Báo cáo khoa học: The calpain 1–a-actinin interaction Resting complex between the calcium-dependant protease and its target in cytoskeleton doc

The calpain 1–a-actinin interaction Resting complex between the calcium-dependant protease and its target in cytoskeleton Fabrice Raynaud 1 , Chantal Bonnal 1 , Eric Fernandez 2 , Laure Bremaud 3 , Martine Cerutti 4 , Marie-Christine Lebart 1 , Claude Roustan 1 , Ahmed Ouali 2 and Yves Benyamin 1 1 UMR 5539 – CNRS, laboratoire de Motilite ´ Cellulaire – EPHE, cc107, USTL, Montpellier, France; 2 Station de Recherches sur la Viande, INRA-Theix, St-Gene ` s-Champanelle, France; 3 Institut ‘Sciences de la Vie et de la Sante ´ ’, Genetique Moleculaire Animale, UR 1061, INRA-Universite ´ de Limoges, Faculte ´ des Sciences, Limoges, France; 4 Laboratoire de Pathologie Compare ´ e, INRA-CNRS URA5087, Saint Christol Le ` s Ale ` s, France Calpain 1 behaviour toward cytoskeletal targets was inves- tigated using two a-actinin isoforms from smooth and skel- etal muscles. These two isoforms which are, respectively, sensitiveand resistant to calpain cleavage, interactwith the protease when using in vitro binding assays. The stability of the complexes in EGTA [K d(–Ca2+) ¼ 0.5 ± 0.1 l M ] was improved in the presence of 1 m M calcium ions [K d(+Ca2+) ¼ 0.05 ± 0.01 l M ]. Location of the binding structures shows that the C-terminal domain of a-actinin and each calpain subunit, 28 and 80 kDa, participates in the interaction. In particular, the autolysed calpain form (76/18) affords a similar binding compared to the 80/28 intact enzyme, with an identified binding site in the cata- lytic subunit, located in the C-terminal region of the chain (domain III–IV). The in vivo colocalization of calpain 1 and a-actinin was shown to be likely in the presence of calcium, when permeabilized muscle fibres were supple- mented by exogenous calpain 1 and the presence of cal- pain 1 in Z-line cores was shown by gold-labelled antibodies. The demonstration of such a colocalization was brought by coimmunoprecipitation experiments of calpain 1 and a-actinin from C2.7 myogenic cells. We propose that calpain 1 interacts in a resting state with cytoskeletal targets, and that this binding is strengthened in pathological conditions, such as ischaemia and dystro- phies, associated with high calcium concentrations. Keywords: calpain; cytoskeleton; alpha-actinin; muscle; calcium. Calpain 1 (Calp1) and calpain 2 (Calp2) are intracellular Ca 2+ -dependent thiol endoproteases [1], expressed through- out the animal kingdom, and recently reported in the plant kingdom [2]. These two proteases are particularly implicated [1] in the selective proteolysis of factors involved in the cell cycle, in myocyte fusion, during apoptosis in association with caspases or in the cleavage of membrane-cytoskeleton complexes during cell motility phases [3]. Many of the substrates are transcription and signalling factors with intracellular presence of less than 2 h [4] or cytoskeletal proteins with long half-lives, generally specialized in the cross-link or the membrane anchorage of fibrillar components [5]. The hypothesis according to which calpains would be released from complexes with calpastatin (its natural inhibitor) to join membrane phospholipids where protease activation is achieved was proposed [6], but the origin of recognition of specific substrates by calpains [7,8] remains unclear. A statistical analysis of the presence of PEST sequences in the target, critical for calpain recognition [9,10], gives valuable scores with short half-life proteins, but is not appropriate in the case of several cytoskeletal actin-binding proteins [1]. For example, filamin, dystrophin and talin are known to be cleaved in vivo by calpains. It should be noted that the accessibility of the calmodulin (CaM)-binding domain in PEST sequences is an important factor to consider [11], as demonstrated for IjBa,aCaMand calpain-binding protein [12]. Moreover, we have shown recently in muscle fibres [13] the presence of a stable Ca 2+ -regulated complex between E64-treated Calp1 and the N-terminal region of the titin located between the Z-band and the N2-line in the I-band of myofibrils. This titin region, rich in PEST sequences, was reported to show a marked calcium binding ability related to acidic sequences [14]. In the absence of Ca 2+ ions, a weak interaction between the Ca 2+ -binding titin fragments and Calp1 was observed. On the other hand, several other calpain substrates deprived of PEST sequences but containing calmodulin- binding domains [15] or EF-hand sequences [16] were Correspondence to Y. Benyamin, UMR 5539, laboratoire de Motilite ´ Cellulaire – EPHE, Bt. 24, cc107, USTL, place E. Bataillon F-34095 Montpellier cedex 5, France. Fax: + 33 4 67144927, Tel.: + 33 4 67143813, E-mail: benyamin@univ-montp2.fr Abbreviations: ask, skeletal muscle a-actinin; asm, smooth muscle a-actinin;Calp1,calpain1(l-calpain); Calp2, calpain 2 (m-calpain); CaM, calmodulin; ELISA, enzyme-linked immunosorbant assay; FITC, fluorescein 5-isothiocyanate; Seph-ask, Sepharose-insolubilized skeletal muscle a-actinin. Note: web pages are available at http://www.dbs.univ-montp2.fr/ umr5539/, http://www.ephe.univ-montp2.fr/ (Received 29 July 2003, accepted 30 September 2003) Eur. J. Biochem. 