Promotion of BR Biosynthesis by miR444 Is Required for Ammonium-Triggered Inhibition of Root Growth

miR444 promotes phytohormone brassinosteroid biosynthesis, conferring ammonium-triggered root elongation inhibition in rice.

Abstract

Rice (Oryza sativa), the staple food for almost half of the world’s population, prefers ammonium (NH₄⁺) as the major nitrogen resource, and while NH₄⁺ has profound effects on rice growth and yields, the underlying regulatory mechanisms remain largely unknown. Brassinosteroids (BRs) are a class of steroidal hormones playing key roles in plant growth and development. In this study, we show that NH₄⁺ promotes BR biosynthesis through miR444 to regulate rice root growth. miR444 targeted five homologous MADS-box transcription repressors potentially forming homologous or heterogeneous complexes in rice. miR444 positively regulated BR biosynthesis through its MADS-box targets, which directly repress the transcription of BR-deficient dwarf 1 (OsBRD1), a key BR biosynthetic gene. NH₄⁺ induced the miR444-OsBRD1 signaling cascade in roots, thereby increasing the amount of BRs, whose biosynthesis and signaling were required for NH₄⁺-dependent root elongation inhibition. Consistently, miR444-overexpressing rice roots were hypersensitive to NH₄⁺ depending on BR biosynthesis, and overexpression of miR444’s target, OsMADS57, resulted in rice hyposensitivity to NH₄⁺ in root elongation, which was associated with a reduction of BR content. In summary, our findings reveal a cross talk mechanism between NH₄⁺ and BR in which NH₄⁺ activates miR444-OsBRD1, an undescribed BR biosynthesis-promoting signaling cascade, to increase BR content, inhibiting root elongation in rice.

Nitrogen (N) is an essential macronutrient for plant growth, development, and productivity. Ammonium (NH₄⁺) is the main and preferred inorganic N form for plant root uptake in flooded or acidic soils, such as those in rice (Oryza sativa) paddies. In addition to a N source, external or internal NH₄⁺ has profound effects on plant growth and development. For example, external NH₄⁺ usually inhibits root elongation and promotes lateral root formation. Plant responses to NH₄⁺, such as modulation of root system architecture (RSA), are key strategies for plants to maximize NH₄⁺ acquisition and utilization or alleviate the stress of excessive NH₄⁺ (Nacry et al., 2013; Giehl et al., 2014; Li et al., 2014a).

NH₄⁺ triggers plant growth and development, potentially as a signal (Lima et al., 2010; Liu and von Wirén, 2017; Straub et al., 2017). However, the NH₄⁺ signal transduction pathway from the sensor to downstream regulators remains largely unknown in plants (Li et al., 2014a; Liu and von Wirén, 2017; Xuan et al., 2017). Several components have been identified to affect NH₄⁺-dependent root growth in Arabidopsis (Arabidopsis thaliana). For example, the GMPase (GDP Man pyrophosphorylase) HSN1 (Hypersensitive to NH₄⁺1) confers NH₄⁺ inhibition of root growth by affecting N-glycosylation of proteins (Qin et al., 2008). The AMMONIUM TRANSPORTER1;3 (AMT1;3) plays a key role in NH₄⁺-triggered lateral root formation and high-order root branching independent of the accumulative uptake of NH₄⁺ (Lima et al., 2010). The CBL-INTERACTING PROTEIN KINASE23-CALCINEURIN B-LIKE1 (CIPK23-CBL1) complex mediates phosphorylation of NH₄⁺ transporters of AMT1s to regulate NH₄⁺-triggered root elongation inhibition by controlling NH₄⁺ accumulation (Straub et al., 2017). In addition, the [Ca²⁺]_cyt-associated protein kinase (CAP1) regulates NH₄⁺ homeostasis-dependent root hair tip growth by modulating tip-focused cytoplasmic Ca²⁺ gradients (Bai et al., 2014). Very little is known about the regulatory mechanisms of NH₄⁺ signaling in rice.

Plant hormones play vital roles in plant growth and development. It has been widely shown that nutrition-adapted growth and development are closely involved in the regulation of phytohormone biosynthesis and signaling (Krouk et al., 2011; Krouk, 2016; Guan, 2017). For example, auxin and ethylene have been suggested to regulate NH₄⁺-triggered lateral root formation in Arabidopsis (Li et al., 2014a). Brassinosteroids (BRs) are a kind of plant-specific steroidal hormone that orchestrate plant growth in a concentration-dependent manner (Chaiwanon et al., 2016; Jaillais and Vert, 2016; Wei and Li, 2016). Low concentrations of BRs promote root growth, and, conversely, high concentrations of BRs inhibit root growth, as shown in rice (Tong, et al., 2014). Like NH₄⁺, BRs inhibit root elongation, but it is still unclear if there is a direct connection between BR and NH₄⁺ signaling during plant growth and development.

MicroRNAs (miRNAs) are a class of small noncoding RNAs and direct target mRNA cleavage and translation repression (Rogers and Chen, 2013). miRNA plays crucial roles in the regulation of plant nutrition signaling and homeostasis (Liu et al., 2014; Chien et al., 2017). miR444 is monocotyledon specific and targets a group of MADS-box transcription factors in rice (Sunkar et al., 2005; Lu et al., 2008; Wu et al., 2009; Li et al., 2010). It has been suggested that miR444 and its target OsMADS-box 25 (OsMADS25) regulate rice nitrate (NO₃⁻) signaling-mediated root growth and development (Yan et al., 2014; Yu et al., 2015; Zhang et al., 2018). OsMADS-box 57 (OsMADS57), another target of miR444, regulates NO₃⁻ translocation from root to shoot (Huang et al., 2019). In addition, the expression of miR444’s MADS-box targets responds to NH₄⁺ in rice roots (Puig et al., 2013; Yu et al., 2014). It is not known if miR444 and its MADS-box targets play a regulatory role in NH₄⁺-triggered growth.

