SEARCH

SEARCH BY CITATION

Keywords:

  • carbon partitioning;
  • central carbon metabolism;
  • developing soybean cotyledon;
  • malic enzyme;
  • metabolic flux analysis;
  • nuclear magnetic resonance;
  • oxidative pentose phosphate pathway

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Metabolic flux maps developed from 13C metabolic flux analysis (13C MFA) are effective tools for assessing the response of biological systems to genetic or environmental perturbations, and for identifying possible metabolic engineering targets. Experimental treatments were designed to distinguish between temperature effects prior to, and during incubation in vitro, on primary metabolism in developing soybeans. Biomass accumulation increased with temperature as did carbon partitioning into lipids. The flux through the plastidic oxidative pentose phosphate pathway (pglP) relative to sucrose intake remained fairly constant [∼56% (±24%)] when cotyledons were transferred from an optimum growth temperature to varying temperatures in in vitro culture, signifying a rigid node under these conditions. However, pglP flux ranged from 57 to 77% of sucrose intake when growth temperature in planta varied and were cultured in vitro at the same temperature (as the plant), indicating a flexible node for this case. The carbon flux through the anaplerotic reactions catalysed by plastidic malic enzyme (meP), cytosolic phosphoenolpyruvate (PEP) carboxylase and the malate (Mal) transporter from the cytosol to mitochondrion varied dramatically with temperature and had a direct influence on the carbon partitioning into protein and oil from the plastidic pyruvate (Pyr) pool. These results of the in vitro culture indicate that temperature during early stages of development has a dominant effect on establishing capacity for flux through certain components of central carbon metabolism.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Metabolic flux, the flow of carbon through metabolic networks, is a crucial indicator of cell physiology (Stephanopoulos 1999). 13C metabolic flux analysis (13C MFA) is the organism-wide quantification of metabolic fluxes by using carbon-labelling experiments and mathematical analysis of the labelling data. Metabolic flux maps, the outcomes of 13C MFA, indicate steady-state fluxes through various reactions and pathways in the organism. A recent review (Ratcliffe & Shachar-Hill 2006) has underscored the significance of 13C MFA and flux maps towards characterization of phenotype in plants.

In plants, 13C MFA and flux maps can be applied to quantify fluxes in known pathways (Rontein et al. 2002; Schwender, Ohlrogge & Shachar-Hill 2003; Sriram et al. 2004), to assess fluxes through parallel pathways in separate compartments (Sriram et al. 2004) or to identify previously unknown pathways (Schwender et al. 2004). In plant metabolic engineering, systemic 13C MFA is an effective tool to assess the impact of an environmental or genetic perturbation on carbon flow. It provides insights on whether specific metabolic nodes are ‘rigid’ or ‘flexible’ (Stephanopoulos & Vallino 1991), and thus the degree of amenability of the node to metabolic engineering. Such results can suggest further manipulations, which on implementation can be followed by 13C MFA in an iterative manner to achieve an optimally engineered biological system (Nielsen 1998). In plant systems biology, 13C MFA in combination with global profiling techniques, such as proteomics, transcriptomics and metabolomics, has significant potential for building predictive models of plant metabolism (Girke et al. 2003; Minorsky 2003; Sweetlove, Last & Fernie 2003; Sweetlove & Fernie 2005; Kauffman, Ogunnaike & Edwards 2006; Ratcliffe & Shachar-Hill 2006).

Although 13C MFA is an established technique to characterize prokaryotic systems, its entry into plant science is fairly recent. This is chiefly because the intricacy and extensive subcellular compartmentation of plant metabolism renders the mathematical techniques of 13C MFA challenging and non-trivial for the casual user. Previously, metabolic flux analysis has been employed to understand metabolism in tomato cell suspension (Rontein et al. 2002), maize root tips (Dieuaide-Noubhani et al. 1995) and maize kernels (Glawischnig et al. 2001, 2002). It has also been applied extensively to probe central metabolism and mitochondrial metabolism in Brassica napus embryos (Schwender et al. 2003, 2004; Schwender, Shachar-Hill & Ohlrogge 2006). Towards the application of flux analysis in plant systems, we developed a generic flux analysis tool, NMR2Flux, (Sriram et al. 2004) incorporating recent advances in efficient flux evaluation (Wiechert & Wurzel 2001; Sriram & Shanks 2004). Together with an extensive two-dimensional (2D) nuclear magnetic resonance (NMR) spectroscopy-based isotopomer data set, extracellular and biomass synthesis fluxes measurements, and statistical analysis, we used NMR2Flux to quantify compartmented fluxes in developing soybean cotyledons (Sriram et al. 2004).

In this paper, we report the application of NMR2Flux to investigate the effect of temperature on central carbon metabolism in developing soybean cotyledons. Soybean is a major source of high-quality protein, edible oil, and food products, and also several industrial non-food applications such as paper coating, textiles and plastics (Myers 1993; Kinney 1998). It is difficult to develop soybeans both with high protein and high oil, however, because of the inverse relationship between these two components (Nakasathien et al. 2000). Consequently, increasing the seed protein concentration without adversely affecting the yield and oil content has been difficult to achieve (Cober & Voldeng 2000).

Studies on the effect of day and night growth temperatures on the seed composition indicated that during the seed-fill and maturation period, protein concentration increased, whereas oil decreased with increasing temperature (Gibson & Mullen 1996). Another study with different soybean cultivars indicated that oil concentration increased with increasing temperature (Piper & Boote 1999). Although various studies have indicated a significant response to temperature in developing soybean, the metabolic pathways contributing to the observed changes in biomass accumulation and composition are not understood clearly. Our study was designed to probe pathway interactions in response to temperature in developing soybean cotyledons, towards the ultimate objective of understanding carbon regulation for improved protein and oil production.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Soybean cotyledon culture

Soybeans (Glycine max cv. Evans) were grown in controlled environment chambers under conditions optimal for seed production and growth: 27 °C/20 °C day/night temperature, 14 h photoperiod, with high-output fluorescent lamps providing approximately 500 µmol photons m−2 s−1 of photosynthetically active radiation at the upper leaves. Sufficient water and nutrients were applied to the soil to ensure vigorous plant growth. At 10 days after flowering (DAF) and prior to rapid seed growth, the growth temperature was adjusted to one of the three conditions −35 °C day/27 °C night (H – high), 27 °C/20 °C (M – medium) and 20 °C/12 °C (L – low). These conditions were maintained for the duration of seed development. Ten days later (20 DAF), pods were harvested from the central nodes and cotyledons were isolated under aseptic conditions for in vitro culture. The cotyledons were approximately 50 mg fresh weight and selected for uniform initial size. They were cultured aseptically for 6 d in separate 25 mL Erlenmeyer flasks containing 4 mL liquid medium. The medium comprised of 146 mm sucrose (10% U-13C sucrose from Isotec, Miamisburg, OH, USA) and 37 mm glutamine as the only carbon sources, and was replaced every 3 d. The flasks were gently shaken at 100 r.p.m. and were maintained at one of the three temperatures (H, M, L) within Percival incubators (Percival Scientific, Inc., Perry, IA, USA) at approximately 100 µmol photons m−2 s−1 light intensity. In all, there were five temperature treatments. Cotyledons collected from plants grown at 35 °C/27 °C were incubated at 35 °C/27 °C (HH). Cotyledons sampled from plants growing at 27 °C/20 °C were incubated at one of the three temperatures: 35 °C/27 °C (MH), 27 °C/20 °C (MM) or 20 °C/12 °C (ML). Cotyledons from plants growing at 20 °C/12 °C were incubated at 20 °C/12 °C (LL). All temperature treatments were performed in triplicate. Cotyledons that were harvested after 6 d in culture remained seed-like and were rinsed with distilled water, frozen in liquid nitrogen and lyophilized at −50 °C and 0.133 mbar for 72 h. The lyophilized cotyledons were finely ground for further processing.

