Mastoparan X altered binding behavior of La3+ to calmodulin in ternary complexes

Available online at www.sciencedirect.com JOURNAL OF

Inorganic Biochemistry Journal of Inorganic Biochemistry 102 (2008) 278–284 www.elsevier.com/locate/jinorgbio

Mastoparan/Mastoparan X altered binding behavior of La3+ to calmodulin in ternary complexes Qin Yang 1, Jian Hu 1, Xiaoda Yang *, Kui Wang Department of Chemical Biology and State Key Laboratories of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing 100083, PR China Received 27 March 2007; received in revised form 19 July 2007; accepted 30 August 2007 Available online 7 September 2007

Abstract Ca2+ binds to calmodulin (CaM) and triggers the interaction of CaM with its target proteins; CaM binding proteins (CaMBPs) can also regulate the metal binding to CaM. In the present paper, La3+ binding to CaM was studied in the presence of the CaM binding peptides, Mastoparan (Mas) and Mas X, using ultrafiltration and titration of fluorescence. Ca2+ binding was used as an analog to understand La3+ binding in intact CaM and isolated N/C-terminal CaM domain of metal-CaM binary system and metal-CaM–CaMBPs ternary system. Mas/Mas X increased binding affinity of La3+ to CaM by 0.5  3 orders magnitude. The metal ions binding affinity to the C-terminal or the N-terminal CaM domain suggested that in the first phase of binding process both Ca2+ and La3+ bind to C-terminal of CaM in the presence of Mas/Mas X. In the presence of CaM binding peptides, La3+ binding preference was substantially altered from the metal-CaM binary system where La3+ slightly preferred binding to the N-terminal sites of CaM. Our results will be helpful in understanding La3+ interactions with CaM in the biological systems.  2007 Elsevier Inc. All rights reserved. Keywords: Calmodulin; Lanthanum; CaM binding peptide; Mastoparan

1. Introduction Calmodulin (CaM) is a ubiquitous Ca2+ binding protein present in most eukaryotic cells. Its function is ‘‘sensing’’ the change of intracellular Ca2+ level and thus Ca2+-mediated signal transduction. The versatile CaM physiological functions result from the interaction of Ca–CaM with binding proteins (CaMBPs) [1,2]. Ca–CaM binding to target proteins lead to the formation of Ca–CaM–CaMBPs ternary complexes, which modulate protein phosphorylation/dephosphorylation and regulate Ca2+ homeostasis and cell mobility. Ca2+ ions bind to the calcium binding sites in C- and Nterminal CaM domain in a stepwise manner [3]. The first

*

1

Corresponding author. Tel.: +86 10 82801539; fax: +86 10 62015584. E-mail address: [email protected] (X. Yang). Contributed equally to the current study.

0162-0134/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2007.08.011

two Ca2+ ions bind to the C-terminal CaM domain and induce conformational change leading to the opening of the local hydrophobic pocket favoring target protein binding. The binding of another two Ca2+ to the N-terminal CaM domain result in further conformational change enabling complete target protein binding to CaM. Some CaM-dependent enzymes are activated only at this Ca2+saturated stage [4]. While Ca2+ binding to CaM facilitate the CaM–CaMBP binding, it was observed in the ternary complex that the existence of CaMBPs to CaM increased the Ca2+ to CaM binding affinity. In the Ca–CaM–CaMBP system, the binding affinity of Ca2+ increased a strong positive cooperation between the two global domains making CaM more sensitive to the change of intracellular Ca2+ concentration [4]. Many other metal ions, including La3+, Al3+, Pb2+ and Cd2+, can bind to CaM [5] and cause conformation and CaM function changes. These changes contribute to the biological effects [6–8] and metal toxicity [9–14]. Among

