Plant, Cell and Environment (2009)
doi: 10.1111/j.1365-3040.2009.02053.x
Extracellular superoxide production, viability and redox poise in response to desiccation in recalcitrant Castanea sativa seeds pce_2053
1..17
THOMAS ROACH1,2, RICHARD P. BECKETT3, FARIDA V. MINIBAYEVA4, LOUISE COLVILLE1, CLAIRE WHITAKER3, HONGYING CHEN1,5, CHRISTOPHE BAILLY6 & ILSE KRANNER1 1
Seed Conservation Department, Royal Botanic Gardens Kew, Wakehurst Place, Ardingly, West Sussex RH17 6TN, UK, Research Department of Genetics, Evolution and Environment, University College London, Gower Street, London WC1E 6BT, UK, 3School of Biological and Conservation Sciences, University of KwaZulu-Natal, Private Bag X01, Pietermaritzburg, Scottsville 3209, Republic of South Africa, 4Kazan Institute of Biochemistry and Biophysics, Russian Academy of Sciences, P.O. Box 30, Kazan 420111, Russian Federation, 5Chinese Academy of Sciences, Kunming Institute of Botany, Kunming 661100, People’s Republic of China and 6UPMC University of Paris 06, EA2388, Physiologie des semences, Site d⬘Ivry, Boîte courrier 152, 4 place Jussieu, F-75005 Paris, France 2
ABSTRACT
INTRODUCTION
Reactive oxygen species (ROS) are implicated in seed death following dehydration in desiccation-intolerant ‘recalcitrant’ seeds. However, it is unknown if and how ROS are produced in the apoplast and if they play a role in stress signalling during desiccation. We studied intracellular damage and extracellular superoxide (O2· -) production upon desiccation in Castanea sativa seeds, mechanisms of O2· - production and the effect of exogenously supplied ROS. A transient increase in extracellular O2· - production by the embryonic axes preceded significant desiccation-induced viability loss. Thereafter, progressively more oxidizing intracellular conditions, as indicated by a significant shift in glutathione half-cell reduction potential, accompanied cell and axis death, coinciding with the disruption of nuclear membranes. Most hydrogen peroxide (H2O2)-dependent O2· - production was found in a cell wall fraction that contained extracellular peroxidases (ECPOX) with molecular masses of ~50 kDa. Cinnamic acid was identified as a potential reductant required for ECPOX-mediated O2· - production. H2O2, applied exogenously to mimic the transient ROS burst at the onset of desiccation, counteracted viability loss of sub-lethally desiccation-stressed seeds and of excised embryonic axes grown in tissue culture. Hence, extracellular ROS produced by embryonic axes appear to be important signalling components involved in wound response, regeneration and growth.
Recalcitrant seeds are desiccation-sensitive (Roberts 1973; Dickie & Pritchard 2002) and maintain high water contents and metabolic rates from seed maturation until germination. It has been suggested that desiccation disrupts their metabolism, leading to the accumulation of potentially harmful reactive oxygen species (ROS) such as hydrogen peroxide (H2O2), superoxide (O2· -), singlet oxygen (1O2) and the hydroxyl radical (·OH) (Hendry 1993; Smirnoff 1993; Côme & Corbineau 1996; Varghese & Naithani 2002; Bailly 2004; Kranner & Birtic´ 2005; Pukacka & Ratajczak 2006; Berjak & Pammenter 2008). Conversely, ROS have essential roles in signalling and stress response (Foyer & Noctor 2005; Bailly, El-Maarouf-Bouteau & Corbineau 2008; Foyer & Noctor 2009). Importantly, extracellular ROS production is a typical component of plant response to abiotic stresses such as mechanical injury (Minibayeva, Kolesnikov & Gordon 1998; de Bruxelles & Roberts 2001; Leon, Rojo & Sanchez-Serrano 2001; Minibayeva et al. 2009). In seeds and seedlings, extracellular ROS formation has been observed in response to avirulent elicitors (Morkunas, Bednarski & Kozlowska 2004), wounding and desiccation (Roach et al. 2008). The latter study showed that excision of embryonic axes from recalcitrant Castanea sativa (sweet chestnut) seeds induced a burst of extracellular O2· - production that was modulated by post-excision desiccation. In orthodox (i.e. desiccation tolerant) seeds, elevated rates of ROS production upon seed imbibition have been suggested to defend the emerging seedling against pathogens (Schopfer, Plachy & Frahry 2001). Furthermore, it has been demonstrated that the ·OH radical is involved in cell wall loosening during cress seed germination and elongation growth (Müller et al. 2009). However, the potential role of extracellular ROS production in stress signalling in recalcitrant seeds has received almost no attention.
Key-words: Castanea sativa; germination; glutathione; hydrogen peroxide; peroxidase; phenolic acid; reactive oxygen species; seed; superoxide.
