Response of cytoplasmic pH to anoxia in plant tissues with altered activities of fermentation enzymes: application of methyl phosphonate as an NMR pH probe

Abstract

Background and Aims

Acidification of the cytoplasm is a commonly observed response to oxygen deprivation in plant tissues that are intolerant of anoxia. The response was monitored in plant tissues with altered levels of lactate dehydrogenase (LDH) and pyruvate decarboxylase (PDC) with the aim of assessing the contribution of the targeted enzymes to cytoplasmic pH (pH_cyt) regulation.

Methods

The pH_cyt was measured by in vivo ³¹P nuclear magnetic resonance (NMR) spectroscopy using methyl phosphonate (MeP) as a pH probe. The potential toxicity of MeP was investigated by analysing its effect on the metabolism of radiolabelled glucose.

Key Results

MeP accumulated to detectable levels in the cytoplasm and vacuole of plant tissues exposed to millimolar concentrations of MeP, and the pH-dependent ³¹P NMR signals provided a convenient method for measuring pH_cyt values in tissues with poorly defined signals from the cytoplasmic inorganic phosphate pool. Pretreatment of potato (Solanum tuberosum) tuber slices with 5 mm MeP for 24 h did not affect the metabolism of [U-¹⁴C]glucose or the pattern of ¹⁴CO₂ release from specifically labelled [¹⁴C]-substrates. Time-courses of pH_cyt measured before, during and after an anoxic episode in potato tuber tissues with reduced activities of LDH, or in tobacco (Nicotiana tabacum) leaves with increased activities of PDC, were indistinguishable from their respective controls.

Conclusions

MeP can be used as a low toxicity ³¹P NMR probe for measuring intracellular pH values in plant tissues with altered levels of fermentation enzymes. The measurements on transgenic tobacco leaves suggest that the changes in pH_cyt during an anoxic episode are not dominated by fermentation processes; while the pH changes in the potato tuber tissue with reduced LDH activity show that the affected isozymes do not influence the anoxic pH response.

INTRODUCTION

Acidification of the cytoplasm is a characteristic feature of the response of many plant tissues to oxygen deprivation (Ratcliffe, 1999; Felle, 2005). The pH change can be measured directly using in vivo ³¹P nuclear magnetic resonance (NMR) spectroscopy (Roberts et al., 1984; Ratcliffe, 1997), and in recent applications NMR has been used to demonstrate the strong resistance to anoxia-induced cytoplasmic acidosis in the anoxia-tolerant shoot tissues of Potamogeton pectinatus (Summers et al., 2000; Dixon et al., 2006), to establish a role for nitrite in the improved pH regulation of nitrate-grown root tissues under anoxia (Libourel et al., 2006), and to compare the metabolic response of root tips to hypoxia and anoxia (Kulichikhin et al., 2007).

The changes in cytoplasmic pH (pH_cyt) that occur during oxygen deprivation are of particular interest because of the pH-sensitivity of pyruvate decarboxylase (PDC) and lactate dehydrogenase (LDH). These fermentation enzymes are central to the biochemical pH-stat model of pH regulation under anoxia (Ratcliffe, 1995, 1999; Greenway and Gibbs, 2003), and following cloning of the relevant genes there have been numerous attempts to explore the functions of PDC and LDH by manipulating their expression in transgenic plants (Dennis et al., 2000; Dolferus et al., 2003). For example, LDH has been over-expressed in tomato root cultures (Rivoal and Hanson, 1994), antisensed in potato (Sweetlove et al., 2000) and a T-DNA knockout line has been characterized in Arabidopsis (Dolferus et al., 2008); while PDC has been over-expressed in tobacco (Bucher et al., 1994), rice (Rahman et al., 2001) and Arabidopsis (Ismond et al., 2003), and a null mutant in PDC1 has been characterized in Arabidopsis (Kürsteiner et al., 2003). Surprisingly, although the metabolic and physiological phenotypes of these plants have been investigated extensively, there have been no reports on the impact of the changes in LDH and PDC level on pH_cyt and pH_cyt regulation.

