Ethylene and jasmonic acid act as negative modulators during mutualistic symbiosis between Laccaria bicolor and Populus roots

Authors

  • Jonathan M. Plett,

    Corresponding author
    1. INRA, UMR 1136 INRA-University Henri Poincaré, Lab of Excellence ARBRE, Interactions Arbres/Microorganismes, INRA-Nancy, Champenoux, France
    2. Hawkesbury Institute for the Environment, University of Western Sydney, Richmond, NSW, Australia
    • Author for correspondence:

      Jonathan Plett

      Tel: +61 4578 1097

      Email: j.plett@uws.edu.au

    Search for more papers by this author
  • Amit Khachane,

    1. Hawkesbury Institute for the Environment, University of Western Sydney, Richmond, NSW, Australia
    Search for more papers by this author
  • Malika Ouassou,

    1. INRA, UMR 1136 INRA-University Henri Poincaré, Lab of Excellence ARBRE, Interactions Arbres/Microorganismes, INRA-Nancy, Champenoux, France
    Search for more papers by this author
  • Björn Sundberg,

    1. Umea Plant Science Centre, Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural Sciences, Umea, Sweden
    Search for more papers by this author
  • Annegret Kohler,

    1. INRA, UMR 1136 INRA-University Henri Poincaré, Lab of Excellence ARBRE, Interactions Arbres/Microorganismes, INRA-Nancy, Champenoux, France
    Search for more papers by this author
  • Francis Martin

    1. INRA, UMR 1136 INRA-University Henri Poincaré, Lab of Excellence ARBRE, Interactions Arbres/Microorganismes, INRA-Nancy, Champenoux, France
    Search for more papers by this author

Summary

  • The plant hormones ethylene, jasmonic acid and salicylic acid have interconnecting roles during the response of plant tissues to mutualistic and pathogenic symbionts.
  • We used morphological studies of transgenic- or hormone-treated Populus roots as well as whole-genome oligoarrays to examine how these hormones affect root colonization by the mutualistic ectomycorrhizal fungus Laccaria bicolor S238N.
  • We found that genes regulated by ethylene, jasmonic acid and salicylic acid were regulated in the late stages of the interaction between L. bicolor and poplar. Both ethylene and jasmonic acid treatments were found to impede fungal colonization of roots, and this effect was correlated to an increase in the expression of certain transcription factors (e.g. ETHYLENE RESPONSE FACTOR1) and a decrease in the expression of genes associated with microbial perception and cell wall modification. Further, we found that ethylene and jasmonic acid showed extensive transcriptional cross-talk, cross-talk that was opposed by salicylic acid signaling.
  • We conclude that ethylene and jasmonic acid pathways are induced late in the colonization of root tissues in order to limit fungal growth within roots. This induction is probably an adaptive response by the plant such that its growth and vigor are not compromised by the fungus.

Introduction

Plants interact with a diversity of organisms, some of which attempt to colonize plant tissues in an effort to access the nutrients stored therein. Growing invasively within plant tissues, these foreign organisms (e.g. mutualistic symbionts, endophytes, and pathogens) may remain within the apoplastic space or can penetrate into plant cells during the process of host colonization. Opposing the invasion of these microbes is a finely tuned defense response network encoded by the plant that detects the presence of nonself cells through the binding of microbe-associated molecular patterns (MAMPs) to plant-encoded pattern recognition receptors (PRRs; reviewed in Hann & Boller, 2011; Chinchilla & Boller, 2012). MAMPs are elicitor proteins that are characteristic of all microorganisms and are intrinsic to their cellular make-up (e.g. chitin; Felix et al., 1993; Kaku et al., 2006; Miya et al., 2007; Boller & Felix, 2009; Millet et al., 2010; Shimizu et al., 2010). PRRs in plants are primarily membrane-bound receptor-like kinase (RLK) proteins that contain an external-facing receptor domain, a transmembrane domain and an intracellular kinase domain. A key component of the signaling induced by PRRs is the triggering of defense pathways controlled by a range of plant hormones (Mishina & Zeier, 2007; Howe & Jander, 2008; Bari & Jones, 2009; Du et al., 2009; Pieterse et al., 2009; Katagiri & Tsuda, 2010; Bernoux et al., 2011). The three plant hormones considered to be primary regulators within this defense network are ethylene (ET), jasmonic acid (JA) and salicylic acid (SA; Glazebrook et al., 2003; Derksen et al., 2013), while abscisic acid (ABA) also plays an important role in biotic interactions (Pieterse et al., 2009; Ton et al., 2009; Martin Rodriguez et al., 2010). ET is synthesized from its precursor 1-aminocyclopropane-1-carboxylic acid (ACC) via the enzyme ACC OXIDASE (ACO) and is perceived by a family of membrane-bound receptors (e.g. ETR1 (ETHYLENE RECEPTOR1)) that control the expression of ETHYLENE RESPONSE FACTOR (ERF) genes (e.g. ERF1). Mutations that block ET binding to receptors (e.g. the etr1-1 mutation) cause plant-wide ET insensitivity (Bleecker et al., 1988; Hua et al., 1995; Kim et al., 2011; Liu & Wen, 2012). The active form of JA, JA-Ile, is perceived by the COI1 (coronatine insensitive1) receptor protein which then binds to JAZ (jasmonate ZIM-domain) proteins that negatively regulate a suite of genes such as PDF1.2 (plant defensin1.2), VSP1 (vegetative storage protein1) and JAR1 (jasmonate response locus) (Katsir et al., 2008; Yan et al., 2009; Fernandez-Calvo et al., 2011). Perception of SA by the cytosolic receptors NPR1 (natriuretic peptide receptor1), NPR3 and NPR4, meanwhile, leads to the expression of a range of PATHOGENESIS RELATED (PR) proteins (e.g. PR1; Fu et al., 2012; Wu et al., 2012). SA is typically associated with defense against biotrophic pathogens (Dewdney et al., 2000; Glazebrook, 2005; Halim et al., 2007; Spoel et al., 2007), while JA and ET typically exhibit coordinated signaling in the defense against necrotrophic pathogens (Berrocal-Lobo et al., 2002; Glazebrook, 2005; Broekaert et al., 2006; Spoel et al., 2007). Because of these specialized roles, and extensive mutant-based evidence in Arabidopsis, SA-induced pathways are generally thought to act antagonistically to the signaling of the JA/ET pathways upon challenge by a pathogen (Pieterse et al., 1998; Gupta et al., 2000; Kondo et al., 2007; Spoel et al., 2007; Spoel & Dong, 2008).

