Biochimica et Biophysica Acta 1804 (2010) 1483–1491

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bbapap

The role of Cys108 in Trigonopsis variabilis D-amino acid oxidase examined throughchemical oxidation studies and point mutations C108S and C108D

Mario Mueller a, Regina Kratzer a, Margaretha Schiller a, Anita Slavica b,1, Gerald Rechberger c,Manfred Kollroser d, Bernd Nidetzky a,b,⁎a Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, Petersgasse 12, A-8010 Graz, Austriab Applied Biocatalysis Research Centre, Petersgasse 14, A-8010 Graz, Austriac Institute of Molecular Biosciences, University of Graz, Humboldtstraße 50, A-8010 Graz, Austriad Institute of Forensic Medicine, Medical University of Graz, Universitätsplatz 4, A-8010 Graz, Austria

Abbreviations: DAO, D-amino acid oxidase; DCIP, 25,5′-dithio-bis(2-nitrobenzoic acid); DTT, dithiothreitollactopyranoside; TvDAO, Trigonopsis variabilis D-amino a⁎ Corresponding author. Institute of Biotechnology

Graz University of Technology, Petersgasse 12/I, A-801873 8400; fax: +43 1 873 8434.

E-mail addresses: mario.mueller@tugraz.at (M. Mue(R. Kratzer), margaretha.schiller@tugraz.at (M. Schiller)gerald.rechberger@uni-graz.at (G. Rechberger), manfred(M. Kollroser), bernd.nidetzky@tugraz.at (B. Nidetzky).

1 Permanent address: Department of Biochemical EngPierottijeeva 6, HR-10000 Zagreb, Croatia.

1570-9639/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.bbapap.2010.02.009

a b s t r a c t

a r t i c l e i n f o

Article history:Received 19 November 2009Received in revised form 19 January 2010Accepted 10 February 2010Available online 1 March 2010

Keywords:D-amino acid oxidaseOxidative inactivationCysteine sulfinic and sulfonic acidSite-directed mutagenesisMass-spectrometric characterization ofchemical oxidationDenaturation pathway

Oxidative modification of Trigonopsis variabilis D-amino acid oxidase in vivo is traceable as the conversion ofCys108 into a stable cysteine sulfinic acid, causing substantial loss of activity and thermostability of theenzyme. To simulate native and modified oxidase each as a microheterogeneity-resistant entity, we replacedCys108 individually by a serine (C108S) and an aspartate (C108D), and characterized the purified variantswithregard to their biochemical and kinetic properties, thermostability, and reactivity towards oxidation byhypochlorite. TandemMSanalysis of tryptic peptides derived fromahypochlorite-treated inactive preparationof recombinant wild-type oxidase showed that Cys108 was converted into cysteine sulfonic acid, mimickingthe oxidative modification of native enzyme as isolated. Colorimetric titration of protein thiol groups revealedthat in the presence of ammonium benzoate (0.12 mM), the two muteins were not oxidized at cysteineswhereas in the wild-type enzyme, one thiol group was derivatized. Each site-directed replacement caused aconformational change in D-amino acid oxidase, detected with an assortment of probes, and resulted in aturnover number for the O2-dependent reaction with D-Met which in comparison with the correspondingwild-type value was decreased two- and threefold for C108S and C108D, respectively. Kinetic analysis ofthermal denaturation at 50 °C was used tomeasure the relative contributions of partial unfolding and cofactordissociation to the overall inactivation rate in each of the three enzymes. Unlike wild-type, C108S and C108Dreleased the cofactor in a quasi-irreversible manner and were therefore not stabilized by external FAD againstloss of activity. The results support a role of the anionic side chain of Cys108 in the fine-tuning of activity andstability of D-amino acid oxidase, explaining why C108S was a surprisingly poor mimic of the native enzyme.

,6-dichloroindophenol; DTNB,; IPTG, isopropyl β-D-1-thioga-cid oxidaseand Biochemical Engineering,0 Graz, Austria. Tel.: +43 316

ller), regina.kratzer@tugraz.at, aslavica@pbf.hr (A. Slavica),.kollroser@medunigraz.at

ineering, University of Zagreb,

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Oxidation is a common posttranslational modification of proteins[1]. It may conclude the normal lifespan of a protein in the cell whereoxidative modification is often used to label proteins for degradation[2–4]. It is a frequently observed epigenetic process induced by

different forms of stress [5,6] and can fulfill a regulatory function (e.g.[7–9]). Oxidation also represents a widespread chemical alterationduring protein production [10] and in isolated proteins [11,12]. Changeor loss of biological activity and decreased stability are primaryfunctional consequences in oxidized protein variants, as observed invarious enzymes, for example subtilisin [13], p-hydroxybenzoatehydroxylase [14,15], chloroperoxidase [16], β-lactamase [17], superox-idedismutase [18], tyrosinekinase [19], alcohol dehydrogenase [20] andD-amino acid oxidase [21,22]. The side chain of cysteine belongs to thesites most easily oxidized in proteins (e.g. [23]). Depending on proteinenvironment and reaction conditions, it may be converted into adisulfide (–S–S–), a sulfenic acid (–SOH), a sulfinic acid (–SO2H), or asulfonic acid (–SO3H) [24–28]. Fenton chemistry was utilized in severalstudies to reproduce cellular oxidation processes on cysteines ofproteins in vitro [29–31]. Likewise, cysteine oxidation by hypochloricacidwas also examined, and themost prominent products formedwerestable sulfinic acids [32] and intramolecular sulfinamide and sulphona-mide crosslinks [33–35]. Many proteins are strongly inactivated by

1484 M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

oxidation of cysteine residues contributed to oneof their functional sites(e.g. [36]). However, oxidative modification of non-functional, yetreadily accessible cysteine residues can also impair activity and stability,typically through its proximally disruptive effects on struc-ture [14,15,37–41]. Oxidation of this latter class of cysteines is oftenpartial, and the structural and functional microheterogeneity thusinduced may hinder further applications of the protein preparation.Hence it represents a relevant issue of quality control during bio-technological protein production [42].