270, 4662–4670 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03859.x reported. It was thus suggested that the calpain domains IV and VI, which have a CaM-like structure and are members of the penta-EF-hand family of proteins [1], could be the binding structures of interfaces with cytoskeleton compo- nents [17] and contribute to link the two subunits in calpains [18]. Indeed, the two subunits of 80 kDa (domains I–IV) and 30 kDa (domains V–VI), contain altogether 10–11 putative EF-hands motifs in domains IV, VI and II [1,19,20], from which five to six are estimated to be functional. A negatively charged loop of domain III also offers Ca 2+ binding capacity [21,22], which maximizes binding to eight equivalents of Ca 2+ in agreement with a previous in vitro evaluation [23]. Domain III, which includes binding sites for calpastatin and phospholipids [1,18], appears as the regulation centre between the CaM-like domains IV and VI and the catalytic domain II, which is in interactionwithdomainIIIthroughCa 2+ -regulated salt bridges [20]. Thus, according to a previous hypothesis, the interaction between a PEST sequence, a CaM-binding domain, or a Ca 2+ -binding motif and a CaM-like domain, would place the catalytic site of calpain in close proximity to the substrate. To test this hypothesis, we have investigated Calp1 interaction with two a-actinin isoforms, either resistant or sensitive to calpain 1 proteolysis, purified from chicken skeletal and smooth muscles, respectively. The a-actinin family displays two EF-hand motifs in the C-terminal domain [24], presents low PEST scores after analysis, and several isoforms are calpain substrates [16,25–27]. Our study of calpain 1–a-actinin interaction suggests the importance of calmodulin-like domains and EF-hand motifs. Finally, it allows dissociation of two aspects in the protease behaviour toward its target, binding and cleavage, in relation to the presence of Ca 2+ ions. Materials and methods Proteins Bovine Calp1 (80/30) was isolated [28] from the bovine skeletal Rectus abdominus muscle, excised within 1 h post- stunning (INRA slaughterhouse, Clermont-Fd, France). Smooth (asm) and skeletal (ask) muscle a-actinins were purified from chicken gizzard and breast muscles, respect- ively, obtained immediately after killing (Avigar slaughter- house, Gard, France). Purification procedures were described previously [29,30]. Human (887 b) cDNA (calpain 28 kDa regulatory subunit) was expressed as C-terminal His-tagged protein in the pET16b vector (Roche Diagnostics). The construct was transformed into competent BL21(DE3) Escherichia coli. Expression was induced by adding 1 m M isopropyl thio-b- D -galactoside for 3 h at 37 °C. Slurry Ni/NTA were added to supernatant after bacteria lysis and gently mixed for 30 min. Solid phase was packed in a column before washing twice with the lysis buffer, adjusted to 20 m M imidazole, pH 8, and the elution performed with the same buffer adjusted to 250 m M imidazole. Eluted fractions were analysed by SDS/PAGE and Coomassie blue staining. Human 80 kDa catalytic subunit (microcalpain) was expressed [31] in Spodoptra frugiperda (SF9) cells using a recombinant 80 kDa subunit baculovirus. Sf9 cell pellets were lysed in 100 m M NaCl, 2 m M EGTA, 0.1% (v/v) Triton X-100, 20 m M Tris/HCl, pH 7.5 buffer, supplemen- ted with antiproteases cocktail (Roche) and cleared at 23 000 g for 15 min. Supernatant was incubated (4 °C, 60 min) with 800 lg of anti-Calp1 Ig (a purified low affinity antibody subfraction [32]), then supplemented with Seph- arose-protein A (Pharmacia, Uppsala, Sweden) and gently mixed for 60 min. Solid phase was washed four times with the lysis buffer, before a batch elution with 0.6 M KI, 2 m M dithiothreitol, 20 m M Tris/HCl, pH 7.5 and dialysis against the interaction buffer. Proteolysis and protein modifications Calp1 autolysis was conducted [1,20,33] during 10 min at 20 °Cin1m M CaCl 2 ,20m M Tris/HCl buffer, pH 7.5, to obtain the autolysed form (76/18) or during 120 min in the same buffer at 20 °C to conduct a more complete degra- dation. Autolysis kinetics were followed by SDS/PAGE and stopped (2 m M EGTA) after an optimal incubation time. Skeletal a-actinin cleavage was performed [30] with thermolysin (1 : 25 enzyme/substrate, w/w). The 30, 55 and 10 kDa domains issued from the cleavage were purified on a PorosHQ/H (Boehringer, Manheim) FPLC column using the procedure previously described with fish ask. ask and asm (1 mgÆmL )1 ) were treated in 1 m M CaCl 2 ,1m M dithiothreitol, 50 m M KCl, 20 m M Tris/HCl buffer, pH 7.5 by Calp1 during 2 h at 20 °C using a protease/substrate ratio of 1 : 10 (w/w) [1,25,30]. Proteolysis was stopped by 2m M EGTA, and the residual Calp1 discarded by the FPLC procedure. Sepharose-insolubilized skeletal muscle a-actinin (Seph-ask) was obtained (1 mg proteinÆmL )1 gel) with BrCN-activated Sepharose 4B (Pharmacia). Protein labelling was performed with biotin succinimide ester [34] or fluorescein isothiocyanate (FITC) [35]. Biotin- amidocaproate N-hydroxysuccinimide ester, E64 calpain inhibitor and fluorescein isothiocyanate were purchased from Sigma Chemical Co. Thermolysin was from Serva (Heidelberg, Germany). Antibody specificities The anti-Calp1 and anti-Calp1,2 Igs are directed against a specific sequence (539–553 in domain IV) and a conserved sequence (330–344 in the subdomain IIb) of the unprocessed 80 kDa subunit (SwissProt, ID number P07384), respect- ively. Sequences were chosen according to their accessibility and helicoidal content criteria before synthesis and coupling to hemocyanin using glutaraldehyde [36]. Rabbit anti- (a-actinin) Igs cross-reacting with ask and asm [30] were fractionated with the 30 kDa, 55 kDa and 10 kDa Seph- arose 4B-insolubilized fragments, issued from ask thermo- lysin cleavage. Anti-rabbit IgG conjugated with alkaline phosphatase was obtained from Biosys (Compiegne, France). Monoclonal (His) 6 antibody was from Qiagen. Binding analysis ELISA was performed [29] in microtitration plates (Poly- sorp, Nalgen Nunc International, Denmark). Incubation steps were carried out at 20 °Cin150m M NaCl, 0.5% gelatine, 3% gelatine hydrolysate, 0.05% Tween 20, 20 m M Ó FEBS 2003 Calpain 1–a-actinin interaction (Eur. J. Biochem. 