Raffinose and Stachyose Metabolism Are Not Required for Efficient Soybean Seed Germination

ARTICLE IN PRESS

Journal of Plant Physiology 166 (2009) 1329—1335

www.elsevier.de/jplph

SHORT COMMUNICATION Rafﬁnose and stachyose metabolism are not required for efﬁcient soybean seed germination

Emily C. Dierkinga, Kristin D. Bilyeub,Ã aUniversity of Missouri-Columbia, Division of Plant Sciences, 110 Waters Hall, Columbia, MO 65211, USA bUSDA-ARS, Plant Genetics Research Unit, 110 Waters Hall, Columbia, MO 65211, USA

Received 14 November 2008; received in revised form 20 January 2009; accepted 20 January 2009

KEYWORDS Summary a-galactosidase Raffinose family oligosaccharides (RFOs), which include raffinose and stachyose, are inhibitor; thought to be an important source of energy during seed germination. In contrast to Carbohydrates; their potential for promoting germination, RFOs represent anti-nutritional units for Raffinose family monogastric animals when consumed as a component of feed. The exact role for oligosaccharides; RFOs during soybean seed development and germination has not been experimen- Seed germination; tally determined; but it has been hypothesized that RFOs are required for successful Soybean germination. Previously, inhibition of RFO breakdown during imbibition and germination was shown to significantly delay germination in pea seeds. The objective of this study was to compare the germination potential for soybean seeds with either wild-type (WT) or low RFO levels and to examine the role of RFO breakdown in germination of soybean seeds. There was no significant difference in germination between normal and low RFO soybean seeds when imbibed/germinated in water. Similar to the situation in pea, soybean seeds of wild-type carbohydrate composition experienced a delay in germination when treated with a chemical inhibitor of a-galactosidase activity (1-deoxygalactonojirimycin or DGJ) during imbibition. However, low RFO soybean seed germination was not significantly delayed or reduced when treated with DGJ. In contrast to the situation in pea, the inhibitor-induced germination delay in wild-type soybean seeds was not partially overcome by the addition of galactose or sucrose. We conclude that RFOs are not an essential source of energy during soybean seed germination. Published by Elsevier GmbH.

Abbreviations: DGJ, 1-deoxygalactonojirimycin; DW, dry Introduction weight; RFO, raffinose family oligosaccharides; STD, standard deviation; WT, wild-type. Although the exact function of raffinose family ÃCorresponding author. Tel.: +1 573 884 0451; fax: +1 573 884 7850. oligosaccharides (RFOs; raffinose, stachyose, and E-mail addresses: [email protected], verbascose) in germinating seeds is largely un- [email protected] (K.D. Bilyeu). known, it is clear that seeds require a large amount

0176-1617/$ - see front matter Published by Elsevier GmbH. doi:10.1016/j.jplph.2009.01.008 ARTICLE IN PRESS

