These authors contributed equally to this article.
Characterization of catalytic activity and structure of selenocysteine-containing hGSTZ1c-1c based on site-directed mutagenesis and computational analysis
Article first published online: 9 JAN 2013
Copyright © 2013 International Union of Biochemistry and Molecular Biology, Inc.
Volume 65, Issue 2, pages 163–170, February 2013
How to Cite
Yu, Y., Song, J., Song, Y., Guo, X., Han, Y. and Wei, J. (2013), Characterization of catalytic activity and structure of selenocysteine-containing hGSTZ1c-1c based on site-directed mutagenesis and computational analysis. IUBMB Life, 65: 163–170. doi: 10.1002/iub.1128
- Issue published online: 28 JAN 2013
- Article first published online: 9 JAN 2013
- Manuscript Accepted: 6 DEC 2012
- Manuscript Received: 21 NOV 2012
- glutathione peroxidase;
- human glutathione transferase zeta 1c-1c;
- cysteine auxotrophic strain;
- site-directed mutagenesis
Human glutathione transferase zeta 1c-1c (hGSTZ1c-1c) is one of the glutathione transferase isoenzymes and considered to be a protein scaffold to imitate glutathione peroxidase (GPX) owing to the natural binding site of glutathione (GSH). In this report, several residues near GSH were mutated to selenocysteine (Sec) or cysteine (Cys) residues and the impacts of the substitutions on different activities were discussed. Mutations of Ser-14 or/and Ser-15 to Cys or Sec residues resulted in dramatic loss of catalytic activity of hGSTZ1c-1c with chlorofluoroacetic acid as substrate, which indicated the importance of the hydroxyl groups in Ser-14 and Ser-15. And subsequent study by molecular modeling suggested that Ser-15 was probably involved in catalysis, while Ser-14 may play a crucial role in binding and orientation of GSH and possibly had a synergistic effect with Ser-15 in catalysis. On the contrary, the result of converting Cys-16 to Ser indicated its trivial role in catalysis. The investigations of the selenocysteine-containing hGSTZ1c-1c (seleno-hGSTZ1c-1c) and the mutant S17C implied that the substitutions of multi-Sec for Cys residues at position 16, 137, and 205 could lead to subtle change in the structure of the protein molecule and concomitant change in catalytic activity as a direct result. This finding provides overwhelming evidence that the protein scaffold containing fewer cysteines should be chosen for imitating GPX using cysteine auxotrophic strain system to avoid unexpected structural changes. © 2013 IUBMB Life, 65(2)163–170, 2013
Human glutathione transferase zeta 1 (hGSTZ1) is one of the glutathione transferase isoenzymes and discovered by use of a bioinformatic approach. Four polymorphic variants of hGSTZ1 were identified, including hGSTZ1a-1a, hGSTZ1b-1b, hGSTZ1c-1c, and hGSTZ1d-1d (1–3). Previous investigations revealed that hGSTZ1 could catalyze the biotransformations of α-haloacids to glyoxylic acid and maleylacetoacetate to fumarylacetoacetate in the presence of glutathione (GSH) (4–6). The SSC motif (Ser14-Ser15-Cys16) highly conserved in most of GSTs was regarded as the active center in catalysis. And further evidence demonstrated that Ser-14 was the only residue found to be essential for catalysis and Cys-16 played an important role in the binding and orientation of GSH in the active site (7).
Because of the natural binding site of GSH, human glutathione transferase zeta 1c-1c (hGSTZ1c-1c) has been considered to be an ideal protein scaffold for imitating glutathione peroxidase (GPX). GPX (EC 184.108.40.206) is an important class of antioxidant enzymes that protects cells from oxidative damage. As radical scavengers, GPXs catalyze the GSH-dependent reaction and thereby reduce different types of peroxides to their respective alcohols. The catalytic center of GPX is known as catalytic triad consisting of a Sec, a Trp, and a Gln residue. Owing to the restrictions of native GPX, increasing researchers have focused on imitating GPX. And after exploring numerous GPX mimics, two factors pivotal to the imitation of GPX have been summed up: (i) the proper binding site of GSH and (ii) the catalytic center (specifically refered to Sec in protein molecule). Sec is regarded as the 21st amino acid and encoded by UGA, which usually functions as a stop codon. And the fact that the mechanism of Sec incorporation differs from prokaryotes to eukaryotes makes it difficult to directly heterologously express recombinant mammalian selenoproteins in E. coli (8). It has been reported that several GPX mimics with high activity are produced by eukaryotic expression system and chemical method (9–13). However, the restrictions, such as low yield and short of specificity, make it inappropriate to produce quantities of selenoproteins and predict the structure of the protein scaffold introduced Sec for imitating GPX. Instead, cysteine auxotrophic strain system could be used to achieve the introduction of Sec more efficiently via tRNACys misloading (14–16).