270) 4663 Tris/HCl buffer, pH 7.4. Each assay monitored at 405 nm was conducted in triplicate and the mean value was plotted after subtraction of nonspecific absorption. In spectrofluorescence experiments, interactions of the fluorescein labelled Calp1 were performed at 20 °Cin 50 m M KCl, 20 m M Tris/HCl buffer, pH 7.4, using a Perkin-Elmer luminescence spectrometer LS 50. The exci- tation was set at 494 nm and the emission spectrum recorded between 510 and 550 nm. Fluorescence changes were deduced from the area of emission spectra [35]. ELISA and fluorescence binding assays were conducted in the presence of 1 m M EGTA or 1 m M CaCl 2 supple- mented with 10 l M E64 to saturate all Ca 2+ binding sites, including the non EF-hand ones [21,22,37] and avoid proteolytic processes. The parameters K d (apparent disso- ciation constant) and A max (maximum effect) were obtained ( CURVEFIT software developed by K. Raner Software, Mount Waverley, Victoria, Australia) by nonlinear least squares fitting of the experimental data points to the following equation A ¼ A max ýL=ðK d þ½LÞ where A is the measured effect and [L] is the ligand concentration. Cell culture C2.7 myoblasts derived from the C2 mouse myogenic cell line [38] were cultured in DMEM (Gibco-BRL/Life Tech- nology) supplemented with 2 m M glutamax (Gibco-BRL), 100 lgÆmL )1 penicillin, 100 lgÆmL )1 streptomycin (Gibco- BRL), and 20% fetal bovine serum (Gibco-BRL). Cells in proliferation (confluence stage) were lysed in 0.1% Triton X-100, 150 m M NaCl, 2 m M EGTA, antiprotease cocktail (Roche), 20 m M Tris/HCl buffer (lysis buffer), then centri- fuged. Protein cosedimentation and immunoprecipitation Cosedimentations were performed at 20 °Cusing2.5lgof Calp1 (80 kDa/28 kDa) mixed with 2.5 lgofthe10min autolysed form (76 kDa/18 kDa), 10 lgofthe120min autolysed form or 10 lg of the calpain 28 kDa subunit, incubated with 50 lg of insolubilized Seph-ask in 200 lLof 50 m M Tris/HCl pH 7.4 buffer supplemented with 150 m M NaCl, 1 m M EGTA, 0.1% NP-40, and 0.25% gelatine. After 60-min incubation, the solid phase was washed four times with 50 m M Tris/HCl, pH 7.4, 1 m M EGTA, 0.1% NP-40 and resuspended in 60 lL of Laemmli loading buffer. Thirty microlitres of the suspension were analysed by SDS/PAGE and Western blotting. The C2.7 line cell lysate supernatant obtained from 10 7 cells was incubated for 1 h at room temperature with 50 lg of anti-Calp1 Ig, then with insolubilized Sepharose-protein A (Sigma) in lysis buffer supplemented by 1% bovine serum albumin. After extensive washing in lysis buffer, the solid phase was treated at 100 °C in Laemmli buffer. A control assay in the same conditions, but without anti-Calp1 Ig, was performed. Electrophoresis (SDS/PAGE) was made [39] using 7.5% resolving gels and stained with Coomassie blue or silver. Molecular mass standards were from Bio-Rad and Phar- macia. Western blotting [40] was performed using the appropriate antibody. Calp1 enrichment of permeabilized muscle fibres Glycerinated fibres were obtained as previously reported [25]. Briefly, small fibre bundles (1 · 5 mm) taken from freshly excised bovine longissiumus muscle, were stretched between two pins and immersed in 30 m M Tris/HCl, pH 7.5, containing 50% glycerol, 5 m M EDTA and anti- proteases cocktail during 5 h, diced into small pieces (0.8 · 0.2 mm), and maintained in the same solution for 18 h, before extensive washing in 200 m M KCl, antiprotease cocktail (Roche), 30 m M Tris/HCl pH 7.5, to flush out endogenous calpains and calpastatin complexes (Western blotting controls). Samples were then incubated under continuous mild stirring, with Calp1 (0.5 mgÆmL )1 )in 30 m M Tris/HCl pH 7.5 containing 5 m M dithiothreitol, 2m M EGTA or in 30 m M Tris/HCl, pH 7.5, containing 5m M dithiothreitol, 1 m M CaCl 2 ,and10l M E64 for 1 h at room temperature. Exogenous Calp1 added to permeabi- lized fibres was located using the postembedding procedure [41,42] performed with a gold (10 nm)-labelled secondary antibody (Sigma) diluted to 1 : 50. Results Interaction of Calp1 with skeletal and smooth muscle a-actinins Specificity of the antibodies directed against the specific sequence 539–553 in subdomain IV (anti-Calp1) and the conserved (Calp1 and Calp2) sequence 330–344 in subdo- main II (anti-Calp1,2) was assessed (Fig. 1A) by Western blotting using bovine skeletal muscle crude extract and the purified Calp1. Calp1 in muscle fibre extract appears as a unique band at the 80 kDa level (Fig. 1Ab,e). In particular, no cross-reactivity of anti-Calp1 was detected toward the purified Calp2 (not shown) or Calp3 (p94) in extract (Fig. 1A,b). As expected, anti-Calp1,2 was able to detect both Calp1 (Fig. 1A,f) and Calp2 (not shown). Chicken a-actinins extracted from skeletal muscle (ask) and gizzard (asm) were assayed as substrates of Calp1. As shown in Fig. 1B, upon Calp1 treatment in the presence of 1m M CaCl 2 , asm is deprived of a segment of about 5 kDa in contrast to the ask isoform which resists to proteolysis. The asm 95 kDa truncated protein did not react with the