2017 年 67 巻 4 号 p. 393-397
Grain size is a trait that is important for rice (Oryza sativa L.) yield potential. Many genes regulating grain size have been identified, deepening our understanding of molecular mechanisms of grain size determination in rice. Previously, we cloned SMALL AND ROUND SEED 5 (SRS5) gene (encoding alpha-tubulin) from a small and round seed mutant and revealed that this gene regulates grain length independently of the brassinosteroid (BR) signaling pathway, although BR-related mutants set small grain. In this study, we showed that overexpression of SRS5 can promote grain length and demonstrated that the overexpression of SRS5 in BR-related mutants rescued the shortened grain length, which is an unfavorable phenotype in the yield potential of BR-related mutants, while preserving the useful semi-dwarf and erect leaf phenotypes.
Grain size is an important trait that determines rice (Oryza sativa L.) yield potential. Many genes regulating grain size in rice were identified by quantitative trait locus analysis, namely, GS3, GW2, qSW5/GW5, GS5, GW8, GL3.1, TGW6, GW6a, and GL7/GW7 (Fan et al. 2006, Ishimaru et al. 2013, Li et al. 2011, Qi et al. 2012, Shomura et al. 2008, Song et al. 2007, 2015, Wang et al. 2012, 2015a, Wang et al. 2015b, Weng et al. 2008, Zhang et al. 2012). Other genes regulating grain size were cloned from short grain rice mutants, d1, brd1, d2, d11, d61, srs1, srs3, Srs5, Sg1, and tud1 (Abe et al. 2010, Ashikari et al. 1999, Fujisawa et al. 1999, Hong et al. 2002, 2003, Hu et al. 2013, Kitagawa et al. 2010, Nakagawa et al. 2012, Segami et al. 2012, Tanabe et al. 2005, Yamamuro et al. 2000), and recently, GLW7 was identified by a genome-wide association study (Si et al. 2016). These genes are classified into two groups, those controlling cell elongation (GL7/GW7, D2, D11, D61, BRD1, SRS1, SRS3, SRS5, and GLW7) and those controlling cell division (GS3, GW2, qSW5/GW5, GS5, GW8, GL3.1, TGW6, D1, SG1, and TUD1). Among these genes, GW8 protein was reported to directly interact with the promoter of GW7, negatively regulating GW7 expression (Wang et al. 2015a). The GLW7 protein was similarly reported to directly interact with the promoter of SRS5 and positively regulate the SRS5 expression (Si et al. 2016). However, Si et al. (2016) did not present genetic evidence that correlates the higher expression of SRS5 with increased grain length that they observed. Previously, we reported that short grain and dwarf phenotypes of the Srs5 (encoding alpha-tubulin) mutant and brassinosteroid (BR)-related mutants were controlled by independent signal transduction pathways (Segami et al. 2012). Semi-dwarfism in BR-related mutants renders them resistant to lodging and their erect leaves improve their photosynthetic ability. Thus, in barley (Hordeum vulgare), a change in one amino acid due to a single nucleotide substitution in HvBRI1, the homologous gene of D61, creates a mutant with low sensitivity to BR and semi-dwarf stature and erect leaves, which is widely used as a high yielding variety (Chono et al. 2003). In rice, a mild mutation in the BR biosynthetic gene OsDWARF4 that does not affect grain size was reported to increase grain yield under conditions of high density and high fertilizer concentrations but not under normal conditions (Sakamoto et al. 2006). Because of the pleiotropic effect of genes on reduced grain size, BR-related mutants have been used only rarely in rice breeding programs.
In the present study, we produced a plant overexpressing SRS5 to test whether higher expression of SRS5 increases grain length and grain yield potential. Further, since SRS5 promotes grain length via a pathway independent from BR-related genes, we hypothesized that overexpression of SRS5 in the BR-related mutant background will improve grain length, while preserving useful dwarfism and erect leaf phenotypes.
A japonica rice cultivar, Taichung 65, was used as wildtype (WT) plant. The BR biosynthesis mutant d2-2 and BR insensitive mutant d61-2 were obtained by MNU treatment of Taichung 65. All transgenic plants were grown in a closed greenhouse under natural sunlight. Room temperature was maintained at 30°C from 09:00 to 18:00 and 25°C from 18:00 to 09:00.
