- © 2015 American Society of Plant Biologists. All Rights Reserved.
Abstract
Legume root nodules convert atmospheric nitrogen gas into ammonium through symbiosis with a prokaryotic microsymbiont broadly called rhizobia. Auxin signaling is required for determinant nodule development; however, the molecular mechanism of auxin-mediated nodule formation remains largely unknown. Here, we show in soybean (Glycine max) that the microRNA miR167 acts as a positive regulator of lateral root organs, namely nodules and lateral roots. miR167c expression was up-regulated in the vasculature, pericycle, and cortex of soybean roots following inoculation with Bradyrhizobium japonicum strain USDA110 (the microsymbiont). It was found to positively regulate nodule numbers directly by repressing the target genes GmARF8a and GmARF8b (homologous genes of Arabidopsis [Arabidopsis thaliana] AtARF8 that encode auxin response factors). Moreover, the expression of miR167 and its targets was up- and down-regulated by auxin, respectively. The miR167-GmARF8 module also positively regulated nodulation efficiency under low microsymbiont density, a condition often associated with environmental stress. The regulatory role of miR167 on nodule initiation was dependent on the Nod factor receptor GmNFR1α, and it acts upstream of the nodulation-associated genes NODULE INCEPTION, NODULATION SIGNALING PATHWAY1, EARLY NODULIN40-1, NF-YA1 (previously known as HAEM ACTIVATOR PROTEIN2-1), and NF-YA2. miR167 also promoted lateral root numbers. Collectively, our findings establish a key role for the miR167-GmARF8 module in auxin-mediated nodule and lateral root formation in soybean.
Legume plants can form unique root lateral organs, called nodules, through symbiosis with a nitrogen-fixing bacterial microsymbiont (broadly called rhizobia). Inside the nodule, the rhizobia differentiate into bacteroids that convert atmospheric dinitrogen gas into ammonium, which can be used as a nitrogen source by the host plant. In turn, the host plant provides energy and resources for the rhizobia. Thus, this relationship is of mutual benefit to both symbiotic partners, particularly in low-nitrate conditions.
Nodule formation is triggered by compatible rhizobia-secreted lipooligosaccharide signals, called Nod factors (NFs). NFs integrate plant developmental pathways to activate the proliferation of root pericycle and cortical cells underlying the infection site opposite xylem poles to form nodule primordia (Timmers et al., 1999; Ferguson et al., 2010; Desbrosses and Stougaard, 2011; Oldroyd, 2013). With the concurrent entrance of rhizobia via curled root hairs (or epidermal cracks as in peanut [Arachis hypogaea]) into the nodule primordia, nitrogen-fixing nodules are formed (Crespi and Frugier, 2008; Gage, 2004). It is apparent that nodulation is a complex process involving interaction between rhizobia and host plants, integrating cross talk between plant defense mechanisms and developmental responses to nitrogen starvation. Phytohormones, such as cytokinin and auxin, have long been speculated to have a role in orchestrating nodule development (Libbenga and Torrey, 1973), with more recent studies demonstrating that they are essential for cell proliferation and differentiation during nodule organogenesis (Gonzalez-Rizzo et al., 2006; Madsen et al., 2010; Suzaki et al., 2012; for review, see Ferguson and Mathesius, 2014).
In recent decades, genetic analyses of numerous nodulation mutants of the legume model plants Medicago truncatula, Lotus japonicus, and soybean (Glycine max) have led to the identification of key components in the NF-activated signaling pathway (for review, see Crespi and Frugier, 2008; Ferguson et al., 2010; Desbrosses and Stougaard, 2011; Oldroyd, 2013). NFs are perceived by Lysin motif receptor-like kinases (called GmNFR1 and GmNFR5 in soybean; Indrasumunar et al., 2010, 2011). After the perception of NFs, a signaling cascade is activated, including ion efflux and calcium spiking events, which are decoded by calcium- and calmodulin-dependent protein kinase (Lévy et al., 2004). Components acting downstream in the NF signaling pathway include a number of transcription factors, such as NODULE INCEPTION (NIN; Schauser et al., 1999), NODULATION SIGNALING PATHWAY1 (NSP1) and NSP2 (Hirsch et al., 2009), NF-YA1 (previously known as HAEM ACTIVATOR PROTEIN2-1), NF-YB1, and NF-YA2 (Combier et al., 2006; Soyano et al., 2013; Laporte et al., 2014), in addition to a number of early nodulin genes, such as ENOD40 (Kouchi and Hata, 1993). Orthologs of genes encoding for many of these components were recently identified in soybean (Hayashi et al., 2012).
Legume plants also have an internal mechanism, termed autoregulation of nodulation (AON), that acts to control the number of nodules they form (Gresshoff and Delves, 1986; Caetano-Anollés and Gresshoff, 1991). The process involves a Leu-rich repeat receptor kinase (Krusell et al., 2002; Nishimura et al., 2002; Searle et al., 2003; Schnabel et al., 2005) that functions in the shoot to perceive root-derived CLAVATA3/embryo-surrounding region (CLE) peptides produced following microsymbiont infection (Okamoto et al., 2009; Reid et al., 2011; Mortier et al., 2010; for review, see Hastwell et al., 2015). Perception of the CLE peptides triggers the production of a shoot-derived inhibitor signal that is transported back to the roots to prevent continued nodule development.
Interestingly, the shoot-derived inhibitor signal was recently postulated to be a shoot-derived 2-isopentenyl adenine cytokinin (Sasaki et al., 2014). This would indicate that cytokinins have a dual role in nodulation, acting positively in nodule development (Gonzalez-Rizzo et al., 2006; Plet et al., 2011; Ariel et al., 2012; Held et al., 2014; for review, see Ferguson and Mathesius, 2014) and negatively through AON (Sasaki et al., 2014). The phytohormone auxin may also play a dual role based on the exogenous application of auxin or auxin transporter inhibitors and on the auxin accumulation/auxin responsiveness at the site of nodule initiation (Allen et al., 1953; Hirsch et al., 1989; Mathesius et al., 1998; Boot et al., 1999; Pacios-Bras et al., 2003; Takanashi et al., 2011; Turner et al., 2013; for review, see Ferguson and Mathesius, 2014). In addition, genetic evidence demonstrates that reduced polar auxin transport from the shoot to the rhizobia-infected root can act to inhibit further nodule formation (Subramanian et al., 2006; van Noorden et al., 2006; Wasson et al., 2006; Zhang et al., 2009).
