The plant species used in the experiments can be arranged in a descending order of anoxia tolerance: I .pseudacorus > I. germanica > Oryza sativa > Avena sativa > Triticum aestivum . Their relative tolerance to anoxia has been estimated in a range of previous investigations according to viability, electrolyte leakage and adenylate energy charge (Chirkova et al. 1991, Hanhijärvi and Fagerstedt 1994, 1995). Only roots of cereals and rhizomes of Iris spp. were used for antioxidant determinations, because under natural conditions these organs are the first to suffer under flooding. Only current year rhizomes were used in the experiments. Before anoxic treatment roots and leaves were removed from the rhizomes .
Seeds of wheat ( Triticum aestivum L. cv. Leningradka), oat ( Avena sativa L. cv. Borrus) or rice ( Oryza sativa L. - cv. VNIIR) were planted in plastic trays and grown at 23 ° /20 ° C (day/night) with a 16-h photoperiod and illumination at 40 molm -2 s -1 for 7 days (wheat, oat and rice). Iris pseudacorus rhizomes were collected locally in a wet riverside meadow. Iris germanica rhizomes were kindly supplied by the Botanical garden of Helsinki University.
Plants (rhizomes or whole seedlings) were placed in glass jars (1.5 L) on moistened filter paper. Anoxic conditions were created with gas generating kits (Oxoid BR 10, Unipath Ltd., Basingstoke, UK) and anaerobic palladium catalysts (Oxoid BR 42). Absence of oxygen was checked with anaerobic indicators (Oxoid BR 55). In addition to releasing hydrogen to remove oxygen, the kit causes carbon dioxide concentration inside the jars to rise to ca. 7- 10%. High CO 2 under anoxia can be considered as a natural condition in soil and is due to root and microbial anaerobic respiration. The oxygen indicators used are sensitive to oxygen concentrations in the aqueous phase only down to 12 m M, which is still enough for the functioning of terminal oxidases and enzymatic production of ROS (Skulachev 1997).
Aerobic control samples were placed into moistened quartz sand (rhizomes) or wrapped in several layers of filter paper (seedlings). Both the anoxic and control samples were kept at room temperature in the dark. In some experiments a two-hour reoxygenation period after the anoxic treatment preceded extractions. Reaeration period (the same for all plant species) was chosen to allow the development the peroxidative reaction, which occurs very quickly after re-admission of oxygen.
This assay is based on the reaction of H 2 O 2 with CeCl 3 to produce electron dense insoluble precipitates of cerium perhydroxides Ce(OH) 2 OOH and Ce(OH) 3 OOH (Bestwick et al. 1997). Experimental plants were treated with 5 mM CeCl 3 and/or inhibitors of H 2 O 2 formation, i.e. 8 m M DPI (diphenyleneiodonium for NADPH oxidase inhibition), 3 mM KCN (peroxidase inhibitor) and 1 mM NaN 3 (catalase inhibitor) and 25 m g/ml (1200 Uml -1 ) of catalase (H 2 O 2 removal). After three washing steps, plants were fixed in 2.5% glutaraldehyde, and H 2 O 2 deposition was visualised by transmission electron microscopy.
Quantification of cerium perhydroxides in electron micrographs was performed by Image Pro ® Plus. The programme differentiated cerium perhydroxide precipitates from the background by the difference in contrast. Calculation of precipitates was based on area and percentage of total area parameters, measured in pixels and as a percentage of the whole image area, respectively. Threshold intensity values and area limits were chosen manually for each calculation, depending on density of cerium perhydroxide accumulation, plant species and electron micrograph quality.
Lipids were extracted by the classical method of Folch et al. (1957) with the modification of Kates (1972) for the extraction of plant tissues. Plant material was homogenized with methanol in a glass Potter homogenizer for 5 min on ice and extracted with two volumes of chloroform (to achieve a chloroform:methanol ratio of 2:1 (v/v)). Contamination with water- and methanol-soluble compounds was avoided by co-extraction with 0.1% (w/v) NaCl followed by centrifugation at 600 g x 10 min. The upper water-methanol phase was aspirated and the procedure repeated three times with 0.1% NaCl:methanol:chloroform (47:48:3, v/v/v) mixture. The final lipid extract was evaporated (+40 o C) under a stream of oxygen-free nitrogen, redissolved immediately in chloroform:methanol (2:1, v/v) and used for conjugated diene and triene assay, second derivative spectrophotometry and inorganic phosphorus determination.
Spectrophotometric detection of conjugated dienes and trienes reflects the presence of fatty acid hydroperoxides in lipid extracts. The procedure was carried out according to Recknagel and Glende (1984). An aliquot of the lipid extract was evaporated under a stream of oxygen-free nitrogen and redissolved in cyclohexane (spectrophotometric grade). UV spectra of lipids were monitored with Shimadzu UV 2100 spectrophotometer (Shimadzu, Kyoto, Japan). The characteristic absorption maxima at 232 nm (conjugated dienes, CD) and 274 nm (conjugates trienes, CT) were measured. Intensity of CD and CT formation was quantified per m g of inorganic phosphorus (P i ) determined by the method of Bartlett modified by Gerlach and Deutike (1963) and expressed as relative units (RU). RU=Abs 233nm (274 nm) / m g P i .
Non-peroxidized lipids exhibit strong absorption in the region of 200-220 nm, and therefore their absorption maximum masks the characteristic absorption maximum of conjugated double bonds. The SD method allows us to extract distinct signals out of shoulders on absorption slopes and to achieve better resolution. In SD spectra of peroxidized lipids signals with a minimum at 233 nm and another minimum at 242 nm have been detected and attributed to trans-trans and cis-trans conjugated dienes, respectively (Corongiu et al. 1986). Lipid extracts were prepared as described above. The following instrument settings were used for SD measurements: bandpass 1 nm; delta wavelength 5 nm; scan speed 11.5 nmmin -1 .
Plant material was extracted with TRIS-HCl buffer (pH 7.4) in the presence of 1.5% Polyclar-AT [w/v] to eliminate polyphenols, which were found to interfere with the assay. After filtration and centrifugation (20 min x 10 000 g, Sorvall) thiobarbituric acid (0.5% [w/v] in 20% [w/v] TCA) was added to an aliquot of the supernatant and the mixture heated in a boiling water bath for 30 min. After cooling and centrifugation (20 min x 10 000 g) absorbance of the supernatant was measured at 532 nm. Malone dialdehyde extinction coefficient (0.156 m M-1cm-1) was used for calculations of the TBARS content (Rubin et al. 1976).
The assay was performed in a photochemical system, containing 1.3 mM riboflavin, 13 mM methionine, 63 m M nitroblue tetrazolium (NBT) and enzyme extract in phosphate buffer, pH 7.6. The method is based on the reduction of NBT by superoxide radicals, produced by photochemistry under constant illumination, and the formation of purple formazan. SOD competes with the photochemical system for superoxide and decreases the amount of NBT reduced. The amount of enzyme, which inhibited NBT reduction by 50% was taken as an activity unit (Giannopolitis et al. 1977).
Vitamin E compounds were extracted by a modified solvent extraction method (Thompson and Hatina 1979). a -, b -, g - and d -tocopherols and a -, b -, g - and d -tocotrienols were determined by normal phase liquid chromatography (NP-HPLC) using the LiChrosorb Si 60 column (5 m m, 250 ´ 4 mm, Merck), an isocratic mobile phase (99.8% hexane and 0.2% 2-propanol), flow rate 1.9 ml/min and column temperature 38 ° C. Sample fluorescence (20 m l injection volume) was detected at 292 nm excitation and 324 nm emission wavelengths. A mixture was prepared from commercial a -, b -, g - and d -tocopherols and a-, b -, g - and d -tocotrienols where the concentration of each isomer was 10 m g/ml hexane. Tocopherol contents were calculated according to the corresponding peak areas.
Identification of the peaks corresponding with tocopherol isomers was verified by preparative HPLC and mass spectrometry (MS). An analytical column (Lichrospher Ò Si 60, 5 m m, 250 ´ 4 mm, Merck), the flow rate of 1.9 ml/min and the injection volume 20 m l and the column temperature of 38 ° C were used during the analysis. The fractions containing probable a-, b - and g -tocopherol were collected and analysed with a semi-preparative column (LiChrospher Ò Si 60, 5 m m, 250 ´ 10 mm, Merck) with the flow rate 11.9 ml/min and the injection volume 200 m l. The fractions containing probable a -, b - and g -tocopherol were concentrated in gaseous argon for MS analysis.
Ascorbic (AA) and dehydroascorbic (DHA) acid content was determined according to Okamura (1980) with the modification of Knörzer et al. (1996). The absorbance at 525 nm was recorded and total ascorbate and AA contents were calculated on the basis of standard curves (AA in the range of 2-16 m g/ml in 5% metaphosphoric acid). DHA content was calculated as the difference between total ascorbate and AA levels.
Glutathione was determined under the following conditions: 100 mM sodium phosphate buffer (pH 7.5), 2.5 mM EDTA, 1 mM 5,5'-dithio-bis(2-nitrobenzoic acid), 0.25 units of GR (from bakers yeast, type III, Sigma; 1 unit = 1 m mol GSSG reduced min -1 at pH 7.6), 0.2 mM NADPH and 20 m l or 50(100) m l of sample (metaphosphoric acid extract) for the GSH and GSSG determinations, respectively (Law et al. 1983). The reaction was initiated by the addition of GR and an increase in the absorbance was followed for 4 min at 412 nm. GSH and GSSG content were calculated from the linear part of the line on the GSH basis, since the reduced and oxidised forms of glutathione have been shown to produce the same standard curves under these assay conditions (Griffith 1980). Standard curves were prepared in the range of 20-200 ng GSHml -1
Since the existing methods, e.g. the one originally designed for potato tubers (Douce et al. 1987), for the separation of mitochondria by fractional centrifugation from wheat roots did not produce required results, the following method was developed. Mitochondria were isolated from the roots of 6-7 day old etiolated wheat seedlings by means of differential centrifugation as follows: Plant material was gently homogenized in 2 volumes of ice-cold extraction medium (Sucrose 0.25 M, EDTA 5 mM, EGTA 1 mM, dithioerythritol 1 mM, BSA 0.1%, Polyclar AT 0.6 % in HEPES-TRIS 10 mM pH 7.4) The homogenate was filtered and squeezed through Miracloth and the mitochondria were immediately separated from the cytoplasmic fraction by centrifugation at 10000 g x 10 min. The resulting crude mitochondrial pellet was resuspended in medium I (Sucrose 0.25 M, EDTA 5 mM, EGTA 1 mM, BSA 0.1% in HEPES-TRIS 10 mM pH 7.4) and centrifuged at 600 g x 5 min to remove nuclei and heavy cell debris. This washing procedure was repeated two times. Washed mitochondria were resuspended in medium II (Sucrose 0.25 M, EGTA 30 m M in HEPES-TRIS 10 mM pH 7.4) and stored on ice. Mitochondrial protein was determined according to Bradford (1976) using BSA as a standard.
O 2 -consumption was monitored at +25 o C by a Clark-type oxygen electrode in 2 ml of continuously stirred medium. The incubation mixture contained 10 mM Hepes/TRIS pH 7.3, 0.25 M sucrose and 30 m M EGTA (the medium was essentially Ca 2+ -free, as checked by EGTA titration in the presence of Arsenazo III. Mitochondrial protein concentration was 0.4 mg/ml, and 5 mM succinate or 1 mM NADH were used as respiratory substrates.
An increase in mitochondrial (or intercristal) volume, caused by permeability transition (PT) or by other physiological reasons, results in diminished light scattering of the mitochondrial suspension at 540 nm, due to osmotic equilibration of solutes and water. The decrease in the light scattering of a suspension of mitochondria was followed at 540 nm (or 570 nm when measured simultaneously with Ca 2+ uptake, since 570 nm is an isosbestic point for Arsenazo III) at +25°C as a function of time. Assay conditions are indicated in the figure legends. In all the experiments an inhibitor (rotenone) of the electron transport chain complex I was used to avoid the effect of pyridine nucleotides on the process studied. A range of factors inducing permeability transition were used in the experiments: high matrix Ca 2+ -concentration, depolarisation of inner mitochondrial membrane (induced by uncouplers: FCCP, DNP) and SH-modifying reagents (phenylarsine oxide). Also inhibitors of mitochondrial ETC and ATP synthesis were used: rotenone (complex I), KCN (cytochrome c oxidase) and oligomycin (ATPase).
A metallochromic indicator of Ca 2+ -concentration, Arsenazo III, was used (10 m M) (Scarpa 1979). Differential absorbance changes (665-685 nm) of Arsenazo III as a function of Ca 2+ -concentration were recorded with a dual beam UV-Visible spectrophotometer (Hewlet-Packard 8452A). Arsenazo III calibration was performed in an incubation medium (Sucrose 0.25 M, EGTA 30 m M in a range of Ca 2+ -concentrations (5-30 m M).
Membrane potential of the inner mitochondrial membrane was measured by recording the spectral shift of the lipophilic cationic dye, safranine O (10 nmol ml -1 ) (Moore et al. 1982) with a dual beam UV-Visible spectrophotometer (HP 8452A). The dye accumulates in mitochondria in response to the generation of a potential (negative inside). The wavelength pair 511-533 nm was used. An increase in absorbance (511-533 nm) correlates positively with the generation of a membrane potential.