The high Mr (~55 kDa) thioredoxin reductases (TR) characteristic of higher eukaryotes are members of the glutathione reductase (GR) family of pyridine nucleotide disulfide oxidoreductases. These homodimeric enzymes catalyze the reduction of a cognate disulfide substrate. During the enzymatic cycle, reducing equivalents pass from NADPH to the conserved active site disulfide via an enzyme-bound FAD and then to the cognate substrate. TRs are unique in the family as electrons are then transferred to the C-terminal active site of the adjacent molecule as part of a 16 amino acid extension (in place of the cognate GR substrate GSSG), prior to transfer to the substrate thioredoxin. Each electron transfer step occurs via thiol-disulfide exchange in a multi-step process mediated by a conserved catalytic acid/base. Mammalian TRs require selenocysteine (Sec) incorporated into the Gly-Cys-Sec-Gly-OH (GCUG) C-terminal tetrapeptide motif, while the TR from Drosophila melanogaster (DmTR) does not, and instead contains a Ser-Cys-Cys-Ser-OH (SCCS) tetrapeptide motif indicating that Sec is not universally necessary to catalyze the reduction of thioredoxin. This project has achieved three major objectives; 1) development of a semisynthetic method for production of mouse mitochondrial TR (mTR3) for structure-function studies, 2) establishment of a new method to study the mechanism of TR by using tetrapeptides in the oxidized form equivalent to the C-terminal active sites as substrates for the truncated forms of both enzymes, 3) determination of the crystal structure of DmTR. The results show that the structure of DmTR explains the biochemical data and has developed a new testable hypothesis in the field for the requirement of Sec in mammalian TR. We demonstrate that the tetrapeptides tested in Aim 2 were all better substrates for DmTR. The data also shows a far greater dependence on Sec for mTR3 than DmTR, which is in agreement with that observed for the collection full-length mutants produced for each enzyme in Aim 1. As this method of investigation is more analogous to the other enzymes of the GR family, the structures of the tetrapeptides determined by NMR spectroscopy were oriented in the active site of the both enzymes using the diglutathione bound in the structure of GR as template. DmTR appears to have a more open active site than observed in the known structure of mTR3. Residues from the helical face of the FAD-domain proximal to the FAD-associated active site are less bulky in DmTR to accommodate the hydroxyls of the serines. This is likely to make the enzyme more amenable for the conformational switching of the SCCS peptide necessary to protonate the leaving group cysteine by the proposed catalytic acid/base. In contrast, mTR3 shows a more restricted interface by incorporating bulkier residues at the interface in conjunction with the smaller Gly residues of the C-terminal sequence GCUG. The tetrapeptides display a conformational preference not suitable for protonation of the first leaving group in mTR3.

published_or_final_version / Biochemistry / Master / Master of Philosophy

Peroxiredoxins are a family of multifunctional enzymes that are able to protect the cell against oxidative stress. Peroxiredoxins form part of a recently discovered peroxide scavenging system along with thioredoxin, thioredoxin reductase and sulfiredoxin. This study describes the purification of a recombinant human peroxiredoxin II from human erythrocytes. The original recombinant clone contained a point mutation at the fourth residue from glycine to valine and a number of problems were encountered with aggregation during purification. Reverting back to the original amino acid sequence allowed the protein to be purified and concentrated without aggregation, as well as leading to over-expression in the same oligomeric state as the native sample from blood. This study also describes the over-expression and purification of the human peroxiredoxin II protein in the intermolecular disulfide form as well as the subsequent crystallisation and X-ray diffraction studies. The crystal structure for this form of the protein was obtained to 3.3 Å resolution revealing the peroxiredoxin to be in the decameric form. In addition conformational changes in the protein that are necessary for formation of the intermolecular disulfide between the peroxidatic (Cys52) and resolving cysteine (Cys172) have been observed. The structure also revealed that these movements did not interfere with the dimer:dimer interface as had been previously suggested. This then allows the disulfide to be seen within the decameric form of peroxiredoxin. The production of covalent complexes formed between peroxiredoxin and sulfiredoxin, and peroxiredoxin and thioredoxin was also investigated. Complexes were stabilised by using DTNB to form a covalent bond between specific cysteine residues. The complex binding results from size exclusion chromatography showed that decameric peroxiredoxin bound to sulfiredoxin in a 1:5 ratio and decameric peroxiredoxin bound to thioredoxin in a 1:10 ratio. Cloning, over-expression and purification of the selenocysteine containing enzyme thioredoxin reductase was achieved. A minimal selenocysteine insertion sequence was added to the 3’ end of the DNA sequence to drive selenocysteine insertion in place of the typical stop UGA codon. The activity of this protein was found to be low but was greatly increased when co-expressed with a plasmid containing the selA, selB and selC genes. Although the activity of this co-expressed thioredoxin reductase was ~20% of the native enzyme activity, it was comparable to the activity of other recombinant forms of the enzyme. These studies report the purification of all of the proteins necessary to reform the peroxiredoxin system and allow the production of a working assay for peroxiredoxin activity. Together with the first report for a structure of a decameric disulfide form of human peroxiredoxin II a greater insight into the peroxiredoxin system has been obtained.

