ca 2+ -dependent failitation of Ca v 2.1 urrents. Neither the assoiation of CaBP1 with the CBD nor the inhiitory modulation y CaBP1 required Ca 2+, in ontrast to the effets of CaM on Ca v 2.1 hannels, whih are stritly dependent on Ca 2+. The oserved assoiation and oloalization of CaBP1 and Ca v 2.1 hannels in neurons in the rain suggest that Ca 2+ hannel regulation y CaBP1 may e an important determinant of Ca 2+ signaling pathways in neurons. Ca 2+ -independent inding and modulation y CaBP1 The Ca 2+ independene of the interation etween CaBP1 and 214 nature neurosiene volume 5 no 3 marh 2002
Ca v 2.1 was unexpeted given the previously oserved Ca 2+ - dependent assoiation of CaBP1 with other CaM targets 25. It is possile that very loal rises in Ca 2+ might have esaped uffering y BAPTA in our experiments, whih ould have een suffiient for inding to CaBP1 and for ausing Ca 2+ -dependent modulation of I Ca. This possiility seems unlikely, however, eause the modulation y CaBP1 did not hange appreialy when Ba 2+ was the permeant ion; Ba 2+ ions ind to EF-hand motifs with relatively low affinity and so should not reprodue Ca 2+ -dependent regulation of target moleules 29. Calium-free CaM an assoiate with and regulate several targets, inluding the ryanodine reeptor RyR1, yli GMP kinase and a CaM-dependent adenylyl ylase from Bordetella pertussis 30 32. In addition, GCAPs photoreeptor Ca 2+ -inding proteins ativate guanylyl ylases in their Ca 2+ -free forms 33,34. Thus, CaBP1 might have a similar flexiility and interat with and regulate some effetors without inding Ca 2+. Although we annot exlude the possiility that CaBP1 may modulate some aspets of Ca 2+ hannel funtion in a Ca 2+ -dependent manner, we found no evidene to support a requirement for Ca 2+ in the effets of CaBP1 on the ativation and inativation of Ca v 2.1 hannels. Thus, despite the Ca 2+ -sensing apaility of CaBP1, we propose that CaBP1 itself does not mediate Ca 2+ -dependent regulation of Ca v 2.1 hannels, ut might indiretly influene feedak regulation y Ca 2+ y ompeting with CaM. Thus, CaBP1 may at more like auxiliary Ca 2+ hannel β-suunits y altering the intrinsi properties of Ca v 2.1 hannels to fine-tune voltagegated Ca 2+ entry in speifi lasses of neurons. Distint modulation of Ca v 2.1 y CaBP1 and CaM We have shown that oth CaM and CaBP1 interat with the CBD of the α 1 2.1 suunit and that this site is essential for full hannel regulation y oth proteins. Confliting evidene indiates that the IQ domain a sequene that is N-terminal to the CBD is involved in the modulation of Ca v 2.1 hannels y Ca 2+ /CaM 15. Our results do not support the importane of the IQ domain in modulation y CaBP1 eause, first, CaBP1 interated with the CBD ut not the IQ domain in yeast two-hyrid assays (Fig. 1 and ); seond, deleting the CBD prevented the o-immunopreipitation of CaBP1 with α 1 2.1 (Fig. 2a); and third, the fast inativation and shifts in the voltage dependene of ativation aused y CaBP1 were aolished in hannels that laked the CBD (Figs. 5 and 6). Notaly, removing the CBD from α 1 2.1 eliminated regulation y CaBP1 more ompletely than it eliminated regulation y CaM 10. Together, our results indiate that the CBD may e the primary determinant for the funtional effets of CaBP1 on Ca v 2.1 hannels. If oth CaM and CaBP1 interat with the CBD, how is it that CaBP1 auses Ca 2+ -independent fast inativation and positively shifted ativation, whereas CaM auses Ca 2+ -dependent failitation and inativation of Ca v 2.1 hannels? One possiility is that key strutural features that distinguish CaBP1 from CaM, suh as its extra-long entral helial domain and N-terminal myristoylation (Fig. 1a), may permit Ca 2+ -independent inding of CaBP1 to the CBD, whih might then lead to its unique inhiitory modulation of I Ca. Future experiments