Introduction
Developing soybean [Glycine max (L.) Merrill] seeds accumulate sucrose, raffinose and stachyose as well as lesser amounts of galactopinitol A, galactopinitol B and fagopyritol B1 in axis and cotyledon tissues during seed maturation (Hsu et al., Reference Hsu, Hadley and Hymowitz1973; Schweizer and Horman, Reference Schweizer and Horman1981; Obendorf et al., Reference Obendorf, Horbowicz, Dickerman, Brenac and Smith1998, Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009, Reference Obendorf, Horbowicz, Lahuta and Preedy2012; Obendorf and Kosina, Reference Obendorf, Kosina and Ng2011). Raffinose series oligosaccharides (RFOs) including raffinose, stachyose and verbascose are α-galactoside derivatives of sucrose. Galactosyl cyclitols, including galactinol, galactopinitol A and galactopinitol B, and fagopyritol B1, are α-galactoside derivatives of the cyclitols myo-inositol, d-pinitol (1d-3-O-methyl-chiro-inositol) and d-chiro-inositol, respectively. The free cyclitols are synthesized in maternal tissues of soybean, transported to the seed coat where they are unloaded into free space surrounding the embryo, and loaded into embryo tissues where they are stored primarily as their respective galactosides (Gomes et al., Reference Gomes, Obendorf and Horbowicz2005; Obendorf et al., Reference Obendorf, Sensenig, Wu, Ohashi, O'Sullivan, Kosina and Schnebly2008a, Reference Obendorf, Zimmerman, Ortiz, Taylor and Schneblyb, Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009; Kosina et al., Reference Kosina, Castillo, Schnebly and Obendorf2009, Reference Kosina, Schnebly and Obendorf2010; Obendorf and Kosina, Reference Obendorf, Kosina and Ng2011; Obendorf and Górecki, Reference Obendorf and Górecki2012) in mature seeds. Soybean galactinol synthase (GmGolS; EC 2.4.1.123) forms galactinol from myo-inositol and UDP-galactose and also forms fagopyritol B1 from d-chiro-inositol and udp-galactose (Frydman and Neufeld, Reference Frydman and Neufeld1963; Obendorf et al., Reference Obendorf, Odorcic, Ueda, Coseo and Vassallo2004). Stachyose synthase (STS; EC 2.4.1.67) forms galactopinitols from d-pinitol and galactinol (Hoch et al., Reference Hoch, Peterbauer and Richter1999; Peterbauer and Richter, Reference Peterbauer and Richter2001). Raffinose, stachyose and sucrose serve as soybean seed reserves contributing to about 15% of the soybean seed dry mass (Hsu et al., Reference Hsu, Hadley and Hymowitz1973). Raffinose and stachyose are one factor that has been proposed to contribute to seed desiccation tolerance and tolerance to heat or cold stresses (Koster and Leopold, Reference Koster and Leopold1988; Blackman et al., Reference Blackman, Obendorf and Leopold1992; Horbowicz and Obendorf, Reference Horbowicz and Obendorf1994; Obendorf, Reference Obendorf1997; Hoekstra et al., Reference Hoekstra, Golovina and Buitink2001; Farrant and Moore, Reference Farrant and Moore2011). Stachyose accumulates concomitantly with desiccation tolerance during seed maturation and drying, including precocious maturation (Blackman et al., Reference Blackman, Obendorf and Leopold1992; Obendorf et al., Reference Obendorf, Horbowicz, Dickerman, Brenac and Smith1998, Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009), but stachyose is not essential for germination (Blackman et al., Reference Blackman, Obendorf and Leopold1992; Dierking and Bilyeu, Reference Dierking and Bilyeu2009). Galactosyl cyclitols have been proposed to function in the same role as raffinose and stachyose in seeds that normally do not accumulate raffinose and stachyose (Horbowicz and Obendorf, Reference Horbowicz and Obendorf1994; Horbowicz et al., Reference Horbowicz, Brenac and Obendorf1998). For this reason, the presence of raffinose and stachyose, or alternatively galactosyl cyclitols, is thought to be of central importance for seed quality, field emergence and agronomic crop yield.
The utilization of soybean as an excellent source of protein and fibre in feed applications is limited by its chemical composition. While raffinose and stachyose may be important for seed viability, these compounds are not digested by non-ruminant animals and are fermented by microflora in the lower gut, causing acidification and production of gas (Gitzelmann and Auricchio, Reference Gitzelmann and Auricchio1965; Ruttloff et al., Reference Ruttloff, Täufel, Krause, Haenel and Täufel1967; Price et al., Reference Price, Lewis, Wyatt and Fenwick1988; Naczk et al., Reference Naczk, Amarowicz and Shahidi1997). Low-raffinose, low-stachyose soybean is desired as a metabolically efficient feed for chickens and pigs (Parsons et al., Reference Parsons, Zhang and Araba2000; Sebastian et al., Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000) and for reduced flatulence in humans (Suarez et al., Reference Suarez, Springfield, Furne, Lohrmann, Kerr and Levitt1999).