Interplay also exists between cytokinin and auxin in the regulation of legume nodule development. Recent evidence using L. japonicus has shown that NF induces cytokinin biosynthesis, which triggers the production of the transcription factor NIN, which then positively regulates auxin accumulation in dividing cortical cells for nodule formation (Suzaki et al., 2012). In soybean, the auxin maximum concentration is in the apex of the nodule primordium, whereas in mature nodules, auxin accumulation is mainly at the nodule periphery (Turner et al., 2013). Interestingly, NIN also directly binds to the promoters of the nodulation-suppressive CLE peptide-encoding genes to activate their expression, leading to the systemic suppression of nodulation via AON and, thus, the reduced expression of NIN and decreased auxin accumulation in dividing cortical cells (Soyano et al., 2013). However, it remains unclear how legumes specifically integrate cytokinin and auxin signaling to modulate the formation and control of nodules.
MicroRNAs (miRNAs) are small, noncoding RNA molecules that negatively regulate the expression of other genes at the posttranscriptional level by cleaving their RNA transcripts in a sequence-specific manner (Hutvágner and Zamore, 2002; Llave et al., 2002). Several profiling studies have revealed that miRNAs are involved in rhizobial infection, nodule formation, and nodule functioning (Subramanian et al., 2008; Lelandais-Brière et al., 2009; Wang et al., 2009; Bazin et al., 2012; De Luis et al., 2012; Dong et al., 2013; for review, see Formey et al., 2014). Indeed, functional analyses have demonstrated an altered number of nodules following miRNA overexpression or knockdown (Combier et al., 2006; Li et al., 2010; De Luis et al., 2012; Turner et al., 2013; Yan et al., 2013; Wang et al., 2014). A recent study reported that ectopic expression of miR160, which negatively regulates AUXIN RESPONSE FACTOR10 (ARF10), increased auxin sensitivity and inhibited nodule development. This suggests that minimal or reduced auxin sensitivity of soybean root cells is required for determinate nodule development (Turner et al., 2013). However, ectopic overexpression of miR393, which targets auxin receptors, influenced indeterminate, but not determinate, nodule development (Mao et al., 2013; Turner et al., 2013) and may implicate a diverse role of auxin in nodulation among different legume species. Previously, a relatively high level of miR167 expression was detected in nitrogen-fixing nodules of soybean (Wang et al., 2009; De Luis et al., 2012), which suggests the possibility of a miR167-mediated regulatory mechanism underlying beneficial soybean-rhizobia symbiosis. In Arabidopsis (Arabidopsis thaliana), the miR167-ARF8 module has been shown to regulate the cell type-specific response to available nitrogen status and the plastic development of lateral roots (Gifford et al., 2008). miR167 also regulates plant development and root plastic development through targeting different genes, such as ARF6 and INDOLE-3-ACETIC ACID-ALANINE RESISTANT3 (IAR3; Ru et al., 2006; Kinoshita et al., 2012). Whether a similar module acts in the development and/or regulation of legume nodules has not been reported previously.
In this study, we analyzed the role of miR167 in nodulation in response to Bradyrhizobium japonicum infection. We found that miR167 was up-regulated in roots after inoculation with B. japonicum and controlled nodulation through two targets, which are homologs of Arabidopsis ARF8 (GmARF8a and GmARF8b). We also established that the miR167-GmARF8 module regulated nodulation efficiency under low microsymbiont density, which is often associated with environmental stress. miR167 and the two target genes were responsive to auxin and NF application. Intriguingly, the function of the miR167-GmARF8 module in nodulation was dependent on NF perception, as demonstrated through the NF receptor gene GmNFR1. Our findings highlight a functional role for the miR167/GmARF8 module in nodulation and further demonstrate a critical role for auxin signaling in legume nodule development. Additionally, we showed that soybean miR167 also targets putative GmARF6 genes but not GmIAR3 genes to modulate plant growth and development, such as lateral root development. Collectively, these results help to establish the mechanism by which legume plants respond to auxin to activate nodule initiation, including the functional integration of auxin- and NF-dependent signaling in nodule organogenesis.
RESULTS
Expression Pattern of miR167
The miR167 family of soybean is composed of 11 highly conserved homologous miRNAs (miR167a–miR167k; Supplemental Fig. S1A). These miR167 preliminary miRNAs (premiRNAs) are generated from different genes, and the lengths of the majority of premiRNAs are similar (78–280 nucleotides), with the exception of miR167c, which has the longest precursor (375 nucleotides; Supplemental Fig. S1A). The phylogenetic analysis of their premiRNAs showed that the miR167 family members can be divided into two subgroups: miR167a/b/d/e/f/g/h/i/j/k and miR167c (Supplemental Fig. S1B). Unexpectedly, soybean miR167c but not miR167a has the closest genetic relationship with Arabidopsis miR167a. These premiRNAs eventually produce highly conserved mature miR167s that have only one or two nucleotide differences located at the 3′ and/or 5′ end of their mature miRNAs (Supplemental Fig. S2).
Solexa sequencing and quantitative real-time (qRT)-PCR analysis revealed that the miR167 members are expressed at different levels in mature soybean nodules (Supplemental Fig. S3). Among the miR167 members, miR167c and miR167j exhibited the highest levels of expression, followed by miR167a, miR167b, miR167d, miR167e, miR167f, and miR167k, which showed the lowest level in mature nodule; miR167g and miR167h were not detected. These results suggest that miR167 family members may differentially mediate soybean nodulation and nodule functionality. Because of the high expression of soybean miR167c and its close relationship with Arabidopsis miR167a, which mediates cell-specific nitrogen response and root plastic development (Gifford et al., 2008), miR167c was the subject of our continued investigation.
Within the 2-kb promoter regions located directly upstream of miR167c, there are many cis-regulatory elements that function as binding sites for transcription factors in hormonal signaling pathways (e.g. auxin and cytokinin) and nodulation. Notably, seven auxin-responsive cis-elements (NTBBF1ARROLB and AUXREPSIAA4)/auxin response factor-binding sites (ARFAT) and 11 nodulin consensus sequence motifs (NODCON1GM and NODCON2GM) were identified (Supplemental Fig. S4). This suggests that miR167c transcription may be regulated by both auxin and rhizobia infection. Using a comprehensive statistical analysis of these cis-elements in miR167 family members (Suzuki et al., 2005), we found that all the miR167 family members contain the nodulin consensus sequence motifs and NTBBF1ARROLB, and the majority of them have the cis-element of ARFAT and AUXREPSIAA4 (Supplemental Table S1), indicating the conserved roles of the miR167 family members in auxin response and nodulation. qRT-PCR showed that miR167c expression was significantly elevated in soybean roots treated with the synthetic auxin 2,4-dichlorophenoxyacetic acid (2,4-D; Fig. 1A) or inoculated with the compatible B. japonicum strain USDA110. Transcript levels were significantly increased by 1 h after inoculation (HAI) and reached their highest level by 6 HAI (Fig. 1B). The levels then decreased somewhat by 12 and 24 HAI but gradually increased at later stages of nodule development (i.e. 3, 5, 10, and 28 DAI; Fig. 1C).
