Preparation and characterization of cobalt-substituted anthrax lethal factor

Biochemical and Biophysical Research Communications 416 (2011) 106–110

Contents lists available at SciVerse ScienceDirect

Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc

Preparation and characterization of cobalt-substituted anthrax lethal factor Crystal E. Säbel, Ryan Carbone, John R. Dabous, Suet Y. Lo, Stefan Siemann ⇑ Department of Chemistry and Biochemistry, Laurentian University, 935 Ramsey Lake Rd., Sudbury, Ontario, Canada P3E 2C6

a r t i c l e

i n f o

Article history: Received 15 October 2011 Available online 10 November 2011 Keywords: Anthrax lethal factor Electronic spectroscopy Metal substitution Thioglycolic acid Zinc proteases

a b s t r a c t Anthrax lethal factor (LF) is a zinc-dependent endopeptidase involved in the cleavage of mitogen-activated protein kinase kinases near their N-termini. The current report concerns the preparation of cobalt-substituted LF (CoLF) and its characterization by electronic spectroscopy. Two strategies to produce CoLF were explored, including (i) a bio-assimilation approach involving the cultivation of LFexpressing Bacillus megaterium cells in the presence of CoCl2, and (ii) direct exchange by treatment of zinc-LF with CoCl2. Independent of the method employed, the protein was found to contain one Co2+ per LF molecule, and was shown to be twice as active as its native zinc counterpart. The electronic spectrum of CoLF suggests the Co2+ ion to be five-coordinate, an observation similar to that reported for other Co2+-substituted gluzincins, but distinct from that documented for the crystal structure of native LF. Furthermore, spectroscopic studies following the exposure of CoLF to thioglycolic acid (TGA) revealed a sequential mechanism of metal removal from LF, which likely involves the formation of an enzyme: Co2+:TGA ternary complex prior to demetallation of the active site. CoLF reported herein constitutes the first spectroscopic probe of LF’s active site, which may be utilized in future studies to gain further insight into the enzyme’s mechanism and inhibitor interactions. Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction Anthrax lethal factor (LF) is a 90 kDa Zn2+-dependent endopeptidase responsible for the excision of an N-terminal peptide segment from most members of the mitogen-activated protein kinase kinase family of signaling proteins [1–3]. The Zn2+ ion in the active site of LF is coordinated to the side chains of His686, His690 and Glu735, and to a water molecule, which serves as the nucleophile in the cleavage of the substrate [4]. In addition, Glu687, which is part of the thermolysin-like HExxH consensus motif, has been shown to be essential for enzyme function [5], presumably by serving as a general base in the catalytic mechanism [4]. In view of the participation of the aforementioned amino acid residues in forming the active site, LF can be classified as a gluzincin [6]. In the case of a considerable number of Zn2+-dependent enzymes, insights into the nature and geometry of metal coordination at the active site, and the type of intermediates involved in the catalytic mechanism have been gained by replacement of the spectroscopically silent Zn2+ with Co2+, which is amenable to spectroscopic investigations [7,8]. Although a recent report has demonstrated the ability of Co2+ to reactivate the apoform of LF to yield an enzyme with slightly higher (by 35%) activity than that noted for the parent enzyme (ZnLF) [9], a Co2+-substituted analog of LF ⇑ Corresponding author. Fax: +1 705 675 4844. E-mail address: [email protected] (S. Siemann). 0006-291X/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2011.11.005

