Studies of an active site mutant of the selenoprotein thioredoxin reductase: The Ser-Cys-Cys-Ser motif of the insect orthologue is not sufficient to replace the Cys-Sec dyad in the mammalian enzyme

Free Radical Biology & Medicine 41 (2006) 649 – 656 www.elsevier.com/locate/freeradbiomed

Original Contribution

Studies of an active site mutant of the selenoprotein thioredoxin reductase: The Ser-Cys-Cys-Ser motif of the insect orthologue is not sufficient to replace the Cys-Sec dyad in the mammalian enzyme Linda Johansson a , L. David Arscott b , David P. Ballou b , Charles H. Williams Jr. b , Elias S.J. Arnér a,⁎ a

Medical Nobel Institute for Biochemistry, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, SE-171 77 Stockholm, Sweden b Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, MI, USA Received 28 January 2006; revised 3 May 2006; accepted 4 May 2006 Available online 12 May 2006

Abstract We have mutated the redox active C-terminal motif, Gly-Cys-Sec-Gly, of the mammalian selenoprotein thioredoxin reductase (TrxR) to mimic the C-terminal Ser-Cys-Cys-Ser motif of the non-selenoprotein orthologue of Drosophila melanogaster (DmTrxR). The activity of DmTrxR is almost equal to that of mammalian TrxR, which is surprising, because Cys mutants of selenoproteins are normally 1-2 orders of magnitude less active than their selenocysteine (Sec) containing counterparts. It was shown earlier that the flanking Ser residues were important for activating the Cys residues in DmTrxR (Gromer, et.al. (2003) PNAS 100, 12618-12623). However, the “Drosophila mimic” mutant of the mammalian enzyme studied herein had <0.5% activity compared to wild-type. Rapid kinetic studies revealed that all of the redox centers of the mutant were active, but that the C-terminal dithiols were not effective reductants of thioredoxin. The charge-transfer complex of the two-electron reduced enzyme slowly disappeared as the N-terminal dithiols reduced the C-terminal disulfide. In wild-type enzyme, the selenenylsulfide is more difficult to reduce and the charge-transfer complex is more stable. These findings suggest that features in addition to the flanking Ser residues are important for facilitating the high activity of the insect enzyme and that the corresponding features are absent in mammalian TrxR. © 2006 Elsevier Inc. All rights reserved. Keywords: Thioredoxin reductase; Selenocysteine; Rapid-reaction kinetics

Introduction Mammalian TrxR is a flavoenzyme belonging to a family of homodimeric pyridine nucleotide–disulfide oxidoreductases that includes glutathione reductase, lipoamide dehydrogenase and mercuric ion reductase [1]. Mammalian TrxR contains selenocysteine (Sec) that is required for its activity [2–4]. Using reducing equivalents from NADPH, TrxR maintains high levels

Abbreviations: Trx, thioredoxin; TrxR, thioredoxin reductase; DmTrxR, TrxR from Drosophila melanogaster; Sec, selenocysteine; DTNB, 5,5′-dithiobis (2-nitrobenzoic acid), Dm-Trx-2, Trx-2 from Drosophila melanogaster. ⁎ Corresponding author. Fax: +46 8 311551. E-mail address: [email protected] (E.S.J. Arnér). 0891-5849/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.freeradbiomed.2006.05.005

of reduced thioredoxin, thereby forming the thioredoxin system, which is important for a wide range of cellular functions [5,6]. The form of TrxR found in archaea, bacteria, plants or yeast is a homodimer of around 35 kDa (referred to as small or low Mr TrxR) containing FAD and a redox active disulfide/dithiol; low Mr TrxR does not contain selenocysteine and has domain features distinct from those of the mammalian enzyme. The large or high Mr TrxR (subunit 54-58 kDa), which is present in higher eukaryotes, has a domain structure similar to that of glutathione reductase but with a third redox active group at the terminus of a C-terminal extension. The third redox active group is a selenenylsulfide/selenolthiol in mammals, or a disulfide/dithiol in insects [5–7]. The characteristic C-terminal motif is –Gly-Cys-Sec-Gly in mammals and –Ser-Cys-Cys-Ser in insects.

650

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

Sec can generally be regarded as taking the role of an extraordinarily reactive cysteine (Cys) homologue, and the direct substitution of Sec to Cys in selenoproteins typically decreases turnover by 10- to 100-fold [8], as also shown with mammalian TrxR [2,4]. The flow of reducing equivalents in high Mr TrxR is from NADPH to the enzyme-bound FAD, from the reduced flavin to the Nterminal disulfide (Cys59-Cys64), and subsequently from the nascent dithiol to the C-terminal selenenylsulfide (Cys497′Sec498′); the selenolthiol formed in the reduced enzyme is the direct reductant of thioredoxin as well as of most other substrates of the enzyme [4,9,10]. The mechanism for both the wild type Drosophila melanogaster thioredoxin reductase (DmTrxR) and the mammalian TrxR has been characterized in detail [4,9–14]. The catalytic cycle shown in Scheme 1 can be regarded to occur in two half-reactions: the reductive half-reaction involves all steps up to the interchange between the reduced enzyme and thioredoxin (k1-k16); the final reduction of Trx constitutes the oxidative half-reaction involving an intermolecular selenenylsulfide-linked intermediate (k17-k19) [4,9,11]. The enzyme cycles in catalysis between the 2-electron reduced form (EH2) and the 4electron reduced form (EH4) [11]. The steps represented by k1-k6 constitute a “priming” reaction that generally should need to occur only once in the cell. Under most physiological conditions where the NADPH concentration

