OsmiR396/growth regulating factor modulate rice grain size through direct regulation of embryo-specific miR408

Abstract

microRNAs (miRNAs) are promising targets for crop improvement of complex agricultural traits. Coordinated activity between/among different miRNAs may fine-tune specific developmental processes in diverse organisms. Grain size is a main factor determining rice (Oryza sativa L.) crop yield, but the network of miRNAs influencing this trait remains uncharacterized. Here we show that sequestering OsmiR396 through target mimicry (MIM396) can substantially increase grain size in several japonica and indica rice subspecies and in plants with excessive tillers and a high panicle density. Thus, OsmiR396 has a major role related to the regulation of rice grain size. The grain shape of Growth Regulating Factor8 (OsGRF8)-overexpressing transgenic plants was most similar to that of MIM396 plants, suggesting OsGRF8 is a major mediator of OsmiR396 in grain size regulation. A miRNA microarray analysis revealed changes to the expression of many miRNAs, including OsmiR408, in the MIM396 plants. Analyses of gene expression patterns and functions indicated OsmiR408 is an embryo-specific miRNA that positively regulates grain size. Silencing OsmiR408 expression (miR408KO) using CRISPR technology resulted in small grains. Moreover, we revealed the direct regulatory effects of OsGRF8 on OsMIR408 expression. A genetic analysis further showed that the large-grain phenotype of MIM396 plants could be complemented by miR408KO. Also, several hormone signaling pathways might be involved in the OsmiR396/GRF-meditated grain size regulation. Our findings suggest that genetic regulatory networks comprising various miRNAs, such as OsmiR396 and OsmiR408, may be crucial for controlling rice grain size. Furthermore, the OsmiR396/GRF module may be important for breeding new high-yielding rice varieties.

Figure 1

Grain characteristics due to the down regulation of OsmiR396 in several japonica subspecies. A, Grain morphology of the MIM396 and corresponding WT ZH11 plants. B, Statistical analysis of the grain length and width of the MIM396 and ZH11 plants (n = 30). C, Statistical analysis of the 1,000 KGW of the MIM396 and ZH11 plants (n = 100). D, Grain morphology of the XM396 plant and the corresponding WT XS134 plants. E, Statistical analysis of the grain length and width of the XM396 plants and XS134 plants (n = 30). F, Statistical analysis of the KGW of the XM396 plants and XS134 plants (n = 100). G, Grain morphology of the RM396 plant and the corresponding WT R8015 plants. H, Statistical analysis of the grain length and width of the RM396 and R8015 plants (n = 30). I, Statistical analysis of the KGW of the RM396 and R8015 plants (n = 100). Comparisons were made compared to the respective WT unless otherwise indicated. Error bars represent ± sd, single and double asterisks in the statistical analyses represent significant difference determined by the Student’s t test at *P < 0.05 and **P < 0.01.

Figure 2

Analysis of the OsGRF1, OsGRF8, and OsGRF9 expression levels and functions. A, Expression profiles of 12 OsGRFs genes after fertilization (n = 3). B, Expression profiles of OsGRF1, OsGRF8, and OsGRF9 in various tissues (n = 3). C, Northern blot analysis of OsmiR396b and OsmiR396d in various tissues. D, Grain morphology of the GRF1OE and WT plants. E, Statistical analysis of the grain length and width of the GRF1OE and WT plants (n = 30). F, Statistical analysis of the KGW of the GRF1OE and WT plants (n = 100). G, Grain morphology of the GRF8OE and WT plants. (H) Statistical analysis of the grain length and width of the GRF8OE and WT plants (n = 30). I, Statistical analysis of the KGW of the GRF8OE and WT plants (n = 100). Comparisons were made compared to the respective WT unless otherwise indicated. Error bars represent ± sd, single and double asterisks in the statistical analyses represent significant difference determined by the Student’s t test at *P <0.05 and **P <0.01.

Figure 3

miRNA microarray analysis of the young panicles of the MIM396 and WT plants as well as verification of OsmiR408. A, Volcano plot of the overall expression of the miRNAs in the MIM396 and WT plants. Green dots indicated the differentially down-regulated miRNAs, whereas red dots represent the differentially up-regulated miRNAs in the MIM396 plants compared with the WT plants. B, Expression of the differentially expressed miRNAs in the MIM396 and WT plants. The bar represents the scale of relative miRNA expression levels on the basis of Log₂ value. C, RT-qPCR analysis of OsmiR408 in GRF8OE, MIM396, and WT plants at 10 DAF. Comparisons were made compared to the respective WT unless otherwise indicated. Error bars represent ± sd, double asterisks in the statistical analyses represent significant difference determined by the Student’s t test at **P < 0.01.