Correspondence: I. Kranner. Fax: +44 1 1444 894110; e-mail:
[email protected] © 2009 Blackwell Publishing Ltd
1
2 T. Roach et al. Previously, we studied extracellular production of O2· - by excised embryonic axes, confirmed the identity of O2· - using electron spin resonance (ESR) and showed that there is a complex interaction between excision and subsequent drying of these isolated axes (Roach et al. 2008). The first aim of the present study was to gain further and more complete insights into ROS and antioxidant metabolism regarding the relationship between the point of desiccationinduced viability loss, cellular redox balance and the pattern of extracellular O2· - production. In contrast to the previous study, embryonic axes were desiccated in intact Castanea sativa seeds such as may naturally occur following harvesting or in the field before suitable germination conditions existed. Changes in the glutathione (GSH)/glutathione disulphide (GSSG) couple were assessed in conjunction with visualization of nuclear membrane integrity, germination testing and assessment of the post-desiccation ability of excised axes to grow in tissue culture. GSH is a major watersoluble antioxidant (Foyer & Noctor 2005). The glutathione half-cell reduction potential (EGSSG/2GSH) is apparently representative of changes in the intracellular ‘redox environment’ and consequently, it is indicative of oxidative stress (Schafer & Buettner 2001; Kranner et al. 2006). Redox changes and nuclear membrane integrity were studied in the cotyledons and embryonic axes separately to assess if viability loss in either or both, in response to progressive desiccation, was responsible for the failure of seeds to germinate. The second aim was to study the mechanisms of O2· production in sweet chestnut seeds. Plants possess a variety of ROS-producing enzymes, including NAD(P)H oxidases (Doke 1985) and cell wall (class III) peroxidases (Bolwell et al. 1995). The role for class I peroxidases (e.g. ascorbate peroxidase) in H2O2 scavenging is well established (Wojtyla et al. 2006), but peroxidases, in particular class III peroxidases, not only scavenge, but can also produce ROS (Passardi et al. 2005). Evidence is emerging that peroxidases may contribute to O2· - production in orthodox radish seeds (Schopfer et al. 2001) and Arabidopsis roots (Dunand, Crevecoeur & Penel 2007). It has also been proposed that NADPH oxidases are involved in ROS production in seed germination (Liu et al. 2007) and dormancy breaking (Oracz et al. 2009). However, the role of extracellular peroxidases (ECPOX) in O2· - production in response to water stress in recalcitrant seeds has not been investigated, and has been given special attention in this paper. O2· --producing enzymes will require reductants such as NAD(P)H. Peroxidase-mediated breakdown of H2O2 involves oxidation of the reactive haem group (Fe+++) of the enzyme by H2O2 to form compound I. Through two single electron reductions, reductants (e.g. NADH) reduce compound I to compound II and re-reduce Fe+++, also producing reductant radicals that can reduce 3O2 to O2· - (Halliwell 1978). However, the presence of NADH has so far not been unequivocally demonstrated in plant apoplasts. Other putative reductants include phenolic acids. Certain ECPOX use H2O2 to oxidatively cross-link lignol molecules when secondary cell
walls are formed (Wallace & Fry 1999). The lignol temporarily becomes a phenoxyl radical that can spontaneously link with another lignol molecule initiating a polymerization event (Ros Barceló et al. 2004), thereby strengthening the cell wall where lignols are often ester-linked to embedded polysaccharides (Ishii 1997). Therefore, seed leachates were analysed by gas chromatography–mass spectrometry (GC-MS) for the presence of phenolic acids, and their ability to stimulate O2· - production was compared with that of NADH. As discussed above, ECPOX are promiscuous enzymes, so most ECPOX can utilize NADH (Halliwell & Gutteridge 1999). We therefore used NADH to compare the effectiveness to stimulate O2· - production of the phenolic acids that likely occur in seed apoplasts. Thirdly, hypothesizing that O2· - production may be beneficial, the effects of exogenous application of H2O2, the product of O2· - dismutation, on viability were investigated in both desiccation-stressed whole seeds and isolated axes grown in tissue culture.
MATERIALS AND METHODS Chemicals, seed material, desiccation and germination tests Analytical grade chemicals were purchased from Sigma (St. Louis, MO, USA), Fisher (Loughborough, Leicestershire, UK) or Fluka (Buchs, Switzerland), and all solutions were made up with distilled deionized water unless indicated otherwise. ‘Broad Range’ molecular mass markers were obtained from Bio-Rad (Hercules, CA, USA). Castanea sativa Mill. seeds grown in Italy were purchased in autumn 2007 (from Corpo Forestale dello Stato Centro Nazionale per lo Studio e la Conservazione della Biodiversità Forestale, Peri-Vr Italy) and kept at 5 °C for up to 12 weeks in a plastic bag until used. The pericarp was removed prior to desiccation. Only seeds with a healthy appearance (i.e. no apparent fungal or bacterial infection, no discoloration and no seed predation, e.g. by maggots, visible) were used in the experiments. Seeds were desiccated at 15% relative humidity and 15 °C for 0 (‘control’), 2, 5, 10 and 21 d. Seed or axis dry weight (DW) was determined gravimetrically (n = 10) after drying at 103 °C for 17 h and water content (WC) expressed on a fresh weight basis. Seed viability was expressed as percentage of total germination. To prevent cross contamination, individual seeds were placed in 40 mm diameter plastic pots containing 20 mL of 0.8% (w/v) agar, pH 7.0. Pots were incubated at 25 °C and an 8 h day (warm white fluorescent light at a photon flux density of 15 mmol m-2 s-1)/16 h night cycle. Successful germination was defined as radicle emergence by at least 10 mm, and scored regularly until all seeds had either germinated or become heavily infected and started disintegrating. Each treatment comprised of five replicates of eight seeds. The viability of embryonic axes was expressed as a percentage of axes that showed root elongation by more than 10 mm after 21 d in aseptic tissue culture rather than callus © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
Superoxide production in Castanea sativa seeds 3 formation alone. In addition, shoot elongation and increase in axis DW were recorded. Embryonic axes were isolated from seeds, sterilized in 50 mm sodium dichloroisocyanurate for 10 min, rinsed three times and then soaked in H2O for 1 h. Castanea sativa was used as model species in this study because it lacks complications (Corredoira et al. 2004) that are often associated with tissue culture protocols of other recalcitrant seeds (Berjak & Pammenter 2008), provided that they are sufficiently sterilized before tissue culturing. The culture medium was McCown’s Woody Plant Basal Salt Mixture with 50 mg L-1 myo-inositol and 0.25 mg L-1 thiamine, pyridoxine and pantothenate (Norstog 1973), 30 g L-1 sucrose and 0.8% (w/v) agar, pH 5.6.Axes were kept in the dark at 25 °C for the first 7 d, then moved into the light (as above) and transferred to new 9 cm diameter Petri dishes after 10 d. Each treatment comprised three replicates of 15 axes, using a 2 cm3 chamber for each individual axis.
