A Novel Stress-Inducible Antioxidant Enzyme Identified from The

Planta (2002) 215: 716–726 DOI 10.1007/s00425-002-0819-0

ORIGINAL ARTICLE

Shaheen B. Mowla Æ Jennifer A. Thomson Jill M. Farrant Æ Sagadevan G. Mundree A novel stress-inducible antioxidant enzyme identiﬁed from the resurrection plant Xerophyta viscosa Baker

Received: 25 March 2002 / Accepted: 30 May 2002 / Published online: 10 July 2002 Springer-Verlag 2002

Abstract A cDNA corresponding to 1-Cys peroxire- which may function to protect nucleic acids within the doxin, an evolutionarily conserved thiol-specific an- nucleus against oxidative injury. tioxidant enzyme, was isolated from Xerophyta viscosa Baker, a resurrection plant indigenous to Southern Af- Keywords Desiccation stress Æ Peroxiredoxin Æ rica and belonging to the family Velloziaceae. The Resurrection plant Æ Xerophyta cDNA, designated XvPer1, contains an open reading frame that encodes a polypeptide of 219 residues with a Abbreviations ABA: abscisic acid Æ Cys: cysteine Æ NLS: predicted molecular weight of 24.2 kDa. The XvPer1 nuclear localisation signal Æ PFD: photon flux densi- polypeptide shows significant sequence identity (approx. ty Æ Prx: peroxiredoxin Æ ROS: reactive oxygen spe- 70%) to other recently identified plant 1-Cys peroxire- cies Æ RWC: relative water content Æ ww: water potential doxins and relatively high levels of sequence similarity (approx. 40%) to non-plant 1-Cys peroxiredoxins. The XvPer1 cDNA contains a putative polyadenylation site. Introduction As for all 1-Cys peroxiredoxins identified to date, the amino acid sequence proposed to constitute the active Xerophyta viscosa Baker (family Velloziaceae) is a site of the enzyme, PVCTTE, is highly conserved in monocotyledonous desiccation-tolerant plant belonging XvPer1. It also contains a putative bipartite nuclear to a small group of angiosperms commonly referred to localization signal. Southern blot analysis revealed that as resurrection plants (Gaff 1971; Bewley and Oliver there is a single copy of XvPer1 in the X. viscosa genome. 1992; Vertucci and Farrant 1995; Ingram and Bartels All angiosperm 1-Cys peroxiredoxins described to date 1996). Desiccation tolerance is defined as the ability to are seed-specific and absent in vegetative tissues even revive from the air-dried state, which is the most severe under stress conditions; therefore, XvPer1 is unique in form of water stress, since under these conditions the that it is expressed in the vegetative tissues of X. viscosa. majority of protoplasmic water is lost from the cell. The XvPer1 transcript was absent in fully hydrated X. X. viscosa can be dehydrated to 5% relative water con- viscosa tissue but levels increased in tissues subjected to tent (RWC) and upon rewatering the plant rehydrates abiotic stresses such as dehydration, heat (42 C), high completely within 80 h, resuming full physiological ac- l –2 –1 light intensity (1,500 mol photons m s ) and when tivities (Sherwin and Farrant 1996). l treated with abscisic acid (100 M ABA) and sodium Many resurrection plants are being studied in an at- chloride (100 mM NaCl). Western blot analyses corre- tempt to understand the mechanisms that enable vege- lated with the patterns of expression of XvPer1 tative tissues to withstand desiccation. The ultimate aim transcripts under different stress conditions. Immuno- is to identify characteristics that may be used to produce fluorescence analyses revealed that XvPer1 is localized in crops with improved tolerance to osmotic stress. X. vis- the nucleus of dehydrated X. viscosa leaf cells. These cosa, being monocotyledonous, can serve as a model for results suggest that XvPer1 is a stress-inducible gene, the identification of characteristics that could be used to improve stress tolerance of cereal crops (Mundree and Farrant 2000). S.B. Mowla Æ J.A. Thomson Æ J.M. Farrant Æ S.G. Mundree (&) In addition to water limitation, the presence of high Department of Molecular and Cell Biology, light intensities during dehydration can be extremely University of Cape Town, Private Bag, Rondebosch, 7701, South Africa damaging to photosynthetically active tissues (Sherwin E-mail: [email protected] and Farrant 1998). Even under mild water-stress con- Fax: +27-21-6897573 ditions, closure of the stomata can result in excitation 717 energy being transferred from chlorophyll to oxygen, functions: as an antioxidant, an endogenous regulator of leading to the formation of oxygen free radicals. If un- apoptosis (Ichimiya et al. 1997), and as an intracellular quenched, these molecules can cause considerable dam- signalling molecule (Kang et al. 1998; Kowaltowski et al. age to the subcellular milieu. Reactive oxygen species 1998). Plant 2-Cys Prx proteins also function as radical (ROS) play a significant role in causing damage to living scavengers coupled to the photosynthetic machinery in cells under severe stress conditions. Some examples of chloroplast (Baier and Dietz 1997, 1999). – ROS are O2 radicals, OH¢ and H2O2 (Foyer et al. 1994). The cellular function of 1-Cys Prx is not completely By themselves, those compounds are relatively unreac- understood. Its protein structure has been elucidated tive and can even be used by the plant in a beneficial and several cDNAs encoding the protein have been manner. For example, H2O2, produced from the oxida- isolated and characterised from yeast, animal, and plant tive burst generated during hypersensitive plant–patho- cells (Goldmark et al. 1992; Aalen et al. 1994; Stacy et al. gen interaction, functions as a local trigger of 1996, 1999). In angiosperms, it is specifically expressed programmed cell death and causes rapid cross-linking of in the nucleus of immature embryos and the aleurone cell wall proteins (Levine et al. 1994). H2O2 also appears layers of the seed (Haslekas et al. 1998). The expression to act as a signal molecule that induces the transcription level increases late in seed development and is main- – of defence-related genes. However, O2 radicals and tained in mature seeds during storage. However, when H2O2 can react to produce ROS that are damaging to non-dormant seeds are imbibed the transcript level de- essential cellular components. Transition metals catalyse creases dramatically and transcripts disappear com- the formation of OH¢ radicals through the Haber-Weiss pletely after seed germination. This suggests that the reaction and this accounts for much of the toxicity of functional role of 1-Cys Prx in angiosperms might be the – O2 radicals and H2O2. They are capable of causing in- maintenance of seed dormancy. However, this sugges- discriminate lipid peroxidation, protein denaturation, tion had to be questioned when it was reported that the and damage to DNA (Inzeánd Van Montagu 1995). transcript levels did not correlate with abscisic acid To combat oxidative stress, complex protective (ABA) levels. ABA is required for the induction of seed mechanisms have been evolved by plants to mitigate and dormancy in ABA-deficient or insensitive mutants of repair the damage initiated by free radicals (Price et al. Arabidopsis (Stacy et al. 1996). 1994). The primary constituents of these protective Recently, transgenic tobacco plants constitutively mechanisms include enzymes such as superoxide expressing rice 1-Cys peroxiredoxin (R1C-Prx) were dismutase (SOD), catalases and peroxidases, and free- generated (Lee et al. 2000). These transgenics showed radical scavengers such as carotenoids, ascorbate, toc- enhanced resistance against oxidative stress imposed by opherols, and oxidized and reduced glutathione (GSSG H2O2 as compared to wild-type control plants. There- and GSH, respectively). fore, the in vivo function of 1-Cys Prx in angiosperms Like most resurrection plants, X. viscosa grows in may not be related to the maintenance of seed dormancy, shallow soils on rocky outcrops at high altitudes where but rather to protective activity against oxidative stress. there is little shade. Upon drying, it loses its chlorophyll Using differential screening of an X. viscosa cDNA (poikilochlorophyllous), and the leaf blades fold in half library (Ndima et al. 2001), a peroxiredoxin homologue along the midrib, with only the abaxial surface being was identified. In this paper we describe the isolation exposed to light. The leaves initially turn yellow, and and characterisation of the XvPer1 cDNA. We investi- then dark purple when in a more advanced dry state. The gate its expression pattern under different stress condi- abaxial surfaces have a reflective sticky coating, which tions in X. viscosa. We also localised the protein to the may serve to reduce light absorbed by the leaf. The ac- nucleus of leaf cells under desiccation stress, which tivities of the three common antioxidant enzymes, makes it unique among angiosperms as other peroxire- ascorbate peroxidase, glutathione reductase and SOD, doxins have been found to be almost entirely confined to increase during dehydration (Sherwin and Farrant 1998). seeds and are absent in vegetative tissues, even upon Peroxiredoxins are one of the most recently discov- dehydration. ered types of enzymatic antioxidant (Chae et al. 1994). They have been shown to be active on substrates such as hydroperoxides and alkyl hydroperoxides. Based on Material and methods their amino acid sequences and immunological properties, the peroxiredoxin (Prx) proteins can be divided into Plant material six distinct groups (Prx I–VI), all of which contain one or two conserved cysteine (Cys) residues, called 1-Cys Xerophyta viscosa Baker plants were collected from Cathedral Peak Nature Reserve, (Kwazulu Natal Province, South Africa). The Prx and 2-Cys Prx (Lyu et al. 1999). All peroxiredoxins plants were grown under glasshouse conditions as described by conserve at least one cysteine residue at the N-terminus. Sherwin and Farrant (1996). They have neither tightly bound metal ions nor pros- thetic groups such as heme or flavin (Kim et al. 1988). The ability of peroxiredoxins to reduce peroxides is at- Sequencing and analysis of XvPer1 tributed to their cysteine residues. 2-Cys Prx exists in a The nucleotide sequence of XvPer1 cDNA was determined on both wide variety of organisms and has diverse cellular forward and reverse strands using the ALFexpress automated 718