Biomass quantification and fractionation

Biomass growth was quantified by measuring the fresh and dry weights of the cotyledon. The total lipid content was quantified gravimetrically using hexane extraction. Protein in the defatted pellet was separated from non-protein nitrogen by precipitation in 240 g kg−1 trichloroacetic acid and was quantified by microkjeldahl procedures. Soluble sugars from the defatted pellet were extracted in 800 g kg−1 ethanol at 60 °C for 20 min. The extraction was repeated four times. Successively, starch in the remaining pellet was digested in 100 mm citrate buffer containing amyloglucosidase in the ratio 0.025:1 mg of dry weight. Starch content was quantified using the starch assay kit (Sigma, St. Louis, MO, USA). Measurement of sucrose and glutamine consumption from the medium has been explained in Supplementary Table S1.

Protein and starch hydrolysis, and NMR sample preparation

The protein and starch were vacuum hydrolysed in hydrolysis tubes (Pierce Endogen, Rockford, IL, USA) using 6 N constant boiling hydrochloric acid (HCl) (Sigma). The acid was added in the ratio 0.5 mL HCl: 1 mg protein/starch. The tube was evacuated, purged with N2 to remove residual air and re-evacuated. The sample (starch/protein) was hydrolysed for 4 h at 150 °C in a heat block. The acid in the hydrolysate was removed with a Rapidvap evaporator (Labconco, Kansas City, MO, USA). The residue was reconstituted in 2 mL of deionized water and lyophilized for 72 h. The sample was reconstituted in 500 µL of deuterium oxide and was transferred into an NMR tube. The sample pH was adjusted to 0.75–1.0 using deuterium chloride. Amino acids in the samples were quantified by high-performance liquid chromatography (HPLC) (Protein facility, Department of Biochemistry, Biophysics and Molecular Biology at Iowa State University) using a reverse phase C18 silica column, with detection at 254 nm (Supplementary Fig. S1).

NMR spectroscopy

[13C, 1H] Heteronuclear single quantum correlation (HSQC) spectroscopy and [1H, 1H] total correlation spectroscopy (TOCSY) were performed on a Bruker Avance DRX 500 MHz spectrometer (NMR facility, Department of Biochemistry, Biophysics and Molecular Biology at Iowa State University) at 298 K. The reference zero p.p.m. was set by the methyl signal of dimethylsilapentanesulfonate (Sigma) as an internal standard. The resonance frequency of 13C and 1H were 125.7 and 499.9 MHz. The number of scans ranged from 16 to 32 and was inversely proportional to the concentration of the NMR samples. The spectral width was 5028.05 and 5482.26 Hz along the 13C (F1) and 1H (F2) dimension, respectively. Peak aliasing was used to minimize the sweep width in the 13C dimension. The number of complex datapoints were 1024 (1H) and 900 (13C). While acquiring the TOCSY spectra, the DIPSI-2 (Decoupling In the Presence of Scalar Interactions) sequence was used for isotropic mixing, with a mixing time of 76 ms. Additionally, 2-D HSQC spectra were obtained using a J-scaling factor of 6 in F1 dimension (Willker, Flogel & Leibfritz 1997). NMR spectra were acquired and processed by using the Xwinnmr (Bruker) software. The HSQC and TOCSY spectra were analysed by using the free software NMRview (OneMoonScientific, Inc., http://www.onemoonscientific.com) (Johnson & Blevins 1994). Deconvolution of multiplet intensities was carried out using a software written in our lab (Sriram et al. 2004) based on a spectral processing algorithm, first proposed by van Winden et al. (2001). Peak assignments were verified using 2-D [1H, 1H] TOCSY and three-dimensional (3-D) [13C, 1H, 1H] TOCSY analyses on a 100% 13C-labelled protein samples.

Flux evaluation methodology and computer program

Fluxes were quantified by using the generic flux evaluation software NMR2Flux reported by Sriram et al. (2004). The working of the tool NMR2Flux was demonstrated by employing the MM temperature condition as a case study (Sriram et al. 2004). NMR2Flux employs isotopomer balancing and a global optimization routine to calculate a stoichiometrically feasible set of fluxes that best explains the observed isotopomer measurements. Approximately 300 simulations were performed for each temperature condition. Statistical analysis using the JMP software (SAS Institute, Cary, NC) was carried out to determine the SDs of the fluxes and reversibilities. The Student's t-test (P < 0.05) was performed to compare the means of different temperature conditions. The error in the isotopomer measurements was used to perform random Monte Carlo simulations to generate probability distributions for the fluxes (Press et al. 1992). For the mathematical details, the reader is referred to our previous work (Sriram et al. 2004).

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

This study was designed to investigate temperature effects on fluxes through central carbon metabolism in developing soybean cotyledons. In most previous studies that documented temperature effects on protein and oil content in soybean (Westgate et al. 1995; Piper & Boote 1999; Westgate et al. 2000; Pipolo, Siclair & Camara 2004), the optimum temperature for seed growth was about 27 °C. A growth temperature higher or lower than this optimum resulted in decreased protein accumulation, decreased oil accumulation, or both. In addition, temperature treatments imposed during the early stage of cotyledon development, when cells are being formed, had a larger impact on final seed composition (Howell & Cartter 1953; Bils & Howell 1963). Therefore, we created two related cases for this study: one to determine whether 13C MFA could distinguish temperature effects on seed metabolism, and a second to investigate whether the growth temperature during cell formation limited the capacity for metabolic flux later in seed development. For case 1, cotyledons were collected from plants growing under moderate conditions (27 °C day/20 °C night – M), in a growth chamber and cultured in vitro at varying temperatures (35 °C/27 °C – H, 27 °C/20 °C – M and 20 °C/12 °C – L). These treatments are designated as MH, MM and ML, respectively. For case 2, the plants were grown under high temperature, moderate temperature or low temperature conditions during seed development, and cotyledons were cultured at the same temperature at which they were growing. These treatments are designated as HH, MM and LL, respectively. In both cases, the cotyledons were selected for uniform initial size of 50 mg fresh weight in all temperature treatments, to ensure that they were in the same stage of metabolic development. Metabolic flux was analysed during the middle of the seed-fill period, when biomass, protein and oil were accumulating rapidly. The intent was to determine how a temperature change when introduced during incubation in vitro affected metabolic fluxes compared with the same temperature treatment imposed during seed formation and development. The metabolic fluxes presented in this paper reflect an average flux of the day/night cycle over a period of 6 d.

