SUGAR-DEPENDENT6 Encodes a Mitochondrial Flavin Adenine Dinucleotide-Dependent Glycerol-3-P Dehydrogenase, Which Is Required for Glycerol Catabolism and Postgerminative Seedling Growth in Arabidopsis

sugar-dependent6 Is Impaired in Gluconeogenesis during Postgerminative Growth

Immediately following germination, Arabidopsis seedlings rely on stored TAG to fuel growth until they can achieve photosynthetic competence. The recessive sugar-dependent6 (sdp6) mutant was selected in a genetic screen designed to identify ethyl methanesulfonate (EMS)-mutagenized lines that are unable to utilize TAG and therefore cannot make this transition without the provision of exogenous Suc (Eastmond, 2006). The sdp6 mutant alleles that were isolated from this screen are able to germinate. However, 100% of the seedlings exhibit growth arrest shortly after germination, even in the presence of normal levels of photosynthetically active radiation (photosynthetic photon flux density [PPFD] = 150 μmol m⁻² s⁻¹). The mutant alleles are therefore conditionally seedling lethal (Fig. 1A ). The growth of sdp6 seedlings in the dark is also severely inhibited (Fig. 1B), suggesting that growth arrest cannot be attributed to a defect in photosynthesis. Seedlings that are rescued with Suc are able to grow and once they develop photosynthetic capacity in the light they can complete their life cycle without exogenous Suc.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Postgerminative growth, TAG breakdown, and sugar content in sdp6 seedlings. A and B, Five-d-old sdp6 and wild-type seedlings grown in the light (A) or in the dark (B) on agar plates containing half-strength MS salts plus or minus 1% (w/v) Suc. Bar = 0.5 cm. C, TAG content of imbibed seeds and 5-d-old sdp6 and wild-type seedlings grown in the light on agar plates containing half-strength MS salts plus 1% (w/v) Suc. Values are the mean ± se of measurements on five batches of 20 seeds or seedlings. D, Sugar content of 3-d-old sdp6 and wild-type seedlings grown in the light on agar plates containing half-strength MS salts. Values are the mean ± se of measurements on four batches of 500 seedlings. [See online article for color version of this figure.]

TAG measurements performed on sdp6 seeds and seedlings grown in the presence of Suc showed that the mutant is capable of storage oil breakdown, despite the strong sugar-dependent phenotype (Fig. 1C). This contrasts with mutants that are impaired in TAG hydrolysis or fatty acid β-oxidation (Hayashi et al., 1998; Eastmond, 2006). In wild-type Arabidopsis seedlings the decline in TAG levels that occurs during postgerminative growth is mirrored by a corresponding increase in sugars (predominantly Glc), which are derived from the breakdown products of TAG via gluconeogenesis (Footitt et al., 2002; Pritchard et al., 2002). In 3-d-old sdp6 seedlings the levels of Suc, Glc, and Fru are approximately 2-, 4- and 3-fold less than the wild type (Fig. 1D). These data suggest that sdp6 is impaired in gluconeogenesis, rather than TAG or fatty acid catabolism (Eastmond et al., 2000; Pritchard et al., 2002).

SDP6 Encodes the Mitochondrial FAD-G3P Dehydrogenase

High-resolution mapping was used to locate the SDP6 locus. The sdp6-1 mutant was out-crossed to wild-type Landsberg erecta, and 768 sdp6-1 lines were selected from the F₂ progeny. PCR-based polymorphic markers were used to localize SDP6 to a 35-kb region on chromosome 3, situated near marker nga126 (Fig. 2A ). This region contains nine genes. Sequencing candidate genes in two independent sdp6 alleles revealed that there were point mutations in locus At3g10370 (Fig. 2B). This gene encodes a FAD-dependent G3P dehydrogenase (FAD-GPDH) that is situated on the mitochondrial inner membrane (Shen et al., 2003). In sdp6-1, a G-to-A mutation at positions +2,272 bp after the ATG causes Glu-546 to become Gly, and in sdp6-2, a C-to-T mutation at +257 bp causes Ser-86 to become Phe. Both amino acid substitutions occur in regions of the polypeptide that are highly conserved, and in sdp6-2 it falls within the putative FAD-binding domain (Guerra et al., 2006).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Positional cloning of SDP6 and characterization of a T-DNA allele. A, A diagram to show how PCR-based SSLP and CAPS markers were used to map SDP6 to a 35-kb region on chromosome 3 that contains nine genes. The positions of markers a to d are marked by bars, and the number of recombination events/total number of chromosomes (1,536) is listed below each. B, A diagram of the SDP6 locus (At3g10370) showing the positions of mutations that were identified in three independent sdp6 alleles. Insertion/substitution positions are numbered relative to the initiation codon. White bars are exons, and the black bars indicate the 5′ and 3′ untranslated regions. C, Phenotype of sdp6-3 seedlings. Five-d-old sdp6-3 and wild-type seedlings grown in the light on agar plates containing half-strength MS salts plus or minus 1% (w/v) Suc. Bar = 0.1 cm. D, Detection of SDP6 transcript in seedlings from the wild type and sdp6 alleles. PCR was performed on cDNA made from total RNA from 2-d-old seedlings. The actin gene ACT2 is shown as a control. [See online article for color version of this figure.]

