• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntcellLink to Publisher's site
Plant Cell. Mar 2012; 24(3): 1034–1048.
Published online Mar 9, 2012. doi:  10.1105/tpc.111.094383
PMCID: PMC3336138

Genetic Control of Seed Shattering in Rice by the APETALA2 Transcription Factor SHATTERING ABORTION1[C][W][OA]


Seed shattering is an important agricultural trait in crop domestication. SH4 (for grain shattering quantitative trait locus on chromosome 4) and qSH1 (for quantitative trait locus of seed shattering on chromosome 1) genes have been identified as required for reduced seed shattering during rice (Oryza sativa) domestication. However, the regulatory pathways of seed shattering in rice remain unknown. Here, we identified a seed shattering abortion1 (shat1) mutant in a wild rice introgression line. The SHAT1 gene, which encodes an APETALA2 transcription factor, is required for seed shattering through specifying abscission zone (AZ) development in rice. Genetic analyses revealed that the expression of SHAT1 in AZ was positively regulated by the trihelix transcription factor SH4. We also identified a frameshift mutant of SH4 that completely eliminated AZs and showed nonshattering. Our results suggest a genetic model in which the persistent and concentrated expression of active SHAT1 and SH4 in the AZ during early spikelet developmental stages is required for conferring AZ identification. qSH1 functioned downstream of SHAT1 and SH4, through maintaining SHAT1 and SH4 expression in AZ, thus promoting AZ differentiation.


Seed shattering is an adaptive trait for seed dispersal in wild plants. However, the seed shattering habit causes yield loss for domesticated crop plants during harvest. Our ancestors began domesticating crop plants by selecting grains that had reduced seed shattering characteristics (Doebley, 2006; Fuller et al., 2009). Seed shattering occurs in the anatomically distinct cell files known as the abscission zone (AZ). Differentiated AZ cells are small, isodiametric, and cytoplasmically dense compared with surrounding cells and are responsive to signals promoting abscission (Hoekstra et al., 2001; Jin, 1986; McKim et al., 2008; Szymkowiak and Irish, 1999). Often these signals are associated with the senescence of the distal organ. However, a spectrum of environmental factors, such as a deficit or surplus of water, extremes of temperature, or pest and pathogen attack, can prematurely precipitate leaf, flower, or fruit fall (Taghizadeh et al., 2009; Taylor and Whitelaw, 2001). Understanding how the process of abscission is regulated in model crops would benefit agriculture.

Previous studies of rice (Oryza sativa) have identified a few of the factors involved in seed shattering. SH4 is a member of the trihelix family of transcription factors and promotes hydrolyzing of AZ cells during the abscission process (Li et al., 2006; Lin et al., 2007). The cultivated rice allele of sh4 severely weakens but does not eliminate shattering (Li et al., 2006). qSH1, a major rice quantitative trait locus on chromosome 1, encodes a BEL1-type homeobox-containing protein. A single nucleotide polymorphism (SNP) in the 5′ upstream regulatory region of qSH1 causes qSH1 expression to disappear from the abscission layer, thus leading to a decline in seed shattering over the history of rice domestication (Konishi et al., 2006). Recently, a recessive shattering locus sh-h, encoding a C-terminal domain phosphatase-like protein, was identified using mutagenesis of cultivated rice and was shown to inhibit the development of AZs in rice (Ji et al., 2010). In addition to the three rice shattering genes, the wheat (Triticum turgidum) Q gene was reported to affect the compaction and fragility of wheat ears and also the ease with which the grain can be separated from the chaff (Doebley, 2006; Simons et al., 2006). Q is a member of the APETALA2 (AP2) family of transcription factors, which have been implicated in a wide variety of plant development roles.

There is a wide range in the degree of seed shattering among worldwide rice cultivars (Konishi et al., 2006), suggesting that the shattering habit is a polygenic and complex trait. To better understand the process of seed abscission and exploit the network of different genes regulating the rice shattering pathway, we investigated suppressor mutants in a genetic background containing known seed shattering–related genes in which the shattering events have in some way been impaired. To this end, we constructed a shattering chromosomal segment substitution line (CSSL), Substitution Line 4 (SL4), by introducing chromosome 4 of wild-rice Oryza rufipogon W1943 (easy shattering) into the recurrent parent O. sativa ssp indica cv Guangluai 4 (GLA4) (reduced shattering). We identified a number of new nonshattering rice mutants by 60Co γ-ray mutagenesis of SL4.

In this study, we report the isolation and characterization of one of these nonshattering mutants, shattering abortion1 (shat1). By map-based cloning, we identified the AP2 domain–containing transcription factor gene SHAT1 as responsible for AZ development. We also identified a new allelic mutant of SH4 that showed a complete loss-of-shattering phenotype, differing from the reduced-shattering habit that results from the sh4 allele in cultivated rice. Genetic analyses demonstrated that SH4 plays a role early in AZ differentiation, and the expression of SHAT1 in the AZ was positively regulated by SH4. We also show that qSH1 functions downstream of SHAT1 and SH4 to maintain their expression in the AZ, thus promoting AZ differentiation. Our results suggest a genetic model in which the persistent and concentrated expression of the active SHAT1 and SH4 in the AZ during early spikelet developmental stages is required for conferring AZ identification.


Characterization of Two Nonshattering Mutants in Rice

To explore the seed shattering regulation pathways, we constructed a CSSL by introgressing chromosome 4 of the shattering donor parent O. rufipogon W1943 into the reduced-shattering recurrent parent O. sativa ssp indica cv GLA4 (Zhu et al., 2011) (see Supplemental Figures 1A and 1B online). This chromosome 4 substitution line was designated as the SL4. SL4 exhibited a very easy shattering phenotype as a result of not only acquiring SH4 from wild rice but also harboring qSH1 in the indica genetic background. SL4 is also referred to as the wild type below. We then generated a rice mutant library by treating the shattering seeds of SL4 with 60Co γ-rays and screened for nonshattering mutants among the T1 plants, isolating two mutant lines (Figure 1A): We named these shat1 and shat2, respectively.

Figure 1.
Characterization of Seed Shattering and Floral AZ Morphology in Wild Type and Different Mutant Lines.

In addition to loss of shattering, the shat1 mutant displayed a range of spikelet and inflorescence developmental defects, including (1) palea degeneration or florets with multiple layers of lemma- and palea-like structures (see Supplemental Figures 2A to 2D and 3H online); (2) changes in numbers, size, appearance, and identities of floral organs, especially carpels and anthers (see Supplemental Figures 2I to 2K and 3D to 3G online); (3) longer grains (see Supplemental Figure 3A online); (4) fewer primary branches (see Supplemental Figure 3B online); and (5) reduced seed set rate (see Supplemental Figure 3C online). Notably, the shat1 mutant displayed a unique phenotype of a crook-neck-like rachilla between the sterile lemmas and the rudimentary glumes in the position where the AZ was usually formed (Figures 1C, 1D, 1I, and 1J). In contrast with shat1, the shat2 mutant showed no significant morphological alterations in spikelets, except for the shattering defect and smaller seeds (Figures 1O and and1P;1P; see Supplemental Figures 2E, 2L, 4A, and 4B online). The shat1 shat2 double mutant had a crook-neck-like rachilla (Figures 1U and and1V;1V; see Supplemental Figure 2F online) and showed similar spikelet defects to those in shat1, including aberrant palea, lemma, and inner floral organs (see Supplemental Figures 2G, 2H, 2M, and 2N online).

