Structural characterization of an iron–sulfur cluster assembly protein IscU in a zinc‐bound form

Abstract

Iron–sulfur (Fe-S) clusters are simple inorganic prosthetic groups widely distributed in nature. Proteins that contain Fe-S clusters play essential roles in diverse biological processes including electron transfer, gene regulation, environmental sensing, and substrate activation.1 Although it is possible to assemble Fe-S clusters into proteins from inorganic sulfide and iron in vitro, the biogenesis of Fe-S cluster in vivo, however, appears to be facilitated by proteins rather than spontaneous formation. Recent studies have led to the discovery that a highly conserved gene cluster iscSUA-hscBA-fdx is essential for the biogenesis of Fe-S cluster proteins in bacteria.2 Proteins encoded by this gene cluster include: IscS, IscU, IscA, HscA, HscB, and Ferredoxin (Fd). Homologies of these proteins have also been identified in eukaryotic organisms, indicating a conserved mechanism for the biogenesis of Fe-S proteins.3, 4 Further biochemical studies have revealed detailed roles of these Fe-S cluster proteins. IscS, a homolog of NifS, is a homodimeric pyridoxal phosphate-dependent cysteine desulfurase. It catalyzes the desulfurization of L-cysteine to L-alanine and provides S2− to IscU for Fe-S cluster assembly.5, 6 IscU provides a scaffold for the assembly of a nascent iron–sulfur cluster prior to its delivery to apo Fe-S proteins.7-9 IscA is proposed to function as an alternative scaffold for Fe-S cluster assembly, but the exact role of this protein is still not determined.10 HscA and HscB are molecular chaperones that can selectively bind IscU and assist in the biogenesis of Fe-S proteins. The detailed roles of these molecular chaperones are also not clear.11 Sequence comparisons suggest IscU is a highly conserved protein and is homologous to the N-terminal domain of NifU (nNifU), an essential protein for nitrogen fixation. IscU/nNifU contains three strictly conserved cysteine residues.12 Site-directed mutagenesis data suggest that all three cysteine residues are essential for the function of IscU/nNifU proteins.13, 14 Biochemical assay showed IscS can form a covalent complex with IscU through residue Cys328 in IscS and residue Cys63 in IscU in Escherichia coli.6 Characterization of Fe-S cluster assembly on IscU or ISU (eukaryotic IscU) from several organisms including Azotobacter vinelandii, E. coli, Thermotoga maritima (Tm) and human have been carried out.4, 6-8 In presence of IscS/NifS, D-cysteine, Fe2+ and reducing agent, IscU/nNifU is able to assemble a transient [Fe2S2]2+ cluster in vitro. All three cysteine residues are proposed to be involved in the coordination of Fe-S cluster with possible one noncysteinyl ligation. Substitution of a highly conserved residue Asp40 (in Tm) or Asp37 (in human) substantially stabilizes the labile IscU-[Fe2S2]2+ complexes, suggesting a critical role of this residue in the Fe-S cluster formation of IscU.8, 15 The interactions between IscU and molecular chaperones (HscB and HscA) have also been extensively studied. IscU has been shown to be a substrate of HscA, with HscA binding IscU specifically through a conserved LPPVK motif in IscU.11 Structural studies of IscU/nNifU proteins can provide critical insight for understanding the molecular functions of these proteins, but have been proved to be challenging. An NMR structural characterization of Tm IscU protein revealed that IscU protein contains six α-helices and a three-β-stranded anti-parallel β sheet, but the detailed three-dimensional structure is not determined due to widely different conformational arrangements.16 At the Berkeley Structural Genomics Center (http://www.strgen.org), about ten homologs of IscU protein from different organisms were selected as targets, and crystals were obtained from Streptococcus pyogenes (Sp). Here we report the crystal structure of the IscU protein from Sp at 2.3 Å resolution in a zinc-bound form. The Sp_IscU crystal structure revealed detailed structural information for the IscU active site. Recently, an NMR structure of IscU protein from Haemophilus influenzae in apo-form has become available in the Protein Data Bank (Hi_IscU).17 Structural comparisons