Visualization of nuclei Sections of embryonic axes and cotyledons were stained with 4′,6′-diamino-2-phenylindole (DAPI) and propidium iodide (PI) to distinguish cells with intact nuclei from cells with impaired nuclear membranes, corresponding to live and dead cells (Trost & Lemasters 1994). Binding of PI to nucleic acids increases fluorescence by 20 to 30-fold. PI is membrane impermeable, so staining is restricted to dead or severely damaged cells with leaky nuclear membranes. DAPI is membrane permeable and stains nucleic acids in live and dead cells, revealing the location and number of nuclei within the sections.
Determination of GSH, GSSG and EGSSG/2GSH For each desiccation interval, four replicates of five seeds each were analysed. From each seed, the embryonic axes (~28 mg DW) and a segment of cotyledon (~35 mg from the periphery of a cotyledon, taking care that always the same area was sampled) were used. Embryonic axes and cotyledon tissues were immediately frozen in liquid nitrogen, stored at -70 °C for up to 4 weeks and then freeze-dried. Freeze-dried cotyledons were ground to a fine powder in a hermetically closed, liquid nitrogen-cooled Teflon vessel using a Retsch MM200 laboratory mill. The powder was stored at -70 °C in humidity-proof vials until use. Embryonic axes were ground in liquid nitrogen using mortar and pestle. Low-molecular-weight (LMW) thiols were measured according to Kranner & Grill (1996). Briefly, ground samples were extracted in ice cold 0.1 m HCl. Total glutathione (GSH + GSSG) was determined after reduction of GSSG by dithiothreitol (DTT) followed by labelling with monobromobimane (MB). GSH was separated on a RP-18 column from other LMW thiols by reversed-phase HPLC, and detected with a fluorescence detector (excitation: 380 nm, emission: 480 nm). GSSG was measured after blocking of GSH with N-ethylmaleimide (NEM), followed by removal of excess NEM, reduction of GSSG to GSH © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
with DTT and labelling of thiol groups with MB as above. The EGSSG/2GSH was calculated using the Nernst equation, following the procedure of Schafer & Buettner (2001) as described by Kranner et al. (2006), using seed or embryonic axis WC to estimate molar concentrations of GSH and GSSG:
RT [GSH ] ln nF [GSSG] 2
EGSSG/2GSH = E 0′ −
where R is the gas constant (8.314 J K-1 mol-1); T, temperature in K; n, number of transferred electrons; F, Faraday constant (9.6485 ¥ 104 C mol-1); E0′, standard half-cell reduction potential at pH 7 [E0′GSSG/2GSH = -240 mV (Rost & Rapoport 1964)]; [GSH] and [GSSG] are molar concentrations of GSH and GSSG, estimated using WCs. The density of water, approximated as 1 g mL-1 and the amount of water per g seed were used in the calculations of molar concentrations GSH and GSSG.
Visualization of O2· - and H2O2 production in embryonic axes tissues Areas of ROS production were identified by incubating sections of embryonic axes in either 5 mm nitroblue tetrazolium (NBT) for 5 min or 3,3′-diaminobenzidine (DAB) for 10 min. Dark purple staining with NBT indicates the presence of O2· -. Brown staining indicates DAB polymerization, requiring H2O2 and peroxidase activity. One cut surface of an embryonic axis was stained with NBT and the opposite surface with DAB so that O2· - and H2O2 production were directly comparable.
Extracellular O2· - production by embryonic axes Extracellular O2· - production was determined colorimetrically at A490 following epinephrine oxidation to adrenochrome, using an extinction coefficient of 4020 m-1 cm-1 (Misra & Fridovich 1972; Takeshige & Minakami 1979). The identity of O2· - in leachates of sweet chestnut embryonic axes was confirmed in our earlier study (Roach et al. 2008) using ESR, so no further ESR studies were conducted here. Embryonic axes excised from seeds desiccated for 0, 2, 5, 10 and 21 d were rinsed for ~20 s in H2O to cleanse the wound surface from cytoplasmic contaminants (including intracellularly produced ROS). The axes were then gently shaken in 1.5 mL of 1 mm epinephrine (pH 7.0) for 1 h. The incubation solution was centrifuged at 13 000 g for 5 min before spectrophotometric measurements. Each desiccation interval comprised four replicates of five axes each. Extracellular O2· - production in relation to axis growth was recorded after 0, 5, 10, 15 and 21 d of tissue culture. Healthy seedling axes grown from non-desiccated embryonic axes were rinsed briefly in H2O to remove all growth medium before incubation in epinephrine for 30 min
4
T. Roach et al.
(n = four replicates of five axes each). Absorption recorded in the presence of 250 units mL-1 superoxide dismutase (SOD) was regarded as non-specific and subtracted from the values measured in the absence of SOD. Inhibition of O2· - production was tested using diphenylene iodonium (DPI), an inhibitor of NAD(P)H oxidase (Henderson & Chappell 1996), and sodium azide (NaN3), an inhibitor of haemoproteins such as peroxidases (Liu et al. 2006). Embryonic axes were excised from nondesiccated seeds and immediately placed in either DPI (1 mm and 10 mm) or NaN3 (0.1 mm and 1 mm) for 10 min, followed by incubation in epinephrine for 10 min. Rates of O2· --production were measured as described above (n = three replicates of five axes each). To assess the involvement of O2· --producing enzymes released from the cell wall into the extracellular medium, leachates were derived by incubating axes in H2O for 1 h and O2· - production was measured for 30 min after removal of the axes. The ability of released 2,2′-azino-bis 3-ethylbenzothiazoline-6-sulfonic acid (ABTS)-dependent enzymes (e.g. peroxidases) to produce O2· - was measured in the presence or absence of 0.5 mm NaN3. The effect of DPI on O2· - production by the leachates was not tested, because the gp91phox subunit of the NAD(P)H oxidase is an integral membrane-bound protein (Keller et al. 1998) that may be involved in in vivo O2· - production in cells on the axis surface but is unlikely to be leaked into the incubation medium.
Cell wall fractionation Cell walls were isolated and proteins fractionated by a modification of the method of Rast et al. (2003). Embryonic axes and cotyledons were separated, freeze-dried and 0.2 g and 1.0 g, respectively, of DW material used for each replicate. The material was finely ground in liquid nitrogen, mixed using a Vortex mixer with 10 mL of ice-cold 0.25 m Tris-HCl buffer at pH 8.0 and centrifuged. All centrifugation steps were carried out at 4000 g for 15 min at 4 °C. The first supernatant contained the crude cell extract (‘C’), and the pellet contained the cell wall fraction. The cell wall pellet (CWP) was re-suspended in 12 mL of 50 mm sodium phosphate buffer at pH 7.0 and centrifuged. The resulting ‘B1’ supernatant contained loosely bound proteins attached by hydrogen bonds. This step was repeated three times, and the supernatants combined. Proteins bound to the cell wall by hydrophobic interactions were then isolated by gently stirring the CWP from the previous step with 10 mL of phosphate buffer containing 0.3% (w/v) digitonin for 3 h over ice, followed by centrifugation (‘B2’; repeated twice). The CWP was re-suspended in 10 mL of phosphate buffer containing 2 m NaCl, gently stirred for 3 h over ice and centrifuged, releasing proteins bound by ionic bonds into the supernatant (‘B3’; repeated twice). Proteins bound by covalent linkages that remained on cell wall fragments (‘B4’) were re-suspended in 5 mL of phosphate buffer.
H2O2-metabolizing enzymes and O2· - production in cell wall fractions, and assessment of cytoplasmic contamination Fractions were tested separately for H2O2-scavenging activity and ability to produce O2· -. All fractions were centrifuged for 10 min at 16 000 g before addition of reagents, except fraction B4, which was centrifuged just before measurement. Five to 100 mL of fraction were made up to 800 mL with 50 mm phosphate buffer at pH 6.0, mixed with 100 mL 10 mm ABTS and 100 mL 100 mm H2O2, and incubated for 10 min at 25 °C. Enzyme activity was determined spectrophotometrically by measuring A420, using an extinction coefficient of 36 mm-1 cm-1. To measure the capability of the cell wall fractions to produce O2· -, 60–420 mL of fraction were made up to 1170 mL with 50 mm phosphate buffer at pH 7.0, mixed with 150 mL of 10 mm epinephrine, 15–150 mL of 10 mm NADH and 150 mL of 100 mm H2O2, and incubated for 10 min at 25 °C. For a control, O2· - was measured in the absence of H2O2 and NADH. O2· - production was quantified as above. The activity of glucose-6-phosphate dehydrogenase (G6PDH) was used as a marker of cytoplasmic contamination of cell wall fractions. The assay (Kranner 1998) contained 400 mL of fraction, 1000 mL of 0.15 m tricine buffer at pH 8.0 containing 0.15 m MgCl2 and 0.15% glucose-6phosphate. The reaction was initiated by the addition of 100 mL of 1.5% NADP. The change in absorbance at A340 was followed for 11 min. Contamination with cytoplasmic enzymes was estimated using the percentage of G6PDH activity in the cell wall fractions relative to the activity in the cytosolic fraction (n = four replicates for each cell wall fraction).
Identification of potential reductants in leachates Embryonic axes were excised, rinsed in double distilled H2O then incubated (~0.5 g mL-1 ultrapure water) at 20 °C for 5 min with constant agitation (75 rpm). The leachate was transferred to a new vial, flash frozen in liquid nitrogen and stored at -75 °C prior to lyophilization. The freeze-dried residue was derivatized [0.2 mL N,Obis(trimethylsilyl)trifluoroacetamide and 0.3 mL pyridine at 70 °C for 30 min] then centrifuged at 13 000 g for 5 min. The supernatant was analysed using GC (Thermo Finnigan Trace GC Ultra) with an Rtx-5MS column (15 m length, 0.25 mm internal diameter, 0.25 mm df; Restek) running a temperature program (1 min at 80 °C, 15 °C min-1 to 240 °C, 6 min hold; helium carrier gas at constant flow rate of 1 mL min-1). An injection volume of 1 mL was used with a split ratio of 10:1. The compounds were detected using MS (Thermo Finnigan Trace DSQ; ionization energy 70 eV, scan frequency range m/z 10–1000 per 0.2 s), and identified through comparison with the National Institute of Standards and Technology mass spectral database. The identity of cinnamic acid was confirmed by co-elution with an analytical standard. © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
Superoxide production in Castanea sativa seeds 5 The ability of ferulic acid, p-coumaric acid and cinnamic acid to stimulate O2· - production was tested in fraction B3. The hydroxycinnamic acids were dissolved in 100% ethanol as 0.1 m stock, and 1.5–15 mL were added instead of NADH in the assay described above, with volume differences made up with phosphate buffer. O2· - production was quantified as above.
Electrophoretic separation and in-gel staining of O2· --producing enzymes Cell wall fractions were dialysed against solid sucrose, followed by reverse dialysis, overnight at 4 °C against 50 mm phosphate buffer pH 7, then microconcentrated using Microcon centrifugal filters (Millipore, Billerica, MA, USA). Samples were aliquoted and used immediately, or stored at -70 °C. Electrophoretic studies on 12% polyacrylamide gels followed a modification of the method of Laemmli (1970). Running buffer and gels contained 0.1% sodium dodecyl sulphate (SDS), but samples were not heated, and mercaptoethanol and SDS were omitted from the loading buffer. Using SDS in the running buffer resulted in sharper bands, while without SDS, some smearing was observed (not shown). ‘Broad Range’ molecular mass markers were stained with Coomassie brilliant blue G250. After electrophoresis, bands were visualized by in-gel staining using guaiacol, 3,3′,5,5′-tetramethyl-benzidine (TMB) or NBT. Staining with 20 mm guaiacol (in the presence of 20 mm H2O2 in 10% glycerol and 0.25 m sodium acetate buffer pH 5.0) is indicative of the presence of peroxidases. In the spectrophotometric assay described above, ABTS was used because it is a very sensitive, non-toxic POX substrate that produced highly reproducible results for cell wall fractions. However, when ABTS was used in gels, the soluble ABTS+ radical produced during the reaction with ECPOX rapidly diffused away, resulting in smearing gels. Conversely, the products of the reaction with guaiacol polymerize, producing long-lasting bands. Hence, guaiacol was used for the visualization of ECPOX activity in gels. To test for haem groups, gels were stained with 6.3 mm TMB in 0.25 m sodium acetate buffer, pH 5.0 and 30 mm H2O2 (Thomas, Ryan & Levin 1976). To study O2· production, gels were pre-equilibrated with 10% glycerol in 50 mm phosphate buffer, pH 7.4, containing 0.1 mm MgCl2 and 1 mm CaCl2 for 30 min, then stained at room temperature for 1 h with 0.5 mm NBT (in the same buffer) and 0.4 mm NADH (Serrano et al. 1994; López-Huertas et al. 1999). Control gels were incubated in the absence of NADH.
Effects of H2O2 on seed germination and growth of seedling axes Non-desiccated seeds were soaked in 0, 0.01, 0.5 or 1 m H2O2 for 1 h. One M H2O2 was used to study the effects of H2O2 treatment on the germination of desiccated seeds. Following desiccation, seeds were germinated, as above, either without pre-treatment or after submerging them for 1 h in H2O or H2O2. © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
Embryonic axes were isolated from seeds that had lost germinability after 5 d of desiccation. After sterilization in 50 mm sodium dichloroisocyanurate for 10 min, the axes were rinsed three times, submerged in 0.1 mm, 1.0 mm or 10.0 mm H2O2 for 1 h and cultured as above. Best growth rates were achieved with 0.1 mm H2O2 and this concentration was used to study the effect of H2O2 on 10 d and 21 d desiccated seeds in comparison with H2O treatment.
Statistical analysis Data were tested for significance using one- or two-way anova, as appropriate, in combination with post-hoc comparison of means using the least significant difference test. Arcsine transformation was applied to total germination and axis viability data. anova relies on the assumption that the data being analyzed is normally distributed. When using percentage values that are by their very nature restricted to the interval of 0 to 100, arcsine transformation is a method of correcting percentage values close to interval limits to simulate a normal distribution of data prior to anova.
RESULTS Seed WC, desiccation and viability Desiccation led to rapid water loss in the cotyledons, and to a lesser extent, in the embryonic axes (Fig. 1a), resulting in progressive viability loss. After 5 d of desiccation, the intact seeds had lost their ability to germinate (Fig. 1b) and were rapidly overgrown by microbes; microbial growth was particularly observed in the cotyledons rather than the embryonic axes (not shown). However, after 5 d of desiccation, 90% of the embryonic axes could be rescued by aseptic tissue culture (Fig. 1b). Longer desiccation also caused progressive viability loss in the embryonic axes (Fig. 1b). Staining with PI and DAPI indicated that cell death occurred when the axis WC fell below 30% (Fig. 2a). Note that PI and DAPI staining were conducted on a single seed basis rather than representing the average response of the population. In embryonic axes with ~30% WC, it was observed that either all or none (Fig. 2a) of the nuclei were stained by PI, so it appears that cell death spread rapidly rather than gradually across the tissue. In the cotyledons of the same seeds, not all nuclei were stained by PI at ~26% WC, implying that cell viability was lost more gradually than in embryonic axes. In contrast, EGSSG/2GSH was measured in samples comprised of multiple seeds in each replicate so that data represent average population responses. On a DW basis, embryonic axes of non-desiccated seeds contained twice as much GSH as cotyledons (Fig. 2b,c). Upon desiccation, GSH concentrations declined in both axes and cotyledons, while GSSG concentrations increased. Correspondingly, EGSSG/2GSH increased towards more positive (more oxidizing) values by approximately 70 mV in embryonic axes and about 25 mV in cotyledons (Fig. 2b,c), indicative of greater oxidative stress in the embryonic axes. EGSSG/2GSH is a function of the molar concentrations of GSH and GSSG,
T. Roach et al.
Water content (% FW)
6
60
(a)
40
20
Extracellular production of ROS by embryonic axes
0 100
(b)
80 Viability (%)
2001) comprised less than 10% of the total LMW thiols and were also converted into their corresponding disulphides during desiccation (data not shown). Overall, EGSSG/2GSH correlated strongly (R2 = 0.916, P < 0.043) with viability in embryonic axes (Figs 1b & 2b) while in cotyledons (Figs 1b & 2c) the correlation was weaker (R2 = 0.678, P < 0.076).
60 40 20 0 0
5
10
15
20
Desiccation (d) Figure 1. Viability loss in response to desiccation of Castanea sativa seeds. Seeds were desiccated at 15% relative humidity and 15 °C; data are means ⫾ SE. (a) Water content in isolated axes (open symbols) and cotyledons (closed symbols); n = 4 replicates of five axes each. (b) Seed viability is shown as a percentage of total germination (closed symbols; n = 5 replicates of eight seeds each), and axis viability is shown as a percentage of axes that had expanded by more than 10 mm after 21 d in aseptic tissue culture (open symbols; n = 3 replicates of 15 axes each). FW, fresh weight.
which were determined by WC that after 21 d was twice as high in the embryonic axes as in the cotyledons. As a consequence, EGSSG/2GSH was more reducing in the cotyledons due to higher molar GSH concentrations and higher GSH/ GSSG ratios than in embryonic axes after 21 d of desiccation. Other LMW thiols such as cysteine, glutamylcysteine and cysteinyl-glycine that also contribute to the ‘intracellular redox environment’ (Schafer & Buettner
Embryonic axis excision induced production of O2· - and of its dismutation product H2O2 (Fig. 3a). Staining was particularly observed at the perimeter of the cut surface, spreading along the uncut external surface, and in the xylem with higher activity at the radical tip than the shoot tip (Fig. 3a). Axes isolated from non-desiccated seeds produced O2· - at a rate of 0.14 ⫾ 0.02 nmol g-1 DW s-1 (Fig. 3b). Although viability was maintained (Fig. 1b), O2· - production significantly increased in embryonic axes desiccated up to 5 d (Fig. 3b). Both viability and O2· - production decreased with further desiccation, showing an overall bell-shaped relationship between WC and O2· - production (Fig. 3b). Incubation of excised axes from non-desiccated seeds with SOD significantly decreased adrenochrome formation by 60% (P < 0.05), confirming the specificity of the assay. Incubation of excised axes with 1 or 10 mm DPI and 1 mm NaN3 decreased O2· - production by 30–40% (P < 0.05) (Table 1). O2· - production was also observed after removal of the axes from the incubation solution (‘leachate control’ in Table 1), suggesting that some ROSproducing enzymes and their substrates were released into the incubation medium. The leachates produced 80% less O2· - in the presence of 0.5 mm NaN3, suggesting that metalloenzymes such as ECPOX were involved in O2· production.
Putative enzymes and reductants involved in extracellular O2· - production Using a cell fractionation technique showed that in embryonic axes and cotyledons, 64% and 28%, respectively, of the activity of ABTS-dependent H2O2-metabolizing enzymes was found in the crude cell fraction (Fig. 4a) with ~15 times
Figure 2. Effect of desiccation on cellular and biochemical viability markers in Castanea sativa seeds. (a) Visualization of live and dead cells using propidium iodide (PI) and 4′,6′-diamino-2-phenylindole (DAPI). Images show sections of embryonic axes (top row) and cotyledons (bottom row) that were isolated from intact seeds and desiccated until they had reached the WCs (%) indicated at the bottom of each image. From each seed, the embryonic axes and a segment of the cotyledon were used so that images in the top and the bottom rows are shown for the same individual seed. PI fluoresces red when bound to nucleic acids, but is unable to cross the nuclear envelope of viable cells, so staining of nuclei only occurs in cells with seriously impaired nuclear membranes, seen here at WCs below 30%. DAPI (green) can permeate intact nuclear envelopes and stain both live and dead cells and was used as a control to show the location and density of nuclei within the sections, showing that embryonic axes had a greater cell density than cotyledons. Two images are shown for axes with 30% WCs and their corresponding cotyledons (26% WC), at which point cell death became apparent in some, but not all embryonic axes. Magnification: 40-fold for all images. (b and c) Glutathione redox state in (b) embryonic axes and (c) cotyledons. White bars represent glutathione (GSH) and black bars glutathione disulphide (GSSG), and the line denotes glutathione half-cell reduction potential (EGSSG/2GSH). Data are means ⫾ SE (n = 4 replicates of five axes or cotyledon segments each). Statistical differences are shown for EGSSG/2GSH. Data labelled with the same letters do not differ significantly (P < 0.05). DW, dry weight. © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
Superoxide production in Castanea sativa seeds 7
(a)
DAPI
49
35
32
30
30
32
WC
45
38
31
26
26
31
WC
5
(b)
c
4
–150
3 GSSG and GSH (µmol g DW –1)
–100
b
2 –200
a a
1 0
–250 (c)
4
EGSSG/2GSH (mV)
Cotyledon
Axis
PI
–150
3 2
c
–200 d
1
b a
0
–250 0
5 10 Desiccation (d)
higher activity in embryonic axes than in cotyledons. In embryonic axes, H2O2-metabolizing activity was also found in three cell wall fractions, 17% of which were loosely bound to the cell wall (B1); 11% were bound by hydrophobic interactions (B2), and 72% were ionically bound (B3). In cell wall fractions of cotyledons, H2O2-metabolizing activity was only detected in fraction B3, displaying half of © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
21
the activity in the B3 axis fraction. Cytoplasmic contamination, assessed by measuring G6PDH activity, was found only in the B1 fraction at 2% of the G6PDH activity in the crude cell fraction. The B3 cell wall fraction of the embryonic axes produced O2· - at significantly higher rates than all other fractions, two times more that the crude cell fraction of the axes
8
T. Roach et al.
(a) 1
1 2 3 4
2
3
1 cm
4
O2•- production (nmol g –1 s–1)
0.4
c
(b)
abc
0.3 0.2
bc
ab a
0.1 0 0
2
5
10
21
Desiccation (d) Figure 3. Production of reactive oxygen species by embryonic axes isolated from Castanea sativa seeds in response to wounding and desiccation. (a) Localization of ROS production in a non-desiccated isolated embryonic axis. The right image shows an unstained whole axis and the dotted lines denote the areas where cross sections (left images) were taken. The far left cross sections show superoxide (O2· -) production after staining with nitroblue tetrazolium (NBT; dark purple stain). The near left sections show hydrogen peroxide (H2O2) production in conjunction with peroxidase activity after incubation in 3,3′-diaminobenzidine (DAB; dark yellow staining). Note the intense NBT staining on the outer edge of the cut surface. (b) Spectrophotometric quantification of O2· - production following axis excision after various periods of desiccation, corresponding to WCs of 48.9 ⫾ 2.2, 34.8 ⫾ 2.2, 31.3 ⫾ 3.6, 25.3 ⫾ 3.6 and 23.5 ⫾ 2.3%, for 0, 5, 10, 15 and 21 d, respectively. Each bar shows the rate of O2· - production during 1 h of imbibition in epinephrine solution. Data show means + SE (n = 4 replicates each comprising five axes that were imbibed in 1.5 mL epinephrine). Bars labelled with the same letters do not differ significantly (P < 0.05).
and almost three times more than that of the cotyledons (Fig. 4b). In the presence of H2O2, NADH stimulated O2· production significantly with a maximum at a concentration of 0.1 mm (Fig. 4c). Of the phenolic acids tested, ferulic acid stimulated O2· - production the most, while cinnamic acid had a weaker, although significant effect and stimulated O2· - production more than NADH (note the different scales in Fig. 4c,d). Cinnamic acid was identified in seed leachates by GC-MS (Table 2) through spectral library
matching and its identity was confirmed by comparison with analytical standards. Another molecule with reductant qualities found in leachates was ascorbic acid (AA). Following electrophoretic separation of cell wall proteins, guaiacol staining revealed that ECPOX-like enzymes with an apparent molecular mass of approximately 50 kDa were present in fractions B1 and B2, and 45 kDa in B3, and TMB staining confirmed the presence of metalloproteins with similar molecular masses (Fig. 5). After NBT staining, two bands that contained enzymes capable of O2· - production with apparent molecular masses of approximately 50 and 93 kDa were identified in all fractions. Clear bands of O2· - production indicated that such O2· - production is not non-specific, but requires a functioning protein.
Effects of H2O2 on seed germination and vigour of isolated embryonic axes, and relationship between O2· - production and axis growth A single treatment of non-desiccated seeds with H2O2 for 1 h had a concentration-dependent effect on subsequent germination. Concentrations of 0.1 m and 0.5 m H2O2 significantly increased total germination (P < 0.05), but higher concentrations (up to 1 m) neither further increased nor decreased total germination (data not shown). Soaking non-desiccated seeds in either H2O or 1 m H2O2 increased WC, vigour and total germination compared with nonsoaked seeds (Fig. 6a). No significant differences were found between the H2O or 1 m H2O2 treatments, indicating that the water uptake rather than H2O2 was responsible for enhanced seed vigour. Compared with values for freshly collected, untreated seeds, mild desiccation for 2 d (Fig. 6b) decreased total germination by 18%. Soaking these mildly desiccated seeds in H2O improved total germination by 15%, whereas incubating them in 1 m H2O2 doubled total Table 1. Effects of superoxide dismutase (SOD), and the enzyme inhibitors diphenylene iodonium (DPI) and sodium azide (NaN3) on O2· - concentrations in leachates from excised C. sativa axes. Axes were excised and incubated with SOD or enzyme inhibitors for 10 min, and O2· - concentrations measured after further incubation with epinephrine for 5 min. ‘Leachate control’ and ‘leachate + NaN3’ show rates of O2· - production over 30 min in the leachates (without axes) in the absence or presence of NaN3, respectively. Leachates were obtained by soaking excised axes in 2 mL of distilled water for 1 h. Data represent mean ⫾ SE (n = 3 replicates of 5 axes each). Data labelled with the same letter do not differ significantly (P < 0.05) Treatment
O2· - (% of control)
Control SOD (250 units mL-1) DPI (1 mm) DPI (10 mm) NaN3 (0.1 mm) NaN3 (1 mm)
100a 40c 64b 67b 97a 61bc
Leachate control Leachate + NaN3 (0.5 mm)
100d 21e
© 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
Enzyme activity (mkat g–1)
40 150
(c)
(a)
a
b
b
30 100 20 d
50
a
a
a
10
e b
b
b c
c
b
c
c
b
0
0 C
B1 B2 B3 Cell wall fraction
B4
O2•- production (nmol g–1 s–1)
Superoxide production in Castanea sativa seeds 9
0.0 0.1 0.2
0.4 1.0
with H 2O 2
0.0 0.2 without H 2O 2
40
(b)
f
(d)
0.5
d
30
0.4 a e
20 d
ab
10
bc
bc c
bc
bc
c
c
a a a b b b
a c b
0.3
d
a c b
d
a c b
c
B1 B2 B3 Cell wall fraction
B4
0.1 0
0 C
0.2
O2•- production (mmol g–1 s–1)
O2•- production (nmol g–1 s–1)
NADH concentration (mM)
0 0.1 0.2 0.4 1.0 Phenolic acid concentration (mM)
Figure 4. Activity of hydrogen peroxide (H2O2)-degrading enzymes and superoxide (O2· -) production in cellular fractions of Castanea sativa embryonic axes and cotyledons. White bars in (a) and (b) represent embryonic axes and black bars denote cotyledons; C, crude cell fraction; B1, enzymes loosely bound to the cell wall, e.g. by hydrogen bonds; B2, enzymes bound by hydrophobic interactions; B3, enzymes bound by ionic bonds; B4, enzymes bound to the cell wall by covalent linkages. (a) Activity of enzymes that catalyze the H2O2-dependent oxidation of 2,2′-azino-bis 3-ethylbenzo-thiazoline-6-sulfonic acid (ABTS) such as extracellular peroxidases. (b) O2· production in the presence of 0.1 mm NADH and 10 mm H2O2, quantified spectrophotometrically after incubation with epinephrine. (c) Stimulation of O2· - production by NADH in fraction B3 of embryonic axes in the presence or absence of 10 mm H2O2. (d) Stimulation of O2· - production by cinnamic acid (white bars), p-coumaric acid (grey bars) and ferulic acid (black bars) in the presence of 10 mm H2O2 in fraction B3. Data show means + SE (n = 4). Bars labelled with the same letters do not differ significantly (P < 0.05).
germination to 80%, significantly higher (P < 0.05) than soaking in H2O. After 5 d of desiccation (Fig. 6c), the seeds reached a critical WC of 28 ⫾ 3% and failed to germinate. In these lethally stressed seeds, soaking in neither H2O nor in H2O2 enhanced viability. However, 87 ⫾ 4% of axes removed from 5 d desiccated seeds (31 ⫾ 4% WC) were still viable and 53 ⫾ 4% also grew shoots after 21 d in tissue culture (Fig. 7c). Snapshots of a typical viable embryonic axis over 21 d in tissue culture are shown in Fig. 7a, with shoot formation starting after 15 d in tissue culture.Treating axes isolated from 5 d desiccated seeds with 0.1 mm H2O2 for 1 h prior to tissue culturing enhanced viability by 17%, expressed as a percentage of embryonic axes that had grown roots, compared with H2O-treated controls (Fig. 7c); concentrations above 0.1 mm of H2O2 decreased viability slightly. No significant differences in shoot production were observed after 21 d tissue culture regardless of H2O2 concentrations tested (Fig. 7c). Treatment with 0.1 mm H2O2 also increased vigour, expressed as root elongation, which increased by 0.27 mm per day more than that of controls, © 2009 Blackwell Publishing Ltd, Plant, Cell and Environment
determined by linear extrapolation of the data shown in Fig. 7b. Treating embryonic axes isolated from 10 d or 21 d desiccated seeds with 0.1 mm H2O2 did not affect viability (not shown) or vigour (Fig. 7b) significantly. Figure 7d shows the relationship between the growth of healthy embryonic axes and O2· - production. The embryonic axes only elongated without gaining DW up to day 10 when maximal O2· - production was observed: DW started to increase between 10 d and 15 d, coinciding with falling rates of O2· - production.
DISCUSSION ROS have been frequently portrayed as the culprits that induce seed death in recalcitrant seeds. In nonphotosynthetic tissues, including seeds, respiration is a major source of intracellular ROS production, and this production is amplified by desiccation through disruption of the mitochondrial electron transport chains (Kranner & Birtic´ 2005). Despite their obvious toxic effects, ROS play
10 T. Roach et al.
No
Putative ID
RSI
SI
Relative abundance (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Glycerol 2-Keto-D-gluconic acid Galactonic acid D-Glucitol (Sorbitol) Myo-Inositol Myo-Inositol L-Ascorbic acid Butan-1,2,3-triol Pentitol 2,3,4-Trihydroxybutanal D-Glucono-1,5-lactone Pentitol Malic acid Ribitol Arabinitol Dihydroxypropylstearate 2,3,4-Trihydroxybutanal 2,3-Dihydroxypropanoic acid (Glyceric acid) 2,3,4-Trihydroxybutanoic acid (Erythronic acid) Myo-Inositol 3,4-Dihydroxy-2(3H)-Furanone Cinnamic acid
894 790 863 913 887 864 853 886 842 894 829 834 843 824 906 802 886 825 826 764 810 581
842 747 855 850 841 840 842 853 821 873 747 778 789 752 843 693 807 773 786 658 685 713
3.246 1.044 0.834 0.818 0.612 0.414 0.273 0.261 0.177 0.104 0.089 0.070 0.045 0.040 0.034 0.038 0.020 0.016 0.015 0.008 0.013 0.004
Table 2. Identification of lowmolecular-weight compounds in leachates of embryonic axes by gas chromatographymass spectrometry. The relative abundance of each compound is expressed as a percentage of the total peak area of all compounds present in the leachate. The leachate of embryonic axes was composed mainly of sugars (91.8%, which are not shown); the remaining 8.2% consisted of polyols and organic acids, including cinnamic acid. Peaks were identified through spectral library matching. Standard index (SI) is a direct matching factor between the spectrum of an unknown peak and the library spectrum, reverse standard index (RSI) is a reverse search matching factor ignoring any unknown peaks that are not present in the library spectrum. Library hits with SI and RSI values above 900 are an excellent match, 800–900 a good match and