Not all organisms seeking to colonize plant tissues are detrimental to plant health. Mutualistic bacteria and fungi enter into symbiotic mutualistic relationships with the plant whereby both partners benefit from the interaction. Colonization of plant tissues by these organisms, however, is no less invasive than colonization by a pathogenic organism, nor do mutualistic organisms lack MAMPs. Therefore, how the plant's immune reaction is controlled in the interaction with mutualistic microbes as opposed to pathogenic organisms is of great interest. ET, JA and SA exhibit contrasting roles in how they modulate mutualistic colonization as opposed to pathogenic colonization (Dewdney et al., 2000; Strack et al., 2003; Glazebrook, 2005; Halim et al., 2007; Hause et al., 2007; Spoel et al., 2007; Herrera-Medina et al., 2008). During the colonization of plant tissues by symbiotic ectomycorrhizal fungi, endogenous ET concentrations increase (Graham & Linderman, 1980; Splivallo et al., 2009), but this increase in ET concentration does not affect the initial stages of root colonization (Graham & Linderman, 1981; Rupp et al., 1989). Colonization of plant tissues by mutualistic arbuscular mycorrhizal (AM) fungi such as Rhizophagus irregularis (formerly Glomus intraradices) is inhibited by ET and JA and delayed by SA (Azcon-Aguilar et al., 1981; Geil et al., 2001; Guinel & Geil, 2002; Medina et al., 2003; Herrera-Medina et al., 2008; Penmetsa et al., 2008; Riedel et al., 2008). Therefore, R. irregularis has evolved at least one countermeasure whereby it produces a secreted protein called SP7 that interacts with a component of the ET signaling pathway (ERF19) to suppress ET signaling that would otherwise be detrimental to fungal colonization (Kloppholz et al., 2011). Similar countermeasures have yet to be found that interact with either the JA or SA pathway. ET and SA have a deleterious effect during the interaction between Rhizobia and its host by interfering in the plant-encoded Nod (nodulation) factor signaling pathway, while JA promotes nodule formation by inducing Nod factor signaling (Grobbelaar et al., 1971; van Spronsen et al., 1995; Penmetsa & Cook, 1997; Martinez-Abarca et al., 1998; Rosas et al., 1998; Mabood & Smith, 2005; Mabood et al., 2006; Stacey et al., 2006; Penmetsa et al., 2008; Oldroyd et al., 2009). Hyphal growth of the beneficial endophytic fungus Piriformospora indica in host tissues is limited by ET, possibly to prevent the fungus from taking over the root tissues and impinging on plant productivity (Waller et al., 2005; Camehl & Oelmüller, 2010; Camehl et al., 2010). These hormones, however, do not always inhibit symbiotic interactions. ET signaling is still needed for nodule formation in Sesbania rostrata and in the initial stages of P. indica colonization, and low concentrations of JA encourage AM fungal colonization (D'Haeze et al., 2003; Khatabi et al., 2012). While a great deal is known concerning the diversity of the effects that these three hormones have on the growth of mutualistic organisms within plant tissues at the physiological level, the molecular events induced by these hormones need further investigation. This is especially true with relation to how the signaling of one hormone aligns with, or opposes, the other hormones at the molecular level.

Using the model interaction between the mutualistic ectomycorrhizal (ECM) fungus Laccaria bicolor Maire Orton S328N and its host plant Populus tremula × Populus alba 717-1B4, we sought to perform an in-depth assessment on the effect of ET on this interaction at the morphological and transcriptomic levels. Further, we sought to determine how the effect of ET on plant tissues undergoing colonization compared with the effect of JA and SA in the same system. Laccaria bicolor colonizes roots in two distinct steps: first, fungal hyphae wrap around the root to form a dense hyphal sheath called the mantle, and then in the second step the hyphae penetrate the apoplastic space of the root to form a complex labyrinthine structure called the Hartig net.

We found that ET (through the application of its metabolic precusor ACC) and JA application inhibited the secondary stage of colonization, that of Hartig net formation. We found that application of ACC, and similarly JA application, induced altered expression of putative PRR and RLK genes as well as genes involved in cell wall biosynthesis and maintenance, suggesting that reduced Hartig net development in roots treated with these hormones is attributable to a rigidification of the plant cell wall and a prevention of the detachment of plant cell walls from each other. As our results demonstrated that in untreated, normal colonization conditions ET- and JA-controlled genes are only significantly regulated in the late stages of colonization of roots by L. bicolor, the role of ET and JA must be to limit Hartig net formation during the late stages of fungal colonization, such that the plant is not overrun by fungal hyphae.

Materials and Methods

Plant and fungal materials

An in vitro assay was used to determine the effects of different chemicals on the ability of L. bicolor S238N to colonize poplar roots. Populus tremula × Populus alba clone 717-1B4 was used for the mycorrhiza formation assays using chemicals to alter hormone signaling. The transgenic lines of 35S::PttACO1 and 35S::Atetr1-1 from Love et al. (2009) were in a P. tremula × P. tremuloides T89 background, and therefore this background was used as a control for these experiments. In all cases, 2-cm-long in vitro cuttings were rooted between two cellophane membranes on the surface of solid Murashige and Skoog (MS) medium (Felten et al., 2009). At the same time, fungal colonies of L. bicolor were grown separately on a cellophane membrane on Pachlewski solid medium. After the plants were rooted, either a membrane bearing 10-d-old L. bicolor colonies (for fungal contacts) or a sterile membrane without fungus (for uncolonized root controls) was placed on top of the roots on low-glucose Pachlewski solid medium (0.1% glucose) or low-glucose Pachlewski solid medium supplemented with 250 μM ACC, 10 nM JA, 10 nM MeJA (methyl jasmonate) or 500 μM SA. Roots were left in contact with the fungus for 2 wk, at which point at least three lateral roots colonized by L. bicolor were harvested from each root system and fixed via vacuum infiltration in 4% paraformaldehyde solution and treated as described below in ‘Microscopic mycorrhizal analysis’ to analyze the development of the intraradical hyphal network within the root, the so-called ‘Hartig net’. Similarly, root samples were sampled and frozen in liquid nitrogen, and their RNA was extracted and treated as described below in ‘Whole-genome oligoarray analysis’ for whole-genome oligoarray or Q-PCR transcriptomic analysis.

DNA recombination procedures

Template cDNA for all cloning procedures was created from RNA extracted using the RNeasy Plant kit (Qiagen, Courtaboeuf, France) according to the manufacturer's instructions with the addition of 25 mg polyethylene glycol 8000 ml−1 RLC buffer to the extraction solution from P. tremula × P. alba clone 717-1B4 roots undergoing colonization by L. bicolor S238N. An on-column DNA digestion step with DNAse I (Qiagen) was also included to avoid DNA contamination. RNA quality was verified using Experion HighSens capillary gels (Bio-Rad, Marnes-la-Coquette, France). Synthesis of cDNA from 100 ng of total RNA was performed using the iScript kit (Bio-Rad) according to the manufacturer's instructions. For cloning procedures, all primers were ordered from Eurogentec (Angers, France) and PCR amplification was performed using Accuprime Pfx Taq (Life Technologies, Illkirch, France) according to the manufacturer's instructions and optimized according to each primer pairing. All primers were designed using L. bicolor genome version 2.0 (http://genome.jgi-psf.org/; Martin et al., 2008) and Populus trichocarpa genome version 3.0 (www.phytozome.net; Tuskan et al., 2006). All vectors used in this study were GATEWAY compatible (Invitrogen) and BP and LR clonase recombination reactions were performed according to instructions provided by the manufacturer (Invitrogen). Escherichia coli strain DH5α was utilized for all subcloning procedures.

Microscopic mycorrhizal analysis

We used a microscopic method to analyze the development of the Hartig net as described by Plett et al. (2011). Briefly, mycorrhizal root samples were fixed via vacuum infiltration on ice in 4% paraformaldehyde for 20 min and left for 24 h at 4°C and then embedded in 6% agarose. Sections, 25–30 μm thick, of the mycorrhizal roots were taken using a Leica 1200 series vibratome (Lognes, France). Attention was paid to always taking sections in the middle of the colonized root (c. 2 mm from the root apex) to ensure the ability to compare the development of the Hartig net between samples. The development of the Hartig net is defined here as the depth of penetration within the apoplastic space between the rhizodermal cells of the root. The depth of penetration was determined using ImageJ analysis of our microscopic images (http://rsb.info.nih.gov/ij/). The data presented in this paper for the development of the Hartig net are the average of at least three biological replicates and a minimum of five technical replicates from each sample.

