CaBP1, a neuronal Ca2+ sensor protein, inhibitsinositol trisphosphate receptors by clampingintersubunit interactionsCongmin Lia,1, Masahiro Enomotob,1, Ana M. Rossic,1, Min-Duk Seob,1,2, Taufiq Rahmanc, Peter B. Stathopulosb,Colin W. Taylorc,3, Mitsuhiko Ikurab,3, and James B. Amesa,3

aDepartment of Chemistry, University of California, Davis, CA 95616; bDivision of Signaling Biology, Ontario Cancer Institute and Department of MedicalBiophysics, University of Toronto, Toronto, ON, Canada M5G 1L7; and cDepartment of Pharmacology, University of Cambridge, Cambridge CB2 1PD,United Kingdom

Edited by Chikashi Toyoshima, University of Tokyo, Tokyo, Japan, and approved April 9, 2013 (received for review December 3, 2012)

Calcium-binding protein 1 (CaBP1) is a neuron-specific member ofthe calmodulin superfamily that regulates several Ca2+ channels,including inositol 1,4,5-trisphosphate receptors (InsP3Rs). CaBP1alone does not affect InsP3R activity, but it inhibits InsP3-evokedCa2+ release by slowing the rate of InsP3R opening. The inhibitionis enhanced by Ca2+ binding to both the InsP3R and CaBP1. CaBP1binds via its C lobe to the cytosolic N-terminal region (NT; residues1–604) of InsP3R1. NMR paramagnetic relaxation enhancementanalysis demonstrates that a cluster of hydrophobic residues(V101, L104, and V162) within the C lobe of CaBP1 that are ex-posed after Ca2+ binding interact with a complementary clusterof hydrophobic residues (L302, I364, and L393) in the β-domainof the InsP3-binding core. These residues are essential for CaBP1binding to the NT and for inhibition of InsP3R activity by CaBP1.Docking analyses and paramagnetic relaxation enhancementstructural restraints suggest that CaBP1 forms an extended tetra-meric turret attached by the tetrameric NT to the cytosolic vesti-bule of the InsP3R pore. InsP3 activates InsP3Rs by initiatingconformational changes that lead to disruption of an intersubunitinteraction between a “hot-spot” loop in the suppressor domain(residues 1–223) and the InsP3-binding core β-domain. Targetedcross-linking of residues that contribute to this interface show thatInsP3 attenuates cross-linking, whereas CaBP1 promotes it. Weconclude that CaBP1 inhibits InsP3R activity by restricting the inter-subunit movements that initiate gating.

EF hand | intracellular Ca2+ channel | ion channel | ryanodine receptor

Dynamic increases in cytosolic free Ca2+ concentration ([Ca2+]c)regulate many cellular events, including acute and long-term

changes in neuronal activity (1–3). Release of Ca2+ from in-tracellular stores is controlled by intracellular Ca2+ channels (4),the most common of which are inositol 1,4,5-trisphosphate re-ceptors (InsP3Rs) (5, 6). Dual regulation of InsP3Rs by InsP3and Ca2+ facilitates regenerative Ca2+ release (6), generatingCa2+ signals of remarkable versatility and spatiotemporal com-plexity (1, 2, 7). The sites through which Ca2+ biphasically regu-lates InsP3Rs are unresolved (5, 8, 9). There is, however, evidence,that proteins with EF-hand Ca2+-binding motifs can regulate gatingof InsP3Rs. These include calmodulin (CaM) (10–12), calmyrin(CIB1) (13), and neuronal Ca2+ sensor (NCS) proteins (2). Thelatter comprise a branch of the CaM superfamily that includesNCS-1 (14) and Ca2+-binding protein 1 (CaBP1) (15, 16).CaBP1–5 proteins (2, 17) have four EF hands that form pairs

within the N lobe (EF1 and 2) and C lobe (EF3 and 4). The twolobes are structurally independent (18) and connected by a flexiblelinker (18). Whereas all four EF hands bind Ca2+ in CaM, EF2 inCaBP1 does not bind Ca2+, and EF1 has reduced selectivity forCa2+ over Mg2+. EF3 and EF4 in the C lobe of CaBP1 exhibitcanonical Ca2+-induced conformational changes (18, 19).Many splice variants and isoforms of CaBPs are expressed in dif-ferent neurons (20–22), and their targets include a variety of ionchannels (2). CaBP1, for example, regulates voltage-gated P/Q-type

(23) and L-type Ca2+ channels (24) and a transient receptorpotential channel, TRPC5 (25). Furthermore, the prevailing viewthat InsP3Rs open only after binding InsP3 was challenged byevidence that CaBP1 (15) and related proteins (13) might, intheir Ca2+-bound forms, gate InsP3Rs. The suggestion that Ca2+,via CaBP1, might directly gate InsP3Rs proved to be contentious,but it spawned further evidence that CaBP1 regulates InsP3Rs(14, 16, 20).InsP3Rs are large tetrameric channels (5, 6). Their activation is

initiated within the N-terminal domain (NT; residues 1–604) bybinding of InsP3 to the InsP3-binding core (IBC; residues 224–604)of each subunit (26). This process leads, via rearrangement of thesuppressor domain (SD; residues 1–223) (27), to opening of anintrinsic pore (28, 29). Despite extensive studies of CaBP1 (2) andof the many proteins that modulate InsP3Rs (5), little is knownabout the structural basis of these protein interactions with InsP3Rsor of CaBP1 with any ion channel. Here we combine NMR, mu-tagenesis, cross-linking, and functional analyses to define, at theatomic level, the interactions between CaBP1 and InsP3Rs.

ResultsCaBP1 Inhibits InsP3-Evoked Ca2+ Release. CaBP1 is found only inneurons (2), and they predominantly express InsP3R1. We there-fore used permeabilized DT40 cells lacking endogenous InsP3Rs,but stably expressing rat InsP3R1, to assess the effects of CaBP1 onCa2+ release from intracellular stores. The two splice variants ofCaBP1 expressed in brain (17) regulate InsP3Rs and Ca2+ channelssimilarly (16). Throughout this study, we used the short variant(Table S1) because its solubility makes it more amenable to NMRanalysis. Across a range of [Ca2+]c, CaBP1 alone had no effecton the Ca2+ content of the intracellular stores of DT40–InsP3R1 cells (Fig. S1 A–D). The lack of effect of Ca2+–CaBP1on InsP3R1 was confirmed by nuclear patch-clamp analysesof single InsP3R1 (Fig. S1E). These results are inconsistentwith the notion that Ca2+–CaBP1 stimulates Ca2+ release viaInsP3R (15), a suggestion that has also been challenged by others,who argue that CaBP1 inhibits InsP3-evoked Ca2+ release (16, 20).CaBP1 caused a concentration-dependent decrease in the

sensitivity of InsP3-evoked Ca2+ release (Fig. 1A and Table S2)without affecting 3H–InsP3 binding (Fig. S1 G–I). Inhibition of

Author contributions: C.W.T., M.I., and J.B.A. designed research; C.L., M.E., A.M.R., M.-D.S.,T.R., and P.B.S. performed research; C.L., M.E., A.M.R., M.-D.S., T.R., P.B.S., C.W.T., M.I., andJ.B.A. analyzed data; and C.L., C.W.T., M.I., and J.B.A. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.1C.L., M.E., A.M.R., and M.-D.S. contributed equally to this work.2Present address: Department of Pharmacy, Ajou University, Suwon 443-749, Korea.3To whom correspondence may be addressed. E-mail: jbames@ucdavis.edu, mikura@uhnres.utoronto.ca, or cwt1000@cam.ac.uk.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1220847110/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1220847110 PNAS | May 21, 2013 | vol. 110 | no. 21 | 8507–8512

BIOCH

EMISTR

Y

InsP3-evoked Ca2+ release by CaBP1 was increased at higher[Ca2+]c, but was evident even at the [Ca2+]c of a resting cell, al-though not at lower [Ca2+]c (Fig. 1B and Table S2). In single-channel analyses recorded under optimal conditions for InsP3Ractivation (30), CaBP1 (10 μM) massively reduced channel activity(NPo) without affecting unitary conductance (γCs) or mean chan-nel open time (τo) (Fig. 1 C and D). Analysis of records that in-cluded only a single functional InsP3R established that an increasein mean channel closed time (τc) from 4.2 ± 0.9 ms to 46 ± 13 msaccounted for the 5.3-fold reduction in NPo in the presence ofCaBP1 (Fig. S1F). The lack of effect on τo and γCs indicates thatCaBP1 inhibits gating rather than blocking the InsP3R pore.