Trigonopsis variabilis D-amino acid oxidase (TvDAO; EC 1.4.3.3)catalyzes O2-dependent oxidative deamination of D-amino acidsubstrates, yielding the corresponding 2-keto acid, ammonia, andhydrogen peroxide as products. TvDAO is widely known for itsapplication in the industrial production of 7-aminocephalosporanicacid where the enzyme is utilized in the first step of the process, that isthe conversion of cephalosporin C into α-ketoadipyl-7-aminocepha-losporanic acid [43]. TvDAO is a functional homodimer and each of itssubunits has one noncovalently-bound flavin adenine dinucleotide(FAD) [44] (for recent reviews on D-amino acid oxidases, see [45,46]).The primary structure of TvDAO contains six cysteine residues (Cys108,Cys145, Cys159, Cys193, Cys257, and Cys298) which have their sidechains reduced in the native enzyme. Of the six, Cys298 was mostsensitive to alkylation [47] and Fenton oxidation [30], and Cys108appeared to be also oxidized under the in vitro conditions used.Modification of Cys298 caused inactivation of TvDAO [30,47]. We haveshown recently that partial oxidation of the enzyme in T. variabilis istraceable as the conversion of Cys108 into a stable cysteine sulfinic acid[22]. Cys108 is presumably located on a surface-exposed protein regionfar away from the catalytic center [21]. However, its modificationseemed to have affected the protein environment of the FAD cofactorand caused about 75% loss of wild-type activity [22]. The oxidized formTvDAO-Cys108(SO2

−) released FAD faster at 50 °C and was thereforetwice as thermally labile as the native enzyme [21]. However, proteinaggregation under the thermal stress conditions was completelysuppressed in the oxidized enzyme whereas it appeared to bekinetically coupled to inactivation of native TvDAO via dissociation ofFAD. If oxidation of Cys108 during enzyme production and under theconditions of the biocatalytic process was prevented efficiently, asignificant improvement of the total turnover number of the oxidase,related to the total mass of the protein preparation, might be obtained.

In this paper, we report on results of experiments by site-directedmutagenesis designed to mimic native and oxidized forms of TvDAO,each as microheterogeneity-resistant protein entity. Therefore,Cys108 was individually replaced by serine (C108S) and aspartate(C108D)whose side chains we considered to be suitable surrogates, interms of both size and charge, of the side chains of cysteine andcysteine sulfinic acid, respectively. The purified point muteins ofTvDAO were characterized and compared to the wild-type enzymewith respect to kinetic properties and thermal stability. To analyzeresistance of the different enzymes to chemical oxidation, wedeveloped a hypochloric acid-based assay, and protein modificationsthus promoted were monitored by thiol group titration.

2. Materials and methods

2.1. Materials

Pfu DNA polymerase was from Promega (Madison, WI, USA), theQIAprep Spin Miniprep kit was from QIAGEN (Hilden, Germany), DpnIrestriction endonuclease and dNTPs were from Fermentas (St. Leon-Rot, Germany). Oligonucleotide synthesis and DNA sequencing wereperformed at VBC Biotech Services GmbH (Vienna, Austria). All otherchemicals were of the highest available purity and purchased fromSigma-Aldrich (St. Louis, MO, U.S.A.). Native and Cys108(SO2

−) formsof TvDAO were isolated from T. variabilis ATCC 10679 as describedelsewhere [21,22].

2.2. Site-directed mutagenesis, and production and purification ofrecombinant wild-type and muteins enzymes

The plasmid vector pTvDAOstrepN was used as the template [42]. Itcontains a gene encoding a chimeric formof TvDAO that has a peptide of12 amino acids (Strep-tag II) fused in-frame to the N-terminus of theenzyme (coding sequence Z80895). Amino acid numbering forTvDAOstrepN includes the initiator Met but not the N-terminal fusionpeptide. The point mutations Cys108→Ser and Cys108→Asp wereintroduced via a two-stage PCR protocol [48]. The following oligonu-cleotide primers were used to amplify the entire vector with Pfu DNApolymerase where the mismatched codons are underlined:

5′-TCGGCCATCTCTCAACGCAACC-3′ (Cys108Ser; forward primer),5′-GGTTGCGTTGAGAGATGGCCGA-3′ (Cys108Ser; reverse primer),5′-TCGGCCATCGATCAACGCAACC-3′ (Cys108Asp; forward primer),5′-GGTTGCGTTGATCGATGGCCGA-3′ (Cys108Asp; reverse primer).

In thefirst step, two separate reactionmixtureswith the forward andreverseprimers for eachmutation, respectively,werepre-heated (95 °C,60 s) and then 4 cycles (95 °C, 50 s/60 °C, 50 s/68 °C, 8 min) of theextension reactions were performed. In the second step, PCR mixtureswere combined and amplification was continued for 18 cycles (95 °C,50 s/60 °C, 50 s/68 °C, 8 min) followed by one extension step (68 °C,7 min). The parental template DNA was digested by DpnI, and themutagenized plasmid vectors were transformed into E. coli BL21 (DE3)cells. The same protocol was used to introduce the substitutionArg110→Ala. The oligonucleotide primers employed were:

5′-CCATGTCGGCCATCTGTCAAGCGAACCCCTGGTTC-3′ (forwardprimer)5′-GAACCAGGGGTTCGCTTGACAGATGGCCGACATGG-3′ (reverseprimer).

Cells harboring expression vectors for wild-type or muteins weregrown over-night at 37 °C in LB-media containing 50 µg ml−1

kanamycin. When the optical density (at 600 nm) of the cultureshad reached a value of about 0.8–1.0, the temperature was decreasedto 25 °C, and gene expression was induced with 500 µM IPTG for 18 h.Fifty mM DL-Ala was added at the time of induction to enhancerecombinant protein production [42]. The harvested cells wereresuspended in 8–12 ml of 10 mM Tris/HCl buffer, pH 7.5, anddisrupted by three passages through a French press. The clearsupernatant, obtained after ultracentrifugation (4 °C/80,000×g/45 min), was applied on a 5 ml Strep-Tactin Superflow Cartrige (IBAGmbH, Göttingen, Germany), and the DAO proteins were elutedaccording to a protocol supplied by themanufacturer. Pooled fractionsof wild-type and muteins (C108S, C108D, R110A) were concentratedwith Vivaspin ultrafiltration tubes (10-kDa cut-off; Vivascience,Hannover, Germany) to ≥2 mg ml−1 in 10 mM Tris/HCl buffer, pH7.5, without addition of DTT. Protein concentrations were determinedusing the Bio-Rad dye binding method with BSA as the standard, andactivity was measured with a reported assay based on reaction of theenzyme with the non-natural electron acceptor 2,6-dichloroindophe-nol (0.15 mM) in a 100 mM potassium phosphate buffer, pH 8.0, at30 °C [49]. Protein purification was checked by SDS-PAGE andnondenaturing anionic PAGE using Coomassie blue for the stainingof protein bands. Stock solutions of wild-type and muteins werealiquoted and stored at −25 °C.