(CoLF) has yet to be isolated and studied in detail. The observation that many Co2+-substituted gluzincins display significantly higher catalytic activities than their native zinc counterparts prompted an investigation into alternative methods of metal replacement. The current report concerns the preparation and characterization of CoLF. The enzyme was prepared in high yields by (i) cultivating LF-expressing cells in the presence of CoCl2 (bio-assimilation approach), and (ii) treatment of ZnLF with excess Co2+ (direct exchange approach). Regardless of the method of preparation, CoLF was found to be twice as active as its native Zn2+-containing counterpart, a finding similar to that documented previously for other gluzincins including thermolysin [10]. The spectroscopic features of CoLF (in the absence and presence of a metal-chelating thiol inhibitor) are discussed in the context of those encountered with other Zn2+-dependent proteases. 2. Materials and methods 2.1. General Chromogenic anthrax lethal factor protease substrate II, S-pNA (Ac-GYbARRRRRRRRVLR-pNA, pNA = para-nitroanilide) was obtained from EMD Biosciences (La Jolla, CA). All other chemicals were purchased from Sigma–Aldrich (St. Louis, MO). All solutions were prepared using MilliQ ultrapure water (P18.2 MX cm resistivity). Hepes buffer (50 mM, pH 7.4) was depleted of trace metals with the aid of Chelex-100 resin.

107

C.E. Säbel et al. / Biochemical and Biophysical Research Communications 416 (2011) 106–110

2.2. Isolation of LF

2.5. Determination of metal content

Wild-type LF was isolated from Bacillus megaterium containing the plasmid pWH1520-LF (MoBiTec, Göttingen, Germany) as documented in the literature [11], with some minor modifications. Cells were grown in Terrific Broth (12 g/L tryptone, 24 g/L yeast extract, 0.4% (v/v) glycerol, 2.31 g/L KH2PO4, 12.54 g/L K2HPO4) supplemented with tetracycline (10 lg/mL) at 37 °C with moderate agitation (900 rpm) using a BioFlo 110 Fermenter (New Brunswick Scientific, Edison, NJ). Upon growth of the culture to an absorbance value of 0.3 at 600 nm (typically after 3 h), LF production was initiated by introduction of 0.5% (w/v) D-xylose to the medium, and growth was allowed to proceed for an additional 6 h. After removal of the cells by centrifugation (12,000g for 30 min), the culture fluid was treated with an equal volume of PEG-8000 (40% (w/v)) to facilitate protein precipitation. The mixture was stirred for 16 h at 4 °C, and the LF-containing protein precipitate was recovered by centrifugation (12,000g for 2 h). Following resolubilization in Tris/HCl buffer (20 mM, pH 8.0), LF was subjected to chromatography on Q-Sepharose equilibrated with the same medium using a discontinuous NaCl gradient. LF was recovered in the 350 mM NaCl fraction, and subsequently concentrated to approximately 1–2 mL using an Amicon Ultra-15 filtration device with a 30 kDa molecular weight cut-off (Millipore, Bedford, MA). The recovered protein preparation was finally subjected to size exclusion chromatography on a HiLoad (16/60) Superdex 200 column using an AKTA FPLC system (Amersham Biosciences, Uppsala, Sweden). Tris/HCl buffer (20 mM, pH 8.0) containing NaCl (200 mM) served both as equilibration and elution media. Fractions harboring LF were collected, pooled and concentrated by Amicon filtration prior to storage at 80 °C. The homogeneity of all LF preparations was assessed by SDS–PAGE according to the method of Laemmli [12]. The concentration of LF was determined spectrophotometrically at 280 nm, using an extinction coefficient of 74,200 M1 cm1 [9].

The Zn2+ content of ZnLF was assessed using the chromophoric chelator 4-(2-pyridylazo)resorcinol (PAR) according to published procedures [9]. The metal content (Co2+ and Zn2+) of CoLF was determined with PAR using the direct spectrophotometric method outlined in the literature [14].