Production of transgenic plantsTotal RNA was extracted from ripening spikelets of WT plants and used to synthesize cDNA. The coding sequence of SRS5 was amplified by PCR using this cDNA as a template and the primers cacc + SRS51stATG-F: 5′-CACCATGAGGGAGTGCATCTCGAT-3′ and SRS5-3UTR-R: 5′-CGCCAACTAAAGGTCACAAT-3′. The amplicon was subcloned into an entry vector pENTR/D-TOPO (Invitrogen, Carlsbad, CA, USA). The DNA fragment containing the full-length SRS5 cDNA in pENTER/D-TOPO was inserted into a binary vector p2KG by the gateway method described in the manual of the pENTR/D-TOPO cloning kit (Invitrogen, USA). The SRS5 cDNA, controlled by the ubiquitin promoter in p2KG, was introduced into the Agrobacterium tumefaciens strain EHA105 (Hood et al. 1993) by electroporation and transformed into the WT, d2-2, and d61-2 as reported previously (Ashikari et al. 2005). The WT, d2-2, and d61-2 containing empty vectors were used as controls. Over 30 plants were obtained from each regenerated plant. Of those, over 10 plants carrying the transgene were selected by PCR using hygromycin resistance gene (HPT)-specific primers, HPT-L: 5′-CGTATATGCTCCGCATTGGT-3′ and HPT-R: 5′-ATTTGTGTACGCCCGACAGT-3′.
RT-PCR analysisTotal RNA was extracted from ripening spikelets using an RNeasy Mini Kit (QIAGEN, Hilden, Germany), and cDNAs were synthesized from the total RNA using a SuperScript III system (Invitrogen, USA). To quantify the SRS5 mRNA, real-time reverse transcription PCR (RT-PCR) was conducted using SYBR Premix Ex Taq II (TAKARA Bio, Inc., Tokyo, Japan). The primer set RT-alpha-tub-F: 5′-ATGAGGGAGTGCATCTCGAT-3′ and RT-alpha-tub-R: 5′-CAAGATCGACGAAGACAGCA-3′ was used to quantify the expression of Srs5 (GenBank accession no. NC_029266), and another primer set, RT-OsUbiquitin1-F: 5′-CTTGGTCGTGTCCCGTTTC-3′ and RT-OsUbiquitin1-R: 5′-TTCTTCCATGCTGCTCTACCAC-3′, was used to quantify the expression of OsUbiquitin1 (GenBank accession no. NC_029261). A Thermal Cycler Dice Real Time System (TAKARA Bio, Inc., Japan) was used for quantification in real-time RT-PCR.
To confirm whether higher expression of SRS5 increases grain length and grain yield potential, we produced plants overexpressing SRS5 controlled by the OsUbi1 promoter and selected at least 10 transgene-inserted plants by PCR using HPT-specific primers (data not shown). All of the selected plants overexpressing SRS5 (SRS5oe) produced longer grains than those of the control plants that were transformed with empty vectors (Fig. 1A, 1B). Real time quantitative RT-PCR analysis detected a twofold higher expression of SRS5 in SRS5oe plants (Fig. 1C). Overexpression of SRS5 affected grain length, but not grain width and grain thickness (Fig. 1D–1F). The 1000-grain weight of SRS5oe was also higher than that of the WT (Fig. 1G). These results support the presumption stated by Si et al. (2016) that the higher expression of SRS5 promotes longitudinal grain elongation.
Grain phenotypes of the wildtype (WT) overexpressing SRS5. Grain morphology of the WT transformed with empty vector (Empty) (A) and the WT overexpressing SRS5 (SRS5oe) (B). Relative expression of SRS5 was calculated using OsUbi1 as internal control (C). Relative amount of SRS5 was normalized by Empty. Comparison of the grain length (D), grain width (E), grain thickness (F), and 1000-grain weight (G) of Empty and SRS5oe. The analysis was based on 10 Empty and 13 SRS5oe. Error bars indicate standard deviation. The results of the Student’s t-test are indicated above the graph bars. **: p < 0.01, n.s.: not significant (p > 0.05).