antibody directed against the C-terminal 10 kDa fragment of a-actinin [25,30], indicating that the deleted segment is located at the C-terminal extremity (not shown). Similar calpain proteolysis was previously reported for fish a-actinin [25,30,43] and for nonmuscle isoforms [16] in contrast to porcine, bovine or rabbit skeletal muscles isoforms [44,45] which resist. Using two independent methods, interaction of a-actinins with Calp1 was investigated. In solid phase assay (ELISA), we observed that ask binds to coated Calp1 in the absence of Ca 2+ ions with a significant affinity (Fig. 2A). Apparent dissociation constant [K d(–Ca2+) ], calculated from five experiments performed in the presence of 1 m M EGTA, corresponds to 0.5 ± 0.1 l M . A similar affinity (0.3 ± 0.1 l M ) was observed when ask was immobilized instead of 4664 F. Raynaud et al. (Eur. J. Biochem. 270) Ó FEBS 2003 Calp1 (Fig. 2A, inset). An equivalent result was obtained with asm [K d(– Ca2+) ] ¼ 1.0 ± 0.4 l M ), except that the 95 kDa fragment generated after Calp1 cleavage did not bind calpain (Fig. 2B). These experiments were confirmed using fluorescent assays in which increasing amounts of ask were added to FITC-labelled Calp1 (Fig. 2C). When the interaction was conducted in the presence of 1 m M calcium (and E64 as calpain inhibitor), a tenfold increase in affinity [K d(+Ca2+) ¼ 0.05 ± 0.01 l M ] was observed (Fig. 2C, Fig. 1. Protein patterns. (A) Specificity of the anticalpain 1 antibodies. Bovine skeletal muscle extract (a,b), purified bovine Calp1 (c,e), and 10-min autolysed Calp1 (d) were stained by Coomassie blue (a), by silver (c,d) or tested by Western blotting (b,e,f) using anti-Calp1 (b,e) and anti-Calp1,2 (f) Igs. (B) Proteolysis of smooth muscle a-actinin (a) cleaved by Calp1 and revealed by anti-(a-actinin) after 30 min (b) and 120 min (c). The arrowhead points to the 95 kDa proteolysis product, and the arrow indicates the position of the rabbit muscle phosphory- lase B (97 kDa). (C), Western blotting of skeletal a-actinin (a) cleaved by thermolysin (b) and the FPLC purified C-terminal 10 kDa frag- ment (c) using anti-(a-actinin). Fig. 2. a-Actinin–Calp1 interactions. (A) Solid phase immunoassay (ELISA) between coated Calp1 (j) or coated 10 min-autolysed Calp1 (h) and increasing ask concentrations or between coated ask and increasing Calp1 concentrations (inset). Binding was monitored at 405 nm using biotin-labelled proteins (ask or Calp1) and streptavidin– alkaline phosphatase-labelled (1 : 2000 diluted). (B) Solid phase immunoassay (ELISA) between coated Calp1 and increasing amounts of intact asm (s) or the 95 kDa cleaved asm (d). The experimental conditions were those described in (A). (C) Decrease in the fluores- cence (DF) of FITC–Calp1 (1 lgÆmL )1 ) in interaction with increasing concentrations of ask in the presence of 1 m M EGTA or 1 m M CaCl 2 (inset). Ó FEBS 2003 Calpain 1–a-actinin interaction (Eur. J. Biochem. 270) 4665 inset) in comparison to the binding [K d(–Ca2+) ¼ 0.5 ± 0.1 l M ] obtained in the absence of calcium. A similar calcium effect [K d(+Ca2+) ¼ 0.2 ± 0.04 l M ; K d(–Ca2+) ¼ 1±0.2l M ] was observed with asm (not shown). In conclusion, ask and ask isoforms are able to bind Calp1 in the absence and in the presence of calcium, with an important increase in the affinity in the presence of calcium. Furthermore, the cleavage of asm by calpain 1 generates a fragment of 95 kDa which is unable to interact with Calp1. This indicates the presence of a calpain binding site in the C-terminal domain of smooth muscle a-actinin. Involvement of the C-terminal domain of ask in Calp1 binding In order to locate the region responsible for the binding of Calp1 on ask isoform, which is not sensitive to calpain cleavage, we tested the three fragments issued from the proteolysis of ask by thermolysin (Fig. 1Ca,b): the 30 (N-terminal actin binding domain), the 60 (spectrin-like repeats, central domain), and the 10 kDa (C-terminal EF-hand domain). In solid phase assay, we observed (Fig. 3A) that the purified 10 kDa fragment (Fig. 1Cc) was the only one to bind Calp1 with a detectable affinity. We confirmed this result using fluorescent assays (Fig. 3B) and found a higher affinity in the presence of calcium [K d(+Ca2+) ¼ 1.0 ±0.1 l M ] in comparison with its absence [K d(–Ca2+) ¼ 2.5 ±0.3 l M ]. Thus, a Calp1 binding site is included in the C-terminal domain of the ask molecule, and this binding is independent of the susceptibility of the a-actinin isoform to calpain proteolysis. Identification of the calpain 1 subunit implicated in a-actinin interaction The regulatory (28 kDa) and the catalytic (80 kDa) sub- units, expressed as recombinant proteins, were assayed for binding activity toward skeletal muscle a-actinin. We observed (Fig. 4) that in the presence of 1 m M Ca 2+ the two subunits interact with ask, the catalytic 80 kDa subunit having a better affinity [K d(+Ca2+) ¼ 0.5 ± 0.1 l M ]than the regulatory 28 kDa chain [K d(+Ca2+) ¼ 1.4 ± 0.2 l M ]. In the absence of calcium (1 m M EGTA), the 28 kDa subunit interaction is very weak [K d(–Ca2+) >10l M ]in contrast to that of the 80 kDa chain. Binding between the 28 kDa subunit and ask in the presence of calcium, was confirmed [K d(+Ca2+) ¼ 2.5 ± 0.5 l M ) using fluorescence assays (not shown) and cosedimentation experiments using Seph-ask (Fig. 5A). Similarly, we have shown that the intact (80 kDa/ 28 kDa) calpain 1 (Fig. 1A) and the 10-min autolysed (76 kDa/18 kDa) form (Fig. 1Ad) have the same binding ability toward ask in the