In this study, we show that NH₄⁺ promotes BR biosynthesis, which is responsible for NH₄⁺-triggered rice root elongation inhibition. Furthermore, we show that NH₄⁺ promotes BR biosynthesis by inducing the accumulation of miR444, which alleviates the direct repression of its MADS-box targets on the transcription of BR biosynthetic gene OsBRD1. Our work thus reveals a NH₄⁺-BR cross talk mechanism controlling root growth in rice.

RESULTS

NH₄⁺ Promotes BR Biosynthesis, Which Is Required for NH₄⁺-Triggered Root Elongation Inhibition in Rice

NH₄⁺ has profound effects on plant growth and development. To investigate the effects of NH₄⁺ on rice growth, wild-type rice plants cultured with 1 mm NO₃⁻ or 1 mm NH₄⁺ as the sole N resource were comparably observed. After 14 d of growth, we observed that compared to the 1 mm NO₃⁻-cultured wild-type roots, 1 mm NH₄⁺-cultured wild-type roots were significantly shorter (Fig. 1, A and B). Root growth under different NH₄⁺ concentrations also was investigated. The results showed that with an increase in NH₄⁺ concentrations, root elongation was inhibited more severely (Fig. 1, C and D), indicating that NH₄⁺ inhibits root elongation in rice. Interestingly, we also observed that wild type cultured with 1 mm NH₄⁺ exhibited obviously larger leaf angles compared to wild type culturing with 1 mm NO₃⁻ (Supplemental Fig. S1, A–C). Further observations showed that NH₄⁺ promoted rice leaf inclination in a dose-dependent manner (Supplemental Fig. S1, D and E). By contrast, NO₃⁻ concentrations did not obviously affect leaf angles (Supplemental Fig. S1, F and G). Together, the above observations indicate that NH₄⁺ inhibits root elongation and increases leaf inclination in rice.

Figure 1.

NH₄⁺ promotes BR biosynthesis to inhibit root elongation. A, Comparison of the root growth of 14-d-old wild-type (WT) rice seedlings cultured with 1 mm NO₃⁻ or 1 mm NH₄⁺ as the sole N source. Bar = 1 cm. B, The root length was statistically analyzed (mean ± se, n = 10, **P < 0.01, Student’s t test). C, Comparison of the root growth of 14-d-old wild-type rice seedlings cultured with a concentration gradient of NH₄⁺ as the sole N source. Bar = 1 cm. D, The root length was statistically analyzed (mean ± se, n = 10). E, Measurement of the contents of BR metabolites (pg/g fresh weight). Rice roots of 14-d-old seedlings cultured with 1 mm NO₃⁻ or 1 mm NH₄⁺ were used for the quantification analysis. (n = 3 for each replicate, **P < 0.01, Student’s t test). nd, No detection. F and G, Comparison of the root growth of wild-type rice plants cultured with NH₄⁺ with or without the addition of BRZ (mean ± se, n = 10, **P < 0.01, Student’s t test). Bar = 2 cm. H and I, Comparison of the root growth of wild-type rice plants cultured with methyl ammonium (MeA) with or without the addition of BRZ (mean ± se, n = 10, **P < 0.01, Student’s t test). Bar = 1 cm. J and K, Phenotypic comparison of the root growth between wild-type and d61 mutants cultured with NH₄⁺. d61-1 and d61-2 showed reduced sensitivity to NH₄⁺ in root elongation inhibition (mean ± se, n = 10, **P < 0.01, Student’s t test). Bar = 1 cm.

We were interested in the underlying mechanisms of NH₄⁺-triggered root elongation inhibition. High concentrations of BRs inhibit root elongation in rice (Tong et al., 2014). In addition, the increase of rice leaf angle is a typical BR signaling-enhanced phenotype (Hong et al., 2004; Tong and Chu, 2012; Sun et al., 2015). Therefore, we tested if NH₄⁺ increases rice BR biosynthesis or activates BR signaling, thus leading to root elongation inhibition. To do this, we measured the amounts of BRs in 1 mm NO₃⁻- and 1 mm NH₄⁺-cultured wild-type roots. The amount of typhasterol (TY) was significantly higher and the amount of castasterone (CS) was slightly higher in NH₄⁺-cultured roots than that in NO₃⁻-cultured roots (Fig. 1E), indicating that NH₄⁺ promotes BR biosynthesis in rice roots, which might be a reason for NH₄⁺-triggered root elongation inhibition.

Then, we used BRZ (a BR biosynthesis inhibitor)-treated rice and BR signaling-related mutants to examine if BR biosynthesis or signaling plays a role in rice root growth in response to NH₄⁺. The results clearly showed that the inhibition of BR biosynthesis by adding BRZ in the medium greatly alleviated the repression of NH₄⁺ as well as the transport analog methylammonium in root elongation (Fig. 1, F–I). In addition, the roots of BR-signaling mutants d61-1 and d61-2 grew longer and showed reduced repression compared to wild type when plants were cultured with NH₄⁺, in which the root length of d61-1 and d61-2 was reduced by 24.1% and 13.1%, respectively, in contrast to the 31.3% reduction of wild type (Fig. 1, J and K). Together, these observations indicate that both BR biosynthesis and signaling are required for rice root growth in response to NH₄⁺.

Thus, the above results collectively indicate that NH₄⁺ triggers root elongation inhibition by promoting BR biosynthesis and signaling in rice.

