Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2026 Jun 25;201(3):kiag435. doi: 10.1093/plphys/kiag435

Loss-of-function of MIR172b and prime editing of SNB reveal a regulatory module underlying cleistogamy in rice

Su-Hyeon Shim 1,#, Rihua Piao 2,3,#, Dong-Yoon Seo 4,#, Sang-Kyu Lee 5,6,#, So Young Park 7, Backki Kim 8,9, Hyun-Sung Leem 10, Dong-Hoon Jeong 11,3,✉,4, Hee-Jong Koh 12,✉,4, Jong-Seong Jeon 13,✉,4
PMCID: PMC13360268  PMID: 42348922

Abstract

Cleistogamy, or self-fertilization of a closed flower, can limit unintended gene flow and may contribute to varietal purity in rice (Oryza sativa), with potential value for transgene containment. We previously identified the natural cleistogamous mutant lodiculeless spikelet (ld). Here, using map-based cloning, we show that ld carries a 4.6-kb deletion encompassing the entire MICRORNA806a (MIR806a) precursor and the upstream region of the MIR172b locus. Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nuclease 9-mediated editing of the MIR806a or MIR172b locus demonstrated that the loss of accumulation of miR172b, but not miR806a, is responsible for the cleistogamy phenotype of ld plants. Small RNA sequencing confirmed that miR172b is nearly absent in ld mutants. Among the five AP2-like genes harboring miR172b target sites, only SUPERNUMERARY BRACT (SNB) transcripts accumulated to significantly higher levels in young ld panicles than the wild type. Prime editing of the miR172-binding site in SNB generated a miR172-resistant SNB transcript isoform that reproduced the vestigial lodicule phenotype, indicating that the derepression of SNB transcript accumulation is sufficient to alter lodicule development. Histological analysis of rice harboring the GUS reporter gene driven by the MIR172b or SNB promoter revealed a strong overlap between the MIR172b and SNB expression domains in developing lodicules, supporting their regulatory relationship. Together, these results identify the miR172b-mediated repression of SNB transcript abundance as an important regulatory module for lodicule development and cleistogamy in rice. The agronomically neutral ld alleles represent valuable genetic resources for developing cleistogamous rice cultivars.

Introduction

Cleistogamy is a mode of pollination in which flowers remain closed, resulting in self-fertilization. In grasses, flower opening is primarily controlled by lodicules. These specialized organs swell rapidly at anthesis and generate the force needed to pry open the spikelet. When lodicules fail to swell, lack proper cellular structure, or are transformed into other organ types (such as palea-like tissue), the flower remains closed, leading to cleistogamy (Yoshida et al. 2007; Nair et al. 2010). Cleistogamy can limit outcrossing between cultivated species and their wild relatives, may contribute to varietal purity in autogamous crops, and has potential applications in reducing transgene flow and pathogen exposure during anthesis.

Cleistogamy has evolved in approximately 300 plant species from 60 families (Lord 1981). This trait is particularly prevalent in grasses, where several crop species such as barley (Hordeum vulgare), wheat (Triticum aestivum), and rice (Oryza sativa) have been studied to elucidate the genetic basis of cleistogamy. In barley, the cleistogamy 1 (cly1) mutant exhibits tightly closed spikelets during anthesis due to defective lodicule function (Kurauchi et al. 1993; Turuspekov et al. 2004). This phenotype is caused by a mutation in the Cly1 gene, an ortholog of Arabidopsis (Arabidopsis thaliana) AP2, which regulates floral organ identity (Nair et al. 2010). In wheat, three orthologs of Cly1, namely TaAP2-A, TaAP2-B, and TaAP2-D, are highly expressed in lodicules, and their transcripts are targeted for cleavage by the microRNA (miR) miR172 (Ning et al. 2013). These findings establish the miR172–AP2 regulatory module as a conserved mechanism controlling lodicule development and cleistogamy in grasses.

miRNAs are a class of small non-coding RNAs that negatively regulate target gene expression at the post-transcriptional level (Rogers and Chen 2013). Functional mature miRNAs, which are 20 to 24 nucleotides (nt) long, are responsible for the cleavage or translational repression of their target transcripts. Plant miRNAs play important roles in growth, development, and environmental responses. In rice, 604 pre-MIRNAs generating 738 mature miRNAs have been reported. Among these, miR172 is a highly conserved miRNA that targets the transcripts of AP2-like transcription factor genes (Aukerman and Sakai 2003). miR172 and its AP2-like targets play crucial roles in developmental phase transitions and floral organ identity (Chen 2004; Zhu and Helliwell 2011; Spanudakis and Jackson 2014; Teotia and Tang 2015). In Arabidopsis, AP2 is predominantly expressed in sepals and petals, whereas miR172 is enriched in anthers and carpels (Chen 2004; Wollmann et al. 2010). The barley cly1 mutant harbors a synonymous substitution at the miR172 target site within the transcript of the AP2-like gene Cly1, suppressing miRNA-guided cleavage and leading to cleistogamy, underscoring the importance of precise miR172-mediated regulation for lodicule development (Nair et al. 2010).

In rice, several AP2-like proteins have been characterized as crucial regulators of flower organ development, flowering time, and grain-related traits. One of these, SUPERNUMERARY BRACT (SNB), mediates the transition from a spikelet meristem to a floral meristem; its dysfunction leads to the formation of extra bract-like organs (Lee et al. 2007). SNB has also been implicated in the regulation of flowering time, grain size, and seed shattering (Lee et al. 2014; Jiang et al. 2019; Ma et al. 2019). INDETERMINATE SPIKELET 1 (IDS1) acts synergistically with SNB to regulate inflorescence architecture and floral meristem development (Lee and An 2012). The snb ids1 double mutant produces more bracts, including rudimentary glumes and palea/lemma, as well as elongated lodicules and a reduced number of stamens compared to the wild type (WT) (Lee and An 2012). RICE STARCH REGULATOR 1 (RSR1) negatively regulates the expression of type I starch biosynthesis genes and affects grain size, amylose content, and heading date (Fu and Xue 2010; Sun et al. 2022). SHATTERING ABORTION 1 (SHAT1) is essential for abscission zone development and seed shattering (Zhou et al. 2012). Another AP2-like protein, SALT AND ABA RESPONSE ERF 1 (SAE1), regulates ABA signaling by directly activating ABA-INSENSITIVE 5 expression, enhancing seed germination and salinity stress tolerance (Li et al. 2022b). Although the genes encoding these AP2-like proteins harbor predicted miR172-binding sites, and recent work has shown that the miR172a–SNB module contributes to disease resistance in rice (Wang et al. 2025), studies of rice cleistogamy mutants have not directly implicated specific MIR172 family members or their AP2-like targets in lodicule-mediated floral opening.

Several cleistogamy mutants have been reported in rice. In cl7 plants, weakened lodicule swelling results in cleistogamy despite otherwise normal floral organ development (Ni et al. 2014). The cl7 phenotype is caused by a mutation in DENSE AND ERECT PANICLE 2, which encodes a conserved plant-specific protein with no domains of known function; however, cl7 exhibits undesirable pleiotropic effects (Li et al. 2010). The superwoman 1-cleistogamy 1 (spw1-cls1) mutant carries a missense mutation in the class-B MADS-box gene SPW1 that converts lodicules into lodicule-glume mosaic organs, but its cleistogamy phenotype is abolished under cold conditions (Yoshida et al. 2007). The temperature-stable allele spw1-cls2 maintains cleistogamy even in cold environments but shows lower fertility than the WT (Lombardo et al. 2017). These mutants provide valuable insights but are of limited use for agriculture and breeding due to pleiotropic defects and environmental instability. Moreover, the contributions of individual MIR172 family members to rice lodicule development, particularly through natural regulatory variation, remain largely unexplored.