absence of calcium (Figs 2A and 5B, 10 min) as in its presence (not shown). Furthermore, the 76, 50 and 30 kDa fragments issued from the 120 min autolysis of Calp1, and recognized by anti-Calp1 (Fig. 5Ba, 120 min), cosedimented with ask (Fig. 5Bb, 120 min). It can be observed (Fig. 5Bb,c, 120 min) that only the 76 kDa fragment is recognized by anti-Calp1 (domain IV) and anti- Calp1,2 (domain II), which locates the 50 kDa and the 30 kDa fragments in the C-terminal region (domains III–IV) of the catalytic subunit. Thus, the 28 kDa subunit (probably its C-terminal 18 kDa region) in a calcium-dependent fashion, and the C-terminal part (domains III–IV) of the 80 kDa subunit, are implicated in the interface linking calpain 1 to skeletal muscle a-actinin. Colocalization of microcalpain and a-actinin in myogenic cells Calpains and a-actinin were previously located in Z-disks [44], and adhesion structures [5], without evidence of strong molecular proximity. To confirm that the a-actinin–Calp1 interaction was physiologically relevant, we first performed coimmunoprecipitation studies. As shown in Fig. 5C, a-actinin was coprecipitated with calpain 1 from a cell Fig. 3. Interactions of ask domains with Calp1. (A) Solid phase immunoassay (ELISA) between coated 30 (u), 60 (j)and10 kDa(s) ask fragments and increasing amounts of biotin-labelled Calp1. Binding was determined at 405 nm using streptavidin–alkaline phosphatase labelling (1 : 2000). (B) Fluorescence decrease (DF) of FITC-labelled Calp1 (1 lgÆmL )1 ) in interaction with increasing concentrations of the 10 kDa fragment in the presence of 1 m M CaCl 2 (s)or1m M EGTA (d). Glutathione S-transferase (Sigma) was used as a negative control (j)oftheinteraction. 4666 F. Raynaud et al. (Eur. J. Biochem. 270) Ó FEBS 2003 lysate issued from C2.7 myoblasts, by using rabbit anti-Calp1 Ig and insolubilized protein A. Furthermore, after incubation of permeabilized muscle fibres with exogenous Calp1 in the presence of 1 m M Ca 2+ ions, we observed that calpain 1 was strikingly concentrated in the Z-disk core (Fig. 6) in comparison with the A- or M-bands. When calcium was omitted, we were unable to detect a preferential location of exogenous Calp1 in myofibrils. Thus, Ca 2+ ions seem to favour the targeting of Calp1 to Z-line and the interaction of the protease with the compo- nents of this anchorage structure. However, in the case of C2.7 cells, coimmunoprecipitation of a-actinin by the anti- Calp1 Ig was also observed in the absence of Ca 2+ ions, which is likely considering the in vitro binding analysis. Discussion We have investigated the hypothesis, according to which calpain 1, as calpain 3 (p94) with titin and glial filaments [46,47], could bind directly to targets in cytoskeleton through specific and stable interactions. This hypothesis involves questions related to the origin of interactions [9,10,12,27] with the targets, the stability of complexes in the resting stage [3,48] and the activation state [7,33] of the binding calpain. The topology of the interface with respect to the catalytic domain II [1] and the cleavage site in target are also underlying. Two muscle a-actinin isoforms (ask and asm) with different calpain cleavage susceptibilities were chosen as targets and their binding with calpain 1 analysed by independent in vitro and in vivo approaches. According to the presented results, Calp1–a-actinin interaction is inde- pendent of the cleavage susceptibility of the target and occurred in the absence of calcium, but is improved in its presence. In the absence of calcium, the apparent dissociation constant of Calp1–ask (or asm) complexes is measured in the micromolar order and decreased to the submicromolar level inthepresenceof1 m M Ca 2+ and E64. The autolysed Calp1 (76/18) form and the intact enzyme (80/28) afforded the same binding ability toward a-actinin. In the experiments performed in the presence of 1 m M calcium, Calp1 was used at 1 lgÆmL (coated or FITC- protein) supplemented with 10 l M E64 to avoid its auto- proteolysis and its aggregation [37]. In fact, the affinity increase in the presence of Ca 2+ ions, was observed from 50 l M CaCl 2 , and the effect increased with increasing calcium concentration. Thus, the conformation changes Fig. 4. Calp1 subunits binding assays. Solid phase immunoassay (ELISA) between the 30 (s,d)andthe80kDa(h,j)-coated Calp1 subunits and biotin-labelled ask in the presence of 1 m M CaCl 2 (open symbols) or 1 m M EGTA (filled symbols). Binding was determined at 405 nm using streptavidin–alkaline phosphatase labelling (1 : 2000). Fig. 5. Cosedimentation of calpain–a-actinin complexes. (A) Cosedi- mentation with Seph-ask of His-tagged 30 kDa subunit (a), in the presence of 1 m M CaCl 2 (b) or 1 m M EGTA (c). Suspensions were centrifuged at 2000 g and the pellet revealed after SDS/PAGE by Western blotting using anti-His 6 Ig (1 : 1000 diluted). (B), cosedi- mentation of the 10 min autolysed Calp1 supplemented by the intact Calp1 (left part, lane a) or the 120 min autolysed Calp1 (right part, lanea)incubatedin1m M EGTA with Seph-ask (see Materials and methods). Pellets (lanes b) are revealed after SDS/PAGE with Coo- massie blue (left part) or by Western blotting using anticalpain anti- bodies (right part, lanes b and c). A negative control (c) using inert Sepharose was included (left part). Anti-Calp1 (lane a,b) and anti- Calp1,2 (lane c) were used (right part). (C) Coimmunoprecipitation of Calp1–ask complexes from C2.7 lysate by anti-Calp1 and Sepharose- protein A. The presence of ask in the pellet was searched in the assay (a) and in the control performed without the anti-Calp1 (b), after SDS/ PAGE and Western blotting, using anti-(a-actinin). Ó FEBS 2003 Calpain 1–a-actinin interaction (Eur. J. Biochem. 