An elite RNA-DEPENDENT RNA POLYMERASE3 allele enhances preharvest sprouting resistance in rice

Abstract

A new elite allele conferring enhanced preharvest sprouting resistance via hormonal pathways is a valuable target for molecular design breeding to improve seed quality and production stability.

Dear Editor,

Preharvest sprouting (PHS) substantially threatens agricultural production by reducing both grain yield and quality, especially in regions with high rainfall or typhoon activity (Xu et al. 2022). Rice (Oryza sativa L.) varieties exhibit variations in PHS resistance. For instance, japonica rice variety Nipponbare (NIP) displays significantly higher PHS resistance compared to indica rice variety Jiafuzhan (JFZ; Fig. 1A and B). To explore the genetic basis of PHS resistance, a set of random crossing-recombinant inbred lines (RC-RILs) were developed from the cross between NIP and JianfuzhanS (JFZS), a dominant nuclear male sterile line with JFZ background (Supplementary Fig. S1). The RC-RILs displayed transgressive inheritance with significant continuous variation in germination rate (Fig. 1C). Bulked segregant analysis by deep sequencing (BSA-seq) revealed 4 major QTLs conferring PHS resistance, with the resistant alleles of qPHS1a and qPHS4a deriving from NIP, while those of qPHS1b and qPHS7a from JFZ (Fig. 1D; Supplementary Table S1). The confidence intervals of qPHS1b and qPHS7a covered Gibberellin (GA) 2-oxidases3 (GA2ox3) and Seed Dormancy4 (Sdr4), respectively, which were previously reported to regulate PHS resistance in rice (Magwa et al. 2016; Zhao et al. 2022; Fig. 1D). To isolate the gene responsible for qPHS1a, a residual heterozygous line (RHL23) carrying a heterozygous region containing qPHS1a was selected from the RC-RIL population. We observed that plants with the homozygous qPHS1a^JFZ and qPHS1a^NIP genotypes in RHL23 exhibited the highest and lowest germination rates, respectively, while those with the heterozygous genotype displayed moderate germination rate close to the mid-point (Fig. 1E). These results suggest that this QTL mainly has additive effect with the qPHS1a^NIP allele enhancing PHS resistance.

Figure 1.