A T-DNA insertion allele of sdp6 was also obtained from the Salk Institute Genomic Analysis Laboratory's T-DNA Express Arabidopsis gene mapping tool (Alonso et al., 2003). The sdp6-3 allele (SALK_080169) contains a T-DNA inserted in the fourth intron of At3g10370 (+1,856 bp). The sdp6-3 mutant exhibits a recessive seedling growth arrest phenotype that cosegregates with the T-DNA insertion in At3g10370. However, the phenotype is more severe than the two EMS alleles. Seed germination in sdp6-3 is delayed and seedlings require several weeks to establish photosynthetic competence, even when provided with exogenous Suc (Fig. 2C). PCR performed on cDNA from homozygous sdp6-3 seedlings, using primers that flank the T-DNA insertion site, showed that SDP6 mRNA is either lacking, or that the presence of the T-DNA in the fourth intron prevents it from being spliced out (Fig. 2D).

FAD-GPDH appears to be a single copy gene in Arabidopsis (Shen et al., 2003). To confirm that sdp6 seedlings are defective in FAD-GPDH activity enzyme assays were performed on whole extracts from 2-d-old seedlings (Eastmond, 2004). No activity was detectable in the sdp6-3 allele, while in sdp6-1 and sdp6-2 activity was reduced by more than 90% (Table I ). Because the germination and growth of sdp6-1 and sdp6-2 seedlings were comparable to the wild type when Suc was provided, the majority of our biochemical analysis was performed on these lines rather than on sdp6-3.

FAD-GPDH Is Required for Glycerol Catabolism

FAD-GPDH has been proposed to play a key role in the pathway of glycerol breakdown in plants (Huang, 1975). This pathway is highly active in oilseeds following germination because the hydrolysis of TAG releases free fatty acids and glycerol, which are both converted to sugars to support seedling growth. The postgerminative growth of many Arabidopsis sugar-dependent mutants, such as the glyoxylate cycle mutant isocitrate lyase-2 (icl-2), can also be rescued using glycerol as a carbon source (Eastmond, 2004). However, glycerol cannot rescue sdp6 alleles (Fig. 3A ). Furthermore, radiolabel feeding experiments show that sdp6 seedlings metabolize at least 10-fold less [U-¹⁴C]glycerol to CO₂ than the wild type (Fig. 3C). These data strongly suggest that FAD-GPDH/SDP6 is required for glycerol catabolism in Arabidopsis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Utilization of glycerol by sdp6 seedlings. A, Hypocotyl length of 5-d-old etiolated wild type, sdp6-1, sdp6-1 gli1-3, gli1-3, and icl-2 seedlings grown in the dark on agar plates containing half-strength MS salts alone or plus 1% (w/v) Suc or 0.2% (w/v) glycerol. Values are the mean ± se of measurements on three batches of 20 seedlings. B, Germination frequency of wild-type, sdp6-1, sdp6-1 gli1-3, gli1-3, and icl-2 seeds grown on agar plates containing half-strength MS salts alone or plus 1% (w/v) Suc or 0.2% (w/v) glycerol. Percent of germination was scored after 3 d using batches of 200 seeds. C, Release of ¹⁴CO₂ by 2-d-old wild-type, gli1-3, sdp6-1, sdp6-2, and sdp6-3 seedlings supplied with [U-¹⁴C]glycerol. Values are the mean ± se of measurements on four batches of 50 seedlings. ND, Not detected.