270) 4667 induced by the saturation of low affinity Ca 2+ -binding sites, changes mainly localized at level of the 28 kDa subunit [37], and the exposure of hydrophobic patches on the surface of the protease which aggregates calpain 1 molecules [37], couldleadtoastickyinteractionwitha-actinin. Neverthe- less, the fact that efficient calcium concentrations (from 50 l M ) are higher than the physiological ones, may be considered in view of several pathological situations such as brain and muscle ischemia [49,50], or several muscle dystrophies [1], but also during necrosis [51], apoptosis [52], or myoblast fusion [53] where Ca 2+ ions concentration increases in unknown proportions. In these cases, the accumulation of calpain 1 on cytoskeleton could explain the rapid intervention of the protease and the quick Ca 2+ -dependent degradation of several cytoskeletal proteins [25,51]. Binding interface between Calp 1 and ask was further located in this study. Calpain binding structures were found within the 10 kDa C-terminal domain of the a-actinin molecule. The inability of the 95 kDa chain (issued from the cleavage of asm by Calp1) to interact with calpain 1, could restrict the location of the binding elements to the last 5 kDa, although we cannot rule out the possibility of conformational changes induced to the 95 kDa by the cleavage. We can thus conclude that calpain 1 displays two distinct behaviours, one consisting of interaction with its target and the other being responsible for the proper cleavage action of the target. The attempt to locate the binding structures on Calp1 implied disposal of the two 28 and 80 kDa isolated subunits in the correct conformation, which was effective by using E. coli [54] and SF9-Bacculo virus[55]asexpression systems, respectively. We have concluded that both subunits display binding abilities, although the regulatory subunit (28 kDa) is strongly controlled by calcium which binds to the CaM-like domain VI. Concerning the catalytic subunit (80 kDa), the restriction was brought by cosedimentation assays to the 50 kDa C-terminal part, bearing domains III and IV. Thus, the ability of the two isolated subunits or the autoproteolysed Calp1 products (18/76 and 55 kDa) to interact with ask indicates that the domains III–IV and VI participate to the interface with the C-terminal region of a-actinin. These domains concentrate 10 EF-hand motifs andanacidicCa 2+ binding sequence [1,20,21]. It is noteworthy that the location of a Calp1 binding site in the C-terminal region of a-actinin [56] situates the protease in the vicinity of titin [57] and CapZ [34], two proteins described as a-actinin partners in the Z-line and known as calpain substrates [1,25,34]. In this context, according to our experimentation of enrichment of permeabilized fibres by exogenous Calp1 on Z-line, one could hypothesize that the two myofibrillar proteins could also bind Calp1, as a-actinin does. Note that these three proteins strongly participate in the Z-disk organization, a compartment rapidly proteolysed during muscle ischemia [25,45] or after a calpain treatment of isolated myofibrils [25]. Targeting of Z-disk by calpains was previously suggested [25,44,48] and a quick ask release from muscles treated by calpains in the presence of calcium was observed. Furthermore, a-actinin is also located in cellular adhesion structures [5], in a colocation with integrin, talin and vinculin [56,58]. Immunoprecipitation of a-actinin from C2.7 lysate by anti-Calp1 proves the association of the protease with a-actinin, either in direct contact or in a complex including the two proteins. In conclusion, our study proves the interaction between a-actinin and calpain 1 and locates binding motifs within regions where the EF-hand domains of the protease and the cytoskeletal protein are concentrated. The behaviour of calpain 1 toward cytoskeletal targets appears dual. In its first state, the protease would oscillate between cytoskeleton components, calpastatin and phospholipids in membrane according to the local calcium concentrations. This would eventually lead to the cleavage of close substrates in cytoskeleton. This equilibrium is currently under investiga- tion by using C2.7 cell line transfection assays with calpain 1 CaM-like subdomains. Acknowledgements This work was supported by grants from the Association Franc¸ aise contre les Myopathies (AFM) and PPF network (EPHE). Authors are grateful to Professor H. Sorimachi for the calpain 1 constructs gift. Fig. 6. Muscle fibres treatment by inactivated Calp1. Permeabilized muscle fibres were enriched with exogenous Calp1 in the presence of either 1 m M Ca 2+ ions (A) or 2 m M EGTA (B) during 1 h, stained with anti-Calp1, then with a gold-labelled secondary anti-rabbit IgG. Z, Z-band; M, M-band; A, A-band. 4668 F. Raynaud et al. (Eur. J. Biochem. 270) Ó FEBS 2003 References 1. Goll,D.E.,Thompson,V.F.,Li,H.,Wei,W.&Cong,J.(2003) The calpain system. Physiol. Rev. 83, 731–801. 2. Margis, R. & Margis-Pinheiro, M. (2003) Phytocalpains: ortho- logous calcium-dependent cysteine proteinases. Trends Plant Sci. 8, 58–62. 3. Dourdin, N., Bhatt, A.K., Dutt, P., Greer, P.A., Arthur, J.S., Elce, J.S. & Huttenlocher, A. (2001) Reduced cell migration and dis- ruption of the actin cytoskeleton in calpain deficient embryonic fibroblasts. J. Biol. Chem. 276, 48382–48388. 4. Hirai, S., Kawasaki, H., Yaniv, M. & Suzuki, K. (1991) Degradation of transcription factors, c-Jun and c-Fos, by calpain. FEBS Lett. 287, 57–61. 5. Bhatt, A., Kaverina, I., Otey, C. & Huttenlocher, A. (2002) Reg- ulation of focal complex composition and disassembly by the calcium-dependent protease calpain. J. Cell Sci. 115, 3415–3425. 6. Tullio, R.D., Passalacqua, M., Averna, M., Salamino, F., Melloni, E. & Pontremoli, S. (1999) Changes in intracellular localization of calpastatin during calpain activation. Biochem. J. 343, 467–472. 7. Johnson, G.V. & Guttmann, R.P. (1997) Calpains: intact and active? Bioessays 19, 1011–1018. 8. Rutledge, T.W. & Whiteheart, S.W. (2002) SNAP-23 is a target for calpain cleavage in activated platelets. J. Biol. Chem. 277, 37009–37015. 9. Barnes, J.A. & Gomes, A.V. (1995) PEST sequences in calmo- dulin-binding proteins. Mol. Cell Biochem. 149, 17–27. 10. Barnes, J.A. & Gomes, A.V. (2002) Proteolytic signals in the primary structure of annexins. Mol. Cell Biochem. 231, 1–7. 11. Molinari, M., Anagli, J. & Carafoli, E. (1995) PEST sequences do not influence substrate susceptibility to calpain proteolysis. J. Biol. Chem. 270, 2032–2035. 12. Shumway, S.D., Maki, M. & Miyamoto, S. (1999) The PEST domain of IjBa is necessary and sufficient for in vitro degradation by l-calpain. J. Biol. Chem. 274, 30874–30881. 13. Fernandez, E., Aubry, L., Benyamin, Y. & Ouali, A. (2000) Co-localization of calpain p94 and calcium ions on N1 and N2 lines of bovine muscle fibers. Partial evidence for a similar locali- zation of calpain 1. In Myologie 2000 conference proceedings, p.160. AFM, Paris, France. 14. Tatsumi, R., Maeda. K., Hattori, A. & Takahashi, K. (2001) Calcium binding to an elastic portion of connectin/titin filaments. J. Muscle Res. Cell Motil. 22, 149–162. 15. Wallace, R.W., Tallant, E.A. & McManus, M.C. (1987) Human platelet calmodulin-binding proteins: identification and Ca 2+ -dependent proteolysis upon platelet activation. Biochemis- try 26, 2766–2773. 16. Selliah, N., Brooks, W.H. & Roszman, T.L. (1996) Proteolytic cleavage of alpha-actinin by calpain in T cells stimulated with anti- CD3 monoclonal antibody. J. Immunol. 156, 3215–3221. 17. Molinari, M., Maki, M. & Carafoli, E. (1995) Purification of l-calpain by a novel affinity chromatography approach: new insights into the mechanism of the interaction of the protease with targets. J. Biol. Chem. 270, 14576–14581. 18. Strobl, S., Fernandez-Catalan, C., Braun, M., Huber, R., Masu- moto, H., Nakagawa, K., Irie, A., Sorimachi, H., Bourenkow, G., Bartunik, H., Suzuki, K. & Bode, W. (2000) The crystal structure of calcium-free human l-calpain suggests an electrostatic switch mechanism for activation by calcium. Proc. Natl Acad. Sci. USA 97, 588–592. 19. Hata, S., Sorimachi, H., Nakagawa, K., Maeda, T., Abe, K. & Suzuki, K. (2001) Domain II of m-calpain is a Ca 2+ -dependent cysteine protease. FEBS Lett. 501, 111–114. 20. Moldoveanu, T., Hosfield, C.M., Lim, D., Elce, J.S., Jia, Z. & Davies, P.L. (2002) A Ca 2+ switch aligns the active site of calpain. Cell 108, 649–660. 21. Tompa, P., Emori, Y., Sorimachi, H., Suzuki, K. & Friedrich, P. (2001) Domain III of calpain is a Ca 2+ -regulated phospholipid- binding domain. Biochem. Biophys. Res. Commun. 280, 1333– 1339. 22. Hosfield, C.M., Moldoveanu, T., Davies, P.L., Elce, J.S. & Jia, Z. (2001) Calpain mutants with increased Ca 2+ sensitivity and implications for the role of the C(2)-like domain. J. Biol. Chem. 276, 7404–7407. 23.Michetti,M.,Salamino,F.,Minafra,R.,Melloni,E.&Pon- tremoli, S. (1997) Calcium-binding properties of human ery- throcyte calpain. Biochem. J. 325, 721–726. 24. Baron, M.D., Davison, M.D., Jones, P. & Critchley, D.R. (1987) The sequence of chick alpha-actinin reveals homologies to spectrin and calmodulin. J. Biol. Chem. 262, 17623–17629. 25. Taylor,R.G.,Papa,I.,Astier,C.,Ventre,F.,Benyamin,Y.& Ouali, A. (1997) Fish muscle cytoskeleton integrity is not depen- dent on intact thin filaments. J. Muscle Res. Cell Motil. 18, 285– 294. 26. Arimura, C., Suzuki, T., Yanagisawa, M., Imamura. M., Hamada, Y. & Masaki, T. (1988) Primary structure of chicken skeletal muscle and fibroblast a-actinins deduced from cDNA sequences. Eur. J. Biochem. 177, 649–655. 27. Parr, T., Waites, G.T., Patel, B., Millake, D.B. & Critchley, D.R. (1992) A chick skeletal-muscle a-actinin gene gives rise to two alternatively spliced isoforms which differ in the EF-hand Ca 2+ - binding domain. Eur. J. Biochem. 210, 801–809. 28. Thompson, V.F. & Goll, D.E. (2000) Purification of l-calpain, m-calpain, and calpastatin from animal tissues. Methods Mol. Biol. 144, 3–16. 29. Lebart, M.C., Mejean, C., Roustan, C. & Benyamin, Y. (1993) Further characterization of the a-actinin–actin interface and comparison with filamin-binding sites on actin. J. Biol. Chem. 268, 5642–5648. 30. Papa, I., Mejean, C., Lebart, M.C., Astier, C., Roustan, C., Benyamin, Y., Alvarez, C., Verrez-Bagnis, V. & Fleurence, J. (1995) Isolation and properties of white skeletal muscle a-actinin from sea trout (Salmo trutta) and bass (Dicentrarchus labrax). Comp. Biochem. Physiol. 112, 271–282. 31. Masumoto, H., Yoshizawa, T., Sorimachi, H., Nishino, T., Ishi- ura, S. & Suzuki, K. (1998) Overexpression, purification, and characterization of human m-calpain and its active site mutant, m-C105S-calpain, using a baculovirus expression system. J. Bio- chem. 124, 957–961. 32. Cong, J., Thompson, V.F. & Goll, D.E. (2002) Immunoaffinity purification of the calpains. Protein Expr. Purif. 