Centre de Recherches sur les Macromolécules Végétales, (CERMAV), Centre National de la Recherche Scientique (CNRS) (affiliated with the Université Joseph Fourier), BP 53, 38041 Grenoble Cedex 9, France
Centre de Recherches sur les Macromolécules Végétales, (CERMAV), Centre National de la Recherche Scientique (CNRS) (affiliated with the Université Joseph Fourier), BP 53, 38041 Grenoble Cedex 9, France
Laboratoire des Interactions Plantes–Microorganismes, UMR 441/2594 Institut National de la Recherche Agronomique (INRA), Centre National de la Recherche Scientique (CNRS), 31326 Castanet-Tolosan Cedex, France
Most land plants live symbiotically with arbuscular mycorrhizal fungi. Establishment of this symbiosis requires signals produced by both partners: strigolactones in root exudates stimulate pre-symbiotic growth of the fungus, which releases lipochito-oligosaccharides (Myc-LCOs) that prepare the plant for symbiosis. Here, we have investigated the events downstream of this early signaling in the roots. We report that expression of miR171h, a microRNA that targets NSP2, is up-regulated in the elongation zone of the root during colonization by Rhizophagus irregularis (formerly Glomus intraradices) and in response to Myc-LCOs. Fungal colonization was much reduced by over-expressing miR171h in roots, mimicking the phenotype of nsp2 mutants. Conversely, in plants expressing an NSP2 mRNA resistant to miR171h cleavage, fungal colonization was much increased and extended into the elongation zone of the roots. Finally, phylogenetic analyses revealed that miR171h regulation of NSP2 is probably conserved among mycotrophic plants. Our findings suggest a regulatory mechanism, triggered by Myc-LCOs, that prevents over-colonization of roots by arbuscular mycorrhizal fungi by a mechanism involving miRNA-mediated negative regulation of NSP2.
Arbuscular mycorrhizal (AM) symbiosis is a widespread root symbiosis that occurs between most land plants and glomeromycotan fungi. Signal molecules produced by both partners are necessary to initiate this symbiosis. Strigolactones present in root exudates activate spore germination and/or hyphal proliferation of the fungus partner (Akiyama et al., 2005; Besserer et al., 2006), whereas lipochito-oligosaccharides (Myc-LCOs) secreted by the AM fungus Rhizophagus irregularis (formerly Glomus intraradices) promote establishment of mycorrhizal symbiosis with various plant species and stimulate lateral root formation on the model legume Medicago truncatula (Maillet et al., 2011). After this early molecular signaling, the fungus forms an hyphopodium on the root surface, penetrates the epidermis cell underneath through a structure called the pre-penetration apparatus (Genre et al., 2005), colonizes the root cortex, and eventually forms highly branched structures in cortical cells called arbuscules. Most nutrient exchanges that take place between the two partners are believed to occur in these arbuscule-containing root cells. AM symbiosis is accompanied by stimulation of lateral root formation (Hodge et al., 2009), which is triggered in M. truncatula by AM fungal factors including Myc-LCOs (Oláh et al., 2005; Maillet et al., 2011). These lateral roots are preferentially colonized by AM fungi (Gutjahr et al., 2009). One infection event generally leads to colonization of a small portion of the root (between 5 and 10 mm from the infection point; Brown and King, 1982; Harley and Smith, 1983), excluding the apical region that comprises the elongation zone and the meristematic zone (Gianinazzi-Pearson et al., 1980; Harley and Smith, 1983).
Apart from the signaling functions of the secreted factors strigolactones and Myc-LCOs, very little is known about the signal transduction mechanisms within root cells that are responsible for root colonization by AM fungi. A recent study implicates a transcription factor called NSP2 in mycorrhization: nsp2 mutants of M. truncatula do not respond to Myc-LCOs and are colonized less than wild-type plants by the AM fungus R. irregularis (Maillet et al., 2011). NSP2 belongs to the GRAS transcription factor family. It is known for its involvement in the Nod factor signaling pathway leading to root nodule formation (Kalóet al., 2005). It therefore belongs to the Sym pathway, the common signaling pathway shared by both symbiotic processes in legumes: mycorrhization and nodulation (Oldroyd et al., 2009).
MicroRNAs are small non-coding RNA molecules of approximately 21 nucleotides that negatively regulate expression of target genes by mRNA cleavage or inhibition of translation (Lanet et al., 2009). Among the genes known to be involved in root symbioses, very few microRNAs have been identified. We previously described the involvement of microRNAs miR169 and miR166 in regulating transcription factors involved in nodulation of M. truncatula (Combier et al., 2006; Boualem et al., 2008), and Li et al. (2010) showed that over-expression of miR482, miR1512 and miR1515 increased nodulation. Evidence from microRNA chip experiments and deep sequencing of small RNAs indicates that dozens of microRNAs are differentially expressed during AM symbiosis (Gu et al., 2010; Devers et al., 2011), but a direct role has never been established unambiguously. Using a degradome approach, it was recently observed that NSP2 is a target of the microRNA miR171h (Branscheid et al., 2011; Devers et al., 2011), but the details of this negative regulation and its consequences for AM symbiosis were not investigated.
In this study, we investigated the regulatory loop involving miR171h and NSP2 during AM symbiosis. We demonstrate that miR171h plays a key role in the control of M. truncatula root colonization by the fungus R. irregularis by controlling NSP2 transcript levels.
MtmiR171h expression is up-regulated in the root elongation zone during AM symbiosis
We measured expression of miR171h in control and mycorrhizal roots. Quantitative RT-PCR amplification of the predicted corresponding precursor (Figure 1a) and Northern blot analysis of mature miR171h (Figure 1b) revealed that miR171h is up-regulated during AM symbiosis. By contrast, pre-miR171h expression remained unchanged in roots nodulated with Sinorhizobium meliloti, as quantified by quantitative RT-PCR (Figure S1a).
To investigate the spatio-temporal expression pattern of miR171h in the roots during AM symbiosis, we analyzed the promoter activity of miR171h by using a β-glucuronidase (GUS) reporter gene assay. Before inoculation of the roots with R. irregularis, we saw no GUS activity by histological staining except for faint staining of the central cylinder of mature roots after prolonged incubation, also visible in inoculated roots (asterisks in Figure 1c,d). Transverse sections of these mature root zones confirmed that GUS staining was located only in the vascular tissue (Figure 1e). In contrast, roots colonized by R. irregularis showed a very different GUS staining pattern (arrow in Figure 1d): the sub-apical part of the roots, corresponding to the elongation zone, exhibited strong GUS staining (Figure 1g,i), which was located in the peripheral root tissues comprising the epidermis and first cortical cell layers (Figure 1g, red arrow). Mature root zones distant from the apex, by contrast, presented a staining pattern similar to control roots (Figure 1f,h). Staining with fluorescein-conjugated wheat germ agglutinin revealed the presence of the fungus exclusively in the mature distal root zones (Figure 1h,j). In the elongation zone, where GUS staining was strong, the fungus was systematically absent (Figure 1i,k). Assays on roots inoculated with S. meliloti did not reveal GUS expression in nodules (Figure S1b).
NSP2 expression is down-regulated during AM symbiosis
To determine which genes were targeted by miR171h, we first performed an in silico analysis. Direct BLAST searches in the M. truncatula genome (version 3.5) (Young et al., 2011) for genes presenting sequences homologous to miR171h led to identification of four putative target genes (Figure 2a), which present up to four mismatches but no mismatch around the cleavage site. Among these four genes, three have been described previously (Devers et al., 2011): NSP2 (Mtr.44789.1.S1_at), another GRAS transcription factor that we called NSP2-like (Medtr5g058860), and a gene encoding a pentatricopeptide-repeat protein (Mtr.25350.1.S1_at). The fourth potential target gene (Mtr.11537.1.S1_at) is predicted to encode a pentatricopeptide-repeat protein. Using quantitative RT-PCR analysis, we observed that the expression levels of the two pentatricopeptide-repeat protein genes and the NSP2-like gene were similar in mycorrhizal and non-mycorrhizal roots (Figure 2b). By contrast, NSP2 expression was significantly lower in mycorrhizal roots than in controls. Among the target genes of miR171h, only NSP2 expression decreased upon infection of roots with R. irregularis, and this correlates negatively with increased miR171h expression. Consistent with our results, the gene expression atlas of M. truncatula (http://mtgea.noble.org/v2/) reveals that NSP2 expression decreases during mycorrhization. These data suggest that miR171h may target NSP2 transcripts for degradation during mycorrhization. To test this possibility, we performed a RACE-PCR analysis on mycorrhized roots. We confirmed cleavage of NSP2 by miR171h (Figure 2c), in agreement with the degradome analysis by Devers et al. (2011).
Over-expression of MtmiR171h affects AM symbiosis
To evaluate the putative role of miR171h in regulating AM symbiosis, we investigated the effect of miR171h precursor over-expression on mycorrhizal development using Agrobacterium rhizogenes root transformation (Boisson-Dernier et al., 2005). When the precursor of miR171h was over-expressed in roots (Figure 3a), the transcript levels established by quantitative RT-PCR analysis for the target genes previously identified were significantly lower than those in the control roots (Figure 3a). The fungal colonization of the composite plants over-expressing miR171h was altered compared to control roots (Figure 3b): the observed reduction in colonization (nearly 50%) was similar to that observed in nsp2 mutant plants (Maillet et al., 2011). However, as in nsp2 mutants, the structure and quantity of arbuscules in colonized root sections appeared normal. We also used quantitative RT-PCR to analyze the effect of miR171h over-expression on expression of various genes involved in AM symbiosis and/or signaling pathways related to Myc-LCOs [Pt4 (Javot et al., 2007), ENOD11 (Boisson-Dernier et al., 2005), BCP1 (Hohnjec et al., 2005), VAPYRIN (Pumplin et al., 2010) and CBF1 (Hogekamp et al., 2011)] and genes responding to Myc-LCO [Mtr.52092.1.S1_s_at, Mtr.37912.1.S1_at, Mtr.38167.1.S1_at, Mtr.35524.1.S1_at and Mtr.156.1.S1_at (Maillet et al., 2011)]. As in the nsp2 mutant, over-expression of miR171h not only reduced fungal colonization but also led to a significant decrease in expression of all of these mycorrhizal marker genes (Figure 3c).
MtmiR171h modulates the spatial colonization of roots by R. irregularis through regulation of NSP2 expression
The similarity of the fungal colonization and host marker gene expression during fungal colonization in roots over-expressing miR171h and the nsp2 mutant (Figure 3) suggests that NSP2 mRNA may be the main target of miR171h during mycorrhization. To investigate this hypothesis, composite plants transformed with an miR171h-resistant NSP2 gene were produced by mutagenesis of the miR171h recognition site in the gene (Figure 2a), while the amino acid sequence of the protein product was unchanged. As Kalóet al. (2005) showed that use of CaMV 35S promoter to drive NSP2 expression is able to complement the nsp2 mutant, we used this strong constitutive promoter to drive NSP2 expression in our miR171h-resistant construct. nsp2 mutant roots over-expressing NSP2 recovered a normal mycorrhizal phenotype (Figure 4b). Using the wild-type A17, we checked that empty vector-transformed (control) roots presented the same behavior during mycorrhization as roots over-expressing NSP2 (Figure 4a,b,d,e and Figure S2a). Using the grid intersect method (Giovannetti and Mosse, 1980) to evaluate overall root colonization by AM fungi, we observed that roots expressing the miR171h-resistant NSP2 showed significantly greater colonization than control roots (Figure 4a). Indeed, at an early stage of root colonization (5 weeks post-inoculation), colonization of miR171h-resistant NSP2 roots was three times greater than in control roots. At later stages, this difference decreased but the colonization level remained significantly higher in miR171h-resistant NSP2 roots compared to control roots (Figure 4a). In addition, complementation assays of the nsp2 mutant with the miR171h-resistant NSP2 gene revealed a similar phenotype pattern to that observed in the wild-type (A17) background (Figure 4b). Expression of most of the mycorrhizal marker genes tested was also strongly up-regulated (Figure 4c), whereas expression of other putative target genes of miR171h remained unchanged (Figure S2b). Further analysis of fungal structures in roots expressing the miR171h-resistant NSP2 gene revealed a higher amount of all mycorrhizal structures (Figures 4d and 5a,b). Strikingly, root tips of plants expressing the miR171h-resistant NSP2 gene were also heavily colonized: 43% of apices showed fungal colonization (Figures 4e and 5d,f), a feature observed in only 4–7% of control roots (Figures 4e and 5c,e). Colonization of root tips in the miR171h-resistant NSP2 roots was observed even after a short period post-inoculation (5 weeks), i.e. when only approximately 10% of control roots were colonized. Root tips of control roots showing fungal colonization exhibited a rounded shape as if their elongation was arrested (Figure 5e). In contrast, most apices of miR171h-resistant NSP2 roots showing fungal colonization appeared normal (Figure 5f). To confirm that this over-colonization of root tips was not due to altered root development in miR171h-resistant NSP2 roots compared to control roots, we measured the total fresh weight, the number of root tips per gram of fresh weight, and the expression of three genes involved in root development or lateral root initiation: CYCB1 (Himanen et al., 2002), LBD and HB1 (Ariel et al., 2010) (Figure S2c–e). No difference in root growth and root branching was detected between control and miR171h-resistant NSP2 roots (Figure S2c,d). Gene expression was also similar between control and miR171h-resistant NSP2 roots (Figure S2e). These data suggest that the presence of the fungus in the apex of miR171h-resistant NSP2 roots was not due to a decrease or arrest of root elongation.
Nodulation assays revealed that miR171h-resistant NSP2 root nodules (Figure S1c) were comparable to control nodules (Figure S1d), indicating that control of NSP2 by miR171h is absent during nodulation.
MtmiR171h is involved in a specific Myc-LCO signaling pathway
As NSP2 is an important protein in Nod factor signaling (Kalóet al., 2005; Hirsch et al., 2009), and Nod factors are very similar in structure to Myc-LCOs, the potential role of NSP2 and miR171h in the Myc-LCO signaling pathway was investigated. We first examined the expression of pre-miR171h in response to Nod factors and Myc-LCOs. Nod factors did not induce expression of pre-miR171h at either 10−7m or 10−8m (Figure S1e,g,i) compared to untreated roots (Figure S1f,h), but Myc-LCO treatment led to significant up-regulation of pre-miR171h expression (Figure 6a). We confirmed this finding by expressing the GUS reporter gene under the control of the miR171h promoter, and observed strong staining of the apical parts of the Myc-LCO-treated roots (Figure 6c) localized in the peripheral root tissues comprising epidermis and first cortical cell layers (arrow in Figure 6e). By contrast, no GUS staining was observed in control roots (Figure 6b,d).
As the pattern of miR171h expression in roots treated with Myc-LCOs was very similar to that observed during AM symbiosis, we next analyzed whether expression of a downstream target gene of NSP2, ENOD11 (Hirsch et al., 2009), was controlled by miR171h during treatment with Myc-LCOs. Plants stably expressing the GUS reporter gene under the control of the ENOD11 promoter were transformed with wild-type NSP2 and miR171h-resistant NSP2 constructs. In addition, nsp2 mutant plants expressing the same GUS reporter gene under the control of the ENOD11 promoter were also tested. Experiments were performed in composite plants and also in root organ cultures, in which the Nod factor signaling pathway is inactive, with similar results. Whereas untreated roots showed no GUS staining (Figure S3a,c,e,g), Myc-LCO-treated control roots revealed GUS staining due to pENOD11 activity (Figure S3b), as observed by Maillet et al. (2011). In parallel, treated nsp2 mutant roots and roots over-expressing miR171h showed no GUS staining (Figure S3d,f). Finally, treated roots expressing the miR171h-resistant NSP2 construct (Figure S3h) revealed much stronger GUS staining than treated control roots (Figure S3b). These results support the role of miR171h in regulation of NSP2, and, as a consequence, in the Myc-LCO signaling pathway.
Overall, these observations reveal that miR171h responds to Myc-LCOs and suggest that exclusion of the fungus from the root tip during mycorrhization may require Myc-LCOs release by the fungus in planta.
The miR171h and its recognition site in NSP2 co-evolved in the green lineage
To further assess the biological relevance of miR171h in establishment of AM symbiosis, we examined the distribution and conservation of the mir171h target site in target genes among various species of embryophytes: mosses, lycophytes and angiosperms. We screened the available gene databases for orthologs of NSP2 and NSP2-like (Medtr5g058860), and the two genes encoding pentatricopeptide-repeat proteins. NSP2-like orthologs were found only in the genomes of two legume species: M. truncatula and Lotus japonicus. Potential orthologs of the two pentatricopeptide-repeat protein genes were found only in a few angiosperms. By contrast, we found putative orthologs of NSP2 in all species examined, and used them, with NSP2-like orthologs, for phylogenetic analysis (Figure 7a). In miRBase (http://www.mirbase.org), we identified sequences homologous to MtMir171h in several mycotrophic plant species (Figure 7b, green arrowheads). However, they were absent from the miR171 families of the non-mycotrophic moss Physcomitrella patens, the lycophyte Selaginella moellendorffii, and the angiosperms Arabidopsis thaliana and Arabidopsis lyrata (Figure 7b, red arrows). The NSP2 mir171h-binding site and the corresponding miR171h sequences from each species were analyzed (Figure 7b). Interestingly, an miR171h-binding site was identified in NSP2 orthologs of all mycotrophic angiosperms tested. By contrast, the mir171h-binding sequences in NSP2 homologs from the non-mycotrophic angiosperm species A. thaliana, A. lyrata, Thelungiella halophila, Cabsella rubella and Brassica rapa were strongly modified (Figure 7b). In summary, this phylogenetic analysis reveals that NSP2 orthologs are found in all land plant species for which genomic data are available. The miR171h recognition site of NSP2 homologs and the corresponding miR171h sequences were strongly conserved only in angiosperms that show AM symbiosis.
Recent studies have investigated the abundance and diversity of small RNAs potentially involved in AM symbiosis (Branscheid et al., 2010; Gu et al., 2010; Devers et al., 2011). Here, we provide evidence that a single microRNA, miR171h, can modulate the pattern of root tissue colonization by the AM fungus Rhizophagus irregularis. Over-expression of miR171h in M. truncatula strongly reduced root colonization by R. irregularis without affecting arbuscule formation, and also reduced expression of several genes involved in AM symbiosis. This microRNA targets the NSP2 gene. The fact that expression of miR171h increases in mycorrhizal roots whereas that of NSP2 decreases, and that miR171h over-expressing roots exhibit a similar mycorrhizal phenotype to the nsp2 mutant, suggests that the main role of miR171h during establishment of AM symbiosis is to target NSP2 transcripts. Of the various target genes, only NSP2 expression was down-regulated in mycorrhizal roots.
Despite the well-established role of NSP2 during Nod factor perception and root nodule formation, miR171h expression is neither induced in nodules nor by Nod factor treatments, suggesting specificity of NSP2 cleavage by miR171h during AM symbiosis. Moreover, although the molecular structures of Myc-LCOs and Nod factors are very similar, we confirm here that the plant is able to distinguish between these two types of signaling molecules (Maillet et al., 2011; Czaja et al., 2012). This work thus establishes miR171h as a specific marker of the Myc-LCO signaling pathway.
As most land plants live symbiotically with AM fungi, we analyzed the prevalence of the NSP2–miR171h couple in the green lineage. Our phylogenetic analysis revealed that NSP2 orthologs are present in all land plant species for which genomic data are available. Interestingly, the miR171h recognition site in NSP2 homologs was strongly conserved only in angiosperms that show AM symbiosis. Similarly, based on available data in miRBase, we identified an miR171h homolog in all mycotrophic angiosperms but not in non-mycotrophic ones. This suggests that regulation of NSP2 by miR171h is important for AM symbiosis and may be a specific trait of mycotrophic angiosperms. However, because of the lack of data on monilophytes and gymnosperms, we cannot exclude a more ancient origin of this regulation in the euphylophyte clade.
A central finding of this work is that MtmiR171h modulates the intensity and spatial colonization of roots by R. irregularis through regulation of NSP2 expression. An obvious explanation for the absence of the fungus in the elongation zone of control roots is that root elongation proceeds faster than intra-radical fungal colonization. However, the presence of the fungus in the apical region of miR171h-resistant NSP2 roots, whose growth rate was equivalent to that of control roots, argues strongly for the occurrence of negative control of fungal colonization and/or penetration in the apical region of control roots. The similarity of the expression patterns of miR171h during mycorrhizal colonization and in response to Myc-LCOs suggests that the expression of miR171h observed during mycorrhization may be induced by Myc-LCOs. This implies that the fungus releases Myc-LCOs in planta, and that these molecules systemically induce expression of miR171h in the apical root region either by direct migration or through intermediate signaling. Such expression would lead to silencing of NSP2, with a negative impact on fungal colonization and/or penetration of the root apical region.
A recent study has demonstrated the involvement of NSP2 in strigolactone biosynthesis (Liu et al., 2011). Moreover, we know that strigolactones stimulate germination and growth of AM fungi (Akiyama et al., 2005; Besserer et al., 2006), and that strigolactone-deficient mutants have a lower mycorrhization rate (Gomez-Roldan et al., 2008; Gutjahr et al., 2012). In roots invaded by the fungus, miR171h induction, which down-regulates NSP2, may decrease strigolactone content and thus control fungal colonization. In contrast, in miR171h-resistant NSP2 roots, strigolactone content would be higher, thus promoting fungal colonization, notably in the root apical region. Development of highly sensitive analytical methods to quantify strigolactone and Myc-LCO contents in various parts of the roots is required to further explore this hypothesis.
In conclusion, our findings suggest that the transcription factor NSP2 plays an important role in AM symbiosis to regulate root colonization by the fungal partner. They also suggest that, in colonized roots, one particular microRNA, miR171h, specifically degrades NSP2 transcripts, particularly in the apical region of the roots. We propose that this post-transcriptional down-regulation of NSP2 prevents the fungus from colonizing the apical root region, perhaps by locally decreasing the content of strigolactone, and that Myc-LCOs may be the fungal signals that trigger this regulatory process of autoregulation.
Rhizophagus irregularis (formerly named Glomus intraradices) DAOM197198 sterile spores were purchased from Agronutrition (www.agronutrition.fr). M. truncatula Gaertn ‘Jemalong’ genotype A17 seeds were surface-sterilized and germinated on agar plates in the dark for 5 days at 4°C. Plants were then cultivated on Oil-Dri US-special substrate (Damolin, www.damolin.fr/) for 5–12 weeks in a growth chamber, and watered every 2 days with modified Long Ashton medium containing a low concentration of phosphate (Balzergue et al., 2011). For inoculation with R. irregularis, we used 450 spores per plant. Nodulation assays were performed as described by Combier et al. (2008).
The Myc-LCOs used in this study are an equimolar mix of the four (Syn)Myc-LCOs LCO-IV(C16:0), LCO-IV(C16:0,S), LCO-IV(C18:1Δ9Z) and LCO-IV(C18:1Δ9Z,S described by Maillet et al. (2011), at a final concentration of 10−7m. Nod factors from S. meliloti were used at a concentration of 10−7 or 10−8m.
DNA fragments of interest were amplified using Pfu polymerase (Promega, www.promega.com) and the primers shown in Table S1. They were cloned using XhoI and NotI restriction enzymes into the pPEX-DsRED plasmid (Combier et al., 2008) for over-expression under the control of the strong constitutive CaMV 35S promoter activity and miR171h-resistance assays, and using KpnI and NcoI into the pPEX GUS plasmid for reporter assays, as described by Combier et al. (2008).
Composite plants with Agrobacterium rhizogenes-transformed roots were obtained as described by Boisson-Dernier et al. (2005). Transformed roots (which were able to grow on kanamycin) were selected by observation of DsRED fluorescence using a fluorescent binocular (leica, www.leica-microsystems.com/). Control roots correspond to roots transformed with A. rhizogenes carrying the empty vector pPEX-DsRED.
Quantitative RT-PCR analyses
RNA was extracted using an RNeasy plant mini kit (Qiagen, www.qiagen.com). Reverse transcription was performed using SuperScript II reverse transcriptase (Invitrogen, www.invitrogen.com) on 500 ng of total plant RNA. For each experiment, six independent transformants were analyzed. Quantitative PCR amplifications were conducted on a Roche LightCycler 480 System (Roche Diagnostics, www.roche-diagnostics.com) under the following conditions: 95°C for 5 min, then 45 cycles of 95°C for 15 sec and 60°C for 1 min. The various primer sets used are described in Table S2.
Histochemical staining and microscopy studies
GUS staining was performed as described by Combier et al. (2008). Root sections (50 μm) were cut using a vibratome (Leica, www.leica-microsystems.com/) from samples embedded in 6% agarose. The samples were observed with an inverted microscope (Leica). After clearing in 10% w/v KOH and rinsing in sterile water, mycorrhizal roots were treated for 30 min with fluorescein-conjugated wheat germ agglutinin (Invitrogen), which binds fungal chitin, then washed three times for ten minutes each in PBS buffer and observed using an inverted light microscope or a confocal microscope (Leica). Alternatively, they were stained with Schaeffer black ink as described by Vierheilig et al. (1998). The percentage mycorrhization was established using the grid intersect method described by Giovannetti and Mosse (1980). More precise phenotyping of mycorrhization was also performed as described by Trouvelot et al. (1986): the frequency (F) of mycorrhiza in the root system and the arbuscule abundance (a) (percentage) were calculated in the colonized root sections using Mycocalc software (http://www2.dijon.inra.fr/mychintec/Mycocalc-prg/download.html). Each mycorrhization experiment was performed at least twice with 15 plants per treatment, each corresponding to an independent A. rhizogenes transformation.
Northern Blot analysis
For RNA gel-blot analyses, frozen tissue was homogenized in buffer containing 0.1 m NaCl, 2% SDS, 50 mm Tris/HCl pH 9, 10 mm EDTA pH 8 and 20 mmβ-mercaptoethanol, and RNAs were extracted twice with phenol/chloroform and recovered by ethanol precipitation. RNA were loaded onto a 15% PAGE gel and transferred to nylon membrane (HybondNX, Amersham, www.gelifesciences.com/webapp/wcs/stores/servlet/Home/fr/GELifeSciences-FR/). Blots were hybridized using radioactively end-labeled oligonucleotide probes for U6 and miR171h detection. Hybridization was performed at 55°C. Hybridization signals were quantified using a Fuji phosphorimager (www.fujifilm.com/products/life_science_systems/) and normalized to those obtained using the U6 oligonucleotide probe.
Sequence collection, phylogenetic analyses and mir171h-binding sequence comparisons
To compare mir171h-binding sequences, NSP2 genes were aligned using Clustal W (Thompson et al., 1994), and the 21 nucleotides corresponding to the mir171h-binding site of MtNSP2 were identified.
The mean values for relative gene expression or mycorrhization rates were compared using the Kruskal–Wallis test, and, when significant, a pairwise comparison was made using the non-parametric Mann–Whitney test. Asterisks indicate significant differences compared to the control (P < 0.05, n = 6 independent transformants for quantitative PCR amplifications, and n = 15 independent transformants for mycorrhization assays). Error bars represent the standard error of the mean (SEM).
This work was performed in the Laboratoire de Recherche en Sciences Végétales in Toulouse, which belongs to the Laboratoire d’Excellence entitled TULIP (ANR-10-LABX-41). Part of the work was supported by the FUI project Neofertil. We thank the Toulouse Réseau Imagerie facility for confocal microscopy analyses and technical advice on the histological analyses. We thank Fabienne Maillet (Laboratoire des Interactions Plantes-Microorganismes, Toulouse, France) for providing Nod factors. We also thank Eric Samain (Centre de Recherches sur les Macromolécules Végétales, Grenoble, France) for providing the chito-oligosaccharides used for LCO synthesis.