Treatment effects on biomass and composition

The MM temperature condition was best suited for maximal relative protein content (as a fraction of dry weight) of the cotyledons (Table 1). To contrast the metabolic rates of different temperature treatments against this optimal condition, we compared the fluxes of the different temperature conditions relative to the MM temperature condition.

Table 1.  Comparison of biomass constituents for case 1 and case 2
Temperature studyGrowth temperatureIncubator temperatureGrowth/incubator temperatureTotal dry weight accumulation (mg/cotyledon)% of biomass
ProteinStarchOil
  • MM was the optimum temperature for protein accumulation and was common between both cases.

  • a

    The values for the MM temperature was published previously (Sriram et al. 2004). The values reported here are exactly three times the values reported in our previous publication, because of an error in assuming that the dry weight accumulation was for three cotyledons, instead of one cotyledon.

  • b

    Implies that ML and LL conditions are statistically different from each other, and

  • c

    implies that MH and HH are different from each other using the Student's t-test (P < 0.05).

  • H, high; M, medium; L, low.

Case 127 °C/20 °C (M)35 °C/27 °C (H)MH55.2 ± 3.4429.3 ± 0.54c5.5 ± 0.18c18.4 ± 0.71c
27 °C/20 °C (M)27 °C/20 °C (M)MMa41.3 ± 7.0531.7 ± 0.326.3 ± 0.1215.6 ± 0.59
27 °C/20 °C (M)20 °C/12 °C (L)ML17.9 ± 0.2230.7 ± 1.51b7.8 ± 0.24b13.6 ± 0.67b
Case 235 °C/27 °C (H)35 °C/27 °C (H)HH56.0 ± 4.1726.3 ± 1.34c5.2 ± 0.58c16.7 ± 2.93c
27 °C/20 °C (M)27 °C/20 °C (M)MMa41.3 ± 7.0531.7 ± 0.326.3 ± 0.1215.6 ± 0.59
20 °C/12 °C (L)20 °C/12 °C (L)LL18.9 ± 0.7928.2 ± 0.66b5.7 ± 0.27b11.6 ± 0.49b

Table 1 presents the biomass composition and dry weight accumulation data for both case 1 and case 2. The soybean cotyledons displayed a trend towards increased dry weight accumulation at higher temperatures. In case 1, the dry weight accumulation was greatest for the MH condition (55.2 mg/cotyledon) and lowest for the ML condition (17.9 mg/cotyledon). The relative protein content of the cotyledons was highest in the MM condition, whereas the relative lipid content was highest in the MH condition, and the relative starch content was highest in the ML condition. In case 2, the protein and starch percentage of the biomass were highest for the MM condition, whereas the lipid percentage was highest for the HH condition. The dry weight accumulation was maximum for the HH condition (56.0 mg/cotyledon) and lowest for the LL condition (18.9 mg/cotyledon). Interestingly, the ML condition accumulated less dry weight than the LL condition. The biomass accumulation was comparable between the MH and HH conditions. The protein and lipid as a percentage of biomass was greater for the MH than the HH condition. Further, the Student's t-test (P < 0.05) indicated that the MH and HH temperature conditions were statistically different from each other for relative protein, starch and lipid contents. This was verified for ML and LL temperature conditions as well. On the basis of this data, we observed that the growth temperature during cell formation had a dominant influence on the biomass accumulation and composition. Additionally, we also expected an increased uptake of carbon and an increased carbon flow through metabolic pathways at higher temperatures compared with lower temperatures.

13C MFA

The soybean cotyledons from both case 1 and case 2 were fractionated into protein, oil and starch, respectively. The protein and starch fractions of cotyledons were hydrolysed to prepare respective NMR samples. 2-D HSQC spectroscopy spectra (of the protein and starch samples) and 2-D TOCSY spectra (of the protein sample) were acquired. Cross-peaks on these spectra displayed peak splitting along the 13C dimension; the intensities of the multiplets so formed were proportional to isotopomer abundances of metabolites in the sample (Harris 1983; Sriram et al. 2004). Isotopomer abundances measured from these spectra are listed in Supplementary Table S2. Measurements of extracellular fluxes (carbon source intake, acid effluxes and biomass synthesis) are listed in Supplementary Table S3.

Our previous work (Sriram et al. 2004) demonstrated how 13C MFA can be used as a powerful tool for identifying active pathways and also for segregating parallel pathways in multiple compartments (Sriram et al. 2007b), and thereby iteratively constructing a metabolic network using the MM temperature condition as a case study. The metabolic network constructed for the central carbon metabolism in developing soybean cotyledons is shown in Fig. 1. The metabolic network consisted of three compartments; cytosol, plastid and mitochondrion, with parallel glycolysis and oxidative pentose phosphate pathway (oxPPP) in the cytosol (pglC) and plastid (pglP). The mitochondrion comprised of the tricarboxylic acid (TCA) cycle and glyoxylate shunt. The glutamine assimilation reactions are considered in the plastid, and the γ-aminobutyric acid (GABA) shunt in the cytosol and mitochondrion. Anaplerotic shunts, such as malic enzyme, are present in the plastid and mitochondrion, and the phosphoenolpyruvate (PEP) carboxylase in the cytosol, respectively. The biosynthetic pathways of protein, lipid and starch are considered in their respective compartments. We have not included the fixation of evolved CO2 by the ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) bypass flux (Schwender et al. 2004) as our isotopomer experimental design cannot identify (Wiechert et al. 2001) a possible flux through the Rubisco pathway. It is to be noted that for the sake of clarity, fructose-6-phosphate (F6PP) (and F6PC) and triose-3-phosphate (T3P)P (and T3PC) are shown separately in the glycolytic and oxPPP, although they represent the same pool in the respective compartments. Further, in the figures, the thickness of the arrows connecting the different metabolite pools is directly proportional to the molar fluxes. For example, for depicting the flux from F6P to T3P catalysed by phosphofructokinase (pfk), the size of the arrow doubles here as 1 mol of F6P gets converted to 2 mol of T3P.