is usually high, the enzyme probably does not become fully oxidized. Reduction of the enzyme by one equivalent of NADPH yields a mixture of the four distinct EH2 forms in which EH2B, a thiolate-flavin charge-transfer complex readily detected at 540 nm, predominates; reduction with two equivalents of NADPH yields a mixture of the two separate EH4 forms in which EH4B predominates [11,12]. Interestingly, while insect DmTrxR is, in principle, a Secto-Cys mutant of mammalian TrxR, it is nearly as active as the Sec-containing mammalian counterpart [15]. In a prior study of DmTrxR, it was shown that the flanking serine residues in the wild-type enzyme motif (SCCS) were critical for activation of the redox active Cys residues so that the enzyme could nearly match the catalytic efficiency of the mammalian enzyme [13]. It was also clear that when Sec was introduced into DmTrxR at the penultimate position, corresponding to that in the mammalian enzyme, the flanking Ser residues were no longer necessary for activity [13]. In the present study we wished to assess whether flanking Ser residues introduced into the C-terminal tetrapeptide of the mammalian enzyme could similarly compensate a Sec-to-Cys mutation and thereby support high activity in direct analogy to the insect orthologue. Therefore we mutated the mammalian TrxR (GCUG) gene from rat to provide a Cys-Cys pair with flanking Ser residues to mimic the C-terminal active site of DmTrxR (SCCS) and studied its activity, as well as the rapid-reaction kinetics of the

Scheme 1. Model of the catalytic cycle of the mammalian TrxR SCCS mutant. The reductive half-reaction involves the steps designated by rate constants k1-k16 and the oxidative half reaction the steps k17-k20. The curved arrows pointing from the thiolate anion, C64, towards the FAD indicate charge-transfer interaction. MC, Michaelis complex. (Principle based on [2,4,9,10,11,12,13,14]). The exchange of selenium in native mammalian TrxR for sulfur in the mutant protein studied here is indicated in parentheses.

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656 Table 1 Spectral properties of selected mammalian TrxR enzyme species Enzyme species a

Eox A EHA 2 or EH4 A EH2 · NADP+ or + EHA 4 · NADP B C EH2 , EH2 or EHB4 d e EHD 2 B EH4 · NADPH

Experimental procedures

Characteristic ε at the ε at distinct wavelengths wavelength characteristic (mM−1cm−1) (nm) wavelength 440 nm 540 nm 670 nm (mM−1cm−1) 464 464 b 670 c

11.3 0.17 2–2.5

540 464 540 f

2.2–2.8 11.3 4.1

8.1 0.65

∼8.7 8.1 7.3

0.14 0.15

651

0.14 0.14

2.2–2.8 0.14 0.14 0.14 4.1 1.3

The spectral features listed here were utilized for characterization of the mammalian TrxR SCCS mutant protein and are based upon either direct determinations of the mutant or were extrapolated from features of wild type enzyme as determined earlier (see text). a For nomenclature of the enzyme species, see text and Scheme 1. b When enzyme FAD is reduced to FADH- a large fraction of the visible absorbance is lost. Note however that the spectrum in Fig. 1, curve 5, has absorbance peaks at 394 nm and 625 nm due to the 10% added methyl viologen used as a mediator to transfer electrons from the added dithionite to the FAD. c The increase in absorbance at this wavelength is transiently observed in stopped-flow kinetic and spectral traces, occurring from ∼5 ms to 50 ms and is not shown in any of the figures. d The spectra shown in Fig. 1, curves 2 and 3, are typical for the enzyme species noted, having the thiolate FAD charge-transfer complex that gives an increased absorbance at 540 nm. It is not possible to distinguish between the three enzyme species with data presently available. e This species was assumed to have the same visible spectrum as oxidized enzyme. The data in Fig. 2, curves 2 and 3, determined from stopped-flow kinetics, suggests that this assumption is correct; that is, at equilibrium with less than one equivalent of reductant, the spectrum is like that of oxidized enzyme. f When more than 2 equivalents of NADPH are added to either wild-type or mutant enzyme, an enhancement of the thiolate-flavin charge transfer is observed, as seen in Fig. 2 curve 6. E. coli TrxR does not form a thiolate-flavin charge transfer complex. Therefore it is possible to determine the spectral characteristics of a NADPH-flavin charge transfer complex, best detected at 570 nm (Lennon, B.W. and C.H. Williams, Jr., Reductive half reaction of thioredoxin reductase from Escherichia coli. Biochemistry 36, 9464-9477, 1997). In all other members of the enzyme family including high Mr TrxR, where a thiolate-flavin charge transfer complex is stabilized, the NADPHflavin interaction is seen as an enhancement of the thiolate-flavin charge transfer (cf. the extinction coefficient in lines 4 and 6).

