Structures of Starches from Rice Mutants Deficient in the Starch

Amylose and amylopectin of rice mutants deficient in a starch synthase (SS) isozyme in the endosperm, either SSI (ΔSSI) or SSIIIa (ΔSSIIIa), were st...
0 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/Biomac

Structures of Starches from Rice Mutants Deficient in the Starch Synthase Isozyme SSI or SSIIIa Isao Hanashiro,*,† Toshiyuki Higuchi,† Satomi Aihara,‡ Yasunori Nakamura,‡ and Naoko Fujita‡ †

Department of Biochemical Science and Technology, Faculty of Agriculture, Kagoshima University, Kagoshima City, Kagoshima 890-0065, Japan ‡ Department of Biological Production, Akita Prefectural University, Akita City, Akita 010-0195, Japan

bS Supporting Information ABSTRACT: Amylose and amylopectin of rice mutants deficient in a starch synthase (SS) isozyme in the endosperm, either SSI (ΔSSI) or SSIIIa (ΔSSIIIa), were structurally altered from those of their parent (cv. Nipponbare, Np). The amylose content was higher in the mutants (Np, 15.5%; ΔSSI, 18.2%; ΔSSIIIa, 23.6%), and the molar ratio of branched amylose and its side chains was increased. The chain-length distribution of the β-amylase limit dextrins of amylopectin showed regularity, which appeared consistent with the generally accepted cluster structure, and the degrees of polymerization found at the intersections were taken as the boundaries of the B-chain fractions. The mole % of the B1B3 fractions was changed slightly in ΔSSI, which is consistent with the proposed role of SSI in elongating the external part of clusters. In ΔSSIIIa, a significant increase in the B1 fraction and a decrease in the B2 and B3 fractions were observed. The internal chain length of the B2 and B3 fractions appeared to be slightly altered, suggesting that the deficiency in SS affected the actions of branching enzyme(s).

’ INTRODUCTION Rice (Oryza sativa L.) is one of the main cereals that serve as a carbohydrate source in the human diet. Improving the eating quality of cooked rice grains has become an important subject in fields related to the food industry. Amylose content is a wellknown factor associated with the gelatinization properties of starch, which inevitably affects the cooking properties of the grains. Thus, the Wx gene and its product, granule-bound starch synthase I (GBSSI), have become primary targets for modifying amylose content. Considering its highly branched structure, it is easily imaginable that the effect of amylopectin on starch properties is more complicated. However, studies have shown a clear relationship between the pasting properties of starch and structural characteristics of amylopectin, such as chain-length (CL) distribution1 or content of extra-long chains (ELCs).2,3 Therefore, one can consider amylose or amylopectin, or both, as candidate targets to design a starch with desired properties, that is, a so-called “tailormade starch”. To achieve that goal, it is essential to elucidate both the function(s) of each biosynthetic enzyme46 and the interactions among the enzymes through reversible phosphorylation.7,8 Recently, rice null mutants, in which both the protein and enzymatic activity of either SSI (mutant line e7/)9 or SSIIIa (ss3a-1)10 was eliminated by inserting the retrotransposon Tos17 into the corresponding gene, were isolated. SSI and IIIa are, respectively, the first and second major SS isozymes responsible for amylopectin biosynthesis in the developing endosperm in japonica rice cultivars.9 Characterization of endosperm starch r 2011 American Chemical Society

biosynthesis in these plants and some structural analyses of their amylopectins were carried out, and functions of rice SSI and SSIIIa have been proposed.9,10 The aim of this study was first to examine the structures of amylose in the mutants because these structures were not subjected to analyses in the previous studies. Second, the structures of amylopectin and its β-amylase limit dextrin were examined to clarify further the structural changes occurred by SS deficiency.