NH₄⁺ Induces the Accumulation of miR444, Which Targets Five Homologous MADS-Box Genes in Rice

The OsMIR444 gene family has six members named from OsMIR444a to OsMIR444f in rice. Each OsMIR444 precursor is processed to two or three kinds of mature miR444s. In total, 12 miR444 isoforms are produced and divided into four subgroups, including miR444.1, miR444.2, miR444a-5p, and miR444d.3, based on the similarity of their sequences (www.mirbase.org; Supplemental Fig. S2). It is well known that miR444 isoforms target a group of genes encoding MADS-box transcription factors in rice, including OsMADS23, OsMADS27, OsMADS57, and OsMADS61 (Sunkar et al., 2005; Lu et al., 2008; Wu et al., 2009; Li et al., 2010). Interestingly, bioinformatics analysis suggests that OsMADS25, encoding another MADS-box transcription factor, also is a potential target of miR444.2 (plantgrn.noble.org; Fig. 2A). OsMADS25 mainly expresses in rice roots (Puig et al., 2013). We found that the mRNA level of OsMADS25 is reduced in roots of previously described OsMIR444a-overexpressed rice plants (Yan et al., 2014; named OsMIR444a-OE rice plants), which show increase of miR444.2 accumulation (Fig. 2, B and C). In addition, OsMIR444a decreased the accumulation of OsMADS25 mRNAs, but not the mRNAs of miR444a.2-resistant OsMADS25 (OsMADS25^R), when they were coexpressed transiently in Nicotiana benthamiana leaves (Fig. 2, D and E), suggesting that miR444a.2 directs OsMADS25 mRNA cleavage in plants. Further, a 5′ rapid amplification of complementary DNA (cDNA) ends assay confirmed that OsMADS25 mRNA is cleaved mainly at the position between the 10th and 11th nucleotides of the miR444.2 binding site in rice roots (Fig. 2A). Together, the above results indicate that OsMADS25 is a real target of miR444.2 and its mRNA is regulated by miR444.2-mediated cleavage in rice. Recently, miR444a.4-3p, a newly identified miR444a isoform, also was shown to target OsMADS25 (Shin et al., 2018). OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 are homologous and form a subclade in the rice MADS-box protein family (Arora et al., 2007). Thus, all above observations together indicate that miR444 targets five homologous MADS-box transcription factors in rice.

Figure 2.

miR444.2 targets OsMADS25 mRNA for cleavage. A, 5′ rapid amplification of cDNA ends assay indicating that OsMADS25 mRNA is cleaved at the miR444.2 target site. The sequence of miR444.2 and its complementary sequences at OsMADS25 mRNA or miR444.2-resistant OsMADS25 (OsMADS25^R) mRNAs are aligned. G-U base pairing and mismatches are shown with dotted lines and circles, respectively. B and C, Reverse transcription-quantitative PCR (RT-qPCR) analysis showing that the expression of OsMADS25 decreased accompanying the increase of miR444.2 accumulation in OsMIR444a-OE rice roots (mean ± se, n = 3, *P < 0.05, Student’s t test). D and E, Coinfiltration assay showed that miR444.2 repressed the accumulation of OsMADS25 mRNA, but not OsMADS25^R mRNA, in N. benthamiana leaves. Schematic structures of the four constructs for coinfiltration assays are shown in D. RT-PCR was used to determine the OsMADS25 mRNA levels. The expression of the hygromycin B phosphotransferase (HPH) gene was used as a miR444.2-uncleaved control (E).

miR444’s targets, OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61, are orthologs of Arabidopsis ANR1, which is a key regulator in NO₃⁻-triggered lateral root elongation (Zhang and Forde, 1998; Arora et al., 2007; Gan et al., 20122). Rice has a preference for NH₄⁺ over NO₃⁻ as a N source. The expression of OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 has been surveyed in response to NH₄⁺. Interestingly, all of the five MADS-box genes down-regulate their expression when N-starved rice roots are resupplied with NH₄⁺ (Puig et al., 2013; Yu et al., 2014), indicating that NH₄⁺ treatment represses their expression in rice roots. We further explored if NH₄⁺-triggered regulation of the expression of these MADS-box genes attributed to the change of miR444 accumulation. We found that compared to the 1 mm NO₃⁻-cultured rice roots, rice roots grown in a medium with 1 mm NH₄⁺ as the sole N source showed much higher miR444 accumulation accompanying the reduced expression of its five MADS-box targets with varied extents (Fig. 3), suggesting that NH₄⁺ promotes miR444 accumulation to down-regulate the expression of its MADS-box target genes in rice roots.

Figure 3.

NH₄⁺ induces miR444 accumulation in rice roots. A and B, RT-qPCR analysis indicating that the accumulation of miR444 (A) and the expression of miR444 targets (B) increased and decreased, respectively, in 1 mm NH₄⁺-cultured rice roots compared to the 1 mm NO₃⁻-cultured rice roots (mean ± se, n = 3, **P < 0.01, Student’s t test).

miR444 Has a Positive Role in NH₄⁺-Triggered Root Elongation Inhibition

We further investigated if miR444 has a role in rice root growth in response to NH₄⁺ using previously described OsMIR444a-OE rice plants and miR444-resistant OsMADS57-overexpressed rice plants (referred to as OsMADS57^R-OE rice plants by Wang et al., 2016). To do this, rice plants of different genotypes were grown in the NH₄⁺-containing media with gradually increased NH₄⁺ concentrations, and root growth was comparably observed. The results showed that OsMIR444a-OE and OsMADS57^R-OE were hypersensitive and hyposensitive, respectively, to NH₄⁺ in triggering root elongation inhibition (Fig. 4), indicating that miR444 plays a positive role in NH₄⁺-triggered root growth by regulating its MADS-box target(s) in rice.

