Identification, cloning, and functional expression of three glutathione transferase genes from Aspergillus fumigatus

Fungal Genetics and Biology 42 (2005) 319–327 www.elsevier.com/locate/yfgbi 1087-1845/$ - see front matter  2005 Elsevier Inc. All rights reserved. doi:10.1016/j.fgb.2005.01.001 IdentiWcation, cloning, and functional expression of three glutathione transferase genes from Aspergillus fumigatus Claire Burns1, Rachel Geraghty1, Claire Neville, Alan Murphy, Kevin Kavanagh, Sean Doyle¤ National Institute for Cellular Biotechnology, Department of Biology, National University of Ireland, Maynooth, Co. Kildare, Ireland Received 13 October 2004; accepted 3 January 2005 Abstract Analysis of the genome of the human pathogen, Aspergillus fumigatus, revealed the presence of several putative glutathione transferase (GST) open reading frames. Three A. fumigatus GST genes, termed gstA, B, and C, were cloned and recombinant proteins expressed in Escherichia coli. Functional analysis of recombinant gstA–C conWrms that the enzymes exhibit GST activity and glutathione peroxidase activity. RT-PCR conWrmed low basal expression of gstA and gstC which was markedly up-regulated (at least 4£– 10£) in the presence of either H2O2 or 1-chloro-2,4-dinitrobenzene (CDNB). GstB expression was only observed in the presence of CDNB. These results demonstrate for the Wrst time the existence of three functional GSTs in A. fumigatus and strongly suggest a role for these enzymes in the response of the organism to both oxidative stress and xenobiotic presence.  2005 Elsevier Inc. All rights reserved. Keywords: GST; MALDI-TOF; Xenobiotic resistance; Oxidative stress; Fungal 1. Introduction Aspergillus fumigatus is a human pathogenic fungus capable of inducing a range of disease states in patients with pre-existing lung damage or immunosuppression following organ transplantation (Daly and Kavanagh, 2001). Three forms of aspergillosis are recognised clinically: saprophytic, allergic, and invasive, with the latter form having a mortality rate of >90% in some patient groups (Denning, 1998). Conventional therapy relies upon the use of amphotericin B and, more recently, on novel azole derivatives and the echinocandin class of anti-fungal agents, but mortality rates remain high. A. fumigatus displays the ability to withstand attack by macrophages and neutrophils and develop in a potentially hostile environment. Toxin-mediated inhibition of oxidative burst in alveolar macrophages and polymorphonuclear leukocytes by conidia and hyphae is well characterised (Bertout et al., 2002; Mitchell et al., 1997; Muryama et al., 1996). In addition, the physical size of developing hyphae prevent phagocytosis by alveolar macrophages and there is emerging evidence that A. fumigatus may be able to tolerate entry of xenobiotics as a result of amphotericin B treatment creating apertures in the fungal cell membrane (Ellis, 2002). Glutathione transferases (GST; EC 2.5.1.18) are dimeric phase II detoxiWcation enzymes with the ability to conjugate a broad range of potentially harmful xenobiotics to glutathione (GSH), thereby rendering them more susceptible to removal from the cell. GSTs have also been shown to exhibit GSH-dependent peroxidase activity and thus may be involved in resistance to oxidative stress. Cytosolic GSTs have been identiWed in almost all organisms, with mammalian GSTs the most clearly characterised. These enzymes have been implicated in pesticide resistance in plants and insects (Sheehan et al., 2001), and ¤Corresponding author. Fax: +353 1 7083845. E-mail address: sean.doy[email protected] (S. Doyle). 1These authors contributed equally to this work. 320 C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 some GST polymorphisms are thought to alter cancer susceptibility in mammals (Hayes and Pulford, 1995). GSTs are divided into several classes based upon substrate speciWcity, sequence similarity (particularly in the N-terminal region which is involved in GSH binding), immunological cross-reactivity and, where available, structure similarity. GST classes include , , , , , , , and  classes, with insect speciWc ( and ⑀), plant speciWc ( and ) and bacterial () classes also described (Sheehan et al., 2001). In addition, it is likely that many more classes have been already characterised exist in the broad ranging GST category; for example, new protozoan and fungal GST classes have been proposed (Cha et al., 2001; Takada et al., 2004). Until recently, relatively little was known about the presence and role of GST in fungi, however it is now clear that GST isoforms exist in a number of fungal species including Schizosaccharomyces pombe, Aspergillus nidulans, Saccharomyces cerevisiae, Issatchenkia orientalis, Yarrowia lipolytica, Cunninghamella elegans, Mucor circinelloides, and Phanerochaete chrysosporium (Cha et al., 2001; Choi et al., 1998, 2002; Dowd et al., 1997; Dowd and Sheehan, 1999; Foley and Sheehan, 1998; Fraser et al., 2002; Kim et al., 2001; Tamaki et al., 1999; Shin et al., 2002; Veal et al., 2002). Fungal GSTs exhibit diVerential expression patterns, with some isoforms shown to be expressed inducibly in the presence of xenobiotics or oxidative stress. For example, of two GSTs identiWed in I. orientalis, only one was constitutively expressed, and both were induced in the presence of o-dinitrobenzene (o-DNB) (Choi et al., 1998). Three