sdp6 Seedlings Accumulate G3P

Glycerol only represents a small proportion (approximately 5%) of the carbon present in TAG, and failure to degrade glycerol alone cannot account for the sugar-dependent phenotype of sdp6 because the glycerol kinase mutant gli1 is not sugar dependent (Eastmond, 2004). Measurements of glycerol and G3P levels in gli1-3 and sdp6 mutant seedlings grown in the presence of Suc showed that while gli1 dramatically overaccumulates glycerol (Eastmond, 2004), sdp6 alleles overaccumulate G3P (Table II ). Studies have previously shown that high concentrations of glycerol are toxic to plants (with the apparent exception of germinated oilseeds) and that this toxicity is likely to be caused by glycerol metabolism rather than glycerol itself (Leegood et al., 1988; Aubert et al., 1994; Eastmond, 2004). The gli1 mutant was originally selected based on its insensitivity to the inhibitory effect of glycerol on seedling root growth (Eastmond, 2004). Growing sdp6 seedlings on medium containing glycerol revealed that germination is strongly inhibited by glycerol (Fig. 3B). Analysis of a sdp6-1 gli1-3 double mutant also showed that gli1 is a suppressor of the sdp6 growth arrest phenotype (Fig. 3A). The hypersensitivity of sdp6 to glycerol, and suppression by gli1, provide genetic evidence that seedling growth arrest is linked to G3P overaccumulation (and/or ATP consumption/phosphate sequestration by GLI1) and not to further downstream metabolism of G3P.

G3P Inhibits PGI Activity

Aubert et al. (1994) have previously used ³¹P NMR to show that when heterotrophic plant cell cultures are fed glycerol, they accumulate high levels of G3P and are unable to convert it to sugars, although they can use it as a substrate for respiration. Aubert et al. (1994) also discovered that G3P is a competitive inhibitor of the glycolytic enzyme PGI (EC 5.3.1.9) in vitro and proposed that glycerol feeding is likely to lead to PGI inhibition, thereby preventing gluconeogenic flux to sugars by blocking Fru-6-P (F6P) conversion to Glc-6-P (G6P). This model could potentially explain the glycerol catabolism-dependent gluconeogenic defect observed in sdp6 seedlings.

Arabidopsis contains a cytosolic and a plastidial isozyme of PGI, whose activities can be discriminated based on their different temperature stability (Yu et al., 2000). Assays performed on extracts from seedlings show that approximately 60% of the activity is present in the cytosol and that there is no significant difference in total activity between the wild type and sdp6-1 (Table III ). To determine whether G3P is an inhibitor of Arabidopsis, PGI activity enzyme assays were performed on extracts from wild-type seedlings, using F6P as a substrate (Aubert et al., 1994). Total PGI activity was competitively inhibited by G3P with a K_i value of 4.7 mm (Table III). Because a substantial proportion of PGI activity is present in both the cytosol and plastid, G3P must be capable of inhibiting both isozymes. The estimated concentration of G3P in 2-d-old sdp6 seedlings is approximately 60 mm (Table II), assuming that G3P can access about 20% of the tissue volume. This concentration of G3P would be sufficient to inhibit almost completely PGI activity in sdp6 seedlings, if both isozymes were exposed to it. Because G3P can be transported across the chloroplast inner membrane (Fliege et al., 1978), inhibition of both isozymes is probable.