2.3. Characterization of wild-type enzyme and muteins

Anion exchange chromatography of purified protein samples usingaMonoQHR 5/5 column (Amersham Biosciences, Freiburg, Germany)was performed as described elsewhere [22]. Absorbance spectra wererecorded with a DU800 UV–VIS spectrophotometer (Beckman-Coulter, Inc., Fullerton, CA, USA). Fluorescence measurements were

1485M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

performed using a software-driven F-4500 fluorescence spectropho-tometer (Hitachi; Tokyo, Japan). Both instruments were equippedwith a temperature-controlled cuvette holder. Far-UV CD spectrawere acquired at room temperature with a J-715 spectropolarimeter(Jasco, Inc., Easton, MD) using a 0.1 cm path-length cylindrical cell andthe following instrument settings. Step resolution was 0.2 nm, scanspeed was 50 nmmin−1, response time was 1 s, and bandwidth was1 nm. Triplicate spectra were recorded in the wavelength range 260–190 nm using enzymes (∼0.10 mgml−1) dissolved in 20 mM potassi-um phosphate buffer, pH 7.5. They were subsequently averaged andcorrected by a blank spectrum lacking enzyme.

Colorimetric titration of protein thiol groups was carried out usingreported procedures [22], adapted from Habeeb [50]. Thermaldenaturation experiments were performed at 50 °C using protocolsdescribed elsewhere [21] whereby loss of enzyme activity, release ofFAD cofactor, and aggregation were monitored in dependence of theincubation time. Kinetic modeling was performed with Athena VisualStudio (Stewart & Associates Engineering Software, Inc.,Madison,WI).

2.4. Oxidative modification with HOCl

Experiments were performed in glass vials soaked in NaOClsolution and exhaustively rinsed with ultrapure H2O. All solutionswere freshly prepared immediately prior to use. The concentration ofHOCl in NaOCl stock solutions (300 mM) was determined spectro-photometrically [51]. To remove all Tris, a primary amine that reactswith HOCl and forms α-monochloramine, the buffer of the enzymestock solutions was replaced with a 50 mM sodium phosphate buffer,pH 7.5, using repeated ultrafiltration (Vivaspin, 3-kDa cut-off).Oxidations of protein samples were carried out in the absence andpresence of protection by ammonium benzoate (0.12 mM). HOCl wasadded to protein solutions, dropwise with constant mixing, until theend concentration was reached. Incubations were then carried out onice or at 37 °C and lasted 0.5 h, 1 h, or 2 h, as indicated. Oxidationswere terminated by removing remaining HOCl through ultrafiltration(Vivaspin), and protein samples were immediately analyzed furtherusing different biochemical methods (see the Results).

2.5. Mass spectrometry

MS was performed using a LCQ Deca XPplus ion trap massspectrometer (Thermo, San Jose, CA, USA) equipped with a nano-electrospray ionization (ESI) source coupled to a nanoLC system (LC-packings, Amsterdam, The Netherlands) consisting of FAMOS™autosampler, SWITCHOS™ loading system, and ULTIMATE™ dualgradient system. For peptide mass fingerprinting, in-gel trypticdigestion of the protein sample was performed using a reportedprotocol [52]. Twenty µl of peptide extract dissolved in 0.1% (v/v)formic acid was trapped on a LC Packings C18 Pep-Map precolumn(5 µm, 100 Å, 300-µm inner diameter×1 mm) with a total loadingtime of 5 min and a sample solution flow rate of 20 µl/min. The loadedmixture was then flushed to a nanoLC separation column of the sametype (75 µm inner diameter×150 mm) with a flow rate of 300 nl/min,and peptides were eluted using a 60-min linear gradient of 5–50% (v/v)acetonitrile in 0.3% (v/v) formic acid. The column effluent was directedto a NanoSpray tip (Pico-Tip™ Emitter, New Objective, Woburn, MA,USA), and positive ions formed by the electrospray under optimizedconditions (spray tip voltage of 1.3 kV; inlet capillary temperature of235 °C) were analyzed with the mass spectrometer using automateddata-dependent acquisitionmode. In tandemMS experiments, a full MSscan between 350 and 1800m/z was followed by data dependant MS/MS scans for the fourmost abundant ions from eachMS scan. Datawereanalyzed with SpectrumMill version 2.7 (Agilent, Santa Clara, CA, USA),employing peptide mass fingerprint and MS/MS search tools, against auser-defined database containing the sequences of wild-type TvDAOand the R110A mutein thereof.

3. Results and discussion

3.1. Site-directed mutagenesis of Arg110 and characterization of R110A

The amino acid sequence of TvDAOstrepN deduced from the codinggene shows residue 110 to be an Arg. Previous MS data had suggestedAla110 [22]. Considering the proximity of Arg/Ala110 to theoxidation-labile Cys108, we reinvestigated native TvDAO by MS. Wealso prepared an Arg110→Ala mutein of TvDAOstrepN (data notshown) and compared the biochemical properties of native enzymeand the mutein.

The peptide mass maps of native TvDAO and TvDAOstrepNcontained an abundant peptide with a monoisotopic mass of 1235.6,corresponding to a mass peak ofm/z 627.0 [2+]. TandemMS analysisof this precursor ion (Fig. 1A) revealed the sequence 100LEGAm-SAICQR

110with Met104 oxidized (m) and Cys108 carboxamidomethy-

lated, in agreement with the deduced primary structure of theenzyme. The previously reported peptide 12 (m/z 912.4 [2+];residues 100–115) containing Ala110 [22] was detected neither inthe mass spectrum of the tryptic digest of native TvDAO nor in thecorresponding mass spectrum of the recombinant enzyme. We canonly speculate that perhaps the different methods of samplepreparation, in-gel tryptic digestion (this work) as compared to thepreviously used in-solution procedure [22], could be a source of theconflict in findings.

Purified R110A displayed a similar specific activity (2.6±0.2 U mg−1; N=2) as wild-type TvDAOstrepN (2.2±0.6 U mg−1;N=7), measured with D-Met as substrate and DCIP as electronacceptor [49]. Fig. 2 compares time courses for loss of activity andcorrelates relative inactivation with FAD release into solution (inset)during incubation of wild-type enzyme and R110A at 50 °C. Theresults show clearly that TvDAOstrepN was more resistant to high-temperature denaturation than R110A. The two enzymes differedsignificantly in the extent to which dissociation of cofactorparticipated in the overall inactivation. Release of FAD (Scheme 1)appeared to have become the dominant path of thermal denatur-ation of R110A. These results imply that Ala110 in native TvDAO [22]is inconsistent with the observed stability of this enzyme at 50 °C[42] and must therefore be considered an MS sequence error. Notethat previous MS data [22] are furthermore conflicting with twopositions in the amino acid sequence, where Val135 and Ala142 shouldappear instead of the reported Glu135 and Arg142, respectively.