’ EXPERIMENTAL SECTION Materials. The rice starches of two SS-null mutant lines (e7, SSIdeficient; ss3a-1 (e1), SSIIIa-deficient) and their parental cultivar Nipponbare were the same specimens used in the previous studies.9,10 In this study, the mutant lines e7/ and ss3a-1 are referred to as ΔSSI and ΔSSIIIa, respectively. Amylose and amylopectin were fractionated from defatted starches according to Takeda, Hizukuri, and Juliano.11 Preparation and debranching of the β-amylase limit dextrin (β-LD) of amylopectin were carried out according to the method previously reported.12 Pseudomonas isoamylase and Klebsiella pullulanase were purchased from Hayashibara Biochemical Laboratories. (Okayama, Japan). Sweet potato β-amylase (Sigma Chemical, St. Louis, MO) was purified by anion-exchange chromatography prior to use. Other reagents were of the highest grade commercially available. Received: January 5, 2011 Revised: February 14, 2011 Published: March 18, 2011 1621

dx.doi.org/10.1021/bm200019q | Biomacromolecules 2011, 12, 1621–1628

Biomacromolecules Size-Exclusion HPLC of Fluorescently Labeled Specimen. Fluorescent labeling and subsequent size-exclusion HPLC (HPSEC) analyses of amylose,13 debranched amylopectin, and debranched β-LD of amylopectin14 were performed according to previous studies. Because fluorescent 2-aminopyridine is introduced via reductive amination to a reducing residue of the glucans, the resulting HPSEC profile of fluorescence detector response gives a molar-based distribution of analyte. Separation of amylose was performed at 37 °C and flow rate of 0.5 mL/min with TSK guard column PWH (7.5  75 mm), TSKgel G6000PW, G4000PW, and G3000PW (connected in series, 7.5  600 mm each, all from Tosoh), using 0.1 M phosphate buffer (pH 6.1) containing 0.02% sodium azide as eluent. Separation of debranched amylopectin or its β-LD was carried out at 50 °C and flow rate of 0.25 mL/min with an HPSEC system equipped with a Shodex OHpak SB-G (6.0  50 mm), a Shodex OHpak SB-803 HQ, and two Shodex OHpak SB-802.5 HQ (8.0  300 mm each, Showa Denko, Tokyo, Japan), using aqueous DMSO (50%) containing 50 mM NaCl as an eluent. The respective HPSEC system was equipped with a fluorescence detector (exiting and measuring wavelengths were 315 and 400 nm, respectively) and a differential refractometer. The ratio of detector responses, refractive index/fluorescence (RI/F), measured as either a given area or height, has been shown to be proportional to (numberaverage) degree of polymerization (DPn or DP) of amylose13 or debranched amylopectin14 and was used to calculate DP(n), using appropriate synthetic amylose standards with narrow size distribution and known DPn (AS110 (DPn, 521) for TSK columns and AS10 (57.7) for OHpak columns, both from IPE, Osaka, Japan). Other Analytical Methods. Capillary electrophoresis (CE) analysis of CL distribution of β-LD of amylopectin was performed as previously described.9 The apparent and actual amylose content was determined on a weight basis by HPSEC of debranched starch and amylopectin.15 The apparent content was obtained as a percentage of area of the peak corresponding to the sum of amylose and ELC of amylopectin in debranched starch, and the actual content was obtained after subtraction of peak area corresponding to ELC of amylopectin. The molar fraction of branched molecules (MFB) of amylose was determined by HPSEC of β-amylolyzate of labeled amylose, as previously described.15 The average number of chains (NCs) of amylose was calculated on the basis of the determination of reducing and nonreducing residues.16

’ RESULTS AND DISCUSSION Apparent and Actual Amylose Content of Starch. The amylose content (% by weight, mean of two measurements) of the rice starches of ΔSSI, ΔSSIIIa, and cv. Nipponbare was, respectively, 20.6, 26.0, and 16.8 (apparent content), and 18.2, 23.6, and 15.5 (actual content). The amylose content, both apparent and actual, of the mutant lines was higher than that of cv. Nipponbare. The amylose content of ΔSSIIIa agreed with that obtained previously by Toyopearl HW55S-HW50S chromatography (apparent, 24.8%; actual, 20.0%).10 For ΔSSI, a slightly higher λmax of the starchiodine complex was previously observed (566 nm, compared with 557 nm for cv. Nipponbare),9 a finding that is consistent with the slightly higher values of amylose content determined in this study. Amylose in the rice endosperm is synthesized by granulebound starch synthase I (GBSSI).17 The SS isozymes deficient in the mutants examined here are generally thought to contribute to elongation of the unit chains of amylopectin. The main SSs responsible for amylopectin biosynthesis in the developing endosperm of cv. Nipponbare, the parent of the two SS mutants, are SSI and SSIIIa.9,10 The activity of SSIIa, another major class of