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

the metal ions binding to CaM, La3+ has been considered as an analog for Ca2+ due to similar ionic radius and coordination chemistry of hard acid metals [15,16]. La3+ binds to CaM with a high affinity [17]. Previous studies have shown that La3+ share the same binding sites with Ca2+ on CaM. Moreover, Ca4CaM, La4CaM and La2Ca2CaM bind to Mas or calcineurin (CaN) with similar affinity [6,7]. The metal ion dissociation from La3+-content ternary complexes was much slower than Ca2+–CaM–Mas complexes [6,17]; but La3+-content CaM complexes exhibit significantly reduced CaN activation [7]. In addition to the biological interest, the CaM metal binding properties have also made it an attractive target for protein engineering [18–21]. How CaMBPs regulates the metal binding of CaM remains a question. CaMBPs may increase the metal binding by transferring the protein binding energy to the overall metal sites, or CaMBPs could exert a preference for CaM metal sites. The CaM inside cells should respond to metal ions in the presence of a variety of CaMBPs. Different CaM response could result in different cellular signaling. If CaMBPs result in different metal binding among CaM metal sites, then a new strategy utilizing protein–protein interactions would be suggested for the design of metalbinding proteins. In this study, the La3+ binding properties to CaM were compared to Ca2+ in the presence of CaMBPs, Mas and Mas X. Mas and Mas X differ in La3+ binding properties to CaM in the ternary complexes from metal-CaM binary system. 2. Materials and methods 2.1. Chemicals Polistes Mastoparan and Mastoparan X were purchased from Sigma, Inc. Lanthanum chloride solution was prepared by dissolving La2O3 (99.9% in purity) in hydrochloric acid. The solution was evaporated to remove the excess HCl, redissolved in 1 mM HCl and the LaCl3 solution was diluted to 50 mM. Other chemicals used were analytical grade. All the solutions used, except for the LaCl3 and CaCl2 solution, were passed through Chelex 100 to remove traces of polyvalent metal ions. 2.2. Proteins The expression and preparation of ApoCaM were conducted as previously described [6]. The CaM N/C-terminal domain was prepared through a site-directed mutagenesis using the QuickChange sitedirected mutagenesis kit using manufacturer’s protocols. For preparation of N-terminal domain of CaM (NCaM), a stop codon was introduced in the middle of the gene at the 82th amino acid position using a pair of primers with the sequence of 5’-CAAGAAAAATGAAAGATATCAA-

279

TAGCGAGAAG-3’ (the underline shows the recognition sequence of EcoR V introduced at the same time, which facilitated selection of positive clone). For preparation of C-terminal domain of CaM (CCaM), two previously phosphorylated PCR primers with the sequence of 5’GATCGTCGACATGAAAGATACAGATAGCGAAGAAG-3’ and 5’-GATCGTCGACGGGTATATCTCCTTCTTAAAG-3’, respectively, were used to prepare the open ring vector with deletion of N-terminal domain of CaM (1–75) from the full length vector encoding CaM gene by the PCR amplification using the TaKaRa LA PCR kit according to the manufacturer’s procedures. The PCR product was ligated by T4 DNA ligase and transformed to JM109. After the colonies were selected, the vectors (for expression of both NCaM and CCaM) were verified by DNA sequencing analysis. The expression and purification of NCaM and CCaM were conducted by using the same protocol as the CaM purification. The NCaM and CCaM contain 81 and 73 amino acids, respectively. The purity of the proteins was determined by electrophoresis and found to be as homogenous as the full length CaM. 2.3. Ultrafiltration titration At room temperature, 5 lM ApoCaM was incubated for 4 h in a pH 7.0 buffered (20 mM HEPES, 100 mM KCl) solution containing La3+ of different concentrations (from 0 to 40 lM), and 5 mM malic acid as competitive ligand. The solution was filtrated in an ultrafiltration tube (Ultra-4, cutoff M.W. = 5000, Millipore). The total concentration of unbounded La3+ in filtrate (CLa) was measured by previously described spectrophotometric method [22]. The free La3+ concentration in the filtrate was calculated using a MatLab program by inputting the following equations: C La ¼ ½La3þ  þ ½LaL þ ½LaL2  T L ¼ ½L þ ½LaL þ 2  ½LaL2  ½LaL ¼ K 1  ½L  ½La3þ  2

