Intramolecular ¹³C pattern in hexoses from autotrophic and heterotrophic C₃ plant tissues

Abstract

The stable carbon isotope ¹³C is used as a universal tracer in plant eco-physiology and studies of carbon exchange between vegetation and atmosphere. Photosynthesis fractionates against ¹³CO₂ so that source sugars (photosynthates) are on average ¹³C depleted by 20‰ compared with atmospheric CO₂. The carbon isotope distribution within sugars has been shown to be heterogeneous, with relatively ¹³C-enriched and ¹³C-depleted C-atom positions. The ¹³C pattern within sugars is the cornerstone of ¹³C distribution in plants, because all metabolites inherit the ¹³C abundance in their specific precursor C-atom positions. However, the intramolecular isotope pattern in source leaf glucose and the isotope fractionation associated with key enzymes involved in sugar interconversions are currently unknown. To gain insight into these, we have analyzed the intramolecular isotope composition in source leaf transient starch, grain storage starch, and root storage sucrose and measured the site-specific isotope fractionation associated with the invertase (EC 3.2.1.26) and glucose isomerase (EC 5.3.1.5) reactions. When these data are integrated into a simple steady-state model of plant isotopic fluxes, the enzyme-dependent fractionations satisfactorily predict the observed intramolecular patterns. These results demonstrate that glucose and sucrose metabolism is the primary determinant of the ¹³C abundance in source and sink tissue and is, therefore, of fundamental importance to the interpretation of plant isotopic signals.

Plants naturally discriminate between ¹²C and ¹³C isotopes during CO₂ fixation by photosynthesis: Thus, C₃ plant organic material is on average ¹³C depleted compared with atmospheric CO₂ (1). (Plant photosynthetic alternative pathways are designated C₃ and C₄. Carbon atom positions in sugars are designated as C-1, C-2, etc.) The amplitude of the isotope fractionation depends on environmental and physiological conditions, notably water deficit, temperature, CO₂ mole fraction, and stomatal conductance (2). Therefore, the natural ¹³C abundance (δ¹³C) in plant organic matter has often been used as a physiological marker to investigate plant traits in the field, such as water-use efficiency, photosynthetic capacity, and carbon allocation. {δ¹³C (‰) = 1,000*[(R_A − R_V-PDB)/R_V-PDB] is the carbon isotope composition expressed as the deviation of the carbon isotopic ratio (R = ¹³C/¹²C) of the analyte (R_A) relative to the international standard (Vienna Pee Dee Belemnite, V-PDB (R = 0.0112372).} However, the differential incorporation of ¹³C into various plant compounds and organs is primarily determined by metabolism, which may be accompanied by isotope fractionation. Typically, most heterotrophic tissues have been demonstrated to be ¹³C enriched by ca. 1‰ compared with photosynthetic tissues. In addition, plant metabolites have contrasting δ¹³C values: Cellulose and leaf starch have been shown to be ¹³C enriched by up to 3‰ (compared with sucrose), whereas lipids are commonly ¹³C depleted by up to 6‰. Exploiting raw plant organic matter in physiological studies thus appears challenging, because isotopic variations unrelated to the trait being investigated may be caused by tissue composition.

Isotopic disparity among metabolites is caused by metabolic reactions, which discriminate between ¹³C and ¹²C isotopes and differentially redistribute C atoms with dissimilar ¹³C-abundance values (see refs. 3 and 4 for details). A heterogeneous intramolecular δ¹³C pattern in source glucose will mean that metabolites derived from glucose metabolism inherit the specific ¹³C abundance of their precursor C atoms. Such a heterogeneity has been shown by partial (bio)chemical degradation of glucose obtained from beet sugar syrup and isotopic analysis by isotope ratio mass spectrometry (IRMS) of the products (5). Thus, a clear ¹³C enrichment in the C-atom positions C-3 and C-4 and a clear depletion in C-6 were detected. This pattern partially explains the ¹³C depletion in fatty acids (derived from C-1 + C-2 and C-5 + C-6) and the enrichment in CO₂ evolved by leaves in the dark (mainly from C-3 and C-4). The ¹³C enrichment of the C-3 and C-4 positions in glucose has been suggested to originate from the fractionation (that favors ¹³C) by aldolase (EC 4.1.2.13) during the condensation of two triose phosphate units into fructose-1,6-bisphosphate during photosynthetic glucose synthesis (6).

