Crystal Structures of the Apo and Gdp-Bound Forms of A Cupin-Like Protein Bbduf985 from Branchiostoma Belcheri Tsingtauense

Yang Du,Yong-Xing He,Saren Gaowa,Xuan Zhang,Yuxing Chen,Shi-Cui Zhang,Cong-Zhao Zhou

DOI: https://doi.org/10.1002/prot.22771

2010-01-01

Proteins Structure Function and Bioinformatics

Abstract:Cupins are ubiquitous proteins existing in all three kingdoms of life and share a conserved β-barrel fold with a characteristic pocket located at the center of the β-barrel.1, 2 Despite sharing the highly conserved topology, cupins show remarkable variations in their sequences, architecture of domains (comprising either one or two cupin domains), quaternary assembly, and the nature of bound metal ion as well.3 In the Pfam database (http://pfam.sanger.ac.uk/), cupins are classified into 35 protein families, with greatly diversified functions such as isomerases, epimerase, dioxygenase, nonenzymatic storage proteins, and it is suggested that cupin is one of the most functionally diverse protein superfamily.2, 4 In recent years, large-scale sequencing of genomes and cDNAs has elicited a great number of hypothetical cupins whose biological functions need to be established. The cephalochordate amphioxus is a modern survivor of an ancient chordate lineage and acts as a living fossil between invertebrates and vertebrates. With the completion of the draft genome of Florida amphioxus Branchiostoma floridae in 2008, it was revealed that they are not only the most primitive chordates but also the genomic features appear to have a great deal in common with vertebrates as well.5 BbDUF985 is a hypothetical protein of 172 residues from Branchiostoma belcheri tsingtauense belonging to the cupin_5 family (Pfam entry: PF06172) of the cupin superfamily.4 Interestingly, although BbDUF985 is a protein from eukaryote, sequence homology analysis indicates that all its closest homologs are prokaryotic and more distantly kindred proteins from higher species. Previously, we solved the structure of YML079w (sharing 33% sequence identity with BbDUF985) from Saccharomyces cerevisiae in the guanosine triphosphate (GTP)-bound form,6 whereas the apo form structure is absent and further biochemical study was not undertook then. Here, we present the crystal structures of BbDUF985 in both the apo and guanosine diphosphate (GDP)-bound forms, which are the first pair of structures (with and without the ligand) from the cupin_5 family. Moreover, using fluorescence spectrometry, we determined the dissociation constant (KD) of GDP and investigated the spectra of BbDUF985 with several proposed ligands binding to the proteins in cupin superfamily. Combined with sequence analysis, we speculate that BbDUF985 and its homologs in the cupin_5 family might be involved in the nucleotide transport or metabolism. The coding sequence of BbDUF985 was cloned into a pET28a-derived vector. The recombinant protein with a hexa-histidine (6 × His) tag at the N-terminus was overexpressed in E. coli Rosetta (DE3) (Novagen, Madison, WI) strain using 2 × YT culture medium (16 g of trypton, 10 g of yeast extract, and 5 g of NaCl per liter). The cells were grown at 37°C up to an A600 nm of 0.6. Expression of recombinant BbDUF985 was induced at exponential phase with 0.2 mM isopropyl-β-D-thiogalactoside and cell growth continued for another 20 h at 16°C before harvesting. Cells were collected by centrifugation at 4000g for 20 min and resuspended in lysis buffer (20 mM Tris-Cl, pH 8.0, 150 mM NaCl). After 5 min of sonication and centrifugation at 12,000g for 30 min, the supernatant containing the soluble target protein was collected and loaded to a Ni-NTA column (GE Healthcare) equilibrated with binding buffer (20 mM Tris-Cl, pH 8.0, 150 mM NaCl). The target protein was eluted with 200 mM imidazole buffer and further loaded onto a Superdex 200 column (Amersham Biosciences) pre-equilibrated with 20 mM Tris-Cl, pH 8.0, 150 mM NaCl. Fractions containing the target protein were collected and concentrated to 15 mg/mL. The purity of protein was determined on SDS-PAGE and then the protein sample was stored at −80°C. The crystals of BbDUF985 were grown at 289 K with hanging drop vapor-diffusion techniques by mixing 1 μL of the 15 mg/mL protein sample with equal volume of reservoir. Apo-form crystals were obtained in the drop containing 2.0M ammonium sulfate, 0.1M sodium acetate trihydate pH 4.6, and reached to a maximal size for X-ray diffraction in 1 week. GDP-bound form crystals were obtained by soaking the apo-form crystal to a 10-μL