Skip to main content
NCBI home page
Plant Biotechnology Journal logoLink to Plant Biotechnology Journal
. 2022 Dec 29;21(3):621–634. doi: 10.1111/pbi.13976

MATE transporter GFD1 cooperates with sugar transporters, mediates carbohydrate partitioning and controls grain‐filling duration, grain size and number in rice

Changhui Sun 1, † ,✉, Yang Wang 1,2, † , Xiaorong Yang 1, † , Lu Tang 3, † , Chunmei Wan 1, † , Jiqing Liu 1, † , Congping Chen 1, Hongshan Zhang 1, Changcai He 1, Chuanqiang Liu 1, Qian Wang 1, Kuan Zhang 1, Wenfeng Zhang 1,2, Bin Yang 1, Shuangcheng Li 1, Jun Zhu 1, Yongjian Sun 1, Weitao Li 1, Yihua Zhou 2, Pingrong Wang 1,✉, Xiaojian Deng 1,✉
PMCID: PMC9946139  PMID: 36495424

Summary

More than half of the world's food is provided by cereals, as humans obtain >60% of daily calories from grains. Producing more carbohydrates is always the final target of crop cultivation. The carbohydrate partitioning pathway directly affects grain yield, but the molecular mechanisms and biological functions are poorly understood, including rice (Oryza sativa L.), one of the most important food sources. Here, we reported a prolonged grain filling duration mutant 1 (gfd1), exhibiting a long grain‐filling duration, less grain number per panicle and bigger grain size without changing grain weight. Map‐based cloning and molecular biological analyses revealed that GFD1 encoded a MATE transporter and expressed high in vascular tissues of the stem, spikelet hulls and rachilla, but low in the leaf, controlling carbohydrate partitioning in the stem and grain but not in the leaf. GFD1 protein was partially localized on the plasma membrane and in the Golgi apparatus, and was finally verified to interact with two sugar transporters, OsSWEET4 and OsSUT2. Genetic analyses showed that GFD1 might control grain‐filling duration through OsSWEET4, adjust grain size with OsSUT2 and synergistically modulate grain number per panicle with both OsSUT2 and OsSWEET4. Together, our work proved that the three transporters, which are all initially classified in the major facilitator superfamily family, could control starch storage in both the primary sink (grain) and temporary sink (stem), and affect carbohydrate partitioning in the whole plant through physical interaction, giving a new vision of sugar transporter interactome and providing a tool for rice yield improvement.

Keywords: carbohydrate partitioning, grain‐filling duration, MATE transporter, sugar transporter, rice

Introduction

All living organisms require organic carbon to grow and survive. Many eukaryotes, including humans, cannot produce carbon supply by themself. More than half of the world's food is provided by cereals, as humans obtain >60% of daily calories from grains (Alexandratos and Bruinsma, 2012; Julius et al., 2017). Assimilating, transporting and distributing carbohydrates from leaves to sink tissues are fundamental to cereals' yield (Durand et al., 2018; Julius et al., 2017; Zhang and Turgeon, 2018). Consequently, understanding carbohydrate partitioning and its genetic regulation mechanism could lead to a breakthrough in crop yield, including rice (Oryza sativa L.), one of the most important food sources (Wei et al., 2017).

Sucrose, the primary carbohydrate form, is produced by photosynthetic source tissues such as leaf blades and transported to other tissues via a continuous mature phloem transport system. This system comprises phloem sieve elements connected end to end, combining the whole plant organs. Generally, sucrose is firstly loaded into the phloem in source tissues, then transported via the long‐distance vascular channels and finally unloaded into sink tissues (Julius et al., 2017; Scofield et al., 2007b; Zhang et al., 2007). The primary sink tissue is grain, where the sucrose is mainly stored as starch. In the grain‐filling process, sucrose flows directly along the pedicel's central vasculature distributing in the whole panicle, then through the rachilla and pericarp dorsal vascular bundle and finally into the endosperm for starch synthesis (Scofield et al., 2007b; Zee, 1972). Besides, temporary starch granules (SGs) stored in the stem can be acted as a temporary sink. These SGs can convert to sucrose again, return to the long‐distance pathway and be transported into the panicle's filling grain. Up to 24–27% of the carbohydrate in the grain originates from stem starch reserves (CocK and Yoshida, 1972). However, the sugar transportation mechanism remains unclear in rice.

In most monocots, the long‐distance transport of sucrose may be involved in the apoplastic pathway that requires sugar transporters SUT (sucrose transporter) and Sugars Will Eventually be Exported Transporters (SWEET) to facilitate sugars movement between cells (Eom et al., 2011; Julius et al., 2017). SUTs load sucrose against its prevailing concentration gradients into the phloem. By comparison, SWEETs appear to function as facilitators that promote transport down the sucrose gradients (Mathan et al., 2021b; Wu et al., 2018). In the genome of rice, there are five SUT genes. OsSUT1, OsSUT3, OsSUT4 and OsSUT5 encode plasma membrane‐localized proteins, and OsSUT2 is localized on the tonoplast membrane (Aoki et al., 2003; Eom et al., 2011; Siao et al., 2011). OsSUT1 is expressed after heading in the filling grain, leaf sheath, stem, nucellus, vascular parenchyma tissue and the nucellar projection (Matsukura et al., 2000; Scofield et al., 2007b). OsSUT1 might take part in sugar assimilation along the long‐distance pathway, from the leaf to the base of the filling grain (Furbank et al., 2001; Hirose et al., 2010; Matsukura et al., 2000; Scofield et al., 2007b). OsSUT2 is highly expressed in leaf mesophyll cells, emerging lateral roots, pedicels of fertilized spikelets and cross‐cell layers of grain coats (Aoki et al., 2003; Eom et al., 2011). The ossut2 mutant exhibits a growth retardation phenotype of tiller number, plant height, 1000‐grain weight and so on (Eom et al., 2011). OsSUT2 is most likely involved in sucrose transport across the tonoplast from the vacuole to the cytosol (Eom et al., 2011; Siao et al., 2011). OsSUT3 transcripts likely accumulate in sink leaves and exposed internode‐1, but not in the enclosed region (Aoki et al., 2003; Scofield et al., 2007b). OsSUT4 shows preferential expression in sink leaves and stems (Aoki et al., 2003; Scofield et al., 2007b). OsSUT5 is expressed most in sink leaves (Aoki et al., 2003; Scofield et al., 2007b; Sun et al., 2010). Among 21 OsSWEET genes in rice, OsSWEET4 and OsSWEET11 have major effects on caryopsis development (Li et al., 2022), and ossweet4 is even defective in grain filling (Sosso et al., 2015). The double‐knockout mutant ossweet11 ossweet15 exhibits unfunctional endosperm and no accumulated starch in the grain (Ma et al., 2017; Yang et al., 2018). However, we still know little about how these sugar transporters work until now.

Multidrug and toxic compound extrusion transporters (MATEs) are a category of cation antiporters in most organisms and constitute one of the largest transporter families (Omote et al., 2006; Takanashi et al., 2014). MATE transporters are involved in various physiological functions in the plant, transporting a broad range of substrates such as organic acids, plant hormones and secondary metabolites (Takanashi et al., 2014; Wang et al., 2016). Members of this family mediate the export of organic substrates with the coupled exchange of Na+ or H+, and are driven by the electrochemical gradient across the membrane (Kuroda and Tsuchiya, 2009; Omote et al., 2006). There are at least 50 MATE family members in rice (Yokosho et al., 2016), but none has been reported to take part in sugar transportation.

Here, we reported a prolonged grain filling duration mutant 1 (gfd1) exhibiting a long grain‐filling duration, less grain number per panicle and bigger grain size without changing grain weight. Map‐based cloning and molecular biological analyses revealed that GFD1 encoded a MATE transporter and expressed high in vascular tissues of the stem, spikelet hulls and rachilla, but low in the leaf, controlling carbohydrate partitioning in the stem and grain but not in the leaf. GFD1 proteins are partially localized on the plasma membrane and in the Golgi apparatus and interact with two sugar transporters, OsSWEET4 and OsSUT2. Genetic analyses further showed that GFD1 might control grain‐filling duration through OsSWEET4, regulate grain size together with OsSUT2 and synergistically modulate grain numbers per panicle with OsSUT2 and OsSWEET4. Together, our study characterized GFD1 as an important regulator in carbohydrate partitioning and revealed a molecular mechanism involving grain‐filling duration, grain size and number per panicle, which may provide a tool for rice yield improvement.

Results

Prolonged grain‐filling duration mutant gfd1 shows pleiotropic phenotypes

By screening the mutant library derived from the chemical mutagenesis of G46B (an indica variety), we identified a prolonged grain‐filling duration mutant gfd1 (Figure 1a). Though the heading date of gfd1 (74 days) was about 4 days later than G46B (70 days), the mature period of gfd1 was more than 15 days longer than G46B. To eliminate the differences in the mature period due to later heading, we examined the grain weight every 3 days from the first day after fertilization (DAF) to the mature stage in gfd1 and G46B. The results showed that the period from fertilization to the largest grain weight day was more than 10 days longer in gfd1 than in the wild type (WT) (Figure 1g,h). The reduced grain weight of gfd1 in the early ripening stage indicated its grain‐filling rate was slowed down (Figure 1g,h). However, the mature grain weight of gfd1 (45 DAF) was comparable with that of the WT due to a longer grain‐filling duration (Figure 1h).

Figure 1.