GSTs in S. pombe were induced by oxidative stress, and mutants lacking gst1+ and gst2+ or gst3+ were more sensitive to the presence of the antifungal drug Xuconazole, thereby indicating a role for GST in mediating anti-fungal drug tolerance (Cho et al., 2002; Kim et al., 2001; Shin et al., 2001; Veal et al., 2002). The identiWcation of a functional theta class GST (gene: gstA) in A. nidulans has further elucidated the role of GST in fungal metabolism. GstA appears to be upregulated by the presence of either 1-chloro-2,4-dinitrobenzene (CDNB)2 or H2O2 in the culture medium and may also play a role in mediating heavy metal resistance in A. nidulans (Fraser et al., 2002). Given the signiWcance of A. fumigatus as a human pathogen and the limited success of anti-fungal agents to treat aspergillosis, particularly in immunocompromised patients, it is surprising that the putative presence and role of GST has merited little attention. In addition, the potential role of fungal GST in allowing A. fumigatus to withstand neutrophil attack may represent a key element in the cell’s ability to survive in the host and colonise pulmonary tissue. Here we describe the identiWcation, cloning, heterologous expression and characterisation of three GST genes from A. fumigatus. We also investigate the response of GST gene expression following exposure of A. fumigatus to both CDNB and H2O2. 2. Experimental 2.1. Genomic DNA isolation Aspergillus fumigatus ATCC 26933 (obtained from the American Type Culture Collection, Maryland, USA) was used in this study. Aspergillus cultures were grown in 5%(v/v) fetal calf serum in minimal essential medium Eagle (MEM) (Sigma–Aldrich, Dorset, UK) for 2 days at 37 °C. Genomic DNA was isolated as described by Nicholson et al. (2001). BrieXy, ca. 4 g A. fumigatus mycelia were crushed in liquid N2 and suspended in 10 ml extraction buVer (10mM Tris–HCl, 10 mM EDTA, 0.5% (w/v) SDS pH 8.0). Phenol:chloroform:isoamyl alcohol (25:24:1, 10ml) was added to the mycelial suspension and mixed gently for 30 min. Phases were separated by centrifugation at 5000g at 4°C. The aqueous layer was removed a fresh tube and phenol extraction repeated until the interface was clear. The Wnal aqueous layer was treated with chloroform:isoamyl alcohol (24:1) and phases separated as before. The remaining aqueous layer was treated with ribonuclease A (20 l; 10 mg ml¡1) at 37°C for 30 min, followed by phenol extraction, then chloroform extraction. The DNA was precipitated from the aqueous layer with 2 volumes of 100% ethanol and 1/10 volume of LiCl (4 M) at ¡20 °C overnight. DNA was recovered by centrifugation at 13,000g for 10 min. The pellet was washed with 70%(v/v) ethanol, air-dried, and resuspended in 1 ml TE buVer (10mM Tris–HCl, 1mM EDTA, pH 8.0). 2.2. DNA sequence and bioinformatic analysis All DNA sequence analysis was performed using a Perkin-Elmer ABI Prism 310 genetic analyser, commercially by MWG Biotech (Milton Keynes, UK) or Lark Technologies (Essex, UK) and sequence similarities were determined using the BLAST algorithm (www.ncbi.nlm. nih.gov/blast/bl2seq/bl2.html). Sequence alignments and neighbor-joined phylogenetic trees were generated using ClustalW (Thompson et al., 1994; http://www.ebi.ac.uk/ clustalw). A bootstrapping value of 1000 was used, with bootstrapping percentages noted at tree branch points. Trees were visualised in Treeview (Page, 1996; http:// taxonomy.zoology.gla.ac.uk/rod/rod.html). Preliminary sequence data was also obtained from The Institute for Genomic Research website at http://www.tigr.org. Sequencing of A. fumigatus genome is near completion with support from the Wellcome Trust and NIH. 2Abbreviations used: CDNB, 1-chloro-2,4-dinitrobenzene; DCNB, 1,2-dichloro-4-nitrobenzene; GST, glutathione transferase; GSH, glutathione; MALDI-TOF, matrix assisted laser desorption ionisation-time of Xight; MEM, minimal essential medium Eagle; CALM, calmodulin. C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 321 2.3. PCR ampliWcation All PCR reagents were obtained from Sigma–Aldrich. PCR was performed using AccuTaq polymerase with 1– 10 ng genomic DNA as template and 1.0 M each of forward and reverse primer (Table 1) in a total volume of 50 l. PCR conditions were as follows: 95 °C denaturation for 5 min; (94°C denaturation for 30 s, 55 °C annealing for 90 s, 72 °C extension for 60 s) £35 cycles; 68 °C extension for 7 min. Optimal cDNA ampliWcation was found to require 45 cycles of PCR. PCR-ampliWed DNA was electrophoresed on 1% (w/v) agarose containing 0.5gml ¡1 of ethidium bromide for 30 min at 100 V. Visualisation of amplicons was performed using an ‘Eagle-Eye II’ digital still video system (Stratagene, CA, USA). 