One natural cleistogamy mutant in rice, lodiculeless spikelet (ld), is characterized by markedly reduced lodicules (Maeng et al. 2006). The ld mutation has no deleterious effects on major agronomic traits, including plant height, panicle structure, and yield (Won et al. 1998). Genetic mapping positioned the LD locus at the distal end of chromosome 1, suggesting that the causal factor is not the rice ortholog to barley Cly1 based on the syntenic relationship between the barley and rice chromosomes (Maeng et al. 2006). In this study, we narrowed down the LD locus to a 23-kb region containing a 4,572-bp deletion that spans the entire MIR806a locus and the promoter region of MIR172b. Analysis of knockout mutants of MIR806a or MIR172b generated via clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nuclease 9 (Cas9)-mediated editing demonstrated that miR172b, but not miR806a, is essential for lodicule development. Among five AP2-like genes with predicted miR172-binding sites, only the transcripts of SNB markedly accumulated in the young panicles of ld plants. A miR172-resistant SNB variant created by prime editing of the miR172 target site in SNB recapitulated the vestigial lodicule phenotype observed in ld plants, and histological staining of rice harboring promoter:GUS revealed strong MIR172b and SNB promoter activity in developing lodicules. Together, these results demonstrate that loss-of-function of MIR172b induces cleistogamy by derepressing SNB, and they reveal a naturally occurring deletion in the MIR172b promoter as a genetic basis for cleistogamy. These findings provide a framework for precision breeding strategies aimed at modulating floral opening in rice.

Results

Isolation of the rice autogamy gene LD

We previously identified ld, a natural rice mutant with spikelets lacking normal lodicules, resulting in cleistogamy (Maeng et al. 2006). To investigate the causal mutation, we performed map-based cloning by constructing a high-resolution map using 1,027 F2 and 1,673 F3 segregating progeny from a cross between ld (in the japonica cultivar Ilpum background) and indica cultivar Milyang23. Fine mapping delimited the ld mutation to an approximately 23-kb genomic interval between STS markers S131588 and S108796 (Ji et al. 2012) (Fig. 1a). In silico analysis revealed that the region contains one predicted gene (LOC_Os01g74130) and two MIRNA loci (MIR172b and MIR806a) in the japonica reference genome (cultivar Nipponbare). We sequenced the 23-kb genomic region in the ld mutant and WT cultivar Ilpum and compared them to the sequence of the corresponding genomic region from the publicly available Nipponbare genome. This analysis identified a 4,572-bp deletion and a 40-bp insertion in ld; this deletion encompasses the entire MIR806a genomic region and the MIR172b promoter region (Fig. 1b). BLAST analysis of the 40-bp insertion sequence against the rice reference genome (IRGSP-1.0) revealed that 38 of 40 bp are identical to sequences within long terminal repeat (LTR) retrotransposon elements present at multiple genomic loci, indicating that the insertion represents filler DNA captured during double-strand break (DSB) repair via illegitimate recombination. The 23-kb region in WT Ilpum is identical to that of Nipponbare.

Figure 1.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Map-based cloning of the LD locus. (a) Fine mapping and diagram of the LD locus. Primary mapping between the markers located at 178.1 Mb and 181.8 Mb on chromosome 1 (n = 397), and fine mapping (n = 2,700) between the markers S108796 and S131588. One annotated gene (LOC_Os01g74130) and two MIRNA loci (MIR806a and MIR172b) within the interval are highlighted. The rectangle indicates the 4,572-bp deleted region in ld identified by DNA walking. (b) Diagram of the deleted region in ld-1 and the corresponding interval in the wild type. Gray boxes labeled “a” and “b” indicate the primers used to amplify across the junction of the deleted fragment. The 40-bp inserted sequence in ld-1 is indicated.

To further characterize the deleted region, we analyzed cis-regulatory elements in the MIR172b promoter (2.5 kb upstream of the pri-miRNA) using the PlantPAN 4.0 platform (https://plantpan.itps.ncku.edu.tw) (Chow et al. 2019; Chow et al. 2024). This analysis identified putative binding motifs for several transcription factor families, including AP2, MADS-box, B3, SBP (SPL), NF-YB, bZIP, NAC, and TCP transcription factors (Figure S1), suggesting that the deleted MIR172b promoter region contains multiple cis-regulatory elements important for proper MIR172b expression during floral organ development.

CRISPR/Cas9-mediated functional validation of the LD gene

To determine which of the two miRNAs is responsible for the ld phenotype, we performed CRISPR/Cas9 multiplex genome editing to delete the sequences corresponding to the precursor of MIR806a or MIR172b. We designed two single guide RNAs (sgRNAs) for each MIRNA locus, one targeting the region immediately upstream of the precursor sequence and the other targeting either the downstream region or the mature miRNA site, to ensure complete functional disruption, potentially resulting in the deletion of the entire genomic region producing each precursor (Fig. 2a). We obtained 14 and 22 transgenic rice plants in the japonica cultivar Dongjin background expressing the CRISPR/Cas9 constructs targeting MIR806a or MIR172b, respectively. Among these, only one plant for MIR806a (designated mir806a-4) and two plants for MIR172b (designated mir172b-4 and mir172b-5) carried large deletions in the respective locus based on Sanger sequencing of PCR products covering the intended targeted region (Fig. 2, b and c; Table S1). The mir806a-4 mutant carries a 502-bp deletion that removes the entire MIR806a precursor sequence, including Target 1 and Target 2 (Fig. 2b). The mir172b-4 mutant carries a 2-bp deletion at Target 1 and a 74-bp deletion covering Target 2, while mir172b-5 carries a 31-bp deletion at Target 1 and a 170-bp deletion at Target 2 (Fig. 2c). In addition, we identified mir172b-6, with 1-bp insertions at both target sites (Fig. 2c). To avoid off-target effects, we analyzed all four sgRNA target sites using offTarget (Doench et al. 2016; http://skl.scau.edu.cn/offtarget) and listed all possible off-target sites (Off-score > 0.05) (Table S2). All MIR806a targets showed relatively low off-target potential, and no possible off-target sites were located within coding regions. By contrast, four putative off-target sites were identified for MIR172b targets, three of which possessed protospacer-adjacent motifs (PAMs) and relatively high Off-score values (> 0.1). We confirmed that no unintended editing occurred at any of the three putative off-target sites in mir172b-4 or mir172b-5 (Table S2; Figure S2).

Figure 2.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Generation of CRISPR/Cas9-mediated rice mutants defective in MIR806a or MIR172b. (a) Diagram showing the target sites for CRISPR/Cas9-mediated editing of the MIR806a or MIR172b locus. The target site upstream of each pri-MIRNA sequence is labeled “Target 1”, and the one downstream is labeled “Target 2”. PAMs are shown in blue. Primers used for genotyping are indicated with arrows. (b) Genomic sequence of the mir806a-4 mutant harboring a deletion of the MIR806a locus. The deleted nucleotides in the mutant are shown as red hyphens. (c) Genomic sequences of the mir172b mutants carrying different deletions of the MIR172b locus. Portions of the MIR172b locus are deleted in mir172b-4 and mir172b-5, and a single nucleotide is inserted within the mature miR172b region in mir172b-6. Deleted nucleotides are shown as red hyphens and the 1-bp insertion as bold red text; the mature miR172b region is highlighted in orange.

All mir172b mutants showed a clear cleistogamy phenotype similar to that of ld, whereas mir806a-4 plants had a normal spikelet morphology indistinguishable from WT Dongjin (Fig. 3). We therefore designated mir172b-4, mir172b-5, and mir172b-6 as ld-2, ld-3, and ld-4, respectively, while the original ld mutant was designated as ld-1. The spikelets of all mir172b mutant lines lacked normal lodicules, whereas Dongjin and mir806a-4 developed intact lodicules (Fig. 3). Cleistogamous flowers release pollen without opening. To confirm the cleistogamous phenotype of the ld mutants, we observed dehulled spikelets at 3 to 6 d after fertilization. Indeed, all mir172b mutant lines retained anthers inside the spikelet, producing normal seed set after anthesis (Fig. 3). These genotypes and phenotypes were stably inherited across subsequent generations.

Figure 3.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Phenotypic analysis of CRISPR/Cas9-mediated mir806a and mir172b mutants. Representative photographs of mature panicles, spikelets, and floral organs at anthesis and developing seeds (top to bottom) from WT Dongjin, ld-2 (mir172b-4), ld-3 (mir172b-5), ld-4 (mir172b-6), and mir806a-4 plants are shown. Arrowheads represent missing lodicules in the ld mutants. Red scale bars, 20 mm; white scale bars, 2 mm.

miR172b is specifically downregulated in ld mutants

To investigate whether the loss of miR172b accumulation underlies the ld phenotype, we sequenced and analyzed small RNAs from the young panicles of WT Ilpum and two mutant alleles (ld-1 and ld-2) at stage Sp5, i.e. during inner organ formation. Four MIR172 loci are annotated within the rice MIR172 family, generating three mature miRNA species: miR172ad, miR172b, and miR172c. Because these mature miRNAs share highly similar sequences and differ in only a few terminal nucleotides, it is difficult to distinguish their individual abundance levels by RNA gel blot analysis or RT-qPCR. We therefore performed small RNA sequencing to accurately quantify the abundance of individual miR172 family members.

