Non-catalytic domains of subunit A negatively regulate the activity of calcineurin

Biochimie 87 (2005) 215–221 www.elsevier.com/locate/biochi

Non-catalytic domains of subunit A negatively regulate the activity of calcineurin Ping Liu a, Chao Huang a, Zongchao Jia b, Fang Yi a, Da-yu Yu a, Qun Wei a,* a

Department of Biochemistry and Molecular Biology, Beijing Normal University, Beijing, 100875, China b Department of Biochemistry, Queen’s University, Kingston ON, Canada, K7 l 3N6 Received 16 October 2004; accepted 21 October 2004 Available online 10 November 2004

Abstract Calcineurin is composed of a catalytic subunit A (CNA) and a regulatory subunit B (CNB). In addition to the catalytic core, CNA further contains three non-catalytic domains—CNB binding domain (BBH), calmodulin binding domain (CBD), and autoinhibitory domain (AI). To investigate the effect of these three domains on the activity of CNA, we have constructed domain deletion mutants CNAa (catalytic domain only), CNAac (CNAa and CBD), and CNAaci (CNAa, CBD and AI). By using p-nitrophenylphosphate and 32P-labeled RII peptide as substrates, we have systematically examined the phosphatase activities, kinetics, and regulatory effects of Mn2+/Ni2+ and Mg2+. The results show that the catalytic core has the highest activity and the order of activity of the remaining constructs is CNAac>CNAaci>CNA. Sequential removal of the non-catalytic domains corresponds to concurrent increases of the phosphatase activity assayed under several conditions. This observation clearly demonstrates that non-catalytic domains negatively regulate the enzyme activity and act as intra-molecular inhibitors, possibly through restraining the conformation elasticity of the catalytic core required for optimal catalysis or interfering with substrate access. The sequential domain deletion favors activation of the enzyme by Mn2+/Ni2+ but not by Mg2+ (except for CNAa), suggesting that enzyme activation by Mn2+/Ni2+ is mainly mediated via the catalytic domain, whereas activation by Mg2+ is via both the catalytic core and noncatalytic domains. © 2004 Elsevier SAS. All rights reserved. Keywords: Calcineurin; Phosphatase activity; Non-catalytic domains; Intra-molecular regulator; Synergism

1. Introduction Calcineurin (CN) functions primarily as a phosphatase and participates in many cellular processes [1]. CN is the only known serine/threonine protein phosphatase whose activity is activated by Ca2+ and other metal ions, such as Mn2+ [2] and Ni2+ [3,4]. The enzyme is composed of catalytic subunit A (CNA) and regulatory subunit B (CNB). CNA can be further divided into two parts, a catalytic domain and a set of three non-catalytic domains. The catalytic domain of CNA is responsible for dephosphorylation of the substrates [1], whereas the non-catalytic domains may be involved in the regulation of the enzyme activity in the presence of CNB and/or calmodulin (CaM). Through mapping of CNA by limited proteolysis and structural analysis, it has been shown that * Corresponding author. Tel.: +86-10-62207365; fax: +86-10-62207365. E-mail address: [email protected] (Q. Wei). 0300-9084/$ - see front matter © 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.biochi.2004.10.009

the three non-catalytic domains are CNB-binding domain (BBH), CaM-binding domain (CBD) and autoinhibitory domain (AI) [5,6]. Furthermore, the increased phosphatase activity has been demonstrated with limited proteolytic cleavage of AI and CBD domains by using p-nitrophenylphosphate (p-NPP) as the substrate [6,7], though the role of BBH domain is not clear. In addition, similar effects have not been reported for other substrates such as the RII peptide (DLDVPIPGRFDRRVSVAAE), a peptide substrate for CN [8]. Up to now, there have been extensive studies looking into the effect of exogenous metal ions on the activity of calcineurin [1]. Li et al. [10] and Martin et al. [9] reported that the activating effect of Mg2+ was different from that of the transition metal ions Mn2+/Ni2+ with respect to Km, Vmax, pH optimum of the phosphatase reaction and their affinity for calcineurin; and it is thought that Mg2+ and Mn2+/Ni2+ may induce different conformational changes [4,9,11,12]. No fur-

216

P. Liu et al. / Biochimie 87 (2005) 215–221

Fig. 1. Schematic representations of CNA and its domain deletion mutants.