Nevertheless, despite the fundamental importance of sugars as the main carbon source of plant metabolism, thus as the determinant of the isotopic composition of nearly all metabolites, little progress in understanding the biochemical parameters that define their site-specific δ¹³C values has been achieved. It is now 20 y since the pioneering work by Rossmann et al. (5) that unequivocally demonstrated a heterogeneous distribution of ¹³C in natural glucose. This is primarily because of the technical difficulties inherent in obtaining such information using (bio)chemical degradation coupled with IRMS measurements. Recently, however, ¹³C NMR spectrometry has been developed that can measure the site-specific ¹³C abundance in natural glucose, fructose, and sucrose to 1‰ or better (7). The data thus obtained on glucose from beet sucrose were consistent with those of Rossmann et al. (5) quoted above, with the notable exception of the C-1 value. Furthermore, it was shown that fructose and glucose have antisymmetrical δ¹³C patterns in C-1 and C-2, thereby suggesting the involvement of isotopic fractionation during their isomerization equilibrium (8). However, the intramolecular pattern of source glucose (starch) or sucrose from leaves rather than a sink tissue has never been measured. Furthermore, the isotope fractionations associated with key enzymes involved in sugar allocation and partitioning within the plant are virtually unknown. Compound-specific analysis with IRMS has suggested that invertase (EC 3.2.1.26), which cleaves sucrose into glucose and fructose, fractionates against ¹³C, so that evolved fructose is ¹³C depleted compared with source sucrose (9). However, the site-specific isotope fractionation of this enzyme has yet to be elucidated. Clearly therefore, the relative roles of glucose biosynthetic reactions and sucrose metabolism as determinants of the intramolecular ¹³C pattern in sugars are unknown. Critically, there is presently no evidence that source leaf glucose has a heterogeneous intramolecular ¹³C pattern as found in glucose from sink tissues.

We have therefore addressed two key questions. First, what are the intramolecular δ¹³C distributions in source sugars? Second, what are the isotopic fractionations associated with the pertinent enzymes for sugar interconversions? To respond to these questions, we have carried out measurements with ¹³C NMR and determined the intramolecular patterns of ¹³C abundance in glucose from leaf, grain, and root. Because the ¹³C NMR technique requires substantial amounts of source material, we used the large leaves of cocklebur (Xanthium strumarium) to obtain pure leaf starch, wheat (Triticum aestivum) grain for storage starch, and sugar beet root (Beta vulgaris) to obtain storage sucrose. Further, we have measured the in vitro isotope fractionations associated with the reactions of commercially available invertase (EC 3.2.1.26) and glucose isomerase (EC 5.3.1.5). We have then integrated these experimentally determined fractionations into a model and show that the activities of the key enzymes aldolase, invertase, and glucose isomerase satisfactorily explain the intramolecular patterns observed in both source and sink organs.

Results

Intramolecular ¹³C Pattern in Starch and Sucrose.

A recent method based on an adapted, nonfractionating derivatization procedure blocking configurational changes, combined with quantitative ¹³C NMR spectrometric analysis (7), gives access to the natural ¹³C pattern in sugars. This technique has allowed us to determine the intramolecular, site-specific δ¹³C in sugars with a precision of 0.9‰ on the δ-scale (7, 8). Due to variations in absolute δ¹³C values caused by different sources of photosynthetic carbon, all of the values given here are expressed with respect to the molecular average as Δδ¹³C values. (Δδ¹³C is defined as the δ¹³C difference between the C-atom position of interest and the molecular average.) The intramolecular ¹³C pattern in glucose residues from leaf transitory starch is shown in Fig. 1A. The observed pattern is similar to that found in sugar beet glucose previously (5, 7), with the exception of a clear difference in C-1, which is ¹³C enriched by ca. 4‰ compared with the molecular average. The C-6 position is substantially ¹³C depleted, by about 8‰.