At least three independent biological replicates were performed for each test outlined in this paper, unless otherwise noted, to ensure reproducibility and significance of data reported. A Student's two-tailed independent t-test performed using Excel was used to determine the significance of the results obtained (< 0.05), unless otherwise noted.

Transgenic over-expression of PtERF1 in Populus

ET induces the expression of a suite of transcription factors called ETHYLENE RESPONSE FACTORs (ERFs). We chose to assay the effect that transgenic alteration of PtERF1 (Potri.001G154100.1) transcription under the control of a cauliflower mosaic virus 35S promotor had on the ability of L. bicolor to colonize poplar roots. We used Agrobacterium rhizogenes strain 15834 to generate transformed roots of P. tremula × P. alba 717-1B4 using a technique similar to that described in Chabaud et al. (2006). Briefly, the coding sequence of PtERF1 (Potri.001G154100.1) was cloned into the vector pH2GW7 for over-expression. These vectors were transformed into A. rhizogenes strain 15834 and selected on LB (Luria broth) medium supplemented with 75 μg ml−1 spectinomycin. To generate transgenic roots, 10 2-cm shoot cuttings of P. tremula × P. alba 717-1B4 were taken from 1.5-month-old plants and a sterile syringe tip coated in transgenic A. rhizogenes was used to puncture the stem of the cutting several times. These cuttings were placed in MS agar medium and left to root at 22°C for 3 wk. In vitro mycorrhiza formation experiments were carried out as above in ‘Plant and fungal materials’. Within this paper, each transformed root from independent plants is considered as an independent transformation event, and the development of the Hartig net between the three mycorrhizal root tips from that main transformed root was used to calculate the standard deviation within each independently transformed root.

Whole-genome oligoarray analysis

For expression experiments, RNA from three biological replicates was extracted as described in ‘DNA recombination procedures’ above and cDNA was synthesized using the SMART PCR cDNA Synthesis Kit (Clontech, Lognes, France) according to the manufacturer's instructions for microarray analysis. Microarray experiments were performed as described previously (Plett et al., 2011). A Student t-test with false discovery rate (FDR) (Benjamini–Hochberg) multiple testing correction was applied to the data using the arraystar software (DNASTAR; DNAStar Inc., Madison, WI, USA). The complete expression data set is available as series (accession number GSE53475) at the Gene Expression Omnibus at NCBI (National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/geo/). Real-time PCR analyses were performed using three biological replicates with a technical replicate for each reaction using the SYBRGreen Supermix following the manufacturer's instructions (Bio-Rad). Fold changes in gene expression between treated and control roots were based on ΔΔCt calculations according to Pfaffl (2001). Heatmaps and associated hierarchical clustering dendrograms were generated using the ‘pheatmap’ library in the R statistics package (Kolde, 2013). Data clustering was performed for both the genes (rows) and treatments (columns) with ‘euclidean’ as the distance measure. Pearson's correlation coefficients and the significance levels between all possible pairs of treated tissues were calculated using the ‘rcorr’ function of the ‘Hmisc’ package in R (Harrell, 2013).

Results

ET limits Laccaria bicolor colonization of poplar root tissues

ET concentrations have been shown to increase during the mycorrhiza formation process (Splivallo et al., 2009). As ET perception has been shown to limit fungal presence within plant tissues (Berrocal-Lobo et al., 2002; Lorenzo et al., 2003; Gutterson & Reuber, 2004; McGrath et al., 2005; Camehl et al., 2010; Lai et al., 2013; Lu et al., 2013), we were interested in determining how ET affected the colonization of root tissues by the mutualistic ECM fungus L. bicolor. Roots colonized by ECM fungi exhibited the different specialized zones typical of colonized roots: the fungal hyphae surrounded the root to form a compact sheath around the root called the mantle (demarked as M in Fig. 1a) under control conditions; and formation of the Hartig net (HN in Fig. 1a). In fully colonized roots, plant rhizodermal cells became longer and thinner (Fig. 1a). During this stage, the connection with neighboring plant cells is lost on two or three sides as the fungal hyphae grow into the apoplastic space and surround the plant cells. This morphology is in stark contrast to rhizodermal cells of poplar roots grown without fungal colonization (Fig. 1b). In this instance, plant cells maintain a more circular morphology and they remain attached on all three sides to adjacent plant cells. To determine how ET affects both stages of this colonization process, we tested the ability of L. bicolor to colonize the roots of transgenic P. tremula × P. tremuloides T89 lines that either produced higher concentrations of ACC (35S::PttACO1 lines 13 and 23), the precursor of ET, or transgenic plant lines that were insensitive to ET (35S::Atetr1-1 lines 1E and 3A; Love et al., 2009). While ET concentrations were not quantified in 35S::PttACO1, the expression of several ERF genes was significantly increased compared with nontransformed controls, indicating that ET was being over-produced (Supporting Information Table S1), while several ERF genes were down-regulated in 35S::Atetr1-1, indicating that these plants were less responsive to ET (Table S2). We observed that neither ACC over-production nor ET insensitivity on the part of the plant affected the formation of the fungal mantle around lateral roots (Fig. 1c). In assessing the formation of the Hartig net, however, we found that 35S::ACO1 lines had significantly reduced development of the Hartig net as determined by the depth of hyphal penetration between the rhizodermal cells after 2 wk of colonization by L. bicolor (Fig. 1d). In the 35S::PttACO1 mutants, we also did not see the alterations to rhizodermal cellular morphology (e.g. cell lengthening and cell detachment) typically seen under control colonization conditions (Fig. 1e). The morphology of the ET-insensitive 35S::Atetr1-1 roots colonized by L. bicolor was identical to that of nontransgenic P. tremula × P. tremuloides T89 with a similar development of the Hartig net (Fig. 1d) and where the rhizodermal cells exhibited a typical elongated morphology (Fig. 1f). Therefore, plants with increases in ET-induced gene expression, as determined by ERF expression analysis, exhibited reduced formation of the Hartig net.

Figure 1.

Increased ethylene signaling reduces the development of the Hartig net. (a) Transverse cross-section of a poplar root colonized by Laccaria bicolor under control conditions after 14 d of contact. M, mantle; E, epidermal cell; PC, parenchyma cell; HN, Hartig net. (b) Transverse cross-section of a poplar root colonized without fungal contact. (c) Percentage of roots exhibiting a fungal mantle. (d) Measurements of Hartig net depth of wild-type Populus tremula × Populus tremuloides clone T89 after 14 d of contact with L. bicolor compared with L. bicolor colonization of root systems of plants insensitive to ethylene (35S::Atetr1-1 line 1E and 35S::Atetr1-1 line 3A) or root systems that produce higher levels of 1-aminocyclopropane-1-carboxylic acid (ACC) (35S::PttACO1 line 13 and 35S::PttACO1 line 23). Error bars represent standard error (± SE); significant difference from wild type: *, < 0.05. (e) Transverse cross-section of a 35S::PttACO1 root colonized by L. bicolor under control conditions after 14 d of contact. (f) Transverse cross-section of a 35S::Atetr1-1 root colonized by L. bicolor under control conditions after 14 d of contact. Bars, 10 μm.