Expression of miR167c following auxin treatment or B. japonicum inoculation. A, miR167c expression in roots treated with 0, 0.2, or 1 μm 2,4-D for 3 d. B and C, miR167c expression in roots inoculated with B. japonicum 0, 1, 3, 6, 12, and 24 HAI (B) and 0, 1, 3, 5, 10, and 28 DAI (C). Expression levels were normalized against the geometric mean of the soybean reference gene miR1520d. All experiments consisted of three independent biological replicates. Error bars indicate sd. Different letters indicate significant differences (Student-Newman-Kuels test, P < 0.05). D and E, GUS activity in uninoculated transgenic miR167cpro::GUS roots (D; bar = 300 μm) and root cross sections (E; bar = 100 μm) grown for 7 d. F to K, GUS activity in similar B. japonicum-inoculated transgenic roots (F; 10 DAI; bar = 300 μm) and root cross sections (G; 10 DAI; bar = 100 μm) in addition to nodulated roots (H; 10 DAI; bar = 600 μm and J; 28 DAI; bar = 1 mm) and nodule cross sections (I; 10 DAI, bar = 200 μm and K; 28 DAI; bar = 700 μm).
To visualize miR167c expression in soybean roots and nodules, the native promoter of miR167c was used to drive the expression of the GUS reporter gene (miR167cpro::GUS) following Agrobacterium rhizogenes-mediated transformation (Kereszt et al., 2007; Jian et al., 2009). Since the GUS gene under different lengths of promoters (2 kb, 1 kb, and 654 bp) that contain the essential cis-regulatory elements like the ARFAT cis-element produced similar expression patterns (data not shown), the shortest promoter construct was used for further analysis. When grown in low-nitrate conditions, GUS was specifically expressed in the pericycle and vascular tissues (Fig. 1, D and E). Following inoculation with B. japonicum, GUS expression was highly abundant in the vascular tissues and pericycle, in addition to cortical cells located close to the endodermis and pericycle (Fig. 1, F and G). During early nodule development, GUS expression extended further into the vascular tissues, cortex, and epidermal cells of the roots (Fig. 1H) and then gradually decreased and became absent at later stages of nodule maturation (Fig. 1J). Within the nodule, GUS was highly expressed in the infected bacteroid zone (Fig. 1, I and K).
miR167 Controls Lateral Root Architecture
To help determine the role of miR167 in lateral root development, transgenic roots overexpressing miR167c were generated. qRT-PCR was used to confirm the overexpression of miR167c in the transgenic roots compared with empty vector controls roots (Supplemental Fig. S5). miR167c-overexpressing roots exhibited significantly different root architecture than the empty vector control roots 10 d after the explants were transplanted (Fig. 2, A–D). Root lengths were comparable between the two root types (Fig. 2B); however, the total lateral root number and the lateral root length of miR167c-overexpressing roots were significantly increased compared with vector control roots (Fig. 2, C and D).
Effects of overexpression or knockdown of miR167c on root system architecture. A, Phenotypes of transgenic empty vector (EV) and 35S::miR167c roots 10 d after transplantation. Bar = 1 cm. B to D, Primary root length (B), lateral root number (C), and total lateral root length (D) per hairy root expressing empty vector and 35S::miR167c. E, Phenotypes of transgenic empty vector, MIM167, and STTM167-88 roots. Bar = 1 cm. F to H, Primary root length (F), lateral root number (G), and total lateral root length (H) per hairy root expressing empty vector, MIM167, and STTM167-88. I and J, Root length (I) and lateral root number (J) of transgenic empty vector and 35S::miR167c roots in response to treatment with 0, 0.2, or 1 μm 2,4-D 10 d after transplantation. Statistically significant differences (Student’s t test) are indicated as follows: *, P < 0.05; ***, P < 0.001; and ns, not significant, P > 0.05. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
To determine whether endogenous miR167 is required for lateral root development, artificial target mimics that compete with endogenous miR167 to reduce their activity were used (Ebert et al., 2007; Todesco et al., 2010). According to the INDUCED BY PHOSPHATE STARVATION1 (IPS1)-based target mimic (Todesco et al., 2010), the miR399-complementary motif of IPS1 was modified to mimic target sites for miR167c (MIM167). A construct containing a short tandem target mimic (STTM167-88) was also generated, which harbors two copies of miR167c partially complementary sequences linked by an 88-bp short spacer (Supplemental Fig. S6; Yan et al., 2012). qRT-PCR analysis confirmed that MIM167 and STTM167-88 were successfully overexpressed in transgenic hairy roots (Supplemental Fig. S7, A and B). Accordingly, the transcript levels of all miR167 family members were significantly reduced, although the rates of reduction in their expression varied (Supplemental Fig. S8). As expected, the reduced miR167 levels, and hence activity, resulted in substantial decreases in both total lateral root numbers and length compared with empty vector control roots (Fig. 2, E, G, and H). Interestingly, the root length of the miR167 knockdown lines was also shorter than that of the vector controls (Fig. 2F). The miR167c overexpression and miR167 knockdown roots also exhibited reduced and enhanced sensitivity to auxin, respectively (Fig. 2, I and J). These data imply a crucial role of miR167 in controlling lateral root development and auxin-mediated root system architecture.
miR167 Positively Regulates Nodulation
Transgenic soybean roots overexpressing miR167c (35S::miR167c) were also used to evaluate nodule numbers 28 d after B. japonicum inoculation (Supplemental Fig. S9). Overexpression resulted in a significant (more than 2-fold) increase in the number of nodules compared with empty vector control roots (note that individual roots were scored to ensure the transgenic nature of the individual root; Fig. 3A). The average number of nodules per root overexpressing miR167c exceeded 30; in contrast, empty vector control roots had an average of 12 nodules per root (Fig. 3B). These data suggest that miR167c stimulates soybean nodulation.
