Mutation of the plastidial alpha-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm

Abstract

Plastidial phosphorylase (Pho1) accounts for approximately 96% of the total phosphorylase activity in developing rice (Oryza sativa) seeds. From mutant stocks induced by N-methyl-N-nitrosourea treatment, we identified plants with mutations in the Pho1 gene that are deficient in Pho1. Strikingly, the size of mature seeds and the starch content in these mutants showed considerable variation, ranging from shrunken to pseudonormal. The loss of Pho1 caused smaller starch granules to accumulate and modified the amylopectin structure. Variation in the morphological and biochemical phenotype of individual seeds was common to all 15 pho1-independent homozygous mutant lines studied, indicating that this phenotype was caused solely by the genetic defect. The phenotype of the pho1 mutation was temperature dependent. While the mutant plants grown at 30 degrees C produced mainly plump seeds at maturity, most of the seeds from plants grown at 20 degrees C were shrunken, with a significant proportion showing severe reduction in starch accumulation. These results strongly suggest that Pho1 plays a crucial role in starch biosynthesis in rice endosperm at low temperatures and that one or more other factors can complement the function of Pho1 at high temperatures.

Figure 1.

Tissue-Specific Expression of Two Types of Phosphorylase in Rice. (A) Activities of two starch Pho isoforms, Pho1 and Pho2, were detected by native-PAGE/enzymatic activity staining analysis in various tissues. The identification of Pho1 and Pho2 was based on our previous study (Yamanouchi and Nakamura, 1992). Lane 1, cv Kinmaze; lane 2, cv Taichung 65 (T65); lane 3, EM583 (a wx mutant derived from the MNU treatment of T65). (B) SDS-PAGE analysis of Pho1 excised from the native-PAGE gel from rice developing endosperm. (C) Primary sequences of a tryptic peptide from the Pho1 protein obtained by liquide chromatography–tandem mass spectrometry analysis and alignment with the unique insertion sequences of Pho1 reported in the database (NCBI nr).

Figure 2.

Subcellular Localization of Pho1 in Developing Endosperm Cells of Rice. Total endosperm proteins and proteins associated with starch granules of mid-developing seeds were resolved by SDS-PAGE and then subjected to immunoblotting for Pho1, BEI, BEIIb, and GBSSI. “Endosperm” denotes the total proteins extracted from the endosperm of developing seeds.

Figure 3.

Effects of pho1 Mutation on Pho1 Content and Activity in Rice Endosperm and on Morphology of the Kernel. (A) SDS-PAGE (top panel) and protein gel blot (bottom panel) profiles of Pho1 protein from the maturing endosperm of five allelic mutant lines induced by independent MNU treatment. Lane 1, EM755; lane 2, EM719; lane 3, EM640; lane 4, EM786; lane 5, EM876. The total proteins extracted from a single brown rice seed for each mutant line and T65 were resolved by SDS-PAGE and analyzed by immunoblotting using anti-Pho1. (B) Zymogram depicting the Pho activities of wild-type T65 and the same five allelic mutant lines as in (A). Plastidial Pho1 activity moved down the gel, while the cytosolic Pho2 activity band is located at the top of the polyacrylamide gel. The activity responsible for the pair of lighter bands below Pho2 activity has not been identified. (C) and (D) pho1-induced changes in grain morphology (C) and grain weight (D). (C) Mature seeds from a single panicle of the pho1 mutant, EM755, grown under field conditions were analyzed. Views of a representative sample of sectioned (top row) and intact (bottom row) seeds are shown. Three distinct grain types were identified based on size and endosperm appearance severely shrunken (lanes 1 to 3), white-core grain (possessing a chalky endosperm core) (lanes 4 to 6), and pseudonormal (lane 7). (D) Distribution of grain weight among the various seed phenotypes. The mean weight for each seed phenotype is shown. White-core grain and pseudonormal seeds were smaller than wild-type T65. Although exhibiting a vitreous appearance much like T65, pseudonormal seeds were ∼8% smaller than white-core seeds. Shrunken seeds weighed only ∼18% of wild-type T65. Nshr, number of shrunken seeds; Npn, number of pseudonormal; Nwcg, number of white-core grains; NT65, number of wild-type T65 seeds.

Figure 4.

Genetic Analysis of the pho1 Mutation in Rice. (A) Gene dosage effects of pho1 mutation on protein amount (top) and activity (bottom) of plastidic α-glucan Pho in rice endosperm. “+” and “p” tentatively represent a wild-type and mutant gene, respectively. Since the endosperm has a gene dosage of 3, the genotypes of the wild type (T65), F1 derived from a cross between a T65 female parent and a EM755 male parent, F1 from a cross between a EM755 female parent and a T65 male parent, and a pho1 mutant (EM755) are represented as “+++,” “++p,” “+pp,” and “ppp”, respectively. Pho1 protein amounts were detected by SDS-PAGE of protein extracted from respective mature seeds. The activity of Pho1 was measured by native-PAGE/enzymatic activity staining analysis. (B) RFLP analysis of the Pho1 gene. The Pho1 digestion products were detected by DNA gel blot analysis using DNA prepared from leaf blades of the pho1 mutant EM755 (p) and of an indica rice cultivar Kasalath (K). (C) Segregation of an RFLP marker in homozygous pho1/pho1 F3 plants. DNA from pho1/pho1 F3 plants (p), which were selected by SDS-PAGE analysis, was digested with ApaI. K represents DNA from the Kasalath cultivar that underwent the same digestion analysis.