reductive and oxidative half-reactions. Scheme 1 provides a model for the catalytic cycle of the mutant enzyme, based upon the previous characterizations of native insect and mammalian enzymes, as described above. Select species of the enzyme have distinct spectral properties (Table 1) and those were analyzed here in order to probe the function of the mammalian TrxR SCCS mutant protein. We found that the mutant showed less than 0.5% activity with its normal substrates, although all redox active centers were functional. These findings provide conclusive evidence that flanking Ser residues in the C-terminal tetrapeptide motif of large TrxRs are not sufficient per se to alleviate the need for Sec in this family of enzymes. DmTrxR must have evolved with additional local active site features, not shared by the mammalian enzyme, that allow the flanking Ser residues to activate the redox active Cys residues in the insect enzyme. Parts of this study have previously been published as a preliminary conference presentation [16].

Construction of mammalian TrxR SCCS mutants The TrxR SCCS mutant was made from the previously published rat TrxR-encoding plasmid pET-TRSTER [17] using a Quik-Change Site-Directed Mutagenesis Kit (Stratagene) and the primers 5′-CCT CCA GTC CAG TTG CTG CAG TTA AGC CCC AG-3′ and 5′-CTG GGG CTT AAC TGC AGC AAC TGG ACT GGA GG-3′. The same kit was used for the additional Cys-mutants with the TrxR SCCS plasmid as template and the following primers; TrxR SCCS C458S: 5′GCC GCA CTC AAG AGC GGG CTG ACC-3′ and 5`-GGT CAG CCC GCT CTT GAG TGC GGC-3′, TrxR C475T: 5′GGC ATC CAC CCG GTC ACT GCA GAG ATA TTT AC-3′ and 5′-GTA AAT ATC TCT GCA GTG ACC GGG TGG ATG CC-3′. The TrxR SCCS C458S+C475T double mutant was constructed from the TrxR SCCS C475T mutant with the C458S primers. Protein expression and purification The recombinant Sec-containing rat TrxR was produced as described previously [17], with induction at late exponential phase to increase the efficiency of selenocysteine incorporation [18]. The enzyme was purified to apparent homogeneity as judged by SDS-PAGE and had a specific activity of approximately 15 U/mg in the classical DTNB (5,5′-dithiobis (2-nitrobenzoic acid) assay [19]. The expression of the nonselenoprotein TrxR SCCS mutant proteins were done in BL21 (DE3) using a standard protocol. In short, expression was induced at exponential phase with 0.5 mM IPTG, and subsequently continued for 16 hours at 24 °C. All TrxR forms were purified to apparent homogeneity using 2′5′-ADPSepharose as described earlier [17,18].

Fig. 1. Titration of the mammalian TrxR SCCS mutant with dithionite. Spectra show oxidized enzyme (1), enzyme reduced with 1 eq dithionite (2), 2 eq (3), 3 eq (4) and 4 eq (5). Peaks at 625 and 394 nm in spectrum 5 are due to methyl viologen radical. Inset shows the effects of the titration with dithionite on the absorbance at 464 nm (dots) and 540 nm (triangles).

652

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

Enzyme assays TrxR activity was measured using both the thioredoxincoupled insulin assay using recombinant human Trx, the direct DTNB assay, or the assay involving reduction of selenite, essentially as described previously [19,20]. For the selenite assay, 100 μM selenite and 150 μM NADPH were used. All assays were conducted at r.t. in a total volume of 0.5 ml. Titration with dithionite The TrxR SCCS mutant enzyme (10 μM subunit) kept in 100 mM phosphate buffer, 2 mM EDTA pH 7.6 was titrated anaerobically with dithionite (in 50 mM pyrophosphate buffer pH 9.0) at r.t. in the presence of 1 μM methyl viologen. The dithionite had been standardized by titration with lumiflavin-3-acetic acid [21]. Rapid reaction kinetics Rapid reaction kinetic studies were all conducted in a Hi-Tech DX-2 stopped-flow spectrophotometer under anaerobic conditions at 4 °C. The enzyme concentration after mixing was approximately 16 μM. The reaction of different amounts of NADPH with oxidized enzyme in 50 mM phosphate buffer, 0.3 mM EDTA, pH 7.6 was followed at 355 nm, 440 nm, 540 nm, and 670 nm. For the oxidative half reaction, the enzyme was pre-reduced with 2.2 equivalents of NADPH and reacted with different amounts of thioredoxin-2 from Drosophila melanogaster (DmTrx-2) in 50 mM phosphate buffer, 0.3 mM EDTA, pH 7.6, following absorbance at 464 and 540 nm. Analysis of mixed-disulfide with TrxR For the attempt to trap a stable intermediate between the TrxR SCCS mutant and Trx, the recombinant enzyme (9 μM of