2.3. Preparation of CoLF CoLF was prepared by two independent methods. The first procedure involved the isolation and purification of the protein in a fashion analogous to that described for ZnLF, except for the inclusion of CoCl2 (at a final concentration of 1 mM) in the growth medium (bio-assimilation approach). In addition, CoLF was prepared by exposing ZnLF to excess Co2+ (direct exchange approach). In brief, ZnLF (3–4 mg), purified as outlined above, was loaded onto a Q-Sepharose column (5 mL bed volume) equilibrated with Hepes buffer. The Zn2+ ion in LF was subsequently allowed to exchange with Co2+ by passing 50 mL of 5 mM CoCl2 in Hepes buffer through the column. Excess Co2+ and any released Zn2+ were removed by washing the column with 50 mL of Hepes buffer. Following elution of CoLF with 10 mL of Hepes buffer containing 350 mM NaCl, the protein was concentrated (to approximately 150 lM) by Amicon Ultra-15 filtration.

2.6. Determination of dissociation constant The dissociation constant of CoLF was determined using metalbuffered media (with dipicolinic acid (DPA) serving as the chelator) in a manner analogous to that employed previously in the estimation of the Kd value for ZnLF [9]. In brief, the Kd value of CoLF was determined by assessing the activity of the enzyme at various concentrations of free Co2+ in the medium using the following equation:

Relative activity ¼

½Co2þ free

ð1Þ

½Co2þ free þ K d

The (total) concentration of DPA ([DPA]tot) required to achieve a particular concentration of free Co2+ was calculated by first solving Eq. (2) before substituting the obtained concentration of free DPA into Eq. (3), 2

b2 ½DPA þ b1 ½DPA þ 1 

½Co2þ tot ½Co2þ free

! ¼0

ð2Þ

½DPAtot ¼ ½DPA þ ½Co2þ free ðb1 ½DPA þ 2b2 ½DPA2 Þ 2+

ð3Þ

2+

where [Co ]tot is the total concentration of Co in the assay, [DPA] is the concentration of the free chelator, and b1 and b2 denote the conditional stability constants of the Co2+(DPA) and Co2+(DPA)2 complexes (at pH 7.4: b1 (Co) = 107; b2 (Co) = 3.2  1012 [15,16]), respectively. The catalytic competence of CoLF (50 nM) was assessed in a typical enzyme assay (using 10 lM S-pNA) after exposure of the protein to Co2+ (50 lM) and DPA for 1 h at 25 °C. The Kd value of CoLF was estimated by fitting the enzyme activities to Eq. (1) by non-linear regression using GraFit 4.0. 2.7. UV–Vis spectroscopy Electronic spectra of CoLF were recorded at 25 °C with an OLIS RSM-1000 spectrophotometer (Bogart, GA) using a 10-mm pathlength micro quartz cell (Hellma, Concord, ON) and 100 lL samples containing approximately 150 lM (13.5 mg LF/mL) protein in Hepes buffer (50 mM, pH 7.4). Since samples containing LF at such high concentrations were slightly turbid, all spectra were corrected for scattering using the reciprocal relationship between the intensity of scattered light and wavelength (I  1/k4). 3. Results 3.1. Preparation of CoLF

2.4. Enzymatic assays The enzymatic activity of LF was assessed spectrophotometrically at 405 nm with the aid of the chromogenic S-pNA substrate according to published protocols [9,13]. Steady-state kinetic parameters (KM, kcat) for the hydrolysis of S-pNA were estimated by a non-linear least squares fit of the initial velocity data to the Michaelis–Menten equation using the Grafit 4.0 software package (Erithacus Software Ltd., Staines, UK) and a De405 nm value of 9920 M1 cm1 [13].

Preliminary studies on the cytotoxicity of Co2+ on B. megaterium cultures revealed an absence of impairment of cellular growth up to a concentration of 1 mM CoCl2 in the growth medium (data not shown). Thus, CoCl2 at a final concentration of 1 mM was added as a supplementary component to the standard growth medium, and cell cultivation and purification of LF were performed according to the protocols employed for the isolation of ZnLF. Protein yields were identical to those obtained for the Zn2+ enzyme (ca. 5 mg of LF/L of cell culture). The metal content (mol of metal per mol of protein) of LF prepared in this manner was found to be 0.98 (±0.10) for

