Author for correspondence: K. G. Black Tel: +353 1706 2332 Fax: +353 1706 1153 Email: firstname.lastname@example.org
• The transition of Nostoc colonies from free-living to symbiotic conditions, in the Nostoc–Gunnera association, involves increased heterocyst frequency and a reliance on carbon imported from the host for metabolic processes, including N2 fixation.
• Here the uptake of a glucose analogue, 3-[14C]-O-methyl-glucose (14C-OMG), in freshly isolated symbiotic and free-living Nostoc cells was characterized. In situ isotope enrichment coupled with GC–MS was used to elucidate the primary pathway(s) of 1-[13C]-glucose metabolism in the Nostoc–Gunnera symbiosis.
• The characteristics of 14C-OMG uptake by symbiotic clusters suggested a respiratory driven process mediated by a hexose transporter. However, uptake by various Nostoc isolates decreased with increasing heterocyst frequency and was specifically associated with vegetative cells. In isolated and symbiotically intact Nostoc cells, 1-[13C]-glucose was imported and converted to various intermediates of the incomplete citric acid cycle, glycolysis and N2-assimilating pathways. Labelling profiles indicated that C metabolism was altered in infected, but not in uninfected, rhizome tissue.
• Although it has been proposed that cyanobacteria such as Nostoc metabolise glucose using enzymes of the oxidative pentose phosphate cycle, our results suggest that glucose is also metabolised via glycolysis as well as the incomplete citric acid cycle in symbiotic cells.
The Gunnera-Nostoc symbiosis is thought to be a highly evolved association in which the cyanobacterium Nostoc exists intracellularly within the rhizome tissue of the host (Bergman et al., 1992). Following invasion through secretory glands in the plant apex (Silvester & McNamara, 1976), the cyanobiont undergoes a number of modifications prior to the development of a functional symbiosis. These include an increase in heterocyst frequency and N2 fixation (Bergman et al., 1992), a loss of photosynthetic activity (Silvester, 1976; Rai et al., 2000) and the down regulation of glutamine synthetase in heterocysts (Söderbäck, 1992). By contrast to the Anabaena–Azolla symbiosis (Grilli-Caiola et al., 1989), Nostoc is excluded from light in both the cycad and Gunnera symbioses and it has been suggested that in these hosts the cyanobiont must be exclusively dependent on heterotrophic mechanisms for generation of reductant and ATP (Lindblad et al., 1991; Söderbäck & Bergman, 1993; Silvester et al., 1996). However, the characteristics of heterotrophic metabolism by Nostoc, within the host tissue of Gunnera, have not been directly examined. The only evidence for heterotrophy in symbiotic Nostoc comes from studies in which addition of glucose to isolated symbiotic clusters supported N2 fixation (Silvester, 1976; Parsons & Sunley, 2001). It might therefore be expected that uptake of C would be higher in symbiotic, than in photoautotrophic isolates of Nostoc, in order to support the metabolic demands associated with N2 fixation in the dark.
In free-living cyanobacteria, heterocyst differentiation occurs in response to signals associated with N starvation, such as hetR (Adams, 2000) and patS (Yoon & Golden, 1998). In the Gunnera symbiosis, however, heterocyst frequency increases despite the fact that the cyanobiont shows no sign of N starvation (Bergman et al., 1992; Rai et al., 2000). Whilst heterocyst differentiation in symbiotic Nostoc is possibly regulated by the C:N ratio (Rai et al., 2000), little attention has been paid to the regulation of heterocyst frequency by C nutrition.
In addition to ATP, cyanobacteria also have a high NADPH demand for N2 fixation (Wolk et al., 1994). It has been suggested that NADPH is generated by glucose metabolism via the oxidative pentose phosphate cycle (OPPC), which is particularly active in heterocysts of free-living cyanobacteria (Böhme, 1998). Although the tricarboxylic acid cycle (TCA) is incomplete in cyanobacteria, the activity of isocitrate dehydrogenase generates α-ketoglutarate and NADPH, which are required for N assimilation and N2 fixation, respectively (Böhme, 1998). In the Nostoc–Gunnera symbiosis, however, the primary location and source of C taken up, the relative flux of C through the OPPC, TCA and glycolysis, and the role of host metabolism in N assimilation are all uncertain (Rai et al., 2000).
