Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B Antje Tru ¨ mpler 1 , Bernhard Schlott 2 , Peter Herrlich 2 , Peter A. Greer 3 and Frank-D. Bo ¨ hmer 1 1 Institute of Molecular Cell Biology, Friedrich Schiller University, Jena, Germany 2 Leibniz Institute for Age Research – Fritz Lipmann Institute, Jena, Germany 3 Division of Cancer Biology and Genetics, Queen’s University Cancer Research Institute, Kingston, Canada Introduction Protein-tyrosine phosphatases (PTPs) are important regulators of cell proliferation, migration and survival [1,2]. Regulation by PTPs depends on modulation of their activity. PTPs are subject to several post-transla- tional modifications, e.g. phosphorylation, sumoy- lation, subcellular localization caused by different interacting partners and, in transmembrane PTPs, ligand binding [2–6]. In addition, the chemical proper- ties of the catalytic center cysteine of classical PTPs allow for yet another mode of regulation, reversible inactivation by oxidation. The chemical environment of this cysteine residue results in a low pK a of  5.4, and thus deprotonation of the sulfhydryl group under physiological conditions [7]. Although this property is a known prerequisite for catalytic activity, it also ren- ders PTPs highly susceptible to oxidation. Sensitivity to oxidation differs from phosphatase to phosphatase, and, possibly, within cells even among different cell types analyzed [8–10]. Interestingly, oxidation of PTPs has been observed upon growth factor stimulation of cells and the subsequent generation of hydrogen perox- ide (H 2 O 2 ) [11], and appears to be required for efficient signal transduction [12]. Insulin induces the oxidation of PTP1B (PTPN1), a physiological negative regulator of insulin receptor signaling [13]. Oxidation of PTP1B by H 2 O 2 converts the catalytic cysteine thiol group Keywords calpain; degradation; insulin signaling; oxidation; protein-tyrosine phosphatase Correspondence F D. Bo ¨ hmer, Institute of Molecular Cell Biology, Hans-Kno ¨ ll-Strasse 2, D-07745 Jena, Germany Fax: +49 3641 939 5602 Tel: +49 3641 939 5631 E-mail: boehmer@med.uni-jena.de (Received 20 January 2009, revised 22 July 2009, accepted 3 August 2009) doi:10.1111/j.1742-4658.2009.07255.x Protein-tyrosine phosphatases (PTPs) are regulated by reversible inactivat- ing oxidation of the catalytic-site cysteine. We have previously shown that reversible oxidation upon UVA irradiation is followed by calpain-mediated PTP degradation. Here, we address the mechanism of regulated cleavage and the physiological function of PTP degradation. Reversible oxidation of PTP1B in vitro strongly facilitated the association with calpain and led to greatly increased calpain-dependent inactivating cleavage. Both oxidation- induced association and cleavage depended exclusively on the presence of the catalytic (reversibly oxidized) cysteine residue 215. A major cleavage site was identified preceding amino acid position Ala77. In calpain-deficient cells, insulin signaling was apparently diminished, consistent with a possible role for calpain in removing a negative regulator of insulin signaling. Reversibly oxidized PTP1B may be a target of calpain in this context. Structured digital abstract l MINT-7234580: Calpain 4 (uniprotkb:P04632) and Calpain 1 (uniprotkb:P07384) physically interact ( MI:0915) with PTP1B (uniprotkb:P18031)byenzyme linked immunosorbent assay ( MI:0411) l MINT-7234562, MINT-7234481: Calpain 4 (uniprotkb:P04632) and Calpain 1 (uniprotkb: P07384) cleave (MI:0194) PTP1B (uniprotkb:P18031)byprotease assay (MI:0435) Abbreviations MEF, murine embryonic fibroblasts; PKB, protein kinase B (or AKT); PTP, protein-tyrosine phosphatase. 5622 FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS initially into sulfenic acid, which then undergoes an intramolecular reaction into a cyclic sulfenyl-amide structure, causing a conformational change in the cata- lytic center [14,15]. Oxidative PTP1B inactivation thus appears to function as a mechanism of positive feed- forward regulation of insulin signaling. We have previously discovered yet another level of PTP regulation: oxidation-dependent cleavage [16]. The requirement for Ca 2+ and inhibitor experiments point to an involvement of calpains in the oxidation- induced PTP degradation. Calpains are known to require Ca 2+ for their activation and exhibit rather broad substrate specificity with some preference for PEST-rich domains [17]. The ubiquitously expressed classical calpains 1 and 2 are heterodimers consisting of a common regulatory subunit (encoded by capn4) and a catalytic subunit (encoded by capn1 or capn2, respectively) [18]. In this study, we have explored the mechanism of oxidation-dependent cleavage of PTP1B in vitro and its possible physiological significance. We show that oxidation of the catalytic cysteine causes an increased association of PTP1B with calpain and trig- gers an inactivating cleavage, predominantly behind amino acid position 76. Results Enhanced calpain-mediated cleavage of PTP1B oxidized at the catalytic cysteine 215 PTP oxidation, including oxidation of PTP1B, upon UVA irradiation of cells or treatment with H 2 O 2 trig- gers proteolytic cleavage [16]. In an early study, C-ter- minal cleavage of PTP1B by calpain in platelets was described, leading to subcellular relocation without a change in activity [19]. To avoid this putative compli- cation, we made use of an engineered PTP1B lacking the C-terminal cleavage site in our in vitro experiments (scheme in Fig. 1A) [20]. Both oxidized and reduced recombinant PTP1B were incubated with increasing amounts of calpain, and degradation was analyzed by SDS ⁄ PAGE followed by Coomassie Brilliant Blue staining. As shown in Fig. 1B, oxidized PTP1B was readily cleaved by calpain. The amount of intact pro- tein decreased progressively with higher amounts of calpain (Fig. 1B), and in a time-dependent manner (Fig. 2D). Distinct proteolytic fragments appeared at intermediate calpain concentrations, notably fragments of  28–30 kDa. Longer incubation with calpain led to complete loss of detectable PTP1B (data not shown). By contrast, reduced PTP1B was less vulnera- ble to cleavage. In particular, the 28 kDa fragment was not produced, although larger fragments were generated from both oxidized and nonoxidized