Nội dung

Crystal structure of Cu ⁄ Zn superoxide dismutase from Taenia solium reveals metal-mediated self-assembly Alejandra Hernández-Santoyo1, Abraham Landa2, Edith González-Mondragón2, Martha Pedraza-Escalona1, Ricardo Parra-Unda2 and Adela Rodrı́guez-Romero1 1 Instituto de Quı́mica, Universidad Nacional Autónoma de México, Mexico 2 Departamento de Microbiologı́a y Parasitologı́a, Facultad de Medicina, Universidad Nacional Autónoma de México, Mexico Keywords crystal structure; metal-mediated self-assembly; superoxide dismutase; Taenia solium Correspondence A. Hernández-Santoyo, A. Rodrı́guezRomero, Instituto de Quı́mica, Universidad Nacional Autónoma de México, México 04510, DF, Mexico Fax: +52 55 56162217 Tel: +52 55 56224568 E-mail: hersan@servidor.unam.mx, adela@servidor.unam.mx (Received 20 July 2010, revised 14 June 2011, accepted 13 July 2011) doi:10.1111/j.1742-4658.2011.08247.x Taenia solium is the cestode responsible for porcine and human cysticercosis. The ability of this parasite to establish itself in the host is related to its evasion of the immune response and its antioxidant defence system. The latter includes enzymes such as cytosolic Cu ⁄ Zn superoxide dismutase. In this article, we describe the crystal structure of a recombinant T. solium Cu ⁄ Zn superoxide dismutase, representing the first structure of a protein from this organism. This enzyme shows a different charge distribution at the entrance of the active channel when compared with human Cu ⁄ Zn superoxide dismutase, giving it interesting properties that may allow the design of specific inhibitors against this cestode. The overall topology is similar to other superoxide dismutase structures; however, there are several His and Glu residues on the surface of the protein that coordinate metal ions both intra- and intermolecularly. Interestingly, one of these ions, located on the b2 strand, establishes a metal-mediated intermolecular b–b interaction, including a symmetry-related molecule. The factors responsible for the abnormal protein– protein interactions that lead to oligomerization are still unknown; however, high metal levels have been implicated in these phenomena, but exactly how they are involved remains unclear. The present results suggest that this structure could be useful as a model to explain an alternative mechanism of protein aggregation commonly observed in insoluble fibrillar deposits. Database The atomic coordinates and structure factors have been deposited in the Protein Data Bank under the accession number 3MND. Structured digital abstract Cu/Zn SOD binds to Cu/Zn SOD by dynamic light scattering (View Interaction 1, 2) l Cu/Zn SOD binds to Cu/Zn SOD by mass spectrometry studies of complexes (View interaction) l Cu/Zn SOD binds to Cu/Zn SOD by molecular sieving (View Interaction 1, 2) l Cu/Zn SOD binds to Cu/Zn SOD by x-ray crystallography (View interaction) l Introduction Superoxide dismutases (SODs, 1.15.1.1) are metalloenzymes that use Cu ⁄ Zn, Mn, Fe or Ni in their active sites to transform superoxide radicals (O2 ) into hydrogen peroxide and molecular oxygen (2O2 + 2H+ fi H2O2 + O2). The metal ions function as cofactors that play important roles in the defence against oxygen-derived Abbreviations Ab, amyloid-b; ALS, amyotrophic lateral sclerosis; PDB, Protein Data Bank; ROS, reactive oxygen species; SOD, superoxide dismutase; Ts, Taenia solium. 3308 FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. free radicals; therefore, these enzymes are important from a pharmaceutical point of view [1,2]. There are two forms of Cu ⁄ Zn-SOD enzymes: one extracellular (ECCu ⁄ Zn-SOD) tetramer composed of 30-kDa subunits, and a cytosolic (Cu ⁄ Zn-SOD) dimer with an Mr value of 16 kDa per subunit. Sequence alignment between the two enzymes shows 50% identity, and both contain a binuclear Cu2+, Zn2+ centre per subunit [3,4]. Cu is involved in the catalytic reaction in two steps: first, Cu2+ reduction by one molecule of O2 produces molecular oxygen, and Cu+ oxidation by another O2 molecule generates H2O2. In contrast, Zn2+ plays a structural role. The structures of Cu ⁄ Zn-SODs from different eukaryotes have been investigated extensively, and each monomer is a flattened Greek-key b-barrel, characterized by eight antiparallel b strands (b1–b8) connected by seven loops (L1–L7) [5,6]. Two of these, the electrostatic (L7) and Zn-binding (L4) loops, with some residues of the b-barrel, form the walls of the active site cavity that steer O2 from the enzyme surface to the active site. Several positively charged residues within the electrostatic loop create a charge gradient, which drives the substrate to the metal site at which catalysis occurs [7]. An important characteristic of human Cu ⁄ Zn-SOD enzymes is that they are involved in inflammation, tumour proliferation, aging, cell growth and neurodegenerative diseases, such as amyotrophic lateral sclerosis (ALS) [1,2,8,9]. In helminths, such as Schistosoma mansoni and Taenia solium, both Cu ⁄ Zn-SOD enzymes have been considered as targets for drug and vaccine development because they are important for the detoxification of reactive oxygen species (ROS) [10]. Recently, the crystal structure of S. mansoni Cu ⁄ Zn-SOD at 1.5 Å resolution has been reported. This enzyme is different from human Cu ⁄ Zn-SOD with regard to two amino acids (Leu132 and Val135) which are localized at the entrance of the channel that leads O2 to the active site [11]. We produced a recombinant T. solium Cu ⁄ Zn-SOD (TsCu ⁄ Zn-SOD) that possesses the classical motifs and biochemical properties of cytosolic enzymes. In vitro studies showed that the enzyme is completely inhibited by 500 mM thiabendazole and 300 mM albendazole; in contrast, neither anthelminthic affected bovine Cu ⁄ Zn-SOD [12]. It is worth mentioning that T. solium is the causal agent of taeniosis in humans and cysticercosis in humans and pigs worldwide; moreover, neurocysticercosis in humans is a debilitating and sometimes mortal disease which requires expensive treatment [13]. In this article, we report the crystal structure of a recombinant TsCu ⁄ Zn-SOD and some of its relevant features. The presence of His and Glu residues on the surface of the protein favours protein–protein interactions through metal coordination. The b sheet formed by the b1, b2, b3 and b6 strands is a key motif that establishes b–b interactions with symmetry molecules. The b2 strand, in particular, promotes protein oligomerization through the metal-mediated self-assembly of dimers of TsCu ⁄ Zn-SOD. We confirmed these aggregation phenomena in the presence of different concentrations of metal ions using gel filtration, mass spectrometry analyses and dynamic light-scattering experiments. Results and Discussion Enzyme preparation, thermal stability and oligomerization analysis After expression in Escherichia coli in the presence of Zn2+ and Cu2+ ions, sequential purification using a DEAE column at pH 7.4 and pH 8.9, respectively, yielded the dimeric holoenzyme (TsCu ⁄ Zn-SOD) with a molecular mass of 32 kDa and a specific activity of 2940 UÆmg)1. This value is comparable with those of other eukaryotic SODs, such as the human and bovine enzymes [14], and is obtained when the protein is completely metallated. The purified holoenzyme presents greater temperature stability in the presence of additional Cu2+ or Zn2+ ions. Figure 1A shows how the activity of TsCu ⁄ Zn-SOD decreases with increasing temperature, but is unaffected in the range 10–37 C. The activity abruptly decreases by 80% at 80 C and this level is then maintained for 30 min. Incubation at 100 C for more than 5 min completely deactivates the enzyme. When 1.0 mM Cu2+ or Zn2+ ions are added to the reaction cell, an increase in the thermostability of the enzyme is observed. At 80 C in the presence of additional metal ions, the enzyme activity is not affected for 30 min (Fig. 1B), whereas incubation of the enzyme at 100 C results in 30% of its original activity, as shown in Fig. 1C. These results show that metal ions diminish the thermal inactivation of the holoenzyme, as has been reported for the porcine and E. coli Cu ⁄ Zn-SOD enzymes [15,16]. Another interesting characteristic of this protein is that its molecular mass distribution changes in the presence of different concentrations of ZnSO4, as determined by gel filtration (Fig. S1). For example, the pure holoprotein (0.05 mgÆmL)1) is dimeric in 0.1 M Tris, pH 8.5, 0.2 M NaCl in the absence of ZnSO4. On addition of 0.5–1.0 mM ZnSO4 to the protein solution, a new broad peak centred at approximately 64 kDa FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS 3309 Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A A. Hernández-Santoyo et al. 15905.24 100 80 Intensity × 104 1.5 60 40 20 1.0 0.5 0 30214.84 0 5 10 15 20 25 30 44375.93 Residual activity (%) B 100 m/z 80 Fig. 2. Matrix-assisted laser desorption ionization time-of-flight mass spectrum of Taenia solium Cu ⁄ Zn superoxide dismutase (TsCu ⁄ Zn-SOD). Several oligomerization species were observed when 1.0 mM ZnSO4 was added to the protein solution. 60 40 20 0 C 0 5 10 0 5 10 15 20 25 30 15 20 25 30 100 80 60 40 20 0 Time (min) Fig. 1. Taenia solium Cu ⁄ Zn superoxide dismutase (TsCu ⁄ Zn-SOD) thermostability. Enzyme activity of the holoenzyme (squares), after the addition of 1.0 mM ZnSO4 (triangles) and of CuSO4 (circles) at 37 C (A), 80 C (B) and 100 C (C). All measurements were performed as triplicates and error bars are presented. appears at the expense of the dimeric species (32 kDa). These results were confirmed by mass spectrometry, which showed several species from monomer to tetramer or even higher molecular weight species, indicating that the protein oligomerizes in the presence of metal ions (Fig. 2). The molecular mass of one monomer of native TsCu ⁄ Zn-SOD is 15 905.24 Da; notably, if we compare this value with the theoretical value obtained from the amino acid sequence (15 588.48 Da), the difference can be explained by the presence of one Zn2+ and one Cu2+ ion, in the active site, and approximately two to three extra Zn2+ ions per monomer. In accordance with these results, the crystal structure 3310 58390.10 0.0 10 000 20 000 30 000 40 000 50 000 60 000 70 000 80 000 90 000 reported in this work showed three metal ions in regions