Effects of overexpression or knockdown of miR167c on nodulation. A, Nodules of transgenic empty vector (EV) and 35S::miR167c roots 28 DAI. Bar = 1.5 mm. B, Nodule number of transgenic empty vector and 35S::miR167c roots 28 DAI. C, miR167c transcript levels in transgenic empty vector, MIM167, and STTM167-88 roots. D, Nodules of transgenic empty vector, MIM167, and STTM167-88 roots 28 DAI. Bar = 4 mm. E, Nodule number per transgenic empty vector, MIM167, and STTM167-88 roots 28 DAI. miR167c expression was normalized against the geometric mean of the soybean reference gene miR1520d. MIM167 and STTM167-88 expression was normalized against the geometric mean of the soybean reference gene elongation factor1-β (ELF1b). Statistically significant differences (Student’s t test) are indicated as follows: ***, P < 0.001. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
A phenotypic analysis of transgenic hairy roots expressing MIM167 and STTM167-88 was conducted to further delineate the role of miR167 in nodulation (Supplemental Fig. S10, A and B). MIM167- and STTM167-88-expressing roots had a significantly reduced level of miR167c transcripts (Fig. 3C). These miR167-knockdown roots formed significantly fewer nodules compared with empty vector control roots at 28 d after B. japonicum inoculation (Fig. 3, D and E). This further suggests that miR167 functions as a positive regulator of soybean nodulation.
miR167 Functions Downstream of the NF Receptor GmNFR1α and Upstream of Multiple Early Nodulation Genes
To determine whether miR167 regulates nodulation through the NF signaling pathway, miR167c expression was determined in roots of the nonnodulating soybean mutant nod49, which carries a defect in GmNFR1α (Indrasumunar et al., 2010). miR167c expression was induced by B. japonicum inoculation in wild-type roots (Fig. 4A); however, this up-regulation was diminished in mutant nod49 roots (Fig. 4B). This suggests that miR167c up-regulation following inoculation is dependent on GmNFR1α. We also tested whether miR167c expression was induced by NF treatment and found that the miR167c transcript level was significantly elevated by NF application in soybean roots. However, the extent of its induction was much lower than that of the early nodulin gene ENOD40 (Fig. 4C), which has previously been shown to be induced by NF (Minami et al., 1996).
Expression of miR167c and various early nodulation genes. A and B, miR167c expression in roots of wild-type cv Bragg (A) and the GmNFR1α mutant nod49 (B) 0, 1, 3, 6, 12, and 24 HAI with B. japonicum. C, miR167c expression in response to NF (10−8 m) treatment, where GmENOD40 was used as a positive control. D to M, Expression of symbiotic genes (ENOD40, NIN, NSP1, HAP2-1, and HAP2-2) in transgenic empty vector (EV) and 35S::miR167c roots (D–H) or MIM167 or STTM167-88 roots (I–M) 28 DAI with B. japonicum. All experiments were conducted three times. Error bars indicate sd. Statistically significant differences (Student’s t test) are indicated as follows: *, P < 0.05; and ***, P < 0.001. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
We then examined the expression of a number of nodulation marker genes (ENOD40, NIN, NSP1, HAP2-1, and HAP2-2) in miR167c overexpression and miR167 knockdown roots during nodulation. The expression of all tested genes in miR167c-overexpressing roots was significantly increased compared with that in empty vector control roots at 10 DAI (Fig. 4, D–H). In contrast, the transcript levels of these genes in miR167 knockdown roots (MIM167 and STTM167-88) were markedly reduced compared with those of the empty vector control roots (Fig. 4, I–M). These results suggest that miR167 acts upstream of these symbiotically related genes in controlling rhizobial infection and nodule development.
GmARF6 and GmARF8 But Not GmIAR3 Are Target Genes of miR167
In Arabidopsis, ARF6, ARF8, and IAR3 have been validated as the targets of miR167 (Ru et al., 2006; Kinoshita et al., 2012). This led us to investigate whether soybean miR167 similarly targets their homolog genes GmARF6, GmARF8, and GmIAR3 in soybean. To this end, bioinformatic analyses were performed to identify two putative orthologs of Arabidopsis ARF6 (At1g30330) in soybean (glyma05g27580 and glyma08g10550); three putative orthologs of Arabidopsis ARF8 (At5g37020) in soybean (glyma02g40650, glyma14g38940, and glyma18g05330); and four putative orthologs of Arabidopsis IAR3 (At1g51760) in soybean (glyma13g42880, glyma15g02560, glyma08g21030, and glyma07g01570; Supplemental Fig. S11). These genes were named GmARF6a and GmARF6b; GmARF8a, GmARF8b, and GmARF8c; and GmIAR3a, GmIAR3b, GmIAR3c, and GmIAR3d, respectively. Each putative GmARF6 and GmARF8 gene contains only one miR167-binding site in its coding sequence with nearly perfect match sequences; by contrast, all GmIAR3 homolog genes do not contain the complementary sequences to miR167 because of high mismatches between GmIAR3 and miR167 sequences (Supplemental Fig. S12), suggesting that GmIAR3 genes are not target genes of miR167 in soybean.
To test whether GmARF6, GmARF8, and GmIAR3 genes are cleaved by miR167 in soybean, we first performed 5′ RACE PCR. Of four GmIAR3 genes, GmIAR3a and GmIAR3b containing sequences with better match to miR167 were used in the experiments. The sequencing results showed that transcripts of GmARF6a/b and GmARF8a/b/c were indeed cleaved by miR167 at the site between the 10th and 11th nucleotides from the 5′ end of miR167, but GmIAR3a and GmIAR3b could not be cleaved by miR167 (Fig. 5A; Supplemental Figs. S13 and S14). Further expression analysis showed that the reduced miR167 activity in the MIM167c and STTM167-88 transgenic knockdown roots resulted in significant up-regulation of GmARF6b, GmARF8a, and GmARF8b transcripts compared with that of control roots (Fig. 5, B–F). Notably, GmARF8a was highly up-regulated in both MIM167c and STTM167-88 transgenic knockdown roots (Fig. 5D), but GmARF6a and GmARF8c transcript levels unexpectedly were even down-regulated by miR167 knockdown. These results suggest that GmARF8a, GmARF8b, and GmARF6b function as direct target genes of miR167, and of them, GmARF8a may be the major target of miR167.
