The txl1+ gene from Schizosaccharomyces pombe encodes a new thioredoxin-like 1 protein that...

YeastYeast 2007; 24: 481–490.Published online 3 May 2007 in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/yea.1483

Research Article

The txl1+ gene from Schizosaccharomyces pombeencodes a new thioredoxin-like 1 protein thatparticipates in the antioxidant defence againsttert-butyl hydroperoxide

Alberto Jimenez1, Laura Mateos1, Jose R. Pedrajas2, Antonio Miranda-Vizuete3 and Jose L. Revuelta1*1Grupo de Ingenier ıa Metabolica, Instituto de Microbiolog ıa Bioquımica y Departamento de Microbiolog ıa y Genetica, CSIC/Universidad deSalamanca, Campus Miguel de Unamuno, 37007 Salamanca, Spain2Grupo de Senalizacion Molecular y Sistemas Antioxidantes en Plantas, Unidad Asociada al CSIC (EEZ), Area de Bioqu ımica y Biolog ıaMolecular, Universidad de Jaen, Spain3Centro Andaluz de Biolog ıa del Desarrollo (CABD-CSIC), Departamento de Fisiologıa, Anatomıa y Biolog ıa Celular, Universidad Pablo deOlavide, 41013 Sevilla, Spain

*Correspondence to:Jose L. Revuelta, Grupo deIngenier ıa Metabolica, Institutode Microbiolog ıa Bioquımica yDepartamento de Microbiolog ıay Genetica, CSIC/Universidad deSalamanca, Campus Miguel deUnamuno, 37007Salamanca, Spain.E-mail: [email protected]

Received: 10 October 2006Accepted: 5 February 2007

AbstractYeasts are equipped with several putative single-domain thioredoxins located in differ-ent subcellular compartments. However, additional proteins containing thioredoxindomains are also encoded by the yeast genomes as described for mammals and othereukaryotic organisms. We report here the characterization of the fission yeast ortho-logue thioredoxin-like 1 (txl1 +), which has been previously identified in mammals.Similarly to the human protein, the fission yeast Txl1 is a two-domain protein com-prising an N-terminal thioredoxin-like domain and a C-terminal domain of unknownfunction. Many other yeasts and fungi species contain homologues of txl1 +; how-ever, there is no evidence of txl1 + orthologues in either Saccharomyces cerevisiae orplants. Txl1 is found in both the nucleus and the cytoplasm of Schizosaccharomycespombe cells and exhibits a strong reducing activity coupled to thioredoxin reductase.In humans, TXL1 expression is induced by glucose deprivation and overexpressionof TXL1 confers resistance against this stress. In contrast, a Sz. pombe �txl1 mutantwas not affected in the response against glucose starvation but the �txl1 mutantstrain showed a clear hypersensitivity to alkyl hydroperoxide. The mRNA levels oftxl1 + in a h20 strain did not change in response to any oxidative insult (hydro-gen peroxide or alkyl hydroperoxide) and the overexpression of an integrated copyof the wild-type txl1 + gene did not confer a significant increased resistance againstalkyl hydroperoxide. Overall, these results indicate that the Txl1 role in the cellulardetoxification of alkyl hydroperoxide is exerted through a constitutive transcriptionof txl1 +. Copyright 2007 John Wiley & Sons, Ltd.

Keywords: thioredoxin-like; antioxidant; tert-butyl hydroperoxide; Schizosaccha-romyces pombe

Supplementary material for this article can be found at http://www.interscience.wiley.com/jpages/0749-503X/suppmat/

Introduction

Thioredoxins (Trx) are redox proteins that func-tion as general protein disulphide oxidoreduc-tases, maintaining the reduced cellular environment

(Hirota et al., 2002; Nakamura, 2005). The redoxactivity of thioredoxins is based on the ability of thecysteines of their active site (Cys–Gly–Pro–Cys)to undergo reversible oxidation from a dithiolto a disulphide form. Thioredoxins activity is

Copyright 2007 John Wiley & Sons, Ltd.

482 A. Jimenez et al.

coupled to the flavoenzyme thioredoxin reduc-tase (TrxR), which maintains thioredoxins in theirreduced active form, using NADPH as the elec-tron donor (Holmgren, 2000). It has been describedthat alterations in the thioredoxin system can leadto several pathological processes (Burke-Gaffneyet al., 2005).

All organisms investigated so far, from prokary-otes to humans, contain different forms of thiore-doxins (Hirota et al., 2002; Nakamura, 2005). Sac-charomyces cerevisiae contains two thioredoxinsystems: one in the cytosol, composed of twothioredoxins (Trx1 and Trx2) and one thioredoxinreductase (Trr1), and the other in the mitochondria,formed by a thioredoxin (Trx3) and a thioredoxinreductase (Trr2) (Gan, 1991; Pedrajas et al., 1999).In contrast, the genome of Schizosaccharomycespombe encodes for only one cytosolic thioredoxin(Trx1), one mitochondrial thioredoxin (Trx2) andone thioredoxin reductase (Trr1), which can belocalized in both in the cytosol and the mitochon-dria (Casso and Beach, 1996; Cho et al., 2001; Leeet al., 2002).