2.4. Cloning and expression of gstA, B, and C The gstA sequence was ampliWed from cDNA and the gstB/gstC sequences were ampliWed from A. fumigatus DNA, using primers incorporating terminal EcoRI and PstI sites to facilitate downstream cloning (Table 1). PCR products were cloned into the pCR2.1 cloning vector (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. gstA, gstB, and gstC were subsequently cloned into the pProEX-Hta expression vector (Invitrogen), which facilitates (His)6 aYnity tag introduction, utilising the engineered restriction sites. Ligations were performed using Quickstick ligase (Bioline, London, UK) according to the manufacturer’s instructions. pPXAgstA, pPXAgstB, and pPXAgstC, the resultant expression vectors containing gstA, gstB, and gstC, respectively, were individually transformed into E. coli strain DH5 by electroporation according to Dower et al. (1988). Expression of all three recombinant GST proteins was induced by the addition of 0.6 mM isopropyl -D-thiogalactoside (IPTG) and monitored by SDS–PAGE and Western blot analysis. For enzyme puriWcation, induced cells were lysed by incubation with lysozyme (90 gml ¡1) and sodium deoxycholate (0.04% (w/v)), in the presence of protease inhibitors (1 gml ¡1 leupeptin and pepstatin, respectively, and 1 mM PMSF). Cell debris was removed by centrifugation at 10,000g for 10 min and Nterminal (His)6-tagged recombinant proteins were puriWed from the supernatant by Ni–NTA chromatography (Qiagen, West Sussex, UK) by elution with 250 mM imidazole in 50mM sodium phosphate/300 mM NaCl. PuriWed GST proteins were dialysed (twice; once overnight, and once for 4 h) against phosphate-buVered saline (PBS) containing 0.02% (w/v) sodium azide for storage at 4 °C. Protein concentrations were determined using the Bradford method (Bradford, 1976) with bovine serum albumin as a standard. 2.5. MALDI-TOF MS Mass spectrometry was carried out using an Ettan MALDI-TOF mass spectrometer (Amersham Biosciences (Europe) GmbH, Freiburg, Germany). Protein samples for peptide mass determination were either (i) separated by SDS–PAGE and digested with trypsin or (ii) obtained following in-solution enzymatic digestion and deposited (1 l) with 1 l -cyano-4-hydroxycinnaminic acid (4-HCCA; 5mg/200l 50%(v/v) acetonitrile in aqueous triXuoroacetic acid) onto mass spectrometry slides and allowed to dry prior to delayed extraction, reXectron TOF analysis at 20 kV. 2.6. GST activity assays Glutathione transferase activity was determined using methods based on those described (Habdous et al., 2002; Habig and Jakoby, 1981) whereby the change in absorbance at 340 nm (A340 nm) was recorded, and enzyme activity calculated as micromoles CDNB utilised/mg GST/min. Activity with 1,2dichloro-4-nitrobenzene (DCNB) was performed at 345 nm in the same way as for CDNB, with some exceptions; 100 mM phosphate buVer (pH 7.5), 100 mM DCNB in 100% ethanol, and 50 mM GSH in phosphate Table 1 Nucleotide sequence of oligonucleotide primers used to amplify Afugst genes A–C from A. fumigatus genomic DNA and cDNA, respectively Nucleotide sequence of control calmodulin primers (Romero et al., 2003) are also given. Oligonucleotide primers were designed based on sequence data obtained from the A. fumigatus genome sequencing eVort (http://www.tigr.org). All Afugst forward primers (F1–F3) contained a 5⬘ EcoRI restriction site, and reverse primers contained 3⬘ PstI sites, to facilitate directional cloning into pProEx-Hta. Gene Primers Sequence (5⬘–3⬘) g stA gstA-F GAGAGAATTCATGGCAAATAGACCTGATATTACACTG gstA-R GAGACTGCAGATTAATGCTTCGCCTATTCG g stB gstB-F GAGAGAATTCATGTCTTTGAAGCCTATCGTC gstB-R GAGACTGCAGTTACTTTTCCTGTGCGGC g stC gstC-F GAGAGAATTCATGCCGGACATCCAACCCATC gstC-R GAGACTGCAGTCAGGTCGAGGGGAAGATGTC Calmodulin LCALM CCGAGTACAAGGAAGCTTTCTC Calmodulin RCALM GAATCATCTCGTCGACTTCGTCGTCAGT 322 C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 buVer were used. Activity with ethacrynic acid was recorded at 270 nm, using 100 mM phosphate buVer, pH 6.5, 20 mM ethacrynic acid in 100% ethanol, and 2.5 mM GSH. Glutathione peroxidase activity was determined as described (Veal et al., 2002). BrieXy, test samples (100 l) were mixed with 880 l of assay buVer (50 mM potassium phosphate, pH 7.0, 1 mM EDTA, 1mM NaN 3, 0.2 mM NADPH, 1 U ml¡1 glutathione reductase and 1 mM GSH) and incubated at 30 °C for exactly 5 min, after which the mixture was transferred to a cuvette and 20 l of 69 mM cumene hydroperoxide added. The depletion of NADPH was measured over 3min at 340nm. 2.7. Induction of GST expression in A. fumigatus and analysis by RT-PCR Aspergillus fumigatus ATCC 26933 was cultured, with agitation, in 500 ml MEM + 5% (v/v) fetal calf serum at 37 °C for 47 h, before addition of either CDNB (Wnal concentration: 200 M) or H2O2 (Wnal concentration: 5 mM). Aliquots (50 ml) were removed both prior to induction and at 1, 2, and 3h post-induction. A. fumigatus mycelia were collected by Wltration, rapidly frozen in liquid nitrogen, and stored at ¡80 °C prior to RNA extraction. RNA was extracted from A. fumigatus mycelia using the RNeasy plant mini kit (Qiagen). QuantiWcation of RNA was performed using Total Lab software (NonLinear Dynamics) to ensure equal amounts of RNA were used subsequently, and ca. 1g RNA was used for cDNA synthesis. RNA was treated with DNase I (Sigma–Aldrich) prior to cDNA synthesis to remove DNA contamination. cDNA synthesis from mRNA was performed using the SuperScript kit (Invitrogen) using oligo(dT) primers. Subsequent PCR of GST cDNA was performed as described above. Control PCRs were performed with primers LCALM and RCALM (Table 1) which amplify 348 and 617 bp regions from A. fumigatus cDNA and genomic DNA, respectively (Romero et al., 2003). Densitiometric quantiWcation of PCR products was performed using Genetools software (Syngene). 