DNA sequencer AMV3.0 (Pharmacia Biotech) using the sequenc- was carried out at an annealing temperature of 55 C, for an ing Fluorescent Labelled Primer cycle sequencing Kit (Amersham extension time of 10 min, and the cycle repeated 15 times. International). The inferred amino acid sequence of XvPer1 was obtained by translation of the cDNA sequence using DNAMAN (Version 3.0, 1997) and used to search for similarities in protein RNA isolation and northern blot analysis sequence databases using the Blast network service (Altschul et al. 1990). Amino acid comparisons were done with the CLUSTAL Total RNA was extracted from control and stress-treated X. viscosa program of DNAMAN (Version 3.0, 1997). leaves using the Trizol method following the manufacturer’s instructions (Life Technologies). For RNA gel blot analysis, samples of approximately 10 lg RNA were electrophoresed in a 1.2% DNA isolation and Southern blot analysis agarose gel stained with EtBr, transferred and UV cross-linked onto nylon membranes (Hybond-XL; Amersham Pharmacia Biotech). Genomic DNA was extracted from leaves of fully hydrated X. vis- To estimate whether equal amounts were loaded, the RNA was cosa plants according to the procedure described by Dellaporta visualized by EtBr staining. Hybridisation was carried out at 65 C et al. (1983). The concentration and purity of the DNA preparation in the same hybridisation buffer as used for Southern hybridisation. were determined spectrophotometrically. Aliquots (15 lg) were The membranes were washed in wash buffer B for 10 min and ex- digested with the restriction enzymes EcoRI, EcoRV and HindIII posed to high-performance film (Amersham Pharmacia Biotech) at (Boehringer-Mannheim) respectively, and electrophoresed on 0.8% –70 C and thereafter developed after 72 h exposure. ethidium bromide (EtBr)-stained agarose gels. The DNA was subsequently transferred and UV cross-linked onto nylon membrane (Hybond-XL; Amersham Pharmacia Biotech), which was Plant stress treatments hybridised at 65 C with radiolabelled probe (PCR on XvPer1 cDNA using primer A 5¢-CCATGCCGGGGCTCACCATT-3¢ All treatments were performed in a phytotron at 25 C, 50% hu- and primer B 5¢-CATTCACTCAGACGTTCGTAAAACG-3¢ midity, a photon flux density (PFD) of 300 lmol m–2 s–1, and a 32 with [a- P]dCTP) in hybridisation buffer (0.5 M NaH2PO4, day/night cycle of 16/8 h, unless otherwise stated. Dehydration/ 0.001 M EDTA, 7% SDS, 1% BSA). The membrane was washed rehydration, heat, high light and cold treatments were done on twice for 10 min in wash buffer B (0.1% SDS, 0.5·SSC) and whole plants while NaCl and ABA treatments were done on excised autoradiographed at –70 C onto high-performance-autoradiog- leaves from healthy plants. Leaf samples were taken by cutting off raphy film (Amersham Pharmacia Biotech). The protocol was leaves (mid-age, middle portion), immediately wrapping them in adapted from Sambrook et al. (1989). The PCR labelling reaction aluminium foil, flash-freezing in liquid nitrogen and subsequently