between the two structures suggest possible conformational changes in the active site of IscU protein upon zinc binding. The S. pyogenes gene, gi:15674463, was amplified by PCR using genomic DNA template [ATCC 12344] and primers [Integrated DNA Technologies] designed for Ligation-Independent Cloning (LIC).18 The N-terminal primer had the sequence: GGCGGTGGTGGCGGCATGGCACTCTCTAAACTGAACCATCTATAC and the C-terminal primer had the sequence: GTTCTTCTCCTTTGCGCCCCTAGACATTTTTCCCTTCCTTTACATTTTG. The amplified PCR product was prepared for vector insertion by purification, quantitation, and treatment with T4 DNA polymerase (NEB) in the presence of 1 mM dTTP. The prepared insert was annealed into the LIC expression vector pB4, a derivative of pET21a (Novagen) that expresses the cloned gene fused with an N-terminal 6-His-Maltose Binding Protein-Tobacco Etch Virus (TEV) cleavage sequence, and transformed into chemical-competent DH5alpha to obtain fusion clones. The seleno-L-methionine (Se-Met) isoform of Sp_IscU protein was expressed in E. coli B834 (DE3). Cell paste was prepared by using Studier's auto-induction method (Dr. William Studier, Brookhaven National Laboratory, personal communication). A single colony of pB4-Sp_IscU was used to inoculate 10 ml of PA-0.5G medium and grow overnight at 37°C. The overnight culture was diluted 1 to 1000 into PASM-5052 auto-inducing medium. For protein purification, 20 g of cell paste was resuspended in 250 ml 50 mM Hepes buffer, pH 7.0, 0.5 M NaCl, 1 mM DTT and 5% glycerol (Buffer A). Cells were opened with a Microfluidizer (Microfluidics, Newton, MA). After centrifugation, the cleared cell lysate was loaded onto a 5-ml HiTrap Chelating HP column (GE Healthcare, Piscataway, NJ). The column was initially washed with Buffer A followed by a second wash of Buffer A containing 40 mM imidazole, then the fusion protein was eluted with an imidazole gradient from 40 to 300 mM in 10 column volumes. To cleave the fusion protein, 3 mg of mTEV was added to the 125 mg of eluted fusion protein. The protein solution was then dialyzed against Buffer A at 4°C overnight. Following mTEV cleavage and dialysis, the protein solution was loaded onto a 5-ml HiTrap Chelating HP column again and washed with Buffer A for 10 column volumes, and the target protein was eluted with Buffer A containing 20 mM imidazole. As a polishing step, the eluted protein was concentrated to 25 mg/ml and loaded onto a Superdex-75 size exclusion column (GE Healthcare). The peak fractions were then pooled and concentrated to 8 mg/ml in 20 mM Hepes, pH 7.5, 300 mM NaCl, and 1mM DTT for crystallization. Inductively Coupled Plasma–Mass Spectrometry (ICP-MS) for this protein was conducted at UGA research center (http://www.ors.uga.edu/cal/). Crystal screening was done by the sparse matrix screening method.19 Optimum conditions were obtained by mixing 1 μl Se-Met protein solution (8 mg/ml) with 1 μl reservoir solution containing 1.9 M Sodium Malonate and equilibrating against 1 ml of reservoir solution. Large crystals were obtained within 2–3 days at room temperature. For data collection, Se-Met crystals of Sp_IscU were transferred to a solution containing 20 mM Hepes, pH 7.5, 300 mM NaCl, and 3.5 M sodium malonate and flash-frozen in liquid nitrogen. Diffraction data were collected at the Advanced Light Source (ALS) Beamline 5.0.2 (Lawrence Berkeley National Laboratory, Berkeley) at 100°K with a Quantum 4 charge-coupled device (CCD) detector and processed with HKL 2000.20 Data statistics are summarized in Table I. The positions of three out of five selenium sites were identified from MAD data using SOLVE21 at 2.5-Å resolution with a figure of merit of 0.63 and Z-score of 16.17. The initial sites were put into SHARP for refinement, phase extension to 2.3-Å resolution, and density modification.22 The initial model building was done by RESOLVE with the build-only option.21 Further model building was done manually with the resultant map from SHARP using the program O.23 A zinc ion was built into the model according to the anomalous difference map and ICP-MS results. After the initial model building, the structure was refined against the merged MAD data to Rfree 25.8% and Rfactor 21.9 using CCP4 refmac5 TLS