Fig. 3. Oxidative half reaction of TrxR SCCS. In (A) the spectra of the TrxR SCCS, pre-reduced with 2.1 eq NADPH, before (2) and after (3) addition of 11 eq DmTrx-2 are shown. The oxidized spectrum (1) is given for comparison. In (B) the time dependence at 464 nm after mixing pre-reduced TrxR SCCS (1) and TrxR wild-type (2), with 5.6 eq and 3 eq DmTrx, respectively, are shown. In (C) the time dependence at 540 nm after mixing pre-reduced TrxR SCCS (3) and TrxR wild-type (4) with 5.6 eq and 3 eq DmTrx, respectively, are shown. Note that the time scale on the X-axis (B) extends to 500 s for curve 1.

TrxR SCCS, or wild-type TrxR as control) was incubated anaerobically at r.t. with 19 μM NADPH for 5 min, before 90 μM of recombinant human Trx C63S C74S mutant protein was added, and incubation was continued for 15 min. Samples were then analyzed with native SDS-PAGE and visualized by Coomassie stain. Results and Discussion Enzymatic activity in the mammalian TrxR SCCS mutant

Fig. 2. Reductive half reaction of TrxR SCCS. The main figure shows the oxidized (1) TrxR SCCS spectra together with the final spectra resulting from reactions with 0.4 eq (2), 0.8 eq (3), 1.7 eq (4), 3.3 eq (5) and 7.7 eq (6) NADPH. The inset shows the kinetics of the reduction with 0.8 eq NADPH at 440 nm (1) and 540 nm (3) and 7.7 eq NADPH at 440 nm (2) and 540 nm (4).

Using either DTNB, Trx, or selenite as substrates, we found that the SCCS mutant form of TrxR had <0.5% activity compared to wild type, although it had full FAD content as judged from the absorbance at 463 nm (not shown). This SCCS active site mutant was thereby even less active than a direct Sec-to-Cys mutant, i.e., having a GCCG active site

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

653

wild type DmTrxR [11,13], various mutants of DmTrxR [11,13], rat recombinant TrxR [12], or the human wild type enzyme [9]. Titration of TrxR SCCS with dithionite

Fig. 4. Analysis of mixed-disulfides. A Coomassie stained native SDS-PAGE show samples of 9 μM TrxR/TrxR SCCS (A) incubated anaerobically with either 19 μM NADPH (B) or 19 μM NADPH followed by incubation with 90 μM Trx (C).

motif, which has an activity of about 1-9% of wild type [2,4,22]. We therefore initiated spectroscopic titration and rapid kinetics analyses to further probe the catalytic features of this mutant, aiming to identify which steps in the catalytic mechanism were inefficient. Using these techniques the separate steps of the reductive and oxidative half reactions can be compared to such analyses previously done with both

Titration to fully reduce the SCCS mutant form of mammalian TrxR required approximately 3 eq of dithionite per enzyme subunit (Fig. 1), which corresponds to the reduction of the three redox active centers in wild-type TrxR, i.e., the FAD, the N-terminal disulfide, and the C-terminal selenenylsulfide [9,12], the latter which was here replaced by a disulfide. Thus, mutant protein contained all redox active centers. Therefore, the very low turnover rate of the TrxR variant can not be explained by lack of one or more of these redox active centers. Reductive half reaction measured in the rapid-reaction spectrophotometer Spectra and kinetics occurring in the reaction of NADPH with the SCCS mutant were compared to those observed with wild-type mammalian TrxR that contains Sec, which has been studied previously [9,10,12,23]. Fig. 2 shows the final spectra resulting from reaction of the SCCS mutant enzyme with various amounts of NADPH. The inset of Fig. 2 shows the kinetics of the reactions with either 0.8 eq of NADPH (Fig. 2, inset, curves 1 and 3) or 7.7 eq of NADPH (Fig. 2, inset,

Fig. 5. Redox active centers in mammalian TrxR and the two accessory Cys residues mutated in this study. In (A) the three redox active motifs of the active site of rat TrxR (the FAD and the N-terminal disulfide of Cys 59 and Cys 64 from one subunit and the C-terminal Cys 497 and Cys 498 of the other subunit of a Sec498Cys mutant), as well as the location of Cys 458 and Cys 475 are shown, as modeled from the crystal structure of rat TrxR U498C [14] (Protein Data bank ID code 1H6V) and visualized using the Swiss PDB viewer. Note that the dimeric holoenzyme has two active sites, while only one is shown here. In (B) an alignment of the C-terminal ends of DmTrxR and rat TrxR is shown, indicating the Cys 458 and Cys 475 by stars and the C-terminal redox active motifs demonstrated by a box.

