Expression, crystallization and preliminary X-ray diffraction analysis of thioredoxin gluta­thione reductase from Schistosoma japonicum in complex with FAD

Thioredoxin glutathione reductase from Schistosoma japonicum (SjTGR), a multifunctional enzyme, plays a vital role in antioxidant pathways and is considered to be a potential drug target for the development of antischistosomal chemotherapy. In this study, two constructs of a truncated form of SjTGR without the last two residues (Sec597–Gly598) were cloned, overexpressed and purified using wild-type and codon-optimized genes. Only SjTGR from the wild-type gene was found to form a complex with flavin adenine dinucleotide (FAD), which could be crystallized in the orthorhombic space group P2₁2₁2₁, with unit-cell parameters a = 84.185, b = 86.47, c = 183.164 Å, at 295 K using the hanging-drop vapour-diffusion method. One dimer was present in the crystallographic asymmetric unit and the calculated Matthews coefficient (V_M) and solvent content were 2.6 ų Da⁻¹ and 52.8%, respectively. Structural determination of SjTGR is in progress using the molecular-replacement method.

1. Introduction

Thioredoxin glutathione reductase (TGR) from Schistosoma japonicum (SjTGR), a unique multifunctional enzyme, performs the reduction of GSSG (oxidized glutathione) and Trx (thioredoxin) in S. japonicum redox systems, which is normally carried out by thio­redoxin reductase (TrxR) and glutathione reductase (GR) in mammals (Alger & Williams, 2002; Johansson et al., 2005; Song et al., 2012). Platyhelminth parasites possess a unique redox system without canonical enzymes in comparison with free-living flatworms. Hence, TGR plays a key role in parasite survival and is normally regarded as a drug target for platyhelminth infections (Otero et al., 2010). Research results have indicated that the multifunctional enzyme TGR meets all of the major criteria for a drug target in the antioxidant pathway (Kuntz et al., 2007; Han et al., 2012). TGR from schistosomes has been used as an essential drug target for the development of antischistosomal chemotherapy (Song et al., 2012; Kuntz et al., 2007).

The efficiency of recombinant protein production in an Escherichia coli system may be diminished by codon-usage bias (Plotkin & Kudla, 2011; Ikemura, 1985; Drummond & Wilke, 2008). Ways to overcome this problem include targeted mutagenesis to change rare codons or the addition of co-expressing genes encoding the rare codon tRNAs in E. coli (Gustafsson et al., 2004; Brinkmann et al., 1989; Springer & Sligar, 1987; Kleber-Janke & Becker, 2000). However, many genes show little bias in codon usage and reflect a lack of selection for mRNA translation. For instance, the expression of Neurospora FREQUENCY (FRQ) protein with an optimized sequence shows non-optimal codon usage for its biological function. Recombinant FRQ expression using the native gene leads to an optimal protein structure and function. However, the expression of the optimized gene leads to conformational changes and functional disorder (Zhou et al., 2013).

In this study, a codon-optimized SjTGR (coSjTGR) gene was synthesized directly and the gene for wild-type SjTGR (wtSjTGR) was amplified from an S. japonicum cDNA library. We cloned and expressed both SjTGR genes in an E. coli system and purified both recombinant SjTGR proteins (wtSjTGR and coSjTGR) for crystallization. Interestingly, the expression and crystallization of wtSjTGR and coSjTGR differed significantly. Only wtSjTGR was found to form a complex with flavin adenine dinucleotide (FAD) and to produce crystals suitable for structure determination.

2. Materials and methods

2.1. Codon optimization and gene cloning

Codon optimization was executed using Optimizer as created by Puigbò et al. (2007, 2008). We selected the codon usage of predicted highly expressed genes and the `one amino acid–one codon' method. All codons of the query sequence were optimized and substituted by the synonymous codon used most frequently in the reference set (Puigbò et al., 2007, 2008). The optimized SjTGR gene was then directly synthesized by chemical methods, and vector construction was completed by Sangon Biotech. The gene of wtSjTGR was amplified via the polymerase chain reaction from a female S. japonicum cDNA library. Sequence information (GenBank AY814219.1) was used to synthesize or to design the forward and reverse oligonucleotide primers. The primer sequences, which contained NdeI and BamHI restriction sites (bold), were 5′-GTGCCGCGCGGCAGC–CATATGCCTCCGATTGAT-3′ and 5′-ACGGAGCTCGAATTC–GGATCCTCAGCAACCG-3′, in which the first 15 bases were designed for subsequent one-step cloning (ligation-independent cloning). The wtSjTGR gene (1791 bp) was cloned into expression vector pET-28a with an N-terminal His₆ tag and a thrombin recognition site (MGSSHHHHHHSSGLVPRGSH). The resulting positive plasmid was confirmed using DNA sequencing and transformed into E. coli BL21 (DE3) strain.

2.2. Expression and purification

Protein expression was induced with 0.2 mM IPTG at room temperature (about 295 K). After overnight incubation, the cells from 1 l of culture were harvested by centrifugation at 10 000g for 15 min (Eppendorf 5804 R) and suspended in 20 ml Ni-affinity lysis buffer [300 mM NaCl, 50 mM NaH₂PO₄ pH 7.5, 10%(v/v) glycerol, 5 mM β-mercaptoethanol]. Cells were disrupted by sonication for 4 × 40 s on ice at maximum power. Cellular debris was removed by centrifugation for 30 min at 12 000g and passage through a 0.45 µm filter.

The protein was purified using a HisTrap Ni-affinity column (GE Healthcare) followed by a HiTrap DEAE column (GE Healthcare) using an ÄKTA FPLC system. The soluble fraction was loaded onto a 5 ml HisTrap Ni-affinity column previously equilibrated with lysis buffer. The column was washed with lysis buffer containing 50 mM imidazole and the protein was eluted using lysis buffer containing 250 mM imidazole. The protein was then dialyzed into a buffer consisting of 50 mM NaCl, 20 mM Tris–HCl pH 8.0, 2 mM EDTA, 5 mM β-mercaptoethanol and loaded onto a DEAE HiTrap column previously equilibrated with the dialysis buffer. The final pure protein was eluted using the linear gradient method using a buffer consisting of 500 mM NaCl, 20 mM Tris–HCl pH 8.0, 2 mM EDTA. The recombinant protein was soluble and the yield was about 20 mg per litre of culture. Protein purity was assessed by SDS–PAGE followed by Coomassie Blue staining, and protein concentration was determined by the Bradford assay with bovine serum albumin as a standard (Bradford, 1976). The protein was concentrated to about 20 mg ml⁻¹ using an Amicon Y5 membrane concentrator (Millipore) and stored at 193 K until use.

2.3. Crystallization

Crystallization trials using both SjTGR samples (wtSjTGR and coSjTGR) were performed at 295 K using the sitting-drop vapour-diffusion method in CrystalEX (Corning) 96-well plates with PEG/Ion 2 from Hampton Research. wtSjTGR crystals were observed in two conditions [0.1 M sodium formate pH 7.0, 12% PEG 3350 and 8%(v/v) Tacsimate pH 6.0, 20% PEG 3350] after 3 d and the conditions were optimized using laboratory-made solutions via the hanging-drop vapour-diffusion method in a 24-well Linbro plate. Finally, wtSjTGR crystals obtained from a condition consisting of 0.2 M potassium sodium tartrate tetrahydrate, 0.1 M HEPES pH 7.0, 20% PEG 3350 were used for data collection. coSjTGR microcrystals were observed using 0.2 M KSCN, 20% PEG 3350 pH 7.0. However, no coSjTGR crystals suitable for X-ray diffraction were obtained even after extensive optimization trials.

2.4. Data collection and processing

Cryoprotection of the crystals was achieved by rapid soaking in the reservoir solution supplemented with 25%(v/v) glycerol. Diffraction data were collected in a stream of cold nitrogen at 100 K on beamline BL17U1 at the Shanghai Synchrotron Radiation Facility (SSRF), People's Republic of China. X-ray diffraction data were collected using an ADSC Q315r CCD detector at a wavelength of 0.9707 Å. The crystal was oscillated by 1.0° per frame over a range of 360° to obtain sufficient multiplicity. The data sets were scaled and merged using the HKL-2000 program package (Otwinowski & Minor, 1997).

2.5. Structure determination

The structure of SjTGR was solved by the molecular-replacement method using PHENIX (Adams et al., 2010). Coot (Emsley & Cowtan, 2004) was used for model building and manual adjustments.

3. Results and discussion

An alignment of the wild-type and the codon-optimized SjTGR genes (wtSjTGR and coSjTGR) indicating the codon changes is shown in Fig. 1. The expression of SjTGR from E. coli BL21 strain with recombinant pET-28a-wtSjTGR and with pET-28a-coSjTGR vectors successfully yielded soluble SjTGRs from the cell lysate. Interestingly, the recombinant proteins produced from different vectors had different colours: wtSjTGR was yellow and coSjTGR was colourless (Fig. 2). Preliminary analysis of the electron density suggests that FAD molecules present in wtSjTGR generate the yellow colour.

Figure 1 The wtSjTGR and coSjTGR genes. Asterisks indicate codon changes.

Figure 2 SDS–PAGE analysis of wtSjTGR and coSjTGR. (a) Purified wtSjTGR shown on a 12% polyacrylamide SDS gel. Lane M1, molecular-weight marker (labelled in kDa); lanes 1 and 2, eluted fraction from DEAE column. (b) Purified coSjTGR shown on a 12% polyacrylamide SDS gel. Lane M2, molecular-weight marker (labelled in kDa); lanes 3 and 4, eluted fraction from DEAE column. I, solution sample of wtSjTGR. II, solution sample of coSjTGR.

Crystals of wtSjTGR (Fig. 3) suitable for X-ray diffraction data collection (0.3 × 0.3 × 0.2 mm) were obtained by mixing 2 µl 20 mg ml⁻¹ protein solution with 2 µl reservoir solution consisting of 0.2 M potassium sodium tartrate tetrahydrate, 20%(w/v) PEG 3350 in bis-tris buffer pH 6.8. Diffraction data were collected to 2.4 Å resolution (Fig. 3). The data-collection and processing statistics are given in Table 1. Auto-indexing was performed with HKL-2000 and the results indicated that the crystals belonged to space group P2₁2₁2₁, with unit-cell parameters a = 84.185, b = 86.47, c = 183.164 Å. One dimer exists in the asymmetric unit, with a corresponding V_M of 2.6 ų Da⁻¹ and solvent content of 52.8% (Matthews, 1968). Only small crystals were observed for coSjTGR using the same methods and screening kit.

Table 1
X-ray diffraction data and processing statistics

Values in parentheses are for the outermost resolution shell.

X-ray source	BL17U1, SSRF
Wavelength (Å)	0.9707
Temperature (K)	100
Detector	ADSC Quantum 315r
Space group	P212121
Resolution (Å)	43.24–2.35 (2.432–2.35)
Unit-cell parameters (Å, °)	a = 84.185, b = 86.470, c = 183.164, α = β = γ = 90
Total reflections	56540
Unique reflections	5506
Completeness (%)	99.9 (99.23)
Rmerge† (%)	14.5 (94.8)
〈I/σ(I)〉	20.44 (4.0)
Multiplicity	7.1 (7.3)

†R_merge = , where I_i(hkl) is the intensity of the ith observation of unique reflection hkl and 〈I(hkl)〉 is the average over symmetry-related observations of unique reflection hkl.

Figure 3 Crystals of coSjTGR and wtSjTGR. (a) Microcrystals of coSjTGR grown with PEG 3350 as precipitant. (b) Typical wtSjTGR crystals with dimensions of 0.3 × 0.3 × 0.2 mm that diffracted to a resolution of 2.4 Å. (c) Diffraction pattern of wtSjTGR; data were detectable to 2.4 Å resolution.

In this study, high-level expression of both wtSjTGR and coSjTGR genes was achieved in an E. coli system. wtSjTGR contained bound FAD, whereas coSjTGR had no bound cofactor. Single crystals suitable for X-ray diffraction data collection were only obtained for wtSjTGR. However, only microcrystals of coSjTGR (Fig. 2) could be grown even after condition optimization (Fig. 2). These results may suggest that codon usage regulates not only the protein expression level but also protein folding. Protein chaperones and sufficient time are required for protein folding. Codon optimization leads to higher translation rates and reduces the time available for folding. Codon optimization is probably the main reason that coSjTGR presents non-optimal codon usage for FAD capture, and affects the protein folding. Previous studies have indicated that codon usage regulates protein folding (Komar et al., 1999; Siller et al., 2010; Zhou et al., 2009), and can change the protein structure and function in some cases (Zhou et al., 2013; Kimchi-Sarfaty et al., 2007).