In fission yeast, the expression of the cytoso-lic Trx1 is moderately induced by hydrogen per-oxide and other environmental stresses, while themitochondrial Trx2 is not. Consequently, trx1 + isconsidered as a part of the ‘core environmentalstress response’ (CESR) in Sz. pombe (Chen et al.,2003). Additionally, the expression of thioredoxinreductase is strongly induced in response to bothhydrogen peroxide and methyl methane sulphonate(MMS) (Chen et al., 2003; Weeks et al., 2006).

The structure of thioredoxins is globular and con-sists of a central core of β-sheets surrounded byα-helices, with the active site situated in a protru-sion emerging from the protein surface (Jeng et al.,1994). While all thioredoxins reported from yeastspecies comprise only one thioredoxin domain,some mammalian members of the thioredoxin fam-ily are composed of additional domains with knownor unknown counterparts in the databases (Cun-nea et al., 2003; Lee et al., 1998; Miranda-Vizueteet al., 1998, 2001; Sadek et al., 2001, 2003).

In a recent study we have characterized thehuman thioredoxin-like 1 (TXL1), which is a two-domain protein composed of a N-terminal thiore-doxin domain followed by a C-terminal domain ofunknown function with no homology to any otherprotein in the databases (Jimenez et al., 2006).Human TXL1 is a cytosolic thioredoxin that can

also translocate to the nucleus; it is predominantlyexpressed in the central nervous system and otherorgans with an elevated metabolic rate and isinvolved in the cellular response to glucose depri-vation (Jimenez et al., 2006).

We have identified a TXL1 orthologue in a Sz.pombe database (http://www.sanger.ac.uk/Pro-jects/S pombe/) and here we show that the Sz.pombe Txl1 is a novel thioredoxin-like protein,which participates in the cellular protection againstoxidative stress induced by alkyl hydroperoxide.

Materials and methods

Strains, growth conditions and chemicals

All the Sz. pombe strains used in this work arelisted in Table 1. Sz. pombe cells were routinelygrown at 28 ◦C in YES rich medium or mini-mal EMM medium with the required supplements(Moreno et al., 1991). Growth was monitored spec-trophotometrically at 595 nm and standard geneticmanipulations were used (Moreno et al., 1991).Sz. pombe transformations were carried out asdescribed elsewhere (Ito et al., 1983).

Amino acids, insulin, NADPH, dithiothreitol(DTT), geneticin (G418), H2O2 and tert-butylhydroperoxide (t-BOOH) were purchased fromSigma (Steinheim, Germany).

Sz. pombe txl1+ protein expression andpurification

The ORF encoding the Sz. pombe Txl1 wasPCR-amplified from cDNA prepared using theSuperScript II RT enzyme (Invitrogen, Carls-bad, USA) and Sz. pombe total RNA as tem-plate. The ORF was cloned into the BamHI–EcoRIsites of the pGEX-4T-1 expression vector (Amser-shan Biosciences, Madrid, Spain), verified bysequencing and used to transform Escherichia

Table 1. Sz. pombe strains used in this work

Strain Genotype Source

h20 h− leu1.32 S. Moreno∗SP-3 h− leu1.32, txl1-GFP This workSP-8 h− leu1.32, txl1::kanMX6 This workSP-14 h− leu1.32, txl1::P3nmt1-txl1 This work

∗ Centro de Investigacion del Cancer, University of Salamanca, Spain.

Copyright 2007 John Wiley & Sons, Ltd. Yeast 2007; 24: 481–490.DOI: 10.1002/yea

Thioredoxin-like 1 from Schizosaccharomyces pombe 483

coli HMS174(DE3). Induction and purification ofthe recombinant protein was achieved as previ-ously reported (Miranda-Vizuete et al., 1998). Anovernight incubation with thrombin (5 U/mg offusion protein) was used to remove the Txl1 recom-binant protein from the glutathione S -transferasedomain. Sz. pombe Txl1 was eluted and proteinconcentration was determined using the Bio-Radprotein assay kit (Bio-Rad, Madrid, Spain), usingBSA as a standard.

Enzymatic thioredoxin activity assays

Enzymatic activity of recombinant Txl1 was per-formed using two different assays. In the DTTassay, DTT was used as the reducing agent andthe assay was carried out as previously described(Wollman et al., 1988). In the thioredoxin reduc-tase assay, recombinant Txl1 activity was deter-mined by its capability to reduce insulin disulphidebonds, using NADPH as electron donor in the pres-ence of the mitochondrial thioredoxin reductase(Trr2) from S. cerevisiae (Pedrajas et al., 1999).The activity assay was performed essentially asdescribed elsewhere (Spyrou et al., 1997) but mon-itoring insulin precipitation at 595 nm. In bothcases, the mitochondrial thioredoxin (Trx3) fromS. cerevisiae (Pedrajas et al., 1999) was used as apositive control.

Gene deletion, overexpression and greenfluorescent protein (GFP) tagging

The entire txl1 + ORF was replaced in a h20strain of Sz. pombe with a kanMX6 cassette,which confers resistance to G418, following themethod described by Bahler et al. (1998). The dele-tion cassette was constructed by PCR using thepFA6kanMX6 vector (Bahler et al., 1998) as tem-plate and 90–95 nucleotide long primers (Supple-mentary Table 1).