654

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

curves 2 and 4). Flavin reduction is seen at 440 nm (Fig. 2, inset, curves 1 and 2), formation of the proximal thiolate-toFAD charge transfer complex, EH2B and EH4B, is observed at 540 nm (Fig. 2, inset, curves 3 and 4), and flavin reoxidation is monitored at 440 nm (Fig. 2, inset, curves 1 and 2). With 0.8 eq of NADPH the flavin was partially reduced in the first ∼0.02 s, and by ∼0.1 s it became partially re-oxidized as the flavin reduced the proximal disulfide. In a much slower reaction extending beyond 500 s, almost complete reoxidation of the flavin is observed as the nascent dithiol proximal to the FAD (species EH2B, Scheme 1) transfers reducing equivalents to the C-terminal disulfide to form EH2D. Complete reoxidation of the FAD is also observed with 0.4 eq NADPH (Fig. 2, spectrum 2). With 7.7 eq of NADPH the first two phases are similar in rate, although the extent of formation of FADH– is considerably greater than with 0.8 eq. The thiolate-to-FAD charge-transfer complex is more intense as EH4B is formed with the additional NADPH and as NADPH binds to EH4B (Fig 2, spectrum 6 and curve 4, inset). The corresponding reaction in the wild-type enzyme, i.e., reduction of the selenenylsulfide, is not as favored thermodynamically as is reduction of the disulfide; thus, with one eq NADPH, EH2B is the predominant species [9]. We conclude that differences in redox potentials between the EH2B and EH2D forms (Scheme 1) result from the mutations, so that the Cterminal disulfide in the mutant was more easily reduced than the selenylsulfide of wild type TrxR.

EH2 level. This phase was approximately 200-fold slower for the SCCS mutant than for the wild-type enzyme. This extremely slow oxidative half reaction explains the very low overall catalytic activity of the SCCS mutant (cf. curves 1 and 3 with curves 2 and 4 in Fig 3B and 3C, respectively). Presumably, the driving force for EH4B or EH2D to reduce DmTrx-2 is not sufficient to promote a rapid reduction of DmTrx-2.

Stopped-flow studies of the oxidative half reaction

Two additional Cys-residues are situated in the C-terminal part of TrxR, but are not present in the Drosophila orthologue, i.e., Cys 458 and Cys 475 (numbers based on the TrxR sequence from rat [17]). Both of these Cys residues could be important for TrxR function. Cys 475 is close to the active site, while Cys 458 is in a more distant surfaceexposed position [14,24]. A model of their juxtaposition to the active site of rat TrxR (based on the crystal structure of rat TrxR U498C [14]) is shown in Fig. 5A and a sequence alignment of the C-terminal region from TrxR and DmTrxR is shown in Fig. 5B. The residue corresponding to Cys 458 in mouse mitochondrial TrxR2 (Cys-483) forms a disulfide bond with Cys 483′ in the other subunit, suggesting there may be a fourth redox active site in that enzyme [24]. We hypothesized that one or both of these accessory Cys residues could interfere with the function of the TrxR SCCS mutant, thereby explaining why this motif is highly reactive in the Drosophila enzyme but has very little activity in the mammalian counterpart. To assess this question, we constructed three additional mutations in the TrxR SCCS mutant, changing the accessory Cys residues to the corresponding amino acids as found in the Drosophila enzyme (Fig. 5B), i.e., making the Cys-458-Ser and Cys-475-Thr mutants, as well as the double mutant C458S+C475T, in what appears to be the periphery of the SCCS active site. The purified mutant enzymes were assayed for TrxR activity using DTNB, Trx, or selenite as substrates, but no gain of activity could be detected and the three mutants were as inactive as the parental SCCS mutant (data not shown).

High Mr TrxR partially reduced with 2.2 eq NADPH typically consists of a mixture made up mostly of EH4B, with lesser amounts of EH4A, EH2A , EH2B, EH2C, and EH2D species. Reaction of EH4B with Trx constitutes the oxidative halfreaction. Pre-reduced SCCS mutant or wild-type TrxR were allowed to react with DmTrx-2. The spectra shown in Fig. 3A are of the reduced SCCS enzyme before and after the addition of 11 eq of DmTrx-2; the spectrum of oxidized enzyme is given for comparison. Spectrum 3, especially at wavelengths greater than 525 nm, indicated that the reaction proceeded only to the EH2 stage; the shape of the spectrum indicates that there is some FADH–NADP+ charge-transfer complex absorbing at 670 nm, and this predominates over any thiolateFAD charge-transfer complex (EH2B) that can best be detected at 540 nm. In addition, there is probably some EH2D, which also has no charge-transfer interaction. The kinetics of the reaction are compared for SCCS and wild-type at 464 nm in Fig. 3B and at 540 nm in Fig. 3C. Spectrum 3 in Fig. 3A indicates that only partial oxidation occurs upon addition of DmTrx-2. The first phase involved only a small change in absorbance relative to the total change, and occurred at similar rates in both enzymes (Fig. 3B); we suggest that this phase represents equilibration of the different enzyme species as Trx forms a mixed disulfide with the enzyme (k17, Scheme 1) and possibly to dissociation of NADP+ from the enzyme. The final phase, where most of the absorbance change took place, reflects formation of Trx(SH)2 and TrxR species at the