One potential problem in such studies is that it can be difficult to perform conventional NMR pH measurements, which exploit the pH-dependent chemical shift of the signal from inorganic phosphate (P_i), on the transgenic material. This problem arises because the value of P_i as a pH marker is compromised if the cytoplasmic P_i signal is weak or poorly resolved, for example in highly vacuolated cells with a small cytoplasmic P_i content, or if the cytoplasmic P_i signal is superimposed on other signals. When this problem was first encountered in bacteria (Slonczewski et al., 1981) and red blood cells (Labotka et al., 1983), it was solved by incubating the cells with methyl phosphonate (MeP). This compound accumulated in the cells, providing a readily detectable pH-dependent ³¹P NMR signal that could be used as an alternative to the endogenous P_i signal. Subsequently, the NMR properties of MeP (Stewart et al., 1986) and numerous other phosphonates (Brénot et al., 1992) were examined in detail, including an analysis of the impact of counterions and ionic strength on the estimation of pH. This led to the routine use of phosphonates as pH probes, particularly in mammalian studies (e.g. Bruynseels et al., 1997). The value of these probes received a further boost with the development of a range of highly sensitive amino-alkylphosphonate derivatives with properties tailored for particular applications (Pietri et al., 2000, 2001; Vidal et al., 2000).

Despite their evident utility, phosphonates have only rarely been used as ³¹P NMR pH probes in plants. In a study on tobacco (Nicotiana tabacum) cells, MeP was used to confirm elicitor-induced changes in pH_cyt deduced from measurements of the cytoplasmic P_i signal (Pugin et al., 1997); while in a study of the anoxic response of sycamore (Acer pseudoplatanus) cells, MeP was used to determine pH_cyt values in phosphate-starved cells that lacked a detectable cytoplasmic P_i pool (Gout et al., 2001). While MeP is generally considered to be non-toxic at low millimolar concentrations (Satre et al., 1989), no evidence has been presented that this is the case in plants.

Accordingly the aims of this study were: (a) to establish the low toxicity of MeP as a pH probe for plants; and (b) to investigate the effect of altered levels of PDC and LDH on the pH response to anoxia in two transgenic plant lines: potato tubers with reduced levels of LDH (Sweetlove et al., 2000) and tobacco leaves over-expressing PDC (Bucher et al., 1994). Measurements of pH_cyt were hindered by the poor definition of the ³¹P NMR signal from the cytoplasmic P_i pool in both tissues, necessitating the use of MeP as a pH probe.

MATERIALS AND METHODS

Plant material

Maize seeds (Zea mays L. cv. B73; ‘Pioneer’) were soaked in deionized water for 30 min and germinated in the dark for 3 d at 25 °C between sheets of absorbent paper soaked in 0·1 mm CaSO₄. Root tips, 5 mm in length, were excised into an oxygenated medium containing 50 mm glucose, 10 mm K-MES (pH 6·0) and 0·1 mm CaSO₄.

Potato tubers (Solanum tuberosum L. ‘Desirée’) were either obtained from local supermarkets or harvested from transgenic plants growing in controlled environment glasshouses. The transgenic lines – two control lines (G111 and G311) and two lines with an antisense construct for LDH (F311 and F312) – were originally generated by Agrobacterium-mediated transformation of stem segments (Sweetlove et al., 2000). Experimental material was obtained by cutting 1-mm cubes of cortex tissue into an oxygenated medium containing 50 mm glucose, 10 mm K-MES (pH 6·0) and 0·1 mm CaSO₄.

Wild-type tobacco (Nicotiana tabacum L. ‘Samsun’) plants and the transgenic line 9204-X were grown for between 1 and 3 months in controlled environment glasshouses. The transgenic line was originally generated by Agrobacterium-mediated transformation of wild-type leaves with a construct that conferred constitutive over-expression of PDC (Bucher et al., 1994). Healthy leaves from positions 3 or 4 were used as experimental material.

Incubation with MeP

Excised maize root tips and potato tuber discs were incubated with MeP in solutions containing 200 mm glucose, 10 mm K-MES (pH 6·0) and 0·1 mm CaSO₄. Excised tobacco leaves were incubated in the light for 16–18 h in solutions containing 5 or 10 mm MeP.

Preparation of plant material for in vivo NMR spectroscopy

Maize root tips (typically 100 tips) or potato tuber fragments (typically 1 g fresh weight) were transferred to a 10 mm diameter NMR tube containing 10 mm K-MES (pH 6·0) and 0·1 mm CaSO₄. A combined air-lift and circulation system was used to circulate the suspending medium through the NMR tube, and to supply the tissue with either oxygen or nitrogen (Fox et al., 1989, 1995). Tobacco leaf strips (typically 10–16 strips, 5 mm wide and 2·5 cm long) were vacuum-infiltrated in 10 mm K-MES (pH 6·0), 0·1 mm CaSO₄ for up to 5 min to eliminate the intercellular air spaces that would otherwise have caused severe line-broadening in the NMR spectra (Loughman et al., 1989). The stacked strips were transferred in a bundle to the 10-mm-diameter NMR tube containing the same buffer, and they were either oxygenated or nitrogenated in the dark using a simple air-lift system (Fox et al., 1989). All tissues were allowed to equilibrate in the NMR tube for 30 min at 21 °C before starting the in vivo NMR measurements.