3.2. Oxidation of Cys108 in TvDAOstrepN and R110A

Incubation of wild-type enzyme and R110A, each at 10 µM, in thepresence of 160 µM HOCl at 37 °C caused complete loss of oxidaseactivity (DCIP assay) in 1 h. Addition of ammonium benzoate(100 mM) afforded full protection to the activity of wild-typeTvDAOstrepN whereas R110A was only partially protected againstHOCl (∼10% retention of specific oxidase activity). Differences amongthe two enzymes regarding the extent to which benzoate affordsprotection are not easily rationalized in terms of the structural modelfor TvDAO [46]. It is conceivable that benzoate binds not as well toR110A as it binds to wild-type enzyme. However, analysis of benzoatebinding affinity was considered to be beyond the scope of this work.Preparations of wild-type enzyme and R110A that had beeninactivated through HOCl treatment were characterized by MS.Comparison of the tryptic peptide mass maps for “native” and“inactive” protein suggested that oxidative modification with HOClwas directed to multiple sites in both wild-type TvDAOstrepN andR110A (data not shown). Fig. 1 shows results of tandemMS analysis ofthe peptide containing Cys108. The HOCl treatment led to conversionof the native thiol group of Cys108 into a sulfonic acid. The particularoxidative modification occurred both in wild-type (Fig. 1A) andmutated oxidase (Fig. 1B), indicating the oxidation-prone state of the

Fig. 1. ESI tandemmass spectra of the peptide containing residue 108 in wild-type TvDAOstrepN (panel A) and R110A (panel B) prior to and after oxidative modification with HOCl.Oxidation was performed with 10 µM protein solutions using 160 µM HOCl and reaction times of 2 h at 37 °C.

1486 M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

Fig. 2. Temperature-induced denaturation of R110A as compared to wild-typeTvDAOstrepN. Relative release of FAD versus relative inactivation (inset). Protein solutions(10 µM) were incubated at 50 °C using an agitation rate of 300 rpm. For the inactivationstudies a 10 mMTris buffer, pH7.5,was usedwhile for the comparative inactivation versusFAD release studies a 100 mM phosphate buffer, pH 8.0, was used. The results are shownfor wild-type TvDAOstrepN enzyme (full circle; S.D.≤10%) and the R110A mutein (opencircle; S.D.≤18%).

Fig. 3. Elution profiles of C108S and C108D muteins in MonoQ anion exchangechromatography, compared with corresponding elution profiles of native and Cys108(SO2

−) forms of TvDAO. Muteins C108S (3.4 mg; S, solid black line) and C108D(3.7 mg; D, broken black line) were applied on the MonoQ HR 5/5 column and elutedas described elsewhere [36]. The elution profiles of native enzyme (1, solid grey line),and TvDAO-Cys108(SO2

−) (2, solid grey line) obtained under exactly identicalconditions are shown for comparison.

1487M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

Cys in either sequence context. While modification by HOCl is clearlynot a perfect mimic of oxidation processes causing microheterogene-ity of TvDAO as isolated from T. variabilis, it underscores Cys108 as aprominent site for oxidative derivatization. Conversion of cysteineinto cysteine sulfonic acid has been seen in chemical models (e.g.[53]) and in proteins (e.g. [16–18,20]).

3.3. Site-directed mutagenesis of Cys108

Isolated C108S and C108D migrated in SDS-PAGE as single proteinbandsof identical apparentmolecularmassof≈41 kDa(SupplementaryFig. 3A), corresponding to the size expected for the full-length proteinsubunit including the N-terminal Strep-tag II. Supplementary Fig. 3Bshows that C108D moved farther than C108S in nondenaturing anionicPAGE, the difference in relative mobility for the two muteins beingsimilar to that observed previously for Cys108(SO2

−) and native forms ofTvDAO [21]. The specific activities of C108S and C108D, measured withDCIP as electron acceptor [49], were 1.2±0.1 U mg−1 and 4.1±0.5 U mg−1, respectively. By way of comparison, the specific activityof the wild-type enzyme was 2.2±0.6 U mg−1.

Fig. 3 shows protein elution profiles during MonoQ anionexchange chromatography for C108S and C108D, superimposed onthe corresponding elution profiles for native and oxidized enzymepreparations isolated from T. variabilis [22]. TvDAOstrepN eluted atthe same position in Fig. 3 as the native enzyme (data not shown; seeRef. [42]). Each mutein eluted as a single protein peak that containedall of the applied protein and enzyme activity. C108S displayed thesame retention time as the native oxidase whereas C108D wasretained more strongly. The interaction of C108D with the MonoQcolumn was intermediate between those of native and C108S

Scheme 1. Proposed mechanism of denaturation of TvDAO at 50 °C. N is the activeenzyme subunit, D is the partially unfolded enzyme, which is inactive but has FADbound, and A is the apo-enzyme, which also lacks activity and undergoes aggregation(Agg).

enzymes (=relatively weak), and Cys108(SO2−) TvDAO (=relatively

strong). Unlike TvDAOstrepNwhich required DTT in the mobile phasefor stability during chromatography on MonoQ, retention of originalspecific activity in eluted protein peaks of C108S and C108D wascomplete, irrespective of the addition of 1 mM DTT to the elutionbuffer. These results strongly implicate the side chain of Cys108 inthe loss of wild-type activity during anion exchange chromatographyon MonoQ, and in its prevention by DTT.

3.4. Conformational properties of C108S and C108D

Absorption spectra of C108S and C108D are shown in Fig. 4A. Thespectra of both proteins exhibited bands at 365 and 455 nm that arecharacteristic for enzyme-bound FAD in the oxidized state. Thespectra were not superimposable on the spectra of TvDAOstrepN(data not shown), native TvDAO and TvDAO-Cys108(SO2

−), or on eachother. CD spectra of wild-type and muteins, enzymes, corrected forminor differences in protein concentration used, are shown in Fig. 4B.The spectra of the native enzyme, TvDAO-Cys108(SO2

−), TvDAOstrepNand C108D were comparable (although not superimposable on eachother). They were, however, clearly different from the spectrum ofC108S. These results imply that the relative secondary structuralcomposition of TvDAOstrepN and native TvDAO is similar, andsubstitution of Cys108 with Ser but not Asp induced a significantchange of it. They suggest that Cys108 influences the structuralconformation in this enzyme.