ther details have been elucidated and it is not even known which metal ion(s) (Mg2+ and Mn2+/Ni2+) interact with which domain(s). In our attempt to fully elucidate the roles of all noncatalytic CNA domains as well as the regulatory mechanism of Mn2+/Ni2+ and Mg2+, we constructed expression vectors of CNA and its domain deletion mutants: CNAa (catalytic domain), CNAac (catalytic domain and CBD domain) and CNAaci (catalytic domain, CBD and AI domains). The activities of CNA and its domain deletion mutants were systematically examined together with several regulators including CaM/CNB and metal ions by using p-nitrophenylphosphate (p-NPP) or 32P-labeled RII peptide as substrates, respectively. 2. Materials and methods 2.1. Bacterial strains, culture medium and vector The bacterial strains HMS174 (kDE3) and BL21 (kDE3), the vector pET-21a (+), and the rat cDNA template of CNAalpha are all conserved by our laboratory. p-NPP was obtained from Sigma Chemical Corp. RII peptide was purchased from BioMoL Research Labs Inc. PPO[OC(C6H5) =NCH=CC6H5] and POPOP{[OC(C6H5)=CHN=C]2C6H4} were obtained from E. Merck Corp. [c-32P]-ATP was obtained from Beijing Furi Biology Engineering Corp. cAMPdependent protein kinase catalytic subunit was purchased from Promega Chemical Corp. All other reagents were of standard laboratory grade and the highest quality available from commercial suppliers.

2.2. Design of truncated mutants and construction of the expression vectors The domain map is shown in Fig. 1. The primers of CNA and its truncated mutants were designed and are shown in Table 1. We constructed the expressing vectors pET21a/CNA, pET-21a/CNAa, pET-21a/CNAac and pET21a/CNAaci from the CNA-alpha cDNA template by using standard PCR and molecular cloning methods. 2.3. Expression and purification The constructed expression vectors were transformed into bacteria line-BL21 (kDE3), pre-cultured in LB media at 37 °C (containing 1‰ Amp of w/v) and then expressed in terrific media (also containing 1‰ Amp of w/v) at 25 °C over night, induced by 100 µM IPTG. For CNA and CNAaci, purification was carried out as previously described [13]. In brief, the harvested cells were resuspended in buffer A (20 mM Mops, 1 mM EGTA, 1% b-ME, 0.4 mM PMSF, pH 7.6), and then disrupted by sonication. Lysate was centrifuged and the supernatant was precipitated in 45% ammonium sulfate, and centrifuged again. The pellets were resuspended in buffer B (20 mM Mops, 1% b-ME, 0.4 mM PMSF, 0.5 mM CaCl2, pH 7.4) and were applied to a CaM-Sepharose 4B column. The desired proteins were eluted by using buffer C (10 mM Mops, 1 mM EGTA, 1% b-ME, 0.4 mM PMSF, pH 7.4). For CNAa and CNAac, purification was performed as described [14,15]. Again in brief, the harvested cells were resuspended in homogenizing buffer (50 mM Mops, 100 mM NaCl, 25 mM sucrose, 2 mM EDTA, 2 mM EGTA, 5%

Table 1 The designed primers and constructed expression vectors CNA CNAa CNAac

CNAaci

PCR primers Primer1: AGGAGATATACATATGTCCGAGCCCAAGGC Primer2:CGCGAAGCTTTCACTGAATATTGCTGC Primer1: GGAGATATACATATGTCCGAGCCCAAGGC Primer2: CGAAGCTTCACATGAAATTTGGGAGCC Primer1: AGGAGATATACATATGTCCGAGCCCAAGGC Primer2:CAGTTCATCATCTGACATGAAATTTGGGAG Primer3:CTCCCAAATTTCATGACTGATGATGAACTG Primer4:GGTGGTAAGCTTAACGCTCTCACTCTCTTCTCT Primer1: AGGAGATATACATATGTCCGAGCCCAAGGC Primer2: CAGTTCATCATCTGACATGAAATTTGGGAG Primer3: CTCCCAAATTTCATGACTGATGATGAACTG Primer4: CGAAGCTTCACATGAAATTTGGGAGCC