For the overexpression of the Sz. pombe txl1 +,we used PCR techniques to replace the native txl1 +promoter by the inducible strong nmt1 + promoter.A pFA6kanMX6–P3nmt1 vector was used as thePCR template (Bahler et al., 1998), using appropri-ate primers (Supplementary Table 1). The nmt1 +promoter allows expression at a relatively highlevel in culture media lacking thiamine. For over-expression experiments using the nmt1 + promoter,cells were grown in EMM medium (without thi-amine).

Gene deletion and integration of the nmt1 +promoter in Sz. pombe was confirmed by Southernblotting and expression analysis was verified bynorthern blotting.

GFP C-terminal tagging was achieved using alsoPCR techniques, employing pFA6–GFP(S65T)–kanMX6 (Bahler et al., 1998) and using the primerslisted in Supplementary Table 1. A fusion fragmentcontaining the txl1 + ORF in-frame to the GFP cod-ing region was constructed by direct genome inte-gration. The fluorescence of the Txl1–GFP(S65T)fusion protein was monitored in living cells as pre-viously described (Niedenthal et al., 1996). Micro-graphs were acquired using a Photometrics SensysCCD camera coupled to a Leica DMXRA micro-scope equipped with Nomarski optics and epifluo-rescence.

Northern blot analysis

Cells grown under different conditions were har-vested and total RNA was prepared as previouslydescribed (Percival-Smith and Segall, 1984). Fornorthern blot analyses, 10 µg of each RNA samplewas used. A 580 bp HindIII fragment from the Sz.pombe txl1 + ORF was labelled with [α-32P] dCTP(Rediprime random primer labelling kit; Amer-sham Pharmacia-Biotech) and used as a radioactiveprobe. The blots were also hybridized with a 1.3 kbSacII–HindII fragment of the Sz. pombe β-actinORF as control. For quantitative analysis, the blotswere scanned and quantified using a BAS1500-Macimage analyzer (Fuji Film Co.).

Glucose deprivation and oxidative treatments

For the glucose deprivation treatment, cells weregrown in YES medium with 2% glucose, 0.5%glucose or 2% glycerol (without glucose). Cultureswere carried out in 96-well microtiter plates, using180 µl culture media/well. The h20 and ∆txl1strains were replicated from a saturated master 96-well plate in which both strains had been previouslygrown in YES medium.

Additionally, 10 ml EMM cultures were achie-ved with h20, ∆txl1 and P3nmt1-txl1 + strains,using 2% glucose (EMM) or 0.5% glucose (EMM-L). In this case, cultures were initiated from satu-rated precultures, using 10–50 µl.

For the treatments with hydrogen peroxide andt-BOOH, cells were grown in EMM medium



(lacking thiamine) to allow the overexpression oftxl1 + in the P3nmt1-txl1 + strain.

Results

Cloning of a novel thioredoxin-like proteinfrom Sz. pombe

A BLAST search (http://www.ncbi.nlm.nih.gov/BLAST/) using the human TXL1 ORF as abait identified several entries, including a Sz.pombe orthologue (ORF No. SPBC577.08c), whichshowed a 38% of identity at the protein level whencompared with that of the human TXL1. The infor-mation related to the SPBC577.08c ORF in the Sz.pombe Database (http://www.genedb.org/genedb/pombe/index.jsp) considered the ORF product tobe a member of the thioredoxin family, infer-ring its role from homology comparisons. TheSPBC577.08c gene comprises 925 bp, maps atchromosome 2 of the Sz. pombe genome and isorganized into two exons, with a spliced length of873 bp.

Surprisingly, there is no apparent orthologuein the S. cerevisiae genome, although the geneis present all along the evolution from lowereukaryotes to humans (see Supplementary Figure 1in Jimenez et al., 2006), indicating that the pro-tein could have an important cellular role. There-fore, we designed specific primers (SupplementaryTable 1) to amplify the SPBC577.08c ORF by PCRfrom a Sz. pombe cDNA library.

Structurally, the SPBC577.08c protein of Sz.pombe (290 residues) comprises a N-terminal

thioredoxin DUF1000human

thioredoxin DUF1000N. crassa

thioredoxin DUF1000C. neoformans

thioredoxin DUF1000Y. lipolytica

thioredoxin DUF1000C. albicans

thioredoxinSz. pombe DUF1000

1 50 100 150 200 250 350300

number of residues

Figure 1. Schematic domain organization of thiore-doxin-like 1 proteins from human and different yeast species(Neurospora crassa, Cryptococcus neoformans, Yarrowia lipolyt-ica, Candida albicans and Schizosaccharomyces pombe)

thioredoxin-like domain and a C-terminal domainof unknown function (DUF1000), as described forthe human TXL1 (Jimenez et al., 2006) and otherorthologues (Figure 1). The predicted chemicalproperties of the SPBC577.08c encoding proteinare also the same as described for other orthologues(Jimenez et al., 2006), with no subcellular local-ization signals within the protein sequence either.Therefore, we decided to consider SPBC577.08c asthe corresponding thioredoxin-like 1 (txl1 +) genein Sz. pombe.

Sz. pombe Txl1 shows reducing thioredoxinactivity

We wanted to confirm whether the Txl1 from Sz.pombe was able to catalyse the reduction of disul-phide bonds, thus having thioredoxin activity. Con-sequently, we expressed a recombinant GST–Txl1fusion protein in E. coli and purified a thrombin-cleaved Txl1 (see Materials and methods).