In vivo NMR spectroscopy

In vivo ³¹P NMR spectra were recorded at 121·49 MHz on a Bruker CXP 300 NMR spectrometer equipped with an Oxford Instruments 7·05 T magnet and a double-tuned ¹³C/³¹P 10-mm-diameter probe. The acquisition conditions were similar to those described elsewhere (Fox et al., 1995) and spectra were recorded in 5-, 10- or 30-min blocks over periods of several hours as appropriate. Chemical shift values were measured relative to the signal at 22·49 ppm from a capillary containing 2 % (v/v) aqueous solution of the tetraethyl ester of methylene diphosphonic acid, or at 16·95 ppm from a capillary containing 50 or 100 mm methylene diphosphonate (MDP) at pH 8·9. Cytoplasmic pH values were determined from the chemical shift of the cytoplasmic P_i signal, using the calibration curve described elsewhere (Spickett et al., 1993), or from the chemical shift of the cytoplasmic MeP signal, using a calibration curved based on a solution containing 100 mm KCl, 5 mm MgSO₄, 2 mm NaH₂PO₄, 3 mm Na₂HPO₄ and 5 mm MeP. pH values were adjusted using HCl and KOH. Best-fit curves were obtained by fitting the experimental data points to theoretical expressions of the form:

1

where pK_a is the acid dissociation constant of the titratable group, δ is the chemical shift of the observed NMR signal, and δ_HA and δ_A⁻ are the limiting chemical shifts of the signal at low and high pH values, respectively.

Respiration measurements

Oxygen consumption was measured at 25 °C using a Clark-type oxygen electrode (Hansatech Instruments, http://www.hansatech-instruments.com/). Mitochondrial respiration was measured using 10 mm succinate as a substrate.

Extraction of phytate

Frozen potato tuber tissue (1 g) was extracted with trichloracetic acid (Sweetlove et al., 2000). The supernatant was partitioned with water-saturated diethyl ether, and after discarding the ether fraction, the pH was adjusted to 5·5 with KOH.

Measurement of LDH and PDC activities

Tobacco leaf extracts were prepared using 2 mL of extraction buffer (Laszlo and St Lawrence, 1983) per gram fresh weight of tissue as described elsewhere (Bucher et al., 1994). Potato tuber extracts were prepared by freezing 200 mg tissue in liquid nitrogen and homogenizing with a chilled pestle and mortar in 1 ml of extraction buffer containing 100 mm K-HEPES (pH 7·5), 1 mm EDTA, 2 mm MgCl₂, 1 mm DTT, 0·1 % (w/v) sodium metabisulphite and 0·1 % (w/v) poly-(vinylpolypyrrolidone). The sample was centrifuged at 13 000 g and 4 °C for 5 min, and then 0·5 mL of extract was desalted on a NAP 5 column (GE Healthcare Life Sciences, http://www.gelifesciences.com/). Enzyme activities were measured spectrophotometrically using procedures described elsewhere for LDH (EC 1·1·1·27; Sweetlove et al., 2000) and PDC (EC 4·1·1·1; Bucher et al., 1994).

Lactate assay

Vacuum-infiltrated leaf strips (0·5–0·8 g f. wt) from 6-week-old tobacco plants were incubated in 10 mm K-MES (pH 6·0), 0·1 mm CaSO₄ for 2 h under oxygenated conditions, and for a further 1 h under nitrogenated conditions. After incubation, samples were blotted dry, freeze clamped, dropped into liquid nitrogen and extracted with 1·5 mL ice-cold 1·4 m perchloric acid (Kruger et al., 2008). After 2 h on ice, the supernatant was collected by centrifugation and the pellet was washed twice by resuspension in 0·5 mL water. The washings were combined with the initial supernatant, the pH was adjusted to the range 7–8 with 5 m potassium carbonate, the precipitated potassium perchlorate was removed by centrifugation, the pellet was washed with 0·5 mL water, and the washing was combined with the supernatant to give the final extract. The recovery of lactate was checked by adding an aliquot of a standard solution to a replicate sample and was found to be 87 %. Lactate was assayed spectrophotometrically by oxidation to pyruvate (Passoneau and Lowry, 1993).