Validation of miR167c target genes. A, Target validation by 5′ RACE-PCR. B and C, GmARF6a and GmARF6b transcript levels in transgenic empty vector (EV), MIM167, and STTM167-88 roots. D to F, GmARF8a, GmARF8b, and GmARF8c transcript levels in transgenic empty vector, MIM167, and STTM167-88 roots. Expression levels were normalized against the geometric mean of the soybean reference gene GmELF1b. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
GmARF8a and GmARF8b Are Negative Regulators of Nodulation
Because ARF8 mRNA is a target of miR167 to mediate lateral root development in response to low nitrate (Gifford et al., 2008; Liang et al., 2012), we chose GmARF8a and GmARF8b as examples for further functional study. Given that miR167c is up-regulated during nodulation (Fig. 1), transcripts of its target genes were expected to be down-regulated following auxin treatment or rhizobia inoculation. Several auxin-responsive elements and nodulin consensus sequences were identified in the promoters of GmARF8a and GmARF8b, although the number of these elements differed between the two genes (Supplemental Fig. S15). Notably, these auxin and nodulin consensus cis-elements appeared on the promoters of all the validated GmARF6 and GmARF8 target genes (Supplemental Table S2) based on the quantitative statistical analysis of these elements (Suzuki et al., 2005), indicating that they are involved in an auxin-mediated process and plant-rhizobia interaction. As expected, the expression of both GmARF8a and GmARF8b was suppressed by 2,4-D treatment (Fig. 6A) or B. japonicum inoculation (Fig. 6B). Their expression remained reduced during the course of nodule development (Fig. 6C). In general, the expression of GmARF8a and GmARF8b was opposite to that of miR167c, which implies that miR167c may regulate nodulation in part by directly repressing GmARF8a and GmARF8b.
GmARF8a and GmARF8b negatively regulate nodule formation. A, Expression of GmARF8a and GmARF8b in roots treated with 0, 0.2, and 1 μm 2,4-D for 3 d. B and C, Expression of GmARF8a and GmARF8b in roots 0, 1, 3, 6, 12, and 24 HAI (B) or 0, 1, 3, 5, 10, and 28 DAI (C) with B. japonicum. D, Expression of GmARF8a and GmARF8b in transgenic empty vector (EV) and RNAi-GmARF8 roots. E, Nodules of transgenic empty vector and RNAi-GmARF8 roots 10 DAI. Bar = 1 mm. F, Nodule number per transgenic empty vector and RNAi-GmARF8 roots. G, qRT-PCR analysis of GmARF8 transcript levels in the transgenic roots expressing empty vector, 35S::GmARF8a, and 35S::GmARF8a m7. H, Nodules of representative roots overexpressing 35S::GmARF8a and 35S::GmARF8a m7. Photographs were taken at 28 DAI. Bar = 4 mm. I, Total number of nodules per transgenic hairy root expressing 35S::GmARF8a and 35S::GmARF8a m7. The nodule number of roots transformed with the empty vector was used as the control. Expression levels were normalized to the soybean reference gene GmELF1b. All experiments were conducted three times. Error bars indicate sd. Statistically significant differences (Student’s t test) are indicated as follows: **, P < 0.01; and ***, P < 0.001. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
To test this hypothesis, RNA interference (RNAi) was used to generate double-knockdown transgenic hairy roots of GmARF8a and GmARF8b. qRT-PCR confirmed that the transcript levels of GmARF8a and GmARF8b were markedly reduced in the transgenic roots compared with empty vector control roots (Fig. 6D). These double-knockdown roots produced 28 nodules per root, whereas the empty vector control roots produced only 13 nodules per root (Fig. 6, E and F). This result supports our previous finding of a role for the miR167 target genes, GmARF8a and GmARF8b, in nodule development.
To confirm whether GmARF8a is the direct target of miR167c in nodulation, we created seven translational silent mutations in the miR167 target sites of the GmARF8a coding sequence (35S::GmARF8a m7) to interrupt the access of miR167c. qRT-PCR analysis confirmed that overexpression of the mutated GmARF8a caused a higher transcript level of GmARF8a (Fig. 6G). GmARF8a m7-overexpressing roots produced a lower number of nodules than the vector control, but total nodule number per GmARF8a m7-overexpressing root was similar to that of GmARF8a-overexpressing roots (Fig. 6, H and I). These results demonstrate that miR167 regulates nodule number through direct cleavage of GmARF8a.
The miR167-GmARF8 Module Regulates Nodulation Efficiency under Low-Rhizobia Titer
The rhizobia population in soil is considered a key determinant of legume nodulation (Thies et al., 1991; Zahran, 1999). To further validate the functional role of miR167 in nodule formation efficiency, we inoculated roots with a low titer (102 CFU mL−1) of B. japonicum. miR167c-overexpressing roots formed roughly 25 nodules per root (Fig. 7A), whereas empty vector control roots produced only two to four nodules per root (Fig. 7B). Moreover, miR167c-overexpressing roots formed only slightly more nodules when inoculated with a more customary titer of B. japonicum (108 CFU mL−1; Fig. 7C). These findings suggest that miR167c can enhance symbiotic nodulation efficiency at suboptimal conditions.
The miR167-GmARF8 module regulates nodulation efficiency. A and B, Expression of miR167c (A) and nodulation phenotype (B) of transgenic empty vector (EV) and 35S::miR167c roots 28 DAI with ultra low B. japonicum titers (102 colony-forming units [CFU] mL−1; LDR). Bar in B = 1 mm. C, Nodule number per transgenic empty vector and 35S::miR167c roots. SDR, Standard B. japonicum titers (107 CFU mL−1). D and E, Expression of GmARF8a and GmARF8b (D) and nodulation phenotype (E) of transgenic empty vector and RNAi-GmARF8 roots 10 DAI with ultra low B. japonicum titers (102 CFU mL−1). Bar in E = 700 μm. F, Nodule number per transgenic empty vector and RNAi-GmARF8 roots. Expression of miR167c, and of GmARF8a and GmARF8b, was normalized to the soybean reference genes, miR1520d and GmELF1b, respectively. All experiments were conducted three times. Error bars indicate sd. Statistically significant differences (Student’s t test) are indicated as follows: **, P < 0.01. Different letters indicate significant differences by Student-Newman-Kuels test (P < 0.05).
To establish whether GmARF8a and GmARF8b function as part of the nodulation efficiency control mechanism mediated by miR167, the nodule number of GmARF8a/b double knockdown roots was determined in the presence of low B. japonicum titer. As expected, these roots formed significantly more nodules than empty vector control roots (23 versus two; Figure 7, D–F). These results are strikingly similar to those of the miR167c-overexpressing lines inoculated with a low titer of B. japonicum, suggesting that miR167 may regulate microsymbiont infection efficiency via its targets, GmARF8a and GmARF8b.