Mature miR172b was the only miR172 isoform showing a statistically significant difference in abundance between the WT and ld mutants (Fig. 4, a to c). The abundance of other known miRNAs, including miR172ad, was similar between the WT and ld mutants. In ld-1 and ld-2, miR172b abundance was very low, with an average of only one or two reads detected, which may reflect trace contamination or sequencing noise (Fig. 4, b to d). Furthermore, miR172ad abundance was relatively high in the WT and ld mutants, whereas we detected no reads for miR172c in any sample regardless of genotype (Fig. 4d). Similarly, miR806a was undetectable in both the WT and ld mutants. Likewise, the abundance of control miRNAs such as miR159ab and miR166a-dfj was comparable between the WT and the two ld mutants (Fig. 4d). Collectively, these data confirm the notion that miR172b is the only mature miRNA that is specifically affected in ld mutants.

Figure 4.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Characterization of miR172 family members accumulating in young panicles. (a) Volcano plot showing the fold-change relative to statistical significance (P-values) of normalized miRNA abundance between the ld mutants and WT Ilpum (IL). Gray dashed vertical lines indicate an absolute log2 (fold-change) threshold of 1, and the orange dashed horizontal line denotes P-value = 0.01. miR172ad and miR172b are shown as blue and red circles, respectively. (b), (c) Scatterplots of miRNA abundance in ld-1 (b) or ld-2 (c) vs. WT IL. miRNA levels were normalized as TP5M. miR172ad and miR172b are shown as blue and red circles, respectively. (d) Abundance, expressed as TP5M values, of selected miRNAs. miR172ad, miR172b, and miR172c represent mature miRNAs produced by different members of the rice MIR172 family. miR159ab and miR166a-dfj served as controls. Red text indicates nucleotide differences among miR172 family members.

SNB is a major target of miR172b during lodicule development

To identify the miR172b target gene responsible for the defective lodicule development in ld mutants, we examined the sequences of the five AP2-like transcription factor genes in rice that contain canonical miR172 complementary sequences: SHAT1 (LOC_Os04g55560), SNB (LOC_Os07g13170), IDS1 (LOC_Os03g60430), RSR1 (LOC_Os05g03040), and SAE1 (LOC_Os06g43220) (Sunkar et al. 2005; Zhu et al. 2009; Lee et al. 2014) (Fig. 5a). Analysis of publicly available Parallel Analysis of RNA Ends (PARE) datasets (Wu et al. 2009) detected clear miR172-guided cleavage signatures for SHAT1, SNB, and RSR1 transcripts. Cleavage signatures for IDS1 and SAE1 could not be reliably detected in the available PARE libraries, likely due to the low transcript abundance of these genes in the tissues profiled, which would limit the sensitivity of cleavage product detection (Fig. 5b).

Figure 5.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Validation of miR172b-mediated regulation of AP2-like genes. (a) Multiple nucleotide sequence alignment of the miR172 target sites complementary to the sequences of the indicated mature miR172 isoforms. Five putative miR172 target genes encoding AP2-like transcription factors and three mature miR172 isoforms are shown. Nucleotide differences among target sites or mature miR172 isoforms are shown in red. Vertical lines indicate perfect matches, and the G:U wobble pair is marked by a circle. The red arrow denotes the predicted miRNA-guided cleavage site. (b) Distribution of normalized PARE reads across the target sites. Red bars represent miR172-mediated cleavage events, while gray bars indicate randomly degraded RNA fragments. (c) Dual-luciferase assay to examine MIR172b-dependent regulation of AP2-like genes. Left, diagrams of the effector and reporter constructs. Each reporter construct consists of a miR172-binding sequence cloned into the 3′ UTR of the firefly luciferase (LUC) gene. Black bars, T-DNA borders; red boxes, 3′ UTRs of the LUC gene where the miR172 target sites were inserted. Right, relative luciferase activity in N. benthamiana leaves co-infiltrated with the effector and reporter constructs. Values are means ± SEM. Student's t-tests were performed (n = 6). *, P < 0.05; ***, P < 0.001. (d) to (h) RT-qPCR of AP2-like transcript levels in young panicles of WT cultivars Ilpum (IL) and Dongjin (DJ) and the corresponding ld mutants ld-1 and ld-3 at stage Sp5. (d) SHAT1; (e) SNB; (f) RSR1; (g) IDS1; (h) SAE1. Expression levels were normalized to those of OsUBQ5. Values are means ± SEM (n = 4; each is a pooled panicle from 3 to 5 plants). Statistical significance of differences was assessed using two-tailed Student's t-tests against the respective WT. **, P < 0.01; ***, P < 0.001; ns, not significant.

To more directly assess whether miR172b specifically regulates transcript abundance for these AP2-like genes, we performed Dual-luciferase reporter assays in Nicotiana benthamiana leaves. We cloned each representative miR172 target site downstream of the firefly luciferase (LUC) reporter gene, driven by the cauliflower mosaic virus 35S promoter (p35S), with Renilla luciferase (REN) driven by the Nopaline synthase (Nos) promoter serving as an internal control. Because SNB, RSR1, and IDS1 transcripts share an identical miR172-binding sequence, we generated three reporter constructs for SHAT1, SNB/RSR1/IDS1, and SAE1, respectively. A non-cleavable (NC) reporter containing deliberate mismatches in the miR172-binding site served as a negative control (Fig. 5, b to c). Co-infiltration of the effector construct p35S:MIR172b with either the empty vector (EV) or NC reporter resulted in relatively high relative LUC/REN values, comparable to that of infiltration with EV alone, confirming that the mutated miR172-binding sequence abolished miR172 recognition (Fig. 5c). By contrast, co-infiltration of 35S:MIR172b with the reporters carrying the SNB/RSR1/IDS1 or SAE1 target site resulted in significantly lower LUC/REN values, indicating effective miR172b-mediated post-transcriptional repression of these transcripts (Fig. 5c). Infiltration of the SHAT1 reporter construct did not result in a significant drop in LUC/REN values, indicating that repression was not detected in the heterologous system used in the Dual-luciferase assay under our conditions (Fig. 5c).

We next examined endogenous transcript levels of the five AP2-like genes in WT Ilpum and Dongjin and the corresponding mutants ld-1 and ld-3 to determine which targets become misregulated upon the loss of miR172b. Because lodicule development is initiated at an early stage of floral organ differentiation, we collected young panicles at the Sp5 stage from ld mutants and the corresponding WT plants (Itoh et al. 2005). RT-qPCR using gene-specific primers (Table S1) revealed that SHAT1, IDS1, RSR1, and SAE1 transcript levels were unchanged in the ld mutants relative to the WT controls. By contrast, SNB transcript levels were markedly elevated in both ld-1 and ld-3 plants (Fig. 5, d to h). These results indicate that SNB is a major downstream target subjected to derepression in miR172b-deficient mutants. The enhanced accumulation of SNB transcripts is strongly associated with reduced lodicule size and the failure of floret opening, resulting in cleistogamy.

miR172-mediated suppression of SNB expression is required for lodicule development

To validate the role of SNB in rice lodicule development, we generated a prime editing–mediated miR172-resistant form of SNB, designated PE-rSNB, in which the miR172-binding site was mutated to prevent complementarity while preserving the intact amino-acid sequence of SNB (Fig. 6a; Figure S3a). Among the five T0 plants obtained, we identified three SNB/rSNB heterozygotes without byproducts and subsequently obtained transgene-free homozygous rSNB mutant plants in the T1 generation (Figure S3b). The progeny of these homozygous plants were used for phenotypic analysis.