Isolation of mitochondria

Mitochondria were isolated from potato tubers using differential centrifugation and Percoll density gradient centrifugation (Considine et al., 2003). The final pellet was suspended in 0·3 m mannitol, 10 mm K-TES (pH 7·5), 0·1 % (w/v) bovine serum albumin, 5 % (v/v) DMSO at a concentration of approx. 10 mg mitochondrial protein mL⁻¹ and stored at –80 °C in 100-μL aliquots after quick freezing in liquid nitrogen.

Incubation of tuber discs with ¹⁴C-labelled substrates

Approximately 1 g of tissue was incubated in the dark in a final volume of 5 mL of 10 mm K-MES (pH 6·0), 0·1 mm CaSO₄ in a 100-mL conical flask. Each flask was sealed with a rubber bung and aerated on a rotary shaker at 100 rpm. The incubation was started by addition of 0·1 mL of [¹⁴C]glucose or [¹⁴C]gluconate to a final concentration of 0·3 mm. The specific activity of [1-¹⁴C]-, [2-¹⁴C]-, [3,4-¹⁴C]- and [6-¹⁴C]glucose, and [1-¹⁴C]gluconate was 2·47 MBq mmol⁻¹, and that of [U-¹⁴C]glucose was 24·7 MBq mmol⁻¹. A vial containing 0·5 mL of 10 % (w/v) KOH was suspended in the flask to capture evolved ¹⁴CO₂. In experiments involving specifically labelled substrate, the KOH solution was replaced every hour for the first 8 h and again after 24 h. Incubation with [U-¹⁴C]glucose was terminated after 4 h, when the tissue was collected, washed with approx. 100 mL of unlabelled incubation medium, blotted dry and frozen in liquid nitrogen. The tissue was stored at –80 °C prior to analysis.

Extraction and fractionation of ¹⁴C-labelled material

Frozen tuber slices were extracted in boiling ethanol (Morrell and ap Rees, 1986). The ethanol-soluble fraction was evaporated to dryness, redissolved in 5·0 mL of water and fractionated into acidic, basic and neutral components by ion-exchange chromatography (Kruger et al., 2007). The ethanol-insoluble residue was homogenized in 10 mL water, autoclaved for 3 h at 121 °C (104 kN m⁻²) and then digested with amylase, amyloglucosidase and pronase (Kruger et al., 2007). Solubilized material was fractionated into acidic, basic and neutral components as above.

Determination of radioactivity

The amount of ¹⁴C in aqueous samples was determined by liquid scintillation counting after addition of four volumes of Optiphase ‘HiSafe’ 3 (Wallac, http://las.perkinelmer.co.uk/). The efficiency of counting was typically >90 %.

Statistical analysis

Student's independent two-sample t-test and repeated measures analysis of variance (ANOVA) based on type III sums-of-squares were performed with SPSS 12·0 for Windows (SPSS, http://www.spss.com/). Analysis was conducted on percentage values following arcsine [arcsin(y/100)^0·5] transformation and on ratios after logarithmic transformation (Wardlaw, 1985). Homogeneity of variance between treatments was assessed using Levene's test, and sphericity of the dependent variable(s) in repeated measures ANOVA was confirmed using Mauchly's test. Statistical comparisons for which P is <0·05 are considered significant.

RESULTS

MeP facilitates ³¹P NMR measurements of pH_cyt

Cytoplasmic and vacuolar MeP signals were readily observed in maize root tips after overnight incubation with millimolar concentrations of MeP (Fig. 1A). The two signals do not overlap the metabolite signals (−20 to 10 ppm) and they were assigned to the cytoplasmic and vacuolar MeP fractions from their pH-dependent chemical shift values. The pH calibration curves (Fig. 1B, C) show that the MeP signal (total displacement 3·78 ppm) is more sensitive to pH than the endogenous P_i signal (total displacement 2·39 ppm). As a result, and despite the higher value for the apparent pK_a (7·53 for MeP and 6·67 for P_i), the acidification of the cytoplasm associated with the switch to anoxia resulted in a larger shift in the MeP signal (Fig. 1A). Moreover, there was good agreement between estimates of pH_cyt based on the chemical shifts of the cytoplasmic P_i and MeP signals in maize root tips: under normoxia both signals in Fig. 1A indicated a pH_cyt of 7·5; while under anoxia the pH_cyt value was estimated to be 6·8 from the P_i signal and 6·9 from the MeP signal. The greater shift in the position of the MeP signal under anoxia (Fig. 1A) suggests that MeP is likely to be a more useful pH probe than P_i since it has the potential to define more detailed time-courses of pH_cyt.

Fig. 1.

