Maurocalcine and domain A of the II-III loop of the dihydropyridine receptor Cav 1.1 subunit share common binding sites on the skeletal ryanodine receptor

Maurocalcine is a scorpion venom toxin of 33 amino acid residues that bears a striking resemblance to the domain A. This domain belongs to the II-III loop of Ca 1.1 which is implicated in excitation-contraction coupling. Besides the structural homology, v maurocalcine also modulates RyR1 channel activity in a manner akin to a synthetic peptide of domain A. Owing to these similarities, we hypothesized that maurocalcine and domain A may bind onto an identical region(s) of RyR1. Using a set of RyR1 fragments, we demonstrate that peptide A and maurocalcine bind onto two discrete RyR1 regions: fragments 3 and 7 encompassing amino acid residues 1021-1631 and 3201-3661, respectively. The binding onto fragment 7 is of greater importance and was thus further investigated. We found that the amino acid region 3350-3501 of RyR1 (fragment 7.2) is sufficient for these interactions. Proof that peptide A and maurocalcine bind onto the same site is provided by competition experiments in which binding of fragment 7.2 to peptide A is inhibited by preincubation with maurocalcine. At the functional level, deletion of fragment 7 abolishes the maurocalcine induced stimulation of H -ryanodine binding onto microsomes of COS-7 cells transfected with RyR1 without affecting the caffeine [3 ] response. MESH


INTRODUCTION
In skeletal muscle cells, the activation of the plasma membrane dihydropyridine receptor induces, through the ryanodine receptor (RyR1), a massive release of the calcium stored in the terminal cisternae of the sarcoplasmic reticulum (SR). The set of events, that starts with the depolarization of the plasma membrane, sensed by the DHPR, and ends with the opening of the RyR1, is called excitation-contraction (EC) coupling. Several results have led to the conclusion that the skeletal muscle EC coupling is based on a mechanical process that is made possible by the physical interactions of the two Ca channels ( ). The transmission of information from 2+ 1 the DHPR to the RyR1 has been termed orthograde signal. In addition, the same physical interactions, which occur between RyR1 and DHPR, have also been proposed to be responsible for a retrograde signal in which the DHPR calcium channel activity is controlled by RyR1 ( ,). The identification of protein domains involved in these interactions between both DHPR and RyR1 and of proteins regulating 2 3 these interactions remains today an important field of research.
The DHPR is composed of four different subunits , , and ( ). The subunit has been shown not only to form the pore region α 1 β α 2 δ γ 4 α 1 and to carry the voltage sensitivity of the channel but also to be directly involved in the functional coupling of the DHPR with RyR1 ( ). 5 Indeed, the cytoplasmic loop that links the domains II and III of the subunit has been proposed to be responsible for the mechanical α 1 coupling between the DHPR and the RyR1 ( ). Within this cytoplasmic loop, two regions (domains A and C, respectively) have been 6 shown to regulate ryanodine binding and channel gating of RyR1 ( ). Other regions of the subunit have been found to interact in vitro 7 α 1 with RyR1 but the functional role of these interactions is less essential or remains to be established ( ). 8 Ryanodine receptors are formed by the homotetramer assembly of a 560 kDa polypeptide. RyR1 activity is regulated by a in vitro number of chemical compounds such as ATP, Ca , Mg , and proteins such as calmodulin and FK506 binding protein (FKBP12) ( ).
Nevertheless, due to i) the extremely large size of RyR1, ii) the small number of high affinity effectors and iii) the absence of atomic resolution structure, the mapping of the functional sites of RyR1 is still far from completion. Expression in dyspedic cells (lack RyR1) of chimeric RyR1 channels, that carry sequences of different RyR isoforms, allowed the identification of domains engaged in the EC coupling process ( ). At the same time, binding sites for different effectors and partner proteins have been identified. As part of the 10-12 search for molecules able to strongly modify RyR1 properties, toxins isolated from scorpion venoms have been described as being able to induce important modifications in RyR1 channel behavior. Nowadays, these toxins represent the most effective effectors of RyR1. Among them, imperatoxin A (IpTx A) and maurocalcine (MCa) have been extensively characterized both in terms of their effects on RyR1 and their three-dimensional structure ( ). Interestingly, these two peptide toxins present some amino acid sequence homology with the 13-18 domain A of DHPR subunit and structural studies strongly suggest that the beta-sheet structure of these toxins could mimic that of the α 1 domain A of the DHPR subunit ( , ). In a previous work, we demonstrated that nanomolar concentrations of MCa induces a dramatic α 1 16 17 conformational change of RyR1, witnessed by the increase in H -ryanodine binding (7-fold) and the induction of long-lasting channel [ 3 ] openings ( , ). The observed homologies in amino acid sequences and in 3-D solution structure have prompted different authors to 13 14 make the hypothesis that IpTx A and MCa could share a common binding site with the domain A of the DHPR subunit on the RyR1.