Increased expression of PtERF1 unlikely to inhibit Hartig net formation in 35S:ACO1

Of the ERF genes up-regulated by more than two-fold in 35S::PttACO1, compared with control tissues, three were annotated as ERF1-like genes (Potri.001G154100, Potri.004G051700, and Potri.011G061700; Table S1). Given the role of ET signaling, and the ERF1 gene in particular, in mediating plant defense responses against fungi (Berrocal-Lobo et al., 2002; Lorenzo et al., 2003; Gutterson & Reuber, 2004; McGrath et al., 2005; Lai et al., 2013; Lu et al., 2013), we wanted to determine what effect over-expression of PtERF1 alone would have on the colonization of poplar root tissues by L. bicolor. For this analysis, we focused our attention on the closest homolog to AtERF1 from these three poplar genes, Potri.001G154100 (Fig. S1). It was found that over-expression of PtERF1 only significantly inhibited the penetration of L. bicolor into the root apoplastic space at very high expression levels (> 120×; Fig. 2). PtERF1 was only found to be 13-fold up-regulated in 35S::ACO1 colonized tissues (Table S1). Therefore, at the expression levels of PtERF1 found in 35S::ACO1 mycorrhizal root tips, this gene probably does not play a large role in limiting the interaction between L. bicolor and Populus roots.

Figure 2.

High transcript abundance of ETHYLENE RESPONSE FACTOR1 (PtERF1) in the roots of poplar impedes the development of the Hartig net by Laccaria bicolor. (a) Depth of the L. bicolor Hartig net in the roots of wild-type P. tremula × P. alba clone 717-1B4 and three independent transformant 717-1B4 lines over-expressing PtERF1 under the control of the cauliflower mosaic virus (CaMV) 35S constitutive promoter (measurments in μm). (b) Fold change of the up-regulated expression of PtaERF1 in the three independent lines of 35S::PtERF1 shown in (a) as determined by Q-PCR. Expression levels are expressed as fold difference from wild-type P. tremula × P. alba clone 717-1B4 roots of the same age. Pt, Populus trichocarpa. *, < 0.05; Error bars represent standard error (± SE).

Over-expression of PttACO1 results in the repression of genes associated with plant–microbe recognition and cell wall biosynthesis

As increased expression of PttACO1 resulted in the repression of Hartig net development during L. bicolor colonization of root tissues, we were interested in determining what genes were significantly regulated in 35S::PttACO1 as opposed to both wild type (P. tremula × P. tremuloides T89) and 35S::etr1-1 using whole-genome oligoarrays (Table S3). Three gene groups that showed increased expression in 35S::PttACO1 compared with the other two host plant lines were receptor-like kinase genes, transcriptional regulators and stress response genes (Table 1). Of note within the genes associated with a kinase function was the up-regulation of a lectin protein kinase (Potri.004G027300). Lectin receptor-like kinases, as part of the PRR family of proteins (De Felice & Wilson, 2010), are transcriptionally induced upon interaction with mutualistic fungi (Gaude et al., 2012; Sanabria et al., 2012), indicating that these receptors form another important component in the interaction of plant tissues with mutualistic organisms. Further, a homolog of CLAVATA3 (Potri.008G191500) and a cysteine-rich receptor-like kinase (Potri.004G024000), which have characterized roles in the perception and signaling during pathogenic plant–microbe interactions (Rayapuram et al., 2012; Schwessinger & Ronald, 2012; Singh et al., 2012), were found to be up-regulated during the interaction. Within the transcription factors, from which we excluded ERF-like genes, we noted the increased expression of a WUSCHEL-related gene (Potri.010G111400), ARGONAUTE 7 (Potri.018G019500) and three ARABIDOPSIS RESPONSE REGULATOR (ARR) genes (Potri.002G152900, Potri.008G181000, and Potri.008G135500). Of the genes normally induced in response to stress, an ARMADILLO (ARM) gene was especially highly regulated (34 ×; Potri.006G240400) and three disease resistance proteins of the NB-ARC (nucleotide-binding adaptor shared by APAF-1, R proteins, and CED-4 domain) or NBS-LRR TIR (nucleotide-binding site/leucine-rich repeat Toll/interleukin-1 receptor) class were also found to be up-regulated (Potri.011G046900, Potri.T014500, and Potri.T047500).

Table 1. Genes that were more highly expressed in 35S::PttACO1 roots compared with either Populus tremula × Populus tremuloides clone T89 or 35S::Atetr1-1 roots undergoing colonization by Laccaria bicolor
 Gene model35S::ACO1 ECM versus controlT89 ECM versus control35S::etr1-1 versus control
  1. AHP5, Arabidopsis histidine phosphotransfer protein 5; AP2, apetala2; ARM, Armadillo; CYP, cyclophilin; ESR, embryo surrounding region; HSFA, Heat Stress Factor A; LRR, leucine-rich repeat; MAPKK, mitogen activated protein kinase kinase; MYB, myeloblastosis; NB-ARC, nucleotide-binding adaptor shared by APAF-1, R proteins, and CED-4 domain; NF-YB3, nuclear factor Y subunit B3; PDR6, ATP-binding cassette superfamily transporters; RLK, receptor-like kinase; SIK1, Salt-inducible kinase 1; TIR, Toll/interleukin-1 receptor; TIR-NBS-LRR, nucleotide-binding site/leucine-rich repeat Toll/interleukin-1 receptor.

Receptor/kinase function
S-locus lectin protein kinase family proteinPotri.004G02730043.01.01.0
AHP5; histidine phosphotransfer kinasePotri.004G18590041.41.01.0
CLAVATA3/ESR-RELATED 25Potri.008G1915003.31.71.5
Serine/threonine protein kinasePotri.010G1183003.01.71.9
SIK1; serine/threonine kinasePotri.010G1574002.91.21.3
CYSTEINE-RICH RLK10Potri.004G0240002.41.21.6
MAPKK10Potri.001G1388002.41.31.7
Transcription factors
WUSCHEL-related homeobox 1Potri.010G1114005.91.00.9
Zinc finger family proteinPotri.001G3096004.21.71.9
AtHASFA6B transcription factorPotri.002G0482004.01.31.1
PHYTOCHROME INTERACTING FACTOR 3Potri.014G1114003.50.90.5
NF-YB3 transcription factorPotri.014G1678003.21.81.9
AP2 domain-containing transcription factor TINYPotri.001G1557003.20.91.0
ARGONAUTE7Potri.018G0195003.21.71.8
ARABIDOPSIS THALIANA HOMEOBOX PROTEIN 40Potri.005G1261003.01.91.9
G-BOX BINDING FACTOR 3Potri.014G0942002.91.41.8
Basic helix-loop-helix (bHLH) family proteinPotri.008G1618002.91.91.9
AtMYB102Potri.004G0331002.91.01.1
NF-YB3 transcription factorPotri.001G3675002.91.81.9
ARABIDOPSIS RESPONSE REGULATOR 2Potri.002G1529002.81.41.8
ARABIDOPSIS RESPONSE REGULATOR 11Potri.008G1810002.41.41.7
ARABIDOPSIS RESPONSE REGULATOR 1Potri.008G1355002.01.71.8
Stress response
ARM repeat superfamily proteinPotri.006G24040034.31.01.6
PDR6 transmembrane movement of substancesPotri.010G2358009.10.30.3
Heat shock protein 70Potri.003G1840003.21.91.8
Glutaredoxin family proteinPotri.014G1342002.91.11.2
Heat shock proteinPotri.009G0502002.81.61.5
CYP714A1Potri.010G1163002.70.30.6
Disease resistance protein (TIR class)Potri.011G0469002.41.61.8
LRR and NB-ARC disease resistance proteinPotri.T0145002.31.81.9
Disease resistance protein (TIR-NBS-LRR class)Potri.T0475002.01.21.4