published_or_final_version / abstract / Obstetrics and Gynaecology / Master / Master of Philosophy

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfillment of the requirements for the degree of Doctor of Philosophy. Johannesburg, 2012 / The thioredoxin-like (Trx-like) fold is preserved in various protein families with diverse functions despite their low sequence identity. Glutathione transferases (GSTs) are characterised by a conserved N-terminal domain with a thioredoxin–like βαβαββα secondary structure topology and an all alpha-helical domain. GSTs are the principal phase II enzymes involved in protecting cellular macromolecules from a wide variety of reactive electrophilic compounds. It catalyses the conjugation of reduced glutathione (GSH) to an electrophilic substrate to form a hydrophilic and non-toxic compound. The binding site for GSH (G-site) is located in the N-terminal domain of GSTs. The sequence identity within members of the Trx-like superfamily is low; however, the members of this family fold into a conserved βαβαββα topology. It, therefore, seems reasonable that there are topologically conserved residues within this fold whose main role is to drive folding and/or maintain the structural integrity of the Trx-like fold. Structural alignments of the N-subdomain (βαβ motif) of the GST family shows that Leu7 in β1 and Leu23 in α1 are topologically conserved residues. The Leu7 side chain is involved in the packing of α1β1α2 and α3, whilst Leu23 is mainly involved in van der Waals interactions with residues in α1 and the loop region connecting α1 and β2. Taking into account the types of interaction that both Leu7 and Leu23 are involved in, as well their location in close proximity to the G-site, it was postulated that both these residues may play a role in the structure, function and stability of the GST family of proteins. Leu7 and Leu23 are not directly involved in the binding of GSH but they could be important in maintaining the G-site in a functional conformation via correct packing of the Nsubdomain. The homodimeric human class Alpha of GST (hGSTA1-1) was used as the representative of the GST family to test this hypothesis. The bulky side chains of Leu7 and Leu23 were replaced with a less bulky alanine residue to prevent altering the hydrophobicity of the βαβ motif. The effect of the mutation on the structure, function and stability of hGSTA1-1 was, therefore, studied in comparison with the wild-type using spectroscopic tools, X-ray crystallography, functional assays and conformational stability studies. The impact of the mutations on the structure of the enzyme was determined using spectroscopic tools and X-ray crystallography. The X-ray structures of the L7A and L23A mutants were resolved at 1.79 Å and 2.2 Å, respectively. Analysis of both X-ray structures shows that the mutation did not significantly perturb the global structure of the protein, which correlates with far-UV CD and intrinsic fluorescence spectroscopic data. In addition, structural alignments using the C-alpha gave root mean square deviation (r.m.s.d) values of 0.63 Å (L7A) and 0.67 Å (L23A) between the wild-type and mutant structures. However, both the L7A and L23A structures showed the presence of a cavity within the local environment of each mutation. The functional properties of the mutants were also similar to those of the wild-type as determined by specific activity and 8-anilino-1-naphthalene sulfonate (ANS)-binding, indicating that Leu7 and Leu23 are not involved in the function of hGSTA1- 1. The conformational stability of L7A and L23A proteins was probed using thermal-induced unfolding, pulse proteolysis and urea-induced equilibrium unfolding studies. The thermal stability of L7A and L23A hGSTA1-1 was reduced in comparison to the wild-type protein. This was consistent with proteolytic susceptibility of L7A and L23A proteins which indicates that both mutants are more prone to thermolysin digestion when compared to wild-type hGSTA1-1. This also correlates with urea-induced equilibrium studies. The ΔG(H2O) value (23.88 kcal.mol-1) for the wild-type protein was reduced to 12.6 and 10.49 kcal.mol-1 in L7A and L23A hGSTA1-l, respectively. Furthermore, the m-values obtained for the L7A and L23A proteins were 1.46 and 1.06 kcal.mol-1.M-1 urea, respectively; these were much lower than that obtained for the wild-type protein (4.06 kcal.mol-1.M-1 urea). The low m-values obtained for the mutant proteins indicated that the cooperativity of hGSTA1-1 unfolding was significantly diminished in both mutations. The results obtained in this study indicate that the topologically conserved Leu7 and Leu23 in the N-subdomain of hGSTA1-1 play a crucial role in maintaining the structural stability of the thioredoxin-like domain and are not involved in the function of the enzyme.