Figure 1

Open in a new tab

Phenotypes characterization of wild‐type (WT) and gfd1 mutant. (a) Plant architecture of WT and gfd1 plants at the grain‐filling stage. Scale bar, 10 cm. (b) Panicle architecture of WT and gfd1 mutant. Scale bar, 3 cm. (c) Decreased grain number per panicle in gfd1. Scale bar, 5 mm. (d) Caryopsis characterization of WT and gfd1. (i, v) Appearance comparison of caryopsis. (ii, vi) Transverse sections of caryopsis. (iii, iv, vii, viii) Scanning electron microscopy analysis of transverse sections of WT (iii, iv) and gfd1 (vii, vii) endosperms. Scale bars, 3 mm (i–ii, v–vi); 20 μm (iii–iv, vii–vii). (e, f) Comparison of grain length (e) and grain width (f) in WT and gfd1. Scale bars, 3 mm. (g) Fresh caryopsis of WT and gfd1 at various development stages. DAF, days after fertilization. Scale bar, 3 mm. (h) Weight statistics of fresh caryopsis. The maximum weight values of WT and gfd1 were indicated with red triangles, respectively. (i–l) Sucrose, glucose, fructose and starch content of WT and gfd1 grains. Data in (h, −l) are shown as means ± SD from three biological replicates. Asterisks indicate statistically significant differences by a Student's t‐test (*P < 0.05; **P < 0.01).

Grain filling is highly correlated with grain quality. We thus compared the grain appearance between the mutant and WT. The unshell mature gfd1 grains and the cross‐sections appeared brown‐pink and opaque, whereas those of WT looked translucent with partial chalk (Figure 1d). Scanning electron microscopy of the endosperms further exhibited that the SGs in gfd1 endosperm were loosely packed, resulting in large sizes of SGs with irregular shapes; those in the WT were quite small and tightly organized (Figure 1d). Chemical examination revealed that gfd1 grains contained higher sucrose, glucose and fructose, but lower starch proportions than the WT (Figure 1i–l). Furthermore, gfd1 displayed additional abnormal agronomic traits, such as bigger grain, less grain number per panicle and a reduced yield (Figures 1b,c,e,f, S1). Therefore, GFD1 mutation compromises grain starch synthesis in the grain‐filling process and negatively regulates plant yield.

Map‐based cloning of GFD1

For genetic analysis, an F2 population was constructed by crossing the mutant gfd1 with its WT variety G46B. In this population, the longer grain‐filling duration and normal grain‐filling duration plants were segregated as 145:476 (P > 0.05), suggesting that a single Mendelian factor controlled the grain‐filling duration of gfd1 (Figure 2a). Moreover, the grain width and length, and the grain number per panicle were perfectly cosegregated with the grain‐filling duration (Figure 2b–d). These results suggested that the pleiotropic variations of gfd1 might be caused by the same locus.

Figure 2.

Figure 2

Open in a new tab

Map‐based cloning of gfd1. (a–d) Distribution statistics of plant numbers in maturity period (a), grain length (b), grain width (c) and the number of grains per panicle (d) in F2 populations (G46B × gfd1). Maturity period was determined by heading date and grain‐filling duration. In gfd1, there was no significant change in the heading date, and the maturity period was mainly dependent on grain‐filling duration. (e) GFD1 was primarily mapped to the short arm of chromosome 3 between InDel marker C1 and simple sequence repeat (SSR) marker RM251. (f) The interval was narrowed to 121 kb between RM3716 and C7. (g) Delimitate GFD1 to a 7.3‐kb region between single‐nucleotide polymorphism (SNP) marker S1 and InDel marker C6 using 1860 homozygous B4F2 plants. (h) Gene structure and mutation site of GFD1. The candidate gene LOC_Os03g12790 comprises two exons and one intron, and its encoding protein contains two domains (MATE1 and MATE2). A single‐nucleotide G‐to‐A substitution at 674 bp at the coding region in gfd1 was indicated with the red triangles.

We applied a map‐based cloning method to identify the responsible gene GFD1 (Figure 2). Sixty‐five individuals with extremely slower grain filling were chosen from the F2 population crossing by gfd1 and ZH11. GFD1 was first located in the short arm of chromosome 3 between the simple sequence repeat (SSR) marker RM218 and InDel marker C1 (Figure 2e). Then, 1 SSR and 9 InDel markers possessing polymorphisms between the two parents were developed (Table S1). Using 232 slower grain‐filling individuals in the B2F2 population (gfd1 × ZH11), we further narrowed down the locus of GFD1 to a 121 kb region between RM3716 and C7 (Figure 2f).

For hyperfine mapping, we constructed a B4F2 generation population (gfd1 × ZH11) and designed 3 single‐nucleotide polymorphism (SNP) markers. Thousand eight hundred sixty individuals with a slower grain‐filling phenotype were carefully selected. The localization interval was finally narrowed to 7.3 kb between S1 and C6 markers, containing only one open reading frame LOC_Os03g12790 (Figure 2g,h). LOC_Os03g12790 encodes a MATE transporter, one of the largest transporter families in the plant (Takanashi et al., 2014; Wang et al., 2016). Full‐length amplification and sequencing of the genome and cDNA of LOC_Os03g12790 showed that gfd1 had a single base mutation (G to A) in the second exon, which changed the amino acid Glu to Lys in the MATE1 domain (Figure 2h).

Complementation and CRISPR/Cas9 knockout of GFD1

The full‐length 6.2 kb genome fragment of LOC_Os03g12790 containing its promoter was jointed to the pCAMBIA1300 vector for the complementation test. Because the indica variety G46B is recalcitrant to transform, we introduced the complementation constructor into the near‐isogenic line of gfd1 in the ZH11 background (NILZH11). We obtained 20 independent‐positive T0 lines holding a normal grain‐filling duration as the WT (Figure 3). In T1 generation lines, normal and slower grain‐filling individuals were segregated, and the positive individuals were all cosegregated with the normal grain filling phenotype. Besides, other agronomic trait variations in NILZH11, such as larger grain size and less grain number per panicle, were also restored to the WT level (Figures 3, S2).

Figure 3.

Figure 3

Open in a new tab

Function verification of GFD1 by complementation and CRISPR/Cas9 knockout. (a) Plant architecture of the NILZH11 (NIL of gfd1 in ZH11 background) and gfd1‐C (complementary lines) at the grain‐filling stage. gfd1‐C1, gfd1‐C2 and gfd1‐C3 are three independent transgenic lines. Scale bar, 10 cm. (b) Panicle architecture of NILZH11 and gfd1‐C1. Scale bar, 3 cm. (c) Comparing grain number per panicle between NILZH11 and gfd1‐C1. Scale bar, 5 mm. (d, e) Comparison of grain length (e) and grain width (f) in NILZH11 and gfd1‐C1. Scale bars, 3 mm. (f) Caryopsis comparison at 9, 15, 21, 27, 33 and 39 day after fertilization (DAF) in (i) the NILZH11 (left) and gfd1‐C1 (right), (ii) ZH11 (left) and KO1 (right), and (iii) Nipponbare (left) and KO3 (right). Scale bars, 3 mm. (g–i) Caryopsis weight at various stages of grain filling in NIL and gfd1‐C (h), ZH11 and KO1 (j), and Nipponbare and KO3 (i). The maximum weight is indicated with a red triangle in (h, i). Data in (h, i) are given as means ± SD from three biological replicates. Asterisks indicate statistically significant differences by a Student's t‐test (*P < 0.05; **P < 0.01).

Further validation was achieved by knockout GFD1 in ZH11 (KO1 and KO2) and Nipponbare (KO3) using the CRISPR/Cas9 system. Two sgRNA targeting sequence sites on the second exon of GFD1 for CRISPR/Cas9 cleavage were applied, respectively (Figure S3). We obtained more than eight independent knockout lines in the two backgrounds. All the knockout lines showed similar phenotypic mutations of gfd1, longer grain‐filling duration, bigger grain size and less grain number per panicle (Figures 3, S4–S8). Taken together, LOC_Os03g12790 was considered to be the responsible gene for the mutation of gfd1.

Expression pattern and subcellular localization

Temporal and spatial expression analysis by quantitative RT‐PCR (qRT‐PCR) showed that GFD1 was constitutively expressed in all tested organs (Figure 4a). However, the transcript levels were higher in the booting stem, heading stem, heading leaf sheath and panicle but lower in leaves. The stem at the heading stage possessed the highest transcription level, indicating the importance of GFD1 in the stem of this stage.

Figure 4.

Figure 4

Open in a new tab

Expression pattern and subcellular localization. (a) GFD1 transcript levels in various tissues detected by qRT‐PCR. Rice Actin1 was used as an internal control. SL, seedling leaf; BL, booting leaf; BS, booting stem; HS, heading stem; HLS, heading leaf sheath; HP, heading panicle. Data are given as means ± SD from three biological replicates. (b) GUS staining in leaf (i), stem (ii, iii), spikelet (iv, v) and developing caryopsis (vi) driven by GFD1 promoter. (vii) Transverse section of caryopsis at 30 day after fertilization (DAF). Ra, rachilla; Es, endosperm; AL, aleurone layer; DV, dorsal vascular bundle. Scale bars, 5 mm (i); 3 mm (iv–vi); 1 mm (ii, iii); 100 μm (vii). (c) GFD1 mRNA accumulation pattern detected by in situ hybridization in transverse sections of the stem at the booting stage (i–iii) and the caryopsis at 9 DAF (iv–vi) in G46B. VBS, vascular bundles; DV, dorsal vascular bundle. Scale bars, 100 μm. (d) Subcellular localization of GFD1. GFP was fused into C‐ or N‐ terminal of GFD1, respectively. OsRAC3‐mRFP was used as a plasma membrane marker. The scale bars represent 10 μm in rice protoplast cells and 20 μm in tobacco leaves.

