|Home | About | Journals | Submit | Contact Us | Français|
Background and aims Our aim was to elucidate how plant tissues under a severe energy crisis cope with imposition of high NaCl, which greatly increases ion fluxes and hence energy demands. The energy requirements for ion regulation during combined salinity and anoxia were assessed to gain insights into ion transport processes in the anoxia-tolerant coleoptile of rice.
Methods We studied the combined effects of anoxia plus 50 or 100 mm NaCl on tissue ions and growth of submerged rice (Oryza sativa) seedlings. Excised coleoptiles allowed measurements in aerated or anoxic conditions of ion net fluxes and O2 consumption or ethanol formation and by inference energy production.
Key Results Over 80h of anoxia, coleoptiles of submerged intact seedlings grew at 100 mm NaCl, but excised coleoptiles, with 50 mm exogenous glucose, survived only at 50 mm NaCl, possibly due to lower energy production with glucose than for intact coleoptiles with sucrose as substrate. Rates of net uptake of Na+ and Cl− by coleoptiles in anoxia were about half those in aerated solution. Ethanol formation in anoxia and O2 uptake in aerobic solution were each increased by 13–15 % at 50 mm NaCl, i.e. ATP formation was stimulated. For acclimation to 50 mm NaCl, the anoxic tissues used only 25 % of the energy that was expended by aerobic tissues. Following return of coleoptiles to aerated non-saline solution, rates of net K+ uptake recovered to those in continuously aerated solution, demonstrating there was little injury during anoxia with 50 mm NaCl.
Conclusion Rice seedlings survive anoxia, without the coleoptile incurring significant injury, even with the additional energy demands imposed by NaCl (100 mm when intact, 50 mm when excised). Energy savings were achieved in saline anoxia by less coleoptile growth, reduced ion fluxes as compared to aerobic coleoptiles and apparent energy-economic ion transport systems.
Coupling ratio: The ratio of ATP (or PPi)/H+ on H+ translocases and the ratio of H+/ion on symporters and antiporters. Thus, low ratios denote low energy expenditure (i.e. ‘energy-economic’ transport systems).
Energy for maintenance: Energy required for maintenance of cells/tissues (i.e. long-term survival without growth). Under anoxia this energy would be mainly required for protein and lipid turnover and membrane transport of ions at flux equilibrium (Edwards et al., 2012; Atwell et al., 2015). This maintenance requirement will increase when coping with an environmental factor, which requires more energy, in the present case 50 or 100 mm NaCl, and/or because cellular compartmentation has to be maintained.
Salinity tolerance requires regulation of Na+ and Cl− net uptake and compartmentation into vacuoles to avoid ion toxicity in the cytoplasm but also to provide solutes for osmotic adjustment to avoid possible water deficits (Greenway and Munns, 1980). In aerobic tissues, ion fluxes can consume considerable energy. For plants in flooded soils which lack O2, however, the anoxic tissues experience an ‘energy crisis’ (Gibbs and Greenway, 2003). Plant responses to combined salinity and low O2 have previously focused on roots, and showed increased Na+ transport to the shoots than saline-aerated roots (Barrett-Lennard, 2003). When the stele became severely hypoxic or anoxic, Cl− transport (Gibbs et al., 1998) and Na+ and K+ regulation to the xylem (Kotula et al., 2015) become impaired, presumably owing to low energy availability for ion transport (Colmer and Greenway, 2011). The present study of anoxic submerged, germinating rice was mainly focused on the balance between energy demand and production during anoxia combined with salinity (50 and 100 mm NaCl), enabling interpretations on possible energy-efficient ion transport mechanisms enabling acclimation to salinity in the anoxia-tolerant coleoptile of rice.
The energetics during anoxia combined with salinity is of particular interest because anoxic tissues already face an ‘energy crisis’ and so the increased energy demand imposed by high NaCl would be a substantial challenge. High external NaCl increases the energy demand as there would be requirements of secondary energy-dependent Na+ effluxes via H+–Na+ antiporters from the cytoplasm across the plasma membrane and tonoplast to avoid toxic Na+ concentrations in the cytoplasm (Munns and Tester, 2008) and if Cl− is to be used as an osmoticum also for energy-dependent Cl− influxes via H+–Cl− symports (Teakle and Tyerman, 2010). The alternative of producing organic solutes, rather than using ions as osmolytes, is about one order of magnitude more energy-expensive than the use of ions (bacteria, Over, 1999; vascular plants, Raven, 1985). Such energy costs may become too taxing in anoxic tissues, which judged on ethanol formation by anoxic rice coleoptiles have rates of glycolysis yielding 7–10 % of the ATP formed by tissues in air with oxidative phosphorylation (Gibbs and Greenway, 2003; Huang et al., 2005).
In our study we used completely submerged rice seeds which can germinate under anoxia, albeit only growing a coleoptile (Atwell et al., 1982; Alpi and Beevers, 1983). As well, under aeration the submerged germinating rice tolerated NaCl, at least up to 200 mm (Kurniasih et al., 2013). In these submerged seedlings, energy formation can be assessed and ion fluxes can be evaluated without the potential complication of xylem transport resulting in accumulation of Na+ in transpiring leaves. Rapid solute exchange occurs between cells of the coleoptile and external solution, as evidenced by transient membrane depolarization followed by rapid repolarizations after adding substrates to the medium (Zhang and Greenway, 1995) and rapid uptake of K+ over the first hour when added to the medium (Colmer et al., 2001). This ready exchange allows measurement of ion fluxes, essential to estimate energy expenditure. The ready exchange is understandable because these coleoptiles had no gas envelopes (as opposed to leaves formed in air, Colmer et al., 2014) and presumably little cuticle development, as they had been submerged since imbibition. Moreover, in the submerged rice seedlings, cell turgor pressure can be assessed as the difference between the osmotic potentials of the protoplasts and the solution (Kurniasih et al., 2013). For interpretation of ion fluxes and energy formation by the coleoptile, excision of this organ and provision of exogenous glucose provides an anoxia-tolerant system for measurements of ion fluxes and ethanol formation (e.g. these excised tissues survive more than 100h of anoxia, Greenway et al., 2012).