3.5. Oxidation of C108S and C108D by HOCl

Modification of cysteine side chains brought about by incubationof the respective enzyme (TvDAOstrepN, cf. Fig. 1; C108S, C108D) inthe presence of HOCl was measured using thiol group titration withDTNB. Fig. 5 compares time courses of the formation of 2-nitro-5-thiobenzoate, related to the molar concentration of oxidase subunits,for untreated enzymes and enzymes oxidized by HOCl in the absenceand presence of protection by 0.12 mM ammonium benzoate.Consistent with changes in primary structure caused by the mutation,both C108S and C108D exhibited only 5 titratable thiol groups/subunit, which is, one group less than the wild-type enzyme.However, the time dependencies of their modification with DTNB(Fig. 5B) were also markedly different from that of the wild-typeenzyme (Fig. 5A), corroborating the notion that site-specific

Fig. 4. Absorption (A) and CD (B) spectra of the purified muteins, compared withcorresponding spectra of TvDAO, TvDAOstrepN, and TvDAO-Cys108(SO2

−). (A) The spectraof C108S (S, solid black line) and C108D (D, broken black line) were recorded at a proteinconcentration of 10 µM in 10 mM Tris/HCl buffer, pH 7.5. Spectra of TvDAO (1, solid greyline) and TvDAO-Cys108(SO2

−) (2, broken grey line) are taken from Slavica et al. [36] andare shown for comparison. (B) CD spectra of C108S (S, solid black line), C108D (D, brokenblack line), native TvDAO (solid grey line), TvDAO-Cys108(SO2

−) (dotted grey line),and TvDAOstrepN (broken grey line) recorded in 20 mM potassium phosphate buffer,pH7.5. (Θ)MRE is themean residualmolar ellipticity. Concentration of proteins in the bufferwas ∼25 µM.

Fig. 5.Oxidativemodification by HOCl of cysteine side chains in wild-type TvDAOstrepN(A) and C108S and C108D muteins thereof (B), monitored by using colorimetrictitration analysis. Solutions of 10 µM protein were incubated with 12 mM DTNB.(A) Comparison of thiol modification in TvDAOstrepN (wt1, solid black line) andTvDAOstrepN oxidized with HOCl in the presence (wt2, solid grey line) and in theabsence of benzoate (wt3, broken grey line). (B) Effect of oxidation by HOCl on thiolmodification in TvDAOstrepN muteins: C108S (S1, solid black line) and C108D (D1,solid grey line); C108S and C108D oxidized in the presence (S2, broken black line; D2,broken grey line) and absence of 120 µM benzoate (S3, dotted black line; D3, dottedgrey line). Oxidation was performed with 10 µM protein solutions using 120 µM HOCland reaction times of 2 h at 37 °C. Lines show averaged results from four parallelexperiments (S.D.=5%).

1488 M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

substitutions of Cys108 translate into slightly altered proteinconformations. In particular, the clear distinction between fast andslow reacting classes of cysteine side chains that was typical of thereaction of TvDAOstrepN with DTNB (Fig. 5A) appeared to have beeneliminated in both muteins (Fig. 5B). In the wild-type enzyme one ofthe available six cysteines was modified immediately after additionof DTNB (Fig. 5A) concomitant with loss of all enzymatic activity.Cysteines of C108D were slightly more reactive kinetically thancorresponding residues of C108S under the conditions tested (Fig. 3B).Muteins exhaustively reacted with DTNB were completely inactive.

C108S and C108D derivatized by HOCl in the absence ofammonium benzoate contained 4 thiol groups/subunit, as shown inFig. 5B, and lacked detectable enzyme activity. In the presence ofbenzoate (0.12 mM), however, both muteins were protected fullyagainst oxidative thiol modification (Fig. 5B) as well as inactivationinduced by HOCl. By contrast, oxidation of TvDAOstrepN in thepresence and absence of benzoate (0.12 mM) caused loss ofapproximately 1 and 1.8 Cys side chains/subunit, respectively(Fig. 5A). Both conditions led to complete inactivation of wild-typeenzyme. Therefore, the concentration of benzoate required forefficient protection against HOCl was higher for TvDAOstrepN thanC108S or C108D. Benzoate is a reversible inhibitor of wild-type TvDAO

that binds with relatively low affinity (Ki 18.8 mM) in competitionwith D-amino acid substrate [44]. It would seem therefore thatformation of an enzyme–benzoate complex (∼84% saturation at100 mM using the reported Ki) prevents an otherwise highly reactivecysteine (not Cys108) that is important for activity or stability of theenzyme from undergoing oxidation by HOCl. Although we have beenunable to track by MS the oxidation state of Cys298 in native andoxidized preparations of TvDAOstrepN (data not shown), there isgood evidence from the work of Schräder and Andreesen [30,47] thatthis Cys is highly susceptible to chemical derivatization in nativeTvDAO. Different types of modification of Cys298 resulted in completeinactivation of the oxidase [30,47]. Observation that in the presence ofa low benzoate concentration (0.12 mM), the extent of HOCl-inducedthiol modification in TvDAOstrepN was smaller than it was in theabsence of benzoate (Fig. 5A) whereas benzoate protection of oxidaseactivity was minimal under these conditions suggests a role forbenzoate in preventing oxidative modification of cysteines other thanCys298, probably Cys108. Results of oxidative modification studies ofthe native enzyme isolated from T. variabilis (Supplementary Figs. 2Aand B) corroborate fully the findings with TvDAOstrepN describedabove (Fig. 5A). Slight differences between the two oxidase

Fig. 6. Temperature-induced inactivation of C108S (A), C108D (B) and TvDAOstrepN (C).Protein solutions (10 µM) were incubated at 50 °C using an agitation rate at 300 rpm.A 10 mM Tris buffer, pH 7.5, was used. Samples were taken at the indicated times, andresidual activity (open circles) and released FAD (open squares) were measured, usingreported methods [5,33]. The DCIP assay was used for activity measurements. Lines aretime courses modeled according to Scheme 1 using parameter values in Table 1. Proteinaggregation was not modeled. Time courses of inactivation recorded under otherwiseidentical conditions in thepresenceof1 mMFADare also shown(full circles), revealing theabsence of stabilization by external cofactor in the C108 muteins. Static light scatteringmeasurements are indicated by open triangles.

1489M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

preparations with regard to kinetics of thiol group derivatization byDTNB were noted but were not further pursued.