Inhibition of InsP3R by CaBP1 Is Enhanced by Ca2+ Binding to bothCaBP1 and InsP3R.Cytosolic Ca2+ enhances the inhibition of InsP3Rby CaBP1 (Fig. 1B and Table S2). We mutated residues withineach of the three functional EF hands of CaBP1 to prevent Ca2+

binding (CaBP1134) (Fig. S2F). CaBP1134 was as effective asCaBP1 in causing Ca2+-dependent inhibition of InsP3-evoked Ca2+

release at the typical [Ca2+]c of a resting cell (230 nM), butless effective than CaBP1 at higher [Ca2+]c (Fig. 1B and Fig. S2 A–D). When only the last pair of EF hands was mutated (CaBP134),the results were similar to those obtained with CaBP1134 (Fig. S2E).Although CaBP1134 does not bind Ca2+ (Fig. S2F), its ability toinhibit InsP3-evoked Ca2+ release was enhanced by increasing[Ca2+]c. However, at the highest [Ca2+]c, the inhibition by mu-tant CaBP1 was less than with native CaBP1 (Fig. 1B andFig. S2). This result suggests that the last pair of EF hands inCaBP1 contributes to the enhanced inhibition of InsP3R at ele-vated [Ca2+]c. We conclude that CaBP1 inhibits InsP3-evokedCa2+ release at resting [Ca2+]c. At the [Ca2+]c of stimulated cells,the inhibition is potentiated by Ca2+ binding to both InsP3R andthe last pair of EF hands in CaBP1. This complex regulation ofCaBP1–InsP3R interactions by cytosolic Ca2+ may have con-tributed to conflicting reports of their Ca2+ dependence (13, 15,16, 20) and requirement for functional EF hands (15, 16).

Local Hydrophobic Interactions Between InsP3R and the C Lobe ofCaBP1. Ca2+–CaBP1 binds via its C lobe to the NT, in both itsapo and InsP3-bound forms, with a 1:1 stoichiometry and anequilibrium dissociation constant (KD) of ∼3 μM (18). In theabsence of Ca2+, CaBP1 binds with 10-fold lower affinity (18),consistent with our functional analyses (Fig. 1B). We used NMR-based approaches, including chemical shift perturbation andparamagnetic relaxation enhancement (PRE) (31), to examine

the structure of the NT–CaBP1 complex. Our PRE experimentsmeasure distances between side-chain methyl groups in CaBP1that are <10 Å away from nitroxide spin labels attached tospecific Cys residues in the NT. The starting point was the NTin which all Cys residues were replaced by Ala (NTCL) (Table S1).Extensive structural and functional studies confirmed thatNTCL mimics the behavior of wild-type NT (29). Isothermaltitration calorimetry demonstrated that Ca2+–CaBP1 binds toNTCL (KD = 16 μM, pKD = 4.8 ± 0.1) (Fig. S3A), although withlower affinity than native NT (KD = 3 μM, pKD = 5.5 ± 0.1) (18).This small difference in Gibbs free energy of binding (ΔΔG° =0.9 kcal/mole) suggests that CaBP1 has a similar structural in-teraction with NT and NTCL, consistent with the similar NMRspectra of Ca2+–CaBP1 bound to NTCL (Fig. S4A, red) and wild-type NT (Fig. S4A, blue). We then introduced single Cys residuesinto strategic sites on the surface of NTCL and used them in PREexperiments. The NMR spectra of CaBP1 bound to wild-typeand mutant NTs are similar, confirming that each NT mutant isfolded and bound similarly to CaBP1.Binding of NTCL to 15N-labeled CaBP1 caused nearly all back-

bone amide resonances to broaden beyond detection in 15N–1Hheteronuclear single quantum coherence spectra, preventing useof backbone amide resonances in the PRE analysis. Only side-chain methyl NMR resonances of CaBP1 were detected withenough sensitivity to be analyzed using PRE. 13C-labeled CaBP1binding to unlabeled NTCL was monitored by using 1H–

13C methyltransverse relaxation-optimized spectroscopy (TROSY) NMR(Fig. S4A). Binding of NTCL had large effects on the NMR res-onances assigned to CaBP1 residues in the C lobe, whereas resi-dues in the N lobe were unaffected. This finding is consistent withthe NT binding via the C lobe of CaBP1 (18). The 13C-labeledmethyl resonances of V101, L104, and V162 in CaBP1 becameseverely broadened after the addition of NTCL, suggesting thatthese residues directly contact NTCL. Mutation of each of theseresidues to Ser massively reduced the affinity of CaBP1 for NTCL

(Fig. S3 B and C). In addition, exposed residues in CaBP1 (I124,L131, L132, and L150) have methyl resonances that show per-turbations in chemical shifts, indicating a change in their magneticenvironment upon binding of the NT. By monitoring the NMRspectral changes of 13C-methyl–labeled CaBP1 complexed withNTCL in the presence (paramagnetic) and absence (diamagnetic)of attached spin label, the proximity of NTCL and CaBP1 wasdefined. A methyl TROSY spectrum of 13C-labeled CaBP1 boundto NTCL with a single Cys insertion, NTCL(E20C), was similar tothat of CaBP1 bound to NTCL (Fig. S4 A and B), indicatingthat NTCL(E20C) is structurally intact. Attachment of a nitroxide

Fig. 1. Inhibition of InsP3R1 by CaBP1. (A) CaBP1inhibits InsP3-evoked Ca2+ release. Permeabilized DT40–IP3R1 cells in CLM with a [Ca2+]c of 1.2 μM were in-cubated with CaBP1 (10 min) before adding InsP3.Results (means ± SEM; n = 3, with duplicate determi-nations in each) show the concentration-dependent re-lease of Ca2+by InsP3. (B) Effects of CaBP1 and CaBP1134(50 μM) on InsP3-evoked Ca2+ release at the indicated[Ca2+]c show that CaBP1 is not the only Ca2+ sensor. Foreach [Ca2+]c, results (percentage of control, means ±SEM; n = 3–4, with duplicate determinations in each)show the Ca2+ release evoked by the concentrationof InsP3 that evoked half-maximal Ca2+ release(EC50) under control conditions. *P< 0.05 relative tocontrol. (C) Typical patch-clamp recordings fromsingle InsP3R1 in medium with [Ca2+]c of 1.5 μMstimulated with InsP3 (10 μM) alone or with CaBP1(10 μM). Bars show the closed state. The holdingpotential was +40 mV. (D) Summary data (mean ±SEM; n given in Fig. S1F) show NPo, mean channelopen (τo) and closed (τc) times and unitary con-ductance (γCs).

8508 | www.pnas.org/cgi/doi/10.1073/pnas.1220847110 Li et al.

spin label to NTCL(E20C) caused a marked decrease in NMRpeak intensity for some CaBP1 residues (I124, L131, L132, V148,and L150), whereas others (L99, V136, L141, L145, and V156)were less affected (Fig. S4B). The ratios of peak intensities in thepresence and absence of spin label (IParamagnetic/IDiamagnetic) weretaken as a measure of the distance between the methyl groupsand spin label. The PRE ratios are listed for all single Cysinsertions in Fig. S4C.We used the PRE restraints and chemical shift perturbation

data within HADDOCK (32) to dock the NMR structure ofCa2+-bound C lobe of CaBP1 (18) onto the crystal structure ofthe NT (29) (Fig. 2A). Within the complex, the Ca2+-boundC lobe of CaBP1 is in the familiar open conformation typical ofCa2+-bound EF hands in CaM (33). Exposed hydrophobic resi-dues in Ca2+–CaBP1 (V101, L104, and V162) interact withclustered hydrophobic residues (L302, I364, and L393) in theIBC-β domain of InsP3R1 (Fig. 2A). These residues are con-served in InsP3Rs but not in ryanodine receptors, consistent withevidence that CaBP1 binds to all three InsP3R subtypes (15, 16)but not to ryanodine receptors (16). Exposed CaBP1 residues inEF4 (I144, M164, and M165) made contacts with H289 in theIBC-β domain, and side-chain atoms of R167 (CaBP1) werewithin 5 Å of conserved residues (N47 and N48) in the SD.These interactions align with evidence that CaBP1 binding toInsP3R is mediated by the NT (15) and with both the IBC andSD being required for high-affinity binding (16, 18). The keyhydrophobic residues in CaBP1 (V101, L104, and V162) are lessexposed in the closed conformation of Ca2+-free CaBP1, con-sistent with the reduced affinity of CaBP1 for NT in the absenceof Ca2+ (18). The importance of the hydrophobic residues withinthe IBC was confirmed by mutagenesis. A triple mutant of NTCL

(L302S/I364S/L393S) bound InsP3, but not CaBP1 (Fig. S5 A–C),confirming that InsP3 and CaBP1 bind to distinct sites (18). In-teraction of the NT with Ca2+–CaBP1 via localized clusters ofhydrophobic residues probably explains why mutation of singleresidues in CaBP1 (V101S, L104S, and V162S) massively attenu-ates its binding to the NT (Fig. S3 B and C).