3. Nomenclature In accordance with recommendations from the A. fumigatus sequencing group (http://www.man.ac.uk), the glutathione transferase genes disclosed here are termed gstA, gstB, and gstC and the corresponding proteins identiWed as gstA, gstB, and gstC. In addition, as recommended, the corresponding S. pombe orthologs are identiWed as superscripts as follows: gstAgst3, gstBgst1, and gstCgst2. Finally, as recommended, the three-letter preWx ‘Afu’, for A. fumigatus gene identiWcation, has only been used when necessary. 4. Results and discussion 4.1. Cloning and sequence analysis of A. fumigatus GST open reading frames Similarity searching of the A. fumigatus genome database with A. nidulans gstA (Genbank Accession No. AAM48104; Fraser et al., 2002) revealed the presence of an A. fumigatus GST (AfugstA) which, following ampliWcation with primers gstA-F/R yielded a PCR product of 909bp when a template of A. fumigatus genomic DNA was used, and 762 bp when cDNA was employed; sequence examination revealed two introns of 95 and 52 bp proximal to the 5⬘ end. The A. fumigatus genome database (http://www.tigr.org) was also interrogated with S. pombe protein sequences corresponding to GSTI (Genbank Accession No. AAK77864; Cho et al., 2002) and GSTII (GenBank Accession No. AAF21054; Kim et al., 2001), and a number of GST-like open reading frames were revealed which exhibited approximately 30% sequence similarity to the query sequences. PCR primers gstB-F/R and gstC-F/R (Table 1) were designed based on these A. fumigatus GST sequences and used to amplify open reading frames of 663 and 675 bp, respectively, from A. fumigatus genomic DNA. GstA, B, and C open reading frames corresponding to protein sequences shown in Fig. 1 were cloned into pProEx-Hta for recombinant expression. Cloned gstA, B, and C sequences were compared with the A. fumigatus genome database; observed diVerences were minimal, resulting in translated amino acid sequences which exhibited 100% sequence identity to the database for gstA and gstB. GstC exhibited a single amino acid diVerence, which was a conservative change from V (http://www.tigr.org) to I (gstC sequence) at position 33. GstC was independently cloned and sequenced several times, indicating a true polymorphism rather than a sequence error. From the amino acid sequence alignment of gstA to C shown in Fig. 1, it is apparent that isolated regions of identity exist between all three proteins throughout their entire sequence, with gstA apparently consisting of a 30 amino acid C-terminal extension relative to gstB and C. However, more Fig. 1. Amino acid sequence alignment of and gstA (254 aa), gstB (221 aa) and gstC (225 aa). Identical residues are highlighted in black boxes while those residues only common to two sequences are shaded grey. Gaps (–) are introduced for optimal alignment. It appears that gstA contains a 30 amino acid C-terminal extension relative to both gstB and C. C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 323 criteria are required to conWdently assign the GSTs to this class, such as immunological cross-reactivity. As relatively few fungal GSTs have been characterised, it is diYcult to determine whether these enzymes are homologues of those found in other organisms or are fungal speciWc. Cha et al. (2001), identiWed a GST from Cunninghamella elegans which bore little resemblance to known classes, and a new fungal class was proposed; it is possible that several classes are yet to be discovered. Fig. 2 shows the relationship between GSTs and many of the fungal GST sequences characterised to date. It can be seen gstB and C are most closely related to the gstA enzymes from A. fumigatus and A. nidulans, respectively. While S. pombe GST1 and 2 also exhibit signiWcant relatedness to the GST proteins from A. fumigatus, those from I. orientalis and Cu. elegans are more distantly related. When the three A. fumigatus GST protein sequences were used in a general BLAST search at http:/ /www.ncbi.nlm.nih.gov/entrez/, similarities to putative GST-like proteins from several fungal species, which have not been fully characterised, were revealed. GstA exhibited the highest similarities, indicating that this GST may have developed prior to speciation. Novel sequences identiWed using gstA included a putative GST from the artichoke pathogen Botryotinia fuckeliana (GenBank Accession No. AAG43132, 69% identity) and putative proteins from the rice blast fungus Magnaporthe grisea (GenBank Accession No. EAA55090, 58% identity), Neurospora crassa (GenBank Accession No. CAD36970, 58% identity), and the cereal pathogen Gibberella zeae (GenBank Accession No. EAA71824, 49% identity). GstB and C also showed similarity to the B. fuckeliana putative GST, at the level of 40% identity. While gstB also exhibits similarity to the N. crassa protein mentioned previously (GenBank Accession No. CAD36970, 44% identity) and gstC shares similarity with the G. zeae protein (39% identity), there is little crossover between similar proteins identiWed by searching with gstA and with the other two A. fumigatus GSTs. GstB and C, however, share similarity with several of the same proteins, although these are less similar than the gstA-like sequences. In addition, both gstB and gstC displayed some similarity with the URE2 group of nitrogen metabolism proteins that have been identiWed from several Saccharomyces species (Fig. 2), but which have not been shown to exhibit GST activity with CDNB (Rai et al., 2003); this similarity was not evident with gstA. URE2 from S. cerevisiae was shown to be involved in defense against heavy metal ions and oxidative