Soybean plants homozygous for the stc1 mutant phenotype have low raffinose synthase (RFS; EC 2.4.1.82) activity (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Dierking and Bilyeu, Reference Dierking and Bilyeu2008) but normal galactinol synthase (GolS; EC 2.4.1.123) and stachyose synthase (STS; EC 2.4.1.67) activities (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002) in seeds and produce seeds with low raffinose and low stachyose (Sebastian et al., Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000; Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Dierking and Bilyeu, Reference Dierking and Bilyeu2008). Plants homozygous for the mips mutant phenotype have reduced myo-inositol phosphate synthase (MIPS; EC 5.5.1.4) activity and produce seeds with low raffinose, low stachyose and low phytin (Sebastian et al., Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000; Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002). Low-raffinose, low-stachyose seeds from plants homozygous for stc1 had similar rates of field emergence as seeds with normal levels of raffinose and stachyose from normal (Stc1) plants (Neus et al., Reference Neus, Fehr and Schnebly2005), and seeds expressing the mutant stc1 phenotype were tolerant to imbibitional chilling (Obendorf et al., Reference Obendorf, Zimmerman, Ortiz, Taylor and Schnebly2008b). By contrast, seeds from plants homozygous for the mutant mips allele had reduced field emergence when compared to seed from normal (Mips) plants (Meis et al., Reference Meis, Fehr and Schnebly2003), and seeds expressing the mutant mips phenotype were sensitive to imbibitional chilling (Obendorf et al., Reference Obendorf, Zimmerman, Ortiz, Taylor and Schnebly2008b).
Sucrose and myo-inositol concentrations affected RFO accumulation in pea (Pisum sativum L.) embryos (Karner et al., Reference Karner, Peterbauer, Raboy, Jones, Hedley and Richter2004) and soybean embryos expressing the mutant mips phenotype (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002). Similarly, feeding d-chiro-inositol or d-pinitol to isolated immature soybean embryos increased the accumulation of fagopyritols and galactopinitols, respectively, during embryo maturation (Obendorf et al., Reference Obendorf, Odorcic, Ueda, Coseo and Vassallo2004). The objective of this research was to determine if feeding myo-inositol, d-chiro-inositol or d-pinitol to isolated immature soybean embryos without maternal tissues results in increased accumulation of galactinol, fagopyritols or galactopinitols in maturing embryos expressing the mutant stc1 phenotype or expressing the mutant mips phenotype in comparison to embryos from seeds expressing the Stc1 and Mips phenotype with normal raffinose, stachyose and phytin.
Materials and methods
Plant material
Seeds for each of four proprietary soybean [Glycine max (L.) Merrill] lines with low raffinose and stachyose (LRS) seeds expressing the mutant stc1 phenotype; low raffinose, stachyose and phytin (LRSP1, LRSP2) seeds expressing the mutant mips phenotype; and normal raffinose, stachyose and phytin (CHECK) seeds expressing the Stc1 and Mips phenotype were provided by Steve Schnebly, Pioneer Hi-Bred. All were advanced breeding lines in related, but not isogenic, Group II maturity agronomic backgrounds developed by traditional breeding. The stc1 and mips alleles in the breeding lines utilized in this study were described by Sebastian et al. (Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000), Hitz et al. (Reference Hitz, Carlson, Kerr and Sebastian2002) and Meis et al. (Reference Meis, Fehr and Schnebly2003). Plants of each line were grown in a greenhouse at 22°C nights (10 h) and 27°C days (14 h) with natural light supplemented 14 h with 640 μmol m− 2s− 1 (photosynthetically active radiation, PAR) incandescent light from Sylvania 1000-watt metal halide lamps at Ithaca, New York, USA, 42° north latitude.
Standards and reagents
Fructose, glucose, maltose, sucrose, raffinose, stachyose, myo-inositol, galactinol, phenyl α-d-glucoside, trimethylsilylimidazole and pyridine were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). d-Pinitol, d-ononitol (1d-4-O-methyl-myo-inositol), and d-chiro-inositol were purchased from Industrial Research Limited (Lower Hutt, New Zealand). Fagopyritols, digalactosyl myo-inositol (DGMI) and trigalactosyl myo-inositol (TGMI) were extracted from buckwheat (Fagopyrum esculentum Moench) seeds and purified by carbon (Mallinckrodt Baker Inc., Phillipsburg, New Jersey, USA)–Celite (Supelco, Bellefonte, Pennsylvania, USA) column chromatography (Whistler and Durso, Reference Whistler and Durso1950) as described by Obendorf et al. (Reference Obendorf, Steadman, Fuller, Horbowicz and Lewis2000). Galactopinitols were extracted from seeds of hairy vetch (Vicia villosa L.) or chickpea (Cicer arietinum L.) and purified following the same procedures.