Fig. 1.

Intramolecular ¹³C patterns in glucose measured by site-specific ¹³C NMR. (A–D) Leaf starch (A), wheat grain starch (B), glucosyl moiety of sugar beet sucrose (C), and fructosyl moiety of sugar beet sucrose (D). Mean ± SE, n = 6. Experimental values (●, ○) are expressed as deviations from the molecular average, Δδ¹³C (see main text). Shaded areas indicate modeled intramolecular patterns with different parameter values used for calculations: starch synthesis within 10–50% of net carbon fixation in leaves (A) or contribution of leaf starch to the buildup of reserves within 50–90% (L = 0.1–0.5) (B–D). In C, values from Rossmann et al. (5) are indicated (○).

The intramolecular ¹³C pattern in glucose from wheat grain storage starch is shown in Fig. 1B. The pattern is seen to be very different from that in leaf transitory starch, with the C-1 and C-2 positions, respectively, ¹³C depleted and ¹³C enriched. In glucose from beet root sucrose (Fig. 1C) the ¹³C pattern is rather similar to that in leaf starch, differing principally in having a more pronounced ¹³C enrichment in C-4. The ¹³C pattern in the fructose moiety of the same sucrose is very different from that of the glucose moiety (Fig. 1D). C-1 and C-2 are relatively ¹³C depleted and ¹³C enriched, implying that the isomerization between glucose and fructose, which involves the C-1 and C-2 positions, causes isotopic alterations.

Isotope Fractionation Introduced by Isomerase and Invertase.

The site-specific ¹³C-isotope fractionations associated with these enzyme activities were measured in vitro. The results are expressed relative to the δ¹³C in the C-4 position (Fig. 2). No isotopic change is anticipated at this position, because it is distal to the chemical bonds involved in the reaction.

Fig. 2.

(A and B) Site-specific ¹²C/¹³C isotope fractionation (‰) of the glucose isomerase (A) and the invertase (B) reactions assayed in vitro. Asterisks indicate fractionations significantly different from zero (P < 0.05). For invertase, there is no thermodynamic measurement because the reaction is irreversible. In both cases, fractionation values are relative to that in C-4; that is, it was assumed that no fractionation occurred at the C-4 position because no net chemical event takes place at this site. Positive values indicate a fractionation against ¹³C, and negative values indicate a fractionation in favor of ¹³C.

Glucose isomerization was found to be accompanied by significant site-specific kinetic and thermodynamic isotope effects. For the former, fractionations were measured over a short 1- to 2-h period on the product in the fructose-to-glucose direction and for the latter when reaction had reached equilibrium (Materials and Methods). When measured in the fructose-to-glucose direction, the reaction was accompanied by a significant normal kinetic isotope effect of ca. 15‰ in C-2 (Fig. 2A). (A normal isotope effect selects against the substrate molecules with ¹³C at that position, leading to a lower ¹³C/¹²C ratio at that position in the product. An inverse isotope effect selects in favor of substrate molecules with ¹³C at that position.) This indicates that, during the isomerization of glucose to fructose (forward reaction), evolved fructose molecules will be ¹³C depleted by 15‰ in C-2 compared with the corresponding position in glucose. The thermodynamic isotope fractionations, which are independent from the enzyme, manifest themselves when the reaction has reached equilibrium. Again relative to fructose as initial substrate, at equilibrium there was a normal thermodynamic isotope effect in C-2 and C-6 (ca. 7‰ and 4‰, respectively) and an inverse thermodynamic isotope effect in C-1 (ca. 13‰) (Fig. 2A). Overall, therefore, the isomerization of fructose to glucose (backward reaction) fractionated against ¹³C by ca. 18‰ in C-1 and ca. 8‰ in C-2. Taken as a whole, glucose isomerization to fructose tends to enrich in ¹³C the C-1 in glucose and the C-2 in fructose. These results are in agreement with thermodynamic potentials that explain the systematic enrichment in carbonyl compared with hydroxyl groups (10, 11).