1330 E.C. Dierking, K.D. Bilyeu of energy during germination (Bewley and Black, contrasting RFO levels that could be used to 1994). This energy is hypothesized to come from produce seeds for germination studies. stored carbohydrates; sucrose and RFOs are the Blo¨chlet al. (2007) directly tested the effect of most abundant of the soluble sugars (Peterbauer inhibition of the breakdown of RFOs on germination and Richter, 2001) but account for only a small in pea. The results of this study indicated that pea portion of the total carbohydrates present in the seeds exhibited a significant delay in germination seeds (Ziegler, 1995). RFOs have previously been when a-galactosidase activity and thus RFO meta- thought to be involved in seed protection during bolism was inhibited by 1-deoxygalactonojirimycin the desiccation process (Black et al., 1996; (DGJ). Further, the germination delay in DGJ- Obendorf, 1997; Bailly et al., 2001) by stabilization treated pea seeds could be partially overcome by of the membrane (Crowe et al., 1987). Recently, the addition of galactinol (Blo¨chlet al., 2007). low raffinose and stachyose and low raffinose, Because of the implication that RFOs play critical stachyose, and phytin lines were tested for their roles in seed development and germination, the sensitivity to imbibitional chilling (Obendorf et al., objective of this project was to determine (without 2008). The results indicated that lines low in any selection for agronomic properties) the germi- raffinose and stachyose were similar to control nation potential of soybean lines with low RFOs as lines, with no sensitivity, while lines low in compared to sibling lines with wild-type RFO raffinose, stachyose, and phytin were sensitive to content. In addition, we sought to examine the imbibitional chilling (Obendorf et al., 2008). How- effect of blocking RFO metabolism in soybean seed ever, many of the proposed functions of RFOs in the germination. developing, stored, and germinating seed have not been confirmed experimentally (Bentsink et al., 2000; Gurusinghe and Bradford, 2001), but are Materials and Methods generally considered valid. Since the breakdown of seed components during germination has been studied extensively, it is known that protein, oils Plant Material and polymeric carbohydrates each break down A ‘Williams 82’ (Bernard and Cremeens, 1988)X slowly over a period of several days (Bewley and PI 200508 cross was made in the summer of 2005 at Black, 1994). This period extends beyond germina- Bradford Research and Extension Center near tion and thus cannot be solely responsible for Columbia, Missouri. Successful crosses were carried fulfilling the energy demands of the germinating through F . A population of approximately 84 F seed. 2 2 plants was grown to maturity in the summer of A soybean plant introduction line, PI 200508, was 2007. The resulting F seeds were phenotyped for identified as having reduced levels of RFOs and 2:3 seed oligosaccharide content using high perfor- elevated levels of sucrose (Kerr and Sebastian, mance liquid chromatography (HPLC). For each 2000). Initial characterization of soybean line PI plant, eight individual seeds were analyzed (Dierking 200508 was carried out by Hitz et al. (2002). Seeds and Bilyeu, 2008). Five lines were selected with of PI 200508, in addition to reduced RFOs, had a 25- wild-type RFO phenotypes and five lines with low fold reduction in raffinose synthase enzyme activity RFO phenotypes. Three hundred F seeds from (Hitz et al., 2002). Each of these characteristics 2:3 each line were used for the germination experi- potentially improves the nutritional quality of the ments described below. The Williams 82 seeds used soybean meal and together highlights the soybean were produced in the summer of 2008. line PI 200508 as an important source for the low RFO trait. In a subsequent study, Neus et al. (2005) reported that the vigor of backcrossed PI 200508- HPLC Oligosaccharide Analysis derived low RFO lines was not affected as there were no significant differences in a number of seed Oligosaccharides were separated and quantified quality traits including field emergence, seed yield, by high performance ion chromatography with maturity, and fatty acid content as compared to pulsed amperometric detection (PAD) employing derived lines with a wild-type (WT) RFO phenotype. an Agilent 1100 series HPLC and an ESA Coulochem We recently determined that the molecular genetic III detector (Agilent Technologies, Chesterfield, basis for the low RFO trait in soybean line PI 200508 MO, USA). A 12.5 mg ground seed sample was was a variant allele of a raffinose synthase gene extracted with 0.5 ml 50% ethanol at 70 1C, (Dierking and Bilyeu, 2008). This discovery allowed 30 min. Samples were then centrifuged 15 min at us to develop closely related soybean germplasm 16,000 g. The supernatant was passed through a without direct agronomic selection, but with 0.2 mm filter. Sugars were separated on a Dionex ARTICLE IN PRESS