Expression vectors pET-21a/CNA pET-21a/CNAa pET-21a/CNAac

pET-21a/CNAaci

P. Liu et al. / Biochimie 87 (2005) 215–221

glycerol, 1% b-ME, 0.4 mM PMSF, pH 7.4), and disrupted by sonication. The lysate was then centrifuged and the supernatant was precipitated with 45% ammonium sulfate, and centrifuged again. The pellet was resuspended in buffer D (20 mM Mops, 20 mM NaCl, 2 mM EDTA, 5% glycerol, 1% b-ME, 0.4 mM PMSF, pH 7.4) and then desalted in a SephadexG-25 column. The fractions were applied to DEAE column and then gradient eluted with 2 M NaCl. The active fractions were collected and precipitated with 45% ammonium sulfate. The precipitate was resuspended and then applied to a Superdex-75 column, eluted with buffer D (50 mM NaCl and no glycerol). The fractions with activity were collected. 2.4. Activity assay The activity assay of enzyme was performed using a protocol [13–15] we have modified. In assaying enzyme activity using substrate p-nitrophenyl phosphate, the enzyme solution (20 µl, containing 0.150 mg/ml–1 16 mg/ml enzyme) was mixed with 180 µl assaying buffer (20 mM p-NPP, 50 mM Tris–HCl, 0.2 mg/ml BSA, 1 mM DTT and 1 mM Mn2+, pH 7.4; with or without 1 mM CaCl2, 2 µM CaM or 2 µM CNB) in 30 °C bath for 20 min. The reaction was terminated by adding 1.8 ml quenching solution (0.5 M Na2CO3, 20 mM EDTA) and the absorbance at 410 nm was measured; a control was done without the enzyme. The units (U) of p-nitrophenyl-phosphatase activity are defined as nanomoles of p-NPP hydrolyzed per milligram enzyme per minute at 30 °C. In assaying enzyme activity using substrate 32P-labeled RII peptide, the enzyme solution (10 µl, containing about 0.00077–0.0154 mg/ml enzyme) was mixed with 10 µl assaying buffer (40 µM 32P-labeled RII peptide, 100 mM Tris– HCl, 0.2 mg/ml BSA, 1 mM DTT and 1 mM Mn2+, pH 7.4; with or without 0.2 mM CaCl2, 0.6 µM CaM or 0.2 µM CNB) in 30 °C bath for 10 min. The reaction was terminated by adding 0.18 ml termination buffer (83.3 mM H3PO4). Finally, released 32P was separated from RII peptide and quantified with liquid scintillation spectrometry. The units (U) of protein phosphatase activity are defined as nanomoles of 32P of RII peptide released per milligram enzyme per minute. In assaying the effect of metal ions, the buffer [20 mM p-NPP, 50 mM Tris–HCl, 0.2 mg/ml BSA, 1 mM DTT, pH 7.4 (to Mn2+/Ni2+) and pH 8.6 (to Mg2+; according to the report [11], the pH of the maximum activation of CN by Mg2+ is around pH 8.5, we found the maximum activation of CNA’s truncated mutants by Mg2+ is about pH 8.6—data not shown)] and exogenous metal ions were used. In the case of CNA, 1 mM CaCl2, 2 µM CaM, 2 µM CNB were added; in the case of CNAaci, 1 mM CaCl2, 2 µM CaM were added; in the cases of CNAac and CNAa, CaCl2, CaM and CNB were all not added (our experiments demonstrated the activity of CNAac was not regulated by CaM). The exogenous metal ion concentrations were 10 mM for Mg2+, 5 mM for Ni2+ and 1 mM for Mn2+. The metal ion concentrations correspond to their maximum activating concentrations.

217

Various concentrations of 32P-labeled RII peptide or p-NPP were used in assaying the kinetic properties of CNA and its truncated mutants.