Disruption of AZ Results in a Nonshattering Phenotype in the shat1 and shat2 Mutants

The shattering defect of shat1 and shat2 mutant was further characterized by measuring pedicel breaking tensile strength (BTS), which is inversely proportional to shattering degree. During the first 6 d after pollination, the BTS force did not differ between the developmental stages in either line (Figure 1B). From 11 d, the BTS value quickly decreased in the wild type. These changes may be due to the progressive degradation of the middle lamellae in the AZ. The BTS value decreased with maturating of seeds. By 13 d, shattering prevailed in the wild type, which left few grains to measure. In indica cv GLA4, the BTS began to decrease from 9 d. The earlier commencement of the decrease in GLA4 might be due to the faster maturating of seeds. However, the decline proceeded at a slower rate in GLA4 than in the wild type. From 9 d onward, the BTS showed no appreciable change in GLA4, and the values were about one-third of those displayed at the earlier stages but did not reach a level permitting grain shattering (Figure 1B). By contrast, neither shat1 nor shat2 mutants showed decreases in BTS with maturating of seeds, suggesting that cell wall adhesion at the seed-pedicel junction did not weaken in these two mutants (Figure 1B). The shat1 shat2 double mutant showed similar changes in BTS values to those in shat1 during seed development.

To distinguish precisely the differences in AZ anatomy among shat1, shat2, and the wild type, longitudinal sections of spikelets at anthesis stage were compared using confocal microscopy. Isodiametrically flattened and thin-walled AZ cells were visible in the basal area near sterile lemmas in the wild type (Figures 1E and and1F).1F). GLA4, however, had an incomplete abscission layer; in the longitudinal section, the line of abscission cells was discontinuous and completely absent near the vascular bundle, where they were replaced by thick-walled cells similar to adjacent pedicel cells (see Supplemental Figure 5A online). In the shat1 (Figures 1K and and1L)1L) and shat1 shat2 mutants (Figures 1W and and1X),1X), such a layer of abscission cells was substituted by peanut-like, thick-walled cortical cells. In the shat2 mutant, although AZ cells had a small and flattened appearance, they were not thin walled, as their cell walls were stained green by Acridine Orange, an indicator for lignin deposition when excited with a 488-nm laser (Briggs and Morris., 2008) (Figures 1Q and and1R).1R). These results suggest that a cytologically distinct and active seed AZ is not properly formed in both shat1 and shat2 mutants.

We then used scanning electron microscopy to examine the interface where a mature grain separates from the pedicel. Consistent with the BTS testing results, there was a smooth fracture surface of rachilla in the wild type as a result of hydrolyzing of the middle lamellae during the seed shattering process (Figures 1G and and1H).1H). Moreover, the fracture surface of the retained rachilla in the wild type was filled with callose that might serve as an ideal barrier to resist pathogen invasion (Figure 1H). In GLA4, the outer surface was smooth, while the center surface was rough, consistent with its discontinuous AZs (see Supplemental Figure 5B online). In shat1 (Figures 1M and and1N),1N), shat2 (Figures 1S and and1T),1T), and shat1 shat2 mutants (Figures 1Y and and1Z),1Z), external force was needed to remove seed from the pedicels when seeds were fully ripened and this left a rough and irregular surface.

SHAT1 Encodes an AP2-Like Transcription Factor, and a Frameshift in the SHAT2 Locus Completely Disrupts SH4 Function

A map-based cloning approach was adopted to isolate the corresponding SHAT1 and SHAT2 genes associated with the mutations in the shat1 and shat2 mutants, respectively. Genetic crossing indicated they were nonallelic mutants. In addition, shat1 was also not allelic to the known shattering genes sh4 and qSH1. As the crook-neck-like rachilla was a typical and visible phenotype resulting from loss of AZ for the shat1 mutant, we therefore adopted it as a standard shat1 phenotype (Figure 1J). All F1 progeny deriving from the cross between shat1 and the wild type showed the same phenotypes of seed shattering as wild-type plants and without the crook-neck-like rachilla. In the F2 population, shattering and nonshattering plants segregated in a 3:1 ratio (314:104, n = 1, χ2 = 0.006, P = 0.936), and the nonshattering phenotype cosegregated with crook-neck-like rachilla, indicating that a single recessive locus was the cause of these two phenotypes in shat1. For the construction of a mapping population, the shat1 mutant was crossed with the japonica cv Nipponbare. A total of 300 F2 plants that had crook-neck-like rachilla were selected for mapping analysis. SHAT1 was initially mapped on the long arm of chromosome 4 between the two insertion or deletion (indels) markers M1056 and M6026 and was subsequently fine-mapped to a 9.0-kb interval between two SNP markers, A and B (Figure 2A; see Supplemental Table 1 online). In this candidate region, only one gene, which encoded an AP2 transcription factor (LOC_Os04g55560) with a 459-residue polypeptide protein, was annotated in The Institute for Genomic Research Rice Genome Annotation database (http://rice.plantbiology.msu.edu/; Figure 2B). Comparison of the nucleotide sequences of the candidate gene in shat1 mutant and wild-type plants revealed a 1-bp deletion in the first exon between the nucleotide sites +41 and +42 in the shat1 mutant, resulting in a frameshift (Figure 2B). A homology search by BLASTP found that the SHAT1 is a rice ortholog of the Arabidopsis thaliana AP2 protein (Figure 2C), which was shown to be involved in the regulation of flower development (Chen., 2004; Wollmann et al., 2010).

Figure 2.
Map-Based Cloning of SHAT1 and SHAT2.

SHAT2 was first mapped on the long arm of chromosome 4 in a 1000-kb region using the F2 population crossed with japonica cv Nipponbare (Figure 2D). We then used another F2 population crossed with indica cv GLA4 for fine mapping the SHAT2. The SHAT2 locus was finally narrowed down to a 9.7-kb region that encompassed two predicted genes (Figure 2D). Comparison of the 9.7-kb sequences between the wild type and shat2 mutant revealed a one-nucleotide insertion in the second exon of the first gene, which was known to be SH4 (LOC_Os04g57530; Figure 2E). Differing from the one–amino acid substitution in cultivated rice sh4 allele, the shat2 mutation resulted in a frameshift before the nuclear localization signal region (Figure 2F). To distinguish these two different sh4 mutant alleles, we renamed the sh4 allele in cultivated rice as sh4-1 and the shat2 allele as sh4-2 in the following text.

Genetic Analysis Confirms that the SHAT1 Affects Rice Seed Shattering

As the shat1 mutant has both indica and wild rice genetic backgrounds, it failed to regenerate shoots from callus. Thus, we could not directly perform a complementation test to identify the function of SHAT1 in the shat1 mutant line. To verify the responsibility of SHAT1 in specifying the AZ, we generated a SHAT1-RNA interference (RNAi) construct (Figure 3A) and transformed it into the indica cv Kasalath, which possesses the wild-type allele of SHAT1 and has a reduced-shattering phenotype. Eight independent SHAT1-RNAi T0 transgenic plants showed various degrees of increased BTS values compared with Kasalath, and these values correlated well with the magnitude of the reduction in SHAT1 mRNA (Figure 3B). Two RNAi lines (lines 2 and 8) with varied levels of suppression were used for AZ anatomical structure studies: Line 8 had a stronger defective AZ phenotype than line 2 (Figure 3C). This observation was consistent with the results that line 8 contained ~17%, while line 2 contained ~51% of the wild-type SHAT1 mRNA levels (Figure 3B).