CaBP1 Forms a Ring Around the Cytosolic Entrance of the InsP3R.Native InsP3R is a tetramer with a central ion-conducting pore(34). Docking crystal structures of the NT (29) onto a cryo-EMstructure of full-length tetrameric InsP3R1 (34) suggests the ar-rangement shown in Fig. 2B, which is similar to that proposed forryanodine receptors (35). Overlaying the structure of the CaBP1/NT complex onto the tetrameric NT generates a structure inwhich four molecules of CaBP1 associate, via the clustered hy-drophobic residues in their C lobes, with the three clusteredhydrophobic residues in each of the four IBC-β domains. Thelatter contribute to the lining of the central cytosolic vestibule,and the four molecules of CaBP1 form a ring-like structurearound it. The position of the N lobe of CaBP1 within the tet-rameric complex could not be defined because NMR signalsassigned to it were unaffected by the NT. Previous studiesshowed that CaBP1 and its isolated C lobe bind to the NT withvery similar affinity (18), consistent with an absence of contactsbetween the NT and N lobe. The location of the N lobe withinthe complex was estimated by first generating an ensemble offull-length CaBP1 structures in which the two lobes are con-nected by a flexible linker and are free to adopt many differentrelative orientations during simulated annealing. This ensembleof full-length CaBP1 structures was then docked into the NTstructure (Fig. S4E). The CaBP1 C-lobe interaction with the NTis well defined in the ensemble of docked structures (cyan in Fig.S4E with rmsd = 0.5 Å), whereas the relative location of theN lobe in the complex is highly variable (Fig. S4E, blue). Eachstructure from the ensemble was then overlaid and docked intothe tetrameric NT structure. The lowest energy model (Fig. 2C)placed the N lobe above the C lobe and projecting into the cy-tosol away from any contact with the NT. This elongated orga-nization of the two lobes of CaBP1 differs from their compact,hexameric arrangement in the CaBP1 crystal structure (19).

Solution NMR has shown that, as with CaM (36) and troponin C(37), the CaBP1 N and C lobes fold independently and do notinteract structurally (18). This finding is consistent with a lackof contact between the lobes of CaBP1 when complexed withInsP3R (Fig. 2C).Functional analyses confirmed the importance of the critical

residues within CaBP1. CaBP1 with mutations to the key hy-drophobic residues [CaBP1(V101S/L104S) or CaBP1(V162S)](Fig. 2A) had no effect on InsP3-evoked Ca2+ release at any [Ca2+]cexamined, even when the CaBP1 concentrations were increasedto 100 μM (Fig. 3A, Fig. S5D, and Table S3). Furthermore, themutant CaBP1s did not affect inhibition of InsP3-evoked Ca2+

release by CaBP1 (Fig. 3B and Fig. S5E), confirming that theydo not compete with CaBP1 for binding to InsP3Rs. Further-more, in patch-clamp analyses of nuclear InsP3R, responses toInsP3 were unaffected by CaBP1(V162S) (Fig. 3C and Fig. S1F).These functional analyses support the proposed structure ofCaBP1 bound to InsP3R (Fig. 2).

CaBP1 Stabilizes Interactions Between the NTs of Tetramic InsP3R.Activation of InsP3R is proposed to begin with InsP3-stimulatedrearrangement of an intersubunit interface between the SD andIBC-β domain. This movement then disrupts interaction of the“hot-spot” (HS) loop (residues 165–180) of the SD with the IBC-βdomain of an adjacent subunit leading to channel gating (29) (Fig.4A). This model predicts key interactions at the intersubunitinterfaces, notably between residues within the SD (K168, L169,

Fig. 2. Structure of CaBP1 bound to InsP3R. (A) The C lobe of Ca2+-boundCaBP1 [cyan; Protein Data Bank (PDB) ID code 2K7D] bound to NT (PDB ID code3UJ4) in a 1:1 complex. Key residues at the binding interface are highlighted inmagenta (CaBP1) and red (InsP3R). NT subdomains are colored pink (SD), yel-low (IBC-β), and gray (IBC-α). (B) Model of tetrameric NT (pink, yellow, andgray) generated by superimposing the NT crystal structure (29) onto a cryo-EMstructure of InsP3R1 (34). NMR structural restraints were used to define con-tacts between eachNTandC lobe of CaBP1 (cyan,withCa2+ in orange). (C) Sideview of tetrameric NT bound to full-length CaBP1 in a 4:4 complex. The N lobeof CaBP1 (blue) is connected to the C lobe (cyan) by a flexible linker (red) thatallows the N lobe to adopt multiple orientations (indicated by the arrow).

Li et al. PNAS | May 21, 2013 | vol. 110 | no. 21 | 8509

BIOCH

EMISTR

Y

and R170) and IBC (T373, T374, D426, K427). We tested thisprediction by inserting pairs of Cys residues into NTCL andassessing their proximity by oxidative cross-linking with copper–phenanthroline (CuP). Three double Cys-substituted mutants(L169C/T373C, L169C/T374C, and K168C/K427C) were engi-neered as candidates for cross-linking in light of the modeledintersubunit interface (Fig. 4A). NTCL(L169C/T373C) pro-vided the most convincing evidence of a concentration- and time-dependent formation of tetramers in the presence of CuP (Fig.S6 A and B). NTCL or NTCL with a single Cys insertion (L169C orT373C) or a pair of Cys that are not expected to be in proximity(A61C/A553C) did not produce tetramers (Fig. S6C). These re-sults confirm the proximity of L169C/T373C to the tetrameric NTinterface and demonstrate the utility of CuP cross-linking for ana-lyses of intersubunit interactions between NTs. We usedNTCL(L169C/T373C) for subsequent analyses and confirmed thatits InsP3-binding affinity was similar to that of NTCL (Fig. S6D).InsP3R activation proceeds via disruption of intersubunit in-

teractions between NT domains (29). Our structure shows CaBP1forming a tetrameric cap tethered to the NTs of the tetramericInsP3R (Fig. 2B). This structure suggests that CaBP1 may lockInsP3Rs in a closed state by restricting the usual InsP3-evokeddisruption of intersubunit interactions. We used CuP cross-linkingand NTCL (L169C/T373C) to assess this possibility. As pre-dicted by our model, InsP3 caused a concentration-dependentinhibition of CuP-mediated formation of cross-linked tetra-meric NTCL(L169C/T373C) (Fig. 4B, Fig. S6E, and Table S4).The biologically inactive isomer of InsP3 (L-InsP3) had no

effect on cross-linking (Fig. 4B). In contrast, the C lobe ofCaBP1, which lacks endogenous Cys, caused a concentration-dependent increase in the rate and extent of formation of cross-linked tetrameric NTCL(L169C/T373C) (Fig. 4C, Fig. S6F, andTable S4). Similar results were obtained with a Cys-less form offull-length CaBP1 (CaBP1CL) (Fig. 4C, Fig. S6F, and Table S4).These results demonstrate that the C lobe of CaBP1 is largelyresponsible for the observed effects on NT cross-linking. TheCaBP1 C lobe in the absence of Ca2+ had no effect on NT tet-ramer cross-linking (Fig. 4D, Fig. S6G, and Table S4), and nei-ther did CaBP1(V162S), which did not bind the NT (Fig. S6Hand Table S4). CaBP1 also partially blocked the inhibition ofcross-linking by InsP3 (Fig. 4E, Fig. S6I, and Table S4). Theseresults (Fig. 4F and Table S4) support our suggestion that InsP3activates InsP3R by disrupting an intersubunit interface betweenthe SD and IBC-β. We suggest that, in the presence of Ca2+,CaBP1 forms a tetrameric cap on the InsP3R that restricts theseintersubunit movements and thereby stabilizes a closed state of thechannel (Fig. 4G).

DiscussionGating of InsP3Rs is regulated by InsP3 binding, but modulated bymany additional signals, notably Ca2+ and a variety of proteins (5),including such Ca2+-regulated proteins as CaM (12), CIB1(13), and CaBP1 (2). These proteins are either highly (CaM) orexclusively (CaBP1) expressed in neurons, where they have beenproposed to attenuate basal InsP3R activity (12), provide the Ca2+sensor for inhibitory feedback of InsP3Rs (5, 11), modulate InsP3Ractivity (5), or, for CaBP1 and CIB1, allow Ca2+ directly to gateInsP3Rs (5, 13). The latter suggested that, within neurons, InsP3Rs,like ryanodine receptors, might mediate regenerative Ca2+ signalswithout the need for coincident production of InsP3. Our resultsdemonstrate that CaBP1 does not directly activate InsP3R1, thepredominant InsP3R subtype in neurons (Fig. S1 A–D). CaBP1does, however, massively reduce InsP3-activated InsP3R activity bystabilizing a closed state of the channel (Fig. 1 C and D), an effectthat is enhanced by Ca2+ binding to both CaBP1 and the InsP3R(or a protein tightly associated with InsP3R). These dual effects ofCa2+ may allow cooperative inhibition of InsP3Rs by increasesin [Ca2+]c. However, even at resting [Ca2+]c, there is detectableinhibition of InsP3Rs by CaBP1 (Fig. 1B and Table S3), sug-gesting that CaBP1 may also contribute to setting the basal sen-sitivity of neuronal InsP3Rs.We identified hydrophobic residues within the C lobe of CaBP1