½LaL2  ¼ K 1  K 2  ½L  ½La3þ  where L is the competitive ligand, malate and dissociation constants of La3+-malate complexes are lgK1 = 4.37 and lgK2 = 2.79 [23]. The average La3+ binding number (n) on CaM was calculated with the following equation: n¼

T La  C La C CaM

where TLa and CCaM are the total La3+ and CaM concentration, respectively. The La3+-CaM complex dissociation constant was calculated by plotting n against free La3+ concentration and fitting the data to a Hill model using a Microcal Origin program.

280

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

2.4. Fluorescent titration For metal-CaM binary system, 5 lM ApoCaM was titrated by La3+ or Ca2+ in a pH 7.0 reaction buffer containing 100 mM KCl and 20 mM HEPES, in the presence of competitive ligands (5 mM malic acid for La3+; 1 mM EGTA for Ca2+). The intensity of intrinsic fluorescence of CaM (kex/em = 273/307 nm) was measured. For metalCaM–Mas/Mas X ternary system, 2 lM ApoCaM or NCaM/CCaM was incubated with equal molar of Mas or Mas X in the reaction buffer containing appropriate competitive ligands. Then, La3+ or Ca2+ aliquots were added gradually and the fluorescent spectrum of the exclusive tryptophan residue of the CaM binding peptides was recorded (310 nm to 400 nm with excitation wavelength at 290 nm and a slit width of 3 nm was set to reduce the interference from CaM intrinsic fluorescence). The calculation and data processing were done by the method similar to that used in the ultrafiltration titration. 3. Results 3.1. Binding of La3+ to CaM Two complementary methods, ultrafiltration and titration of fluorescence, were used to measure the La3+ binding affinity to CaM expressed as the dissociation constant Kd. The curve fitting results with the two methods (Fig. 1) were similar: ultrafiltration titration, Kd = 12.3 ± 0.7 nM, Hill coefficient n = 3.6 with maximal binding number of 4.5 ± 0.3; fluorescence titration, Kd = 7.5 ± 0.1 nM with n = 3.2. For comparison, the Kd for Ca2+ to CaM complexes was 7.24 · 107 M with a Hill coefficient of 1.7 (Supplementary material). 3.2. Binding of metal ions to intact CaM in metal-CaM– Mas/Mas X ternary systems Upon Ca2+ binding, the exposure of the CaM hydrophobic pocket facilitated specific binding to the target peptides, such as Mas and Mas X. In the metal-CaM-peptide ternary complex, the exclusive Trp residue on the CaM binding peptide was embedded in a hydrophobic environment [24,25]. As described previously [6,26], the change of fluorescence emission at 325 nm could be used for monitoring the metal binding process. As shown in Fig. 2a,b, La3+ induced similar fluorescence spectra changes to those induced by Ca2+ (Supplementary material), suggesting the La3+ binding to CaM could also promote CaM binding to its target peptides. During the Ca2+ titration of CaM, the intensity of fluorescence at 325 nm also exhibited a threestate transition [26] (Fig. 3). The La3+ binding curve was similar to Ca2+ (Fig. 3b) in the presence of Mas X; however, some differences occurred in the presence of Mas (Fig. 3a). The metal ion dissociation constants to CaM-Mas/Mas X ternary system were determined based on the fluorescent

Fig. 1. Binding functions versus the free La3+ concentration (malic acid as the competitive ligand). (a) Curve for La3+ binding to CaM obtained by ultrafiltration titration. (b) Curve for La3+ binding to CaM obtained by titration of fluorescence. All data were average of triplicate measurements.