stress (Rai et al., 2003). Recently, a putative GST sequence was identiWed from the A. fumigatus genome as part of a coregulated gene cluster, which is postulated as responsible for production of gliotoxin (Gardiner et al., 2004). While this GST has not been characterised and is not the same as any of the GSTs described here, it suggests a metabolic role for GSTs in A. fumigatus, as has been postulated in other organisms (Hayes and Pulford, 1995), and also suggests there may be further, as yet undiscovered, GSTs in this fungus. Expression, puriWcation, and activity analysis of recombinant GST proteins Protein expression plasmids pPXAgstA, pPXAgstB, and pPXAgstC, consisting of the vector pProEx-Hta containing the open reading frames of gstA (cDNA), gstB, and gstC respectively, were transformed into E. coli DH5 and expression induced by the addition of Fig. 2. Phylogenetic analysis of the three A. fumigatus GST proteins compared to 11 other characterised fungal GSTs. Sequences were aligned and a neighbor-joined tree generated using ClustalW, with bootstrapping of 1000. Percentage bootstrapping values are shown at branch points. Sequence GenBank Accession numbers are as follows: Cunninghamella elegans GST2, AAL02369; Cu. elegans GST1, AAL02368; S. p ombe GST2, AAF21054; S. pombe GST1, AAK77864; A. nidulans GSTA, AAM48104; S. cerevisiae URE2, A39609; S. cerevisiae GSTI, P40582; S. pombe GST3, AAK59430; S. cerevisiae GSTII, Q12390; I. orientalis GSTY1, BAA77459; and I. orientalis GSTY2, S16178. Fig. 3. SDS–PAGE and Western blot analysis of recombinant GST expression in E. coli. Duplicate SDS–PAGE gels were loaded as follows: M, protein size marker; lane 1, uninduced E. coli harbouring expression plasmid pPXAgstC; lane 2, E. coli harbouring expression plasmid pPXAgstC induced with 0.6 mM IPTG; lane 3, insoluble proteins extracted from induced E. coli; lane 4, soluble proteins extracted from induced E. coli; lane 5, (His)6-puriWed recombinant gstC (5 g); lane 6, (His)6-puriWed recombinant gstB (5 g); and lane 7, (His)6-puriWed recombinant gstA (4 g). One gel was Coomassie stained (A) and the other was probed with anti-(His)6 murine monoclonal antibody to identify recombinant proteins (B). GstB was puriWed in the same manner as gstC. Both gstB and gstC were detected with anti-(His)6 monoclonal antibody, but not gstA. However, the identity of this protein was conWrmed by mass spectrometry. 324 C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 0.6mM IPTG (Fig. 3). All three recombinant proteins were present in cell lysate supernatants, indicating solubility, and recombinant gstB and gstC were puriWed using the N-terminal (His)6-tag, with a yield of approximately 17 and 18mg per gram of E. coli cells cultured, respectively (Fig. 3). Detection of recombinant gstA using the (His)6 tag was not possible, and large-scale metal chelate aYnity chromatography was required to purify suYcient enzyme for activity analysis (Fig. 3; Table 2). PuriWed recombinant proteins were analysed by MALDI-TOF MS and peptides (following tryptic digestion) were identiWed corresponding to the theoretical amino acid sequence for all three proteins whereby 4/ 83 peptides (20% sequence coverage), 7/57 peptides (34% sequence coverage), and 4/46 peptides (15% sequence coverage) were observed for recombinant gstA, B, and C, respectively. SDS–PAGE data conWrm molecular masses of 26, 27, and 30kDa for gstA, B, and C, respectively (Fig. 3). These are consistent with theoretical molecular masses of 28.99 and 28.72 kDa for the (His)6tagged gstB and gstC proteins, but smaller than the theoretical mass of 32.76 kDa for gstA. Protein molecular mass analysis via FPLC gel Wltration chromatography (Superose 6) conWrmed the dimeric status of puriWed gstB (56kDa) and C (61 kDa), respectively (data not shown). No binding to glutathione–Sepharose aYnity columns was observed for any recombinant GST (data not shown). Furthermore, attempts to purify native gstA to C from A. fumigatus extracts by GSH aYnity chromatography were unsuccessful (data not shown), which has also been noted in the puriWcation of some theta GST enzymes (Hayes and Pulford, 1995). Crystal structures of some theta GSTs have indicated that the GSH-binding site may be sited further inside the protein than in other classes of GST, and that conventional GSH-aYnity matrices may be unable to bind the active site (Hayes and Pulford, 1995). The enzymatic activities of puriWed, recombinant gstA–C were assessed with several substrates. CDNB, DCNB, and ethacrynic acid are substrates for glutathione transferase activity, whereas cumene hydroperoxide is a substrate for glutathione peroxidase activity (Veal et al., 2002). GstA–C exhibited low level GST and glutathione peroxidase activities, which were detectable due to the high concentrations of recombinant enzyme obtained (Table 2). GstB exhibited a four to six times higher speciWc activity against CDNB than either gstA or C, respectively. No activity was observed for either enzyme when DCNB or ethacrynic acid were used as substrates. The speciWc activity observed against cumene hydroperoxide was almost six times higher for gstB then gstA/C (Table 2). Although relatively lower, the actual ratio of relative activity against CDNB was similar to that previously found (Veal et al., 2002) whereby S. pombe gst1 showed three times greater activity towards CDNB