The crystal structure and identification of NQM1/YGR043C, a transaldolase from Saccharomyces cerevisiae.
Hua Huang,Hui Rong,Xu Li,Shuilong Tong,Zhiqiang Zhu,Liwen Niu,Maikun Teng
DOI: https://doi.org/10.1002/prot.22237
2008-01-01
Abstract:Transaldolase belongs to the class I aldolase family, whose members are the pivotal enzymes catalyzing the most important carbon-carbon bond-forming reactions and aldol condensation reactions.1 Different from the members of the class II adolase family, transaldolase is characterized by the formation of a covalent Schiff base intermediate between a lysine residue within the active site and the substrate.2-4 As a key enzyme in the pentose phosphate pathway, transaldolase catalyzes the reversible transfer of a dihydroxyacetone moiety from fructose 6-phosphate to erythrose 4-phosphate and forms sedoheptulose 7-phosphate and glyceraldehyde 3-phosphate, respectively.4-6 The vital functions of this enzyme are reported to balance the levels of NADPH and reactive oxygen intermediates (ROI) in the pentose phosphate pathway of glucose metabolism7 and also to maintain the mitochondrial trans-membrane potential and sperm fertility.8 Dysfunction of transaldolase attracts a great deal of attention since it could lead to several severe diseases, such as liver cirrhosis9 and neonatal multi-organ disease.10 In the Saccharomyces cerevisiae genome, one open reading frame, named NQM1/YGR043C, exhibits more than 50% sequence identity to a group of transaldolases both from prokaryotic and eukaryotic organisms, such as Escherichia coli transaldolase B (TALB_ECOLI), Mus musculus transaldolase (TAL_MOUSE), and Homo sapiens transaldolase (TAL_HUMAN). According to the sequence alignment, NQM1/YGR043C was proposed to be a putative transaldolase in yeast (TAL2_YEAST).11 However, no experimental evidence has been reported to prove its transaldolase activity to date. In our studies, the NQM1/YGR043C gene from the cDNA library of S. cerevisiae was cloned into a pET22b(+) vector and expressed in Escherichia coli strain BL21 (DE3). The purified protein was crystallized using the hanging-drop vapor diffusion method and the crystal structure was determined by the molecular replacement method and refined to a 1.9 Å resolution. The NQM1/YGR043C protein folds into an eight-stranded α/β barrel, which is conserved among all reported transaldolase structures. The enzyme activity assays further demonstrated its transaldolase activity. Based on these data, we concluded that the open reading frame (ORF) of NQM1/YGR043C does encode a transaldolase, which could be named as TAL2_YEAST. The corresponding catalytic sites and the proposed substrate entry channel in TAL2_YEAST were predicted based on the findings in E. coli transaldolase B6, 12 and human transaldolase.13, 14 As the first structure of transaldolase in yeast, our study may deepen the understanding of the catalytic mechanisms of transaldolase in metabolic pathways. On the basis of the sequence of NQM1/YGR043C (GeneID:852934), two oligonucleotide primers, 5′ CTG GAT CCC ATA TGT CAG AAC CTT CAG AG 3′ (forward) and 5′ CAG CTC GAG CAT TTT TTC TTC AAC CAG 3′ (reverse) were synthesized and used to clone the corresponding gene from the cDNA library of S. cerevisiae. By using a standard PCR protocol, an approximately 1000-bp fragment was amplified. The sequence of this fragment was analyzed by DNA sequencing in Sangon Co., Ltd (Shanghai, China) and proved to be the same as that of the target gene in the Saccharomyces Genome Database. The gene was then cloned into the NdeI and XhoI sites of a pET22b (+) vector (Novagen). The plasmid was transformed into E. coli strain BL21 (DE3) for protein expression. Cells were grown at 310 K in 3 L LB medium (10 g Bacto tryptone, 5 g yeast extract, and 10 g NaCl per liter of solution) containing ampicillin (50 μg/mL) until OD600 = 0.6. At this stage, isopropyl-β-D-thiogalactoside (IPTG) was added to a final concentration of 0.5 mM and the cells were grown for an additional 6 h at 310 K. Cells were then harvested by centrifugation at 8000 rpm for 8 min and resuspended in the binding buffer (20 mM Tris-HCl, pH 8.0, 300 mM NaCl). The crude cell extract was prepared by sonication and centrifugation at 15,000 rpm for 30 min. The supernatant was loaded onto a Ni2+-affinity column equilibrated with binding buffer. The column was washed with 6 column volumes of the binding buffer, 8 column volumes of the washing buffer (20 mM Tris-HCl, pH 8.0, 300 mM NaCl, 100 mM imidazole), and finally the protein was eluted in 5 column volumes of the elution buffer (20 mM Tris-HCl, pH 8.0, 300 mM NaCl, 500 mM imidazole). The purified protein was analyzed by SDS-PAGE, concentrated and desalted by ultrafiltration equipment to approximately 15 mg/mL (measured with a Bio-Rad Protein Assay kit) in buffer (10 mM Tris-HCl, pH 8.0, 30 mM NaCl) to be used for crystallization. Crystallization of the protein was performed at 283 K using the hanging-drop vapor diffusion method and initial trials were carried out by using screening Crystal Screen I and Crystal Screen II reagent kits (Hampton Research). Each hanging drop was a mixture of 1 μL of protein and an equal volume reservoir solution, which was then equilibrated against 400 μL reservoir solution. The polyhedral crystal was obtained from the potassium citrate condition. After optimization and adding some small molecular compounds, the crystals suitable for X-ray diffraction were grown in 1.3 M potassium citrate, 10% polyethylene glycol 200, 1% ethylene glycol, and 0.1 mM HEPES, pH 7.3, with a protein concentration of 10 mg/mL. X-ray diffraction data were collected at beam line 3W1A of the Beijing Synchrotron Radiation Facility at the Institute of High Energy Physics, Chinese Academy of Sciences. During the data collection, the crystal was maintained at 100 K using nitrogen gas with cryoprotection (1.2 M potassium citrate, 8% polyethylene glycol 200, 1% ethylene glycol, 0.09 mM HEPES, pH 7.3, and 15% glycerol). The diffraction data were processed and scaled with the HKL2000 suite of program.15 The crystal belongs to the space group I212121, and the asymmetric unit contains two molecules with a solvent content of ∼52%. The processing statistics are summarized in Table I. The structure was determined by molecular replacement program PHASER16 using the monomer structure of human transaldolase (PDB code: 1F05) as the search model. The complete model was built manually using Coot17 with 2Fo-Fc and Fo-Fc electron-density maps as references, and refinement was carried out with REFMAC5.18 Water molecules, glycerol, ethylene glycol and polyethylene glycol 200 were added near the end of the refinement, followed by manual modification in the graphics program Coot. The quality of the final model was checked by PROCHECK.19 The final detailed refinement statistics and geometry are listed in Table I. The coordinate and structure factors have been deposited in the Protein Data Bank (accession code: 3CQ0). Transaldolase activity assays were carried out at 298 K in 40 mM triethanolamine, 5 mM EDTA, pH 7.6, with 5 mM fructose-6-phosphate, 0.2 mM erythrose-4-phosphate, 0.1 mM NADH, and 10 U/mL each of the auxiliary enzymes triosephosphate isomerase and glycerol-3-phosphate dehydrogenase (Sigma), essentially as described previously.13 0.32 nmol protein of NQM1/YGR043C was added to the 200 μL reaction solution to initiate the reaction. Control experiments were run in an identical manner, except that the protein storage buffer was substituted for the putative enzyme. The decrease in A340 of NADH concentration, which indicates the formation of glyceraldehyd-3-phosphate from the conversion of fructose-6-phosphate and erythrose-4-phosphate, was monitored every 6 sec. In the crystal of NQM1/YGR043C, each asymmetric unit contains two