MeP as a pH probe for maize root tips. (A) ³¹P NMR spectra recorded under oxygenated and nitrogenated conditions after incubation with 5 mm MeP for 18 h. Peak assignments: 1, vacuolar MeP; 2, cytoplasmic MeP; 3, MDP chemical shift reference; 4, cytoplasmic P_i; 5, vacuolar P_i. The imposition of anoxia reduces pH_cyt, causing peak 2 to move to the left and peak 4 to move to the right. (B, C) Calibration curves for the P_i (B) and MeP (C) signals showing the pH dependence of the chemical shifts.

MeP is also useful when the cytoplasmic P_i signal is poorly defined, for example when the cytoplasmic pool is small or when the cytoplasmic P_i signal is hidden by other signals. Figure 2A illustrates this point with spectra of potato tuber tissue. Here the cytoplasmic P_i signal is poorly resolved, because of the overlapping phytate signals (Fig. 2B; Kime et al., 1982), making it difficult to measure pH_cyt. In contrast, the cytoplasmic and vacuolar MeP signals are well resolved (Fig. 2C) permitting accurate estimates of the pH values.

Fig. 2.

MeP as a pH probe for potato tuber tissue. (A) ³¹P NMR spectra recorded under oxygenated and nitrogenated conditions after incubation with 5 mm MeP for 24 h. Peak assignments: 1, vacuolar MeP; 2, cytoplasmic MeP; 3, MDP chemical shift reference. (B) ³¹P NMR spectrum of a potato tuber extract (upper trace) and a deconvolution of the same spectrum (lower trace). Peak assignments: 1, 3, 4, 6, phosphate moieties of phytate showing the expected 1 : 2 : 2 : 1 intensity pattern; 2, glucose 6-phosphate; 5, P_i. (C) Expansion of the spectra in (A), showing the MeP signals and the shift in the cytoplasmic MeP signal in response to the anoxia-induced acidification of the cytoplasm.

MeP has only minor effects on metabolism

Several observations show that incubation with MeP at a level sufficient to generate usable ³¹P NMR signals had only minor effects on the tissue (Fig. 3). Incubation with 5 mm MeP had no effect on the phosphorylated metabolites observed in the in vivo ³¹P NMR spectra of excised maize root tips (Fig. 3A), or on the position of the pH-dependent cytoplasmic P_i signal at 2·8 ppm, implying that the treatment had no effect on bioenergetic status. MeP also had no effect on the respiration of maize root tips (Fig. 3B), potato tuber tissue (Fig. 3C) or tobacco leaves (Fig. 3D) at concentrations up to 20 mm (Fig. 3B), and no effect on the elongation of excised maize root tips at concentrations up to 5 mm (Fig. 3E). Addition of 5 mm MeP to isolated potato mitochondria had no effect on their respiration rate (Fig. 3F); and pretreatment of tobacco leaves with 5 mm MeP had no effect on extractable PDC activity, although an inexplicable increase in this activity was observed at 10 mm (Fig. 3G). Overall these observations suggest that incubation with MeP at concentrations up to 5 mm is benign.

Fig. 3.

Physiological and metabolic effects of MeP. (A) ³¹P NMR spectra of excised maize root tips after 24-h incubation with 0 or 5 mm MEP. Peak assignments: 1, several phosphomonoesters, including (a) glucose 6-phosphate and (c) phosphocholine, with (b) unassigned; 2, cytoplasmic P_i; 3, vacuolar P_i; 4, 5 and 8, the γ-, α- and β-phosphates, respectively, of nucleoside triphosphate; 6, UDP-glucose and NAD(P)(H); and 7, UDP-glucose. (B–D) Respiration rates of (B) excised maize root tips, (C) potato tuber tissue and (D) tobacco leaves after 18 h incubation with MeP. Each value is the mean ± s.e. (n = 3–4) and * indicates values that differ significantly from the 0 mm treatment in Student's t-test. (E) Elongation of excised maize root tips incubated in 0, 1, 5, 20 or 100 mm MeP as indicated. Sets of measurements were taken at the same time but they are staggered on the graph for clarity. (F) Respiration rates of potato tuber mitochondria measured before and after the addition of 5 mm MeP. (G) PDC activity of freshly harvested transgenic tobacco leaves, and after 16–18 h incubation with MeP. Each value is the mean ± s.e. (n = 3) and * indicates values that differ significantly from the 0 mm treatment in Student's t-test.