108

C.E. Säbel et al. / Biochemical and Biophysical Research Communications 416 (2011) 106–110

Co2+ and 0.03 (±0.02) for Zn2+ (Table 1). It is important to note that the Zn2+ content of the growth medium was found to be 62 lM (determined with PAR). It thus appears that a P500-fold excess of Co2+ over other metal ions is sufficient to obtain a preparation of CoLF which is virtually devoid of Zn2+. The bio-assimilation route appears to be a simple and straightforward procedure for the isolation of CoLF in high yields. However, such methodology may not be suitable for the generation of CoLF using other commonly employed LF expression systems (e.g., Escherichia coli). Thus, the possibility of generating CoLF directly from ZnLF by metal ion exchange was explored. In preliminary experiments, the propensity of Co2+ to displace LF’s native Zn2+ ion was examined by assessing changes in the enzymatic activity of ZnLF upon its exposure to Co2+. As shown in Fig. 1, supplementation of the assay with Co2+ at concentrations greater than 100 lM led to a marked increase in LF’s catalytic competence, with maximal (2-fold) activity being noted at concentrations P1 mM. This result clearly suggests that Co2+ is capable of directly replacing LF’s active site Zn2+ to yield a hyperactive enzyme, an observation similar to that reported for Co2+-substituted thermolysin [10,17]. To obtain CoLF in the large quantities required for spectroscopic investigations, ZnLF was loaded onto Q-Sepharose prior to treatment of the immobilized protein with Co2+ (see Section 2). Following removal of any unbound metal ions and elution of the enzyme, the metal content of LF was found to be virtually identical to that obtained for CoLF generated by bio-assimilation (see Table 1). 3.2. Steady-state kinetic parameters and dissociation constant of CoLF Steady-state kinetic parameters for the hydrolysis of S-pNA were obtained for both CoLF and ZnLF (Table 1). While both proteins exhibited similar KM values for S-pNA, CoLF displayed a 2-fold higher kcat value with respect to that of the native Zn2+ enzyme. The dissociation constant of CoLF was estimated by determining the activity of the enzyme in a series of metal-buffered media using a 1000-fold molar excess of Co2+ with respect to the concentration of CoLF in the assay, and excess DPA as the chelator. As depicted in Fig. 2, the Kd value of CoLF was found to be 75 (±4) pM, and was thus 60 times higher than that observed for its Zn2+-containing analog [9].

Fig. 1. Influence of Co2+ on the activity of ZnLF. Assays were performed by preincubating ZnLF (50 nM) for 30 min in Hepes buffer in the presence of Co2+ prior to monitoring the protein’s activity with S-pNA (10 lM). Activities are expressed relative to that of ZnLF. Error bars represent ±1 s.d. of three independent experiments.

absorption below 350 nm is a consequence of absorption by the tyrosine and tryptophan residues of LF. Exposure of CoLF to thioglycolic acid (TGA) at a concentration of 1 mM for up to 30 min was found to only marginally increase the absorption in the visible region of the spectrum with no apparent change in band shape (Fig. 3A). However, incubation of the enzyme in the presence of 10 mM TGA for 2 min was found to significantly broaden the absorption band in the 500–600 nm region with concomitant shift of kmax to 570 nm and an increase in the molar extinction coefficient to 200 M1 cm1. Furthermore, the marked increase in absorbance below 350 nm upon addition of TGA can be attributed to a sulfur-to-Co2+ ligand-to-metal charge transfer (LMCT) band. Indeed, subtraction of the spectra of CoLF (unexposed to TGA) and unbound TGA (9.85 mM, assuming the formation of a enzyme: Co2+:TGA ternary complex) from that of CoLF treated with 10 mM TGA revealed an intense (e = 3500 M1 cm1) absorption band at 310 nm (Fig. 3B). Both, the magnitude of the extinction coefficient and the position of kmax are consistent with this absorption band arising from an LMCT transition [10]. Thus,

