Rice aleurone layer specific OsNF-YB1 regulates grain filling and endosperm development by interacting with an ERF transcription factor

Abstract

Grain yield and quality of rice mainly depend on grain filling and endosperm development. Here we report that a rice NUCLEAR FACTOR Y (NF-Y) transcription factor, OsNF-YB1, is specifically expressed in the aleurone layer of developing endosperm and regulates grain filling and endosperm development. Knockdown of OsNF-YB1 expression by RNAi significantly retarded grain filling, leading to small grains with chalky endosperm as well as altered starch quality. Whereas OsNF-YB1 shows subcellular localization in both the cytosol and the nucleus in roots, it was specifically targeted to the nucleus of aleurone layer cells, which was facilitated by interacting with OsNF-YC proteins preferentially expressed in the aleurone layer. RNA sequencing analysis revealed that genes related to membrane transport and ATP biosynthesis were enriched in the down-regulated category in OsNF-YB1 RNAi plants, which is consistent with the crucial role of OsNF-YB1 in rice grain filling and endosperm development. Identification of the genome-wide targets of OsNF-YB1 by ChIP sequencing showed that OsNF-YB1 directly regulates genes involved in the transport of nutrients such as sugar and amino acids. Interestingly, different from the binding sites reported for other NF-Y complexes, the GCC box, the binding motif of ERF transcription factors, was enriched in the binding peaks of OsNF-YB1. Indeed, further analyses confirmed the interaction of OsERF#115 with OsNF-YB1, and OsERF#115 directly binds to the GCC box. It is proposed that OsNF-YB1 specifically regulate the transcription of downstream genes during rice endosperm development by forming protein complexes consisting of OsNF-YB1, OsNF-YC and ERF, providing informative insights into the molecular functional mechanisms of the NF-Y factor.

Keywords: Aleurone layer; ERF; GCC box; OsNF-YB1.; endosperm development; grain filling.

Fig. 1.

OsNF-YB1 is specifically expressed in the aleurone layer of rice endosperm. RNA in situ hybridization analysis of the OsNF-YB1 transcripts during seed development. Transverse (upper, bars: 500 µm) and longitudinal (bottom, bars: 1 mm) sections of seeds at 4, 6, 8, and 10 DAF were analyzed. Sense probe was used as control.

Fig. 2.

Suppressed expression of OsNF-YB1 results in reduced grain filling. (A) Observation and calculation showed that brown grains of OsNF-YB1 RNAi plants were smaller compared with ZH11 (left, bar: 1 cm) and 1000-brown-grain weight was reduced (right). Data are shown as mean±SD from ten replicates. (B) Analysis of the time course of seed dry weight revealed reduced grain filling of OsNF-YB1 RNAi plants. Data are shown as mean±SE (n=10). (C) Observation and measurement showed that endosperm of OsNF-YB1 RNAi plants displayed higher chalkiness including higher percentage of grain with chalkiness (PGWC) and degree of endosperm chalkiness (DEC), and lower apparent amylose content (AAC). Data are shown as mean±SD from three replicates. Bar: 1 cm. (D) Observation of transverse sections of seeds revealed a normal aleurone layer of OsNF-YB1 RNAi plants. Dorsal (left) and ventral side (right) of seeds at 10 DAF are shown. Bars: 200 µm.

Fig. 3.

OsNF-YC11 and OsNF-YC12 interact with OsNF-YB1 in vivo and mediate the nuclear localization of OsNF-YB1. (A) Observation of OsNF-YB1 subcellular localization showed a dual cytosolic–nuclear localization in root cells and nucleus-specific localization in aleurone layer cells. Root of young seedlings and aleurone layer cells of 10 DAF seeds expressing pUbi:OsNF-YB1-GFP were observed. Bars: 50 µm (upper) or 20 µm (bottom). (B) OsNF-YB1-GFP protein showed a dual cytosolic–nuclear localization in the presence of red fluorescent protein (RFP) or OsNF-YC2-RFP, and was translocated to nucleus of rice protoplast cells in the presence of OsNF-YC11-RFP or OsNF-YC12-RFP. Bars: 5 µm. (C, D) Co-immunoprecipitation analysis revealed the interaction of OsNF-YB1-GFP and OsNF-YC11-mCherry (C), and OsNF-YB1-GFP and OsNF-YC12-mCherry (D) in tobacco cells. Total protein extracts (Input) or immunoprecipitated (IP) fractions using an anti-GFP antibody were analyzed using anti-GFP or anti-mCherry antibodies.

Fig. 4.

Genome-wide analyses of the OsNF-YB1-regulated network. (A) Selected enriched GO terms of the 960 down-regulated genes by RNA-seq using aleurone layer cells at 8 DAF of OsNF-YB1 RNAi plants. Hypergeometric test was used with subsequent Benjamini and Hochberg false discovery rate corrections. Only GO terms with a corrected P-value <0.05 and at least five annotated genes were kept. Length of bars represents negative logarithm (base 10) of the corrected P-value. (B) qRT-PCR analysis confirmed the down-regulation of genes in aleurone layer cells of the OsNF-YB1 RNAi plants. Relative expression of examined genes (listed in Supplementary Table S2) was calculated (expression of the corresponding gene in ZH11 was set as 1.0) and data are shown as mean±SE (n=3).

Fig. 5.

Overview of ChIP-seq data and verification of OsNF-YB1 binding sites. (A) Distribution of the OsNF-YB1 binding peaks relative to gene structure. (B) Selected enriched GO terms of genes associated with OsNF-YB1 binding by ChIP-Seq analysis. Only GO terms with a corrected P-value <0.05 and at least five annotated genes were kept. Length of bars represents negative logarithm (base 10) of the corrected P-value. (C) ChIP-qPCR validation of the selected OsNF-YB1 targets. Enrichment of the ten candidate OsNF-YB1 binding targets (listed in Supplementary table S2) was examined. OsNF-YB1, which has no enrichment in ChIP-seq data, was used as the negative control. Fold enrichment of DNA was calculated as the ratio between αGFP and IgG. Data are presented as means±SD from three replicates. (D) The most enriched motifs identified by DREME in OsNF-YB1 binding peaks that appeared within 3 kb upstream regions of genes. (E) Enrichment of GCC box and CCAAT box compared with the promoter regions (0–3 kb) of reference rice genome. CCAAT box was not enriched in the binding sites.

Fig. 6.

OsNF-YB1 interacts with transcription factor OsERF#115. (A) Yeast two-hybrid assays revealed the interactions between OsNF-YB1 and OsERF#115. Serial dilutions (10×) of yeast cells expressing the indicated proteins were plated onto the non-selective medium (SD/–Leu/–Trp) (left) or selective medium (SD/–Leu/–Trp/–His/–Ade) (right) and cell growth were observed. OsERF members and OsNF-YB1 were fused to activation domain (AD) and binding domain (BD), respectively. (B) Yeast two-hybrid assays showed that there was no interaction between OsERF#115 and OsNF-YC11 or OsNF-YC12. Serial dilutions (10×) of yeast cells expressing the indicated proteins were plated onto the non-selective medium (SD/–Leu/–Trp) (left) or selective medium (SD/–Leu/–Trp/–His) (right) and cell growth was observed. OsERF#115 and OsNF-YC11/OsNF-YC12 were fused to AD and BD, respectively. (C) Yeast one-hybrid assays revealed that OsERF#115 binds to, while OsNF-YB1 does not, the GCC box. OsERF#115 or OsNF-YB1 was fused to GAL4 transcriptional activation domain (AD) and used as prey. DNA fragment of Os07g19070 promoter containing the GCC box (or with mutated bases) was repeated three times and fused to pHIS2 as baits. Yeast cells co-transformed with indicated prey or bait vectors were grown on non-selective medium (SD/–Leu/–Trp) (left) or selective medium (SD/–Leu/–Trp/–His, 200 mM 3-amino-1,2,4-triazole (3-AT)) (right) and cell growth was observed. (D) Analysis showed that OsERF#115, which was fused to the GAL4 binding domain (BD) of pGBKT7, does not present transactivational activity. LAX1-BD was used as positive control. (E) There was no transactivational activity of yeast cells expressing OsERF#115, OsERF#115/OsNF-YB1, OsERF#115/OsNF-YC11, OsERF#115/OsNF-YC12, or OsERF#115/OsNF-YB1/OsNF-YC11, OsERF#115/OsNF-YB1/OsNF-YC12. OsERF#115 was fused to the GAL4 binding domain (BD, pBridge vector). OsNF-YB1 was subcloned into pBridge-OsERF#115 construct. OsNF-YC11 or OsNF-YC12 was subcloned into modified pGADT7 without the GAL4-activation domain. LAX1-BD was used as a positive control. (F) A hypothetical model showing how OsNF-YB1 functions during grain filling. In aleurone layer cells, OsNF-YB1 is imported into nucleus through heterodimerization with two aleurone-preferentially expressed factors, OsNF-YC11 or OsNF-YC12. The heterodimer OsNF-YB1–YC11 (or YC12) interacts with OsERF#115 to form a transcriptional complex, which regulates the expression of target genes bearing the GCC box, to regulate the nutrient transportation into endosperm and further the rice filling and seed size/quality. OsERF#115 binds to the GCC box and provides the sequence specificity of the transcriptional complex.