To elucidate the characteristics of ion transport during an energy crisis, in the anoxia-tolerant coleoptile of rice, we: (1) evaluated the effects of NaCl combined with anoxia on net fluxes of Na+, Cl− and K+, tissue concentrations of these ions and growth; (2) evaluated whether energy production via glycolysis linked to ethanol formation increases in response to additional energy requirements associated with NaCl exposure; and (3) constructed an energy budget to compare any increase in energy production and/or reduction in energy consumption owing to reduced growth, and to possible increased demand for energy related to increased ion transport. This energy budget elucidated here for anoxic rice coleoptiles showed that the ATP available for acclimation to high NaCl in anoxia was much lower than in aerobic coleoptiles. We propose that use of energy-efficient transport systems and reduced permeability to Cl− and Na+ of the plasma membrane and tonoplast are key factors in acclimation of anoxia-tolerant cells/tissues to combined anoxia and high NaCl.
A series of five experiments were conducted on rice seedlings grown submerged in nutrient solution culture, using the intact seedlings or excised coleoptile tips (8–12mm in length), in order to measure responses to NaCl salinity during anoxia for growth, ion net uptake or loss, tissue ion concentrations, tissue sap osmotic potential and ethanol production; these parameters were then interpreted as related to energy balance. Recovery upon re-aeration after anoxia and NaCl treatments, as well as the effects of NaCl on respiration by aerobic coleoptile tips, were also evaluated. The objectives of the various experiments, plant materials used and measurements taken are listed in Table 1. There were only comparisons between anoxic with aerated tissues for the excised coleoptile tips. For the intact seedlings such a comparison would be rather meaningless because when in aerated conditions the coleoptile soon enters senescence, whereas in anoxia the coleoptile continues to grow (Kurniasih et al., 2013).
Dehulled rice (Oryza sativa L. ‘Amaroo’) seeds (caryopses) were washed with 0·1 % (v/v) sodium hypochlorite for 20min, rinsed thoroughly with deionized water and submerged in aerated nutrient solution in the dark at 30°C. ‘Amaroo’ is regarded as an anoxia-tolerant genotype (Huang et al., 2003) and it tolerated at least up to 200 mm NaCl when grown submerged in aerated nutrient solution (Kurniasih et al., 2013). For the responses elucidated here for ‘Amaroo’ to combined anoxia and salinity, future experiments should also assess other genotypes of rice that vary in anoxia tolerance (e.g. Huang et al., 2003) and/or germination and early growth when seeded in a flooded soil (Ismail et al., 2009) and in salinity resistance (e.g. Yeo and Flowers, 1983; Rahman et al., 2016).
Experiments were either with intact seedlings (seed and coleoptile) or with excised coleoptile tips (8–12mm in length, see Huang et al., 2005), dependent on the purpose of the experiment. Both the seedlings for use intact and the seedlings grown to excise coleoptiles were always submerged first in nutrient solution continuously bubbled with air for 36h, then an N2/air mix so that dissolved O2 was 0·06 mm (hypoxia) for 30h, followed by imposition of anoxia, in the dark at 30°C. The period of hypoxia (i.e. hypoxic pretreatment) was given to avoid any possible ‘anoxic shock’ (see Gibbs and Greenway, 2003). Treatments (aerated or anoxic) started at 66–72h after imbibition, NaCl treatment was imposed 18h after anoxia commenced, and experiments terminated at different times ranging between 156 and 216h after imbibition. All experiments had 3–4 replicates for each treatment (see figure and table captions); for intact seedlings one replicate was 24 seedlings in an individual flask, and for excised coleoptile tips one replicate was an individual flask containing 20–30 of these tips. All flasks were sealed and had continuous gas-flow via inlet and outlet ports. Experiments with time-series had sets of flasks, so that each flask was only opened and sampled once (i.e. the sampled flasks were discontinued and only unopened flasks were continued).
For intact seedlings, the nutrient solution contained (mm): Ca2+ 2·0, Cl− 3·0, 0·3, 0·3, 0·9, K+ 0·3, Mg2+ 0·3, H2PO4 0·1; and (μm): Fe-EDTA 12·5, H3BO3 6·25, Mn2+ 0·5, Zn2+ 0·5, Cu2+ 0·125, Mn2+ 0·125, Ni2+ 0·25. For experiments with excised coleoptiles the solution contained (mm): Ca2+ 1·0, 1·6, K+ 0·3, Mg2+ 0·3; plus glucose and antibiotic (see below). Both types of solutions contained 0·5 mm 2-[N-morpholino]ethanesulfonic acid (MES), which was added after autoclaving, and the pH was adjusted to 6·5 using Ca(OH)2. All glassware had been autoclaved and coleoptile excision and handling were in a laminar-flow hood. The reason why these solutions differed in composition was that for the intact seedlings evaluation of growth was a main feature, so complete nutrition was provided; whereas glucose was added for excised coleoptiles and then a complete nutrient composition would tend to foster bacterial growth, so a minimal nutrient solution was then used.
For intact seedlings, each experimental unit consisted of 24 seeds placed in 30mL of nutrient solution in a 50-mL glass vessel (conical flask) in which the seedlings remained for the whole experiment. In all experiments with intact seedlings, the small developing first leaf, which is enclosed by the coleoptile, was excluded during sampling.
To produce seedlings used to excise coleoptiles, the dehulled and washed seeds were sown on netting across a PVC insert submerged in cylindrical vessels containing 4 litres of nutrient solution. The tips of coleoptiles were excised, length 8–12mm, therefore excluding any enclosed leaf tissue. The excised coleoptiles were collected in the nutrient solution with 20 mm glucose and 50mg L−1 ampicillin also added. This antibiotic was added to all solutions containing glucose and was adequate to maintain low populations of bacteria, provided solutions were refreshed every day. Following excision, sub-samples of about 0·1g of fresh weight of coleoptiles were each transferred to a 50-mL flask and allowed to recover/heal for 5h while maintaining hypoxia (0·06 mm O2) in the nutrient solution with 20mm glucose. This ‘healing’ is required because during the first 1–2h after excision there were K+ losses from rice coleoptiles (Colmer et al., 2001). Glucose in the nutrient solution was then changed to 5 mm for the aerated and 50 mm for the anoxic treatments with excised coleoptiles (see below); these concentrations prevented excessive sugar accumulation in aerated tissues and ensured adequate glucose uptake in anoxic tissues (Huang et al., 2005).
For seedlings in flasks, at the end of the hypoxic pretreatment the nutrient solution in each flask was refreshed and then all flasks were continuously flushed with high-purity N2 to impose anoxia. For the excised and healed coleoptiles, the nutrient solution in each flask was refreshed and either continuously flushed with air or with high-purity N2 to impose anoxia, for 72–78h. The air and N2 were pre-humidified. NaCl treatment (50 or 100 mm) was imposed at 18h after anoxia started and increased by steps of 50 mm every 24h.