Figure 6.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Characterization of an rSNB line generated by prime editing. (a) Diagram of the SNB locus and location of the edited sequence in the rSNB line. Green boxes, exons; gray boxes, UTRs. Red letters mark nucleotide substitutions introduced by prime editing. (b) Relative SNB transcript levels in young panicles of the indicated genotypes at stage Sp5. Values are means ± SEM. Different letters indicate statistically significant differences based on ANOVA (P < 0.05). (c) Magnified views of reproductive organs from WT Dongjin, ld-2, and PE-rSNB plants. Red arrowheads indicate lodicules. Black scale bars, 200 μm; white scale bars, 5 mm. (d) Proposed model for the miR172b–SNB-mediated regulation of lodicule development and cleistogamy in rice. In the wild type, miR172b represses SNB expression via mRNA cleavage in developing lodicules, enabling normal cell expansion and lodicule-driven floret opening at anthesis. In ld mutants lacking miR172b, or in PE-rSNB plants producing a miR172-resistant SNB transcript, SNB mRNA escapes miR172-mediated cleavage and accumulates ectopically, leading to smaller or vestigial lodicules. The failure of lodicule swelling prevents palea/lemma separation, resulting in cleistogamy. In addition, stronger or spatially expanded SNB derepression (as observed in PE-rSNB) disrupts panicle exsertion.

To confirm the accumulation of rSNB transcripts, we isolated total mRNA from young panicles at the Sp5 stage and performed RT-qPCR using SNB-specific primers (Table S1). SNB transcript levels were significantly higher in PE-rSNB than in WT Dongjin, suggesting that miR172-mediated post-transcriptional repression was disrupted in the mutant (Fig. 6b). Examination of mature flowers revealed that the lodicules from PE-rSNB plants were much smaller than the WT and vestigial in nature, similar to those of the ld mutants (Fig. 6c; Figure S4). Although lodicules were not visible in the ld mutants at low magnification (Fig. 3), closer inspection confirmed the presence of vestigial lodicules with an abnormal morphology in ld plants, similar to those in PE-rSNB (Fig. 6c; Figure S4). In addition to reduced lodicules, PE-rSNB plants exhibited a panicle enclosure phenotype in which panicles failed to emerge from the flag leaf sheath (Figure S5f). This phenotype resembles that of the shortened uppermost internode 4 (sui4) mutant, which carries a missense substitution (Ser to Thr) in the miR172 target site of SNB (Ji et al. 2019). This panicle enclosure phenotype was not observed in any ld mutant, potentially due to the greater accumulation of SNB transcripts in PE-rSNB compared to ld-2 (Figs. 3 and 6b). This additional phenotype in PE-rSNB plants suggests that stronger or spatially extended SNB derepression may occur and that the loss of regulation by other miR172 isoforms in tissues where they are normally active may also contribute to the enhanced defects.

To investigate the tissue-specific expression patterns of MIR172b and SNB, we generated pMIR172b:GUS and pSNB:GUS reporter lines containing 2.5-kb and 2-kb promoter fragments, respectively, driving the GUS reporter and performed GUS staining of homozygous T3 lines. We detected no GUS signal in the leaves or internodes of either line, although weak signals appeared in the nodes of pSNB:GUS plants (Figure S6). However, broad but weak GUS signals were detected throughout young panicles in both lines at stages Sp5 and Sp7, corresponding to the initiation and completion of lodicule primordium development, respectively (Figure S6). Strong GUS signals were observed specifically in the lodicules of mature flowers (Figure S6). These results indicate that MIR172b and SNB are co-expressed in expanding lodicules. These spatial expression patterns are consistent with a regulatory relationship between miR172b and SNB during lodicule development.

Agronomic traits of ld and PE-rSNB mutants

Because cleistogamy limits floral opening and may reduce unintended outcrossing, cleistogamous rice has potential value for certain breeding applications. Previous studies have shown that the grain yield of the original ld (now ld-1) mutant was comparable to that of the WT (Won et al. 1998; Maeng et al. 2006). To evaluate the agronomic suitability of all ld mutants, we examined plant height, tiller number, mature grain number per panicle, grain morphology, and thousand-grain weight (TGW) (Figure S5). All plants were of similar height regardless of genotype, and tiller number and mature grain number per panicle were also unchanged among lines (Figure S5, a to c, e, and f). Grains were slightly shorter in all ld mutants compared to the respective WT plants; however, this decrease was offset by a modest increase in grain width, resulting in no significant change in TGW (Figure S5, d and g). The slight change in grain shape may reflect physical confinement of developing grains within closed spikelets.

Discussion

In this study, we demonstrated that the cleistogamous phenotype of rice ld mutants results from loss-of-function of MIR172b caused by a deletion in its promoter and that derepression of SNB underlies their reduced lodicule growth. Fine mapping and sequencing analysis revealed a ∼4.6-kb deletion in the ld-1 mutant that removes the entire MIR806a locus and the promoter region upstream of MIR172b (Fig. 1; Figure S1). Through the generation of CRISPR/Cas9-mediated loss-of-function mutants of MIR806a or MIR172b, we confirmed that the loss of miR172b, but not miR806a, underlies the cleistogamy phenotype of the mutants (Figs. 2 and 3). Small RNA sequencing further demonstrated that miR172b abundance was reduced to near-background levels (∼157-fold reduction) in ld mutants (Fig. 4), while all other miRNAs remained unchanged, providing strong evidence that the promoter deletion effectively abolishes MIR172b expression in vivo. Although the specific cis-regulatory elements removed by the ∼4.6-kb deletion remain to be defined, the strong reduction of miR172b abundance implies a direct causal link between promoter loss and miR172b deficiency. Functional dissection of individual regulatory elements within the deleted region will be an important direction for future investigation. In the same analysis, miR806a was undetectable even in WT plants (Fig. 4). This is consistent with the previous suggestion that MIR806a behaves more like a short-interacting RNA-generating locus rather than a canonical MIRNA gene; in fact, miR806a is no longer annotated as a bona fide miRNA in miRBase (Jeong et al. 2011; Kozomara et al. 2019).

Notably, the ∼4.6-kb deletion in ld-1 is accompanied by a 40-bp insertion at the breakpoint junction. Of these 40 base pairs, 38-bp are identical to LTR retrotransposon sequences dispersed throughout the rice genome. The co-occurrence of a large deletion with a short LTR-derived filler DNA insertion is a structural hallmark of illegitimate recombination (Gorbunova and Levy 1997; Ma et al. 2004). During DSB repair, exposed 3′ single-stranded overhangs can transiently invade ectopic genomic templates sharing short microhomology via synthesis-dependent strand annealing, resulting in the incorporation of short filler sequences at the repair junction (Gorbunova and Levy 1997; Salomon and Puchta 1998). Given that LTR retrotransposon sequences constitute approximately 22% of the rice genome (Ma et al. 2004), these sequences represent highly accessible ectopic templates for filler DNA capture. This analysis indicates that the deletion in the ld-1 mutant arose through illegitimate recombination rather than through precise transposon excision or homologous recombination between flanking repeats.

Previous work in barley revealed that cleistogamy can result from mutations in the miR172-binding sites of transcripts from the AP2-like gene Cly1 (Nair et al. 2010), suggesting that disrupted miR172−AP2-mediated regulation impairs lodicule development. Because miR172 and AP2-like targets are highly conserved in grasses, we examined five AP2-like genes in rice that harbor canonical miR172-binding sequences. Analysis of PARE data identified in vivo miR172-guided cleavage between the adenosine at position 10 and uracil at position 11 (of miR172b) for the SHAT1, SNB, and RSR1 transcripts (Fig. 5, a and b). Dual-luciferase assays showed that LUC activity derived from the p35S:LUC reporter with an added miR172-binding site from SNB/RSR1/IDS1 or SAE1 at the 3′ UTR of LUC declined upon MIR172b expression, whereas a reporter with the miR172-binding site from SHAT1 did not respond to the presence of miR172b, despite its high sequence complementarity to miR172b (Fig. 5, a and c). Among the five AP2-like genes, SHAT1 is also the closest rice ortholog to barley Cly1, further suggesting that it might be a strong miR172 target. Analysis of PARE data confirmed that SHAT1 cleavage occurs in vivo, indicating that the heterologous assay system may not fully recapitulate endogenous regulation. Importantly, SHAT1 transcript levels did not increase in the ld mutants, indicating that SHAT1 is not the primary mediator of cleistogamy (Fig. 5d). The absence of clear cleavage signatures for IDS1 and SAE1 in the analyzed PARE datasets does not exclude the possibility of miR172-mediated regulation, as detection sensitivity depends on transcript abundance and tissue specificity. In addition, plant miRNAs can repress the expression of their target genes through translational inhibition without producing cleavage signatures. By contrast, only SNB transcripts significantly accumulated in the young panicles of ld mutants (Fig. 5e). Collectively, these findings support the notion that SNB is a major functional target of miR172b during lodicule development.