In this paper, we report on the results of an investigation of glucose uptake and metabolism in intact and isolated symbiotic Nostoc. In addition to characterising the kinetics of glucose uptake by free-living and symbiotic cultures, the influence of heterocyst frequency and the age of cyanobiont colonies was also examined. In addition, isotope enrichment and GC-MS were used to elucidate the primary pathway(s) of 1-[13C]-glucose metabolism in the Nostoc–Gunnera symbiosis.
Materials and Methods
Isolation and growth of Nostoc cells
Nostoc puntiforme PC9229 (an axenic, free-living strain originally isolated from Gunnera; B. Bergman, pers. comm.) was obtained from photoautotrophic cultures grown in continuous light in BG medium (Rippka et al., 1979) at an irradiance of 200 µmol m−2 s−1 (λ 400–700 nm), in either the absence (BG11o) or presence of 1.5 g l−1 NaNO3 (BG11). Purity of cultures was checked by streaking on nutrient agar followed by incubation at 25°C.
Symbiotic Nostoc cells were isolated from rhizome tissue of Gunnera tinctoria using homogenation, filtration and centrifugation protocols, as previously used to measure O2-sensitive nitrogenase activity in isolated Nostoc clusters from Gunnera magellanica (Silvester et al., 1996). Isolated cells were then subjected to discontinuous zone density centrifugation to remove fine plant material and phenolics. Resuspended cells (5 cm3) were layered on top of a 10 : 30 : 50% (w/v) sucrose gradient in 50 cm3 centrifuge tubes and centrifuged at 200 g for 15 min The zone, at the 30 : 50% sucrose interface, containing Nostoc cells was removed, suspended in a buffered osmoticum (250 mM mannitol, 1 mM CaCl2 and 25 mM MES, pH 5.7) and washed six times by centrifugation at 5000 g for 5 min.
Nostoc cell suspensions were examined using a compound microscope at × 400 magnification and heterocyst frequency was determined by counting the number of heterocysts per 100 vegetative cells (Kumar et al., 1986).
Measurement of 3-O-methyl-glucose uptake in cell suspensions
Most of the uptake work was based on the use of the glucose analogue, 3-O methyl-D-glucose (OMG), which has been used as a nonmetabolised substrate for examining uptake systems in Nostoc (Beuclerk & Smith, 1978). Free-living and symbiotic isolates were centrifuged at 5000 g for 5 min and cells were resuspended to a final cell density of 1–1.5 mg d. wt cm−3 in a buffered osmoticum of 250 mM mannitol, 1 mM CaCl2 and 25 mM MES (pH 5.7). For all uptake experiments, cell suspensions were incubated in the presence of 3-[14C]-O-methyl-glucose (14C-OMG) under an atmosphere of ambient air in the dark at 25°C. For the time course experiment, cyanobacteria were incubated in 14C-OMG at a final concentration of 5 mM (1.42 GBq mmol−1, Sigma, St Louis, MI, USA), in the presence or absence of unlabelled glucose, fructose or sucrose (10 mM) for 2–45 min. The kinetics and inhibition of OMG uptake by other sugars were determined using a range of 14C-OMG concentrations (0.1–1 mM), with an incubation period of 15 min.
Once the kinetics of 14C-OMG uptake were established, maximal uptake rates by symbiotic and free-living Nostoc isolates, cultured in the presence (BG11) or absence (BG11o) of nitrogen, were determined using a substrate concentration of 5 mM and an incubation period of 15 min. Light- and dark-dependent mechanisms for glucose uptake were determined by adding a specific inhibitor of PSII (3-(3,4 dichlorophenyl)-1,1-dimethylurea, DCMU; 50 µM) and an uncoupler of cytosolic membrane electron transport (carbonyl cyanide p-trifluoromethoxyphenylhydrazone, FCCP; 20 µM) to the reaction mixture. The maximum rate of 14C-OMG uptake was measured under aerobic conditions at 25°C in the dark or light and in the presence or absence of these inhibitors.