The solution volumes in flasks varied between 10 and 30mL, depending on time periods between samplings, and this ensured K+ changes in the solution could be monitored because there was measurable but not excessive depletions of exogenous K+. Solution samples were obtained by a needle inserted via the exit port of the flask and drawing up the liquid into a gas-tight syringe, while the flow of high-purity N2 (or air) was continued. The solutions in the tubes were completely replaced every 24h, while maintaining gas flow, to prevent ethanol build up in the anoxic solution. The fresh solution for the anoxic treatment was pre-flushed for several hours with high-purity N2 as a stock before being injected into the individual flasks. Tissue samplings required opening of the flasks, entailing substantial entry of O2 into anoxic solutions, so the sampled flasks were discontinued and only unopened duplicate flasks were continued.
For anoxia, rates of ethanol formation were measured for excised coleoptiles in treatments of 0·3 and 50 mm NaCl at: 0–2, 4–6, 34–36, 70–72 and 88–90h (Experiment 5). Incubation solution was renewed at the start of each period and the outlet of each flask was connected to another flask containing deionized water and in an ice bath to trap ethanol. Solutions in the flasks containing tissues, as well as the ice-cold deionized water in traps on the outlets of each flask, were sampled and stored in sealed tubes at−20°C. The recovery of ethanol spiked into this system was 95 %. Using the same time-frame, aerated, excised coleoptiles were used to measure O2 consumption in air-equilibrium solution (Experiment 5).
The tissue samplings were at 0, 24, 30, 48, 54, 72 and 78h after starting NaCl treatments, while samplings of tissues at 0·3 mm NaCl were at 0, 24 and 72h (Experiment 4 on intact seedlings; Experiment 5 on excised coleoptiles). In both these experiments, osmotic potential of expressed tissue sap (πsap) was measured for samples at 72h after the start of combined anoxia and high NaCl.
Tissues used for ion measurements were free-space washed under the same gas regime as during the prior treatment. The wash procedure was for 3min in 2 mm CaSO4 and iso-osmotic mannitol; 3min will remove almost all the solutes from the free-space and this time period also mitigates the risks that ions from the protoplasts are lost to the solution. For tissues used for total sugars and Pi measurements, the washing was three periods each of 3min (NaCl still present in those treatments) and the solution contained iso-osmotic mannitol rather than glucose. Tissues used to measure πsap were blotted to remove surface water, without a free-space wash.
Excised coleoptiles were exposed to anoxia combined with 50 mm NaCl for 72h and then reaerated in nutrient solution with 0·3 mm NaCl, and net K+ uptake (or loss) was measured by taking solution samples at 1, 4, 12, 24, 48 and 72h (Experiment 3). Tissue concentrations of Na+, K+ and Cl− were measured at 0, 48 and 72h after reaeration. The reaeration after exposure to 50 mm NaCl and anoxia was carried out using non-saline solutions with or without mannitol (iso-osmotic to the previous NaCl) to test for possible hypo-osmotic shock.
Tissue Na+, Cl− and K+ were extracted in 0·5 m HNO3 for 2 days. Cl− was measured using a Buchler-Cotlove chloridometer (Buchler Instruments, Model 4-2008, Fort Lee, NJ, USA). Na+ and K+ were measured using a flame photometer (Corning Medical and Scientific, Model 410, Cambridge, UK). Recoveries of Na+, Cl− and K+ from a reference tissue were 87–95 %; the data were not adjusted. Tissue ion concentrations were calculated on a fresh weight basis (μmol g−1 f. wt) and for Experiment 4 on a tissue water basis (mm), in the latter case using the measured fresh and dry weights of the coleoptile tissue to calculate tissue water content and assuming 10 % water in the free-space in the coleoptile.
Osmotic potential, of treatment solutions (πsol) and πsap, were measured using a calibrated freezing-point depression osmometer (Model 210, Fiske Associates, Norwood, MA, USA). Tissues which had been blotted to remove surface solution were sealed in cryo-vials, frozen in liquid N2, thawed while in the vials and then crushed in a stainless steel press to extrude sap which was immediately analysed. Turgor pressure was assessed as the difference between πsap and πsol, a suitable approach for submerged tissues (Kurniasih et al., 2013).
Ethanol was assayed in a 1-mL cuvette containing 100 mm glycylglycine buffer and containing 300mm KCl, 1·7 mm NAD+, aldehyde dehydrogenase and alcohol dehydrogenase (adapted from Beutler, 1983). The reaction was monitored at 340nm using a UV-visible spectrophotometer (Shimadzu, model UV-1601, Tokyo, Japan). Respiration (O2 consumption rate of excised and healed coleoptiles) was measured by O2 depletion starting with air-equilibrated solution stirred in a chamber containing tissues and the tip of a Clark-type mini-electrode (Unisense Microrespiration Equipment, Unisense A/S, Aarhus, Denmark; details in Colmer and Pedersen, 2008).
Soluble sugars in tissues were extracted twice in 80 % ethanol by boiling under reflux for 20min. Total sugars (hexose equivalents) were estimated using anthrone (Yemm and Willis, 1954) with absorbance measured at 620nm in a glass cuvette and a spectrophotometer (as described above). Recovery of glucose spiked into tissues prior to extraction was 86–92 % (data were not adjusted). The pellet remaining after the 80 % ethanol extraction was rinsed with water and digested in 1 m KOH and protein was measured according to Lowry et al. (1951).
Pi was extracted from the excised coleoptiles using ice-cold 5 % perchloric acid. Pi in the extracts and also in nutrient solution samples was assayed using the molybdate/malachite green method (Motomizu et al., 1983) and absorbance at 650nm using a spectrophotometer (as described above). The nutrient solution used for excised coleoptiles did not contain Pi, so that even small losses from the tissues could be detected.
Data sets were analysed using the program of general analysis of variance of Genstat 14th edition. Pairwise comparisons used Duncan’s multiple range test.
Growth rates of the coleoptiles of anoxic intact seedlings (fresh weight basis) were reduced by ~15 % at 50 mm NaCl and by ~40 % at 100 mm NaCl, compared with seedlings at 0·3 mm NaCl (Fig. 1). This reduction at 100 mm NaCl in anoxia is somewhat more severe than the ~25 % reduction at 100 mm NaCl in aerated submerged rice seedlings (Kurniasih et al., 2013). Protein concentrations in the coleoptiles of anoxic, intact seedlings were 6·6–8·1mg g−1 f. wt (0·3, 50 and 100 mm NaCl treatments) and the concentration of total sugars (hexose units) was 25–26μmol g−1 f. wt (0·3 and 50 mm NaCl treatment only); in both cases there were no significant differences between NaCl treatments.