crystallization reservoir containing 2 mM GDP molecule for about 30 min and then mounted for X-ray diffraction immediately. The diffraction image of the apo and GDP-bound forms were recorded at 100 K in a liquid nitrogen stream using beamline 3W1A at Beijing Synchrotron Radiation Facility (λ = 1.0000 Å) with MAR 165 mm CCD (MARresearch, Germany) and Rigaku MM007 X-ray generator (λ = 1.5418 Å) with MarRearch 345 image-plate detector (USTC, Hefei, China), respectively. Data were processed with MOSFLM 7.0.47 and scaled with SCALA.8 The crystal structures of BbDUF985 were determined by the molecular replacement method with MOLREP9 using the coordinates of homologous protein from Shewanella oneidensis [protein data bank (PDB) code 1yud], which shares 37% sequence identity (156 residues aligned) with BbDUF985 as the search model. The root-mean-square deviation (RMSD) between 1yud and the apo form is 1.5 Å and that between 1yud and the ligand-bound form is 1.4 Å using pairwise Dali server. The initial model was refined by using the maximum likelihood method implemented in REFMAC510 as part of CCP4 program suite11 and rebuilt interactively by using the σA-weighted electron density maps with coefficients 2mFO-DFC and mFO-DFC in the program COOT.12 During the later stage, the restrained positional and B-factor refinement was performed using the program phenix.refine13 and tight non-crystallographic symmetry (NCS) restraints over the two subunits were applied during the refinement. The final models were evaluated with the programs MOLPROBITY14 and PROCHECK.15 The final coordinates and structure factors were deposited in the PDB under the accession code of 3LOI and 3LZZ, respectively. The data collection and structure refinement statistics were listed in Table I. All structure figures were prepared with the program PyMOL.16 Crystal structures and sequence alignment of BbDUF985. A: Schematic representation of BbDUF985 homodimer comprising of subunits A and B with one GDP molecule binding to each subunit. All β-strands and helices are labeled sequentially in subunit A. The black arrow denotes the 2mFO-DFC map of GDP that is contoured at 1.5 σ. B: Interactions between BbDUF985 and GDP. C: Structural superposition of the apo and GDP-bound forms, which are colored in red and chartreuse, respectively. D: Multiple sequence alignments of BbDUF985 (the first row) against the homologs from the prokaryotes Shewanella oneidensis, Clostridium phytofermentans, and Pseudoalteromonas atlantica, fungus Saccharomyces cerevisiae, protozoan Dictyostelium discoideum, plant Arabidopsis thaliana, respectively. The cupin_5 domain of NUP from Pseudoalteromonas atlantica are indicated in orange background and the conserved residues involving in GDP binding are marked by red up-triangle arrow. The crystal structure of BbDUF985 in the apo form was refined to the resolution of 2.1 Å. It belongs to the I4122 space group with one molecule in an asymmetric unit (Table I). Most residues are well fitted in the electron density map except for the N-terminal residues Met1-Ser9. The overall structure of BbDUF985 monomer exhibits a typical cupin-like β-barrel topology which comprises an eight-strand β-barrel surrounded by four α-helices [Fig. 1(A)]. Besides, a long loop between β1 and β2 resembles a handle protruding from the core domain. Two monomers are further assembled into a dimer via reciprocal interactions from strands β1, β2, β3, and β8. The dimeric interface is about 1500 Å2, as shown in Figure 1(A), which involves eight hydrogen bonds and extensive hydrophobic interactions analyzed by PDBsum.19 In addition, BbDUF985 also exists as a dimer in solution confirmed by gel filtration using Superdex 75 column (GE healthcare), indicating that the protein probably functions as a dimer form (data not shown). Comparative structural analysis using Dali server (http://www.ebi.ac.uk/dali/)20 revealed a group of structurally similar proteins, such as auxin binding protein (PDB code 1LR5, Z-score 11.1), canavalin (PDB code 1CAX, Z-score 9.5), sugar phosphate isomerase (PDB code 3I7D, Z-score 9.5), dTDP sugar epimerase (PDB code 3EJK, Z-score 9.1), and other hypothetical cupins. Combined with the information provided by Pfam database (cupin clan NO. CL0029), we attempted to soak the apo-form crystal with several chemical molecules such as GDP, glucose-6-phosphate, and ectoine. In the end, only GDP can be bound in the crystal structure of BbDUF985. The structure was refined to the resolution of 2.5 Å. It belongs to the I41 space group with