The influence of colony age and heterocyst frequency on 14C-OMG uptake was also determined on freshly isolated Nostoc from young (0.5–5 mm diameter), intermediate (5–8 mm diameter) and old (10–20 mm diameter) colonies, which were excised from the rhizome of Gunnera plants that were 18 months old. Isolated cyanobacteria (1–1.5 mg d. wt cm−3) were incubated aerobically in the dark at 25°C for 15 min in the buffered osmoticum containing 5 mM 14C-OMG.
For all uptake experiments, cells were harvested by centrifugation at 10 000 g for 30 s followed by five washes in buffered osmoticum containing 25 mM unlabelled glucose in order to remove bound 14C-OMG from cell walls. Radioactivity in samples was analysed using a scintillation counter (LKB Rackbeta model 1211, Wallac OY, Finland). The absolute amount of 14C (Bq) in each sample was corrected using a 14C quench curve (Black et al., 2000). The total amount of 14C added to each assay and the amount of radiolabel recovered from cells, incubation buffer and cell washings was then used to determine the percentage recovery of 14C. The percentage recovery of 14C from cells, incubation buffer and cell washings after labelling varied from 96 to 99% in the four experiments.
Uptake and distribution of 14C-OMG in intact rhizome tissue discs
14C-Labelling of tissue discs
Infected and uninfected tissue discs were obtained by removing 3 mm diameter cores from rhizome tissue with a cork borer. Transverse sections (1.5 mm) were cut from the cores with a hand microtome and washed in ice cold buffered osmoticum (250 mM mannitol, 1 mM CaCl2 and 25 mM MES, pH 5.7) as previously described (Whittaker & Botha, 1997). Infected and uninfected discs (0.8–1 g f. wt) were then transferred to 40 cm3 crimp vials containing 5 cm3 buffered osmoticum and 14C-OMG at a final concentration of 5 mM. In order to trap 14CO2, a 12% (w/v) KOH solution was added to a 2 cm3 centre well. Vials were then sealed and incubated under normal atmospheric conditions in the dark at 25°C in an orbital incubator at 100 rpm. After an incubation time of 5–250 min, tissue discs were removed and washed three times in a buffered osmoticum containing 25 mM unlabelled glucose, in order to remove bound label. Infected and uninfected tissue was then homogenised in 6 cm3 buffered osmoticum and kept on ice. The amount of 14C-OMG taken up and the quantity of label recovered from CO2 traps was determined using scintillation counting, as described earlier.
Isolation and fractionation of vegetative cells and heterocysts
Cellular extracts from labelled tissue discs, containing Nostoc, were subjected to zone density centrifugation (see above), followed by six washes in 20 mM HEPES (pH 7.5) containing 0.5 mM mannitol, in order to isolate symbiotic Nostoc cells. Heterocysts were isolated using a combined lysozyme and French Press treatment (10 MPa), which results in the selective disruption of vegetative cells (Tel. Or & Stewart, 1977). Intact heterocysts were removed from the extract of vegetative cells by centrifugation at 300 g for 10 min, followed by two to five washes in buffered osmoticum, until heterocyst preparations were about 90% free of vegetative cell material. Isolated heterocysts were then ruptured by passing them twice through a French press at 110 MPa. The amount of 14C-OMG taken up by vegetative cells and heterocysts was determined using scintillation counting as described earlier.
Determination of the cell volume of Nostoc within intact rhizome tissue discs
The amount of label taken up by Nostoc cells within the intact labelled discs from rhizome tissue was calculated using a correction factor based on the concentration of chlorophyll a in infected plant tissue extracts. Chlorophyll a was extracted by adding 1 cm3 of tissue extract to 9 cm3 of 90% (v/v) acetone and boiled for 2 min The absorbance of the solution was measured at 663 nm using a Novaspec spectrophotometer (Model 4409, Pharmica LKB, Biochem Ltd, Cambridge, UK) and chlorophyll a concentrations calculated according to Arnon (1949). The concentration of chlorophyll a in each extract was converted to a mass per unit volume, based on a chlorophyll a standard curve derived from Nostoc suspensions of varying cell density (0.05–0.8 mg d. wt cm−3).