Moreover, a vector with a GUS reporter gene driven by the GFD1 promoter was constructed and transformed into ZH11. Histochemical analysis revealed that strong GUS signals were detected in the stem, spikelet hull and rachilla (Figure 4b). Consistent with the qRT‐PCR results, there were weak GUS signals in the leaf, suggesting a feeble function of GFD1 in the leaf. During the grain filling and mature process, GUS signals were dynamically changed in the caryopsis: gradually increased from 6 DAF to 12 DAF, maintained at a high level from 12 DAF to 21 DAF, then decreased from 21 DAF to 27 DAF and finally shrunk in the pericarp dorsal vascular bundle of the caryopsis which was also reconfirmed under a microscope (Figure 4b, vi and vii). In situ hybridization also showed that GFD1 preferred to express in the pericarp and dorsal vascular bundle of caryopsis (Figure 4c, iv–vi). Besides, both GUS histochemical (Figure 4b, iii) and in situ hybridization (Figure 4c, i–iii) analyses showed that GFD1 could express in sclerenchyma cell, parenchymal cell and vascular bundles in the stem. Consequently, GFD1 was mainly expressed in the vital organization for substance transport in stem and grain.

Plant MATE transporters are located in different compartments, such as plasma membrane, vacuolar membrane and Golgi complex (Upadhyay et al., 2019; Wang et al., 2016). The GFD1 protein was predicted as a MATE transporter containing two domains (MATE1 and MATE2) with no apparent signal peptide (www.ncbi.nlm.nih.gov, www.cbs.dtu.dk). For subcellular localization detection, GFD1‐GFP and GFP‐GFD1 constructors driven by the CaMV35S promoter were transiently transformed into rice protoplasts. As shown in Figure 4d, GFP signals of both GFD1‐GFP and GFP‐GFD1 fusion proteins might partly distribute on the plasma membrane. Therefore, OsRAC3, a plasma membrane‐localized protein, was used as a plasma membrane marker (Chen et al., 2010). GFD1‐GFP was co‐transformed with OsRAC3‐mRFP into rice protoplasts and Nicotiana benthamiana leaves. The results showed that the fluorescences of OsRAC3‐mRFP and GFD1‐GFP could be merged as white light in the plasma membrane (Figure 4d). Besides, uneven spot‐like distribution of fluorescent signals of GFD1 was also observed in the cytoplasm in rice protoplasts and N. benthamiana leaves.

By analysing the spot‐like signals' characteristics, GFD1 is not likely to localize in the nucleus, chloroplast and endoplasmic reticulum (Nelson et al., 2007). As a result, the punctate signals of GFD1 are similar to the Golgi apparatus. Therefore, we co‐transformed GFD1‐GFP with Golgi marker Man49‐mRFP or ERD2‐mRFP, respectively (Li et al., 2009a; Montesinos et al., 2014). The results showed that the punctate signals of GFD1 could partially merge with Man49 or ERD2 as white light (Figure S9). However, some punctate signals are still out of the Golgi apparatus, suggesting the complex subcellular localization of GFD1. In summary, the GFD1 protein could partially localize on the plasma membrane and in the Golgi apparatus.

gfd1 disordered carbohydrate distribution in stem and grain but not leaf

In rice, the primary form of the assimilated carbon is sucrose, which is transported starting from the leaf (source), through long‐distance transport in the stem (flow), and finally reaches the vascular trace of seed (sink) (Krishnan and Dayanandan, 2003; Sturm and Tang, 1999). And then, the sucrose is hydrolysed in the extracellular space into monosaccharides and transported into the endosperm by sugar transporters such as OsSWEET4 for starch synthesis (Wang et al., 2008). Each of the steps above changes could affect the grain‐filling results. Therefore, to probe the physiological basis of the gfd1 mutant phenotype, we detected the carbohydrate distribution of gfd1 in leaf, stem or grain under four development stages, including the heading stage before pollination, 9 DAF, 15 DAF and after the grain‐filling stage (Figure 5).

Figure 5.

Figure 5

Open in a new tab

Carbohydrate partitioning was disordered in gfd1. (a, b) Starch and sucrose content in leaf, stem and grain at the heading stage, 9 day after fertilization (DAF) and 15 DAF. (c, d) Glucose and fructose content in grain at the heading stage, 9 DAF and 15 DAF. (e) Starch content in leaf and stem after the grain‐filling stage. At the grain‐filling stage is about 5 days after heading. After, the grain‐filling stage is about 47 days after heading and with mature grain. Data in (a–e) are given as means ± SD at least three independent assays. DW means dry weight. Asterisks indicate statistically significant differences by Student's t‐test (*P < 0.05; **P < 0.01). (f) Starch granules (SGs) were detected in the stem of WT and gfd1 at the grain‐filling stage. Scanning electron microscopy images showed the cross‐sections of internode‐1 (i, ii, iii, iv), internode‐2 (v, vi, viii, ix), internode‐3 (xi, xii, xiv, xv). Starch iodine staining images showed the cross‐sections of internode‐2 (vii, x) and internode‐3 (xiii, xvi). Scale bars, 20 μm. (g) SGs were detected in the stem of WT and gfd1 after the grain‐filling stage. Scanning electron microscopy images showed the cross‐sections of internode‐1 (i, ii, iii, iv), internode‐2 (v, vi, vii, viii) and internode‐3 (ix, x, xi, xii). Scale bars, 5 μm. (h) Ultrastructure of chloroplasts in mesophyll cells at and after the grain‐filling stages. There was no significant difference in SG accumulation in the chloroplasts between WT and gfd1. Both WT and gfd1 exhibited more SG accumulation at the grain‐filling stage and less SG accumulation after the grain‐filling stage. Red arrows indicate SGs. Scale bars, 2 μm.

In the immature grain of gfd1, the starch content was significantly decreased in the 9 and 15 DAF (Figure 5a). However, sucrose, glucose and fructose were increased (Figure 5b–d). It was consistent with the results in the mature grain of gfd1 (Figure 1i–l). More sugar but less starch in the immature and mature grain of gfd1 suggested that GFD1 could participate in some steps of sugar transport and thus affect the starch synthesis rate in the grain.

In the gfd1 stem, the sucrose flow was sharply decreased after pollination (Figure 5b), along with a significant starch accumulation, especially in 9 and 15 DAF (Figure 5a). As described previously, starch accumulation tended to be higher in the basal but lower in the apical internodes in rice (Bihmidine et al., 2015; Julius et al., 2017). Scanning electron microscope data demonstrated that this conclusion still worked in our study. At the grain‐filling stage (about 5 days after heading), internode‐2 and internode‐3 but not internode‐1 displayed a higher SGs density in gfd1, which was further confirmed by iodine‐stained (Figure 5f). As a result, sucrose in gfd1 stem is more likely to convert into starch instead of being transported to grain at the grain‐filling stage. However, such SG accumulation was significantly reduced in WT and gfd1 after the grain‐filling stage (about 47 days after heading, Figure 5g), which was reconfirmed by the starch content measurement (Figure 5e), implying that most of the SGs in the stem had been transported into seeds after complete filling.

Interestingly, the carbohydrate distribution in the leaf was not the same as that in the stem, since the starch or sucrose contents were unchanged under the four development stages in gfd1 (Figure 5a,b,e). In the leaf, starch accumulation usually occurs in the chloroplast. Thus, we detected the accumulation of SGs in the chloroplast of the leaf in WT and gfd1 by transmission electron microscopy. Consistent with the measurement results of starch content in the leaves (Figure 5a,e), there was no significant difference in SG accumulation between WT and gfd1 at or after the grain‐filling stage (Figure 5h). Accordingly, GFD1 mediates sugar transport in both stem and grain but not in the leaf in rice.

GFD1 affects the expression of starch synthesis and sugar transporter genes

For transcriptional basis exploration of gfd1, starch synthesis genes (OsAGPL2, OsBE1, OsBEIIb, OsSSI, OsSSIIa, OsSSIIIa and OsGBSSI) and sugar transporters (OsSUTs and OsSWEETs) were selected and detected in the booting stem, and 3 DAF stem and grain (Durand et al., 2018; Li et al., 2017). The qRT‐PCR results showed that lesions in GFD1 altered the expression balance of starch synthesis and sugar transport genes. In gfd1, most sugar transporters' expression levels were down‐regulated in the stem (Figure 6b), but increased in the grain 3 DAF (Figure 6d). Meanwhile, consisting of the higher starch content in the stem and lower starch synthesis rate in the grain in gfd1, all the starch synthesis genes were up‐regulated in the stem (Figure 6a), while most of them were down‐regulated in the grain (Figure 6c).

Figure 6.

Figure 6

Open in a new tab

Expression analysis of starch synthesis and sugar transporter genes. OsAGPL2 is an AGP large subunits gene; OsBE1 and OsBEIIb are starch branching enzyme genes; OsSSI, OsSSIIa and OsSSIIIa are amylopectin synthesis genes; OsGBSSI is an amylose starch synthase gene; OsSUT1‐5, OsSWEET4, OsSWEET5, OsSWEET11, OsSWEET13, OsSWEET14 and OsSWEET15 are sugar transporter genes. (a) Expression analyses of starch synthesis genes in the stem at the booting stage. (b) Expression analyses of sugar transporter genes in the stem at 3 days after fertilization (DAF). (c, d) Expression analyses of starch synthesis (c) and sugar transporter (d) genes in the panicle at 3 DAF. Data are given as the means ± SD from at least three replications. Asterisks indicate statistically significant differences between WT and gfd1 by Student's t‐test analysis (*P < 0.05; **P < 0.01).

GFD1 interacts with OsSUT2 and OsSWEET4 in vitro and in vivo

GFD1 affects carbohydrate distribution, and its expression pattern and subcellular localization are similar to some sugar transporters (Chen et al., 2012; Hirose et al., 2010; Matsukura et al., 2000; Scofield et al., 2007a,2007b). Thus, we wanted to determine whether GFD1 could interact with sugar transporters for sugar transportation. Then, we designed several experiments. Firstly, we screened the candidate sugar transporters in OsSUTs and OsSWEETs by yeast two‐hybrid assays. The results revealed that the MATE1 but not MATE2 domain of GFD1 could interact with OsSUT2 and OsSWEET4 in yeast (Figures 7a, S10, S11). Moreover, we also amplified the mutant MATE1 gfd1 fragment from gfd1, inserted it into the AD vector and co‐transformed it with OsSUT2 and OsSWEET4, fusing BD vectors. The results showed that the mutant MATE1 gfd1 domain could not interact with OsSUT2 and OsSWEET4 in yeast, suggesting that a single mutant from Glu to Lys could block the interaction between GFD1 and OsSUT2/OsSWEET4 (Figure 7a).

Figure 7.

Figure 7

Open in a new tab

GFD1 could interact with OsSUT2 and OsSWEET4. (a) Yeast two‐hybrid assays (Y2H) showed that MATE1 could interact with OsSUT2 and OsSWEET4. Serial dilutions (1, 1/10, 1/100) of yeast transformant were plated onto the medium of QDO/X‐α‐gal for further screening. Co‐transformation of pGADT7‐T (T) and pGBKT7‐53 (53) was used as the positive control. The transformant of pGADT7‐T (T) and pGBKT7‐lam (lam), pGADT7 (AD) and pGBKT7 (BD) were used as the two negative controls. DDO, SD/‐Leu/‐Trp; QDO, SD/‐Leu/‐Trp/‐Ade/‐His. (b) LCI assay of GFD1 with OsSUT2 (up) or OsSWEET4 (down) in tobacco leaves. Coloured scale bars indicate the luminescence intensity in counts per second (cps). (c) Bimolecular fluorescence complementation assay analysis. nYFP‐GFD1 with cYFP‐OsSWEET4, cYFP‐GFD1 with nYFP‐OsSWEET4 were co‐transformed into tobacco leaves. nYFP, N‐terminal YFP; cYFP, C‐terminal YFP. nYFP‐GFD1 with cYFP, cYFP‐GFD1 with nYFP served as negative controls. Scale bars, 20 μm.

Next, we tested whether GFD1 interacted with OsSUT2 or OsSWEET4 in plant cells. LCI assays were applied in N. benthamiana leaves (Hu et al., 2019). Apparent LUC activity was observed when cLUC‐GFD1 was co‐transformed with OsSUT2‐nLUC or OsSWEET4‐nLUC (Figure 7b). Moreover, bimolecular fluorescence complementation assay (BiFC) assays showed that OsSWEET4 and GFD1 could be associated with the plasma membrane in N. benthamiana leaf cells (Figure 7c). Consequently, our data suggested that GFD1 could interact with OsSUT2 and OsSWEET4 in vitro and in vivo.

Genetic relationship analysis of GFD1 and OsSWEET4 / OsSUT2

GFD1 could interact with OsSWEET4 and OsSUT2, mediate carbohydrate partitioning and affect grain‐filling duration, grain size and number per panicle in rice. These results lead us to find out the genetic relationship between GFD1 and OsSWEET4/OsSUT2. Therefore, we generated single knockout mutants ossweet4 and ossut2, and double‐mutant gfd1 ZH11 ossweet4 and gfd1 ZH11 ossut2 in ZH11 and gfd1 ZH11 (KO2‐1) by CRISPR/Cas9 technology (Figure S12). Then, we planted ZH11, gfd1 ZH11, ossweet4, ossut2, gfd1 ZH11 ossweet4 and gfd1 ZH11 ossut2 in the experimental fields in Wenjiang, Chengdu (Figure 8).

Figure 8.

Figure 8

Open in a new tab

Genetic analysis of GFD1, OsSUT2 and OsSWEET4. (a) Comparison of seed setting rate of ossweet4 and gfd1 ossweet4. Scale bars, 2 cm. (b, e, f and g) Comparison of grain numbers per panicle of ZH11, gfd1 ZH11, ossut2, gfd1 ZH11 ossut2, ossweet4 and gfd1 ZH11 ossweet4. Scale bars, 2 cm. (c, d, h and i) Comparison of grain length and grain width of ZH11, gfd1 ZH11, ossut2 and gfd1 ZH11 ossut2. Scale bars, 3 mm. (j) Caryopsis comparison at 9, 15, 21, 27, 33 and 39 day after fertilization (DAF) in ZH11 (up, left), gfd1 ZH11 (up, right), ossut2 (down, left) and gfd1 ZH11 ossut2 (down, right). Scale bars, 3 mm. (k) Caryopsis weight at various stages of grain filling in ZH11, gfd1 ZH11, ossut2 and gfd1 ZH11 ossut2. Details of KO4‐3 (ossut2), KO5‐7 (gfd1 ZH11 ossut2), KO6‐1 (ossweet4), KO7‐3 (gfd1 ZH11 ossweet4) are showed in Figure S12. The letters a, b, c, d and e in (b–d) indicate significant differences at P < 0.05 according to one‐way ANOVA test with Tukey correction. Asterisks in (k) indicate statistically significant differences Student's t‐test analysis (**P < 0.01).

It was reported that OsSWEET4 acted as a switch controlling sugar transport from the maternal phloem into the endosperm at the basal endosperm transfer layer (Sosso et al., 2015). Our ossweet4 mutant was also defective in seed filling. Moreover, the double‐mutant gfd1 ZH11 ossweet4 showed a similar inferior grain‐filling variation with ossweet4, indicating that the regulating function of GFD1 might be OsSWEET4‐dependent in sugar transport and grain‐filling process (Figure 8a). Additionally, ossweet4 and ossut2 could reduce grain number per panicle in ZH11 and gfd1 ZH11 background (Figure 8b,e–g), suggesting that the three transporters coordinate controlled grain number per panicle in rice.

Consistent with the previous report, ossut2 mutant exhibited a growth retardation phenotype (Eom et al., 2011). The single‐mutant ossut2 did not change the grain‐filling duration (Figure 8k), while the double‐mutant gfd1 ZH11 ossut2 showed a longer grain‐filling duration as gfd1, and the seed at 33 DAF was greener than ossut2 (Figure 8j). Furthermore, a substantial reduction in grain weight of double‐mutant gfd1 ZH11 ossut2 was found during the whole grain‐filling process (Figure 8k). As the double‐mutant gfd1 ZH11 ossut2 showed a medial grain size between gfd1 ZH11 and ossut2, GFD1 and OsSUT2 were antagonists in grain size regulation (Figure 8c,d,h,i). Therefore, the less grain weight was completely dependent on the grain‐filling ability loss in the double‐mutant gfd1 ZH11 ossut2. These results suggested that, in the gfd1 background, ossut2 severely affected grain‐filling ability, implying an enhanced relationship between GFD1 and OsSUT2 in the grain‐filling process.

Discussion

MATE transporter GFD1 controls the grain‐filling duration in rice

In rice, grain filling is essential to determine the accumulation rate and duration of storage compounds in the grain and also has a crucial influence on rice's final yield and quality (Nagata et al., 2015; Takai et al., 2005; Upadhyay et al., 2019; Wang et al., 2008). So far, only a few grain‐filling genes have been identified, but most of these genes do not remarkably affect the grain‐filling duration (Hirose et al., 2002; Liu et al., 2019; Wang et al., 2008; Wei et al., 2017; Xiong et al., 2019). Besides, grain‐filling duration is also an important agronomic trait that defines maturity duration for seasonal and regional adaptation after the heading date in rice (Chen et al., 2022; Fang et al., 2019; Sun et al., 2014, 2021; Zhao et al., 2018; Zhou et al., 2021). On the contrary, MATE transporters belong to one of the largest transporter families and are involved in various physiological and developmental functions (Takanashi et al., 2014; Upadhyay et al., 2019; Wang et al., 2016). In our study, though the grain weight of gfd1 was nearly the same as that of the WT, the grain‐filling duration was largely prolonged. Hence, we build a relationship between the grain‐filling duration and a MATE transporter GFD1.

GFD1 might tune the transport rate of OsSWEET4

When arrived at the grain, the apoplastic sucrose is split by cell wall‐bound invertases (such as OsGIF1) into glucose and fructose, which are subsequently uptake into the endosperm by hexose transporters. OsSWEET4, which was proven as a hexose transporter at the basal endosperm transfer layer and localized on the plasma membrane, is required for this uptake (Sosso et al., 2015). Our study verified that GFD1 could also express in the basal endosperm transfer layer and localized on the plasma membrane. Besides, ossweet4 exhibited defective grain‐filling phenotypic variation, while gfd1 showed longer grain‐filling duration and lower grain‐filling rate. Therefore, GFD1 and OsSWEET4 might interact for hexose transportation together. We designed four experiments and proved our assumption. First, yeast two‐hybrid assays showed that GFD1 could interact with OsSWEET4. Second, the interaction was further confirmed by the LCI assay. Third, BiFC results showed that GFD1 interacted with OsSWEET4 in the plasma membrane. Finally, the double‐mutant gfd1 ZH11 ossweet4 showed a similar grain‐filling defect just as ossweet4. Therefore, it seems that OsSWEET4 is epistatic to GFD1. OsSWEET4 acts as an on/off switch of grain filling, while GFD1 is the regulator of this switch through physical interaction with OsSWEET4, controlling the valve of the switch and tuning the flow rate of hexose in the basal endosperm transfer layer.

GFD1 involves in stem starch reserves

The plant stem can act as an intermediate storage sink organ of starch. In rice, 24–27% of grain carbohydrate originates from stem starch reserves (CocK and Yoshida, 1972), making stem starch a significant carbohydrate source in the grain‐filling process (Scofield et al., 2007b). Interestingly, cultivated rice Nipponbare utilizes stem starch more than Oryza australiensis at the grain‐filling stage and produces a higher grain yield (Mathan et al., 2021a; Wang et al., 2020). However, the mechanisms for sucrose and starch assimilating into and out of the stem are rarely understood (Scofield et al., 2007b). Our research on GFD1 provided some clues. Compared with WT, starch content and SG density of gfd1 was higher in the stem but not in the leaf, especially at the grain‐filling stage, suggesting that GFD1 could participate in starch accumulation in the stem. However, though the starch content was still a little higher in gfd1 stem after the grain‐filling stage, most SGs in WT and gfd1 stems disappeared after the completion of grain filling, suggesting that the SGs accumulated in gfd1 stems might be due to its slow grain‐filling rate.

GFD1 may be an enhancer of OsSUT2

A previous study showed that ossut2 significantly reduced sugar exportability and likely interfered with sucrose translocation to sink organs (Sun et al., 2008). Furthermore, OsSUT2 could be involved in unloading an enormous amount of sucrose being received in the stem (Wang et al., 2020). In our study, the single‐mutant ossut2 did not exhibit grain‐filling defect phenotypes. However, the double‐mutant gfd1 ZH11 ossut2 exhibited severer grain‐filling variants, a prolonged grain‐filling duration and a substantial reduction in grain weight during the whole grain‐filling process. Compared with ossut2, the grain size in double‐mutant gfd1 ZH11 ossut2 is larger. Still, the grain weight is significantly lower, which is different from both ossut2 and gfd1 single mutants, suggesting that functional GFD1 could partly compensate for the loss of function OsSUT2. Therefore, GFD1 and OsSUT2 may act on the same pathway, and GFD1 may be an enhancer of OsSUT2 in grain‐filling process.

Some clues on sugar transporter interactome

Membranes contain thousands of proteins whose biochemical or physiological functions have not been identified experimentally. The transport activity of many transporters depends on the interactions of membrane proteins (Lalonde et al., 2010). However, most of the putative protein–protein interactions were previously unknown. Identifying genetic and molecular interactions is a hopeful way to identify membrane protein functions (Boone et al., 2007; Jones et al., 2014; Jonikas et al., 2009; Lalonde et al., 2008). MATEs are initially classified as members of the Na+‐ or H+‐coupled transporters in the major facilitator superfamily (MFS), which is one of the two largest membrane transporter families on the earth (Pao et al., 1998; Upadhyay et al., 2019). Plant sugar SWEETs and SUTs proteins are initially attributed to MFS (Marger and Saier, 1993; Wipf et al., 2021). Interestingly, we provided evidence that MATE transporter GFD1 could interact with OsSUT2 and OsSWEET4 and assist them in sugar transport. Though GFD1 and OsSWEET4 are associated with the plasma membrane, the subcellar locations of OsSUT2 and GFD1 are different. GFD1 protein was localized on the plasma membrane and in the Golgi apparatus, while OsSUT2 was tonoplast‐localized.

In eukaryotic cells, proteins transport across organelles such as the plasma membrane, the Golgi apparatus and vacuoles by membrane trafficking, which is essential for normal cellular functions (Shimizu et al., 2021). In our subcellular localization experiments, we found a large variation in the number and size of the punctate signals of GFD1. As some punctate signals are localized in Golgi, others are still out of the Golgi, suggesting other possible subcellular localizations of GFD1. These punctate apparatuses might translocate GFD1 to the tonoplast where OsSUT2 is localized. Exploring this cross‐organelle interaction mechanism will give a new vision to sugar transporter interactome research.

GFD1 takes part in multiple regulatory pathways

Interacting with OsSWEET4 and OsSUT2 could not fully explain all the varied phenotypes of gfd1. More regulation pathways need to be explored in future studies. Sucrose metabolism plays a pivotal role in synthesizing essential compounds for plant growth. On the contrary, sugars could also act as signals to regulate meristem activity, flowering, inflorescence branching, tillering and so on (Ruan, 2014). The disordered sugar distribution in gfd1 might also affect sugar signals distributing for plant development regulation. Additionally, we cannot rule out that GFD1 might involve other substances' transportation.

Materials and methods

Plant materials and growth conditions

The gfd1 mutant was obtained by screening the EMS mutagenesis library of indica rice cv. Gang46B (G46B). F2 population (gfd1 × G46B) was used for genetic analysis. F2 population of gfd1 × ZH11 (Zhonghua11, a japonica variety), and B2F2 and B4F2 populations by backcrossing with ZH11 were used for gene mapping.

Rice seedlings for generating protoplast cells were grown in the dark at 28°C, and N. benthamiana plants were grown in a culture room at 23°C with a 16‐h light and 8‐h dark photoperiod. Other plant materials were grown under normal conditions in Sichuan Agriculture University's experimental fields in Wenjiang, Chengdu, China (Wang et al., 2010).

Grain‐filling duration investigation

To obtain fertilized spikelets on the same day, we cut off the pollinated spikelets in the early morning and unpollinated spikelets at dusk in 1 day. The spikelets still in the panicle were used for grain‐filling duration determination. For WT and gfd1, fresh grains were randomly collected every 3 days since 3 DAF. For complementation, knockout of GFD1 analysis and the double‐mutant gfd1 ZH11 ossut2 analysis, the fresh grains were randomly collected every 6 days since 9 DAF. Then, the fresh caryopses were weighed. Three biological replicates were performed with no less than 30 grains per replicate.

Microscopy

Scanning electron microscopy was performed according to a modified method using an Apreo S scanning electron microscope (Thermo Fisher, Waltham, Massachusetts, United States) (Kang et al., 2006). For SG detection in the internode, paraffin sections were made as described by Li et al. (2009a,2009b). The paraffin sections were stained with I2‐KI and observed under a light microscope (Nikon DS‐U3; Nikon, Minato City, Tokyo, Japan) (Peng et al., 2014). For transmission electron microscopy analysis, leaves were treated in 3% glutaraldehyde and then fixed in 1% osmium tetroxide. After dehydrating in a gradient acetone series, the leaf sections were embedded in Epon812 medium for thin sectioning. Uranyl acetate and Reynolds' lead citrate were used to stain the thin sectioning, and an H‐600 IV transmission electron microscope was used to assess the picture (Hitachi, Chiyoda City, Tokyo, Japan) (Li et al., 2015).

Measurement of sugar and starch

Mature grains of WT and gfd1 were shelled and ground into powder, respectively. Leaves (the top‐three leaves), stems or developing grains of WT and gfd1 were harvested at the heading stage, 9 DAF, 15 DAF, and after the grain‐filling stage, and then dried at 65°C for heat‐inactivation. The above samples' total starch content was measured according to the manufacturer's protocol with a starch assay kit (BC0705; Solarbio, Fengtai District, Beijing, China). Sucrose, glucose and fructose contents were determined using soluble sugar assay kits BC2465, BC2505 and BC2455 (Solarbio, Fengtai District, Beijing, China). All analyses were repeated with three biological replicates.

Map‐based cloning

For map‐based cloning, F2 population crossing by gfd1 and ZH11, and subsequent B2F2, B4F2 generation populations were constructed. A total of 2157 slower grain‐filling homozygous individuals were collected in 3 years (2011–2014) in Wenjiang, Chengdu, China. We screened and got more than 100 polymorphic SSR markers distributed over the 12 chromosomes between gfd1 and ZH11. Insertion/deletion (InDel) and SNP markers were developed based on nucleotide polymorphisms between Nipponbare and indica rice cv 9311 reference genomes in the corresponding regions. All the primers are listed in Table S1.

Vector construction and rice transformation

For complement of gfd1, the whole genomic fragment of WT (G46B) GFD1 containing 3‐kb native promoter was inserted into pCAMBIA1300 vector (Table S1) and then transformed into the near‐isogenic line of gfd1 in the ZH11 background (NILZH11).

According to the previous method, CRISPR/Cas9 vector construction was performed (Cong et al., 2013). The targeting sequences of GFD1 (two targets, SG1: GGACGCGGCGCAGACGTTCG; SG2: CCGGCTACTCGGTGCTCTCC), OsSUT2 (SG3: CAAGTCTGCCTTTCTACTTC) and OsSWEET4 (SG4: ACGTTCATACGGATCTGGA) were synthesized and annealed to form the oligo adaptors. Agrobacterium‐mediated transformation was performed as described previously (Hiei et al., 1994). Finally, GFD1 was knockout in ZH11 (two targets, SG1 and SG2) and Nipponbare (one target, SG2). OsSUT2 and OsSWEET4 were knockout in ZH11 and KO2‐1 (gfd1 ZH11, Figures S3, S12) to generate single and double mutants. All the primers are listed in Table S1.

RNA isolation and quantitative RT‐PCR

According to the product manual, total RNA was extracted using an RNA isolater (Total RNA Extraction Reagent, R401‐01; Vazyme, Red Maple Technology Industrial Park, Nanjing, China). Reverse transcription of total RNA (~2 μg) was performed using HiScript III RT SuperMix for qPCR (+gDNA wiper) (R323‐01; Vazyme, Red Maple Technology Industrial Park, Nanjing, China). The qRT‐PCR analysis was performed on a CFX96 real‐time PCR system (Bio‐Rad, Hercules, California, United States) with ChamQ Universal SYBR qPCR Master Mix (Q711‐03; Vazyme, Red Maple Technology Industrial Park, Nanjing, China). The Actin1 gene was used as the internal control (Li et al., 2015). The primers used here are listed in Table S1.