PTP1B (Fig. 1B and Coomassie Brilliant Blue stain of Fig. 2D). In vitro cleavage was not complete, but could be enhanced by increasing the amounts of calpain. We conclude that oxidized PTP1B exhibits enhanced susceptibility to calpain-mediated degradation. We then investigated which oxidation was relevant for the calpain-dependent cleavage. Previous studies using MS had shown that reversible oxidation of PTP1B under similar conditions as used here resulted in nearly exclusive oxidation of the catalytic cysteine, A B C Fig. 1. Calpain cleavage of PTP1B is strongly enhanced upon oxida- tion and inactivates the phosphatase. (A) Schematic drawing of full-length PTP1B (upper) indicating the known C-terminal calpain cleavage site, and of the truncated 37 kDa PTP1B (aa 1-321) used for the in-vitro studies. (B) Recombinant 37 kDa PTP1B (10 lg) was pretreated with or without 300 l M H 2 O 2 , excess of H 2 O 2 was inac- tivated by catalase and the samples were subjected to increasing amounts of calpain and subsequently analyzed by SDS ⁄ PAGE and Coomassie Brilliant Blue-staining. Arrows indicate oxidation-specific fragments (28–30 kDa) or undegraded PTP1B, respectively. (C) Aliquots of in vitro cleavage reactions of wild-type (WT) and cata- lytic center mutant (C215S) were analyzed with immunoblotting against the catalytic domain-epitope of PTP1B. Representative examples of at least three independent experiments are shown. A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS 5623 Cys215 [9,14]. In order to test whether oxidation of Cys215 is important for the enhanced cleavage by calpain, we compared cleavage of wild-type PTP1B and a PTP1B Cys215Ser mutant upon oxidative treat- ment. Fragments were detected by immunoblotting with an antibody directed against the PTP1B catalytic domain. Calpain treatment of oxidized wild-type PTP1B led to the formation of small fragments, including the major 28 kDa peptide (Fig. 1C). Oxi- dant-treated PTP1B Cys215Ser mutant protein was not cleaved, however. Thus, oxidation at Cys215 is indeed necessary to trigger calpain-dependent degradation. A C B D + H 2 O 2 (µM) 300010003001003010 Phosphatase activity (% of untreated) 0% 20% 40% 60% 80% 100% + H 2 O 2 + H 2 O 2 + DTT 3000 1000 300 100 30 10 – – H 2 O 2 (µM) Reversibly oxidized PTP1B DTT 10 30 100 300 H 2 O 2 (µM) – 1000 3000 300 –+ Calpain IB: anti PTP1B (N-terminus) IB: anti PTP1B (catalytic domain) Undegraded PTP1B 28–30 kDa fragments Undegraded PTP1B >30 kDa fragments IB: anti oxPTP Undegraded PTP1B 28–30 kDa fragments Fig. 2. Irreversible inactivation of PTP1B by oxidation-specific cleavage. (A) Recombinant PTP1B was incubated with increasing amounts of hydrogen peroxide (H 2 O 2 ), and subjected to phosphatase assay (black bars, upper) using p-nitrophenyl phosphate (pNPP) as substrate. Aliqu- ots were back-reduced with 5 m M dithiothreitol (gray bars) before pNPP assay. The activity of mock-treated PTP1B was set to 100%. Measurements were performed in triplicate, and the experiment was repeated three times with consistent results. In parallel, aliquots of H 2 O 2 -treated PTP1B were alkylated using 15 mM sodium iodoacetate and subjected to radioactive in-gel assay (lower). (B) Recombinant PTP1B (10 lg) was treated with increasing concentrations of H 2 O 2, excess H 2 O 2 was inactivated by catalase and the samples were subse- quently subjected to calpain cleavage (1 lg). Aliquots (corresponding to 0.5 lg PTP1B input) were analysed by immunoblotting using PTP1B antibodies against different epitopes: the N-terminus (upper), the catalytic domain (middle) and irreversibly oxidized PTPs (‘oxPTP antibody’, lower). (C) Recombinant PTP1B was treated with 300 l M H 2 O 2 followed by catalase, as under (B), and was then either directly subjected to calpain digestion, or was first treated with dithiothreitol (DTT; 5 m M for 30 min), subsequently catalase and then subjected to calpain diges- tion, as indicated. Controls were exposed to calpain alone or to calpain in presence of dithiothreitol. (D) Recombinant 37 kDa PTP1B (10 lg) was pretreated with or without 300 l M H 2 O 2 , excess of H 2 O 2 was inactivated by catalase, and the samples were subjected to digestion with 1 lg calpain for different times as indicated. Samples were then subjected to SDS ⁄ PAGE and Coomassie Brilliant Blue staining (upper), and aliquots to radioactive in-gel phosphatase assay (lower). Representative examples of at least three independent experiments are shown. Note that no PTP activity is detectable for the 28–30 kDa fragment, consistent with the lack of a part of the catalytic core (see Fig. 3). Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B A. Tru ¨ mpler et al. 5624 FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS Identification of the cleavage site in PTP1B: a major cleavage site in the N-terminus To analyze the fragments for residual PTP activity and determine the site of fragmentation, we oxidized recombinant PTP1B by treatment with H 2 O 2 and obtained calpain-dependent fragments for in vitro anal- ysis. Using dose-titration with H 2 O 2 , the concentration for optimal reversible oxidation was identified. Revers- ibility was assessed by two different types of PTP assay: enzymatic activity after reduction by dithiothrei- tol and in-gel phosphatase assay after prior incubation with the alkylating agent sodium iodoacetate. In the latter assay, only the reversibly oxidized PTP species is detected because it is protected from alkylation and subsequently reactivated by dithiothreitol [21]. Both assays yielded similar results indicating complete and up to 80% reversible inactivation at 300 lm H 2 O 2 (Fig. 2A). Higher H 2 O 2 concentrations caused substan- tial irreversible oxidation, although reversibly oxidized species could be readily detected even at 3000 lm H 2 O 2 . The modified in-gel assay resulted in an appar- ently better recovery of PTP activity at 1000–3000 lm H 2 O 2 , possibly caused by more extensive reduction. Subsequent to H 2 O 2 pretreatment and removal of H 2 O 2 by catalase treatment, PTP1B was subjected to calpain digestion and fragment formation was moni- tored with different antibodies. Pronounced cleavage of PTP1B and formation of the 28 kDa fragment could be seen upon pretreatment with 300–3000 lm H 2 O 2 using detection