Embryo feeding
Embryos from each of four soybean lines were isolated from immature seeds (mid pod fill; 30–35 d after flowering) by surgical removal of seed coat and nucellus–endosperm tissues. Initial fresh weight averaged 237 mg per embryo (N= 192). Three isolated embryos were placed in 20-ml bottles containing 3 ml solutions for each of four feeding treatments: (1) 100 mM d-chiro-inositol plus 100 mM sucrose; (2) 100 mM myo-inositol plus 100 mM sucrose; (3) 100 mM d-pinitol plus 100 mM sucrose; or (4) 100 mM sucrose (a control solution without cyclitols) for 24 h at 25°C under light (300 μmol m− 2s− 1 PAR, fluorescent) on oscillating shakers following the procedures described by Obendorf et al. (Reference Obendorf, Odorcic, Ueda, Coseo and Vassallo2004). After 24 h incubation, embryos were blotted to remove adhering solution and placed in small plastic Petri dishes. Embryos were precociously matured by slow drying for 14 d at 22°C by daily transfer to successive lower relative humidities (RH) controlled by saturated salt solutions (Blackman et al., Reference Blackman, Wettlaufer, Obendorf and Leopold1991, Reference Blackman, Obendorf and Leopold1992): day 1, 92% RH (KNO3); days 2 and 3, 87% RH (Na2CO3); day 4, 75% RH (NaCl); day 5, 51% RH [Mg(NO3)2]; day 6, 45% RH (K2CO3); day 7, 32% RH (MgCl2); day 8, 12% RH (LiCl); and remained at 12% RH days 8–14. Precociously matured dry embryos (6% moisture) were separated into axis and cotyledons, weighed and stored at − 80°C. For determination of soluble carbohydrates before incubation, immature embryos that were not placed in solutions were separated into axis and cotyledons, weighed and stored at − 80°C. The experiment was replicated four times with three embryos per replication (N= 12) for each of four lines and four treatments.
Soluble carbohydrate extraction and analysis
Cotyledons and axes were frozen in liquid nitrogen and ground to a fine powder with a cold mortar and pestle chilled with liquid nitrogen. Soluble carbohydrates in the frozen powder of cotyledon tissues were extracted in 1800 μl ethanol:water (1:1, v/v) containing 300 μg of phenyl α-d-glucoside (internal standard) in a ground glass homogenizer. The frozen powder of axis tissues was extracted in 800 μl ethanol:water (1:1, v/v) containing 100 μg of phenyl α-d-glucoside (internal standard). Extracts were centrifuged at 14,000 × g for 20 min. Supernatants (500 μl) were passed through 10,000 molecular weight (MW) cutoff filters (NANOSEP 10K Omega, Pall Corp., East Hills, New York, USA) by centrifugation (14,000 × g). Filtrates (200 μl for cotyledons, 400 μl for axes) were dried under a stream of nitrogen gas and stored overnight above P2O5 in desiccators to remove traces of water. Dry residues were derivatized with 100 μl of 1-(trimethylsilyl)-imidazole (TMSI):pyridine (1:1, v/v) for 45 min at 80°C. After cooling to room temperature, soluble carbohydrates were analysed by high-resolution gas chromatography (Horbowicz and Obendorf, Reference Horbowicz and Obendorf1994). Results are expressed as mg per g dry weight of cotyledon or axis tissues. Values below the level of detection are presented as zero. Significant differences (P< 0.05) after a Tukey correction for multiple comparisons were determined between lines, feeding treatments, and line by feeding treatment interactions using JMP Statistical Discovery Software (SAS Institute Inc., Cary, North Carolina, USA). Statistical analyses were performed using a square root transformation of the response to correct for non-constant residual variance.
Gas chromatography
Trimethylsilyl-derivatives of soluble carbohydrates were analysed on a Hewlett-Packard 6890 Gas Chromatograph (Agilent Technologies, Palo Alto, California, USA) equipped with a flame ionization detector, split-mode injector (1:50) and a HP-1MS capillary column (15 m length, 0.25 mm internal diameter, 0.25 μm film thickness). Oven temperature was programmed to an initial temperature of 150°C, adjusted to 200°C at 3°C min− 1, adjusted to 325°C at 7°C min− 1, and held at 325°C for 20 min. The injector port was operated at 335°C and the detector at 350°C. The carrier gas was nitrogen at 2.5 ml min− 1. Characterization of seed soluble carbohydrates by gas chromatography has recently been reviewed (Obendorf et al., Reference Obendorf, Horbowicz, Lahuta and Preedy2012).