Three major classes of genes were found to be down-regulated in 35S::PttACO1 compared with both wild type and 35S::Atetr1-1 (Table 2). There were a number of down-regulated transcription factors, with MYB3 (myeloblastosis3) (Potri006G275900) being the most repressed gene in the list. There were also a number of genes implicated in pollen–pistle recognition, including a series of three serine/threonine kinases (Potri.010G017300, Potri.011G128700 and Potri.T022200). Two receptor-like lectin kinases (Potri.009G035500 and Potri.009G036500), receptors that are important in plant–microbial cellular recognition (Singh et al., 2012; Bouwmeester et al., 2013; Singh & Zimmerli, 2013), were also found to be significantly down-regulated. The third major class of genes found to be down-regulated in the 35S::PttACO1 mutant were genes related to plant cell wall modification (Table 2). These included a number of cellulose synthases (e.g. Potri.003G142300), an expansin precursor (Potri.018G098200) and a number of pectin esterases (e.g. Potri.013G013200).

Table 2. Genes that exhibited lower expression in 35S::PttACO1 roots compared with either Populus tremula × Populus tremuloides clone T89 or 35S::Atetr1-1 roots undergoing colonization by Laccaria bicolor
 Gene model35S::ACO1 ECM versus controlT89 ECM versus control35S::etr1-1 versus control
  1. ARK, A. thaliana receptor kinase; MYB, myeloblastosis; WOX, WUSCHEL-related homeobox ; BGLU, B glucosidase; CSLG3, cellulose synthase-like protein G3; CHIB, chitinase B; LAC, Laccase; RGP1, retrograde golgi transport homolog protein1.

Cellular recognition
Serine/threonine kinasePotri.010G0255000.11.35.5
ARK3 serine/threonine kinasePotri.001G4138000.21.12.2
ARK3 serine/threonine kinasePotri.010G0179000.21.32.4
RLK (receptor lectin kinase); kinasePotri.009G0355000.31.10.6
Serine/threonine kinasePotri.010G0173000.31.28.2
Serine/threonine kinasePotri.011G1287000.30.90.6
Serine/threonine kinasePotri.T0222000.40.70.8
Lectin protein kinase family proteinPotri.009G0365000.41.00.6
Transcription factors
MYB3Potri.006G2759000.01.81.3
SCARECROW-LIKE 21Potri.019G0856000.10.60.9
AtMYB103Potri.001G4705000.30.60.6
WOX11Potri.013G0669000.20.71.2
Cell wall synthesis
EXPANSIN-LIKE B2 PRECURSORPotri.018G0982000.01.10.7
BGLU17Potri.001G2233000.00.80.5
BGLU12Potri.001G2229000.00.90.6
CSLG3; cellulose synthasePotri.003G1423000.10.70.5
CHIB; chitinasePotri.017G0156000.10.90.7
CSLG3; cellulose synthasePotri.003G1423000.10.70.5
Cellulose synthasePotri.003G1423000.10.80.6
PectinesterasePotri.013G0132000.30.60.6
PectinesterasePotri.005G0227000.30.60.6
PectinesterasePotri.010G2477000.40.80.8
LAC12; laccasePotri.010G1835000.41.10.9
RGP1; cellulose synthasePotri.015G0993000.40.60.7
LAC5; laccasePotri.008G0738000.51.10.8

JA and SA treatments of poplar roots show opposing effects on L. bicolor colonization

As ET-induced pathways typically interact with the pathways controlled by JA and SA, we wished to determine how exogenous application of JA and SA would affect colonization of roots by L. bicolor. This was compared with exogenous treatment with ACC (a precursor of ET) which was expected to phenocopy the results obtained with transgenic P. tremula × P. tremuloides T89 over-expressing the ACC oxidase gene ACO1 (Fig. 1). Transcriptomic analysis of plant tissues treated with these three chemicals showed up-regulation of marker gene transcription associated with each pathway (Table S4), and therefore we can be confident that the compounds and concentrations used were sufficient to induce JA-, SA- and ET-controlled pathways.

In all cases, mantle formation was not significantly affected by hormone treatment (Fig. 3a). As found in 35S::PttACO1 lines, treatment with ACC significantly inhibited Hartig net formation (Fig. 3b). Similarly, treatment with JA resulted in an almost complete arrest of fungal penetration between rhizodermal cells, while SA treatment had no affect on Hartig net formation (Fig. 3b). In both ACC- and JA-treated roots, we did not see alterations in rhizodermal cellular morphology (Fig. 3c–e). The morphology of SA-treated root cells colonized by L. bicolor (Fig. 3f) was identical to that of untreated colonized control roots (Fig. 3g), with rhizodermal cells exhibiting elongated morphology and fungal hyphae penetrating between the cells.

Figure 3.

Treatment of developing ectomycorrhizal root tips with 1-aminocyclopropane-1-carboxylic acid (ACC) and jasmonic acid (JA) affects the development of the Hartig net. (a) Percentage of roots exhibiting a fungal mantle under control, untreated conditions and ACC-, JA-, methyl jasmonate (MeJA)- or salicylic acid (SA)-treated conditions. (b) Depth (in μm) of the Hartig net of Laccaria bicolor colonizing roots of Populus tremula × Populus alba 717-1B4 under control conditions (Cont.) and under treatment with ACC, JA, MeJA or SA. All values are the average of three biological replicates + SE. Significant difference from control conditions: *, < 0.05. (c) Transverse cross-sections of colonized roots under the same conditions: ACC (c), JA (d), MeJA (e), SA (f) and control (g). Bars, 10 μm.

A subset of genes regulated by ACC, JA and SA are similarly regulated in late-stage colonized root tissues

Given the effects of ET, JA and SA on fungal colonization (Figs 1, 3), and given previous work that has demonstrated that all three of these hormones are produced during the interaction between mutualistic fungi and their host plants (Hause et al., 2002; Luo et al., 2009; Splivallo et al., 2009), we were interested in determining the timing with which plant genes normally regulated by ET, JA or SA were induced during colonization of Populus by L. bicolor. To determine this, we compared genes regulated in P. tremula × P. alba 717-1B4 roots at early (3 d post plant–fungal contact; gene regulation as reported by Felten et al., 2009) and late (14 d post plant–fungal contact) stages of colonization by L. bicolor with gene expression in roots treated with exogenous application of ACC (the precursor of ET in plants), JA or SA (Fig. 4).

Figure 4.

Treatment of developing ectomycorrhizal (ECM) root tips with exogenous 1-aminocyclopropane-1-carboxylic acid (ACC) or jasmonic acid (JA) induces alterations to the root transcriptome that mimic transcriptomic alterations of late-stage mycorrhizal root tips. (a) Heat-map of log2 fold change in gene expression of colonized roots versus control roots (i.e. no fungal contact) after 3 d of contact between Laccaria bicolor and poplar roots (results obtained from Felten et al., 2009) compared with gene regulation in ACC colonization conditions versus control roots, JA colonization conditions versus control roots and salicylic acid (SA) colonization conditions versus control roots. Hierarchical clustering analysis is depicted above the heat-map. (b) Heat-map of log2 fold change in gene expression of colonized roots versus control roots (i.e. no fungal contact) after 14 d of contact between L. bicolor and poplar roots. To be included in the heat-map, the gene must be significantly regulated in at least one of the conditions considered: control colonization conditions versus control roots; ACC colonization conditions versus control roots; JA colonization conditions versus control roots; SA colonization conditions versus control roots. Hierarchical clustering analysis is depicted above the heat-map. Supporting Information Tables S5 and S6 contain expression values for all genes included in the heat-maps.