with an antibody recognizing the PTP1B catalytic domain. Under these conditions, PTP1B is both reversibly and irreversibly oxidized as described above. Either of the two species or both could potentially exhibit enhanced susceptibility to cal- pain. To distinguish between these possibilities, we first used an antibody detecting only irreversibly oxidized PTPs (‘oxPTP antibody’) [10]. This antibody reacted readily with non-degraded PTP1B which was pretreat- ed with 300–3000 lm H 2 O 2 . However, it gave only very weak signals with comparable amounts of the 28 kDa fragment (based on anti-PTP1B reactivity), indicating that this fragment contained only minor amounts of irreversibly oxidized PTP1B (Fig. 2B, low- est panel), and suggesting that it was derived from reversibly oxidized PTP1B. To further address this question, PTP1B was first oxidized with H 2 O 2 under conditions which result in the formation of reversibly and irreversibly oxidized species, and was subsequently treated with dithiothreitol to back-reduce the reversibly oxidized fraction. Upon calpain digestion under these conditions, degradation was greatly impaired and only the > 30 kDa species were formed, identical to the case of reduced PTP1B (Fig. 2C). These data clearly indicate that reversibly oxidized PTP1B, but not irre- versibly oxidized PTP1B, exhibits enhanced calpain sensitivity. An important question was how the observed oxida- tion-dependent fragmentation would affect PTP activ- ity. This was assessed by PTP in-gel assays, omitting prior alkylation. Under the reducing conditions of the in-gel-phosphatase assay only the full-length non- degraded PTP1B and fragments > 30 kDa retained abundant activity, whereas the 28 kDa fragment, char- acteristic of the oxidation-dependent cleavage, was inac- tive (Fig. 2D). Semiquantitative evaluation of the in-gel assay yielded relative activities of 100, 84, and 0.1 (arbi- trary units) of the undegraded PTP1B, fragments > 30 kDa and the 28 kDa fragment, respectively. Immunoblot experiments revealed that the 28 kDa fragment resulted from an N-terminal cleavage: the 28 kDa fragment did not react with an antibody recog- nizing an N-terminal epitope, although well recognized by an antibody specific for the Cys215-carrying catalytic domain (Fig. 2B), despite the fact that the fragment was inactive. The exact cleavage site was determined by N- terminal sequencing of the 28 kDa fragment. The first amino acid of the fragment was Ala77. The calculated size of the peptide corresponding to amino acids 77–321 is 28.2 kDa, which matches the observed size. Sequenc- ing of some of the minor fragments of lower molecular mass gave the same N-terminus (not shown). Thus, cal- pain-mediated cleavage of oxidized PTP1B occurs, at least predominantly, within the N-terminus, possibly in a step-wise fashion. The PTP fragment of 28 kDa lacks part of the PTP catalytic core including the invariant resi- dues Asn44, Arg45, Tyr46, Tyr66 and Asn68 (Fig. 3A) [20], consistent with its lack of activity. The chemical con- stitution of the > 30 kDa fragments, which occur in the absence of oxidation, was analyzed by MS. Consistent with the immunoblotting experiments (Fig. 2B) and with retained PTP activity (Fig. 2D), these fragments possess an intact N-terminus and lack peptides from the C-termi- nus downstream of amino acid 279 (details in Fig. S1). Augmented association of oxidized PTP1B with calpain in vitro We considered two possibilities for the mechanism of oxidation-enhanced, calpain-mediated cleavage. Enhanced cleavage could be either caused by better accessibility of the cleavage site in oxidized PTP because of an alteration in the PTP1B structure, or result from enhanced interaction of PTP with calpain. The first possibility is unlikely, however, based on the overlay of the crystal structures of both reduced and A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS 5625 reversibly oxidized PTP1B: this comparison showed no difference in the area of the cleavage site which is located in the loop connecting b-sheets b2 and b3as indicated in Fig. 3B [14,20]. The alternative explanation for enhanced degra- dation by calpain could be an increased affinity of calpain to oxidized PTP1B. Oxidation creates confor- mational changes in PTP1B in the vicinity of the cata- lytic site (Fig. 3B) [14]. The altered conformation may affect an interacting domain, although not the cleavage site directly, and may facilitate association with calpain, and in turn cleavage of PTP1B. This possibil- ity was experimentally assessed. We analyzed binding of oxidized versus nonoxidized PTP1B to inactive calpain in an ELISA-based assay. To this end, calpain was immobilized in 96-well plates in the presence of an irreversible calpain inhibitor. Thereafter, it was incu- bated with oxidized or untreated PTP1B, and the amount of bound PTP1B was detected by an immune reaction. As shown in Fig. 4A, pretreatment of PTP1B with increasing concentrations of H 2 O 2 strongly increased the amount bound to calpain. Importantly, the nonoxidizable Cys215Ser mutant of PTP1B showed considerably less association. Thus, Cys215 oxidation induced both calpain–PTP association (Fig. 4A) and, as shown above, PTP cleavage. Further, the lack of H 2 O 2 -induced association of the Cys215Ser mutant also excludes that the binding is merely caused by some nonspecific effect of H 2 O 2 such as increased charges. We considered that generation of a PTP1B mutant which was resistant to oxidation-dependent cleavage by calpain would be an excellent tool with which to investigate the role of this degradation for signal transduction. Mutation of Ala77 itself was excluded from this approach because this residue is buried in the structure and mutation was expected to destabilize the protein (D. Barford, personal communi- cation). Instead, mutation of amino acid residues sur- rounding the identified cleavage site (Lys73Ala, Glu75Ala, Glu76Gln, Glu76Ala, Gln78Ala and the double mutant Lys73Ala ⁄ Gln78Ala) was carried out and the corresponding recombinant proteins were ana- lyzed. These mutations had no impact on the associa- tion of PTP1B or on cleavage: oxidized PTP1B mutated in these positions was efficiently cleaved (data not shown). This is consistent with the notion that the interaction site and cleavage site