that determine how suh differenes etween CaBP1 and CaM ontriute to speifi forms of Ca v 2.1 regulation may reveal how ion hannels and other signaling moleules are differentially modulated y CaM and related Ca 2+ -inding proteins. Modulation of neuronal Ca v 2.1 hannels y CaBP1 Our immunopreipitation and immunofluoresene studies showed that CaBP1 and α 1 2.1 assoiate physially in extrats of rat ereellum and that their suellular distriutions overlap in this rain region, whih indiates that CaBP1 may have a physiologial role in the regulation of Ca v 2.1 hannels. Beause oth CaM and CaBP1 interated with the same site on the α 1 2.1 suunit, an important issue is whether Ca v 2.1 would interat funtionally with CaM and/or CaBP1 in neurons in whih oth Ca 2+ -inding proteins are expressed. Although we do not know whether CaM and CaBP1 ind simultaneously to Ca v 2.1 hannels, our eletrophysiologial studies suggested that CaM and CaBP1 might ompetitively regulate the hannel. CaBP1 more strongly enhaned inativation of I Ca when the influene of Ca 2+ /CaM was suppressed either with extraellular Ba 2+ or intraellular BAPTA, or with test voltages that eliited sumaximal Ca 2+ influx. These results imply that when intraellular Ca 2+ onentrations are high Ca v 2.1 hannels may e failitated predominantly y CaM, and that the inativating effets of CaBP1 eome most prominent when ytoplasmi Ca 2+ onentrations deline. In this way, CaM and CaBP1 may oordinately at as a moleular swith to intensify neuronal Ca 2+ influx in response to ativity-dependent alterations in intraellular onentrations of Ca 2+. NCBPs and synapti transmission Emerging evidene supports a role for NCBPs in the regulation of synapti transmission. In partiular, neuronal Ca 2+ sensor-1 (NCS-1), whih is more distantly related to CaM than is CaBP1, regulates neurotransmitter release 35, synapse formation 36 and neuronal iruits that ontrol assoiative learning 37. Notaly, NCS-1 has een impliated in the negative regulation of Ca 2+ hannels in hromaffin ells 38, whih suggests that Ca v 2.1 hannels may e modulated y NCBPs in addition to CaBP1. Given the widespread distriution of Ca v 2.1 hannels throughout the nervous system, the ell type speifi modulation of Ca v 2.1 y CaBP1, CaM or other NCBPs may fundamentally determine the nature of presynapti and postsynapti Ca 2+ signals and the funtional onsequenes of synapti ativity. METHODS Yeast two-hyrid assays. We amplified DNAs enoding the long isoform of human CaBP1 (ref. 25) and the ytoplasmi domains of α 1 2.1 y polymerase hain reation and suloned them into the yeast twohyrid vetors pact2 and pas2-1, respetively (Clonteh, Palo Alto, California). To test for interations etween CaBP1 and speifi domains of α 1 2.1, the orresponding plasmids were otransformed into yeast strain Y190. We assayed growth on medium laking histidine and β-galatosidase to identify interating proteins as desried 9. Cell ulture and transfetion. We grew tsa-201 ells to 70% onflueny and transfeted them y the alium phosphate method with an equimolar ratio of DNAs enoding the rat rain Ca 2+ hannel suunits α 1 2.1 (ra), β 2a and α 2 δ (ref. 26). The α 1 2.1 onstrut that laks amino aids 1969 2000 (α 1 2.1 CBD ) has een desried 10. The long isoform of human CaBP1 (ref. 25) was suloned into the BamHI sites of pdna3.1+ (Invitrogen, Carlsad, California) and transfeted at a 5:1 molar exess with Ca 2+ hannel suunits. For eletrophysiologial experiments, we plated ells on 35-mm dishes and transfeted them with 5 µg of total DNA, inluding 0.3 µg of a CD8 expression plasmid to allow the detetion of transfeted ells. For o-immunopreipitation assays, we plated ells on 150-mm dishes and transfeted them with 50 µg of total plasmid DNA. Co-immunopreipitation assays. At least 48 h after transfetion, tsa-201 ells were homogenized in ie-old lysis uffer (1% Nonidet P-40 in TBS (20 mm Tris-HCl, ph 7.3, 150 mm NaCl), 10 mm EGTA and protease inhiitors) and entrifuged at 1,000g for 5 min. To maintain Ca 2+ - nature neurosiene volume 5 no 3 marh 2002 215