that become more exposed when CaBP1 binds Ca2+ (V101, L104,and V162) and showed by both NMR and functional analyses thatthey make essential contacts with hydrophobic residues in theIBC-β domain (L302, I364, and L393) (Fig. 2A). Additional minorcontacts between the CaBP1 C lobe and residues within the SDcontribute further to high-affinity binding of CaBP1 (18). Thehydrophobic interactions between CaBP1 and the IBC are es-sential for CaBP1 binding and inhibition of InsP3R (Fig. 3 andFigs. S1F and S5). Docking the NT–CaBP1 complex into thestructure of a full-length InsP3R reveals an arrangement in whichtetrameric CaBP1 is anchored by its hydrophobic contacts to theunderlying NT domains. CaBP1 thereby forms a ring-like struc-ture around the cytosolic vestibule that leads to the InsP3R pore(Fig. 2B). This arrangement has the Ca2+-binding sites of CaBP1lining the route through which Ca2+ passes via the InsP3R to thecytosol (Fig. 2 B and C).The InsP3-induced conformational change that initiates InsP3R

activation involves rearrangement of an interface between the SDand IBC-β domain. This intrasubunit rearrangement then disruptsan interaction between subunits mediated by the HS loop of theSD (29). This loop includes a residue (Y167 in InsP3R1) that isimportant for gating of InsP3Rs (38) and ryanodine receptors(39). Our cross-linking analyses support this scheme becauseresidues that contribute to the intersubunit interface are lessreadily cross-linked in the presence of InsP3 (Fig. 4B and Fig.S6E). Ca2+–CaBP1 has the opposite effect: It increases cross-linking (Fig. 4) and inhibits InsP3R gating by stabilizing a closed

Fig. 3. Hydrophobic residues in CaBP1 are essential for inhibition of InsP3R.(A) Inhibition of InsP3-evoked Ca2+ release by CaBP1 is abolished after mu-tation of its key hydrophobic residues. Permeabilized DT40–InsP3R1 cells inCLM with [Ca2+]c of 3.5 μM were incubated with the indicated concen-trations of CaBP1, CaBP1(V162S), or CaBP1(V101S/L104S) (10 min) beforeadding InsP3. Results show the concentration-dependent release of Ca2+ byInsP3. (B) Ca

2+ release evoked by 1 or 3 μM InsP3 alone, with CaBP1 (50 μM),or with CaBP1 (50 μM) and mutant CaBP1 (100 μM). Results (A and B) aremeans ± SEM; n = 4, with duplicate determinations in each. Similar resultsperformed in CLM with 1.2 μM [Ca2+]c are shown in Fig. S5 D and E. Summaryresults are in Table S3. (C) Typical patch-clamp recordings from single InsP3R1in medium with [Ca2+]c of 1.5 μM stimulated with InsP3 (10 μM) alone or incombination with CaBP1(V162S) (10 μM) shows that the mutant CaBP1 isinactive. Bars show the closed state. The holding potential was +40 mV.Summary results are shown in Fig. S1F. Fig. 2A shows the positions of mu-tated residues.

8510 | www.pnas.org/cgi/doi/10.1073/pnas.1220847110 Li et al.

state of the channel (Fig. 1). We suggest that CaBP1 counteractsthe InsP3-induced conformational change by “clamping” theunderlying InsP3R subunits and restricting their relative motion.We speculate that CaBP1 held loosely to neuronal InsP3R atresting [Ca2+]c tightens its grip as Ca2+ passing through an openInsP3R binds to CaBP1 to cause rapid feedback inhibition(Fig. 4G).

Materials and MethodsExpression and Purification of CaBP1, NT, and Their Mutants. The short formof CaBP1 was used throughout. CaBP1 and its mutants were expressed andpurified from Escherichia coli strain BL21(DE3) as described (40). NT (resi-dues 1-604) and NTCL (in which native Cys are replaced by Ala) from ratInsP3R1 were expressed and purified as described (29, 41). Individual Cysresidues were introduced into NTCL using the QuikChange site-directedmutagenesis kit. Sequences of all plasmids were confirmed. Table S1 liststhe proteins used.

NMR Spectroscopy. Samples were prepared by dissolving perdeuterated anduniformly 15N/13C-labeled CaBP1 (0.2 mM) containing protonated methylgroups (for Val, Leu, and Ile) (42) in 0.3 mL of 95% [2H]H2O containing 10 mM[2H11]Tris, pH 7.4, 0.1 mM KCl, and either 5 mM EDTA or 5 mM CaCl2 in thepresence of 0.2 mM NTCL or NTCL with single Cys substitutions (E20C, A61C,R170C, H289C, N300C, A394C, and K424C). Methyl TROSY experiments on13C-labeled CaBP1 bound to unlabeled NTCL were performed as described

(43). NMR-PRE experiments were performed on samples that containedisotopically labeled CaBP1 bound to NTCL with a single Cys insertion with orwithout an attached nitroxide spin label. Spin labeling was performed asdescribed (31). All NMR experiments were performed at 30 °C on a BrukerAvance 800 MHz spectrometer equipped with triple-resonance cryoprobeand z axis gradient. NMR assignments were described (18).

Molecular Docking. Atomic coordinates for the NT (PDB ID code 3UG4) andCaBP1 C lobe (PDB ID code 2K7D) were used to generate the docked structurein Fig. 2A. For docking of full-length CaBP1 to NT (Fig. 2C), an ensemble ofstructures of full-length CaBP1 (with a flexible linker between the two lobes)was generated by a simulated annealing protocol within CYANA using dis-tance restraints derived for the CaBP1 C lobe and N lobe (PDB ID code 2K7B).All docking calculations were performed by using the HADDOCK Guru in-terface (http://haddock.science.uu.nl/services/HADDOCK/haddock.php) (44).Mutagenesis data and chemical-shift perturbation data were used as inputsto define active and passive residues to generate ambiguous restraints (44).The PRE ratios (Fig. S4C) were converted into unambiguous restraints anddocking calculations were performed as described (45). The final dockedstructure in Fig. 2A is the average of 148 calculated structures that con-verged within a single cluster (red dots in Fig. S4D).

InsP3-Evoked Ca2+ Release and 3H-InsP3 Binding. DT40 cells stably expressingonly rat InsP3R1 (DT40–InsP3R1 cells) were loaded with a low-affinity luminalCa2+ indicator, permeabilized in cytosol-like medium (CLM), and the Ca2+

content of the endoplasmic reticulum was continuously monitored during

Fig. 4. Opposing effects of CaBP1 and InsP3 oninteractions between NTs. (A) Top view of the tet-rameric structure of NTs, and close-up view of theboxed area of the intersubunit interface betweenthe HS loop of the SD (magenta) and IBC-β (yellow)(29). The two residues that were replaced by Cys forCuP cross-linking analyses are boxed. (B) InsP3weakens the interactions between NT subunits andthereby the rate of CuP-mediated cross-linking oftetrameric NT. NTCL(L169C/T373C) (75 μM) in me-dium containing 5 mM CaCl2 was incubated withCuP (100 μM) alone or with the indicated concen-trations of InsP3 (the naturally occurring D-isomer)or biologically inactive L-InsP3. Results are expressedas fractions of the average intensity of the tetramerband detected at 60 min in the control incubation(no InsP3). The results with NTCL(L169C/T373C) alone(black) are shown for comparison. (C and D) Similarcross-linking experiments show that CaBP1 has theopposite effect to InsP3. Effects on tetramer for-mation of the indicated concentrations of theC lobe of CaBP1 (CaBP-C) and of the Cys-less form offull-length CaBP1 (CaBP1CL) in medium containing5 mM CaCl2 (C) or in Ca2+-free medium (D). (E) TheC-terminal of CaBP1 substantially blocks the de-stabilization of NT subunit interactions by InsP3.Effects of the indicated concentrations of InsP3 withCaBP1 C lobe (300 μM) in medium containing 5 mMCaCl2. Results are expressed as fractions of the av-erage intensity of the tetramer band detected at 60min in the control incubation (no InsP3 or CaBP1).The results with NTCL(L169C/T373C) alone (black) areshown for comparison. (F) Summary data showamounts of cross-linked NTCL(L169C/T373C) tetra-mer detected at 60 min relative to NTCL(L169C/T373C) alone at 60 min. Results in B–F show means ±standard deviation from three independent ex-periments. InsP3 denotes D-InsP3 unless indicatedotherwise. The data from which these analyses de-rive are shown in Fig. S6, and the rate constants andnormalized band intensities in Table S4. (G) Inter-actions between adjacent NTs mediated by IBC-β(yellow) and the HS loop of the SD (magenta) holdthe tetrameric InsP3R in a closed state. InsP3 bindingcloses the clam-like IBC, disrupting these inter-subunit interactions, and allowing the channel to open. The cytosolic vestibule of the InsP3R with four CaBP1s (cyan) bound is probably at least 5 Å across andunlikely to impede the flow of ions. Instead, we suggest that CaBP1 clamps the intersubunit interactions and thereby inhibits channel opening.

