Discussion
The grain size and weight of rice are complex traits, which involves in grain development and filling process (mainly the sucrose metabolism and starch biosynthesis), and is controlled by both the genetic and the environmental factors. In last decades, several research groups have provided genetic and molecular evidence that grain size and filling processes are controlled by many genes (see review by Zuo & Li, 2014; Sun et al., 2018). All these researches provide rice breeders opportunities to improve rice grain yield by manipulating these genes. However, it still has a long way to go before understanding the complex regulatory networks involving grain development and filling processes. The heterotrimeic G protein (hereafter G protein) is well known to play important roles in plant growth and development (Botella, 2012; Sun et al., 2018). Recently, several rice genes/QTLs encoding G protein subunits have been shown to control grain size and shape of rice, including: qPE9-1/DEP1 , GS3 , RGB1 and RGA1(Oki et al., 2005; Huang et al., 2009; Zhou et al. , 2009; Mao et al., 2010; Liu et al., 2018; Miao et al., 2018; Sun et al., 2018).qPE9-1/DEP1 encoding a Gγ subunit, positively regulates grain size (Huang et al., 2009; Zhou et al., 2009); GS3, another Gγ subunit, negatively regulate grain size and shape (Mao et al., 2010; Sun et al., 2018); RGB1 and RGA1 , encoding Gβ and Gα subunit, respectively, positively regulates grain size and shape (Fujisawa et al. , 1999; Utsunomiya et al., 2011; Sun et al., 2018). However, the functions and molecular mechanisms of G protein to regulate grain development and starch biosynthesis are largely unknown.
Rice grain filling is a very complicated process that involves photoassimilate translocation from photosynthetic sources (i.e., leaves and stem-sheaths), sucrose degradation, transmembrane transport and starch synthesis in the grains (Liang et al., 2001; Lü et al., 2008; Tang et al., 2009). Approximately twenty enzymes/proteins have been reported to be involved in these biochemical processes (Zhu et al. , 2003; Tetlow et al., 2004; Ohdan et al., 2005). The results of our lab and several other groups showed that sucrose synthase (SUS), invertase (INV), ADP-glucose pyrophosphorylase (AGP), soluble starch synthase (SSS) and grannuel-bound starch synthase (GBSS), branching enzyme (BE) and debranching enzyme (DBE) play key roles in the regulation of sucrose metabolism and starch biosynthesis (Liang et al., 2001; Lü et al. , 2008; Tang et al., 2009; also see reviewed by Tetlow et al., 2004; Keeling & Myers, 2010; Zeeman et al., 2010). The grain-filling process is also a highly regulated process in which both genetic and environmental factors are involved. It is well known that plant hormones play important roles in grain development and filling process (Liang et al. , 2001; Yang et al., 2006; Tang et al., 2009; see also reviewed by Basunia & Nonhebel, 2019). Understandably, rice grain size, not like the wheat grains, is physically limited by the spikelet hull, of which the final size is almost determined upon flowering. In this sense, the endosperm cell proliferation and elongation as well as the accumulation of starch and other storage compounds after flowering are of crucial importance for final grain yield and quality. Our results showed that grain size, including the length, width and thickness of grains reduced, and as a result, the final grain weight and starch content reduced significantly inRGB1 knock-down lines (Figure 1). This reduction was mainly due to the delay of caryopsis development and the lower starch accumulation at the early stage of grain filling. As we known, the rice grain filling process is in fact the process of sucrose metabolism and starch biosynthesis occurred in endosperm cells, in which many enzymes/proteins are involved. So, it is expected that the expression levels of genes encoding enzymes catalyzing sucrose metabolism and starch biosynthesis during grain filling are closely related to the starch accumulation and grain weight. In RGB1 knock-down lines, the expression of genes encoding sucrose metabolism and starch biosynthesis was either initiated later or much lower at the early filling stage, which can well explain the results observed here (Figure 1 and 2) and imply that suppression ofRGB1 expression could down-regulate the expression of genes encoding enzymes catalyzing sucrose metabolism and starch biosynthesis. However, we still unknown the molecular mechanisms of RGB1regulation on the expression of these genes.
