The OsZHD1 and OsZHD2, Two Zinc Finger Homeobox Transcription Factor, Redundantly Control Grain Size by Influencing Cell Proliferation in Rice

Abstract

Grain size is vital determinant for grain yield and quality, which specified by its three-dimensional structure of seeds (length, width and thickness). The ZINC FINGER-HOMEODOMAIN (ZHD) proteins play critical roles in plant growth and development. However, the information regarding the function in reproductive development of ZHD proteins is scarce. Here, we deeply characterized the phenotype of oszhd1, oszhd2, and oszhd1oszhd2. The single mutants of OsZHD1/2 were similar with wild type. Nevertheless, the double mutant displayed dwarfism and smaller reproductive organs, and shorter, narrower, and thinner grain size. oszhd1oszhd2 revealed a significant decrease in total cell length and number, and single cell width in outer parenchyma; reducing the average width of longitudinal epidermal cells, but the length were increased in outer and inner glumes of oszhd1oszhd2 compared with wild-type, oszhd1-1, oszhd2-1, respectively. OsZHD1 and OsZHD2 encoded the nucleus protein and were distributed predominately in stem and the developing spikelets, asserting their roles in grain size. Meanwhile, yeast two-hybrid, bimolecular fluorescence complementation, and Co-immunoprecipitation assay clarified that OsZHD1 could directly interacted with OsZHD2. The differential expression analysis showed that 839 DEGs, which were down-regulated in oszhd1oszhd2 than wild type and single mutants, were mainly enriched in secondary metabolite biosynthetic, integral component of membrane, and transporter activity pathway. Moreover, it is reliable that the altered expression of cell cycle and expansion-related and grain size-related genes were observed in RNA-seq data, highly consistent with the qRT-PCR results. Altogether, our results suggest that OsZHD1/2 are functional redundancy and involved in regulating grain size by influencing cell proliferation in rice.

Keywords: OsZHD1; OsZHD2; Cell proliferation; Grain size; Rice; Zinc finger-homeodomain.

Fig. 1

Genotyping analysis of oszhd1, oszhd2, and oszhd1oszhd2. (A) Identification of three CRISPR/Cas9-mediated mutations in OsZHD1. (B) Identification of three CRISPR/Cas9-mediated mutations in OsZHD2. (C) Identification the double mutant mediated by CRISPR/Cas9. (D) The expression levels of OsZHD1 and OsZHD2 in the mutants. The gene structure of OsZHD1 (A), OsZHD2 (B), and both (C) are respectively shown at the top. The primers of P1 and P2, P3 and P4 were used for identifying oszhd1, oszhd2, and oszhd1oszhd2 in the T2 generation. The diagrams show the sequences of the CRISPR mutations in OsZHD1 and OsZHD2. The black letters indicated the base pairs with WT in the target site. The dotted lines indicated nucleotide deletions, and dark blue letters indicated the amino acids encoded by the WT, oszhd1, oszhd2, and oszhd1oszhd2 sequences. Data are given as means ± SD. Different letters denote significant difference at P < 0.05 according to ANOVA in combination with Duncan’s multiple range test

Fig. 2

The oszhd1oszhd2 exhibited dwarfism and smaller spikelets. (A) The phenotype of wild-type (left, ZH11), oszhd1-1, oszhd2-1, and oszhd1oszhd2 plants at maturity. Bar = 15 cm. (B-C) The internodes length and panicle of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2. Bar = 10 cm. (D-E) The spikelets of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2. Bar = 1 mm. (F-G) Anthers and pistils of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2, respectively. Bar = 500 μm. (H-J) The average plant height, internode length, panicles length of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2, respectively. (n = 20 per sample). Data are given as means ± SD. Different letters denote significant difference at P < 0.05 according to ANOVA in combination with Duncan’s multiple range test

Fig. 3

Functional verification of OsZHD1 and OsZHD2 in grain size. (A-B) Comparation on grain size of oszhd1-1, oszhd2-1, and oszhd1oszhd2 with ZH11, respectively. Bar = 5 mm. (C-F) Grain length, width, thickness, and 100-grain weight of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2. (n = 60 per sample). Data are given as means ± SD. Different letters denote significant difference at P < 0.05 according to ANOVA in combination with Duncan’s multiple range test