Previous studies analyzing SNB function through knockout or constitutive overexpression did not report a cleistogamy phenotype (Lee et al. 2007, 2014; Jiang et al. 2019; Ma et al. 2019), suggesting that simple loss- or gain-of-function approaches may not fully capture the regulatory context of SNB during lodicule development. Therefore, we designed a prime-edited miR172-resistant SNB allele (PE-rSNB) that disrupts the miR172-binding site while preserving the intact protein sequence to clarify the molecular function of SNB under post-transcriptional regulation by miR172 (Fig. 6a; Figure S3a). PE-rSNB reproduced the vestigial lodicule phenotype seen in ld mutants, but also exhibited panicle enclosure, characterized by a failure of panicle exsertion from the flag leaf sheath. This phenotype was absent from ld mutants and resembled that of the sui4 mutant, which carries a missense mutation in the miR172-binding site of SNB (Fig. 6, a and c; Figure S5, e and f) (Ji et al. 2019). These contrasting phenotypes indicate that the magnitude, spatial pattern, and timing of SNB derepression differ between ld and PE-rSNB plants. SNB transcript levels were moderately higher (1.2–2-fold) in ld mutants but were significantly higher (∼3-fold) in PE-rSNB plants than in the WT (Figs. 5e and 6b), suggesting a dosage effect, whereby mild derepression specifically reduces lodicule growth, whereas stronger or ectopic derepression disrupts panicle exsertion. Additionally, SNB transcripts produced by PE-rSNB plants likely escape repression by other miR172 isoforms (miR172ad and miR172c), resulting in derepression in tissues where these isoforms normally function, an escape mechanism that is absent from ld mutants. Although IDS1 and other AP2-like genes harbor canonical miR172-binding sites and have been implicated in floral development, our genetic data indicate that derepression of SNB alone is sufficient to induce vestigial lodicule formation and cleistogamy. The PE-rSNB lines reproduced the lodicule defects observed in ld mutants without modifications of other AP2 loci, supporting SNB as a major functional mediator of the phenotype. Nevertheless, we cannot exclude potential contributions from additional miR172 targets, such as IDS1, under specific developmental contexts, which warrant further investigation.

SNB is known to regulate spikelet meristem determinacy, and our results suggest that it also contributes to panicle architecture or stem elongation (Lee et al. 2007, 2012). This broader developmental role is consistent with the panicle enclosure phenotypes observed in both PE-rSNB and sui4 plants (Ji et al. 2019). While miR172b-mediated regulation of SNB expression is clearly important for lodicule development, its downstream molecular mechanisms remain largely unresolved. The reduced lodicule size observed in ld mutants indicates impaired lodicule expansion; however, the cellular basis underlying this phenotype, such as altered cell proliferation, cell expansion, or cell identity, remains unresolved. Because AP2-type transcription factors can regulate developmental programs in diverse contexts (Zhao et al. 2019), SNB may influence pathways affecting lodicule growth. Detailed histological and cytological analyses will be required to distinguish among these possibilities. Identifying direct SNB target genes through chromatin immunoprecipitation followed by sequencing or DNA affinity purification followed by sequencing, together with transcriptome profiling of ld mutants across developmental stages, will also be essential in order to elucidate the SNB-dependent developmental program of lodicules. In addition, although the GUS reporter assays supported spatial overlap between MIR172b and SNB expression domains (Figure S6), they did not resolve the precise temporal dynamics of their expression during lodicule development. Higher-resolution stage-specific or quantitative analyses will be required to clarify the timing and hierarchy of this regulatory interaction.

A conceptual model summarizing the regulatory relationship between miR172b, SNB expression levels, and the resulting floral phenotypes in WT, ld, and PE-rSNB plants is shown in Fig. 6d. This model illustrates how the degree and spatial pattern of SNB derepression determine whether plants undergo normal anthesis (autogamy) or closed-flower selfing (cleistogamy) and explains the phenotypic differences between naturally occurring ld alleles and engineered rSNB mutants. Importantly, our findings point to the potential applications of these ld mutants for breeding cleistogamous rice cultivars. The ld alleles represent naturally occurring MIR172b loss-of-function mutations that result in a complete cleistogamy phenotype while maintaining normal plant growth, panicle exsertion, and grain yield across multiple backgrounds (Figure S5). By contrast, engineered SNB gain-of-function alleles, such as with PE-rSNB plants, exhibit undesirable pleiotropic defects. Thus, ld alleles provide a promising genetic resource for introducing cleistogamy into commercial rice cultivars.

In conclusion, our study demonstrates how the deletion of a single MIR172 family member, MIR172b, induces cleistogamy through specific post-transcriptional derepression of the AP2-like gene SNB. Because ld alleles maintain normal agronomic traits while providing complete cleistogamy, these naturally occurring variants represent valuable genetic resources for developing cleistogamous rice cultivars, with potential applications for transgene containment and genetic purity management. However, understanding the extent to which cleistogamy reduces outcrossing under field conditions will require direct evaluation in agronomic settings.

Materials and methods

Plant materials and growth conditions

Seeds of japonica rice (O. sativa) cultivars Ilpum and Dongjin, the indica cultivar Milyang23, and mutant lines were surface sterilized and cultured on half-strength Murashige and Skoog medium for 10 d under a 14-hour light/10-hour dark photoperiod at 25℃. The seedlings were transferred to soil. Rice plants were grown in either a growth chamber under a 14-hour light/10-hour dark photoperiod and a 30/20 ℃ (day/night) cycle or in an experimental field under natural conditions.

Map-based cloning

The ld-1 mutant used in this study was originally identified as CL-SNU, a natural mutant from japonica rice germplasm in the Ilpum background (Won et al. 1998; Maeng et al. 2006). The LD locus was previously mapped to the distal region on the long arm of chromosome 1 (Maeng et al. 2006). In the current study, the ld mutant was crossed with indica cultivar “Milyang23′, and 1,027 F2 plants were used for fine mapping. To further narrow down the LD region, 1,673 F3 plants derived from F2 lines carrying recombination events within the interval were analyzed.

Genomic DNA was isolated from young leaves using a simple miniprep method (Chen and Ronald 1999). Ten STS markers were developed for fine mapping of the LD locus based on the DNA sequence differences between Milyang23 and the japonica rice subspecies (Ji et al. 2012). The full-length sequence of the approximately 23-kb mapping region was divided into segments to design specific primers to amplify genomic DNA from the ld mutant and the WT cultivar Ilpum. To identify the region harboring the mutation, a DNA walking strategy was employed using a DNA Walking SpeedUp Premix Kit (Seegene, Korea). The primers qLD-F1 and qLD-R1 were used for amplification (Table S1). The resulting amplicons were cloned and subjected to Sanger sequencing. The deleted region was confirmed using primers qLD-F2 and qLD-R2 (Table S1).

Generation of additional ld mutants by CRISPR/Cas9 editing

To identify effective PAM sites and to avoid off-target effects, 20-bp gene-specific target sequences within the presumptive LD locus were selected using the CRISPRdirect web tool (Naito et al. 2015; https://crispr.dbcls.jp). To induce large deletions in each MIRNA locus, two sets of sgRNAs were designed to target MIR806a (5′-AGCTAGTGAGATCGAACTAG-3′ and 5′-GTAATTAGTATACTTTAATT-3′) or MIR172b (5′-GGCATCACACTCTCTACTGC-3′ and 5′-GAATCTTGATGATGCTGCAT-3′). To synthesize the tRNA-sgRNA1-tRNA-sgRNA2 cassettes, PCR fragments were amplified using primer sets L5AD5-F/806a-T1-R, 806a-T1-F/806a-T2-R, and 806a-T2-F/L5AD5-R for MIR806a and L5AD5-F/172b-T1-R, 172b-T1-F/172b-T2-R, and 172b-T2-F/L5AD5-R for MIR172b, with the pGTR plasmid (#63143, Addgene, MA, USA) used as the template (Table S1). The resulting DNA fragments were assembled using the Golden Gate method, as described by Xie et al. (2015), and subcloned into pRGEB32, a binary vector for multiplex genome editing. The final constructs were transformed into Agrobacterium tumefaciens strain LBA4404 and introduced individually into japonica rice cultivar “Dongjin” via Agrobacterium-mediated transformation (Jeon et al. 2000). Mutations at the sgRNA target sites were identified by Sanger sequencing of PCR amplicons generated with the primer sets 806a-seq-F/806a-seq-R and 172b-seq-F/172b-seq-R (Table S1). Transgene-free homozygous T1 mutants were isolated from the progeny of self-fertilized T0 plants, and T2 or T3 plants were used for all experiments.

