J Biol Chem, Vol. 274, Issue 43, 30451-30458, October 22, 1999

Cloning and Characterization of AOEB166, a Novel Mammalian Antioxidant Enzyme of the Peroxiredoxin Family^*

Bernard Knoops^§¶, André Clippe^§, Cédric Bogard, Karim Arsalane^§, Ruddy Wattiez^**, Cédric Hermans^§, Elee Duconseille, Paul Falmagne, and Alfred Bernard^§§§

From the  Laboratory of Cell Biology, Department of Biology, Université Catholique de Louvain, 1348 Louvain-la-Neuve, Belgium, the ^§ Unit of Industrial Toxicology and Occupational Medicine, School of Medecine, Université Catholique de Louvain, 1200 Brussels, Belgium, and the  Department of Biological Chemistry, Université de Mons-Hainaut, B-7000 Mons, Belgium

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Using two-dimensional electrophoresis, we have recently identified in human bronchoalveolar lavage fluid a novel protein, termed B166, with a molecular mass of 17 kDa. Here, we report the cloning of human and rat cDNAs encoding B166, which has been renamed AOEB166 for antioxidant enzyme B166. Indeed, the deduced amino acid sequence reveals that AOEB166 represents a new mammalian subfamily of AhpC/TSA peroxiredoxin antioxidant enzymes. Human AOEB166 shares 63% similarity with Escherichia coli AhpC22 alkyl hydroperoxide reductase and 66% similarity with a recently identified Saccharomyces cerevisiae alkyl hydroperoxide reductase/thioredoxin peroxidase. Moreover, recombinant AOEB166 expressed in E. coli exhibits a peroxidase activity, and an antioxidant activity comparable with that of catalase was demonstrated with the glutamine synthetase protection assay against dithiothreitol/Fe3+/O₂ oxidation. The analysis of AOEB166 mRNA distribution in 30 different human tissues and in 10 cell lines shows that the gene is widely expressed in the body. Of interest, the analysis of N- and C-terminal domains of both human and rat AOEB166 reveals amino acid sequences presenting features of mitochondrial and peroxisomal targeting sequences. Furthermore, human AOEB166 expressed as a fusion protein with GFP in HepG2 cell line is sorted to these organelles. Finally, acute inflammation induced in rat lung by lipopolysaccharide is associated with an increase of AOEB166 mRNA levels in lung, suggesting a protective role for AOEB166 in oxidative and inflammatory processes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In cells and organisms that have evolved to live in an atmosphere rich in oxygen, the incomplete reduction of oxygen generates potent oxidizing agents (1). These include reactive oxygen species (ROS)¹ and their toxic by-products, which may react with various cellular components such as lipids, proteins, and nucleic acids, leading to cell damage and possibly cell death (2, 3). In eukaryotes, two major intracellular sources of ROS are the mitochondrion, where electron transport coupled to oxidative phosphorylation takes place (4), and the peroxisome in which high amounts of hydrogen peroxide or superoxide anions are generated during -oxidation of fatty acids and by the activity of various oxidases (5, 6). Moreover, other oxidative pathways in different subcellular compartments may account for ROS production, and ROS may also be generated extracellularly in the course of inflammatory processes (7).

Mammalian cells have developed complex mechanisms to protect themselves against oxidative attacks but also to maintain a redox balance in their different subcellular compartments (1). These antioxidant defense systems include nonenzymatic antioxidants (vitamin E, vitamin C, vitamin A, and uric acid), enzymes with antioxidant properties (catalase, superoxide dismutase, and glutathione peroxidase) as well as low molecular weight reducing agents (glutathione and thioredoxin). Recently, a new family of antioxidant enzymes, the AhpC/TSA peroxiredoxin family, has been discovered in prokaryotes and eukaryotes (8). These enzymes exhibit hydrogen peroxide and alkyl hydroperoxide reductase activities (9-12). Peroxiredoxins are considered to be involved in oxidative stress protection mechanisms but also in cell differentiation (13, 14), proliferation (14, 15), immune response (16), and apoptosis (17, 18).

Here, we report the cloning and initial characterization of AOEB166, a novel member of the mammalian peroxiredoxin family with mitochondrial and peroxisomal sorting signals. Although this new peroxiredoxin was first identified in human bronchoalveolar lavage fluid, AOEB166 presents the features of a highly conserved and widely expressed protein that might play an important antioxidant protective role in various tissues under nonpathological conditions but also during inflammatory processes.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