dependent interations, we pretreated some groups with 5 µm A23187 and 2 mm CaCl 2 for 15 min ut did not inlude EGTA in the lysis uffer. The postnulear supernatant (300 400 µg of memrane protein) was inuated with 15 µg of α 1 2.1-speifi antiodies (raised against CNA5) 39 for 2 h at 4 C. Immune omplexes were separated on protein A Sepharose, resolved y SDS PAGE and transferred to nitroellulose. For immunolotting, we loked nitroellulose filters for 30 min in 5% milk/tbs and inuated them with the CaBP1 antiserum UW72 (ref. 25; 1:1,000 dilution) or with CNA5-speifi antiodies (2.5 µg/ml) for 1 h. Blots were washed three times in TBS with 0.05% Tween 20 (TBST) and inuated with horseradish peroxidase linked protein A (Amersham, Pisataway, New Jersey; 1:2,000) for 40 min. We used ECL western lotting reagent (Amersham) for detetion of hemiluminesene. For o-immunopreipitations from rat rain, we homogenized ereellar tissue from two adult male rats in 0.3 M surose, 75 mm NaCl, 10 mm Tris-HCl, ph 7.4, and 10 mm EGTA. We inluded protease inhiitors in the homogenization uffer and in uffers used at all susequent steps. Homogenates were entrifuged for 10 min at 1,000g, and memrane frations were separated from the postnulear supernatant at 100,000g for 30 min. Memrane proteins were soluilized with 4 ml of uffer A (1% Triton X-100, 10 mm Tris, ph 7.4, and 10 mm EGTA) and insolule material was removed y further entrifugation (100,000g for 30 min). Ca 2+ hannels were immunopreipitated with 15 µg of CNA5-speifi antiodies per ml of soluilized memrane protein. We isolated immune omplexes on protein A Sepharose and deteted the assoiated CaBP1 y immunolotting as desried aove. Immunoytohemistry. Anesthetized adult Sprague Dawley rats were perfused intraardially with 4% paraformaldehyde in 0.1 M sodium phosphate uffer, ph 7.4. The rain was post-fixed and ryoproteted in 30% (w/v) surose, and tissue setions (35 µm) were ut on a sliding mirotome in 0.1 M phosphate uffer. Tissue setions were rinsed with 0.1 M Tris-uffered saline (TBS) and loked sequentially with 2% avidin and 2% iotin. For the doule-laeling of CaBP1 and α 1 2.1, we inuated tissue setions in UW72 antiserum (diluted 1:100) for 36 h at 4 C, iotinylated goat antiody against rait IgG (Vetor Laoratories, Burlingame, California; 1:300) for 1 h at 37 C, and avidin D fluoresein (Vetor Laoratories; 1:300) for 1 h at 37 C, with rinsing etween eah step. The tissue was then loked with 5% normal rait serum in TBS for 1 h and inuated with affinity-purified Fa fragments for 1 h at 37 C. After rinsing, tissue setions were inuated with antiodies speifi for CNA5 (1:15) for 36 h at 4 C, iotinylated goat antiody against rait IgG (1:300) for 1 h at 37 C, and avidin D Texas Red (1:300) for 1 h at 37 C. Tissue setions were mounted on gelatin-oated slides, proteted with overslips, and viewed with a Bio-Rad MRC 600 mirosope in the W.M. Kek Imaging Faility at the University of Washington. All proedures onformed to protools approved y the Animal Welfare Committee of the University of Washington. Eletrophysiology and data analysis. At least 48 h after transfetion, tsa-201 ells were inuated with CD8-speifi antiody oated mirospheres (Dynal, Oslo, Norway) to permit detetion of transfeted ells. We reorded whole-ell Ca 2+ urrents with a List EPC-7 pathlamp amplifier and filtered them at 5 khz. Leak and apaitive transients were sutrated using a P/ 4 protool. Extraellular reording solutions were omposed of 150 mm Tris, 1 mm MgCl 2 and 10 mm CaCl 2 or BaCl 2 ; intraellular solutions were omposed of 120 mm N- methyl-d-gluamine, 60 mm HEPES, 1 