ARTICLE

SS isozyme, is significantly reduced in japonica rice cultivars.18,19 The increased amylose content in the SS mutants was considered to result from both direct and indirect effects of the mutations. The direct effect would be a relative increase in amylosesynthesizing ability in the mutant endosperm due to the loss of activity of the respective SS isozyme. In addition, the loss of SS activity would result in more preferable conditions for GBSS. Although ADP-glucose (ADPG) serves as a common glucosyl donor for both the granule-bound and soluble starch synthases, the Michaelis constant for the substrate is much higher for GBSS than for SS.20 Furthermore, increased activity of ADPG pyrophosphorylase has been found in the rice SS mutants.9,10 Thus, a rational expectation is that a deficiency in either SSI or SSIIIa, that is, leading to decreased consumption of ADPG by SS and increased amounts of ADPG, could result in an increase in the amount, concentration, or both of ADPG available for GBSSI, which would lead to a higher amylose content. A similar situation was reported for potato tubers of SSIII antisense lines, which showed both increased ADPG in tubers and starch with a higher amylose content.21 In the case of ΔSSIIIa, the increased amount of GBSSI protein caused by a pleiotropic effect of the SSIIIa mutation10 is regarded as an indirect effect of the SS mutation on the amylose content. High-amylose phenotypes associated with SS deficiency have been previously reported, including SSIIIdeficient maize,22 and SSII-deficient maize,23 wheat,24 barley,25 and pea.26 Structure of Amylose. Some structural characteristics of amylose isolated from the rice starches are summarized in Table 1. The number-average degree of polymerization (DPn) of amylose was determined by both colorimetric (reducing residues by a modified Park-Johnson method) and labeling methods. There were no large differences between the two methods or between the values obtained for the cv. Nipponbare and the SS mutants. Nevertheless, high-performance size-exclusion chromatography (HPSEC) showed that the size distribution of amylose was altered by SS deficiency (Figure 1). ΔSSI and ΔSSIIIa showed a fluorescence peak that eluted slightly faster and had a more pronounced shoulder at 7580 min as compared with cv. Nipponbare; in addition, ΔSSI showed a peak that was shifted leftward to a retention time of ca. 8590 min (Figure 1A). cv. Nipponbare and ΔSSI showed refractive index (RI) profiles that were similar to each other except for a slightly higher peak in ΔSSI (Figure 1B). The size distribution of ΔSSIIIa amylose showed a decrease in high-molecular-weight molecules and an increase in midrange molecules as compared with cv. Nipponbare and ΔSSI. These differences in elution profile were consistently reproducible. (Multiple elution profiles for each amylose are presented in the Supporting Information.) Table 1 also indicates that the branching characteristics of amylose were altered by SS mutation; namely, more branched amylose was produced in the mutants. The decrease in numberaverage chain length (CLn) in the ΔSSI and ΔSSIIIa mutants was attributed to a 1.5-fold increase in the number of side chains of branched amylose molecules (NCB) in ΔSSI; furthermore, the molar ratio of branched-to-linear amylose molecules (MFB) was increased in addition to NCB in ΔSSIIIa. Such an increase in the degree of branching of amylose has been found to accompany SSII deficiency in wheat27 and increased amylose content in some transgenic lines of Wxa-expressing transgenic rice.12,15 Chain-Length Distribution of Amylopectin. In previous studies, the CL distribution of the mutant amylopectins was examined by CE after reducing residues were labeled with a 1622

dx.doi.org/10.1021/bm200019q |Biomacromolecules 2011, 12, 1621–1628

Biomacromolecules

ARTICLE

Table 1. Structures of Rice ΔSSI and ΔSSIIIa Amylosea DPn source

colorimetric

CLn

labeling

NC

MFB

NCB

Nipponbare

1060 ( 46

1350* ( 50

261 ( 6

4.1

0.34 ( 0.01

10.1

ΔSSI

1240* ( 60

1280** ( 36

205* ( 9

6.0

0.37* ( 0.01

14.5

ΔSSIIIa

1120** ( 47

1290*,** ( 12

167** ( 6

6.7

0.42** ( 0.01

14.6

DPn (labeling) was determined from RI/F ratio of total peak area of corresponding chromatogram by refractive index and fluorescence detectors. NC and NCB were calculated as DPn (colorimetric)/CLn and [NC  (1  MFB)]/MFB, respectively, where NC, NCB, and MFB are the average number of chains, the average number of chains of branched molecules, and the molar fraction of branched molecules, respectively. Values are the means ( SD of four to six replicates. A different number of asterisks in the same column indicates that the values are significantly different (by t test with Bonferroni correction, P < 0.017). a