As a more searching test of this conclusion, the effect of incubating excised potato tuber tissue with 5 mm MeP on the pathways of central carbon metabolism was investigated by radiolabelling. Table 1 summarizes the effect of 5 mm MeP on the metabolism of [U-¹⁴C]glucose. Although MeP pretreatment caused a small increase in the uptake of glucose, there was no change in the redistribution of label over the 4 h experiment. The observed redistribution is similar to that seen in other experiments on aged potato tissue, with a notably high percentage of metabolized label released as ¹⁴CO₂ (Laties, 1964). In a further experiment, the impact of MeP on the pathways of carbohydrate oxidation was investigated by monitoring the release of ¹⁴CO₂ from specifically labelled glucose or gluconate (Fig. 4). As expected for aerobic plant tissues (Stitt and ap Rees, 1978; Malone et al., 2006; Kruger et al., 2007), ¹⁴CO₂ was released in the order C3, 4 > C1 > C6 > C2 from specific positions in glucose (data not shown), reflecting the combined activities of glycolysis, the oxidative pentose phosphate pathway, the tricarboxylic acid cycle, recycling of triose phosphates to hexose phosphates, and pentan synthesis (ap Rees, 1980). The ratio of the release from specific carbon positions in glucose and gluconate was unaffected by preteatment with MeP at 5 or 20 mm (Fig. 4) and a statistical analysis using repeated measures ANOVA confirmed that incubation with 5 mm MeP had no significant effect on metabolism (data not shown).

Table 1.

Effect of MeP on metabolism of [U-¹⁴C]glucose by potato tubers

	Radioactivity in specified fraction in tuber disks
Fraction	Control	5 mm MeP
Ethanol soluble	43·4 ± 3·0	43·3 ± 2·6
(1) Neutral	32·9 ± 2·2	32·5 ± 1·9
(2) Acidic	9·6 ± 0·7	9·1 ± 0·2
(3) Basic	0·9 ± 0·2	1·6 ± 0·4
Ethanol insoluble	28·7 ± 4·5	28·5 ± 4·3
(1) Neutral	14·3 ± 2·6	13·1 ± 1·3
(2) Acidic	11·2 ± 1·2	14·0 ± 2·4
(3) Basic	3·2 ± 0·8	1·3 ± 0·5
Carbon dioxide	28 ± 2·4	28·2 ± 2·2
Total 14C uptake
dpm	346 591 ± 11 104	435 171 ± 17 811*
as % applied	15·6 ± 0·5	19·6 ± 0·8*

Segments of mature tuber parenchyma were incubated for 24 h in buffer with or without 5 mm MeP. The tissue was incubated for 4 h in 5 mL of buffer containing 37 kBq [U-¹⁴C]glucose prior to extraction in ethanol and subsequent metabolic fractionation. The amount of radioactivity in each fraction was determined and expressed as a percentage of total ¹⁴C uptake. Each value is the mean ± s.e. from four independent replicates.

* Values that differ significantly (P < 0·05) from the corresponding values of the control samples.

Fig. 4.

Influence of MeP on the oxidation of respiratory substrates by potato tubers. Segments of mature tuber parenchyma previously incubated for 18 h in buffer containing either 0 (control), 5 or 20 mm MeP as indicated, were incubated for 24 h in 0·3 mm [1-¹⁴C]-, [2-¹⁴C]-, [3,4-¹⁴C]- or [6-¹⁴C]glucose, or [1-¹⁴C]gluconate (2·47 MBq mmol⁻¹). Cumulative ¹⁴CO₂ release was expressed as a proportion of applied label, and then used to calculate the ratio of release of carbon from specific positions within glucose and gluconate. Shown are the ratios of ¹⁴CO₂ release for: (A) [1-¹⁴C]glucose : [3,4-¹⁴C]glucose; (B) [2-¹⁴C]glucose : [3,4-¹⁴C]glucose; (C) [6-¹⁴C]glucose : [3,4-¹⁴C]glucose; (D) [1-¹⁴C]gluconate : [3,4-¹⁴C]glucose; (E) [1-¹⁴C]glucose : [6-¹⁴C]glucose; (F) [1-¹⁴C]gluconate : [6-¹⁴C]glucose; (G) [2-¹⁴C]glucose : [1-¹⁴C]glucose; (H) ( [1-¹⁴C]glucose – [6-¹⁴C]glucose) : [3,4-¹⁴C]glucose. Each value is the mean ± s.e. for ratios determined from three separate tuber samples.