DISCUSSION
We identified soybean miR167-GmARF8 as a key regulatory module of root nodule and lateral root development. In Arabidopsis, miR167 and ARF8 are expressed in the pericycle of roots, where they mediate a pericycle-specific response to nitrate treatment and subsequent development of lateral roots (Gifford et al., 2008). We show that miR167c of soybean is also highly expressed in pericycle and vascular tissues of uninfected roots and exhibits the cell type-specific expression pattern required for lateral root development. Thus, to our knowledge, miR167 represents the first noncoding miRNA that mediates cell type responses to nitrogen and rhizobial infection and root system architecture control in plants.
Rhizobia infection induced the expression of miR167c in inner cortex cells of soybean roots, where root nodules originate, which is regulated by NF receptors. Indeed, the specific expression pattern of miR167c is consistent with it having a positive role in both lateral root and nodule formation. miR167c levels are tightly regulated spatially and temporally during nodulation and root development. Overexpression of miR167c did not affect root growth, but its knockdown resulted in shortened root length. The inconsistent phenotypes indicate that, in addition to miR167c, other miR167 family members may regulate soybean root growth, as current IPS and STTM technologies cannot specifically knock down miR167c. The fact that all the miR167 family members contain similar auxin-responsive cis-elements and nodulin consensus motifs in their promoters supports that they may function coordinately to fine-tune plant development, including the root system, and reproductive growth as well, as in other plants (Liu et al., 2014).
In Arabidopsis, miR167 controls lateral root development through the suppression of ARF8, an activator of auxin signaling (Ulmasov et al., 1999), in response to nitrate availability (Gifford et al., 2008). Here, we provide molecular and functional evidence demonstrating that GmARF8a and GmARF8b, which are putative orthologs of AtARF8 in Arabidopsis, are targets of miR167 in soybean. GmARF8a and GmARF8b encode proteins with high sequence similarity to the Arabidopsis gene, indicating that they may have similar functions to the auxin signaling activator. Our results demonstrate that miR167 and its target genes are responsive to auxin and that overexpression of miR167c can reduce auxin sensitivity in soybean roots. This suggests that miR167 may control nodule formation and root architecture by suppressing auxin signaling, supporting the notion that attenuated auxin signaling favors nodulation (Mathesius, 2008; Turner et al., 2013).
The expression of GmARF8a and GmARF8b was tightly regulated by miR167c, and silencing of GmARF8a and GmARF8b was required for optimum nodule formation. Thus, our work demonstrates the existence of a miR167-GmARF8 module that mediates a regulatory mechanism of nodulation and root system architecture in soybean. Although both GmARF8 genes were repressed to regulate nodulation, and knockdown of the genes exhibited a similar result, the two GmARF8 genes might not function equally during nodule and root development. Stronger repression and nodule phenotypes of GmARF8a overexpression and overexpression of the GmARF8a-resistant version to miR167 cleavage favor the notion that GmARF8a may act as a major target gene of miR167 in nodulation. Moreover, due to imperfect target recognition for miRNAs, we cannot exclude the possibility that additional targets may exist to regulate miR167-mediated nodulation and root system architecture in soybean. These putative targets include the soybean homologs of Arabidopsis ARF6 genes, because GmARF6a/b genes can be cleaved by miR167c, and further functional analysis of the GmARF6 genes will uncover their roles in soybean nodulation. These results prove that miR167-GmARF6/8 modules may be a highly conserved regulatory mechanism for the auxin-mediated biological processes during plant organ or de novo organ development. Notably, although the IAR3 gene has been proved to be a direct target of miR167a in Arabidopsis root plastic development in response to osmotic stress (Kinoshita et al., 2012), it is apparent that GmIAR3 encoding soybean homologs of the Arabidopsis IAR3 were not targeted by miR167 in soybean. These results indicate that the homologous IAR3 nucleic acid sequences, including the putative cleavage sites for miR167 and regulation of the IAR3 genes, are evolutionarily species diverged.
In Arabidopsis, miR167 and miR160 have opposing roles in regulating lateral root development in response to low nitrogen (Gifford et al., 2008; Liang et al., 2012). In soybean, previous studies have shown that miR160 acts as a positive regulator of auxin signaling to negatively regulate nodule formation (Turner et al., 2013). Our findings demonstrate the opposite role for miR167, suggesting that two miRNAs may function antagonistically to regulate nodule formation and root system architecture. The contrasting expression patterns of an array of nodulation genes (NIN, NSP1, ENOD40, HAP2-1, and HAP2-2) in soybean roots having altered transcript levels of miR167c (shown here) and miR160 (Turner et al., 2013) support this hypothesis.
Low Rhizobium spp. population density is a major constraint of optimal nodule formation and related nitrogen fixation and yield (Herridge et al., 1987; Thies et al., 1991). The application of rhizobia has become a routine technique for soybean cultivation in fields with low Rhizobium spp. populations. However, it is not yet known how nodulation efficiency under low Rhizobium spp. population is controlled. We have shown that miR167c-GmARF8 acts as a key module in controlling rhizobial infection and nodulation efficiency under a low soil Rhizobium spp. population. Increased transcript levels of miR167c or reduced levels of GmARF8 enhanced soybean nodulation. Since miR167c targets auxin response factors, the level of auxin and/or auxin sensitivity in root cells may be key in determining effective nodulation under low Rhizobium spp. density. Moreover, miR167c-GmARF8 may be a possible target for the genetic improvement of nodulation efficiency under low Rhizobium spp. populations.
In summary, our findings indicate that miR167 is a pivotal gene that mediates nodule formation and development via its target GmARF8 genes, which are important for an active cellular response to local auxin accumulation in root cortical cells. It is likely that rhizobia inoculation or NF treatment induces miR167 expression to repress GmARF8 target genes. This would subsequently attenuate auxin sensitivity in dividing root cortical cells while promoting nodule development through downstream signals in the NF signaling pathway (Fig. 8). The findings reported here shed light on the role of a critical miRNA in plant development and highlight its manipulation of auxin signaling to regulate nodule organogenesis and root system architecture. Clearly, our findings are not consistent with the previous research on the minor role of auxin signaling in determinant nodule development (Mao et al., 2013). A possible explanation for this contradictory result of auxin effects on determinant nodule development is that the particular miR393 family member used in the previous study may not be the one that plays a critical role in the development of determinant nodules. Based on our observation, the miR393 family contains many members that are usually differentially expressed during soybean nodulation (Z. Cai and X. Li, unpublished data). Further systematic analysis of miR393 and their targets will help to clarify the important role of auxin in determinant nodule development. Furthermore, it is worthy to note that the cis-regulatory elements responsive to other hormones, such as cytokinin, are also identified in the promoter of miR167c, indicating that it is likely to be regulated by other hormones, such as cytokinin. Therefore, we do not exclude the possibility that miR167 and the target genes also mediate multiple hormonal responses during nodulation.