Figure 5.

Schematic Representation of Mutations in the Pho1 Gene of BMF136 and EM640. (A) Schematic representation of the Pho1 gene and the locations of single nucleotide substitutions in pho1 mutant lines BMF136 and EM640. ATG and TAG indicate the initiation and termination codons, respectively. Gray boxes, lines, and arrowheads indicate exons, introns, and mutation sites, respectively. a.a., amino acid. (B) Alignment of the Pho1 gene from T65 with the pho1 gene from the BMF136 mutant. The splice site border sequence between the fifth exon and intron (at nucleotide 2626) was disrupted by a G-to-A mutation, which results in intron retension and generates a stop codon 66 nucleotides downstream of the mutation site. (C) Alignment of the Pho1 gene from T65 with the pho1 gene from the EM640 mutant. In EM640, a C-to-T transition occurred at nucleotide 4716 in the tenth exon, and this resulted in a codon change from CAG (Gln) to a stop codon (TAG).

Figure 6.

Effects of pho1 Mutations on Activities of Starch Biosynthetic Enzymes. Native-PAGE/activity staining analysis was performed using enzymes from three developing endosperms of pho1 mutants BMF136 and EM640 and of their wild-type parent, T65. The numbers above the lanes indicate the volume (μL) of crude enzyme extract applied to each gel lane. The activity bands for isoamylase, pullulanase, and Pho1 (A), for branching enzymes, BEI, BEIIa, and BEIIb, and for Pho1 (B), and for SSI and SSIIIa (C) are indicated with arrowheads. Note that the activity bands for BEIIb and Pho1 are not resolved and overlap on the polyacrylamide gel as denoted by the asterisk in (B), as previously described (Yamanouchi and Nakamura, 1992). Note that the staining intensity of this band is mainly due to Pho1 activity, which is absent in BMF136 and EM640.

Figure 7.

Effects of pho1 Mutation on Starch Granules as Viewed by Scanning Electron Microscopy. Starch granules were prepared from plump or shrunken seeds from the pho1 mutant lines BMF136 and EM640 and their wild-type parental cultivar, T65. Bars = 5 μm.

Figure 8.

The Change in Chain Length Distribution of Amylopectin in Rice Endosperm between Various pho1 Mutants and Their Wild-Type Parents (T65 or Kinmaze) as Determined by the APTS-Capillary Electrophoresis Method. The change in the molar amounts (y axis) of specific size chains (x axis) between various pho1 mutants and their wild-type parent, T65 or Kinmaze, are depicted. For pho1 mutants BMF136 and 134, differences in chain length distribution for pseudonormal (pn), shrunken (sh), and white-core grain (wcg for BMF 134 only) from wild-type T65 are shown. BMF136 and BMF134 are near isogenic lines derived from crosses between the parent cultivar, T65, and EM755 and EM719, respectively. BMF10 is a near isogenic line from a cross between cv Kinmaze and EM141. AMF46 is an amylose-free pho1 double mutant line (pho1/pho1 wx/wx) derived from a cross between a waxy mutant (wx, EM583) and pho1 mutant EM755.

Figure 9.

Comparison of Chain Elongation Reaction Products of rPho1 and rSSIIa Using MOS as Primers. Specific MOS products were generated by rPho1 (35 ng protein, closed triangles) in 5 min and by rSSIIa (10.8 μg protein, open circles) in 60 min, using maltohexaose as a substrate. The percentage of product formed for each MOS was plotted on a logarithmic scale. Note that under these conditions similar amounts of maltohexaose were metabolized by rPho1 and rSSIIa (12.9 and 13.5%, respectively) and that no substantial amounts of MOS with DP ≥10 were synthesized by rSSIIa.

Figure 10.

Effects of Temperature on Kernel Morphology during Development of pho1 Mutant Seeds. Mutant plants of EM755 (A) and BMF136 (B) were removed from the field plot at the maximum flowering stage and grown at temperatures of either 30°C (top rows) or 20°C (bottom rows) until they reached maturity. S-shr, severely shrunken grains; shr, shrunken grains; wcg, white-core endosperm grains; pn, pseudonormal grains.

Figure 11.

Model Representing the Possible Role of Pho1 in Starch Biosynthesis in the Rice Endosperm. In this model, Pho1 is suggested to play a major role in one or more events of starch initiation. One possible role for Pho1 is to extend the chain length of the initial primer so that it is acted upon by SSs. This involvement of Pho1 in the starch initiation process would account for the shrunken phenotype and loss of starch in pho1 mutants grown at 20°C. A hypothetical X factor is also proposed that can partially complement the function of Pho1 in starch initiation at 30°C. The X factor is suggested to have low activity or be present at low amounts at 20°C but to be fully active at 30°C. Hence, at 20°C, the loss of Pho1 and low net activity of X factor results in a high incidence of shrunken seeds, while at higher temperature the X factor can partially complement Pho1 role in starch initiation leading to the production of a high percentage of seeds that accumulate near normal starch levels.