3.3. Electronic spectra of CoLF As shown in Fig. 3, the electronic spectrum of CoLF displayed an absorption band at 555 nm (e = 150 M1 cm1), indicative of a Co2+ d–d transition. The band was found to be accompanied by a prominent shoulder at 520 nm and a smaller shoulder at 580 nm. A second absorption band, centered at 415 nm in the spectrum of CoLF suggests the presence of a small amount of Co3+ (a consequence of air-oxidation of some of the Co2+ centers), a feature documented previously for Co2+-substituted VanX [18]. The steep increase in

Table 1 Kinetic parameters of LF-mediated hydrolysis of S-pNA. Enzyme

ZnLF CoLFa CoLFb a b

Metal content (mol/mol of LF) Zn2+

Co2+

1.00 (±0.10) 0.03 (±0.02) 0.05 (±0.03)

– 0.98 (±0.10) 1.02 (±0.05)

Prepared by bio-assimilation. Prepared by direct exchange.

KM (lM)

kcat (s1)

kcat/KM (M1 s1)

3.5 (±1.4) 2.9 (±1.6) 2.8 (±1.2)

3.74 (±0.48) 7.15 (±0.92) 7.10 (±0.91)

1.07  106 2.47  106 2.54  106

Fig. 2. Determination of the dissociation constant of CoLF. LF (50 nM) in Hepes buffer was exposed to Co2+ (50 lM), and excess DPA (calculated using Eqs. (2) and (3)) for 1 h at 25 °C prior to initiation of the assay by introduction of S-pNA (10 lM). The relative activities for CoLF (squares) were fit to Eq. (1) by non-linear regression (solid line). Error bars represent the standard error of the mean of three independent experiments. For comparison, the data reported previously for the determination of ZnLF’s Kd value [9] are included in the figure (circles).

C.E. Säbel et al. / Biochemical and Biophysical Research Communications 416 (2011) 106–110

Fig. 3. Electronic spectra of CoLF. Panel A: absorption spectra of CoLF (152 lM) in Hepes buffer in the absence (black) and presence of 1 mM (magenta) and 10 mM (blue) TGA. CoLF was exposed to TGA for 2 min prior to recording absorption spectra. The spectrum of 10 mM TGA in Hepes buffer is depicted as a dashed line. Panel B: difference spectrum of CoLF. The spectrum was obtained by subtracting the spectra of 152 lM CoLF and 9.85 mM (unbound) TGA from that of CoLF (152 lM) exposed to 10 mM TGA. The magnified difference spectrum showing changes in the d–d transitions is depicted in the inset. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the appearance of an LMCT band provides clear evidence of a direct interaction of LF-bound Co2+ with TGA through its sulfur atom. As outlined above, alterations in extinction coefficient(s) in the spectral region of the d–d transitions are small compared to those in the near-UV region (see inset of Fig. 3B). In view of the well-documented ability of TGA to serve as a bidentate chelator for Co2+ and other transition metal ions [19], it was of interest to determine whether TGA is capable of demetallating CoLF following prolonged exposure of the enzyme to TGA. As shown in Fig. 4A, treatment of CoLF with TGA (10 mM) for 120 min led to a dramatic change in the absorption spectrum as evidenced by the emergence of two intense bands at 370 and 475 nm. To establish whether these alterations are a consequence of TGAmediated removal of Co2+ from CoLF, absorption spectra of Co2+ (150 lM) exposed to various concentrations of TGA (0–1 mM) were recorded. As depicted in Fig. 4B, addition of TGA to Co2+ led to a progressive increase in the absorbance at 370 nm, a finding similar to that reported recently [20]. More importantly, a second distinct band at 475 nm became prominent only when the [TGA]:[Co2+] ratio approached a value of two (see inset of Fig. 4B). These observations are indicative of the feature at 475 nm originating from a TGA:cobalt complex of 2:1 stoichiometry. Indeed, previous studies on the complexation of Co2+ by TGA have shown that the absorption bands at 370 and 475 nm can be attributed to (significantly red-shifted) charge-transfer transitions originating from CoIIITGA2 [19], a finding not surprising in view of the well-documented ability of Co2+ to undergo rapid air oxidation in the presence of two equivalents of TGA [21]. Hence, the appearance of the absorption bands at 370 and 475 nm in the spectrum of CoLF is a clear reflection of the removal the enzyme’s metal ion and