than gst2. Conversely, the glutathione peroxidase activity of S. pombe gst2 was about twice that observed for gst1 albeit at a six times (approx.) lower level to that found for A. fumigatus gstB. GST speciWc activity determinations for S. pombe gst1 (Kim et al., 2001) were carried out on cell lysates using impure gst1, thus making exact comparison diYcult. However, gst1 activity was detectable against CDNB with no glutathione peroxidase activity evident. S. pombe gst2 also possesses glutathione transferase activity, as measured by CDNB conjugation (Cho et al., 2002). The speciWc activities of the A. fumigatus GSTs are low in comparison with other characterised GSTs, however, any loss of activity due to the presence of the (His)6 tag is unlikely as TEV protease removal of the (His)6 tag (from gstC) did not result in enhanced activity (data not shown). Ultimately, activity analysis of the native GSTs is required to fully assess activity. 4.2. Induction of A. fumigatus GST expression with CDNB and H2O2 GSTs are thought to be involved in the response against external and cellular toxins, and may also aid the cell when challenged by oxidative stress. After initial growth in the absence of either CDNB or H2O2, A. fumigatus was further cultured in the presence of 200 M CDNB or 5 mM H2O2 and expression levels of gstA, gstB, and gstC determined by RT-PCR (Fig. 4). Moreover, the diVerence in amplicon size for gstA can be seen thereby conWrming the removal of intronic sequence. Primers speciWc for the calmodulin gene (Romero et al., 2003) were used to conWrm absence of genomic DNA and RNA equivalence between time-points. Low basal expression of gstA and gstC was detected, but gstB expression was not detectable under basal conditions. Upon induction with CDNB, all three gst genes were upregulated within one hour, with gstA showing a 10-fold induction (approximately) and gstC showing at least a 4fold induction at all time-points, and gstB exhibiting only weak induction, with a 20% drop in RNA expression at Table 2 SpeciWc activities of puriWed recombinant gstA, gstB, and gstC with both glutathione transferase and glutathione peroxidase substrates GST activity was assayed with CDNB, DCNB and ethacrynic acid, and glutathione peroxidase activity was assayed with cumene hydroperoxide. Units are expressed as micromoles substrate utilised per minute. SpeciWc activities stated are the average of Wve replicates and were corrected against blank reactions that had been performed in triplicate. N.D., no activity detected. Substrate SpeciWc activity (U mg¡1) (§SD) gstA gstB gstC CDNB 0.004 §0.0001 0.025 §0.0028 0.006 §0.0005 DCNB N.D. N.D. N.D Ethacrynic acid N.D. N.D. N.D Cumene hydroperoxide 0.019 §0.0009 0.145 §0.0088 0.025 §0.006 C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 325 3 h post induction (Fig. 4A). This result is quite interesting as it is somewhat at variance with the above observation that gstB exhibits greater activity towards CDNB than does gstA or gstC, however it is possible that the higher speciWc activity of gstB may necessitate production of smaller amounts of actual protein to conjugate available CDNB. Previous work, whereby -galactosidase expression was placed under the control of the upstream regulatory regions associated with gst1 and gst2 of S. pombe, has shown that gst1+ gene expression was enhanced by mercuric chloride and menadione (generates superoxide radicals), whereas gst2 expression was only signiWcantly induced by the presence of o-dinitrobenzene (o-DNB) (Cho et al., 2002; Kim et al., 2001; Shin et al., 2002). o-DNB also induces GST gene expression in I. orientalis (Tamaki et al., 1999). In the work presented here, exposure of cells to H2O2 did not result in gstB induction. However, weak induction of gstC was detected within 1 h post-induction (approx. 3-fold increase over T0) which stabilised at 2 and 3 h (approx. 7-fold increase over T0). In addition, strong expression of gstA at 1 h post-induction was evident (at least 5-fold increase) which appeared to further increase at 2 h (approx. 10-fold increase over T0) before reducing at 3 h post-induction (approx. 4.5-fold with respect to T0), possibly due to depletion of added H2O2 (Fig. 4B). GstB and gstC were shown to be diVerentially inducible by H2O2, whereby up-regulation of gstC expression only was observed. This indicates the possibility of diVerent roles for these proteins within the organism when subjected to environmental stress. Previous work has also indicated that expression of gst1, as well as gst2, in S. pombe is induced by H2O2. This observation suggests potentially diVerent mechanisms are involved in the response to oxidative stress for both organisms and that the A. fumigatus gstB plays a diVerent role than gst1 in S. pombe—an hypothesis supported by greater sequence divergence between gstB and C in A. fumigatus compared to the corresponding genes in S. pombe (Fig. 2). It is intriguing that gstB expression was not induced in the presence of H2O2, yet it exhibited the greatest glutathione peroxidase activity against cumene hydroperoxide. In vivo, it is postulated that GST enzymes may be involved in detoxiWcation of secondary oxidation products produced by initial conversion of reactive oxygen species and H2O2 by enzymes such as glutathione peroxidase and superoxide dismutase (Hayes and McLellan, 1999). These secondary oxidation products include lipid and DNA oxidation products, similar to cumene hydroperoxide, and it is possible that gstB expression is induced by the secondary oxidation products rather than directly by H2O2. Thus, a putative secondary compound required for induction of gstB may not have been produced when H2O2 was used as the inducer. A. fumigatus gstA was identiWed via homology searching of the A. fumigatus genome using A. nidulans gstA. In fact, these proteins exhibit a very high degree of sequence similarity (83%) which indicates that they may share a common function in both Aspergillus spp. Fraser et al. (2002) have shown that gstA encodes a GST which is involved in xenobiotic and metal ion resistance in A. nidulans. A. fumigatus gstA was strongly up-regulated in the presence of both CDNB and H2O2. In addition, gstA harbours sequences upstream of the initial ATG with similarity to the xenobiotic (XRE) and antioxidant responsive elements (ARE) involved in regulation of mammalian GSTs (Hayes and Pulford, 1995; Rushmore et al., 1991; Rushmore and Pickett, 1993). Although such responsive elements have not been demonstrated to be functional in the regulation of fungal GSTs to date, XRE and ARE have been identiWed in the promoter regions of gst genes in S. cerevisiae (Choi et al., 1998). DNA sequences from the A. fumigatus genome (http:// www.tigr.org) 2 kb upstream of each gst gene were examined for the consensus XRE (TNGCGTG) and ARE (TGACNNNGC) regions. Several XRE and ARE-like regions were observed upstream of all three gst genes. Examination of transcriptional regulation via promoter analysis linked to a reporter gene would be necessary to conWrm whether these responsive elements were functional and involved in regulation of gene expression in the presence of CDNB and H2O2. The expression of all three gst genes in response to organism exposure to CDNB and of gstA and C following exposure of Fig. 4. RT-PCR of A. fumigatus cDNA isolated from cultures induced with either CDNB (A) or H2O2(B). A. fumigatus RNA was isolated prior to induction (T0), 1 h post-induction (T1), 2 h post-induction (T2), and 3 h post-induction (T3). PCR was performed on cDNA using gstA, B, and C speciWc primers and primers speciWc for the calmodulin gene (CALM) (Romero et al., 2003). Optimal cDNA ampliWcation was found to require 45 cycles of PCR. Agarose gels were loaded as follows: M, DNA size marker; T0; T1; T2; and T3; gDNA, genomic DNA control; and H2O, no DNA control. Expression of all three gst genes was observed when induced with CDNB, and induction of gstA and g stC only was seen with H2O2 induction. 326 C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 A. fumigatus to H2O2 indicates a high likelihood that GSTs play a role in the response of A. fumigatus to both xenobiotic presence and oxidative stress and, as in other fungal species (Veal et al., 2002), may be involved in mediating anti-fungal drug resistance. 5. Summary Three previously unidentiWed open reading frames encoding glutathione transferases have been identiWed in A. fumigatus, which are either constitutively expressed under experimental conditions employed (gstA and C) or inducible in response to xenobiotic (gstA, B, and C) or oxidative stress (gstA and C). Recombinant proteins corresponding to all three genes have been expressed and enzymatic activity characteristic of glutathione transferase and glutathione peroxidase deWned. The availability of this information will facilitate further exploration of the possible role(s) played by GST in xenobiotic tolerance and mediating response to oxidative stress in A. fumigatus. 6. Accession numbers Sequences for the three Aspergillus fumigatus glutathione transferases described are deposited in GenBank under Accession Nos. AY770045, AY770043, and AY770044 (gstA to C) and AY770046 (gstA cDNA). Acknowledgments This work was funded under the IRCSET/Enterprise Ireland Basic Research Grant Programme and the Irish Government Programme for Research in Third Level Institutions. Rachel Geraghty was funded by Dublin City Council studentship scheme, Claire Neville was a recipient of a Daniel O’Connell Fellowship from NUI Maynooth. Preliminary sequence data was obtained from The Institute for Genomic Research website at http://www.tigr.org. Sequencing of A. fumigatus was funded by the National Institute of Allergy and Infectious Disease U01 AI 48830 to David Denning and William Nierman, the Wellcome Trust, and Fondo de Investicagiones Sanitarias. References Bertout, S., Badoc, C., Mallie, M., Giamimis, J., Bastide, J., 2002. Spore diVusate isolated from some strains of Aspergillus fumigatus inhibits phagocytosis by murine alveolar macrophages. FEMS Immunol. Med. Microbiol. 33, 101–106. Bradford, M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Cha, C.J., Coles, B.F., Cerniglia, C.E., 2001. PuriWcation and characterization of a glutathione S-transferase from the fungus Cunninghamella elegans. FEMS Microbiol. Lett. 203 (2), 257–261. Cho, Y.W., Park, E.H., Fuchs, J.A., Lim, C.J., 2002. A second stressinducible glutathione S-transferase gene from Schizosaccharomyces pombe. Biochim. Biophys. Acta Gene Struct. Expr. 1574 (3), 399– 402. Choi, J.H., Lou, W., Vancura, A., 1998. A novel membrane-bound glutathione S-transferase functions in the stationary phase of the yeast Saccharomyces cerevisiae. J. Biol. Chem. 273 (45), 29915– 29922. Daly, P., Kavanagh, K., 2001. Pulmonary Aspergillosis: clinical presentation, diagnosis and therapy. Br. J. Biomed. Sci. 58, 197–205. Denning, D.W., 1998. Invasive aspergillosis. Clin. Inf. Dis. 26, 781–805. Dowd, C.A., Buckley, C.M., Sheehan, D., 1997. Glutathione S-transferases from the white-rot fungus, Phanerochaete chrysosporium. Biochem. J. 324, 243–248. Dowd, C.A., Sheehan, D., 1999. Variable expression of glutathione Stransferase isoenzymes in the fungus, Mucor circinelloides. FEMS Microbiol. Lett. 170 (1), 13–17. Dower, W.J., Miller, J.F., Ragsdale, C.W., 1988. High-eYciency transformation of Escherichia coli by high-voltage electroporation. Nucleic Acids Res. 16 (13), 6127–6145. Ellis, D., 2002. Amphotericin B: spectrum and resistance. J. Antimicrob. Chemother. 49, 7–10. Foley, V., Sheehan, D., 1998. Glutathione S-transferases of the yeast Yarrowia lipolytica have unusually large molecular mass. Biochem. J. 333, 839–845. Fraser, J.A., Davis, M.A., Hynes, M.J., 2002. A gene from Aspergillus nidulans with similarity to URE2 of Saccharomyces cerevisiae encodes a glutathione S-transferase which contributes to heavy metal and xenobiotic resistance. Appl. Environ. Microbiol. 68 (6), 2802–2808. Gardiner, D.M., Cozijnsen, A.J., Wilson, L.M., Pedras, M.S.C., Howlett, B.J., 2004. The sirodesmin biosynthetic gene cluster of the plant pathogenic fungus Leptosphaeria maculans. Mol. Microbiol. 53 (5), 1307–1318. Habdous, M., Vincent-Viry, M., Visvikis, S., Siest, G., 2002. Rapid spectrophotometric method for serum glutathione S-transferases activity. Clin. Chim. Acta 326 (1-2), 131–142. Habig, W.H., Jakoby, W.B., 1981. Assays for diVerentiation of glutathione-S-transferases. Methods Enzymol. 77, 398–405. Hayes, J.D., McLellan, L.I., 1999. Glutathione and glutathione-dependent enzymes represent a coordinately regulated defence against oxidative stress. Free Radic. Res. 31 (4), 273–300. Hayes, J.D., Pulford, D.J., 1995. The glutathione S-transferase supergene family: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance. Crit. Rev. Biochem. Mol. Biol. 30 (6), 445–600. Kim, H.G., Park, K.N., Cho, Y.W., Park, E.H., Fuchs, J.A., Lim, C.J., 2001. Characterization and regulation of glutathione S-transferase gene from Schizosaccharomyces pombe. Biochim. Biophys. Acta Gene Struct. Expr. 1520 (2), 179–185. Mitchell, C.G., Slight, J., Donaldson, K., 1997. DiVusible component from the spore surface of the fungus Aspergillus fumigatus which inhibits the macrophage oxidative burst is distinct from gliotoxin and other hyphal toxins. Thorax 52, 796–801. Muryama, T., Amitani, R., Ikegami, Y., Awada, R., Lee, W.J., Kuze, F., 1996. Suppressive eVects of Aspergillus fumigatus culture Wltrates on human alveolar macrophages and polymorphonuclear leucocytes. Eur. Respir. J. 9, 293–300. Nicholson, T.P., Rudd, B.A.M., Dawson, M., Lazarus, C.M., Simpson, T.J., Cox, R.J., 2001. Design and utility of oligonucleotide gene probes for fungal polyketide synthases. Chem. Biol. 8, 157–178. Page, R.D.M., 1996. TREEVIEW: an application to display phylogenetic trees on personal computers. Comput. Appl. Biosci. 12, 357– 358. C. Burns et al. / Fungal Genetics and Biology 42 (2005) 319–327 327 Rai, R., Tate, J.J., Cooper, T.G., 2003. Ure2, a prion precursor with homology to glutathione S-transferase, protects Saccharomyces cerevisiae cells from heavy metal ion and oxidant toxicity. J. Biol. Chem. 278, 12826–12833. Romero, B., Turner, G., Olivas, I., Laborda, F., De Lucas, J.R., 2003. The Aspergillus nidulans alcA promoter drives tightly regulated conditional gene expression in Aspergillus fumigatus permitting validation of essential genes in this human pathogen. Fungal Genet. Biol. 40 (2), 103–114. Rushmore, T.H., Morton, M.R., Pickett, C.B., 1991. The antioxidant responsive element-activation by oxidative stress and identiWcation of the DNA consensus sequence required for functional activity. J. Biol. Chem. 266 (18), 11632–11639. Rushmore, T.H., Pickett, C.B., 1993. Glutathione S-transferases, structure, regulation, and therapeutic implications. J. Biol. Chem. 268 (16), 11475–11478. Sheehan, D., Meade, G., Foley, V.M., Dowd, C.A., 2001. Structure, function and evolution of glutathione transferases: implications for classiWcation of non-mammalian members of an ancient enzyme superfamily. Biochem. J. 360, 1–16. Shin, Y.H., Park, E.H., Fuchs, J.A., Lim, C.J., 2002. Characterization, expression and regulation of a third gene encoding glutathione S-transferase from the Wssion yeast. Biochim. Biophys. Acta Gene Struct. Expr. 1577 (1), 164–170. Takada, Y., Uda, K., Kawamura, K., Matsuoka, T., 2004. Molecular cloning and characterization of a novel glutathione S-transferase gene induced by light stimulation in the protozoan Blepharisma japonicum. FEMS Microbiol. Lett. 231 (2), 185–189. Tamaki, H., Yamamoto, K., Kumagai, H., 1999. Expression of two glutathione S-transferase genes in the yeast Issatchenkia orientalis is induced by o-dinitrobenzene during cell growth arrest. J. Bacteriol. 181 (9), 2958–2962. Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-speciWc gap penalties and weight matrix choice. Nucleic Acids Res. 22 (22), 4673–4680. Veal, E.A., Toone, W.M., Jones, N., Morgan, B.A., 2002. Distinct roles for glutathione S-transferases in the oxidative stress response in Schizosaccharomyces pombe. J. Biol. Chem. 277 (38), 35523–35531.