Li et al. PNAS | May 21, 2013 | vol. 110 | no. 21 | 8511

BIOCH

EMISTR

Y

additions of ATP (to allow Ca2+ uptake), CaBP1, and InsP3 as described(46). Ca2+ release evoked by InsP3 is expressed as a percentage of theATP-dependent Ca2+ uptake. 3H-InsP3 binding to rat cerebellar membranesor purified NT was performed in CLM at 4 °C as described (28). Results werefitted to Hill equations.

Patch-Clamp Recording. Currents were recorded from patches excised from theouter nuclear envelope of DT40–InsP3R1 cells by using symmetrical cesiummethanesulfonate (140 mM) as charge carrier. The composition of recordingsolutions and methods of analysis were otherwise as described (47).

Cross-Linking of Cys Residues. For CuP cross-linking, amixture of 50 mMCuSO4

and 65 mM 1,10-phenanthroline (Sigma) was freshly prepared. Concen-trations in the text refer to final CuP concentration (≥50 μM). NTCL or itsmutants (75 μM) were incubated on ice with CuP in medium containing360 mM NaCl, 20 mM Tris·HCl, pH 8.4, 2.5% (vol/vol) glycerol, and 0.2 mMTris(2-carboxyethyl)phosphine (TCEP). Ca2+-free buffer was prepared byusing Chelex 100 resin (Bio-Rad Laboratories). NTCL(L169C/T373C) andCaBP1 C lobe were used after dialysis in Ca2+-free buffer. Reactions were

quenched by addition of 10 mM N-ethylmaleimide and 10 mM EDTA (finalconcentrations). Samples were mixed with 4× nonreducing SDS loadingbuffer, heated at 55 °C for 15 min, and subjected to SDS/PAGE by usingNuPAGE 3–8% Tris·acetate gels (Invitrogen). After Coomassie Brilliant Bluestaining, band intensities of tetrameric NT were quantified by densitometryusing ImageJ. For analyses of time courses, the intensity of the tetramerband at each time for the experimental condition was expressed relative tothe intensity of the tetramer band at 60 min for control condition (no InsP3or CaBP1). These normalized intensities were fitted with a single exponentialtime course by using IGOR Pro-6 (WaveMetrics).

ACKNOWLEDGMENTS. We thank Dr. Jerry Dallas for help with NMR experi-ments. This work was supported by National Institutes of Health GrantsEY012347 and NS045909 (to J.B.A.) and RR11973 (to the University ofCalifornia at Davis NMR Facility); Canadian Institutes of Health Research(CIHR) and the Canada Foundation for Innovation grants (to M.I.); WellcomeTrust Grant 085295; Biotechnology and Biological Sciences Research CouncilGrant BB/H009736/1 (to C.W.T.). A.M.R. is a Fellow of Queens’ College. T.R.was a Fellow of Pembroke College.

1. Berridge MJ, Bootman MD, Roderick HL (2003) Calcium signalling: Dynamics,homeostasis and remodelling. Nat Rev Mol Cell Biol 4(7):517–529.

2. Haynes LP, McCue HV, Burgoyne RD (2012) Evolution and functional diversity of theCalcium Binding Proteins (CaBPs). Front Mol Neurosci 5:e9.

3. Barbara JG (2002) IP3-dependent calcium-induced calcium release mediates bidirectionalcalcium waves in neurones: Functional implications for synaptic plasticity. BiochimBiophys Acta 1600(1-2):12–18.

4. Taylor CW, Dale P (2012) Intracellular Ca2+ channels—a growing community. Mol CellEndocrinol 353(1-2):21–28.

5. Foskett JK, White C, Cheung KH, Mak DO (2007) Inositol trisphosphate receptor Ca2+

release channels. Physiol Rev 87(2):593–658.6. Taylor CW, Tovey SC (2010) IP3 receptors: Toward understanding their activation. Cold

Spring Harb Perspect Biol 2(12):a004010.7. Konieczny V, Keebler MV, Taylor CW (2012) Spatial organization of intracellular Ca2+

signals. Semin Cell Dev Biol 23(2):172–180.8. Miyakawa T, et al. (2001) Ca2+-sensor region of IP3 receptor controls intracellular

Ca2+ signaling. EMBO J 20(7):1674–1680.9. Taylor CW, Laude AJ (2002) IP3 receptors and their regulation by calmodulin and

cytosolic Ca2+. Cell Calcium 32(5-6):321–334.10. Adkins CE, et al. (2000) Ca2+-calmodulin inhibits Ca2+ release mediated by type-1, -2

and -3 inositol trisphosphate receptors. Biochem J 345(Pt 2):357–363.11. Michikawa T, et al. (1999) Calmodulin mediates calcium-dependent inactivation of

the cerebellar type 1 inositol 1,4,5-trisphosphate receptor. Neuron 23(4):799–808.12. Patel S, Morris SA, Adkins CE, O’Beirne G, Taylor CW (1997) Ca2+-independent in-

hibition of inositol trisphosphate receptors by calmodulin: Redistribution of cal-modulin as a possible means of regulating Ca2+ mobilization. Proc Natl Acad Sci USA94(21):11627–11632.

13. White C, Yang J, Monteiro MJ, Foskett JK (2006) CIB1, a ubiquitously expressedCa2+-binding protein ligand of the InsP3 receptor Ca2+ release channel. J BiolChem 281(30):20825–20833.

14. Schlecker C, et al. (2006) Neuronal calcium sensor-1 enhancement of InsP3 receptoractivity is inhibited by therapeutic levels of lithium. J Clin Invest 116(6):1668–1674.

15. Yang J, et al. (2002) Identification of a family of calcium sensors as protein ligands ofinositol trisphosphate receptor Ca2+ release channels. Proc Natl Acad Sci USA 99(11):7711–7716.

16. Kasri NN, et al. (2004) Regulation of InsP3 receptor activity by neuronal Ca2+-bindingproteins. EMBO J 23(2):312–321.

17. Haeseleer F, et al. (2000) Five members of a novel Ca2+-binding protein (CABP) sub-family with similarity to calmodulin. J Biol Chem 275(2):1247–1260.

18. Li C, et al. (2009) Structural insights into Ca2+-dependent regulation of inositol 1,4,5-trisphosphate receptors by CaBP1. J Biol Chem 284(4):2472–2481.

19. Findeisen F, Minor DL, Jr. (2010) Structural basis for the differential effects of CaBP1 andcalmodulin on CaV1.2 calcium-dependent inactivation. Structure 18(12):1617–1631.

20. Haynes LP, Tepikin AV, Burgoyne RD (2004) Calcium-binding protein 1 is an inhibitorof agonist-evoked, inositol 1,4,5-trisphosphate-mediated calcium signaling. J BiolChem 279(1):547–555.

21. Menger N, Seidenbecher CI, Gundelfinger ED, Kreutz MR (1999) The cytoskeleton-associated neuronal calcium-binding protein caldendrin is expressed in a subset ofamacrine, bipolar and ganglion cells of the rat retina. Cell Tissue Res 298(1):21–32.

22. Seidenbecher CI, Reissner C, Kreutz MR (2002) Caldendrins in the inner retina. AdvExp Med Biol 514:451–463.

23. Lee A, et al. (2002) Differential modulation of Cav2.1 channels by calmodulin andCa2+-binding protein 1. Nat Neurosci 5(3):210–217.

24. Zhou H, Yu K, McCoy KL, Lee A (2005) Molecular mechanism for divergent regulationof Cav1.2 Ca2+ channels by calmodulin and Ca2+-binding protein-1. J Biol Chem280(33):29612–29619.

25. Kinoshita-Kawada M, et al. (2005) Inhibition of TRPC5 channels by Ca2+-bindingprotein 1 in Xenopus oocytes. Pflugers Arch 450(5):345–354.

26. Bosanac I, et al. (2002) Structure of the inositol 1,4,5-trisphosphate receptor bindingcore in complex with its ligand. Nature 420(6916):696–700.

27. Bosanac I, et al. (2005) Crystal structure of the ligand binding suppressor domain oftype 1 inositol 1,4,5-trisphosphate receptor. Mol Cell 17(2):193–203.

28. Rossi AM, et al. (2009) Synthetic partial agonists reveal key steps in IP3 receptor ac-tivation. Nat Chem Biol 5(9):631–639.

29. Seo MD, et al. (2012) Structural and functional conservation of key domains in InsP3and ryanodine receptors. Nature 483(7387):108–112.

30. Rahman T, Taylor CW (2010) Nuclear patch-clamp recording from inositol 1,4,5-trisphosphate receptors. Methods Cell Biol 99:199–224.

31. Clore GM, Tang C, Iwahara J (2007) Elucidating transient macromolecular interactionsusing paramagnetic relaxation enhancement. Curr Opin Struct Biol 17(5):603–616.