109

Fig. 4. Absorption spectrum of CoLF after prolonged exposure to TGA. Panel A: absorption spectra of CoLF (152 lM) following its incubation in the presence of TGA (10 mM) for 2 min (blue) and 120 min (black). Panel B: absorption spectra of CoCl2 (150 lM) in Hepes buffer in the absence and presence of increasing concentrations of TGA (up to 1 mM). The dependence of the ratio of absorbances at 370 and 475 nm on the TGA:Co2+ concentration ratio is depicted in the inset. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the formation of a CoTGA2 complex. This view gains further support from the observation that the metal ion could be removed quantitatively from the enzyme by Amicon filtration (data not shown). 4. Discussion Previous reports on the ability of apoLF to regain its catalytic competence upon exposure to Co2+ and other transition metals [5,9] have demonstrated that the degree of reactivation falls significantly below that encountered with other zinc proteases such as thermolysin and carboxypeptidase A [10,17,22]. In light of these observations, alternative methods of metal replacement were explored. The current study indicates that CoLF can be prepared by two approaches which are less commonly employed to generate Co2+-substituted zinc enzymes. Both the bio-assimilation and direct exchange strategies were found to produce CoLF which contained one Co2+ per protein molecule and was virtually devoid of Zn2+. In view of the propensity of Co2+ to displace Zn2+ from LF, the question arises as to whether this phenomenon might form the basis of the preparation of CoLF via bio-assimilation. It is important to note in this context that when Co2+ was added to the growth medium at the time of induction of LF expression (i.e., 3 h after initiation of cell growth), the active site occupation by Co2+ was found to be only 0.5 Co2+/LF. Indeed, if ZnLF were secreted from the cells with subsequent metal exchange in the external milieu, addition of Co2+ at the onset of LF expression should have resulted in a Co2+ content of 1.0 Co2+/LF (see Table 1). Consequently, it appears likely that (for at least a significant proportion of LF) Co2+ is incorporated into LF intracellularly.

110

C.E. Säbel et al. / Biochemical and Biophysical Research Communications 416 (2011) 106–110

Analysis of the kinetic parameters of substrate hydrolysis has revealed a 2-fold higher kcat value of CoLF compared to ZnLF, an observation similar to that noted for Co2+–thermolysin [10] and other zinc proteases [8]. In addition, the Kd value of CoLF was found to be significantly higher than that previously reported for ZnLF [9]. This observation is in accordance with those documented for Co2+–thermolysin [10] and in general agreement with the position of Co2+ and Zn2+ in the Irving–Williams series [23]. The electronic spectrum of CoLF was found to display an absorption band at 555 nm (e = 150 M1 cm1). With respect to band shape and position, the spectrum of CoLF is similar to that documented for Co2+–thermolysin [10] and virtually identical to that recorded for the Co2+-substituted zinc peptidase VanX [18]. The magnitude of the extinction coefficient of Co2+ complexes has been shown to depend on the coordination number, such that the e value of pentacoordinate arrangements is typically between 50 and 300 M1 cm1, and that of tetrahedral geometry above 300 M1 cm1 [7,24]. Consequently, the observed d–d transition in the spectrum of CoLF suggests Co2+ in the active site of the protein to be five-coordinate. This situation is similar to that encountered with Co2+–thermolysin [25], for which an extinction coefficient of 80 M1 cm1 has been observed [10]. In addition, the electronic spectra of a variety of other Co2+-substituted zinc proteases, including carboxypeptidase A [26], angiotensin-converting enzyme [27], VanX [18], and tetanus neurotoxin [28] have been shown to display extinction coefficients between 75 and 150 M1 cm1 for their major d–d absorption bands, consistent with pentacoordination of Co2+. It is important to note that the crystal structure of ZnLF has revealed Zn2+ to be tetrahedrally coordinated [4]. As has been pointed out previously, however, it is unlikely that substitution of Zn2+ by Co2+ causes an increase in the metal’s coordination number [18]. Thus, it is not inconceivable that Zn2+ in LF is rather five- than four-coordinate. The observation of spectral changes upon exposure of CoLF to TGA (shift of kmax with concomitant increase in emax; emergence of LMCT transitions) attests to the suitability of such enzyme preparation not only as a spectroscopic tool to probe the metal ion’s coordination environment, but also to gain insight into enzymeinhibitor interactions. In the case of CoLF, spectroscopic evidence indicates that TGA initially interacts with the enzyme to form a ternary complex, followed by removal of the metal ion to yield apoLF and CoTGA2. As such, the Co2+-substituted form of LF reported herein constitutes the first direct spectroscopic probe of the protein’s active site, which should prove useful in future investigations on mechanistic aspects of LF-mediated substrate hydrolysis, and on the molecular details of the interaction of inhibitors with this enzyme.