In anoxia, the growth of excised coleoptiles was slow at 0·3 mm NaCl and reduced further by ~75 % at 50 mm NaCl. At 100 mm NaCl, the tissues became flaccid and water loss was confirmed by the loss of 50mg g−1 f. wt d−1 over 72h (Fig. 1). Consistently, the assessed turgor pressure was a mere 0·1MPa (Table 2) and tissue K+ concentration was only 6 mm (Table 3). Thus, the excised coleoptiles exposed to anoxia combined with 100 mm NaCl were fatally injured.
Turgor pressure is an important characteristic to evaluate as the associated solute accumulation may cost energy and a consequent reduction in turgor pressure is a possible contributing factor to any growth reduction. In the coleoptiles of intact seedlings at 0·3 mm NaCl, assessed turgor pressure was not substantially affected by anoxia (Table 2), as found before (Atwell et al., 1982). In the anoxic excised coleoptiles turgor pressure was already lower than in aerated solution at 0·3 mm NaCl, and further declined in the NaCl treatments (Table 2). The more negative πsap at 50 and 100 mm NaCl was associated with the uptake of Na+ and Cl− (Table 3), so that even though tissue K+ concentration had declined (Table 3) the sum of these three ions to πsap had still substantially increased (Table 2); the individual contributions of these three ions to πsap are presented in Supplementary Data Table S1.
In experiments involving exposure to adverse factors it needs to be established whether the tissues are able to recover when returned to optimal conditions; otherwise, the data obtained during the exposure to anoxia and 50 or 100 mm NaCl might merely reflect the consequences of injury rather than acclimation.
In the intact seedlings, which had been at 0·3, 50 and 100 mm NaCl, all treatments doubled in fresh weight over 50h after return to aerated 0·3 mm NaCl in nutrient solution (data not shown) showing an ability to resume growth, but not excluding possible repair of any injury upon return to aerated nutrient solution. The recovery by the submerged rice seedlings previously in anoxia and NaCl in our experiments compares favourably with the 50 % death of leaves of rice after 6 d with roots at 50 mm NaCl in nutrient solution (shoots in air, Yeo and Flowers, 1983).
For the excised coleoptiles it was particularly important to evaluate recovery, because the excised coleoptiles at 50 mm NaCl were used in the assessments of energy formation and possible energy requirements for acclimation to NaCl when in anoxia; by contrast, the excised coleoptiles in anoxia with 100 mm NaCl were severely injured and were not used further. During the first hour after return of excised coleoptiles from anoxia with 50 mm NaCl, to aerated solution at 0·3 mm NaCl, there was a rapid resumption of aerobic metabolism as shown by substantial net K+ uptake at 2·5–3·0μmol g−1 f. wt h−1 (Fig. 2). These rates of net K+ uptake during the recovery phase were similar to those measured for rice coleoptiles continuously in aerated solutions (Colmer et al., 2001). Rates of net K+ uptake were similar, irrespective of whether the coleoptiles had been in anoxia at 0·3 or 50 mm NaCl, and without or with mannitol during the recovery phase to avoid any possible hypo-osmotic shock at the time of transfer from 50 to 0·3 mm NaCl (Fig. 2). This rapid recovery upon reaeration and removal of NaCl occurred even though the coleoptiles had experienced small net K+ losses during the preceding 40h in anoxia with 50 mm NaCl (Fig. 2); because injury was unlikely (see also next paragraph) such K+ net losses during long-term anoxia are most simply interpreted as providing some energy via catabolism of organic acids, with release of the balancing cation K+ (referred to as a ‘battery’, see Atwell et al., 2015).
Rapid K+ accumulation upon reaeration indicates the tissue had not suffered serious injury, as accumulation of K+ and its counter anion requires membrane integrity and unimpaired membrane functions. Substantial Pi loss to the medium was a good indicator for serious damage in wheat shoots when exposed to anoxia, as shown by loss of 4·8 % h−1 of the tissue Pi content over 9h in anoxia (Menegus et al., 1991). The Pi loss to the medium of the present excised rice coleoptiles was about 0·19 % h−1 of the initial tissue Pi content over 72h in anoxia combined with 50 mm NaCl (data not shown), supporting that any possible injury was slight. In conclusion, membrane integrity and function were not compromised, or at least rapidly restored, after return of excised coleoptiles from anoxia combined with 50 mm NaCl to aerated, non-saline solution.
Transfer of the excised coleoptiles from anoxic 50 mm NaCl to aerated 0·3 mm NaCl also showed that tissue Na+ concentrations only decreased by 35 % over 48h (Fig. 3). This decrease must be mainly via efflux, rather than from dilution by growth, because these excised coleoptiles increased only 15 % in fresh weight. This slow decrease in tissue Na+ concentration is consistent with the very slow Na+ re-translocation from leaves of intact barley plants with shoots in air, which was attributed to an inability to absorb large amounts of Na+ into the phloem, because there was substantial Na+ net loss from the sheath (Greenway et al., 1965). However, because in the submerged rice coleoptiles Na+ efflux could occur to the external solution without the need for phloem transport, the present data indicate that the slow declines in shoot tissue Na+ concentrations might be more a question of Na+ retention by the rice coleoptile cells and also by the leaf cells in the earlier case for barley, presumably mainly in the vacuoles.
Excised coleoptiles were used to assess rates of catabolism, which in turn were required to draw up an energy budget; this estimate cannot be derived when using the whole seedlings, which would give the sum of the coleoptile and seed. The detailed observations on energy formation and ion uptake with time after addition of NaCl are presented for the 50 mm NaCl treatment, because as shown previously the excised coleoptiles at 100 mm NaCl were fatally injured; therefore, no meaningful assessment of energy production can be made for that NaCl level, at least during experiments lasting ~ 80h. Total sugar concentrations (hexose units, μmol g−1 f. wt) in the excised coleoptiles at 72h of NaCl treatments were: aerobic tissues 94 ± 9 at 0·3 mm NaCl and 66 ± 3 at 50 mm NaCl; anoxic tissues 103 ± 7 at 0·3 mm NaCl and 86 ± 5 at 50 mm NaCl.