Assessment of the cleistogamy phenotype

To visualize plant growth and flowering phenotypes, field-grown plants and their panicles were harvested from each genotype at the anthesis stage and photographed using a Canon EOS 100D camera (Canon, Japan). Individual spikelets, florets, and lodicules from dehulled flowers were observed under a stereomicroscope (SZX16, Olympus, Japan). For detailed morphological observations, lodicules were examined under a TM-20B stereomicroscope (Sunny Optical Technology, Korea).

Small RNA library construction and sequence analysis

Total RNA was isolated from the young panicles of WT Ilpum and ld plants using TRIzol reagent (Invitrogen, MA, USA), and 20–30-nt small RNAs were purified from a 15% (w/v) denaturing PAGE gel (Accerbi et al. 2010). Small RNA libraries were constructed using a NEXTFlex Small RNA-seq Kit v3 (Bioo Scientific, TX, USA) and sequenced on the Illumina MiSeq platform. Small RNA sequencing data were processed using the CLC Genomics Workbench version 11.0.1 (Qiagen, Germany). Rice miRNAs registered in miRBase Release 22 were used to identify mature miRNAs with perfect matches and to quantify their abundance (Kozomara and Griffiths-Jones 2014). Their abundance was normalized as transcripts per 5 million reads (TP5M). Differentially abundant mature miRNAs between WT and ld plants were identified using the R package EdgeR with the thresholds fold-change > 2 and P-value < 0.01. Raw reads from small RNA sequencing have been deposited in the Gene Expression Omnibus at the National Center for Biotechnology Information (NCBI) under accession number GSE174230.

Analysis of PARE library data

To validate miR172-mediated target cleavage, publicly available PARE data were retrieved from the NCBI Gene Expression Omnibus (accession number GSM455939) and analyzed. PARE reads that mapped to target genes were obtained from the website https://mpss.meyerslab.ucdavis.edu/oryza-sativa.

Dual-luciferase assay

For the effector construct, a binary vector containing the MIR172b primary MIRNA under the control of the cauliflower mosaic virus 35S promoter (p35S) was generated. An attB1-MIR172b-attB2 fragment was amplified from Dongjin genomic DNA and cloned into the entry vector pDONR201 using Gateway BP Clonase II Enzyme mix (Invitrogen, MA, USA). The resulting entry plasmid was recombined with the Gateway-compatible binary vector pGWB502 using Gateway LR Clonase II Enzyme mix (Invitrogen, MA, USA).

For the reporter constructs, the pGreen:dual-luc vector was used (#55206, Addgene, MA, USA), containing the REN reporter gene driven by the Nopaline synthase promoter (pNos) and a p35S:firefly luciferase (LUC) expression cassette. The 21-bp miR172-binding sites for five AP2-like genes (SNB, RSR1, IDS1, 5′-CUGCAGCAUCAUCAGGAUUCU-3′; SHAT1, 5′-CUGCAGCAUCAUCACGAUUCC-3′; SAE1, 5′-CUGCAGCAUCAUCAGGAUUCC-3′) and a non-cleavage control sequence (5′-CUGCUGCGAGUGACGGCUUUU-3′) were individually inserted into the 3′ untranslated region (3′ UTR) of LUC at the AvrⅡ and AgeⅠ restriction sites. Effector and reporter constructs were individually introduced into Agrobacterium strain GV3101.

Effector–reporter pairs were co-infiltrated into the leaves of Nicotiana benthamiana plants following the method of Ma et al. (2012). Dual-luciferase assays were performed using leaf samples collected three days after infiltration as described by Liu et al. (2014). A Dual-Glo Luciferase Assay System (Promega, WI, USA) was used for lysis and luminescence measurements, and LUC and REN activities were quantified using a VICTOR Nivo Multimode Plate Reader (PerkinElmer, MA, USA). Three biological replicates, each with three technical replicates, were analyzed per genotype.

mRNA isolation and transcript analysis

To quantify the expression levels of the five AP2-like genes predicted to be miR172b targets, total RNA was extracted from young panicles at the Sp5 stage using a FastPure Universal Plant Total RNA Isolation Kit (Vazyme, China). First-strand cDNA was synthesized using ReverTra Ace qPCR RT Master Mix (Toyobo, Japan). Quantitative PCR was performed using gene-specific primers (Table S1) and Prime Q-Master Mix (Genet Bio, Korea) on a Rotor-Gene 6000 Real-Time PCR system (Qiagen, Germany). The rice UBIQUITIN 5 gene (OsUBQ5; LOC_Os01g22490) was used as an internal control (Table S1). For RT-qPCR, four independent biological replicates were analyzed, with each biological replicate consisting of pooled young panicles from 3 to 5 plants of the same genotype. Each biological replicate was assayed in three technical replicates, and the mean value of the technical replicates was used for statistical analysis. Statistical significance was assessed using two-tailed Student's t-tests comparing mutants to their respective WT controls. P-values < 0.05 were considered statistically significant. Data are presented as means ± standard error of the mean (SEM).

Generation of a miR172-resistant SNB line by prime editing

To modify the miRNA172-binding site in the endogenous SNB locus while preserving the native amino-acid sequence, a 21-bp target region (5′-CTTCGTGGTCATGGCAAATGC-3′) was selected using the CRISPRdirect program (Naito et al. 2015; https://crispr.dbcls.jp). A pegRNA was designed to include a primer binding site (5′-TTTGCCATGA-3′) and a reverse transcription template (5′-AAAATCCGCTACTCGCAGCAGTAGGGAGTAATGGCAAAGGGGAGCCCTGCA-3′), conjugated to the evopreQ1 pseudoknot RNA motif (Figure S3a). The pegRNA was cloned into the enhanced pPE2 binary vector following the method of Li et al. (2022a). The resulting plasmid was introduced into japonica cultivar Dongjin by Agrobacterium-mediated transformation (Jeon et al. 2000). T0 transgenic plants were screened for editing by PCR using the primers SNB-qRT-F and SNB-qRT-R, followed by Sanger sequencing of the amplicons (Table S1). Transgene-free homozygous plants were subsequently isolated from the progeny of self-fertilized T0 plants, and T1 lines derived from these plants were used for analysis.

Analysis of GUS transgenic rice

To analyze tissue-specific expression patterns, a 2.5-kb promoter fragment of MIR172b and a 2-kb promoter fragment of SNB were individually amplified using the primer pairs pMIR172b-F/pMIR172b-R and pSNB-F/pSNB-R, respectively (Table S1). The PCR fragments were cloned upstream of the GUSplus reporter gene in the binary vector pCAMBIA1305, and transgenic plants were generated by Agrobacterium-mediated transformation (Jeon et al. 2000). The resulting pMIR172b:GUS and pSNB:GUS transgenic plants were advanced for three generations to obtain homozygous lines. Various tissues (leaves, stems, young panicles, and flowers) were collected, fixed, and stained with 1 mM X-Gluc (Thermo Fisher Scientific, MA, USA) for GUS histochemical analysis following Shim et al. (2022). Stained samples were examined under an SZX16 stereomicroscope (Olympus, Japan).

Analysis of agronomic traits

To analyze agronomic traits, eight individual plants per genotype grown in the paddy field were used to measure plant height and tiller number at the mature stage before harvest. Grain number per panicle was counted from three individual plants using three representative panicles per plant. Thousand-grain weight was measured from randomly selected grains after drying. Statistical comparisons were performed using two-tailed Student's t-tests.