25, 283– 290. 33. Cong, J., Goll, D.E., Peterson, A.M., Kapprell, H. & P0 (1989) The role of autolysis in activity of the Ca 2+ -dependent proteinases (l-calpain and m-calpain). J. Biol. Chem. 264, 10096–10103. 34. Papa, I., Astier, C., Kwiatek, O., Raynaud, F., Bonnal, C., Lebart, M.C., Roustan, C. & Benyamin, Y. (1999) a)Actinin-CapZ, an anchoring complex for thin filaments in Z-line. J. Muscle Res. Cell Motil. 20, 187–197. 35.Renoult,C.,Blondin,L.,Fattoum,A.,Ternent,D.,Maciver, S.K., Raynaud, F., Benyamin, Y. & Roustan, C. (2001) Binding of gelsolin domain 2 to Actin: An Actin interface distinct from that of gelsolin domain 1 and from ADF/Cofilin. Eur. J. Biochem. 268, 6165–6175. 36. Benyamin, Y., Roustan, C. & Boyer, M. (1986) Anti-actin anti- bodies: chemical modification allows the selective production of antibodies to the N-terminal region. J. Immunol. Methods 86, 21–29. 37. Dainese, E., Minafra, R., Sabatucci, A., Vachette, P., Melloni, E. & Cozzani, I. (2002) Conformational changes of calpain from human erythrocytes in the presence of Ca 2+ . J. Biol. Chem. 277, 40296–40301. Ó FEBS 2003 Calpain 1–a-actinin interaction (Eur. J. Biochem. 270) 4669 38. Pinset, C., Montarras, D., Chenevert, J., Minty, A., Barton, P., Laurent, C. & Gros, F. (1988) Control of myogenesis in the mouse myogenic C2 cell line by medium composition and by insulin: characterization of permissive and inducible C2 myoblasts. Dif- ferentiation 38, 28–34. 39. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 40. Astier,C.,Raynaud,F.,Lebart,M.C.,Roustan,C.&Benyamin, Y. (1998) Binding of a native titin fragment to actin is regulated by PIP2. FEBS Lett. 429, 95–98. 41. Bendayan, M., Nanci, A. & Kan, F.W. (1987) Effect of tissue processing on colloidal gold cytochemistry. J. Histochem. Cyto- chem. 35, 983–996. 42. Stirling, J.W. (1990) Immuno- and affinity probes for electron microscopy: a review of labeling and preparation techniques. J. Histochem. Cytochem. 38, 145–157. 43. Tsuchiya, H. & Seki, N. (1991) Action of calpain on a-actinin within and isolated from carp myofibrils. Nippon Suisan Gakkaishi 57, 1133–1139. 44. Goll, D.E., Dayton, W.R., Singh, I. & Robson, R.M. (1991) Studies of the a-actinin/actin interaction in the Z-disk by using calpain. J. Biol. Chem. 266, 8501–8510. 45. Taylor, R.G., Geesink, G.H., Thompson, V.F., Koohmaraie, M. & Goll, D.E. (1995) Is Z-disk degradation responsible for post- mortem tenderization? J. Anim. Sci. 73, 1351–1367. 46. Sorimachi, H., Kinbara, K., Kimura, S., Takahashi, M., Ishiura, S., Sasagawa, N., Sorimachi, N., Shimada, H., Tagawa, K. & Maruyama, K. (1995) Muscle-specific calpain, p94, responsible for limb girdle muscular dystrophy type 2A, associates with connectin through IS2, a p94-specific sequence. J. Biol. Chem. 270, 31158– 31162. 47. Konig, N., Raynaud, F., Feane, H., Durand, M., Mestre-Frances, N., Rossel, M., Ouali, A. & Benyamin, Y. (2003) Calpain 3 is expressed in astrocytes of rat and Microcebus brain. J. Chem. Neuroanat. 25, 129–136. 48. Delgado, E.F., Geesink, G.H., Marchello, J.A., Goll, D.E. & Koohmaraie, M. (2001) Properties of myofibril-bound calpain activity in longissimus muscle of callipyge and normal sheep. J. Anim. Sci. 79, 2097–2107. 49. Rami, A. (2003) Ischemic neuronal death in the rat hippocampus: the calpain-calpastatin-caspase hypothesis. Neurobiol. Dis. 13, 75–88. 50. Sandmann,S.,Prenzel,F.,Shaw,L.,Schauer,R.&Unger,T. (2002) Activity profile of calpains I and II in chronically infarcted rat myocardium – influence of the calpain inhibitor CAL 9961. Br.J.Pharmacol.135, 1951–1958. 51. Papa, I., Taylor, R., Astier, C., Ventre, F., Lebart, M.C., Roustan, C., Ouali, A. & Benyamin, Y. (1997) Dystrophin cleavage and sarcolemme detachment are early post mortem changes on bass (Dicentrarchus labrax)whitemuscle.J. Food Sci. 62, 917–921. 52. Neumar, R.W., Xu, Y.A., Gada, H., Guttmann, R.P. & Siman, R. (2003) Cross-talk between calpain and caspase proteolytic systems during neuronal apoptosis. J. Biol. Chem. 278, 14162–14167. 53. Kwak, K.B., Kambayashi, J., Kang, M.S., Ha, D.B. & Chung, C.H. (1993) Cell-penetrating inhibitors of calpain block both membrane fusion and filamin cleavage in chick embryonic myo- blasts. FEBS Lett. 323, 151–154. 54. Graham-Siegenthaler, K., Gauthier, S., Davies, P.L. & Elce, J.S. (1994) Active recombinant rat calpain II. Bacterially produced large and small subunits associate both in vivo and in vitro. J. Biol. Chem. 269, 30457–30460. 55. Yoshizawa, T., Sorimachi, H., Tomioka, S., Ishiura, S. & Suzuki, K. (1995) A catalytic subunit of calpain possesses full proteolytic activity. FEBS Lett. 358, 101–103. 56. Blanchard, A., Ohanian, V. & Critchley, D. (1989) The structure andfunctionofa-actinin. J. Muscle Res. Cell Motil. 10, 280–289. 57. Ohtsuka, H., Yajima, H., Maruyama, K. & Kimura, S. (1997) Binding of the N-terminal 63 kDa portion of connectin/titin to a-actinin as revealed by the yeast two-hybrid system. FEBS Lett. 401, 65–67. 58. Dewitt, S. & Hallett, M.B. (2002) Cytosolic free Ca 2+ changes and calpain activation are required for b integrin-accelerated phago- cytosis by human neutrophils. J. Cell Biol. 159, 181–189. 4670 F. Raynaud et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . The calpain 1–a-actinin interaction Resting complex between the calcium-dependant protease and its target in cytoskeleton Fabrice. calpain binding site in the C-terminal domain of smooth muscle a-actinin. Involvement of the C-terminal domain of ask in Calp1 binding In order to locate the