Elite OsRDR3 allele enhances PHS resistance in rice. A, B) Germination performance (A) and germination rate (B) of seeds in freshly harvested panicles of NIP and JFZ (Shaxian). Scale bar, 1 cm. Data are mean ± s.e.m. (n = 10). Significant differences were calculated using two-tailed Student's t-tests. **P < 0.01. C) Frequency distribution of germination rate in RC-RILs. D) QTL analysis (qPHS1a, PVE: 3.59%; qPHS1b, PVE: 5.81%; qPHS4a, PVE: 4.04%; qPHS7a, PVE: 3.75%). E) Germination rates of individual plants from the RHL23, categorized by genotypes: qPHS1a^NIP/qPHS1a^NIP, qPHS1a^NIP/qPHS1a^JFZ and qPHS1a^JFZ/qPHS1a^JFZ (Sanya). Data are mean ± s.e.m. (n = 10). Different letters denote significant differences (P < 0.05) from a Duncan's multiple range test. F) Fine mapping of qPHS1a. Chromosomal segments homozygous for qPHS1a^JFZ and qPHS1a^NIP alleles are represented by filled and open bars, respectively (Sanya). Data are mean ± s.e.m (n = 6). G) The allelic variations in the coding region of OsRDR3 between NIP and JFZ. H) Amino acid substitution and deletion of OsRDR3 protein from JFZ. In the diagram, the small box in the middle indicates the catalytic domain. I) Germination performance of seeds in freshly harvested mature panicles of wild type, 3 osrdr3 mutant lines generated by CRISPR/Cas9 and transgenic osrdr3-c1 plants carrying the p35S::OsRDR3^NIP or p35S::OsRDR3^JFZ constructs. Scale bar, 1 cm. J)OsRDR3 mutations generated via CRISPR/Cas9-based genome editing. The osrdr3-c1 and osrdr3-c2 possess a 33-bp deletion or a 30-bp deletion in the splice site of the intron 14/exon 15 junction, respectively. And both osrdr3-c1 and osrdr3-c2 produce the same alternatively spliced proteins, in which a frameshift causes a premature stop codon resulting in a truncated protein with 898 amino acids. The osrdr3-c3 contains an insertion of a single adenine resulting in a frameshift mutation causing a premature protein with 885 amino acids. In the diagram, the small box in the middle indicates the catalytic domain. K) Germination rate of plants in I (Shaxian). Data are mean ± s.e.m. (n = 10). Different letters denote significant differences (P < 0.05) from a Duncan's multiple range test. L) Heatmap showing the log2 fold-change of relative expression levels of OsGRF genes in osrdr3-c1 relative to the WT. Data used in heatmap were derived from 3 independent RNA-seq biological replicates. M) Relative transcript abundances of OsGRF genes. Data are mean ± s.e.m. (n = 3). Significant differences were calculated using two-tailed Student's t-tests. **P < 0.01, *P < 0.05, and n.s. is not significant. N) Heatmap shows the TPM (transcripts per million) ratios of miR396 in the osrdr3-c1 relative to the NIP (WT), with the TPM values in WT set as 1. O) Germination performance of seeds in freshly harvested mature panicles of WYJ7, WYJ7-osgrf4, and transgenic WYJ7 plants carrying the p35S::OsGRF4^ngr2-GFP constructs. Scale bar, 1 cm. P) Time-course germination percentage of seeds in freshly harvested mature panicles of WYJ7, WYJ7-osgrf4, and WYJ7 p35S::OsGRF4^ngr2-GFP. Q) Distribution of germination rate among 98 rice varieties for the 2 OsRDR3 haplotypes defined by a 139-bp deletion. The center line represents the average value. The upper and lower box limits indicate the medians of the data greater than or less than the average value, respectively. The ends of the whiskers represent the absolute maximum and minimum values. Significant differences were calculated using two-tailed Student's t-tests. **P < 0.01. R) Genotyping analysis of OsRDR3. S) Germination rate of seeds in freshly harvested panicles of plants showed in R (Shaxian). Data are mean ± s.e.m. (n = 10). Different letters denote significant differences (P < 0.05) from a Duncan's multiple range test. T) Plants of TaifengB, TaifengB-OsRDR3^NIP, and TaifengB-OsRDR3^NIP-qPHS4a^NIP. Scale bar, 20 cm. U) The PHS phenotype of TaifengB, TaifengB-OsRDR3^NIP, and TaifengB-OsRDR3^NIP-qPHS4a^NIP (observed in field condition after 3 consecutive days of rain in Shaxian, 2024). Scale bar, 1 cm. V) Germination rate of seeds in freshly harvested mature panicles of TaifengB, TaifengB-OsRDR3^NIP, and TaifengB-OsRDR3^NIP-qPHS4a^NIP (Sanya, 2024, Shaxian, 2023 and 2024). Data are mean ± s.e.m. (n = 10). Different letters denote significant differences (P < 0.05) from a Duncan's multiple range test. W) The germination rate of seeds stored for 6 months after being harvested. Data are mean ± s.e.m. (n = 10). Different letters denote significant differences (P < 0.05) from a Duncan's multiple range test. Germination tests in (A, B, I, K, O, and P) were repeated with 10 panicle replicates. Photographs in (A, I, and O) were taken after 6-d imbibition. The arrowheads in (A, I, O, and U) indicated the sprouted seeds.

Fine mapping using 2,389 progenies from RHL23 narrowed down qPHS1a to a ∼20-kb region harboring a single gene, RNA-DEPENDENT RNA POLYMERASE3 (OsRDR3; Fig. 1F). The RDR family is divided into α, β, and γ subtypes, with 5 RDRs identified in rice. OsRDR1 (Wang et al. 2016), OsRDR2 (Wang et al. 2022), and OsRDR6 (Wagh et al. 2016; Liu et al. 2020) belong to the α subtype; OsRDR3 and OsRDR4 belong to the γ-clade (Jha et al. 2021). Although the roles of RDRs in viral defense as well as rice development and growth have been investigated, their specific function in regulating PHS remains unclear. Sequence comparison revealed that the OsRDR3^JFZ harbored multiple nucleotide substitutions and a 139-bp deletion spanning an exon-intron junction (Supplementary Fig. S2), leading to exon skipping and a C-terminal truncated protein with a premature stop codon (Fig. 1G and H). To validate whether OsRDR3 is the causal gene at qPHS1a, we first examined the expression levels of OsRDR3 in 2 RHL23 progeny plants: one harboring qPHS1a^JFZ and the other qPHS1a^NIP. Our analysis revealed no significant difference in OsRDR3 expression level between the 2 plants (Supplementary Fig. S3). We next generated 3 ososrdr3 mutants using CRISPR/Cas9 (Fig. 1I), which yielded C-terminally truncated OsRDR3 proteins (Fig. 1J). Although previously reported RNAi-mediated knockdown of OsRDR3 severely affects rice plant growth and development (Jha et al. 2021), our CRISPR/Cas9-generated osrdr3 mutations primarily accelerated the germination of freshly harvested seeds (Fig. 1K), with minimal impact on other aspects of rice plant growth and development (Supplementary Fig. S4A to F). This discrepancy may stem from the CRISPR target site lying downstream of the catalytic-domain-encoding region of OsRDR3 (Fig. 1J), thus preserving most catalytic function, whereas RNAi-mediated knockdown reduces OsRDR3 function more globally. Moreover, we found that the accelerated germination in ososrdr3-c1 mutant was restored to the wild-type (NIP) level by introducing OsRDR3^NIP, but not by OsRDR3^JFZ (Fig. 1I and K). These results demonstrate that OsRDR3 plays a previously uncharacterized role in seed germination, and that the elite OsRDR3^NIP allele confers enhanced PHS resistance.

Although RDRs synthesize double-stranded RNAs, which are processed into small RNAs and mediate gene silencing, the absence of RDRs can result in complex downstream effects, including both upregulation and downregulation of target genes (Liu et al. 2020). To understand how OsRDR3 regulates PHS, RNA-seq assays were performed using freshly harvested seeds of NIP and osrdr3-c1 after 24 h of imbibition. The knockout of OsRDR3 resulted in the upregulation of 1,622 genes and the downregulation of 1,562 genes (log2 fold-change > 0.5, P < 0.05; Supplementary Fig. S5A and Table S2). Among these differentially expressed genes, several were associated with ABA biosynthesis and signaling. For instance, the expression level of β-OsLCY, a gene involved in ABA biosynthesis, was reduced in osrdr3-c1, and this reduction was reported to an increased PHS (Supplementary Fig. S5B; Fang et al. 2008). Similarly, OsPP2C51, a positive regulator of seed germination acting downstream of the receptor in ABA signaling, showed increased expression in osrdr3-c1 (Supplementary Fig. S5B; Bhatnagar et al. 2017). In addition, we identified several genes associated with seed germination from our RNA-seq data (Supplementary Fig. S5C). For example, OsZDS, which inhibits seed germination, was downregulated in osrdr3-c1 (Supplementary Fig. S5C; Fang et al. 2008), while SD6, which promotes seed germination, was upregulated (Supplementary Fig. S5C; Xu et al. 2022). These genes are also linked to ABA biosynthesis and signaling. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) assays further confirmed that these seed germination-related genes are regulated by OsRDR3 (Supplementary Fig. S5D to K). Moreover, the endogenous ABA content was lower in osrdr3-c1 grains compared to NIP (Supplementary Fig. S5L). These results indicated that OsRDR3 regulates PHS via modulating the ABA biosynthesis and signaling pathway.

Notably, several OsGRF genes, including OsGRF4, OsGRF5, OsGRF7, and OsGRF10, were downregulated in the osrdr3-c1 mutant (Fig. 1L). Growth-regulating factors (GRFs) are plant-specific transcription factors that play essential roles in controlling various developmental processes and responses to both biotic and abiotic stimuli (Piya et al. 2020). Many GRF genes contain a target site for microRNA miR396, which can suppress GRF expression in mature tissues (Debernardi et al. 2012). RT-qPCR assays of all the OsGRF genes showed that the mRNA abundances were reduced in the osrdr3-c1 mutant, except for OsGRF9 (Fig. 1M). Correspondingly, miRNA-seq assays showed that the abundance of miR396 was increased in the osrdr3-c1 mutant (Fig. 1N; Supplementary Table S3). Our previous research results showed that increasing the expression of OsGRF4 enhances nitrogen use efficiency (NUE) and grain yield (Li et al. 2018). However, the role of OsGRF4 in PHS regulation remains unknown. Using wild-type plants WYJ7 (Wuyunjing 7, a japonica rice variety with the OsRDR3^NIP allele), osgrf4 mutant (WYJ7-osgrf4), and OsGRF4 overexpression transgenic plant (WYJ7 35S::OsGRF4^ngr2-GFP; Li et al. 2018), we observed that the osgrf4 mutant showed faster germination in freshly collected panicles compared to WYJ7, whereas the overexpression of OsGRF4 delayed germination (Fig. 1O and P). RT-qPCR assays revealed that seed germination-related genes regulated by OsRDR3 were also influenced by OsGRF4 (Supplementary Figs. S5D to K and S6A to H). These results indicate that OsGRF4 is involved in OsRDR3-mediated PHS resistant regulation, and increased OsGRF4 activity simultaneously enhances both NUE and PHS resistance in rice.

In hybrid rice seed production, PHS can result in average yield losses of 10% to 20%, with losses reaching up to 50% in certain years (Hu et al. 2003). Genotyping analysis of 98 rice varieties revealed that the 139-bp deletion characteristic of OsRDR3^JFZ is predominantly observed in indica varieties and is associated with reduced PHS resistance (Fig. 1Q; Supplementary Tables S4 and S5). Examination of OsRDR3 genotypes in the maintainer lines of several prominent indica sterile lines crucial for hybrid rice breeding in China showed that nearly all carried the OsRDR3^JFZ allele (Fig. 1R) and exhibited relatively lower PHS resistance (Fig. 1S). TaifengA and YexiangA are 2 dominant sterile lines in high-quality hybrid rice breeding in China. We introduced the OsRDR3^NIP allele into the maintainer lines (TaifengB and YexiangB) through backcrossing (Fig. 1T; Supplementary Fig. S7A). Both TaifengB-OsRDR3^NIP and YexiangB-OsRDR3^NIP exhibited markedly enhanced resistance to PHS compared to TaifengB and YexiangB (Fig. 1U and V; Supplementary Fig. S7B and C). Importantly, the introduction of OsRDR3^NIP improved PHS resistance without affecting the germination of stored seeds (Fig. 1W; Supplementary Fig. S7D) or other agronomic traits related to yield and grain shape (Supplementary Figs. S7E to K and S8A to H). As mentioned above, the qPHS4a^NIP allele from NIP also contributes to improving resistance to PHS (Fig. 1D). By introducing qPHS4a^NIP into TaifengB-OsRDR3^NIP, the PHS resistance of TaifengB was further enhanced (Fig. 1T to W).

In summary, our work suggests that OsRDR3 regulates PHS resistance partly by modulating ABA biosynthesis and signaling pathway. Additionally, OsGRF4, a positive regulator of NUE and grain yield (Li et al. 2018), is identified as a target of OsRDR3 that also represses PHS. We also propose a breeding strategy utilizing the elite OsRDR3^NIP allele in indica hybrid rice parents to reduce PHS risk, thereby enhancing the safety of hybrid rice seed production.

Author contributions

K.W., W.W., and X.F. designed the experiments; H.G., Z.C., S.W., B.J., J.Z., Y.G., S.Z., S.L., D.M., L.L., W.F., Q.L., W.S., L.H., X.X., C.W., and Y.Y. carried out the experiments and analyzed the data; S.C. and J.C. detected ABA content; K.W., S.W., and W.W. wrote the manuscript. All authors read and approved of the manuscript.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1. Flowchart of building the RC₄F₆ population for QTL analysis.

Supplementary Figure S2. Sequence variations between JFZ and NIP at OsRDR3.

Supplementary Figure S3. The relative expression levels of OsRDR3.

Supplementary Figure S4. Characterization of NIP and osrdr3-c1.

Supplementary Figure S5.  OsRDR3 regulates PHS via modulating the ABA-related pathway.

Supplementary Figure S6. Seed-germination-related genes were regulated by OsGRF4.

Supplementary Figure S7. The OsRDR3^NIP allele enhances PHS resistance in YexiangB.

Supplementary Figure S8. Characterization of TaifengB and TaifengB-OsRDR3^NIP.

Supplementary Table S1. Linkage analysis between INDEL markers and candidate QTLs of PHS.

Supplementary Table S2. RNA-Seq gene expression levels and differential expression analysis.

Supplementary Table S3. miRNA expression profile of NIP and rdr3-c1.

Supplementary Table S4. Germination rates of 98 varieties selected from 3,000 Rice Genomes Project.

Supplementary Table S5. The primer sequences used for map-based cloning, genotyping, transgene constructs and RT-qPCR assays.

Supplementary Materials and Methods.

Funding

This work was supported by funding from National Key Research and Development Program of China (2023YFF100501); The Strategic Priority Research Program of the Chinese Academy of Sciences (XDB0630102); Jiangsu Province Government (BE2023331); The National Natural Science Foundation of China (U24A20392, 3222206, 32301862); The New Cornerstone Investigation Program (NCI202234); The China Postdoctoral Science Foundation (2022M713360); Shanghai Agriculture Applied Technology Development Program (T20220106).

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• OsRDR1 Gramene: Os02g0736200

• OsRDR1 Araport: Os02g0736200

• OsRDR2 Gramene: Os04g0465700

• OsRDR2 Araport: Os04g0465700

• OsRDR6 Gramene: Os01g0527600

• OsRDR6 Araport: Os01g0527600

• OsRDR3 Gramene: Os01g0197900

• OsRDR3 Araport: Os01g0197900

• OsRDR4 Gramene: Os01g0198000

• OsRDR4 Araport: Os01g0198000