Figure 4.

miR444 positively regulates NH₄⁺-triggered root elongation inhibition. A and B, Phenotypic comparison between wild-type (WT) and OsMIR444a-OE or OsMADS57^R-OE rice plants of 14-d-old cultured with a concentration gradient of NH₄⁺ as the sole N source. C and D, Measurement of the root length of 14-d-old rice seedlings described in A and B (mean ± se, n = 10).

miR444 Promotes BR Biosynthesis by Up-regulating the Expression of OsBRD1

Interestingly, we observed that OsMIR444a-OE rice plants showed bigger leaf angles grown in Murashige and Skoog (MS) medium (Supplemental Fig. S3, A and B) and in soils, implying a potential role of miR444 in positive regulation of BR biosynthesis or signaling. BRs are positively involved in the growth of coleoptiles and internodes (Yamamuro et al., 2000). We further observed that the growth of coleoptiles and internodes was more elongated in OsMIR444a-OE than that in wild type under dark conditions (Supplemental Fig. S3, E–G). By contrast, OsMADS57^R-OE showed BR signaling-reduced phenotypes with smaller leaf angles and shortened coleoptiles and internodes when grown in dark compared with wild type (Supplemental Fig. S3, C, D, and H–J). Together, the above results suggest that miR444 positively regulates BR signaling via its MADS-box target(s) in rice.

Enhanced BR signaling usually represses the expression of BR-biosynthetic and -signaling genes as a negative feedback mechanism (He et al., 2005; Sun et al., 2010; Yu et al., 2011). Indeed, the expression of BR-biosynthetic genes such as OsDWARF11 (OsD11), OsDWARF4 and BR-deficient dwarf 2 (OsBRD2) and BR-signaling genes BRASSINOSTEROID INSENSITIVE 1 (OsBRI1) and BRASSINAZOLE-RESISTANT 1 (OsBZR1) were significantly down-regulated in OsMIR444a-OE (Supplemental Fig. S4, A–F). Unexpectedly, OsBRD1, a key BR-biosynthetic gene (Hong et al., 2002; Mori et al., 2002), was expressed ∼8-fold higher and was accompanied by a significant increase in the contents of TY and CS in OsMIR444a-OE than in wild type (Fig. 5, A and D), indicating that overexpression of OsMIR444a promotes BR biosynthesis by increasing the expression of OsBRD1, which then enhances BR signaling in rice. Reversely, OsMADS57^R-OE showed decreased OsBRD1 expression and reduced amounts of BRs (TY and CS; Fig. 5, B and D). Thus, the above results indicate that miR444 positively regulates OsBRD1 expression to promote BR biosynthesis through decreasing the expression of its MADS-box target(s) in rice, which might be a mechanism involved in the miR444-mediated regulation of rice root response to NH₄⁺.

Figure 5.

miR444 positively regulates BR biosynthesis. A and B, RT-qPCR analysis showed that the relative expression of OsBRD1 increased and decreased in OsMIR444a-OE and OsMADS57^R-OE rice plants, respectively (mean ± se, n = 3, **P < 0.01, Student’s t test). C, RT-qPCR analysis showed that the relative expression of OsBRD1 and OsAMT1.3 increased in 1 mm NH₄⁺-cultured rice roots compared with the 1 mm NO₃⁻-cultured rice roots (mean ± se, n = 3, **P < 0.01, Student’s t test). D, Measurement of the contents of BR metabolites (pg/g fresh weight) in wild-type (WT), OsMIR444a-OE, and OsMADS57^R-OE rice plants. Leaf laminas of 2-week-old seedlings were used for the quantification analysis (n = 3 for each replicate, **P < 0.01, Student’s t test).

miR444 Regulates NH₄⁺-Triggered Root Growth Dependent on BR Biosynthesis

The above results showed that NH4⁺ induces the accumulation of miR444, which promotes BR biosynthesis by up-regulating OsBRD1 expression (Figs. 3A and 5D) and positively regulates NH₄⁺-dependent root elongation inhibition (Fig. 4 ). We have shown that NH₄⁺ promotes BR biosynthesis conferring NH₄⁺-dependent root elongation inhibition (Fig. 1). These observations together suggest that miR444 positively regulates rice root growth in response to NH₄⁺ by promoting BR biosynthesis. Further experiments supported this notion as shown by the fact that BR biosynthesis inhibition (by adding BRZ) caused OsMIR444a-OE being no longer hypersensitive to NH₄⁺ (Fig. 6, A and B); in agreement with this, extra brassinolide (BL) addition reduced the insensitive responses of OsMADS57^R-OE to NH₄⁺ (Fig. 6, C and D).

Figure 6.

BR biosynthesis is required for miR444-mediated increase of NH₄⁺ sensitivity. A and B, Comparison of the root growth of wild-type (WT) and OsMIR444a-OE rice plants cultured with 1 mm NH₄⁺ with or without the addition of BRZ (mean ± se, n = 10, *P < 0.05, **P < 0.01, Student’s t test). C and D, Comparison of the root growth of wild-type and OsMADS57^R-OE rice plants cultured with 1 mm NH₄⁺ with or without the addition of BL (mean ± se, n = 10, *P < 0.05, **P < 0.01, Student’s t test). Bar = 1 cm.

Thus, the above observations collectively indicate that NH₄⁺ activates the miR444-OsBRD1 signaling cascade to promote BR biosynthesis, resulting in root elongation inhibition. Consistently, the expression of OsBRD1 was ∼6-fold higher in 1 mm NH₄⁺-cultured roots than that in 1 mm NO₃⁻-cultured roots (Fig. 5C) accompanying the increase in miR444 accumulation and the decrease in the expression of its MADS-box target genes (Fig. 3).