are not identical. We also attempted, but failed, to consistently detect complexes of reversibly oxidized PTP1B and calpain in intact cells by co-immunoprecipitation (data not shown). Although complexes of calpain-regulatory subunits with target proteins have previously been shown [22,23], calpain–substrate interactions appear gener- ally difficult to monitor by co-immunoprecipitation. Calpain activity requires Ca 2+ and conformational changes upon Ca 2+ binding are necessary for the rec- ognition of some substrates [24–26]. We therefore tested whether Ca 2+ is important for binding to PTP1B. As shown in Fig. 4B, the H 2 O 2 -induced asso- ciation of PTP1B with calpain was Ca 2+ independent. Insulin signaling to protein kinase B (PKB/AKT) and p70-S6 kinase appears attenuated in calpain-deficient cells PTP1B is a major negative regulator of insulin signal- ing [11,13,27,28]. Insulin receptor stimulation is also Reduced conformation Oxidized conformation Cys215 / sulphenyl amide Identified cleavage site β3 β2 A B Fig. 3. Calpain removes an N-terminal fragment from oxidized PTP1B. (A) Amino acid sequence of the 37 kDa PTP1B (aa 1-321). Amino acids of the phosphatase domain are underlined, and invari- ant residues within the N-terminal part are in bold and boxed. The identified cleavage site between Glu76 and Ala77 is indicated by a gray arrowhead. The removed N-terminus upon cleavage is in gray letters, and the remaining C-terminus is black. Asterisks indicate the amino acids that have been substituted with the aim of achiev- ing calpain resistance. (B) Overlay of crystal structures of reduced PTP1B (2HNQ), and of the sulfenyl amide, reversibly oxidized PTP1B (1OEM) has been carried out using PYMOL 0.99rc6 (DeLano Scientific LLC, San Francisco, CA, USA). The cleavage site loop located between b-sheets b2 and b3, as well as regions showing conformational differences among the two structures are coloured in yellow (reduced form) and green (sulfenyl amide form). Other areas are kept in light and dark gray, respectively. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B A. Tru ¨ mpler et al. 5626 FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS known to cause reversible oxidation of PTP1B, which is considered to be an important mechanism of signal enhancement [11]. According to our data, reversibly oxidized PTP1B should be prone to calpain-mediated degradation. Because insulin-dependent oxidation pre- sumably reaches only a small fraction of PTP1B in proximity to activated insulin receptors, bulk degrada- tion is unlikely to occur. Absence of calpain should, however, cause a reduction in insulin signaling pro- vided the PTP1B cleavage was relevant for enhanced signal transduction. To address this question we used immortalized fibroblasts from wild-type or capn4 knockout mice, which have no detectable calpain 1 or 2 activity [29]. The cells were stimulated with insulin and analyzed with respect to signal transduction. As shown in Fig. 5, insulin stimulation did not signifi- cantly alter total PTP1B levels in these cells, and had no obvious effects on cellular distribution of PTP1B detected by immunocytochemistry (Fig. S2). Also, we could not detect effects of insulin stimulation on total levels of T-cell PTP, a close relative of PTP1B which likewise negatively controls insulin signaling [30] (not shown). Interestingly, however, the activation of two critical downstream signaling mediators, AKT and p70-S6 kinase, was significantly reduced in capn4 knockout fibroblasts, compared with wild-type fibro- blasts. Importantly, lentiviral rescue by Capn4 of the capn4 knockout cells was able to almost fully rescue insulin-induced AKT and S6 kinase activation. It should be noted that in unstimulated wild-type cells, P-AKT and P-p70-S6Kinase levels were higher than in calpain 4 knockout or rescued cells, possibly reflecting a role for calpain in the regulation of basal activity of these pathways. Phosphorylation of a protein of  120 kDa, which reacted with antibodies against phosphorylated insulin receptor, but is apparently not representing the insulin receptor itself, was also strongly diminished in the calpain negative cells (Fig. 5), whereas using the same antibody no obvious differences could be detected in phosphorylation of im- munoprecipitated insulin receptor b subunit in absence or presence of Capn4 (Fig. S3). Activation of ERK1 ⁄ 2 by insulin stimulation was very low under the employed conditions, and not reduced or even some- what higher in the absence of calpain (Fig. 5). Appar- ently, calpain affects the strength of insulin signaling to AKT, and p70-S6 kinase. The detailed mechanism and the further biological consequences of Capn4- deficiency for insulin signaling require further investi- gation. Discussion We have shown previously that PTPs are subject to calpain-dependent degradation under conditions which cause PTP oxidation and calpain activation, such as UVA irradiation [16]. In this study, we reconstituted the oxidation-dependent degradation of PTP1B by cal- pain in vitro in order to prove that calpain is responsi- ble for the cleavage reaction, to identify the site of cleavage and the mechanism of the reaction. Oxidation of PTP1B at the catalytic Cys215 is required for both AB Fig. 4. Oxidation enhances binding of PTP1B to calpain. (A) Preoxidized recombinant PTP1B wild-type (black bars) and the PTP1B C215S mutant (gray bars) were allowed to bind to calpain-coated 96-well plates in absence of Ca 2+ . Bound PTP1B was detected with a PTP1B anti- body, and horseradish peroxidase-coupled secondary antibody. Calpain was preincubated with an irreversible inhibitor to prevent degradation of bound PTP1B. The values were calculated relative to the association of wild-type PTP1B in the absence of H 2 O 2 treatment (100%). (B) Association of PTP1B pretreated with (white bars) or without (black bars) 300 l M H 2 O 2 to calpain was assessed as in (A). Experiments were performed either in the absence of Ca 2+ (EGTA) or in the presence of 1 mM CaCl 2 (Ca 2+ ). Data present means of triplicate measures and consistent results have been obtained in at least three independent experiments. A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS 5627 the association of calpain with its substrate and for the cleavage reaction. Cleavage