32. Dominguez C, Boelens R, Bonvin AM (2003) HADDOCK: A protein-protein dockingapproach based on biochemical or biophysical information. J Am Chem Soc 125(7):1731–1737.

33. Ikura M (1996) Calcium binding and conformational response in EF-hand proteins.Trends Biochem Sci 21(1):14–17.

34. Ludtke SJ, et al. (2011) Flexible architecture of IP3R1 by cryo-EM. Structure 19(8):1192–1199.

35. Tung CC, Lobo PA, Kimlicka L, Van Petegem F (2010) The amino-terminal diseasehotspot of ryanodine receptors forms a cytoplasmic vestibule. Nature 468(7323):585–588.

36. Zhang M, Tanaka T, Ikura M (1995) Calcium-induced conformational transitionrevealed by the solution structure of apo calmodulin. Nat Struct Biol 2(9):758–767.

37. Gagné SM, Tsuda S, Li MX, Smillie LB, Sykes BD (1995) Structures of the troponin Cregulatory domains in the apo and calcium-saturated states. Nat Struct Biol 2(9):784–789.

38. Yamazaki H, Chan J, Ikura M, Michikawa T, Mikoshiba K (2010) Tyr-167/Trp-168 intype 1/3 inositol 1,4,5-trisphosphate receptor mediates functional coupling betweenligand binding and channel opening. J Biol Chem 285(46):36081–36091.

39. Amador FJ, et al. (2009) Crystal structure of type I ryanodine receptor amino-terminalbeta-trefoil domain reveals a disease-associated mutation “hot spot” loop. Proc NatlAcad Sci USA 106(27):11040–11044.

40. Wingard JN, et al. (2005) Structural analysis of Mg2+ and Ca2+ binding to CaBP1,a neuron-specific regulator of calcium channels. J Biol Chem 280(45):37461–37470.

41. Chan J, et al. (2007) Ligand-induced conformational changes via flexible linkers in theamino-terminal region of the inositol 1,4,5-trisphosphate receptor. J Mol Biol 373(5):1269–1280.

42. Tugarinov V, Kay LE (2004) An isotope labeling strategy for methyl TROSY spectros-copy. J Biomol NMR 28(2):165–172.

43. Tugarinov V, Sprangers R, Kay LE (2004) Line narrowing in methyl-TROSY using zero-quantum 1H-13C NMR spectroscopy. J Am Chem Soc 126(15):4921–4925.

44. de Vries SJ, van Dijk M, Bonvin AM (2010) The HADDOCK web server for data-drivenbiomolecular docking. Nat Protoc 5(5):883–897.

45. Battiste JL, Wagner G (2000) Utilization of site-directed spin labeling and high-reso-lution heteronuclear nuclear magnetic resonance for global fold determination oflarge proteins with limited nuclear overhauser effect data. Biochemistry 39(18):5355–5365.

46. Tovey SC, Sun Y, Taylor CW (2006) Rapid functional assays of intracellular Ca2+

channels. Nat Protoc 1(1):259–263.47. Taufiq-Ur-Rahman, Skupin A, Falcke M, Taylor CW (2009) Clustering of InsP3 receptors

by InsP3 retunes their regulation by InsP3 and Ca2+. Nature 458(7238):655–659.

8512 | www.pnas.org/cgi/doi/10.1073/pnas.1220847110 Li et al.

Supporting InformationLi et al. 10.1073/pnas.1220847110

Fig. S1. Ca2+-binding protein 1 (CaBP1) alone does not stimulate Ca2+ release via inositol 1,4,5-trisphosphate receptor (InsP3R) or affect3H–InsP3 binding to

InsP3R1. (A–C) CaBP1 has no effect on the Ca2+ content of the intracellular stores of permeabilized DT40–InsP3R1 cells. Typical traces from populations ofpermeabilized DT40–InsP3R1 cells show fluorescence (RFU, relative fluorescence units) recorded from a luminal Ca2+ indicator (Mag-fluo 4) during addition of

Legend continued on following page

Li et al. www.pnas.org/cgi/content/short/1220847110 1 of 10

ATP (1.5 mM) to allow active Ca2+ sequestration, and then addition of the indicated concentrations of CaBP1. The code shown in A, defining the concentrationsof CaBP1, applies to A–C. The experiments were performed in cytosol-like medium (CLM) with the cytosolic free Ca2+ concentration ([Ca2+]c) buffered as shown.Ca2+-free CLM had the following composition: 20 mM NaCl, 140 mM KCl, 1 mM EGTA, 20 mM Pipes, 2 mM MgCl2, pH 7.0; it was supplemented with CaCl2 togive appropriate free [Ca2+], which was measured using fluo-4 (KD

Ca = 325 nM at 22 °C) or Mag-fluo-4 (KDCa = 22 μM at 22 °C). (D) Summary results show the

Ca2+ content of the intracellular stores measured 30 s after addition of CaBP1 for the two highest concentrations of CaBP1 used. Results are means ± SEM fromthe number of independent experiments shown in parentheses. (E) Typical records from interleaved patch-clamp recordings of excised nuclei from DT40–InsP3R1 cells show that CaBP1 alone does not activate InsP3R. The pipette solution contained InsP3 (10 μM) with a free [Ca2+]c of 200 nM (a; 7/16 responded) or1.5 μM (c; 5/14 responded), or with InsP3 replaced by CaBP1 (10 μM), when 0/10 (b) and 0/28 (d) patches responded. The closed state in each recording is shownby a bar. The holding potential was +40 mV. (F) Summary results show single-channel properties of InsP3R1 recorded in cytosolic medium containing InsP3(10 μM), ATP (0.5 mM), and [Ca2+]c of 1.5 μM alone or with 10 μM CaBP1 or CaBP1(V162S). Sample sizes are shown in parentheses, where channel activity (NPo)is reported for only those patches in which an active InsP3R was detected. (G) CaBP1 has no effect on InsP3 binding to InsP3R1. Specific binding of 3H–InsP3(1.5 nM) to rat cerebellar membranes, which express almost entirely InsP3R1, in CLM with 250 nM [Ca2+]c is shown in the presence of the indicated concen-trations of unlabeled InsP3. The free [Ca2+] of CLM was measured using fluo-4 (KD

Ca = 456 nM at 4 °C) or Mag-fluo-4 (KDCa = 32 μM at 4 °C). Results were fitted

to Hill equations from which half-maximal inhibitory concentrations (IC50) and thereby pKDs were calculated. The KD for InsP3 is 14 nM (pKD = 7.84 ± 0.15; Hillcoefficient = 0.8 ± 0.1; n = 4). (H and I) Specific binding of 3H–InsP3 (1.5 nM) to cerebellar membranes in CLM is shown for the indicated [Ca2+]c alone or in thepresence of 10 μM (H) or 50 μM (I) CaBP1. Results (means ± SEM; n = 4–5, with triplicate determinations for each) are expressed as percentages of the specificbinding determined under identical conditions in the absence of CaBP1. Nonspecific binding was defined by addition of 1 μM InsP3.

Li et al. www.pnas.org/cgi/content/short/1220847110 2 of 10

Fig. S2. Ca2+-modulated inhibition of InsP3-evoked Ca2+ release by CaBP1134 and CaBP134. (A–D) The effects of normal CaBP1 or CaBP1 that cannot bind Ca2+

(CaBP1134) demonstrate that Ca2+ binding to CaBP1 is not essential for inhibition of InsP3-evoked Ca2+ release. Populations of permeabilized DT40–InsP3R1 cellswere incubated in CLM containing the indicated [Ca2+]c, with 50 μM CaBP1 or CaBP1134 (10 min) before addition of InsP3. Ca

2+ release (percentage of ATP-dependent Ca2+ uptake) is shown as means ± SEM from three or four experiments, each performed in duplicate. The code shown in A applies also to B–D. (E)Experiments similar to those shown in A–D were used to examine the effects of CaBP134 on InsP3-evoked Ca2+ release in CLM with [Ca2+]c of 3.5 μM. Results(means ± SEM, n = 3–7, with duplicate determinations in each) show the Ca2+ release evoked under each condition expressed as a percentage of the responseevoked by the concentration of InsP3 that evoked half-maximal Ca2+ release (EC50) under control conditions. *P < 0.05 relative to the response evoked by InsP3alone. The residues mutated in CaBP1134 and CaBP134 are described in Table S1. (F) There is no detectable binding of Ca2+ to CaBP1134 measured by isothermaltitration calorimetry (ITC) using the procedures and conditions described (1).

1. Li C, et al. (2009) Structural insights into Ca2+-dependent regulation of inositol 1,4,5-trisphosphate receptors by CaBP1. J Biol Chem 284(4):2472–2481.