Results
Cotyledons and axis prior to substrate feeding
Cotyledons of immature soybean embryos, before incubation in cyclitol solutions, contained sucrose, free cyclitols and reducing sugars; α-galactosides (galactosyl cyclitols or raffinose-family oligosaccharides) were detected in minute amounts (Table 1). Cotyledons of LRS, LRSP1 and LRSP2 initial embryos contained significantly more d-pinitol than the CHECK cotyledons (Table 1). LRSP1 cotyledons contained significantly less myo-inositol, galactinol and DGMI than CHECK cotyledons or LRS cotyledons, while LRSP2 embryos contained significantly less myo-inositol (Table 1). Cotyledons of LRSP1 initial embryos had less d-chiro-inositol than cotyledons of CHECK, LRS or LRSP2 (Table 1). Axes of immature soybean embryos, before incubation in cyclitol solutions, contained sucrose, free cyclitols and reducing sugars, but only minute amounts of RFOs and total α-galactosides (raffinose and galactinol) (Table 1).
Table 1 Soluble carbohydrates [mg (g dry weight)−1] in cotyledon or axis tissues of CHECK, LRS, LRSP1 and LRSP2 soybean embryos before feeding
d-chiro-inos, d-chiro-inositol; α-gal, α-galactoside; sol carb, soluble carbohydrate; suc, sucrose.
† For comparisons between columns within a row for cotyledons or axes, means not connected by the same letter are significantly different (P< 0.05; N= 12) after a Tukey correction for multiple comparisons. Statistical analyses were performed using a square root transformation of the response.
‡ DGMI, digalactosyl myo-inositol; TGMI, trigalactosyl myo-inositol; TGPA, trigalactosyl pinitol A; RFOs, raffinose family oligosaccharides. Total d-chiro-inositol, total myo-inositol and total d-pinitol are calculated from each free cyclitol plus their respective galactosyl cyclitols.
Substrate feeding, cotyledons
RFOs and galactosyl cyclitols accumulated during precocious maturation after incubation in the control (sucrose without cyclitols) and cyclitol (sucrose plus cyclitol) feeding solutions. Feeding the control solution resulted in large accumulations of stachyose, total RFOs, and total α-galactosides in cotyledons of CHECK embryos, but significantly lower amounts in cotyledons of LRS, LRSP1 and LRSP2 embryos (Table 2). Cotyledons of the LRS line had 90% less RFO than the CHECK line, but accumulated significantly higher concentrations of total myo-inositol, galactinol, DGMI and ciceritol than CHECK cotyledons after feeding the control solution. LRS cotyledons accumulated significantly higher total myo-inositol, galactinol and DGMI concentrations but significantly lower raffinose and stachyose concentrations after all feeding treatments, consistent with low raffinose synthase activity (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Dierking and Bilyeu, Reference Dierking and Bilyeu2008) in LRS embryos (Table 2, Fig. 1). LRSP1 and LRSP2 cotyledons had 85% and 75% less total RFOs than CHECK cotyledons after feeding a control solution without cyclitols (Table 2). LRSP1 and LRSP2 cotyledons accumulated significantly more sucrose and d-pinitol but significantly less total myo-inositol, galactinol, DGMI, and galactopinitol A than CHECK cotyledons after feeding the control solution (Table 2).
Table 2 Soluble carbohydrates in cotyledon tissues [mg (g dry weight)–1] after precocious maturation following feeding free cyclitols to isolated embryos of CHECK, LRS, LRSP1 and LRSP2 soybean
d-chiro-inos, d-chiro-inositol; α-gal, α-galactoside; sol carb, soluble carbohydrate; suc, sucrose.
† For comparisons between columns within a row, means not connected by the same letter are significantly different (P < 0.05; N= 12) after a Tukey correction for multiple comparisons.
‡ DGMI, digalactosyl myo-inositol; TGMI, trigalactosyl myo-inositol; TGPA, trigalactosyl pinitol A; RFOs, raffinose family oligosaccharides. Total d-chiro-inositol, total myo-inositol and total d-pinitol are calculated from each free cyclitol plus their respective galactosyl cyclitols.