No stable mixed disulfide forms between Trx and the SCCS mutant Because the oxidative half reaction was extremely slow in the mutant enzyme, we asked whether a stable mixed disulfide intermediate between Trx and the SCCS mutant TrxR protein could be formed, which might further contribute to the low activity of the enzyme. We tried to detect the presence of such an intermediate using non-reducing SDS-PAGE, but were unable to detect any evidence for formation of a stable intermolecular mixed disulfide intermediate (Fig. 4). Interestingly, the band corresponding to a TrxR subunit dimer, commonly seen in trace amounts using SDS-PAGE analysis with the wild-type enzyme, was barely detectable with the SCCS mutant (Fig. 4). Analysis of the possible involvement of the two additional Cys residues of the TrxR SCCS mutant

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

Concluding remarks It has previously been shown that the flanking Ser residues in the C-terminal active site of DmTrxR are necessary for activating the active site Cys residues [13]. Thus, without both of these Ser residues present, significant activity was only realized when a Sec was substituted for the penultimate Cys residue [13]. Those results lead to questions about the role and necessity of the Sec residue in the mammalian selenoprotein TrxR. For example, could flanking Ser residues alone obviate the need for Sec in the mammalian enzyme? Here we have shown that inclusion of the Ser residues is not sufficient to make such Sec-to-Cys mutants of mammalian TrxR active. Although the rat SCCS mutant form of TrxR contained all of its redox active centers in at least a minimally functional state, it had less than 0.5% activity compared to that of wild-type enzyme. Rapid kinetics studies of the SCCS mutant enzyme demonstrated that the initial phases of the reductive halfreaction proceeded at rates similar to those of the wild-type enzyme. However, when <1 eq of NADPH was used, the charge-transfer complex of the two-electron reduced mutant (equivalent to EH2B in Scheme 1) was unstable, since reducing equivalents were slowly transferred to the C-terminal disulfide to form EH2D. A similar transfer of reducing equivalents was not observed in the wild-type selenoprotein when ∼1 eq of NAPH was used, probably because the selenenylsulfide is more difficult to reduce than the disulfide. This transfer of reducing equivalents to the C-terminal disulfide in the SCCS mutant protein was very slow, which partially accounts for the low activity of the enzyme. Even with 1.7 eq of NADPH, the final spectrum (Fig. 2, spectrum 4) looked like that of EH4A in complex with NADP+ (giving rise to the long wavelength absorbance) plus some EH2D . It had characteristics of both oxidized and reduced flavin, but very little (if any) thiolate-toFAD charge-transfer interaction. Only when >3 eq NADPH was used did significant thiolate-FAD charge-transfer interaction accrue at the end of the reaction (Fig. 2, spectra 5 and 6). These results indicate that transfer of reducing equivalents from the N-terminal cysteines to the C-terminal cysteines is thermodynamically favorable, but that this process is slow. The dependence of the extent of flavin reduction on NADPH concentration indicates that the reduction potential of the flavin is comparable to that of NADPH. In the wild type enzyme, it would appear that excess NADPH would be required to reduce the selenenylsulfide so that reduction of thioredoxin could occur. Finally, the oxidative half-reaction of the SCCS mutant enzyme was 200-fold slower than that of wild-type enzyme, and this alone should largely account for the very low overall activity, indicating that the C-terminal dithiol was inefficient in reducing thioredoxin. Our findings that mutations to the mammalian TrxR that mimic the C-terminal Ser-Cys-Cys-Ser motif of the Drosophila enzyme do not give active enzyme indicate that the active site microenvironment of DmTrxR must have been “fine-tuned” in evolution to support the Ser-mediated activation of its active site Cys residues; alternatively, the

655

features supporting such activation have been lost in the case of the mammalian enzyme. Why, then, has the mammalian TrxR evolved as an enzyme that must depend upon the presence of Sec rather than “activated Cys” in its active site, considering the energetically costly machinery for synthesis of selenoproteins? As proposed earlier, one reason could be a broader pH range and broader substrate specificity [13]. Another advantage accruing to a Sec-containing enzyme could be related to non-enzymatic activities of this protein, depending upon the reactivity of a Sec residue. In such context, it is intriguing to note that selenium-compromised forms of TrxR can induce apoptosis in cancer cells by a gain of function [25]. It is clear that Sec-containing proteins, especially TrxR with the Sec at an accessible C-terminal position, are easily targeted by electrophilic agents [12,26,27]. Because the enzymatic activity of most selenoprotein oxidoreductases can be substituted in Nature by evolution of Cys containing orthologues [13,28–31], could it be that the high reactivity of Sec as target for electrophiles has a physiological importance that has dictated its presence in mammalian TrxR? Acknowledgments DmTrx-2 was kindly provided by Stephan Gromer and Holger Bauer, Heidelberg University, Germany. Recombinant human thioredoxin was kindly provided by Arne Holmgren, Karolinska Institutet, Stockholm. Support was received from the National Institute of General Medical Sciences grant GM11106 to D.P.B. and from Karolinska Institutet, the Swedish Cancer Society and the Swedish Research Council for Medicine to E.S.J.A. References [1] Williams, C. H., Jr.; Lipoamide dehydrogenase, glutathione reductase, thioredoxin reductase and mercuric ion reductase - A family of flavoenzyme transhydrogenases. Vol. 3. 1992: CRC Press. [2] Lee, S. R.; Bar-Noy, S.; Kwon, J.; Levine, R. L.; Stadtman, T. C.; Rhee, S. G. Mammalian thioredoxin reductase: oxidation of the Cterminal cysteine/selenocysteine active site forms a thioselenide, and replacement of selenium with sulfur markedly reduces catalytic activity. Proc. Natl. Acad. Sci. U. S. A. 97:2521–2526; 2000. [3] Zhong, L.; Arnér, E. S. J.; Ljung, J.; Åslund, F.; Holmgren, A. Rat and calf thioredoxin reductase are homologous to glutathione reductase with a carboxyl-terminal elongation containing a conserved catalytically active penultimate selenocysteine residue. J. Biol. Chem. 273:8581–8591; 1998. [4] Zhong, L.; Holmgren, A. Essential role of selenium in the catalytic activities of mammalian thioredoxin reductase revealed by characterization of recombinant enzymes with selenocysteine mutations. J. Biol. Chem. 275:18121–18128; 2000. [5] Arnér, E. S. J.; Holmgren, A. Physiological functions of thioredoxin and thioredoxin reductase. Eur. J. Biochem. 267:6102–6109; 2000. [6] Gromer, S.; Urig, S.; Becker, K. The thioredoxin system–from science to clinic. Med. Res. Rev. 24:40–89; 2004. [7] Williams, C. H., Jr.; Arscott, L. D.; Muller, S.; Lennon, B. W.; Ludwig, M. L.; Wang, P. F.; Veine, D. M.; Becker, K.; Schirmer, R. H. Thioredoxin reductase two modes of catalysis have evolved. Eur. J. Biochem. 267:6110–6117; 2000. [8] Johansson, L.; Gafvelin, G.; Arner, E. S. J. Selenocysteine in proteinsproperties and biotechnological use. Biochim. Biophys. Acta 1726:1–13; 2005.

656

L. Johansson et al. / Free Radical Biology & Medicine 41 (2006) 649–656

[9] Arscott, L. D.; Gromer, S.; Schirmer, R. H.; Becker, K.; Williams Jr., C. H. The mechanism of thioredoxin reductase from human placenta is similar to the mechanisms of lipoamide dehydrogenase and glutathione reductase and is distinct from the mechanism of thioredoxin reductase from Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 94:3621–3626; 1997. [10] Zhong, L.; Arnér, E. S. J.; Holmgren, A. Structure and mechanism of mammalian thioredoxin reductase: the active site is a redox-active selenolthiol/selenenylsulfide formed from the conserved cysteine-selenocysteine sequence. Proc. Natl. Acad. Sci. U. S. A. 97:5854–5859; 2000. [11] Bauer, H.; Massey, V.; Arscott, L. D.; Schirmer, R. H.; Ballou, D. P.; Williams, C. H., Jr.; The mechanism of high Mr thioredoxin reductase from Drosophila melanogaster. J. Biol. Chem. 278:33020–33028; 2003. [12] Cenas, N.; Nivinskas, H.; Anusevicius, Z.; Sarlauskas, J.; Lederer, F.; Arnér, E. S. J. Interactions of quinones with thioredoxin reductase: a challenge to the antioxidant role of the mammalian selenoprotein. J. Biol. Chem. 279:2583–2592; 2004. [13] Gromer, S.; Johansson, L.; Bauer, H.; Arscott, L. D.; Rauch, S.; Ballou, D. P.; Williams, C. H., Jr.; Schirmer, R. H.; Arnér, E. S. J. Active sites of thioredoxin reductases: why selenoproteins? Proc. Natl. Acad. Sci. U. S. A. 100:12618–12623; 2003. [14] Sandalova, T.; Zhong, L.; Lindqvist, Y.; Holmgren, A.; Schneider, G. Three-dimensional structure of a mammalian thioredoxin reductase: implications for mechanism and evolution of a selenocysteine-dependent enzyme. Proc. Natl. Acad. Sci. U. S. A. 98:9533–9538; 2001. [15] Kanzok, S. M.; Fechner, A.; Bauer, H.; Ulschmid, J. K.; Muller, H. M.; Botella-Munoz, J.; Schneuwly, S.; Schirmer, R.; Becker, K. Substitution of the thioredoxin system for glutathione reductase in Drosophila melanogaster. Science 291:643–646; 2001. [16] Johansson, L.; Arscott, L. D.; Ballou, D. P.; Williams, C. H., Jr.; Arnér, E. S. J. Characterization of a “Drosophila-mimic” active site mutant of mammalian thioredoxin reductase. 15th International Symposium on Flavins and Flavoproteins. Hayama, Japan; 2005. [17] Arnér, E. S. J.; Sarioglu, H.; Lottspeich, F.; Holmgren, A.; Böck, A. Highlevel expression in Escherichia coli of selenocysteine-containing rat thioredoxin reductase utilizing gene fusions with engineered bacterial-type SECIS elements and co-expression with the selA, selB and selC genes. J. Mol. Biol. 292:1003–1016; 1999. [18] Rengby, O.; Johansson, L.; Carlson, L. A.; Serini, E.; Vlamis-Gardikas, A.; Kårsnäs, P.; Arnér, E. S. J. Assessment of production conditions for efficient use of Escherichia coli in high-yield heterologous recombinant selenoprotein synthesis. Appl. Environ. Microbiol. 70:5159–5167; 2004.

[19] Arnér, E. S. J.; Holmgren, A. Measurement of thioredoxin and thioredoxin reductase. In: Sassa, S. ed. Current protocols in toxicology. New York: John Wiley and Sons Inc, pp. 7.4.1–7.4.14; 2000. [20] Arnér, E. S. J.; Zhong, L.; Holmgren, A. Preparation and assay of mammalian thioredoxin and thioredoxin reductase. Meth. Enzymol. 300:226–239; 1999. [21] Williams, C. H., Jr.; Arscott, L. D.; Matthews, R. G.; Thorpe, C.; Wilkinson, K. D. Vitamins and Coenzymes. In: McCormick, D.B., Wright, L.D. ed. Methods in Enzymology, Vol. 62 Part D. New York: Acedemic, pp. 185–198; 1979. [22] Bar-Noy, S.; Gorlatov, S. N.; Stadtman, T. C. Overexpression of wild type and SeCys/Cys mutant of human thioredoxin reductase in E. coli: the role of selenocysteine in the catalytic activity. Free Radic. Biol. Med. 30:51–61; 2001. [23] Luthman, M.; Holmgren, A. Rat liver thioredoxin and thioredoxin reductase: purification and characterization. Biochemistry 21:6628–6633; 1982. [24] Biterova, E. I.; Turanov, A. A.; Gladyshev, V. N.; Barycki, J. J. Crystal structures of oxidized and reduced mitochondrial thioredoxin reductase provide molecular details of the reaction mechanism. Proc. Natl. Acad. Sci. U. S. A. 102:15018–15023; 2005. [25] Anestål, K.; Arnér, E. S. J. Rapid induction of cell death by seleniumcompromised thioredoxin reductase 1 but not by the fully active enzyme containing selenocysteine. J. Biol. Chem. 278:15966–15972; 2003. [26] Johansson, L.; Chen, C.; Thorell, J.; Fredriksson, A.; Stone-Elander, S.; Gafvelin, G.; Arnér, E. S. J. Exploiting the 21st amino acid-purifying and labeling proteins by selenolate targeting. Nature Methods 1:61–66; 2004. [27] Witte, A. B.; Anestål, K.; Jerremalm, E.; Ehrsson, H.; Arnér, E. S. J. Inhibition of thioredoxin reductase but not of glutathione reductase by the major classes of alkylating and platinum-containing anticancer compounds. Free Radic. Biol. Med. 39:696–703; 2005. [28] Kim, H. Y.; Gladyshev, V. N. Different Catalytic Mechanisms in Mammalian Selenocysteine- and Cysteine-Containing Methionine-RSulfoxide Reductases. PLoS Biol. 3:e375; 2005. [29] Kryukov, G. V.; Castellano, S.; Novoselov, S. V.; Lobanov, A. V.; Zehtab, O.; Guigo, R.; Gladyshev, V. N. Characterization of mammalian selenoproteomes. Science 300:1439–1443; 2003. [30] Zhang, Y.; Fomenko, D. E.; Gladyshev, V. N. The microbial selenoproteome of the Sargasso Sea. Genome Biol. 6:R37; 2005. [31] Zhang, Y.; Gladyshev, V. N. An algorithm for identification of bacterial selenocysteine insertion sequence elements and selenoprotein genes. Bioinformatics 21:2580–2589; 2005.