3.6. Kinetic consequences in C108S and C108D

Initial-rate measurements for oxidative deamination of D-Metcatalyzed by TvDAOstrepN, C108S, and C108D were carried out underconditions in which the concentration of O2 was varied in the range50–1000 µMwhile the substrate concentrationwas constant at 20 mMand saturating. Kinetic parameters were obtained from nonlinear fitsof the Michaelis–Menten equation to the data and have S.D. of ≤30%:wild-type, kcat=200 s− 1, kcat/KmO2=274 mM− 1 s− 1; C108S,kcat=110 s−1, kcat/KmO2=153 mM−1 s−1; C108D, kcat=75 s−1,kcat/KmO2=258 mM−1 s−1. A molecular mass of 41 kDa for theenzyme subunit was assumed for calculation of kcat. By wayof comparison, apparent kcat values for DCIP-dependent conversionof D-Met calculated from specific activities recorded with 0.15 mMelectron acceptor were 1.5 s−1, 0.84 s−1, and 2.8 s−1 for wild-typeenzyme, C108S and C108D respectively. The results show thatsubstitution of Cys108 caused moderate changes in the kineticproperties of the wild-type enzyme. However, observation that thesignificant increase in apparent kcat for the DCIP-dependent reactioncatalyzed by C108D was accompanied by a decrease in both apparentkcat and Km for dioxygen in the same enzyme leads to the interestingsuggestion that replacement of Cys108 by Asp affected the catalyticcycle of TvDAO at steps involved in the reduction of the electronacceptor. An important ramification of the results is that kineticconsequences of mutation of Cys108 into Asp and oxidation of thecysteine to a cysteine sulfinic acid [22]were clearly different. Likewise,C108S appeared to be not a good mimic, in kinetic terms, of the nativeenzyme.

3.7. Effects of mutation of Cys108 on thermal stability

Fig. 6 summarizes time courses of inactivation of C108S (panel A)and C108D (panel B) at 50 °C. The corresponding time course forTvDAOstrepN(panelC) is shownasa reference. Thermal denaturationofTvDAOcanbe formally described by Scheme1 [21].While loss of activityreflects the overall process of denaturation, the concentrations ofreleased cofactor and protein aggregation serve as selective reporters ofthe path involving dissociation of FAD. Fig. 6 displays the relevant timecourses for wild-type and recombinant muteins. Scheme 1 predictsthat concentrations of external FAD sufficiently high to reduce thenet rate of cofactor release (kb′=kb kagg /(k−b [FAD]+kagg)) to a nearzero value should not only stabilize the enzyme activity but alsoenable determination of the thermal unfolding rate constant ka as akinetically isolated denaturation step. We therefore measuredinactivation of TvDAOstrepN and the two muteins in the presenceof 1 mM FAD, and the time courses obtained were added to Fig. 6.Unlike the wild-type enzyme, external FAD did not confer any extrastability to C108S and C108D, which was unexpected becausethermal denaturation of both muteins went along with release ofFAD. However, if in Scheme 1 kagg wasmuch larger than k−b, kb′wouldbe independent of the concentration of free FAD and approximatelyequal kb.We estimated that from the complete lack of stabilization byexternal FAD that kagg/k−b must be 100 µM or greater. The coupledordinary differential and algebraic equations derived for thecomplete mechanism in Scheme 1 were employed to fit the wholeset of data for each enzyme, except the values from light scatteringmeasurements (see below), using numerical integration and non-linear least squares regression analysis in which kagg /k−b of themuteins was chosen to have a constant value of 100 µM. Table 1summarizes estimates of ka and kb′ for TvDAOstrepN and muteins. Itreveals that the relative contributions of thermal unfolding (ka) andFAD release (kb) steps to the overall rate of denaturation at 50 °Cwere enhanced and reduced, respectively, in the muteins compared

with the wild-type. C108D showed the largest changes in ka (1.6-foldincrease) and kb′ (3.6-fold decrease). Alterations of ka and kb′ causedby the mutations combine to yield enzymes more thermally labilethan the wild-type (Fig. 6). However, note that decreased rates ofFAD releasemight be practically useful under conditions inwhich theenzyme is stabilized against thermal unfolding by immobilization.The ratio kagg/k−b was changed from a value of ≈0.9 in the wild-type to ≥100 in the muteins.

At face value, aggregation of the TvDAOstrepN seemed to have beenstrongly reduced as a result of the site-directed replacement of Cys108(see Fig. 6), calling into question that the ratio kagg/k−b should be highin the muteins. However, the intensity of static light scattering (iS) maynot be an appropriate measure of relative aggregation in the three

Table 1Rate constants for the thermal inactivation of wild-type enzymes, C108S, and C108D.The constants were obtained by fitting the experimental time courses shown in Fig. 6with the system of ordinary differential equations derived from Scheme 1. Their S.D. is≤25%.

Enzyme variant ka [h−1] kb [h−1] k−b [µM−1 h−1] kagg [h−1]

TvDAOa 0.260 0.431 0.849 0.774TvDAOstrepN 0.246 0.413 =0.849 =0.774C108S 0.327 0.282 kagg/k−b=100C108D 0.397 0.113 kagg/k−b=100

a Data from Ref. [21]; inactivation time courses of TvDAO and TvDAOstrepN weresuperimposable but the limited set of data collected for TvDAOstrepN did not allowestimation of k−b and kagg as individual parameters; the fit was performed with k−b

and kagg and ratio fixed to the values of TvDAO [21].

1490 M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

enzymes. Variation of iS among TvDAOstrepN and muteins could alsoindicate that aggregate structures derived from the three enzymeswerenot the same. It was therefore problematic to utilize iS for a comparativeestimation of kagg under these circumstances. It is clear, however, thatC108D did not eliminate thermally induced aggregation, as seen withTvDAO-Cys108(SO2

−) [13,42].Summarizing, we have utilized point mutations Cys→Ser and

Cys→Asp to examine the role of Cys108, an oxidation-sensitiveresidue of TvDAO that undergoes partial conversion to a cysteinesulfinic acid in vivo [22]. Irreversible oxidation of Cys108 wasconfirmed using in vitro experiments where HOCl served as thechemical oxidant. In contrast to the working hypothesis that hadinitially driven this study, it was found that replacements of Cys108 bySer and Asp could not mimic the functional properties of, respectively,the native enzyme where the side chain of Cys108 is not derivatized,and the in vivo oxidation product of the enzyme, TvDAO-Cys108(SO2

−). Substitution of Cys108 with Asp was conformationally andfunctionally more conservative than the Cys108→Ser replacement.Ser and Cys are almost isosteric, and it is therefore suggested that theside chain of Cys108 is present in an ionized, negatively charged formin TvDAO. Considering that the protonated thiol group is generallymuch less reactive than the counterpart thiolate anion, the highsusceptibility of Cys108 to oxidative modification in vivo [22] and invitro (this work) is consistent with the proposed ionization state of itsside chain. If properties of charge of residue Cys108 are more relevantthan molecular size, Asp will be a closer mimic of Cys than Ser.Altogether, the results support a role of Cys108 in fine-tuning ofactivity and thermal stability of TvDAO. Finally, it was shown by MSthat native TvDAO as isolated from T. variabilis contains Arg110, notAla110 as previously thought [22]. Mutagenesis data reveal thatsubstitution of Arg110 by Ala results in accelerated release of FADcofactor, hence decreased stability at elevated temperature. Therefore,even if Ala110 existed in a (minor) isozymic form of TvDAO that mayhave been captured in the earlier study [22], its presence is notcompatible with native properties of the enzyme.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.bbapap.2010.02.009.