Figure 1 Proposed pathways for synthesis of cyclitols, cyclitol galactosides and raffinose family oligosaccharides (modified from Ma et al., Reference Ma, Horbowicz and Obendorf2005; Obendorf and Górecki, Reference Obendorf and Górecki2012). Parentheses (unknown) by an arrow indicate that an enzyme catalysing the reaction has not been identified. Some reactions may be reversible. Ciceritol, digalactopinitol A; DGMI, digalactosyl myo-inositol; gol, galactinol; GolS, galactinol synthase (EC 2.4.1.123); GPA, galactopinitol A; GPB, galactopinitol B; IMP, myo-inositol-phosphate monophosphatase (EC 3.1.3.25); IMT, myo-inositol 4-O-methyltransferase (EC 2.1.1.129); IPK, inositol polyphosphate kinases; MIPS, myo-inositol-phosphate synthase (EC 5.5.1.4); myo, myo-inositol; RFS, raffinose synthase (EC 2.4.1.82); STS, stachyose synthase (EC 2.4.1.67); (STS?), indicates STS is proposed by extrapolation but has not been demonstrated experimentally; TGMI, trigalactosyl myo-inositol; TGPA, trigalactopinitol A; UDP, uridine diphosphate; UDP-gal, uridine diphosphate galactoside; VBS, verbascose synthase. For chemical structures, see Obendorf (Reference Obendorf1997) and Obendorf and Górecki (Reference Obendorf and Górecki2012).
As expected (Ma et al., Reference Ma, Horbowicz and Obendorf2005), feeding d-chiro-inositol significantly increased the total d-chiro-inositol (free d-chiro-inositol plus that in fagopyritols) in cotyledons of CHECK, LRS, LRSP1 and LRSP2 embryos compared to feeding a control solution without cyclitols (Table 2). Feeding d-chiro-inositol to embryos increased fagopyritol B1 six- to eightfold in cotyledons of all lines (Fig. 2) compared to feeding a control solution, but did not significantly alter the concentration of RFOs (Table 2). Feeding d-chiro-inositol to LRS embryos increased fagopyritol B1 sixfold and fagopyritol B2 fivefold in cotyledons compared to feeding a control solution (Fig. 2). There were no significant line by feeding treatment interactions for total fagopyritols (Fig. 2). When the data were pooled across lines, feeding d-chiro-inositol to embryos resulted in significantly higher concentrations of fagopyritol B1 and total fagopyritols in cotyledons compared to feeding myo-inositol, d-pinitol or a control solution without cyclitols (Fig. 2). Raffinose and stachyose concentrations were significantly lower, but galactinol and DGMI concentrations were significantly higher in LRS cotyledons than in other lines after all feeding treatments (Table 2). Feeding d-chiro-inositol to LRSP1 and LRSP2 embryos increased free d-chiro-inositol and fagopyritol B1 eightfold, fagopyritol B2 sixfold, and total α-galactosides twofold in cotyledons compared to feeding a control solution (Table 2).
Figure 2 Fagopyritol B1, fagopyritol B2, fagopyritol B3 and total fagopyritols (fagopyritol B1+ fagopyritol B2+ fagopyritol B3) in cotyledon tissues of isolated embryos as a function of line (CHECK, LRS, LRSP1 and LRSP2) by feeding treatment (100 mM d-chiro-inositol plus 100 mM sucrose; 100 mM myo-inositol plus 100 mM sucrose; 100 mM d-pinitol plus 100 mM sucrose; or a control solution of 100 mM sucrose without cyclitols) followed by precocious maturation of embryos by slow drying. Bars not connected by the same letter are significantly different (P < 0.05, N= 12) after a Tukey correction for multiple comparisons. Use lower-case letters to compare fagopyritol B1, fagopyritol B2 or fagopyritol B3 across treatments. Use upper-case letters to compare total fagopyritols across treatments.
Feeding d-pinitol to embryos significantly increased free d-pinitol and total d-pinitol concentrations in cotyledons of all lines. Feeding d-pinitol to CHECK embryos increased free d-pinitol ninefold and galactopinitol A twofold in cotyledons compared to feeding a control solution (Table 2). Feeding d-pinitol to LRS embryos increased free d-pinitol tenfold and more than doubled galactopinitol A and galactopinitol B in cotyledons compared to feeding a control solution. Feeding d-pinitol to LRSP1 and LRSP2 embryos increased accumulation of free d-pinitol three- to fourfold and also increased total d-pinitol in cotyledons, but concentrations of galactopinitols were not increased (Table 2).
Feeding myo-inositol to CHECK embryos increased free myo-inositol threefold and free d-chiro-inositol 11-fold, but total myo-inositol and total d-chiro-inositol were not significantly different. Feeding myo-inositol caused significantly greater accumulation of galactinol in LRS cotyledons than after feeding d-chiro-inositol or d-pinitol (Table 2). Feeding myo-inositol to LRSP1 embryos increased stachyose eightfold, free myo-inositol 18-fold, and total RFOs fourfold in LRSP1 cotyledons compared to feeding a control solution (Table 2). Feeding myo-inositol to LRSP2 embryos increased stachyose, total RFOs and total α-galactosides more than twofold.