To test whether overexpression of SRS5 in the BR-related mutant background can improve grain size while preserving the useful dwarf and erect leaf traits, we overexpressed SRS5 in d2-2 and d61-2 mutants. The overexpression of SRS5 in the WT slightly increased plant height but did not affect the leaf angle (Fig. 2A, 2B, 2G, 2H, 2N, 2O). Although the overexpression of SRS5 in d2-2 and d61-2 resulted in 2.43- and 4.21-fold higher transcription of SRS5, respectively (Fig. 2M), the height of d2-2 plants did not differ significantly from that in the control and it was only slightly increased in d61-2 (Fig. 2C–2F, 2M). The leaf angle in d2-2 and d61-2 overexpressing SRS5 was not affected (Fig. 2I–2L, 2O). These results indicate that overexpression of SRS5 in BR-related mutants slightly increased plant height and preserved dwarfism and the erect leaf phenotype.
Gross morphology and leaf angle of SRS5oe in brassinosteroid (BR)-related mutant background. Gross morphologies of the wildtype (WT) (A), d2-2 (C), and d61-2 (E) background transformed with empty vector used as control (Empty). Gross morphologies of SRS5 overexpression (SRS5oe) in WT (B), d2-2 (D), and d61-2 (F) background. Leaf angle of the flag leaf at ripening stage of Empty in WT (G), d2-2 (I), and d61-2 (K) background. Leaf angle of the flag leaf at ripening stage of SRS5oe in WT (H), d2-2 (J), and d61-2 (L) background. Relative expression of SRS5 was calculated using OsUbi1 as internal control (M). Relative amount of SRS5 was normalized by Empty. Comparison of plant height (N) and leaf angle (O). The analyses were based on 10 Empty and 13 SRS5oe in the WT background, 10 Empty and 13 SRS5oe in the d2-2 background, and 10 Empty and 11 SRS5oe in the d61-2 background. Error bars indicate standard deviation. The results of the Student’s t-test are indicated above the graph bars. **: p < 0.01, *: 0.01 < p < 0.05, n. s.: not significant (p > 0.05).
The grain length increased in d2-2 and d61-2 overexpressing SRS5 (Fig. 3A–3E). These grains were almost the same length as those of WT plants (Figs. 1A, 1C, 3B, 3D, 3E). Further, with increasing grain length, the 1000 grain-weight of SRS5oe in the d2-2 and d61-2 backgrounds was also restored significantly (Fig. 3F). Thus, overexpression of SRS5 in the BR-mutant background is a useful combination that produces normal grain size with semi-dwarf and erect leaf phenotypes.
Grain phenotypes of plants overexpressing SRS5 in brassinosteroid (BR)-related background. Grain morphology of d2-2 (A) and d61-2 (C) background transformed with empty vector used as control (Empty). Grain morphology of SRS5 overexpression (SRS5oe) in d2-2 (B) and d61-2 (D) background. Comparison of grain length (E) and 1000-grain weight (F). The analysis was based on 10 Empty and 13 SRS5oe in the d2-2 background and 10 Empty and 11 SRS5oe in the d61-2 background. Error bars indicate standard deviation. The results of the Student’s t-test are indicated above the graph bars. **: p < 0.01.
In this study, we demonstrated that overexpression of SRS5 in the WT as well as in the BR-related mutant background increases grain length and yield potential. These results are consistent with the hypothesis that higher expression of SRS5 has a potential to promote grain size (Si et al. 2016) and that SRS5 regulates grain length via cell elongation independently from the BR signaling transduction pathway (Segami et al. 2012). Therefore, when novel genes regulating grain size are identified, it is important to analyze genetic interactions between novel and known genes. If the genes are completely epistatic and one masks the effect of the other, their interaction would contribute to our understanding of the molecular mechanisms of grain size regulation. If the genes are not epistatic, or they are partially epistatic, we can utilize these genes simultaneously in the genepyramiding approach, as we demonstrated in this study. Although a number of genes regulating grain size are identified, their genetic interactions are rarely demonstrated. Therefore, it is important to clarify genetic interactions of known genes regulating grain size for cloning of novel genes in future studies.
We thank Dr. H. Kitano for providing seeds of mutant plants and Dr. K. Shimamoto for providing the p2KG vector. This study was supported by the JSPS KAKENHI, Grant Numbers 26712001 and 26450008.