Statistical analysis

All datasets were generated from at least three biological replicates, each with three technical replicates, and error bars represent SEM or SD. Student's t-test or one-way ANOVA was used for statistical analysis as appropriate.

Accession numbers

Raw small RNA sequencing reads generated in this study have been deposited in the NCBI Sequence Read Archive under accession number GSE174230.

Supplementary Material

kiag435_Supplementary_Data
kiag435_supplementary_data.zip (3.5MB, zip)

Acknowledgments

We thank Professor Pengcheng Wei of Anhui Agricultural University, Hefei, China, for providing the enpPE vector used to generate the PE-rSNB mutant line.

Contributor Information

Su-Hyeon Shim, Graduate School of Green-Bio Science and Crop Biotech Institute, Kyung Hee University, Yongin 17104, Republic of Korea.

Rihua Piao, Department of Plant Science, Seoul National University, Seoul 08826, Republic of Korea; Rice Research Institute, Jilin Academy of Agricultural Sciences, Gongzhuling 136100, China.

Dong-Yoon Seo, Graduate School of Green-Bio Science and Crop Biotech Institute, Kyung Hee University, Yongin 17104, Republic of Korea.

Sang-Kyu Lee, Graduate School of Green-Bio Science and Crop Biotech Institute, Kyung Hee University, Yongin 17104, Republic of Korea; Division of Life Science, Gyeongsang National University, Jinju 52828, Republic of Korea.

So Young Park, Department of Life Science and Multidisciplinary Genome Institute, Hallym University, Chuncheon 24252, Republic of Korea.

Backki Kim, Department of Plant Science, Seoul National University, Seoul 08826, Republic of Korea; Department of Plant Bioscience and Life and Industry Convergence Research Institute, Pusan National University, Miryang 50463, Republic of Korea.

Hyun-Sung Leem, Graduate School of Green-Bio Science and Crop Biotech Institute, Kyung Hee University, Yongin 17104, Republic of Korea.

Dong-Hoon Jeong, Department of Life Science and Multidisciplinary Genome Institute, Hallym University, Chuncheon 24252, Republic of Korea.

Hee-Jong Koh, Department of Plant Science, Seoul National University, Seoul 08826, Republic of Korea.

Jong-Seong Jeon, Graduate School of Green-Bio Science and Crop Biotech Institute, Kyung Hee University, Yongin 17104, Republic of Korea.

Author Contributions

J-.S.J. and H-.J.K. designed the research; S-.H.S., R.P., D-.Y.S., S-.K.L., S.Y.P., B.K., and H-.S.L. performed the experiments and analyzed the data; S-.H.S., R.P., D-.Y.S., D-.H.J., S-.K.L., H-.J.K., and J-.S.J. wrote the manuscript.

Supplementary material

Supplementary material is available at Plant Physiology online.

Funding

This work was supported by grants from the National Research Foundation of Korea (2023R1A2C1003142 for J-.S.J., RS-2024-00355816 for D-.H.J., and RS-2025-02263262 for S-.K.L.), and the Rural Development Administration (RS-2024-00322378 for J-.S.J.).

Data availability

The data underlying this article will be shared upon reasonable request to the corresponding author.

References

  1. Accerbi  M  et al.  2010. Methods for isolation of total RNA to recover miRNAs and other small RNAs from diverse Species. In: Meyers  BC, Green  PJ, editors. Plant MicroRNAs: methods and protocols. Humana Press. p. 31–50. [DOI] [PubMed] [Google Scholar]
  2. Aukerman  MJ, Sakai  H. 2003. Regulation of flowering time and floral organ identity by a MicroRNA and its APETALA2-like target genes. Plant Cell.  15:2730–2741. 10.1105/tpc.016238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chen  DH, Ronald  PC. 1999. A rapid DNA minipreparation method suitable for AFLP and other PCR applications. Plant Mol Biol Rep.  17:53–57. 10.1023/A:1007585532036. [DOI] [Google Scholar]
  4. Chen  X. 2004. A MicroRNA as a translational repressor of APETALA2 in Arabidopsis flower development. Science. 303:2022–2025. 10.1126/science.1088060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chow  C-N  et al.  2019. PlantPAN3.0: a new and updated resource for reconstructing transcriptional regulatory networks from ChIP-seq experiments in plants. Nucleic Acids Res.  47:D1155–D1163. 10.1093/nar/gky1081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chow  C-N  et al.  2024. PlantPAN 4.0: updated database for identifying conserved non-coding sequences and exploring dynamic transcriptional regulation in plant promoters. Nucleic Acids Res.  52:D1569–D1578. 10.1093/nar/gkad945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Doench  JG  et al.  2016. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nat Biotechnol.  34:184–191. 10.1038/nbt.3437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fu  F-F, Xue  H-W. 2010. Coexpression analysis identifies rice starch Regulator1, a rice AP2/EREBP family transcription factor, as a novel rice starch biosynthesis regulator. Plant Physiol.  154:927–938. 10.1104/pp.110.159517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gorbunova  V, Levy  AA. 1997. Non-homologous DNA end joining in plant cells is associated with deletions and filler DNA insertions. Nucleic Acids Res.  25:4650–4657. 10.1093/nar/25.22.4650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Itoh  J-I  et al.  2005. Rice plant development: from zygote to spikelet. Plant Cell Physiol.  46:23–47. 10.1093/pcp/pci501. [DOI] [PubMed] [Google Scholar]
  11. Jeon  J-S  et al.  2000. T-DNA insertional mutagenesis for functional genomics in rice. Plant J.  22:561–570. 10.1046/j.1365-313x.2000.00767.x. [DOI] [PubMed] [Google Scholar]
  12. Jeong  D-H  et al.  2011. Massive analysis of rice small RNAs: mechanistic implications of regulated MicroRNAs and variants for differential target RNA cleavage. Plant Cell.  23:4185–4207. 10.1105/tpc.111.089045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ji  H  et al.  2012. Development of rice molecular genetic and physical map using PCR-based DNA markers with the recombinant inbred population derived from Milyang23/Gihobyeo cross. Kor J Breed Sci. 44:273–281. [Google Scholar]
  14. Ji  H  et al.  2019. Mutations in the microRNA172 binding site of SUPERNUMERARY BRACT (SNB) suppress internode elongation in rice. Rice. 12:62. 10.1186/s12284-019-0324-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jiang  L  et al.  2019. The APETALA2-like transcription factor SUPERNUMERARY BRACT controls rice seed shattering and seed size. Plant Cell.  31:17–36. 10.1105/tpc.18.00304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kozomara  A, Birgaoanu  M, Griffiths-Jones  S. 2019. miRBase: from microRNA sequences to function. Nucleic Acids Res.  47:D155–D162. 10.1093/nar/gky1141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kozomara  A, Griffiths-Jones  S. 2014. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res.  42:D68–D73. 10.1093/nar/gkt1181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kurauchi  N, Makino  T, Hirose  S. 1993. Inheritance of cleistogamy-chasmogamy in barley. Barley Genet Newsletter. 23:19. https://scholar.google.co.kr/scholar?hl=ko&as_sdt=0%2C5&q=Kurauchi+N%2C+Makino+T%2C+Hirose+S.+1993.+Inhieritance+of+cleistogamy-chasmogamy+in+barley.+Barley+Genetics+Newletter.+23%3A19&btnG=. [Google Scholar]
  19. Lee  D-Y, An  G. 2012. Two AP2 family genes, SUPERNUMERARY BRACT (SNB) and OsINDETERMINATE SPIKELET 1 (OsIDS1), synergistically control inflorescence architecture and floral meristem establishment in rice. Plant J.  69:445–461. 10.1111/j.1365-313X.2011.04804.x. [DOI] [PubMed] [Google Scholar]
  20. Lee  D-Y, Lee  J, Moon  S, Park  SY, An  G. 2007. The rice heterochronic gene SUPERNUMERARY BRACT regulates the transition from spikelet meristem to floral meristem. Plant J.  49:64–78. 10.1111/j.1365-313X.2006.02941.x. [DOI] [PubMed] [Google Scholar]
  21. Lee  Y-S, Lee  D-Y, Cho  L-H, An  G. 2014. Rice miR172 induces flowering by suppressing OsIDS1 and SNB, two AP2 genes that negatively regulate expression of Ehd1 and florigens. Rice. 7:31. 10.1186/s12284-014-0031-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Li  F  et al.  2010. Rice DENSE AND ERECT PANICLE 2 is essential for determining panicle outgrowth and elongation. Cell Res.  20:838–849. 10.1038/cr.2010.69. [DOI] [PubMed] [Google Scholar]
  23. Li  J  et al.  2022a. Development of a highly efficient prime editor 2 system in plants. Genome Biol.  23:161. 10.1186/s13059-022-02730-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li  Y  et al.  2022b. SALT AND ABA RESPONSE ERF1 improves seed germination and salt tolerance by repressing ABA signaling in rice. Plant Physiol.  189:1110–1127. 10.1093/plphys/kiac125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu  Q, Wang  F, Axtell  MJ. 2014. Analysis of complementarity requirements for plant MicroRNA targeting using a Nicotiana benthamiana quantitative transient assay. Plant Cell.  26:741–753. 10.1105/tpc.113.120972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lombardo  F  et al.  2017. The superwoman1-cleistogamy2 mutant is a novel resource for gene containment in rice. Plant Biotechnol J.  15:97–106. 10.1111/pbi.12594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lord  EM. 1981. Cleistogamy: a tool for the study of floral morphogenesis, function and evolution. Bot Rev. 47:421–449. 10.1007/BF02860538. [DOI] [Google Scholar]
  28. Ma  J, Devos  KM, Bennetzen  JL. 2004. Analyses of LTR-retrotransposon structures reveal recent and rapid genomic DNA loss in rice. Genome Res.  14:860–869. 10.1101/gr.1466204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ma  L, Lukasik  E, Gawehns  F, Takken  FLW. 2012. The use of agroinfiltration for transient expression of plant resistance and fungal effector proteins in Nicotiana benthamiana leaves. In: Bolton  MD, Thomma  BPHJ, editors. Plant fungal pathogens: methods and protocols. Humana Press. p. 61–74. [DOI] [PubMed] [Google Scholar]
  30. Ma  X  et al.  2019. A novel rice grain size gene OsSNB was identified by genome-wide association study in natural population. PLoS Genet.  15:e1008191. 10.1371/journal.pgen.1008191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Maeng  J-Y  et al.  2006. Molecular mapping of a gene ‘ld(t)’ controlling cleistogamy in rice. Theor Appl Genet.  112:1429–1433. 10.1007/s00122-006-0244-z. [DOI] [PubMed] [Google Scholar]
  32. Nair  SK  et al.  2010. Cleistogamous flowering in barley arises from the suppression of microRNA-guided HvAP2 mRNA cleavage. Proc Natl Acad Sci U S A.  107:490–495. 10.1073/pnas.0909097107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Naito  Y, Hino  K, Bono  H, Ui-Tei  K. 2015. CRISPRdirect: software for designing CRISPR/cas guide RNA with reduced off-target sites. Bioinformatics. 31:1120–1123. 10.1093/bioinformatics/btu743. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Ni  D-H  et al.  2014. Identification and utilization of cleistogamy gene cl7(t) in rice (Oryza sativa L.). J Exp Bot.  65:2107–2117. 10.1093/jxb/eru074. [DOI] [PubMed] [Google Scholar]
  35. Ning  S  et al.  2013. Structure, transcription and post-transcriptional regulation of the bread wheat orthologs of the barley cleistogamy gene Cly1. Theor Appl Genet.  126:1273–1283. 10.1007/s00122-013-2052-6. [DOI] [PubMed] [Google Scholar]
  36. Rogers  K, Chen  X. 2013. Biogenesis, turnover, and mode of action of plant MicroRNAs. Plant Cell.  25:2383–2399. 10.1105/tpc.113.113159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Salomon  S, Puchta  H. 1998. Capture of genomic and T-DNA sequences during double-strand break repair in somatic plant cells. EMBO J.  17:6086–6095. 10.1093/emboj/17.20.6086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Shim  S-H  et al.  2022. Rice β-glucosidase Os12BGlu38 is required for synthesis of intine cell wall and pollen fertility. J Exp Bot.  73:784–800. 10.1093/jxb/erab439. [DOI] [PubMed] [Google Scholar]
  39. Spanudakis  E, Jackson  S. 2014. The role of microRNAs in the control of flowering time. J Exp Bot.  65:365–380. 10.1093/jxb/ert453. [DOI] [PubMed] [Google Scholar]
  40. Sun  B  et al.  2022. qHD5 encodes an AP2 factor that suppresses rice heading by down-regulating Ehd2 expression. Plant Sci.  324:111446. 10.1016/j.plantsci.2022.111446. [DOI] [PubMed] [Google Scholar]
  41. Sunkar  R, Girke  T, Jain  PK, Zhu  J-K. 2005. Cloning and characterization of MicroRNAs from rice. Plant Cell.  17:1397–1411. 10.1105/tpc.105.031682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Teotia  S, Tang  G. 2015. To bloom or not to bloom: role of MicroRNAs in plant flowering. Mol Plant.  8:359–377. 10.1016/j.molp.2014.12.018. [DOI] [PubMed] [Google Scholar]
  43. Turuspekov  Y  et al.  2004. Identification and mapping of cleistogamy genes in barley. Theor Appl Genet.  109:480–487. 10.1007/s00122-004-1673-1. [DOI] [PubMed] [Google Scholar]
  44. Wang  H  et al.  2025. The miR172a–SNB module orchestrates both induced and adult-plant resistance to multiple diseases via MYB30-mediated lignin accumulation in rice. Mol Plant.  18:59–75. 10.1016/j.molp.2024.11.015. [DOI] [PubMed] [Google Scholar]
  45. Wollmann  H, Mica  E, Todesco  M, Long  JA, Weigel  D. 2010. On reconciling the interactions between APETALA2, miR172 and AGAMOUS with the ABC model of flower development. Development. 137:3633–3642. 10.1242/dev.036673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Won  YJ, Koh  HJ, Heu  MH. 1998. Inheritance of cleistogamy and its interrelationship between other agronomic characters in rice. Kor J Breed (Korea Republic). 30:42–46. https://scholar.google.co.kr/scholar?hl=ko&as_sdt=0%2C5&q=Won+YJ%2C+Koh+HJ%2C+Heu+MH.+1998.+Inheritance+of+cleistogamy+and+its+interrelationship+between+other+agronomic+characters+in+rice.+Kor+J+Breed+&btnG=. [Google Scholar]
  47. Wu  L  et al.  2009. Rice MicroRNA effector complexes and targets. Plant Cell.  21:3421–3435. 10.1105/tpc.109.070938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Xie  K, Minkenberg  B, Yang  Y. 2015. Boosting CRISPR/Cas9 multiplex editing capability with the endogenous tRNA-processing system. Proc Natl Acad Sci U S A.  112:3570–3575. 10.1073/pnas.1420294112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yoshida  H  et al.  2007. superwoman1-cleistogamy, a hopeful allele for gene containment in GM rice.  Plant Biotechnol J.  5:835–846. 10.1111/j.1467-7652.2007.00291.x. [DOI] [PubMed] [Google Scholar]
  50. Zhao  Y  et al.  2019. Genome-wide identification and analysis of the AP2 transcription factor gene family in wheat (Triticum aestivum L.). Front Plant Sci.  10:1286. 10.3389/fpls.2019.01286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zhou  Y  et al.  2012. Genetic control of seed shattering in rice by the APETALA2 transcription factor SHATTERING ABORTION1. Plant Cell.  24:1034–1048. 10.1105/tpc.111.094383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhu  Q-H, Helliwell  CA. 2011. Regulation of flowering time and floral patterning by miR172. J Exp Bot.  62:487–495. 10.1093/jxb/erq295. [DOI] [PubMed] [Google Scholar]
  53. Zhu  Q-H, Upadhyaya  NM, Gubler  F, Helliwell  CA. 2009. Over-expression of miR172 causes loss of spikelet determinacy and floral organ abnormalities in rice (Oryza sativa). BMC Plant Biol.  9:149. 10.1186/1471-2229-9-149. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

kiag435_Supplementary_Data
kiag435_supplementary_data.zip (3.5MB, zip)

Data Availability Statement

The data underlying this article will be shared upon reasonable request to the corresponding author.

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES