Anal. Chem. 2009, 81, 1543–1548
Electrochemical Determination of Thioredoxin Redox States Vlastimil Dorc ˇ a´k and Emil Palecˇek* Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i, Kra´lovopolska´ 135, CZ-612 65 Brno, Czech Republic Thioredoxin (TRX) is a general protein disulfide reductase with a large number of biological functions, including its roles in human diseases. The TRX redox mechanism is based on reversible oxidation of two cysteine thiol groups to a disulfide, accompanied by the transfer of two protons. Using constant-current chronopotentiometric stripping analysis (CPSA) and the electrocatalytic TRX peak H, we have determined redox states of TRX at submicromolar TRX concentrations. A concentration of 1 nM TRX produces a well-developed peak H at moderate accumulation time without stirring. On the basis of this peak, interactions of 4-hydroxy-2-nonenal (HNE, product of lipid peroxidation) with TRX and the formation of TRX-HNE adducts were studied. CPSA of TRX at a carbon electrode is less sensitive and does not discriminate between reduced and oxidized forms of TRX. Thioredoxin (TRX) is an ubiquitous protein occurring in all species from bacteria to man.1 It has a large number of biological functions, including its role in human diseases such as cancer, cardiac conditions, viral diseases, aging, etc.2 In most of its functions TRX serves as a general protein disulfide reductase.1 Its seemingly simple redox mechanism is based on reversible oxidation of two Cys thiol groups to a disulfide, accompanied by the transfer of two protons. Reduction of TRX is manifested by an increase of Trp fluorescence,3 originating from a decreased quenching of Trp28. Fluorescence of this amino acid is thus a useful probe for direct determination of thiol-disulfide exchange reactions in TRX. To our knowledge no electrochemical method for determination of TRX redox states has been reported. Polarography of proteins was applied in biochemistry and medicine and particularly in oncology for several decades around the middle of the 20th century (reviewed in refs 4-7). At the * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Holmgren, A. Structure 1995, 3, 239–243. (2) Burke-Gaffney, A.; Callister, M. E. J.; Nakamura, H. Trends Pharmacol. Sci. 2005, 26, 398–404. (3) Stryer, L.; Holmgren, A.; Reichard, P. Biochemistry 1967, 6, 1016–1020. (4) Zuman, P.; Palecek, E. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; pp 755-771. (5) Brezina, M.; Zuman, P. Polarography in Medicine, Biochemistry and Pharmacy; Interscience: New York, 1958. (6) Heyrovsky, M. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; Vol. 1, pp 657-687. 10.1021/ac802274p CCC: $40.75 2009 American Chemical Society Published on Web 01/26/2009
beginning of the 1970s electrochemists turned their attention to some enzymes and to direct electrochemistry of a limited number of conjugated proteins containing redox-active centers (such as metal ions), creating a new important research area. It has been shown that such proteins can yield a reversible electrochemical process at solid electrodes (reviewed in refs 8-11) which can be related to biological functions of these proteins. On the other hand, the potentialities of electrochemical methods as tools for analysis of the majority of proteins were neglected for several decades. Electrochemical analysis of proteins not containing any nonprotein redox center for fast reversible electrochemistry is limited to studies of their adsorption/desorption behavior at different electrodes,7,12 oxidation of tyrosine (Tyr) and tryptophan (Trp) residues at carbon electrodes,13-15 and reduction processes at mercury electrodes.7,12 Carbon electrodes have been increasingly applied in bioelectrochemistry16 and particularly in the analysis of nucleic acids and proteins.7,12,17,18 Mercury electrodes, which were almost abandoned in the 1980s, are reappearing now as highly sensitive tools in protein research.7,19-25 (7) Palecek, E. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; Vol. 1, pp 690-750. (8) Armstrong, F. A. In Bioelectrochemistry; Wilson, G. S., Ed.; Wiley-VCH: Weinheim, Germany, 2002; Vol. 9, pp 11-29. (9) Ferapontova, E. E.; Shleev, S.; Ruzgas, T.; Stoica, L.; Christenson, A.; Tkac, J.; Yaropolov, A. I.; Gorton, L. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; Vol. 1, pp 517-598. (10) Wackerbarth, H.; Zhang, J.; Grubb, M.; Hansen, A. G.; Ooi, B. L.; Christensen, H. E. M.; Ulstrup, J. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; Vol. 1, pp 485-516. (11) Warsinke, A.; Stocklein, W.; Leupold, E.; Micheel, E.; Scheller, F. W. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; pp 451-484. (12) Palecek, E.; Ostatna, V. Electroanalysis 2007, 19, 2383–2403. (13) Brabec, V. Bioelectrochem. Bioenerg. 1980, 7, 69–82. (14) Palecek, E.; Jelen, F.; Teijeiro, C.; Fucik, V.; Jovin, T. M. Anal. Chim. Acta 1993, 273, 175–186. (15) Cai, X. H.; Rivas, G.; Farias, P. A. M.; Shiraishi, H.; Wang, J.; Palecek, E. Anal. Chim. Acta 1996, 332, 49–57. (16) McCreery, R. L. Chem. Rev. 2008, 108, 2646–2687. (17) Wang, J.; Kawde, A. N.; Musameh, M.; Rivas, G. Analyst 2002, 127, 1279– 1282. (18) Wang, J.; Liu, G. D.; Jan, M. R.; Zhu, Q. Y. Electrochem. Commun. 2003, 5, 1000–1004. (19) Palecek, E.; Ostatna, V.; Masarik, M.; Bertoncini, C. W.; Jovin, T. M. Analyst 2008, 1, 76–84.
Analytical Chemistry, Vol. 81, No. 4, February 15, 2009
1543
Scheme 1. Sequence of E. coli Thioredoxina
a One-letter code of amino acid residues is used. Electroactive residues are black; nonelectroactive are gray. Residues containing H+ donors are bold. Nonelectroactive residues with no H+ donor are thin. Italic style indicates hydrophobic side chain. G refers to glycine residue with no side chain.
The ability of proteins to catalyze hydrogen evolution at the dropping mercury electrode (DME) was discovered in 1930.26 The catalytic dc polarographic presodium wave of proteins (obtained with the DME) appeared, however, too close to the background discharge, and its shape was poorly developed and difficult to measure; this wave was thus considered not suitable for analytical purposes (reviewed in refs 5, 27, and 28). About 10 years ago we found that peptides and proteins produce a well-developed peak at highly negative potentials at the hanging mercury drop electrode (HMDE), if constant-current chronopotentiometric stripping analysis (CPSA) was used.7,29 Similar to the presodium wave is catalytic hydrogen evolution that is responsible for this peak, which we designated as peak H because of Heyrovsky´, hydrogen evolution, and high sensitivity.7 Peak H differs from other electrochemical protein signals particularly (i) by its ability to detect peptides and proteins down to nanomolar and subnanomolar concentrations and (ii) by its remarkable sensitivity to local and global changes in protein structure.21,23,24 Recently, we showed using peak H that reduced and oxidized peptides can be discriminated.30 TRX from Escherichia coli is a well-characterized protein.31 Among 108 amino acid residues forming the TRX molecule (Scheme 1), only 6 contain electroactive side chains: two cysteines (Cys32, Cys35), two tyrosines (Tyr49, Tyr70), and two tryptophans (Trp28, Trp31). In addition, amino acid residues with labilized protons such as lysine or arginine and cysteine can be involved in the electrocatalysis of the hydrogen evolution responsible for peak H (Scheme 1). Their participation in the electrode process strongly depends on the experimental conditions and particularly on pH, nature and concentration of ions in the background electrolyte, amino acid composition, and conformation of the protein or peptide. For example, peptides, which do not contain cysteine residues, usually produce peak H at acid pHs but not in alkaline media.7 (20) Palecek, E.; Masarik, M.; Kizek, R.; Kuhlmeier, D.; Hassmann, J.; Schulein, J. Anal. Chem. 2004, 76, 5930–5936. (21) Ostatna, V.; Palecek, E. Electrochim. Acta 2008, 53, 4014–4021. (22) Ostatna, V.; Kuralay, F.; Trnkova, L.; Palecek, E. Electroanalysis 2008, 12, 174. (23) Ostatna, V.; Uslu, B.; Dogan, B.; Ozkan, S.; Palecek, E. J. Electroanal. Chem. 2006, 593, 172–178. (24) Masarik, M.; Stobiecka, A.; Kizek, R.; Jelen, F.; Pechan, Z.; Hoyer, W.; Jovin, T. M.; Subramaniam, V.; Palecek, E. Electroanalysis 2004, 16, 1172–1181. (25) Kizek, R.; Trnkova, L.; Palecek, E. Anal. Chem. 2001, 73, 4801–4807. (26) Heyrovsky, J.; Babicka, J. Collect. Czech. Chem. Commun. 1930, 2, 370– 378. (27) Palecek, E. In Topics in Bioelectrochemistry and Bioenergetics; Milazzo, G., Ed.; J. Wiley: London, 1983; Vol. 5, pp 65-155. (28) Heyrovsky, J.; Kuta, J. Principles of Polarography; Czechoslovak Academy of Science: Prague, 1965. (29) Tomschik, M.; Havran, L.; Fojta, M.; Palecek, E. Electroanalysis 1998, 10, 403–409. (30) Dorcak, V.; Palecek, E. Electroanalysis 2007, 19, 2405–2412. (31) Holmgren, A. Annu. Rev. Biochem. 1985, 54, 237–271.