We next checked the enzymatic activity of therecombinant Sz. pombe Txl1, using DTT or thiore-doxin reductase coupled to NADPH as reductants,using the mitochondrial thioredoxin (Trx3) fromS. cerevisiae as a positive control (Pedrajas et al.,1999). As shown in Figure 2, the recombinant Txl1from Sz. pombe is able to reduce the insulin disul-phide bonds either with DTT or thioredoxin reduc-tase. Remarkably, the kinetics of S. cerevisiae Trx3and Sz. pombe Txl1 were quite similar and thespecific thioredoxin activity of both protein prepa-rations was also comparable. Conversely, we havepreviously reported that human TXL1 thioredoxinactivity shows a long latency phase and muchlower reducing activity than that of human TRX1(Jimenez et al., 2006). This discrepancy betweenthe human and Sz. pombe Txl1 thioredoxin activi-ties might reflect other functional divergences.

Subcellular localization of the Sz. pombe Txl1

In order to investigate the subcellular localizationof Txl1, we carried out C-terminal GFP tagging ofthe Sz. pombe Txl1 protein. Fluorescence could beseen in both the nucleus and the cytoplasm of cellsexpressing the Txl1–GFP fusion (Figure 3), indi-cating that Txl1 can translocate into the nucleuseven though a nuclear localization signal is lack-ing in its sequence. The same localization pat-tern has been previously reported for the human



0

0.1

0.2

0 30 60 90

Time (min)

O.D

. (59

5nm

)

Trx3Txl1

A

0

0.1

0.2

0.3

0 30 60 90

Time (min)

O.D

. (59

5nm

)

Trx3Txl1TR control

B

Figure 2. Thioredoxin activity assay of Txl1 from Sz. pombe.(A) Purified recombinant Txl1 (2 µM) from Sz. pombe wasassayed for its ability to reduce insulin disulphide bondsin the presence of NADPH and DTT. (B) Thioredoxinactivity assay of Sz. pombe Txl1 (2 µM), using NADPH and S.cerevisiae thioredoxin reductase (Trr2, 0.5 µM). Trx3 (2 µM)from S. cerevisiae was used as positive control. In the TRcontrol, the assay was performed without any thioredoxin.Identical results were obtained from three independentexperiments

TXL1 (Jimenez et al., 2006) and, more recently,a comprehensive global analysis of the Sz. pombeORFome has also confirmed these localizations(Matsuyama et al., 2006). Nuclear translocationhas also been reported for other thioredoxins lack-ing nuclear targeting signals (Hirota et al., 1999).

We also checked whether the pro-oxidants t-BOOH and H2O2 (see below for details) couldchange the subcellular localization of Txl1. How-ever, the treatments with t-BOOH or H2O2 at dif-ferent time points did not alter the localizationpattern of Txl1 (Figure 3).

Figure 3. Txl1 from Sz. pombe localizes in both thenucleus and the cytosol. Sz. pombe cells expressing anintegrated txl1+ –GFP(S65T) fusion were monitored underepifluorescence microscopy. Cells were grown on YESmedium and treatments with 0.5 mM t-BOOH or 2 mMH2O2 at 1, 4, 8 and 12 h were assayed. Images ofdifferential interference contrast microscopy (DIC) andepifluorescence microscopy of the same field representingthe GFP channel (GFP) are shown. For t-BOOH and H2O2treatments, pictures corresponding to 4 h incubation areshown

Sz. pombe Txl1 is not involved in the cellularresponse to glucose starvation

Our previous results indicate that human TXL1could play a protective role against sugar starvationstress (Jimenez et al., 2006). Thus, we decided tocheck whether Sz. pombe Txl1 could be functioningin a way similar to human TXL1 regarding theresponse against glucose deprivation.

Initially we looked for the Sz. pombe txl1 +transcription pattern in h20 cells cultured underglucose starvation stress. A northern blot analy-sis revealed that the txl1 + mRNA levels did not



0.6

0.5

0.4

0.3

0.2

0.1

0

O. D

. 595

nm

2% glucose

h20∆txl1

0 12 24 36 48 60 72 84 96 108

time (hours)

0.4

0.3

0.2

0.1

0

O. D

. 595

nm

0.5% glucose

h20∆txl1

0 12 24 36 48 60 72 84 96 108

time (hours)

0.6

0.5

0.4

0.3

0.2

0.1

0

O. D

. 595

nm

2% glycerol

glucose

h20∆txl1

0 12 24 36 48 60 72 84 96 108120

time (hours)