Purification of RyR1
The skeletal heavy SR vesicles were prepared from rabbit leg and back muscles by differential centrifugation, as previously described ( ). Vesicles were then solubilized in the presence of CHAPS, and RyR1 was purified by centrifugation of the solubilized proteins on a 19 sucrose density gradient ( ) and stored in liquid nitrogen. 20

Peptide syntheses
Peptides used in this work were synthesized as previously described ( , ), with the addition of an exogenous biotin on the 14 21 C-terminal amino acid of i) MCa synthetic peptide, ii) peptide A , corresponding to residues Thr -Lys of the DHPR 1s subunit, and

Pull-down experiments
Polystyrene magnetic beads (500 g) coated with streptavidine (Dynal) were incubated for 30 min at room temperature in the presence μ of biotinylated peptide (100 g/ml in buffer A; 150 mM NaCl, 2 mM EGTA, 2 mM CaCl , pCa5 and 20 mM HEPES, pH 7.4). Beads were μ 2 then washed three times with buffer A and incubated for 2 h at room temperature with purified RyR1 (100 nM in buffer A). After washing twice with buffer A, proteins bound to the immobilized peptide were eluted by boiling the beads for 5 min in 2.5 SDS and analyzed by % Western blot using antibodies directed against RyR1 after separation on 8 SDS-PAGE.

Cloning and expression of RyR1 fragments
RyR1 fragments F1-F8 were cloned in pSG5 vector (Stratagene) by a PCR-based method using specific primers with a linker for further enzymatic digestion and insertion into the plasmid backbone (

Expression of His-tagged RyR1 fragments in bacteria
RyR1-F7 fragment was sub-cloned into three shorter segments (F7.1, F7.2 and F7.3) by a PCR-based approach. Briefly, the cDNA was amplified using primers containing a linker encompassing a I site (in the 5 end of the forward primer) and a I (in the 3 end of Nco

Transfection of COS-7 cells with RyR1 constructs and cytoimmunofluorescence
Wild-type RyR1 (RyR1wt) construct was cloned into a modified pCI-neo (Promega

Microsomes preparation
COS-7 cells were harvested in ice cold buffer containing 320 mM sucrose, 5 mM Hepes pH 7.4 and protease inhibitors, using a cell-scraper. Cells were then homogenized using a Teflon potter and centrifuged for 5 min at 7,000 at 4 C. Microsomes were then g°c ollected by centrifugation at 100,000 for 1 h at 4 C and resuspended in the same buffer. After quantification of protein concentration by g°B radford assay, samples were stored in liquid nitrogen. -ryanodine (10 nM) and caffeine (0 or 6 mM) or MCa (0 or 30 nM). Free and bound H -ryanodine were separated by filtration of the microsomes on Whatmann GF/B filters and H -ryanodine measured by liquid scintillation. Non-specific H -ryanodine binding was  detected when beads were covered with peptide-C or biotin. These results confirm the specificity of the interaction between the skeletal Sk sequence of domain A and RyR1 and represent the first biochemical evidence of a direct interaction of MCa with RyR1. 4 9 In order to identify the amino acid sequences of RyR1 involved in the interaction with MCa and peptide-A , we generated a contig of b sk nine clones encompassing the full length RyR1 cDNA in plasmids ( ). Each of these clones was expressed separately in an Figure 1C in cell-free translation system in the presence of radiolabeled S -methionine. RyR1 fragments (F1 9) were incubated in the presence vitro     ). Figure 3B In order to establish the functional relevance of the F7 region in RyR1 regulation by MCa or peptide-A , we characterized the binding sk of H -ryanodine on RyR1wt or RyR1 F7 expressed in COS-7 cells and tested the effects of both caffeine and MCa on this binding. For [ 3 ] Δ this purpose, we prepared microsomes from cells transfected with either RyR1wt or RyR1 F7. Immunoblot analysis of the different Δ microsome preparations using antibodies directed against RyR1 ( upper panel) supports the localization of expressed RyR1wt Figure 3B and RyR1 F7 in reticulum membrane of transfected cells. These results also show that RyR1wt and RyR1 F7 are expressed at a similar Δ Δ level and remain stable in transfected cells. Finally, endogenous RyR1, if present, is barely measurable. Using these microsome preparations, we then measured H -ryanodine binding on expressed RyR1wt and RyR1 F7 in the presence or absence of caffeine or Figure 3B addition, no effect of caffeine or MCa was observed on these control microsomes (data not shown binding by caffeine and MCa is presented in . In the presence of 6 mM caffeine, we observed a 197 5 and 218 32 Figure 3B ± % ± % increase of H -ryanodine binding on RyR1wt and RyR1 F7 microsomes, respectively. In contrast, 30 nM MCa induced a 202 28