The invertase reaction was accompanied by a kinetic isotope fractionation against ¹³C (ca. 7‰) in the C-2 position of the fructosyl moiety: There was no significant effect in any other site. Because invertase is not reversible, there is no possible thermodynamic isotope fractionation. The fractionation in C-2 obtained here agrees with the chemical mechanism of the reaction, which is predicted to involve a primary ¹³C isotope effect only in this position (9).

We recognize that the enzymes studied here were not from plants but from microorganisms (Streptomyces, Saccharomyces). In particular, plant glucose isomerase uses glucose-6-phosphate as substrate, not glucose. Nevertheless, the chemical mechanisms of the plant counterparts are the same. Although some differences in the ¹²C/¹³C isotope effect of enzymes from different biological origins have been reported (12, 13), it seems reasonable to anticipate isotope effects of the same degree for the plant and microbial enzymes.

Isotope ¹³C-Distribution Patterns Obtained by Modeling.

The model previously described for analyzing ¹³C allocation in plant metabolism in the steady state (14) was extended to integrate the metabolic reactions of sugar interconversions (depicted in Fig. 3) (SI Text). The output of the model is the site-specific δ¹³C pattern in glucose, fructose, and sucrose. Due to the difference in intramolecular pattern in leaf cytoplasmic day sucrose and in starch-derived night sucrose, the model depends upon the nature of source sucrose (day or night) used to accumulate grain starch or root sucrose. The robustness of the model was then explored, using different levels of prevalence of day or night sucrose (shaded area in Fig. 1 B–D).

Fig. 3.

(A–D) Simplified scheme showing the important steps of sugar metabolism in the illuminated leaf (A), the darkened leaf (B), wheat grain (C), and beet root (D). 1, ribulose-1,5-bisphosphate carboxylase/oxygenase; 2, (trans)aldolase; 3, transketolase; 4, phosphoglucose isomerase; 5, invertase; 6, sucrose synthase. In A–D, the major pathway is indicated with thick arrows. Known partition factors are given (%); for example, in B, the commitment of fructose to glycolysis represents about 20% of total fructose utilization; in D, 35% of glucose produced by invertase accumulates in the vacuole whereas 65% reforms sucrose. In C and D, source sucrose comes from A and B, with variable proportions (main text and Fig. 1 legend). Abbreviations: DHA-P, dihydroxyacetone phosphate; Gald-3-P, 3-phosphoglyceraldehyde; PGA, 3-phosphoglyceric acid. For clarity, bis-phosphorylated intermediates are not shown.

Discussion

It has previously been demonstrated that the ¹³C distribution within plant sugars and primary storage carbohydrates is not homogeneous and that this heterogeneity is maintained in metabolites derived from these primary carbon sources (3, 4). It is now apparent that an analysis of the mean δ¹³C distribution (molecular average) in source sugars is insufficient to link effectively δ¹³C values to plant metabolism. First, it is clear that there is sugar-specific variation in the intramolecular δ¹³C distribution in source sugars, which will vary dependent on the source involved. Second, it is evident that this variation will be governed by isotopic fractionations associated with enzymes involved in sugar interconversions. It remains to be unraveled to what extent the intrinsic enzyme properties and metabolic-pathway–related C partitioning will impact the δ¹³C distribution determined in metabolites.

Involvement of Enzymes in Defining Intramolecular ¹³C Patterns in Primary Sugars.