Figure 4

Expression profile of OsmiR408. A, miRNA northern blot analysis of OsmiR408 in young embryos at 1–10 DAF. B, miRNA northern blot analysis of OsmiR408 in young embryos at 6–14 DAF. C, The OsmiR408 LNA antisense probe detected OsmiR408 in the embryos at 8 DAF. D, The OsmiR408 LNA antisense probe detected OsmiR408 in the embryos at 12 DAF. E, Analysis with the OsmiR408 LNA sense probe. F–M, GUS staining of the following p408::GUS parts: the root (F), the shoot apical meristem (G), the leaf (H), the spikelet (I), the germinating seeds (J), the isolated young embryo (K), the seed (cross section) (L), and the enlarged embryo (M). Abbreviations: ra, radicle; pa, plumular axis; pl, plumule; co, coleoptile. Bars in C–E were 100 µm, and those in F–M were 1 mm.

Figure 5

Functional analysis of OsmiR408 in rice. A, Overall morphology of the 35S::miR408 transgenic and WT plants. Bar was 10 cm. B, miRNA northern blot analysis of OsmiR408 in the 35S::miR408 and WT plants. C, Leaf angle of the 35S::miR408 and WT plants. Bar was 1 cm. D, Sequence deletions in homozygous miR408KO-1 and miR408KO-2 plants. Red bases represent the mature OsmiR408 sequences, whereas red dashes indicate the deletions in the 408KO-1 and 408KO-2 lines. E, miRNA northern blot analysis of OsmiR408 transcripts in the miR408KO-1 and miR408KO-2 plants. F, Grain morphology of the WT, miR408KO-1, and miR408KO-2 lines. G, Statistical analysis of the grain length and width of the WT and the 408KO lines (n = 3). H, Statistical analysis of the KGW of the WT and 408KO lines (n = 3). Comparisons were made compared to the respective WT unless otherwise indicated. Error bars represent ± sd, single and double asterisks in the statistical analyses represent significant difference determined by the Student’s t test at *P < 0.05 and **P < 0.01.

Figure 6.

Morphological analysis of the cross between MIM396 and 408KO-2 as well as biochemical analysis of the binding of OsGRF8 to the OsMIR408 promoter. A, Expression analysis of the IPS gene in the ZH11, MIM396, 408KO-2, and cross plants (n = 3). B, Grain morphology of the ZH11, MIM396, 408KO-2, and cross plants. C, Grain length and width of the ZH11, MIM396, 408KO-2, and cross plants (n = 3). D, Statistical analysis of the KGW of the ZH11, MIM396, 408KO-2, and cross plants (n = 3). E, Schematic representation of the 2-kb OsMIR408 promoter including the positions of the putative GRF-binding motifs as well as the relative sites for the YOH, EMSA, and ChIP assays. The black line represents the OsMIR408 promoter, with the gene coding direction indicated by the arrowhead. The red bar indicates the mature OsmiR408 region. The blue bars indicate the predicted GRF-binding motifs. The pink and blue squares, respectively, indicate the normal and truncated regions used for the EMSA. The orange squares indicate the regions used for the YOH. The green squares indicate the regions used for the ChIP assay. F, The YOH assay of OsGRF8 and the OsMIR408 promoter. The fragments used for the YOH assay (YOHP1 and YOHP2) are indicated in (E). Mutations were introduced into the respective primers (Supplemental Table S2). G, EMSA assay of the OsGRF8 WRC domain and the normal OsMIR408 promoter fragments containing the putative motifs and the truncated fragments lacking the putative motifs. H, ChIP analysis of the GRF8OE and WT plants. The ChIP1, 2 and 3 sites ware indicated in (E) (n = 3). ChIP0 refers to a site upstream of the OsMIR408 promoter where there is no putative GRF-binding motif (internal reference). Comparisons were made compared to the respective WT unless otherwise indicated. Error bars represent ± sd, single and double asterisks in (A), (C), and (D) represent significant difference determined by the Student’s t test at *P < 0.05 and **P < 0.01.