Figure 2. Chain-length distribution of rice ΔSSI and ΔSSIIIa amylopectin. (A) HPSEC chromatograms: solid line, fluorescence; dashed line, RI; dash-dot-dash line, DP determined from RI/F (height) at a given elution position. Numbers with arrowheads indicate peak-top DP. (B) Difference obtained by subtraction of detector response of cv. Nipponbare from that of mutant amylopectin: dashed line, ΔSSI; solid line, ΔSSIIIa. Numbers indicate peak-top DP. Figure 1. Size distribution of rice amylose. The amyloses were fractionated from rice starch with butanol precipitation, followed by purification with ultracentrifugation. (A) Fluorescence profiles: pale gray, cv. Nipponbare; gray, ΔSSI; black, ΔSSIIIa. (B) RI profiles: gray dashed line, cv. Nipponbare; dashed line, ΔSSI; solid line, ΔSSIIIa. The peak areas of the six profiles were normalized by RI response. Numbers indicate peak-top DP. DP plot (dashed-dotted-dashed line in pale gray) of cv. Nipponbare amylose was shown for reference, which was calculated from RI/F ratio (response measured as height) of a given elution position.

charged fluorophore, 8-aminopyrene-1,3,6-trisulfonic acid (APTS).9,10 Although this method provides high resolution of unit chains with different degrees of polymerization (DP), the longer the CL, the lower the detector response because only one residue per chain is labeled regardless of the CL. Therefore, detection with an RI detector is expected to counteract the disadvantage of molar-based analyses. Figure 2A shows the chromatograms of debranched amylopectin obtained by HPSEC with fluorescence and RI detectors. The elution profiles were divided at inflection points into four fractions, as shown in Figure 2A, and the relative amounts of the unit-chain fractions are summarized in Table 2. Figure 2B shows the differences in CL distribution between the mutants and cv. Nipponbare. On both molar and weight bases, SS deficiency caused changes in CL distribution (Figure 2), as compared with cv. Nipponbare.28 The changes appeared specific for each SS-deficient mutant and were more pronounced in ΔSSIIIa.

On a molar basis, ΔSSI amylopectin showed a CL distribution that was essentially similar to the previously reported CL distribution of cv. Nipponbare (Figure 3C in Fujita et al.28), although the peaks of the A and B1 fractions were slightly shifted leftward (Figure 2A). Partial compensation of function of SSI by other SS isozymes is thought to be the reason for the relatively modest changes in CL distribution.9 More evident changes in CL distribution were observed for ΔSSIIIa amylopectin, namely, increased and decreased amounts of the A and B2þB3 fractions, respectively (Figure 2 and Table 2). The molar ratio of the unitchain fractions, (AþB1)/(B2þB3), was 10.5 for cv. Nipponbare, 10.8 for ΔSSI, and 16.8 for ΔSSIIIa. The higher ratio in ΔSSIIIa was mainly due to the decrease in the B2þB3 fraction, indicating a more densely branched cluster structure. The fluorescence profiles essentially showed the same characteristic differences as those detected by CE.9,10 On a weight basis, in addition to the changes corresponding to those detected on a molar basis, the content of ELCs was higher in ΔSSI and ΔSSIIIa than in cv. Nipponbare. Because ELCs are a product of the amylosesynthesizing enzyme GBSSI,15,29 the increase in ELCs is thought to be an indirect effect that accompanies SS deficiency, as described above for the increased amylose content. The number-average chain length (CLn) of amylopectin and its unit-chain fractions did not significantly differ among the three amylopectins (Table 2), indicating that SS deficiency affects the amount but not the average length of each group of unit chains, that is, A, B1, B2, and B3. 1623

dx.doi.org/10.1021/bm200019q |Biomacromolecules 2011, 12, 1621–1628

Biomacromolecules

ARTICLE

Table 2. Chain-Length Distributions of Rice ΔSSI and ΔSSIIIa Amylopectina amount (mole %) source Nipponbare ΔSSI ΔSSIIIa

ELC b

c