miR444’s MADS-Box Targets Directly Repress OsBRD1 Transcription

Plant MADS-box transcription factors regulate gene expression by binding to a highly conserved CC(A/T)6GG or C(A/T)8G (CArG) motif (de Folter and Angenent, 2006; Ito et al., 2008; Fujisawa et al., 2013). It has been shown that miR444’s targets OsMADS23, OsMADS27, and OsMADS57 repress gene expression through directly binding to the CArG motifs of gene promoters (Guo et al., 2013; Wang et al., 2016; Chen et al., 2018). Cis-element scanning found that the OsBRD1 promoter contained a typical CArG motif (CTATATAAAG) close to its transcriptional start site (Supplemental Fig. S5A), implying that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and/or OsMADS61 might repress OsBRD1 expression by directly binding to the CArG motif of its promoter, and the miR444-mediated increase in OsBRD1 expression might be due to the attenuation of the repressive roles of its MADS-box targets on OsBRD1 transcription. Indeed, transient expression assays by coexpressing OsMADS23, OsMADS25, OsMADS27, OsMADS57, or OsMADS61 with OsBRD1 promoter-fused GUS in N. benthamiana leaves indicated that the five MADS-box transcription factors were capable of repressing the activity of the OsBRD1 promoter (Fig. 7, A and B). In addition, electrophoretic mobility shift assays (EMSA) confirmed that each of the five MADS-box proteins could form protein-DNA complexes with the CArG motif of OsBRD1 promoter in vitro but not with the CArG motif from the promoter of OsDWARF2 (OsD2; CAATAAAAAG; Supplemental Fig. S5B), another BR biosynthesis gene (Hong et al., 2003; Fig. 7C). Further, yeast one-hybrid (Y1H) assays showed that each of the five MADS-box proteins could bind to the region of the OsBRD1 promoter containing the CArG motif in yeast (Saccharomyces cerevisiae; Fig. 7, D and E). More importantly, these MADS-box proteins interacted with the CArG motif-containing region of the OsBRD1 promoter in rice roots as shown by chromatin immunoprecipitation (ChIP)-PCR assays using protein-specific antibodies (Fig. 7F). Together, these observations suggest that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 directly bind to the CArG motif of the OsBRD1 promoter for transcriptional repression in rice. Thus, miR444 promotes the expression of OsBRD1 by alleviating the direct repression of its MADS-box targets on OsBRD1 transcription, revealing a signaling cascade from miR444 to OsBRD1 through the five MADS-box proteins controlling BR biosynthesis in rice.

Figure 7.

miR444’s MADS-box targets directly bind to the OsBRD1 promoter and repress its expression. A, Schematic diagram of the constructs for transient expression assays. GUS with an intron driven by the OsBRD1 promoter was used as a reporter system. The Renilla luciferase (LUC) gene driven by the 35S promoter was used as an internal reference. B, Transient expression assays showed that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 repressed the activity of the OsBRD1 promoter in N. benthamiana leaves through measuring the expression of the OsBRD1 promoter-fused GUS reporter gene (mean ± se, n = 3, **P < 0.01, Student’s t test). C, EMSA analysis showed that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 bound to the CArG motif from OsBRD1, but not from OsD2. The biotinylated DNA probe containing the CArG motif sequence from the OsBRD1 or OsD2 promoter was incubated with GST-OsMADS23, GST-OsMADS25, GST-OsMADS27, GST-OsMADS57, GST-OsMADS61, or GST protein. The nonlabeled DNA probe was used as the cold competitor. D, Schematic diagram of the constructs for Y1H assays. OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 were fused to the GAD, and the reporter gene LacZ was driven by the OsBDR1 promoter fragment containing the CArG motif. E, Y1H assays showed that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 bound to the OsBRD1 promoter region containing the CArG motif in yeast. F, ChIP-PCR assays showed that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 bound to the OsBRD1 promoter region containing the CArG motif in rice roots. The amplified PCR band from the OsUbiquitin promoter was used as a negative control. The top represents a schematic diagram of the CArG site (red rectangle) and the amplified region of ChIP-PCR (labeled by two green arrows). G, A proposed model for the NH₄⁺-triggered signaling cascade regulating root growth in rice. miR444 promotes BR biosynthesis by releasing OsBRD1 expression from the repression of its MADS-box protein targets. NH₄⁺ promotes BR biosynthesis, then enhances BR signaling, through activating the miR444-MADS-box-OsBRD1 signaling cascade, resulting in root elongation inhibition.

To interact with the CArG motif, MADS-box proteins need to form homologous or heterogeneous complexes (Riechmann et al., 1996). OsMADS23, OsMADS27, and OsMADS57 have the ability to form homodimers and heterodimers with each other in rice (Guo et al., 2013; Wang et al., 2016; Chen et al., 2018). EMSAs showed that OsMADS25 and OsMADS61 formed high-order binding complexes with the CArG motif of the OsBRD1 promoter (Fig. 7C), suggesting that they could form a homologous complex. Further yeast two-hybrid (Y2H) assays confirmed that OsMADS25 and OsMADS61 interacted with themselves in yeast (Supplemental Fig. S6, A and B). In addition, Y2H assays displayed that OsMADS25 interacted with OsMADS27, and OsMADS61 interacted with OsMADS23 and OsMADS27 in yeast (Supplemental Fig. S6, A and B). Furthermore, the five MADS-box genes show similar expression patterns, especially in roots (Puig et al., 2013). These results together suggest that the five homologous MADS-box proteins potentially form homologous or heterogeneous complexes in rice. So, it is possible that OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 bind to the OsBRD1 promoter as high-order homologous or heterogeneous complexes in rice.

DISCUSSION

N is a fundamental macronutrient for plant growth and development. NO₃⁻ and NH₄⁺ are two major inorganic N forms for plant absorption from soil. In addition to providing nutrition, NO₃⁻ and NH₄⁺ usually act as signals to regulate plant growth and development, thus affecting crop grain yields (Ho and Tsay, 2010). For example, NO₃⁻ and NH₄⁺ availability fundamentally influence rice root system architecture, which directly determines the utilization efficiency of nutrition and water (Li et al., 2017). The underlying mechanisms of NO₃⁻-regulated root architectures have been widely studied in Arabidopsis (Kiba and Krapp, 2016). In contrast, little is known about the physiological and molecular mechanisms for NH₄⁺-triggered plant root growth (Li et al., 2014a; Liu and von Wirén, 2017). In this study, we reveal a molecular mechanism in which regulation of BR biosynthesis by the miR444-OsBRD1 signaling cascade plays an important role in NH₄⁺-triggered root growth in rice.

miR444 Positively Regulates Rice BR Biosynthesis via Its MADS-Box Targets by Activating OsBRD1 Expression

BRs do not undergo long-distance transportation within plants and are biosynthesized at tissue and cellular levels (Symons and Reid, 2004). BR homeostasis is critical for BR-mediated regulation in plants and is tightly controlled by positively or negatively regulating BR biosynthesis and metabolism (Zhao and Li, 2012). Transcription factors, such as BZR1, BRI1 EMS SUPPRESSOR 1 (BES1), TEOSINTE-LIKE 1, CYCLOIDEA, and PROLIFERATING CELL FACTOR 1 transcription factor TCP1, CESTA, and Phytochrome-interacting transcription factor 4/5 (PIF4/5) in Arabidopsis and OsBZR1 and Related to ABI3/VP1-Like1 (RAVL1) in rice, play important roles in the regulation of BR homeostasis by directly binding to the promoters of BR biosynthetic genes (He et al., 2005; Guo et al., 2010; Je et al., 2010; Sun et al., 2010; Poppenberger et al., 2011; Yu et al., 2011; Wei and Li, 2016).

We here show that the five homologous MADS-box transcription factors (MADS23/25/27/57/61) negatively regulate BR biosynthesis by directly binding to the promoter of BR biosynthetic gene OsBRD1 and that miR444 targets the five MADS-box transcription factors to positively regulate BR biosynthesis in rice, thus revealing that miR444-MADS-box-OsBRD1 controls BR homeostasis as an important signaling cascade in rice. Several key observations support this notion. First, MIR444a-OE rice showed increased the expression of OsBRD1, altered profiles of bioactive BRs with increased TY and CS and typical BR signaling-enhanced phenotypes (Fig. 5, A and D; Supplemental Fig. S3). Second, overexpression of one of the five MADS-box genes, OsMADS57, resulted in down-regulation of the expression of OsBRD1, a decrease in the bioactive contents of BRs (TY and CS), and a reduction in BR sensitivity in rice (Fig. 5, B and D; Supplemental Fig. S3). Third, each of the five MADS-box proteins was able to form a homodimer and bind to the CArG motif of the OsBRD1 promoter to repress its transcription, as shown by a series of assays in vitro and in planta (Fig. 7). In addition, among the five MADS-box proteins, some interacted with others in yeast (Supplemental Fig. S6; Wang et al., 2016), suggesting that they potentially form heterocomplexes to play the regulatory role. Collectively, the above observations revealed that miR444-MADS-box-OsBRD1 is an undescribed signaling cascade controlling BR biosynthesis in rice. It has been shown that miR444 or its target plays a regulatory role in plant growth and response to biotic and abiotic stresses (Guo et al., 2013; Li et al., 2014b; Yan et al., 2014; Wang et al., 2016; Chen et al., 2018; Chu et al., 2019; Huang et al., 2019). Future studies need to illustrate how the miR444-MADS-box-OsBRD1 signaling pathway is exploited to control BR biosynthesis during various growth and developmental processes and environmental conditions.

Promotion of BR Biosynthesis by miR444 Is Required for NH₄⁺-Triggered Root Elongation Inhibition in Rice

Phytohormones such as cytokinin, auxin, abscisic acid, and ethylene play key roles in N-mediated regulation of plant growth and development in which N signals regulate the biosynthesis, transportation, or signaling of these phytohormones (Krouk et al., 2011; Krouk, 2016; Guan, 2017). Recently, a genome-wide association analysis revealed that BR signaling kinase3 (BSK3) alleles contribute to root elongation in Arabidopsis accessions under low N conditions, indicating that BR signaling plays a role in limited N-induced root elongation (Jia et al., 2019). The regulatory mechanism of NH₄⁺ signaling is poorly understood, and the connection between NH₄⁺ and BRs has not been well established (Wei and Li, 2016).

Here, we demonstrate that NH₄⁺ activates the miR444-MADS-box-OsBRD1 signaling cascade to promote the biosynthesis of BR, which is required for NH₄⁺-triggered root growth in rice. We showed that miR444 accumulation increased accompanying the reduction of the expression of its MADS-box target genes, which then led to enhancement of OsBRD1 expression and BR contents (TY and CS) in NH₄⁺-cultured rice roots compared with that in NO₃⁻-cultured rice roots (Figs. 1E, 3, and C5C). We also showed that BR biosynthesis and signaling were required for NH₄⁺-triggered rice root growth by analyzing NH₄⁺ responses of the BRZ-treated rice plants and rice BR signaling mutants (Fig. 1, F–K). We further showed that OsMIR444a-OE rice plants responded stronger to NH₄⁺ in the inhibition of root elongation dependent on BR biosynthesis (Fig. 6, A and C), and OsMADS57^R-OE rice plants had reduced response to NH₄⁺ in the inhibition of root elongation but recovered their sensitivity to NH₄⁺ under addition of BL (Fig. 6, B and D). These observations together indicate that increased BR biosynthesis by the miR444-MADS-box-OsBRD1 signaling cascade plays a key role in NH₄⁺-triggered root growth in rice. Further study needs to illustrate how miR444-MADS-box-OsBRD1 signaling cascade-mediated BR biosynthesis confers NH₄⁺-triggered root growth in rice. OsMADS23, OsMADS25, OsMADS27, and OsMADS57 specifically express in the rice root central cylinder (Puig et al., 2013), suggesting that the change of the contents of BRs in the root central cylinder may play an important role in NH₄⁺-triggered effects in rice root growth. Among the BR derivatives, the content of TY, but not TE or CS, was increased the most in NH₄⁺-cultured rice roots (Fig. 1E). Whether TY is the most biologically active BR functioning in NH₄⁺-regulated root growth in rice needs to be further examined.

miR444 and its target, OsMADS25, negatively and positively, respectively, regulate NO₃⁻-triggered lateral root elongation in rice (Yan et al., 2014; Yu et al., 2015). Another miR444 target, OsMADS57, regulates long-distance NO₃⁻ transport (Huang et al., 2019). It is interesting to know if control of BR biosynthesis by the miR444-MADS-box-OsBRD1 signaling cascade mediates NO₃⁻-triggered regulation and NO₃⁻- and NH₄⁺-signaling cross talk in rice.

In addition, nutrition, such as N and phosphate (Pi), modulates rice leaf inclination angle, which is a key agronomic trait and affects rice light capture (Mghase et al., 2011). Very recently, Ruan et al. revealed that the Pi starvation-induced protein SPX1(SPX domain-containing protein1) interacts with REGULATOR OF LEAF INCLINATION1 (RLI1) to regulate leaf inclination in response to Pi availability by modulating the expression of BR signaling-related genes BRASSINISTEROID UPREGULATED1 (BU1) and BUL1 COMPLEX1 (BC1) in rice (Tanaka et al., 2009; Jang et al., 2017; Ruan et al., 2018). We show that NH₄⁺ has an effect on rice leaf inclination (Supplemental Fig. S1, D and E). Further study needs to illustrate the regulatory mechanisms and whether BR biosynthesis or signaling plays a role in NH₄⁺-triggered leaf inclination increase in rice.

CONCLUSION

In summary, our study demonstrates that miR444 promotes BR biosynthesis by releasing OsBRD1 expression from the repression of its MADS-box protein targets. By exploiting this mechanism, NH₄⁺ promotes BR biosynthesis through activating the miR444-MADS-box-OsBRD1 signaling cascade, resulting in root elongation inhibition in rice (Fig. 7G). Though excess NH₄⁺ is toxic to plants, rice prefers NH₄⁺ as the form of N nutrition. This mechanism could be manipulated to engineer yield-improved crops by coordinating N utilization with BR biosynthesis.

MATERIALS AND METHODS

Plant Materials and Growth Conditions

OsMIR444a-OE and OsMADS57^R-OE rice (Oryza sativa) lines were previously produced in our laboratory and have been used in previous studies (Yan et al., 2014; Wang et al., 2016). Rice d61-1 and d61-2 mutants were kindly provided by Dr. Chengcai Chu. For phenotypic observations, plants were grown in a 25°C incubator with a 16-h light/8-h dark photoperiod or under dark and approximately 70% humidity. All of the nutrition cultures for plant growth were solid one-half strength MS medium (PhytoTechnology Laboratories), except when described specifically, with 2% (w/v) Glc and 1% (w/v) agar. To produce cultures with different NO₃⁻ or NH₄⁺ concentrations, N-free one-half strength MS medium (PhytoTechnology Laboratories) was used with the addition of the desired KNO₃ or NH₄Cl. K⁺ concentration adjusted to the same level in all cultures using KCl. For chemical treatment, BL, BRZ, or methylammonium with desired concentrations was added to the medium.

BR Measurement

For wild-type rice plants grown with NO₃⁻ or NH₄⁺, rice roots of 2-week-old seedlings cultured with 1 mm NO₃⁻ or 1 mm NH₄⁺ were used for the quantification analysis. For wild-type, OsMIR444a-OE, and OsMADS57^R-OE rice plants, ∼2 cm second leaf laminas of 2-week-old seedlings were used for the quantification analysis. BR quantifications were performed based on the ultra-performance liquid chromatography-mass spectrometry method as described previously (Xin et al., 2016).

RT-qPCR Analysis

To analyze coding-gene expression, total RNA (1 μg) treated with genomic DNA remover was subjected to reverse transcription to produce cDNA products using RT III reaction mixture (TOYOBO) following the supplier’s protocol. To detect miRNA accumulation, total RNA (1 μg) underwent the reverse transcription reaction with miRNA-specific hairpin primers using RT III reverse transcriptase (Invitrogen). RT-qPCR was performed by adding SYBR green real-time PCR master mix (TOYOBO) to the reaction system and run on a DNA engine Opticon2 real-time PCR detection system (Bio-Rad) in accordance with the manufacturer’s instructions. Three replicates were performed for each gene or miRNA. Relative quantification of each sample was determined by normalization against the internal control gene actin or miR528. The sequences of the primers used for RT-qPCR are listed in Supplemental Table S1.

Transient Expression in Nicotiana benthamiana

The transient expression assays were performed as previously described (Wang et al., 2016) to examine the repressive activities of OsMADS23, OsMADS25, OsMADS27, OsMADS57, and OsMADS61 on OsBRD1 transcription and miR444-mediated cleavage of OsMADS25 mRNA. In brief, after overnight growing, Agrobacterium tumefaciens (GV3101) transformants harboring the designated construct (pCAMBIA1300-OsMADS23/25/25^R/27/57/61/OsMIR444a, pBI121-pro::OsBRD1::GUS, or an internal control construct 35S:Riluc) were harvested by centrifugation and resuspended in MMA buffer (10 mm MgCl₂, 10 mm MES [pH 5.6], and 100 μm acetosyringone) to an OD600 of 1.0. Then, the agrobacterial cells harboring the designated constructs were mixed with a desired ratio. After incubation at room temperature for 3 h, the agrobacterial cell suspension containing different construct combinations was pressure-infiltrated into N. benthamiana leaves. Two days after infiltration, the leaves were harvested for the analysis of GUS and Renilla luciferase activities or OsMADS25 mRNA level.