Fig. 1. Nucleotide and deduced amino acid sequence of XvPer1 from Xerophyta viscosa. The cDNA is 849 bp long and has two putative polyadenylation sites at the 3¢-end (under- lined). The amino acid sequence has an open reading frame from 1 to 657, 219 amino acids and a molecular weight of 24,223 Da. The highly conserved active site PVCTTE is boxed at position 44–49 and the putative nuclear localisation signal (NLS) is highlighted at positions 195, 196; 210–211; 214–217. The stop codon is indicated by an asterisk (*) 719 storing at –70 C until nucleic acid isolation. Dehydration stress values (determined in triplicate) for each treatment were plotted on was carried out by withholding water from the whole plant over a separate axes using the Microsoft Excel software package (Micro- period of 25 days, at the end of which the RWC was 5%. Rehy- soft Office 2000). dration took place when the plant was re-watered and returned to full turgor 5 days after watering. Heat stress was imposed by increasing the temperature in the phytotron to 42 C, leaving all Protein extraction and western blot analysis other parameters constant. The plant was watered every day to eliminate the chance of dehydration stress. Cold stress was carried Proteins were extracted using the Trizol reagent according to the out in a chamber kept at 4 C with a PFD of approx. 250 lmol m–2 manufacturer’s instructions. Following sodium dodecyl sulphate– s–1 and a 16/8 h day/night cycle. High-light stress was imposed by polyacrylamide gel electrophoresis (SDS–PAGE), proteins were placing plants close to a high-light source in the phytotron. The transferred onto nylon membranes (Hybond-P; Amersham Phar- PFD was approx. 1,500 lmol m–2 s–1, while temperature and hu- macia Biotech) using a western transfer apparatus (Bio-Rad). Fil- midity were kept constant. The salinity and ABA stresses were ters were blocked with 2.5% skimmed milk in 1·PBS (pH 7.4) for imposed on leaves excised from a healthy plant. Mid-age leaves 2 h, labelled overnight at 4 C with XvPer1 antiserum diluted were excised from a whole plant and the cut ends were submerged 1:2,000, washed for 2·10 min in blocking solution and incubated in either 150 mM NaCl or 100 lM ABA solution. At specific time for 2 h with secondary antibody (anti-rabbit IgG, peroxidase- points the leaves were removed and flash-frozen. A wounding linked whole antibody from goat; Sigma) diluted 1:5,000, prior to control was performed as well using sterile distilled water as the the final wash (2·10 min in 1·PBS, pH 7.4). Antibody binding was solution. The RWC was determined at each sampling point for detected by using the ECL detection system adapted from the ECL each treatment. The initial weight (Wi) of each sample was taken detection kit from Amersham. To confirm equal loading in all before immersing it in Milli-Q water. The weight at full turgor (Wt) lanes, filters were stained with Ponceau red prior to detection. was taken and leaf samples were then dried at 95 C for 24 h and Filters were exposed to Kodak BioMax ML Film (Sigma). the dry weight (Wd) was recorded. The RWC was calculated using the formula previously determined by Jin et al. (2000): RWC=[(Wi–Wd)/Wd]/[(Wt–Wd)/Wd]·100. The water potential Immunolocalisation (ww) was also determined at each sampling point using a thermo- couple psychrometer (Aqualab 1.5; Decagon, Washington, USA) Paraffin-embedded sections prepared from hydrated and dehy- following the manufacturer’s instructions. Average RWC and ww drated X. viscosa leaf tissue were incubated with 1:200 dilutions of