is induced by reversible but apparently not by irreversible oxidation, and a major cleavage occurs at amino acid position 77. The resulting enzyme fragments are inactive, consistent with the fact that important positions of the catalytic center are removed. What determines calpain-substrate association? In addition to modulation of the activity of an enzyme, altered interaction affinity is frequently used in cellu- lar regulation. For example, Jun N-terminal kinase finds its substrate Jun through a domain remote from the phosphorylation sites [31]. The serine phos- phatase MYPT-1-PP1 associates with and acts on its substrate merlin only after its own dephosphoryla- tion, and again substrate binding and dephosphoryla- tion of the substrate occur at different sites [32]. Often the substrate appears to be made susceptible for interaction with an enzyme. In our report, the PTP is modified by oxidation to become an inter- action partner for calpain, although we still lack evidence that this complex formation can occur in intact cells. Another substrate of calpain, a-actinin, is influenced by binding to phosphoinositides: phos- phatidylinositol-3,4,5-trisphosphate enhances and phosphatidylinositol-4,5-diphosphate inhibits the cleavage [33]. The subcellular localization or specific cellular conditions may determine the cleavage, as with PTP1B in platelets [19]. The interaction in vitro is apparently independent of Ca 2+ (Fig. 4B) suggest- ing that it may occur through a calpain region which is distant from the active center. The interaction domain in PTP1B or the conforma- tional determinants have not yet been mapped. They may define the cleavage site. Reversible Cys215 oxida- tion generates a conformational change [14,15], which apparently catalyzes the association with calpain and leads to proteolytic action around amino acid position 77. For example, oxidation causes exposure of the amino acid residues Tyr46 of the phosphoTyr loop and Ile219 of the PTP loop, which could be important for calpain binding [14]. Oxidation-specific cleavage does not occur with nonoxidized PTP1B. Nonoxidized PTP1B is instead cleaved C-terminally and activity is retained in the fragments occurring under these condi- tions (see Figs 1B, 2D and S1). Similarly, in activated platelets, calpain splits an 8 kDa fragment off the PTP1B C-terminus [19], and the truncated PTP1B retains the enzymatic activity. The lost C-terminal fragment carries an endoplasmic reticulum localization signal causing relocalization of truncated PTP1B to the cytosol. It thus appears that the state of the sub- strate defines the association with calpain, the cleavage site and the resulting enzymatic activity of the substrate. In capn1-negative mouse platelets, PTP1B protein levels and enzyme activity were increased, resulting in reduced overall protein tyrosine phosphor- ylation levels and impaired platelet aggregation. This phenotype was rescued in capn1 and PTP1B doubly deficient mice [34]. With respect to this platelet pheno- type, calpain thus seems to negatively regulate PTP1B activity. It is not currently known whether PTP1B oxidation is involved in this context. P-AKT P-p70-S6 kinase AKT p70-S6 kinase Ø 5 10 15 30 Wild-typecapn4 neg. Lenti-rescued Ø 5 10 15 30Ø 5 1015 30 (min) Insulin Vinculin PTP1B 72 95 130 P-InsR InsR-β 72 95 130 Calpain 4 26 panERK P-ERK Fig. 5. Calpain-deficiency results in reduced activation of AKT upon insulin stimulation of cells. Immortalized murine fibroblasts, either calpain 4 knockout (capn4 neg.), wild-type or calpain 4 knockout with rescued calpain 4 expression by lentiviral transduction (lenti- rescued), were starved in 0.5% serum overnight, and stimulated with 10 n M insulin for the indicated times. Lysates (100 lg) were subjected to immunoblotting and analyzed for levels of PTP1B as well as for the phosphorylation state of AKT (Ser473), p70-S6Kinase (Thr389), the insulin receptor (Tyr1162,1163) and ERK1 ⁄ 2. Note that pEKR1 ⁄ 2 phosphorylation became detectable only after prolonged exposure. The antibody against phosphorylated insulin receptor detects an obviously unrelated phosphoprotein in the lysates. Detection of insulin receptor phosphorylation requires prior immunopreciptitation (see Fig. S3). Controls for total amounts of the corresponding proteins, and for calpain 4 expression are also shown. Note that re-expression of calpain 4 was performed with a construct encoding a somewhat truncated form. Total levels of PTP1B were consistently somewhat higher in the Capn4-rescued cells and total ERK1 ⁄ 2 levels in calpain 4 knockout cells from unknown reasons, presumably reflecting an unrelated property of the independently immortalized cell pools. The blots in panels 1–6, 11 (from top) are representative of 5, in panels 7–10 (from top) for three independent experiments. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B A. Tru ¨ mpler et al. 5628 FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS PTP1B is not the only PTP accessible to calpain. For example, SHP-1 and LAR-PTP were targets upon UVA irradiation of cells [16]. Accordingly, oxidation- dependent degradation of PTPs by calpain may be of more general importance. The sensitivity to oxidation may vary with the PTP. For example, the lower sensi- tivity of SHP-1 for UVA-induced degradation than that of PTP1B may reflect its comparatively lower sen- sitivity to reversible oxidation [8,9]. The sequence of the loop harboring the oxidation-induced cleavage site identified in PTP1B is quite different among different PTPs. Cleavage may nevertheless occur at correspond- ing positions, given the relatively low sequence require- ments for calpain cleavage evidenced by efficient proteolysis of our PTP1B mutants around position 77. Mutants of Lys73, Glu75, Glu76 and Gln78 are not expected to exhibit grossly altered PTP structure [17]. They were similarly active, similarly reversibly oxidized and, when oxidized, showed in vitro similar calpain- sensitivity as wild-type PTP1B (data not shown). It has recently been shown by Groen et al. [4] that oxidation leads to enhanced and altered degradation of four different receptor-like PTPs by trypsin. In addition to conformational changes in the catalytic domains upon oxidation, the authors link changes in PTP dimerization to the altered