pH_cyt measurements on tissues with altered levels of PDC or LDH

The in vivo ³¹P NMR spectra of tobacco leaves were poorer than might have been expected from those of other leaf tissues (Loughman et al., 1989; Bligny et al., 1997) and the cytoplasmic P_i signal was unsuitable for monitoring pH_cyt during oxygen deprivation (data not shown). In contrast, the chemical shift of the cytoplasmic MeP signal observed in the NMR spectra of tobacco leaves incubated with 5 mm MeP was easily measured under oxygenated and nitrogenated conditions (Fig. 5A). This allowed a direct comparison of the anoxic response of wild-type leaves, and leaves over-expressing PDC (Fig. 5B). Although the extractable PDC activity in the over-expressing leaves under normoxia (66 ± 3 nmol min⁻¹ mg⁻¹ protein; mean ± s.e., n = 4) was over 20 times higher than in the wild type, in agreement with the high PDC activities in the transgenic leaves reported earlier (Bucher et al., 1994), the constitutive over-expression of PDC had no effect on the pH response to anoxia, and on the subsequent recovery of pH_cyt under normoxia (Fig. 5B). In parallel with this observation, it was found that lactate levels in the leaves were low (Fig. 5C), that there was no difference in these levels between the wild-type and transgenic leaves, and that there was also no difference between the extractable LDH activity (Fig. 5D). The LDH activities were comparable to those found in potato leaves (8–15 nmol min⁻¹ g⁻¹ f. wt), but substantially lower than the activity in developing tubers (170–200 nmol min⁻¹ g⁻¹ f. wt; Sweetlove et al., 2000).

Fig. 5.

Effect of an anoxic episode on transgenic tobacco leaves over-expressing PDC. (A) ³¹P NMR spectra recorded: (i) after 2-h normoxia; (ii) after 30-min anoxia; (iii) after 1-h anoxia; and (iv) 2 h after the return to normoxia. Peak assignments: 1, vacuolar MeP; 2, cytoplasmic MeP, with the asterisks emphasizing the changing position of this signal during the time-course. The unlabelled signal in (iv) is from the MDP chemical shift reference. (B) Time-course (mean values; n = 3) for pH_cyt showing the effect of a 1 h anoxic interlude after equilibration under normoxia. pH_cyt was constant in both lines over the last 15 min of anoxia, showing that the pH had stabilized before the switch back to normoxia. The spectra in (A) correspond to the points at 120, 150, 180 and 300 min. (C, D). Lactate content (C) and LDH activity (D) of freshly harvested (control) leaves, after a 2-h incubation under normoxic conditions, and after a further 1 h under anoxia. Each value is the mean ± s.e. (n = 3) and * indicates values that differ significantly from the control in Student's t-test.

MeP pretreatment also provided a convenient route for monitoring pH_cyt in wild-type and transgenic potato tubers (Figs 2A and C and 6). Two lines with an antisense construct for LDH (F311 and F312) had lower LDH activities than the control transgenic lines (G111 and G311; Fig. 6A), but a comparison of the response of pH_cyt to the withdrawal and resupply of oxygen showed no difference between the lines (Fig. 6B).

Fig. 6.

Effect of an anoxic episode on transgenic potato tubers with reduced levels of LDH. (A) LDH activity of the control lines (G111 and G311) and the antisense lines (F311 and F312). Each value is the mean ± s.e. (n = 3). (B) Time-course for pH_cyt in the four transgenic lines showing the effect of a 3-h anoxic interlude after equilibration under normoxia. Each value is the mean ± s.e. (n = 3).

DISCUSSION

As expected, MeP accumulated in the cytoplasm and vacuole of plant tissues incubated with millimolar concentrations of MeP, and the pH-dependent chemical shift of the ³¹P NMR signal from the cytoplasmic MeP pool provided a convenient method for estimating pH_cyt. There was no particular advantage in using MeP in maize root tips, since the cytoplasmic P_i signal was well resolved, but in potato tuber tissue (Fig. 2), where the cytoplasmic P_i signal was swamped by the intensity from the phytate signals, and tobacco leaves, where the cytoplasmic P_i signal was weak, pH_cyt could only be measured reliably from the cytoplasmic MeP signal. Thus incubation with MeP extends the range of plant tissues for which intracellular pH values can be measured by ³¹P NMR.