Isotopic enrichment of cells with 1-[13C]-glucose
Isolated symbiotic Nostoc clusters were extracted from infected Gunnera rhizome tissue as described earlier. Cell density was adjusted to a final concentration of 2 mg d. wt cm−3, after which 1-[13C]-glucose was added to 3 cm3 cell suspensions in order to give a final concentration of 25 mM. Cells were then incubated in the dark for 0–60 min at 25°C and harvested every 5 min, followed by centrifugation and four washes in the assay buffer containing 25 mM unlabelled glucose. The final pellet was resuspended in a boiling solution of 70% (v/v) ethanol containing 1% (v/v) HCl before freezing in liquid N.
Infected and uninfected rhizome tissue discs (3 mm) were excised from Gunnera plants (3 month old) and washed in ice-cold buffered osmoticum for 15 min. Discs (0.8–1 g f. wt) were placed in labelling vials (see above) containing 5 cm3 buffered osmoticum and 1-[13C]-glucose at a final concentration of 25 mM, and then incubated at 25°C in the dark. Discs were removed after incubation for 5, 15 and 30 min. After 30 min, 1-[13C]-glucose was replaced with 25 mM unlabelled glucose and incubated for a further 5, 15 and 30 min After each labelling period the discs were removed from the vials, washed in buffer containing 25 mM unlabelled glucose and frozen in liquid N.
Extraction of metabolites and GC-MS
Rhizome tissue and Nostoc cells were extracted in 70% (v/v) ethanol containing 1% (v/v) HCl and freeze-dried overnight. Dried extracts were resuspended in 400 µl of distilled water and a 40 µl subsample was dried using rotary evaporation under vacuum at 60°C. Dried samples were then resuspended in 20 µl pyridine and derivatised using either 20 µl of N-(tert-butyldimethlysilyl)-N-methyltrifloroacetamide (MTBSTFA) or trimethychlorolsilane (TMS). A 1-µl sample solution was then subjected to GC-MS (VG Organic MD800) using single ion monitoring (SIM). Parent peaks M, M + 1 and M + 2 for glucose, fructose, sucrose, fructose-6-phosphate, glucose-6-phosphate, pyruvate, citrate, α-ketoglutarate, oxaloacetate, malate, aspartate, glutamate, asparagine and glutamine were automatically computed and integrated and the isotopic ratios calculated based on their enriched ratio of standards (Silvester et al., 1996).
Kinetics of 14C-OMG uptake
Symbiotic Nostoc cells displayed a time- and substrate-dependent uptake of 14C-OMG, although, in all cases, maximum uptake was achieved within 20–30 min (Fig. 1a). Whilst fructose and glucose markedly inhibited the uptake of 14C-OMG by isolated symbiotic Nostoc cells under aerobic conditions in the dark, sucrose did not inhibit uptake of the analogue (Fig. 1a). The concentration-dependent uptake of 14C-OMG conformed to the classical pattern of saturation kinetics when analysed by the Lineweaver-Burk double-reciprocal method (Fig. 1b). Whilst the maximum velocity of 14C-OMG uptake (Vmax) was not affected by the presence of other sugars, the substrate concentration required to achieve half the maximum velocity of uptake (Km) increased from 0.7 mM to 1.9 mM in the presence of glucose or fructose (Fig. 1b). Once the kinetics of substrate uptake was established, the maximum rates of 14C-OMG uptake were determined using a substrate concentration of 5 mM.