B

txl1+

Actin

2% glucose 0.5% glucose

0h 1h 2h 4h 8h 0h 1h 2h 4h 8h 0h 1h 2h 4h 8h

glycerolA

Figure 4. Sz. pombe Txl1 is not involved in the cellular response against glucose deprivation. (A) h20 cells growing inYES medium (OD595nm = 0.4) were washed and transferred to fresh media containing different carbon sources. Aliquotswere taken for 8 h and RNA levels of txl1+ were determined by northern blotting. (B) h20 and ∆txl1 strains were grownin YES media containing the indicated carbon sources. Cultures were made in 96-well microtitre plates and growth wasmonitored at 595 nm for 108 h. Exogenous glucose (2%) was added to the glycerol cultures after 108 h

txl1+ txl1+EcoRI

EcoRI

EcoRI

EcoRI

EcoRI

EcoRI

EcoRI EcoRI

probe

1.7 kb

h20

kanMx6

0.4 kb1.8 kb

∆txl1

txl1+ txl1+P3nmt1+kanMX6

3.9 kb0.2 kb

P3nmt1-txl1

A B

h20 ∆txl1P3nmt1-txl1+ h20 ∆txl1

P3nmt1-txl1+

-thiamine +thiamine

txl1+

28SrRNA

C

h20

∆txl1+

P3nmt1-

txl1+

kb

2

0.5

4

0.1

Figure 5. Construction of mutant ∆txl1 knock-out and overexpressing P3nmt1-txl1 strains. (A) Schematic representationof the wild-type txl1+, txl1 :: kanMX6 and txl1 :: P3nmt1-txl1 loci. The genomic fragment used as radioactive probe in theSouthern blot analysis is indicated. (B) Southern blot analysis of EcoRI-digested genomic DNA from the h20, ∆txl1 andP3nmt1- txl1+ strains. (C) northern blot analysis of total RNA (10 µg) from the h20, ∆txl1 and P3nmt1-txl1+ strains grownin EMM (without thiamine) and YES (with thiamine) media

change during the treatment when glucose or glyc-erol were used as carbon sources (Figure 4A).Hence, our results seem to indicate that txl1 +transcription is not affected by glucose depriva-tion.

To confirm this assumption, we decided to inves-tigate whether the absence of TXL1 might influ-ence the growth pattern of Sz. pombe under glu-cose starvation stress. Thus, we constructed a ∆txl1

null mutant by replacing the TXL1 gene with thekanMX6 dominant marker, which confers resis-tance to G418 (see Materials and methods fordetails). Gene deletion was verified by Southernand northern blot (Figure 5).

The ∆txl1 strain did not show any apparentmorphological defect and the growth pattern inrich medium (2% glucose) was identical to theh20 parental strain. As shown in Figures 4B, 6A,



there were no differences among the ∆txl1 and theh20 strains with low (0.5%) glucose concentrationsin the culture media. Glycerol did not supportthe growth of either the ∆txl1 or the h20 strains(Figure 4B) and the addition of exogenous glucoseafter 108 h immediately induced the same growthenhancement in both strains. Therefore, we canconclude that Txl1 is not involved in the glucosestarvation response in Sz. pombe.

The ∆txl1 strain of Sz. pombe showshypersensitivity to an alkyl hydroperoxide

The results shown above indicate that human andSz. pombe txl1 + orthologues might accomplishdifferent tasks in the cell. Sz. pombe Txl1 shows

a high disulphide-reducing activity, which mightreflect a possible role for Txl1 in cellular defenceagainst oxidants. Therefore, to further characterizethe ∆txl1 strain we decided to explore the effectof two pro-oxidants, hydrogen peroxide (H2O2)

and tert-butyl hydroperoxide (t-BOOH), on thegrowth of the ∆txl-1 strains in liquid cultures.The h20 and ∆txl1 strains were grown in minimalmedia with 2 mM H2O2 or 0.5 mM t-BOOH for40 h. Our results showed that hydrogen peroxidedoes not affect the growth of the ∆txl1 strainwhen compared with the h20 strain (Figure 6A).However, t-BOOH did induce a marked delayin the growth of the ∆txl1 strain (Figure 6A),suggesting that Txl1 might be involved in the

time (hours)

EMM

h20∆txl1P3nmt1-txl1+

A 10.90.80.70.60.50.40.30.20.1

00 4 8 24 32 40

O. D

. 595

nm

EMM + H2O2


time (hours)

0 4 8 24 32 40

10.90.80.70.60.50.40.30.20.1

0

O. D

. 595

nm

EMM + t-BOOH

time (hours)

0 4 8 24 32 40


10.90.80.70.60.50.40.30.20.1

0

O. D

. 595

nm

EMM-L


time (hours)

0 4 8 24 32 40

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

O. D

. 595

nm

H2O2 t-BOOH

txl1+

Actin

0h 4h 8h 12h 0h 4h 8h 12h

B

Figure 6. Sz. pombe Txl1 is required for the cellular resistance against t-BOOH. (A) h20, ∆txl1 and P3nmt1-txl1+ strainswere grown on minimal EMM media with 2% glucose (EMM), 0.5% glucose (EMM-L) or supplemented with H2O2 (2 mM)or t-BOOH (0.5 mM). (B) h20 cells growing in YES medium (OD595nm = 0.4) were washed and transferred to fresh mediacontaining H2O2 (2 mM) or t-BOOH (0.5 mM). Aliquots were taken for 12 h and RNA levels of txl1+ were determined bynorthern blotting



detoxification of t-BOOH and probably other alkylhydroperoxides.

We assumed that Txl1 protection against alkylhydroperoxides might be mediated by an inductionof txl1 + transcription. However, when we testedthe effect of the two pro-oxidants previously usedon the levels of the txl1 + transcripts in a h20 strain,we found that neither t-BOOH nor H2O2 were ableto induce an increase on the mRNA levels of txl1 +(Figure 6B). To confirm this result, we expressedan integrated copy of txl1 + under the control ofthe strong promoter nmt1 + (Figure 5) and checkedthe growth rate of the txl1 + overexpressing strainin the presence of t-BOOH or H2O2. As shown inFigure 6A, no significant increase in the resistanceto t-BOOH was achieved through the overexpres-sion of txl1 +, indicating that non-transcriptionalmechanisms might underlie the protective effect ofTxl1 against t-BOOH.

Discussion

Redox balance is important in the cellular phys-iology and is maintained by means of a steady-state situation between oxidants and antioxidants(Kondo et al., 2006; Nakamura, 2005). However,it has been proposed that oxidative stress may beconsidered as a disruption of redox signalling andcontrol, and that specific redox mechanisms deter-mine discrete signalling and control events (Jones,2006).

Thioredoxin like-1 was first identified and char-acterized in humans (Lee et al., 1998; Miranda-Vizuete et al., 1998). More recently, we pro-posed that human TXL1 participates in the cellu-lar response against glucose deprivation-mediatedstress in a TRX1-independent manner (Jimenezet al., 2006). In this work, we have identified thefission yeast orthologue of the human TXL1 andwe have characterized it in order to complementour data with the human TXL1.

We have demonstrated that the thioredoxin-like1 protein from Sz. pombe has a strong reducingactivity coupled to thioredoxin reductase, differingsignificantly from that of the human TXL1, whichshows a much lower activity. A possible expla-nation for this difference may be the fact that theactive site of the fission yeast protein is preceded bya tryptophan residue (WCGPC), as occurs in mostthioredoxins (Holmgren and Bjornstedt, 1995). In

humans and other mammals, the tryptophan residueis substituted in the TXL1 protein by a glycineresidue (GCGPC) (Jimenez et al., 2006). Indeed,the presence of the tryptophan in the thioredox-ins active site has been proposed to regulate theircatalytic activity (Krause and Holmgren, 1991).

The domain organization of the Sz. pombethioredoxin-like 1 protein is identical to theother reported orthologues, having a thioredoxin-like N-terminal domain and a C-terminal domain(DUF100), which has been suggested to bea regulatory domain (Jin et al., 2002). Thesubcellular localization of the Sz. pombe Txl1protein also coincides with the nuclear andcytosolic localization of the human TXL1.Furthermore, neither Sz. pombe Txl1 nor humanTXL1 have any nuclear translocation signal in theirrespective sequences, so the translocation into thisorganelle must be mediated by co-transport withother proteins, rather than passive diffusion, givenits size (Jimenez et al., 2006).

The transcription of txl1 + remained unaltered inresponse to the stress induced by glucose starva-tion or hydroperoxides. This feature constitutes asignificant difference between the human and thefission yeast orthologues. Indeed, a mutant ∆txl1strain did not show any growth defect when cul-tured in media lacking glucose, with low glucoseconcentrations or in the presence of hydrogen per-oxide. However, the deletant ∆txl1 displayed anotorious growth defect when an alkyl hydroperox-ide (t-BOOH) was present in the culture mediumbut, surprisingly, a mutant strain overexpressingtxl1 + did not exhibit higher resistance to t-BOOHcompared to that of the Sz. pombe h20 strain,indicating that induction of txl1 + transcription isnot responsible for the Txl1 protection against t-BOOH. A genome-wide analysis of S. cerevisiaeviable deletion strains showed that most genesinvolved in resistance to oxidative stress are nottranscriptionally induced (Thorpe et al., 2004). Ithas also been shown that stress-mediated induc-tion in many genes is evolutionarily conserved inS. cerevisiae and Sz. pombe; and the same workhas reported that only eight genes with annotatedantioxidant function are included in the core envi-ronmental stress response (CESR) in Sz. pombe(Chen et al., 2003). Indeed, most of the genes thatare induced in response to hydrogen peroxide inSz. pombe do not code for antioxidant enzymes(Chen et al., 2003). Therefore, we can consider



that Sz. pombe is equipped with a set of con-stitutively expressed genes that participate in theantioxidant housekeeping defence, such as txl1 +in response to t-BOOH or grx2 + against paraquat(Chung et al., 2004). Additionally, Sz. pombe hasthe CESR set of genes that are transcriptionallyinduced by oxidative stress, as described for thetrx1 + and trr1 + genes (Chen et al., 2003). A recentreport has shown that posttranslational processingcontributes to the function of the cytosolic thiore-doxin 1 (Haendeler, 2006), and we cannot excludethe possibility that Txl1 might suffer some post-translational modifications to regulate its function.

Another question was the possibility that Txl1function might be conditioned through its localiza-tion in either the nucleus or the cytosol. However,neither H2O2 nor t-BOOH induced a translocationof the GFP-tagged Txl1 protein.

A S. cerevisiae ahp1 mutant lacking an alkylhydroperoxide reductase has been described tobe hypersensitive to t-BOOH (Lee et al., 1999).The fission yeast database at the Sanger Institu-te (http://www.genedb.org/genedb/pombe/index.jsp) displays the entry number SPCC330.06cas a sequence similar to the S. cerevisiaeAHP1, although the identity at protein level isonly 25%. The SPCC330.06c gene product isannotated as a thioredoxin peroxidase and thesubcellular localization has been shown to beboth cytosolic and nuclear (Matsuyama et al.,2006). Whether SPCC330.06c is involved inthe resistance against alkyl hydroperoxides isstill unknown but, considering that SPCC330.06cand Txl1 co-localize within the same subcellularcompartments in Sz. pombe, and that the proposedSPCC330.06c orthologue in S. cerevisiae functionsas an alkyl hydroperoxide reductase, we mightspeculate that SPCC330.06c and Txl1 couldparticipate in the same specific mechanism againstalkyl hydroperoxides.

A comparison of proteins among Sz. pombeand S. cerevisiae revealed that gene duplicationhad occurred more in S. cerevisiae than in Sz.pombe, thus accounting for the high number ofextra proteins found in S. cerevisiae (Wood et al.,2002). However, there are a number of Sz. pombegenes (17%) with no homologues in S. cerevisiae(Wood et al., 2002). The budding yeast genomedoes not encode for a txl1 + orthologue. Never-theless, other yeast species and higher eukaryoticorganisms retain a gene encoding a thioredoxin-like

protein. Neofunctionalization mechanisms in whicha duplicated gene may lose or retain functions,or acquire novel functions (He and Zhang, 2005),might explain the functional differences betweenthe human and Sz. pombe txl1 + orthologues. Suchfunctional divergence among eukaryotic specieshas been reported in other families of proteins asthe exportin-5 orthologues (Shibata et al., 2006).In the thioredoxin-like 1 family, mutations in thesequence of the active site leading to the substi-tution of a tryptophan residue (present in lowereukaryotes) by a glycine (present in mammals)may strongly affect the reducing thioredoxin activ-ity. In this regard, the human TXL1 gene mighthave acquired the ability to participate in the cellu-lar response against glucose deprivation; a func-tion that is not accomplished by the Sz. pombethioredoxin-like 1 protein.

Acknowledgements

This work was supported by Grants GEN2001-4707-C08-01 and AGL2005-07245-C03-03 from the Ministerio deEducacion y Ciencia, Spain. A.J. is a recipient of apostdoctoral contract (Programa Juan de la Cierva) fromthe Spanish Ministerio de Educacion y Ciencia. L.M. wassupported by a predoctoral fellowship from the Universidadde Salamanca, Spain. A.M.-V. is supported by a researchcontract under the Ramon y Cajal Programme of theSpanish Ministerio de Educacion y Ciencia. We thank M.D. Sanchez for excellent technical help and N. Skinner forcorrecting the manuscript.

References

Bahler J, Wu JQ, Longtine MS, et al. 1998. Heterologous modulesfor efficient and versatile PCR-based gene targeting inSchizosaccharomyces pombe. Yeast 14: 943–951.

Burke-Gaffney A, Callister ME, Nakamura H. 2005. Thioredoxin:friend or foe in human disease? Trends Pharmacol Sci 26:398–404.

Casso D, Beach D. 1996. A mutation in a thioredoxin reductasehomolog suppresses p53-induced growth inhibition in thefission yeast Schizosaccharomyces pombe. Mol Gen Genet 252:518–529.

Chen D, Toone WM, Mata J, et al. 2003. Global transcriptionalresponses of fission yeast to environmental stress. Mol Biol Cell14: 214–229.

Cho Y, Shin YH, Kim Y, et al. 2001. Characterization andregulation of Schizosaccharomyces pombe gene encodingthioredoxin. Biochim Biophys Acta 1518: 194–199.

Chung WH, Kim KD, Cho YJ, Roe JH. 2004. Differentialexpression and role of two dithiol glutaredoxins Grx1 and Grx2in Schizosaccharomyces pombe. Biochem Biophys Res Commun321: 922–929.



Cunnea PM, Miranda-Vizuete A, Bertoli G, et al. 2003. ERdj5, anendoplasmic reticulum (ER)-resident protein containing DnaJand thioredoxin domains, is expressed in secretory cells orfollowing ER stress. J Biol Chem 278: 1059–1066.

Gan ZR. 1991. Yeast thioredoxin genes. J Biol Chem 266:1692–1696.

Haendeler J. 2006. Thioredoxin-1 and posttranslational modifica-tions. Antioxid Redox Signal 8: 1723–1728.

He X, Zhang J. 2005. Rapid subfunctionalization accompanied byprolonged and substantial neofunctionalization in duplicate geneevolution. Genetics 169: 1157–1164.

Hirota K, Murata M, Sachi Y, et al. 1999. Distinct roles ofthioredoxin in the cytoplasm and in the nucleus. A two-stepmechanism of redox regulation of transcription factor NF-κB.J Biol Chem 274: 27891–27897.

Hirota K, Nakamura H, Masutani H, Yodoi J. 2002. Thioredoxinsuperfamily and thioredoxin-inducing agents. Ann N Y Acad Sci957: 189–199.

Holmgren A. 2000. Antioxidant function of thioredoxin andglutaredoxin systems. Antioxid Redox Signal 2: 811–820.

Holmgren A, Bjornstedt M 1995. Thioredoxin and thioredoxinreductase. Methods Enzymol 252: 199–208.

Ito H, Fukuda Y, Murata K, Kimura A. 1983. Transformation ofintact yeast cells treated with alkali cations. J Bacteriol 153:163–168.

Jeng MF, Campbell AP, Begley T, et al. 1994. High-resolutionsolution structures of oxidized and reduced Escherichia colithioredoxin. Structure 2: 853–868.

Jimenez A, Pelto-Huikko M, Gustafsson J-A, Miranda-Vizuete A.2006. Characterization of human thioredoxin-like-1: potentialinvolvement in the cellular response against glucose deprivation.FEBS Lett 580: 960–967.

Jin J, Chen X, Zhou Y, et al. 2002. Crystal structure of thecatalytic domain of a human thioredoxin-like protein. Eur JBiochem 269: 2060–2068.

Jones DP. 2006. Redefining oxidative stress. Antioxid Redox Signal8: 1865–1879.

Kondo N, Nakamura H, Masutani H, Yodoi J. 2006. Redoxregulation of human thioredoxin network. Antioxid Redox Signal8: 1881–1890.

Krause G, Holmgren A. 1991. Substitution of the conservedtryptophan 31 in Escherichia coli thioredoxin by site-directedmutagenesis and structure–function analysis. J Biol Chem 266:4056–4066.

Lee J, Spector D, Godon C, Labarre J, Toledano MB. 1999. Anew antioxidant with alkyl hydroperoxide defence properties inyeast. J Biol Chem 274: 4537–4544.

Lee K-K, Murakawa M, Takahashi S, et al. 1998. Purification,molecular cloning and characterization of TRP32, a novelthioredoxin-related mammalian protein of 32 kDa. J Biol Chem273: 19160–19166.

Lee YJ, Cho YW, Kim D, et al. 2002. Characterization andregulation of a second gene encoding thioredoxin from thefission yeast. Biochim Biophys Acta 1575: 143–147.

Matsuyama A, Arai R, Yashiroda Y, et al. 2006. ORFeomecloning and global analysis of protein localization in the

fission yeast Schizosaccharomyces pombe. Nat Biotechnol 24:841–847.

Miranda-Vizuete A, Gustafsson J-A, Spyrou G. 1998. Molecularcloning and expression of a cDNA encoding a humanthioredoxin-like protein. Biochem Biophys Res Commun 243:284–288.

Miranda-Vizuete A, Ljung J, Damdimopoulos AE, et al. 2001.Characterization of Sptrx, a novel member of the thioredoxinfamily specifically expressed in human spermatozoa. J BiolChem 276: 31567–31574.

Moreno S, Klar A, Nurse P. 1991. Molecular genetic analysis offission yeast Schizosaccharomyces pombe. Methods Enzymol194: 795–823.

Nakamura H. 2005. Thioredoxin and its related molecules: update2005. Antioxid Redox Signal 7: 823–828.

Niedenthal RK, Riles L, Johnston M, Hegemann JH. 1996.Green fluorescent protein as a marker for gene expression andsubcellular localization in budding yeast. Yeast 12: 773–786.

Pedrajas JR, Kosmidou E, Miranda-Vizuete A, et al. 1999.Identification and functional characterization of a novelmitochondrial thioredoxin system in Saccharomyces cerevisiae.J Biol Chem 274: 6366–6373.

Percival-Smith A, Segall J. 1984. Isolation of DNA sequencespreferentially expressed during sporulation in Saccharomycescerevisiae. Mol Cell Biol 4: 142–150.

Sadek CM, Damdimopoulos AE, Pelto-Huikko M, et al. 2001.Sptrx-2, a fusion protein composed of one thioredoxin and threetandemly repeated NDP-kinase domains, is expressed in humantestis germ cells. Genes Cells 6: 1077–1090.

Sadek CM, Jimenez A, Damdimopoulos AE, et al. 2003. Char-acterization of human thioredoxin-like 2 (Txl-2): a novelmicrotubule-binding thioredoxin predominantly expressed in thecilia of lung airway epithelium and spermatid manchette andaxoneme. J Biol Chem 278: 13133–13142.

Shibata S, Sasaki M, Miki T, et al. 2006. Exportin-5 orthologuesare functionally divergent among species. Nucleic Acids Res 34:4711–4721.

Spyrou G, Enmark E, Miranda-Vizuete A, Gustafsson J-A. 1997.Cloning and expression of a novel mammalian thioredoxin.J Biol Chem 272: 2936–2941.

Thorpe GW, Fong CS, Alic N, Higgins VJ, Dawes IW. 2004.Cells have distinct mechanisms to maintain protection againstdifferent reactive oxygen species: oxidative-stress-responsegenes. Proc Natl Acad Sci USA 101: 6564–6569.

Weeks ME, Sinclair J, Butt A, et al. 2006. A parallel proteomicand metabolomic analysis of the hydrogen peroxide- and Sty1p-dependent stress response in Schizosaccharomyces pombe.Proteomics 6: 2772–2796.

Wollman E, d’Auriol L, Rimsky L, et al. 1988. Cloning andexpression of a cDNA for human thioredoxin. J Biol Chem 263:15506–15512.

Wood V, Gwilliam R, Rajandream MA, et al. 2002. The genomesequence of Schizosaccharomyces pombe. Nature 415: 871–880.


The txl1+ gene from Schizosaccharomyces pombe encodes a new thioredoxin-like 1 protein that...

Documents

Transcript of The txl1+ gene from Schizosaccharomyces pombe encodes a new thioredoxin-like 1 protein that...