The aim of this study was to investigate the changes of catalytic activity from hGSTZ1c-1c to seleno-hGSTZ1c-1c produced using cysteine auxotrophic strain system. Furthermore, the role of different residues in the SSC motif on binding GSH and the effect of introducing Sec into hGSTZ1c-1c were explored by site-directed mutagenesis and computational analysis.
MATERIALS AND METHODS
Bacterial Strains, Plasmids, and Growth Conditions
The bacterial strains and plasmids used in this study are listed in Table 1. The medium used in site-directed mutagenesis and subcloning experiments was Luria-Bertani broth. In the selenoprotein expression experiments, a modification of the minimal medium was used as described previously and the medium was fortified with 250 mg/L of L-methionine (15).
Construction of Plasmids
Expression plasmid pCZ1c coding for hGSTZ1c-1c was constructed by subcloning the BamH I/Hind III fragment from pQEZ1 into the BamH I/Hind III sites of pCold I vector (TaKaRa). Plasmid DNA of pCZ1c was prepared using the MiniBEST Plasmid Purification Kit (TaKaRa). Mutations were created using the QuikChange™ site-directed mutagenesis kit (Stratagene) following the manufacturer's instructions and the mutagenic primers used in this study are listed in Table 2. PCR was carried out in a 50 μL mixture using 50 ng template plasmid pCZ1c, 125 ng of each mutagenic primer, 0.2 mM of dNTP, 2.5 U of PfuTurbo DNA polymerase in 1× reaction buffer. The PCR program was as follows: 95 °C for 30 Sec and then 18 cycles of 95 °C for 30 Sec, 55 °C 1 Min and 68 °C for 14 Min, and a final incubation at 4 °C. Add 1 μL of the Dpn I restriction enzyme (10 U/μL) directly to the amplification reaction. The reaction mixture was incubated at 37 °C for 1 H and transformed into E. coli DH5α. All mutations were verified by automated Sanger sequencing.
Overexpression and Purification of Seleno and Nonseleno hGSTZ1c-1c Variants
Plasmids DNA of pCZ1c and its various mutants were prepared using the MiniBEST Plasmid Purification Kit (TaKaRa) and then were transformed into E. coli BL21(DE3) for the production of nonseleno proteins. Plasmids DNA of pCZ1c and its various mutants were transformed into BL21(DE3)cys for the production of selenoproteins. Overexpression of seleno-hGSTZ1c-1c and its various mutants was performed using the method described previously (15). The recombinant proteins were purified by the immobilized metal affinity chromatography purification system using standard Ni2+ charged beads and confirmed by SDS-PAGE and Western blotting.
The substitution ratio of Sec residue was determined by measuring the molecular masses of seleno and nonseleno hGSTZ1c-1c using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS).
Assay of Enzyme Activities
The GST activities of hGSTZ1c-1c and its various mutants were measured by the spectrometric method quantifying glyoxylic acid formation as described previously (4). The specific activity is expressed in μmol/Min per mg of enzyme.
The GPX activities of the hGSTZ1c-1c and its various mutants were determined using a previously described method (17). The activity unit (U) is defined as the amount of enzyme that uses 1 μmol of NADPH per Min. The specific activity is expressed in U/μmol.
All computations were performed with Insight II package, version 2000 (Accelrys, San Diego, CA). The X-ray crystallographic structure of hGSTZ1c-1c from Protein Data Bank (PDB ID code 1FW1) was used for the starting coordinates for calculations. The 3D structures of hGSTZ1c-1c and its various mutants were constructed using the Homology program, and the models with the best value of discrete optimized potential energy statistical potential were selected (18). The obtained models were energy minimized using the method of conjugate gradient (300 steps) minimization, and then a molecular dynamics (MD) simulation was performed for 1 ns simulations at a constant temperature 298 K to examine the quality of the model structures. TIP3P water was used to construct a 20 Å water cap from the center of mass of each model, and then the models were subsequently energy minimized until the root mean-square gradient energy was lower than 0.001 kcal/mol Å. All calculations were performed using the Discover 3 software package (19, 20). The final models were analyzed with Profile-3D (21) and Procheck (22).
The Affinity program was used to perform molecular docking (23). The crystal structure of hGSTZ1c-1c obtained from Protein Data Bank contains the ligand (GSH) and hence the binding site of the model was determined easily. Molecular docking was performed with the aid of the consistent-valence force field. The centered enzyme–ligand complexes were solvated in a sphere of TIP3P water molecules with radius 20 Å. The docked complexes were selected by the criteria of interacting energy combined with the geometrical matching quality. These complexes were used as the starting conformation for further energy minimization and geometrical optimization before the final models were achieved (11).
Overexpression and Purification of Seleno and Nonseleno hGSTZ1c-1c Variants
SDS-PAGE analysis showed a single band of about 26,000 Da as judged on Fig. 1, and the same band was detected by Western blotting, which suggested that the seleno and nonseleno hGSTZ1c-1c and its variants were successfully expressed and purified.
MALDI-TOF MS analysis revealed that Sec residues were introduced into seleno-hGSTZ1c-1c successfully. The molecular mass of seleno-hGSTZ1c-1c was 26,694 ± 27 Da, and that of the wild-type hGSTZ1c-1c was 26,399 ± 98 Da from the results of MALDI-TOF MS. And then, the selenium content of seleno-hGSTZ1c-1c was determined to be 3.37 mol selenium per mol selenoenzyme. The results illustrated that the average substitution ratio of Sec for Cys residues was 75% in this study and the selenoproteins prepared using cysteine auxotrophic strain system were seleninic acid (RSeO2H) forms.
Assay of Enzyme Activities
The GST activities of seleno and nonseleno hGSTZ1c-1c and its various mutants are listed in Table 3, and only seleno-hGSTZ1c-1c (S17C) shows insignificant GPX activity. It was noted that the GST activities of some mutants (S14C, S15C, and S14C/S15C) decreased sharply following the mutation of Ser-14 or/and Ser-15 to Cys, but no significant changes were observed for two of the mutants (C16S and S17C) in nonseleno condition (Fig. 2). Therefore, the hydroxyl groups of Ser-14 and Ser-15 may play a fundamental role in the catalytic reaction of chlorofluoroacetic acid (CFA) by hGSTZ1c-1c, while Cys-16 and Cys-17 was probably not involved in the reaction. Moreover, the GST activities declined more significantly following the substitution of multi-Sec for Cys in seleno-hGSTZ1c-1c and its various mutants (Fig. 2). The results above indicated that some Cys in hGSTZ1c-1c may be necessary to retain the structure of the protein.
Different 3D structures of hGSTZ1c-1c variants were built using the X-ray crystallographic structure coordinates of hGSTZ1c-1c (PDB ID code 1FW1) as templates as described previously (11). And after refined by MD simulations, the resulting models were obtained. The Profile-3D score of each model lied between expected high score and low score, and Procheck Ramachandran plot analysis showed that more than 90% of residues in each model lied in the core region (data not shown). The results indicated that the conformations of hGSTZ1c-1c and its various mutants are reliable.