titrations, as described in Section 2 (Fig. 4). The constants corresponding to the increasing phase (i) and declining phase (d) were calculated, Table 1. For comparison, the dissociation constants for Ca2+ were also determined (Supplementary material). Results were in good agreement with previous report [26]. In the presence of Mas/Mas X, there was a less than 10-fold for Ca2+ and 10  1000-fold for La3+ increase in metal binding affinity. 3.3. Bindings of metal ions to the NCaM/CCaM in the ternary systems To clarify the metal ion binding preference to CaM in ternary systems, the CaM N-terminal domain (NCaM) and the CaM C-terminal domain (CCaM) were expressed and purified to homogeneity. As shown in Fig. 5, the NCaM to Mas/Mas X binding in the presence of Ca2+ could induce slight blue shift of Trp spectra and increase in the intensity of fluorescence; while CCaM to Mas and Mas X binding in the presence of Ca2+ induced blue spectral shifts and large increase of fluorescence at 325 nm.

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

Fig. 2. The fluorescence spectra of the CaMBPs with La3+ binding to CaM in the titration process. (a) The Mas spectra with La3+ titration. (b) The Mas X spectra with La3+ titration. The arrow indicated the increase in the metal ions molar ratio to CaM from zero to 4 in steps of 0.5.

The Ca2+/La3+ to NCaM/CCaM dissociation constants in the metal-CaM- Mas/Mas X ternary systems were determined. The binding curves were shown in Fig. 6. The calculated dissociation constants have been summarized in Table 1. The Ca2+ binding curves were shown in Supplementary material. The La3+ to CCaM dissociation constants in the presence of Mas X or La3+ to NCaM and Mas were not obtainable due to the limited fluorescent change during La3+ titration. It is noteworthy that the Ca2+/La3+ to CCaM dissociation constants were close to the metal affinities in the first binding phase (fluorescence increasing phase) of the full length CaM in the ternary systems while the Ca2+/La3+ to NCaM dissociation constants agreed with the second phase of the metal binding (fluorescence declining phase).

281

Fig. 3. Fluorescence intensity change at 325 nm of the CaMBPs upon binding to CaM in the titration process by metal ions (La3+/Ca2+). (a) Mas. (b) Mas X. All data were the average of triplicate measurements.

4. Discussion

Fig. 4. CaMBPs fluorescence at 325 nm versus free La3+ concentration. The free La3+ concentration was calculated using a MatLab program based on the data shown in Fig. 3. For comparison, the intrinsic fluorescence intensity change at 307 nm of CaM was also plotted (the solid diamond).

The mechanism of protein/peptide binding in the regulation of CaM metal binding is important. In this study, we investigated La3+ binding to CaM in the presence of two

CaM binding peptides by comparing with Ca2+ binding in both binary and ternary systems.

282

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

Table 1 The dissociation constants of metal ions in the metal-calmodulin proteins (CaM/NCaM/CCaM)-binding peptide (Mas/Mas X) systems Kd (M) Ca2+ CaM CaM–MasX NCaM–MasX CCaM–MasX CaM–Mas NCaM–Mas CCaM–Mas a b c

La3+ 7

a

(7.24 ± 0.15) · 10 (C-terminal domain) (2.95 ± 0.07) · 108(i)b, (2.59 ± 0.03) · 107(d)b (1.67 ± 0.01) · 107 (5.49 ± 0.13) · 108 (5.19 ± 0.11) · 108(i) (1.98 ± 0.05) · 107(d) (2.20 ± 0.02) · 107 (2.19 ± 0.02) · 107

(1.20 ± 0.01) · 108 (1.70 ± 0.09) · 1011(i), (3.55 ± 0.04) · 109(d) (3.13 ± 0.03) · 109 n.o.c (2.69 ± 0.15) · 109(i), (1.43 ± 0.03) · 108(d) n.o. (2.20 ± 0.10) · 109

The Kd for Ca2+ binding to the CaM N-terminal domain sites is 8.13 · 106 M [3]. i, the increasing (first) phase; d, the declining (second) phase. n.o.: Not obtained.