Of the poplar genes significantly regulated by the presence of L. bicolor after 3 d post contact with P. tremula × P. alba 717-1B4 (= 507 genes; as reported in Felten et al., 2009), we found a significant negative correlation to gene expression regulated by both JA (= −0.27) and SA treatments (= −0.12), while there was no significant correlation to gene expression in ACC-treated roots (= −0.07; < 0.05; Fig. 4a; Table S5). We found the opposite to be true in fully colonized roots. Using a hierarchical cluster analysis based on the Euclidean distance of all genes significantly up- and down-regulated after 14 d of contact between L. bicolor and P. tremula × P. alba 717-1B4 (= 3515), we found that there was a significant positive correlation with gene expression in ACC-treated tissues (= 0.78), JA-treated tissues (= 0.73) and SA-treated tissues (= 0.64; < 0.05; Fig. 4b; Table S6). In the significantly regulated genes that showed similar expression between nontreated and hormone-treated root tissues, we found a large number of genes encoding signaling proteins (e.g. Potri.0003G222000, Potri.0011G028700, Potri.0008G166000, Potri.0003G213300, Potri.0018G151200 and Potri.0001G153200), NAC domain proteins (e.g. Potri.0004G107400 and Potri.0017G031600) and metal-binding proteins (e.g. Potri.0010G206300, Potri.0002G021000 and Potri.0008G220500; Table 3). Therefore, our results demonstrate that a subset of genes whose expression is controlled by ET, JA and SA are differentially regulated during the later stages of root colonization by L. bicolor. As these hormones control plant defense responses against microbes, this indicates that during the last stages of colonization the plant is mounting a defense against the fungus, possibly to limit its growth within root tissues.

Table 3. Most similarly expressed genes in untreated Populus tremula × Populus alba 717-1B4 roots colonized by Laccaria bicolor and colonized root tissues treated with 1-aminocyclopropane-1-carboxylic acid (ACC), jasmonic acid (JA) or salicylic acid (SA)
 Control ECM versus control rootP-value (control ECM versus control root)ACC ECM versus control rootP-value (ACC ECM vs control root)JA ECM versus control rootP-value (JA ECM vs control root)SA ECM versus control rootP-value (SA ECM versus control root)Arabidopsis homolog
  1. AIR3, AUXIN-INDUCED IN ROOT CULTURES 3; ATL2, ARABIDOPSIS TOXICOS EN LEVADURA2; SKS18, SKU5 SIMILAR 18; SKU5, Extracellular Glycosyl Phosphatidylinositol–Anchored Glycoprotein.

Potri.011G050000.2170.61.6E-12226.64.9E-1377.71.4E-036.58.81E-01AIR3; serine-type endopeptidase
Potri.003G222000.182.62.4E-0498.25.3E-0517.93.7E-01405.62.20E-09Receptor-Like Protein 46 (AtRLP46); kinase/protein binding
Potri.011G028700.178.21.5E-0494.02.0E-04545.05.4E-1393.31.30E-01CYSTEINE-RICH RLK10 (CRK10); ATP-binding/kinase/protein kinase/protein serine/threonine kinase/protein tyrosine kinase
Potri.012G117000.171.82.4E-0468.13.1E-0452.04.9E-0336.92.07E-01Root hair defective 3 GTP-binding (RHD3) family protein
Potri.017G127100.167.57.2E-0656.34.0E-06112.43.6E-08327.11.63E-07Glutamine synthetase 1;5 (GLN1;5); glutamate-ammonia ligase
Potri.004G124900.164.97.5E-0684.82.6E-0724.58.9E-0382.39.47E-05Embryo defective 2076 (emb2076)
Potri.008G166000.160.16.3E-1386.54.7E-1375.41.0E-12112.75.49E-11Ethylene-responsive factor, putative
Potri.001G153200.159.82.4E-0556.14.0E-05125.62.5E-08641.24.65E-13DNA binding
Potri.001G099500.155.14.9E-0245.06.0E-021.09.9E-013.21.00E+00Heavy-metal-associated domain-containing protein
Potri.004G107400.149.96.5E-0668.23.5E-0727.71.4E-02207.88.69E-12ARABIDOPSIS NAC-DOMAIN PROTEIN 101 (ANAC101); transcription activator/transcription factor/transcription regulator
Potri.002G141800.146.31.5E-0940.44.1E-0926.85.8E-07591.94.64E-13Malate dehydrogenase, cytosolic, putative
Potri.009G170200.145.31.5E-0354.68.1E-0519.34.2E-02178.22.04E-07Lectin protein kinase family protein
Potri.009G104200.142.96.0E-0438.51.1E-0340.52.0E-03316.97.67E-11Hydrolase
Potri.011G050200.141.06.1E-1253.96.0E-1327.93.0E-0614.52.53E-02AIR3; serine-type endopeptidase
Potri.006G157100.140.86.3E-0330.31.8E-0212.62.0E-01290.03.19E-09Transferase family protein
Potri.010G100000.140.04.5E-0328.91.9E-0243.55.0E-03293.04.00E-08ATA1 (ARABIDOPSIS TAPETUM 1); binding/catalytic/oxidoreductase
Potri.003G106800.138.61.9E-0436.72.3E-041.04.9E-0153.01.17E-03SWI1 (SWITCH1); phospholipase C
Potri.010G206300.134.54.5E-0436.82.2E-0431.61.2E-03346.63.03E-11ATL2; zinc ion binding
Potri.002G021000.132.74.8E-0326.01.3E-021.08.9E-0184.42.11E-03sks18 (SKU5 Similar 18); copper ion binding
Potri.014G183700.132.33.2E-0328.66.5E-0371.41.6E-06240.22.32E-11Catalytic/cation binding/hydrolase, hydrolyzing O-glycosyl compounds
Potri.004G229400.130.53.2E-0330.22.3E-0320.13.8E-02271.77.56E-09ATP-dependent Clp protease ATP-binding subunit ClpX, putative
Potri.008G220500.129.52.3E-0234.79.1E-041.08.8E-0152.21.48E-02Zinc ion binding
Potri.019G023000.128.72.4E-0220.12.5E-01280.53.3E-11177.01.28E-06ATTPS-CIN (terpene synthase-like sequence-1,8-cineole); (E)-beta-ocimene synthase/myrcene synthase
Potri.004G124100.128.44.0E-0242.75.7E-0310.53.5E-01237.47.36E-09Lyase/pectate lyase
Potri.006G277300.127.22.8E-0423.06.3E-041.81.9E-0140.74.24E-02BETA-HYDROXYISOBUTYRYL-COA HYDROLASE 1 (CHY1); 3-hydroxyisobutyryl-CoA hydrolase
Potri.T018300.126.44.1E-0224.04.7E-0210.77.6E-0171.74.96E-03Transferase/transferase, transferring acyl groups other than amino-acyl groups
Potri.007G122900.125.12.9E-0225.92.5E-021.05.1E-011.01.62E-02ISOCITRATE LYASE (ICL); catalytic/isocitrate lyase
Potri.003G213300.125.05.4E-0320.91.1E-023.46.6E-0125.94.92E-02Zinc finger (GATA type) family protein
Potri.014G166400.124.68.9E-0421.52.7E-0354.31.8E-05202.86.30E-13STERILE APETALA (SAP); transcription factor/transcription regulator
Potri.018G151200.123.62.5E-0232.86.4E-0324.34.9E-0259.85.37E-03Zinc finger (CCCH-type) family protein
Potri.001G308100.122.83.0E-0416.03.5E-039.68.2E-0252.82.71E-04CRUCIFERIN 3 (CRU3); nutrient reservoir
Potri.017G118600.122.81.7E-0729.06.7E-091.49.3E-0111.03.22E-01Hydrolase
Potri.004G016400.117.93.6E-0219.52.4E-023.49.7E-016.08.65E-01MATE efflux family protein
Potri.017G111200.116.21.0E-0322.51.2E-054.06.3E-012.59.37E-01Aminotransferase class I and II family protein
Potri.017G031600.116.03.5E-0212.48.9E-025.26.2E-0112.14.13E-01ARABIDOPSIS NAC DOMAIN CONTAINING PROTEIN 83 (ANAC083); transcription factor

