miR156-Targeted SBP-Box Transcription Factors Interact with DWARF53 to Regulate TEOSINTE BRANCHED1 and BARREN STALK1 Expression in Bread Wheat

miR156-regulated SBP-box transcription factors function antagonistically with DWARF53 to regulate TEOSINTE BRANCHED1 and BARREN STALK1 expression during bread wheat tillering and spikelet development.

Abstract

Genetic and environmental factors affect bread wheat (Triticum aestivum) plant architecture, which determines grain yield. In this study, we demonstrate that miR156 controls bread wheat plant architecture. We show that overexpression of tae-miR156 in bread wheat cultivar Kenong199 leads to increased tiller number and severe defects in spikelet formation, probably due to the tae-miR156-mediated repression of a group of SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) genes. Furthermore, we found that the expression of two genes TEOSINTE BRANCHED1 (TaTB1) and BARREN STALK1 (TaBA1), whose orthologous genes in diverse plant species play conserved roles in regulating plant architecture, is markedly reduced in the tae-miR156-OE bread wheat plants. Significantly, we demonstrate that the strigolactone (SL) signaling repressor DWARF53 (TaD53), which physically associates with the transcriptional corepressor TOPLESS, can directly interact with the N-terminal domains of miR156-controlled TaSPL3/17. Most importantly, TaSPL3/17-mediated transcriptional activation of TaBA1 and TaTB1 can be largely repressed by TaD53 in the transient expression system. Our results reveal potential association between miR156-TaSPLs and SL signaling pathways during bread wheat tillering and spikelet development.

Bread wheat (Triticum aestivum) is a major staple crop worldwide. Global demand for bread wheat is increasing with world population growth. To guarantee global food security, people have been seeking elite agronomic traits of bread wheat to improve its yield. Genetic and environmental factors affect plant architecture, which strongly influences crop productivity. The identification and characterization of the regulatory genes associated with wheat plant architecture is indispensable for both understanding the genetic basis of phenotypic variation and facilitating the breeding of elite varieties with ideal wheat plant architecture. However, in addition to the wheat “green revolution” gene Reduced height 1 (Peng et al., 1999), the main genes that determine wheat plant architecture remain to be identified.

MicroRNA156 (miR156) targets members of the SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) gene family for cleavage and/or translational repression, which encodes transcription factors that, in turn, regulate a large network concerning plant growth and development (Cardon et al., 1999; Wu and Poethig, 2006; Schwarz et al., 2008; Wang et al., 2009; Wu et al., 2009; Yamaguchi et al., 2009; Yu et al., 2010; Gou et al., 2011). SPL transcription factors share a highly conserved DNA binding domain named the SQUAMOSA PROMOTER BINDING PROTEIN (SBP)-box and have been shown to associate with motifs with the consensus sequence TNCGTACAA (N represents any base) to promote the downstream target genes transcription (Cardon et al., 1999; Yamasaki et al., 2004). In the past few years, a number of key genes controlling important agronomic traits have been cloned from rice (Oryza sativa), among which are several transcription factors from the SPL gene family (Chuck et al., 2007; Jiao et al., 2010; Miura et al., 2010; Wang et al., 2012; Xie et al., 2012; Si et al., 2016). The rice IDEAL PLANT ARCHITECTURE1 (IPA1) gene, isolated by the map-based approach, encodes OsSPL14, and the ipa1 mutant shows ideal plant architecture with decreased tiller number and increased plant height and panicle branches (Jiao et al., 2010). WEALTHY FARMER’S PANICLE, another overexpression allele of OsSPL14, resulted from an epigenetic change in the OsSPL14 promoter and shows a similar phenotype (Miura et al., 2010). The quantitative trait locus GW8, encoding OsSPL16, regulates rice grain size, shape, and quality (Wang et al., 2012). The major quantitative trait locus GLW7, encoding OsSPL13, positively regulates cell size in the grain hull, resulting in enhanced rice grain length and yield (Si et al., 2016). Therefore, it has been suggested that elite SPLs alleles have great potential for improving crop agronomic traits and enhancing grain yield.

In addition to SPLs, several key genes associated with crop architecture have been identified from maize (Zea mays) and rice. For example, the maize TEOSINTE BRANCHED1 (TB1) and its orthologous genes rice OsTB1 (FINE CULM1) and Arabidopsis (Arabidopsis thaliana) BRANCHED1, belonging to members of the TEOSINTE BRANCHED1, CYCLOIDEA AND PCF TRANSCRIPTION FACTOR (TCP) gene family, function as conserved negative regulators of lateral branching (Doebley et al., 1997; Takeda et al., 2003; Aguilar-Martínez et al., 2007). Also, the maize BARREN STALK1 (BA1) gene encodes a noncanonical bHLH protein, and the tassel of ba1 mutants is unbranched, shortened, and predominantly sterile owing to the often complete lack of spikelets (Gallavotti et al., 2004).

Strigolactones (SLs) are a group of newly identified plant hormones that suppress plant shoot branching (Gomez-Roldan et al., 2008; Umehara et al., 2008). In recent years, great advances have been made in SL signaling in the model plants Arabidopsis and rice. In rice, SL signaling requires degradation of the SL signaling repressor DWARF53 (D53), which is mediated by a receptor complex including D14 and D3 (Jiang et al., 2013; Zhou et al., 2013). In Arabidopsis, the SL receptor complex including AtD14 and the D3 ortholog MORE AXILLARY GROWTH2 mediates degradation of the D53-like proteins, SUPPRESSOR OF MORE AXILLARY GROWTH 2-LIKE6 (SMXL6), SMXL7, and SMXL8 (Soundappan et al., 2015; Wang et al., 2015; Yao et al., 2016). However, the downstream transcription factors repressed by the D53 transcriptional repressors in the SL signaling remain to be identified.

Here, we report the characterization of tae-miR156 as a critical regulator of bread wheat plant architecture. We first showed that overexpression of miR156 in bread wheat plants triggers a bushy phenotype and severe defects in spikelet formation. Meanwhile, a group of putative SPL genes were identified to be targeted and repressed by tae-miR156. Next, we showed that the expression of two genes, TaBA1 and TaTB1, was markedly reduced in the miR156-overexpression (miR156-OE) wheat plants. Furthermore, our assays revealed that the SL signaling repressor TaD53 functions as a transcriptional repressor through interaction with the transcriptional corepressor TOPLESS (TaTPL). Significantly, we found that TaD53 could physically interact with the miR156-controlled SPL proteins TaSPL3/17 and largely repress TaSPL3/17-mediated transcriptional activation of TaBA1 and TaTB1 expression. Our findings provide new insight into the miR156-TaSPLs module in regulating bread wheat plant architecture.

