FLOURY ENDOSPERM16 encoding a NAD-dependent cytosolic malate dehydrogenase plays an important role in starch synthesis and seed development in rice

Abstract

Starch is the most important form of energy storage in cereal crops. Many key enzymes involved in starch biosynthesis have been identified. However, the molecular mechanisms underlying the regulation of starch biosynthesis are largely unknown. In this study, we isolated a novel floury endosperm rice (Oryza sativa) mutant flo16 with defective starch grain (SG) formation. The amylose content and amylopectin structure were both altered in the flo16 mutant. Map-based cloning and complementation tests demonstrated that FLO16 encodes a NAD-dependent cytosolic malate dehydrogenase (CMDH). The ATP contents were decreased in the mutant, resulting in significant reductions in the activity of starch synthesis-related enzymes. Our results indicated that FLO16 plays a critical role in redox homeostasis that is important for compound SG formation and subsequent starch biosynthesis in rice endosperm. Overexpression of FLO16 significantly improved grain weight, suggesting a possible application of FLO16 in rice breeding. These findings provide a novel insight into the regulation of starch synthesis and seed development in rice.

Keywords: energy supply; floury endosperm mutant; grain weight; malate; redox regulation; rice; starch synthesis.

Figure 1

Phenotypes of the flo16 mutant. (a) Images of mature seeds of wild type (left) and flo16 (right). Scale bar, 2 mm. (b) Cross sections of the mature seeds of wild type (left) and flo16 (right). Scale bar, 1 mm. (c) One‐week‐old seedlings of wild type (left) and flo16 (right). Scale bar, 4 cm. (d) Wild type (left) and flo16 (right) plants at the booting stage. Scale bar, 20 cm. (e) Grain weights of wild type and flo16 at various stages post‐fertilization. Grain weight indicates the weight of 100 dehulled dry grains, values are means ± SD, n = 3. (f) 1000‐grain weights of wild type and flo16. Values are means ± SD, n = 5. (g) Grain size comparisons between wild type and flo16. Values are means ± SD, n = 10. Asterisks indicate the statistical significance between the wild type and mutants determined by Student's t‐tests (*P < 0.05; **P < 0.01).

Figure 2

Abnormal starch grain (SG) formation in flo16 endosperm. (a‐d) Semi‐thin sections of the wild type (a, b) and flo16 (c, d) endosperms at 12 days after flowering (DAF). (b) and (d) are enlargements of the boxed areas in (a) and (c), respectively. Small and scattered SGs are displayed in lower left of (b, d). Triangles indicate compound starch grains in (b) and (d); a asterisk indicates a single starch grain in (d). (e–j) SEM of endosperm of wild type (e‐g) and flo16 (h‐j). Scale bars, 40 μm in (a, b), 20 μm in (c, d), 1 mm in (e, h); 200 μm in (f, i); 10 μm in (g, j).

Figure 3

Properties of grains and physicochemical characteristics of starch in the flo16 mutant. (a‐e) The contents of total starch (a), amylose (b), protein (c), lipid (d) and sucrose (e) were proportions of grain dry weight in wild type and flo16. Values are means ± SD, n = 3. (f) Differences in amylopectin chain length distributions between the wild type and the flo16 mutant. (g) Swollen volumes of wild‐type and flo16 starch in urea solutions at various concentrations (0–9 M). Values are means ± SDs, n = 3. (h) Pasting properties of endosperm starch of the wild type and flo16 mutant. Grey line indicates temperature changes during measurements. Asterisks indicate the statistical significance between the wild type and mutant determined by Student's t‐tests (*P < 0.05; **P < 0.01).

Figure 4

Map‐based cloning of FLO16. (a) Fine mapping of the FLO16 locus. The FLO16 locus was mapped to an 88 kb region by markers 188–21 and 188–2 on chromosome 10 (Chr.10), containing six predicted genes. Numbers of recombinants are indicated below the map. (b) The flo16 mutant has a 4 bp deletion in the fifth exon of Os10 g0478200. White boxes indicate untranslated regions; black boxes indicate exons; lines indicate the introns. (c‐f) Functional complementation lines of flo16 restore normal seed appearance. Complemented seeds became translucent (c), and SGs were restored to normal (d‐f). W‐3, W‐4 and W‐5 are representative positive transgenic lines. Scale bars, 2 mm in (c), 25 μm in (d‐f).

Figure 5

Spatial expression patterns of FLO16. (a) Expression levels of FLO16 in various tissues. Developing seeds were sampled at 6, 9, 12, 15 and 18 days after flowering (DAF), and other tissues were sampled at heading. The value of Actin I mRNA was used as an internal control for data normalization. Values are means ± SD, n = 3. (b‐g) GUS activity in panicles (b), spikelet before heading (c), during heading (d), and at 3 DAF, and grains at 3 (e), 6 DAF (f) and 9 DAF (g), leaf (h), node (i, j). (j) Cross section of the node in (i). Scale bars, 10 mm in (b), 2 mm in (c‐j). (k) Subcellular localization of FLO16. Free GFP served as a control (lower panel). The fusion construct FLO16‐GFP was expressed in the rice protoplasts (upper panel). Green fluorescence signals, red chlorophyll autofluorescence, and bright field and merged images are shown in each panel. Scale bars, 10 μm.

Figure 6

Responses of the metabolic pathways to deletion of FLO16 activity. (a) Native PAGE profiles of malate dehydrogenase (MDH) activities in developing endosperms at 6 and 9 DAF. (b) Malate contents in developing endosperms. (c) NAD‐MDH activity in wild‐type and flo16 developing endosperm. (d) Sucrose contents in developing and mature endosperm. (e) Starch contents in wild‐type and flo16 endosperm. All values are means ± SD, n = 3. Asterisks indicate statistical significance between the wild type and mutant, determined by Student's t‐tests (*P < 0.05; **P < 0.01).

Figure 7

Gene expression and protein activity analyses of starch synthesizing enzymes. (a) Expression levels of starch synthesis‐related enzymes in developing endosperm 10 DAF. Data represent ratios of expression levels in flo16 to that of wild type. Values are means ± SD, n = 3. (b) AGPase activities in developing endosperm of wild type and flo16. Values are means ± SD, n = 3. Asterisks indicate the statistical significance determined by Student's t‐tests (*P < 0.05; **P < 0.01). (c) Activity bands for SSI, and SSIII. (d) Activity bands for PHO1 and PHO2.

Figure 8

Influence of exogenous malate and sucrose on starch synthesis enzymes in wild‐type and flo16 endosperm. (a) AGPase activities in wild‐type and flo16 developing endosperms at 9–12 DAF treated with malate. (b) Expression levels of starch synthesis enzymes in the developing endosperm at 9–12 DAF incubated in malate. (c) The Expression levels of starch synthesis enzymes in developing endosperm at 9–12 DAF treated with sucrose. Values are means ± SD, n = 3.

Figure 9

Effects of endosperm‐specific FLO16 overexpression on grains. (a, b) Images of hulled (a) and dehulled (b) seeds of the recipient and transgenic lines. EO‐27, EO‐29 and EO‐31 are independent transgenic lines. Scale bars, 1 cm. (c) Enhanced expression levels of FLO16 in overexpression endosperms. Expression level in Zhonghua 11 was considered as sample control. Values are means ± SDs, n = 3. (d–f) Size comparisons between the wild‐type and transgenic seeds. Values are means ± SDs, n = 10. (g) 1000‐grain weights of overexpression lines. Values are means ± SDs, n = 3. Asterisks indicate the statistical significance between the wild type and the mutant, determined by Student's t‐tests (**P < 0.01).