Figure 3.
Genetic Identification of SHAT1.

More evidence supporting the involvement of SHAT1 in AZ development was obtained from the characterization of another allelic SHAT1 mutant line (2B70080) identified in the T-DNA (T-DNA) insertion line database (Jeon et al., 2000). This mutant line had a T-DNA insertion in the 3′ untranslated region (UTR) of SHAT1 (Figures 3D and and3E).3E). The expression levels of SHAT1 were dramatically decreased in mutant 2B70080 compared with its wild-type japonica cv Huayong. This result also correlated closely with the increasing BTS value in 2B70080 (Figure 3F). Further investigation of the AZ anatomical structure using confocal microscopy indicated that the thin-walled property of AZ cells was altered in the 2B70080 mutant (Figure 3G). Overall, these results confirmed that the loss of shattering in the shat1 mutant was caused by disruption of the SHAT1 gene.

SHAT1 Is Localized in the Nucleus and Has Transactivational Activity

SHAT1 protein was predicted to have a nuclear localization signal by the PSORT program (http://psort.hgc.jp/). To confirm this prediction, we made a construct constitutively expressing the full-length SHAT1 fused to the C terminus of green fluorescent protein (GFP). We transiently expressed the SHAT1-GFP fusion protein in onion epidermal cells and observed GFP signal to be exclusively localized in the nucleus (see Supplemental Figure 6 online).

Although SHAT1 acts as a transcription factor and is localized in the nucleus, it is unknown whether SHAT1 has transcriptional activation activity. We applied the yeast one-hybrid system to detect transcriptional activation activity of SHAT1. The yeast GAL4 DNA binding domain was fused to different segments of SHAT1 (Figure 4). The full-length of rice bZip72 was used as a positive control (Lu et al., 2009). All transformants grew well on synthetic dropout (SD) medium lacking Trp (Figure 4). Then, yeast colonies of each transformant were transferred onto either Trp- and adenylate (Ade)-negative SD media or Trp- and His (His)-negative SD media with 0.5 mM 3-amino-1,2,4-triazole (3AT), respectively. On the SD/Trp-/Ade- plate, all types of transformants failed to grow, except for the positive control. However, on the SD/Trp-/His- plate, transformants of the positive control, pGBKT7-SHAT1 and pGBKT7-SHAT1ΔC, grew well, whereas pGBKT7-SHAT1ΔN and the negative control displayed cell growth inhibition (Figure 4). These results revealed that the N terminus of SHAT1 peptide that included the AP2 domain exhibited weak transactivational activity.

Figure 4.
Transactivation Tests of SHAT1 in Yeast.

SHAT1 Is Universally Expressed in Tissues from Seedlings to Booting Panicle

We examined the expression pattern of SHAT1 using quantitative RT-PCR and the β-glucuronidase (GUS) reporter gene under the control of the SHAT1 promoter. SHAT1 transcripts were universally present in all tested vegetative and reproductive tissues, from seedlings to the booting panicle by quantitative RT-PCR (Figure 5A). The GUS signal correlated well with the RT-PCR results. It was not only found in vegetative tissues, including leaf, stem, node, and leaf sheath (Figures 5B to to5E),5E), but also in young spikelet (Figure 5F). In addition, SHAT1 exhibited an intense signal in AZ and the inner floral organs of 2-mm-long spikelets (Figure 5G). When spikelet length was around 5 mm, GUS staining was weaker in the apiculus as well as in the palea and lemma (Figure 5H). With spikelet development, the GUS signals in spikelets gradually decayed (Figures 5I and and5J)5J) and completely disappeared when spikelet length was around 8 mm.

Figure 5.
The Expression Pattern of SHAT1 by Quantitative RT-PCR and GUS Assay.

The Expression of SHAT1 in AZ Is Positively Regulated by SH4

To exploit the effect of SHAT1 on AZ differentiation, we examined SHAT1 expression during early floral development using in situ hybridization. The inflorescence and spikelet developmental stages used in this study were according to the criteria reported before (Itoh et al., 2005). Our analyses of SHAT1 gene expression in the wild type showed that when lemma and palea primordia were first visible on the flanks of the floral meristem at sp6 stage, SHAT1 mRNA showed hotspots of expression in palea and lemma primordia (Figure 6A; see Supplemental Figure 7A online). Subsequently, SHAT1 signal declined in lemma and palea but appeared in the inner floral organ primordia, such as stamens and carpels during stage sp7, when carpel primordia began to differentiate (Figure 6B). When the ovule was first visible, SHAT1 transcripts were restricted to the AZ and anthers from early stage sp8 (Figure 6C). Afterwards, during late sp8 stage, SHAT1 expression accumulated to higher levels in the AZ than during early stage sp8 (Figure 6D). We also characterized SHAT1 expression patterns in GLA4. Its genetic background is very close to the wild type, as mentioned above, but its sh4-1 allele is the cultivated rice type. Therefore, we used GLA4 as the sh4-1 mutant line in this study. The SHAT1 signal in GLA4 was very similar to that in the wild type, showing converging expression in AZ from early stage sp8 and becoming more intense during late stage sp8 (Figures 6E to to6H).6H). By contrast, the sh4-2 mutation completely disrupted SHAT1 expression in AZ, with no signals observed during stages sp6-sp8 (Figures 6I to to6L),6L), similar to that in the shat1 mutant (Figures 6M to to6P6P).

Figure 6.
In Situ Hybridization of SHAT1 during Spikelet Developmental Stages Sp6 to Sp8.

The altered expression patterns of SHAT1 in the AZ of the sh4-2 mutant suggests that SHAT1 might act downstream of, or be positively regulated by, SH4 during AZ development. To verify this hypothesis, we examined SH4 expression using in situ hybridization. In the wild type, the convergent expression of SH4 in the AZ was first visible from early stage sp6 (Figure 7A; see Supplemental Figure 7B online). Notably, SH4 expression commenced earlier in the AZ than did that of SHAT1. Then, SH4 expression gradually increased in the AZ and anthers during stage sp7 (Figure 7B) and increased further at stage sp8 (Figures 7C and and7D).7D). The sh4-1 mutation did not change the SH4 expression pattern, as shown by the expression pattern in sh4-1 being similar to the wild type (Figures 7E to to7H).7H). By contrast, the sh4-2 mutation completely disrupted SH4 expression in AZs (Figures 7I to to7L).7L). These results confirmed that the sh4-1 allele is different from the sh4-2 allele. Whereas sh4-1 still plays a role in specifying AZ identity, sh4-2 appears to be a null-function mutation. We then determined SH4 expression in the shat1 mutant. SH4 exhibited a nearly intact expression pattern and expression level in the shat1 mutant until early sp8 stage (Figures 7M to 7O). However, SH4 transcripts were not detected in most cases from late sp8 stage when SH4 was expected to concentrate in AZ (Figure 7P). The disappearance of SH4 expression in the shat1 mutant suggests that SHAT1 functions in maintaining SH4 expression in the AZ during late stage sp8.