MCa. No significant specific H -ryanodine binding was observed on microsomes obtained from non-transfected cells ( ). In
increase of H -ryanodine binding on RyR1wt microsomes but no significant change of H -ryanodine binding on RyR1 F7 microsomes

DISCUSSION
Recently, we characterized the effect of MCa on RyR1 ( , ). MCa is a 33 amino acid toxin initially isolated from the venom of the scorpion , and subsequently produced by chemical peptide synthesis ( ). Based on sequence similarities, MCa Scorpio maurus palmatus 24 and IpTxA (another scorpion toxin interacting with RyR1) were proposed to share with the domain A of the skeletal DHPR subunit a α 1 common binding site(s) on RyR1. In this work, we identified two sequences of RyR1, fragment F3 encompassing amino acids 1021-1631 and fragment F7 comprising the residues 3201-3661, that are able to interact with both MCa and domain A of the II-III loop of the in vitro DHPR subunit. Among these two sequences, fragment F7 displays the strongest interaction with MCa and peptide-A . We then α 1 sk generated a RyR1 construct that lacks the amino acid region corresponding to fragment F7. This RyR1 construct with its internal deletion displays an expression profile and a caffeine sensitivity that is perfectly similar to the wild-type RyR1. In contrast, deletion of the F7 region completely abolished the stimulatory effect of MCa on ryanodine binding, highlighting the critical role of this domain for MCa regulation. A molecular dissection of fragment F7 allowed us to identify a discrete domain corresponding to amino acids 3351-3501 (fragment F7.2) responsible for the interaction of MCa and peptide-A with RyR1. These results clearly demonstrate that, in agreement sk with earlier expectations, MCa and domain A interact with the same RyR1 region within the 3351-3501 stretch of amino acid residues.
Binding of MCa and peptide-A to the same RyR1 site was confirmed by the complete inhibition of the interaction of peptide-A with the sk sk fragment F7.2 in the presence of an excess of MCa.
Using an affinity chromatography approach, Leong  properties ( , ), the RyR1 sequences that we identified in this report ought to play an important function in RyR1 calcium channel 13 14 behavior. The fact that domain A of the II III loop of DHPR binds onto the same RyR1 sequences re-opens the question of the role of the domain A in the cont rol of RyR1 calcium channel activity in the context of the DHPR/RyR1 complex. Our results suggest that the two sequences of RyR1 (F3 and F7), that are involved in the binding of MCa, could fold together to create a unique MCa binding site.
Cryoelectron microscopy studies of the RyR1-MCa complex should permit to test this hypothesis and to refine the correspondence between primary structure and 3-D reconstruction images of RyR1.

Figure 3
Deletion of the F7 region abolishes the effect of maurocalcine on H -ryanodine binding onto RyR1 expressed in COS-7 cells , Functional effect of the F7 deletion on C RyR1 binding properties. H -ryanodine binding experiments on purified cell microsomes showed that, while both constructs responded to [ 3 ] caffeine administration (6 mM), the deletion of the F7 fragment abolished MCa effect (30 nM).