The phosphatidylethanolamine-binding proteins OsMFT1 and OsMFT2 regulate seed dormancy in rice

Abstract

Seed dormancy is crucial for optimal plant life-cycle timing. However, domestication has largely diminished seed dormancy in modern cereal cultivars, leading to challenges such as preharvest sprouting (PHS) and subsequent declines in yield and quality. Therefore, it is imperative to unravel the molecular mechanisms governing seed dormancy for the development of PHS-resistant varieties. In this study, we screened a mutant of BASIC HELIX-LOOP-HELIX TRANSCRIPTION FACTOR4 (OsbHLH004) with decreased seed dormancy and revealed that OsbHLH004 directly regulates the expression of 9-CIS-EPOXYCAROTENOID DIOXYGENASE3 (OsNCED3) and GIBBERELLIN 2-OXIDASE6 (OsGA2ox6) in rice (Oryza sativa). Additionally, we determined that two phosphatidylethanolamine-binding proteins, MOTHER OF FT AND TFL1 and 2 (OsMFT1 and OsMFT2; hereafter OsMFT1/2) interact with OsbHLH004 and Ideal Plant Architecture 1 (IPA1) to regulate their binding capacities on OsNCED3 and OsGA2ox6, thereby promoting seed dormancy. Intriguingly, FT-INTERACTING PROTEIN1 (OsFTIP1) interacts with OsMFT1/2 and affects their nucleocytoplasmic translocation into the nucleus, where OsMFT1/2-OsbHLH004 and OsMFT1/2-IPA1 antagonistically modulate the expression of OsNCED3 and OsGA2ox6. Our findings reveal that OsFTIP1-mediated inhibition of nuclear translocation of OsMFT1/2 and the dynamic transcriptional modulation of OsNCED3 and OsGA2ox6 by OsMFT1/2-OsbHLH004 and OsMFT1/2-IPA1 complexes in seed dormancy in rice.

Figure 1.

OsbHLH004 inhibits seed dormancy in rice. A) CRISPR/Cas9-mediated target mutagenesis of OsbHLH004. Exons and introns of OsbHLH004 are indicated by black boxes and lines, respectively. Each alignment between wild-type and mutated sequences containing the target sites is shown below the schematic diagram. The target sequences adjacent to the underlined protospacer adjacent motifs (PAMs) are indicated in red in wild-type sequences. B) Germination phenotypes of fresh seeds from panicles of wild type, Osbhlh004 mutants and Osbhlh004-1 gOsbHLH004-4HA at 30 d after pollination (DAP), after imbibition in water for 5 d. Scale bars, 1 cm. C) The time-course germination percentage of fresh seeds collected in different genetic backgrounds. Error bars indicate Sd (n = 3). D and E) Expression patterns of OsbHLH004 in various tissues (D) or developing seeds (E) in rice detected by quantitative real-time PCR (RT-qPCR). Error bars indicate Sd (n = 3). The expression values were normalized to the Ubq expression and shown relative to the highest expression in each experiment, which was set to 100%. F) The subcellular localization of OsbHLH004-GFP in rice protoplasts. GFP, GFP fluorescence; NLS-mCherry, mCherry fluorescence of a nuclei marker; Merge, merge of GFP, mCherry, and bright field. Scale bars, 10 μm.

Figure 2.

OsbHLH004 directly regulates the expression of OsNCED3 and OsGA2ox6. A) Volcano plots of total gene expression profiles in Osbhlh004-1 compared with wild type. B) RT-qPCR analysis of OsNCED3 and OsGA2ox6 expression in wild type, Osbhlh004 and Osbhlh004-1 gOsbHLH004-4HA. Seeds at 20 to 25 DAP of each genotype were collected for RT-qPCR analysis. The levels of each gene expression normalized to Ubq expression are shown relative to the level in the wild type, which was set to 1. Error bars denote Sd (n = 3). Asterisks indicate significant differences between Osbhlh004-1 and wild type (Student's t-test, P < 0.01). C) ABA contents in wild-type and Osbhlh004-1 seeds at 20 to 25 DAP. Error bars represent Sd (n = 3). The asterisk indicates a significant difference between Osbhlh004-1 and wild type (Student's t-test, P < 0.01). D) Endogenous GA levels in wild-type and Osbhlh004-1 seeds at 20 to 25 DAP. #, undetectable GAs. Error bars represent Sd (n = 3). Asterisks indicate significant differences between Osbhlh004-1 and wild type (Student's t-test, P < 0.01). E) Representative photographs of firefly luciferase fluorescence signals reveal the OsbHLH004 repressive activity to OsNCED3 and OsGA2ox6 in N. benthamiana leaves. F and G) ChIP analysis of OsbHLH004 association with the genomic regions of OsNCED3F) and OsGA2ox6G). Seeds at 20 to 25 DAP of wild type and Osbhlh004-1 gOsbHLH004-4HA #6 were collected for ChIP analysis. The enrichment fold of each fragment was calculated by first normalizing the amount of a target DNA fragment against a genomic fragment of Ubq as an internal control and then normalizing the values of Osbhlh004-1 gOsbHLH004-4HA against that of wild-type plants. Error bars denote Sd (n = 3). Asterisks indicate the CANNTG motifs in the promoters of OsNCED3 and OsGA2ox6, respectively. Black asterisks bars represent E-boxes, while red asterisks denote G-boxes.

Figure 3.

OsbHLH004 interacts with OsMFT1 and OsMFT2. A) In vitro MBP pull-down assay of the interactions between OsbHLH004 and OsMFT1/2. OsbHLH004-4HA extracted from the 20 to 25 DAP seeds of Ubi:OsbHLH004-4HA was incubated with immobilized MBP, MBP-OsMFT1, or MBP-OsMFT2. Immunoblot analyses were performed using anti-HA or anti-MBP antibody. Input, 5% of the OsbHLH004-4HA extracted from the 20 to 25 DAP seeds of Ubi:OsbHLH004-4HA. B) BiFC analysis of the interactions between OsbHLH004 and OsMFT1/2 in rice protoplasts. MBP was used as a negative control. Merge, merge of EYFP, NLS-CFP, and bright-field (BF) images. Scale bars, 10 µm. C) LCI assay of interactions between OsbHLH004 and OsMFT1/2. Indicated fusion pairs were coexpressed in N. benthamiana leaves. D and E) CoIP assay showing the in vivo interaction between OsbHLH004 and OsMFT1 D) or OsMFT2 E). Total proteins extracted from 20 to 25 DAP seeds of gOsMFT1-3FLAG and gOsbHLH004-4HA gOsMFT1-3FLAG or gOsMFT2-3FLAG and gOsbHLH004-4HA gOsMFT2-3FLAG were used to perform CoIP assay by anti-HA magnetic beads. The input and coimmunoprecipitated protein were detected by anti-FLAG (upper) antibody or anti-HA (lower) antibody.

Figure 4.

OsMFT1/2 represses the binding capacity of OsbHLH004 on OsNCED3 and OsGA2ox6. A) Germination phenotypes of fresh seeds from panicles of wild type, Osmft1, Osmft2, Osmft1/2, and OsMFT1&2-Cas9. Panicles at 30 DAP were imbibed in water for 5 d. Scale bar, 1 cm. B) The time-course germination percentage of fresh seeds collected from the above genotypes, respectively. Error bars indicate Sd (n = 3). C) RT-qPCR analysis of OsNCED3 and OsGA2ox6 expression in wild type, Osmft1, Osmft2, Osmft1/2, and OsMFT1&2-Cas9. Seeds at 20 to 25 DAP of each genotype were collected for RT-qPCR analysis. The levels of each gene expression normalized to Ubq expression are shown relative to the level in wild-type plants, which was set to 1. Error bars denote Sd (n = 3). Asterisks indicate significant differences between single or double mutants and wild type (Student's t-test, P < 0.01). D) The schematic representation of the effector and reporter constructs used in the transfection assays in E) and F). E and F) OsMFT2 alleviates the transcriptional inhibition of OsbHLH004 on the promoters of OsNCED3E) and OsGA2ox6F) in rice protoplasts. The relative luciferase activity was first calculated by LUC/REN and then normalized by the control vector, which was set to 1. Error bars denote Sd (n = 3). Different letters indicate significant differences (P < 0.05, one-way ANOVA). G and H) ChIP analysis of OsbHLH004 association with the chromatins of OsNCED3 and OsGA2ox6 in the presence or absence of OsMFT1/2. Developing seeds at 20 to 25 DAP of Osbhlh004-1 gOsbHLH004-4HA, Osmft1/2 Osbhlh004-1 gOsbHLH004-4HA, and wild type were collected for ChIP analysis. The enrichment fold of each fragment was shown as the relative of the value of a target DNA in Osmft1/2 Osbhlh004-1 gOsbHLH004-4HA to that in Osbhlh004-1 gOsbHLH004-4HA. Error bars represent Sd (n = 3). Asterisks indicate significant differences between the indicated pairs (Student's t-test, P < 0.01).

Figure 5.

OsMFT1/2 interacts with IPA1. A) In vitro pull-down assay of the interactions between OsMFT1/2 and IPA1. IPA1-4HA generated from rice protoplasts was incubated with immobilized MBP, MBP-OsMFT1, or MBP-OsMFT2. Immunoblot analysis was performed using anti-HA antibody or anti-MBP antibody. Input, 5% of the IPA1-4HA generated from rice protoplasts. B) BiFC analysis of the interaction between OsMFT1/2 and IPA1 in rice protoplasts. MBP was used as a negative control. Scale bars, 10 µm. C) LCI assay of interaction between OsMFT1/2 and IPA1. Indicated fusion pairs were coexpressed in N. benthamiana leaves. D and E) CoIP assays between OsMFT1/2 and IPA1. Nuclear proteins extracted from seeds at 20 to 25 DAP of Osmft1-1 gOsMFT1-3FLAGD) or Osmft2-1 gOsMFT2-3FLAGE) were incubated with anti-IPA1 or preimmune serum (anti-IgG) antibody. The input and coimmunoprecipitated protein were detected by anti-FLAG (upper) antibody or anti-IPA1 (lower) antibody.

Figure 6.

OsMFT1/2 enhances the transcriptional activity of IPA1 on OsNCED3 and OsGA2ox6. A) Germination phenotypes of fresh seeds from panicles of ZH11 (wild type) and ipa1-10. Panicles at 30 DAP were imbibed in water for 4 d. Scale bar, 2 cm. B) The time-course germination percentage of fresh seeds of wild type and ipa1-10. Error bars represent Sd (n = 3). C) RT-qPCR analysis of OsNCED3 and OsGA2ox6 expression in 20 to 25 DAP seeds of wild type and ipa1-10. Error bars denote Sd (n = 3). Asterisks indicate significant differences in the expression levels between ipa1-10 and wild type (Student's t-test, P < 0.001). D and E) OsMFT2 enhances the transcriptional activation activities of IPA1 on the promoter of OsNCED3D) or OsGA2ox6E) in rice protoplasts. Different letters indicate significant differences (P < 0.05, one-way ANOVA). Error bars indicate Sd (n = 3). F) ChIP-qPCR analysis of IPA1 binding to the promoter regions of OsNCED3 and OsGA2ox6. Developing seeds at 20 to 25 DAP of wild type were precipitated by anti-IPA1 or anti-IgG antibodies. In the scheme of OsNCED3 or OsGA2ox6 promoter, black triangles indicate the GTAC motifs, while the red one represents TGGGCC/T motif. Error bars denote Sd (n = 3). G) ChIP analysis of IPA1 binding to OsNCED3 or OsGA2ox6 in the presence or absence of OsMFT1/2. Developing seeds at 20 to 25 DAP of wild type or Osmft1/2 were precipitated by anti-IPA1 or anti-IgG antibodies. Error bars denote Sd (n = 3). Asterisks indicate significant differences of OsNCED3 and OsGA2ox6 enrichments in Osmft1/2 and wild type (Student's t-test, P < 0.01).

Figure 7.

OsFTIP1 represses seed dormancy by restricting the nucleocytoplasmic distribution of OsMFT1/2. A) Germination phenotypes of fresh seeds from panicles of wild type, Osftip1-1, and Osftip1-1 gOsFTIP1-HA. Panicles at 30 DAP were imbibed in water for 5 d. Scale bar, 1 cm. B) The time-course germination percentage of fresh seeds of the wild type, Osftip1, and Osftip1-1 gOsFTIP1-HA. Error bars indicate Sd (n = 3). C) RT-qPCR analysis of OsNCED3 and OsGA2ox6 expression in wild type, Osftip1, and Osftip1-1 gOsFTIP1-HA. Seeds at 20 to 25 DAP of each genotype were collected for RT-qPCR analysis. Error bars denote Sd (n = 3). Asterisks indicate significant differences in the expression levels between Osftip1 and wild type (Student's t-test, P < 0.01). D) In vitro MBP pull-down assay of the interaction between OsFTIP1 and OsMFT1/2. OsFTIP1-HA extracted from seeds of Osftip1-1 gOsFTIP1-HA was incubated with immobilized MBP, MBP-OsMFT1, or MBP-OsMFT2. Immunoblot analysis was performed using anti-HA antibody or anti-MBP antibody. Input, 5% of the OsFTIP1-HA extracted from Osftip1-1 gOsFTIP1-HA. E) CoIP assays between OsFTIP1 and OsMFT1/2. Total proteins were extracted and immunoprecipitated with IgG or anti-FLAG magnetic beads. The input and coimmunoprecipitated protein were detected by anti-HA antibody. F) OsMFT2 abundance in whole-cell extracts, cytoplasmic and nuclear fractions of seeds of wild type, Osftip1-1, Osftip1-2, and Osftip1-1 gOsFTIP1-HA in the gOsMFT2-3FLAG background. OsMFT2 abundance was determined by anti-FLAG antibody. The purity of the cytosolic fractions and nuclear was examined by western blot analysis using anti-Actin and anti-Histone 3 (H3) antibodies, respectively.

Figure 8.

A proposed working model describing the functions of OsFTIP1, OsMFT1/2, OsbHLH004, and IPA1 during seed dormancy. During seed maturation, OsFTIP1 expression is downregulated and leads to the release of OsMFT1/2 from the cytosol to the nucleus. After being transported to the nucleus, on the one hand, OsMFT1/2 interacts with OsbHLH004 to alleviate the inhibition of OsbHLH004 on the transcription of OsNCED3 and OsGA2ox6; on the other hand, OsMFT1/2 interacts with IPA1 and enhances its transcriptional binding capacity to the above genes. These two effects result in the expression incrementation of OsNCED3 and OsGA2ox6, thereby promoting seed dormancy.