Efﬁcient germination of low rafﬁnose and stachyose soybean seeds 1331

Carbo Pac PA 10 analytical column (250 mm Â 4 mm, tose. The experimental time points and data 10 mm) connected to a Carbo Pac PA 10 guard collection were carried out exactly as described column (50 mm Â 4 mm). The mobile phase was above except three replicates of 50 seeds were 90 mM NaOH with flow rate of 1.5 mL minÀ1, used for each treatment. maintained at 30 1C. Detection settings were: time 0, 0.1 v, time 0.41, À2.0 v, time 0.42, 0.6 v, and time 0.44, À0.1 v. Results and Discussion Germination Experiments Soybean seeds were produced in a field environ- Germination experiments for wild-type and low ment from plants which stably inherited wild-type RFO soybeans included two treatments, using water or low RFO phenotypes depending on their inheri- (control) or DGJ (Industrial Research, Wellington, tance of homozygous variant alleles of a raffinose New Zealand), and were carried out essentially as synthase gene (RS2) from the wild-type (Williams 82) described (Blo¨chl et al., 2007). Seeds were imbibed or low RFO (PI 200508) parent (Dierking and Bilyeu, at 25 1C in the dark in Petri dishes (50 seeds per 2008). The plants were related F2 individuals 15 mm Petri dish) with 25 mL of either water that produced F2:3 seeds, but were not selected (control) or 50 mM DGJ. Seeds were imbibed for for any particular agronomic characteristics. Five 16 h and then transferred to containers lined with F2:3 lines were selected with wild-type RFO profiles wetted paper towels (DGJ treatment was wetted and homozygous wild-type RS2 alleles, and five with 50 mM DGJ). One hundred fifty seeds were used lines were selected with low RFO profiles and per line, per treatment. The seeds were placed in homozygous variant alleles of RS2 (Table 1). In the 25 1C in the dark for germination. Germination was lines with the low RFO profile, sucrose was 165% defined as the point at which the radical pierced of the wild-type level, and raffinose and stachyose the seed coat. The number of germinated seeds decreased to 18% and 33% of the wild-type levels, was counted at six different time intervals: 23, 40, respectively, consistent with a reduction in raffi- 47, 64, 71, and 88 h; where 0 h is the time at which nose synthase enzyme activity in developing seeds seeds are introduced to either water or DGJ in the (Hitz et al., 2002). Five F2:3 lines of each RFO Petri dish. Germination experiments were per- profile were chosen to overcome the effect of other formed for five lines (F2:3 seeds) for each pheno- unknown segregating genes which might influence type, either wild-type RFO levels or low-RFO levels. germination. Germination experiments for Williams 82 seeds Germination was monitored for the lines with included six treatments: water (control), 50 mM contrasting RFO profiles to determine if reduced DGJ, 25 mM sucrose (Sigma, St. Louis, MO, USA), RFOs would have a negative impact on standard 50 mM galactose (Sigma, St. Louis, MO, USA), 50 mM germination. Standard germination was efficient DGJ+25 mM sucrose, and 50 mM DGJ+50 mM galac- for both types of lines, and reached a plateau of

Table 1. Oligosaccharide content of wild-type and low RFO dry seeds.

Line Genotype Sucrose (%DW) Rafﬁnose (%DW) Stachyose (%DW)

Mean STD Mean STD Mean STD

1 WT 4.85 1.76 1.13 0.27 5.99 0.69 2 WT 4.86 0.89 0.99 0.25 5.79 0.43 3 WT 5.05 1.12 1.13 0.13 4.77 0.64 4 WT 3.79 0.54 1.07 0.09 5.96 0.57 5 WT 3.51 0.71 0.85 0.14 4.65 0.81 Mean 4.41 1.00 1.03 0.18 5.43 0.14

6 Low RFO 6.06 1.08 0.19 0.04 2.26 0.34 7 Low RFO 6.19 1.36 0.20 0.03 1.47 0.44 8 Low RFO 9.35 1.03 0.18 0.03 1.78 0.24 9 Low RFO 7.74 0.61 0.18 0.07 1.87 0.60 10 Low RFO 7.00 1.40 0.18 0.05 1.58 0.41 Mean 7.27 1.09 0.19 0.04 1.79 0.41

DW, dry weight; STD, standard deviation. WT, wild type; RFO, rafﬁnose family oligosaccharide. ARTICLE IN PRESS

1332 E.C. Dierking, K.D. Bilyeu approximately 95% germination by 64 h after signiﬁcantly delay germination; at 42 h after imbibition (Figure 1A). There were no signiﬁcant imbibition, only 25% of DGJ-treated pea seeds differences for standard germination in the control germinated compared to 90% for the control treatment between wild-type and low RFO lines. treatment (Blo¨chlet al., 2007). Since DGJ inhibits Given that standard germination was not de- a-galactosidase activity, and the addition of galac- creased in low RFO lines, the germination potential tose was able to partially overcome the germina- of soybean seeds treated with the a-galactosidase tion delay in pea, we investigated the effect on inhibitor DGJ was evaluated. Treatment of imbibing germination for DGJ-treated soybean seeds. Similar pea seeds with DGJ was previously shown to to the situation in pea, soybean lines with wild-type

100 90 80 70 60 50 40 30 20 germination (percent) WT - control 10 Low RFO - control 0 0 16 23 40 47 64 71 88 time (hours)

100 90 80 70 60 50 40 30

20 WT - control germination (percent) 10 WT - DGJ 0 0 16234047647188 time (hours)

100 90 80 70 60 50 40 30 20

germination (percent) Low RFO - control 10 Low RFO - DGJ 0 0 16234047647188 time (hours)