RESULTS

Identification of a Tandem tae-microRNA156 Gene in Bread Wheat

Previous analysis on the whole-genome shotgun draft sequence of the bread wheat A-genome progenitor Triticum urartu uncovered a large group of scaffolds that contain putative microRNA precursors in T. urartu genome (Ling et al., 2013), which makes it easier to identify miRNA genes in bread wheat plants underlying important agronomic traits. Among these scaffolds, we noticed that the scaffold14333 potentially contains a tandem tae-microRNA156 (tae-miR156) gene, in which three MIR156 precursors (annotated as MIR156a, b, and c in this study) were aligned in tandem (Supplemental Fig. S1). To identify the homologous sequences of this tandem tae-miR156 gene in hexaploid bread wheat, we used scaffold14333 as query to search the wheat survey sequences, which include the chromosome-based draft sequence of the hexaploid bread wheat (https://urgi.versailles.inra.fr/blast/; Deng et al., 2007; International Wheat Genome Sequencing Consortium, 2014). Finally, three highly conserved sequences separately located on chromosomes 3A, 3B, and 3D were obtained (Fig. 1A). Similar to that in T. urartu genome, the three sequences include potential MIR156a/b/c precursors in tandem (separately marked by red, blue, and orange boxes in Fig. 1A), indicating a high-level conservation of this tandem miR156 gene in bread wheat during the evolutionary process from its progenitor species. We further validated the tandem miR156 gene by predicting the RNA secondary structures of these precursors. Certainly, each of these sequences can give rise to three tandem stable stem-loop structures with extremely low free energies (described as ΔG in Fig. 1B and Supplemental Fig. S2), which are reminiscent of miRNA precursors. Importantly, all of the mature tae-miR156 sequences are located on the stem regions of the stem-loop structures (highlighted by red in Fig. 1B and Supplemental Fig. S2) and are completely identical to the known miR156 sequences in barley (Hordeum vulgare), purple false brome (Brachypodium distachyon), maize, rice, and Arabidopsis annotated by miRBase (http://www.mirbase.org/; Fig. 1C). To further test whether the identified tandem MIR156 precursors could generate the mature tae-miR156 in vivo, we generated transgenic bread wheat lines using the bread wheat cultivar Kenong199 (KN199) as the background, in which a 1-kb nucleotide sequence from chromosome A of KN199 containing the tandem MIR156 cluster was overexpressed by using the ubiquitin promoter (Supplemental Fig. S3). We next checked the accumulation of mature tae-miR156 in transgenic plants using the stem-loop quantitative real-time PCR (qRT-PCR) and observed 5- to 12-fold increase of tae-miR156 compared with the wild-type KN199 plants (Fig. 1D). Meanwhile, as a negative control, we also determined the level of another bread wheat endogenous miRNA, miR319, by the stem-loop qRT-PCR. Our results confirmed that miR319 was similarly expressed in wild-type and transgenic bread wheat plants (Fig. 1D), further confirming the specific up-regulation of tae-miR156. In summary, these data strongly support the conclusion that the identified tandem MIR156 cluster encodes functional tae-miR156 gene in bread wheat.

Figure 1.

Identification of a tandem microRNA156 gene in bread wheat. A, The nucleotide sequences from chromosomes 3A, 3B, and 3D of the bread wheat cultivar KN199 encoding tandem miR156 genes. The three MIR156 precursors (MIR156a, b, and c) aligned in tandem are separately marked by red, blue, and orange boxes. The black shade boxes represent the positions of mature miR156. B, Predicted secondary structure of tandem tae-MIR156a/b/c precursors from chromosome 3A. ΔG (kcal mol⁻¹) calculated by Mfold represents the minimum free energy of the RNA secondary structure. nt, Nucleotides. C, Nucleotide sequences of miR156 from different plant species. tae, Triticum aestivum; hvu, Hordeum vulgare; bdi, Brachypodium distachyon; zma, Zea mays; osa, Oryza sativa; ath, Arabidopsis thaliana. D, Stem-loop qRT-PCR quantification of tae-miR156 and tea-miR319 in wild-type bread wheat KN199 as well as in transgenic bread wheat plants overexpressing the tandem tae-MIR156-3A cluster driven by the ubiquitin promoter. Three independent transgenic lines (1#, 2#, and 3#) were analyzed. The tae-miR156 and tae-miR319 levels were all normalized against TaU6, and the mean values in wild-type KN199 were set to 1. Error bar represents sd among three independent biological replicates; asterisks above the bars denote significant differences compared with wild-type KN199 plants at P < 0.01 (Student’s t test).

tae-miR156 Controls Bread Wheat Plant Architecture

To our knowledge, the roles of miR156 in controlling bread wheat plant development have not been reported up to now. By sowing the tae-miR156-OE transgenic bread wheat lines under field conditions (Beijing) and investigating their vegetative and reproductive development phenotypes, we characterized the biological roles of tae-miR156 in regulation of bread wheat plant architecture in detail in this study. The most conspicuous aspect of the phenotypes of tae-miR156-OE plants is the enhanced tillering. When the bread wheat plants were at the vernalization stage, we already observed more visible tiller buds in tae-miR156-OE seedlings compared with wild-type KN199 (Fig. 2A), suggesting that overexpression of tae-miR156 led to more easier and earlier production of lateral branching. After turning green in spring, the tae-miR156-OE transgenic lines generated more tillers than those in wild-type KN199 and showed bushy phenotype at all their developmental stages (the jointing stage as shown in Fig. 2B and the heading and mature stages as shown in Fig. 2C). Statistical analyses showed that overexpression of tae-miR156 caused a dramatic increase of tillers from ∼10 in the wild type to more than 40 in transgenic lines (Fig. 2E). Moreover, we observed that a certain number of axillary buds in leaf axil in the tae-miR156-OE transgenic plants could outgrow and thus produce high-node tillers that rarely occur in wild-type plants (Supplemental Fig. S4). These observations indicate that tae-miR156 plays a positive role in promoting the outgrowth of axillary buds. In addition, the stem of tae-miR156-OE plants became thinner than that in wild-type KN199 (Supplemental Fig. S4).

Figure 2.

Morphological characters of tae-miR156-OE bread wheat plants. A, Appearance of the visible tiller buds in wild-type KN199 and tae-miR156 overexpression (tae-miR156-OE) bread wheat seedlings at the vernalization stage (in the middle of November, Beijing). The red arrows point to the positions of tiller buds. Column diagram shows the statistical analyses of the numbers of visible tiller buds of wild-type KN199 and three independent transgenic lines (n ≥ 5). B, Vegetative phenotypes of tae-miR156-OE transgenic bread wheat plants at jointing stage grown in the field. The whole seedling (top) and the stem bases (bottom) from three independent transgenic bread wheat lines as well as wild-type KN199 were observed in early April. Bars = 10 cm. C, The bushy phenotype of tae-miR156-OE bread wheat plants at heading (top, in late April) and mature (bottom, in late June) stages. Bars = 20 cm. D, Spike phenotypes of wild-type KN199 and tae-miR156-OE bread wheat plants at the heading (top left), grain filling (top right, in early May), or mature (middle and bottom) stages. Bars = 5 cm. E, Statistical analyses of the plant architecture characters of wild-type KN199 and tae-miR156-OE bread wheat plants. Three independent transgenic lines were analyzed at the mature stage, and error bar represents sd (n ≥ 5). Asterisks above the bars in A and E denote significant differences compared with wild-type KN199 plants at P < 0.01 (Student’s t test).

In addition to the enhanced tillering phenotype, overexpression of tae-miR156 also led to severe defects in spike morphology. A typical spike from wild-type KN199 includes a central long rachis (8–10 cm in length) on which several tightly packed rows of spikelets (including 20–30 spikelets) are located (Fig. 2, D and E). However, in the tae-miR156-OE lines, an abnormally small spike was observed containing the extremely shortened rachis (1–2 cm in length) and a single spikelet located on the top (Fig. 2, D and E; Supplemental Fig. S4). By comparing the spikelets from the tae-miR156-OE plants and wild-type KN199, we noticed that both spikelets equally contained at least three florets, and each floret had a lemma and a palea (Supplemental Fig. S4). However, the main difference is that unlike a pair of glumes in wild-type spikelets, only one glume was observed in the tae-miR156-OE spikelets (Supplemental Fig. S4). The reduced spikelet number and defective spikelet morphology of tae-miR156-OE plants suggest that tae-miR156 is a key regulator of bread wheat domestication-related traits on spike architecture.

tae-miR156 Targets a Group of Putative TaSPL Genes

The plant microRNAs regulate plant growth and development mainly through the down-regulation of their corresponding target genes. To further elucidate the molecular bases of the defects in plant architecture caused by overexpression of tae-miR156, we focused on the analyses of tae-miR156 targeting genes. We predicted the potential tae-miR156 target genes based on the whole wheat cDNA library (T. aestivum, cDNA, Ensemblplants, release-31) using the Web-based plant small RNA target analysis tool psRNATarget (Dai and Zhao, 2011). Meanwhile, 10 previously cloned and annotated bread wheat SPL genes were also included for the prediction (Zhang et al., 2014). Finally, 16 potential miR156-targeting genes were obtained with the maximum expectation as 2.0 (Supplemental Table S1). Consistent with previous studies (Jones-Rhoades and Bartel, 2004; Wang et al., 2009; Wu et al., 2009), all the predicted tae-miR156 target genes encode putative plant specific SPL transcription factors (Supplemental Table S1). Among these targets, T8 and T9 had already been annotated as TaSPL3 and TaSPL17, respectively, in bread wheat in a previous study (Zhang et al., 2014).