In general, there is a relative enrichment of the C-3 and C-4 positions compared with the average of the other positions. This enrichment is particularly clear in the glucose residue of cocklebur starch (Fig. 1A) and the glucosyl moiety of sugar beet sucrose (Fig. 1C). During photosynthetic carbon metabolism, these positions are derived from triose phosphates through condensation by aldolase to form fructose-1,6-bisphosphate (Fig. 3, reaction 2), a reaction that is known to favor ¹³C (6). Indeed, the equilibrium isotope fractionation for the rabbit muscle aldolase has been shown to be, respectively, 3.6‰ and 4.9‰ in favor of ¹³C in C-3 and C-4. Back calculation from the isotope distribution in glucose obtained by Rossmann et al. (5) and from β-factors (thermodynamic potential) has arrived at larger values, up to 18‰ (14). Here, the comparison of positional average values (C-3 + C-4 vs. C-1 + C-2 + C-5 + C-6) suggests an effective fractionation in favor of ¹³C in C-3 + C-4 of ca. 10‰.

However, the aldolase reaction is certainly not the sole actor in the determination of intramolecular ¹³C patterns, because the C-6 is systematically ¹³C depleted and the C-1 and C-2 positions vary antisymmetrically (Fig. 1 C and D). Hexose metabolism is certain to be influenced by the isomerization equilibrium between glucose-6-phosphate and fructose-6-phosphate, because such a conversion is required to synthesize sucrose or starch (Fig. 3). We show here that glucose isomerase is accompanied by both a thermodynamic and a kinetic isotope effect (Fig. 2A). When equilibrium is reached, the C-1 in glucose and the C-2 in fructose are ¹³C enriched.

Other enzymes are also likely to influence the ¹³C pattern in sugars. Notable among these is invertase, which cleaves sucrose into glucose and fructose and fractionates against ¹³C, thereby depleting the C-2 position in fructose (Fig. 2D). However, the impact of fractionation introduced by invertase will vary depending on the tissue. Thus, in beet roots the effect is rather small because the proportion of sucrose molecules hydrolyzed by invertase is only ca. 3%, the rest being cleaved by sucrose synthase (EC 2.4.1.13) (Fig. 3C) (15, 16). The chemical mechanism of the latter does not involve the cleavage or creation of a C-C bond and therefore is unlikely to discriminate significantly between carbon isotopes at the level of the fructose moiety (secondary isotope effect) (17). (A primary isotope effect is associated with the atoms involved in the cleavage or formation of a bond. Secondary isotope effects occur due to isotopic substitution in adjacent atoms that are only indirectly involved in the reaction.) In addition, the possible fractionation against ¹³C in C-1 of the glucose moiety is counterbalanced by the very large commitment of this reaction (97%; Fig. 3C, reaction 6). Similarly, the sucrose phosphate synthase (EC 2.4.1.4) that forms sucrose phosphate from UDP-glucose and fructose-6-phosphate cannot fractionate in C-1 of the glucosyl moiety, because UDP-glucose is essentially fully committed to sucrose production.

Triose phosphate isomerization, the interconversion of 3-phosphoglyceraldehyde and dihydroxyacetone phosphate, may also introduce isotope fractionation. However, the kinetic fractionation in carbon is certain to be small, because the limiting step of the reaction [proton transfer to the active site (18)] again does not involve a primary carbon isotope effect. The thermodynamic fractionation is probably also negligible, due to the free energy associated with the reaction, which leads to a ratio of 3-phosphoglyceraldehyde to dihydroxyacetone phosphate of 3:97 (19): that is, to a nearly complete conversion of 3-phosphoglyceraldehyde into dihydroxyacetone phosphate. In photosynthesis, dihydroxyacetone phosphate is synthesized extremely rapidly from 3-phosphoglyceraldehyde by the very efficient enzyme triose phosphate isomerase (k_cat/K_m ∼ 10⁵ L⋅mol⁻¹⋅s⁻¹), thereby preventing any ¹²C/¹³C isotope fractionation with respect to source phosphoglycerate. In addition, should the fractionation be large or significant, then the C-2 position in leaf glucose would be considerably enriched: The evidence is quite to the contrary (Fig. 1A).

Metabolic Pathways and Partitioning in Carbohydrate Metabolism.

The combination of the known metabolic pathways (Fig. 3) and the isotope fractionation values discussed above allows us to predict the ¹³C patterns in sugars from equations based on steady-state conditions (shaded areas in Fig. 1; see SI Text for calculation details). Calculations nevertheless depend on physiological parameters, such as the proportion of day sucrose vs. night sucrose (denoted as L in the model) (SI Text) used to accumulate starch or sucrose in wheat and beet, respectively, because the modeled ¹³C patterns in day and night sucrose are different. Regrettably, specific data on diurnal kinetics of grain starch synthesis or beet sucrose accumulation are scarce. Available data indicate that sucrose accumulation and growth are more intense in beet root during the night (20, 21) and that in wheat grains, starch accumulation is privileged in the night, whereas cell division and cell wall synthesis occur in the light (22, 23). Therefore, the calculations were carried out with a prevalence of night sucrose (larger utilization of sucrose synthesized from leaf starch remobilization: L < 0.5). The predicted values are not very sensitive to L, however, as seen from the relatively narrow shaded areas in Fig. 1 B–D. Similarly, in leaves the predicted ¹³C pattern in starch is not very sensitive to variations in the starch synthesis rate (Fig. 1A). The agreement between modeled and measured ¹³C patterns is generally very satisfactory, with no Δδ¹³C difference of more than 4‰. Particularly noteworthy is that the patterns generated by the model match very closely the ¹³C depletion in the C-6 position of glucose compared with the C-1, a difference that up to now has been poorly understood (5, 14, 24). In the framework of the model, this clearly demonstrates that the isotopic difference between these two positions principally originates from the ¹³C enrichment in C-1 caused by the isomerization equilibrium between glucose and fructose (Fig. 2). In the steady state, invertase further tends to ¹³C deplete free fructose accumulated in the vacuole and to cause a ¹³C enrichment in fructose recycled to sucrose. Accordingly, the modeled Δδ¹³C values in C-1 and C-2 of fructose from sugar beet sucrose that result from the combined action of invertase and isomerase (Fig. 3D) reflect perfectly the observed values (Fig. 1D).

Of course, the present calculations cannot be considered to be fully representative in that they are derived from a model that does not integrate all of the complexities of sugar metabolism and partitioning in plants, such as pentose phosphate metabolism. For example, there is currently considerable uncertainty as to the specific source of carbon used to synthesize starch in cereals and the possible involvement of stem and culm reserves (mainly in the form of fructans) (25, 26). Should fructans be involved in wheat grain filling, their impact on the isotopic pattern in starch would be a ¹³C enrichment (nonfractionating remobilization of “old” fructans), a ¹³C depletion (conversion of fructose into fructans), or no effect at all (steady fructan pool) in the C-2 position. Further data on the δ¹³C in fructans and the dynamics of fructan pools would thus be necessary to improve our current model of the intramolecular ¹³C pattern in wheat sugars. In addition, in both beet and wheat, carbon partitioning to sucrose and starch, respectively, seems to be highly sensitive to environmental conditions, such as day and night temperature (20, 27) and water stress (28, 29). Plant responses to environmental conditions encompass changes in enzyme activity including invertase (e.g., in maize, ref. 30). We recognize that our analysis is based on typical documented partition factors and commitments largely obtained from plants grown under experimental conditions. In the field, intramolecular ¹³C patterns may vary with different plant culture conditions or crop varieties.

Consequences of Intramolecular ¹³C Patterns in Primary Sugars on Plant Metabolites.

The intramolecular heterogeneous distribution of ¹³C in glucose and fructose has pervading consequences for ¹³C isotopic distribution in plants, because all of the downstream metabolites inherit the isotopic composition of their metabolic source. Well-known examples are the ¹³C enrichment in leaf-respired CO₂ and the overall ¹³C depletion in fatty acids (see ref. 24 for a discussion), which derive from C-3 + C-4 and C-1 + C-2 + C-5 + C-6 positions of glucose, respectively. It is interesting to note that the intramolecular variation in ¹³C isotopic distribution in glucose also plays a role in defining the well-established alternating site-specific ¹³C distribution in fatty acids, described both for bacteria (31) and for plants (32) and previously assigned principally to kinetic isotope effects in the pyruvate dehydrogenase reaction (31, 33).

In this study, we show that source glucose used in darkened leaves (remobilized from starch) is ¹³C enriched in the C-1 position. The decarboxylation of this carbon atom by the pentose phosphate pathway should produce ¹³C-enriched carbon dioxide, unless an effective isotope fractionation against ¹³C by the associated decarboxylase [phosphogluconate dehydrogenase (decarboxylating) EC 1.1.1.44] occurs (34, 35). In contrast, the ¹³C depletion in C-6 should impact on C atoms in metabolites: For example, carboxylic groups in histidine and tryptophan or the formyl (or methylene) group exchanged in C₁ metabolism, which come from the C-1 and C-6 positions. We further postulate that such isotopic consequences are dissimilar between plant organs because the Δδ¹³C values are different. For instance, CO₂ evolved by pyruvate dehydrogenase (EC 1.2.4.1), the first decarboxylation step of respiration, should be relatively more ¹³C enriched in sugar beet root than in wheat grains (Fig. 1). Similarly, metabolites formed from glycolytic acetyl-CoA should be more ¹³C depleted in beet root than in wheat grains. In other words, in addition to tissue composition being variable between organs, the intramolecular pattern may cause changes in the natural ¹³C abundance of organic matter (3, 36–38). Our results show that the metabolism of sucrose taking place in sink organs has clear effects on the ¹³C pattern: The less heterogeneous pattern in wheat grain should lead to smaller isotopic differences among downstream metabolites than in sugar beet roots.

Other metabolic pathways can have a significant influence on the ¹³C signal of plant tissues. Glucose derived from C₄ plant metabolism does not seem to have the same ¹³C distribution as that from C₃ plants (5, 7). In addition, anaplerotic carbon fixation by phosphoenolpyruvate carboxylase (PEPc; EC 4.1.1.31) leads to a positional ¹³C enrichment in organic acids and amino acids (3). The PEPc activity has been shown to vary between plant parts and this might correlate with the well-documented ¹³C enrichment in sink organs (39, 40). Because organic acids are key intermediates involved in respiration and in the biosynthesis of amino acids, their intramolecular ¹³C pattern is thus of importance to better understand the ¹³C distribution within plants. This needs to be addressed in the future.

Materials and Methods

Chemicals.

Silica (63- to 200-µm mesh), TLC plates (aluminum sheets, silica gel 60 F₂₅₄), acetic acid (99% vol/vol), ethanol (99% vol/vol), magnesium sulfate pentahydrate, charcoal, and celite were purchased from VWR. Glucose isomerase (EC 5.3.1.5) immobilized from Streptomyces murinus, invertase (EC 3.2.1.26) from baker’s yeast (Saccharomyces cerevisiae), amyloglucosidase (EC 3.2.1.3) from Aspergillus niger, and acetonitrile (HPLC quality) were purchased from Sigma Aldrich. Sodium acetate was from Fluka.

Plant Growth Conditions.

Xanthium strumarium L. were grown from seeds in the greenhouse as described in ref. 41. After 2 mo, mature leaves (ranks 5 and 6 from the apex) were collected in the light and immediately frozen in liquid nitrogen for starch extraction.

Starch Extraction and Hydrolysis.

Starch was purified from leaves as described in ref. 42 with acid gelatinization (HCl) and precipitation in cold methanol. Samples were then lyophilized for further analysis. One gram of pure starch (representing 30 leaves) was placed in 10 mL of acetate buffer (pH 4.6). The mixture was then autoclaved (120 °C; 1 bar) for 30 min to gelatinize the starch. Ten milliliters of acetate buffer and 20 mg of amyloglucosidase from Aspergillus niger were added. The medium was heated at 60 °C and the reaction was followed by TLC on silica plates (AcOEt/MeOH/H₂O 7:2:1). After starch disappearance (ca. 2 h), the solution was filtered on a celite/silica bed (1 cm). The filtrate was then evaporated and coevaporated with ethanol to obtain a syrup that was used without further purification for glucose derivatization.

Sugar Quantitation by HPLC.

Sugars were analyzed by high pressure liquid chromatography on an NH₂ column (Phenomenex) equipped with a loop injector (20 µL) and a refractometer. Elution was made with a mixture of CH₃CN/H₂O (80:20) at a flow rate of 1mL/min at room temperature.

Isomerase Assay.

The reaction medium was prepared with 500 mL of distilled water containing 1.5 g of MgSO₄⋅5H₂O at pH 7.5 (adjusted with NaHCO₃). To 100 mL of the medium was added 17 g of either glucose or fructose and the medium was stirred at 60 °C for 30 min. Immobilized glucose isomerase (400 mg, 350 units/mg) was then added and the reaction was followed by HPLC (see above). The chemical equilibrium was reached in ca. 24 h and the ratio of fructose to glucose was 1:1. To stop the reaction, the medium was filtered on a 0.22-µm filter. The resulting filtrate was evaporated to yield a syrup, which was used without further purification for glucose and fructose derivatization. For the determination of equilibrium isotope effects, the reaction was followed for 3 or 10 d and both fructose and glucose were used as substrates. For the determination of kinetic isotope effects, the reaction was run with fructose as the substrate and was stopped at different times (1–2 h).

Invertase Assay.

One milliliter of a 5-mg/mL (300 units/mg) invertase solution in water was added to 100 mL of sucrose solution (0.43 M) at room temperature. The reaction was quenched at different times (between 50 and 120 min) by adding 0.5 mL of 1 mM NaOH. Sucrose, glucose, and fructose concentrations were followed by HPLC.

Sucrose Isolation.

To the quenched invertase assay was added 15 g of activated charcoal. The mixture was stirred for 15 min at room temperature and then placed at 4 °C for 1 h. The mixture obtained was deposited on a charcoal celite column (40 g charcoal; 40 g celite, conditioned with distilled water) equipped with a vacuum flask (flow rate of ca. 10 mL/min). Elution was carried out with pure water, ethanol/water (1:99), ethanol/water (5:95), and then pure ethanol until all of the sucrose had eluted. The fractions were then evaporated and dried over P₂O₅ in a desiccator overnight before derivatization.

Carbohydrate Derivatization and NMR Analysis.

The carbohydrates (glucose, fructose, and sucrose) were converted to glucose and fructose acetonide derivatives and were analyzed by quantitative ¹³C NMR spectrometry as described previously (7, 8).

Calculation of Isotope Effects.

Kinetic isotope effects were calculated from experimental data, using equations describing isotopic Rayleigh effects (43). The positional isotope effect values were calculated assuming a negligible isotope effect on the C-4 position of both glucose and fructose (Fig. 2).

Modeling.

The calculation of the site-specific carbon isotope composition (¹³C pattern) was carried out using the model of Tcherkez et al. (14), completed with the isotope fractionations of glucose isomerization found here (Fig. 3 A and B). Day and night sucrose (output of the model) were used as source sucrose to compute the ¹³C pattern in starch and sucrose in wheat grain and sugar beet root, respectively, with the metabolic pathways considered in Fig. 3 C and D and the isotope fractionation by aldolase, invertase, and isomerase. In all cases, the fractionation by starch synthase, glycolysis, and sucrose synthase (SuSy) was assumed to be negligible and the metabolism was assumed to be in the steady state (see SI Text for further details).