method.24 The refined model of this protein contains 135 amino acids, one zinc ion, and 22 water molecules. The density for the N-terminal six residues and the C-terminal 18 residues is missing and these residues are not included in the current model. A ribbon diagram of the Sp_iscU monomer structure is shown in Figure 1(a). The structure of Sp_iscU is a single domain structure with overall dimensions of 31 Å × 40 Å × 27 Å. The overall fold of this protein belongs to αβ fold with one layer of β sheet on one side and an α-helical structure on the other side. The β-sheet is composed of three anti-parallel β strands (β1↑ β2↓ β3↑). In these three β-strands, β1 (residues 33 to 36) is shorter and more solvent-exposed than β2 (residues 43 to 49) and β3 (residues 57 to 63). The α-helical structural part is composed of an N-terminal α-helix (α1), a three α-helix bundle (α2, α3, and α6), and two short α-helices (α4 and α5). The N-terminal α-helix (α1, residues 8–19) is located perpendicularly to the beginning of α2 and is connected to β1 through a long flexible loop (Loop 1, Residues 20–32) which wraps the outside of the β-sheet. In the three α-helix bundle, α2 (residues 66–79) and α6 (residues 122–139) form extensive interactions with the three β-strands and form the core structural region in the protein. The two short α-helices (α4, α5) are located between α3 and α6, and the sequence for this part of the structure is conserved in a few bacterial organisms such as Tm, but not present in eukaryotic organisms and most bacteria. a: Ribbon diagram of the Sp_IscU monomer. Helices and β-strands are shown in blue and green, respectively. The polypeptide termini are indicated with N and C. Zinc ion is depicted as a sphere in red. b: Active site of Sp_IscU. Zinc ion is shown as a dark red sphere. Residues Cys40, Asp42, Cys65, Arg124, and Cys127 are shown as ball-and-stick models. The anomalous difference electron density map around the zinc ion is shown in pink at 16.0 σ. The presence of zinc ion in the protein was determined based on ICP-MS results and anomalous difference electron density map. The ICP-MS results showed that the possible metal element in the Sp_IscU protein is zinc ion with a ratio of signal over background of 33. One data set was collected on Sp_IscU Se-Met crystals at the zinc peak wavelength of 1.2828 Å. The statistics for this data set are shown in Table I. The anomalous difference electron density map indicates the presence of zinc ion bound at the active site [Fig. 1(b)]. This zinc binding site is located at the top of the structural core which is composed of three anti-parallel β strands and two α-helices (α2 and α6) [Fig. 1(b)]. This zinc binding site is also on the protein surface and is solvent accessible. The bound zinc ion is coordinated by the three conserved cysteine residues (Cys40, Cys65, and Cys127) and another conserved residue, Asp42. The coordination between zinc ion and the S atoms of the three cysteine residues and OD1 group of the Asp42 is tetrahedral with an average distance of approximately 2.5 Å. Regions that participate in the zinc ion binding site include Loop 2 which connects β1 and β2, Loop 3 which connects β3 and α2, and the upper part of α6. Loop 2 consists of residues NNPTCGD (Residues 36–42) in which PXCGD is a highly conserved motif among IscU/NifU protein family (Fig. 2). In this motif, Cys40 is one of the three cysteine residues that are essential for function. Asp42 is another invariant residue in this motif which has been shown to substantially stabilize IscU-[Fe2S2]2+ complexes in both Tm and human IscU proteins when this residue is mutated to Ala.8, 15 Loop 2 is a partially disordered flexible loop. The flexibility of this loop can be accounted for by the high temperature factors of this loop. The electron density for the side chains of Asn36 and Asn37 in Loop 2 is missing and is broken between residue Pro38 and Thr39. Residues Cys40, Gly41, and Asp42 are reasonably well defined. Loop 3 is composed of residues Gly64 and Cys65, both are conserved residues among IscU/NifU protein families. Cys65 (Cys63 in E. coli) is another functionally essential cysteine residue that has been shown to form a disulfide bond with a cysteine residue in the active site of IscS in E coli.6 The electron density for the two residues in Loop 3 is very well defined. Cys127 is the third functionally essential cysteine residue. This residue is located on the upper helix of α6. This residue is also well defined in Sp_IscU structure. Multiple sequence alignment of IscU proteins from the three kingdoms of life. The three invariant cysteine residues are labeled in dark grey, other invariant residues are labeled in red, highly conserved residues are labeled in tan and generally conserved residues are labeled in blue. The secondary structures are labeled on top of the sequence alignment (arrows stand for β strands, squares stand for α-helices and dashes stand for loops). Another well-defined and potentially important residue around the zinc binding site is Arg124. This residue is highly conserved among IscU/NIfU protein families and is located at the beginning of α6. In the Sp_IscU structure, the side chain of Arg124 is positioned at one side of the zinc binding site and is at a close distance to Cys65 [Fig. 1(b)]. The distance between NH1 group of Arg124 and S atom of Cys65 is 3.65 Å. In E. coli IscU, lys103 (Arg124 in Sp) has been reported to be involved in the interactions between Ec_IscU and Ec_HscA.11 Mutating this residue to alanine has led to significant loss of binding affinity between Ec_IscU and Ec_HscA. Another possible role for Arg124 is the stabilization of the S2− during sulfur transferring from IscS to IscU because of its close distance to residue Cys65.6 The Dali program25 was used for searching for structural homologs of Sp_IscU protein. The search results showed that Sp_IscU is structurally homologous to a recently solved NMR structure of IscU protein from Haemophilus influenzae (Z score 7.6 and RMSD 2.4 Å over 95 Cα atoms). Since the Hi_IscU structure is in a metal free form, this structure is therefore referred to as an apo form of the IscU structure in this article. Sp_IscU and Hi_IscU share 31% sequence identity and 47% sequence similarity (Fig. 2). Closer scrutiny of the Dali results suggests these two IscU proteins share the same kind of structural fold and the structural conservation is greatest over the core structural region. Superpositions between these two structures using 70 Cα atoms in the core structural elements lead to a smaller RMSD of 1.69 Å [Fig. 3(a)]. a: Superposition of the Sp_IscU and Hi_IscU structures. Sp_IscU and Hi_IscU are shown in green and purple, respectively. b: Structural comparison of the active site between zinc-bound Sp_IscU and apo Hi_IscU. Sp_IscU is shown in green and Hi_IscU in purple. The arrow shows the possible movement of Loop 2 in Hi_IscU structure. According to these superpositions, the biggest structural difference between the zinc-bound Sp_IscU and apo Hi_IscU structures lies at the N-terminal region. In the IscU/NifU protein family, the sequence for the N-terminal region (Fig. 2, residues 1 to 21 for Hi_IscU and residues 6 to 29 for Sp_iscU) is highly conserved, but the structures for these two IscU proteins in this region are quite different. In Sp_IscU, residues 8–19 in the N-terminal region form an α-helix (α1) located at the same side as the other α-helices. This α-helix is connected to β1 by a long sequence-conserved loop that wraps the outside of the β sheet. In addition, the N-terminal region of Sp_IscU makes close contact with α2. In Hi_IscU structure, the N-terminal region forms a long loop. The direction of this loop is almost in the opposite direction of α1 of Sp_IscU and is entirely disassociated from other parts of the protein. Other differences between zinc bound Sp_IscU and apo Hi_IscU include the long α-helix α3, α4, and α5, and the C-terminal missing part. In the Hi_IscU structure, α3 of the Sp_IscU structure is replaced with two short α-helices and one of these two helices (residues 90–97) does not superpose well with α3 [Fig. 3(a)]. This difference could be accounted for by the sequence diversity in this region (Fig. 2). As mentioned before, the sequence for α4 and α5 in the Sp_IscU structure is not present in the Hi_IscU protein, so this part of the structure is not present in the Hi_IscU structure. In the Sp_IscU structure, the C-terminal 18 residues are missing. In