Evidence That PGI Is Inactive in sdp6 Seedlings

To investigate whether the interconversion of G6P and F6P, catalyzed by PGI, is blocked in sdp6 seedlings, feeding experiments were performed with [U-¹⁴C]Glc or [U-¹⁴C]Fru (Fig. 4 ), and the proportion of radiolabel incorporated into the two hexosyl moieties of Suc was determined (Geigenberger and Stitt, 1993; Fernie et al., 2001). In 2-d-old wild-type seedlings about 30% of the radiolabel was redistributed from Glc to the fructosyl moiety of Suc and, in reciprocal experiments, from Fru to the glucosyl moiety of Suc (Fig. 4B). Preferential labeling of the fructosyl moiety of Suc by Fru has been observed in many plant tissues and is largely attributed to some direct incorporation via Suc synthase, which bypasses hexose phosphate isomerization (Geigenberger and Stitt, 1993). In contrast to the wild type, less than 5% of the radiolabel was redistributed to the opposing hexosyl moiety when Glc or Fru was supplied to sdp6-1 seedlings (Fig. 4B). These data suggest that sdp6-1 seedlings exhibit a 6-fold reduction in their capacity to interconvert G6P and F6P and therefore PGI is very likely to be inhibited in vivo.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Incorporation of ¹⁴C-label from Glc and Fru into the hexosyl moieties of Suc in sdp6 seedlings. A and B, Two-day-old wild-type and sdp6-1 seedlings were incubated with [U-¹⁴C]-labeled Glc or Fru, either alone or with unlabeled Glc or Fru and the amount of total label in Suc (A), and percentage redistribution to the opposite hexosyl moiety of Suc (B) were determined. Values are the mean ± se of measurements on three separate incubations.

The same experiment also showed that sdp6-1 seedlings are able to incorporate radiolabel from Glc into Suc without the addition of unlabeled Fru, whereas the incorporation of radiolabel from Fru into Suc requires the supply of unlabeled Glc (Fig. 4A). One interpretation of these data is that endogenous metabolism in germinated sdp6-1 seedlings is able to provide F6P (and/or Fru) for Suc synthesis, but not G6P. If this is so, the reaction catalyzed by PGI is the only step in gluconeogenesis that is severely compromised in sdp6-1 seedlings. G6P can also be converted to F6P via the oxidative pentose phosphate pathway (Hatzfeld and Stitt, 1990). However, the 6-fold reduction in the transfer of radiolabel from Glc into the fructosyl moiety of Suc in sdp6-1 seedlings suggests that the oxidative pentose phosphate pathway is unlikely to supply sufficient F6P to account for the level of incorporation of radiolabel into Suc when the mutant is supplied with Glc alone (Fig. 4).

The different abilities of various exogenous carbon sources to rescue postgerminative growth of sdp6-1 (measured as hypocotyl elongation in the dark) also provide corroborating evidence to support the hypothesis that gluconeogenesis is blocked at the level of PGI in this mutant. Suc and Glc can rescue sdp6-1 growth, but Fru cannot (Fig. 5 ). In contrast, all three sugars can rescue postgerminative growth of the glyoxylate cycle mutant icl-2 (Fig. 5). Although Fru can be converted to Suc by plants (Fig. 4; Geigenberger and Stitt, 1993; Fernie et al., 2001), our data suggest that in sdp6 seedlings the accumulation of G3P leads to PGI inhibition and that this impairs the synthesis of G6P, which is also required for Suc synthesis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Rescue of sdp6 seedling growth by various exogenous carbon sources. Hypocotyl length of 5-d-old etiolated wild-type, sdp6-1, sdp6-1 gli1-3, gli1-3, and icl-2 seedlings grown in the dark on agar plates containing half-strength MS salts alone or plus 1% (w/v) Suc, 1% (w/v) Glc, or 1% (w/v) Fru. Values are the mean ± se of measurements on three batches of 20 seedlings.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

The effect of ABA and salt stress on sdp6-3 (white circles) and wild-type (black circles) seed germination. Seeds were germinated on agar plates containing half-strength MS salts and various concentrations of ABA or NaCl. The percent of germination frequency was recorded over the course of 10 d.

sdp6 Seedlings Have an Altered NADH/NAD⁺ Ratio and Seed Germination Is Hypersensitive to Abscisic Acid and Salt Stress