Fig. 2. Multiple sequence alignment of peroxiredoxin homologue sequences from Bromus secalinus (p52571), Hordeum vulgare (p52572), Oryza sativa (p52573), Arabid- opsis thaliana (0040005) and Fagopyrum esculentum (AAF12782) with XvPer1. Asterisks (*) denote identities and dots (Æ) denote similarities 720

Fig. 4. Southern blot analysis of XvPer1. Genomic DNA (20 lg) isolated from leaf tissues of X. viscosa was digested with EcoR1 (lane 1), EcoRV (lane 2) and HindIII (lane 3), fractionated by electrophoresis, transferred onto a nylon membrane and hybridised 32 XvPer1 Fig. 3. Phylogenetic tree representing the relationship of the with a P-labelled cDNA probe. The molecular sizes of the X. viscosa 1-Cys Prx protein sequence to 16 other peroxiredoxin bands observed are indicated on the right margin homologues. The accession numbers of the sequences used in the comparison are: Sulfolobus metallicus (Archea), AF007757; Pseudo- monas putida (bacterium), AF075709; Saccharomyces cerevisiae (yeast), Z23261; A. thaliana, Y12089; B. secalinus (brome grass), sequence. The cysteine residue constitutes the active site X63202; H. vulgare (barley), X96551; O. sativa (rice), D63917; of the enzyme and is characteristic of all 1-Cys Prx X. viscosa, AF484696; F. esculentum (buckwheat), AF191099; proteins isolated thus far. A putative bipartite nuclear Tortula ruralis (moss), U40818; Bos taurus (cow), AF090194; Homo localisation signal (NLS) is also present, which is con- sapiens (human), D14662; Rattus norvegicus (rat), Y17295; served in all plant 1-Cys Prx proteins (Raikhel 1992). O. volvulus (nematode), U31052; Plasmodium falciparum (protozo- an), AB020595 Two hexanucleotide sequences, ATATAT, found at the 3¢ end of the sequence, represent a putative polyadenylation signal of this gene (Wu et al. 1995). The deduced amino acid sequence of XvPer1 exhibited considerable anti-XvPer1 serum and pre-immune serum. Immunoreactivity was visualized using fluorochrome-tagged secondary goat anti-rabbit similarity to other plant 1-Cys Prx homologues (i.e. 72– IgG antibody (Alexa fluor 568; Molecular Probes) at a dilution of 77%, Fig. 2) from Bromus secalinus (Goldmark et al. 1:1,000. Fluorescence was detected and viewed with an inverted 1992, p52571), Hordeum vulgare (Aalen et al. 1994, fluorescent microscope (Nikon) under oil emersion with a B2 p52572), Oryza sativa (Fujino et al. unpublished, DM510 epi-fluorescence filter (Zeiss). The protocol was adapted from Vander Willigen (2001). p52573); Arabidopsis thaliana (AtPer1) (Haslekas et al. 1998, 0040005) and Fagopyrum esculentum (FePer1) (Lewis et al. 2000, AAF12782). Over 50 members of the peroxiredoxin family have been identified in organisms Results from all kingdoms. In an attempt to determine the evolutionary relationships between 1-Cys Prx family Isolation and analysis of XvPer1 cDNA members, a phylogenetic tree was generated which revealed that XvPer1 is most closely related to its rice The XvPer1 cDNA was isolated by differential screening homologue (Fig. 3). of a X. viscosa cDNA library (Ndima et al. 2001). This cDNA is up-regulated when the plant is dehydrated. The nucleotide sequence of 849 bp, representing the full- Southern blot analysis of XvPer1 length cDNA, was sequenced and the sequence analysed at the nucleotide and amino acid levels (Fig. 1). The Southern blot analysis of total genomic DNA from open reading frame (ORF) of XvPer1 coded for a pro- hydrated (100% RWC) X. viscosa leaves revealed that tein composed of 219 amino acid residues. XvPer1 XvPer1 is a single-copy gene (Fig. 4). The restriction contains the highly conserved PVCTTE amino acid enzymes EcoRI and EcoRV do not cut inside the 721 cDNA transcript and give single bands of 9 kb and level was also increased by high-temperature stress 14 kb respectively. HindIII cuts the cDNA transcript (Fig. 5B). Salinity stress did not appear to affect twice and hence three bands were observed (0.9 kb, XvPer1 abundance, although low levels were observed 0.4 kb and 0.2 kb), with the 0.4-kb band corresponding at days 4–7 (Fig. 5C). The XvPer1 transcript level was to the internal XvPer1 fragment. Fainter bands were increased by cold, with expression starting at day 3 also observed in all three lanes, which might be due to and increasing until day 7 (Fig. 5D). High light re- non-specific hybridisation or caused by partial diges- sulted in an increase in the XvPer1 steady-state mRNA tion of the genomic DNA by these enzymes. XvPer1 is level (Fig. 5E), so did the plant stress hormone ABA therefore likely to be a single-copy gene in the X. vis- (Fig. 5F). RWC and ww of plants for all treatments, cosa genome. apart from the dehydration treatment, remained high enough not to impose additional stress besides the in- tended one (Fig. 6). XvPer1 expression