Of the genes that are expressed during the normal course of the interaction between L. bicolor and Populus (i.e. ‘control’ colonization conditions), a number were also differentially regulated by exogenous ACC, JA or SA treatment. As seen in 35S::PttACO1 plants, gene regulation after ACC treatment (as compared to untreated colonized tissues) led to an increase in the transcription of genes encoding NBS-LRR disease resistance proteins, self-incompatibility proteins, CLE (CLV3/ESR-related) peptides and wound-responsive genes, while genes encoding expansins, those encoding receptor-like kinases, and a number of MYB genes were found at a lower abundance (Table S6). JA treatment led to the increased transcript abundance of several stress-related genes and a repression of cell wall active enzymes. Of genes with elevated transcription in JA-treated roots, we found seven pathogenesis-related thaumatin family proteins, three terpene synthases, three receptor-like kinases, 11 leucine-rich receptor-like kinases, three NBS-LRR class disease resistance proteins, three chitinase genes and six extensin genes (Table S6). Of those genes repressed by JA treatment, we found seven genes encoding pectin esterases, a number encoding xyloglucan transferases, three encoding expansins and three encoding pectinases. SA induced the heightened expression of genes encoding receptor-like kinases, xyloglucan transferases, a number of pathogenesis-related thaumatin family proteins and six NBS-LRR disease resistance proteins. We found that SA treatment also induced the expression of three pectin esterase genes, two COBRA genes, two expansin genes, six glucosyl hydrolase genes and five lectin protein-kinase genes, while the expression of genes encoding senescence-related proteins, 12 leucine-rich receptor-like kinases and three chitanases were repressed (Table S6). It is of note that SA treatment also resulted in the repression of eight CLE family genes.

ET and JA alter the expression of a transcriptional core that is opposed by SA

Given the extensive body of literature supporting the theory that ET and JA treatments induce similar plant gene networks, and that these genes are regulated in an opposing manner by SA, we searched within our expression data sets to determine whether there were genes whose transcription was affected in a similar manner by ET and JA treatments but in an opposite manner by SA treatment during L. bicolor colonization of poplar tissues. We found that 133 genes were regulated in this manner, including 106 genes significantly up-regulated in JA- and ET-treated tissues and down-regulated in SA-treated tissues (Fig. 5; Table S7). Of those genes co-induced in JA- and ET-treated tissues, we found that there were a number of transcription factors, such as MYB61 and MYB66 (Potri.13G001000 and Potri.018G049000), and two bZIP (Basic Leucine Zipper Domain) transcription factor family proteins (Potri.019G091900 and Potri.013G124400). There were also two lectin receptor-like protein kinases (Potri.001G411100 and Potri.001G411400) that were down-regulated by JA and ET but were significantly up-regulated by SA treatment (Table S7), as well as a number of leucine-rich receptor-like kinases (e.g. Potri.019G021700) and NAC domain-containing proteins (Potri.003G113000 and Potri.016G076000) differentially regulated between the hormone treatments. Therefore, while ET, JA and SA all induce a higher level of defense-related signaling genes they have opposing effects on a number of receptor-like kinases and cell wall active enzymes.

Figure 5.

Ethylene (ET) and jasmonic acid (JA) treatment of roots undergoing colonization by Laccaria bicolor induces gene regulation that is opposed by salicylic acid (SA) treatment. A heat-map is shown of poplar genes that exhibit similar expression patterns following ET and JA treatments of L. bicolor colonized roots compared with control roots but that are oppositely regulated by SA treatment. All values are log2 fold change in gene expression of colonized roots versus control roots (i.e. no fungal contact) after 14 d of contact between L. bicolor and poplar roots. Gene identities and fold change of genes included in the heat-map analysis can be found in Supporting Information Table S7. ECM, ectomycorrhizal.

Discussion

Plants encode multiple systems of perception that are utilized to recognize self versus nonself cells. These mechanisms of perception become critical in instances where microorganisms, especially pathogenic ones, attempt to colonize plant tissues. Three common plant-based defense mediators induced by microbial colonization are the hormones ET, JA and SA (Lamb et al., 1998; Ellis & Turner, 2001; Berrocal-Lobo et al., 2002; Lorenzo et al., 2003; Berrocal-Lobo & Molina, 2004; van Loon et al., 2006; Spoel & Dong, 2008; Mersmann et al., 2010). ET is an interesting compound as exogenous ET can increase the spread of certain pathogenic organisms (Bent et al., 1992; Lund et al., 1998; Hoffman et al., 1999) or it can suppresses growth of other mutualistic organisms within plant tissues (Waller et al., 2005; Camehl et al., 2010). Using the model system of the mutualistic ECM fungus L. bicolor and its host Populus, we sought to identify the role of ET signaling in controlling fungal colonization of plant tissues and to further our knowledge concerning the molecular events that underpin the role of ET in planta during the ECM symbiosis. We then compared and contrasted the impact of ET on this system with that of JA and SA to gain a better understanding of the interaction between the three hormonal systems.