1544
Analytical Chemistry, Vol. 81, No. 4, February 15, 2009
In this report we show that peak H and oxidation peaks of tyrosine and tryptophan residues at a pyrolytic graphite electrode (PGE) can be used to analyze TRX. Peak H yields, however, better sensitivity of the analysis than oxidation peaks obtained with PGE. Using peak H we are able to discriminate between reduced and oxidized forms of TRX, and this peak can be applied to studies of TRX chemical reactivity toward 4-hydroxy-2-nonenal (HNE). EXPERIMENTAL SECTION Reagents. Thioredoxin from E. coli as well as tris(2-carboxyethyl)phosphine hydrochloride and azodicarboxylic acid bis(dimethylamide) were purchased from Sigma. HNE was from Calbiochem. All chemicals for preparation of buffers were ACS reagents from Sigma. Water (ACS reagent) was purchased from Aldrich. Apparatus. All chronopotentiometric measurements were carried out with an Autolab analyzer (Eco Chemie, Utrecht, The Netherlands) in connection with VA-stand 663 (Metrohm, Herisau, Switzerland). The three-electrode system consisted of an HMDE or PGE (basal plane; its surface was renewed using adhesive tape; more details were described in ref 32) as working electrode (areas of 0.4 mm2 or ∼12.0 mm2, respectively), Ag|AgCl|3 M KCl as a reference electrode, and Pt wire as a counter electrode. Our previous analysis of nucleic acids and proteins at carbon electrodes showed that usual PGE and highly oriented pyrolytic graphite electrode yielded in principle the same results.14 Procedures. To avoid oxidation of reduced protein by atmospheric oxygen, the analyte was dissolved in deaerated water and kept in an argon atmosphere. All measurements were carried out at 25 °C. The working volume in the electrolytic cell was 2 mL. Analytes were adsorbed on the electrode from unstirred solutions for an accumulation time (tA) at open circuit or at a given accumulation potential (EA). After accumulation, stripping current (Istr) was applied to record the E-t curve [which was automatically converted to a (dE/dt)-1-E curve] from the set initial potential (Ei). Deoxygenation of measured solution was necessary to avoid oxidation of reduced TRX. To keep the measured solution oxygen-free, argon was passed through the solution in the cell before the measurement. During the measurement argon was passed above the solution. RESULTS AND DISCUSSION TRX Oxidation at Carbon and Reduction at Mercury Electrodes. Pyrolytic Graphite Electrode. Using CPSA in combination with a PGE at accumulation time tA of 60 s and stripping current Istr of +25 µA, we found that 1 µM TRX yields at pH 9.5 two partially separated peaks at 0.63 and 0.73 V (Figure 1A). These two peaks might be due to partially resolved oxidation of Tyr (peak Y) and Trp (peak W) residues. Well-separated peaks for these residues were earlier obtained with luteinizing hormone releasing hormone.15 At pH 7.5 only a single TRX peak at 0.77 V is seen (Figure 1A); at pH 5.5 the highest peak is observed at potentials almost 100 mV more positive. Peaks were baselinecorrected; without baseline correction only inflections were obtained. (32) Tomschik, M.; Jelen, F.; Havran, L.; Trnkova, L.; Nielsen, P. E.; Palecek, E. J. Electroanal. Chem. 1999, 476, 71–80.