3. Results 3.1. Purification of CNA and its truncated mutants Using a rat cDNA library as template, the open reading frame (ORF) for the CNA was amplified by PCR, and PCR products were cloned into the pET-21a vector. A series of truncated mutants of CNA were also generated and verified by sequencing analyses (data not shown). The targeted proteins were purified using chromatography (see Section 2) and analyzed with SDS-PAGE (Fig. 2). 3.2. Effects of the three CNA domains In the absence of the activators-CaM and CNB, the effects of three functional domains (BBH, CBD and AI) on the enzyme activity are shown in Fig. 3. The activity orders are CNAa>>CNAac>CNAaci>CNA for both substrates, p-NPP and 32P-labeled RII peptide. For p-NPP, the activities of CNAa, CNAac and CNAaci are about 43-fold, 14-fold and sevenfold of that of CNA, respectively. For 32P-labeled RII peptide, the activity of CNAa is about 68-fold, CNAac about ninefold and CNAaci about 2.5-fold of that of CNA (Table 2). The relative ratio of CNAa/CNAac is about 3.06, CNAac/CNAaci 1.93, and CNAaci/CNA 7.34 for p-NPP; and for 32P-labeled RII peptide the relative ratio of CNAa/CNAac is about 7.90, CNAac/CNAaci 3.43, and CNAaci/CNA 2.52. 3.3. Regulation in the presence of CaM and CNB In the presence of CaM and/or CNB, enzyme activity for p-NPP increased with the sequential cleavage of functional domains (CNA
Fig. 2. SDS-PAGE analysis of purified targeted proteins. (a) CNAa; (b) CNAaci; (c) CNA; (d) CNAac.

218

P. Liu et al. / Biochimie 87 (2005) 215–221

Fig. 3. In the absence of CaM/CNB, the negative effects of three domains. (a) p-NPP as the enzyme substrate; (b) 32P-labeled RII peptide as the enzyme substrate. The assaying concentrations of enzymes are: CNAa 0.150 mg/ml, CNAac 0.530 mg/ml, CNAaci 0.810 mg/ml and CNA 1.16 mg/ml to p-NPP; CNAa 0.00149 mg/ml, CNAac 0.00684 mg/ml, CNAaci 0.0154 mg/ml and CNA 0.00077 mg/ml to 32P-labeled RII peptide (the same below). All values are the average of three tests.

tivities for 32P-labeled RII peptide also increased with the removal of the functional domains, but significantly the activity of CNA was the highest in the presence of both CNB and CaM. The activity of CNAa for both substrates was much higher than that of CNAac and CNAaci in the absence or Table 2 The relative activities of CNA and its splicing domain mutants in the absence of CaM/CNB

CNA CNAaci CNAac CNAa

p-NPP as the enzyme substrate 1.00 7.34 ± 0.186 14.19 ± 0.06 43.41 ± 2.22

32 P-labeled RII peptide as the enzyme substrate 1.00 2.52 ± 0.49 8.65 ± 0.34 68.32 ± 1.76

All activity values are the average values of three tests.

Fig. 4. The effects of CNB and CaM on the enzyme activity of CNA and its splicing domain mutants. (a) p-NPP as the enzyme substrate; (b) 32P-labeled RII peptide as the enzyme substrate. All values are average of three test values.

presence of CaM/CNB, especially for 32P-labeled RII peptide (Fig. 4). For substrate 32P-labeled RII peptide, as shown in Table 3 there is 2.35-fold activation of CNA by CaM and 45.35-fold activation by CNB. When CNB and CaM are both present, the activation on CNA is about 154-fold. CNA activation by CaM, CNB or CaM/CNB for 32Plabeled RII peptide is all higher than that for p-NPP (Table 3). In comparison, activation of CNAaci by CaM for 32P-labeled RII peptide is also higher than that for p-NPP. In addition, CaM can activate CNAaci but not CNAac for both substrate p-NPP and 32P-labeled RII peptide. 3.4. Enzyme kinetics As shown in Table 4, the Km value of CNAa is the largest with 32P-labeled RII peptide as a substrate, followed by CNAac, CNAaci and CNA. In addition, the Km values of truncated mutants for 32P-labeled RII peptide are all larger than those for p-NPP (data not shown).