Our hierarchical cluster analysis demonstrates that a portion of the gene repertoire whose transcription is controlled by ET is normally regulated during the later stages of the interaction between L. bicolor and poplar roots. This conclusion was reached in light of the significant correlation of gene expression between ACC-treated roots and late-stage mycorrhizal root tissues (= 0.78; < 0.05; Fig. 4b). During initial stages of fungal colonization, this correlation was not observed (= −0.07; > 0.05; Fig. 4a). As either exogenous application or transgenic modulation of ACC concentrations interferes with proper Hartig net formation, but not mantle formation, and because gene expression in both ACC-treated tissues and 35S::PttACO1 tissues indicated that ERF genes were induced above levels in control tissue (Tables S1, S4), we can infer that ET-induced gene transcription in the late stages of colonization results in a curtailing of the growth of ECM fungal hyphae within plant tissues. Our results are also similar to those of previous studies with another class of mutualistic fungi, the AM fungi, which also showed that intraradical fungal growth of Rhizophagus clarus, Endogone versiformis, R. irregularis and Glomus aggregatum is inhibited by ET (Geil et al., 2001; Penmetsa et al., 2008; Riedel et al., 2008; Zsögön et al., 2008). Similarly, hyphal growth in planta of the growth-promoting fungus Piriformospora indica is impeded by altered ET signaling in root tissues (Camehl et al., 2010). Interestingly, however, while ET insensitivity results in uncontrolled growth of infection threads during nodulation (Penmetsa & Cook, 1997; Penmetsa et al., 2003, 2008), ET insensitivity in 35S::Atetr1-1 lines did not significantly increase L. bicolor penetration into Populus root tissues. As JA signaling was found to closely mirror ET signaling in colonized root tissues, it is possible that JA signaling could mitigate the loss of ET signaling in 35S::etr1-1 lines and thus curtail uncontrolled growth of L. bicolor in those lines. Therefore, our results further the theory that ET signaling, as opposed to its role in aiding fungal colonization during pathogenic biotrophic interactions (Bent et al., 1992; Lund et al., 1998; Hoffman et al., 1999), physically limits the interaction between mutualistic organisms and their plant hosts (Waller et al., 2005; Camehl & Oelmüller, 2010; Camehl et al., 2010).

The ability of increased ET signaling to stop the formation of the Hartig net would suggest that ET has the ability to alter host physiology to physically restrain fungal growth in plant tissues in addition to its ability to induce systemic acquired resistance (Pieterse et al., 2009). This morphological phenotype is underpinned by the altered regulation in 35S::PttACO1 of a whole suite of genes associated with cell wall synthesis and modification, including the reduction in cellulose synthase genes, pectin esterases and expansins, to name a few (Tables 1, 2). In the early stages of the interaction between L. bicolor and Populus, these genes are highly up-regulated (Felten et al., 2009), probably to soften the connections between rhizodermal cells and to permit the elongation and detachment of plant cells in contact with L. bicolor hyphae. Therefore, the misregulation of these cell wall-restructuring genes by increased ET concentrations after ACC application or increased expression of ACO1 in the 35S::PttACO1 plants must inhibit the proper restructuring of plant cell walls that would normally allow the establishment of the Hartig net. These results further the concept that plant cell walls are a key aspect of plant defenses against microbes (Ellis et al., 2002; Humphrey et al., 2007) and that one of the main roles of ET in curtailing the growth of a mutualistic organism within plant tissues, be it an ECM fungus, rhizobial bacterium or AM fungus, is in controlling cell wall composition. The results are consistent with the role of ET in altering cell wall properties in the defense against Botrytis cinerea (Ferrari et al., 2007; Rowe et al., 2010; Lloyd et al., 2011; Ramirez et al., 2011). It is interesting, however, to note that the role of ET in altering root cellular walls is opposite to the role of ET signaling in fruit tissues. During the ripening process, increased ET biosynthesis results in the softening of cell walls via the activity of polygalacturonases and as a result of pectin solubilization and polyuronide depolymerization (Sanudo-Barajas et al., 2009). Our results therefore suggest a tissue-specific role for ET in cell wall dynamics. Therefore, limiting of L. bicolor growth within the root apoplastic space of 35S::PttACO1 lines is a result, in part, of the transcriptional alteration of genes associated with microbial perception as well as genes that impact the cell wall properties of Populus root cells.

ET is not the only plant hormone that has a role in limiting ECM hyphal penetration within host roots. Our results demonstrate that JA treatment of roots undergoing colonization by L. bicolor also inhibits the formation of the Hartig net, while SA treatment does not affect colonization. We found evidence not only for dual action between ET and JA at the morphological level but also for extensive cross-talk between the ET- and JA-induced transcriptomic pathways during colonization – cross-talk that was often in contrast to gene regulation by SA treatment (Fig. 5; Table S7). These include the regulation of CLE genes, leucine-rich receptor-like kinase genes, and lectin protein kinase genes. The CLAVATA family of leucine-rich repeat receptor-like kinases signal to inhibit microbial colonization of plant tissues (Schnabel et al., 2005; Mortier et al., 2010, 2012; Reid et al., 2011; Saur et al., 2011). As we found that one or more CLV (CLAVATA)-like genes were highly up-regulated in plants over-expressing PttACO1 and in ACC- and JA-treated tissues, these proteins could contribute to the increased resistance of host plants to L. bicolor colonization. We also found that JA and SA had a significant impact on the transcription of genes encoding many cell wall active enzymes, many of which were different from those affected in tissues treated with ACC. We found that JA treatment led to the increase in transcription of a number of extensin genes and reduced expression of three expansin genes. Expansins act to loosen cell walls to allow growth (Cho & Cosgrove, 2000; Choi et al., 2003; Lee et al., 2003), while extensins are proteins that integrate into the cell wall after expansion and lend rigidity to the structure (Kieliszewski & Lamport, 1994; Borner et al., 2002). Increased production of extensins and loss of expansin activity can lead to loss of cell elongation (Cho & Cosgrove, 2000; Hall & Cannon, 2002; Humphrey et al., 2007) and increased resistance to pathogenic microbes (Wei & Shirsat, 2006). Therefore, the differential regulation of expansins and extensins by JA correlates well with our morphological data, whereby JA treatment resulted in reduced penetration of L. bicolor hyphae into the root as well as the lack of rhizodermal cellular extension in response to the presence of L. bicolor (Fig. 3). SA treatment, meanwhile, resulted in dramatic growth changes to poplar rhizodermal cells in contact with L. bicolor, a re-orientation that was accompanied by an increase in the expression of genes correlated with cell growth and extension. Specifically, two genes encoding homologs of the Arabidopsis COBRA gene were up-regulated in SA-treated roots, as were two expansin genes. Like expansins, COBRA proteins are implicated in the growth and extension of plant cells (Schindelman et al., 2001; Roudier et al., 2005) and their induction correlates with the increased rhizodermal expansion induced by L. bicolor colonization of SA-treated roots. Therefore, ET/JA and SA show contrasting roles in the maintenance of cell walls during fungal challenge and colonization.

Taking these results together, we have found that JA and ET limit Hartig net development through a combination of ET/JA transcriptional cross-talk. Particularly in the early stages of Hartig net formation, root treatments that induce genes controlled by ET and JA can prevent symbiosis from being achieved. One interesting topic of future research is in the adaptations of mycorrhizal fungi for avoiding or controlling these plant defense responses. AM fungi, for example, use effector proteins (e.g. SP7 (Secreted Protein7); Kloppholz et al., 2011) to attenuate the ET-controlled pathway and support colonization. While no such countermeasure has been identified in L. bicolor to date, the key role of the effector protein MiSSP7 (Mycorrhizal induced Small Secreted Protein7) in Hartig net development (Plett et al., 2011) may suggest that it, like SP7, also has a role in controlling plant hormonal pathways to foster mutualistic associations.

Acknowledgements

This work was supported by the European Commission within the Project ENERGYPOPLAR (FP7-211917), and the ANR project FungEffector (through a grant to F.M.). This research was also sponsored by the Genomic Science Program (project ‘Plant-Microbe Interactions’), US Department of Energy, Office of Science, Biological and Environmental Research under the contract DE-AC05-00OR22725. We would like to thank M. Buée and Y. Dessaux for kindly supplying the strain 15834 of A. rhizogenes and J. Felten and V. Legué. This work was supported by the French National Research Agency through the Clusters of Excellence ARBRE (ANR-11-LABX-0002-01).

Ancillary