cDNA Cloning and Sequence Analysis-- A reverse cloning approach was used based on peptide microsequencing informations for cloning human AOEB166 cDNA. First strand cDNA was obtained with Moloney leukemia virus reverse transcriptase (Superscript II, Life Technologies, Inc.) from 2 µg of human lung RNA using oligo(dT)₂₅ as primer according to the manufacturer's instructions. A partial cDNA fragment was PCR-amplified by GoldStar DNA polymerase (Eurogentec) during 40 cycles of PCR with 5'-ATCAAGGTGGGNGAYGC-3' and 3'-TTYCCNTTCTTCCCNCA-5' degenerate primers. PCR cycles were as follows: denaturing for 30 s at 94 °C, annealing for 45 s at 50 °C, extension for 1 min at 72 °C, and, after 40 cycles, a final extension step for 5 min at 72 °C. The PCR product of 98-bp was gel purified with Qiaquick (Qiagen) and cloned into the pCR2.1 vector with the TA cloning kit (Invitrogen). Clones of the PCR product were sequenced on ABI 377 automatic DNA sequencer (Perkin-Elmer). Subsequently, the presumed full-length human AOEB166 cDNA was identified by the rapid amplification of cDNA ends (RACE) with Expand High Fidelity DNA polymerase (Roche Molecular Biochemicals) using a Marathon-ready human lung cDNA kit (CLONTECH) and AOEB166-specific primers 3'-CCACTTGGACCGTCTCGACAAGTT-5' and 5'-GCCATCCCAGCAGTGGAGGTGTTTG-3' for 5'-amplification and 3'-amplification, respectively. PCR products were cloned and sequenced as described above. Based on the new sequence, a 738-bp sequence of AOEB166 mRNA was amplified using 5'-GGGTATGGGACTAGCTGGCG-3' and 3'-GACGTTAACCTTACAACCGGTC-5' primers. The corresponding sequence of the PCR product including the coding sequence was proven to be identical to RACE product sequences. Human AOEB166 cDNA sequence was used to perform a BLAST search at the NCBI web site. Homologies with mouse and rat AOEB166 expressed sequence tag clones were found and used to design primers to amplify the presumed full-length rat AOEB166 cDNA with a Marathon-ready rat liver cDNA kit (CLONTECH). Gene-specific primers 3'-CTTCCGTTCCAAGCCGAGGACCGACT-5' and 5'-GCATTTACACCTGGCTGTTCCAAGACC-3' were used in RACE for 5'-amplification and 3'-amplification, respectively. PCR products were cloned and sequenced as described above. Multiple alignments of deduced protein sequences and the phylogenic analysis, according to the neighbor joining method (19), were performed with CLUSTAL W (version 1.7) (20).

Chromosomal Assignment-- AOEB166 gene localization on human chromosomes was performed by PCR using the Genebridge 4 radiation hybrid panel according to the protocol of Research Genetics with two different pairs of primers specific for AOEB166 human gene: 5'-ATGTTATGCAACCCTTTGCGACAC-3' and 3'-CTCGGTCCCTTGTTCCACTTGGAC-5' primers, and 5'-GTGTTTGAAGGGGAGCCAGGGAAC-3' and 3'-GGTTCTACCACTTTGGGACAGAGA-5' primers.

Northern and Dot Blotting-- Human multiple tissue Northern blot (CLONTECH) and human master dot blot (CLONTECH) were hybridized according to the manufacturer's instructions with a 738-bp ³²P-labeled AOEB166 cDNA fragment (Rediprime, Amersham Pharmacia Biotech) amplified by PCR as described before. After stripping, the membranes were reprobed with ³²P-labeled -actin probe (CLONTECH) for the multiple tissue Northern blot. Northern blotting of RNA isolated from human cell lines was performed as described previously (21), and ³²P-labeled AOEB166 and -actin probes were hybridized with ExpressHyb hybridization buffer (CLONTECH).