Figure 7.
In Situ Hybridization of SH4 during Spikelet Developmental Stages Sp6 to Sp8.

The Persistent and Concentrated Expression of SHAT1 and SH4 in AZ Is Required for AZ Identification

To explore further the relevance of SHAT1 and SH4 to AZ development in rice, we examined the genotypes of SHAT1 and SH4 and the AZ phenotype in a number of cultivated varieties. SH4 is highly conserved in cultivated rice varieties, with almost all varieties possessing the sh4-1 allele (Zhang et al., 2009). Similarly, SHAT1 was also conserved in cultivated rice, with few functional variants found in the coding region in genome sequences of 944 rice cultivars (http://www.ncgr.ac.cn/RiceHapMap). However, characterizing the shattering/nonshattering phenotypes showed that some varieties still showed nonshattering and had no AZ, for which the genotypes were SHAT1 and sh4-1. For example, japonica cv Nipponbare harbored SHAT1 and sh4-1 as did indica cv GLA4 (Table 1); however, confocal microscopy and scanning electron microscopy analysis showed that Nipponbare had no AZs (see Supplemental Figures 8A and 8B online).These results suggest the wild-type SHAT1 and sh4-1 genotypes do not form AZs.

Table 1.
Expression of SH4, SHAT1, and qSH1 and the Shattering Phenotypes

To explore whether the defect in AZ in Nipponbare was due to the temporal and spatial expression pattern changes of SHAT1 and sh4-1, we performed in situ hybridization of these two genes in Nipponbare. SHAT1 exhibited obvious signals in the AZ region during early stage sp8, although they were not so convergent compared with that in the wild type (Figures 8A to to8C).8C). However, these signals disappeared from this region during late stage sp8 (Figure 8D). There was a similar situation for SH4 with nearly intact expression pattern in AZ region from stage sp6 to early stage sp8 (Figures 8E to to8G)8G) but no signals during late stage sp8 (Figure 8H).Therefore, we proposed two questions: Did AZ specification require the continuous expression of active SHAT1 and sh4-1 in AZs? Were some functions of a genetic partner in maintaining SHAT1 and sh4-1 expression in AZs impaired in Nipponbare? Previous study revealed that qSH1 was responsible for AZ identification, so the incompetent qsh1 in Nipponbare may be such a defective genetic partner. To verify our hypothesis, we used a qSH1 near-isogenic line N52 in which the qSH1 locus was derived from the indica cv 93-11 but the other genetic background was from the japonica cv Nipponbare (see Supplemental Figure 9 online). The expression of qSH1 in N52 exhibited a wild-type like pattern (Figure 9H), in striking contrast with the absent expression in Nipponbare AZs (Konishi et al., 2006). We then determined SHAT1 and SH4 expression in N52 and, as expected, expression of both was maintained in AZs during late stage sp8 (Figures 8I to 8L and 8M to 8P) and finally resulted in AZ formation in N52 (see Supplemental Figures 8C and 8D online). These results suggest the identification of a competent AZ requires not only the wild-type SHAT1 and sh4-1 genotypes but also their persistent expression in AZ during the AZ differentiation process. qSH1 had its effect by maintaining SHAT1 and SH4 expression in AZ, thus promoting AZ differentiation.

Figure 8.
Effect of Persistent and Concentrated Expression of SHAT1 and SH4 on AZ Differentiation.
Figure 9.
In Situ Hybridization of qSH1.

qSH1 Functions Downstream of SHAT1 and SH4

The functions of qSH1 in maintaining SHAT1 and SH4 expression in AZs led us to investigate whether qSH1 activity was dependent on SHAT1 and SH4. To address this, qSH1 expression was examined in the wild type and shat1 and sh4-2 mutants. In the wild type, the qSH1 signal commenced in AZ from early stage sp8 (Figures 9A to 9C and 9E). During late stage sp8, the qSH1 signal became more intense in AZs and anthers (Figure 9D) and closely resembled that of SHAT1 in the wild type. By contrast, no qSH1 signal was detected in AZs of either the shat1 mutant (Figure 9F) or the sh4-2 mutant (Figure 9G). These results indicated that qSH1 functions downstream of SHAT1 and SH4 in maintaining their expression in the AZ, thus promoting AZ differentiation.


SHAT1 Is a Member of the AP2 Family of Genes

The AP2 domain defines a large gene family of mostly plant specific DNA binding proteins called AP2/Ethylene Response Factors (Riechmann and Meyerowitz, 1998). Those carrying two AP2 domain genes are classified into two groups: AP2 and AINTEGUMENTA. The AP2 group members harbor a miR172 target site (Magnani et al., 2004; Kim et al., 2006). Phylogenetic analysis showed that SHAT1 belongs to the AP2 group and is closely related to barley (Hordeum vulgare) Cleistogamy1 (Cly1) and Arabidopsis AP2 (see Supplemental Figure 10, Supplemental References 1, and Supplemental Data Set 1 online). Cly1 is known to regulate lodicule development. A synonymous nucleotide substitution at the miR172 targeting site in Cly1 abolishes miR172-mediated cleavage of Cly1 and results in the failure to develop normal lodicules, thus producing the cleistogamous phenotype (Nair et al., 2010). In the shat1 mutant, enlarged and/or an increased number of lodicules was also frequently observed (see Supplemental Figures 2J and 3E online), and, occasionally, lodicules were elongated and transformed into lemma/palea-like organs (see Supplemental Figure 2M online); consequently, these spikelets were unable to close after flowering and resulted in naked grains. These phenotypes were previously observed in transgenic rice plants overexpressing miR172b (Zhu et al., 2009). These results suggest that miR172-mediated cleavage of SHAT1 might also be essential for normal flower opening in rice. However, the seed shattering defects we observed in the rice shat1 mutant were not reported in the barley cly1 mutant (Nair et al., 2010). Similarly, AP2 in Arabidopsis has been shown to influence many critical aspects of development, from regulating reproductive organ morphogenesis (Jofuku et al., 2005; Ohto et al., 2005; Würschum et al., 2006; Yant et al., 2010) to determining flowering time (Yant et al., 2010); however, none of them are related to cell separation processes, such as petal abscission or fruit dehiscence.

As far as we know, the sole AP2 gene reported to affect seed shattering is the wheat Q gene; this affects a repertoire of characters important for domestication, such as threshability, glume shape, and glume tenacity (Simons et al., 2006). Although Q may represent a duplicate paralog distinct from rice SHAT1, the defects of the AZ phenotype in q greatly resembled those of shat1. A single spikelet of wheat is composed of three fertile florets and a pair of glumes. The AZ is close to the base of the glumes. The mutant q spikelet also has three fertile florets and a pair of glumes, but the region destined to be the AZ is substituted by a longer rachilla (Simons et al., 2006). The rice spikelet is similar to that of wheat; however, in rice, the second and third florets degenerate into sterile lemmas and the original glumes degenerate into rudimentary glumes (Yoshida and Nagato, 2011). The AZ is located between the sterile lemmas and the rudimentary glumes. In the shat1 mutant, the region destined to be the AZ was substituted by a longer and crook-neck-like rachilla. The comparable AZ phenotype caused by the paralog shat1 and q genes suggests that rice and wheat may share similar mechanisms in seed AZ development.