In addition to a role in glycerol catabolism, it has been proposed that GPDH also participates in a G3P shuttle that channels reducing equivalents from the cytosol to the mitochondrion (Shen et al., 2003, 2006). Arabidopsis gpdhc1 mutants, which are deficient in the other proposed component of this shuttle (a cytosolic G3P dehydrogenase), exhibit an altered redox state of their NAD(H) pool (Shen et al., 2006). Measurement of total NADH and NAD⁺ levels in 5-d-old sdp6-1 and sdp6-2 seedlings grown in the presence of Suc show that NADH content is doubled, while NAD⁺ content is decreased slightly (Table II). These data suggest that the cytosolic NADH/NAD⁺ ratio is likely to be higher in sdp6 seedlings than in the wild type. Shen et al. (2006) reported a similar shift in the NADH/NAD⁺ ratio in leaves from gpdhc1, which they further corroborated by estimating the ratio of free NADH/NAD⁺ in the cytosol according to the concentrations of metabolites considered to be in near equilibrium with the NADH/NAD⁺ couple in the cytosolic compartment. The total NADH/NAD⁺ ratio in 5-d-old wild-type Arabidopsis seedlings was found to be approximately 0.09 (Table II), which is lower than has been reported for photosynthetically active leaf tissue in the light (Queval and Noctor, 2007; Wang and Pichersky, 2007). However, Wang and Pichersky (2007) have previously shown that NADH/NAD⁺ ratios are lower in Arabidopsis seedlings (approximately 0.09) and roots (approximately 0.03) than in leaves (approximately 0.4).

Shen et al. (2006) have shown that gpdhc1 mutant seedlings can grow normally under optimal growth conditions but exhibit hypersensitivity to abscisic acid (ABA) and salt stress. The transcript abundance of both GPDHC1 and FAD-GPDH are increased by certain stress treatments (Shen et al., 2003, 2006). To investigate whether sdp6 is also stress intolerant, seeds of the wild type and the null sdp6-3 allele were germinated on medium containing a range of concentrations of ABA or NaCl. In the absence of exogenous ABA or NaCl, the germination of sdp6-3 already shows a delay relative to the wild type, but after 7 d the majority of the seeds have germinated (Fig. 6). The inclusion of ABA or NaCl in the growth medium was found to have a more pronounced inhibitory effect on sdp6-3 germination than on the wild type (Fig. 6).

sdp6 Seeds Have Reduced Weight and Fatty Acid Content

Increasing G3P levels by glycerol feeding or overexpression of NAD-GPDH has recently been shown to promote TAG synthesis in developing embryos of oilseed rape (Brassica napus) by driving G3P acyltransferase activity (Vigeolas and Geigenberger, 2004; Vigeolas et al., 2007). To determine whether disrupting FAD-GPDH activity significantly influences seed oil content in Arabidopsis, between eight and 10 wild-type, sdp6-1, sdp6-2, and sdp6-3 plants were grown in the glasshouse under the same growth conditions (Li et al., 2006). Seed fresh weight and fatty acid measurements performed on separate batches of seeds from each plant suggest that sdp6 mutants produce significantly smaller seeds with less total fatty acids per seed than the wild type (Table IV ).

Plant Material and Growth Conditions

Wild-type Arabidopsis (Arabidopsis thaliana; ecotype Columbia and Landsberg erecta) were obtained from the European Arabidopsis Stock Centre (University of Nottingham, UK). The isolation of sdp6-1 and sdp6-2 is described in Eastmond (2006), and the isolation of gli1-3 in Eastmond (2004). The sdp6-3 mutant (SALK_080169) was obtained from T-DNA express (Alonso et al., 2003). Seeds were surface sterilized, applied to agar plates containing half-strength Murashige and Skoog (MS) salts (Sigma-Aldrich), and imbibed in the dark for 4 d at 4°C. The plates were then transferred to a growth chamber set to 21°C (16 h light/8 h dark; PPFD = 150 μmol m⁻² s⁻¹). For experiments on etiolated seedlings, the plates were kept in darkness in the growth chamber after 1 h of light. In some experiments Suc, Glc, Fru, glycerol, ABA, or NaCl were added to the agar medium. For measurements of seed weight and total fatty acid content, plants were grown on soil in the glasshouse with supplemental lighting (16 h light/8 h dark; PPFD = 150 μmol m⁻² s⁻¹).

Metabolite Analysis and Radiolabel Feeding

The fatty acid and TAG content of seeds and seedlings were measured using the methods described in Eastmond (2006). Glycerol and G3P were measured as described in Eastmond (2004) and Wei et al. (2004), respectively. Sugars were determined by a sequential series of assays linked to NADPH production (Pritchard et al., 2002). Pyridine nucleotides were extracted and the concentration of NAD⁺ and NADH determined as described by Shen et al. (2006). Sugars, glycerol, G3P, and pyridine nucleotide recoveries were determined to be within the range of 82% to 105% for all experiments.

For ¹⁴C-radiolabel feeding experiments, seedlings were incubated in 50 mm MES/NaOH (pH 5.2) plus 0.1 mm [U-¹⁴C]glycerol (20 MBq mmol⁻¹) or 50 mm [U-¹⁴C]Glc or [U-¹⁴C]Fru (1.25 MBq mmol⁻¹) for 4 h at 22°C. Incubations using glycerol were performed with 50 seedlings in 0.2-mL reactions and were stopped by the addition of 0.5 mL of 6 m formic acid. Each was conducted in a sealed 2-mL vial and ¹⁴CO₂ was collected in a well within the vial containing 0.1 mL of 15% (w/v) KOH. The ¹⁴C content of the KOH was determined by liquid scintillation counting. Incubations using Glc and Fru were performed with 100 mg fresh weight of seedlings in 2-mL reactions. Suc was extracted and the proportion of ¹⁴C-label in glucosyl and fructosyl moieties was determined according to the methods described in Fernie et al. (2001). In some experiments, unlabeled 50 mm Glc or Fru was also included in the incubation medium.

Mapping

The sdp6-1 mutant was out-crossed to the wild-type ecotype Landsberg erecta. F₁ plants were allowed to self-fertilize and the F₂ progeny were screened for the sugar-dependent phenotype. Genomic DNA was isolated from 768 F₂ sdp6-1 lines using the Extract-N-Amp plant PCR kit (Sigma-Aldrich). Mapping was carried out using simple sequence polymorphisms (Bell and Ecker, 1994) and cleaved amplified polymorphic sequences (Konieczny and Ausubel, 1993) utilizing information from the Monsanto Arabidopsis polymorphism collection (Jander et al., 2002). New markers used in this study (Fig. 2A, a–d) are: SSLP, 5′-TGCCCAATGGTATTCAGGTTAG-3′, 5′-GCAATGAATATAAACACCAAGGACA-3′; CAPS, 5′-GGATCCGAATCATTTTGCAC-3′, 5′-CCTCTGAGCTTGTGCCTTCT-3′, HindIII; dCAPS, 5′-GTTTCTCGATATTCTTTTCTCTCTAGATCTA-3′, 5′-CATGAAAAATTCAACTGATGACG-3′, ClaI; and dCAPS, 5′-TCACCCATCAACTGCCTAAGTTT-3′, 5′-GTGGATGGTAAGCCGGTTG-3′, HpaI. Candidate genes within the mapping interval were amplified from sdp6-1 and sdp6-2 genomic DNA by PCR and sequenced to identify mutations.

Transcript Analysis

DNAse-treated total RNA was isolated from Arabidopsis seedlings using the RNeasy kit from QIAGEN. The synthesis of single-stranded cDNA was carried out using SuperScript II RNase H− reverse transcriptase from Invitrogen. SDP6 and the actin (ACT2) transcripts were detected by PCR using primers SDP6F (5′-TCTTAACATTAAGGTTCGACGTAC-3′), SDP6R (5′-TCACAACAAATCATCCGAGAAACATGAG-3′), ACT2F (5′-GTTGGGATGAACCAGAAGGA-3′), and ACT2R (5′-CTTACAATTTCCCGCTCTGC-3′). The product sizes are 878 and 500 bp, respectively.

Enzyme Assays

Glycerol kinase and FAD-GPDH activity were assayed using the methods described in Eastmond (2004). PGI activity was assayed by spectrophotometry at 340 nm by coupling to the reduction of NAD (Aubert et al., 1994). Leaves of ecotype Columbia of Arabidopsis plants were homogenized in 10× (w/v) of extraction buffer (100 mm Tris-HCl [pH8.3], 10 mm MgCl₂, 10% [w/v] glycerol, 5 mm DTT, 1 mm EDTA) and centrifuged at 15,000 rpm for 10 min at 4°C. Fifty microliters of the supernatant were used in 1 mL of reaction buffer containing 100 mm Tris-HCl (pH8.3), 10 mm MgCl₂, 0.75 mm NAD, 3 units of G6P dehydrogenase, and 0.05 to 10 mm F6P. The inhibitory effect of G3P was tested by adding 2 to 20 mm to the reaction buffer. Plastidial PGI was inactivated by heating the extract at 50°C for 10 min (Yu et al., 2000).