Figure 1. Germination of wild-type (WT) and low RFO soybean seeds in water (control) or DGJ from 0 to 88 h after imbibition. Data points represent the means of 5 independent lines, 150 seeds each. Error bars represent mean7standard deviation (STD): (A) control germination for WT and low RFO soybean seeds, (B) control and DGJ treatment for WT oligosaccharide soybean seeds, and (C) control and DGJ treatment for low RFO soybean seeds. ARTICLE IN PRESS

Efﬁcient germination of low rafﬁnose and stachyose soybean seeds 1333

RFO levels exhibited a delay in germination for the phenotypes was not affected by the a-galactosidase DGJ-treatment compared to the control treatment inhibitor; soybeans with wild-type RFO phenotypes (Figure 1B). Although germination was statistically exhibited a significant but less drastic delay in significantly lower for the DGJ-treated seeds for germination compared to that observed in pea every time point beyond 23 h after imbibition, at seeds (Blo¨chlet al., 2007). Soybean seeds appear to 88 h the difference between control treated and germinate efficiently with reduced RFO content, DGJ-treated seeds was only 98–92%, respectively. and mobilization of RFOs during soybean germina- The maximum difference in average germination tion has different responses to DGJ-treatment than between control and DGJ-treated wild-type RFO mobilization of RFOs during pea germination soybean seeds was at 40 h and only reduced (Blo¨chlet al., 2007). germination from 82% to 61%, a less drastic In order to further understand the delay in reduction than was previously observed for pea germination of wild-type RFO soybeans treated (Blo¨chlet al., 2007). Germination for DGJ-treated with DGJ, we carried out additional experiments low RFO soybeans was not significantly different at using seeds of the cultivar Williams 82, the source any time point (Figure 1C). This result indicates of the wild-type RFO phenotype. The mean that the germination of soybeans with low RFO oligosaccharide content as determined by HPLC

100 90 80 70 60 sucrose 50 sucrose + DGJ 40 control 30 DGJ germination (percent) 20 10 0 0 16234047647188 time (hours)

100 90 80 70 60 galactose 50 galactose + DGJ 40 control DGJ 30

germination (percent) 20 10 0 0 16 23 40 47 64 71 88 time (hours)

Figure 2. Germination of Williams 82 soybean seeds in water (control), sucrose, sucrose+DGJ, galactose, galactose+DGJ, or DGJ from 0 to 88 h after imbibition. Data points represent the means of three replicates of 50 seeds. Error bars represent mean7STD: (A) germination of Williams 82 soybean seeds in water (control), sucrose, sucrose+DGJ, or DGJ. (B) Germination of Williams 82 soybean seeds in water (control), galactose, galactose+DGJ, or DGJ. ARTICLE IN PRESS

1334 E.C. Dierking, K.D. Bilyeu

50 45 40 35 30 25 20 15

germination (percent) 10 5 0 control sucrose sucrose + galactose galactose DGJ DGJ + DGJ