Li et al. www.pnas.org/cgi/content/short/1220847110 3 of 10

Fig. S3. Binding of CaBP1 and its mutants to N-terminal domain (NT; residues 1–604) and the NT in which all Cys residues were replaced by Ala (NTCL). ITC wasused to measure affinities of CaBP1 binding to NT using described methods (1). (A) Typical ITC data showing binding of CaBP1 to NTCL in 5 mM CaCl2. The 25additions of CaBP1 (0.5 mM in 10 μL) to NTCL (50 μM in 1.5 mL) (Upper) and integrated binding isotherm (Lower) are shown. The binding isotherm was fit toa one-site model with 1:1 stoichiometry. (B) Similar ITC analyses showing the binding of NT to CaBP1 or its mutants. The 25 additions of CaBP1 (1.0 mM in 10 μL)to NT (100 μM in 1.5 mL) (Upper) and integrated binding isotherm (Lower) are shown. The very weak signals for the mutant CaBP1s are similar to backgroundsignals (heat of dilution). Traces are representative of three independent experiments. (C) Summary ITC results (means ± standard deviation; n = 3; showing KD

and ΔH for the indicated interactions). ND denotes no detectable specific heat signal.

1. Li C, et al. (2009) Structural insights into Ca2+-dependent regulation of inositol 1,4,5-trisphosphate receptors by CaBP1. J Biol Chem 284(4):2472–2481.

Li et al. www.pnas.org/cgi/content/short/1220847110 4 of 10

Fig. S4. NMR analysis of Ca2+-bound C lobe of CaBP1 binding to NTCL. The sample conditions for each NMR experiment are defined in Materials and Methods.(A) Overlay of 13C–1H methyl transverse relaxation-optimized spectroscopy (TROSY) NMR spectra of 13C-labeled CaBP1 (0.2 mM) in the absence of NT (green) orin the presence of unlabeled NTCL (0.2 mM; red) or NT (0.2 mM; blue). The NMR resonance assignments were obtained from ref. 1. (B) Overlay of 13C–1H methylTROSY NMR spectra of 13C-labeled CaBP1 bound to NTCL(E20C) in the absence (black) and presence (red) of attached nitroxide spin label. (C) Paramagneticrelaxation enhancement (PRE) values (IParamagnetic/IDiamagnetic; mean ± standard deviation; n = 3) are listed for the indicated side-chain methyl resonances fromCaBP1 bound to NTCL with a spin label attached at either E20C, R107C, A394C, or K424C. ND denotes a PRE value that is not detectable because the NMRintensity in the absence of spin label (IDiamagnetic) could not be detected above background noise. (D) Statistics for the docking calculation of CaBP1 C lobebound to NT in a 1:1 complex (Fig. 2A). A total of 200 calculations were performed as described inMaterials and Methods. The relative energy of each structure(HADDOCK score) is plotted vs. the interface rmsd (i-rmsd). Nearly 75% of the calculated structures (148 of 200) have similar i-rmsd values, indicating that eachstructure within this cluster (red dots) has a similar docked arrangement. The structures in the lowest energy cluster have an overall rmsd of 0.5 Å relative to themean structure. The HADDOCK score was 65 ± 8 for the converged cluster, and the intermolecular energy ranged from 375 to 600 kcal/mole. (E) Moleculardocking of full-length CaBP1 into the crystal structure of NTCL. An ensemble of structures for full-length CaBP1 (C and N lobes connected by a flexible linker)was first generated by simulated annealing. The ensemble of structures for full-length CaBP1 (C lobe, cyan; N lobe, blue) was then docked into the crystalstructure of NTCL (gray; Protein Data Bank ID code 3UJ4). In the lowest energy structures, the C lobe of CaBP1 interacts directly with NTCL as it does in Fig. 2A.The N lobe of CaBP1 occupies many different relative positions in the ensemble (it is dynamically disordered) and does not make direct contact with NTCL.

1. Li C, et al. (2009) Structural insights into Ca2+-dependent regulation of inositol 1,4,5-trisphosphate receptors by CaBP1. J Biol Chem 284(4):2472–2481.

Li et al. www.pnas.org/cgi/content/short/1220847110 5 of 10

Fig. S5. Critical hydrophobic residues within the NT and CaBP1 are required for CaBP1-mediated inhibition of InsP3R. (A) Specific binding of 3H–InsP3 (1.5 nM)to NTCL and NTCL(L302S/I364S/L393S) in CLM with 250 nM [Ca2+]c (B) Summary results confirm that both proteins bind InsP3 with the same affinity. Results showKD, pKD (/M, mean ± SEM), and Hill coefficient (mean ± SEM) from three independent determinations. (C) NTCL(L302S/I364S/L393S) binds InsP3 but not CaBP1.ITC analysis shows binding of CaBP1 to NTCL, but no detectable binding to NTCL(L302S/I364S/L393S) with 100 μM CaBP1 (Fig. S3C). (D) CaBP1s with the criticalhydrophobic residues mutated do not inhibit InsP3-evoked Ca2+ release. The experiments are similar to those shown in Fig. 3 A and B, but with CLM containinga [Ca2+]c of 1.2 μM. Permeabilized DT40–InsP3R1 cells were incubated with the indicated concentrations of CaBP1, CaBP1(V162S), or CaBP1(V101S/L104S)(10 min) before addition of InsP3. Results (means ± SEM; n = 4; with duplicate determinations in each) show the concentration-dependent release of Ca2+ by InsP3.(E) Similar experiments show that mutant CaBP1 does not block inhibition of InsP3-evoked Ca2+ release by wild-type CaBP1. Ca2+ release evoked by 1 or 3 μMInsP3 alone, with CaBP1 (50 μM), or with CaBP1 (50 μM) in combination with mutant CaBP1 (100 μM) is shown. Results are means ± SEM from three independentexperiments. The results (summarized in Table S3) demonstrate that the mutant CaBP1s do not bind to the site through which CaBP1 inhibits InsP3R activity.

Li et al. www.pnas.org/cgi/content/short/1220847110 6 of 10

Fig. S6. Cross-linking of NTCL Cys-substituted mutants with CuP. (A and B) NTCL(L169C/T373C) was incubated on ice (30 min) with the indicated concentrationsof CuP (A) or with CuP (100 μM) for the indicated times (B). Typical results show the cross-linking assessed after SDS/PAGE and Coomassie Brilliant Blue staining.Molecular mass markers are shown on the left (in kilodaltons). Arrowheads show tetrameric, trimeric, dimeric, and monomeric NT. Trimers were not detectedwith concentrations of CuP ≤100 μM. Small amounts of larger oligomers were also detected, but only with the highest concentrations of CuP (≥200 μM) (A) orthe most prolonged incubations (≥45 min) (B). These scarce higher-order oligomers do not affect our analyses of tetramer cross-linking in the presence of InsP3and CaBP1 because this cross-linking was maximal within 45 min. We note that CuP never cross-links NTCL(L169C/T373C) to completion because the medium weoptimized for resolving the time courses includes a reducing agent [Tris(2-carboxyethyl)phosphine (TCEP)] that allows cleavage of disulfide bonds. Cross-linkingreactions therefore reach a steady-state, rather than proceeding to completion. (C) Similar analyses of NTCL(L169C), NTCL(T373C), and NTCL(A61C/A553C) using300 μM CuP for 0 or 60 min. (D) Specific binding of 3H–InsP3 (1.5 nM) is shown for NTCL and NTCL(L169C/T373C) in CLM with 250 nM [Ca2+]c. Results show KD

(mean), pKD, and Hill coefficient (h; means ± SEM) from three independent determinations. (E) Effects of the indicated concentrations of InsP3 (the activeD-isomer) or L-InsP3 (the biologically inactive isomer) on cross-linking of NTCL(L169C/T373C) in medium containing 5 mM CaCl2. C denotes the band detectedafter 60 min without InsP3. (F and G) Typical gels showing only the tetrameric NT bands after cross-linking of NTCL(L169C/T373C) (75 μM) incubated with theindicated concentrations of the C lobe of CaBP1 (CaBP1-C) or full-length Cys-less CaBP1 (CaBP1CL; 300 μM) in medium containing 5 mM CaCl2 (F) or no Ca2+ (G).(H) Similar analysis, in medium containing 5 mM CaCl2, of the effects of CaBP-C(V162S) (300 μM) and the normalized time course (mean ± standard deviation;n = 3) of tetramer cross-linking in the presence (red) or absence (black) of CaBP-C(V162S). (I) Effects of the indicated concentrations of InsP3 on cross-linking ofNTCL(L169C/T373C) with 300 μM CaBP1-C in medium containing 5 mM CaCl2. C denotes the band detected after 60 min without InsP3. Incubations (E–I) included100 μM CuP. InsP3 denotes D-InsP3 unless specified otherwise.