References

[1] M.J. Davies, The oxidative environment and protein damage, Biochim. Biophys.Acta 1703 (2005) 93–109.

[2] T. Grune, K. Merker, G. Sandig, K.J. Davies, Selective degradation of oxidativelymodified protein substrates by the proteasome, Biochem. Biophys. Res. Commun.305 (2003) 709–718.

[3] T. Nyström, Aging in bacteria, Curr. Opin. Microbiol. 5 (2002) 596–601.[4] E.R. Stadtman, H. Arai, B.S. Berlett, Protein oxidation by the cytochrome P450

mixed-function oxidation system, Biochem. Biophys. Res. Commun. 338 (2005)432–436.

[5] N. Le Moan, G. Clement, S. Le Maout, F. Tacnet, M.B. Toledano, The Saccharomycescerevisiae proteome of oxidized protein thiols: contrasted functions for thethioredoxin and glutathione pathways, J. Biol. Chem. 281 (2006) 10420–10430.

[6] V.I. Perez, R. Buffenstein, V. Masamsetti, S. Leonard, A.B. Salmon, J. Mele, B.Andziak, T. Yang, Y. Edrey, B. Friguet, W. Ward, A. Richardson, A. Chaudhuri,Protein stability and resistance to oxidative stress are determinants of longevity inthe longest-living rodent, the naked mole-rat, Proc. Natl. Acad. Sci. U. S. A. 106(2009) 3059–3064.

[7] J.V. Cross, D.J. Templeton, Thiol oxidation of cell signaling proteins: controlling anapoptotic equilibrium, J. Cell. Biochem. 93 (2004) 104–111.

[8] R.L. vanMontfort, M. Congreve, D. Tisi, R. Carr, H. Jhoti, Oxidation state of the active-site cysteine in protein tyrosine phosphatase 1B, Nature 423 (2003) 773–777.

[9] L.B. Poole, K.J. Nelson, Discovering mechanisms of signaling-mediated cysteineoxidation, Curr. Opin. Chem. Biol. 12 (2008) 18–24.

[10] J.O. Konz, J. King, C.L. Cooney, Effects of oxygen on recombinant proteinexpression, Biotechnol. Prog. 14 (1998) 393–409.

[11] D.Houde, P. Kauppinen,R.Mhatre, Y. Lyubarskaya,Determinationofproteinoxidationby mass spectrometry and method transfer to quality control, J. Chromatogr. A 1123(2006) 189–198.

[12] S. Orru, A. Amoresano, R. Siciliano, R. Napoleoni, O. Finocchiaro, A. Datola, E. De Luca,A. Sirna, P. Pucci, Structural analysis of modified forms of recombinant IFN-betaproduced under stress-simulating conditions, Biol. Chem. 381 (2000) 7–17.

[13] C.E. Stauffer, D. Etson, The effect on subtilisin activity of oxidizing a methionineresidue, J. Biol. Chem. 244 (1969) 5333–5338.

[14] W.J. Van Berkel, F. Müller, The elucidation of the microheterogeneity of highlypurified p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens byvarious biochemical techniques, Eur. J. Biochem. 167 (1987) 35–46.

[15] F.J. van der Bolt, M.C. Drijfhout, M.H. Eppink,W.R. Hagen,W.J. van Berkel, Selectivecysteine→serine replacements in p-hydroxybenzoate hydroxylase from Pseudo-monas fluorescens allow the unambiguous assignment of Cys211 as the site ofmodification by spin-labeled p-chloromercuribenzoate, Protein Eng. 7 (1994)801–804.

[16] C.E. Grey, F. Rundbäck, P. Adlercreutz, Improved operational stability ofchloroperoxidase through use of antioxidants, J. Biotechnol. 135 (2008) 196–201.

[17] I. Garcia-Saez, J.D. Docquier, G.M. Rossolini, O. Dideberg, The three-dimensionalstructure of VIM-2, a Zn-β-lactamase from Pseudomonas aeruginosa in its reducedand oxidised form, J. Mol. Biol. 375 (2008) 604–611.

[18] N. Fujiwara, M. Nakano, S. Kato, D. Yoshihara, T. Ookawara, H. Eguchi, N.Taniguchi, K. Suzuki, Oxidative modification to cysteine sulfonic acid of Cys111in human copper–zinc superoxide dismutase, J. Biol. Chem. 282 (2007)35933–35944.

[19] D.J. Kemble, G. Sun, Direct and specific inactivation of protein tyrosine kinases in theSrc and FGFR families by reversible cysteine oxidation, Proc. Natl. Acad. Sci. U. S. A.106 (2009) 5070–5075.

[20] L. Men, Y. Wang, The oxidation of yeast alcohol dehydrogenase-1 by hydrogenperoxide in vitro, J. Proteome Res. 6 (2007) 216–225.

[21] I. Dib, A. Slavica, W. Riethorst, B. Nidetzky, Thermal inactivation of D-amino acidoxidase from Trigonopsis variabilis occurs via three parallel paths of irreversibledenaturation, Biotechnol. Bioeng. 94 (2006) 645–654.

[22] A. Slavica, I. Dib, B. Nidetzky, Single-site oxidation, cysteine 108 to cysteinesulfinic acid, in D-amino acid oxidase from Trigonopsis variabilis and its structuraland functional consequences, Appl. Environ. Microbiol. 71 (2005) 8061–8068.

[23] R. Sanchez, M. Riddle, J. Woo, J. Momand, Prediction of reversibly oxidized proteincysteine thiols using protein structure properties, Protein Sci. 17 (2008) 473–481.

[24] A. Claiborne, T.C. Mallett, J.I. Yeh, J. Luba, D. Parsonage, Structural, redox, andmechanistic parameters for cysteine-sulfenic acid function in catalysis andregulation, Adv. Protein Chem. 58 (2001) 215–276.

[25] C. Jacob, J.R. Lancaster, G.I. Giles, Reactive sulphur species in oxidative signaltransduction, Biochem. Soc. Trans. 32 (2004) 1015–1017.

[26] T. Kinumi, Y. Shimomae, R. Arakawa, Y. Tatsu, Y. Shigeri, N. Yumoto, E. Niki,Effective detection of peptides containing cysteine sulfonic acid using matrix-assisted laser desorption/ionization and laser desorption/ionization on poroussilicon mass spectrometry, J. Mass Spectrom. 41 (2006) 103–112.