Hi_IscU, this part of the structure is an extended α-helix of α6 and a C-terminal loop. Although Sp_IscU and Hi_IscU share high structural conservation in the core structural region, the detailed structure in the active center is quite different. One notable difference is that the three highly conserved cysteine residues in the apo Hi_IscU structure are much further apart from other to of the zinc bound Sp_IscU structure. In the Hi_IscU structure, the between the S atoms of and and and Cys63 and are and [Fig. In the Sp_IscU structure, these three are and Å. Closer of these two of suggests that these differences are most by the movement of Loop 2 due to the in the apo Hi_IscU structure is much apart from and Cys63 than that of the zinc-bound Sp_IscU structure Å the distance difference between Cys63 and is between these two structures Å the flexibility of Loop 2 this loop to the of residues and Cys63 conformational is located in the upper helix of the C-terminal α-helix and Cys63 is located in a This conformational the conformational flexibility of Loop 2 in the active site. Figure 2 shows the sequence alignment between Sp_IscU and other IscU proteins from three kingdoms including IscU is one of the most conserved proteins in the highly conserved and invariant residues onto the Sp_IscU structure suggests that most of the highly conserved residues the N-terminal in the IscU protein family are located in the around the zinc binding site and the core structural region. residues to form the core or are involved in The high sequence conservation in these structural that all IscU proteins and proteins are to the same kind of structural core and Fe-S cluster assembly of the surface of Sp_IscU structure revealed a highly conserved surface This surface is by conserved residues not Since IscS proteins have high this conserved be involved in IscS binding. IscU/nNifU proteins are to as a for the Fe-S cluster assembly in Fe-S protein it is not the Fe-S cluster onto the IscU/nNifU proteins. studies have shown that IscU/nNifU proteins are able to assemble a labile Fe-S cluster and results the that the labile [Fe2S2]2+ is coordinated by three cysteine residues and one 6-8 As to which residue in IscU/nNifU is involved in the [Fe2S2]2+ coordination is still not Biochemical showed that a highly conserved residue in to in Tm and human IscU can substantially stabilize the IscU-[Fe2S2]2+ complexes, suggesting a possible between the residue and [Fe2S2]2+ In the crystal structure of the Sp_IscU, a zinc ion was bound in the active site and coordinated by the three highly conserved cysteine residues and the residue results suggest the zinc binding site in Sp_IscU is to be the [Fe2S2]2+ cluster in IscU/nNifU proteins. This is also by the presence of a flexible and highly conserved loop (Loop that could the [Fe2S2]2+ cluster to bind in the zinc binding site. This active site structure of Sp_IscU also the that a single of the IscU protein is able to assemble a [Fe2S2]2+ This is with the reported biochemical data that a E. coli IscU protein was able to assemble a [Fe2S2]2+ cluster in The role of zinc binding on iron–sulfur cluster assembly on IscU proteins has not been studies have revealed that the three highly conserved cysteine residues in the NifU protein can a The for iron binding on IscU proteins are The Tm IscU has been to a high affinity to a report showed E. coli IscU does not bind or Further biochemical studies are essential to be able to zinc binding on IscU proteins has biological or it is an iron According to the sequence the N-terminal sequence of IscU/nNifU is highly The conserved include part of α1 and part of Loop 1 in the Sp_IscU structure. The structure of this region in zinc bound Sp_IscU and apo Hi_IscU structures are quite different. of surface of the Sp_IscU protein suggest that both and are involved in a highly conserved in the Sp_IscU protein. this is essential for IscS binding to the IscU protein, then the N-terminal of apo Hi_IscU protein has to over in to bind Further biochemical are to this for and for expression studies and cell paste for search of the gene, and for with also are to the at the Advanced Light Source which is by the of of of the of at Berkeley National to for to this