Substrate feeding, axes
During precocious maturation after feeding a control solution (sucrose without cyclitols), CHECK axes accumulated high concentrations of stachyose and total RFOs and detectable but lower concentrations of galactopinitols and fagopyritols compared to stachyose (Table 3). Axes of the LRS line had 75% less RFO but substantially more galactinol and DGMI than axes of the CHECK line after all feeding treatments (Table 3). LRS axes accumulated significantly more di- and trigalactosides of myo-inositol and d-pinitol than CHECK line axis tissues after feeding a control solution without cyclitols. Axes of the LRSP1 line had 80% less RFO than the CHECK line after feeding a control solution (Table 3). Axes of the LRSP2 line had RFO values between those of LRSP1 axes and CHECK axes after feeding a control solution (Table 3). LRSP1 and LRSP2 axes had significantly less galactinol than axes of the LRS line after all feeding treatments. Galactinol concentrations were similar in axes of CHECK, LRSP1 and LRSP2.
Table 3 Soluble carbohydrates in axis tissues [mg (g dry weight)–1] after precocious maturation following feeding free cyclitols to isolated embryos of CHECK, LRS, LRSP1 and LRSP2 soybean
d-chiro-inos, d-chiro-inositol; α-gal, α-galactoside; sol carb, soluble carbohydrate; suc, sucrose.
† For comparisons between columns within a row, means not connected by the same letter are significantly different (P < 0.05; N= 12) after a Tukey correction for multiple comparisons.
‡ DGMI, digalactosyl myo-inositol; TGMI, trigalactosyl myo-inositol; TGPA, trigalactosyl pinitol A; RFOs, raffinose family oligosaccharides. Total d-chiro-inositol, total myo-inositol and total d-pinitol are calculated from each free cyclitol plus their respective galactosyl cyclitols.
Feeding d-chiro-inositol to CHECK embryos increased fagopyritol B1 sixfold in CHECK axes compared to feeding a control solution without d-chiro-inositol (Table 3, Fig. 3). Feeding d-chiro-inositol to embryos increased fagopyritol B1 in axes of all lines compared to the other feeding treatments (Fig. 3). Feeding d-chiro-inositol to LRS embryos increased fagopyritol B1 fourfold in axes of precociously matured embryos compared to feeding a control solution. Feeding d-chiro-inositol to LRSP1 embryos increased fagopyritol B1 and fagopyritol B2 nine- to tenfold compared to feeding a control solution (Table 3). Feeding d-chiro-inositol to LRSP2 embryos increased fagopyritol B1 sixfold and fagopyritol B2 fivefold in precociously matured axes of LRSP2 compared to feeding a control solution (Table 3).
Figure 3 Fagopyritol B1, fagopyritol B2, fagopyritol B3 and total fagopyritols (fagopyritol B1+ fagopyritol B2+ fagopyritol B3) in axis tissues of isolated embryos as a function of line (CHECK, LRS, LRSP1 and LRSP2) by feeding treatment (100 mM d-chiro-inositol plus 100 mM sucrose; 100 mM myo-inositol plus 100 mM sucrose; 100 mM d-pinitol plus 100 mM sucrose; or a control solution of 100 mM sucrose without cyclitols) followed by precocious maturation of embryos by slow drying. Bars not connected by the same letter are significantly different (P < 0.05, N= 12) after a Tukey correction for multiple comparisons. Use lower-case letters to compare fagopyritol B1, fagopyritol B2 or fagopyritol B3 across treatments. Use upper-case letters to compare total fagopyritols across treatments.
Feeding d-pinitol to CHECK embryos increased free d-pinitol sixfold, galactopinitol A threefold and galactopinitol B threefold in CHECK axes compared to feeding a control solution without d-pinitol (Table 3). RFOs and galactinol series oligosaccharides were not significantly increased in axis tissues after feeding the free cyclitols d-pinitol or myo-inositol (Table 3). Feeding d-pinitol had no significant effect on the accumulation of soluble carbohydrates in LRSP2 axes compared to feeding a control solution, and feeding d-pinitol to LRSP1 embryos significantly increased only free d-pinitol in LRSP1 axes compared to feeding a control solution (Table 3).
Feeding myo-inositol to CHECK, LRS, LRSP1 and LRSP2 embryos resulted in no significant changes in soluble carbohydrates in precociously matured axes compared to feeding a control solution (Table 3).