Li et al. www.pnas.org/cgi/content/short/1220847110 7 of 10

Table S1. CaBP1 and NT proteins used

Protein Description

CaBP1 Full-length, wild-type short form of CaBP1, residues 1–167CaBP1 C lobe Residues 96–167 of CaBP1CaBP1(V101S) CaBP1 with V101 replaced by SerCaBP1(L104S) CaBP1 with L104 replaced by SerCaBP1(V162S) CaBP1 with V162 replaced by SerCaBP1(V101S/L104S) CaBP1 with V101 and L104 replaced by SerCaBP1(V101S/V162S) CaBP1 with V101 and V162 replaced by SerCaBP1 (V101S/L104S/V162S) CaBP1 with V101, L104 and V162 replaced by SerCaBP134 CaBP1 with Ca2+ binding to EF3 and EF4 disabled by replacing E123 and E160 with GlnCaBP1134 CaBP1 with Ca2+ binding to EF1, EF3, and EF4 disabled by replacing D35 with Ala and E123 and E160 with GlnCaBP1CL CaBP1 with all Cys (4, 44, and 50) replaced by Ser (4, 44) and Ala (50)NT Residues 1–604 of InsP3R1NTCL Residues 1–604 of InsP3R1 with all Cys replaced by AlaNTCL(E20C) NTCL with E20 replaced by CysNTCL(A61C) NTCL with A61 replaced by CysNTCL(R170C) NTCL with R170 replaced by CysNTCL(H289C) NTCL with H289 replaced by CysNTCL(N300C) NTCL with N300 replaced by CysNTCL(A394C) NTCL with A394 replaced by CysNTCL(K424C) NTCL with K424 replaced by CysNTCL(A553C) NTCL with A553 replaced by CysNTCL(L302S/I364S/L393S) NTCL with L302, I364 and L393 replaced by SerNTCL(L169C) NTCL with L169 replaced by CysNTCL(T373C) NTCL with T373 replaced by CysNTCL(L169C/T373C) NTCL with L169 and T373 replaced by CysNTCL(L169C/T374C) NTCL with L169 and T374 replaced by CysNTCL(K168C/K427C) NTCL with K168 and K427 replaced by CysNTCL(A61C/A553C) NTCL with A61 and A553 replaced by Cys

EF hands (EF1, EF2, EF3, and EF4) are the Ca2+-binding motifs of CaBP1.

Li et al. www.pnas.org/cgi/content/short/1220847110 8 of 10

Table

S2.

EffectsofCa2

+oninhibitionofInsP

3-evo

kedCa2

+releasebyCaB

P1

[CaB

P1],

μM

[Ca2

+] c

50nM

230nM

1.2μM

3.5μM

pEC

50

EC50,

μMMax

imal

release,

%n

ΔpEC

50

pEC

50

EC50,

μMMax

imal

release,

%n

ΔpEC

50

pEC

50

EC50,

μMMax

imal

release,

%n

ΔpEC

50

pEC

50

EC50,

μMMax

imal

release,

%n

ΔpEC

50

06.62

±0.19

0.24

77±

114

7.05

±0.06

0.09

69±

66

6.29

±0.04

0.51

69±

23

5.67

±0.09

2.11

74±

25

106.54

±0.14

0.29

74±

74

0.08

±0.15

6.42

±0.20

*0.38

77±

86

0.63

±0.23

5.89

±0.15

*1.29

52±

63

0.40

±0.15

4.78

±0.29

*16

.60

68±

55

0.89

±0.31

506.61

±0.26

0.24

63±

74

0.01

±0.11

6.40

±0.18

*0.40

74±

66

0.65

±0.20

5.51

±0.23

*3.10

63±

63

0.78

±0.23

4.32

±0.47

*47

.86

56±

64

1.35

±0.42

Simila

rex

perim

ents

tothose

shownin

Fig.1

Awereusedto

mea

sure

theeffectsofCaB

P1onInsP

3-evo

kedCa2

+releasein

CLM

withtheindicated

[Ca2

+] c.R

esultsshow

pEC

50(/M;m

eans±SE

M),EC

50(m

eans),

andthemax

imal

response

(aspercentages

oftheCa2

+co

ntentoftheintracellularstores)(m

eans±SE

M)ofnindep

enden

tex

perim

ents

each

withduplicatedeterminations.ΔpEC

50isthedifference

betwee

nthe

pEC

50va

lues

determined

intheab

sence

andpresence

ofCaB

P1(Δ

pEC

50=

pEC

50co

ntrol−

pEC

50CaBP1).*P

<0.05

relative

toresultsobtained

intheab

sence

ofCaB

P1.

Li et al. www.pnas.org/cgi/content/short/1220847110 9 of 10

Table S3. Inhibition of InsP3-evoked Ca2+ release by CaBP1, but not its mutants

μM

CaBP1 CaBP1(V162S) CaBP1(V101S/L104S)

pEC50

Maximalrelease ΔpEC50 pEC50

Maximalrelease ΔpEC50 pEC50

Maximalrelease ΔpEC50

[Ca2+]c, 1.2 μM0 6.18 ± 0.18 (15) 65 ± 2 (15) 6.18 ± 0.18 (15) 65 ± 2 (15) 6.18 ± 0.18 (15) 65 ± 2 (15)10 6.17 ± 0.44 (5) 55 ± 4 (5) 0.01 ± 0.43 ND ND ND ND ND ND50 4.94 ± 0.30* (6) 64 ± 2 (6) 1.24 ± 0.35 6.26 ± 0.26 (4) 71 ± 2 (4) −0.08 ± 0.38 6.09 ± 0.36 (5) 64 ± 2 (5) 0.09 ± 0.38100 ND ND ND 6.33 ± 0.17 (4) 71 ± 2 (4) −0.15 ± 0.37 6.19 ± 0.37 (5) 61 ± 2 (5) −0.01 ± 0.38

[Ca2+]c, 3.5 μM0 5.69 ± 0.06 (12) 67 ± 2 (12) 5.69 ± 0.06 (12) 67 ± 2(12) 5.69 ± 0.06 (12) 67 ± 2 (12)10 4.78 ± 0.29* (5) 68 ± 5 (5) 0.91 ± 0.20 ND ND ND ND ND ND50 4.32 ± 0.47* (4) 56 ± 6 (4) 1.37 ± 0.27 5.66 ± 0.08 (4) 61 ± 1(4) 0.03 ± 0.12 5.52 ± 0.12 (4) 62 ± 5 (4) 0.17 ± 0.13100 ND ND ND 5.48 ± 0.16 (4) 69 ± 5 (4) 0.21 ± 0.14 5.68 ± 0.13 (4) 70 ± 7 (4) 0.01 ± 0.13

Summary results show the effects of the indicated concentrations of CaBP1 and its mutants on InsP3-evoked Ca2+ release in CLM with a [Ca2+]c of 1.2 or 3.5μM. Results show pEC50 (/M; means ± SEM) and maximal release (as percentages of the Ca2+ content of the stores; means ± SEM) for InsP3-evoked Ca2+ releasefrom (n) independent experiments, each with duplicate determinations. ΔpEC50 = pEC50

control − pEC50CaBP1. *P < 0.05 relative to results obtained in the absence

of CaBP1. ND, not determined.

Table S4. Effects of CaBP1 and InsP3 on kinetics of CuP-mediated cross-linking of tetramericNTCL(L169C/T373C)

[CaBP1], mM [InsP3], mM k, min−1 k/[CaBP1], min−1/mMNormalized maximal

intensity

Ca2+-CaBP1-C0 0 0.047 ± 0.006 — 1.000.075 0 0.071 ± 0.010 0.94 ± 0.13 1.32 ± 0.020.15 0 0.137 ± 0.017 0.92 ± 0.11 1.77 ± 0.030.3 0 0.271 ± 0.032 0.90 ± 0.11 2.04 ± 0.03

Ca2+-CaBP1CL

0.3 0 0.234 ± 0.022 0.78 ± 0.07 1.99 ± 0.02CaBP1-C0 0 0.047 ± 0.004 — ND0.3 0 0.056 ± 0.006 0.19 ± 0.02 1.01 ± 0.02

Ca2+-CaBP1C(V162S)0.053 0 0.053 ± 0.003 0.18 ± 0.02 1.03 ± 0.01

Ca2+-CaBP1-C0 0.05 0.042 ± 0.006 — 1.02 ± 0.020 0.5 0.090 ± 0.012 — 0.47 ± 0.020 1.5 0.118 ± 0.026 — 0.26 ± 0.030 1.5 L-InsP3 0.049 ± 0.007 0.96 ± 0.020.3 0.05 0.094 ± 0.007 0.31 ± 0.02 2.02 ± 0.010.3 0.5 0.062 ± 0.008 0.21 ± 0.03 1.40 ± 0.030.3 1.5 0.064 ± 0.010 0.21 ± 0.03 0.75 ± 0.02

Results show the rate constant (k), k/[CaBP1], and normalized maximal intensity [ratio of tetramericNTCL(L169C/T373C) band intensity to that of NTCL(L169C/T373C) alone]. Results are means ± standard deviationfrom three independent measurements. InsP3 refers to D-InsP3 unless indicated otherwise. ND, not determined.

Li et al. www.pnas.org/cgi/content/short/1220847110 10 of 10