Transfection with Human AOEB166 in Fusion with Green Fluorescent Protein and Immunostaining-- The pcDNA3.1/NT-GFP-TOPO and pcDNA3.1/CT-GFP-TOPO vectors (Invitrogen) were used to generate a GFP fusion product under control of the cytomegalovirus promoter. The construct contained both the complete sequence of hAOEB166 (including the predicted mitochondrial presequence) at the N terminus and the hAOEB166 peroxisomal targeting sequence SQL at the C terminus of GFP (see Fig. 4A). First, two 22-mers (5'-AGCCAATTGTGATAGAGATCTA-3' and 5'-AGATCTCTATCACAATTGGCTA-3') were allowed to hybridize to generate a DNA insert coding for SQL (containing also two stop codons and a BglII site) and ligated into pcDNA3.1/NT-GFP-TOPO vector. Second, hAOEB166 cDNA fragment was amplified by PCR with 5'-GGGTATGGGACTAGCTGGCG-3' and 3'-TGGGTTATAGTAGAGTGTCG-5' primers and ligated into pcDNA3.1/CT-GFP-TOPO vector. Escherichia coli TOP10 (Invitrogen) were transformed with the plasmids. Third, the 1280-bp BstB1-AvrII restriction product of pcDNA3.1/NT-GFP-TOPO containing the 3'-coding sequence of the GFP fused to the insert coding for SQL was ligated into the 5653-bp BstB1-AvrII restriction product of pcDNA3.1/CT-GFP-TOPO containing the insert coding for hAOEB166 and the 5'-coding sequence of GFP. The construct in the resulting plasmid was sequenced. HepG2 human hepatoblastoma cell line cultured on coverslips was transiently transfected with the plasmid encoding the GFP fusion protein by using LipofectAMINE Plus reagent (Life Technologies, Inc.). The fusion protein was expressed at detectable levels 48 h after transfection. For immunostaining, transfected cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min, rinsed three times with Tris-buffered solution (0.05 M, pH 7.6) containing 0.9% NaCl and 0.1% Triton X-100 (TBS-T), and immersed 30 min in TBS-T containing 10% nonfat milk. Coverslips were incubated sequentially overnight in TBS-T containing 1% nonfat milk and 1:1000 rabbit anti-bovine catalase antiserum (Rockland) or mouse monoclonal anti-human cytochrome c oxidase subunit I (5 µg/ml, Molecular Probes), and 1 h with TRITC-conjugated swine anti-rabbit IgG (Dako) or TRITC-conjugated rabbit anti-mouse IgG (Dako) diluted 1/20 in TBS-T. Cells were washed twice 10 min with TBS-T after each incubation. The coverslips were mounted in Mowiol with anti-fading (1,4-diazabicyclo[2.2.2]octane, 25 mg/ml, Sigma) and examined by fluorescence microscopy with standard fluorescein isothiocyanate filters for GFP and TRITC filters.

Bacterial Expression of Human AOEB166 and Purification-- IMPACT kit (New England Biolabs) was used to express and purify E. coli recombinant AOEB166 without its predicted mitochondrial presequence. The human AOEB166 cDNA was PCR-amplified using the following primers: 5'-CTGCACATATGGCCCCAATCAAGGTG-3' (NdeI site underlined) and 3'-GTTATAGTAGAGTGTCGAGACGCCTTCTCGCGGTCT-5' (SapI site underlined). The PCR product was digested with NdeI and SapI and ligated into vector pTYB1. The insert was sequenced, and the fusion of AOEB166 with the intein tag at its C-terminal end was confirmed. The resulting plasmid was used to transform ER2566 E. coli. Expression of AOEB166-intein was induced with 1 mM isopropyl--D-thiogalactopyranoside. The recombinant protein was purified through New England Biolabs affinity chitin column, and intein was cleaved at 4 °C in presence of DTT according to the manufacturer's instructions. The eluted AOEB166 protein was dialyzed against PBS and analyzed by SDS-polyacrylamide gel electrophoresis (see Fig. 5A), and identity was confirmed by N-terminal microsequencing.

Antioxidant Assay and Peroxidase Assay-- The protection of glutamine synthetase by recombinant human AOEB166 against dithiothreitol/Fe3+/O₂ oxidation was performed essentially as described previously (Ref. 22; see also Fig. 5). Peroxidase activity of recombinant protein was measured as described by Kang et al. (11) with minor modifications (see Fig. 6).

Cell Cultures-- Human cell lines (see Fig. 3) were maintained in Dulbecco's modified Eagle's medium containing penicillin (100 units/ml) and streptomycin (100 units/ml) and supplemented with 10% fetal calf serum. The cells were grown to 90% confluence, and about 2 × 10⁷ cells were washed twice with PBS before RNA isolation with Trizol reagent (Life Technologies, Inc.) according to the manufacturer's instructions.

In Vivo Experiments-- Adult male Harlan Sprague-Dawley rats weighing approximately 250-300 g were used. Animals were anesthetized by intraperitoneal injection of sodium pentobarbital (50 mg/kg), and lung inflammation was induced by intratracheal instillation of LPS (E. coli 055:B5, Sigma) at the dose of 100 µg/100 g of body weight. LPS was dissolved in 200 µl of sterile PBS. Control rats were instillated with 200 µl of PBS. Rat lungs were collected at indicated times, and total RNA was isolated with Trizol reagent.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cloning of Human and Rat AOEB166 cDNAs-- Human AOEB166 was initially identified during the two-dimensional electrophoresis mapping of proteins from pooled human bronchoalveolar lavage fluid. This unknown protein was shown to have a molecular mass of 17 kDa and a pI of 6.9 (23). Thus, a partial amino acid sequence of 37 residues at the N terminus (Fig. 1A) was used to design degenerate oligonucleotide primers which allowed the PCR amplification of a human lung cDNA fragment. Based on the sequence of this amplicon, 5'- and 3'-RACE was performed, and the longest PCR products were cloned and sequenced. The composite cDNA sequence contained a poly(A) tract at the 3'-end, 36 bases of a 5'-leader sequence, an open reading frame of 645 bases, and a 116-base-long 3'-trailer sequence containing a AATAAA polyadenylation signal (Fig. 1A). Two Kozak consensus sequences for translation initiation (24) were found in the same reading frame. The open reading frame (ORF) of the longest sequence encoded a polypeptide of 214 residues (GenBank^TM accession number AF110731). The screening of 5'- and 3'-amplicons obtained by RACE on rat liver cDNA led to the identification of a composite cDNA sequence containing a poly(A) tract at the 3'-end, 39 bases of a 5'-leader sequence, an ORF of 642 bases, and a 135-base-long 3'-trailer sequence containing a AATAAA polyadenylation signal. The ORF encoded a polypeptide of 213 residues (GenBank^TM accession number AF110732).

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Nucleotide and deduced amino acid sequences of human AOEB166 (A), alignment of human and rat AOEB166 amino acid sequences (B), and alignment of human AOEB166 and S. cerevisiae YLR109W proteins (C). A, nucleotides are numbered (left margin) beginning with the first base of ATG initiator codon for AOEB166, and the nucleotides on the 5' sides of residue 1 are indicated by negative numbers. Deduced amino acid residues are numbered (right margin, in parentheses). Underlined amino acid residues were determined by amino acid sequencing of AOEB166 spot in two-dimensional electrophoresis of bronchoalveolar lavage fluid (23). The consensus polyadenylation signal (AATAAA) is in bold type and underlined. The stop codon is indicated with asterisks. The predicted mitochondrial presequence appears in bold type. B, Clustal W (version 1.7) alignment of AOEB166 from human and rat shows highly conserved amino acid sequences. Identical residues (asterisks) and conservative substitution (dots) are indicated. The consensus SQL putative peroxisomal targeting sequence conserved in both species appears in bold italic type. The three conserved cysteines are in bold type. The human and rat nucleotide and amino acid sequences have been submitted to GenBankTM under accession numbers AF110731 (human AOEB166) and AF110732 (rat AOEB166). C, Clustal W (version 1.7) alignment of human AOEB166 without its predicted mitochondrial presequence and S. cerevisiae YLR109W (accession number Z73281) also named thioredoxin peroxidase typeII (TypeII TPX) or alkyl hydroperoxide reductase (AHP1). Identical residues (asterisks) and conservative substitution (dots) are indicated.

Chromosomal Assignment-- According to the Genebridge 4 radiation hybrid panel, the human AOEB166 gene was mapped to chromosome 11q13, about 7 cR from marker D11S913 and between markers D11S1963 and D11S4407.

Deduced Amino Acid Sequences-- Analysis of the amino acid sequences of human and rat AOEB166 reveals several interesting features (Fig. 1B). First, the amino acid sequences are well conserved between the two species downstream from the second methionine (Met⁵³ for human AOEB166 and Met⁵² for rat AOEB166) because they are 90% identical. However, amino acid sequences diverge upstream from Met^52-53. A more precise analysis of human and rat amino acid sequences between Met¹ and Met^52-53 showed that although these sequences are different in amino acid composition, they both display mitochondrial presequence features (25, 26). Indeed, this sequence in human and rat is composed of abundant amino acid residues with positive charges, very few negative charges, and frequent hydroxylated residues. The existence of a SQL peroxisomal targeting sequence of the peroxisomal targeting signal 1 family (27-29) was noted at the C-terminal of human and rat AOEB166. Also, three cysteines in AOEB166 (Fig. 1B) were identified, and the functional significance of these residues in AOEB166 for its antioxidant activity will be discussed below.

AOEB166 Is a Novel Mammalian Peroxiredoxin-- Protein data bases were screened using BLAST 2.0 (gapped BLAST at the NCBI), and a search for identical or homologous polypeptides revealed that AOEB166 is a novel mammalian protein not yet characterized. Interestingly, sequence homology was noted with several proteins of different phyla, but none were from vertebrates. Among proteins with significant homology and known function or subcellular localization, we found that human AOEB166 (without its predicted mitochondrial presequence) had 68-65% similarity (36-35% identity) with, respectively, PMP20A and PMP20B peroxisomal membrane proteins of yeast Candida boidinii (GenBank^TM accession numbers J04984 and J04985), 66% similarity (26% identity) with YLR109w ORF of Saccharomyces cerevisiae (Fig. 1C; GenBank^TM accession number Z73281) recently identified as a thioredoxin peroxidase/alkyl hydroperoxide reductase (12, 30), and 63% similarity (25% identity) with E. coli alkyl hydroperoxide reductase AhpC22 protein (GenBank^TM accession number D13187). To identify homologies between human AOEB166 and the known members of the human peroxiredoxins, we selected one member of the five known subfamilies of human peroxiredoxins, and we performed an amino acid alignment (Fig. 2A). Notably, AOEB166 conserved amino acids especially around Cys¹⁰⁰ of human AOEB166, which has been directly implicated in catalysis of peroxides in peroxiredoxins (9, 10, 31-35). However, as for the so-called one-cysteine peroxiredoxin ORF06 (36), many residues in human AOEB166 differ from the consensus found for the other peroxiredoxins. Interestingly, human AOEB166 possesses two other cysteines at positions 125 and 204 (Fig. 1) that could be involved in the catalysis of peroxides and in dimerization because most of peroxiredoxins exist as homodimers or heterodimers (36). However, AOEB166 seems to diverge phylogenetically from known mammalian peroxiredoxins as illustrated by the alignment but also by the phylogenetic tree presented in Fig. 2. For these reasons, we propose that AOEB166 proteins represent a new peroxiredoxin subfamily named peroxiredoxin V. As discussed by Jin et al. (37) and as illustrated in the phylogenetic tree of Fig. 2B, the other subfamilies are subfamilies I, II, III, and IV and 1-Cys.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   AOEB166 defines a new peroxiredoxin subfamily. A, amino acid sequence alignment of human AOEB166 (the form identified in the bronchoalveolar lavage fluid without its predicted mitochondrial presequence) with five human antioxidant enzymes representing the known subfamilies of the peroxiredoxin family (37). The alignment was generated by CLUSTAL W (version 1.7) (20). Residues that are present in human AOEB166 and more than half of aligned peroxiredoxins are included in the consensus and highlighted with bold type. Conservative amino acids are also indicated (dots). The most highly conserved block is localized around the peroxiredoxin antioxidant domain defined by the conserved cysteine. B, evolutionary analysis of identified mammalian peroxiredoxins from Homo sapiens, Rattus norvegicus, and Mus musculus species. Identical sequences named differently and likely allelic variants were not considered for the analysis. Phylogenetic tree was constructed using CLUSTAL W (version 1.7) (20), and the unrooted tree reconstruction was performed with DRAWTREE from the PHYLIP package (46). GenBankTM accession numbers of sequences are as follows: human proliferation-associated gene (hPAG), X67951; human natural killer enhancing factor B (hNKEFB), L19185; human antioxidant enzyme 37-2 (hAOE37-2), U25182; human antioxidant protein 1 (hAop1), D49396; human ORF06 protein (hORF06), D14662; human antioxidant enzyme B166 (hAOEB166), AF110731; human thiol-specific antioxidant (hTSA), Z22548; rat antioxidant enzyme B166 (rAOEB166), AF110732; rat peroxiredoxin III (rPrxIII), AF106944; rat peroxiredoxin IV (rPrxIV), AF106945; rat thiol-specific antioxidant (rTSA), U06099; rat heme-binding 23-kDa protein (rHBP23), D30035; rat acidic calcium-independent phospholipase A2 (raiPLA2), AF014009; mouse thiol-specific antioxidant (mTSA), X82067; mouse antioxidant protein 1 (mAOP1), M28723; mouse nonselenium glutathione peroxidase (mGPx), Y12883; mouse osteoblast-specific factor 3 (mOSF-3), D21252.

Northern and Dot Blot Analysis of Human AOEB166-- Northern blot analysis of AOEB166 mRNA expression in human tissues and cell lines revealed a hybridizing region at approximately 1 kilobase (Fig. 3, A and B). AOEB166 mRNA is ubiquitously expressed in all tissues examined as well as in the cell lines. Master dot blots (CLONTECH) normalized for eight housekeeping genes were used to estimate the levels of AOEB166 mRNA in 30 different human tissues (Fig. 3C). Interestingly, expression was significantly different among the tissues. The highest levels of expression were detected in thyroid gland, trachea, kidney, lung, adrenal gland, heart, and colon. Lower but still detectable levels were observed in pancreas, peripheral leukocytes, lymph node, and whole brain.

View larger version (52K): [in this window] [in a new window]   Fig. 3.   Analysis of AOEB166 gene expression in human tissues and cell lines. A, human multiple tissue Northern blot (CLONTECH) containing approximately 2 µg of poly(A+) RNA of different tissue origins was analyzed for the presence of human AOEB166 mRNA. B, Northern blot containing 40 µg/lane of total RNA from human cell lines of different tissue origins was also hybridized with human AOEB166 cDNA probe. Cell lines were Caco-2 human colonic adenocarcinoma cell line, MRC-5 human fetal lung fibroblast cell line, ECV-304 human endothelial cell line, MDA-231 human breast cancer cell line, MCF-7 human breast cancer cell line, THP-1 human monocytic leukemia cell line, HT-29 human colonic adenocarcinoma cell line, LS174T human colonic adenocarcinoma cell line, T47D human breast cancer cell line, and HepG2 human hepatoblastoma cell line. Positions for 4.4 and 1.35 kilobases as marked by ribosomal markers are indicated. Both blots were stripped of the hAOEB166 probe and rehybridized with a human -actin probe to assess integrity and levels of RNA between lanes (small panels below each blot). C, quantitation of AOEB166 mRNA levels in human master dot blot (CLONTECH). The dot blot, containing poly(A+) RNA normalized to the mRNA expression levels of eight different housekeeping genes, was hybridized with AOEB166 cDNA probe. mRNA levels were quantitated using phosphorimaging technology. Higher levels are detected in thyroid gland and at lower levels in pancreas. Yeast S. cerevisiae mRNA was used as negative control.