This conclusion assumes that MeP does not perturb tissue pH values, either directly as a result of MeP uptake, or indirectly through an effect of MeP on metabolism. Several lines of evidence establish that low concentrations of MeP are non-toxic to plant cells. First, a diverse set of measurements (Fig. 3) all suggested that incubation with 5 mm MeP had a negligible effect on maize root tips, potato tuber mitochondria and tobacco leaves. There was some evidence for perturbation at higher concentrations, in agreement with earlier work on Dictyostelium discoideum (Satre et al., 1989), but such high concentrations were not necessary to generate NMR-detectable pools of MeP in root tips, potato tubers and tobacco leaves. Secondly, radiolabelling experiments showed that pretreatment of potato tuber tissue with 5 mm MeP had no effect on the metabolism of glucose (Table 1) and no effect on the pattern of ¹⁴CO₂ release from respiratory substrates labelled at specific carbon atoms (Fig. 4). These experiments provide a direct assessment of the activity of the pathways of central metabolism and carbohydrate oxidation, and the results provide compelling evidence that exposure to low millimolar concentrations of MeP for periods of several hours has no deleterious effects on the tuber tissue. This conclusion is supported by work on other organisms, where it has been shown that MeP and other phosphonates are generally non-toxic (Satre et al., 1989), and by the good agreement between pH measurements based on MeP and P_i signals for sycamore cells (Gout et al., 2001) and maize root tips (Fig. 1A). Thus the assumption that MeP is a benign pH probe in plant tissues is justified by the experimental evidence, and this gives confidence that the pH measurements on the transgenic plant material are valid.

At first sight both transgenic tissues gave unexpected results within the context of the biochemical pH-stat model for pH_cyt regulation under anoxia. In this model, the fall in pH_cyt activates PDC and inhibits LDH, triggering a switch from lactate to ethanol production, that allows the pH to stabilize, for the short term at least, at a value that may, according to species, be up to 1 pH unit lower than the normal aerobic pH_cyt (Ratcliffe, 1999; Felle, 2005). Thus increased expression of PDC, and reduced expression of LDH, might be expected to reduce the extent of the pH fall and/or alter the kinetics of the anoxic response. In the event, while both experimental systems showed the expected acidification of the cytoplasm in response to anoxia, neither transgenic manipulation had any discernable effect on the time-course of the response (Figs 5B and 6B).

The original analysis of the tobacco lines showed an 8- to 20-fold increase in ethanol production during the first 2–4 h of anoxia in leaves over-expressing PDC (Bucher et al., 1994). However, as pointed out subsequently (Gibbs and Greenway, 2003), the anoxic response in tobacco leaves is not well characterized, and the rate of ethanol production in the transgenic leaves is actually very low in comparison with the rates observed in more commonly studied tissues. In fact the data reported here show that the over-expression of PDC had no effect on the pH response to anoxia, indicating either that the contribution of ethanol production to the overall proton balance in the leaves was very small or that the proton-neutral fermentation pathway to ethanol did not replace a significant proton-generating fermentation process, such as that to lactate. Some support for the latter interpretation can be found in the low levels of lactate and LDH activity in the leaves (Fig. 5C, D). Perhaps more importantly the data suggest that the acidification of the cytoplasm is not dominated by the onset of fermentation, and thus provide evidence that the origin of the fall in pH lies in some other metabolic process, for example ATP hydrolysis (Gout et al., 2001) or in biophysical processes related to ion transport. This conclusion does not rule out the likelihood that the balance between the low levels of lactate and ethanol production is determined by the pH sensitivity of PDC and LDH, but it does emphasize, as argued elsewhere (Ratcliffe, 1999), that these processes are not necessarily the only or predominant determinant of the anoxic pH response.

In contrast to tobacco leaves, lactate and ethanol are produced in substantial quantities by anoxic potato tuber tissue and the reduction in LDH activity might reasonably be expected to alter the pH response to anoxia. However, reducing the activity of the two isozymes had no effect on the time-course of pH_cyt under anoxia (Fig. 6) in apparent contradiction of the biochemical pH-stat model for pH regulation. In fact the molecular characterization of the transgenic lines used here revealed that the transgenic manipulation resulted in the preferential reduction in the activity of only two of the five isozymes of LDH (Sweetlove et al., 2000). Moreover the transgenic tubers showed increased levels of lactate under normoxia and no alteration in the accumulation of lactate under anoxia. This suggested that the antisensed isozymes were responsible for the oxidation of lactate to pyruvate in vivo, and that their activity would be irrelevant to the anoxic response, a prediction that is strongly supported by the results reported here (Fig. 6B). Thus the different in vivo functions of the LDH isozymes in potato tubers complicate the apparently simple task of testing the biochemical pH-stat and the role of LDH in anoxic pH regulation.

In conclusion, MeP is a low-toxicity ³¹P NMR probe that can be used to measure intracellular pH values in plant tissues with altered levels of fermentation enzymes. A large number of transformed and mutant lines with altered levels of these enzymes are now available, and it is expected that the MeP pH probe will facilitate the analysis of anaerobic pH regulation in these plants.