Fig. 4

Histological analysis of spikelet hulls. (A) Young spikelet hulls of ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2. The white line indicates the position of the cross-section. Bar = 1 mm. (B) Cross-sections of spikelet hulls from ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2, respectively. Bar = 500 μm. (C) Magnified view of the cross-section area boxed in (B). Bar = 20 μm. (D) Statistical data of the total length, cell number and width in the outer parenchyma layer (n = 12). (E) Scanning electron microscope (SEM) analysis of the outer surface of glumes. Bar = 100 μm. (F) Statistical analysis of cell length and width in outer glumes (n = 40). (G) Scanning electron microscope (SEM) analysis of the inner surface of glumes. Bar = 100 μm. (H) Statistical analysis of cell length and width in inner glumes (n = 60). Data are given as means ± SD. Different letters denote significant difference at P < 0.05 according to ANOVA in combination with Duncan’s multiple range test

Fig. 5

The expression profile of OsZHD1 and OsZHD2. (A) Subcellular localization of OsZHD1 and OsZHD2 proteins were respectively shown to target the nuclear by transient expression of 35::OsZHD1-GFP and 35::OsZHD2-GFP in tobacco. The upper row is the 35 S::GFP as the control. Bar = 50 μm. (B-E) Expression pattern detected in transgenic plants carrying the pOsZHD1::GUS and pOsZHD2::GUS vector in ZH11 background. GUS signal was detected in the stem node of pOsZHD1::GUS (B) and pOsZHD2::GUS (C), respectively. Developing spikelets in turn 1, 2, 3, 4, 5, 6,7 mm, and mature spikelet of pOsZHD1::GUS (D) and pOsZHD2::GUS (E), respectively

Fig. 6

OsZHD1 interacts with OsZHD2. (A) The OsZHD1-OsZHD2 interaction as revealed by Y₂H assays in yeast cells. SD, yeast dropout culture medium; PGADT7, activation domain; PGBKT7, DNA-binding domain. (B) BiFC analysis in N.benthamiana leaves of the interaction between OsZHD1-cYFP and OsZHD2-nYFP. The split YFP system was used. cYFP and nYFP are empty vectors. The YFP signals were detected with a confocal microscope. Bar = 50 μm. (C) In vivo Co-IP assays of OsZHD1-GFP and OsZHD2-Myc. The fusion proteins were transiently co-expressed N.benthamiana (tobacco) leaves. Protein extracts (input) were immunoprecipitated using an anti-GFP antibody (IP)

Fig. 7

Global expression levels, DEGs, GO, and KEGG pathway analysis. (A) Percentage of gene numbers in each tissue according to their expression levels based on FPKM values. (B) Hierarchical clustering of the twelve panicle samples using Pearson correlation coefficients. (C) Numbers of differentially expressed genes among ZH11, oszhd1-1, oszhd2-1, and oszhd1oszhd2 tissues. (D) Numbers of down-regulated genes in oszhd1oszhd2 versus ZH11, oszhd1-1, and oszhd2-1, respectively. (E-G) GO classifications for all down-regulated genes in oszhd1oszhd2 versus ZH11, oszhd1-1, and oszhd2-1. The results were classified into three main categories: biological process (E), cellular component (F), and molecular function (G). (H) KEGG pathway analysis for all down-regulated genes in oszhd1oszhd2 versus ZH11, oszhd1-1, and oszhd2-1

Fig. 8

Validation of the expression patterns of cell cycle and expansion-related and grain size-related genes in the panicle of OsZHD1 and OsZHD2 mutants by qRT-PCR. (A) Expression profiling based on the FPKM values from RNA-seq analysis. (B) Expression patterns of genes encoding cell cycle and expansion-related and grain size-related genes by qRT-PCR. Data are given as means ± SD. Different letters denote significant difference at P < 0.05 according to ANOVA in combination with Duncan’s multiple range test