Subcellular Localization of GFP Fusion Protein-- To demonstrate that identified mitochondrial and peroxisomal targeting sequences of AOEB166 are able to sort the protein into mitochondria and peroxisomes, a vector was made with a GFP fusion construct containing both the complete coding sequence of hAOEB166 (including the sequence encoding the predicted mitochondrial presequence) upstream from GFP and a sequence coding for AOEB166 peroxisomal targeting sequence SQL downstream from GFP (Fig. 4A). After transfection of HepG2 cells with pcDNA3.1 containing such a construct, GFP was detected in subcellular compartments with reticular or globular patterns that were proven to be mitochondria as demonstrated by co-localization with mitochondrial cytochrome c oxidase subunit I (Fig. 4, B and C) or in subcellular compartments with punctate patterns that were demonstrated to be peroxisomes by co-localization with peroxisomal catalase (Fig. 4, D and E).

View larger version (129K): [in this window] [in a new window]   Fig. 4.   Subcellular localization of human AOEB166 in fusion with GFP. A, structure of the construct cloned in mammalian expression vector pcDNA3.1 and transfected into HepG2 cells. The two potential ATG initiation codons and the TGA termination codon are indicated. 48 h after transfection, HepG2 cells were examined for GFP fusion expression and showed reticular (but also globular) and punctate subcellular fluorescent patterns typical of, respectively, mitochondria and peroxisomes (B and D). Subsequently, transfected cells were processed for immunofluorescence with antibodies directed either to mitochondrial cytochrome c oxidase subunit I or to peroxisomal catalase (see "Experimental Procedures"). Localization of human AOEB166 in fusion with GFP in mitochondria was confirmed by co-localization of GFP (B) with cytochrome c oxidase subunit I (C). Localization of human AOEB166 in fusion with GFP in peroxisomes was confirmed by co-localization of GFP (D) with catalase (E). BGH poly(A), bovine growth hormone polyadenylation signal; CoxI, cytochrome c oxidase subunit I; SQL, AOEB166 peroxisomal targeting sequence Ser-Gln-Leu; mito, predicted mitochondrial presequence of hAOEB166; pCMV, promoter of cytomegalovirus. Bars, 10 µm.

Antioxidant and Peroxidase Activity of Human AOEB166-- Antioxidant activity of human AOEB166 was measured on E. coli recombinant protein without its predicted mitochondrial presequence (Fig. 5). The protection of glutamine synthetase from inactivation by thiol-dependent metal-catalyzed oxidation has been extensively used previously to determine antioxidant properties of different peroxiredoxins (8-10). As shown in Fig. 5B, inactivation of glutamine synthetase in presence of DTT as sulfhydryl reductor is completely prevented in presence of 0.2 mg/ml of AOEB166. However, when DTT is replaced by ascorbate, a reductor lacking thiol, recombinant AOEB166 is unable to protect glutamine synthetase from inactivation as expected for a peroxiredoxin (Fig. 5C). To compare the potency of catalase to that of AOEB166, we measured protection activity at various concentrations of the proteins (Fig. 5D). Bovine catalase and AOEB166 exhibited 50% of protection at the same concentration of about 0.04 mg/ml. Peroxidase activity of recombinant human AOEB166 was also demonstrated by time-dependent removal of H₂O₂ or tert-butyl hydroperoxide (TBHP) in presence of DTT as reductor of AOEB166 (Fig. 6). The peroxidase activity of recombinant human AOEB166 measured by the consumption of H₂O₂ was very similar to the previously reported activity of hORF06, the human 1-Cys member of the mammalian peroxiredoxins (36).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Antioxidant activities of human AOEB166. A, SDS-polyacrylamide gel electrophoresis analysis of E. coli purified recombinant human AOEB166 (without its predicted mitochondrial presequence) was done on a Amersham Pharmacia Biotech Phast System, and the proteins were visualized by Coomassie Blue staining according to the manufacturer's recommendations. The recombinant protein was used for antioxidant activity assays. Identity of recombinant human AOEB166 was also confirmed by N-terminal microsequencing of 10 residues. Molecular mass markers (MW) are in the first lane. B, glutamine synthetase protection activity of recombinant human AOEB166. The protection assay was performed at 37 °C in a reaction mixture (100 µl) containing 0.6 units of E. coli glutamine synthetase (Sigma), FeCl3 3 µM, 10 mM DTT, and 20 µg of recombinant human AOEB166 in 50 mM imidazole, pH 7.0. Aliquots of 15 µl were removed at the indicated times and assayed for glutamine synthetase activity. 1 mM EDTA was used as control to chelate the catalyst Fe3+ and consequently prevents oxidative inactivation of glutamine synthetase. C, glutamine synthetase protection activity as described in B but in presence of ascorbate instead of DTT. D, variable amounts of recombinant human AOEB166 or bovine catalase were added into the inactivation mixture as described before in presence of DTT. After 20 min, glutamine synthetase activity was measured and normalized against the activity protected by 1 mM EDTA. Data are the means of duplicates from a representative experiment.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Peroxidase activity of recombinant hAOEB166. Time-dependent removal of H2O2 (A) or TBHP (B) by recombinant hAOEB166. The reaction mixture (200 µl) contained 50 mM Hepes-NaOH, pH 7.4, 2 mM DTT, 1 mM H2O2 or 800 µM TBHP, and recombinant hAOEB166 (0.2 mg/ml). At the indicated times, the remaining concentration of H2O2 or TBHP was measured in 20 µl of reaction mixture with ferrous ammonium sulfate/potassium thiocyanate and compared with standards. The reduction of H2O2 or TBHP in absence of AOEB166 was also measured (none). Data are the means of duplicate experiments.

AOEB166 Gene Expression during Lung Inflammation-- AOEB166 was first identified at high levels in bronchoalveolar lavage fluids of patients suffering from various lung diseases (23). We therefore assessed the possibility that AOEB166 is regulated at the transcriptional level in rat lung during inflammation induced by LPS instillation. As shown in Fig. 7, AOEB166 mRNA levels increased in rat lungs with inflammation. Higher expression was reached after 24 h and was still high 72 h after LPS instillation.

View larger version (42K): [in this window] [in a new window]   Fig. 7.   Rat AOEB166 gene expression is up-regulated in acute lung inflammation induced by LPS. AOEB166 and -actin mRNA levels in rat lungs were quantitated at different times in animals instillated with 100 µg of LPS/100 g of body weight or PBS as control. Northern blots containing 20 µg/lane of total RNA from rat lungs were hybridized with rat AOEB166 probe. The same blots were stripped of the AOEB166 probe and rehybridized with a rat -actin probe. Northern blots were quantitated using phosphorimaging technology (Molecular Dynamics), and AOEB166 mRNA data were normalized to -actin mRNA levels as shown in the histogram. The results represent the means of the values obtained in triplicate experiments. The error bars indicate the S.D. The Northern blots shown at the top in are representative of three experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have recently identified the AOEB 166 protein in human bronchoalveolar lavage fluid as a novel protein (23). Here, we show that, structurally and functionally, AOEB166 is a new member of the AhpC/TSA peroxiredoxin family, a recently identified group of antioxidant enzymes evolutionarly conserved in all phyla (8). Structurally, the antioxidant function of peroxiredoxins is dependent upon conserved cysteine residues responsible for peroxide reduction and dimerization (9, 38). In the peroxiredoxin prototype TSA from S. cerevisiae, two active cysteines are present in positions 47 and 170. These cysteines and their neighboring residues are highly conserved in four peroxiredoxin subfamilies (Prx I, II, III, and IV)³⁷ (see also Fig. 2). By contrast, the 1-Cys subfamily is defined by few peroxiredoxins that have conserved only the corresponding Cys⁴⁷ and its surrounding residues. AOEB166 does not fit perfectly these subfamilies. Indeed, like Prx I, II, III, IV and 1-Cys, AOEB166 possesses a cysteine corresponding to Cys⁴⁷ of S. cerevisiae TSA (Cys¹⁰⁰ for human AOEB166) but has no cysteine residue corresponding to Cys¹⁷⁰ of Prx I, II, III, and IV. AOEB166 does not enter either into the 1-Cys subfamily because it contains two other cysteines, Cys¹²⁵ and Cys²⁰⁴, lacking in 1-Cys subfamily and that may be involved in antioxidant activities and/or dimerization. Also, in AOEB166 the amino acids surrounding Cys¹⁰⁰ are much less conserved than in the other peroxiredoxins (Fig. 2A). For these reasons, AOEB166 represents the prototype for a new mammalian peroxiredoxin subfamily (Prx V in Fig. 2B).

Functionally, the antioxidant activity of AOEB166 has been confirmed in vitro by testing the ability of the recombinant protein to protect glutamine synthetase from the dithiothreitol/Fe3+/O₂ oxidation. Like other peroxiredoxins, AOEB166 requires a thiol-containing reductor (DTT) to exert its antioxidant activity and is inactive or less active in presence of other electron donors such as ascorbate. The cellular thiol-containing reductor is still to be identified, but two good candidates as direct electron donors for AOEB166 are thioredoxin and glutathione, which are physiological reductors of several members of the peroxiredoxin family (9, 12, 30, 37, 39). The antioxidant activity of recombinant AOEB166 was quantitatively comparable with that of catalase, which suggested that hydrogen peroxide was indeed a substrate for AOEB166. This was corroborated by the time-dependent removal of hydrogen peroxide by recombinant AOEB166 in the in vitro peroxidase assay. Of interest, tert-butyl hydroperoxide is also consumed by the recombinant protein, which demonstrates that AOEB166 is able to reduce organic peroxides like its S. cerevisiae orthologue (12, 30). Thus, these data suggest that AOEB166 might afford a protection not only against hydrogen peroxide but also against alkyl hydroperoxides in mammalian cells.

Analysis of deduced amino acid sequences of both human and rat AOEB166 reveals the presence of a predicted mitochondrial presequence at the N terminus as well as a SQL peroxisomal targeting signal type 1 at the C terminus in the same protein. Furthermore, we demonstrate that in fusion with the green fluorescent protein, these targeting sequences are functional and sort the protein to mitochondria and peroxisomes in HepG2 cells. Members of the Prx III subfamily have been also identified in mitochondria (40). AOEB166 represents therefore the second peroxiredoxin subfamily to be localized in mitochondria. Interestingly, AOEB166 is the only peroxiredoxin reported so far to be addressed to the peroxisomes. The functional significance of AOEB166 localization in organelles to which other antioxidant proteins with similar enzymatic activities (glutathione peroxidase, catalase) are sorted is still to be investigated.

The fact that AOEB166 is well conserved among species and is expressed in all tissues and cell lines examined in this study is consistent with an important physiological function for that protein. In that respect, it is interesting to note that AOEB166 expression is highest precisely in those tissues, such as thyroid gland, lung, or kidney, that are particularly exposed to oxidative stress (41, 42). Although at this stage the implication of AOEB166 in physiopathological processes remains speculative, the significant increase of AOEB166 gene expression in rat lung with lipopolysaccharide-induced inflammation suggests that the protein may play in vivo a protective role against oxidative damage. Furthermore, several observations indicate that AOEB166 might be implicated in various pathophysiological situations by mechanisms that do not necessarily imply directly its antioxidant activity but rather its high conservation during evolution. In particular, the AOEB166 gene is located to human chromosome 11q13, which is a region of genetic linkage for atopic hypersensitivity (asthma, hay fever, and eczema) (43) and AOEB166 presents a high homology to a major allergen of Aspergillus fumigatus (GenBank^TM accession number U58050) (44). An attractive hypothesis would be that IgE antibodies directed to the allergen would cross-react with AOEB166 and therefore initiate autoimmunity. This mechanism has been postulated for allergen manganese superoxide dismutase of A. fumigatus, which also exhibits a high homology to human manganese superoxide dismutase (45).

In conclusion, our data show that AOEB166 represents a new subfamily of the peroxiredoxin mammalian antioxidant enzymes with functional mitochondrial and peroxisomal targeting signals. The protein, highly conserved throughout species and widely distributed in the body, presents several features, suggesting that it may play an important protective role against oxidative damages caused by peroxides in organelles that are major sources of ROS.

	ACKNOWLEDGEMENTS

The skillful technical assistance of Sebastien Boulanger is gratefully acknowledged. We thank Dr. François Huaux for guidance with in vivo experiments. We also thank Drs. René Rezsohazy and Pascal Hols for helpful discussions during the preparation of the manuscript. Nathalie Havaux, Nathalie Hellen, and Drs. Fabienne Kinard, Vincent Dubois, and Jean-François Rees are also acknowledged for providing human cell line cultures. We thank Prof. Frank Roels and Dr. Marc Espeel for suggestions concerning anti-catalase immunocytochemistry.

	Addendum

Since this article was submitted, a third paper describing S. cerevisiae YLR109w thioredoxin peroxidase function has been published (47).

	FOOTNOTES

^* This work was supported by a grant from the National Research Funds of Belgium, European Union Environment and Climate Program Grant EV4-CT96-0171, and a grant from the French Association of Lead and Cadmium.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF110731 and AF110732.

^¶ To whom correspondence should be addressed: Laboratory of Cell Biology, Dept. of Biology, Catholic University of Louvain, Place Croix du Sud, 5, B-1348 Louvain-la-Neuve, Belgium. Tel.: 32-10-47-37-60; Fax: 32-10-47-35-15; E-mail: knoops@bani.ucl.ac.be.

^** Scientific Collaborator of the National Research Funds of Belgium.

Research Fellow of the National Research Funds of Belgium.

^§§ Research Director of the National Research Funds of Belgium.

	ABBREVIATIONS

The abbreviations used are: ROS, reactive oxygen species; AhpC, alkyl hydroperoxide reductase subunit C; DTT, dithiothreitol; GFP, green fluorescent protein; LPS, lipopolysaccharide; ORF, open reading frame; PBS, phosphate buffered saline; PCR, polymerase chain reaction; Prx, peroxiredoxin; RACE, rapid amplification of cDNA ends; TBHP, tert-butyl hydroperoxide; TSA, thiol-specific antioxidant; bp, base pair(s); TRITC, tetramethylrhodamine isothiocyanate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES