Introduction

Starch is the main storage carbohydrate in higher plants and not only is the primary source of calories in human diet but also has great value in industrial applications. Because its starch content, rice (Oryza sativa L.) is one of the most important staple foods for more than half of the world’s population. Starch synthesis and accumulation occur in the amyloplasts that exist in tubers, cotyledons, and endosperms, and is also an important determinant for crop yield1,2.

Starch comprises two major components, amylose and amylopectin. A series of key enzymes involved in starch biosynthesis have been identified. ADP-glucose (ADPG) is the substrate for starch synthesis, which is converted from glucose-1-phosphate by ADP-glucose pyrophosphorylase (AGPase). Granule-bound starch synthase I (GBSSI), which is encoded by the Waxy gene is required for the synthesis of amylose3,4. On the other hand, a series of starch synthases (SSI, SSII, SSIII, and SSIV), starch branching enzymes (BEI and BEII), and starch debranching enzymes (ISA and Pullulanase) control the synthesis of amylopectin5,6,7. Mutants defective in these genes show abnormal characteristics of endosperm stored starch. For instance, mutations in OsAGPL2 and OsAGPS2b (encoding a large and small AGPase subunits, respectively) cause a shrunken endosperm, due to a significant reduction in starch synthesis8. Loss-of-function mutations of the rice GBSSI (Waxy) gene produce an amylose-free endosperm9,10. A deficiency of SSIIIa leads to abnormal amylose content, amylopectin structure, and starch granule properties11,12. Loss-of-function mutation of ISA1 produces a sugary endosperm and a dramatic alteration of polyglucan structure from amylopectin to phytoglycogen13,14. Mutation of BEIIb in the amylose-extender (ae) mutant generates a chalkiness endosperm and alterations in the structure of amylopectin and in the gelatinization properties of starch granules15.

Besides these key enzyme genes, several regulators are known to indirectly regulate starch biosynthesis in rice endosperm. For instance, FLO2, a tetratricopeptide (TPR) domain-containing protein, regulates the expression of genes involved in the production of storage starch and storage proteins in the endosperm via interaction with a bHLH protein16. FLO6, encoding a CBM48 domain-containing protein, is involved in compound granule formation and starch synthesis putatively through interaction with ISA117. Grain Incomplete Filling 1 (GIF1) encodes a cell-wall invertase that regulates sucrose unloading, and the mutation of GIF1 leads to significantly reduced starch content and grain weight18. Mutant for the OsAlaAT1 gene, encoding an alanine-aminotransferase, also displays altered amylose content and amylopectin structure19. The bZIP transcription factor OsbZIP58 can regulate the expression of six starch-synthesizing genes by directly binding to their promoters20. Rice Starch Regulator (RSR1), encoding an AP2/EREBP family transcription factor, negatively regulates the expression of type I starch synthesis genes, and RSR1 deficiency results in enhanced expression of starch synthesis genes in seeds21. Despite all these advances, the identification of novel genes involved in starch synthesis is necessary to understand the molecular network underlying starch synthesis and seed quality formation in rice.

As the substrate for starch synthesis, ADPG needs to be transported into amyloplasts from the cytosol by an ADPG transporter located in the envelope membrane of amyloplasts. Sullivan et al. first proposed BT1 proteins as ADPG transporters in cereal endosperms22. These BT1 proteins belong to the mitochondrial carrier family (MCF)23,24. Although MCF proteins are presumed to be targeted to the mitochondrial inner membrane in plants23, some also occur in peroxisomes, glyoxysomes, plasma membrane, and plastids and are involved in the transport of diverse solutes across these organelle membranes25,26,27,28,29. Transport studies have shown that maize ZmBT1, barley HvNST1, and wheat TaBT1 are able to transport ADPG in counter exchange with ADP30,31,32. OsBT1 corresponds to the rice orthologue32 which may also function to transport ADPG.

In this study, we characterize a white-core endosperm mutant, osbt1. Map-based cloning showed that OsBT1 encodes an ADPG transporter which locates in the amyloplast envelope membrane and functions in the regulation of starch biosynthesis and compound starch granule formation during rice seed development.

Results

The rice osbt1 mutant has a defect in endosperm appearance and grain weight

Numerous mutants with chalky endosperm induced by tissue culture were screened and a mutant with white core was obtained and later named osbt1. The endosperm of osbt1 was opaque and presented a white-core in the centre region of the seed, while the periphery region was translucent compared to that of wild-type (Fig. 1a,b,e,f). Scanning electron microscopy (SEM) analysis of transverse sections indicated that the starch grains (SGs) in both central and periphery regions of the wild-type endosperm cells were polyhedral and densely packed (Fig. 1c,d). The SGs in the endosperm peripheral cells of the osbt1 mutant were similar to those of the wild-type (Fig. 1g), while those in the inner endosperm cells of the mutant were loosely packed with spherical or oval and smooth surface (Fig. 1h). Quantification of seed size showed no significant differences between wild-type and osbt1 in terms of seed length, width, and thickness (Fig. 1i–k), though 1000-grain weight in osbt1 was significantly decreased compared to wild-type (Fig. 1l). Besides the endosperm defect and lower grain weight, osbt1 plants did not display any visible differences from wild-type during the whole growth period (Fig. S1).

Figure 1: Characterization of the osbt1 mutant.
figure 1

(a,e) Appearance of mature seeds of wild-type (WT) (a) and osbt1 (e); (b,f) Cross-sections of mature seeds of WT (b) and osbt1 (f); (c,g) SEM analysis of the periphery area of mature endosperm of WT (c) and osbt1 (g). The periphery areas shown are indicated as blue squares in (b,f); (d,h) SEM analysis of the central area of mature endosperm of WT (d) and osbt1 (h). The central areas shown are indicated as red squares in (b,f); Scale bars: 1 mm in (a,b,e,f); 20 μm in (c,d,g,h); (i–l) Quantification of WT and osbt1 seed size and grain weight including grain length (i), grain width (j), grain thickness (k) and 1000-grain weight (l). Data are given as means ± SD from three replicates. Statistical comparisons were performed using Student’s t-test; all data were compared with WT (*P < 0.05, **P < 0.01).

Full size image

The compound starch granules in osbt1 endosperm cells are abnormal

Semi-thin sections of developing endosperms at 9 days after fertilization (DAF) were prepared to compare the formation of compound starch granules in endosperm cells between osbt1 and wild-type (Fig. 2). The results showed that the amyloplasts in wild-type endosperm cells were composed of several polyhedral, sharp-edged, and easily separable granules, forming well developed compound starch granules. The entire central region of the wild-type endosperm was almost filled with compound starch granules (Fig. 2a,b). However, abundant single starch granules were observed in the centre of the osbt1 mutant endosperm. In addition, some abnormal weakly stained granules were also observed in the mutant (Fig. 2d,e). By contrast, the endosperm peripheral cells of osbt1 were similar to those of wild-type, both having fewer well developed compound starch granules and some immature amyloplasts (Fig. 2c,f).

Figure 2: Abnormal compound granules formation in osbt1 seeds.
figure 2

Semi-thin sections of WT (a–c) and osbt1 mutant (d–f) endosperm at 9 DAF. (a,b,d,e) The central region of endosperm cells; (c,f) The peripheral region of endosperm cells. Stars indicate aleurone cells in (c,f). Arrowheads in (e) indicate smaller, abnormal starch granules in cytosol. Scale bars: 10 μm.

Full size image

Transmission electron microscope (TEM) analysis was further carried out to observe the development of compound starch granules. The amyloplasts in the endosperm peripheral cells of both wild-type and osbt1, and the central endosperm cells of wild-type were filled with about a dozen of polyhedral starch granules and the space between the starch granules was gradually occupied during seed development, to eventually form normal compound starch granules (Fig. 3a–e). However, in the osbt1, the amyloplasts were disintegrated at early developmental stages and the starch granules were dispersive through the central endosperm cells (Fig. 3f–h), which was consistent with the results observed in the semi-thin sections (Fig. 2d,e).

Figure 3
figure 3

Electron micrographs depicting amyloplast development in WT (a–d) and the osbt1 mutant (e,h) endosperms. (a,e) A representative amyloplast in peripheral endosperm cells of WT and osbt1 at 6 DAF; (b–d,f–h) An amyloplast in central endosperm cells of WT and osbt1 at 6 (b,f), 9 (c,g), and 12 DAF (d,h), respectively. White arrows indicate the stroma inside the amyloplast. Bars: 2 μm.

Full size image

Starch physicochemical properties are altered in the osbt1 mutant

An analysis of the physicochemical characteristics of the starch was performed in wild-type and the osbt1 mutant. The total starch and amylose contents were significantly decreased in osbt1 compared to wild-type (Fig. 4a,b). The protein content was lower, while the lipid content was clearly higher than those in the wild-type (Fig. 4c,d). To further examine the fine structure of amylopectin, the chain length distribution of amylopectin was also determined. Compared to wild-type, the proportion of short chains with degree of polymerization (DP) values between 7 and 17 were elevated in the mutant, whereas the proportion of intermediate chains with DP values between 18 and 30 were decreased (Fig. 4e).

Figure 4: Seed properties and starch physicochemical characteristics in the osbt1 mutant.
figure 4

(a–d) Quality trait parameters of the WT and osbt1. Values are means ± SD (n = 3). The asterisks indicate statistical significance between WT and the mutant, as determined by a Student’s t–test (*P < 0.05; **P < 0.01). (e) Differences in the amylopectin chain length distributions between the WT and osbt1. (f) Pasting properties of endosperm starch of WT (blue line) and osbt1 (red line). The viscosity value at each temperature is the average of three replicates. The gray line indicates the temperature changes during the measurements. (g) Gelatinization temperature of endosperm starch. To, Tp, and Tc represent the onset, peak, and conclusion gelatinization temperatures, respectively. All data are presented as means ± SD from three replicates. Two-tailed unpaired t-tests indicate the significant differences: *P < 0.05, **P < 0.01. (h) Gelatinization characteristics of starch from osbt1 mutant seeds. Starch powder was mixed with different concentrations (1–9 M) of urea solution. The most significant difference was observed for 3–4 M urea.

Full size image

The starch pasting properties were analyzed with a rapid visco analyzer (RVA; Fig. 4f). The starch pasting characteristics in wild-type showed distinct parameters represented by a peak viscosity, a fast viscosity increase when the temperature decreased and a high final viscosity value. The viscosity pattern of the osbt1 starch presented a trend similar to wild-type, but was maintained at lower levels, with a final viscosity of 68% of that of wild-type. The thermal gelatinization temperature of endosperm starch from wild-type and osbt1 were analyzed and the result showed that the onset (To), peak (Tp), and conclusion (Tc) temperatures of gelatinization of endosperm starch in the osbt1 mutant were all about 3 °C lower than those in the wild-type (Fig. 4g). The gelatinization properties of the SGs were also examined by measuring solubility in urea solutions15. Various concentrations (0–9 M) of urea solutions were added to powdered starch. The result showed that the gelatinization characteristics of osbt1 starch have most significant differences in 3–4 M urea from that of wild-type (Fig. 4h). In summary, the starch physicochemical characteristics of osbt1 were obviously different from those of the wild-type.

Map-based cloning and complementation of the OsBT1 gene

Map-based cloning was carried out to identify the gene affected in the mutant. An F2 segregating population was created by crossing the osbt1 mutant (japonica) with NJ-11 (indica). The mutated locus was first mapped to a short arm in chromosome 2 between the simple sequence repeat (SSR) markers RM12669 and RM12679. Based on three SSR markers and three newly-developed insertion/deletion (Indel) markers, the osbt1 locus was further narrowed down to a 39.5-kb region between markers Ind3 and Ind10 by using a total of 1,222 recessive homozygous individuals. Seven open reading frames (ORFs) were included in this region (Fig. 5a and Table S1). Sequence analysis revealed that there was a 18-bp repeat insertion in ORF2 (Os02g0202400) in the mutant, which lead to a six amino acid increase in the encoded protein compared to the wild-type sequence (Fig. 5b,c). The three-dimensional structures of OsBT1 in wild-type and osbt1 were predicted using Protein Homology/analogy Recognition Engine V 2.0 (Phyre2) web portal, the result showed that the OsBT1 protein cannot be normally folded due to six amino acids added in osbt1 (Fig. S2).

Figure 5: Map-based cloning, complementation of the osbt1 mutant and phenotype of transgenic lines.
figure 5

(a) Fine mapping of the OsBT1 locus. The OsBT1 locus was mapped to a 39.5 kb region by markers Indel1-3 and Indel1-10 on chromosome 2 (Chr.2), which contained seven predicted genes. The molecular markers and the number of recombinants are shown. cM, centimorgan; ORF, open reading frame. (b) Gene structure and mutation site in osbt1. The cDNA sequence comparison shows the 18 bp insertion in osbt1 leading to six amino acid addition. (c) Identification of the 18 bp insertion between the WT and osbt1 using an Indel marker. M, Marker. (d) Complementation of the osbt1 mutation in transgenic lines (S1–S2) and overexpressing the OsBT1 in osbt1 (OX-1, OX-2) showing the completely restored WT seed appearance (d) and the starch granules morphology (e). RNAi of OsBT1 in WT genotype produce abnormal seed appearance. RNAi seeds (Ri-1, Ri-2) became chalkiness (d), and the starch granules became abnormal (e). Insets in (d) represent the transverse sections of representative grains. Bars: 1 mm in (d); 20 μm in (e).

Full size image

To test whether Os02g0202400 corresponds to the candidate OsBT1 gene, a vector bearing the OsBT1 genomic sequence including its native promoter and a vector bearing the OsBT1 coding region driven by the Actin1 promoter were constructed and introduced into the osbt1 mutant. The expression of OsBT1 in complementary and overexpressing lines was significant higher than that in the wild-type (Fig. S3). Seeds produced by independent positive transgenic T1 lines showed to the wild-type appearance restored (Fig. 5d). The 1000 brown grains weight in complementary and overexpressing lines also restored to wild-type level and the brown grain length, width, and thickness had no significant differences compared to the wild-type (Fig. S4). The SGs in endosperm cells of these positive transgenic lines also presented the polyhedral and densely packed granules morphology, similar to those of wild-type (Fig. 5e). A further validation was obtained by analyzing Os02g0202400 RNAi knock-down plants (Ri) generated in a wild-type genetic background which showed reduced expression of OsBT1 (Fig. S3). Grains harvested from the independent Ri transgenic T1 lines showed the osbt1 mutant appearance (Fig. 5d). The 1000 brown grains weight in Ri lines significantly decreased and the brown grain length, width, and thickness had no significant differences compare to the wild-type (Fig. S4). Moreover, the SGs morphology in the Ri endosperm cells was similar to those of the osbt1 mutant (Fig. 5e). All these results demonstrated that Os02g0202400 is the gene responsible for the osbt1 phenotypes.

Expression pattern and subcellular localization of OsBT1

The temporal and spatial expression patterns of OsBT1 were analyzed by quantitative real-time PCR (qRT-PCR). The results showed that OsBT1 was specifically expressed in seeds and that its relative expression level gradually increased after fertilization, with a maximum expression level at 21 DAF (Fig. 6a). Moreover, GUS histochemical assay showed that the OsBT1 promoter directed the GUS reporter gene to express predominantly in the endosperm but not in roots, stems, leaves, leaf sheaths, and panicles (Fig. 6b–h).

Figure 6: Expression pattern of OsBT1.
figure 6

(a) qRT-PCR analysis of OsBT1 expression level in various tissues and in developing endosperms of the WT and osbt1. Values are means ± SD (n = 3). The asterisks indicate statistical significance between the WT and the mutant, as determined by a Student’s t-test (*P < 0.05; **P < 0.01). (b–h) GUS expression in root (b), steam (c), leaf (d), leaf sheath (e), panicle (f), spikelet (g) and brown rice (h) driven by the OsBT1 promoter. Bars: 2 mm.

Full size image

Subcellular localization of the OsBT1 protein was carried out using two GFP fusions, OsBT1-GFP and osbt1-GFP that were transiently expressed in tobacco protoplasts. The free GFP control protein was widely expressed in the cytoplasm and nuclei (Fig. 7a), whereas OsBT1-GFP and osbt1-GFP both co-localized with the chlorophyll autofluorescence signals of the plastids (Fig. 7b,c). Another vector harboring OsBT1-GFP driven by the maize UBIQUITIN1 promoter (Ubi) was generated and transformed into the Kitaake rice cultivar. Laser scanning confocal microscope showed that OsBT1-GFP was exclusively localized at the amyloplast envelope membranes in endosperm cells (Fig. 7d). All these results demonstrated that OsBT1 localizes to the envelope membrane of amyloplasts, and that the mutation in osbt1 did not affect the protein localization.

Figure 7: Subcellular localization of OsBT1.
figure 7

(a) Free GFP used as a control. (b,c) Full-length coding region fusion protein OsBT1-GFP (b) and osbt1-GFP (c) in front of the GFP. 16 h after transformation, tobacco protoplasts were observed using a confocal laser scanning microscope. Green fluorescence signals (GFP), red chlorophyll autofluorescence (red), bright-field images, and an overlay of green and red signals are shown in each panel. (d) OsBT1-GFP fusion protein was expressed in rice under the control of the maize UBIQUITIN1 promoter and then the localization of OsBT1-GFP in developing endosperm was observed by confocal laser scanning microscope. The outer limit envelope of the amyloplasts is indicated by white arrowhead. Scale bars: 10 μm.

Full size image

The expression of starch synthesis related genes is changed in the osbt1 mutant

The expression pattern of starch synthesis related genes was examined in wild-type and osbt1 endosperms at 6, 9, 12 and 15 DAF via qRT-PCR (Fig. 8). On the whole, results showed that expression of the 14 starch synthesis related genes was higher in osbt1 than that in the wild-type. The expression levels of OsGBSSI, OsISA1, and OsPHOL were always higher in the mutant compared to the wild-type from 6 to 15 DAF. The transcript levels of other genes including two AGP genes (OsAGPL2, and OsAGPS2b), and five amylopectin synthesis related genes (OsSSI, OsSSIIa, OsSSIVb, OsISA2, and OsPUL) were lower at 6 DAF, but were elevated at 9 to 15 DAF compared to wild-type. The expression levels of OsAGPL1 and OsBEIIb were higher in osbt1 than in the wild-type, but no significant differences were detected at 6 or 9 DAF. OsAGPS1 expression was higher in osbt1 than in the wild-type at the early grain filling stage (6–12 DAF), but maintained a similar expression level at 15 DAF. The expression pattern of the OsBE1 gene was opposite to OsAGPS1, which was lower than in the wild-type at the early grain filling stage (6–12 DAF), but higher at 15 DAF (Fig. 8). Additionally, a zymogram analysis indicated that the activities of Pho (Pho1 and Pho2) and SSI were increased in the osbt1 mutant while the activity of SSIIIa has no significant differences in the mutant compared to wild-type (Fig. S5).

Figure 8: Expression profiles of rice starch synthesis genes during seed development in WT and osbt1.
figure 8

Total RNA extracted from developing seeds at 6, 9, 12 and 15 DAF was used for qRT-PCR analysis. The expression of each gene in the 6 DAF WT seeds was set as reference value of 1. All data are shown as means ± SD from three replicates. The asterisks indicate statistical significance between the WT and the mutant, as determined by a Student’s t-est (*P < 0.05; **P < 0.01).

Full size image

Discussion

Starch biosynthesis is a complex metabolic process and the molecular regulatory machinery remains largely unknown. Characterization of new rice endosperm-defective mutants is useful to further elucidate the starch synthesis process. Previous studies have identified at least 29 endosperm-defective mutants, including six chalky endosperm mutants (flo4, flo5, rsr1, gif1, bzip58, and nfyb1)12,18,20,21,33,34, 16 floury endosperm mutants (flo1-flo3, ae, pdil1-1, gpa1 to gpa3, osagpl2–3, flo6, ssg4, osalaat1, flo7, ssg6, pfp, and ss3a/ss4b)16,17,18,19,35,36,37,38,39,40,41,42,43,44,45,46,47, four shrunken endosperm mutants (isa1/sugary1, osagpl2, osagps2, and pho1)8,14,48, and three dull/waxy endosperm mutants (dull1, dull3, and waxy)10,49,50. In addition, the quantitative trait locus Chalk5 was also identified to influence grain chalkiness in rice51. Most of these mutants show decreased amylose content and changed amylopectin structure, and some also exhibit slower grain-filling and decreased 1000-grain weight.

In this study, we isolated and characterized the osbt1 mutant, which exhibits white-core endosperm, abnormal compound starch granules formation, decreased starch content, altered amylopectin structure and physiochemical properties and significant reduced final grain weight. These data indicate that starch synthesis in osbt1 mutant endosperm is seriously affected. In addition, the morphology of the compound starch granules in amyloplasts of some endosperm-defective mutants has also been reported to be abnormal. For example, mutations in SSG4 and SSG6 develop enlarged amyloplasts37,45, while flo6 endosperm contains smaller and abnormal compound granules17. In our case, the amyloplasts were disintegrated at early developmental stages and the starch granules were dispersive (not compound) in the central endosperm cells of osbt1 (Fig. 2). We propose that OsBT1 plays a crucial role in maintaining the stability of the structure of the amyloplast envelope membrane.

Usually, regulators of starch synthesis affect the expression level of genes encoding key enzymes of the starch synthesis pathway. For example, RSR1 negatively regulates the expression of type I starch synthesis genes (i.e. the expression of most starch synthesis genes is higher in the rsr1 mutant and repressed in RSR1-overexpressing lines)21. When another starch biosynthesis regulator, OsbZIP58, was inactivated, the expression of OsAGPS2b, OsAGPL2, OsSSI, OsSSIIIa, OsSSIVb, OsBEIIb, and OsISA2 was distinctly up-regulated, while that of OsAGPL3, OsPHO1, Wx, and SBE1 was down-regulated, leading to a reduced content of total starch and amylose. Also, the fine structure of amylopectin is altered in the osbzip58 mutant20. When Chalk5, which encodes a vacuolar H+-translocating pyrophosphatase influencing grain chalkiness in rice, is over-expressed, the transcript levels of eight genes involved in starch metabolism are significant elevated in seeds at 5 DAF, implying a feedback regulation of Chalk5 expression in controlling grain chalkiness51. In our study, the expression of several key starch synthesis genes was significantly enhanced in the endosperm of osbt1 (Fig. 8). Expression changes of some starch synthesis related genes may alter the structure of amylopectin that further changes the crystalline structure of the starch granule, consequently affecting the physicochemical properties of starch. For example, SSI appears to be primarily responsible for synthesizing shorter chains of amylopectin52, whereas BEIIb functioned in the branching of short chains, as revealed by in vitro analysis53. So the increased short chains in osbt1 may result from enhanced OsSSI and OsBEIIb expression. In our study, the increased short chains lead to a lower gelatinization temperature (Fig. 4g), which was consistent with what was observed in the sbe1 mutant15,54. For these starch synthesis related genes located downstream of OsBT1, mutation of OsBT1 results in lack of enough substrate for starch synthesis, so the elevated expression of starch synthesis related genes likely involve a physiological feedback regulation, which is similar to that in the osbzip58 and rsr1 mutant20,21.

As the predominantly product of photosynthesis, sucrose occurs in source leaves, which is then transported into endosperm cells to be used as carbon source for starch synthesis. Sucrose is divided and converted into ADPG, G6P, and Glc and then imported into amyloplasts by their corresponding transporters (Fig. S6). The glucose 6-phosphate/phosphate translocator (GPT), which is located on the endosperm plastid envelope membranes, is responsible for transporting G6P into plastids55. The plastidic glucose transporter (pGlcT) located in the envelope of chloroplasts has a role in the export of glucose, and also is located in the envelope of endosperm amyloplast with the role to import glucose56. As substrate of starch synthesis in endosperm cells, ADPG needs to be transported into the amyloplast by the ADPG transporter located in its envelope from the cytosol. The designation of BRITTLE-1 (BT1) protein as the ADPG transporter in cereal endosperms was first proposed by Sullivan et al. (1991)22, and then characterized in maize, barley, and wheat30,31,32. In this study, the rice OsBT1 gene was identified using a forward genetics approach. Subcellular localization analysis showed that OsBT1 is located to the envelope membrane of amyloplasts. Thus, we proposed that OsBT1 is the ADPG transporter in rice endosperm cells. To validate our hypothesis, we carried out import studies using [14C] ADPG into Escherichia coli cells harboring recombinant OsBT1, however this assumption was also recently confirmed by Cakir et al.57. Together, these data demonstrate that OsBT1 is the rice ADPG transporter which transports the substrate ADPG into amyloplasts from the cytosol of endosperm cells.

In summary, our results indicate that OsBT1 has an essential role in starch synthesis and compound starch granule formation. Ongoing efforts are still needed to identify new regulatory steps of starch biosynthesis as means to aid breeders and biotechnologists to improve quality in crops.

Materials and Methods

Plant materials and growth conditions

The osbt1 mutant was generated from Oryza sativa japonica cv. Nipponbare induced by tissue culture in vitro. An F2 population was produced by crossing the osbt1 mutant with an indica rice cv. Nanjing11 (NJ-11) for gene mapping. The rice plants were grown under natural conditions at China National Rice Research Institute experimental field plots, in Hangzhou, China.

Microscopy

The brown rice of wild-type and the osbt1 mutant was cut transversely with the back of a knife, and the ruptured transverse surface was coated with gold to prepare samples. The ruptured transverse surface was observed by scanning electron microscope (SEM) which was performed as described previously using a HITACHI S-3400N scanning electron microscope (http://www.hitachi-hitec.com) 33. For analyzing the development of compound starch granules, transverse sections (approximately 1 mm in thickness) of wild-type and osbt1 endosperms at 9 DAF were used to make samples of semi-thin sections. Samples were treated as described by Peng et al.17. Semi-thin sections (800 nm) were stained with I2-KI for 5 s and subsequently examined under a light microscope (Nikon Eclipse 80i; http://www.nikon.com). For the ultrastructure observation of amyloplasts, developing seeds (6–12 DAF) were fixed over 12 h in 0.2 M phosphate buffer (PH 7.2) with 2.5% glutaraldehyde. Samples were treated as described by Takemoto et al.58 and sectioned using an ultramicrotome (Power Tome-XL; RMC, http://www.rmcproducts.com). A transmission electron microscope (H-7650; Hitachi, http://www.hitachi.com) was used for observation.

Physicochemical properties of rice grains

The total starch content of the rice flour was measured according to the manufacturer’s protocol, with a starch assay kit (Megazyme, Wicklow, Ireland, http://www.megazyme.com/). Amylose content was determined as described by Liu et al.59, and the lipid and protein contents in the seeds of osbt1 mutant and wild-type were measured following the method described by Kang et al.33. To determine the starch pasting properties, 3 g of milled rice flour (0.5 mm or less, 14% moisture basis) was transferred into a vessel containing 25 ml of distilled water. The sample was mixed and measured with a Rapid Visco Analyzer (RVA Techmaster, Newport Scientific, Narrabeen, Australia), applying the temperature program described by the America Association of Cereal Chemists, Approved Method 61–02.01 (AACC, 1995). To determine the chain length distribution of amylopectin, 5 mg of rice powder was digested with Pseudomonas amyloderamosa isoamylase (Megazyme) and then analyzed by capillary electrophoresis (PA800 plus pharmaceutical analysis system, Beckman Coulter, USA; http://www.beckmancoulter.com/). For the gelatinization temperature analysis, 5 mg of rice powder was placed in an aluminum sample cup, mixed with 10 μl of distilled water, and sealed, and then the samples were analyzed by a differential scanning calorimeter (DSC 1; METTLER-TOLEDO). The heating rate was 10 °C min−1 over a temperature range of 35 °C to 100 °C. The swelling modes of endosperm starch in urea solution were measured according to a previous report15.