The higher rate of uptake (c. 80%) of 14C-OMG by free living Nostoc cells grown in the presence of N (BG11), compared with that in both symbiotic and photoautotrophic cells grown without N (BG11o, Table 1), was predominantly associated with a decrease in heterocyst frequency. The heterocyst frequency of free-living cells, grown in the presence of N, was 6%, compared with 19 and 22% for symbiotic and free living (–N) Nostoc isolates incubated in BG11o.
Table 1. Effect of different incubation conditions and inhibitors on uptake of 3-[14C] O-methyl-glucose
14C-OMG uptake (nmol mg−1 min−1 d. wt)
Suspensions of photoautrophic and freshly isolated Nostoc cells were incubated aerobically in buffered osmoticum, containing 3-[14C]-O-methyl-glucose at a final concentration of 5 mM (1.42 GBq mmol−1), for 15 min at 25°C in the dark or light and in the presence or absence of inhibitors (DCMU and FCCP) as indicated. Axenic cultures were grown photoautotrophically in BG liquid media with 1.5 g l−1 NaNO3 (BG11) or without N (BG11o). All values represent means (± SD, n= 4).
FCCP (20 µM)
DCMU (50 µM)
FCCP + DCMU
When Nostoc isolates were incubated in the light in the absence of inhibitors, the uptake of 14C-OMG increased by c. 34%, when compared with that for cyanobacteria incubated in the dark (Table 1). In the light, 50 µM DCMU inhibited the rate of uptake by 20–40%, and these rates were comparable with those achieved by cyanobacteria incubated in the dark (Table 1). Since the PSII-driven uptake of substrate was similar for all isolates (based on the effects of treatment with DCMU) differences in uptake between isolates were not, apparently, associated with PSII activity. In addition, the PSII-related uptake of 14C-OMG represented only a small component of total uptake. In contrast, addition of 20 µM FCCP to cell suspensions inhibited the uptake of 14C-OMG by 80–90%, when cyanobacteria were incubated in either the dark or the light (Table 1).
The influence of heterocyst frequency on the rate of 14C-OMG uptake was also examined using symbiotic cells isolated from young, intermediate and old rhizome colonies, using size and location as an indicator of age (Fig. 2). Young colonies (0.5–2 mm diameter) were obtained from host tissue near the meristematic region of the rhizome, while older colonies (10–20 mm diameter) were excised from tissue near the base of the rhizome. The uptake of 14C-OMG decreased by 60% as heterocyst frequency increased from 18% in younger to 52% in older colonies (Fig. 2).
Uptake and distribution of labelled substrate in intact tissue discs
The initial and maximum uptake rates of 14C-OMG were significantly higher (23–30%) in infected, compared with uninfected, rhizome tissue discs (Fig. 3). The increased uptake by infected tissue was also broadly consistent with the calculated and measured uptake of OMG by symbiotically intact Nostoc cells.
Whilst it was apparent that a small amount of 14C-OMG was converted to CO2 in tissue discs, after an incubation period of 250 min, the percentage of substrate metabolised was similar for both infected and uninfected tissue (Table 2). Nearly all of the radioactivity taken up (99%) by uninfected tissue discs was present in the fraction soluble in ethanol-water. In contrast, only 80% of the radioactivity taken up by infected tissue was detected in the host fraction. Based on the d. wt percentage of Nostoc cells in infected tissue (30% of infected tissue) when chlorophyll concentration was used as a biomass marker, it was estimated that 19% of the radioactivity in infected tissue was taken up by Nostoc. Most of this radioactivity (98%) was recovered from the vegetative cell fraction (Table 2).
Table 2. Partitioning of 3-[14C]-O-methyl-glucose into different fractions isolated from infected and uninfected rhizome tissue discs
% Distribution (kBq)
Tissue discs were incubated aerobically in a buffered osmoticum, containing 3-[14C]-O-methyl-glucose at a final concentration of 5 mM (1.42 GBq mmol−1), for 240 min at 25°C in the dark. All values represent means (± SD, n= 4).