Figure 1. Constant-current chronopotentiometric stripping (CPS) responses of 1 µM thioredoxin (TRX) (A) at a pyrolytic graphite electrode (PGE) and (B) at a hanging mercury drop electrode (HMDE). (A) Oxidation peaks at PGE are smoothed and baseline-corrected. Gray lines show responses of the background electrolyte (thin line) and of TRX (thick line) at pH 5.5 without smoothing and baseline correction. (B) Peaks H at different pHs; neither smoothing nor baseline correction was used. Chronopotentiograms were recorded from an initial potential (Ei) of +0.1 V at a stripping current (Istr) of +25 µA (PGE) or -20 µA (HMDE) after 60 s of accumulation time (tA) at open circuit in 50 mM McIlvain buffer, pH 5.5 and 7.5, and in 50 mM sodium borate, pH 9.3.
Hanging Mercury Drop Electrode. In McIlvain buffer at pH 5.5 1 µM TRX yields a complex response showing three peaks: the smallest one at -1.81 V (H1) and two more negative at -1.84 (H2) and -1.88 V (H3) (Figure 1B). When the pH is increased to 7.5, peaks at -1.82 (H1) and -2.04 V (H2) appear. At pH 9.3 much smaller peaks are obtained at -1.84 (H1) and -2.00 V (H2) (Figure 1B). The heights of peak H increase with decreasing pH, which is in good agreement with the catalytic nature of these peaks.6,33,34 Around -0.7 V, TRX yields another peak, which is considerably smaller than peaks H; this peak is due to reduction of the Hg-S bond (peak S), and at pH 7.5 it is about 200-fold smaller than peak H. Peak S will not be considered in this paper because, under the conditions optimum for peak H, it is too small (for more details see the Supporting Information, Figure S-1). Our results suggest that both the oxidation peaks (Figure 1) and peak H can be of use in the electrochemical study of TRX. Redox States of TRX. Earlier we found that using peak H at pH 9.3 (but not oxidation peaks at carbon electrodes) that the redox state of cysteine-containing peptides can be determined.30 Here we show that redox states of about a 10-fold larger TRX protein molecule can be determined using the same approach. For 250 nM TRX two well-separated peaks H are observed: a small peak H1 and a much larger one H2 (Figure 2). Only the reduced form of TRX provides well-developed peaks H1 and H2 at EA +0.1 V (Figure 2A), whereas almost the same response as its oxidized form is obtained when EA is -0.8 V (Figure 2B). We measured the dependence of TRX peaks H1 and H2 on accumulation potential, EA, at pH 9.3 in a potential range from +0.15 to -1.7 V. Adsorption of TRX at a positively charged (33) Banica, F. G.; Ion, A. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed.; John Wiley and Sons: Chichester, U.K., 2000; pp 11115-11144. (34) Heyrovsky, M. Electroanalysis 2004, 16, 1067–1073.
HMDE (EA close to the anodic dissolution of mercury) resulted in significant differences between reduced and oxidized TRX peak areas (Ap) and peak potentials (Ep) (Figure 3). In this EA range both peaks H1 and H2 of reduced TRX are much higher than those of oxidized TRX. At more negative EA both peaks of the reduced and oxidized TRX were almost the same. Our results thus show that like the peptides,30 at carefully chosen positive EA, reduced and oxidized forms of TRX can be easily distinguished (Figures 2 and 3). On the other hand, at negative EA, CPS peaks H of reduced and oxidized TRX are almost the same, making it possible to determine the concentration of TRX regardless of its redox form (Figure 3). In contrast to peak H, using TRX oxidation signals and PGE we were not able to discriminate between reduced and oxidized TRX (not shown). Reaction of TRX with HNE. Studies of cytotoxic molecules involved in peroxidation reactions led to the discovery of a group of conjugated aldehydes with toxic potential.35-37 Among them the most abundant member, and a major end product of lipid peroxidation, was identified as HNE. This aldehyde derives from n-6 polyunsaturated fatty acids and is permanently formed at background levels in the physiological environment. At physiological concentrations (