molecules, molecule A and B. The final model displayed excellent geometry except for the first eleven amino acids (amino acids 1–11 at the N-terminus, both in molecule A and B) and a region of an outside loop (203–205 in molecule A and 202–206 in molecule B), which were disordered in the crystal. Some regions of NQM1/YGR043C revealed a weaker electron density in molecule B when compared with molecule A, which was consistent with what has been found in the structure of human transaldolase.4 The core structure of NQM1/YGR043C adopted a modified (α/β)8 barrel fold. Seven α-helices (with absence of α1 helix) embraced the β-barrel. The core structure was further surrounded by seven additional helices (αA–αG helices) [Fig. 1(A)]. This topology was consistent with the typical core structure among the superfamily of transaldolases.4, 6, 20 (A) Ribbon figure of the monomer structure of NQM1/YGR043C (stereo view). The protein is colored by secondary structure elements (α2–8 helices are in yellow, αA–G helices are in magenta and β1-8 strands are green) and numbered from N terminus to the C terminus according to the way used by TAL_HUMAN. (B) Comparison of homodimer interfaces of TAL_HUMAN and NQM1/YGR043C. When molecule A of TAL_HUMAN (blue) was perfectly aligned to molecule A of NQM1/YGR043C (green), positional divergence of the two B molecules was observed. This suggested that the homodimer interface of NQM1/YGR043C is different from that of TAL_HUMAN. (C) Homodimer interface of NQM1/YGR043C. 19 residues from each molecule are shown as magenta and orange respectively, pg4 molcule is shown as green, the two Trp149 residues from each molecule whose main chain nitrogen atoms form hydrogen bonds with pg4 are shown as slate, the hydrogen bonds are shown as red dashed lines. The residues forming hydrogen bonds are labeled and shown as sticks. (D) Stick presentation of the catalytic site (Lys144), the correlative sites for the formation of the reduced Schiff-base intermediate complex (Asp29, Asn47, Glu108, Asn167, Thr169, Ser189 and Arg194) and the recognizing and binding sites for substrate (Ser173) of NQM1/YGR043C (green) are compared with the corresponding sites in TALB_ECOLI (magenta), TAL_HUMAN (slate) and the reduced Schiff-base intermediate complex of TALB from Escherichia coli (orange). Residue numbers are labeled respectively. These figures were prepared using the PyMol program (http://pymol.sourceforge.net/). A homodimer of NQM1/YGR043C [Fig. 1(B)] was observed in crystal, as well as in solution demonstrated by size exclusion chromatography. Of approximately 6.5% of the total accessible surface area in the dimer, nearly 1710 Å2, made up the subunit-subunit interface and 19 amino acids (Arg112, Phe115, Asp116, Glu150, Glu294, Pro295, Arg298, Tyr299, Asn302, Glu303, Gln305, Thr308, Glu309, Ser312, Asp313, Ile315, Arg316, Ser319 and Glu323) of each monomer contributed to the homodimer interface [Fig. 1(C)]. Within these amino acids, 18 residues, except for Glu323, were identical to those involved in the homodimer interface in human transaldolase. These data were prepared by the Protein-Protein Interaction Server.21 Different from the human transaldolase crystallization conditions, the diffraction quality of NQM1/YGR043C crystal had improved remarkably while the polyethylene glycol 200 was used as additive. Considering the differences between the homodimer interfaces in human transaldolase and NQM1/YGR043C [Fig. 1(B)], the pg4 molcule (main component of PEG200) may form hydrogen bonds directly with the Trp149 nitrogen atom in each monomer and indirectly with nearby residues by some water molecules, which could make a substantial contribution to the homodimer interface formation [Fig. 1(C)]. Sequence alignments revealed that the NQM1/YGR043C protein is fairly similar to other transaldolases. NQM1/YGR043C exhibited 52% of overall sequence identity to Escherichia coli transaldolase B (TALB_ECOLI), 54% to Escherichia coli transaldolase A (TALA_ECOLI), 57% to Mus musculus transaldolase (TAL_MOUSE), 57% to Homo sapiens transaldolase (TAL_HUMAN), and 68% to S. cerevisiae transaldolase 1 (TAL1_YEAST). NQM1/YGR043C contained all residues that are highly conserved and invariant in the transaldolase superfamily,4 for instance, the Lys on β4 strand of the core α/β barrel, where the reduced Schiff-base intermediate is produced.6 Three dimensional structure comparisons revealed that both the overall structure and the core α/β barrel of NQM1/YGR043C are similar to those in other transaldolases. Superimpositions of the core α/β barrel of NQM1/YGR043C with those of TAL_HUMAN and TALB_ECOLI showed a r.m.s.d of 0.407Å on all 119 Cα atom pairs and a r.m.s.d of 0.453 Å on all 118 Cα atom pairs, respectively. A few solvent-exposed loops, remote from the active sites, however exhibited notable differences, for example, the loop connecting αE helix to αF helix. Alignments of the sequence and structure of NQM1/YGR043C with other transaldolases indicated that this protein contained all residues carrying out the transaldolase activity in the catalytic sites that had been reported in Escherichia coli transaldolase B (PDB ID: 1ucw)6, 12 and human transaldolase (PDB ID: 1f05).4 All of the catalytic residues of NQM1/YGR043C: Lys144, Asp29, Asn47, Glu108, Asn167, Thr169, Ser189 and Arg194, were located at the same spatial positions within the active sites identified in other transaldolases [Fig. 1(D)]. Meanwhile, the region around Ser173 of NQM1/YGR043C was also structurally conservative to the proposed important substrate binding site in human transaldolase (region around Ser171).13, 14 All these evidences indicated that NQM1/YGR043C may possess the same catalytic activity as other members of transaldolase superfamily. In the transaldolase activity assay, a dramatic decrease of A340 has been detected after adding the protein into the assay solution (see Fig. 2), suggesting a great deal of glyceraldehyde-3-phosphate was produced during the transferring and reducing reactions between fructose 6-phosphate and erythrose 4-phosphate. These results demonstrated that the putative transaldolase does have transaldolase activity, with the Vmax value as 201.2 ± 6.5 (s.d.) μM/min and the specific activity as 3.4 ± 0.1 (s.d.) U/mg. The change of ▵A340 monitored in the transaldolase activity assay. The values of ▵A340 during the reaction solution which contained the protein of NQM1/YGR043C are labeled as squares, whereas those in the control are labeled as diamonds. Data were collected from three independent experiments. See Materials and methods for details. In this study, we cloned the NQM1/YGR043C gene from the S. cerevisiae cDNA library, expressed and crystallized the recombinant protein, and finally determined its structure. Additionally, we demonstrated its transaldolase activity by using an enzyme catalytic assay. Based on both the structural information and the enzymatic assay results, we concluded that the ORF of NQM1/YGR043C does encode the transaldolase in yeast, named as TAL2_YEAST, which belongs to the transaldolase superfamily. As the first crystal structure of yeast transaldolase, our results will be helpful for further exploring the biological functions and molecular mechanisms of both transaldolases in yeast, TAL1_YEAST22 and TAL2_YEAST. We will perform experiments to elucidate the complete details of the substrate entry channel from Ser173 to Lys144 in our future studies. The authors thank Prof. Congzhao Zhou in School of Life Science, University of Science and Technology of China for the cDNA of S. cerevisiae. The authors also thank Profs. Yuhui Dong, Peng Liu and their colleagues on beam line 3W1A of the Beijing Synchrotron Radiation Facility at the Institute of High Energy Physics, Chinese Academy of Sciences for diffraction data collection.