SHAT1 Expression and AZ Differentiation

A previous anatomical study of AZs in rice revealed that the AZ could be observed by elongation of cells in the pedicel and rachilla when the panicle length was 20 to 30 mm and spikelet length was around 2 mm (Jin, 1986). Studies on oat (Avena sativa), another monocotyledon, also showed that prior to cell separation the site where cell wall breakdown will take place was well defined (Hoekstra et al., 2001). In this study, SHAT1 mRNA accumulated in AZs as early as sp8 stage when carpel primordia began to differentiate. However, we cannot rule out whether the morphological differentiation process of AZ cells starts at sp8 stage, since provisional AZ cells generally appear very similar to adjacent cells during this period. Our observations showed that the time when AZ cells were first distinguishable from adjacent cells in the wild type was around 15 d before heading (see Supplemental Figures 11A and 11D online). During this time, the length of spikelet was around 2 to 3 mm, whereas in contemporaneous shat1 mutant, such one or two layers of flattened and small AZ cells were not observed (see Supplemental Figures 11B and 11C online). Another interesting finding was that the expression intensity of SHAT1 increased with development of the spikelet (Figures 6C and and6D).6D). One possible explanation is that SHAT1 may be positively regulated by its gene products. Arabidopsis AP2 was reported to participate in a self-feedback loop through negatively regulating its inhibitor miR172 (Yant et al., 2010), and a similar mechanism may exist in rice.

New Insight into SH4 Function in AZ Specification

SH4 has previously been identified to affect AZ formation as well as the hydrolysis process (Li et al., 2006). Here, we identified a new SH4 allelic mutation, sh4-2. Confocal microscopy and scanning electron microscopy analysis showed that phenotypes of the AZ in sh4-2 were very different from that resulting from the cultivated rice sh4-1 allele. Compared with the isodiametrically flattened and thin-walled appearance in sh4-1, AZ cells in sh4-2 had a flattened shape but lost the thin-walled property (Figure 1R; see Supplemental Figure 5A online). Moreover, in situ hybridization results showed that the expression of SH4 in the sh4-2 mutant was completely abolished (Figure 7L), whereas it showed a wild-type-like pattern in sh4-1 (Figure 7H). Therefore, sh4-2 was assumed to be a null-function mutant. However, one of the most unexpected findings in this study was that the sh4-2 mutation suppressed the sh4-1 phenotype. Plants with these intragenic suppressor mutations had a similar seed-shattering habit to wild-type plants (see Supplemental Figure 12 online). This originally led us to the conclusion that sh4-2 could be a different locus from the cultivated rice sh4-1. Since sh4-2 had a one-nucleotide insertion in the second exon of SH4, whereas sh4-1 had a one-nucleotide substitution in the first exon, the functional SH4 may be a hybrid multimeric protein formed from different domains produced by different mutant alleles. This hypothesis is supported by the finding that SH4 protein is predicted to have a coiled-coil structure that probably forms a homopolymer. Examples of intragenic complementation through forming a tetrameric protein include ASL in human (Turner et al., 1997). However, in situ hybridization did not detect SH4 expression in the sh4-2 mutant, leading to the question of how a heterodimer could be generated if the sh4-2 allele lacked any mRNA. Since the in situ technique could not quantify the gene expression levels, we performed quantitative RT-PCR to detect any expression of SH4 in sh4-1 and sh4-2 mutants (see Supplemental Figure 13A online). Compared with the wild type, SH4 showed a decrease in both sh4-1 and sh4-2, with a greater decline in sh4-2. The presence of some expression of SH4 in sh4-2 seems inconsistent with the in situ results, although this may reflect the higher resolution of the quantitative RT-PCR technique.

Dissecting the Relative Contribution of SHAT1, SH4, and qSH1 to Seed AZ Differentiation

SHAT1, SH4, and qSH1 are three important genes involved in AZ differentiation. We summarized the relationship between their expression and shattering phenotype in Table 1. SH4 seemed to play a role early in AZ formation. The earliest accumulation of SH4 in AZ and the inability to detect SHAT1 and qSH1 expression in AZs of the sh4-2 mutant suggest that SH4 acts largely upstream of SHAT1 and qSH1 (Figure 10). The highly accumulated SH4 may confer the provisional AZ cells with some properties that distinguish them from their neighbors, thus providing clues for SHAT1 and qSH1 accumulation in a narrow strip of AZ cells. The in situ hybridization results showed that only after SH4 accumulated in the AZ did SHAT1 and qSH1 expression begin to be limited to the AZ, then becoming more intense in the AZ with the development of spikelet (Table 1). Our results provide a striking contrast to previous reports that qSH1 had a stronger effect on seed shattering than did sh4-1 in the genetic background of japonica cv Nipponbare (Onishi et al., 2007; Ishikawa et al., 2010). We interpret this discrepancy as suggestive of different effects from the different mutation alleles of SH4. Since the sh4-1 allele in cultivated rice was still functional, it is difficult to decipher the genetic interaction among SH4 and other shattering-related genes. By contrast, sh4-2 is a presumed null-function mutation, which afforded us a good opportunity to explore the function of SH4 in the shattering pathway.

Figure 10.
A Genetic Model of Regulatory Network Specifying AZ Development in Rice.

SHAT1 appeared to play a dual-functional role in AZ specification: functioning downstream of SH4 to activate qSH1 expression and maintaining the expression of SH4 in the AZ (Figure 10). The inability to detect qSH1 expression in the shat1 mutant suggests that qSH1 is activated by SHAT1 rather than by SH4, given that SH4 still exhibited intact expression in shat1 during early stage sp8, when qSH1 was expected to be expressed in the AZ (Table 1). However, we cannot rule out that this activation could be fulfilled by SHAT1 alone or may require the help of SH4. More experiments should be performed to examine qSH1 expression in the sh4-2 background and how SHAT1 expression was maintained in AZ. On the other hand, the disappearance of SH4 expression in the shat1 AZ during late-stage sp8 suggests that SHAT1 plays an important role in maintaining SH4 expression in the AZ, which is required for AZ identity (Table 1). The SHAT1-dependent activity of SH4 was much more apparent when comparing AZ phenotypes of shat1, sh4-2, and shat1 sh4-2 mutants. The AZ phenotype of shat1 sh4-2 was similar to that of shat1, which developed a crook-neck-like rachilla (Figure 1V) and showed peanut-like cortical cells in the AZ position (Figures 1W and and1X).1X). This is compatible with the result that shat1 sh4-2 showed similar changes in BTS values to shat1 rather than to sh4-2 (Figure 1B). Although both shat1 and sh4-2 did not develop AZs, we observed different changes in BTS values among these two mutant lines, with sh4-2 showing higher values before 13 d (Figure 1B). We explain these results as the crook-neck-like structure being able to endure only around 150-g pull-off forces without fracturing. Therefore, the BTS values of shat1 and shat1 sh4-2 were even lower than for the wild type before 9 d (Figure 1B).

In addition to these two functions, SHAT1 seemed to play other roles independent of SH4. This hypothesis is supported by two findings: First, there were various spikelet defects seen in the shat1 mutant but not the sh4-2 mutant (see Supplemental Figures 2B to 2E online); second, SHAT1 maintained some expression in the sh4-2 mutant. Although we did not detect SHAT1 expression in the sh4-2 mutant using in situ hybridization, SHAT1 showed a slight increase in sh4-1 and a modest repression in the sh4-2 mutant (see Supplemental Figure 13B online) using quantitative RT-PCR. The reduced, yet persistent, expression of SHAT1 in the sh4-2 mutant suggests that SHAT1 expression is not simply regulated by SH4. The sh4-2 mutation disrupted SHAT1 expression in AZs, whereas the other expression domains, such as anthers, might be unaffected.

qSH1 affects the maintenance of SHAT1 and SH4 expression in the AZ, thus promoting the AZ differentiation process (Figure 10). In japonica cv Nipponbare, the expression of SHAT1 and SH4 receded from the AZ during late stage sp8 and ultimately the AZ was not formed (Table 1). Conversely, the defective AZ phenotype in Nipponbare was suppressed by sustained expression of SHAT1 and SH4 in the AZ after introgression of a functional qSH1 locus from indica cv 93-11. These results suggest that qSH1 functions in maintaining SHAT1 and SH4 expression in the AZ, thus promoting AZ identity specification. However, qSH1 may not be the only genetic partner determining the expression of SHAT1 and SH4 in the AZ, as some qSH1-defective rice subspecies appeared to show a reduced-shattering habit (Konishi et al., 2006). Since we are not sure whether AZs are present in these rice subspecies and how the qSH1 behaves in them, more experiments should be performed to test the possibility if one or more factor(s) (in addition to qSH1) are able to sustain the expression of SH4 and SHAT1 in AZs, thus promoting AZ differentiation.

Selection for Shattering Genes in Domestication

Selection of reduced seed shattering is still an important challenge in many agricultural crops worldwide (Gan et al., 2008). In the process of rice domestication, artificial selection is likely to have favored mutation that reduced but did not completely eliminate grain shattering. The cultivated rice sh4-1 allele achieved an ideal balance between shattering and threshing during rice domestication. Zhang et al. (2009) demonstrated that this reduced-shattering sh4-1 allele was quickly fixed in all rice cultivars, with levels of sequence polymorphism significantly reduced in both indica and japonica cultivars relative to the wild progenitors. To explore whether mutations in the SHAT1 gene also occur in the natural population, we determined the polymorphism of SHAT1 in different rice landraces. Our high-throughput sequencing results (http://www.ncgr.ac.cn/RiceHapMap) of 614 accessions of landraces from China and 330 accessions of international varieties showed 19 SNPs located in the SHAT1 genic region (Huang et al., 2010): Eight were located in the UTR regions, four in introns, and seven in exons. Those SNPs generated few functional variants, with most cultivated varieties having the same protein sequence as that in O. rufipogon W1943. Therefore, differing from SH4, SHAT1 might not be subject to artificial selection during domestication.


Mutant Material and Growth Conditions

Firstly, we constructed a CSSL designated as SL4 in the Oryza sativa ssp indica cv GLA4 background with the whole chromosome 4 substituted by Oryza rufipogon W1943 with the average density of detecting markers of ~2.5 Mb (see Supplemental Figures 1A and 1B online). We then treated ~10,000 T0 generation seeds of SL4 with 60Co γ-rays (65 Gy) and generated 4800 lines in the T1 generation. We screened for nonshattering mutants in the T1 generation. The mutant line 2B70080 was identified in the T-DNA insertion line database (Jeon et al., 2000). The indica cv GLA4 and the japonica cv Nipponbare were used as sh4-1 mutant line and qsh1 mutant line, respectively, according to the previous sequencing results (Konishi et al., 2006; Li et al., 2006). All plants (O. sativa) were grown in the paddy field of Shanghai Plant Physiology and Ecology, Shanghai, China.

Characterization of Mutant Phenotype

Plants materials were photographed with a Nikon E5400 digital camera and a Nikon SMZ1000 dissecting microscope. For scanning electron microscopy observation, the spikelets of SL4 and shat1 at ~35 d after heading were collected and processed essentially as described by Keijzer et al. (1996) and observed with a JSM-6360LV scanning electron microscope (Jeol). Confocal microcopy was performed as described by Li et al. (2006). For shattering degree tests, the BTS upon detachment of seeds from the pedicels by pulling was measured by a digital force gauge (Qin et al., 2010). For an individual plant, a total of 100 flowers or grains from five panicles were measured. For transgenic plants, measurement was made on panicles ~30 d after anthesis.

Cloning of SHAT1 and SHAT2

For the positional cloning of SHAT1, shat1 was crossed with O. sativa ssp japonica cv Nipponbare. A total of 300 crook-neck-like F2 plants were selected for mapping analysis. DNA was extracted from fresh leaves according to the cetyl-trimethylammonium bromide method (Murray and Thompson., 1980) with minor modifications. The molecular markers used in this study are listed in Supplemental Table 1 online. Mutation sites in shat1 were determined by PCR amplification and sequencing analysis.

For the positional cloning of SHAT2, shat2 was first crossed with japonica cv Nipponbare. A total of 96 F2 plants that exhibited a nonshattering phenotype and had SH4 and qSH1 loci from shat2 were selected for primary mapping analysis. After SHAT2 was mapped to the long arm of chromosome 4, shat2 was then crossed with indica cv GLA4 to construct a new F2 population. There were 907 nonshattering F2 plants selected for fine mapping.

RNAi Experiment

To generate the SHAT1-RNAi construct for SHAT1 gene suppression, a 426-bp fragment of SHAT1 cDNA was PCR amplified with primer SHAT1-RNAi-F (5′-cggggtaccactagtCAACCGCTACAGCAGCTGCA-3′, KpnI, SpeI) and SHAT1-RNAi-R (5′-cgcggatccgagctcACTGCTTGAGGCGACGCTTG-3′, BamHI, SacI), which harbors restriction sites (set in lowercase letters) for cloning. The resulting PCR products were first digested by SpeI and SacI and ligated into vector pTCK303 (Wang et al., 2004) to get the transitional vector. Then, the PCR products were digested by KpnI and BamHI and ligated into the transitional vector. The resulting RNAi construct was used for knockdown of gene expression.

RNA Extraction and Quantitative RT-PCR Analysis

Total RNA was extracted using the Trizol reagent (Invitrogen) according to the manufacturer’s instructions. After treatment with DNaseI (NEB), 5 μg of total RNA was used to synthesize the oligo(dT) primed first-strand cDNA using SuperScript II reverse transcriptase (Invitrogen). Quantitative RT-PCR was performed on the Applied Biosystems 7500 real-time PCR System. Diluted cDNA was amplified using SYBR Premix Ex Taq (TaKaRa). The levels of SHAT1 transcripts were normalized by endogenous Ubiquitin transcripts. Each set of experiments was repeated three times. Primers used for quantitative real-time PCR are listed in Supplemental Table 2 online.

Nuclear Localization Analysis and Transactivation Assay

To construct the SHAT1-GFP fusion plasmid, the full-length SHAT1 coding region without stop codon was PCR amplified with primers containing XhoI and SpeI site (see Supplemental Table 2 online) and was then in-frame cloned into the XhoI-SpeI site of pA7 (Peng et al., 2008), which contained a GFP coding sequence under the control of 35S promoter (kindly provided by Zhang Jingliu). Transient expression of the pA7- SHAT1-GFP fusion in onion epidermal cells was performed as previously described (Scott et al., 1999) using a helium biolistic device (PDS-1000; Bio-Rad). The samples were observed with a confocal laser microscope (LSM510; Zeiss).

We performed the transactivation activity assay using the yeast one-hybrid system (Clontech). To construct pGBKT7-SHAT1, pGBKT7-SHAT1ΔC (1 to 326), and pGBKT7-SHAT1ΔN (112 to 459), the full-length coding sequence and the N and C termini of SHAT1 were amplified by PCR (see Supplemental Table 2 online). The PCR products were digested with EcoRI and BamHI and cloned into pGBKT7 to fuse to the GAL4 binding domain. We transformed all vectors into yeast strain AH109 using electroporation and selected on SD/Trp- plates. After 2 d of incubation at 30°C, the yeast colonies were diluted to an OD600 of 0.5 and dropped on either SD/Trp-/Ade- plates or SD/Trp-/His- plates with 0.5 mM of 3AT. These plates were incubated at 30°C until yeast cells grew to form colonies.

Construction of the SHAT1 Promoter-GUS Fusion and GUS Assay

For constructing the SHAT1 promoter:GUS fusion plasmid, a 2.3-kb region (from −2393 to −66 bp from the translation start site) was PCR amplified from japonica cv Nipponbare with primers containing SalI and BamHI (see Supplemental Table 2 online). After digestion, the released segment was ligated upstream of the GUS in the pCAMBIA1300GN:GUS (Ren et al., 2005) (kindly provided by Lin Hongxuan).

For detection of GUS activity, tissue was fixed in 90% acetone for 1 h at 4°C and then rinsed with GUS buffer (100 mM NaH2PO4/Na2HPO4, pH 7.0, 10 mM EDTA, 0.1% Triton X-100, 0.5 mM potassium ferricyanide, and 0.5 mM potassium ferrocyanide). Samples were incubated with GUS buffer supplemented with 0.05% 5-bromo-4- chloro-3-indolyl β-d-glucuronide cyclohexylamine salt (Rose Scientific) at 37°C overnight. Then, tissues were cleared overnight in destaining buffer (ethanol:acetic acid = 84:16) and washed several times with 70% ethanol before observation under a dissecting microscope.

In Situ Hybridization

Young panicles were fixed in 4% paraformaldehyde, dehydrated through an ethanol series, embedded in paraplast (Sigma-Aldrich), and sectioned at 8-μm thickness using a rotary microtome (Leica). A 426-bp gene-specific region of SHAT1 cDNA, amplified by PCR reaction (primers 5′-CAACCGCTACAGCAGCTGCA-3′ and 5′-ACTGCTTGAGGCGACGCTTG-3′); a 446-bp gene-specific region of SH4 cDNA, amplified by PCR reaction (primers 5′-ATCATCGGCCGGAGGAGTCG-3′ and 5′-GCACCACCATCACGGCCATC-3′); and a 227-bp gene-specific region of qSH1 cDNA, amplified by PCR reaction (primers 5′-CGAAGCTCATCTCCATGATG-3′ and 5′-TGCAGGAAGTGTTCGAACAG-3′), were used as the probes. The amplified DNA fragments were subcloned into a pGEM-T easy vector (Promega) in two orientations; the sense and antisense probes were synthesized and used to generate the RNA probe. In situ hybridization was performed as described (Luo et al., 1996).

Supplemental Data

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


We thank Hongquan Yang (Shanghai Jiao Tong University) and Zhenbiao Yang (University of California, Riverside) for critical reading of the manuscript. We thank Jiqin Li for resin sections, Zhiping Zhang for scanning electron microscopy observation, and Hongxuan Lin for providing the vector pCAMBIA1300GN at the Institute of Plant Physiology and Ecology, Chinese Academy of Sciences. We thank Kang Chong for providing the vector pTCK 303 at the Institute of Botany, Chinese Academy of Sciences. We thank Qiaoquan Liu for providing us the CSSL line N52 (Yangzhou University), and Gynheung An for providing the T-DNA insertion mutant line 2B70080 (Kyung Hee University). This work was supported by the Ministry of Science and Technology of China (2011CB100205, 2012AA10A302 and 2012AA10A304), and the Ministry of Agriculture of China (2011ZX08001-004 and 2011ZX08009-002), the National Natural Science Foundation of China (31121063), and the Chinese Academy of Sciences to B.H.


Y.Z., D.L., B.Z., and B.H. designed the experiment. Y.Z., D.L., C.L., J.L., B.Z., B.-F.Z., Z.W., Y.S., and J.Z. performed the experiment and analyzed the data. T.S. supervised measuring pedicel BTS and gene cloning. Y.Z. and B.H. wrote the article.