Fig. 5. The CaMBPs fluorescence spectra with binding to NCaM/CCaM in presence of metal ions (4 lM Ca2+/4 lM La3+). (a), (b): Mas or Mas X binding to NCaM. (c), (d): Mas or Mas X binding to CCaM.

4.1. The La3+ to CaM binding affinity in binary complexes The lanthanide ion binding affinity to several calciumbinding proteins has been reported to be in the lM range [18–20,27]; however, there was no competitive ligand included in these studies. Critical to the following discussion, the La3+ to CaM binding parameters were determined with two methods

(Fig. 1) using malate as the competitive ligand. The Kd (1.20 · 108 M) was 2-fold less than the previous studies. This result was consistent with the hydrolysis La3+ ions at neutral solution that would result in reduced free La3+ ions concentration. The La3+ binding affinity was much higher than Ca2+, which was consistent with previous report [17]. The Hill coefficient (n = 3.5  3.8) suggested a strong cooperation

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

Fig. 6. The CaMBPs fluorescence intensity change at 325 nm with binding to NCaM/CCaM in the La3+ titration process. (a) Mas X binding to NCaM in the La3+ titration process. (b) Mas binding to CCaM with La3+ titration. For comparison, the curve for peptide binding to full length CaM was also plotted. All data were the average of triplicate measurements.

among all four binding sites. This pattern of cooperation was quite different from that of Ca2+ [25,26], in which the two sites at the C-terminal bind Ca2+ with a Kd of 106 M, but the N-terminal domain sites exhibited a Kd of 105 M. 4.2. Mas/Mas X altered the La3+ to CaM binding properties in the ternary systems Based on the previous [6,7,26] and current results (summarized in Table 1), Mas/Mas X altered the La3+ to CaM binding properties in the ternary systems. La3+-CaM-CaMBP complexes formation is a three-state transition process similar to that of Ca2+ complexes. It is proposed the CaM binding of Mas X to N/C-terminal domain [26] that the first fluorescence increasing phase corresponds to the Mas X binding on the C-terminal CaM domain with Ca2+ binding on the same domain. The second fluorescence decreasing phase corresponds to the

283

Mas X binding on both the C-terminal and N-terminal domains with Ca2+ binding on all the four binding sites. The titration curves (Figs. 3 and 4) demonstrate that La3+ share the same sequence of binding. However, in the presence of Mas, the Ca2+ to NCaM dissociation constants and CCaM domain were both close to the second phase. We postulate that only in the context of full-length CaM could Mas induce conformational changes of the C-terminal domain in order to increase Ca2+ affinity from 106 M in the binary system to 4 · 108 M in the ternary system, because Mas binds to CaM with much lower affinity than Mas X and could interact with both Ca4CaM globular domains [24]. These observations agree with the the previous postulation that protein structures/interactions are key to determining metal calcium binding protein function [28]. Mas/Mas X caused the La3+ increase in CaM binding affinity. Compared with the binary complexes (Kd  108), the La3+ binding affinity increased 10  1000-fold in the ternary system (Fig. 4, Table 1). Importantly, the binding affinity increase was observed with both La3+ and Ca2+. The increase in Ca2+ binding affinity to CaM in Ca2+-content ternary complexes was previously explained in terms of the smaller change of DG upon peptide binding to the opened hydrophobic pocket in Ca2+-bound CaM because of the reduced hydrophobic surface exposed to polarized solvent [29,30]. Studies of La3+ binding to an EF-hand protein motif [31] and Eu3+ binding to CaM [27] suggested that La3+ ions would have coordination water due to high coordination numbers. Therefore, it is plausible that this peptide binding resulted in more significant increase of binding affinity for La3+ than Ca2+. Additionally, the La3+ high affinity may strengthen the previous postulation that CaM is a major La3+ target in cells. Mas/Mas X changed the La3+ preference for CaM binding sites. In binary complexes, the La3+ preferred binding was at the N-terminal site; but this priority over the C-terminal domain was small [6]. Comparing the dissociation constants for the two phases of La3+ to NCaM/CCaM (Table 1), in the formation of ternary complexes, La3+ could bind in the first phase to the CaM C-terminal sites together with the peptide binding, then bind to the N-terminal domain and fulfill the metal-CaM–CaMBP complexes formation. Now the La3+ binding preference became the same as that of Ca2+ as described above. The possible reason for Mas/Mas X facilitated both Ca2+ and La3+ ions binding to C-terminal domain of CaM might be due to the strong interaction of the CaMBPs with the CaM C-terminal domain. It has been reported that the Mas/Mas X affinity to the CaM C-terminal domain was several orders of magnitude higher than to the N-terminal counterpart [26]. In the binary complex formation, no strong preference was observed in La3+ binding between the binding sites of the two global domains, but in the ternary system, the action of Mas/Mas X in La3+ binding to the C-terminal domain was then enhanced preferentially. A similar mechanism might explain a higher