Discussion
LRS soybean embryos expressing the mutant stc1 phenotype have low raffinose synthase (RFS) activity (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Dierking and Bilyeu, Reference Dierking and Bilyeu2008), but normal myo-inositol phosphate synthase (MIPS) activity and normal stachyose synthase activity (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002), resulting in reduced raffinose and stachyose accumulation but normal to elevated concentrations of galactopinitols and fagopyritols during precocious maturation (Sebastian et al., Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000; Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Dierking and Bilyeu, Reference Dierking and Bilyeu2008; Obendorf et al., Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009) (Fig. 1) compared to CHECK embryos after feeding a control solution without cyclitols. LRS embryos also have higher galactinol and digalactosyl myo-inositol (DGMI), probably due to the decreased accumulation of RFOs resulting in higher accumulations of galactinol (Fig. 1). Low raffinose synthase activity in LRS embryos results in less accumulation of raffinose and stachyose (Fig. 1) which, in turn, results in an increase in galactinol and a higher concentration of DGMI (Fig. 1). LRSP1 and LRSP2 soybean embryos expressing the mutant mips phenotype have low myo-inositol phosphate synthase (MIPS) activity (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002) but normal raffinose synthase and stachyose synthase activities (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002), resulting in reduced galactinol, raffinose, stachyose, total RFOs and phytin (Sebastian et al., Reference Sebastian, Kerr, Pearlstein, Hitz and Drackley2000; Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002) (Fig. 1). CHECK soybean embryos expressing the normal Stc1 and normal Mips phenotype have normal enzyme activities and accumulate normal concentrations of raffinose, stachyose, total RFOs and phytin, and smaller but normal concentrations of galactopinitols and fagopyritols (Fig. 1).
Sucrose, myo-inositol, d-chiro-inositol and d-pinitol are synthesized in soybean leaf (maternal) tissues, transported to the seed, unloaded by the seed coat to the embryo, and may be stored as sucrose, RFOs, fagopyritols and galactopinitols in mature seeds (Gomes et al., Reference Gomes, Obendorf and Horbowicz2005; Obendorf et al., Reference Obendorf, Sensenig, Wu, Ohashi, O'Sullivan, Kosina and Schnebly2008a, Reference Obendorf, Zimmerman, Ortiz, Taylor and Schneblyb, Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009; Kosina et al., Reference Kosina, Castillo, Schnebly and Obendorf2009, Reference Kosina, Schnebly and Obendorf2010). About 70% of RFOs accumulate in soybean seeds after physiological maturity (maximum dry mass) during the period of seed drying (Obendorf et al., Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009). The development of desiccation tolerance occurs concomitantly with the accumulation of stachyose (Blackman et al., Reference Blackman, Obendorf and Leopold1992; Buitink et al., Reference Buitink, Thomas, Gissot and Leprince2003; Rosnoblet et al., Reference Rosnoblet, Aubry, Leprince, Vu, Rogniaux and Buitink2007; Obendorf et al., Reference Obendorf, Zimmerman, Zhang, Castillo, Kosina, Bryant, Sensenig, Wu and Schnebly2009). Galactosyl cyclitols, such as fagopyritols, have been proposed to function in the same role as raffinose and stachyose in seeds that normally do not accumulate raffinose and stachyose (Horbowicz and Obendorf, Reference Horbowicz and Obendorf1994; Horbowicz et al., Reference Horbowicz, Brenac and Obendorf1998). Kosina et al. (Reference Kosina, Schnebly and Obendorf2010) proposed that upregulation of maternally synthesized d-chiro-inositol may be effective for increasing unloaded d-chiro-inositol to embryos and increasing the accumulation of fagopyritols during seed maturation of soybean lines with reduced raffinose and stachyose or with reduced raffinose, stachyose and phytin. Soybean plants normally unload sucrose, myo-inositol, d-chiro-inositol and d-pinitol from seed coats to the embryos during seed development (Gomes et al., Reference Gomes, Obendorf and Horbowicz2005; Kosina et al., Reference Kosina, Castillo, Schnebly and Obendorf2009). Feeding d-chiro-inositol to stem–leaf–pod explants of CHECK, LRS, LRSP1 and LRSP2 soybean increased the unloading of d-chiro-inositol from seed coats (Kosina et al., Reference Kosina, Schnebly and Obendorf2010) and increased the accumulation of fagopyritol B1 in mature, dry seeds of all four lines in the presence of maternal tissues (Obendorf et al., Reference Obendorf, Sensenig, Wu, Ohashi, O'Sullivan, Kosina and Schnebly2008a). The results of the present study demonstrate that feeding d-chiro-inositol to CHECK, LRS, LRSP1 and LRSP2 isolated embryos (free of maternal tissues) significantly increased the accumulation of fagopyritol B1 in both cotyledon and axis tissues during precocious maturation of embryos of all four lines. These results support the proposal (Kosina et al., Reference Kosina, Schnebly and Obendorf2010) that increasing the supply of maternal d-chiro-inositol to soybean embryos may increase the accumulation of fagopyritol B1 in mature seeds expressing the mutant stc1 phenotype with low raffinose and stachyose, expressing the mutant mips phenotype with low raffinose, stachyose and phytin, or expressing the normal Stc1 and Mips phenotype with normal raffinose, stachyose and phytin.