Ethanol formation in anoxic excised coleoptiles was 12 % higher at 4–6, 34–36 and 70–72h after addition of NaCl than at 0–2h and 88–90h (Fig. 4A). O2 consumption by excised coleoptiles in aerobic solution was stimulated by 13–15 % at 4–6h onwards following exposure to 50 mm NaCl (Fig. 4B). In contrast to anoxic coleoptiles in which ethanol production again declined at 88–90h after expose to 50 mm NaCl, the stimulation of O2 consumption in aerobic coleoptiles persisted (Fig. 4B) even when these aerobic tissues had reached flux equilibrium (Figs 6A, B and 7). The difference in catabolic rates between coleoptiles in 0·3 and 50 mm NaCl was assessed to increase ATP production in anoxia via glycolysis by about 0·8μmol ATP g−1 f. wt h−1 and in aeration via respiration by 13–15μmol ATP g−1 f. wt h−1 (Fig. 4, Table 4).
The large differences in stimulation of assessed ATP production in excised coleoptiles in anoxic and aerobic solutions have now set the scene to consider the changes in tissue Na+, Cl− and K+ concentrations and the net fluxes of these ions with time.
The tissue Na+ and Cl− concentrations increased hyperbolically with time in the coleoptile and endosperm of intact seedlings, and in excised coleoptiles (Figs 5 and 6). The step up from 50 to 100 mm NaCl resulted in further increases in tissue Na+ and Cl− concentrations (Fig. 5). The coleoptiles’ quasi steady-state Na+ and Cl− concentrations were above the external concentrations (Figs 5 and 6). In contrast, the endosperm reached only external Na+ and Cl− concentrations (Fig. 5). An aerated comparison is available for the excised coleoptiles, showing similar pattern as in the anoxic treatments, but with rates of uptake and quasi steady-state concentrations 2–3 times higher in aerated than in anoxic coleoptiles (Fig. 6).
In the coleoptile and endosperm of the intact, anoxic seedlings, K+ concentrations decreased from ~140 to 60–70 mm (tissue water concentrations), with only slight differences between the 50 and 100 mm NaCl treatments (Fig. 5). The excised coleoptiles contained only ~ 50 mm K+ at the start of the experiment (Fig. 6). Under aeration, tissue K+ concentration increased, more so in coleoptiles at 0·3 than at 50 mm NaCl (Fig. 6). Under anoxia, K+ concentrations were similar for excised coleoptiles at 0·3 and 50 mm NaCl and tissue K+ remained at ~ 50 mm over the 90h of the treatments (Fig. 6). The exception was the earlier discussed case of anoxic excised coleoptiles at 100 mm NaCl, which had very low K+ concentrations in the tissue (Table 3).
Net uptake rates of Na+ and Cl− by excised coleoptiles in anoxia were initially as high as 6·5 and 3·5μmol g−1 f. wt h−1, respectively, during the first few hours (Fig. 7). These rates of ion net uptake in anoxia were about half (Na+) and one-third (Cl−) those in excised coleoptiles in aerated solution. Both in anoxia and in aerated solutions, the rates of net uptake of both ions declined with time and were very small by 70–90h after adding 50 mm NaCl (Fig. 7).
The key question posed in this paper is by what mechanisms anoxic rice seedlings, which are in an ‘energy crisis’, cope with the additional energy demands imposed by high NaCl. We first established the anoxic, submerged seedlings grew even at 100 mm NaCl, reinforcing the reputation of very high anoxia tolerance of germinating rice and the coleoptile tissues. Results on energy production by excised, anoxic coleoptiles at 50 mm NaCl, which showed no sign of injury, revealed that there must have been large economies in energy expenditure in the anoxic coleoptiles compared to the aerated ones (Table 4), we suggest by energy-efficient transporters and reductions in membrane permeability (discussed below).
The higher tolerance of anoxia combined with high NaCl of coleoptiles of intact seedlings than of excised coleoptiles (Fig. 1) is probably due to the intact seedlings having sucrose from their seeds as substrate, compared to exogenous glucose for the excised coleoptiles; use of sucrose can yield 25–50 % more ATP than glucose (Atwell et al., 2015). Moreover, in intact seedlings energy spent in the seed is recouped in the coleoptile (Atwell et al., 2015; Supplementary Data Appendix S1). For the excised coleoptiles in anoxia, 100 mm NaCl was fatal presumably because the energy requirement for cell maintenance exceeded the energy that could be produced (Table 4). The energy deficit would lead to collapse of transmembrane gradients (Felle, 2005), essential for K+ homeostasis in the cytoplasm (Leigh and Wyn Jones, 1984) and to sequester Na+ in the vacuole (Munns and Tester, 2008). ‘Ion toxicity’ from high Na+/K+ ratio in the cytoplasm might be a secondary effect, triggered by the collapse of the transmembrane gradients.
For intact seedlings, the most plausible cause for the growth reduction when anoxic and combined with 100 mm NaCl, compared with 0·3 mm NaCl, is that under anoxia processes essential for survival have priority for energy resources over growth (Greenway et al., 2012; Atwell et al., 2015). The energetic costs of sustaining ion fluxes and cellular compartmentation are presumably the main cause of the poorer growth of anoxic seedlings at 100 mm than at 0·3 and 50 mm NaCl. Another possibility for the reduced growth of coleoptiles at high NaCl is reduced turgor pressure (Table 2). However, a decrease in turgor pressure may be compensated for by an increased cell-wall extensibility (see Cosgrove, 1988); if so, a reduced turgor pressure would be of acclimative advantage in anoxia as less energy would need to be spent on accumulating ions in the vacuole and energy-expensive organic solutes in the cytoplasm (for these energy costs see: Raven, 1985; Over, 1999). In the anoxic coleoptiles at 50 mm NaCl, total sugars showed no increase above non-saline controls (Table S1), but redistribution of sugars, or of other organic solutes, between vacuole and cytoplasm is possible and such transport costs would be small relative to the high energy cost of synthesis of organic solutes. The need for organic solutes increases with higher salinity, because NaCl and KCl above 100 mm slow rates of enzyme reactions and therefore compatible solutes in the cytoplasm are required (Flowers et al., 1977; Greenway and Munns, 1980), but the energy cost varies widely depending on the particular organic solute and other conditions (discussed in Supplementary Data Appendix S2). Importantly, in the coleoptile of anoxic rice seedlings, Na+ and Cl− as high as 140 mm can function as ‘cheap’ osmolytes. Consistently, anoxic rice coleoptiles can transport Cl− against its free energy gradient (Zhang and Greenway, 1995; Greenway et al., 2012).