susceptibility to prote- olysis. It is tempting to speculate that susceptibility to cell-endogenous proteinases such as calpain may also be affected by oxidation-associated dimerization. Calpain is a cysteine-proteinase and has been reported to be prone to inactivating oxidation [35,36]. Under the conditions used here for PTP1B oxidation, calpain activity appeared, however, unaffected. Also, PTP1B cleavage occurred irrespective of the addition of catalase to the in vitro experiments (not shown). Thus, calpain seems to be less sensitive to H 2 O 2 than PTP1B. Is regulated proteolysis by calpain a physiologically relevant process? The association of calpain with a-actinin appears to be important for actin dynamics in tumor cells [33]. Calpain has been reported to par- ticipate in integrin b3 inside-out signaling [37,38]. As alluded to above, reduced levels of tyrosine phosphory- lation and reduced platelet aggregation have been observed in calpain 1-negative platelets and have been attributed to a specific PTP1B cleavage by calpain [34]. Furthermore, calpain-negative primary murine embryonic fibroblasts (MEF) exerted an altered response towards a variety of cell death stimuli (e.g. tumour necrosis factor a, staurosporine, puromycin) when compared with wild-type cells [39]. In the absence of calpain, AKT phosphorylation was reduced and the apoptosis rate was increased, indicating an involvement of calpain in AKT activation possibly by removing a phosphatase. Very recently, reduced AKT activity in calpain 1-negative fibroblasts was attributed to enhanced stability of the B56alpha regulatory sub- unit of PP2A, leading to enhanced association of PP2A with AKT [40]. The data obtained suggest a possible role of calpain in insulin signaling. Target phosphorylation of insulin was reduced in calpain-negative cells (capn4 knockout MEFs) and restored upon lentiviral introduction of calpain. Because insulin stimulation has been shown to induce oxidation of PTP1B [41], one could speculate that the here-described increased susceptibility of oxidized PTP1B for an inactivating degradation by calpain may play a role in this context. However, the currently available data do not allow us to establish a causal link between insulin signaling and potential PTP1B degradation yet. For this, it will be necessary to show inactivating fragmentation of PTP1B upon insulin stimulation, and interference with insulin sig- naling by PTP1B mutants which are less susceptible to degradation. It is possible that the above described mechanism of calpain-mediated positive control of AKT [40] also contributes to the effects we have observed in our setting of insulin stimulation. Further, it is tempting to speculate that the require- ment of calpain for insulin signaling described here may have implications for pathology. There have been a number of reports linking calpain dysfunction to type II diabetes mellitus and type II diabetes mellitus- related phenotypes [42–48]. We have obtained indications for a previously unrecognized positive role of calpain for insulin signaling, prompting further investigation of the underlying mechanisms. Materials and methods Materials Hydrogen peroxide (#H1009), catalase (#C9322), p-nitro- phenyl phosphate (#S0942), sodium iodoacetate (#I9148), 3,3¢,5,5¢-tetramethylbenzidine (#T2885), poly(Glu,Tyr) (#P0275), E64 (#E3132), insulin (#I0516) and poly-l-lysine (#P2636) were purchased from Sigma-Aldrich (Taufkirchen, Germany). Calpain (#208713) and PD150606 (#513022) were from Merck Biosciences (Nottingham, UK). Complete protease inhibitor cocktail (#11836170001) was from Roche Diagnostics (Mannheim, Germany). Anti-PTP1B against the catalytic domain (#PH01) was from Merck Biosciences; against the N-terminus of PTP1B (sc1718) and anti-insulin Rb (#sc711) were from Santa Cruz (Santa Cruz, CA, USA); anti-PTP1B against the N-terminus of PTP1B (AP8411c) was from Abgent (San Diego, CA, USA); and anti-calpain (C0355) was from Sigma-Aldrich. Antibodies A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS 5629 against phospho-Ser473-AKT (#4058S), panAKT (#9272), phospho-Thr389-S6 kinase (#9202), S6-kinase (#2217) and phospho-p44 ⁄ 42 MAPK (Erk1 ⁄ 2) (Thr202 ⁄ Tyr204) (#9106S) were from Cell Signaling (Danvers, MA, USA); against murine PTP1B (#07-088) was from Millipore (Sch- walbach, Germany); anti-insulin ⁄ insulin-like growth factor- 1 receptor (IR ⁄ IGF1R) [pYpY1162 ⁄ 1163] (#44804G) and Alexa Fluor 488 goat anti-(rabbit IgG) (#A11034) were from Invitrogen (Carlsbad, CA, USA). Anti-vinculin (V248) was from Upstate (Biomol, Hamburg, Germany), horseradish peroxidase-coupled secondary antibodies [anti- mouse and anti-(rabbit IgG)] were from KPL (Gaithers- burg, MD, USA) or Pierce (#1858413; Rockford, IL, USA). Polyclonal oxPTP antibody was a kind gift from Arne O ¨ stman (Stockholm, Sweden) [10]. Wild-type (PZ ⁄ PZ), capn4 knockout (P ⁄ P), and lentivirally rescued P ⁄ P immortalized mouse embryonic fibroblasts have been described [29]. Cells were routinely kept in DMEM supplied with 10% fetal bovine serum and passaged every 2–3 days by trypsinization. Immortalized murine fibroblasts derived from mice with disruption of the PTP1B gene were kindly provided by Ben Neel and Fawaz Haj and have been described earlier [49]. These cells were routinely kept in DMEM supplied with 10% newborn bovine serum. Plasmids and mutants preparation A plasmid encoding the C-terminally truncated variant of wild-type PTP1B (aa 1-321) for recombinant expression was a kind gift from D. Barford (Institute of Cancer Research, London, UK) [20]. Mutants were created by site-directed mutagenesis using Pfu turbo polymerase (Stratagene, La Jo- lla, CA, USA). The forward primer for C215S mutant was: 5¢-CCCGTTGTGGTGCACTCC AGTGCAGGCATCG GC A-3¢ (other primer sequences can be obtained on request). Expression was performed in Escherichia coli Rosetta, and purification to homogeneity by chromatography was carried out according to a protocol slightly modified from the previously published procedure [20]. Briefly, separation was initially performed using a HiTrap Q anion exchange column (GE-Amersham, Freiburg, Germany), and the pooled fractions were directly subjected to separation with a phenyl- superose column (GE-Amersham). In vitro degradation assay Recombinant PTP1B (10 lg in