Feeding d-pinitol to isolated embryos increased free d-pinitol in cotyledons of all four lines and increased galactopinitol A in cotyledons of CHECK and LRS embryos, but galactopinitols were not increased in LRSP1 and LRSP2 embryos, probably due to low availability of galactinol in these embryos (Fig. 1). Feeding myo-inositol (with sucrose) to LRSP1 and LRSP2 embryos, expressing a mutant mips gene, resulted in increased stachyose accumulation, most likely by increasing the supply of galactinol for use as the galactosyl donor to form stachyose (Fig. 1), in cotyledons, as previously reported (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Karner et al., Reference Karner, Peterbauer, Raboy, Jones, Hedley and Richter2004).
The myo-inositol feeding experiments are complicated and difficult to interpret. myo-Inositol is synthesized in soybean leaves (and other maternal tissues) and transported to seeds (Gomes et al., Reference Gomes, Obendorf and Horbowicz2005; Kosina et al., Reference Kosina, Castillo, Schnebly and Obendorf2009, Reference Kosina, Schnebly and Obendorf2010). Since the Mips gene is also expressed in embryos of soybean seeds (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Chappell et al., Reference Chappell, Scaboo, Wu, Nguyen, Pantalone and Bilyeu2006), it is likely that myo-inositol is also synthesized in embryos. Additionally, myo-inositol is used to form many products (Loewus and Murthy, Reference Loewus and Murthy2000; Raboy, Reference Raboy2009) including membranes (phosphatidyl inositol phosphates), cell walls (pectin), phytic acid (phytin), galactinol, RFOs, cyclitols (including d-chiro-inositol, d-ononitol and d-pinitol) and galactosyl cyclitols (Fig. 1). We did not analyse all possible products of myo-inositol in cotyledons and axis tissues, including membranes, cell walls and phytic acid; this work focused on the low molecular weight soluble carbohydrates present in maturing soybean seeds. Therefore, it is not surprising that the concentrations of free myo-inositol or total myo-inositol measured after feeding myo-inositol and sucrose to soybean embryos were not as markedly increased as noted for d-chiro-inositol or d-pinitol after feeding these cyclitols. However, LRSP1 and LRSP2 embryos expressing a mutant mips gene accumulated significantly higher concentrations of stachyose and total RFOs in cotyledons (Table 2) after myo-inositol feeding, as expected (Fig. 1); this result has also been reported by others (Hitz et al., Reference Hitz, Carlson, Kerr and Sebastian2002; Karner et al., Reference Karner, Peterbauer, Raboy, Jones, Hedley and Richter2004).
Fagopyritol B1 is formed by galactinol synthase (GolS) using d-chiro-inositol as the galactosyl acceptor and UDP-galactose as the galactosyl donor (Fig. 1; Obendorf et al., Reference Obendorf, Odorcic, Ueda, Coseo and Vassallo2004). Enhanced accumulation of fagopyritol B1 in maturing embryos after feeding d-chiro-inositol has been demonstrated widely in different species (Obendorf and Górecki, Reference Obendorf and Górecki2012) and in soybean lines expressing different mutant genes (Obendorf and Kosina, Reference Obendorf, Kosina and Ng2011). Embryos of smooth tare [Vicia tetrasperma (L.) Schreb.] (Lahuta et al., Reference Lahuta, Górecki and Horbowicz2005) and garden pea (Pisum sativum L.) (Lahuta and Dzik, Reference Lahuta and Dzik2011), which do not normally have d-chiro-inositol or fagopyritols, can take up d-chiro-inositol into immature embryos and form fagopyritol B1 during precocious maturation. In soybean, we have demonstrated an enhanced accumulation of fagopyritol B1 after feeding d-chiro-inositol to immature embryos expressing the mutant stc1 gene and the mutant mips gene (Tables 2 and 3). Because excessively high concentrations of d-chiro-inositol may result in shrivelled seeds (Gomes et al., Reference Gomes, Obendorf and Horbowicz2005), increasing the conversion of myo-inositol to d-chiro-inositol in maternal tissues, followed by transport to and unloading by seed coats, may be a preferred option for increasing fagopyritol B1 in mature seeds toward the improvement of agronomic performance of soybean lines with low raffinose, stachyose and phytin in seeds.
Acknowledgements
This research was conducted as part of Multistate Projects W-1168 and W-2168 (NY-C 125–802; NY-C 125–852) and supported in part by Morley and Hatch-Multistate Undergraduate Research Grants (E.M.S., E.M.B., A.B.O., M.O.), federal work-study funds, and Cornell University Agricultural Experiment Station federal formula funds received from Cooperative State Research, Education and Extension Service, US Department of Agriculture. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the US Department of Agriculture. We gratefully acknowledge Hanna S. Kubica, Angela D. Zimmerman, Jennifer Wu, Harpartap Singh and Timothy E. O'Sullivan for assistance with extractions, Suzanne M. Kosina for helpful suggestions and Françoise Vermeylen for advice on statistical analysis.