[2]

[3]

[4]

[5] [6] [7] [8] [9]

[10] [11]

[12] [13] [14]

[15]

[16]

[17] [18]

[19]

[20]

[21] [22] [23] [24]

Acknowledgments We are indebted to Dr. J. Mogridge (Dept. of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada) for providing us with the expression system for LF, and to NSERC for financial support. References [1] G. Vitale, R. Pellizzari, C. Recchi, G. Napolitani, M. Mock, C. Montecucco, Anthrax lethal factor cleaves the N-terminus of MAPKKs and induces tyrosine/

[25]

[26] [27]

[28]

threonine phosphorylation of MAPKs in cultured macrophages, Biochem. Biophys. Res. Commun. 248 (1998) 706–711. G. Vitale, L. Bernardi, G. Napolitani, M. Mock, C. Montecucco, Susceptibility of mitogen-activated protein kinase kinase family members to proteolysis by anthrax lethal factor, Biochem. J. 352 (2000) 739–745. N.S. Duesbery, C.P. Webb, S.H. Leppla, V.M. Gordon, K.R. Klimpel, T.D. Copeland, N.G. Ahn, M.K. Oskarsson, K. Fukasawa, K.D. Paull, G.F. Vande Woude, Proteolytic inactivation of MAP-kinase–kinase by anthrax lethal factor, Science 280 (1998) 734–737. A.D. Pannifer, T.Y. Wong, R. Schwarzenbacher, M. Renatus, C. Petosa, J. Bienkowska, D.B. Lacy, R.J. Collier, S. Park, S.H. Leppla, P. Hanna, R.C. Liddington, Crystal structure of the anthrax lethal factor, Nature 414 (2001) 229–233. S.E. Hammond, P.C. Hanna, Lethal factor active-site mutations affect catalytic activity in vitro, Infect. Immun. 66 (1998) 2374–2378. W.N. Lipscomb, N. Sträter, Recent advances in zinc enzymology, Chem. Rev. 96 (1996) 2375–2434. I. Bertini, C. Luchinat, High spin cobalt(II) as a probe for the investigation of metalloproteins, Adv. Inorg. Biochem. 6 (1984) 71–111. D.S. Auld, Removal and replacement of metal ions in metallopeptidases, Methods Enzymol. 248 (1995) 228–242. C.E. Säbel, S. St-Denis, J.M. Neureuther, R. Carbone, S. Siemann, Alkaline earth metals are not required for the restoration of the apoform of anthrax lethal factor to its holoenzyme state, Biochem. Biophys. Res. Commun. 403 (2010) 209–213. B. Holmquist, B.L. Vallee, Metal substitutions and inhibition of thermolysin: spectra of the cobalt enzyme, J. Biol. Chem. 249 (1974) 4601–4607. A. Kassam, S.D. Der, J. Mogridge, Differentiation of human monocytic cell lines confers susceptibility to Bacillus anthracis lethal toxin, Cell. Microbiol. 7 (2005) 281–292. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680–685. F. Tonello, M. Seveso, O. Marin, M. Mock, C. Montecucco, Screening inhibitors of anthrax lethal factor, Nature 418 (2002) 386. C.E. Säbel, J.L. Shepherd, S. Siemann, A direct spectrophotometric method for the simultaneous determination of zinc and cobalt in metalloproteins using 4(2-pyridylazo)resorcinol, Anal. Biochem. 391 (2009) 74–76. K. Fukasawa, K.M. Fukasawa, M. Kanai, S. Fujii, J. Hirose, M. Harada, Dipeptidyl peptidase III is a zinc metallo-exopeptidase. Molecular cloning and expression, Biochem. J. 329 (1998) 275–282. R.M. Tichane, W.E. Bennett, Coordination compounds of metal ions with derivatives and analogs of ammoniadiacetic acid, J. Am. Chem. Soc. 79 (1957) 1293–1296. Y. Hashida, K. Inouye, Kinetic analysis of the activation-and-inhibition dual effects of cobalt ion on thermolysin activity, J. Biochem. 141 (2007) 843–853. R.M. Breece, A. Costello, B. Bennett, T.K. Sigdel, M.L. Matthews, D.L. Tierney, M.W. Crowder, A five-coordinate metal center in Co(II)-substituted VanX, J. Biol. Chem. 280 (2005) 11074–11081. D.L. Leussing, The stabilities and absorption spectra of complexes of some divalent metal ions of the 1st transition series with the thioglycolate ion, J. Am. Chem. Soc. 80 (1958) 4180–4183. Y. Yamaguchi, K. Imamura, A. Sasao, E. Murakami, Y. Arakawa, H. Kurosaki, Metal preference of Zn(II) and Co(II) for the dinuclear metal binding site of IMP-1 metallo-b-lactamase and spectroscopic properties of Co(II)-substituted IMP-1 with mercaptoacetic acid, Med. Chem. Commun. 2 (2011) 720–725. L. Michaelis, M.P. Schubert, Cobalt complexes of thioglycolic acid, J. Am. Chem. Soc. 52 (1930) 4418–4426. M.W. Makinen, G.B. Wells, S.O. Kang, Structure and mechanism of carboxypeptidase A, Adv. Inorg. Biochem. 6 (1984) 1–69. H. Irving, R.J.P. Williams, The stability of transition-metal complexes, J. Chem. Soc. (1953) 3192–3210. N. Shapir, J.P. Osborne, G. Johnson, M.J. Sadowsky, L.P. Wackett, Purification, substrate range, and metal center of AtzC: the N-isopropylammelide aminohydrolase involved in bacterial atrazine metabolism, J. Bacteriol. 184 (2002) 5376–5384. D.R. Holland, A.C. Hausrath, D. Juers, B.W. Matthews, Structural analysis of zinc substitutions in the active site of thermolysin, Protein Sci. 4 (1995) 1955– 1965. S.A. Latt, B.L. Vallee, Spectral properties of cobalt carboxypeptidase. The effects of substrates and inhibitors, Biochemistry 10 (1971) 4263–4270. R. Bicknell, B. Holmquist, F.S. Lee, M.T. Martin, J.F. Riordan, Electronic spectroscopy of cobalt angiotensin converting enzyme and its inhibitor complexes, Biochemistry 26 (1987) 7291–7297. F. Tonello, G. Schiavo, C. Montecucco, Metal substitution of tetanus neurotoxin, Biochem. J. 322 (1997) 507–510.