Figure 3. Germination of Williams 82 seeds in water (control), sucrose, sucrose+DGJ, galactose, galactose+DGJ, and DGJ at 23 h after imbibition. Each bar represents the mean of three replicates of 50 seeds each. Error bars represent mean7STD. for four Williams 82 seeds is 4.53% dry weight (DW) DGJ delayed the germination of wild-type oligo- sucrose, 0.45% DW raffinose, and 3.40% DW sta- saccharide soybeans, but DGJ had no affect on chyose. Six treatments were employed: water soybeans with a reduced raffinose and stachyose (control), DGJ, sucrose, galactose, DGJ+sucrose, oligosaccharide content. Further, the reduction at and DGJ+galactose (Figure 2). As expected, treat- 23 h in germination of the DGJ-treated Williams 82 ment with DGJ induced a minor reduction in early seeds could not be overcome by the addition of germination, but by 88 h after imbibition there was sucrose or galactose. Our results demonstrate that no significant difference between any treatment. soybean and pea have a fundamental difference in The maximum difference in average germination oligosaccharide requirements during germination. between control and DGJ-treated Williams 82 This data refutes the long held assumption that soybean seeds was at 23 h and reduced germination accumulated seed RFOs are required for successful from 42% to 16%. Unlike the situation in pea (Blo¨chl soybean seed germination. We have shown that et al., 2007), the addition of either sucrose or reducing raffinose and stachyose in soybean seeds, galactose also induced a significant reduction in without selection for agronomic characteristics, soybean germination at 23 h, although there was no does not reduce or delay soybean seed germina- significant difference as compared to the control tion. This neutral germination impact provides treatment from 40 h through 88 h. The addition of additional incentive for soybean breeding programs sucrose or galactose to DGJ-treated seeds did not to incorporate the low RFO trait into elite varieties overcome the reduction in germination. At 23 h all to increase the value of soybean meal used for treatments significantly reduced germination as consumption by monogastric animals. compared to the control treatment (Figure 3). The control treated seeds had significantly higher germination at this time point; there were no Acknowledgements significant differences between any of the remain- ing treatments. This result is in contrast with The authors would like to acknowledge M. germination in pea seeds where the addition of Reinerd and A. Tetlow for technical assistance and sucrose and galactose partially overcame the experimental design. reduced germination effect of DGJ on germinating seeds (Blo¨chlet al., 2007). Overall, the results of this study indicate that wild-type levels of raffinose and stachyose are not References required for efficient germination of soybean Bailly C, Audigier C, Ladonne F, Wagner MH, Coste F, seeds. The suggested necessity of RFO mobilization Corbineau F, Come D. Changes in oligosaccharide for efficient seed germination is shown here to be content and antioxidant enzyme activities in devel- inconsequential, at least for soybeans. There was oping bean seeds as related to acquisition of drying no significant difference between germination of tolerance and seed quality. J Exp Bot 2001;52:701–8. control treated wild-type RFO and low RFO soy- Bentsink L, Alonso-Blanco C, Vreugdenhil D, Tesnier K, beans; the addition of the a-galactosidase inhibitor Groot SP, Koornneef M. Genetic analysis of ARTICLE IN PRESS

Efﬁcient germination of low rafﬁnose and stachyose soybean seeds 1335

seed-soluble oligosaccharides in relation to seed stor- Hitz WD, Carlson TJ, Kerr PS, Sebastian SA. Biochemical ability of Arabidopsis. Plant Physiol 2000;124:1595–604. and molecular characterization of a mutation that Bernard RL, Cremeens CR. Registration of ’Williams 82’ confers a decreased raffinosaccharide and phytic acid soybean. Crop Sci 1988;28:1027–8. phenotype on soybean seeds. Plant Physiol 2002;128: Bewley JD, Black M. Seeds: Physiology of Development 650–60. and Germination, second ed. New York: Plenum Press; Kerr PS, Sebastian SA. US Patent 6147193. Date Issued: 1994. November 14, 2000. Black M, Corbineau F, Grzesikit M, Guyi P, Come D. Neus JD, Fehr WD, Schnebly SR. Agronomic and seed Carbohydrate metabolism in the developing and characteristics of soybean with reduced raffinose and maturing wheat embryo in relation to its desiccation stachyose. Crop Sci 2005;45:489–592. tolerance. J Exp Bot 1996;47:161–9. Obendorf RL. Oligosaccharides and galactosyl cyclitols in Blo¨chlA, Peterbauer T, Richter A. Inhibition of raffinose seed desiccation tolerance. Seed Sci Res 1997;7: oligosaccharide breakdown delays germination of pea 63–74. seeds. J Plant Physiol 2007;164:1093–6. Obendorf RL, Zimmerman AD, Ortiz PA, Taylor AG, Crowe JH, Crowe LM, Carpenter JF, Aurell Wistrom C. Schnebly SR. Imbibitional chilling sensitivity and Stabilization of dry phospholipid bilayers and proteins soluble carbohydrate composition of low raffinose, by sugars. Biochem J 1987;242:1–10. low stachyose soybean seed. Crop Sci 2008;48: Dierking EC, Bilyeu KD. Association of a soybean raffinose 2396–403. synthase gene with low raffinose and stachyose seed Peterbauer TR, Richter A. Biochemistry and physiology of phenotype. Plant Genome 2008;1:135–45. raffinose family oligosaccharides and galactosyl cycli- Gurusinghe S, Bradford KJ. Galactosyl-sucrose oligosac- tols in seeds. Seed Sci Res 2001;11:185–97. charides and potential longevity of primed seeds. Ziegler P. Seed development and germination. New York: Seed Sci Res 2001;11:121–34. Marcel Dekker Inc.; 1995.