P. Liu et al. / Biochimie 87 (2005) 215–221

219

Table 3 Activating extents of CaM/CNB on CNA and its splicing domain mutants

0 CaM CNB CaM + CNB

CNA 1.00 2.20 ± 0.11 1.75 ± 0.02 3.98 ± 0.10

p-NPP as the enzyme substrate CNAaci CNAac 1.00 1.00 1.41 ± 0.07 0.99 ± 0.03 1.03 ± 0.04 1.01 ± 0.01 1.42 ± 0.04 0.98 ± 0.00

CNAa 1.00 0.99 ± 0.03 0.99 ± 0.04 0.97 ± 0.01

32 P-labeled RII peptide as the enzyme substrate CNA CNAaci CNAac CNAa 1.00 1.00 1.00 1.00 2.35 ± 0.09 2.33 ± 0.07 0.86 ± 0.00 0.90 ± 0.04 45.35 ± 5.86 1.00 ± 0.00 1.07 ± 0.03 1.01 ± 0.06 153.84 ± 7.94 2.37 ± 0.22 1.10 ± 0.03 0.97 ± 0.05

The activity of CNA, CNAaci, CNAac and CNAa in the absence of CaM/CNB as one unit, and the activating extents are the ratios of the activities of CNA in the presence of CaM, CNB or CaM/CNB with that in the absence of CaM, CNB or CaM/CNB. All activity values are the average values of three tests. Table 4 The enzyme kinetics analysis of CNA and its splicing domain mutants (32P-labeled RII peptide as substrate) Km (µM) Vmax (U/mg) Kcat (s–1) Kcat/Km (s–1 M–1)

CNA 31.9 ± 1.0 1250 ± 0.0 1.27 ± 0.0 39812 ± 1124

CNAaci 146.6 ± 54.8 208.3 ± 66.5 0.20 ± 0.07 1364 ± 51

CNAac 265.3 ± 33.5 416.7 ± 67.0 0.31 ± 0.05 1168 ± 44

CNAa 499.5 ± 97.8 5000 ± 833.5 3.58 ± 0.60 7167 ± 252

The different assaying concentrations of substrate 32P-labeled RII peptide are: 2.5, 5, 10, 20, 40, and 80 µM; assaying CNA and CNAaci in the presence of CaM and CNB, assaying CNAac and CNAa in the absence of CaM and CNB.

The Vmax/Kcat value of CNAa is much higher than that of CNAac and is the largest one among all constructs, with the order being CNAa>CNA>CNAac>CNAaci. Furthermore, the Vmax and Kcat values of CNA are also much higher than those of CNAaci.

The activation effect of Mg2+ on the enzyme is almost unchanged from CNA to CNAaci, but abruptly decreased from CNAaci to CNAac, followed by a dramatic increase from CNAac to CNAa (Fig. 5).

3.5. Effect of metal ions

4. Discussion

When using p-NPP as the substrate, the effect of metal ions on CNA is Mn2+CNA>CNAaci>CNAac. The activation level of Mn2+/Ni2+ on CNAa is much higher than that on CNAac.

It has been known that the regulatory fragment of CNA mediates the activations of CNA by CNB and CaM [5,6], in which Klee et al. have identified that there are three functional domains (BBH, CBD and AI domains) in the regulatory fragment. Other reports [12,16] have shown that the AI domain is critical in regulating enzyme activity, and there is activation synergism between AI and CBD domains in vivo [17]. In the absence of CaM/CNB, our results clearly show that the enzyme activities increase with the sequential deletion of three domains and the CNAa activity is the highest, as assayed against both substrates p-NPP and 32P-labeled RII peptide (Fig. 3). This evidence indicates that the dephosphorylation activity of the CNA catalytic domain itself is very high, similar to PP1 which is a kind of Ser/Thr protein phosphatase and its structure is very similar to the catalytic domain of CNA [18]. But the three functional domains, individually or together, act as negative regulators of the enzyme and result in a significant decrease of the phosphatase activity. Elucidated by the activity comparison (Table 2), the negative regulation of regulatory fragment is different for different substrates. When 32P-labeled RII peptide is used as the substrate, CBD is the most effective intra-molecular suppressor, followed by AI domain and BBH domain. But for substrate p-NPP, the strongest intramolecular suppressor is BBH domain, then CBD domain and AI domain. As analyzed in crystal structure studies [19,20], these inhibitory domains like a coat covering on the active site of enzyme, though positioned differently in CNA, can

Fig. 5. Effects of metal ions on the activity of CNA and its splicing domain mutants. Metal ion is added into assaying buffer; and the substrate is p-NPP.

220

P. Liu et al. / Biochimie 87 (2005) 215–221

exert direct influence on the catalytic core by either restricting its conformational adaptability to accommodate substrate or partially interfere with substrate access. In the presence of CaM and/or CNB (Fig. 4), the enzyme activities also increase with the removal of three functional domains, further confirming that these domains are negative regulators with or without CaM and/or CNB. The activity of CNAac is much lower than that of CNAa, and Vmax and Kcat of CNAac are all much lower than those of CNAa. Therefore, CBD is in this case again the most important domain in suppressing the enzyme activity. CNA activation by CNB is much higher than that by CaM for 32P-labeled RII peptide, and the activation by CaM/CNB is also much higher than that by CNB only (Table 3). The results that CaM could activate CNAaci but not CNAac demonstrate that CBD domain alone cannot mediate the activation of CaM. This an interesting observation in light of the CBD name—calmodulin binding domain—which tends to give rise to the impression that this is the only structural element through which CaM activates CNA. Obviously, our data now indicate that AI domain, in concert with CBD, must also play an indispensable role in mediating the activation by CaM. However, this is also not completely surprising. In the absence of CN–CaM complex structure, we do not current have detailed knowledge with respect to the interaction between the phosphatase and CaM. In the present study, activation levels by CaM/CNB on enzyme activities of various constructs are different for different substrates (Fig. 4). It is reasonable to envisage that the access and subsequent interaction of substrate with the catalytic core are different for different substrates with varying size and structure. For example, the large size of the peptide substrate would almost certainly have different interaction with the phosphatase than the smaller substrate, p-NPP. The effects of exogenous metal ions on the enzyme activity of calcineurin have been investigated for many years. These studies include enzyme activity, the change of enzyme structure, dynamic analysis, energetic and spectral analysis [4,10–12,21,22]. But it has not been established through which domain or domains Mg2+ and Mn2+/Ni2+ exert their regulation effects. Evident from our results, in the presence of Mn2+/Ni2+, the enzyme activity increases upon sequential deletion of the three non-catalytic domains. The highest activation by Mn2+/Ni2+ is observed for CNAa, suggesting that the Mn2+/Ni2+ activation is primarily mediated through the catalytic core. Upon gradual removal of the inhibitory domains, the effects of Mn2+/Ni2+ activation become stronger. Comparing the activity of CNA with that of CNAaci/ CNAac/CNAa in the presence of Mg2+ (Fig. 5), the activity of CNA and CNAa are higher than that of CNAac. In the presence of Mg2+, removal of BBH domain results in little effect on enzyme activity but further deletion of AI domain leads to sharp decrease of the enzyme activity. This activity is sharply increased upon deletion of all three domains. All these results demonstrate that the activation of Mg2+ is not

only mediated through the catalytic domain, but also is closely related to the non-catalytic domains of CNA, in particular to CBD and AI domains. In summary, we have shown unequivocally that the noncatalytic domains of CNA act as inhibitors. They negatively regulate the phosphatase activity of CN, presumably through restricting the conformational adaptability and/or hindering substrate access. Calcineurin has a diverse role in a variety of cellular processes and its activity is tightly controlled by numerous factors such as calmodulin, inhibitors, subunit B and metal ions etc. Our experiment evidence now adds yet another set of intra-molecular regulators—the non-catalytic domains of CNA. Furthermore, the domains through which metal ions regulate the phosphatase activity are elucidated. Acknowledgments This work was supported in part by grants from the National Nature Science Foundation of China; The Research Fund for Doctoral Program of Higher Education; and The National Important Basic Research Project. References [1]

F. Rusnak, P. Mertz, Calcineurin: form and function, Physiol. Rev. 80 (2000) 1483–1521. [2] C.B. Klee, H. Ren, X. Wang, Regulation of the calmodulin-stimulated protein phosphatase, calcineurin, J. Biol. Chem. 273 (1998) 13367– 13370. [3] H. Matsui, C.J. Pallen, A.M. Adachi, J.H. Wang, P.H. Lam, Demonstration of different metal ion-induced calcineurin conformations using a monoclonal antibody, J. Biol. Chem. 260 (1985) 4174–4179. [4] C.J. Pallen, J.H. Wang, Stoichiometry and dynamic interactions of metal ion activators with calcineurin phosphatase, J. Biol. Chem. 261 (1986) 16115–16120. [5] B.A. Perrino, Y. Ng, R.S. Lilly, Thomas, Calcium regulation of calcineurin phosphatase activity by its B subunit and calmodulin, J. Biol. Chem. 270 (1995) 340–346. [6] M.J. Hubbard, C.B. Klee, Functional domain structure of calcineurin A: mapping by limited proteolysis, Biochemistry 28 (1989) 1868– 1874. [7] Q. Wei, E.Y.C. Lee, Mutagenesis of the L7 loop connecting beta strands 12 and 13 of calcineurin: evidence for a structural role in activity changes, Biochemistry 36 (1997) 7418–7424. [8] D.K. Blumenthal, K. Takio, R.S. Hansen, E.G. Krebs, Dephosphorylation of cAMP-dependent protein kinase regulatory subunit (Type II) by calmodulin-dependent protein phosphatase, J. Biol. Chem. 261 (1986) 8140–8145. [9] B.L. Martin, L.A. Jurado, A.C. Hengge, Comparison of the reaction progress of calcineurin with Mn2+ and Mg2+, Biochemistry 38 (1999) 3386–3392. [10] H.-C. Li, W.W.S. Chan, Activation of brain calcineurin towards proteins containing thr(P) and ser(P) by Ca2+, calmodulin, Mg2+ and transition metal ions, Eur. J. Biochem. 144 (1984) 447–452. [11] H.-C. Li, Activation of brain calcineurin phosphatase towards nonprotein phosphoesters by Ca2+, calmodulin, and Mg2+, J. Biol. Chem. 259 (1984) 8801–8807. [12] A.B. Carol, J.S. Douglas, W.S. Clark, L.C. Darryl, L.L. Karen, Effect on protein phosphatase activity of peptide backbone modification and truncation of the autoinhibitory domain peptide of calcineurin, Int. J. Peptide Res. 47 (1996) 98–102.

P. Liu et al. / Biochimie 87 (2005) 215–221 [13] Q. Wei, E.Y.C. Lee, Expression and reconstitution of calcineurin A and B subunits, Biochem. Mol. Bio. Inter. 41 (1997) 169–177. [14] S.J. Yang, L. Zhang, Q. Wei, Activities and properties of calcineurin catalytic domain, Chin. Sci. Bull. 45 (2000) 1394–1398. [15] B.Q. Xiang, P. Liu, G.H. Jiang, K. Zou, F. Yi, S.J. Yang, Q. Wei, The catalytically active domain in the A subunit of calcineurin, Biol. Chem. 384 (2003) 1429–1434. [16] N. Yokoyama, J.H. Wang, The role of the autoinhibitory domain in differential metal ion activation of calmodulin-stimulated phosphatase, FEBS Lett. 337 (1994) 128–130. [17] K. Tokoyoda, Y. Takemoto, T. Nakayama, T. Arai, M. Kubo, Synergism between the calmodulin-binding and autoinhibitory domains on calcineurin is essential for the induction of their phosphatase activity, J. Biol. Chem. 27 (2000) 11728–11734.

221

[18] Z. Zhang, G. Bai, S. Deans-Zirattu, M.F. Browner, E.Y.C. Lee, Expression of the catalytic subunit of phosphorylase phosphatase (protein phosphatase-1) in Escherichia coli, J. Biol. Chem. 267 (1992) 1484–1490. [19] C.R. Kissinger, H.E. Parge, D.R. Knighton, C.T. Lewis, L.A. Pelletier, A. Tempczyk, et al., Crystal structure of human calcineurin and the human FKBP12–FK506–calcineurin complex, Nature 378 (1995) 641–644. [20] J.P. Griffith, J.L. Kim, E.E. Kim, M.D. Sintchak, J.A. Thomson, M.J. Fitzgibbon, et al., X-ray structure of calcineurin inhibited by the immunophilin-immunosuppressant FKBP12–FK506 complex, Cell 82 (1995) 507–522. [21] C.J. Pallen, J.H. Wang, Regulation of calcineurin by metal ions, J. Biol. Chem. 259 (1984) 6134–6141. [22] B.L. Martin, B. Li, C.X. Liao, D.J. Rhode, Differences between Mg2+ and transition metal ions in the activation of calcineurin, Arch. Biochem. Biophys. 380 (2000) 71–77.