The thioredoxin-like fold is a well conserved fold that is present in many families of proteins. One such superfamily of proteins include the GSTs which are involved in phase II detoxification. GSTs primarily catalyse the metabolism of xenobiotics but are also involved in transporting non–substrate ligands and reactive compounds. The GST fold comprises an N-terminal thioredoxin domain and an all alpha helical C-terminal domain and is present in at least 18 classes of proteins. The N-terminal thioredoxin domain is characterised by the βαβαββα topology and can be further divided into two structural motifs, an N-terminal (βαβ) and a C-terminal (αββα) motif. A well conserved hydrophobic network exists between these two motifs and the role of the C-terminal motif is elucidated in this study using class Alpha GST as a model protein. A topologically conserved valine (Val58) and an isoleucine (Ile75) located on β3 and α3, respectively, were mutated to alanine. Secondary and tertiary structural characterisation as well as ligandin function of the mutant enzymes displayed no major structural alteration with respect to the wild-type enzyme. This was confirmed with high resolution crystal structures obtained. Enzymatic activity was maintained indicating that no structural alterations have occured that affects the active site dynamics and the domain interface as a result of the induced mutations. Thermal denaturation studies, however, indicated a slight destabilisation in the enzyme in the case of the valine mutation, but a large destabilisation was witnessed as a result of the isoleucine mutation. This is further observed in denaturant-induced equilibrium studies where the thermodynamic stability of proteins can be determined. Furthermore, as a result of the isoleucine mutation, the enzyme unfolds via a populated intermediate in contrast to the wild-type which globally unfolds via a two-state mechanism with no stable intermediates being populated. Pulse-proteolysis was employed as an additional probe for thermodynamic stability where the enzyme was digested by thermolysin at varying denaturant concentrations. Pulse-proteolysis results were in agreement with the thermal and denaturant-induced stability studies further confirming that the isoleucince substitution causes a large destabilisation. Thus these conserved hydrophobic residues of the thioredoxin C-subdomain play a crucial stabilising role in the GST fold.