The impact of SSC motif on ligand (GSH) binding was investigated by docking GSH to the active sites of different hGSTZ1c-1c mutants using the Insight II/Affinity module (23). It is shown that the thiol of GSH is orientated toward to the hydroxyl groups of Ser-14 and Ser-15, and the sulfur atom of GSH is within hydrogen-bonding distance to the oxygen atom of Ser-15 (3.38 Å) in the wild-type hGSTZ1c-1c (Fig. 3A). But the thiol of GSH is toward the other side when Ser-14 is mutated to Cys and the distances between the sulfur/oxygen atoms of Cys-14/Ser-15 and the sulfur of GSH increase from 4.55 Å/3.38 Å to 7.95 Å/8.88 Å (Fig. 3B). Conversely, the distance between sulfur atoms of Cys-15 and GSH is beyond hydrogen-bonding distance in the mutant S15C (4.26 Å), and Ser-14 is far apart from the thiol of GSH (4.55–7.20 Å) because of the mutation (Fig. 3C). These results suggested that the mutations of Ser-14 or Ser-15 to Cys residues could have effect on the binding of GSH to the protein. However, unlike the two mutants above, the complex of hGSTZ1c-1c (C16S) with GSH does not change much in comparison with that of the wild-type hGSTZ1c-1c with GSH (Fig. 3D), which may be the reason why there is no dramatic decrease of GST activity.
A pseudocatalytic center consisting of Sec-17, Trp-18, and Gln-207 near the GSH binding site is detected in seleno-hGSTZ1c-1c (S17C) produced by cysteine auxotrophic strain system (Fig. 3E). The structural geometry of the pseudocatalytic cluster is similar to that of the active site of natural GPX and the distance between selenium atom of Sec-17 and sulfur of GSH is 10.65 Å.
A structural comparison between seleno and nonseleno hGSTZ1c-1c revealed that the introduction of multi-Sec residues into hGSTZ1c-1c could lead to structural changes in the protein. There are five Cys residues (Cys-16, Cys-137, Cys-154, Cys-165, and Cys-205) in the wild-type hGSTZ1c-1c, and a putative 3D structure of seleno- hGSTZ1c-1c was created with all the Cys residues replaced by Sec residues (Fig. 3F). The conformations of Cys residues were coincident with the Sec residues at positions 154 and 165 when the substitutions of multi-Sec for Cys residues were carried out, while those differed from the Sec residues at positions 16, 137, and 205. This implied that the substitutions of some Sec for Cys may cause the conformational change of the protein, although no disulfide bond was detectable in hGSTZ1c-1c.
It has been reported that the Sec-14 and Sec-15 of selenium-containing hGSTZ1-1 (Se-hGSTZ1-1) produced by chemical converting Ser to Sec residue contributed significantly to the GPX activity (9). But it was difficult to prepare selenoenzyme in large amount due to the restrictions of chemical method. Instead, cysteine auxotrophic strain system was used to produce GPX mimics (14). However, after the recombinant selenoprotein of seleno-hGSTZ1c-1c (S14C/S15C) was prepared and purified successfully, no GPX activity was detected in this study. Therefore, it is worthwhile to figure out the reason for the loss of GPX activity in seleno-hGSTZ1c-1c (S14C/S15C).
Mutations of the SSC motif (Ser14-Ser15-Cys16) in hGSTZ1c-1c were performed in this experiment, and the substitutions of the first two Ser residue(s) at positions 14 or/and 15 caused dramatic loss of catalytic activity with CFA as substrate although the Ser-to-Cys mutation is thought to be conservative. This can be explained from the molecular modeling results. Figure 3B shows that the thiol of GSH is orientated away from both Cys-14 and Ser-15 as the mutation of Ser-14 to Cys is performed, indicating a significant role of the hydroxyl group in Ser-14 in orientation and binding of GSH. But the participation of Ser-14 in catalysis could not be excluded due to the fact that the distance extension between Ser-14 and GSH in the mutant S15C (Fig. 3C) may preclude the nucleophilic attack of GSH on CFA. Compared with the mutant S14C, the loss of GST activity of hGSTZ1c-1c from mutation of Ser-15 to Cys may be attributed to the lack of the catalytic group. According to Fig. 3C, the distance between the sulfur atoms of Cys-15 and GSH does not extend much in the mutant S15C than in the wild-type protein (from 3.38 to 4.26 Å). And combined with the obvious GPX activity of the Ser-15 to Sec mutant produced in eukaryotes in our previous study (13), it is conceivable that the Ser/Sec-15 probably involves little in binding GSH but has a great role in catalysis for GST/GPX activity. Unlike the importance of Ser-14 and Ser-15, the GST activity is mostly retained when Cys-16 is mutated to Ser. It is shown that the relative position of the residues in the SSC motif and GSH does not change much in the mutant C16S (Fig. 3D). And taking into account the dramatic loss of catalytic activity when hydroxyl group in Ser-15 was replaced by sulfhydryl group, Cys-16 may have little or no effect in catalysis.