Mapping of the OsBT1 gene

Seeds which have similar white-core mutant phenotype were selected from the F2 population which derived from a cross between the osbt1 mutant and an indica rice cv. NJ-11. Then these white-core seeds were grown until seedling stage for DNA extraction. Twenty-two F2 plants with mutant phenotype and more than 160 polymorphism simple sequence repeat markers evenly distributed over the whole genome were selected to identify markers linked to the OsBT1 locus. Further molecular markers were developed based on the nucleotide polymorphisms in the corresponding regions between the Nipponbare and NJ-11 (Table S2). A total of 1,222 recessive individuals with the mutant phenotype were selected from the F2 population to fine-map the OsBT1 locus.

Plasmid construction and rice transformation

The genomic fragment of the OsBT1 gene, included its native promoter from Nipponbare was amplified by PCR (all primers used for plasmid construction are listed in Table S2) and then subcloned into the binary vector pCAMBIA2300 to generate the complementation vector. A vector bearing the OsBT1 coding region driven by Actin1 promoter was constructed and then sub-cloned into the binary vector pCAMBIA2300 to generate the overexpressing vector. A specific sequence from the OsBT1 cDNA from Nipponbare was cloned into the binary vector LH-FAD2-1390RNAi under the control of the maize UBIQUTIN1 promoter to obtain the RNAi vector. The resulting plasmids were introduced into osbt1 and the wild-type plants, respectively, by the Agrobacterium tumefaciens-mediated method. The genotype of transgenic plants was determined by PCR amplification of the specific transgenic fragment. A 2.1 kb sequence of OsBT1 promoter region was isolated by PCR amplification from Nipponbare and then sub-cloned into the pCAMBIA1305.1-GUS vector. The resulting plasmid was transformed into rice through Agrobacterium-mediated transformation, and the transgenic plants were analyzed by GUS staining assay.

RNA extraction and real-time RT-PCR analysis

Total RNA from roots, stems, leaves, leaf sheaths, panicles, and developing seeds at 6, 9, 12, 15, 18, 21, 24 DAF were extracted using the RNAprep pure Plant kit (TIANGEN Biotech, Beijing, China). 2 μg of total RNA was reverse-transcribed by priming with oligo(dT18) in a 40 μl reaction based on the PrimeScript Reverse Transcriptase kit (TaKaRa, http://www.takara-bio.com). Gene expression was measured by quantitative real-time PCR (qRT-PCR) using the Ubiquitin gene (GenBank accession AF184280) as reference. The qRT-PCR primers used to evaluate the expression of starch synthesis-related genes have been described previously16. PCR was performed with an ABI 7500 Fast Real-Time PCR System using the following program: 95 °C for 30 s, 40 cycles of 95 °C for 5 s, 60 °C for 34 s and 95 °C for 15 s. Changes in gene expression were calculated using the 2−ΔΔCT method.

Subcellular localization of OsBT1 protein

The OsBT1 coding sequences without a termination codon of wild-type and osbt1 were cloned in-frame in front of the GFP coding region in the TR2-GFP vector to create the fusion constructs TR2-OsBT1-GFP and TR2-osbt1-GFP. The two fusion constructs including the empty control vector (TR2-GFP) were introduced into the Agrobacterium strain EHA105 and used to infiltrate N. benthamiana leaves. Plants were then incubated at 28 °C for 2 days before examination according to protocols described previously60. N. benthamiana protoplasts were isolated as described previously45. GFP fluorescent signals were detected using a Zeiss LSM710 confocal laser scanning microscope. The OsBT1 coding sequence without a termination codon was also cloned in-frame in front of GFP in the pCAMBIA1305-GFP vector to generate the 1305-OsBT1-GFP construct under the control of the maize UBIQUTIN1 promoter. The 1305-OsBT1-GFP plasmid was transformed into an indica rice cv. Kitaake. Endosperms from T1 transgenic seeds were used for observation.

Additional Information

How to cite this article: Li, S. et al. OsBT1 encodes an ADP-glucose transporter involved in starch synthesis and compound granule formation in rice endosperm. Sci. Rep. 7, 40124; doi: 10.1038/srep40124 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.