A proposed model of the miR167-GmARF8 module that regulates nodulation. When rhizobia are absent, miR167 members are maintained at a low level and their targets, such as GmARF8, are able to activate the transcription of auxin-responsive genes to inhibit root nodulation. In the presence of rhizobia, miR167 members are induced and inhibit GmARF8 activity, thus inhibiting the transcription of auxin-responsive genes to promote nodulation.
MATERIALS AND METHODS
Plant and Bradyrhizobiumjaponicum Growth Conditions
Soybean (Glycine max) ‘Williams 82’ was used for cloning miRNAs, 5′ RACE to validate target genes, and functional analysis of miR167c and its gene targets. cv Bragg and its isogenic mutant line, nod49, containing a mutation in GmNFR1α, were used for miR167c expression studies associated with the NF receptor. B. japonicum strain USDA110 was used for all rhizobia inoculation, and associated nodulation, studies. Plant and B. japonicum growth conditions were similar to those reported by Wang et al. (2009). For gene expression analysis in response to B. japonicum inoculation, different soybean tissues were collected at various times following inoculation (30 mL per pot) with a suspension of B. japonicum USDA110 (107 CFU mL−1, unless stated otherwise). For RNA extraction studies, seedlings were rinsed briefly in phosphate-buffered saline (pH 7.5) to remove vermiculite and perlite particles. Harvested tissues were immediately frozen in liquid nitrogen and stored at −80°C. The Solexa sequencing technology method has been described by Dong et al. (2013).
Auxin Response Assays
Composite plants (empty vector control or 35S::miR167c) were transplanted into pots (13 cm × 10 cm × 8.5 cm) containing vermiculite and perlite (3:1) and watered with low-nitrate solution (Wang et al., 2009) with or without 0.2 or 1 μm 2,4-D. Ten days later, the root length, lateral root number, and total lateral root length per hairy root were measured. To determine the expression of miR167c, GmARF8a, and GmARF8b in response to auxin treatment, seeds of cv Williams 82 were sown in a mixture of vermiculite and perlite (3:1) and watered with low-nitrate solution [0.127 mm Ca(NO3)2] with or without 0.2 or 1 μm 2,4-D. After 3 d, the roots were harvested, frozen in liquid nitrogen, and stored at −80°C.
Extraction and Application of B. japonicum NF
Lipooligosaccharide NFs were induced and purified from B. japonicum strain USDA110 as described previously (Sanjuan et al., 1992; Carlson et al., 1993). To induce B. japonicum Nod gene expression, soybean seed extract was added to cultures of the bacteria (Banfalvi et al., 1988; Smit et al., 1992). For application studies, soybean seeds were germinated for 4 d, following which 10 mL of distilled, deionized water containing 10−8 m NF was used to irrigate the seedlings. After 24 h, root samples were collected and used to analyze miR167c expression. Root samples collected from seedlings irrigated with 10 mL of distilled, deionized water were used as a control.
RNA Extraction and Quantitative PCR Analysis
Total RNA and small RNA were extracted from leaves, roots, and nodules using Trizol reagent (Tiangen; http://www.tiangen.com/). Total RNA was treated with DNase I (Invitrogen; www.invitrogen.com) to remove genomic DNA contamination. The first complementary DNA (cDNA) strand was synthesized from total RNA using the FastQuant RT Kit (Tiangen). qRT-PCR assays were performed using SuperReal PreMix Plus (SYBR Green; Tiangen) with gene-specific primers. GmELF1b was used as a housekeeping gene (Jian et al., 2008). Primers designed across the miRNA target sites are provided in Supplemental Table S3.
Stem-Loop qRT-PCR
Stem-loop specific reverse transcription for miRNAs was performed according to Chen et al. (2005). gma-miR1520d was used as a reference miRNA gene for normalization, as described by Kulcheski et al. (2010). All primers used in stem-loop qRT-PCR are listed Supplemental Table S3.
5′ RACE Mapping of miRNA Target Cleavage Sites
Putative targets of miRNAs were predicted using psRNATarget (http://plantgrn.noble.org/psRNATarget/). Total RNA was isolated from a mixture of root, leaf, and nodule tissues collected from 4-week-old soybean plants using Plant RNA Purification Reagent (Invitrogen) according to the manufacturer’s recommended protocol. The GeneRacer Kit (Invitrogen) was used to process the total RNA and map the 5′ terminus of the primary transcript. The resulting cDNA samples were amplified with nested PCR according to the manufacturer’s protocols. All gene-specific primers were designed by Invitrogen (Supplemental Table S3).
Vector Construction
For the promoter::GUS reporter fusion construct, the region located directly upstream of miR167c (654 bp) was amplified from cv Williams 82 genomic DNA and inserted immediately upstream of the GUS gene of the pCAMBIA3301 vector using BamHI and NcoI. For the miR167c overexpression construct, the premiRNA fragment of miR167c (890 bp) was amplified and inserted into the pEGAD vector directly following the cauliflower mosaic virus 35S promoter sequence using AgeI and HindIII. For the RNAi construct, a 435-bp segment of the GmARF8a coding sequence was amplified by PCR from cv Williams 82 cDNA and inserted into the pTCK303 vector in the sense and antisense orientations using KpnI/SpeI and BamHI/SacI. To generate miR167c knockout roots, the STTM167-88 construct was made using a method modified from Yan et al. (2012). The MIM167 construct was generated using methods modified from Todesco et al. (2010), where the gma-miR399-complementary motif of GmIPS1 (glyma10g07180) was modified to mimic target sites for miR167c (MIM167). Site-directed mutagenesis was performed by sequential PCR using complementary oligonucleotides containing the mutations (Supplemental Table S3). According to the site-directed mutagenesis of AtARF8a in Arabidopsis (Arabidopsis thaliana) as described (Wu et al., 2006), seven nucleotide mutations in the miRNA-binding site in GmARF8a were designed. All primers used for plasmid construction are listed in Supplemental Table S3.