EMSA

GST-OsMADS23, GST-OsMADS27, and GST-OsMADS57 proteins were purified using the constructs described previously (Wang et al., 2016). To express and purify the recombinant proteins of GST-OsMADS25 and GST-OsMADS61, OsMADS25 and OsMADS61 were separately cloned to pGEX-3X/GST vector. EMSA was performed using the Light Shift Chemiluminescent EMSA kit following the supplier’s protocol (Thermo Scientific). Briefly, 2 nm biotin-labeled probes and purified GST fusion proteins with a concentration gradient (0.2, 2.5, and 5 pm) were incubated in a 20-μL reaction mixture at room temperature for 30 min. Then, the reaction mixtures were separated on a 6% (w/v) native polyacrylamide gel and transferred electrophoretically to Hybond-N⁺ membranes (Amersham BioScience). For the competition assays, unlabeled probe sequences (5-, 10-, or 50-fold of the labeled probe sequences) were added to the EMSA reactions. Labeled fragments and their shifted complexes with proteins were visualized using the kit of Chemiluminescent Nucleic Acid Detection Module (Thermo Scientific). The probe sequences are listed in Supplemental Table S1.

ChIP-PCR

ChIP-PCR assays were performed according to the method described previously (Wang et al., 2016) with all operations carried out at 4°C. In brief, 2-week-old wild-type roots (1 g) were harvested and ground to powder in liquid nitrogen after cross linking with 1% (v/v) formaldehyde under vacuum. The powders were suspended in ChIP extraction buffer I (0.4 m Suc, 10 mm Tris-HCl [pH 8.0], 10 mm MgCl₂, 1mm dithiothreitol [DTT], 0.1 mm phenylmethylsulfonyl fluoride [PMSF], and cocktail protease inhibitor [Sigma]). After filtering through two layers of miracloth, the mixture was centrifuged 20 min with 4000 rpm. Then, the pellet was resuspended gently and precipitated (10 min, 12,000 rpm) two times in 1 mL ChIP extraction buffer II (0.25 m Suc, 10 mm Tris-HCl [pH 8.0], 10 mm MgCl₂, 1% [v/v] TritonX-100, 1 mm DTT, 0.1 mm PMSF, and cocktail protease inhibitor [Sigma]). Then, the pellet was resuspended gently in 300 µL ChIP extraction buffer III (1.7 m Suc, 10 mm Tris-HCl [pH 8.0], 2 mm MgCl₂, 0.15% [v/v] TritonX-100, 1 mm DTT, 0.1 mm PMSF, and cocktail protease inhibitor [Sigma]). After centrifuging (1 h, 13,000 rpm), the pellet was resuspended gently in 100 µL nuclear lysis buffer (50 mm Tris-HCl [pH 8.0], 10 mm EDTA [pH 8.0], 1% [w/v] SDS, 1 mm DTT, 0.1 mm PMSF, and cocktail protease inhibitor [Sigma]) to get chromatin complexes. After 30 min on ice, the chromatin complexes were sonicated by adding desired ChIP dilution buffer (16.7 mm Tris-HCl [pH 8.0], 1.2 mm EDTA [pH 8.0], 1% [v/v] TritonX-100, 167 mm NaCl, 1 mm DTT, 0.1 mm PMSF, and cocktail protease inhibitor [Sigma]). After keeping 50 µL as the input, the sonicated chromatin complexes were used for immunoprecipitation assays by incubating with protein A/G dynal beeds and purified polyclonal antibodies (anti-OsMADS23, anti-OsMADS25, anti-OsMADS27, anti-OsMADS57, and anti-OsMADS61, BGI). The precipitated DNA was recovered and dissolved in water as the templates for PCR analysis. The primer sequences are listed in Supplemental Table S1.

Y1H Assay

The pGAD424 plasmid harboring GAL4-activation domain-fused MADS-box gene (pGAD424::OsMADS23, pGAD424::OsMADS25, pGAD424::OsMADS27, pGAD424::OsMADS57, or pGAD424::OsMADS61) and the plasmid pLacZi2µ containing the LacZ reporter gene driven by the OsBRD1 promoter fragment (proOsBRD1::LacZ) were cotransformed into the yeast (Saccharomyces cerevisiae strain EGY48). The transformants were grown on proper drop-out plates (lacking Ura and Leu) containing X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) for blue color selection. The primer sequences for plasmid construction are listed in Supplemental Table S1.

Y2H Assay

For Y2H assays, the designated plasmids of pDEST-32 and pDEST-22 harboring OsMADS23, OsMADS25, OsMADS27, OsMADS57, or OsMADS61 were combined and transformed into the yeast strain MaV203. The transformed yeast cells containing different construct couples were spread on the medium lacking Leu and Trp at 28°C for 2 d. To analyze the interactions, yeast transformants were screened by growing on selective medium excluding Leu, Trp, and His plus 5 mm 3-amino-1,2,4-triazole. The primer sequences for plasmid construction are listed in Supplemental Table S1.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL or RiceGE data libraries under the following accession numbers: LOC4327329 (D2), LOC4336116 (D11), LOC4332134 (OsDWARF4), LOC4324691 (OsBRI1), LOC4343719 (OsBZR1), LOC4333399 (OsBRD1), LOC4348555 (OsBRD2), LOC4345644 (OsMADS23), LOC4335427 (OsMADS25), LOC4329771 (OsMADS27), LOC4330621 (OsMADS57), LOC_Os04g38770 (OsMADS61), and LOC4338914 (Actin).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Comparison of the leaf inclination of rice plants grown with sole NH₄⁺ or NO₃⁻.

• Supplemental Figure S2. miR444 family in rice.

• Supplemental Figure S3. Phenotypic comparison between wild-type and OsMIR444a-OE or OsMADS57^R-OE rice plants.

• Supplemental Figure S4. Expression of BR biosynthesis and signaling genes in OsMIR444a-OE rice plants.

• Supplemental Figure S5. Schematic diagram of the position of the CArG motifs of OsBRD1 and OsD2 promoters.

• Supplemental Figure S6. Interactions between miR444’s MADS-box target proteins in Y2H assays.

• Supplemental Table S1. Primers and probes used in this study.