reaction volume of 50 lL) was pretreated with or without 300 lm H 2 O 2 at room tem- perature in the dark for 1 h. Afterwards, catalase was added to a concentration of 0.3 lg Æ mL )1 and incubation continued for 30 min. The cleavage reaction was started with addition of 1 mm CaCl 2 and 3 lg calpain achieving final volume of 70 lL per reaction (for alternate amounts of calpain see figure legends), and allowed to proceed for another 30 min at 20 °C. The reaction was stopped by addition of SDS ⁄ PAGE sample buffer and boiling at 95 °C for 5 min. Aliquots were separated on 13% SDS ⁄ PAGE and either Coomassie Brilliant Blue-stained or transferred to a poly(vinylidene difluoride) membrane. Membranes were either probed with antibodies against different PTP1B epitopes or Coomassie Brilliant Blue-stained. For identifica- tion of fragments, the Coomassie Brilliant Blue-stained oxi- dation-specific bands on the poly(vinylidene difluoride) membrane were cut out and sequenced by EDMAN degradation with 494A Procise protein sequencer (Applied Biosystems, Foster City, CA, USA). In vitro association assay Assays were performed according to a published protocol [50]. Briefly, 96-well plates were coated with 0.1 lg calpain per well in the presence of the irreversible inhibitor 1 lm E64, either with 1 mm Ca 2+ or with 1 mm EGTA overnight. All washings were carried out using NaCl ⁄ P i ⁄ Tween (0.5 m NaCl, 2.7 mm KCl, 4.3 mm Na 2 HPO 4 , 1.4 m m KH 2 PO 4 , pH 7.4, 0.2% Tween-20). Plates were blocked against unspe- cific binding of proteins with blocking solution consisting of 2.5% milk and 1% BSA in NaCl ⁄ P i for 3 h. PTP1B was preincubated with or without H 2 O 2 as described above and, after washing the plates twice, allowed to bind to immobi- lized calpain for 2 h. Washing was repeated three times and primary PTP1B antibody as well as secondary antibody (PH01 1 : 1000 and anti-mouse IgG–horseradish peroxidase 1 : 10 000, respectively) diluted in blocking solution were incubated 2 h each. Detection was done using 3,3¢,5,5¢-tetra- methylbenzidine as substrate and 1 m HCl to stop reaction after  30 min. Quantification was carried out by measuring A at 405 nm. Insulin stimulation Cells were starved overnight in 0.5% serum, washed twice with serum-free DMEM, and stimulated for the indicated times with 10 nm insulin at 37 °C. After washing twice with ice-cold NaCl ⁄ P i , cells were lysed in ice-cold lysis buffer (50 mm Hepes pH 7.5, 150 mm NaCl, 0.5% NP-40, 10 mm NaPP, 10 mm b-glycerophosphate, 1 mm NaVO 4 ,50mm NaF and protease inhibitors). Protein concentration was determined using BCA kit (Pierce). Adjusted volumes were boiled in SDS ⁄ PAGE sample buffer, separated on 7.5% SDS ⁄ PAGE, transferred onto poly(vinylidene difluoride) membrane and probed with respective antibodies. Stripping of membranes was performed in 62.5 mm Tris pH 6.8, 2% SDS, and 100 mm b-mercaptoethanol for 30 min at 50 °C. Immunhistochemistry Cells were plated on polylysine-coated cover slips in 24-well plates and starved overnight in 0.5% serum. After washing Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B A. Tru ¨ mpler et al. 5630 FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS twice with serum-free DMEM cells were incubated for 30 min with or without 10 nm insulin in DMEM, washed twice with ice-cold NaCl ⁄ P i and fixed with 3.7% formalde- hyde in NaCl ⁄ P i for 10 min at room temperature. Permea- bilization was carried out after washing with NaCl ⁄ P i with 5 min incubation with 0.2% Triton X-100 in NaCl ⁄ P i and followed by blocking using 1.5% BSA in NaCl ⁄ P i . Incuba- tion with primary antibody against murine PTP1B (1 : 200 dilution in BSA ⁄ NaCl ⁄ P i ) was performed overnight at 4 °C. After washing with BSA ⁄ NaCl ⁄ P i , slides were incu- bated with secondary antibody against rabbit IgG (1 : 500 dilution in BSA ⁄ NaCl ⁄ P i , Alexa Fluor 488 coupled IgG) for 20 min in the dark. Slides were washed, mounted and kept in the dark until further analysis. Images were acquired using a LSM510 Meta from Carl Zeiss (Jena, Germany). Mass spectrometry Appropriate gel sections were subjected to in gel digestion with trypsin according to Shevchenko et al. [51] with one modification: reduction and oxidation of thiol goups was performed with a mixture of 0.1 mm tributylphosphine (Sigma-Aldrich) and 0.82 mm 4-vinylpyridine (Sigma- Aldrich). Sequencing grade modified trypsin was purchased from Promega (Mannheim, Germany). The tryptic in gel protein digests were analyzed with a LC-ESI-MS equipment consisting an Ettan MDLC Ô -HPLC (GE Healthcare, Frei- burg, Germany) coupled online to a Finnigan LTQ mass spectrometer (Thermo Electron Corp., Dreieich, Germany). The HPLC was equipped with a Zorbax 300SB, 5 lm, 5 · 0.3 mm trapping column and a Zorbax 300SB, 5 lm, 150 · 0.075 mm separation column (Agilent, Boeblingen, Germany). The separation of peptides on the HPLC occurred by applying a linear gradient running from 0% to 47% acetonitrile, followed by a stepwise elution with 84% acetontrile in 0.1% formic acid each under control of the unicornÔ software (GE Healthcare). The LTQ was oper- ated under control of the xcalibur 1.4Ô software (Thermo Electron Corp.). For processing of the CID-spectra (MS ⁄ MS-spectra spectra) with final protein identification the bioworks 3.2Ô software (Thermo Electron Corp.) and the NCBI human protein database were used. Acknowledgements This work was supported by grants DFG He 551⁄ 11 (to PH) and DFG Bo 1043 ⁄ 6-2 (to FDB) and from the European Community (MRTN-CT-2006-035830) (to FDB). We thank David Barford for discussions and provision of PTP1B expression constructs, and Benja- min Neel, Fawaz Haj and Arne O ¨ stman for reagents. Furthermore, we thank Andrea Uecker for technical assistance with a part of the experiments. References 1 Hendriks WJ, Elson A, Harroch S & Stoker AW (2008) Protein tyrosine phosphatases: functional inferences from mouse models and human diseases. FEBS J 275, 816–830. 2 Tonks NK (2006) Protein tyrosine phosphatases: from genes, to function, to disease. Nat Rev Mol Cell Biol 7, 833–846. 3 Flint AJ, Gebbink MF, Franza BR Jr, Hill DE & Ton- ks NK (1993) Multi-site phosphorylation of the protein tyrosine phosphatase, PTP1B: identification of cell cycle regulated and phorbol ester stimulated sites of phos- phorylation. EMBO J 12, 1937–1946. 