If these TaSPL genes are indeed the tae-miR156 target genes, whose transcripts can be cleaved by tae-miR156, their transcripts should have lower levels when tae-miR156 is overexpressed. In support of our hypothesis, qRT-PCR assays showed that all the tested miR156-targeting genes (T1 to T9 in Fig. 3A) were significantly downregulated in three independent tae-miR156-OE transgenic bread wheat lines, compared with those in wild-type KN199 (Fig. 3A), suggesting the repression of these putative TaSPL genes by the overexpression of tae-miR156. To further confirm the specificity of tae-miR156 in regulation of these putative TaSPL genes, we simultaneously examined the expression of two other TaSPL genes, TaSPL20 and TaSPL21, which were predicted to be not regulated by tae-miR156 due to their low-level complementarity with tae-miR156 (Supplemental Fig. S5). Indeed, our results showed that the transcription levels of TaSPL20 and TaSPL21 were similarly expressed in both wild-type KN199 and tae-miR156-OE transgenic bread wheat lines (Supplemental Fig. S5), also supporting our prediction that TaSPL20 and TaSPL21 are not the targets of tae-miR156.

Figure 3.

tae-miR156 targets a group of putative TaSPL genes in bread wheat. A, Determination of the transcript levels of putative tae-miR156 targets in tae-miR156-OE bread wheat lines. The transcript levels of indicated genes were first normalized against TaGAPDH, and the mean values in wild-type KN199 were set to 1. Error bars represent sd among three independent biological replicates; asterisks above the bars denote significant differences compared with wild-type KN199 plants at **P < 0.01 or *P < 0.05 (Student’s t test). B, Mapping of cleavage sites of tae-miR156 by 5′RACE. Nucleotide sequences of potential tae-miR156 target sites on TaSPL3 and TaSPL17, together with the Watson-Crick base pairings to the tae-miR156 are shown. The free energies of duplex structures represented by ΔG (kcal mol⁻¹) are calculated by Mfold. The red arrows indicate the cleavage sites, and numbers below the arrows show the frequency of clones with matching 5′RACE product from this site out of total clones confirmed by sequencing.

Next, we validated the tae-miR156-mediated cleavage on its targeting TaSPL genes in vivo through a previously described RACE procedure (Liu et al., 2014). Here we employed T8_6DS/TaSPL3 and T9_7DS/ TaSPL17 for the RACE assays. Our results showed that tae-miR156 could predominantly cleave the TaSPL3 and TaSPL17 mRNAs at the position 10 to 11 of tae-miR156 from its 5′ end (Fig. 3B), which has been considered as a canonical miRNA cleavage site (Peters and Meister, 2007; Höck and Meister, 2008); as negative controls, no cleavage product of TaSPL20 and TaSPL21 mRNAs was detected.

It has been proved in Arabidopsis that miR156 levels are higher in young seedlings and decrease in an age-dependent manner; meanwhile, the expression patterns of miR156-targeting SPLs are inverse to that of miR156 (Wang et al., 2009; Wu et al., 2009). Therefore, we wondered whether tae-miR156 and its targets also display similar temporal expression patterns in bread wheat plants. As expected, our results confirmed that tae-miR156 was highly accumulated in bread wheat juvenile leaves that were collected from KN199 seedlings at trefoil stage, but significantly decreased in flag leaves at heading stage (Supplemental Fig. S6), indicating a decrease tendency of tae-miR156 accompanying with the transition from juvenile to adult phases. The lowest level of tae-miR156 was detected in KN199 young spikes at heading stage when compared with the leaf tissues (Supplemental Fig. S6). On the contrary, the transcription levels of tae-miR156-regulated TaSPL3 and TaSPL17 displayed an increase tendency. qRT-PCR assays showed that the expression levels of TaSPL3 and TaSPL17 were much lower in juvenile leaves and increased in flag leaves (Supplemental Fig. S6). Strikingly, their maximum transcript levels were observed in young spikes at heading stage (Supplemental Fig. S6). The inverse expression patterns of tae-miR156 and TaSPL3/17 further support our conclusion that TaSPL3/17 are targeted and cleaved by tae-miR156 in vivo.

In summary, we define a group of TaSPL genes as targets of tae-miR156 in bread wheat, whose expression can be repressed by tae-miR156.

The tae-miR156-Targeted TaSPLs Positively Regulate TaTB1 Expression

To further explore the underling mechanism of the enhanced tillering in tae-miR156-OE transgenic plants, we focused on the TB1 gene. Previous studies have revealed that TB1 genes in maize and rice act as negative regulators of lateral branching (tillering) and repress the outgrowth of axillary buds (Doebley et al., 1997; Wang et al., 1999; Takeda et al., 2003). The enhanced tillering phenotype of the tae-miR156-OE transgenic bread wheat lines prompted us to determine whether the TB1 ortholog in bread wheat is transcriptionally affected. First, we used the coding sequence of ZmTB1 (GenBank accession no. JQ900502) as query to search the wheat survey sequences using BLAST and obtained three nucleotide sequences with high similarity. Annotation and BLAST analyses revealed that the three homologous coding sequences are separately located on chromosomes 4A, 4B, and 4D (Supplemental Fig. S7) and encode proteins sharing ∼56% identities with OsTB1 (LOC_Os03g49880) and ∼54% with ZmTB1 (Supplemental Fig. S8), suggesting that they are most likely the candidates of TaTB1-4A, -4B, and -4D in bread wheat. Then, we designed specific primers to determine the transcription of TaTB1 by qRT-PCR. Considering that TB1 gene is involved in lateral branching control, we used the stem bases from bread wheat seedlings grown at the vernalization stage for our analyses. Our results showed that in all detected tae-miR156-OE lines, TaTB1 was dramatically downregulated (Fig. 4A).

Figure 4.

TaSPL3 positively regulates the expression of TaTB1 in bread wheat. A, Quantification of the TaTB1 transcript levels in stem base tissues of KN199 and tae-miR156-OE bread wheat seedlings by qRT-PCR. The stem base of the bread wheat seedling were collected at the vernalization stage (in the middle of November, Beijing), and the transcript levels of TaTB1 were quantified by normalizing against TaGAPDH. Mean values of TaTB1 in wild-type KN199 were set to 1. **P < 0.01 (Student’s t test). B, Schemes illustrating the 2-kb promoters of TaTB1-4A, 4B, and 4D. The black boxes display the positions of SBP-box binding core motifs. C, LUC activity assays showing the promoter-driving activities of TaTB1-4A_pro/4B_pro/4D_pro. The 2-kb promoter sequences of TaTB1-4A_pro, -4B_pro, and -4D_pro were used to drive the LUC gene expression, and the LUC activities were determined by A. tumefaciens-mediated infiltration in N. benthamiana 48 h postinfiltration (hpi). Left, A representative leaf image; right, quantification of the relative luminescence intensities (n = 10); the colored scale bar indicates the luminescence intensity (CPS). The mean values in EV were set to 1, and asterisks above the bars represent significant differences against EV control at P < 0.01 (Student’s t test). D, Transient expression assays illustrating the activation of TaTB1 transcription by TaSPL3. TaTB1-4A_pro and -4D_pro were employed for the analyses. Left panel shows representative leaf images, and the right column diagram represents the quantification of the relative luminescence intensities (n = 15). The mean values in combinations 1 and 3 were all set to 1. Asterisks above the bars represent significant differences against combinations 1 or 3 at P < 0.01 (Student’s t test). Error bars in A, C, and D represent sd among three independent biological replicates.