284

Q. Yang et al. / Journal of Inorganic Biochemistry 102 (2008) 278–284

preference for Ca2+ binding to the CaM N-terminal domain in the presence of a peptide from the gating domain of small conductive K+ channel [32]. Therefore, we speculated that the La3+ binding mode would be more dependent on the presence of CaMBPs. The variety of CaMBPs may suggest different La3+ binding mode from that of Ca2+ in the biological systems. Compared with Ca2+, it could be seen that Mas and Mas X increased overall Ca2+ affinity to both domains, but the enhancement for the N-terminal was slightly larger. The two peptides exhibited different effects on La3+ binding. Mas increased La3+ binding to C-terminal domain by 4-fold. Strikingly, Mas X increased La3+ affinity 700-fold to the C-terminal domain in contrast to a 3fold enhancement to the N-terminal domain. These results clearly indicate that the CaM protein interaction with its binding protein/peptides could alter the CaM site specificity to metal ions, suggesting that protein interactions might be utilized for metal recognizing in addition to binding site modulation in metal protein engineering. In summary, the present study showed that CaM intereaction with CaMBPs Mas and Mas X could influence the La3+ to CaM binding in two ways: 3 orders of magnitude of increase in binding affinity and the alteration of the binding site preference. Protein interactions could play significant role in regulating the metal binding affinity in a metal-protein–protein ternary complex. Previously it was recognized that the metal binding would influence the sequential protein–protein interaction [7]. Our studies suggested that different La3+ binding mode from Ca2+ in the biological systems should be expected due to the variety of CaMBPs. The alteration of metal binding could be achieved by protein interaction. Our results would be helpful for understanding the interactions of La3+ with CaM in the presence of CaMBPs in the biological systems as well as provide useful information for the design of metal binding proteins based on CaM. 5. Abbreviations CaM calmodulin Mas mastoparan Mas X mastoparan X CaMBPs CaM binding proteins/peptides NCaM N-terminal domain of CaM CCaM C-terminal domain of CaM EGTA ethylene glycol bis(b-aminoethyl ether)-N, N, N, N 0 -tetraacetic acid HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid Acknowledgements This work is supported by National Natural Science Foundation of China (20331020 & 20637010) and 985 program. We thank Prof. John J. Hefferren for editing the manuscript.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jinorgbio. 2007.08.011. References [1] D. Chin, A.R. Means, Trends Cell Biol. 10 (2000) 322–328. [2] W.Y. Cheung, Science 207 (1980) 19–27. [3] J.A. Cox, M. Comte, A. Mamar-Bachi, M. Milos, J.-J. Schaer, Cation Binding to Calmodulin and Relation to Function, Springer-Verlag, Berlin Heidelberg New York, 1988. [4] O.B. Peersen, T.S. Madsen, J.J. Falke, Protein Sci. 6 (1997) 794–807. [5] J.S. Mills, J.D. Johnson, J. Biol. Chem. 260 (1985) 15100–15105. [6] J. Hu, X. Jia, Q. Li, X.D. Yang, K. Wang, Biochemistry 43 (2004) 2688–2698. [7] J. Hu, X. Yang, K. Wang, J. Biol. Inorg. Chem. 10 (2005) 704–711. [8] J. Hu, S. Yu, X. Yang, K. Wang, Z.M. Qian, Biometals 19 (2006) 13– 18. [9] B. Solomon, R. Koppel, J. Jossiphov, Brain Res. Bull. 55 (2001) 253– 256. [10] C. Weis, A. Haug, Arch. Biochem. Biophys. 254 (1987) 304–312. [11] J.M. Cordeiro, V.S. Silva, C.R. Oliveira, P.P. Goncalves, J. Inorg. Biochem. 97 (2003) 132–142. [12] M. Kern, M. Wisniewski, L. Cabell, G. Audesirk, Neurotoxicology 21 (2000) 353–363. [13] R. Sandhir, K.D. Gill, Exp. Mol. Pathol. 61 (1994) 69–75. [14] S.S. Powlin, P.C. Keng, R.K. Miller, Toxicol. Appl. Pharmacol. 144 (1997) 225–234. [15] J. Ni, Bioinorganic Chemistry of Rare Earth Elements, Science Press, Beijing, 2002. [16] Evans, Biochemistry of Lanthanides, Plenum Press, New York, 1990. [17] J.M. Buccigross, C.L. O’Donnell, D.J. Nelson, Biochem. J. 235 (1986) 677–684. [18] Y. Ye, H.-W. Lee, W. Yang, J.J. Yang, J. Inorg. Biochem. 99 (2005) 1376–1383. [19] Y. Ye, H.-W. Lee, W. Yang, S.J. Shealy, A.L. Wilkins, Z.-R. Liu, I. Torshin, R. Harrison, R. Wohlhueter, J.J. Yang, Protein Eng. 14 (2001) 1001–1013. [20] Y. Ye, H.-W. Lee, W. Yang, S. Shealy, J.J. Yang, J. Am. Chem. Soc. 127 (2005) 3743–3750. [21] F. Arnesano, L. Banci, I. Bertini, F. Capozzi, S. Ciofi-Baffoni, S. Ciurli, C. Luchinat, S. Mangani, A. Rosato, P. Turano, M.S. Viezzoli, Coord. Chem. Rev. 250 (2006) 1419–1450. [22] W.F. Bowers, S. Fluton, J. Thompson, Clin. Pharmacokinet. 9 (1984) 49–60. [23] Department of Metallurgy, Sun Yat-Sen Zhongshan University(Ed.), The Physical and Chemical Constants of Rare Earth Elements, Metallurgy Press, Beijing, 1978. [24] S. Linse, T. Drakenberg, S. Forsen, FEBS Lett. 199 (1986) 28–32. [25] G. Sanyal, L.M. Richard, K.L.r. Carraway, D. Puett, Biochemistry 27 (1988) 6229–6236. [26] T. Murase, T. Iio, Biochemistry 41 (2002) 1618–1629. [27] P. Mulqueen, J.M. Tingey, W.D. Horrocks, Biochemistry 24 (1985) 6639–6645. [28] E. Pidcock, G.R. Moore, J. Biol. Inorg. Chem. 6 (2001) 479–489. [29] M. Yazawa, M. Ikura, K. Hikichi, L. Ying, K. Yagi, J. Biol. Chem. 262 (1987) 10951–10954. [30] S. Linse, S. Forsen, Adv. Second Messenger Phosphoprotein Res. 30 (1995) 89–151. [31] J.T. Welch, W.R. Kearney, S.J. Franklin, Pnas 100 (2003) 3725–3730. [32] M.A. Schumacher, A.F. Rivard, H.P. Bachinger, J.P. Adelman, Nature 410 (2001) 1120–1124.