Excised rice coleoptiles were quite tolerant to at least 76h exposure to anoxia combined with 50 mm NaCl, as shown by the high rates of net K+ uptake during the first hour after return from this treatment to aerated non-saline solution (Fig. 2). So, there was no detectable injury, either during exposure to 50 mm NaCl in anoxia or by possible free radical formation after return to aerated solution; the latter is claimed to be a major problem in other species (Supplementary Data Appendix S3). Furthermore, after return to aerated non-saline solution, Na+ concentration in these excised coleoptiles decreased only slowly from 89 to 56 mm over 72h (Fig. 3). These slow decreases support the suggestion by Kurniasih et al. (2013) that submerged rice can be used to evaluate tissue tolerance to Na+. As well, tissue Cl− concentrations even increased (Fig. 3). Thus, there is a prima facie case that these tissues did not suffer Na+ and Cl− toxicities. At the end of anoxia K+ was at least 50 mm in the tissue as a whole, which should ensure 100 mm K+ in the cytosol of the cells (Leigh and Wyn Jones, 1984; assessed for aerated tissues). Thus, the excised coleoptiles at 0·3 and 50 mm NaCl provide an experimental system to evaluate energy production in glycolysis and measure ion net fluxes as affected by salinity in an anoxia-tolerant plant organ/tissues.
The increase of 12 % in ethanol formation in response to 50 mm NaCl by anoxic excised rice coleoptiles (Fig. 4A) supports the suggestions by Colmer et al. (2001) and Huang et al. (2005) that under anoxia without additional adverse factors, rice coleoptiles down-regulate ethanol formation, once acclimation with its high energy demands is completed. The acclimative advantage of such down-regulation of ethanolic fermentation would be to prolong the availability of substrates needed to survive extended periods of anoxia (cf. Drew, 1997; Gibbs and Greenway, 2003). NaCl in the present experiments was applied to anoxia-acclimated coleoptiles; in these anoxic coleoptiles glycolysis, as estimated from ethanol formation, was stimulated by exposure to 50 mm NaCl, presumably because of increased energy requirements during NaCl exposure, associated with Cl− uptake, Na+ extrusion across the plasma membrane, and Na+, Cl− and K+ compartmentation between cytoplasm and vacuoles.
We have estimates of rates of energy production for excised coleoptiles exposed to 0·3 and 50 mm NaCl (Table 4). Importantly, tissue sugars were at 86μmol hexose equivalents g−1 f. wt in the 50 mm NaCl treatment, which would be adequate for fast rates of glycolysis (Huang et al., 2005), and these tissues did not suffer injury (see Results). The estimates in Table 4 are approximate owing to various uncertainties; as one example PPi formed during macromolecule synthesis may be recouped in glycolysis (Plaxton and Tran, 2011; Atwell et al., 2015), but in the present case this is likely to be small because most of the PPi is formed during macromolecule synthesis and growth was slow in the excised coleoptiles.
The estimate for the ATP required for survival of the coleoptiles in anoxia combined with 50 mm NaCl is 6·5μmol ATP g−1 f. wt h−1 (row 4, Table 4); this is the sum of 0·8 + 2·4 + 3·3, being the increase in ATP synthesis in glycolysis, ATP diverted from growth reduction and ‘basic’ cell maintenance, respectively (Table 4). This amount is slightly lower than the assessed ATP production in sugar catabolism during glycolysis of 7·3μmol ATP g−1 f. wt h−1 in these coleoptiles (row 3a, Table 4). At 100 mm NaCl, the ATP requirement is assessed at 9·7μmol ATP g−1 f. wt h−1 (row 4, Table 4), which exceeds the amount of ATP produced (row 3a, Table 4). Although based on several assumptions, this estimate is large enough to support the notion that at 100 mm NaCl, the fatal injury to the anoxic excised coleoptiles is due to the ATP demand exceeding the ATP produced. In the anoxic coleoptiles at 50 mm NaCl, the increased ATP required during exposure to 50 mm NaCl is for 27 % derived from the stimulation in ethanol formation and for 73% by ATP saved due to the reduction in growth. Experiments on the dose–responses of anoxic coleoptiles to NaCl between 0·3 and 50 mm are needed to determine whether growth is reduced in preference to an increase in glycolysis, or vice versa.
For the anoxic coleoptiles at 50 mm NaCl, the ATP required for coping with 50 mm NaCl was 3·2μmol ATP g−1 f. wt h−1 (footnote to Table 4), which was much less than the apparent increase for energy in the aerobic coleoptiles of 13μmol ATP g−1 f. wt h−1 based on the stimulation of respiration (Table 4), i.e. the anoxic coleoptiles were more economical in use of energy. This already substantial economy in energy requirement during anoxia as compared with aerobic tissues is a minimum estimate, because the value for the aerobic tissues is based only on the increased respiration and there might also be some ATP diverted from growth upon exposure to 50 mm NaCl in aerated conditions (data on growth are not available so this is not known). Although the absolute amount of ATP used for acclimation to NaCl was lower in anoxic than in aerated coleoptiles, as proportions of the amounts of ATP produced these were 45 % in anoxic and 12 % in aerated coleoptiles. The substantially lower amount of ATP required to acclimate to 50 mm NaCl during anoxia than in aerated solution raised questions on the types of membrane ion transporters which may become engaged, as discussed in the next section and considered further in the General Discussion.
Assessments of energy requirements for Na+ and Cl− fluxes are made relative to the free energy gradients of ions, which have both membrane potential and ion concentration components. The usual negative membrane potential of the plasma membrane of plant cells will promote Na+ influx along its free energy gradient and require energy to sustain a Cl− influx against its free energy gradient. The plasma membrane potential of anoxic rice coleoptiles is only known for low NaCl conditions in which it stabilizes at−110 to−120mV (Zhang and Greenway, 1995). Upon exposure of barley roots to NaCl, plasma membrane potentials increased (became less negative); after 5 d at 80 mm NaCl the potentials were –60mV compared to –120mV at low NaCl (Bose et al., 2014). Even if such depolarization occurred in anoxic rice coleoptiles, at 50 mm NaCl, Na+ influx will be along a free energy gradient and Cl− against a free energy gradient.