Histochemical GUS analysis

The 1.6‐kb promoter of GFD1 was amplified from G46B and cloned into the pCAMBIA1391Z (Table S1). The resulting vector ProGFD1:GUS was transformed into ZH11 by the Agrobacterium‐mediated transformation method (Hiei et al., 1994). Histochemical GUS assay was performed as described previously (Jefferson et al., 1987). The tissues for detection were soaked in GUS staining solution, incubated at 37°C for 12–15 h and faded by Alcohol: acetic mixture (3:1) for observation. For microscopic examination, the developing caryopsis in 30 DAF was fixed in FAA after GUS staining for paraffin section making (Li et al., 2009b) and then observed using a Nikon DS‐U3 light microscope (Nikon).

In situ hybridization

The stem at the booting stage and the caryopsis at 9 DAF were used for in situ hybridization detection following the method depicted by Kouchi and Hata (Kouchi and Hata, 1993). In brief, the GFD1‐specific probe (Table S1) was labelled using Digoxigenin. After fixation, dehydration, sectioning, pre‐hybridization, hybridization, anti‐DIG‐AP and BCIP/NBT chromogenic solution addition, rinsing and sealing, the samples were observed and photographed by a Nikon DS‐U3 light microscope (Nikon).

Subcellular localization of GFD1

Green fluorescent protein (GFP) was fused to N‐ and C‐terminus of full‐length GFD1 CDS (coding sequence) driven by cauliflower mosaic virus (CaMV) 35 S promoter (Table S1). Then, the two fusion constructs, GFP‐GFD1 and GFD1‐GFP, were transformed into rice protoplasts separately following the method described by Chen et al. (2006). Meanwhile, OsRAC3‐mRFP (monomeric red fluorescent protein) was used as a plasma membrane marker (Chen et al., 2010). Man49‐mRFP (Nelson et al., 2007) and ERD2‐mRFP (Montesinos et al., 2014) were used as Golgi markers. These makers were co‐transformed into rice protoplasts or N. benthamiana leaves with GFD1‐GFP (Li et al., 2009a). Fluorescence signals were observed under a confocal laser scanning microscope (Nikon A1; Nikon).

Yeast two‐hybrid assay

Yeast two‐hybrid (Y2H) assay was performed using the Y2H Gold‐Gal4 system (Clontech, http://www.clontech.com). Full‐length CDS, the containing domains (MATE1, MATE2 and the mutant MATE1 gfd1 ) of GFD1, and the sugar transporters CDS (OsSUT1‐5, OsSWEET4, OsSWEET11 and OsSWEET15), were cloned into pGADT7 (AD) and pGBKT7 (BD), respectively. Yeast transformations were completed following the manufacturer's instructions (Clontech) and cultured on SD/‐Trp‐Leu or SD/‐Trp‐Leu‐His‐Ade medium containing X‐ɑ‐gal at 30°C in the dark for about 3 days. Primers used are given in Table S1.

LCI assay

Split‐luciferase complementation (LCI) assay was performed as described previously (Hu et al., 2019). The full‐length CDS of GFD1, OsSUT2 and OsSWEET4 was amplified and fused with luciferase (Table S1). The final constructs were introduced into Agrobacterium tumefaciens strain EHA105 and pairwise infiltrated the N. benthamiana leaves. Then, the leaves were stained using Beetle Luciferin, Potassium Salt kit (E1601; Promega, Madison, Wisconsin, United States) and placed into NightOWLIILB 983 in vivo imaging system (Berthold, Bad Wildbad, Germany). The interaction was determined based on the bioluminescence signal intensity acquired by IndiGO software.

BiFC assay

For BiFC, fusion vectors of GFD1 or OsSWEET4 were constructed using the binary BiFC vectors pSPYNE (nYFP) and pSPYCE (cYFP), respectively (Table S1). Bimolecular fluorescence complementation assay analysis was performed in the leaf epidermal cells of N. benthamiana, as previously described (Waadt and Kudla, 2008). The YFP fluorescence was observed using a Nikon A1 laser scanning confocal microscope (Nikon, Minato City, Tokyo, Japan).

Conflicts of interest statement

The authors declare no conflict of interest.

Author contributions

CS, PW and XD planned and designed the research. CS, YW, XY, LT, CW, JL, CC, HZ, CH, CL, QW, KZ, WZ and BY performed experiments and conducted fieldwork. CS, YW, XY, LT, CW, JL, SL, JZ, YS, WL, PW and XD analysed data. CS, YW, YZ and XD wrote the manuscript. CS, YW, XY, LT, CW and JL contributed equally.

Supporting information

Figure S1 Comparison of major agronomic traits of wild‐type (WT) and gfd1 mutant.

Figure S2 Comparison of major agronomic traits of NILZH11 and gfd1‐C1.

Figure S3 CRISPR/Cas9 knockout of GFD1 in ZH11 and Nipponbare background.

Figure S4 Phenotypes characterization of ZH11 and KO1.

Figure S5 Phenotypes characterization of ZH11 and KO2.

Figure S6 Comparison of major agronomic traits of ZH11, KO1 and KO2.

Figure S7 Phenotypes characterization of Nipponbare and KO3.

Figure S8 Comparison of major agronomic traits of Nipponbare and KO3.

Figure S9 GFD1‐GFP was agroinoculated into Nicotiana benthamiana leaves together with the Golgi marker Man49‐mRFP or ERD2‐mRFP.

Figure S10 Y2H screening the interaction proteins of GFD1.

Figure S11 Y2H screening the interaction proteins of GFD1.

Figure S12 CRISPR/Cas9 knockout of OsSUT2 and OsSWEET4.

PBI-21-621-s002.pdf (2.5MB, pdf)

Table S1 Primers used in this study.

PBI-21-621-s001.docx (34.5KB, docx)

Acknowledgements

This study was supported by grants from the National Natural Science Foundation of China (91335107, 91735303, 32172022, 31371602 and 31401358) and the Sichuan Science and Technology Program (2020YJ0408).

Contributor Information

Changhui Sun, Email: sunhui0307@163.com.

Pingrong Wang, Email: prwang@sicau.edu.cn.

Xiaojian Deng, Email: xjdeng@sicau.edu.cn.

References

  1. Alexandratos, N. and Bruinsma, J. (2012) World agriculture towards 2030/2050: the 2012 revision.
  2. Aoki, N. , Hirose, T. , Scofield, G.N. , Whitfeld, P.R. and Furbank, R.T. (2003) The sucrose transporter gene family in rice. Plant Cell Physiol. 44, 223–232. [DOI] [PubMed] [Google Scholar]
  3. Bihmidine, S. , Baker, R.F. , Hoffner, C. and Braun, D.M. (2015) Sucrose accumulation in sweet sorghum stems occurs by apoplasmic phloem unloading and does not involve differential Sucrose transporter expression. BMC Plant Biol. 15, 186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boone, C. , Bussey, H. and Andrews, B.J. (2007) Exploring genetic interactions and networks with yeast. Nat. Rev. Genet. 8, 437–449. [DOI] [PubMed] [Google Scholar]
  5. Chen, S. , Tao, L. , Zeng, L. , Vega‐Sanchez, M.E. , Umemura, K. and Wang, G.L. (2006) A highly efficient transient protoplast system for analyzing defence gene expression and protein‐protein interactions in rice. Mol. Plant Pathol. 7, 417–427. [DOI] [PubMed] [Google Scholar]
  6. Chen, L. , Shiotani, K. , Togashi, T. , Miki, D. , Aoyama, M. , Wong, H.L. , Kawasaki, T. et al. (2010) Analysis of the Rac/Rop small GTPase family in rice: expression, subcellular localization and role in disease resistance. Plant Cell Physiol. 51, 585–595. [DOI] [PubMed] [Google Scholar]
  7. Chen, L.Q. , Qu, X.Q. , Hou, B.H. , Sosso, D. , Osorio, S. , Fernie, A.R. and Frommer, W.B. (2012) Sucrose efflux mediated by SWEET proteins as a key step for phloem transport. Science 335, 207–211. [DOI] [PubMed] [Google Scholar]
  8. Chen, R. , Deng, Y. , Ding, Y. , Guo, J. , Qiu, J. , Wang, B. , Wang, C. et al. (2022) Rice functional genomics: decades' efforts and roads ahead. Sci. China Life Sci. 65, 33–92. [DOI] [PubMed] [Google Scholar]
  9. CocK, J.H. and Yoshida, S. (1972) Accumulation of 14C‐labelled carbohydrate before flowering and its subsequent redistribution and respiration in the rice plant. Japanese J. Crop Sci. 41, 226–234. [Google Scholar]
  10. Cong, L. , Ran, F.A. , Cox, D. , Lin, S. , Barretto, R. , Habib, N. , Hsu, P.D. et al. (2013) Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Durand, M. , Mainson, D. , Porcheron, B. , Maurousset, L. , Lemoine, R. and Pourtau, N. (2018) Carbon source‐sink relationship in Arabidopsis thaliana: the role of sucrose transporters. Planta 247, 587–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Eom, J.S. , Cho, J.I. , Reinders, A. , Lee, S.W. , Yoo, Y. , Tuan, P.Q. , Choi, S.B. et al. (2011) Impaired function of the tonoplast‐localized sucrose transporter in rice, OsSUT2, limits the transport of vacuolar reserve sucrose and affects plant growth. Plant Physiol. 157, 109–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fang, J. , Zhang, F. , Wang, H. , Wang, W. , Zhao, F. , Li, Z. , Sun, C. et al. (2019) Ef‐cd locus shortens rice maturity duration without yield penalty. Proc. Natl. Acad. Sci. U. S. A. 116, 18717–18722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Furbank, R.T. , Scofield, G.N. , Hirose, T. , Wang, X.‐D. , Patrick, J.W. and Offler, C.E. (2001) Cellular localisation and function of a sucrose transporter OsSUT1 in developing rice grains. Funct. Plant Biol. 28, 1187–1196. [Google Scholar]
  15. Hiei, Y. , Ohta, S. , Komari, T. and Kumashiro, T. (1994) Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T‐DNA. Plant J. 6, 271–282. [DOI] [PubMed] [Google Scholar]
  16. Hirose, T. , Takano, M. and Terao, T. (2002) Cell wall invertase in developing rice caryopsis: molecular cloning of OsCIN1 and analysis of its expression in relation to its role in grain filling. Plant Cell Physiol. 43, 452–459. [DOI] [PubMed] [Google Scholar]
  17. Hirose, T. , Zhang, Z. , Miyao, A. , Hirochika, H. , Ohsugi, R. and Terao, T. (2010) Disruption of a gene for rice sucrose transporter, OsSUT1, impairs pollen function but pollen maturation is unaffected. J. Exp. Bot. 61, 3639–3646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hu, B. , Jiang, Z. , Wang, W. , Qiu, Y. , Zhang, Z. , Liu, Y. , Li, A. et al. (2019) Nitrate‐NRT1.1B‐SPX4 cascade integrates nitrogen and phosphorus signalling networks in plants. Nat Plants 5, 401–413. [DOI] [PubMed] [Google Scholar]
  19. Jefferson, R.A. , Kavanagh, T.A. and Bevan, M.W. (1987) GUS fusions: beta‐glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901–3907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jones, A.M. , Xuan, Y. , Xu, M. , Wang, R.S. , Ho, C.H. , Lalonde, S. , You, C.H. et al. (2014) Border control—a membrane‐linked interactome of Arabidopsis. Science 344, 711–716. [DOI] [PubMed] [Google Scholar]
  21. Jonikas, M.C. , Collins, S.R. , Denic, V. , Oh, E. , Quan, E.M. , Schmid, V. , Weibezahn, J. et al. (2009) Comprehensive characterization of genes required for protein folding in the endoplasmic reticulum. Science 323, 1693–1697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Julius, B.T. , Leach, K.A. , Tran, T.M. , Mertz, R.A. and Braun, D.M. (2017) Sugar transporters in plants: new insights and discoveries. Plant Cell Physiol. 58, 1442–1460. [DOI] [PubMed] [Google Scholar]
  23. Kang, J. , Jain, M. , Wilkinson, D.S. and Embury, J.D. (2006) Microscopic strain mapping using scanning electron microscopy topography image correlation at large strain. J. Strain Analysis Eng. Design 40, 559–570. [Google Scholar]
  24. Kouchi, H. and Hata, S. (1993) Isolation and characterization of novel nodulin cDNAs representing genes expressed at early stages of soybean nodule development. Mol. Gen. Genet. 238, 106–119. [DOI] [PubMed] [Google Scholar]
  25. Krishnan, S. and Dayanandan, P. (2003) Structural and histochemical studies on grain‐filling in the caryopsis of rice (Oryza sativa L.). J. Biosci. 28, 455–469. [DOI] [PubMed] [Google Scholar]
  26. Kuroda, T. and Tsuchiya, T. (2009) Multidrug efflux transporters in the MATE family. Biochim. Biophys. Acta 1794, 763–768. [DOI] [PubMed] [Google Scholar]
  27. Lalonde, S. , Ehrhardt, D.W. , Loque, D. , Chen, J. , Rhee, S.Y. and Frommer, W.B. (2008) Molecular and cellular approaches for the detection of protein‐protein interactions: latest techniques and current limitations. Plant J. 53, 610–635. [DOI] [PubMed] [Google Scholar]
  28. Lalonde, S. , Sero, A. , Pratelli, R. , Pilot, G. , Chen, J. , Sardi, M.I. , Parsa, S.A. et al. (2010) A membrane protein/signaling protein interaction network for Arabidopsis version AMPv2. Front. Physiol. 1, 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Li, M. , Xiong, G. , Li, R. , Cui, J. , Tang, D. , Zhang, B. , Pauly, M. et al. (2009a) Rice cellulose synthase‐like D4 is essential for normal cell‐wall biosynthesis and plant growth. Plant J. 60, 1055–1069. [DOI] [PubMed] [Google Scholar]
  30. Li, S. , Qian, Q. , Fu, Z. , Zeng, D. , Meng, X. , Kyozuka, J. , Maekawa, M. et al. (2009b) Short panicle1 encodes a putative PTR family transporter and determines rice panicle size. Plant J. 58, 592–605. [DOI] [PubMed] [Google Scholar]
  31. Li, C. , Hu, Y. , Huang, R. , Ma, X. , Wang, Y. , Liao, T. , Zhong, P. et al. (2015) Mutation of FdC2 gene encoding a ferredoxin‐like protein with C‐terminal extension causes yellow‐green leaf phenotype in rice. Plant Sci. 238, 127–134. [DOI] [PubMed] [Google Scholar]
  32. Li, S. , Wei, X. , Ren, Y. , Qiu, J. , Jiao, G. , Guo, X. , Tang, S. et al. (2017) OsBT1 encodes an ADP‐glucose transporter involved in starch synthesis and compound granule formation in rice endosperm. Sci. Rep. 7, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Li, P. , Wang, L.H. , Liu, H.B. and Yuan, M. (2022) Impaired SWEET‐mediated sugar transportation impacts starch metabolism in developing rice seeds. Crop J. 10, 98–108. [Google Scholar]
  34. Liu, E. , Zeng, S. , Zhu, S. , Liu, Y. , Wu, G. , Zhao, K. , Liu, X. et al. (2019) Favorable alleles of grain‐filling RATE1 increase the grain‐filling rate and yield of rice. Plant Physiol. 181, 1207–1222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ma, L. , Zhang, D. , Miao, Q. , Yang, J. , Xuan, Y. and Hu, Y. (2017) Essential role of sugar transporter OsSWEET11 during the early stage of rice grain filling. Plant Cell Physiol. 58, 863–873. [DOI] [PubMed] [Google Scholar]
  36. Marger, M.D. and Saier, M.H. (1993) A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem. Sci. 18, 13–20. [DOI] [PubMed] [Google Scholar]
  37. Mathan, J. , Singh, A. and Ranjan, A. (2021a) Sucrose transport and metabolism control carbon partitioning between stem and grain in rice. J. Exp. Bot. 72, 4355–4372. [DOI] [PubMed] [Google Scholar]
  38. Mathan, J. , Singh, A. and Ranjan, A. (2021b) Sucrose transport in response to drought and salt stress involves ABA‐mediated induction of OsSWEET13 and OsSWEET15 in rice. Physiol. Plant. 171, 620–637. [DOI] [PubMed] [Google Scholar]
  39. Matsukura, C. , Saitoh, T. , Hirose, T. , Ohsugi, R. , Perata, P. and Yamaguchi, J. (2000) Sugar uptake and transport in rice embryo. Expression of companion cell‐specific sucrose transporter (OsSUT1) induced by sugar and light. Plant Physiol. 124, 85–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Montesinos, J.C. , Pastor‐Cantizano, N. , Robinson, D.G. , Marcote, M.J. and Aniento, F. (2014) Arabidopsis p24δ5 and p24δ9 facilitate Coat Protein I‐dependent transport of the K/HDEL receptor ERD2 from the Golgi to the endoplasmic reticulum. Plant J. 80, 1014–1030. [DOI] [PubMed] [Google Scholar]
  41. Nagata, K. , Yoshinaga, S. , J‐i, T. and Terao, T. (2015) Effects of dry matter production, translocation of nonstructural carbohydrates and nitrogen application on grain filling in rice cultivar Takanari, a cultivar bearing a large number of spikelets. Plant Prod. Sci. 4, 173–183. [Google Scholar]
  42. Nelson, B.K. , Cai, X. and Nebenfuhr, A. (2007) A multicolored set of in vivo organelle markers for co‐localization studies in Arabidopsis and other plants. Plant J. 51, 1126–1136. [DOI] [PubMed] [Google Scholar]
  43. Omote, H. , Hiasa, M. , Matsumoto, T. , Otsuka, M. and Moriyama, Y. (2006) The MATE proteins as fundamental transporters of metabolic and xenobiotic organic cations. Trends Pharmacol. Sci. 27, 587–593. [DOI] [PubMed] [Google Scholar]
  44. Pao, S.S. , Paulsen, I.T. and Saier, M.H., Jr. (1998) Major facilitator superfamily. Microbiol. Mol. Biol. Rev. 62, 1–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Peng, C. , Wang, Y. , Liu, F. , Ren, Y. , Zhou, K. , Lv, J. , Zheng, M. et al. (2014) FLOURY ENDOSPERM6 encodes a CBM48 domain‐containing protein involved in compound granule formation and starch synthesis in rice endosperm. Plant J. 77, 917–930. [DOI] [PubMed] [Google Scholar]
  46. Ruan, Y.L. (2014) Sucrose metabolism: gateway to diverse carbon use and sugar signaling. Annu. Rev. Plant Biol. 65, 33–67. [DOI] [PubMed] [Google Scholar]
  47. Scofield, G.N. , Aoki, N. , Hirose, T. , Takano, M. , Jenkins, C.L. and Furbank, R.T. (2007a) The role of the sucrose transporter, OsSUT1, in germination and early seedling growth and development of rice plants. J. Exp. Bot. 58, 483–495. [DOI] [PubMed] [Google Scholar]
  48. Scofield, G.N. , Hirose, T. , Aoki, N. and Furbank, R.T. (2007b) Involvement of the sucrose transporter, OsSUT1, in the long‐distance pathway for assimilate transport in rice. J. Exp. Bot. 58, 3155–3169. [DOI] [PubMed] [Google Scholar]
  49. Shimizu, Y. , Takagi, J. , Ito, E. , Ito, Y. , Ebine, K. , Komatsu, Y. , Goto, Y. et al. (2021) Cargo sorting zones in the trans‐Golgi network visualized by super‐resolution confocal live imaging microscopy in plants. Nat. Commun. 12, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Siao, W. , Chen, J.Y. , Hsiao, H.H. , Chung, P. and Wang, S.J. (2011) Characterization of OsSUT2 expression and regulation in germinating embryos of rice seeds. Rice 4, 39–49. [Google Scholar]
  51. Sosso, D. , Luo, D. , Li, Q.B. , Sasse, J. , Yang, J. , Gendrot, G. , Suzuki, M. et al. (2015) Seed filling in domesticated maize and rice depends on SWEET‐mediated hexose transport. Nat. Genet. 47, 1489–1493. [DOI] [PubMed] [Google Scholar]
  52. Sturm, A. and Tang, G.Q. (1999) The sucrose‐cleaving enzymes of plants are crucial for development, growth and carbon partitioning. Trends Plant Sci. 4, 401–407. [DOI] [PubMed] [Google Scholar]
  53. Sun, A.J. , Xu, H.L. , Gong, W.K. , Zhai, H.L. , Meng, K. , Wang, Y.Q. , Wei, X.L. et al. (2008) Cloning and expression analysis of rice sucrose transporter genes OsSUT2M and OsSUT5Z. J. Integr. Plant Biol. 50, 62–75. [DOI] [PubMed] [Google Scholar]
  54. Sun, Y. , Reinders, A. , LaFleur, K.R. , Mori, T. and Ward, J.M. (2010) Transport activity of rice sucrose transporters OsSUT1 and OsSUT5. Plant Cell Physiol. 51, 114–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sun, C. , Chen, D. , Fang, J. , Wang, P. , Deng, X. and Chu, C. (2014) Understanding the genetic and epigenetic architecture in complex network of rice flowering pathways. Protein Cell 5, 889–898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Sun, C. , Zhang, K. , Zhou, Y. , Xiang, L. , He, C. , Zhong, C. , Li, K. et al. (2021) Dual function of clock component OsLHY sets critical day length for photoperiodic flowering in rice. Plant Biotechnol. J. 19, 1644–1657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Takai, T. , Fukuta, Y. , Shiraiwa, T. and Horie, T. (2005) Time‐related mapping of quantitative trait loci controlling grain‐filling in rice (Oryza sativa L.). J. Exp. Bot. 56, 2107–2118. [DOI] [PubMed] [Google Scholar]
  58. Takanashi, K. , Shitan, N. and Yazaki, K. (2014) The multidrug and toxic compound extrusion (MATE) family in plants. Plant Biotechnol. 31, 417–430. [Google Scholar]
  59. Upadhyay, N. , Kar, D. , Deepak Mahajan, B. , Nanda, S. , Rahiman, R. , Panchakshari, N. , Bhagavatula, L. et al. (2019) The multitasking abilities of MATE transporters in plants. J. Exp. Bot. 70, 4643–4656. [DOI] [PubMed] [Google Scholar]
  60. Waadt, R. and Kudla, J. (2008) In planta visualization of protein interactions using bimolecular fluorescence complementation (BiFC). CSH Protoc 2008, pdb prot4995. [DOI] [PubMed] [Google Scholar]
  61. Wang, E. , Wang, J. , Zhu, X. , Hao, W. , Wang, L. , Li, Q. , Zhang, L. et al. (2008) Control of rice grain‐filling and yield by a gene with a potential signature of domestication. Nat. Genet. 40, 1370–1374. [DOI] [PubMed] [Google Scholar]
  62. Wang, P. , Gao, J. , Wan, C. , Zhang, F. , Xu, Z. , Huang, X. , Sun, X. et al. (2010) Divinyl chlorophyll(ide) a can be converted to monovinyl chlorophyll(ide) a by a divinyl reductase in rice. Plant Physiol. 153, 994–1003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang, L. , Bei, X. , Gao, J. , Li, Y. , Yan, Y. and Hu, Y. (2016) The similar and different evolutionary trends of MATE family occurred between rice and Arabidopsis thaliana . BMC Plant Biol. 16, 207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Wang, G.Q. , Li, H.X. , Gong, Y.L. , Yang, J.C. , Yi, Y.K. , Zhang, J.H. and Ye, N.H. (2020) Expression profile of the carbon reserve remobilization from the source to sink in rice in response to soil drying during grain filling. Food Energy Security 9, e204. [Google Scholar]
  65. Wei, X. , Jiao, G. , Lin, H. , Sheng, Z. , Shao, G. , Xie, L. , Tang, S. et al. (2017) GRAIN INCOMPLETE FILLING 2 regulates grain filling and starch synthesis during rice caryopsis development. J. Integr. Plant Biol. 59, 134–153. [DOI] [PubMed] [Google Scholar]
  66. Wipf, D. , Pfister, C. , Mounier, A. , Leborgne‐Castel, N. , Frommer, W.B. and Courty, P.E. (2021) Identification of putative interactors of Arabidopsis sugar transporters. Trends Plant Sci. 26, 13–22. [DOI] [PubMed] [Google Scholar]
  67. Wu, Y. , Lee, S.K. , Yoo, Y. , Wei, J. , Kwon, S.Y. , Lee, S.W. , Jeon, J.S. et al. (2018) Rice transcription factor OsDOF11 modulates sugar transport by promoting expression of sucrose transporter and SWEET genes. Mol. Plant 11, 833–845. [DOI] [PubMed] [Google Scholar]
  68. Xiong, Y. , Ren, Y. , Li, W. , Wu, F. , Yang, W. , Huang, X. and Yao, J. (2019) NF‐YC12 is a key multi‐functional regulator of accumulation of seed storage substances in rice. J. Exp. Bot. 70, 3765–3780. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Yang, J. , Luo, D. , Yang, B. , Frommer, W.B. and Eom, J.S. (2018) SWEET11 and 15 as key players in seed filling in rice. New Phytol. 218, 604–615. [DOI] [PubMed] [Google Scholar]
  70. Yokosho, K. , Yamaji, N. , Fujii‐Kashino, M. and Ma, J.F. (2016) Functional analysis of a MATE gene OsFRDL2 revealed its involvement in Al‐induced secretion of citrate, but a lower contribution to Al tolerance in rice. Plant Cell Physiol. 57, 976–985. [DOI] [PubMed] [Google Scholar]
  71. Zee, S.Y. (1972) Vascular tissue and transfer cell distribution in the rice spikelet. Aust. J. Biol. Sci. 25, 411–414. [Google Scholar]
  72. Zhang, C. and Turgeon, R. (2018) Mechanisms of phloem loading. Curr. Opin. Plant Biol. 43, 71–75. [DOI] [PubMed] [Google Scholar]
  73. Zhang, W.H. , Zhou, Y. , Dibley, K.E. , Tyerman, S.D. , Furbank, R.T. and Patrick, J.W. (2007) Review: nutrient loading of developing seeds. Funct. Plant Biol. 34, 314–331. [DOI] [PubMed] [Google Scholar]
  74. Zhao, C. , Xu, W. , Song, X. , Dai, W. , Dai, L. , Zhang, Z. and Qiang, S. (2018) Early flowering and rapid grain filling determine early maturity and escape from harvesting in weedy rice. Pest Manag. Sci. 74, 465–476. [DOI] [PubMed] [Google Scholar]
  75. Zhou, S. , Zhu, S. , Cui, S. , Hou, H. , Wu, H. , Hao, B. , Cai, L. et al. (2021) Transcriptional and post‐transcriptional regulation of heading date in rice. New Phytol. 230, 943–956. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1 Comparison of major agronomic traits of wild‐type (WT) and gfd1 mutant.

Figure S2 Comparison of major agronomic traits of NILZH11 and gfd1‐C1.

Figure S3 CRISPR/Cas9 knockout of GFD1 in ZH11 and Nipponbare background.

Figure S4 Phenotypes characterization of ZH11 and KO1.

Figure S5 Phenotypes characterization of ZH11 and KO2.

Figure S6 Comparison of major agronomic traits of ZH11, KO1 and KO2.

Figure S7 Phenotypes characterization of Nipponbare and KO3.

Figure S8 Comparison of major agronomic traits of Nipponbare and KO3.

Figure S9 GFD1‐GFP was agroinoculated into Nicotiana benthamiana leaves together with the Golgi marker Man49‐mRFP or ERD2‐mRFP.

Figure S10 Y2H screening the interaction proteins of GFD1.

Figure S11 Y2H screening the interaction proteins of GFD1.

Figure S12 CRISPR/Cas9 knockout of OsSUT2 and OsSWEET4.

PBI-21-621-s002.pdf (2.5MB, pdf)

Table S1 Primers used in this study.

PBI-21-621-s001.docx (34.5KB, docx)

Articles from Plant Biotechnology Journal are provided here courtesy of Society for Experimental Biology (SEB) and the Association of Applied Biologists (AAB) and John Wiley and Sons, Ltd

RESOURCES