[27] A. Poljak, R. Grant, C.J. Austin, J.F. Jamie, R.D. Willows, O. Takikawa, T.K. Littlejohn,R.J. Truscott, M.J. Walker, P. Sachdev, G.A. Smythe, Inhibition of indoleamine 2,3dioxygenase activity by H2O2, Arch. Biochem. Biophys. 450 (2006) 9–19.

[28] D. Seth, J. Rudolph, Redox regulation of MAP kinase phosphatase 3, Biochemistry45 (2006) 8476–8487.

[29] J. Rua, J. Soler, F. Busto, D. de Arriaga, Specific and reversible inactivation ofPhycomyces blakesleeanus isocitrate lyase by ascorbate-iron: role of two redox-active cysteines, Fungal Genet. Biol. 35 (2002) 223–234.

[30] T. Schräder, J.R. Andreesen, Studies on the inactivation of the flavoprotein D-amino acid oxidase from Trigonopsis variabilis, Appl. Microbiol. Biotechnol. 45(1996) 458–464.

[31] N. Stadler, L. Váchová, A. Krasowska, M. Höfer, K. Sigler, Role of strategic cysteineresidues in oxidative damage to the yeast plasma membrane H+-ATPase causedby Fe- and Cu-containing Fenton reagents, Folia Microbiol. (Praha) 48 (2003)589–596.

[32] X. Fu, S.Y. Kassim, W.C. Parks, J.W. Heinecke, Hypochlorous acid oxygenates thecysteine switch domain of pro-matrilysin (MMP-7). A mechanism for matrixmetalloproteinase activation and atherosclerotic plaque rupture by myeloperox-idase, J. Biol. Chem. 276 (2001) 41279–41287.

[33] X. Fu, D.M. Mueller, J.W. Heinecke, Generation of intramolecular and intermolec-ular sulfenamides, sulfinamides, and sulfonamides by hypochlorous acid: apotential pathway for oxidative cross-linking of low-density lipoprotein bymyeloperoxidase, Biochemistry 41 (2002) 1293–1301.

1491M. Mueller et al. / Biochimica et Biophysica Acta 1804 (2010) 1483–1491

[34] J.M. Pullar, M.C. Vissers, C.C. Winterbourn, Glutathione oxidation by hypochlorousacid in endothelial cells produces glutathione sulfonamide as a major product butnot glutathione disulfide, J. Biol. Chem. 276 (2001) 22120–22125.

[35] C.C. Winterbourn, S.O. Brennan, Characterization of the oxidation products ofthe reaction between reduced glutathione and hypochlorous acid, Biochem. J. 326(1997) 87–92.

[36] N.M. Giles, G.I. Giles, C. Jacob, Multiple roles of cysteine in biocatalysis, Biochem.Biophys. Res. Commun. 300 (2003) 1–4.

[37] M. Fürlinger, D. Haltrich, K.D. Kulbe, B. Nidetzky, A multistep process isresponsible for product-induced inactivation of glucose–fructose oxidoreductasefrom Zymomonas mobilis, Eur. J. Biochem. 251 (1998) 955–963.

[38] J.L. Kimmel, P.A. Tipton, Inactivation of GDP-mannose dehydrogenase fromPseudomonas aeruginosa by penicillic acid identifies a critical active site loop, Arch.Biochem. Biophys. 441 (2005) 132–140.

[39] V.I. Tishkov, V.O. Popov, Protein engineering of formate dehydrogenase, Biomol.Eng. 23 (2006) 89–110.

[40] A.P. Vintem, N.T. Price, R.B. Silverman, R.R. Ramsay, Mutation of surface cysteine374 to alanine in monoamine oxidase A alters substrate turnover and inactivationby cyclopropylamines, Bioorg. Med. Chem. 13 (2005) 3487–3495.

[41] H. Yamamoto, K. Mitsuhashi, N. Kimoto, Y. Kobayashi, N. Esaki, Robust NADH-regenerator: improved alpha-haloketone-resistant formate dehydrogenase, Appl.Microbiol. Biotechnol. 67 (2005) 33–39.

[42] I. Dib, D. Stanzer, B. Nidetzky, Trigonopsis variabilis D-amino acid oxidase: controlof protein quality and opportunities for biocatalysis through production inEscherichia coli, Appl. Environ. Microbiol. 73 (2007) 331–333.

[43] W. Riethorst, A. Reichert, An industrial view on enzymes for the cleavage ofcephalosporin C, Chimia 53 (1999) 600–607.

[44] L. Pollegioni, S. Butò, W. Tischer, S. Ghisla, M.S. Pilone, Characterization of D-aminoacid oxidase from Trigonopsis variabilis, Biochem. Mol. Biol. Int. 31 (1993) 709–717.

[45] P.F. Fitzpatrick, Oxidation of amines by flavoproteins, Arch. Biochem. Biophys. 493(2010) 13–25.

[46] L. Pollegioni, G. Molla, S. Sacchi, E. Rosini, R. Verga, M.S. Pilone, Properties andapplications of microbial D-amino acid oxidases: current state and perspectives,Appl. Microbiol. Biotechnol. 78 (2008) 1–16.

[47] T. Schräder, J.R. Andreesen, Evidence for the functional importance of Cys298 inD-aminoacidoxidase from Trigonopsis variabilis, Eur. J. Biochem.218 (1993)735–744.

[48] W.Wang, B.A. Malcolm, Two-stage PCR protocol allowing introduction ofmultiplemutations, deletions and insertions using QuikChange Site-Directed Mutagenesis,Biotechniques 26 (1999) 680–682.

[49] C. Trampitsch, A. Slavica, W. Riethorst, B. Nidetzky, Reaction of Trigonopsisvariabilis D-amino acid oxidase with 2,6-dichloroindophenol: kinetic character-ization and development of an oxygen-independent assay of the enzyme activity,J. Mol. Catal. B Enzym. 32 (2005) 271–278.

[50] A.F.S.A. Habeeb, Reaction of protein sulfhydryl groups with Ellman's reagent,Methods Enzymol. 25 (1972) 457–464.

[51] J.C. Morris, The acid ionization constant of HOCl from 5 to 35°, J. Phys. Chem. 70(1966) 3798–3805.

[52] R. Birner-Gruenberger, H. Susani-Etzerodt, M. Waldhuber, G. Riesenhuber, H.Schmidinger, G. Rechberger, M. Kollroser, J.G. Strauss, A. Lass, R. Zimmermann, G.Haemmerle, R. Zechner, A. Hermetter, The lipolytic proteome of mouse adiposetissue, Mol. Cell. Proteomics 4 (2005) 1710–1717.

[53] P. Nagy, M.T. Ashby, Reactive sulfur species: kinetics and mechanisms of theoxidation of cysteine by hypohalous acid to give cysteine sulfenic acid, J. Am.Chem. Soc. 129 (2007) 14082–14091.