Here we will consider the energy requirements for ion transport during the first 3h of exposure to 50 mm NaCl of anoxic coleoptiles, with a rate of uptake of 3μmol Cl− g−1 f. wt h−1 (Fig. 7). First, assuming this import is via an H+–Cl− symport with coupling ratio of 2:1, the H+ influx would be 3 × 2 = 6μmol H+ g−1 f. wt h−1. Furthermore, the H+-ATPase at the plasma membrane usually has a coupling ratio (ATP/H+) of 2 (Ullrich and Novacky, 1990; Felle, 1996), which would give for the observed Cl− uptake a requirement of 12μmol ATP g−1 f. wt h−1 to pump these H+ ions out of the cytoplasm. Yet, during the first 2h of anoxia, the ATP production assessed from ethanol formation was only ~ 8μmol ATP g−1 f. wt h−1 (Fig. 4A, first 2h when there is as yet no stimulation of ethanol formation). Hence, the classical assumptions on the cost of Cl− transport are quite implausible. Similar implausible high energy costs for active Cl− fluxes in aerobic animal tissues at high NaCl have been called ‘Ussing’s conundrum’ and led to a search for alternative transport systems costing no energy, or at least less energy, than the conventional transport systems (Britto and Kronzucker, 2009). In the General Discussion we will argue that for anoxic rice coleoptiles at 50 mm NaCl during an energy crisis there are at least three plausible mechanisms which might minimize energy expenditure related to ion transport: (1) Cl− uptake along a free energy gradient, which is feasible, but only after the initial period following NaCl addition, i.e. at a high / and depolarized plasma membrane potentials; (2) low coupling ratios of the H+-ATPase, H+–Cl− symporters and H+–Na+ antiporters, i.e. low energy costs of transport; and (3) during the quasi steady-state, low membrane permeability for Cl− and Na+. These possibilities will be considered in the General Discussion in relation to the literature on ion transport.
In this General Discussion we suggest a number of mechanisms by which energy expenditure on ion transport can be minimized during the ‘energy crisis’ during anoxia.
In the first hours after exposure to high NaCl, Cl− may be transported along a free energy gradient as may be substantially lower than (Teakle and Tyerman, 2010). As well, the plasma membrane potential may depolarize due to Na+ influx via a Na+-permeable uniport (Teakle and Tyerman, 2010). Then, influx of both Na+ and Cl− would be along a free energy gradient (Teakle and Tyerman, 2010) as considered further in Supplementary Data Appendix S4. Consistently, in the present experiments there was no stimulation of ethanol formation, and by inference energy production, during the first 2h after an increase to 50 mm NaCl. The short term of the period that the Cl− influx may cost no energy should not depreciate its acclimative value, because substantial Cl− uptake may be involved during this early period at 50 mm NaCl, providing a solute for turgor regulation. The energy saved by energy-independent Cl− uptake would be quite substantial; even assuming the most favourable coupling ratios (discussed below), energy-dependent Cl− transport would cost 3μmol ATP g−1 f. wt h−1. These economies would be particularly useful, as they occur during the early phase of acclimation, when energy economy is particularly important because there may be additional energy requirements for adjustments to the more negative, external, osmotic potentials. It should be noted that depolarized membrane potentials in anoxic rice coleoptiles at high NaCl are a reasonable speculation, but these have not been measured.
An energy-efficient coupling ratio of 1 or even lower has been found for the plasma membrane H+-ATPase in energy-depleted Neurospora (Warncke and Slayman, 1980), for vesicles from the plasma membrane of Beta vulgaris (Briskin and Reynolds-Niesman, 1991) and during a study on mutants of the plasma membrane H+-ATPase of yeast (Baunsgaard et al., 1996). Whether this potential energy saving is possible depends on the free energy gradient for the ‘uphill’ H+ extrusion, so would depend on conditions as further considered in Appendix S4.
Substantial Cl− uptake in anoxic intact rice coleoptiles, at 0·3 mm NaCl, must have occurred as these accumulated 45 mm Cl− over 75h (Fig. 5C). Furthermore, excised coleoptiles at 0·3 mm NaCl also accumulated Cl− at 0·5μmol g−1 f. wt h−1 between 60 and 84h of anoxia (Greenway et al., 2012), although this did not occur in the present experiments (Fig. 6B). This discrepancy indicates it is worthwhile to test further possible energy-dependent transport, which may be unique for extremely anoxia-tolerant species. When energy is required for Cl− transport across the plasma membrane, the energy cost would depend on the coupling ratio of the H+–Cl− symporter. The theoretically possible ratio can be deduced from the difference in free energy of H+ (‘downhill’) and Cl− (‘uphill’) across the plasma membrane. For a membrane potential of−100mV and cells with 30–50 mm cytoplasmic Cl−, the H+–Cl− coupling ratio could readily be 1 rather than the 2 which is assumed for Cl− transport during aeration and low NaCl (Appendix S4).
The combination of low coupling ratios of the H+-ATPase and the H+–Cl− symporter would reduce the energy requirement for Cl− uptake from 12 to 3μmol ATP g−1 f. wt h−1. This amount is about the same as the extra ATP becoming available from the combination of the reduction in growth and increase in glycolysis (Table 4).
Transport of Na+ and Cl− across the tonoplast into the vacuole may require energy, depending on the concentration gradient and tonoplast membrane potential. Teakle and Tyerman (2010) considered that at a ‘normal’ tonoplast potential, the ratio of Cl− in the vacuole/cytoplasm would be 3 when at diffusion equilibrium (tonoplast potential of+30mV, based on the Nernst equation). That is, Cl− flux from cytoplasm into the vacuole could occur along a free energy gradient (‘downhill’) until / of 3 was reached. Importantly, compartmentation of Cl− into vacuoles would keep cytoplasmic Cl− concentration low, which could prolong the period of Cl− influx across the plasma membrane along a free energy gradient, which would be a considerable energy saving for Cl− uptake for osmotic adjustment, as compared with Cl−:H+ symport and associated engagement of the H+-ATPase (Appendix S4). This ‘energy-economical’ Cl− uptake could occur even in tissues that contain 30 mm Cl− (e.g. at the tissue concentrations in rice seedlings, Fig. 5C), if vacuolar compartmentation is as described above. Na+ transport across the tonoplast would require energy because it would be ‘uphill’ against an electrochemical gradient as the tonoplast potential is positive inside and cytoplasmic Na+ is maintained at low concentrations (Munns and Tester, 2008). Energy requirements for ion transport at the tonoplast would increase with rising vacuolar Na+ concentrations and for Cl− when / increased above 3 as then Cl− too would be ‘uphill’ (if tonoplast potential is+30mV).