  • Briggs C.L., Morris E.C. (2008). Seed-coat dormancy in Grevillea linearifolia: Little change in permeability to an apoplastic tracer after treatment with smoke and heat. Ann. Bot. (Lond.) 101: 623–632 [PMC free article] [PubMed]
  • Chen X.M. (2004). A microRNA as a translational repressor of APETALA2 in Arabidopsis flower development. Science 303: 2022–2025 [PubMed]
  • Doebley J. (2006). Plant science. Unfallen grains: How ancient farmers turned weeds into crops. Science 312: 1318–1319 [PubMed]
  • Fuller D.Q., Qin L., Zheng Y.F., Zhao Z.J., Chen X., Hosoya L.A., Sun G.P. (2009). The domestication process and domestication rate in rice: spikelet bases from the Lower Yangtze. Science 323: 1607–1610 [PubMed]
  • Gan Y., Malhi S.S., Brandt S.A., McDonald C.L. (2008). Assessment of seed shattering resistance and yield loss in five oilseed crops. Can. J. Plant Sci. 88: 267–270
  • Hoekstra G.J., Darbyshire S.J., Mather D.E. (2001). Anatomical features at the disarticulation zone in florets of fatuoid and nonfatuoid oat (Avena sativa L.). Can. J. Bot. 79: 1409–1416
  • Huang X.H., et al. (2010). Genome-wide association studies of 14 agronomic traits in rice landraces. Nat. Genet. 42: 961–967 [PubMed]
  • Ishikawa R., Thanh P.T., Nimura N., Htun T.M., Yamasaki M., Ishii T. (2010). Allelic interaction at seed-shattering loci in the genetic backgrounds of wild and cultivated rice species. Genes Genet. Syst. 85: 265–271 [PubMed]
  • Itoh J., Nonomura K., Ikeda K., Yamaki S., Inukai Y., Yamagishi H., Kitano H., Nagato Y. (2005). Rice plant development: from zygote to spikelet. Plant Cell Physiol. 46: 23–47 [PubMed]
  • Jeon J.S., et al. (2000). T-DNA insertional mutagenesis for functional genomics in rice. Plant J. 22: 561–570 [PubMed]
  • Ji H., et al. (2010). Inactivation of the CTD phosphatase-like gene OsCPL1 enhances the development of the abscission layer and seed shattering in rice. Plant J. 61: 96–106 [PubMed]
  • Jin I.D. (1986). On the formation and development of abscission layer in rice plants, Oryza sativa. Jpn. J. Crop. Sci. 55: 451–457
  • Jofuku K.D., Omidyar P.K., Gee Z., Okamuro J.K. (2005). Control of seed mass and seed yield by the floral homeotic gene APETALA2. Proc. Natl. Acad. Sci. USA 102: 3117–3122 [PMC free article] [PubMed]
  • Keijzer C.J., Reinders M.C., LeferinkTenKlooster H.B. (1996). The mechanics of the grass flower: The extension of the staminal filaments and the lodicules of maize. Ann. Bot. (Lond.) 77: 675–683
  • Kim S., Soltis P.S., Wall K., Soltis D.E. (2006). Phylogeny and domain evolution in the APETALA2-like gene family. Mol. Biol. Evol. 23: 107–120 [PubMed]
  • Konishi S., Izawa T., Lin S.Y., Ebana K., Fukuta Y., Sasaki T., Yano M. (2006). An SNP caused loss of seed shattering during rice domestication. Science 312: 1392–1396 [PubMed]
  • Li C.B., Zhou A.L., Sang T. (2006). Rice domestication by reducing shattering. Science 311: 1936–1939 [PubMed]
  • Lin Z., Griffith M.E., Li X., Zhu Z., Tan L., Fu Y., Zhang W., Wang X., Xie D., Sun C. (2007). Origin of seed shattering in rice (Oryza sativa L.). Planta 226: 11–20 [PubMed]
  • Lu G., Gao C.X., Zheng X.N., Han B. (2009). Identification of OsbZIP72 as a positive regulator of ABA response and drought tolerance in rice. Planta 229: 605–615 [PubMed]
  • Luo D., Carpenter R., Vincent C., Copsey L., Coen E. (1996). Origin of floral asymmetry in Antirrhinum. Nature 383: 794–799 [PubMed]
  • Magnani E., Sjölander K., Hake S. (2004). From endonucleases to transcription factors: evolution of the AP2 DNA binding domain in plants. Plant Cell 16: 2265–2277 [PMC free article] [PubMed]
  • McKim S.M., Stenvik G.E., Butenko M.A., Kristiansen W., Cho S.K., Hepworth S.R., Aalen R.B., Haughn G.W. (2008). The BLADE-ON-PETIOLE genes are essential for abscission zone formation in Arabidopsis. Development 135: 1537–1546 [PubMed]
  • Murray M.G., Thompson W.F. (1980). Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res. 8: 4321–4325 [PMC free article] [PubMed]
  • Nair S.K., et al. (2010). Cleistogamous flowering in barley arises from the suppression of microRNA-guided HvAP2 mRNA cleavage. Proc. Natl. Acad. Sci. USA 107: 490–495 [PMC free article] [PubMed]
  • Ohto M.A., Fischer R.L., Goldberg R.B., Nakamura K., Harada J.J. (2005). Control of seed mass by APETALA2. Proc. Natl. Acad. Sci. USA 102: 3123–3128 [PMC free article] [PubMed]
  • Onishi K., Takagi K., Kontani M., Tanaka T., Sano Y. (2007). Different patterns of genealogical relationships found in the two major QTLs causing reduction of seed shattering during rice domestication. Genome 50: 757–766 [PubMed]
  • Peng L.T., Shi Z.Y., Li L., Shen G.Z., Zhang J.L. (2008). Overexpression of transcription factor OsLFL1 delays flowering time in Oryza sativa. J. Plant Physiol. 165: 876–885 [PubMed]
  • Qin Y., Kim S.M., Zhao X.H., Jia B.Y., Lee H.S., Kim K.M., Eun M.Y., Jin I.D., Sohn J.K. (2010). Identification for quantitative trait loci controlling grain shattering in rice. Genes Genomics. 32: 173–180
  • Ren Z.H., Gao J.P., Li L.G., Cai X.L., Huang W., Chao D.Y., Zhu M.Z., Wang Z.Y., Luan S., Lin H.X. (2005). A rice quantitative trait locus for salt tolerance encodes a sodium transporter. Nat. Genet. 37: 1141–1146 [PubMed]
  • Riechmann J.L., Meyerowitz E.M. (1998). The AP2/EREBP family of plant transcription factors. Biol. Chem. 379: 633–646 [PubMed]
  • Scott A., Wyatt S., Tsou P.L., Robertson D., Allen N.S. (1999). Model system for plant cell biology: GFP imaging in living onion epidermal cells. Biotechniques 26: 1125–1132 [PubMed]
  • Simons K.J., Fellers J.P., Trick H.N., Zhang Z.C., Tai Y.S., Gill B.S., Faris J.D. (2006). Molecular characterization of the major wheat domestication gene Q. Genetics 172: 547–555 [PMC free article] [PubMed]
  • Szymkowiak E.J., Irish E.E. (1999). Interactions between jointless and wild-type tomato tissues during development of the pedicel abscission zone and the inflorescence meristem. Plant Cell 11: 159–175 [PMC free article] [PubMed]
  • Taghizadeh M.S., Crawford S., Nicolas M.E., Cousens R.D. (2009). Water deficit changes the anatomy of the fruit abscission zone in Raphanus raphanistrum (Brassicaceae). Aust. J. Bot. 57: 708–714
  • Taylor J.E., Whitelaw C.A. (2001). Signals in abscission. New Phytol. 151: 323–339
  • Turner M.A., Simpson A., McInnes R.R., Howell P.L. (1997). Human argininosuccinate lyase: A structural basis for intragenic complementation. Proc. Natl. Acad. Sci. USA 94: 9063–9068 [PMC free article] [PubMed]
  • Wang M., Chen C., Xu Y.Y., Jiang R.X., Han Y., Xu Z.H., Chong K. (2004). A practical vector for efficient knockdown of gene expression in rice (Oryza sativa L.). Plant Mol. Biol. Rep. 22: 409–417
  • Wollmann H., Mica E., Todesco M., Long J.A., Weigel D. (2010). On reconciling the interactions between APETALA2, miR172 and AGAMOUS with the ABC model of flower development. Development 137: 3633–3642 [PMC free article] [PubMed]
  • Würschum T., Gross-Hardt R., Laux T. (2006). APETALA2 regulates the stem cell niche in the Arabidopsis shoot meristem. Plant Cell 18: 295–307 [PMC free article] [PubMed]
  • Yant L., Mathieu J., Dinh T.T., Ott F., Lanz C., Wollmann H., Chen X.M., Schmid M. (2010). Orchestration of the floral transition and floral development in Arabidopsis by the bifunctional transcription factor APETALA2. Plant Cell 22: 2156–2170 [PMC free article] [PubMed]
  • Yoshida H., Nagato Y. (2011). Flower development in rice. J. Exp. Bot. 62: 4719–4730 [PubMed]
  • Zhang L.B., Zhu Q.H., Wu Z.Q., Ross-Ibarra J., Gaut B.S., Ge S., Sang T. (2009). Selection on grain shattering genes and rates of rice domestication. New Phytol. 184: 708–720 [PubMed]
  • Zhu B.F., et al. (2011). Genetic control of a transition from black to straw-white seed hull in rice domestication. Plant Physiol. 155: 1301–1311 [PMC free article] [PubMed]
  • Zhu Q.H., Upadhyaya N.M., Gubler F., Helliwell C.A. (2009). Over-expression of miR172 causes loss of spikelet determinacy and floral organ abnormalities in rice (Oryza sativa). BMC Plant Biol. 9: 149–161 [PMC free article] [PubMed]

Articles from The Plant Cell are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...