Soybean Hairy Root Transformation and Subsequent B. japonicum Inoculation
Soybean hairy root transformation was performed according to Jian et al. (2009) and Kereszt et al. (2007). For nodulation assays, transgenic composite plants were transferred to pots (13 × 10 × 8.5 cm) containing a mixture of vermiculite (grade 3) and perlite (3:1), grown for 1 week (16 h of light, 25°C, and 50% relative humidity), and then inoculated with a suspension of B. japonicum USDA110 at either standard or low titer (107 or 102 CFU mL−1, 30 mL per pot). Ten or 28 d after inoculation, roots and nodules were assayed and harvested.
Histochemical Localization of GUS Expression
Histochemical localization of GUS expression was performed according to Jefferson et al. (1987). Uninoculated or B. japonicum-inoculated transgenic hairy roots and nodules grown in low-nitrate solution were examined (multiple roots from at least six independent lines per construct).
Bioinformatics Analysis
Using ClustalX 1.83 (Thompson et al., 1997) and MEGA6 (Tamura et al., 2013), a phylogenetic tree was constructed based on mature sequences (http://www.mirbase.org/) of miR167 family members. Promoter analyses for analyzing cis-elements of miR167c, GmARF8a, and GmARF8b were performed in PLACE (http://www.dna.affrc.go.jp/PLACE/). The 2,000-bp region located directly upstream of miR167c and the start codon (ATG) of GmARF8a and GmARF8b were used as the promoter sequences (http://www.phytozome.net/). For conserved alignment of the ARF8, ARF6, and IAR3 genes in Arabidopsis and putative homologous genomic sequences of soybean, a sequence of 200 kb was examined on either side of Arabidopsis ARF8, ARF6, and IAR3 and the putative homologous soybean regions. Alignments were obtained from the Plant Genome Duplication Database (http://chibba.agtec.uga.edu/duplication).
Statistical Analysis
All data were analyzed using Student’s t tests or one-way ANOVAs using SigmaPlot 10.0 or GraphPad Prism 5 software. In all tables and figures, statistically significant differences are marked as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001; or ns, not significant, P > 0.05. Moreover, different letters indicate significant differences at P < 0.05.
Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers gma-miR167a (MI0001777), gma-miR167b (MI0001778), gma-miR167c (MI0007210), gma-miR167d (MI0010571), gma-miR167e (MI0010725), gma-miR167f (MI0010726), gma-miR167g (MI0016548), gma-miR167h (MI0017914), gma-miR167i (MI0017926), gma-miR167j (MI0018686), gma-miR167k (MI0031050), gma-miR1520d (MI0007233), GmARF8a (glyma02g40650), GmARF8b (glyma14g38940), GmARF8c (glyma18g05330), GmARF6a (glyma05g27580), GmARF6b (glyma08g10550), GmIAR3a (glyma13g42880), GmIAR3b (glyma15g02560), GmENOD40 (glyma02g04180), GmNIN (glyma04g00210), GmNSP1 (Glyma16g01020), GmHAP2-1 (Glyma02g35190), GmHAP2-2 (Glyma10g10240), and GmELF1B (glyma02g44460).
Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Analysis of premiRNA sequences of all miR167 family members in soybean.
Supplemental Figure S2. Analysis of mature miRNA sequences of all miR167 family members in soybean.
Supplemental Figure S3. Expression analysis of miR167 family members in nodules.
Supplemental Figure S4. Promoter analysis of miR167c.
Supplemental Figure S5. qRT-PCR analysis of miR167c transcript levels.
Supplemental Figure S6. STTM167c-88 structure illustrating the design strategy.
Supplemental Figure S7. qRT-PCR analysis of IPS and STTM transcript levels in hairy roots at 10 d after transplantation.
Supplemental Figure S8. qRT-PCR analysis of miR167 family member transcript levels.
Supplemental Figure S9. qRT-PCR analysis of miR167c transcript levels.
Supplemental Figure S10. qRT-PCR analysis of IPS and STTM transcript levels in hairy roots at 28 d after inoculation.
Supplemental Figure S11. Conserved alignment of the ARF8, ARF6, and IAR3 genes in Arabidopsis and putative homologous genomic sequences of soybean.
Supplemental Figure S12. Alignment between predicted target genes and gma-miR167c.
Supplemental Figure S13. Agarose gel electrophoresis of the 5′ RACE products amplified by RT-PCR.
Supplemental Figure S14. Validation of GmIAR3a and GmIAR3b targets by 5′ RACE analysis.
Supplemental Figure S15. Promoter analysis of GmARF8a and GmARF8b.
Supplemental Table S1. Statistically enriched motifs from miR167 family members.
Supplemental Table S2. Statistically enriched motifs from the targets of miR167.
Supplemental Table S3. Primers used in this study.
Acknowledgments
We thank Dr. Kan Wang (Iowa State University) and Dr. Guiliang Tang (University of Kentucky) for providing the pTF101.1 vector and for advice regarding the miR167c deletion; Wenxin Chen (China Agricultural University) for providing B. japonicum USDA110; and Tianfu Han (Chinese Academy of Agricultural Sciences) for assistance with some experimental protocols.
Footnotes
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Xia Li (xli{at}genetics.ac.cn).
Y.W. and X.L. conceived the original screening and research plans; Y.W., K.L., P.M.G., and X.L. supervised the experiments; K.L., L.C., and H.L. performed the miRNA and target gene analyses and the miR167 cleavage of target gene experiments; Y.W., Y.Z., D.L., and R.W. performed the phenotypic analyses of transgenic roots; F.Z. and Y.T. conducted the hairy root transformation; Y.W. designed the experiments and analyzed the data; Y.W., B.J.F., and X.L. conceived the project and wrote the article with contributions of all the authors; X.L. and P.M.G. supervised and complemented the writing.
↵1 This work was supported by the National Science Foundation of China (grant nos. 31230050 and 30971797), the Ministry of Agriculture of the People’s Republic of China (grant nos. 2009ZX08009–132B and 2014ZX08009–29B), the Youth Innovation Promotion Association of the Chinese Academy of Sciences, and the Australian Research Council (Discovery grant nos. DP130102266 and DP130103084 to P.M.G. and B.J.F.).
↵[OPEN] Articles can be viewed without a subscription.
Glossary
- AON
- autoregulation of nodulation
- miRNA
- microRNA
- qRT
- quantitative real-time
- 2,4-D
- 2,4-dichlorophenoxyacetic acid
- HAI
- hours after inoculation
- DAI
- days after inoculation
- CFU
- colony-forming units
- cDNA
- complementary DNA
- Received February 19, 2015.
- Accepted May 1, 2015.
- Published May 4, 2015.
REFERENCES
