During the last decade the role of ROS in tissue disorders and later in signalling cascades has been intensively studied in connection with various environmental and biotic stresses. The fundamental role of activated oxygen as a non-specific messenger in stress physiology has been widely accepted. Under natural conditions anoxia-induced changes represent a combined effect of two stresses: Anoxic stress itself and oxidative stress, which occurs after the restoration of normoxic conditions. Experimental determination of anoxia-induced changes necessarily implies a reoxygenation period, and hence oxidative stress. Indications of ROS formation under oxygen deprivation ( IV ) lead to the question of interaction and priority between the physiological processes taking place under hypoxia and during reaeration. The investigation of ROS formation (H 2 O 2 ) and LP (TBARS) under hypoxic conditions (10 -5 M of oxygen), suggests that ROS concentration increases already under a very low oxygen concentration, and subsequent normoxia affects only the rate of the process . Indeed, the functioning of plant terminal oxidases (K m 10 -6 M for oxygen) (Skulachev 1997) is not limited by oxygen concentration in the experimental system and the formation of ROS is possible. Non-enzymatic one electron O 2 reduction can also occur at about 10 -4 M and higher oxygen concentrations (Skulachev 1997).
Incubation of the anoxia intolerant plant tissues under oxygen deprivation caused changes in the cell ultrastructure such as plasmolysis and disintegration of organelles. Such changes have been described in detail earlier by Vartapetian and Zakhmilova (1990). The anatomy and ultrastructure of Iris pseudacorus rhizome parenchyma containing masses of storage fructan have been described earlier by Hanhijärvi and Fagerstedt (1994). In this study I. pseudacorus rhizome parenchymal cells did not show any signs of damage on the ultrastructural level even after 45 days of anoxic incubation. According to our investigation and earlier biochemical studies (Chirkova, 1988; Hanhijärvi and Fagerstedt, 1994; Chirkova et al., 1998; Blokhina et al., 1999), the plant species used can be placed in the following order of increasing anoxia tolerance: wheat -.oat - rice - I. germanica - I. pseudacorus .
Under normal metabolic conditions the formation of ROS and peroxidation of membrane lipids are in dynamic equilibrium with the activity of the antioxidant systems. Any changes in environmental conditions may alter this equilibrium and lead to oxidative stress. Such alteration may be achieved either by enhanced ROS formation, or by slowing down its decomposition, i.e. inhibition of antioxidant systems.
H 2 O 2 visualised by transmission electron microscopy ( IV ) is probably of enzymatic origin, considering oxygen concentration in our experiments and the positive effects of the various inhibitors of H 2 O 2 producing enzymes on the amount of H 2 O 2 . An investigation on H 2 O 2 accumulation under root hypoxia in the roots and leaves of Hordeum vulgare has shown two times higher levels of H 2 O 2 in the roots of hypoxic plants as compared with the controls, however, in leaves this parameter was only 18% higher under stress conditions (Kalashnikov et al. 1994). In contrast to the chemical reduction of oxygen, an enzymatic process implies a high level of metabolic regulation, which can be accomplished, in the case of oxygen deprivation stress, by anoxia-specific metabolic changes. The latter can inhibit or stimulate the formation of ROS via indirect mechanisms. A decline in ATP content observed under anoxia (Chirkova et al. 1984, Hanhijärvi and Fagerstedt 1995) increases the probability of ROS formation in the ETC of mitochondria and, at the same time inhibits the energy-dependent step of GSH biosynthesis. The key enzyme in GSH biosynthesis, g -glutamylcysteine synthase (EC 22.214.171.124), requires ATP as a cofactor and has an alkaline pH optimum 8-8.4 (Noctor et al. 1998). The regulatory role of ATP concentration and a threshold dependence of membrane integrity on ATP concentration have been shown (Rawyler et al. 1999) . If the rate of ATP synthesis falls below 10 m molg -1 FWh -1 membrane lipid hydrolysis occurs, thus providing free fatty acids as the substrates for LP. Possible involvement of metabolic parameters in sensing the lack of oxygen (or ROS) and in the induction of the stress response has been proposed (Clement et al. 1998). The intracellular superoxide anion concentration and an increase in the chemical reduction state and acidification of the intracellular environment have been considered as signals for H 2 O 2 -mediated apoptosis (Clement et al. 1998). This particular combination of conditions constitutes a signal and is an indication of reductive as opposed to oxidative stress (Clement et al. 1998).
A decrease in ATP availability during anoxia can slow down ATP-dependent plasma membrane H+-ATPase and lead to the acidification of the cytoplasm and possibly to the alkalinisation of the cell wall. Localisation of H 2 O 2 predominantly in the cell wall under anoxia and reoxygenation ( IV ) (taking into consideration the control treatments with inhibitors) suggests disturbance of the enzymatic decomposition of H 2 O 2 in this compartment. Another reason is the exhaustion of the reductants, especially ascorbate, under anoxia and reoxygenation ( III ), which are necessary for H 2 O 2 utilisation by peroxidases. Acidification of the cytoplasm under anoxia can affect the activity of enzymes involved in the antioxidative defence system. Thus the very first step in antioxidative defence - detoxification of ROS can be affected by anoxia-induced metabolic changes.
In respect to anoxia tolerance the formation of ROS (H 2 O 2 ) appearance was less pronounced and postponed in time in anoxia-tolerant in comparison to anoxia-intolerant species. In the rhizomes of I. pseudacorus , an extremely anoxia-tolerant plant, H 2 O 2 was detected only after 45 days of anoxia, indicating that no damage to the membranes had occurred. This observation correlated positively with conjugated diene formation measured under the same experimental conditions: There are no minima at characteristic wavelengths in I. pseudacorus rhizome samples ( I ) . Only after 45 days of anoxia both cis-trans and trans-trans diene minima appeared in the SD spectrum. Similar results have been obtained for I. pseudacorus rhizomes using EPR (electron paramagnetic resonance) (Crawford et al. 1994). In this species the signal corresponding to free radicals could not be observed in the spectrum, while in the anoxia-intolerant I. germanica the signal was clear. The data suggests that there was no detectable free radical formation, and hence neither diene nor TBARS production emerged in the rhizomes of the anoxia tolerant I. pseudacorus . In our experiments the CD and CT content corresponded with resistance to anoxia. The anoxia-tolerant oat and I. pseudacorus both exhibited very low accumulation of dienes and the highest tolerance to oxidative stress as compared with wheat and I. germanica , respectively.
Data gained by SD spectrophotometry on the formation of different stereoisomers permit us to evaluate the impact of chemical and enzymatic processes in LP development. Hydroperoxides, formed as a result of the autooxidation of lipids, represent a mixed population of stereoisomers with the cis-trans/trans-trans ratio depending on the temperature and substrate concentration (Porter et al. 1980). Another source of hydroperoxides is the lipoxygenase reaction, which leads to specific positional and configurational isomers. The enzyme converts the cis,cis -1,4-pentadiene structure, present in PUFA, to cis, trans conjugated hydroperoxydiene (Vick and Zimmerman 1987). Also tocopherols have been shown to affect the isomeric composition of hydroperoxides. The addition of 5% a -tocopherol led to an absolute formation of cis, trans isomers in the autooxidation of methyl linolenate (Kamal-Eldin and Appelqvist 1996). In our experiments the absence of peaks corresponding to cis, trans dienes in SD spectra of aerated controls in the dark, can be explained by the slowing down of constitutive LP under the lack of light (Tremolieres 1985). According to our calculations from the peak heights in the SD spectrum (Corongiu and Banni 1994), cis, trans dienes are formed in a larger amount under anoxia in wheat and rice than under aeration. It is known that anoxia leads to cytoplasmic acidification (Richard et al. 1994). The pH optimum for many characterized lipoxygenases is acidic (Vick and Zimmerman 1987, Reddy 1991), and hence the probability of cis, trans diene formation increases.
Conditions favouring free radical generation such as low energy charge, high reducing equivalent levels and saturated ETC components, typically exist in anoxic tissue (Van Toai and Bolles 1991). A burst of ROS occurs when aerobic conditions are restored. In terms of membrane damage the post-anoxic burst of oxidative reactions have a more pronounced effect than anoxia itself. In our experiments the degree of such damage correlated positively with the duration of anoxic exposure (Table 1) and negatively with the plant's tolerance to anoxia as estimated by other means than LP parameters i.e. maintenance of AEC and electrolyte leakage (Chirkova et al. 1991). Hence, anoxically treated membranes were more susceptible to ROS attack and lipid peroxidation than aerobic controls ( I, IV ). Since de novo synthesis of lipids needs molecular oxygen, desaturases and large amounts of ATP, it stops under anoxia (Crawford and Brändle 1996). Therefore, the most efficient way to maintain functional membranes under anoxia is to preserve PUFA and lipids instead of their synthesis (Pfister-Sieber and Brändle 1994).
It is a common feature that anoxia induces the formation of free fatty acids. It has been suggested that an increased amount of free fatty acids and a simultaneous decrease in polar lipids may be a common feature to all anoxia-intolerant species under oxygen deprivation stress (Hetherington et al. 1982). It has also been suggested that long-term anoxia-tolerance is dependent on the ability to maintain membrane function and membrane integrity (Chirkova, 1988). In the anoxia-sensitive I. germanica large amounts of free fatty acids are formed under anoxia and simultaneously the amount of polar lipids decreases due to membrane collapse. In contrast, in the anoxia-tolerant Acorus calamus formation of free fatty acids does not increase and the amount of polar lipids does not decrease in long-term anoxia (Henzi and Brändle 1993).
The factors underlying higher membrane stability of anoxia-tolerant plants can include both structural properties of the membranes (the degree of PUFA unsaturation, fluidity, permeability, incorporation of tocopherols) and anoxia-induced metabolic changes such as cytoplasmic acidification, lowered energy charge and elevation of cytosolic Ca 2+ .
Effect of LP on membrane properties and the reasons underlying membrane stability can be approached through a variety of methods: Biochemical characterisation of LP products ( I ), measurements of membrane function, such as barrier and transport properties and direct investigation of membrane structure. LP products have a pronounced effect on membrane orientational order, packing and interaction of phospholipids (Van Ginkel and Sevanian 1994). The formation of hydroperoxides causes the migration of the affected acyl chain from the membrane and tends to increase the area occupied by the phospholipid. Among other effects of lipid peroxides the decrease in molecular orientation order and a tendency for the bilayer structure to collapse should be named. Interestingly, LP products do not affect the fluidity of the membranes (Van Ginkel and Sevanian 1994), but membrane fluidity has been shown to affect the rate of LP propagation stage (Shewfelt and Purvis 1995). Both lipids and proteins are the targets of peroxidative processes, although LP is kinetically favoured. Thus, if the process is regulated strictly by kinetics, peroxidation of PUFA may serve a protective function to prevent a direct attack on proteins (Shewfelt and Purvis 1995). Lipid hydroperoxides increase the hydrophilic properties of the lipid bilayer (Frenkel 1991). A similar change in hydrophilicity has been shown under anoxia in the mitochondrial proteins. The transition of phenylalanyl residues to tyrosyl, providing an increase in hydrophilicity, has been detected under anoxia (Chirkova et al. 1993). This transition could affect synergetic action of tocopherol and ascorbate (Beyer 1994), which occurs on the hydrophobic-hydrophilic interface.
Still another aspect of LP role in stress metabolism should be discussed as the generally accepted approach of LP as a detrimental process is not complete. A role in stress signalling has been recently suggested not only for ROS but also for the oxygenated derivatives of LP (Tarchevskii 1992). 13-hydroperoxylinolenic acid, which is formed by lipoxygenase, is a precursor of jasmonic acid. Jasmonic acid has phytohormone-like activities and is involved in the regulation of developmental processes and in the plant's response to wounding and pathogens (Rosahl 1996). Peroxidation products have been proposed also to act as primary mediators in plants during heat stress (Kurganova et al. 1999).
Examples of cross-resistance can be found between other environmental stresses and oxidative stress. These observations have been used to propose a common basis for resistance against ROS (Alsher et al. 1997). According to Alsher et al. (1997) a first hypothesis could be that cross-resistance occurs between stresses, which originate in the same subcellular compartment, but not between stresses, which have disparate sites of action. However, cell membranes are usually on the first line of defence, and hence they form the most probable compartment for both sensing and resisting the stress.
The antioxidant status of the cells is probably species and tissue specific. The differential response of the antioxidant systems in the plants studied here, indicates the importance of interaction between multiple parameters for the provision of protection: The rate of turnover of redox active ascorbate and glutathione, anoxia-induced metabolic changes and structural and functional properties of the membranes. Superimposed factors such as the duration of the anoxic treatment and the reoxygenation period are significant for the estimation of the defense efficiency, and hence anoxia tolerance. With respect to antioxidant systems, different rates of ROS formation in anoxia-tolerant and -intolerant species should be noted.
The investigation of SOD activity in our experimental plants resulted in contradictory observations ( I and Monk et al. 1989), which can be explained by the different anoxia-tolerance of the species studied and experimental setup. In cereals the activity of SOD declined and was dependent on the duration of the anoxic treatment, while in Iris pseudacorus . a 14-fold increase was observed. The reason for the difference in results possibly originates from the particular experimental conditions, e.g. a prolonged reoxygenation period in the case of Iris spp. In cereal roots activity of the enzyme was determined immediately after anoxia. The formation of ROS already under hypoxic conditions and during the oxidative burst after re-admission of oxygen caused rapid substrate overload of constitutive resources of SOD, while induction was hindered by other factors (e.g. time, activity of downstream enzymes in the ROS-detoxification cascade, inhibition by the product (H 2 O 2 ) and consequences of anoxic metabolism). Observations on SOD activity in different plant species under another stress conditions (drought and salinity) suggest that different mechanisms may be involved in oxidative stress injury (Yu and Rengel 1999). Activation of oxygen may proceed through different mechanisms, not necessarily producing a substrate for SOD. Changes in O 2 electronic configuration can lead to the formation of highly reactive singlet oxygen 1 O 2 . Comparison of drought and water stress effects on tolerant and intolerant wheat genotypes suggests that different mechanisms can participate in ROS detoxification. Water stress did not affect SOD activity, while under drought conditions a significant increase was detected (Sairam et al. 1998). Under oxidative stress conditions combined with cold acclimation of cold-resistant and unresistant wheat cultivars, SOD activity was unaffected by low temperature treatment both in the leaves and in the roots, but plants exhibited higher guaiacol peroxidase activity (Scebba et al. 1998). Inefficiency of ROS detoxificating enzymes (SOD, catalase, ascorbate peroxidase and non-specific peroxidase) has been shown under water deficit-induced oxidative stress in rice (Boo and Jung 1999). A decrease in enzymatic activity was accompanied by LP, chlorophyll bleaching, loss of AA, GSH, a -tocopherol and carotenoids in stressed plants. The authors suggest the formation of a certain strong pro-oxidant, which is neither superoxide nor H 2 O 2 under the conditions of water deficit (Boo and Jung 1999). However, in the roots and leaves of Hordeum vulgare under root hypoxia the activity of SOD increased by 40-60% (Kalashnikov et al. 1994). The ability of plants to overcome oxidative stress only partly relies on the induction of SOD activity because of the following reasons: Diversification of the pathways of ROS formation, compartmentalisation of oxidative processes (charged ROS can not penetrate the membrane) and compartmentalization of SOD isozymes. It is also possible that in different plant species and tissues different mechanisms are involved in the protection against oxidative stress.
The intensity of the oxidative stress, as mentioned above, relies on the interaction of several factors. One of them, antioxidant status of the tissue, is in turn regulated by the following parameters: Constitutive levels of antioxidants, redox state of the cell metabolites, the activity of antioxidant-regenerating enzymes and the conditions for successful antioxidant interaction (e.g. ascorbate-glutathione cycle).
Initial levels of AA and GSH, measured in the experimental plants before treatment did not correlate with the tolerance to anoxia. In cereal roots the differences were minor, while the rhizomes of Iris spp. differed significantly in their constitutive antioxidant levels ( III ). AA content in the rhizomes of the anoxia-tolerant I. pseudacorus was 7 times higher than that in the intolerant I. germanica , while the comparison of GSH content revealed the opposite situation: Anoxia-intolerant I. germanica contained 3 times more GSH than I. pseudacorus . However, these differences did not affect the initial level of LP in these species ( I ). In the roots of cereals the main form of ascorbic acid was DHA, which is probably due to its transport role in these species (discussed in III ). However, the higher content of DHA than AA in the roots affects the redox state of the tissue, antioxidant properties and, probably, anoxia tolerance of cereals. In order to elucidate the uniform relationship between antioxidant status and the intensity of LP the data were calculated as per cent of the control level in plants before treatment (Table I). During the time course of anoxia increased LP, measured as CD, CT and TBARS formation, resulted in a significant decrease of both reduced and oxidised forms of ascorbic acid and glutathione, despite the differences in the initial antioxidant content. The situation reveals severe oxidative damage to anoxically treated membranes and utilisation of antioxidants for ROS scavenging and disturbance of AA and GSH turnover via enzymatic mechanisms (discussed below). Calculation of the correlation coefficients between the parameters of LP and reduced and oxidised forms of the antioxidants was performed only for the treatments with identical experimental setup (anoxia and 2h reoxygenation) (Table 2). A negative correlation has been shown between LP levels and antioxidant content for all anoxic treatments with the exception of the anoxia-tolerant I. pseudacorus . In this species the correlation coefficient between the data sets on TBARS and ascorbate, and TBARS and glutathione was positive, and revealed the suppression of peroxidative processes before the termination stage of LP (TBARS accumulation) ( I ). In the roots of wheat, despite of the relatively low AA and GSH content in comparison with Iris spp., the correlation coefficients were close to -1 in the case of conjugated TBARS/AA (-0.999) and TBARS/GSH (-0.905), emphasising the importance of this redox couple for protection against LP. Similarly, the correlation coefficients calculated for I. pseudacorus between Dienes/GSH (-0.171) and Trienes/GSH (-0.414), probably indicate that these parameters do not significantly depend on each other and that another factor (e.g. high polyphenol content) is more important for the regulation of LP in I. pseudacorus rhizomes.
Imposition of anoxic stress causes several non-specific and anoxia-specific metabolic changes, some of which can interfere with the antioxidant turnover and synthesis, as well as affect non-antioxidant functions of the compounds studied. Consequences of ATP depletion on antioxidant system and LP are discussed later
An increase in the NADH/NAD + -ratio during the first hours of anoxia can play an adaptative role by supplying reducing equivalents for the antioxidant turnover. The participation of NADH in MDHA reduction to ascorbate has been demonstrated (Noctor and Foyer 1998). AA and GSH metabolism (reduction to an active antioxidant form) may provide an additional sink mechanism for excess protons and NADH produced during the first stages of anaerobiosis (Chirkova et al. 1992), therefore explaining the increase in AA and GSH contents observed under anoxia by other authors (Biemelt et al. 1996, 1998). Our results on the depletion of the reduced forms of ascorbate and glutathione suggest an imbalance in the glutathione-ascorbate cycle under anoxia, a condition which affects the redox state of the cell and redox-dependent reactions. A decrease in oxidised forms, DHA and GSSG, suggests that there is no turnover of antioxidants under these experimental conditions, despite the fact that tocopherol (necessary for AA regeneration) was maintained ( III Iris spp. and I rice roots) or even enhanced ( I wheat and oat roots). AA can be eliminated from the Halliwell-Asada cycle via oxidation to DHA and further decomposition to diketogulonic acid (Smirnoff 1995). Enzymatic turnover of ascorbate and glutathione is problematic for the plants under anoxic stress, since a decrease in corresponding enzymatic activities has been observed under prolonged flooding (anoxia) by several authors. Monodehydroascorbate reductase (MDHAR), dehydroascorbate reductase (DHAR) and glutathione reductase (GR) activities decreased slightly or remained unaltered under hypoxia, while anoxia caused a significant inhibition of enzyme activities (Biemelt et al. 1998). Upon the restoration of normoxic conditions, however, the activity of antioxidative enzymes returned to the control level, which was interpreted by the authors as activation of the antioxidative defence. Inhibition of GR, ascorbate peroxidase, catalase and SOD activities has been shown also by Yan et al. (1996) in corn leaves under prolonged flooding. In contrast, many investigators report the induction of active oxygen-processing enzymes under oxygen deprivation stress. In an investigation on the effect of root hypoxia on H 2 O 2 accumulation and the activity of the antioxidative enzymes the induction of SOD, GR, peroxidase and catalase has been shown in the roots and leaves of Hordeum vulgare (Kalashnikov et al. 1994). In these experiments higher levels of enzymatic activity have been detected under hypoxia (120-160% of the control) as compared with the same activities measured after the restoration of normoxic conditions (85-106 % of the control). A systematic study to characterise the specific effects of various environmental stressors (high temperature, enhanced light intensity, water deficiency, chilling and UV-B radiation) on the activities of several antioxidant enzymes in Arabidopsis thaliana under comparable experimental conditions has been undertaken by Kubo et al. (1999). The results reveal differential response of the antioxidant enzymes, although all of the above mentioned stressors are known to induce the production of ROS. The authors suggest that particular antioxidative enzymes have different threshold values of stress needed to increase their activity. This phenomenon may be explained by assuming that there are differences in major ROS produced under specific stress conditions, or there are differences in the cellular sites where ROS are formed (Kubo et al. 1999).
The activity of the antioxidant enzymes and the antioxidant turnover depends to a great extent on experimental conditions, especially in the case of oxygen deprivation stress. Indeed, O 2 availability is a necessary condition for ROS formation and for the induction of protective reactions i.e. ROS-scavenging and antioxidant-regenerating enzymes. Such induction is probably achieved through redox changes and is sensed by an unknown sensor. From that point of view, a decrease in AA/DHA and GSH/GSSG ratios in out experiments represents one of the possible redox signals in anoxically damaged tissue, undergoing oxidative stress. Another reason for the diversification of data on antioxidant content and corresponding enzymatic activity may originate from species- and tissue-specific factors underlying stress tolerance. These factors can include the concentration and compartmentalisation of a particular antioxidant, its functioning at the expense of other antioxidants, properties of the membranes and metabolic parameters, which can vary to a great extent under the same stress situation in different tissues and species.
Under anoxic conditions de novo synthesis of lipids is abolished and the protection against ROS is maintained via preservation of membrane structures and of antioxidant resources. The concentration of a -tocopherol, which has numerous functions in stress protection (i.e. quenching and scavenging of ROS, stabilisation of membrane structures and formation of complexes with free fatty acids), is extremely important for plant survival. High efficiency of the free-radical scavenging function of tocopherols should be supported by its turnover via reduction by ascorbate. During short-term anoxia in our experiments tocopherol levels were maintained ( III Iris spp. and I rice roots) or enhanced ( I wheat and oat roots). At the same time rapid utilisation of AA and GSH was observed ( III ). The formation of LP products detected after anoxia ( II, Table 1) suggests intensive involvement of all antioxidant resources in membrane protection. A clear negative correlation between levels of LP products (CD, CT and TBARS) and antioxidants' reduced forms has been shown in anoxic, but not in control samples (Table 2). Therefore, tocopherol as an efficient and polyfunctional compound was maintained at the expense of AA and GSH, the latter at the same time being involved in radical scavenging in the aqueous phase of the cell. A similar temporal pattern of antioxidant interaction has been shown under ischemia-reperfusion in an isolated rat heart: Ascorbate and glutathione were depleted selectively prior to vitamin E (Haramaki et al. 1998). These results were considered as evidence of the antioxidant network under ischemia-reperfusion.
Lipid peroxidation reactions are dependent on molecular oxygen by definition. The reaction between a tocopherol molecule and PUFA is inhibited in the absence of oxygen (Saucy et al. 1990). A significant decrease in the tocopherol content observed after prolonged anoxia in Iris spp. ( III ) could be attributed to the membrane breakdown rather than antioxidative reactions per se . Indeed, the liberation of free fatty acids observed by Henzi and Brändle (1993) in I. germanica rhizomes under anoxia was interpreted as breakdown of membrane structures. It is possible that a -tocopherol molecules localised in membranes could be destroyed simultaneously with the collapse of the membranes. The absence of the same effect on b -tocopherol content could be partly explained by different localisation of a - and b -tocopherols. The significance of tocopherols for stress protection, and an indirect indication of the possibility of tocopherol synthesis under anoxic conditions, have been demonstrated in the experiments on anaerobically grown rice seedlings (Ushimaru et al. 1994). Their tocopherol level was 3 times higher than that of the aerobic controls and LP products (TBARS) were present at one third of aerobic levels.
However, other than oxygen-centered radicals can be formed and interact with unsaturated lipids. Under anaerobic conditions carbon-centered radicals (pentadienyl or epoxyallylic) can undergo radical-radical coupling with tocopheryl radical (TO · ), thus forming adducts and terminating the reaction (Kamal-Eldin and Appelqvist 1996). When oxygen is present in trace amounts and hydroperoxides are present in negligible concentrations tocopherols can react directly with alkyl radicals (L · ):
|TOH + L · ® TO · + LH||(Kamal-Eldin and Appelqvist 1996).|
Hence, under hypoxic conditions the possibility of free radical reactions exists, and tocopherol can carry out its antioxidant function. a -Tocopherol has been shown also to form stable complexes with free fatty acids (increasing under stress conditions) and lysophospholipids, thus stabilising the membrane structure (Kagan 1989). Therefore, the decrease in tocopherols under anoxia has a detrimental effect on the membranes through a dual mechanism: The deprivation of antioxidant protection and destabilisation of membranes. Oxidative stress causes a decrease in a -tocopherol concentration in a cell because of its radical scavenging function. The decrease in a -tocopherol concentration in turn acts as a signal of the cell redox state in sensitive cells. Thus, a -tocopherol can act as a sensor of the oxidant status of the cell and a transducer of the stress signal (Azzi et al. 1995).
In general, close consideration of anoxia-induced metabolic changes and PTP-triggering factors suggest PTP induction under anoxia. This phenomenon has been described for mammalian mitochondria (Kuzminova et al. 1998) but it has not been shown in any plant mitochondria.
Permeability transition, which was not dependent on a high Ca 2+ - and Pi-concentration, has been observed in yeast mitochondria, and it was cyclosporin A (CsA) insensitive (Jung et al. 1997). High matrix Ca 2+ is generally considered vital for PTP induction and the ability of mitochondria to take up this cation from the surrounding medium can be crucial for pore functioning. Our results on Ca 2+ -uptake by plant mitochondria are in accordance with other observations obtained on plant mitochondria. In purified mitochondria isolated from pea ( Pisum sativum L.) and Jerusalem artichoke ( Heliantus tuberosus L.) accumulation of external free Ca 2+ by means of an electrophoretic uniporter with a K m of approximately 150 nM was observed. This uptake was Pi dependent in the Jerusalem artichoke, but was Pi independent in pea stem mitochondria (Zottini and Zannoni 1993). Mitochondrial Ca 2+ -fluxes have been detected also under anoxia in maize suspension culture cells: A majority of mitochondria released Ca 2+ in response to anoxia and took it up upon reoxygenation. These changes were prevented by ruthenium red, an inhibitor of intracellular Ca 2+ channels (Subbaiah et al. 1998). However, these investigations were not aimed at the Ca 2+ fluxes in connection with the possible induction of PTP, and hence other important mitochondrial parameters were not measured.
The swelling of mitochondria observed in our experiments after Ca 2+ -uptake and energization of the mitochondria with a respiratory substrate (succinate or NADH) was not inhibited by CsA. The Ca 2+ -dependent Ca 2+ -release, characteristic of the pore opening in animal mitochondria, was not detected. These two parameters routinely used to detect permeability transition in animal physiology, failed to reveal the same phenomenon in plant mitochondria. In a similar investigation on yeast mitochondria and the CsA-insensitive, Ca 2+ - and Pi-independent permeability transition, the pore was characterised by means of size calibration (Jung et al. 1997). The changes observed in yeast mitochondria were considered a result of PTP functioning, which was regulated in a different manner from the mammalian PTP. To our knowledge there are no reports on the functioning of PTP in plant mitochondria, however some sensitivity of membrane potential to CsA has been demonstrated in pea stem mitochondria (Petrussa et al. 1992). According to our results PTP was not induced in wheat root mitochondria under the conditions studied here. Swelling occurred after the energisation of mitochondria but Ca 2+ -uptake was not accompanied by membrane potential loss and Ca 2+ -release. This type of swelling can be defined as 'high energy swelling' and is due to electrophoretic uptake of monovalent cations, K + in particular. It is characterised by the preservation of the membrane potential and low inner membrane permeability (Bernardi et al. 1999).
Changes in the mitochondrial volume according to recent findings on mitochondrial ultrastructure (Mannella et al. 1997) and the role of mitochondria in programmed cell death via cytochrome c and apoptogenic protein release (Skulachev 1998), can be considered a signalling event. Passage of water and solutes through the PTP is one of the possible mechanisms of mitochondrial volume regulation. However, other mechanisms can be involved in this process. The contraction of mitochondria, observed in our experiments upon the onset of anoxic conditions, which was preceeded by swelling (i.e. osmotic influx of water and electrophoretic uptake of ions), was probably due to passive extrusion of osmotically active compounds, accumulated during high energy swelling. This process is energy independent, since mitochondrial contraction occurred simultaneously with the membrane depolarisation. However, the physiological role of mitochondrial volume changes under anoxia is unclear and requires further investigation.