The correlation between the compromised transcription of TaTB1 and the down-regulation of TaSPL3 and TaSPL17 followed by the overexpression of tae-miR156 in transgenic bread wheat plants led us to ask whether TaTB1 is a potential downstream gene regulated by TaSPLs. To confirm this hypothesis, we first cloned the 2-kb promoter sequences of TaTB1-4A, -4B, and -4D from the KN199 genome using specific primers. As expected, we identified a number of SBP-box binding motifs (including CGTACAA, GTACAA, CGTACA, GTACA, and GTAC) on each of the three identified TaTB1 promoters (Fig. 4B; Supplemental Fig. S9), in support of our hypothesis that TaSPLs might regulate TaTB1 genes through binding to their promoter regions. To experimentally confirm the regulatory effect of the miR156-regulated TaSPLs on TaTB1 genes, we carried out transient transcriptional activity assays by Agrobacterium tumefaciens-mediated transient infiltration in Nicotiana benthamiana leaves. First, we generated the reporters by fusing the 2-kb identified TaTB1 promoters (TaTB1-4A_pro, -4B_pro, and -4D_pro) with firefly luciferase (LUC) gene to produce constructs TaTB1-4A_pro:LUC, -4B_pro:LUC, and -4D_pro:LUC. To check the basal activities of these reporters, the constructs of TaTB1-4A_pro:LUC, -4B_pro:LUC, and -4D_pro:LUC were separately infiltrated into N. benthamiana leaves, followed by the determination of LUC activity. Interestingly, LUC activity was exclusively observed in the TaTB1-4A_pro:LUC infiltrated samples (with luminescence intensity ∼6,000 counts per second [CPS], as shown in Fig. 4C, infiltration 2), whereas no obvious LUC signal was detected either in TaTB1-4B_pro:LUC or -4D_pro:LUC expressed samples (the luminescence intensities were <2,000 CPS; Fig. 4C, infiltrations 3 and 4), compared with the empty vector (EV) control. We attributed these variations to divergent driving activities among the three types of TaTB1 promoters. Subsequently, we selected TaTB1-4A_pro:LUC (with high driving activity) and TaTB1-4D_pro:LUC (with low driving activity) constructs for further activation analyses. Our results confidently confirmed that coexpression of 35S:TaSPL3 with TaTB1-4A_pro:LUC and TaTB1-4D_pro:LUC all led to almost 30- to 40-fold increases of the LUC reporter activity compared with that in the EV controls (with luminescence intensities higher than 30,000 CPS; Fig. 4D), suggesting that TaSPL3 could dramatically elevate the driving activities of TaTB1-4A_pro and TaTB1-4D_pro. Parallel experiments were performed using TaSPL17, and similar results were observed (Supplemental Fig. S10), indicating that both TaSPL3 and TaSPL17 are activators of TaTB1 expression.

In summary, the above data imply that TaTB1 is probably involved in the miR156-TaSPLs module-mediated signaling through transcriptional activation by TaSPL3 and TaSPL17 transcription factors.

The tae-miR156-Targeted TaSPLs Are Essential for the Transcription of TaBA1

Numerous genes have been reported to be essentially involved in the inflorescence/panicle/spike development in diverse plant species (Pelaz et al., 2000; Honma and Goto, 2001; Chuck et al., 2002; Ritter et al., 2002; Gallavotti et al., 2004, 2010; Bortiri et al., 2006; Shitsukawa et al., 2006; Paolacci et al., 2007; Derbyshire and Byrne, 2013; Liu et al., 2013; Ren et al., 2013; Lin et al., 2014; Dobrovolskaya et al., 2015; Guedira et al., 2016). To dissect the molecular basis of defective spike architecture in the tae-miR156-OE lines, we made efforts to determine whether the expressions of these genes are altered in our transgenic lines, especially in young spikes before the heading stage. According to our results, the expression of some inflorescence/panicle/spike development-related genes, such as TaVERNALIZATION1 (GenBank accession no. AY747604), TaWFL (GenBank accession no. AB231888), TaSEPALLATA1 (GenBank accession no. AM502866), and TaSEP2 (GenBank accession no. AM502867), were not affected by the overexpression of tae-miR156, indicating that these genes might not participate in the miR156-TaSPL-regulated signaling pathway in bread wheat (Supplemental Fig. S11). Interestingly, TaBA1 was markedly downregulated by tae-miR156 overexpression in bread wheat (Fig. 5A). The ZmBA1 gene, an ortholog of TaBA1, was initially isolated from the maize recessive mutant ba1, in which the tassel (apical male inflorescence in maize) is seriously defective in branching and spikelet formation (Ritter et al., 2002; Gallavotti et al., 2004), suggesting that ZmBA1 is a positive regulator of spikelet formation in maize. Three copies of intronless TaBA1 genes (Supplemental Fig. S12), which are separately located on 3A, 3B, and 3D chromosomes in hexaploid bread wheat, all encode transcription factors with a conserved bHLH domain and share ∼58% identity with ZmBA1 protein (GenBank accession no. NP_001105271; Fig. 5B). Significantly, the bHLH domain in TaBA1, which is involved in DNA binding activity, is 100% identical to that in ZmBA1 (Fig. 5B), implying that BA1 proteins in maize and bread wheat might be involved in similar signaling pathways in regulating spike architecture.

Figure 5.

TaSPL3 positively regulates the expression of TaBA1 in bread wheat. A, Determination of the TaBA1 transcript levels by qRT-PCR in KN199 and tae-miR156-OE bread wheat plants. The expression levels of TaBA1 were normalized against TaGAPDH, and the mean values of TaBA1 in wild-type KN199 were set to 1. **P < 0.01 (Student’s t test). B, Sequence comparison of TaBA1-3A/3B/3D and ZmBA1 proteins. The bHLH domains are indicated by red lines. The black shade represents the identity, and gray means the similarity. C, Schemes represent the promoters of TaBA1-3A, -3B, and -3D. The black boxes display the positions of SBP-box binding core motifs. D, LUC activity assays illustrating the promoter-driving activities of TaBA1-3A_pro/3B_pro/3D_pro. The 2-kb promoter sequences of TaBA1-3A_pro, -3B_pro, and -3D_pro were constructed to drive the LUC gene, and the LUC activities were determined by A. tumefaciens-mediated infiltration in N. benthamiana. Left panel, a representative leaf image; right panel, the quantification of the relative luminescence intensities by using n = 10 independent leaves; the scale bar illustrates the luminescence intensity of CPS by color. The mean values in EV were set to 1, and asterisks above the bars represent significant differences against EV control at P < 0.01 (Student’s t test). E, Transient expression assays illustrating the activation of TaBA1 transcription by TaSPL3. TaBA1-3B_pro and -3D_pro were employed for the assays. Left panel shows representative leaf images, and the right column diagram represents the quantification of the relative luminescence intensities using n = 15 independent leaves. The mean values in combinations 1 and 3 were all set to 1. Asterisks above the bars represent significant differences against combinations 1 or 3 at P < 0.01 (Student’s t test). Error bars in A, D, and E represent sd among three independent biological replicates.

Based on the finding that the transcription of TaBA1 was drastically compromised, we tested whether TaBA1 is a potential target gene controlled by the tae-miR156-regulated TaSPLs. First, 2-kb promoter sequences corresponding to the TaBA1-3A, -3B, and -3D genes were identified from bread wheat KN199 based on the information of wheat survey sequences. Analyses further revealed that several SBP-binding motifs (such as GTACA and GTAC) were included in TaBA1-3A_pro and TaBA1-3D_pro, and one (GTACAA) in TaBA1-3B_pro (Fig. 5C; Supplemental Fig. S13). The transient expression assays in N. benthamiana were carried out by using the promoters of TaBA1-3A, -3B, and -3D fused with the LUC gene as reporters (TaBA1-3A_pro:LUC, -3B_pro:LUC, and -3D_pro:LUC) to identify the driving activities of TaBA1 promoters. Results showed that only TaBA1-3D_pro could constitutively drive the expression of LUC gene (with luminescence intensity ∼5,000 CPS), compared with the EV negative control, while no obvious LUC signal was observed in TaBA1-3A_pro:LUC or -3B_pro:LUC expression samples (with luminescence intensity <1,000 CPS; Fig. 5D). These results suggest that TaBA1-3D_pro has the highest driving activity among the three types of TaBA1 promoters. We further carried out transient transcriptional activation assays in N. benthamiana using TaBA1-3B_pro:LUC (with low driving activity) and TaBA1-3D_pro:LUC (with high driving activity) as the reporters to determine the effects of TaSPL3 on the transcription of TaBA1 genes. In contrast to the low levels of LUC activity in the TaBA1-3B_pro:LUC or TaBA1-3D_pro:LUC expressed samples (Fig. 5E, coinfiltrations 1 and 3), dramatically enhanced luminescence intensities were observed in all samples coexpressing 35S:TaSPL3 (with luminescence intensities higher than 20,000 CPS; Fig. 5E, coinfiltrations 2 and 4), indicating that TaSPL3 could efficiently activate the transcription of TaBA1 genes. This conclusion was further confirmed by using TaSPL17 as the activator (Supplemental Fig. S14).

Taken together, the above data lead to the conclusion that the tae-miR156-targeted TaSPLs, including TaSPL3 and TaSPL17, are potential activators of TaBA1 expression.

TaD53 Directly Interacts with TaSPL3 and TaSPL17

The enhanced tillering of tae-miR156-OE bread wheat lines is reminiscent of the phenotypes of some rice tillering dwarf mutants (d mutants) defective in SL biosynthesis and signaling (Ishikawa et al., 2005; Zou et al., 2006; Arite et al., 2007, 2009; Lin et al., 2009; Jiang et al., 2013; Nakamura et al., 2013). Among them, the d53 mutant, which shows increased tillering and less branched panicle, was well characterized (Jiang et al., 2013; Zhou et al., 2013). Previous studies revealed that D53 shares predicted features with the class I Clp ATPase proteins and functions as a repressor of SL signaling in rice (Jiang et al., 2013; Zhou et al., 2013). Based on the plant architecture phenotypes of rice d53 mutant, we supposed a hypothesis that the bread wheat D53 homologs TaD53 might be associated with the miR156-TaSPLs signaling module in controlling bread wheat plant architecture. To test this hypothesis, we first identified the bread wheat TaD53 genes (Supplemental Fig. S15) based on the coding sequence of OsD53 (LOC_Os11g01330). BLAST analyses revealed that three copies of TaD53 encode proteins sharing 70% identity with OsD53 and 42% with the D53-like protein AtSMXL7 in Arabidopsis (Supplemental Fig. S16). We further asked whether TaD53 physically interacts with the tae-miR156-regulated TaSPL transcription factors. To this end, we fused TaD53 to the N-terminal part of LUC (nLUC) to generate nLUC-TaD53, and TaSPL3 and TaSPL17 to the C-terminal part of LUC (cLUC) to produce cLUC-TaSPL3 and cLUC-TaSPL17. We than performed LUC complementation imaging (LCI) assays in N. benthamiana leaves. Interestingly, obvious LUC activities were observed both in nLUC-TaD53/cLUC-TaSPL3 and nLUC-TaD53/cLUC-TaSPL17 coexpression samples (Fig. 6A, coinfiltrations 4 and 6), demonstrating the physical association of TaD53 with TaSPL3 and TaSPL17. The direct interaction between TaD53 and TaSPL3/TaSPL17 was further confirmed by bimolecular fluorescence complementation (BiFC) assays in N. benthamiana, by which the fluorescence signal of yellow fluorescent protein (YFP) could be observed exclusively in the nuclei of nYFP-TaSPL3/cYFP-TaD53 and nYFP-TaSPL17/cYFP-TaD53 coexpression samples, but not in the negative controls (Fig. 6B). Therefore, we proposed that the SL signaling repressor TaD53 might be functionally associated with the tae-miR156-regulated TaSPLs transcription factors, such as TaSPL3 and TaSPL17.

Figure 6.

The miR156-regulated TaSPL3 and TaSPL17 physically interact with TaD53. A, Firefly LCI assay detecting the interaction between TaSPL3/17 and TaD53. The signals were collected 48 hpi. B, BiFC assays confirming the interaction between TaSPL3/17 and TaD53. The YFP fluorescence signals were collected 48 hpi. BF, Bright field. Bars = 50 µm. In A and B, five independent N. benthamiana leaves were used for the assays and three biological replications were performed with similar results.

The N Terminus of TaSPL3 Mediates the Interaction with TaD53, While Its C Terminus Has the Transcriptional Activation Activity

To further define the interaction domains of TaSPL3/17 with TaD53, the full-length TaSPL3 and TaSPL17 proteins were divided into three truncated parts according to the positions of the highly conserved SBP domains, i.e. NT (N terminus), MD (middle domain, containing the intact SBP domain), and CT (C terminus; Fig. 7A; Supplemental Fig. S17). The LCI results showed that the strongest interaction signals were observed between TaD53 and NT parts of TaSPL3 and TaSPL17 (Fig. 7B; Supplemental Fig. S17), suggesting that the N-terminal domains of TaSPL3/17 mainly mediate the physical interaction with TaD53.

Figure 7.

The N-terminal domain of TaSPL3 is sufficient for the interaction with TaD53 and its C-terminal domain has transcriptional activation activity. A, Schemes display full length as well as truncated versions of TaSPL3 protein. NT, N-terminal domain; MD, middle domain; CT, C-terminal domain; aa, amino acids. B, LCI assay showing the interaction between the truncated TaSPL3 versions and the full-length TaD53. The LUC signals were observed at 48 hpi (n ≥ 5). Three biological replications were performed, and similar results were observed. C, Transcriptional activation activity determination of TaSPL3 in yeast AH109 (S. cerevisiae). SD-W, Synthetic dextrose medium lacking Trp; SD-W/H, synthetic dextrose medium lacking Trp and His; 10⁻¹, 10⁻², 10⁻³, and 10⁻⁴ denote the different dilution series.

As is well known, the SBP domains of SPL transcription factors are responsible for the DNA binding to their target genes. We here performed transcriptional activity assays in AH109 yeast (Saccharomyces cerevisiae) cells for the full length as well as truncated forms of TaSPL3 protein to define its transcriptional regulation domain. We separately fused the full-length as well as the NT, MD, and CT parts of TaSPL3 with pGBKT7 (BD) vector and transformed into yeast strain AH109. The transformants were selected on synthetic dextrose medium lacking Trp and His (SD-W/H) to check their growth status. As expected, the yeast cells harboring the full-length TaSPL3 grew well on the selective medium, suggesting that TaSPL3 could efficiently activate the transcription of the report gene (Fig. 7C); more importantly, the yeast cells containing TaSPL3-CT showed similar growth status with those containing the full-length TaSPL3 on the selective medium (Fig. 7C), indicating that the C-terminal part of TaSPL3 is responsible for its transcriptional activation activity.

In summary, our analyses suggest that the three different regions of TaSPL3 protein have distinct functions, i.e. MD (the SBP domain) for the DNA binding ability, CT for the transcriptional activation activity, and NT for the physical interaction with the transcriptional repressor TaD53.

Interaction of TaD53 with TaTPL

A previous study has shown that OsD53 could physically associate with the TPL/TOPLESS-RELATED PROTEIN (TPR) class of transcriptional corepressors through ethylene-responsive element binding factor-associated amphiphilic repression (EAR) motifs (Jiang et al., 2013). Like OsD53, we found that TaD53 also contains three putative EAR motifs: ⁵⁷¹Leu-Val-Leu-Asn-Leu-575 (EAR1), ⁷⁹²Leu-Asp-Leu-Ser-Leu-796 (EAR2), and ⁹⁷⁰Phe-Asp-Leu-Asn-Leu-974 (EAR3; Supplemental Fig. S16). Thus, we speculated that TaD53 might also be associated with the TPL orthologs in bread wheat. In this context, we identified a TaTPL gene (Supplemental Fig. S18) from bread wheat based on the sequence of OsTPR1 (LOC_Os01g15020). Analysis revealed that the TaTPL protein candidate shares high level identities with OsTPR1 (62% identity), OsTPR2 (LOC_Os08g06480; 67% identity), OsTPR3 (LOC_Os03g14980; 94% identity), and AtTPL (At1g15750; 77% identity) (Supplemental Fig. S19). Next, we conducted LCI assays in N. benthamiana and confirmed that TaTPL indeed could interact with TaD53 (Fig. 8A). The interaction was further supported by the BiFC assays, in which an obvious interaction signal was observed exclusively in the cell nucleus of nYFP-TaTPL/cYFP-TaD53 coexpression sample (Fig. 8B). These results confirm the conservation of D53-TPL regulatory module in both rice and bread wheat.

Figure 8.

Physical interaction of TaD53 with TaTPL. A, LCI assay showing the interaction of TaD53 with TaTPL. B, BiFC confirming the physical interaction between TaD53 and TaTPL. The LUC and YFP fluorescence signals were all collected 48 hpi. BF, Bright field. Bars = 50 µm. In A and B, five independent N. benthamiana leaves were used for the assays and three biological replications were performed with similar results.

TaD53 Represses the Transcriptional Activation Activity of TaSPL3 on TaBA1 and TaTB1 Transcription

To evaluate the effects of TaD53 on the transcriptional activities of TaSPLs, we coexpressed 35S:TaD53 and 35S:TaSPL3 together with the reporter TaTB1-4A_pro:LUC in a 1:1:1 ratio in N. benthamiana (Fig. 9A, coinfiltration 4). Meanwhile, the coexpression of 35S:TaSPL3/35S:GFP, 35S:TaD53/35S:GFP with the LUC reporter in 1:1:1 ratio (Fig. 9A, coinfiltrations 2 and 3) and the expression of 35S:GFP with the reporter in a 2:1 ratio (Fig. 9A, coinfiltration 1) were simultaneously carried out as controls. Similar to the above results, coexpression of 35S:TaSPL3 dramatically elevated the luminescence intensity (Fig. 9A, coinfiltration 2), compared with the 35S:GFP expressed negative control (Fig. 9A, coinfiltration 1). However, after the coexpression of 35S:TaD53 with 35S:TaSPL3, the induction of LUC activity by TaSPL3 was dramatically compromised by more than 70% (Fig. 9A, coinfiltration 4), suggesting that the transcriptional activation activity of TaSPL3 was largely suppressed by TaD53 in regard to the activation of TaTB1 transcription. Parallel experiments using TaBA1-3D_pro:LUC as the reporter were carried out. As expected, the transcriptional activation activity of TaSPL3 in enhancing TaBA1 transcription was largely blocked by TaD53 (Fig. 9B, coinfiltrations 5–8). These results suggest a repressive effect of TaD53 on the activation activities of tae-miR156-targeted TaSPLs in regulating the transcription of their downstream target genes.

Figure 9.

TaD53 represses the transcriptional activation activity of TaSPL3 for TaBA1 and TaTB1 transcription. Transient expression assays separately illustrating the repressive effect of TaD53 on TaSPL3-mediated activation of TaBA1 (A) and TaTB1 (B) expression. The coinfiltrations of 35S:GFP, 35S:TaSPL3-GFP/35S:GFP, 35S:TaD53-GFP/35S:GFP, and 35S:TaD53-GFP/35S:TaSPL3-GFP with the reporters TaTB1-4A_pro:LUC or TaBA1-3D_pro:LUC in 2:1 or 1:1:1 ratios were carried out as indicated. Left panels show representative leaf images, and the right column diagrams represent quantification of the relative luminescence intensities (n = 15). The values in combinations 3 and 7 were all set to 1. Asterisks above the bars represent significant differences against combinations 3 or 7 at P < 0.01 (Student’s t test). Error bar represents sd among three independent biological replicates.

DISCUSSION

miR156 Controls Bread Wheat Plant Architecture

Previous study on the maize Corngrass1 (Cg1) mutant proved that overexpression of miR156 in maize led to repeated initiation of tillers and transformed maize into bush (Chuck et al., 2007). Overexpression of miR156 in rice also led to enhanced tillering phenotype (Luo et al., 2012; Xie et al., 2012). Similarly, in the model plant Arabidopsis, a bushy phenotype was observed upon the overaccumulation of miR156 (Schwab et al., 2005). These studies, coupled with the bushy phenotype in our tae-miR156-OE transgenic bread wheat plants (Fig. 2), confidently illustrate that miR156 is functionally conserved in promoting tillering/branching in different plant species.

However, the potential molecular function of miR156 in inflorescence/panicle/spike development still needs to be elucidated. In this study, we identified a potential relationship between tae-miR156 and bread wheat spike morphology. Interestingly, unlike its positive effect on tillering initiation, tae-miR156 was shown to act as a repressor of bread wheat spikelet development, according to the observation that overdose of tae-miR156 led to defective spikes with single or few spikelet(s) (Fig. 2; Supplemental Fig. S4). In contrast to less affected panicle morphology in rice miR156-OE transgenic plants (Xie et al., 2012), miR156-OE in bread wheat appears to trigger an extremely severe defect in spike shape, including dramatically shortened rachis (1–2 cm) and single spikelet (Fig. 2; Supplemental Fig. S4). Previous studies showed that miR156 is highly accumulated in young seedlings but subsequently decreases along with the vegetative-reproductive phase transitions (Xie et al., 2006; Chuck et al., 2007; Wang et al., 2008, 2009; Xie et al., 2012). A similar expression pattern was observed in this study that tae-miR156 is highly accumulated in juvenile but decreases in flag leaves and young spikes (Supplemental Fig. S6). However, how plants precisely control miR156 levels in different developmental stages remains unclear. Future works on this issue should be of benefit to answer why and how tae-miR156 is simultaneously involved in the regulation of bread wheat plant architecture through modulation of two distinct aspects, including the tiller number per plant and the spikelet number per spike.

Based on bioinformatics prediction and experimental confirmation, we further identified a group of bread wheat putative TaSPLs as potential tae-miR156 targeting genes (Fig. 3; Supplemental Table S1). Upon the overexpression of tae-miR156, the expression of these TaSPL genes was significantly repressed (Fig. 3), while the other TaSPLs, such as TaSPL20 and TaSPL21, that are not targeted by tae-miR156 were not affected (Supplemental Fig. S5). Previous studies in rice have revealed that IPA1, encoding the miR156-controlled SPL14 transcription factor, could profoundly determine rice ideal plant architecture, including decreased tillers but increased panicle branching (Jiao et al., 2010; Miura et al., 2010; Luo et al., 2012; Wang et al., 2017). More importantly, among our identified tae-miR156 targets, TaSPL17 is the orthologous gene of rice IPA1, which shares high identity with IPA1 at the protein level. These results promote us to assume that the phenotypes of tae-miR156-OE plants, including enhanced tillering and decreased spikelet number, might be partially due to the compromise of these tae-miR156-targeted TaSPL genes. Although our current data could not support the conclusion that all these miR156 targets are functional in regulation of bread wheat plant architecture, further detailed investigations, such as generation of the dominant and miR156-resistant mutations for specific tae-miR156 targeted TaSPL genes, may provide clues for understanding the potential roles of these TaSPLs in bread wheat plant architecture modulation.

The miR156-Controlled TaSPLs Regulate the Expression of TaTB1 and TaBA1 Genes

The TB1 homologous genes from diverse plant species were reported to play a conserved role in suppressing lateral branching, especially tillering (Doebley et al., 1997; Takeda et al., 2003; Aguilar-Martínez et al., 2007), while ZmBA1 acts as a pivotal regulator in modulating maize plant architecture, especially the spike morphology (Ritter et al., 2002; Gallavotti et al., 2004). In this study, we noticed that the expression of TaTB1 was significantly reduced in the stem bases of tae-miR156 overexpression bread wheat plants (Fig. 4A); similarly, TaBA1 gene expression was also largely compromised in young spikes of tae-miR156-OE bread wheat lines (Fig. 5A). These observations correlated with the extremely enhanced tillering and defective spike phenotype in tae-miR156-OE transgenic bread wheat plants, which led us to ask whether these genes are potentially involved in the miR156-TaSPL-mediated signaling pathway. Analyses of TaTB1 and TaBA1 promoter sequences identified a certain number of SBP-binding motifs (Supplemental Figs. S9 and S13), raising the possibility that TaSPL transcription factors might bind to the promoters of these genes to regulate their expression. In vitro assays indeed showed that both TaSPL3 and TaSPL17 could significantly activate the transcription of TaTB1 and TaBA1 (Figs. 4 and 5; Supplemental Figs S10 and S14). According to these analyses, we assume that TaTB1 and TaBA1 might be potentially located downstream of miR156-TaSPLs signaling, and their transcriptional expression should partially require the activation activities of miR156-controlled TaSPLs. In agreement with our findings, a functional link between the miR156-SPLs module and TB1 has been established by a previous study in rice showing that IPA1/OsSPL14 directly regulates the expression of OsTB1 to suppress rice tillering, thus placing the miR156-SPLs module upstream of OsTB1 (Lu et al., 2013). Together, we propose that the miR156-SPLs-TB1 signaling cascade might be highly conserved in diverse crop species. In the future, it would be intriguing to confirm the functions of TaTB1 and TaBA1 in suppressing tillering and promoting spike/panicle development in bread wheat and thus provide new potential genetic resources for crop molecular breeding.

Repression of TaSPL Activities by the SL Signaling Repressor TaD53

In recent years, several studies greatly advanced our understanding on SL signaling, especial the signal perception coupled with the degradation of SL signaling repressors D53 proteins. As a relatively new class of plant hormone, more components involved in SL signaling remain to be identified. For example, the downstream transcription factors regulated by the repressor D53 in SL signaling remain largely unknown up to now. In this study, we showed that TaD53 interacts with the transcription corepressor TaTPL (Fig. 8), which supports the conclusion that TaD53 functions as a transcriptional repressor. Significantly, we found that TaD53 directly interacts with tae-miR156-controlled TaSPL transcription factors, such as TaSPL3 and TaSPL17 (Fig. 7; Supplemental Fig. S17). Furthermore, our results showed that the transcriptional activity of TaSPL3 could be largely suppressed by TaD53 in regard to the activation of its downstream target genes TaTB1 and TaBA1 (Fig. 9). These data support the hypothesis that the tae-miR156-controlled TaSPLs might also be direct targets of TaD53 at the protein activity layer, which provides new insight into the SL signaling pathway in plants. However, it should be more complex in real situations. A previous study in rice showed that overexpression of the OsSPL14 gene could largely suppress the tillering phenotype of SL biosynthesis/signaling mutants d10-2 and d3-2 (Luo et al., 2012). Meanwhile, Luo et al. (2012) also showed that the treatment with GR24 (a synthetic SL analog) still could suppress the tiller outgrowth in miR156-OE transgenic rice plants. These results support a hypothesis that SPLs might act independently or downstream of SL signal in the control of tillering. In this study, we showed that the SL signaling repressor TaD53 could physically interact with the miR156-regulated TaSPLs and repress their activities (Figs. 6, 7, and 9), indicating that SL might also act, at least in part, through D53-mediated regulation of SPLs transcription factors to control tillering. However, we could not rule out the possibility that SL might also participate in the control of bread wheat plant architecture through other transcription factors besides the miR156-regulated SPLs, which may to some extent explain the observation that GR24 is still effective in miR156-OE transgenic rice lines in suppressing rice tillering (Luo et al., 2012). The findings in rice and bread wheat together suggest that SL might act through the SPL-dependent pathway via direct repression of the SPLs activities by D53 and the SPL-independent pathway via other types of transcription factors.

Based on our findings in this study, we propose a working model on the miR156-TaSPL regulatory module and TaD53 in controlling bread wheat plant architecture (Fig. 10). In this model, the miR156-controlled TaSPLs, such as TaSPL3 and TaSPL17, act as transcriptional activators to enhance the expression of some critical downstream genes including TaTB1 and TaBA1, which might potentially contribute to the bread wheat plant architecture with limited tillers and normal spike morphology; meanwhile, several types of repressors are involved in this signaling pathway, like tae-miR156 that may down-regulate TaSPL mRNAs at the posttranscriptional level, and the TaD53-TaTPL module that may directly interact with and suppress the transcriptional activation activities of TaSPLs. However, besides the tae-miR156/TaD53-controlled TaSPL signaling pathway, other TaSPL-independent pathways might also exist, in which the SL-controlled TaD53 might functionally antagonize the activities of some other unidentified transcription factors to regulate bread wheat tillering (Fig. 10). Finally, we might be inclined to assume that instead of complete antagonism, the complex mutual regulation is more likely the theme among these positive and negative regulators in vivo, aiming to fine tune the establishment of appropriate bread wheat plant architecture.

Figure 10.

A proposed working model of the miR156-TaSPLs module and TaD53 in controlling bread wheat plant architecture.

MATERIALS AND METHODS

Plant Materials and Growth Conditions

For gene transformation, bread wheat (Triticum aestivum) cultivar wild-type Kenong199 was selected as the receptor plant. The indicated tae-miR156 gene was ligated to the BamHI/KpnI double-digested pUbi:cas vector and transformed into the 1-month-old embryogenic calli of KN199 using a PDS1000/He particle bombardment system (Bio-Rad) with a target distance of 6.0 cm from the stopping plate at helium pressure 1100 p.s.i., as described previously (Shan et al., 2013).

The wild-type KN199 as well as tae-miR156-OE transgenic lines (T2 generation) were planted in an experimental field (39°96’N, 116°33’E) of the Institute of Crop Science, the Chinese Academy of Agricultural Sciences, Beijing. Briefly, the seeds were sown at the beginning of October before the winter and harvested in mid-June next year. Nicotiana benthamiana was grown in a greenhouse at 22°C with a 16-h-light/8-h-dark cycle.

RNA Extraction and Gene Expression Analyses

For gene expression assays, leaves, stem bases, or young panicles were collected form wild-type KN199 or tae-miR156-OE lines, and total RNA was extracted using Trizol (Invitrogen) reagent. For detection of miRNAs by stem-loop qRT-PCR, specific reverse transcription primer for mature miRNAs with stem-loop structure was designed as previously described (Chen et al., 2005). About 2 μg of total RNA was used for reverse transcription. For quantification of coding genes, about 2 μg of total RNA and M-MLV reverse transcriptase (Promega) were further used for reverse transcription. SYBR Premix Ex Taq (Perfect Real Time; TaKaRa) was used for qRT-PCR assay, and expression levels of target genes were normalized to TaU6 or TaGAPDH. The statistical significance was evaluated by using data from three independent biological replications (Student’s t test).

The miRNA Gene Identification, Cloning, and Secondary Structure Prediction

The scaffold14333 from wheat A-genome progenitor Triticum urartu was used as the query sequence to search the wheat survey sequences using BLAST (https://urgi.versailles.inra.fr/blast/), and three scaffolds, i.e. scaffold211901, 225387, and 272705 from chromosomes 3A, 3B, and 3D, were obtained based on wheat TGAC (The Genome Analysis Centre) whole-genome shotgun assembly database. The cloning primers then were designed according to the flanking sequences of the tandem hairpin structures, and PCR were performed by using wild-type KN199 genome DNA as the template. The obtained PCR product was ligated to cloning vector pEASY-Blunt (Transgen Biotech; CB101). The sequences containing the tandem miR156 derived from chromosomes 3A, 3B, and 3D were identified by sequencing. The primer sequences are shown in Supplemental Table S2.

The secondary structures of the tandem miR156 gene containing three MIR156 precursors (MIR156a/b/c) from chromosomes 3A, 3B and 3D were separately predicted using the RNA-folding program Mfold (Zuker, 2003).

Validation of miRNA Cleavage Site by 5′RACE Assay

The RLM-RACE kit (TaKaRa; code D315) was used for 5′RACE assay according to the manufacturer’s instruction. About 2 μg of wild-type KN199 total RNAs were used for the ligation of RNA Oligo adaptor without calf intestinal phosphatase treatment. For the first round PCR, the 5′RACE outer primer together with gene-specific outer primers were used; a nested PCR amplification was then carried out using the 5′RACE inner primer together with gene-specific inner primers. The obtained PCR products were then were ligated to cloning vector for sequencing. The primer sequences are shown in Supplemental Table S2.

Gene and Promoter Cloning and Sequence Analyses

For coding gene cloning, a BLASTn search against the database of wheat survey sequences (https://urgi.versailles.inra.fr/blast/) was initially conducted by using the coding sequence of orthologs from other plant species as the query. According to the obtained sequences from the hexaploid bread wheat, specific primers were designed for the cloning in wild-type KN199 cDNAs. The obtained cloning products were further verified by sequencing. Encoded protein sequences were aligned with orthologous proteins from other plant species by MegAlign using the ClustalW method, and the identities among the protein sequences were calculated.

For promoter cloning, the corresponding scaffolds from the hexaploid bread wheat containing the target genes were first identified by screening the database of wheat survey sequences. Specific cloning primers were designed based on the 2-kb sequences upstream of the genes, and the promoters were cloned in KN199 by using its genome DNA as the template. Sequences were verified by sequencing. The primer sequences are shown in Supplemental Table S2.

Generation of DNA Constructs

The constructs used in this study are based on several expression vectors.

The constructs used for LCI assays were based on the vectors p1300-35S-nLUC and p1300-35S-cLUC (Chen et al., 2008). Briefly, the target genes were PCR amplified and separately ligated into the KpnI/SalI-digested p1300-35S-nLUC and KpnI/BamHI digested p1300-35S-cLUC using ligation free cloning mastermix (Applied Biological Materials; E011-5-A) according to the manufacturer’s instruction.

The vectors for BiFC assays were generated based on the Gateway vectors pEarleygate201-YN (nYFP) and pEarleygate202-YC (cYFP) (Lu et al., 2010). The constructs for transcriptional activation assays were based on the Gateway vector pGWBs (Nakagawa et al., 2007). All the gene sequences were cloned into the entry vector pQBV3 and subsequently introduced into the destination vectors pEarleygate201-YN (nYFP), pEarleygate202-YC (cYFP), or pGWBs following Gateway technology (Invitrogen).

All the primers used for generation of constructs are shown in Supplemental Table S2.

Transcriptional Activity Assays in N. benthamiana

The transcriptional activity assays were performed in N. benthamiana leaves as previously described (Sun et al., 2012). The 2-kb promoter sequences were fused with the luciferase reporter gene LUC through Gateway reactions (Invitrogen) into the plant binary vector pGWB35 (Nakagawa et al., 2007) to generate the reporter constructs. For effector construction, the coding sequences of indicated genes were cloned into the plant binary vector pGWB5 (Nakagawa et al., 2007) using pQBV3 as the entry. The reporter and effector constructs were separately introduced into Agrobacterium tumefaciens strain GV3101 (pMP90), to carry out the coinfiltration in N. benthamiana leaves. LUC activities were observed and quantified 48 h after infiltration using NightSHADE LB 985 (Berthold). In each experiment, 10 independent N. benthamiana leaves were infiltrated and analyzed, and three biological replications were performed with quantification.

LCI Assays

The LCI assays for the protein interaction detection was performed in N. benthamiana leaves as described previously (Sun et al., 2013). In brief, the indicated genes were separately fused with the N- and C-terminal parts of the luciferase reporter gene LUC and separately introduced into A. tumefaciens strain GV3101. A. tumefaciens cells harboring the nLUC and cLUC derivative constructs were coinfiltrated into N. benthamiana leaves, and the LUC activities were analyzed 48 h after infiltration using NightSHADE LB 985 (Berthold). In each analysis, five independent N. benthamiana leaves were infiltrated and analyzed, and three biological replications were performed with similar results.

BiFC Assays

For infiltration, A. tumefaciens cells harboring the nYFP and cYFP derivative constructs were coinfiltrated into N. benthamiana leaves, and YFP signal was imaged 48 h after infiltration using the confocal microscope (Carl Zeiss; LSM880). Five independent N. benthamiana leaves were observed for analyses, and totally three biological replications were performed with similar results.

Yeast Experiments

For transcriptional activation activity assays in yeast (Saccharomyces cerevisiae) cells, AH109 strain was used. In brief, the full-length TaSPL3 as well as its truncated versions (including the NT, MD, and CT parts) were fused with GAL4-BD vector pGBKT7 and separately transformed into yeast AH109 strain. The transformants were first selected on synthetic dextrose growth medium lacking Trp (SD-W). Then the yeast strains were dropped on synthetic dextrose selection medium lacking Trp and His (SD-W/H) for transcriptional activation activity evaluation according to their growth status.

Accession Numbers

Sequence data from this study can be found in the NCBI data library under the following accession numbers: TaSPL3, KF447885; TaSPL17, KF447877; TaSPL20, KF447878; TaSPL21, KF447882; TaTB1-4A_pro, KY363317; TaTB1-4B_pro, KY363318; TaTB1-4D_pro, KY363319; TaBA1-3A_pro, KY363320; TaBA1-3B_pro, KY363321; TaBA1-3D_pro, KY363322; TaD53, KY363316; and TaTPL, KY363324.

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Nucleotide sequence of partial scaffold14333 from Triticum urartu.

• Supplemental Figure S2. Predicted secondary structure of tandem tae-MIR156a/b/c precursors from chromosomes 3B and 3D.

• Supplemental Figure S3. Nucleotide sequence of tae-MIR156-3A cluster used for gene transformation in bread wheat plants.

• Supplemental Figure S4. Morphological characterization of wild-type KN199 and tae-miR156-OE transgenic bread wheat plants (3#).

• Supplemental Figure S5. Quantification of the expression levels of TaSPL20 and TaSPL21 in tae-miR156-OE bread wheat lines.

• Supplemental Figure S6. Expression of tae-miR156, TaSPL3 and TaSPL17 in wild-type KN199 in different development phases/tissues.

• Supplemental Figure S7. The coding sequences of TaTB1 genes.

• Supplemental Figure S8. Sequence comparison of TaTB1-4A/4B/4D, ZmTB1, and OsTB1 proteins.

• Supplemental Figure S9. Nucleotide sequences of TaTB1-4A/4B/4D promoters.

• Supplemental Figure S10. Transient expression assays illustrating the activation of TaTB1 transcription by TaSPL17.

• Supplemental Figure S11. Quantification of the expression levels of some inflorescence/panicle/spike development relevant bread wheat genes in tae-miR156-OE transgenic lines.

• Supplemental Figure S12. The coding sequences of TaBA1 genes.

• Supplemental Figure S13. Nucleotide sequences of TaBA1-3A/3B/3D promoters.

• Supplemental Figure S14. Transient expression assays illustrating the activation of TaBA1 expression by TaSPL17.

• Supplemental Figure S15. The coding sequences of TaD53 genes.

• Supplemental Figure S16. Sequence alignment of TaD53, OsD53, and AtSMXL7 proteins.

• Supplemental Figure S17. The N-terminal domain of TaSPL17 physically interacts with TaD53.

• Supplemental Figure S18. The coding sequence of TaTPL gene.

• Supplemental Figure S19. Alignment of TaTPL with TPL-related proteins in rice and Arabidopsis.

• Supplemental Table S1. Predicted target genes of tae-miR156.

• Supplemental Table S2. Primers used in this study.

Glossary

SL

strigolactone

KN199

Kenong199

CPS

counts per second

EV

empty vector

LCI

LUC complementation imaging

BiFC

bimolecular fluorescence complementation

hpi

hours postinfiltration