According to our previous research, it was speculated that Sec-17 would be another potential catalytic group contributing significantly to GPX activity because of the pseudocatalytic triad consisting of Sec-17, Trp-18, and Gln-207 detected in Se-hGSTZ1-1 (Ser17Sec) (13). Besides, the result shown above indicated that Ser-17 in hGSTZ1c-1c was not involved in binding GSH, its mutation may thereby have low impact on the GSH-binding ability of seleno-hGSTZ1c-1c (S17C) and GPX activity would probably be introduced subsequently. However, this hypothesis is not supported by our finding that seleno-hGSTZ1c-1c (S17C) produced using cysteine auxotrophic strain system only shows insignificant GPX activity (Table 3). A possible explanation for the result is that the distance between the selenium atom in Sec-17 and sulfur atom of GSH is so far (10.65Å) that it may go against the redox reaction (Fig. 3E). It is notable that the mutant S17C had almost no change in the catalytic activity with CFA as substrate, but the activity diminished most when multi-Sec residues were introduced into the protein. Similar situations were observed for the wild-type hGSTZ1c-1c and the mutant C16S. And combined with the loss of the GPX activity in seleno-hGSTZ1c-1c (S15C), it could be proposed that the substitutions of the multi-Sec for Cys residues, specially at positions 16, 137, and 205, may cause the structural change in hGSTZ1c-1c and affect binding of GSH to the protein as a result. That would be another reason for the low GPX activity of seleno-hGSTZ1c-1c (S17C) although the Sec, Gln, and Trp residues were even closer in the pseudocatalytic center than in seleno-hGSTZ1-1 (Ser15Sec) produced in eukaryotic cells (13), and in Se-hGSTZ1-1 produced by chemical method (9). And the loss of the GST activity of seleno-hGSTZ1c-1c (S17C) may be attributed to that either. Accordingly, the effect of introductions of multi-Sec residues on binding GSH to the selenoproteins may be another contributor to the loss of the GST activity and the expected GPX activity for seleno-hGSTZ1c-1c (S14C, S15C, and S14C/S15C) besides the importance of Ser-14 for binding GSH.
In conclusion, we explored the different role of the residues in SSC motif, and a new insight was gained from the change of structure and activity after introducing Sec residues into hGSTZ1c-1c by site-directed mutagenesis and computational simulation. The conclusions outlined here could be used to understand the residues involved in GSH binding and maintaining the structure of the whole complex in GPX mimics. And due to the influence of multi-Sec substitutions on protein structure, the protein containing fewer cysteines should be chosen as the ideal protein scaffold for imitating GPX using cysteine auxotrophic strain system. Although seleno-hGSTZ1c-1c (S17C) and (S14C/S15C) do not show significant GPX activity, Sec site-directed-substitution and the clear structure information encourage us to make a further study to screen the ideal protein scaffold and new mutants with high GPX activity and develop their applications in industry and medicine.
The authors thank Prof Marie-Paule Strub and August Böck for providing the E. coli cysteine auxotrophic strain, BL21(DE3)cys, and Philip G. Board for providing the plasmid pQEZ1(human GST Zeta 1c-1c). This work is supported by the National Natural Science Funds, China (Nos. 30870540, 30970633, and 31270851).
- 19Accelrys Inc. ( 1998) Insight II, Version 98.0. Accelrys Inc., San Diego, CA.
- 20Accelrys Inc. ( 1999) Discover 3 User Guide. Accelrys Inc., San Diego, CA.
- 21Accelrys Inc. ( 1999) Profile-3D User Guide. Accelrys Inc., San Diego, CA.
- 23Accelrys Inc. ( 1999). Affinity User Guide. Accelrys Inc., San Diego, CA.