"December 2002" / Errata inside front cover. / Bibliography: leaves 148-176. / 176 leaves : ill. (some col.), plates ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Plant Science, 2003

Reporter gene transactivation by human p53 is inhibited in budding yeast lacking the TRR1 gene encoding thioredoxin reductase. Thioredoxin reductase specifically catalyzes the NADPH-dependent reduction of thioredoxin. Thioredoxin provides a source of electrons for disulfide reduction in various cellular processes. Reduction of disulfides within the cell can be accomplished by the separate but partially overlapping glutathione reductase - glutathione - glutaredoxin pathway. The basis for p53 inhibition was investigated by measuring the redox state of thioredoxin and glutathione in wild-type and Δtrr1 yeast lacking the gene encoding thioredoxin reductase. The Δtrr1 mutation caused an increased in oxidation in both molecules. Highcopy expression of the GLR1 gene encoding glutathione reductase in Δtrr1 yeast restored the redox state of glutathione to wild-type levels, but did not restore p53 activity. Also, p53 activity was unaffected by be a Δglr1 mutation, even though the mutation was known to result in glutathione oxidation. These results indicate that p53 activity has a specific requirement for an intact thioredoxin system, rather than a general dependence on the intracellular reducing environment. In order to test if p53 activity requires an intact thioredoxin system in mammalian cells, dominant-negative and RNAi approaches designed to suppress thioredoxin reductase activity were used in a breast adenocarcinoma cell which contains an endogenous wild-type p53. In cells stably transformed with a plasmid encoding a dominant-negative form of thioredoxin reductase, thioredoxin reductase activity was inhibited 4.3-fold and p53 reporter gene expression was inhibited by 2-fold. In cells stably transformed with a RNAi plasmid designed to target thioredoxin reductase mRNA, thioredoxin reductase activity was inhibited by 1.7-fold and p53 reporter gene expression was inhibited by 1.6-fold. A decrease in the protein levels of the p53 endogenous target genes p21 and Bax was also observed in both dominant-negative and RNAi transformants. Additionally, thioredoxin was shown to bind p53 in vitro (Kd=0.9 μM), and a LexA-thioredoxin fusion protein was shown to bind p53 in vivo. These results suggest that p53 activity is regulated by thioredoxin reductase in mammalian cells through a direct interaction with thioredoxin. / Graduation date: 2004

E. coli thioredoxin (TRX) was modified by the episulfonium ion derived from S-(2-chloroethyl)glutathione (CEG) or S-(2-chloroethyl)cysteine (CEC). The alkylation site was located at Cys-32, which was confirmed by tandem mass spectrometry. Two forms of native TRX, Oxi- and Red-TRX, and two modified TRXs, GS- and Cys-TRX, were examined by hydrogen/deuterium (H/D) exchange reactions using electrospray ionization mass spectrometry (ESI-MS) for the analysis. Conformational dynamics during thermal denaturation were probed by H/D exchange-in experiments. Under conditions in which the folded conformational state is marginally stable, H/D exchange-in experiments resulted in mass spectra differing in the number of incorporated deuteriums which indicates the presence of two distinct populations of molecules. As the exchange-in time increased, the population representing the unfolded state increased and the population for the folded state decreased. The rate of conversion was used to estimate the rate constant of unfolding. ESI mass spectra were also recorded as a function of temperature without H/D exchange, and the observed bimodal charge state distributions were analyzed in order to estimate melting temperatures. GS-TRX showed increased resistance to hydrogen isotope exchange in comparison with Red-TRX indicating that there were enhanced intramolecular interactions in the former protein. Pepsin digestion was performed on deuterated TRXs to analyze different structural regions. The amount of deuterium incorporated was monitored with peptic peptides from deuterated TRXs with different exchange-in incubation periods. Deuterium levels of each peptide were plotted versus the exchange time and fitted with a series of first-order rate terms. The regions 59-80 and 81-108 of Oxi- and Red-TRX showed an EX1 mechanism as evidenced by two distinct mass envelopes that appeared after a short incubation time in deuterated solvent. Tandem mass spectrometry (MS/MS) was carried out to obtain the information on individual amide linkages. MS/MS data showed generally excellent correlations with the exchange rate constants from published NMR data on Oxi- and Red-TRXs. Two residues, Ile-75 and Ala-93 in GS-TRX indicated the most probable sites responsible for induced H-bonding by the attached glutathionyl group, which was consistent with the energy minimized structure predicted by AMBER force field constants. / Graduation date: 2001

Thesis (M. Phil.)--University of Hong Kong, 2008. / Also available in print.