X. viscosa plants were subjected separately to dehy- Western blot analysis of XvPer1 dration/rehydration, heat, NaCl, cold, high light intensity and ABA stresses. The XvPer1 steady-state Western blot analysis was performed on protein iso- mRNA transcript abundance was higher when the lated from X. viscosa tissues subjected to dehydration/ plant was dehydrated from 63% to 4% RWC relative rehydration, heat, NaCl, cold, high light and ABA. to the fully hydrated state (Fig. 5A). XvPer1 tran- When X. viscosa was hydrated (95%–75% RWC), scripts were present during the early phase of rehy- XvPer1 was absent (Fig. 7A). The protein could be dration and disappeared when 92% RWC was detected as the RWC dropped to 51% and was present reached. The XvPer1 steady-state mRNA transcript throughout the dehydration process (to 6% RWC). As

Fig. 5A–F. Northern blot analysis of XvPer1 expression under diﬀerent environmental stress conditions. A Dehydra- tion/rehydration; B heat (42 C); C NaCl (100 mM); D cold (4 C); E high light (PFD 1,500 lmol m–2 s–1); F ABA (100 lM). Upper panels Results of hybridisation using radiolabelled full-length XvPer1 cDNA as a probe. Lower panels Ribosomal RNA bands on an ethidium bromide-stained 1.2% agarose gel for comparison of RNA loading in gels 722

Fig. 6A–F. Relative water content (squares) and water potential (circles) for each stress treatment. A Dehydration/ rehydration; B heat (42 C); C NaCl (150 mM); D ABA (100 lM); E high light (PFD 1,500 lmol m–2 s–1); F cold (4 C). In most cases, the error bars (±SE) are wholly contained within the symbol

the plant rehydrated, the protein level started to Discussion decrease to a non-detectable level at 95% RWC. When X. viscosa was subjected to high-temperature stress (42 C), XvPer1 was detected at 72 h and 96 h By differential screening of a cDNA library from (Fig. 7B). Under salinity stress, XvPer1 expression was X. viscosa, a stress-inducible gene named XvPer1 was detected at 48 h and 72 h but not at 96 h (Fig. 7C). isolated. The data presented in this study indicate that ABA substantially induced XvPer1 expression with XvPer1 encodes a 1-Cys Prx, which is a newly dis- high levels accumulating at 48 h and 72 h (Fig. 7D). covered class of thiol-dependent antioxidant enzyme. No expression was detected under cold and high-light The full-length cDNA sequence is 849 bp, containing stresses. one major reading frame of 660 bp (Fig. 1). Two putative polyadenylation signals, ATATAT, were identified 155 bp and 172 bp downstream from the Immunofluorescence localisation of XvPer1 stop codon. The amino acid sequence deduced from in dehydrated X. viscosa tissues the full-length cDNA indicates that XvPer1 encodes a polypeptide of 219 amino acid residues with a pre- The putative NLS present in the XvPer1 amino acid dicted molecular mass of 24.2 kDa. This coincides sequence indicated that the protein might be localised to with other 1-Cys Prx proteins characterised in both the nucleus of dehydrated X. viscosa cells. An immu- plants and mammals (Haslekas et al. 1998; Kang et al. nofluorescence study was undertaken in order to 1998; Stacy et al. 1999). XvPer1 has more than 70% examine the cellular localisation of the protein in leaf identity with other plant 1-Cys Prx proteins (Fig. 2) tissues. It is clear that XvPer1 is localised to the nucleus (Aalen et al. 1994; Fujino et al. unpublished; Haslekas (Fig. 8). et al. 1998; Lewis et al. 2000). 723

Fig. 7A–D. Western blot analysis of X viscosa exposed to diﬀerent environmental stress treatments. A Dehydration/ rehydration; B heat (42 C); C NaCl (100 mM); D ABA (100 lM)