Once reaching the quasi steady-state, maintaining low Na+ and Cl− but high K+ concentrations in the cytoplasm is likely to be the main energy cost associated with exposure to high NaCl for non-growing cells. First considering K+, reasonable cytoplasmic K+ concentrations can be maintained without energy expenditure. Taking a tonoplast potential of+30mV, the ratio of / concentration at equilibrium would be 3·3 (calculated using the Nernst equation), so with at least 50 mm K+ in the cells (Figs 5 and 6), no energy is required to maintain high cytoplasmic K+ concentrations. Therefore, Na+ fluxes are likely to be the main energy-requiring ion transport process, with influxes into the cytoplasm due to more negative membrane potential than the external solution and the vacuole. These Na+ influxes require equivalent effluxes from the cytoplasm to both the external solution and the vacuole; both would be enacted by H+–Na+ antiporters. Low energy costs to maintain Na+ compartmentation were proposed for aerobic tissues at high NaCl (Niu et al., 1995), but during the ‘energy crisis’ of anoxia even small additional costs could present a challenge. The energy cost comes from pumping the H+ moved by the H+–Na+ antiporter back into the vacuole, which requires 12·5kJ mol−1 (with +30mV and pH difference of 2 across the tonoplast) so based on the free energy of hydrolysis of PPi of 27kJ mol−1 (Davies et al., 1993) the H+-PPiase could, in principle, have a coupling ratio of PPi-H+ of 0·5. If the coupling ratio for the H+–Na+ antiporter is 1 then Na+ transport across the tonoplast may cost 0·5mol PPi per mol of Na+, but if it were 0·5 then the cost would be 0·25mol PPi per mol of Na+; these possibilities for the H+–Na+ antiporter coupling ratio will also depend on the Na+ concentration gradient across the tonoplast (i.e. on the amounts of Na+ in the cytoplasm and in the vacuole) (Appendix S4). Taking examples of 20 and 50 mm Na+ in the cytoplasm, with 90 mm Na+ in the vacuole, we highlight that the costs of Na+ compartmentation could differ under these two scenarios as theoretically the coupling ratio of the Na+–H+ antiporter might differ owing to the lower free energy gradient for ‘uphill’ Na+ transport into the vacuole with Na+ at 50 mm rather than 20 mm in the cytoplasm (Appendix S4). For consideration of Na+ concentration in the cytoplasm at which possible adverse effects might occur, see Munns and Tester (2008) and Cheeseman (2013). Finally, the likely switch from dominant use of the tonoplast H+-ATPase to the H+-PPiase during anoxia cannot be considered as an energy-saving device, as in anoxic rice coleoptiles any ‘excess’ PPi can conserve equivalent amounts of ATP by use of PPi-dependent glycolytic enzymes (Plaxton and Tran, 2011; Atwell et al., 2015); instead, the H+-PPiase would be acclimative because it would direct energy to the crucial maintenance of cellular compartmentation (Atwell et al., 2015).
The present data on assessed ATP formation by anoxic rice coleoptiles indicate the energy costs during the quasi steady-state in anoxia may be slight, as there was no stimulation of ethanol formation (Fig. 4A) once ion flux equilibrium was reached (Fig. 7). This low amount of ATP required during the quasi steady-state of Na+ and Cl− fluxes is consistent both with the ‘channel arrest’ in animal tissues during an energy crisis (Hochachka, 1986) and with the hypothesis by Felle (2005) that in plant tissues H+-co-transport during an energy crisis is down-regulated, lest the plasma membrane H+-ATPase would consume too much energy. The decrease in K+ permeability of membranes has been assessed at 17-fold for anoxic rice coleoptiles (Colmer et al., 2001) and at 10-fold for hypoxia-tolerant Vitis roots (Manusco and Marras, 2006). Evidence for down-regulation of Na+-permeable channels in the tonoplast was shown in mature leaves of the salt-tolerant species Chenopodium quinoa at 400 mm NaCl; by contrast, young leaves which keep Na+ low by salt excretion show no such regulation (Bonales-Alatorre et al., 2013). So, for an energy deficit it is quite conceivable that fluxes through both Na+- and Cl−-permeable uniports, or channels, can also be down-regulated. Otherwise, there would not only be substantial Na+ fluxes into the cytoplasm, but also a Cl− efflux from cytoplasm to outside, along its free energy gradient, so that a compensatory Cl− influx would then be needed to maintain cellular Cl− needed for osmoticum and the likely H+–Cl− symport involved in uptake would cost precious energy. The suggestion that during flux equilibrium, i.e. the quasi steady-state, the Na+ and Cl− fluxes are much higher in aerobic than in anoxic tissues is consistent with the continued stimulation of O2 uptake by 50 mm NaCl for coleoptiles in aerated solution, even when the tissues had reached flux equilibrium (Fig. 4C). In aerated solution this energy may be needed not only for the H+–Na+ antiporters but also for the H+–Cl− symporters, needed to counter Cl− effluxes along their free energy gradient across the plasma membrane; these were, for example, ~ 4μmol g−1 f. wt h−1 for aerated roots of intact barley at 10 and 100 mm NaCl (Britto et al., 2004).
This study demonstrates that anoxia-tolerant coleoptiles of rice survive anoxia and there is even no injury when additional energy expenditure is imposed by exposure to 100 mm NaCl for intact seedlings and to 50 mm NaCl for excised coleoptile tips. The increase in energy production as judged from increased ethanol formation by excised coleoptile tips in the NaCl treatment supports the notion that in the anoxia-tolerant coleoptile of rice, glycolysis is down-regulated after acclimation to anoxia (Colmer et al., 2001), and that glycolysis can be stimulated again by factors which impose additional demand for energy. Apart from this increased energy production there was also re-allocation of energy ‘saved’ by reduced growth, and the latter was of greater magnitude than the energy assessed to have been derived from the stimulation in glycolysis linked to ethanol formation; these two amounts together would have accommodated the higher energy requirement associated with Na+ and Cl− fluxes (established at 50 mm NaCl). The energy requirements associated with these ion fluxes were 4·5-fold lower in anoxic than in aerobic coleoptiles. This economy might be achieved by engagement of energy-efficient ion transporters, while reductions in permeability to Cl− and Na+ of the plasma membrane and tonoplast might play an equally important role.
Supplementary data are available online at www.aob.oxfordjournals.org and consist of the following. Table S1: contributions of K+, Na+, Cl− and total sugars (hexose equivalents) to πsap in the coleoptile of intact rice seedlings in anoxia, and in excised coleoptile tips in anoxia and in aerated solution. Appendix S1: possible causes for collapse of anoxic, excised coleoptiles at 100 mm NaCl. Appendix S2: comparison of the energy costs of organic solutes and Na+ and Cl− as osmolytes. Appendix S3: evidence for the negligible effect of free radicals of oxygen in rice coleoptiles previously in anoxia with 50 mm NaCl. Appendix S4: detailed discussion on Na+ and Cl− transport during the initial period after exposure of anoxic coleoptiles to NaCl.
B.K. acknowledges The University of Gadjah Mada, Indonesia, for the opportunity to undertake PhD studies at The University of Western Australia. We thank the reviewers for their constructive criticisms which helped us to improve the paper.