two molecules in an asymmetric unit (Table I). GDP is bound at the center of the β-barrel with the ribose ring and phosphate group moiety exposed to the solvent. The GDP molecule fits well in electron density map (2mFO-DFC map) contoured at 1.5 σ [Fig. 1(A)], when compared with the YML079w structure in the GTP-bound form, in which only the purine base moiety fits the electron density map.6 As shown in Figure 1(B), GDP is stabilized by both hydrophobic interactions and hydrogen bonds. The hydrophobic interaction is mainly contributed by Phe31, Phe140, and Phe142 and five hydrogen bonds are formed involving three water molecules. In detail, N1 of purine base forms H-bond with Wat211 that is further stabilized by Oγ1 of Thr54, while N7 of purine base interacts with Nε2 of His24. Furthermore, O3 of ribose moiety makes H-bond with Wat180 and O1A of α-phosphate group interacts with Wat244 that can fix the phosphate group tail of GDP molecule probably avoiding from wobbling in solvent environment. Besides, N2 of purine base interacts with Oε1 of Asp33 through attractive electrostatic force. To perform a multiple sequence alignment with Multalin21 and ESPript22 [Fig. 1(D)], we selected a set of representative homologs in the cupin_5 family. Remarkably, all the residues participating in ligand binding are conserved in the cupin_5 family even between BbDUF985 and relative distant homologs from plant, implying that functional convergence and conservation of GDP binding pattern of these proteins from the cupin_5 family. Although superposition of the apo and GDP-bound forms yields a RMSD of only 0.5 Å, obvious conformational changes are still observed in the pocket on GDP binding, which is mainly contributed by the conserved ligand binding residues [Fig. 1(C)]. As a result of induced fit, residues His24 and Asp33 shift inward to form hydrogen bonds with GDP, and the side chains of Phe31, Phe140, and Phe142 flip toward the purine and ribose rings of GDP to make hydrophobic interactions. The cupin superfamily is a very large group of functionally diverse proteins that are found in all three kingdoms of life: archaea, eubacteria, and eukaryota, and lots of cupin proteins have not been functionally characterized yet.23 As mentioned above, BbDUF985 belongs to the uncharacterized cupin_5 family of entire 35 cupin families, the residues of which involving in ligand binding are well conserved [Fig. 1(D)], implying that proteins in the cupin_5 family probably share a similar molecular function differing from that of other cupin families. Given the feature of hydrophobic interaction with GDP contributed by aromatic residues in binding pocket, protein fluorescence spectrometry was applied. Change of fluorescence intensity/wavelength can be used to compare binding affinity of different chemical molecules. Based on the Dali output, four representative compounds (200 μM each) were assayed, listed as GDP, dTDP (nucleotide moiety of substrate for dTDP sugar isomerase), glucose-6-phosphate (substrate for gluocose-6-phosphate isomerase), and ectoine (pyrimidine-like molecule). The fluorescence spectra revealed that GDP and dTDP are able to trigger significant decrease of the fluorescence intensity, whereas gluocose-6-phosphate or ectoine is not [Fig. 2(A)]. In addition, using the spectra with a gradient of GDP concentrations [Fig. 2(B)], we determine the KD of GDP, which is 146 ± 11 μM, approximately equivalent to the average physiological concentration of GDP (159 ± 51 μM) in most species except for Homo sapiens.24 The GDP binding assay. A: Fluorescence spectra of BbDUF985 with each selected molecule in contrast to that of the apo form. B: The plot of decrease in maximal fluorescence intensity (Y) against the GDP concentration (X) for determination of KD of GDP. Moreover, there are five kinds of domain organization in the cupin_5 family,4 one of which was defined by a protein of 575 amino acids from Pseudoalteromonas atlantica. It comprises both purine nucleoside permease (NUP) domain (N-terminal 53–385 residues) and cupin_5 domain [N-terminal 414–555 residues, as shown in Fig. 1(D)]. This evidence also suggests that the proteins in the cupin_5 family might be related to the nucleotide transport or metabolism. Despite the definite biological function of BbDUF985 remains unclear, however, our structural and biochemical data of BbDUF985 provided hints for further functional characterization of proteins in the cupin_5 family. The authors thank Beijing Synchrotron Radiation Facility (BSRF) for X-ray data collection.