Introduction
Seeds in storage age and lose viability with time (Priestley et al., Reference Priestley, McBride and Leopold1980; Gidrol et al., Reference Gidrol, Serghini, Noubhani, Mocouot and Mazliak1989; Walters, Reference Walters1998; Corbineau et al., Reference Corbineau, Gay-Mathieu, Vinel and Côme2002; Walters et al., Reference Walters, Hill and Wheeler2005). Among the processes that may lead to a decrease in seed viability during storage, oxidative damage has been considered a major contributor (Bailly, Reference Bailly2004; Kranner and Birtić, Reference Kranner and Birtić2005). Reactive oxygen species (ROS) play important roles in signalling and stress response, but their production must be strictly controlled by antioxidants (Foyer and Noctor, Reference Foyer and Noctor2005, Reference Foyer and Noctor2009; Bailly et al., Reference Bailly, El-Maarouf-Bouteau and Corbineau2008). Oxidative stress arises if ROS production prevails due to a failure of the antioxidant machinery, and ROS can then cause considerable damage to macromolecules, including lipids, proteins and nucleic acids. Damage to lipids involves peroxidation of membrane lipids, leading to a loss of membrane integrity, increased electrolyte leakage and cellular damage (Stewart and Bewley, Reference Stewart and Bewley1980; Tammela et al., Reference Tammela, Salo-Väänänen, Laakso, Hopia, Vuorela and Nygren2005), and the peroxidation of storage lipids (Gidrol et al., Reference Gidrol, Serghini, Noubhani, Mocouot and Mazliak1989).
Major intracellular water-soluble antioxidants are glutathione (γ-glutamyl-cysteinyl-glycine; GSH) and ascorbate, while tocopherol and carotenoids are lipid-soluble and thus preferentially membrane-bound antioxidants (Noctor and Foyer, Reference Noctor and Foyer1998; Munné-Bosch and Alegre, Reference Munné-Bosch and Alegre2002). Ascorbate is absent in dry, mature orthodox seeds (De Gara et al., Reference De Gara, de Pinto and Arrigoni1997; Tommasi et al., Reference Tommasi, Paciolla and Arrigoni1999; De Tullio and Arrigoni, Reference De Tullio and Arrigoni2003), but GSH is maintained (Kranner and Birtić, Reference Kranner and Birtić2005; Kranner et al., Reference Kranner, Birtić, Anderson and Pritchard2006). In unstressed tissues, GSH is at high concentration and can donate electrons to detoxify free radicals, forming glutathione disulphide (GSSG); GSSG can be re-reduced to GSH by the enzyme glutathione reductase. However, stresses, including seed ageing, are often accompanied by decreasing GSH levels (Noctor and Foyer, Reference Noctor and Foyer1998).
The redox state of the concentration-dependent redox couple GSSG/2GSH is more accurately defined by the glutathione half-cell reduction potential (EGSSG/2GSH), which considers both the concentration of GSH and ratio of GSH/GSSG (Schafer and Buettner, Reference Schafer and Buettner2001; Kranner et al., Reference Kranner, Birtić, Anderson and Pritchard2006). Whereas EGSSG/2GSH appears to be a reliable biochemical marker of seed viability (Kranner et al., Reference Kranner, Birtić, Anderson and Pritchard2006 and citations therein), there is contradictory evidence in the literature as to whether seed tocopherol concentration correlates with viability, even within the same species. The majority of studies have shown that viability loss is correlated with a decrease in tocopherol (particularly α-tocopherol), for example in the seeds of Fagus sylvatica (Ratajczak and Pukacka, Reference Ratajczak and Pukacka2005), Quercus robur (Hendry et al., Reference Hendry, Finch-Savage, Thorpe, Atherton, Buckland, Nilsson and Seel1992), Pinus sylvestris (Tammela et al., Reference Tammela, Salo-Väänänen, Laakso, Hopia, Vuorela and Nygren2005), Acer platanoides (Pukacka, Reference Pukacka1991) and Glycine max (Senaratna et al., Reference Senaratna, Gusse and McKersie1988), and studies on Arabidopsis mutants have verified the importance of tocopherols for seed longevity (Sattler et al., Reference Sattler, Gilliland, Magallanes-Lundback, Pollard and DellaPenna2004). In contrast, other studies have shown no correlation between tocopherol and viability loss in Glycine max (Priestley et al., Reference Priestley, McBride and Leopold1980) or that tocopherol content increased as seed viability declined in Acer pseudoplatanus and Acer platanoides (Greggains et al., Reference Greggains, Finch-Savage, Quick and Atherton2000). Similarly, there are also discrepancies in the literature as to whether electrolyte leakage is a useful marker of seed viability. For example, some studies have shown that increased electrolyte leakage is correlated with seed viability loss (for example, Goel et al., Reference Goel, Goel and Sheoran2003; Fessel et al., Reference Fessel, Vieira, da Cruz, de Paula and Panobianco2006; Mirdad et al., Reference Mirdad, Powell and Matthews2006), others that no correlation exists between electrolyte leakage and seed viability at all (Thapliyal and Connor, Reference Thapliyal and Connor1997) or not at temperatures below 30°C (Fessel et al., Reference Fessel, Vieira, da Cruz, de Paula and Panobianco2006).
Suaeda maritima is a coastal halophyte that has been studied extensively for its ability to tolerate salinity (for example, Flowers, Reference Flowers1972; Yeo and Flowers, Reference Yeo and Flowers1980; Wetson et al., Reference Wetson, Cassaniti and Flowers2008), but its seed biochemistry has received little attention. When assessing the potential of halophytes as alternative crops (Glenn et al., Reference Glenn, Brown and Blumwald1999), knowledge of the seed biochemistry and viability during storage is important. The seeds appear to be oily, but no precise data for seed oil content have been reported so far. In this paper, we provide data for seed oil content and tested whether tocopherol and glutathione concentrations in conjunction with electrolyte leakage are useful biochemical markers of seed viability during the long-term storage of S. maritima seeds.
Materials and methods
Seed material
Seeds of S. maritima (L.) Dumort were collected from Cuckmere Haven, East Sussex, UK, during the years 2005, 2003, 2000, 1999 and 1990 and were stored in plastic boxes at 4°C until experimental work commenced in 2006. Seed moisture content (MC) was determined gravimetrically by drying for 17 h at 103°C (ISTA, 2007) and MC was expressed on a fresh weight (FW) basis: ![MC = [(FW - DW)/FW]\times 100](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20151024042926365-0890:S0960258509990250_inline1.gif?pub-status=live){kind=small}
, where DW is dry weight.
Viability of S. maritima seeds
Seed viability was assessed through germination testing. Four replicates of 25 seeds from each year of collection were germinated on 1% agar-water and incubated at alternating temperatures of 20/5°C and a photoperiod of 8 h light/16 h dark, with the higher incubation temperature during the light period (warm white fluorescent light at a photon flux density of 15 μmol m− 2 s− 1). Germination was measured as radicle protrusion of ≥ 2 mm.
Electrolyte leakage
Electrolyte leakage was measured using a conductivity meter (CM100, Reid and Associates, South Africa). Five replicates of 15 seeds from each collection year were placed into 1.5 ml distilled water and measurements were carried out at 20°C for 18 h. Seeds were then dried for 17 h at 103°C (ISTA, 2007) to calculate the MC as before and results expressed on a DW basis.
Tocopherol analysis
Tocopherols were isolated and separated following the procedure of Pfeifhofer et al. (Reference Pfeifhofer, Willfurth, Zorn, Kranner, Kranner, Beckett and Varma2002). Briefly, three replicates of 50 mg of freeze-dried seeds from each collection year were ground to a fine powder in liquid nitrogen using a pestle and mortar. Ground seeds were extracted in 0.5 ml ice-cold acetone and centrifuged at 13,000 g and 4°C for 40 min. The pellet was resuspended in 0.5 ml acetone and centrifuged as above. The supernatants from the two centrifugation steps were combined, the exact volume recorded, and centrifuged immediately prior to high performance liquid chromatography (HPLC) analysis. α-, γ- and δ-Tocopherols were separated by isocratic reversed-phase HPLC (Jasco, Great Dunmow, Essex, UK), using an RP18 column [HiQsil C18V from KyaTech, Tokyo, Japan, 250 × 4.6 mm internal diameter (i.d.), 5 μm particle size] and methanol as a mobile phase at a flow rate of 1 ml min− 1. Tocopherols were detected with a fluorescence detector (excitation: 295 nm; emission: 325 nm) and identified and quantified using standards (Sigma Aldrich, Poole, Dorset, UK) prepared in acetone.
Glutathione analysis
Three replicates of 30 mg of freeze-dried seeds were ground using a pestle and mortar to a fine powder with liquid nitrogen, extracted in 1 ml of 0.1 M HCl with 0.5% (v/v) Triton X-100 and 30 mg polyvinylpolypyrrolidone and centrifuged at 4°C for 40 min. The supernatant was then used in the procedure detailed by Kranner and Grill (Reference Kranner and Grill1996a) which determines GSH and GSSG based on derivatization with monobromobimane (mBBr). Low-molecular-weight thiols were separated by reversed-phase HPLC on an HiQsil RP18 column (150 × 2.1 mm i.d., 3 μm particle size; KyaTech), and detected fluorimetrically (excitation: 380 nm; emission: 480 nm) with a gradient elution of 0.25% (v/v) acetic acid in distilled water at pH 3.9/methanol. Glutathione was separated from other low-molecular-weight thiols cysteine, cysteinyl-glycine and γ-glutamyl-cysteinyl. Standards of these low-molecular-weight thiols (Sigma Aldrich) at different concentrations were prepared to construct calibration curves. Calculation of EGSSG/2GSH followed the formulas given in Schafer and Buettner (Reference Schafer and Buettner2001) and Kranner et al. (Reference Kranner, Birtić, Anderson and Pritchard2006) using the Nernst equation:
![\begin{eqnarray} E_{GSSG/2GSH} = E^{0\prime } - \frac {RT}{nF}\,ln\,\frac {[GSH]^{2}}{[GSSG]} \end{eqnarray}](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20151024042926365-0890:S0960258509990250_eqnU1.gif?pub-status=live){kind=small}
where R is the gas constant (8.314 J K− 1 mol− 1); T, the temperature in K; n, the number of transferred electrons; F, the Faraday constant (9.6485 × 104 C mol− 1); E0′, standard half-cell reduction potential at pH 7 [E0′GSSG/2GSH = − 240 mV]; [GSH] and [GSSG] are molar concentrations of GSH and GSSG, estimated using MCs. The density of water, approximated as 1 g ml− 1, and the amount of water per gram of seed were used in the calculations of molar concentrations of GSH and GSSG.
Determination of oil content
Seed oil was extracted with supercritical carbon dioxide using an ISCO SFX 3560 fat analyser (ISCO Inc., Lincoln, Nebraska, USA) as described by Seal et al. (Reference Seal, Kranner and Pritchard2008). Three replicates of 0.5 g seeds from each collection year were ground with diatomaceous earth, to a ratio of 1 part seed to 4 parts diatomaceous earth, in an IKA-WERKE A11 basic grinding mill (Staufen, Germany) to a particle size of < 1 mm. Ground samples were analysed for oil content using a two-step extraction at a pressure of 6000 psi and temperature of 80°C, with a control (sunflower oil) for every two samples. Oil was collected on to glass wool and dried to constant weight under vacuum at 70°C for 1 h to remove residual carbon dioxide and water, and oil contents were calculated on a DW basis.
Statistical analysis
Germination data were arcsine square root transformed before testing for significance using one-way analysis of variance (ANOVA) with least significant difference (LSD) post-hoc analysis of means. Pearson's correlation coefficient (R 2) values were also calculated. P values ≤ 0.05 were considered to be statistically significant in both analyses.
Results
Seeds of S. maritima collected at five time points over a 16-year period and stored at 4°C had MCs of between 10 and 13%. Seeds were oily, with values ranging from 18.5 to 24.5% (Table 1), which did not change significantly with increasing storage (one-way ANOVA, P>0.05). After 3 years in storage, seed viability declined significantly (one-way ANOVA, P < 0.05): overall, total germination (mean ± SD) declined from 98 ± 4.0% to 1 ± 2.0% during storage, although the total germination of the 16-year-old seed lot was higher than that of the preceding collection date (Fig. 1). In seeds from all collection dates, no further germination was observed after 63 days from the onset of imbibition.
Table 1 Oil content of S. maritima seeds collected between the years 1990 and 2005. Seeds in storage for 6 and 7 years were pooled due to small sample sizes of seeds. Data are mean±SD of three replicates of 0.5 g of seeds and expressed as % oil (w/w) of seed on a DW basis. There was no significant difference in oil content with time in storage (one-way ANOVA, P>0.05)
Figure 1 Viability (black bars), expressed as total germination, and electrolyte leakage (white bars) of S. maritima seeds during long-term storage at 4°C. Data are mean values and SD. Viability was assessed through germination testing of four replicates of 25 seeds, incubated at 20/5°C with 8 h light during the warm incubation. Electrolyte leakage was measured as conductivity on five replicates of 15 seeds.
With the loss in seed viability, changes in electrolyte leakage (Fig. 1) and the concentration of α-, γ- and δ-tocopherols (Fig. 2) were recorded. The concentration of α-tocopherol declined with time in storage and was positively correlated with viability (R 2 = 0.95, P < 0.01). Across all storage periods, electrolyte leakage was negatively correlated with seed viability (R 2 = 0.64, P = 0.05) and with α-tocopherol (R 2 = 0.64, P = 0.05). δ-Tocopherol was present in only low quantities [ranging from 2 to 15 nmol (g DW)− 1] and both δ- and γ-tocopherol were poorly correlated with viability (R 2 = 0.29 and R 2 = 0.01). α-Tocopherol and seed oil content were not significantly correlated (R 2 = 0.45, P>0.05).
Figure 2 Concentrations of α- (black bars), γ- (grey bars) and δ-tocopherol (white bars) in seeds of S. maritima stored for up to 16 years at 4°C, measured by HPLC in combination with fluorescence detection. Results show mean values and SD of three replicates of 50 mg of seeds.
The concentration of glutathione and EGSSG/2GSH also changed with length of storage time (Fig. 3). The concentration of GSH and GSSG was high in seeds stored for 1 year, with a total glutathione (GSH+GSSG) concentration of 796 nmol (g DW)− 1. However, this vastly reduced to 56 nmol (g DW)− 1 by 3 years in storage, at which stage viability remained high at 99%, and thus neither the concentration of GSH nor of GSSG correlated with viability (both R 2 = 0.39). The proportion of GSSG, expressed as a percentage of total glutathione, did not significantly change during storage (32–36%). In contrast, EGSSG/2GSH correlated well with viability throughout storage (R 2 = 0.72, P < 0.05). GSH, GSSG and EGSSG/2GSH did not correlate with electrolyte leakage (R 2 = 0.04, 0.04, 0.23, respectively).
Figure 3 Concentration of glutathione (GSH; white bars) and glutathione disulphide (GSSG; black bars), measured by HPLC in combination with fluorescence detection, and the glutathione half-cell reduction potential (EGSSG/2GSH; closed dots) during prolonged seed storage at 4°C. Data are mean values and SD of three replicates of 30 mg of seeds.
Discussion
Four isoforms of tocopherol (α, γ, β, δ) are found in plants, of which α-tocopherol, ‘vitamin E’, is essential for human nutrition, while the other isoforms are considered α-tocopherol precursors. The role and distribution of tocopherol isomers in plant organs is, however, poorly understood. Seeds of many species, such as Arabidopsis, Glycine max and Scots pine, contain higher concentrations of γ-tocopherol than α-tocopherol (Priestley et al., Reference Priestley, McBride and Leopold1980; Simontacchi et al., Reference Simontacchi, Sadovsky and Puntarulo2003, Sattler et al., Reference Sattler, Gilliland, Magallanes-Lundback, Pollard and DellaPenna2004; Tammela et al., Reference Tammela, Salo-Väänänen, Laakso, Hopia, Vuorela and Nygren2005). In contrast, α-tocopherol was the most abundant form in the seeds of S. maritima, followed by γ-tocopherol and δ-tocopherol. Seeds were also found to be high in oil, equivalent to values found in seeds of Glycine max (Seal et al., Reference Seal, Kranner and Pritchard2008). Their high oil content and α-tocopherol concentration suggests that the seeds of S. maritima may have useful nutritional and commercial value as an alternative crop, warranting further investigations, including those into the seed oil composition and toxicity, to assess suitability for use.
During the storage of S. maritima seeds at a MC of 10–13%, electrolyte leakage was lower in seeds stored for 16 years than for the preceding storage time, but this was accompanied by a higher viability and concentration of α-tocopherol and a lower EGSSG/2GSH. The negative correlation between electrolyte leakage and viability (Fig. 1), indicative of changes in membrane permeability, was as reported for Shorea robusta (Chaitanya and Naithani, Reference Chaitanya and Naithani1994), Pinus sylvestris (Tammela et al., Reference Tammela, Salo-Väänänen, Laakso, Hopia, Vuorela and Nygren2005) and Fagus sylvatica (Pukacka and Wójkiewicz, Reference Pukacka and Wójkiewicz2003; Pukacka and Ratajczak, Reference Pukacka and Ratajczak2007) seeds. The concentration of α-tocopherol was significantly correlated with electrolyte leakage (Fig. 2) in agreement with the role of α-tocopherol in stabilizing the phospholipid bilayer and scavenging lipid peroxyl radicals, thus protecting from lipid peroxidation (Munné-Bosch and Alegre, Reference Munné-Bosch and Alegre2002). In contrast, GSH (both concentration and EGSSG/2GSH) was poorly correlated with electrolyte leakage, as would be expected from an intracellular water-soluble antioxidant that is less likely to be involved in the protection of membranes than lipid-soluble antioxidants.
With a loss of membrane integrity, subsequent cellular damage may become lethal and negatively impact upon seed viability. Oily seeds, such as S. maritima, are more prone to lipid peroxidation (Wilson and McDonald, Reference Wilson and McDonald1986; Bailly, Reference Bailly2004), so a decrease in the concentration of α-tocopherol may lower the ability to protect the storage lipids from degradation, thus diminishing the quality of this vital energy reserve for germination (Sattler et al., Reference Sattler, Gilliland, Magallanes-Lundback, Pollard and DellaPenna2004) and resulting in poor germination. In S. maritima, a severe loss of α-tocopherol resulted in extremely poor viability, in agreement with data reported on Fagus sylvatica embryonic axes and cotyledons during natural and artificial ageing (Pukacka and Wójkiewicz, Reference Pukacka and Wójkiewicz2003; Ratajczak and Pukacka, Reference Ratajczak and Pukacka2005; Pukacka and Ratajczak, Reference Pukacka and Ratajczak2007), Pinus sylvestris during natural ageing (Tammela et al., Reference Tammela, Salo-Väänänen, Laakso, Hopia, Vuorela and Nygren2005), and Quercus robur embryonic axes during desiccation (Hendry et al., Reference Hendry, Finch-Savage, Thorpe, Atherton, Buckland, Nilsson and Seel1992) which all documented lipid peroxidation and a decrease in α-tocopherol with viability loss.
It is apparent from the literature that a general correlation between tocopherol concentration and seed viability cannot be formulated, which may be explained by the analytical methods or seed treatments used, i.e. different seed moisture contents or conditions of seed desiccation and storage (Wilson and McDonald, Reference Wilson and McDonald1986; Greggains et al., Reference Greggains, Finch-Savage, Quick and Atherton2000; Pukacka and Ratajczak, Reference Pukacka and Ratajczak2007), producing variation in the severity of oxidative stress within the seed. Here, care was taken to minimize the temperature effects of the analytical method that may promote lipid peroxidation in vitro (Chappell and Cohn, Reference Chappell and Cohn2007). It is also unlikely that the biochemical changes seen are a reflection of the maturity status of the seed, as no significant change in oil content was detected between the collection years and no significant correlation was apparent between α-tocopherol concentration and oil content. Thus the data provide strong evidence that α-tocopherol is a good marker of viability in the long-term storage of S. maritima seeds at 4°C.
The concentrations of either GSH or GSSG alone, or percentage GSSG did not correlate with viability. However, these measures do not necessarily represent the oxidative state of the tissue correctly if taken independently (Schafer and Buettner, Reference Schafer and Buettner2001), because GSSG/2GSH is a concentration-dependent redox couple. Hence, when the concentrations of GSH and GSSG were considered as part of EGSSG/2GSH, a close correlation was seen with viability. Seed viability was lost at EGSSG/2GSH between − 180 and − 160 mV, which has been described as the zone of EGSSG/2GSH where plant, fungal and human tissues lose viability (Kranner et al., Reference Kranner, Birtić, Anderson and Pritchard2006). However, the concentrations of both GSH and GSSG decreased considerably, accompanied by an increase in EGSSG/2GSH by 18 mV, even in seeds that had been stored for 3 years despite maintained viability and the concentration of α-tocopherol remaining high. This decrease in GSH and GSSG prior to seed viability loss may be interpreted as an early warning signal diagnostic of a failing protective system. It has been suggested that changes in EGSSG/2GSH by less than 70 mV are reversible, while changes above this threshold correlate with cell death (Schafer and Buettner, Reference Schafer and Buettner2001; Kranner et al., Reference Kranner, Birtić, Anderson and Pritchard2006). In addition, low GSSG concentrations may also increase signalling sensitivity and GSSG may act as an oxidative signal through protein interactions and modifications such as glutathionylation (Kranner and Grill, Reference Kranner and Grill1996b; Rouhier et al., Reference Rouhier, Lemaire and Jacquot2008; Foyer and Noctor, Reference Foyer and Noctor2009). Although α-tocopherol also has an important role in plant signalling in response to stress conditions, this seems to be largely associated with photosynthetic tissue (Munné-Bosch et al., Reference Munné-Bosch, Weiler, Alegre, Müller, Düchting and Falk2007) and more research is needed to understand the signalling role of tocopherols in seeds.
In conclusion, viability of S. maritima seeds stored long-term at 4°C was strongly correlated with changes in α-tocopherol concentrations and EGSSG/2GSH, and weakly correlated with changes in electrolyte leakage. Previous studies with mutants have demonstrated the cause–effect relationship between a lack of tocopherol and seed viability (Sattler et al., Reference Sattler, Gilliland, Magallanes-Lundback, Pollard and DellaPenna2004), and between low concentrations of GSH and embryo lethality (Cairns et al., Reference Cairns, Pasternak, Wachter, Cobbett and Meyer2006) and inhibited root and shoot development during postembryonic development after germination (Vernoux et al., Reference Vernoux, Wilson, Seeley, Reichheld, Muroy, Brown, Maughan, Cobbett, Van Montagu, Inzé, May and Sung2000; Reichheld et al., Reference Reichheld, Khafif, Riondet, Droux, Bonnard and Meyer2007). Conversely, it has been shown that overexpression of GSH paradoxically causes oxidative stress (Creissen et al., Reference Creissen, Firmin, Fryer, Kular, Leyland, Reynolds, Pastori, Wellburn, Baker, Wellburn and Mullineaux1999). Therefore, more studies are needed to clearly distinguish between the cause–effect relationship of antioxidants and seed viability in S. maritima.
Both glutathione and α-tocopherol are closely linked with a range of antioxidants and ROS-processing enzymes that are involved in plant responses to abiotic stress (Kanwischer et al., Reference Kanwischer, Porfirova, Bergmüller and Dörmann2005; Munné-Bosch, Reference Munné-Bosch2005). For halophytes, which have to withstand extreme environmental conditions, these antioxidants are likely to have an equally important role during the development of the plant in addition to maintaining seed viability. Regarding their use as seed viability markers, in S. maritima changes in EGSSG/2GSH preceded viability loss, suggesting that EGSSG/2GSH is an earlier stress marker than α-tocopherol concentrations and electrolyte leakage. Our data suggest that the GSSG/2GSH redox couple is involved in early stress perception, signalling and response, while changes in α-tocopherol concentrations and electrolyte leakage, together with a further shift in EGSSG/2GSH towards more positive (i.e. more oxidizing) values, appear to be linked to progressive seed deterioration.
Acknowledgements
The authors would like to thank Dr Anne Wetson for collecting and supplying seeds. The Millennium Seed Bank Project is supported by the Millennium Commission, The Wellcome Trust, Orange PLC and Defra. The Royal Botanic Gardens, Kew, receive grant-in-aid from Defra.