exposed to the solvent (see below). Dynamic light-scattering measurements allow us to analyse the aggregation processes in the presence of Zn2+, Cu2+ and Ni2+ ions. As the scattering intensity is proportional to the second power of the particle mass, the contribution to scattering from larger particles dominates the scattering signal [17], as shown in Fig. S2. It is interesting to note that, in Tris buffer, pH 8.0, the protein at a concentration of 1.0 mgÆmL)1 is a tetramer. This result differs from that obtained in gel filtration, where the protein behaves as a dimer, as a consequence of the lower protein concentration in the latter experiment. After addition of the ions, distinct species appeared in the size distribution extrapolated from the dynamic lightscattering data using MALVERN DTS software (Malvern Instruments, Malvern, Worcestershire, UK). Figure S2 shows the holoenzyme behaviour in the presence of different ions. When 1 mM ZnSO4 was added to the dynamic light-scattering cell, we observed oligomerization with a peak centred at approximately 300 kDa (Fig. S2A). When CuSO4 was used, oligomerization was also observed; however, twice the concentration of Cu2+ ions was needed to observe a peak centred at 200 kDa (Fig. S2B). Conversely, 1 mM NiCl2 produced the opposite effect and the dimeric state was obtained (Fig. S2C), suggesting that this ion could favour a monodisperse state in crystallization conditions. There exist in the literature conflicting statements about the role of Cu2+ and Zn2+ ions as promoters of oligomerization; nonetheless, several reports have demonstrated that Zn2+ efficiently induces the aggregation FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. of synthetic amyloid-b (Ab) peptide under conditions similar to physiological conditions in the normal brain [18,19]. Moreover, Stellato et al. [20] have demonstrated, using X-ray absorption spectroscopy, that Zn2+ favours Ab peptide aggregation, supporting our results. In the case of ALS, where Homo sapiens SOD is involved, it has been reported that the immature nascent enzyme is prone to aggregation as a result of the absence of metal ions or mis-metallation [21]. Nonetheless, for TsCu ⁄ Zn-SOD, oligomerization occurs through a different mechanism. Crystal structure overview TsCu ⁄ Zn-SOD crystals grew in about 1 week with a rod-shaped morphology and diffracted at 2.2 Å resolution. The analysis of the diffraction pattern showed that the crystals belonged to the space group P212121 with unit cell parameters of a = 42.17 Å, b = 53.80 Å and c = 117.26 Å. The calculated Matthews’ coefficient [22] for two monomers per asymmetric unit is 2.13 Å3ÆDa)1 and gives an estimated solvent content of 42.2%. The refined structure contains two monomers, each consisting of 152 amino acid residues and 109 ordered water molecules, with a final Rwork of 0.192 and Rfree of 0.249. Details of the refinement statistics are listed in Table 1. The TsCu ⁄ Zn-SOD structure shows the canonical features conserved throughout the phyla, including the Greek-key b-barrel motif (Fig. 3A). The active site includes a catalytically active Cu2+ ion and a structural Zn2+ ion. Figure 3B shows secondary structure elements with their canonical nomenclature indicated [23]. The Cu2+ ion in both monomers is coordinated by four His residues in a distorted tetrahedral geometry. Solvent molecules are observed at 2.48 and 2.91 Å from the Cu2+ ion in the active site of monomers A and B, respectively (Fig. 4A, C). In general, when compared with Homo sapiens SOD [Protein Data Bank (PDB) entry 2V0A], water molecules around the Cu2+ ion occupy similar positions and are similar in number (four); only monomer A in TsCu ⁄ Zn-SOD lacks one of these molecules. It is worth mentioning at this point that the catalytic activity of Cu ⁄ Zn SOD seems to be unrelated to the presence of water and that the electron transfer is not water mediated [24]. The Cu– His60–Zn imidazolate bridge that is intact in both monomers is consistent with Cu2+. In the reduced form, the imidazolate bridge is ruptured and the catalytic metal is three coordinated [25], whereas, if Cu is oxidized, it is coordinated to four His residues and is also connected to Zn through a bridging His residue [26]. In the Taenia enzyme, His60 makes a bridge Table 1. Data collection and structure refinement statistics. Data collection X-Ray source Wavelength (Å) Space group Unit cell (Å, deg) Resolution range (Å)a VM (Å3ÆDa)1) Rmerge (%)a,b Completeness (%)a Total reflections Unique reflections Redundancya I ⁄ r(I)a B-factors of data from Wilson plot (Å2) Refinement Resolution range (Å)a Number of reflectionsa Rcryst (%) ⁄ Rfree (%)a,c,d Rmsd bond length (Å) Rmsd bond angle (deg) Protein atoms Metal ions Cu ⁄ Zn Glycerol molecules Water molecules Average B-factor (Å2) Protein ⁄ solvent ⁄ metal ions ⁄ glycerol Ramachandran plot (%)e Preferred Allowed regions Beamline X12B, NSLS 0.9795 P212121 a = 42.17, b = 53.80, c = 117.26, a = b = c = 90 53.8–2.2 (2.32–2.2) 2.08 11.6 (35.6) 98.8 (97.7) 44634 13934 3.2 (3.0) 7.0 (2.8) 20.9 39.6–2.2 (2.28–2.2) 13829 (1864) 19.2 ⁄ 24.5 0.007 1.082 2200 2⁄5 2 109 20.9 ⁄ 21.8 ⁄ 28.7 ⁄ 26.4 95.2 4.8 a Values in parentheses correspond to the last resolution shell. Rmerge = RjRh(|Ij,h – |) ⁄ RjRh(), where subscript h is the unique reflection index, Ij,h is the intensity of the symmetry-related P reflection and is the mean intensity. c R = h||Fo|h – P |Fc|h| ⁄ h|Fo|h for all reflections, where Fo and Fc are the observed and calculated structure factors, respectively, and h defines unique reflections. d Rfree is calculated analogously for the test reflections, randomly selected and excluded from the refinement. e Ramachandran plots were prepared for all residues other than Gly and Pro. b between Cu2+ and Zn2+ ions, spanning 6.12 and 6.33 Å for monomers A and B, respectively. In oxidized SODs, a typical Cu–Zn separation should be around 6.0 Å, whereas, for the reduced enzyme, this distance should be > 6.5 Å [26]; therefore, in this work, both TsCu ⁄ Zn-SOD monomers have been captured in the oxidized form. The Zn2+ ion is four coordinated with three His and one Asp (Fig. 4B, D) in a tetrahedral geometry. Interestingly, electron density maps show that TsCu ⁄ Zn-SOD contains three additional ions, one in monomer A and two in monomer B, coordinated to residues exposed to the solvent (Figs 3A and 5). We included Zn2+ ions in the latter FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS 3311 Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. Zn2 A Zn2 Zn1 Zn1 Cu Zn3 Cu Monomer A Monomer B L2 B C N Metal-binding loop L4 β1 β6 β3 β5 β4 α2 L5 Electrostatic loop L7 Greek-key loop L6 β2 β7 β8 α1 Interface analysis L3 L1 Fig. 3. Structure of Taenia solium Cu ⁄ Zn superoxide dismutase (TsCu ⁄ Zn-SOD). (A) Ribbon representation of the dimeric TsCu ⁄ ZnSOD where the metals are shown as spheres in yellow (copper) and grey (zinc). (B) Superposition of TsCu ⁄ Zn-SOD monomers. Each one contains an eight-strand b sandwich (b1–b8) with seven loops (L1–L7) (canonical nomenclature is included). positions on the basis of metal coordination, results from dynamic light-scattering experiments with different ions and appropriate behaviour during refinement. Space group symmetry expansion shows that, in monomers A and B, additional metal ions (Zn2) are coordinated intramolecularly by two His residues (64 and 107) and a solvent molecule. These ions stabilize two very flexible regions: the metal binding (L4) and the Greek-key (L6) loops, as shown in Figs 3A and 5C. This could explain the higher temperature stability observed when metal ions are added to the protein solutions. The third ion, which is located in monomer B (Zn3), is pentacoordinated intermolecularly by residues His29, Glu31 and His29 of a symmetry-related molecule, and O1 and O3 of a glycerol molecule which is present as the fourth ligand. These results are in line with the proposal that physiological Zn binding by 3312 metal-sequestering proteins is necessary to ensure cell homeostatic control [27]. Another interesting feature of the protein is the entrance of the channel connecting the active site, which shows that T. solium and S. mansoni Cu ⁄ Zn-SODs differ from the human enzyme. This area, known as the electrostatic loop, contains several highly conserved charged residues and has been proposed to be responsible for the long-range routing of superoxide towards the catalytic site [23]. In human Cu ⁄ Zn-SOD, this area is positively charged, whereas, in TsCu ⁄ Zn-SOD, nonpolar residues predominate. These amino acids play an important role in the conformation and charge distribution for substrate attraction. Indeed, in vitro studies have shown that the TsCu ⁄ Zn-SOD enzyme is completely inhibited by 300 mM albendazole, whereas this compound does not affect the bovine enzyme [12]. This could be explained by the hydrophobic nature of albendazole, which interacts more favourably with the hydrophobic amino acid residues at the entrance of the electrostatic loop. These studies suggest that TsCu ⁄ Zn-SOD could be used as a target protein to design agents for the treatment of cysticercosis. Eukaryotic Cu ⁄ Zn-SODs are dimeric structures with conserved subunit interfaces. The interaction of the two monomers is based on a contact region defined by four clusters. The first consists of the b1 and b2 strands of each monomer, the second is located in the loop formed between the b4 strand and helix 1, the third is formed by residues located in the L6 (Greekkey) loop and the fourth is formed by residues in the b8 strand (C-terminal). The Taenia (this work), Schistosoma (PDB entry 1TO4) and human (PDB entry 2V0A) Cu ⁄ Zn-SOD interfaces are very similar and consist of the same clusters; however, they vary slightly in their amino acid composition. The Ca atoms of the three structures were superimposed, showing rmsd values ranging from 0.37 to 0.49 Å in monomers A and B, respectively. The most important difference is observed when the dimers of these three SODs (Fig. 6) are compared. They superimpose with rmsd values of 0.64 and 1.02 Å for human [28] and Schistosoma [11] Cu ⁄ Zn-SODs, respectively. Interestingly, although the sequence identities of TsCu ⁄ Zn-SOD with the Schistosoma and human Cu ⁄ Zn-SODs are 71% and 57.9%, respectively, structurally the Taenia enzyme is similar to the human enzyme. The three enzymes present four conserved hydrogen bonds among residues Ile148, Gly48 and Gly111 FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. His43 A H2O 2.48 Cu 2.28 His117 His60 B His68 2.21 2 07 2.07 2.33 Zn 2.12 His60 2.23 2.09 2.31 Asp80 His45 His77 C His60 D His43 2.09 Fig. 4. Close-up view of the active site in the Taenia solium Cu ⁄ Zn superoxide dismutase (TsCu ⁄ Zn-SOD). (A, C) Coordination environment of Cu2+ ions in monomers A and B. (B, D) Zinc-binding residues in monomers A and B. The interatomic distances between coordination residues are shown. His60 2.21 Cu 2.47 2 91 2.91 His117 Asp80 2.04 2.18 Zn 2.21 2.05 His68 2 2.22 22 H2O His77 His45 A B His64A His107A 2.13 Zn His64B 2.13 His107B 2.34 2.33 2.15 Zn 1.94 H2O H2O C Fig. 5. Details of the additional metal-binding sites. 2Fo ) Fc (blue) and Fo ) Fc (green) electron density maps of these sites. (A, B) Zinc ion is coordinated by His107, His64 and a water molecule in monomers A and B. (C) Stabilization of L4 and L6 loops by metal ion coordination. (D) Zn2+ ion coordinated with His29, Glu31 of monomer B, His29 of a symmetry-related molecule A and a glycerol molecule. The interatomic distances between coordinating residues and solvent water molecules are indicated. D Glycerol H2O His64A His29A-Symmetry H 1.99 Zn His107A 2.23 2.23 His64A Metal-binding loop (L4) (numbering of TsCu ⁄ Zn-SOD) at their interfaces. In all cases, monomer A has a greater contact area relative to monomer B, which is more mobile. These differences are highlighted in Fig. 6, in which it is clear that the orientation between the two monomers in the TsCu ⁄ Zn-SOD and Schistosoma Cu ⁄ Zn-SOD dimers 2.46 Zn 2.19 His64A Greek-key loop (L6) differs by about 17, whereas this difference for human Cu ⁄ Zn-SOD is only about 6. Hough et al. [29] have reported this dimer interface alteration in the crystal structures of two mutants (Ala4Val and Ile113Thr) of Homo sapiens Cu ⁄ Zn-SOD, confirming that they are significantly destabilized in comparison with wild-type FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS 3313 Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. Monomer B Monomer A symmetry 17° Fig. 6. Structural alignment of Cu ⁄ Zn superoxide dismutases from Taenia solium (blue), Schistosoma (red) and Homo sapiens (green). Crystal structures were superimposed onto one monomer (left). Trp32 SOD. The Ala4Val mutant accounts for approximately 50% of the SOD-linked ALS cases [30]. Similar results have been observed for other amyloidogenic proteins, such as light chains, which, after alteration of their interfaces, show an increased tendency to form amyloid fibre [31,32]. Ala30 His29 Gly33 Zn Glu31 Ser34 Intermolecular packing interactions Analysis of the intermolecular packing contacts reveals interesting interactions (Fig. 7). The most relevant contact is observed with monomer B, in which a parallel b interaction with a symmetry-related molecule is established involving Zn2+ ion coordination (top inset). The strands implicated in this interaction generate a periodic stack with strands that are perpendicular to the direction of the longitudinal array (Fig. 7). Notably, when we superimpose this region on the human enzyme, the latter presents a b-bulge-like structure in strand b2 (Fig. 7, bottom inset), and a difference of about 2.8 Å is observed between the Ca atoms of these strands in both proteins. Several reports have indicated that proteins can acquire structural adaptations that enable them to avoid undesired protein aggregation and fibril formation, and b bulges are considered to be anti-aggregation motifs [33]. It is worth mentioning that the native human SOD has less tendency to aggregate than ALS mutants [28], whereas the TsCu ⁄ Zn-SOD oligomerizes, probably because the latter does not present the b bulge (Fig. 7, bottom inset) and therefore establishes the parallel b interactions mediated by metal ions coordinated to several residues (His and Glu) exposed to the solvent. This type of interaction has not been described previously 3314 Fig. 7. Crystal packing interactions of Taenia solium Cu ⁄ Zn superoxide dismutase (TsCu ⁄ Zn-SOD). b–b interactions through ion metal coordination are established with symmetry-related molecules. The Zn2+ ion involved is shown as a purple sphere and the residues implicated in the metal ion coordination are shown as sticks. The top inset shows a close-up view of the b–b arrangement. The bottom inset shows the b-bulge-like structure for the human enzyme (green), but not for TsCu ⁄ Zn-SOD (orange). for other SODs. Nonetheless, the metal-mediated self-assembly of natural [34] and engineered [35] proteins has already been reported in several examples in which blocks with noninteracting surfaces were assembled into aggregates by metal coordination. Therefore, mutants of human SOD containing His, Glu or Asp residues exposed to the solvent could present this behaviour, and such chemical control of protein–protein interactions might be physiologically relevant or be involved in neurodegenerative diseases. Protein oligomerization, aggregation and the formation of insoluble amyloid deposits are commonly observed in neurodegenerative diseases, but the factors initiating and modulating the abnormal interactions that lead to oligomerization remain unknown. Metal ions have been implicated in these phenomena, but the FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. structural basis for their involvement remains unclear [36]. Huang et al. [37] have shown evidence that metals are the initiators of the oligomerization and amyloid fibril formation of Ab in Alzheimer’s disease. Interestingly, Cherny et al. [38] have demonstrated that homogenization of Alzheimer’s disease brains in buffer containing chelators liberates more Ab than buffer alone, indicating that metal ions are key components in maintaining the structural integrity of amyloid deposits. Later, Dong et al. [39], using Raman spectroscopy, demonstrated that Cu2+ and Zn2+ bind Ab subunits in brain plaque amyloid through His residues. In addition, it has been observed that tissues affected by ALS are rich in metal ions, including Zn2+ ions [40]. In humans, ALS is a progressive neurodegenerative disorder selectively affecting motor neurones, in which 2% of the total cases are associated with mutations in the gene coding for the enzyme Cu ⁄ Zn-SOD. The causes of motor neurone death in ALS are poorly understood in general, but, for Cu ⁄ Zn-SOD-linked familial ALS, aberrant oligomerization of SOD mutant proteins has been strongly implicated. Several authors have suggested that metal-free human SOD is prone to aggregation as a result of conformational changes that can even form abnormal disulfide bridges [21,36]. Based on our findings, we suggest that TsCu ⁄ Zn-SOD could be useful as an alternative model for the interpretation of the mechanisms involved in protein aggregation and the formation of insoluble fibrillar deposits that are commonly observed in neurodegenerative diseases. Materials and methods Protein expression, purification and crystallization Recombinant TsCu ⁄ Zn-SOD protein was expressed and purified as described previously [12]. Briefly, transformed bacteria containing the pRSET vector with the coding region for TsCu ⁄ Zn-SOD were induced using 1.0 mM isopropyl thio-b-D-galactoside (IPTG), 0.2 mM CuSO4 and 0.17 mM ZnSO4. Cells were harvested by centrifugation, and the bacterial pellet was sonicated in 10 mM Tris ⁄ acetate, pH 7.5, with 0.1 mM phenylmethanesulfonyl fluoride and 0.75 M sucrose. The suspension obtained was centrifuged at 11 000 g to give a clear supernatant that was applied to a HiPrep 16 ⁄ 10 DEAE FF column. Bound proteins were eluted using a linear saline gradient. Fractions with Cu ⁄ Zn-SOD activity were pooled, dialysed in 50 mM Tris ⁄ HCl, pH 8.9, and applied to the same column. Final fractions were dialysed against Tris buffer and concentrated. Crystallization experiments were carried out at room temperature by the hanging-drop vapour diffusion technique. Drops contained a 1 : 1 volume ratio of the recombinant TsCu ⁄ Zn-SOD (4 mgÆmL)1) and precipitant solution 45 from Crystal Screen 2 (Hampton Research, Aliso Viejo, CA, USA), which includes 20% w ⁄ v polyethylene glycol monomethyl ether 2000 in hexahydrate 0.1 M Tris buffer, pH 8.5, and 0.01 M nickel(II) chloride. Determination of Cu ⁄ Zn-SOD activity Cu ⁄ Zn-SOD activity was determined indirectly by the inhibition of cytochrome c reduction at 25 C. This method uses the xanthine–xanthine oxidase system to generate superoxide radical. The total volume of the reaction was 1 mL in 50 mM phosphate buffer, pH 7.8, containing 1.0 mM EDTA and 57 mU of xanthine oxidase, 10 lM cytochrome c and 50 lM xanthine. The concentrations of Cu ⁄ Zn-SOD used for the assay were from 0 to 5 lgÆmL)1. One unit of SOD activity is defined as the amount causing 50% inhibition of the reduction of cytochrome c [41]. Effects of temperature and metal ions (Cu2+ or Zn2+) on the thermal stability of TsCu ⁄ Zn-SOD The effect of temperature on enzyme stability was determined in 50 mM Tris ⁄ HCl, pH 7.8, incubating TsCu ⁄ ZnSOD (100 UÆmL)1) in a water bath at temperatures of 10, 25, 37, 80 and 100 C. We added 1.0 mM CuSO4 or ZnSO4 to the protein solution to determine their effect on protein thermal stability. Aliquots of 200 lL at different times (0, 5, 10, 20 and 30 min) were transferred to a 4 C bath and assayed by the xanthine–xanthine oxidase method. Residual activity was determined after 30 min. As controls for the assay, TsCu ⁄ Zn-SOD with buffer and without metals was used. Monitoring of TsCu ⁄ Zn-SOD aggregation by gel filtration TsCu ⁄ Zn-SOD (0.5 mgÆmL)1) aliquots were incubated with different ZnSO4 concentrations from 0.5 to 1.0 mM. Onehundred-microlitre aliquots of the enzyme samples were applied to a size exclusion Superdex 75 HL 16 ⁄ 60 column (GE Healthcare, Sweden) in a fast protein liquid chromatography system. The column was equilibrated with 0.1 M Tris, 200 mM NaCl buffer, pH 8.5, and the flow rate was set to 1.0 mLÆmin)1. Elution of the species formed during incubation was detected by monitoring the absorbance at 280 nm. The column was calibrated with standards of known molecular mass (ubiquitin, 8.5 kDa; cytochrome C, 12 kDa; hen white lysozyme, 14.3 kDa; thaumatine, 24 kDa; bovine serum albumin, 66.4 kDa). FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS 3315 Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. Characterization of TsCu ⁄ Zn-SOD using matrixassisted laser desorption ionization time-of-flight mass spectrometry The enzyme alone or in the presence of ZnSO4 at different concentrations was mixed with sinapinic acid in 30% acetonitrile, 70% water and 0.1% trifluoroacetic acid, and analysed in a MICROFLEX matrix-assisted laser desorption ionization time-of-flight instrument (Bruker Daltonik GmbH, Leipzig, Germany) equipped with a 20-Hz nitrogen laser. Spectra were recorded in the positive linear mode for the mass range 2000–70 000 Da. Dynamic light-scattering measurements These data were obtained using a Zetasizer Nano S dynamic light-scattering device from MALVERN Instruments at 20 C and fitted using DTS software from Malvern Instruments. The TsCu ⁄ Zn-SOD filtered samples (pore size, 0.22 lm) were placed in a quartz cuvette (50 lL) and used to test the effect of metal ion (CuSO4, ZnSO4 and NiCl2) addition on the oligomerization behaviour. The buffer was filtered (pore size, 0.22 lm) immediately before use and care was taken to reduce contamination of the samples by dust. At least 25 measurements were collected for each sample. Size distributions in percentage volume were calculated using MALVERN Instruments software by approximating the protein as a spherical object. Data collection and processing Diffraction data were collected using synchrotron radiation at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory (Upton, NY, USA) on beamline X12B, with an ADSC Quantum-4 CCD detector at 100 K. The crystal was cryoprotected with 15% glycerol in the mother liquor. The dataset was indexed and integrated with XDS [42] and scaled with SCALA [43], contained in the CCP4 crystallographic package [44]. A summary of data collection and processing is given in Table 1. Structure determination and refinement Analysis of the unit cell content suggested the presence of two protein molecules in the asymmetric unit, consistent with a solvent content of 42.78% [22]. The structure was solved by molecular replacement using the program Phaser [45] as implemented in PHENIX [46], and, as a template, the crystal structure of S. mansoni Cu ⁄ Zn-SOD at 1.55 Å resolution (PDB entry code 1TO4 [11]) was used, which shares 71% identity with TsCu ⁄ Zn-SOD. The solution of the molecular replacement gave a final Z score of 36.5 and a log likelihood gain of 1405. Refinement was carried out with PHENIX [46] using a random test set of 10% of the reflections for cross-validation. Briefly, we used a rigid 3316 body refinement, followed by simulated annealing and successive rounds of Cartesian and temperature factor minimization with manual model building in Coot [47]. All of the active-site metals, as well as three additional metal sites, were visible in the electron density maps from the first stages of refinement. Noncrystallographic symmetry constraints were not imposed to ensure that potential structural differences between the monomers were not doubtful. Water molecules were added to the model near the end of the refinement by a search procedure based on peaks observed in the difference maps and bond distance criteria. PROCHECK [48] was used for the analysis of the stereochemistry of the model and validation. Molecular comparisons with other SODs were performed with the ALIGN program [49] and figures prepared with Pymol [50] and Chimera [51]. Statistics on data collection and refinement are reported in Table 1. PDB accession number The atomic coordinates and structure factors have been deposited in the Protein Data Bank with accession number 3MND. Acknowledgements This work was supported in part by grants from Consejo Nacional de Ciencia y Tecnologı́a (contracts 82947 to A.R.-R. and 80134 to A.L.) and Dirección General de Asuntos del Personal Académico, Universidad Nacional Autónoma de México (IN207507-3 to A.L.). The characterization of crystals was performed at Laboratorio Nacional de Estructura de Macromoléculas, Instituto de Quı́mica, Universidad Nacional Autónoma de México. X-Ray data were measured at beamline X12B of the National Synchrotron Light Source. We thank Alexei Soares for his help during data collection. The National Synchrotron Light Source is supported by the US Department of Energy through contract no. DEAC02-98CH10886. We also thank M. C. Georgina Espinosa for technical assistance. References 1 Fridovich I (1998) Oxygen toxicity: a radical explanation. J Exp Biol 201, 1203–1209. 2 Fattman CL, Schaefer LM & Oury TD (2003) Extracellular superoxide dismutase in biology and medicine. Free Radic Biol Med 35, 236–256. 3 Tainer JA, Getzoff ED, Beem KM, Richardson JS & Richardson DC (1982) Determination and analysis of the 2 A-structure of copper, zinc superoxide dismutase. J Mol Biol 15, 181–217. FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS Metal-mediated self-assembly of T. solium Cu ⁄ Zn-SOD A. Hernández-Santoyo et al. 4 Henkle-Dührsen K & Kampkötter A (2001) Antioxidant enzyme families in parasitic nematodes. Mol Biochem Parasitol 114, 129–142. 5 Antonyuk SV, Strange RW, Marklund SL & Hasnain SS (2009) The structure of human extracellular copper– zinc superoxide dismutase at 1.7 Å resolution: insights into heparin and collagen binding. J Mol Biol 388, 310– 326. 6 Richardson JS, Thomas KA, Rubin BH & Richardson DC (1975) Crystal structure of bovine Cu,Zn superoxide dismutase at 3Å resolution: chain tracing and metal ligands. Proc Natl Acad Sci USA 72, 1349–1353. 7 Tainer JA, Getzoff ED, Richardson JS & Richardson DC (1983) Structure and mechanism of copper, zinc superoxide dismutase. Nature 306, 284–287. 8 Shaw BF & Valentine JS (2007) How do ALS-associated mutations in superoxide dismutase 1 promote aggregation of the protein? Trends Biochem Sci 32, 78–85. 9 Valdivia A, Pérez-Alvarez S, Aroca-Aguilar JD, Ikuta I & Jordán J (2009) Superoxide dismutases: a physiopharmacological update. J Physiol Biochem 65, 195–208. 10 Vaca-Paniagua F, Torres-Rivera A, Parra-Unda R & Landa A (2008) Taenia solium: antioxidant metabolism enzymes as targets for cestocidal drugs and vaccines. Curr Top Med Chem 8, 393–399. 11 Cardoso RM, Silva CH, Ulian de Araújo AP, Tanaka T, Tanaka M & Garratt RC (2004) Structure of the cytosolic Cu,Zn superoxide dismutase from Schistosoma mansoni. Acta Crystallogr D Biol Crystallogr 60, 1569– 1578. 12 Castellanos-González A, Jiménez L & Landa A (2002) Cloning, production and characterisation of a recombinant Cu ⁄ Zn superoxide dismutase from Taenia solium. Int J Parasitol 32, 1175–1182. 13 Montresor A & Palmer K (2006) Taeniasis ⁄ cysticercosis trend worldwide and rationale for control. Parasitol Int 55, Suppl–S301. 14 Marklund SL (1982) Human copper-containing superoxide dismutase of high molecular weight. Proc Natl Acad Sci USA 79, 7634–7638. 15 Lin CW, Yang JH & Su LC (1997) The extraction and properties of superoxide dismutase from porcine blood. Meat Sci 46, 303–312. 16 Benov L, Sage H & Fridovich I (1997) The copper-and zinc-containing superoxide dismutase from Escherichia coli: molecular weight and stability. Arch Biochem Biophys 340, 305–310. 17 Lomakin A, Benedek GB & Teplow DB (1999) Monitoring protein assembly using quasielastic light scattering spectroscopy. Methods Enzymol 309, 429–459. 18 Bush AI, Pettingell WH, Multhaup G, Paradis MD, Vonsattel JP, Gusella JF, Beyreuther K, Masters CL & Tanzi RE (1994) Rapid induction of Alzheimer Ab amyloid formation by zinc. Science 265, 1464–1467. 19 Esler WP, Stimson ER, Jennings JM, Ghilardi JR, Mantyh PW & Maggio JE (1996) Zinc-induced aggregation of human and rat b-amyloid peptides in vitro. J Neurochem 66, 723–732. 20 Stellato F, Menestrina G, Dalla-Serra M, Potrich C, Tomazzolli R, Meyer-Klaucke W & Morante S (2006) Metal binding in amyloid b-peptides shows intra- and inter-peptide coordination modes. Eur Biophys J 35, 340–351. 21 Seetharaman SV, Prudencio M, Karch C, Holloway SP, Borchelt DR & Hart PJ (2009) Immature copper–zinc superoxide dismutase and familial amyotrophic lateral sclerosis. Exp Biol Med 234, 1140–1154. 22 Matthews BW (1968) Solvent content of protein crystals. J Mol Biol 33, 491–497. 23 Getzoff ED, Tainer JA, Weiner PK, Kollman PA, Richardson JS & Richardson DC (1983) Electrostatic recognition between superoxide and copper, zinc superoxide dismutase. Nature 306, 284–287. 24 Banci L, Bertini I, Hallewell RA, Luchinat C & Viezzoli MS (1989) Water in the active cavity of copper ⁄ zinc superoxide dismutase. A water 1H-nuclear-magneticrelaxation-dispersion study. Eur J Biochem 184, 125– 129. 25 Ascone I, Castañer R, Tarricone C, Bolognesi M, Stroppolo ME & Desideri A (1997) Evidence of His61 imidazolate bridge rupture in reduced crystalline Cu,Zn superoxide dismutase. Biochem Biophys Res Commun 241, 119–121. 26 Hough MA & Hasnain SS (2003) Structure of fully reduced bovine copper zinc superoxide dismutase at 1.15 Å. Structure 11, 937–946. 27 Palmiter RD & Findley SD (1995) Cloning and functional characterization of a mammalian zinc transporter that confers resistance to zinc. EMBO J 14, 639–649. 28 Strange RW, Yong CW, Smith W & Hasnain SS (2007) Molecular dynamics using atomic-resolution structure reveal structural fluctuations that may lead to polymerization of human Cu–Zn superoxide dismutase. Proc Natl Acad Sci 104, 10040–10044. 29 Hough MA, Grossman JG, Antonyuk SV, Strange RW, Doucette PA, Rodriguez JA, Whitson LJ, Hart PJ, Hayward LJ, Valentine JS et al. (2004) Dimer destabilization in superoxide dismutase may result in diseasecausing properties: structures of motor neuron disease mutants. Proc Natl Acad Sci 101, 5976–5981. 30 Cudkowicz ME, McKenna-Yasek D, Sapp PE, Chin W, Geller B, Hayden DL, Shoenfeld DA, Hosler BA, Horvitz H & Brown RH (1997) Epidemiology of mutations in superoxide dismutase in amyotrophic lateral sclerosis. Ann Neurol 41, 210–221. 31 Baden EM, Owen BA, Peterson FC, Volkman BF, Ramirez-Alvarado M & Thompson JR (2008) Altered dimer interface decreases stability in an amyloidogenic protein. J Biol Chem 283, 15853–15860. FEBS Journal 278 (2011) 3308–3318 ª 2011 The Authors Journal compilation ª 2011 FEBS 3317