mm MgCl 2, 2 mm Mg-ATP and 0.5 mm EGTA or 10 mm BAPTA. The ph of all solutions was adjusted to 7.3 with methanesulfoni aid. The time ourse of I Ca deay was fit y either A[exp( t/τ)] or A slow [exp( t/τ slow )] + A fast [exp(t/τ fast )], where t is time; A slow and A fast are the amplitudes of the slow and fast exponentials, respetively, at t = 0; and τ slow and τ fast are the time onstants of the deay of the two proesses. Normalized tail urrent voltage urves were fit with a single Boltzmann funtion: A/{1 + exp[(v V 1/2 )/k] + }, where V is test pulse voltage, V 1/2 is the midpoint of the ativation urve, k is a slope fator, A is the amplitude and is the aseline. Curve fits and data analysis were done with Igor Pro software (Wavemetris, Lake Oswego, Oregon). All averaged data are the mean ± s.e.m. We determined the statistial signifiane of differenes etween groups y Student s t-test (SigmaPlot, SPSS Siene, Chiago, Illinois). Aknowledgements This work was supported y NIH Researh Grant R01 NS22625 to W.A.C, a NSRA postdotoral researh fellowship from NIH (F32 NS10645) to A.L., NIH Researh Grant R01 EY08061 to K.P. and researh grants from Researh to Prevent Blindness, In., the Alon Researh Institute and the E.K. Bishop Foundation to K.P. RECEIVED 24 SEPTEMBER; ACCEPTED 21 DECEMBER 2001 1. Miller, R. J. Multiple alium hannels and neuronal funtion. Siene 235, 46 52 (1987). 2. Catterall, W. A. Struture and funtion of neuronal Ca 2+ hannels and their role in neurotransmitter release. Cell Calium 24, 307 323 (1998). 3. Ikeda, S. R. Voltage-dependent modulation of N-type alium hannels y G-protein βγ suunits. Nature 380, 255 258 (1996). 4. Herlitze, S. et al. Modulation of Ca 2+ hannels y G protein βγ suunits. Nature 380, 258 262 (1996). 5. De Waard, M. et al. Diret inding of G-protein βγ omplex to voltagedependent alium hannels. Nature 385, 446 450 (1997). 6. Bezprozvanny, I., Sheller, R. H. & Tsien, R. W. Funtional impat of syntaxin on gating of N-type and Q-type alium hannels. Nature 378, 623 626 (1995). 7. Zhong, H., Yokoyama, C. T., Sheuer, T. & Catterall, W. A. Reiproal regulation of P/Q-type Ca 2+ hannels y SNAP-25, syntaxin and synaptotagmin. Nat. Neurosi. 2, 939 941 (1999). 8. Wiser, O., Bennett, M. K. & Atlas, D. Funtional interation of syntaxin and SNAP-25 with voltage-sensitive L- and N-type Ca 2+ hannels. EMBO J. 15, 4100 4110 (1996). 9. Lee, A. et al. Ca 2+ /almodulin inds to and modulates P/Q-type alium hannels. Nature 339, 155 159 (1999). 10. Lee, A., Sheuer, T. & Catterall, W. A. Ca 2+ /almodulin-dependent failitation and inativation of P/Q-type Ca 2+ hannels. J. Neurosi. 20, 6830 6838 (2000). 11. Qin, N., Olese, R., Bransy, M., Lin, T. & Birnaumer, L. Ca 2+ -indued inhiition of the ardia Ca 2+ hannel depends on almodulin. Pro. Natl. Aad. Si. USA 96, 2435 2438 (1999). 12. Peterson, B. Z., DeMaria, C. D. & Yue, D. T. Calmodulin is the Ca 2+ sensor for Ca 2+ -dependent inativation of L-type alium hannels. Neuron 22, 549 558 (1999). 13. Zühlke, R. G., Pitt, G. S., Deisseroth, K., Tsien, R. W. & Reuter, H. Calmodulin supports oth inativation and failitation of L-type alium hannels. Nature 399, 159 161 (1999). 14. Pate, P. et al. Determinants for almodulin inding on voltage-dependent Ca 2+ hannels. J. Biol. Chem. 275, 39786 39792 (2000). 15. DeMaria, C. D., Soong, T., Alseikhan, B. A., Alvania, R. S. & Yue, D. T. Calmodulin ifurates the loal Ca 2+ signal that modulates P/Q-type Ca 2+ hannels. Nature 411, 484 489 (2001). 16. Forsythe, I. D., Tsujimoto, T., Barnes-Davies, M., Cuttle, M. F. & Takahashi, T. Inativation of presynapti alium urrent ontriutes to synapti depression at a fast entral synapse. Neuron 20, 797 807 (1998). 17. Cuttle, M. F., Tsujimoto, T., Forsythe, I. D. & Takahashi, T. Failitation of the presynapti alium urrent at an auditory synapse in rat rainstem. J. Physiol. (Lond.) 512, 723 729 (1998). 18. Borst, J. G. & Sakmann, B. Failitation of presynapti alium urrents in the rat rainstem. J. Physiol. (Lond.) 513, 149 155 (1998). 19. Dunlap, K., Lueke, J. I. & Turner, T. J. Exoytoti Ca 2+ hannels in mammalian entral neurons. Trends Neurosi. 18, 89 98 (1995). 20. Wheeler, D. B., Randall, A. & Tsien, R. W. Roles of N-type and Q-type Ca 2+ hannels in supporting hippoampal synapti transmission. Siene 264, 107 111 (1994). 21. Polans, A., Baehr, W. & Palzewski, K. Turned on y Ca 2+! The physiology and pathology of Ca 2+ -inding proteins in the retina. Trends Neurosi. 19, 547 554 (1996). 22. Burgoyne, R. D. & Weiss, J. L. The neuronal alium sensor family of Ca 2+ - inding proteins. Biohem. J. 353, 1 12 (2001). 23. Sallese, M. et al. The G-protein-oupled reeptor kinase GRK4 mediates homologous desensitization of metaotropi glutamate reeptor 1. FASEB J. 14, 2569 2580 (2000). 24. Shaad, N. C. et al. Diret modulation of almodulin targets y the neuronal alium sensor NCS-1. Pro. Natl. Aad. Si. USA 93, 9253 9258 (1996). 25. Haeseleer, F. et al. Five memers of a novel Ca 2+ -inding protein (CABP) sufamily with similarity to almodulin. J. Biol. Chem. 275, 1247 1260 (2000). 26. Stea, A. et al. The loalization and funtional properties of a rat rain α 1A alium hannel reflet similarities to neuronal Q- and P-type hannels. Pro. Natl. Aad. Si. USA 91, 10576 10580 (1994). 27. Seideneher, C. I. et al. Caldendrin, a novel neuronal alium-inding protein onfined to the somato-dendriti ompartment. J. Biol. Chem. 273, 21324 21331 (1998). 28. Westenroek, R. E. et al. Immunohemial identifiation and suellular distriution of the α 1A suunits of rain alium hannels. J. Neurosi. 15, 6403 6418 (1995). 216 nature neurosiene volume 5 no 3 marh 2002
29. Chao, S. H., Suzuki, Y., Zysk, J. R. & Cheung, W. Y. Ativation of almodulin y various metal ations as a funtion of ioni radius. Mol. Pharmaol. 26, 75 82 (1984). 30. Rodney, G. G., Williams, B. Y., Strasurg, G. M., Bekingham, K. & Hamilton, S. L. Regulation of RYR1 ativity y Ca 2+ and almodulin. Biohemistry 39, 7807 7812 (2000). 31. Yamaki, T. & Hidaka, H. Ca 2+ -independent stimulation of yli GMPdependent protein kinase y almodulin. Biohem. Biophys. Res. Commun. 94, 727 733 (1980). 32. Greenlee, D. V., Andreasen, T. J. & Storm, D. R. Calium-independent stimulation of Bordetella pertussis adenylate ylase. Biohemistry 21, 2759 2764 (1982). 33. Rudnika-Nawrot, M. et al. Changes in iologial ativity and folding of guanylate ylase-ativating protein 1 as a funtion of alium. Biohemistry 37, 248 257 (1998). 34. Haeseleer, F. et al. Moleular haraterization of a third memer of the guanylyl ylase-ativating protein sufamily. J. Biol. Chem. 274, 6526 6535 (1999). 35. MFerran, B. W., Graham, M. E. & Burgoyne, R. D. Neuronal Ca 2+ sensor 1, the mammalian homologue of frequenin, is expressed in hromaffin and PC12 ells and regulates neuroseretion from dense-ore granules. J. Biol. Chem. 273, 22768 22772 (1998). 36. Chen, X. L. et al. Overexpression of rat neuronal alium sensor-1 in rodent NG108-15 ells enhanes synapse formation and transmission. J. Physiol. (Lond.) 532, 649 659 (2001). 37. Gomez, M. et al. Ca 2+ signaling via the neuronal alium sensor-1 regulates assoiative learning and memory in C. elegans. Neuron 30, 241 248 (2001). 38. Weiss, J. L., Arher, D. A. & Burgoyne, R. D. Neuronal Ca 2+ sensor- 1/frequenin funtions in an autorine pathway regulating Ca 2+ hannels in ovine adrenal hromaffin ells. J. Biol. Chem. 275, 40082 40087 (2000). 39. Sakurai, T., Westenroek, R. E., Rettig, J., Hell, J. & Catterall, W. A. Biohemial properties and suellular distriution of the BI and ra isoforms of α 1A suunits of rain alium hannels. J. Cell Biol. 134, 511 528 (1996). nature neurosiene volume 5 no 3 marh 2002 217