Recently, several research groups have reported that an accumulation of auxin immediately before the starch biosynthesis in rice grains (Abu-Zaiton et al., 2012). Furthermore, application of exogenous auxin also had a positive effect on starch accumulation. These results suggested that auxin may involve in the regulation of starch biosynthesis in rice grains. RNA-seq assay and auxin quantitation also showed a great difference in the expression of auxin biosynthesis related genes and IAA contents in grains between RGB1 knocking down lines and wildtype during filling stage. The expression of endosperm-specific genes for auxin biosynthesis was down-regulated and endogenous IAA content significantly reduced in the grains ofRGB1 knocking down lines (Figure 3). The assumption that RGB1 involvement in the regulation of starch biosynthesis is through changing auxin homeostasis of grains was further validated according to the results of exogenous application of IAA on starch accumulation and the expression of sucrose metabolism and starch biosynthesis related genes during grain filling stage (Figure 3).
Auxin biosynthesis in higher plants is catalyzed by a large number ofTARs (encoding tryptophan aminotransferase) and YUCs(encoding indole-3-pyruvate mono-oxygenases) with differing patterns of spatiotemporal expression, which allows for multiple roles. Different genes may be responsible for the auxin production in different time and/or in different tissues, and therefore, play various roles in regulating plant growth and development. In rice, tissue-specific expression of these genes showed that OsTAR1 , OsYUC9 andOsYUC11 genes expressed highly in the endosperm cells (Figure 4a), suggesting these three genes might play important roles in controlling auxin biosynthesis in grains. However, only the expression of OsYUC11 was well correlated with the grain IAA content (Figure 4c). Based on these results, we hypothesize that OsYUC11 were mainly and especially responsible for the auxin biosynthesis in endosperm cells of rice and other auxin biosynthesis related genes may participate in separate signaling processes.
It is clear that the delay of caryopsis development and the lower starch accumulation and grain weight in RGB1 knocking down lines are due to the lower auxin content in grains caused by the lower expression ofOsYUC11 during grain filling stage. Furthermore, OsYUC11also plays a positive regulatory role in the starch biosynthesis pathway by up-regulating the expression of several sucrose metabolism and starch biosynthesis related genes. Collectively, our study suggests thatOsYUC11 is of crucial importance in regulating grain development and starch biosynthesis by controlling auxin content during grain filling stage. However, we still need direct evidence to verify that OsYUC11 is the key enzyme in controlling the level of IAA in rice grain. We tried to knock out OsYUC11 in WYJ8, but failed to obtain the regeneration plantlet from callus, likely because of disrupting the balance of auxin and cytokinin supplemented in medium. Next, we will modify the ratio of auxin and cytokinin in regeneration medium to create the null mutant of OsYUC11 .
In eukaryotes, transcription of genes is regulated by various transcription factors. Our results showed that OsYUC11 promoter may interact with several families of transcript factors, including MADS, MYB, CCAAT, etc. (Table S1), suggesting that the regulation ofOsYUC11 expression is very complicated and involves signaling networks. However, if we considered the results of RNA-seq analysis that showed the differences of the expression of various transcription factors between RGB1Ri line and WYJ8 plant during grain filling stage, and the results of tissue-specific analysis of gene expression of these transcript factors, it is reasonable to assume thatOsNF-YB1 might involve in the regulation of the expression ofOsYUC11 . Our present results provide biochemical evidence to support the conclusion that OsNF-YB1 was crucial importance in regulating OsYUC11 expression, and finally the auxin content in grains. OsNF-YB1 has been reported to activate the expression of sucrose transporters and waxy gene, and finally regulate the endosperms development. Knockout of OsNF-YB1 led to defective grains with chalky endosperms and significantly decreased grain weight (Bai et al. , 2016; Bello et al. , 2018). Our results provide a new insight that OsNF-YB1 regulates grain development not only through directly activating the sucrose metabolism and starch biosynthesis genes, but also through controlling the auxin accumulation.
In summary, according to our present results, a working model is proposed to illustrate the roles of RGB1 in regulating grain development and grain filling process. RGB1 may positively regulate\souts expression of transcription factor OsNF-YB1 , which activates theOsYUC11 transcription by interacting with its promoter, and leads to an increase in auxin level. The increased auxin then stimulates the expression of sucrose metabolism and starch biosynthesis related genes in endosperm cells, as a consequence, the biosynthesis of starch and grain filling. However, the mechanism underlying how RGB1 regulates the expression of OsNF-YB1 remains to be further studied (Figure 6).