4 Groen A, Overvoorde J, van der Wijk T & den Hertog J (2008) Redox regulation of dimerization of the recep- tor protein-tyrosine phosphatases RPTPalpha, LAR, RPTPmu and CD45. FEBS J 275, 2597–2604. 5 Brautigan DL & Pinault FM (1993) Serine phosphory- lation of protein tyrosine phosphatase (PTP1B) in HeLa cells in response to analogues of cAMP or diacylglyc- erol plus okadaic acid. Mol Cell Biochem 127–128, 121–129. 6 Dadke S, Cotteret S, Yip SC, Jaffer ZM, Haj F, Ivanov A, Rauscher F III, Shuai K, Ng T, Neel BG et al. (2007) Regulation of protein tyrosine phosphatase 1B by sumoylation. Nat Cell Biol 9, 80–85. 7 Lohse DL, Denu JM, Santoro N & Dixon JE (1997) Roles of aspartic acid-181 and serine-222 in intermedi- ate formation and hydrolysis of the mammalian protein-tyrosine-phosphatase PTP1. Biochemistry 36, 4568–4575. 8 Weibrecht I, Bo ¨ hmer SA, Dagnell M, Kappert K, O ¨ stman A & Bo ¨ hmer FD (2007) Oxidation sensitivity of the catalytic cysteine of the protein-tyrosine phosphatases SHP-1 and SHP-2. Free Radic Biol Med 43, 100–110. 9 Groen A, Lemeer S, van der Wijk T, Overvoorde J, Heck AJ, O ¨ stman A, Barford D, Slijper M & den Her- tog J (2005) Differential oxidation of protein-tyrosine phosphatases. J Biol Chem 280, 10298–10304. 10 Persson C, Sjo ¨ blom T, Groen A, Kappert K, Engstrom U, Hellman U, Heldin CH, den Hertog J & O ¨ stman A (2004) Preferential oxidation of the second phosphatase domain of receptor-like PTP-alpha revealed by an anti- body against oxidized protein tyrosine phosphatases. Proc Natl Acad Sci USA 101, 1886–1891. 11 Meng TC, Buckley DA, Galic S, Tiganis T & Tonks NK (2004) Regulation of insulin signaling through reversible oxidation of the protein-tyrosine phosphata- ses TC45 and PTP1B. J Biol Chem 279, 37716– 37725. 12 Sundaresan M, Yu ZX, Ferrans VJ, Irani K & Finkel T (1995) Requirement for generation of H 2 O 2 for platelet- derived growth factor signal transduction. Science 270, 296–299. A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS 5631 [...]... L & Goldstein BJ (2001) Insulin-stimulated hydrogen peroxide reversibly inhibits protein-tyrosine phosphatase 1b in vivo and enhances FEBS Journal 276 (2009) 5622–5633 ª 2009 The Authors Journal compilation ª 2009 FEBS A Trumpler et al ¨ 42 43 44 45 46 47 48 Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B the early insulin action cascade J Biol Chem 276, 21938–21942.. .Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B 13 Elchebly M, Payette P, Michaliszyn E, Cromlish W, Collins S, Loy AL, Normandin D, Cheng A, HimmsHagen J, Chan CC et al (1999) Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase- 1B gene Science 283, 1544–1548 14 Salmeen A,... Goldstein BJ (1995) Osmotic loading of neutralizing antibodies 5632 28 29 30 31 32 33 34 35 36 37 38 39 40 41 A Trumpler et al ¨ demonstrates a role for protein-tyrosine phosphatase 1B in negative regulation of the insulin action pathway J Biol Chem 270, 20503–20508 Kenner KA, Anyanwu E, Olefsky JM & Kusari J (1996) Protein-tyrosine phosphatase 1B is a negative regulator of insulin- and insulin-like growth... regulation of protein tyrosine phosphatase 1B involves a sulphenyl-amide intermediate Nature 423, 769–773 15 van Montfort RL, Congreve M, Tisi D, Carr R & Jhoti H (2003) Oxidation state of the active-site cysteine in protein tyrosine phosphatase 1B Nature 423, 773–777 16 Gulati P, Markova B, Gottlicher M, Bohmer FD & ¨ ¨ Herrlich PA (2004) UVA inactivates protein tyrosine phosphatases by calpain-mediated degradation. .. Barford D, Flint AJ & Tonks NK (1994) Crystal structure of human protein tyrosine phosphatase 1B Science 263, 1397–1404 21 Markova B, Gulati P, Herrlich PA & Bohmer FD ¨ (2005) Investigation of protein-tyrosine phosphatases by in-gel assays Methods 35, 22–27 22 Rosenberger G, Gal A & Kutsche K (2005) AlphaPIX associates with calpain 4, the small subunit of calpain, and has a dual role in integrin-mediated... silver-stained polyacrylamide gels Anal Chem 68, 850–858 Supporting information The following supplementary material is available: Fig S1 Chemical constitution of PTP1B fragments obtained by calpain-mediated cleavage in the absence of oxidation Fig S2 Detection of PTP1B in wild-type and calpaindeficient cells by immunocytochemistry Fig S3 Insulin receptor phosphorylation in Capn4deficient cells This supplementary... Friedrich P (2004) On the sequential determinants of calpain cleavage J Biol Chem 279, 20775–20785 18 Goll DE, Thompson VF, Li H, Wei W & Cong J (2003) The calpain system Physiol Rev 83, 731–801 19 Frangioni JV, Oda A, Smith M, Salzman EW & Neel BG (1993) Calpain-catalyzed cleavage and subcellular relocation of protein phosphotyrosine phosphatase 1B (PTP -1B) in human platelets EMBO J 12, 4843–4856 20... through inhibition of a merlin phosphatase Nature 442, 576–579 Sprague CR, Fraley TS, Jang HS, Lal S & Greenwood JA (2008) Phosphoinositide binding to the substrate regulates susceptibility to proteolysis by calpain J Biol Chem 283, 9217–9223 Kuchay SM, Kim N, Grunz EA, Fay WP & Chishti AH (2007) Double knockouts reveal that protein tyrosine phosphatase 1B is a physiological target of calpain1 in platelets... 19810–19816 Tan Y, Dourdin N, Wu C, De Veyra T, Elce JS & Greer PA (2006) Conditional disruption of ubiquitous calpains in the mouse Genesis 44, 297–303 Galic S, Hauser C, Kahn BB, Haj FG, Neel BG, Tonks NK & Tiganis T (2005) Coordinated regulation of insulin signaling by the protein tyrosine phosphatases PTP1B and TCPTP Mol Cell Biol 25, 819–829 Karin M & Gallagher E (2005) From JNK to pay dirt: jun... exposure of pancreatic islets to calpain inhibitors impairs mitochondrial fuel metabolism and the exocytosis of insulin Metabolism 52, 528–534 Smith LK, Rice KM & Garner CW (1996) The insulininduced down-regulation of IRS-1 in 3T3-L1 adipocytes is mediated by a calcium-dependent thiol protease Mol Cell Endocrinol 122, 81–92 49 Haj FG, Markova B, Klaman LD, Bohmer FD & Neel ¨ BG (2003) Regulation of receptor . peroxide reversibly inhibits protein-tyrosine phosphatase 1b in vivo and enhances Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase. the overlay of the crystal structures of both reduced and A. Tru ¨ mpler et al. Calpain-mediated degradation of reversibly oxidized protein-tyrosine phosphatase 1B FEBS