Nostoc vegetative cells
The uptake and fate of 1-[13C]-glucose by symbiotic Nostoc cells
The 13C isotope of glucose was taken up and metabolised by isolated Nostoc clusters and infected tissue discs (Figs 4 and 5). The maximum amount of 1-[13C]-glucose taken up after 10 min was 20% higher in infected tissue discs than in uninfected tissue discs (Fig. 5). These findings were consistent with the studies using 14C-OMG, where uptake increased by 23–30% in infected tissue (Fig. 3).
Metabolism of glucose
When derivatised samples were analysed using GC and multiple ion monitoring (m/z 0–800), all target metabolites were detected, except for fructose-6-phosphate and glucose-6-phosphate (data not shown). Subsequent single ion monitoring of parent peaks revealed the presence of 13C in fructose, malate, aspartate and glutamate in labelled symbiotic cells (Fig. 4). No isotope enrichment was detected in the parent peaks of citrate, oxaloacetate, α-ketoglutarate, pyruvate, glutamine or asparagine in isolated symbiotic cells (Fig. 4). However, problems associated with peak shifts and the narrow retention time windows used to detect isotope enrichment in parent peaks of oxaloacetate, α-ketoglutarate and asparagine, prevented us from analysing 13C enrichment in these compounds.
Whilst glucose was converted to fructose in both infected and uninfected tissue, the isotopic enrichment of fructose was significantly higher in uninfected rhizome discs (Fig. 5). However, the fructose component may also represent a combined dephosphorlylated pool of fructose-6-phosphate and fructose-1,6-bisphosphate, due to the mild acidic ethanol extraction protocol. The 13C-enrichment of pyruvate and malate showed similar trends for both infected and uninfected tissue discs over the labelling period, but 13C-enrichment into these organic acids was about 50% greater in infected discs, particularly 5–30 min into the chase period with unlabelled glucose (Fig. 5b). The isotopic enrichment of M/M + 1 parent peaks for aspartate and glutamate was two–fivefold higher in infected, compared with uninfected discs (Fig. 5c). Whilst the isotopic enrichment of these two precursors of N assimilation was followed by an increase in the labelling of glutamine in infected discs, this was not observed for asparagine (Fig. 5c). However, comparisons of chromatograms obtained from infected and uninfected tissue discs show that asparagine is present at higher concentrations in infected tissue (data not shown). No significant isotope enrichment of glutamine was detected in labelled uninfected discs, at a detection limit of 0.5% atom excess (Fig. 5c).
In this study, the uptake of glucose by isolated symbiotic Nostoc cell suspensions was characterised using 14C-OMG, a nonmetabolisable analogue of glucose that is taken up by the same transport system (Beuclerk & Smith, 1978). The kinetic properties and Michaelis constant (Km = 0.7 mM) for 14C-OMG uptake by symbiotic Nostoc were similar to those reported for free-living Nostoc by Beuclerk & Smith (1978), where the Km for 14C-OMG uptake was 0.4 mM.
It has been proposed that C could be taken up by both vegetative cells and heterocysts in plant-cyanobacterial symbioses (Rai et al., 2000). Our results, however, show that glucose is almost exclusively (98%) taken up by vegetative cells (Table 2). These findings suggest that the primary function of vegetative cells, in the symbiosis, may be C uptake to provide the energy and reductant for N2-fixation in heterocysts. This is consistent with the inverse relationships between heterocyst frequency and OMG uptake (Table 2, Fig. 2). In addition, the previously reported increase in heterocyst number, coupled with a decrease in N2-fixation in Gunnera magellanica tissue (Bergman et al., 1986, 1992) and coralloid roots of cycads (Lindblad et al., 1991), could be due to C limitation (Bergman et al., 1986).
This study provided no evidence that symbiotic cells have an increased capacity to take up glucose, despite the expected greater demand on carbon uptake to support N2-fixation. Symbiotic and free-living cells with the same heterocyst frequency had comparable rates of OMG uptake. This may be due to numerous factors, including a slower growth rate of symbiotic (heterotrophic) cells compared with free-living photoautotrophic cultures (Rai et al., 2000), a shorter filament length, which may restrict the movement of C between cells in the symbiosis (Rai et al., 2000), filament sheering during the isolation of symbiotic cells or a high proportion of degenerate heterocysts, particularly in older symbiotic colonies (Bruce, 1997), resulting in a lower demand for inorganic C. Alternatively, the uptake process may not be the limiting step in the supply of C for metabolic processes in symbiotic Nostoc. In particular, a lack of direct contact between many vegetative cells and heterocysts or an increase in multiple heterocysts in symbiotic tissue may significantly limit the supply of C for N2 fixation (Bergman et al., 1986; Bergman et al., 1992).
The increase in Km with no change in Vmax for OMG uptake in the presence of glucose and fructose suggest competitive inhibition for a binding site and that the transporter involved in OMG uptake may be a hexose transporter. Although these results show that both glucose and fructose are taken up by symbiotic Nostoc, these may not be the only sources of C supplied by the host. Previous studies have shown that nitrogenase activity of symbiotic cells, in the dark, is supported by an exogenous supply of various compounds, including sucrose, glucose and fructose (Silvester et al., 1996; Parsons & Sunley, 2001).
It is evident from the sensitivity of 14C-OMG uptake to FCCP that most of the energy required for active transport is provided by respiratory electron flow (Raboy & Padan, 1978). Although the uptake of substrate was also inhibited by DCMU in the light, this inhibition was small and the influence of PSII electron transport on glucose uptake may be significant only during the early colonisation events, where Nostoc cells are exposed to light.
Although freshly isolated symbiotic clusters were isolated under strict anaerobic conditions during the extraction procedure, the presence of O2 during the glucose uptake assays may have influenced nitrogenase activity in vitro (Lindblad et al., 1991; Silvester et al., 1996). It is possible therefore that the rate of glucose uptake in vitro may have been underestimated due to a lower C demand for N2 fixation (Silvester et al., 1996; Rai et al., 2000). However, glucose uptake studies based on the uptake of the nonmetabolisable anolouge 14C-OMG are likely to be governed only by the uptake mechanism per se (Raboy & Padan, 1978) and not by C-dependent N2 fixation and assimilation pathways. When 1-[13C]-glucose was used as a substrate for uptake by intact tissue discs, it was unlikely that the uptake and metabolism of glucose was affected by O2 since nitrogenase activity in tissue discs has been shown to be insensitive to O2 for up to 3 h after excision (Silvester et al., 1996). In addition, evidence for N2 fixation and assimilation in tissue discs was manifested in this study by 13C enrichment of end products of N assimilation (Fig. 5).
The higher initial and maximal uptake of OMG in infected tissue, compared with uninfected tissue, confirmed that 14C-OMG was taken up by Nostoc in situ. In addition, the somewhat slower uptake of OMG by Nostoc clusters in situ, compared with that in vitro (Fig. 3), suggests that the transport of assimilates to Nostoc via the host symplast is regulated by the host. This would be consistent with the hypothesis that regulation of the symbiosis may involve control by the plant over the availability or uptake of C by Nostoc and that this may be altered to match the N requirement of the plant (Parsons & Sunley, 2001).
In this study, isotope enrichment and metabolic profiling using GC-MS confirmed that glucose is taken up and rapidly metabolised in both isolated and intact symbiotic clusters. Based on the differential labelling patterns associated with infected and uninfected tissue discs, it is evident that C metabolism is altered in the presence of the cyanobiont. Whilst glucose is rapidly converted to fructose in both infected and uninfected tissue, isotope enrichment of fructose was significantly higher in uninfected rhizome discs. This may result from alterations in a number of metabolic steps. For example, there could be an increased flux into fructose-6-phosphate, via hexokinase and hexose-phosphate isomerase, which is then converted to fructose. Alteratively, a continuous cycle of sucrose synthesis and degradation could lead to randomisation of label into fructose via hexose isomerase in host sink tissue (Giegenberger & Stitt, 1991), which may be higher in uninfected discs. A third possibility is the presence of dephosphorlylated fructose-6-phosphate and fructose 1,6-phosphate in the fructose pool, caused by the acid ethanolic extraction. This could suggest a higher glycolytic flux in uninfected tissue, which would also be consistent with the absence of any hexose phosphate peaks when samples were analysed by GC-MS. Finally, a higher flux of 1-[13C]-glucose into the oxidative pentose phosphate cycle (OPPC) could result in the loss of 13CO2 from the C-1 position of 6-phosphogluconate during the synthesis of ribulose-5-phosphate in infected tissue. In three of these four scenarios, C metabolism of the host tissue may be altered via the regulation of hexokinase. Similarly, in sink tissue, it has been shown that the control of C metabolism can be mediated by sugar sensing mechanisms, based on signal transduction and gene regulation of hexokinase (Kock, 1996; Yu, 1999). Alternatively, the apparent higher flux into the OPPC (scenario 3, above) would be consistent with glucose-stimulated expression in Nostoc of the gene encoding glucose-6-phosphate dehydrogenase (Sundaram et al., 1998), the first enzymatic step in the OPPC.
Whilst it has been proposed that the OPPC is the primary pathway for glucose metabolism in Nostoc (Böhme, 1998), the results from these experiments suggest that intermediates of the incomplete citric acid cycle, glycolytic and N assimilation pathways were not synthesised exclusively via the OPPC since 13C, from 1-[13C]-glucose, was detected in these metabolite pools (Figs 4 and 5). The only possible pathway for 13C incorporation into intermediates of the citric acid cycle exclusively via the OPPC would be through re-incorporation of 13CO2, released by the decarboxylation of 6-phosphogluconate, into oxaloacetate through the activity of phosphenolpyruvate carboxylase (PEPC). Previous studies have shown that 14CO2 can be fixed by symbiotically intact Nostoc in the dark, suggesting that PEPC is active in the Nostoc-Gunnera symbiosis (Silvester, 1976). Although these results provide no evidence that glucose is not metabolised via the OPPC, it is evident that glucose is also metabolised via glycolysis. Reductant for N2 fixation via feredoxin: NADP reductase could also be provided by NADP-isocitrate dehydrogenase, which is present at high activity in heterocysts (Martin-Figueroa et al., 2000). Therefore, the results suggest that nitrogenase activity may not be solely dependent on the supply of NADPH from metabolism of glucose via the OPPC in heterocysts of symbiotic cells. However, the isotopic enrichment of pyruvate, glutamine and asparatate in infected tissue and the absence of labelled pyruvate in freshly isolated Nostoc suggests that most of the C skeletons required for N assimilation are synthesised in host tissue (Silvester et al., 1996; Rai et al., 2000).
Whilst significant isotopic enrichment of amino acids occurs in situ, the absence of 13C enrichment of glutamine in freshly isolated symbiotic cells could suggest that N2 may not be fixed or assimilated to the same extent in isolated symbiotic clusters. This may be due to the sensitivity of nitrogenase activity to inactivation by O2 in Nostoc isolated from cyanobacterial symbioses (Lindblad et al., 1991; Silvester et al., 1996). Since most free-living Nostoc isolates and cyanobionts continue to fix N2 under aerobic conditions, it is uncertain what changes occur in heterocysts, isolated from Gunnera, to make them O2 sensitive (Silvester et al., 1996). It is possible there could have been some damage to heterocysts during the isolation procedure or that the heterocyst cell wall is modified in the symbiotic state to allow the diffusion of NH3 into the host cell (Silvester et al., 1996; Rai et al., 2000). It has been proposed that NH3 is released from heterocysts and that its assimilation takes place entirely within the host cell via GS-GOGAT, with N being subsequently transported to uninfected tissue as asparagine (Silvester et al., 1996; Rai et al., 2000).
We would like to thank Bergitta Bergman for providing the isolated Nostoc strain PC9229 for these experiments and the European Science Foundation ‘CYANOFIX program’ for the travel grant for the GC-MS analyses at the University of Dundee. We are also grateful to Kevin O’Connor, Department of Industrial Microbiology, UCD, for the use of the French Press.