Peroxiredoxins represent a highly conserved family of The next, B15C (later re-named PER1), was identified in proteins found in organisms as widely divergent as the barley (Hordeum vulgare L.) seeds (Aalen et al. 1994). Archea bacteria, plants and humans (Fig. 3) (Kang et al. Thereafter, a 1-Cys Prx gene, AtPer1, was isolated from 1998). the seeds of the dicotyledonous plant, A. thaliana The conservation of the amino acid sequence among (Haslekas et al. 1998). In all of the above, transcripts of 1-Cys Prx proteins identified from widely divergent the gene disappeared after germination and were absent species indicates its importance in function. XvPer1 in vegetative tissues even under dehydration stress. The contains the highly conserved PVCTTE amino acid se- transcript was also always specific to the aleurone layer quence (44–49) where the cysteine residue at position 46 and the embryo of the seeds. The authors postulated that constitutes the active site of the enzyme and is conserved the protein performs a protective function in the only in all 1-Cys Prx proteins isolated. For the human 1-Cys two tissues that survive desiccation. A second role pos- Prx, site-directed mutagenesis studies have demonstrated tulated was its involvement in the maintenance of seed that this conserved cysteine residue is essential for the dormancy. X. viscosa is unique in that it is the first function of the protein, constituting the site of oxidation angiosperm in which a 1-Cys Prx has been found in (Kang et al. 1998). vegetative tissues. The homologue from the resurrection Stacy et al (1999) showed that the dormancy-related moss Tortula ruralis (bryophyte), Tr155, is expressed in barley peroxiredoxin homologue, PER1, is localized in the gametophyte in response to desiccation and rehy- the nucleus of immature embryo and aleurone cells. In dration (Wood et al. 1999). No published data have been XvPer1, a putative bipartite NLS was identified in the found on this gene in relation to its activity or localisa- carboxyl-terminal end of the protein sequence, which is tion. Rab24, the rice homologue, was found in callous conserved in all the plant 1-Cys Prx proteins (Lewis et al. tissue (reported in the GenBank entry), but there is no 2000). This would suggest that this enzyme is targeted to published evidence of expression in the vegetative tissues the nucleus. However, a similar core sequence is also of Oryza sativa plants. XvPer1 is absent from healthy present in the human 1-Cys Prx homolog, which has unstressed plants, but is transcribed as soon as the plant been localised to the cytosol (Kang et al. 1998; Stacy is exposed to various environmental and abiotic stresses et al. 1999). XvPer1 appears to be encoded by a single- such as dehydration, heat, cold, high light and ABA copy gene (Fig. 4). Again, this is characteristic of all (Fig. 5). X. viscosa, being a resurrection plant, has cer- plant 1-Cys Prx proteins isolated (Stacy et al. 1996; tain seed-specific properties. For example, it can remain Haslekas et al. 1998). The first plant 1-Cys Prx was dormant in a dry state for long periods of time without isolated from the seed of Bromus secalinus, a grass spe- sustaining major tissue damage (Scott 2000), and hence cies, while the authors were investigating the mecha- it is not surprising that it should express certain ‘‘tradi- nisms involved in seed dormancy (Goldmark et al. 1992). tionally’’ seed-specific genes in times of stress. 724

Fig. 8A, B. Subcellular immunofluorescence localization of XvPer1 in hydrated and dehydrated X. viscosa leaf cells. Paraffin-embedded sections were incubated in XvPer1 antiserum and secondary antibody conjugated to Alexa fluor 568 for detection by inverted fluorescent microscopy. Red label indicates a XvPer1-specific reaction. A Transverse sections as seen under the inverted light microscope; B the same sections under fluorescent light. a, b Dehydrated X. viscosa leaf cells incubated with anti-XvPer1. c, d Hydrated X. viscosa leaf cells incubated with anti-XvPer1. N Nucleus, C chloroplast. ·10,000

Previous functional assays on 1-Cys Prx have re- promoters and mutants having altered responses to vealed that it is a DNA-protecting enzyme (Stacy et al. ABA provided evidence of multiple signalling pathways 1996). A second role was in the maintenance of dor- and changes in gene expression that are ABA-dependent mancy. However, a rice 1-Cys Prx over-expressed in and/or- independent (Bray 2002). Analysis of the pro- transgenic tobacco showed that the gene did not main- moter of AtPer1, the Arabidopsis homologue, identified tain dormancy in seeds but rather enhanced resistance an ABRE (ABA response element), and an ARE against oxidative stress in transgenic plants, suggesting (antioxidant response element) (Haslekas et al. 1998). It that antioxidant activity may be its primary function is hence likely that XvPer1 expression is ABA-depen- (Lee et al. 2000). In this study, XvPer1 steady-state dent. ROS are known to cause damage to nucleic acids. mRNA transcript abundance as well as protein levels The immunofluorescence study undertaken here to were found to increase under drought and heat stresses. examine the cellular localization of XvPer1 in the leaf These two stresses are known to induce, through stom- tissues revealed that the protein is concentrated in the atal closure, a reduced CO2 availability at the Rubisco nucleus (Fig. 8). This suggests that it is involved in the site (Broin et al. 2000). Consequently, an excess of ab- protection of nucleic acids against ROS. There are other sorbed light energy occurs and provokes a deviation of antioxidant enzymes that plants express to combat oxi- reducing power to oxygen and an increased production dative stress, including superoxide dismutase (SOD), of ROS. Oxidative stress resulting from ROS has also catalases and peroxidases, and free-radical scavengers been shown to occur in response to low temperature, such as carotenoids, ascorbate, tocopherols, and oxi- high light intensities and herbicides (Inzeánd Montague dized and reduced glutathione (GSSG and GSH, re- 1995). This could explain the increase in XvPer1 steady- spectively) (Price et al. 1994). However, the newly state mRNA transcripts and protein in X. viscosa sub- discovered 1-Cys Prx is believed to have a more spec- jected to these stresses (Figs. 5, 7). ABA is involved in ialised role in combating oxidative stress, although this the generation of stress-inducible genes and is required has not been elucidated. Only a few of the desiccation- for changes in gene expression in response to water- associated proteins described have been found localised deficit stress (Bray 1997). Both the steady-state mRNA to the nucleus (RAB17, Goday et al. 1994; TAS14, transcript level and protein level of XvPer1 increase Godoy et al. 1994; RAB28, Niogret et al. 1996; GP47, upon treatment with ABA (Figs. 5, 7). The analysis of Chiantante et al. 1995; PER1, Stacy et al. 1999). Of the 725 XvPer1 homologs, antioxidant activity has been shown Godoy JA, Lunar R, Torres-Schumann S, Moreno J, Rodrigo RM, for the human 1-Cys Prx, (Netto et al. 1996), the bovine Pintor-Toro JA (1994) Expression, tissue distribution and subcellular localization of dehydrin TAS14 in salt-stressed to- eye 1-Cys Prx (Peshenko et al. 2001) and the yeast 1-Cys mato plants. Plant Mol Biol 26:1921–1934 Prx (Kim et al. 1988). A better understanding of the Goldmark PJ, Curry J, Morris CF, Walker-Simmons MK (1992) function of 1-Cys Prx proteins and their genetic ma- Cloning and expression of an embryo-specific mRNA up-reg- nipulation would allow us to determine the usefulness of ulated in hydrated dormant seeds. Plant Mol Biol 19:433–441 Haslekas C, Stacy RAP, Nygaard V, Culianez-Macia FA, Aalen XvPer1 in the development of stress-tolerant transgenic RB (1998) The expression of a peroxiredoxin antioxidant gene, plants. AtPer1,inArabidopsis thaliana is seed-specific and related to dormancy. Plant Mol Biol 36:833–845 Acknowledgements We thank Ms. Di James (Molecular and Cell Ichimiya S, Davis JG, O’Rourke DM, Katsumata M, Greene MI Biology Department, University of Cape Town) for sequencing (1997) Murine thioredoxin peroxidase delays neuronal apop- XvPer1, Ms. Faeza Davids for assisting with the anti-XvPer1 tosis and is expressed in areas of the brain most susceptible to production, the staff of the Electron Microscope Unit at UCT for hypoxic and ischemic injury. DNA Cell Biol 16:311–321 assisting with the immunofluorescence work as well as Ms. Clare Ingram J, Bartels D (1996) Molecular basis of dehydration toler- Vander Willigen for her kind assistance with immunofluorescence. ance in plants. Annu Rev Plant Mol Biol 47:377–403 Finally we thank Mr. Desmond Barnes for organising and setting Inze´D, Van Montagu M (1995) Oxidative stress in plants. Curr up the phytotrons in the Department of Botany, UCT where the Opin Biotechnol 6:153–158 stress treatments were carried out. We thank the Rockefeller Jin S, Chen CCS, Plant AL (2000) Regulation by ABA of osmotic- Foundation for financial support. stress-induced changes in protein synthesis in tomato roots. Plant Cell Environ 23:51–60 Kang SW, Chae HZ, Seo MS, Kim K, Ivan CB, Rhee SG (1998) References Mammalian peroxiredoxin isoforms can reduce hydrogen peroxide generated in response to growth factors and tumor ne- crosis factor-alpha. J Biol Chem 273:6297–6302 Aalen RB, Opsahl-Ferstad H-G, Linnestad C, Olsen O-A (1994) Kim K, Kim IH, Lee K-Y, Rhee SG, Stadtman ER (1988) The Transcripts encoding an oleosin and a dormancy-related pro- isolation and purification of a specific ‘‘protector’’ protein tein are present in both the aleurone layer and the embryo of which inhibits enzyme inactivation by a thiol/Fe (III)/O2 mixed- developing barley (Hordeum vulgare L.) seeds. Plant J 5:385– function oxidation system. J Biol Chem 263:4704–4711 396 Kowaltowski AJ, Netto LE, Vercesi AE (1998) The thiol-specific Altschul SF, Gish W, Miller W, Meyers EW, Lipman DJ (1990) antioxidant enzyme prevents mitochondrial permeability tran- Basic local alignment search tool. J Mol Biol 215:403–410 sition. Evidence for the participation of reactive oxygen species Baier M, Dietz K-J (1997) The plant 2-Cys peroxiredoxin BAS1 is a in this mechanism. J Biol Chem 273:12766–12769 nuclear-encoded chloroplast protein: its expressional regula- Lee KO, Jang HH, Jung BG, Chi YH, Lee JY, Choi YO, Lee JR, tion, phylogenetic origin, and implications for its specific Lim CO, Cho MJ, Lee SY (2000) Rice 1Cys-peroxiredoxin physiological function in plants. Plant J 12:179–190 over-expressed in transgenic tobacco does not maintain dor- Baier M, Dietz K-J (1999) Protective function of chloroplast 2- mancy but enhances antioxidant activity. FEBS 486:103–106 cysteine peroxiredoxin in photosynthesis. Evidence from Levine A, Tenhaken R, Dixon R, Lamb C (1994) H2O2 from the transgenic Arabidopsis. Plant Physiol 199:1407–1414 oxidative burst orchestrates the plant hypersensitive disease Bewley JD, Oliver MJ (1992) Desiccation-tolerance in vegetative response. Cell 79:583–593 plant tissues and seeds: protein synthesis in relation to desic- Lewis ML, Miki K, Veda T (2000) FePer1, a gene encoding an cation and a potential role for protection and repair mecha- evolutionary conserved 1-Cys peroxiredoxin in buckwheat nisms. In: Somero GN, Osmond CB, Bolis CL (eds) Water and (Fagopyrum esculentum Moench), is expressed in a seed-specific life (1992). A comparative analysis of water relationships at the manner and induced during seed germination. Gene 246:81–91 organismic, cellular and molecular levels. Springer, Berlin Lyu MS, Rhee SG, Chae HZ, Lee TH, Adamson MC, Kang SW, Heidelberg New York, pp 141–160 Jin DY, Jeang KT, Kozak CA (1999) Genetic mapping of six Bray EA (1997) Plant responses to water deficit. Trends Plant Sci mouse peroxiredoxin genes and fourteen peroxiredoxin related 2:48–54 sequences. Mamm Genome 10:1017–1019 Bray EA (2002) Abscisic acid regulation of gene expression during Mundree SG, Farrant JM (2000) Some physiological and molec- water-deficit stress in the era of the Arabidopsis genome. Plant ular insights into the mechanisms of desiccation tolerance in the Cell Environ 25:153–161 resurrection plant Xerophyta viscosa Baker. In: Cherry JH et al Broin M, Cuine S, Peltier G, Rev P (2000) Involvement of (eds) Plant tolerance to abiotic stresses in agriculture: role of CDSP 32, a drought-induced thioredoxin, in the response to genetic engineering, pp 201–222 oxidative stress in potato plants. FEBS Lett 467:245–248 Ndima T, Farrant J, Thomson J, Mundree S (2001) Molecular Chae HZ, Chung SJ, Rhee SG (1994) Thioredoxin-dependent characterization of XVT8, a stress-responsive gene from the peroxide reductase from yeast. J Biol Chem 269:27670–27678 resurrection plant Xerophyta viscosa Baker. Plant Growth Re- Chiantante D, Onelli E, Patrignani G, Scippa GS (1995) Local- gul 35:137–145 ization of a nuclear protein (QP47) in embryonic meristems Netto L, Chae HZ, Kang SW, Rhee SG, Stadtman ER (1996) during seed maturation and germination and its distribution Removal of hydrogen peroxide by thiol-specific antioxidant among crop plants. J Exp Bot 46:815–84 enzymes (TSA) is involved with its antioxidant properties – Dellaporta SL, Wood J, Hicks JB (1983) A plant DNA mini- TSA possesses thiol peroxidase activity, J Biol Chem preparation: version II. Plant Mol Biol Rep 1:19–21 271:15315–15321 Foyer CH, Lelandais M, Kunert KJ (1994) Photo-oxidative stress Niogret MF, Culianez-Macia FA, Goday A, Alba MM, Pages M in plants. Physiol Plant 92:696–717 (1996) Expression and cellular localization of rab28 mRNA and Gaff DF (1971) Desiccation tolerant flowering plants in Southern Rab28 protein during maize embryogenesis. Plant J 9:549–557 Africa. Science 174:1033–1034 Peshenko IV, Singh AK, Shichi H (2001) Bovine eye 1-Cys per- Goday A, Jensen AB, Culinez-Macia FA, Alba MM, Figuetas M, oxiredoxin: expression in E. coli and antioxidant properties. Serrotosa J, Torrent M, Pages M (1994) The maize abscisic J Ocul Pharmacol Ther 17:93–99 acid-responsive protein Rab 17 is located in the nucleus Price AH, Taylor A, Ripley SJ, Griffiths A, Trewavas AJ, Knight and interacts with nuclear localization signal. Plant Cell 6:351– MR (1994) Oxidative signals in tobacco increases cytosolic 360 calcium. Plant Cell 6:1301–1310 726

Raikhel RAP (1992). Nuclear targeting in plants. Plant Physiol Stacy RAP, Nordeng TW, Culianez-Macia FA, Aalen RB (1999) 100:1627–1632 The dormancy-related peroxiredoxin anti-oxidant, PER1, is Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a localized to the nucleus of barley embryo and aleurone cells. laboratory manual, vol 1–3. Cold Spring Harbor Laboratory Plant J 19:1–8 Press, Cold Spring Harbor, NY Vander Willigen C (2001) Comparisons of the resurrection grass, Scott P (2000) Resurrection plants and the secrets of eternal leaf. Eragrostis nindensis, with the related desiccation-sensitive spe- Ann Bot 85:159–166 cies, E. curvula. PhD thesis, University of Cape Town Sherwin HW, Farrant JM (1996) Diﬀerences in rehydration of Vertucci CW, Farrant JM (1995) Acquisition and loss of desicca- three desiccation-tolerant angiosperm species. Ann Bot 78:703– tion tolerance. In: Kigel J, Galilli G (eds). Seed development 710 and germination, Dekker, New York, pp 237–271 Sherwin HW, Farrant JM (1998) Protection mechanisms against Wood AJ, Duﬀ RJ, Oliver MJ (1999) Expressed sequence tags excess light in the resurrection plants Craterostigma wilmsii and (ESTs) from desiccated Tortula ruralis identify a large number Xerophyta viscosa. Plant Growth Regul 24:203–210 of novel plant genes. Plant Cell Physiol 40:361–368 Stacy RAP, Munthe E, Steinum T, Sharma B, Reidunn AB (1996) Wu L, Ueda T, Messing J (1995) The formation of mRNA 3¢-ends A peroxiredoxin antioxidant is encoded by a dormancy-related in plants. Plant J 8:323–329 gene, Per1, expressed during late development in the aleurone and embryo of barley grains. Plant Mol Biol 31:1205–1216