image

Figure 1. Metabolic flux map representing central carbon metabolism in developing soybean cotyledons for MH temperature condition. The thickness of the arrows is proportional to the difference in the relative fluxes (standardized with respect to 100 mol of sucrose intake) between MH and MM condition. Red and green arrows indicate that the flux is greater or lower than the MM (optimum) temperature condition, respectively. A dashed dark grey line indicates that the flux is statistically not different from each other. Supplementary Table S5 lists all relative fluxes for the MH and MM temperature condition. Intracellular metabolites are shown in white ovals, and gray ovals represent sink metabolites (amino acids, nucleotides, etc.). Note: 1. For the sake of clarity, F6PP (and F6PC) and T3PP (and T3PC) are shown separately in the glycolytic and oxPPP, though they represent the same pool in the respective compartments. 2. The malT1 transporter channels malate from cytosol to mitochondrion in MM condition and from mitochondrion to cytosol in the case of MH condition. Further, even though total mass is conserved, moles are not conserved. For example, for depicting the flux from F6P to T3P catalysed by (pfk), the size of the arrow doubles here as 1 mol of F6P gets converted to 2 mol of T3P. H, high; M, medium; L, low; T3P, triose-3-phosphate; oxPPP, oxidative pentose phosphate pathway; malT1, malate transporter from mitochondrion to cytosol; pfk, phosphofructokinase.

Download figure to PowerPoint

The reactions included were determined from biochemical texts (Dey & Harborne 1997; Singh 1998), online databases (Soybase and KEGG encyclopedia) and literature (Chollet, Vidal & O'Leary 1996; Casati et al. 1999).

Metabolic trends observed in case 1 and case 2

13C MFA of the five temperature conditions was carried out using the tool NMR2Flux. The NMR multiplet intensities of experimental and simulated data for all the five temperature conditions compared well (except for the outlier Leuδ1 points; Supplementary Fig. S2). Further analysis of the intracellular fluxes and flux ratios are discussed later.

The metabolic flux map for the MH temperature condition (Fig. 1) depicts the metabolic trends discussed further. As observed in Sriram et al. (2004), parallel pathways of glycolysis and oxPPP were distinguished between the cytosol and the plastid. The T3P pools in the plastid and cytosol rapidly exchanged between the two compartments. The hexose isomerase flux was in the direction of F6P [RIGHTWARDS ARROW] glucose-6-phosphate (G6P), channelling carbon flow through the oxPPP in a cyclic manner. The flux through oxPPP was considerable in both the cytosol and plastid, but higher in the plastid (pglP). The flux through non-oxidative reactions of the oxPPP catalysed by transketolase (tkt) and transaldolase (tal) were greater (>4.0-fold) in the plastid compared with the cytosol. Flux from T3P to F6P was substantial in the plastid (where it is catalysed by fructose-1,6-bisphosphatase, f16bpP) but negligible in the cytosol (where it is catalysed by pyrophosphatase). Large enrichments of the γ2 carbon atom of Val and δ2 atom of Leu (Supplementary Table S2) indicated that the plastidic pyruvate (Pyr) pool was generated mainly from sucrose (Sriram et al. 2004). This result implies negligible fixation of CO2 from photosynthetically derived T3P or dilution by glutamine, which is supplied as a nitrogen source. Although the carboxylation by Rubisco bypass reaction (Schwender et al. 2004) cannot be identified directly by our current experimental design using uniformly labelled 13C sucrose, stoichiometric analysis indicates that flux through this bypass is negligible (Supplementary Table S8).

Table 2 summarizes the flux ratios of key reactions in both case 1 and case 2. Some common trends observed among all the temperature conditions are listed further. The contribution of the plastidic malic enzyme (meP) reaction to the Pyr pool was substantially less than the overall pyruvate kinase (pyktotal) flux in the glycolytic pathway. The flux through the cytosolic PEP carboxylase reaction (ppc) was also negligible compared with the net pyk flux, but considerable compared with the flux through pyruvate dehydrogenase (pdh) in the mitochondrion. The glyoxylate shunt, which is associated with lipid catabolism, was negligible in both cases as expected as the cotyledons predominantly accumulate lipids and proteins. There was a substantial flux through the GABA shunt. Thus, our metabolic flux map establishes compartmentation for primary metabolism, delineates extent of flux through parallel reactions in multiple compartments, and elucidates compartment interactions including the direction of transport of various metabolites. Together, this information reveals the quantitative nature of partitioning of carbon into multiple branches at various nodes of primary carbon metabolism.

Table 2.  Comparison of flux ratios on a carbon mol basis and contribution of pglP flux towards plastidic NADPH consumption for both case 1 and case 2
Temperature [RIGHTWARDS ARROW]Case 1Case 2
MHMMaMLHHMMaLL
Flux ratio [DOWNWARDS ARROW]RatioRatio
  • a

    The values for the MM temperature condition has been published previously (Sriram et al. 2004).

  • The numbers in bold indicate that they are not statistically different from the MM temperature condition. Superscripts C, P and M represent cytosol, plastid and mitochondrion, respectively.

  • H, high; M, medium; L, low; pgl, oxidative pentose phosphate pathway; pyk, pyruvate kinase; me, malic enzyme; pdh, pyruvate dehydrogenase; ppc, PEP carboxylase.

Glycolysistotal/sucrose intake0.52 ± 0.030.56 ± 0.040.55 ± 0.050.50 ± 0.030.56 ± 0.040.47 ± 0.04
 pglC/sucrose intake0.57 ± 0.220.48 ± 0.220.17 ± 0.170.37 ± 0.220.48 ± 0.220.34 ± 0.20
 pglP/sucrose intake0.59 ± 0.230.57 ± 0.260.53 ± 0.240.65 ± 0.290.57 ± 0.260.77 ± 0.24
Glycolysistotal/pgltotal0.46 ± 0.130.58 ± 0.220.90 ± 0.440.53 ± 0.220.58 ± 0.220.45 ± 0.16
 pyktotal/meP11.41 ± 2.1337.5 ± 16.47.82 ± 3.0315.89 ± 8.8137.5 ± 16.410.85 ± 2.03
 pyktotal/ppcC11.14 ± 0.798.83 ± 0.865.53 ± 1.1411.63 ± 1.408.83 ± 0.8612.05 ± 1.23
 pdhM/ppcC3.15 ± 0.282.60 ± 0.343.07 ± 0.501.91 ± 0.412.60 ± 0.345.24 ± 0.53
 pdhP/ppcC8.01 ± 0.915.52 ± 0.853.07 ± 0.849.01 ± 1.305.52 ± 0.856.88 ± 1.31
 pdhP/meP8.12 ± 1.1023.02 ± 9.654.23 ± 1.4512.04 ± 5.6823.02 ± 9.656.10 ± 0.94
 Respiration/sucrose intake0.46 ± 0.020.43 ± 0.020.48 ± 0.040.38 ± 0.020.43 ± 0.020.48 ± 0.03
pglP contribution to NADPH consumption1.050.981.61.050.981.07

Absolute and relative fluxes for case 1 and case 2

The absolute flux data (µmol day−1 cotyledon−1) for all the metabolic steps are listed in Supplementary Table S4 for case 1 and case 2. The temperature imposed during in vitro culture had the expected effect on cotyledon growth and metabolism as explained further. There was more sucrose consumed and more amino acids, lipids and starch synthesized in the HH, MH and MM conditions than in the ML and LL temperature conditions (Supplementary Table S4). Also, most absolute fluxes for the HH and MH conditions were greater than the MM temperature condition. Conversely, the absolute fluxes were lesser for the ML and LL conditions compared with the MM temperature condition. Thus, greater carbon uptake and carbon flow occurred through the central carbon metabolism with an increase in temperature. Supplementary Fig. S3 illustrates the absolute flux map for the MH compared with the MM temperature condition.

Although absolute fluxes indicate the most active pathways, relative fluxes provide better insight into the regulation of the reaction network because in comparison of relative fluxes, the total substrate entering into the system is constant across the different temperature conditions. To that end, the relative fluxes were standardized with respect to 100 mol of sucrose intake. All comparisons made henceforth are in terms of relative flux values and analysed with respect to the MM temperature condition. Supplementary Table S5 lists the relative fluxes of the metabolic network (standardized with respect to 100 mol of sucrose intake) for case 1 and case 2. Supplementary Fig. S4 represents the relative flux map for the HH temperature condition relative to the MM temperature condition.

Nodal analysis was carried out to analyse the partitioning of the carbon into multiple branches in both case 1 and case 2. The ‘flexibility’ or ‘rigidity’ of a node can be insightful in identifying the preferred direction of carbon flow and regulation of metabolism (Stephanopoulos & Vallino 1991). To elucidate, if the ratio of the carbon partitioning into multiple reactions at a node remains constant, irrespective of changes in the total incoming flux, then the node is said to be ‘rigid’. Hence, the genetic manipulation of the reactions involved in a flexible node is likely to be more productive in channelling the precursor metabolite in the preferred direction (Stephanopoulos & Vallino 1991). Conversely, for rigid nodes, it is more difficult to modify the intrinsic regulation of flux through the node, and hence would likely require multiple genetic manipulations of the enzymes involved. For example, the ratio of the pdhP/meP in the plastid was statistically different on a carbon (C) mol basis (Supplementary Table S6) for both case 1 (MH and ML) and case 2 (HH and LL) compared with the MM temperature condition (Table 2). This indicates that the pdhP/meP node is flexible; this metabolic node alters the direction of carbon flow to meet the metabolic demand of downstream reactions. Two other significant results were observed from the nodal analysis of the temperature conditions; the rigidity of the important pglP node with respect to the sucrose intake, and the flexibility of the anaplerotic reactions contributing to the Pyr node, which is the crossroad for partitioning of carbon into protein and lipid synthesis.

Ratio of the pglP flux to the total sucrose intake

The pglP reaction is a source of reductant (NADPH) for lipid and protein biosynthesis, glutamine assimilation, and countering oxidative stress within the plastid (Hauschild & von Schaewen 2003). Flux through pglP was substantial for both case 1 and case 2 cotyledons. The ratio of the pglP flux to the total sucrose intake rate on a C mol basis was not statistically different [approximately 56% (±24%)] for either MH or ML compared with the MM temperature condition (case 1) (Table 2). This finding indicates that the relative amount of carbon entering the pglP remained ‘rigid’ relative to sucrose intake regardless of temperature in vitro (Fig. 2a). In contrast, the pglP flux to the sucrose intake ratio varied among the LL, MM and HH temperature treatments tested in case 2 (Table 2). This latter result implies that carbon flow through the pglP reaction can be ‘flexible’ in response to temperature early (but not later) in the seed development. The relative flux maps for the pglP node for case 1 and case 2 are shown in Fig. 2.

image

Figure 2. Plastidic oxPPP node for the (a) MH temperature condition and (b) HH temperature condition. The thickness of the arrows is proportional to the relative fluxes (standardized with respect to 100 mol of sucrose intake). Red and green arrows indicate that the flux is greater or lower than the MM (optimum) temperature condition, respectively. Dashed brownish yellow arrows indicate that the flux is not statistically different from the MM temperature condition. Supplementary Table S5 lists the relative fluxes with SDs for the two cases. Table 2 lists the flux ratios of key nodes for both case 1 and case 2. H, high; M, medium; L, low; oxPPP, oxidative pentose phosphate pathway.

Download figure to PowerPoint

Carbon partitioning at plastidic Pyr node

An important node in the central carbon metabolism that signifies the focal point of carbon partitioning into protein and oil is the Pyr pool in the plastid. The inverse relationship between protein and oil is a fundamental limitation towards generating high-protein and high-oil-content soybeans (Cober & Voldeng 2000). Hence, we analysed the metabolic flux through the plastidic Pyr node to understand the regulation of carbon allocation towards these two primary biomass components in the seed (protein and oil).

The primary reactions that direct carbon flux into the plastidic Pyr pool are the pyruvate kinase (pyktotal) and the meP. The plastidic Pyr is then channelled into the biosynthesis of protein and oil. Figure 3a depicts the Pyr node for the ML temperature condition with key flux values. The pyktotal flux was much greater than the meP flux (Table 2). For case 1 (MH and ML), the pyktotal flux decreased and the meP flux increased relative to the MM condition. This result is shown graphically in Fig. 3a for the ML treatment and Fig. 3b for MH treatment. The flux values through specific enzymes relative to the MM are presented in Supplementary Table S7. The pyktotal fluxes for the MH and ML were 5–8% less than the MM temperature condition. The meP flux for the MH treatment was 177% of MM, and for the ML treatment it was 325% of the MM values (Supplementary Table S7). Thus, for case 1, the meP flux varied dramatically in response to temperature; there was much less variation in pyktotal flux.

image

Figure 3. Plastidic Pyr node for the (a) ML and (b) MH temperature condition. The thickness of the arrows is proportional to the relative fluxes (standardized with respect to 100 mol of sucrose intake). Red and green arrows indicate that the flux is greater or lower than the MM (optimum) temperature condition, respectively. Flux values of key metabolic reactions are indicated in the figure. Supplementary Table S5 lists all relative fluxes with SDs for the MH temperature condition, and Table 2 lists the flux ratios of key nodes for both case 1 and case 2. H, high; M, medium; L, low.

Download figure to PowerPoint

We also considered how temperature treatments affected flux through the ppc reaction in the cytosol and the malate (Mal) transporter from mitochondrion to cytosol (malT1). These two components are responsible for directing the carbon flow from the cytosol to the plastid (malT2) (Fig. 3a). The malT2 transporter replenishes oxaloacetate (OAA) and Pyr pool in the plastid by directing the carbon flow into the plastidic Mal dehydrogenase (mdhP) and meP reactions (Fig. 3a). Flux through cytosolic ppc increased by approximately 50% (Supplementary Table S7) under the ML condition relative to MM (Fig. 3a). Additionally, the OAA produced from the ppc reaction was converted to Mal. The Mal in the cytosol apparently met with two fates; some was transported into the mitochondrion to feed the TCA cycle, and the remaining was transported to the plastid (Fig. 3a). The MH temperature condition, however, decreased flux through cytosolic ppc reaction by about 25% (Supplementary Table S7) relative to MM (Fig. 3b), which effectively reversed the net flux through the malT1 transporter relative to the MM condition. In MH cotyledons, net Mal transport was from the mitochondrion to the cytosol, and then to the plastid (Fig. 3b).

In case 2, the pyktotal flux for HH and LL was approximately 11–12 times greater than the meP flux (Table 2). Supplementary Fig. S5 depicts the Pyr node for the HH temperature condition with key flux values. The pyktotal flux decreased whereas the meP increased for the HH and LL relative to the MM temperature condition. Similar to case 1, the percentage change in flux was greater for meP than for pyktotal (Supplementary Table S7). Furthermore, the direction of the Mal transporters was from the mitochondrion to cytosol and then to the plastidic Mal pool for both the HH and LL temperature conditions. Thus, the anaplerotic reactions and the Mal transporters contributing carbon to the plastidic Pyr node varied largely in response to the temperature treatments, with the meP and Mal transporters between mitochondrion, cytosol and plastid responding to the greatest extent to temperature.

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Applying 13C MFA to soybean cotyledons

The metabolic architecture responsible for the changes in physiology caused by temperature was examined in this study using carbon-labelling experiments and a recently developed generic 13C MFA tool NMR2Flux (Sriram et al. 2004). The assumption of steady-state conditions in the labelling studies of soybean embryos was verified. The cotyledon volume was approximately 0.3 mL, and the intracellular sucrose concentration in a soybean cotyledon was found from literature to be approximately 37 mm (Lichtner & Spanswick 1981). For the HH temperature condition, the sucrose uptake rate was 1.6 µmol h−1 cotyledon−1. The residence time of sucrose was calculated to be 7 h. Isotopic steady state is attained within five residence times, and thus was estimated to be 35 h, which was much less than the labelling time (6 d) in culture. Isotopic steady state also was verified for all the temperature treatments (data not shown). The metabolic fluxes presented in this paper are a time average over a period of 6 d.

We observed that simulated and experimental isotopomer abundances compared well except for Leuδ1 (Supplementary Fig. S2). According to well-established biochemical pathways for their synthesis, the multiplet intensities of Leuδ1, Valγ1 and Alaβ peaks should agree with each other, as they reflect the same carbon atom of the precursor metabolite Pyr (Szyperski 1998). However, a discrepancy between the Leuδ1 and Valγ1/Alaβ peaks at all temperature conditions was observed (Supplementary Fig. S6). The consistency between Valγ1 and Alaβ peaks indicates that Leuδ1 discrepancy would not affect the fluxes of the primary carbon metabolism. Rather, this isotopomeric discrepancy suggests that one or more of the intermediate reactions in the biosynthesis of Leu from 2-oxoisovalerate (Dey & Harborne 1997) might be reversible. A similar discrepancy between the Leuδ1 and Valγ1/Alaβ peaks has been detected in another biological system (Catharanthus roseus) being studied in our lab (Sriram, Fulton & Shanks 2007a).

There is negligible external CO2 assimilation indicated by the large enrichments of the γ2 carbon atom of Val and δ2 atom of Leu. The assimilation of internal CO2 by Rubisco bypass flux cannot be identified by our current experimental design using uniformly labelled 13C sucrose. However, we expect that the contribution of the Rubisco bypass flux is very small as our evaluated fluxes accounted for the experimental isotopomer abundances well. The negligible contribution of Rubisco bypass flux has been demonstrated in detail in Supplementary Table S8. This result is likely an average and could be higher during the day. Nonetheless, the best currently available technology for flux analysis cannot process a system where the fluxes are transient (Shastri & Morgan 2007), and hence, the time average is the best representation of the fluxes in our study.

Dominant effect of the early development temperature

Growth temperature is a key parameter influencing the biomass and composition of the soybean seed (Dornbos & Mullen 1992; Gibson & Mullen 1996; Kane et al. 1997). In case 1, the early development temperature in the growth chamber was maintained the same before the cotyledons were transferred to the in vitro culture at varying conditions (MH, MM and ML; Table 1). However, in case 2, the temperature in the growth chamber and in the in vitro culture was maintained the same but at varying levels (HH, MM, LL; Table 1). The effect of the temperature during in vitro culture on the biomass composition was more pronounced (Table 1) when a temperature change was imposed (case 1), compared with those grown at a constant temperature (case 2).

Although the biomass accumulation in culture is linear (data not shown), it is approximately 25% higher for the MM treatment than that found in vivo. The protein content for the MM treatment is about 10% lower in the in vitro culture as compared with the plant (data not shown). However, because it is not easy to measure fluxes in vivo, in vitro cultures are extremely important to elucidate quantitative differences between treatments and to generate working hypothesis. From our in vitro culture study, we observed that the ratio of pglP flux to total sucrose intake was rigid in case 1 but flexible in case 2 in response to temperature in vitro. Together, these results indicate that the growth temperature during the early development stage temperature has a dominant influence on establishing constraints for primary metabolism.

The pglP flux is an important source of reductant for lipid and protein biosynthesis, glutamine assimilation, and countering oxidative stress by maintaining the redox potential in the plastid (Hauschild & von Schaewen 2003). The NADPH requirement in the plastid was completely satisfied by pglP flux for all the temperature conditions (Supplementary Table S9 illustrates this for the MH temperature treatment). Additionally, the production of NADPH by pglP flux was 1.6 times the NADPH requirement for the plastidic biosynthetic and glutamine assimilation reactions for the LL temperature condition. The greater NADPH production was attributed to a 36% greater pglP flux in LL compared with the MM temperature condition. The increase in the pglP flux could result from the decreased glycolytic carbon flux (∼15%) observed in the LL condition relative to the MM temperature condition. The increase in pglP flux could also be a response to the cold stress to maintain the redox state of the plastid. This possibility is consistent with observed increases in the activity of glucose 6-phosphate dehydrogenase (an enzyme that exerts significant metabolic control in the oxPPP) in cold-treated rape plants (Maciejewska & Bogatek 2002).

Crossroad for protein and oil partitioning

The plastidic Pyr node in the metabolic network provides a focal point for carbon partitioning into protein and oil in soybeans. The overall glycolytic flux catalysed by pyruvate kinase (pyktotal) was substantially greater than the meP flux for both case 1 and case 2 (Table 2). Our results corroborate the observations in B. napus embryos, in which the label incorporated into fatty acids was primarily from glycolysis; negligible carbon was derived from Mal (Schwender et al. 2003). Our results clearly demonstrate that the activity of the cytosolic ppc and meP varies greatly depending on temperature. Further, the large variation in the fluxes around the plastidic Pyr node indicates that this metabolic node is ‘flexible’.

The importance of ppc activity in regulating carbon partitioning in soybean seeds has been previously emphasized (Smith, Rinne & Seif 1989). Moing, Maucourt & Renaud (2004), however, questioned whether ppc influenced the direction of carbon flux based on metabolite profiling of Arabidopsis transformants. Our direct measurements of carbon flux through the anaplerotic reactions (cytosolic ppc and meP) indicate they have a fundamental impact on carbon partitioning into protein and oil in soybeans. It is also to be noted that we have represented all comparisons in terms of relative fluxes (standardized with respect to 100 mol of sucrose intake per day). Thus, relative fluxes allow us to elucidate if the same amount of sucrose (carbon source) was taken up under different temperature treatments, what are the levels of activity of various enzymes through the metabolic network. Subsequently, the large variation in the fluxes through meP and ppc also indicates that these enzymes could be potential metabolic engineering targets. Overexpression of the meP and ppc could potentially help increase the carbon flow into the plastidic Pyr pool, thus enabling increased lipid and protein production.

CONCLUDING REMARKS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Flux analysis is a powerful technique for analyzing the interaction among multiple metabolic compartments inherent in plant systems. Our detailed in vitro culture study on the effects of temperature on the fluxes of central carbon metabolism in developing soybean cotyledons show that the early development temperature is dominant in regulating certain metabolic aspects of the system, which was reflected in the rigidity of the important pglP node. 13C MFA has also proven useful in identifying regulatory nodes for carbon partitioning within the system. Our results indicate that temperature has a large impact on the direction of flux through cytosolic ppc and meP. The large variation suggests that meP and subsequently ppc could be possible targets for genetic manipulation, in an effort towards increased protein and lipid production in soybeans. Finally, it would be advantageous to integrate flux analyses with other global analytical approaches, such as metabolomics, transcriptomics and proteomics, to better understand carbon partitioning into protein and lipid in soybean cotyledons.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

We would like to thank Dr. M.H. Spalding (Department of Genetics, Development and Cell Biology, Iowa State University) for useful discussions on this work. This work was funded by the Iowa Soybean Promotion Board, Plant Sciences Institute at Iowa State University, and an Endowment to the Department of Agronomy at Iowa State University.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information
  • Bils R.F. & Howell R.W. (1963) Biochemical and cytosolic changes in developing soybean cotyledons. Crop Science 3, 304308.
  • Casati P., Drincovich M.F., Edwards G.E. & Andreo C.S. (1999) Malate metabolism by NADP-malic enzyme in plant defense. Photosynthesis Research 61, 99105.
  • Chollet R., Vidal J. & O'Leary M.H. (1996) Phosphoenolpyruvate carboxylase: A ubiquitous, highly regulated enzyme in plants. Annual Review of Plant Physiology and Plant Molecular Biology 47, 273298.
  • Cober E.R. & Voldeng H.D. (2000) Developing high-protein, high-yield soybean population and lines. Crop Science 40, 3942.
  • Dey P.M. & Harborne J.B. (1997) Plant Biochemistry. Academic Press, San Diego, CA, USA.
  • Dieuaide-Noubhani M., Raffard G., Canioni P., Pradet A. & Raymond P. (1995) Quantification of compartmented metabolic fluxes in maize root tips using isotope distribution from 13C or 14C-labeled glucose. Journal of Biological Chemistry 270, 1314713159.
  • Dornbos D.L.J. & Mullen R.E. (1992) Soybean seed protein and oil contents and fatty acid composition adjustments by drought and temperature. Journal of American Oil Chemists' Society 69, 228231.
  • Gibson L.R. & Mullen R.E. (1996) Soybean seed composition under high day and night growth temperatures. Journal of American Oil Chemists' Society 76, 733737.
  • Girke T., Ozkan M., Carter D. & Raikhel N.V. (2003) Towards a modeling infrastructure for studying plant cells. Plant Physiology 132, 410414.
  • Glawischnig E., Gierl A., Tomas A., Bacher A. & Eisenreich W. (2001) Retrobiosynthetic nuclear magnetic resonance analysis of amino acid biosynthesis and intermediary metabolism. Metabolic flux in developing maize kernels. Plant Physiology 125, 11781186.
  • Glawischnig E., Gierl A., Tomas A., Bacher A. & Eisenreich W. (2002) Starch biosynthesis and intermediary metabolism in maize kernels. Quantitative analysis of metabolite flux by nuclear magnetic resonance. Plant Physiology 130, 17171727.
  • Harris R.K. (1983) Nuclear Magentic Resonance Spectroscopy: A Physiochemical View. Pitman Books, London.
  • Hauschild R. & Von Schaewen A. (2003) Differential regulation of glucose-6-phosphate dehydrogenase isoenzyme activities in potato. Plant Physiology 133, 4762.
  • Howell R.W. & Cartter J.L. (1953) Physiological factors affecting composition of soybeans. I. Correlation and temperature during certain portions of the pod filling stage with oil percentage in mature bean. Agronomy Journal 45, 526528.
  • Johnson B.A. & Blevins R.A. (1994) NMRView: a computer program for the visualization and analysis of NMR data. Journal of Biomolecular NMR 4, 603614.
  • Kane M., Steele C., Grabau L., Mackown C. & Hildebrand D. (1997) Early-maturing soybean cropping system: III. Protein and oil contents and oil composition. Agronomy Journal 89, 464469.
  • Kauffman K.J., Ogunnaike B.A. & Edwards J.S. (2006) Designing experiments that aid in the identification of regulatory networks. Briefings in Functional Genomics and Proteomics 4, 331342.
  • Kinney A.J. (1998) Plants as industrial chemical factories – new oils from genetically engineered soybeans. Fett/Lipid 100, 173176.
  • Lichtner F.T. & Spanswick R.M. (1981) Sucrose uptake by developing soybean cotyledons. Plant Physiology 68, 693698.
  • Maciejewska U. & Bogatek R. (2002) Glucose catabolism in leaves of cold-treated winter rape plants. Journal of Plant Physiology 159, 397402.
  • Minorsky P.V. (2003) Achieving the in silico plant. Systems biology and the future of plant biological research. Plant Physiology 132, 404409.
  • Moing A., Maucourt M., Renaud C., et al. (2004) Quantitative metabolic profiling by 1-dimensional 1H-NMR analyses: application to plant genetics and functional genomics. Functional Plant Biology 31, 889902.
  • Myers D.J. (1993) Industrial applications for soy protein and potential for increased utilization. Cereal Foods World 38, 355360.
  • Nakasathien S., Israel D.W., Wilson R.F. & Kwanyuen P. (2000) Regulation of seed protein concentration in soybean by supra-optimal nitrogen supply. Crop Science 40, 12771284.
  • Nielsen J. (1998) Metabolic engineering: techniques for analysis of targets for genetic manipulations. Biotechnology and Bioengineering 58, 125132.
  • Piper E.L. & Boote K.J. (1999) Temperature and cultivar effects on soybean seed oil and protein concentrations. Journal of American Oil Chemists' Society 76, 12331241.
  • Pipolo A.E., Siclair T.R. & Camara G. (2004) Effects of temperature on oil and protein concentration in soybean seeds cultured in vitro. Annals of Applied Biology 144, 7176.
  • Press W.H., Teukolsky S.A., Vellerling W.T. & Flannery B.P. (1992) Numerical recipes in C, 2nd edn. Cambridge University Press, Cambridge, UK.
  • Ratcliffe R.G. & Shachar-Hill Y. (2006) Measuring multiple fluxes through plant metabolic networks. The Plant Journal 45, 490511.
  • Rontein D., Dieuaide-Noubhani M., Dufourc E.J., Raymond P. & Rolin D. (2002) The metabolic architecture of plant cells. Journal of Biological Chemistry 277, 4394843960.
  • Schwender J., Ohlrogge J. & Shachar-Hill Y. (2003) A flux model of glycolysis and oxidative pentose phosphate pathway in developing Brassica napus embryos. Journal of Biological Chemistry 278, 2944229453.
  • Schwender J., Goffman F., Ohlrogge J. & Shachar-Hill Y. (2004) Rubisco without the Calvin cycle improves the carbon efficiency of developing green seeds. Nature 432, 779782.
  • Schwender J., Shachar-Hill Y. & Ohlrogge J. (2006) Mitochondriol metabolism in developing embryos of Brassica napus. Journal of Biological Chemistry 281, 3404034047.
  • Shastri A.A. & Morgan J.A. (2007) A transient isotopic labeling methodology for 13C metabolic flux analysis of photoautotrophic microorganisms. Phytochemistry 68, 23022312.
  • Singh B. (1998) Plant Amino Acids (Books in Soils, Plants, & the Environment). Marcel-Dekker, New York, NY, USA.
  • Smith A.J., Rinne R.W. & Seif R.D. (1989) Phosphoenolpyruvate carboxylase and pyruvate kinase involvement in protein and oil biosynthesis during soybean seed development. Crop Science 29, 349353.
  • Sriram G. & Shanks J.V. (2004) Improvements in metabolic flux analysis using carbon labeling experiments: bondomer balancing and boolean function mapping. Metabolic Engineering 6, 116132.
  • Sriram G., Fulton D.B., Iyer V.V., Peterson J.M., Zhou R., Westgate M.E., Spalding M.H. & Shanks J.V. (2004) Quantification of compartmented metabolic fluxes in developing soybean embryos by employing biosynthetically directed fractional 13C labeling, two-dimensional [13C, 1H] nuclear magnetic resonance, and comprehensive isotopomer balancing. Plant Physiology 136, 30433057.
  • Sriram G., Fulton D.B. & Shanks J.V. (2007a) Flux quantification in central carbon metabolism of Catharantus roseus hairy roots by 13C labeling and comprehensive bondomer balancing. Phytochemistry 68, 22432257.
  • Sriram G., Iyer V.V., Fulton D.B. & Shanks J.V. (2007b) Identification of hexose hydrolysis products in metabolic flux analytes: a case study of levulinic acid in plant protein hydrolysate. Metabolic Engineering 9, 442451.
  • Stephanopoulos G. (1999) Metabolic fluxes and metabolic engineering. Metabolic Engineering 1, 111.
  • Stephanopoulos G. & Vallino J.J. (1991) Network rigidity and metabolic engineering in metabolite overproduction. Science 252, 16751681.
  • Sweetlove L.J. & Fernie A.R. (2005) Regulation of metabolic networks: understanding metabolic complexity in the systems biology era. New Phytologist 168, 924.
  • Sweetlove L.J., Last R.L. & Fernie A.R. (2003) Predictive metabolic engineering: a goal for systems biology. Plant Physiology 132, 420425.
  • Szyperski T. (1998) 13C-NMR, MS and metabolic flux balancing in biotechnology research. Quarterly Reviews of Biophysics 31, 41106.
  • Westgate M.E., Orf J.H., Schussler J.R. & Shumway C. (eds) (1995) Temperature Regulation of Uptake and Metabolism of Protein and Oil Precursors by Developing Soybean Embryos. American Society of Agronomy, Madison, WI, USA.
  • Westgate M.E., Piper E.L., Batchelor W.D. & Hurburgh C. (2000) Effects of cultural and environmental conditions during soybean growth on nutritive value of soy products. In Soy in Animal Nutrition (ed. J.K.Drackley), pp. 7589. Federation of Animal Science Societies, Savoy, IL, USA.
  • Wiechert W. & Wurzel M. (2001) Metabolic isotopomer labeling systems. Part I: global dynamic behavior. Mathematical Biosciences 169, 173205.
  • Wiechert W., Mollney M., Petersen S. & De Graaf A.A. (2001) A universal framework for 13C metabolic flux analysis. Metabolic Engineering 3, 265283.
  • Willker W., Flogel U. & Leibfritz D. (1997) Ultra-high-resolved HSQC spectra of multiple-13C-labeled biofluids. Journal of Magnetic Resonance 125, 216219.
  • Van Winden W., Schipper D., Verheijen P. & Heijnen J. (2001) Innovations in generation and analysis of 2D [13C, 1H] COSY NMR spectra for metabolic flux analysis purposes. Metabolic Engineering 3, 322343.

Supporting Information

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUDING REMARKS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Figure S1. Comparison of amino acid profile for case 1 and case 2.

Figure S2. Comparison of simulated and experimental NMR isotopomer abundances.

Figure S3. Absolute flux map of MH temperature condition.

Figure S4. Relative flux map of HH temperature condition.

Figure S5. Pyr node for HH temperature condition.

Figure S6. Comparison of multiplet intensities of Leu δ1, Val γ1 and Ala β.

Table S1. Calculation of sucrose and glutamine consumption.

Table S2. Relative multiplet intensities from protein and starch hydrolysate.

Table S3. Extracellular and biomass synthesis flux data.

Table S4. Absolute flux data.

Table S5. Relative flux data.

Table S6. Carbon mol basis calculations.

Table S7. Percentage change for various reactions contributing towards Pyr node for case 1 and case 2.

Table S8. Analysis of Rubisco bypass in soybean embryos.

Table S9. NADPH balance calculations for MH temperature condition.

Table S10. Demonstration of flux balance at multiple metabolite pools in soybean central carbon metabolism.

FilenameFormatSizeDescription
PCE_1781_sm_Figs_S1-S6_Tables_S1-S10.pdf944KSupporting info item
PCE_figs_s1-s6_tables_s1-s10.pdf944KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.