Calcium channel CNGC19 mediates basal defense signaling to regulate colonization by Piriformospora indica in Arabidopsis roots

The CNGC19 calcium channel plays an important role in regulating the colonization of Arabidopsis roots by Piriformospora indica via modulation of systemic and defense-related pathways of MAMP-triggered immunity.


Introduction
Piriformospora indica (syn. Serendipita indica) is a cultivable, root-colonizing endophytic fungus belonging to Sebacinales (Basidiomycota) (Verma et al., 1998;Weiß et al., 2016). It colonizes many plant species including Arabidopsis and promotes their growth (Varma et al., 1999;Peškan-Berghöfer et al., 2004;, enhances nutrient uptake (Yadav et al., 2010;Rani et al., 2016;Bakshi et al., 2017;Prasad et al., 2018), and imparts tolerance to abiotic and biotic stresses to a wide range of its hosts (Waller et al., 2005;Baltruschat et al., 2008;Jogawat et al., 2013Jogawat et al., , 2016Sun et al., 2014). It colonizes the root epidermal and cortex cells without penetrating the central cylinder, and displays a biphasic colonization strategy (Deshmukh et al., 2006;Zuccaro et al., 2011). However, the establishment of a beneficial plant-microbe interaction is not always harmonious, and rejection of the invading symbiont or control of its colonization can occur due to active plant defense ). Basal plant defense relies on the recognition of conserved microbial structures called microbe-associated molecular patterns (MAMPs), and is termed MAMP-triggered immunity (MTI) (Millet et al., 2010). In soil, plant roots perceive MAMPs using specific pattern-recognition receptor (PRR) proteins (Choi and Klessig, 2016). Recognition of MAMPs triggers downstream early plant-defense responses such as elevation of cytosolic calcium (Ca 2+ cyt ) and a burst of reactive oxygen species (ROS), which further activates mitogen-activated protein (MAP) kinase and various phytohormone pathways that stimulate defense-related pathways (Harper and Harmon, 2005;Ranf et al., 2011;Steinhorst and Kudla, 2013). Similar to plant pathogens, mutualists such as P. indica are also confronted with an effective innate immune system in roots, and the colonization success depends on the evolution of strategies for immunosuppression (Van Wees et al., 2008;Jacobs et al., 2011). During colonization by Serendipita indica on Arabidopsis, eATP, which acts as a damage-associated molecular pattern (DAMP) accumulates in the apoplast. Serendipita indica secrets SiE5′NT, an enzymatically active nucleotidase capable of hydrolysing eATP, in the apoplast and thus suppresses immunity (Nizam et al., 2019). During the early stages of mycorrhiza formation and P. indica colonization, H 2 O 2 is produced and its production declines when a mutualistic interaction is established (Fester and Hause, 2005;Matsuo et al., 2015). Piriformospora indica also actively represseses ROS accumulation by activating ROS-scavenging genes (Matsuo et al., 2015). To achieve a harmonious interaction with plants, P. indica also regulates biosynthesis and signaling of several phytohormones such as jasmonic acid (JA), gibberallins (GA), and ethylene (Camehl et al., 2010;Sun et al., 2014;Vahabi et al., 2015;Pan et al., 2017;Xu et al., 2018). Piriformospora indica association also alters callose deposition and defense-related metabolites, such as phytoalexins and glucosinolates (GS) (Jacobs et al., 2011;Lahrmann et al., 2015). Indole glucosinolates (iGS) are an important part of MTI in plants (Clay et al., 2009;Böhm et al., 2014) and they are found to be essential in balancing the beneficial interaction between P. indica and Arabidopsis (Nongbri et al., 2012;Lahrmann et al., 2015). In addition, P. indica also suppresses innate immunity upon encountering the flagellin 22 elicitor from bacteria (Jacobs et al., 2011). An active plant immunity and its suppression is thus critical for controlled P. indica colonization. The early-activated plant defense genes that are responsible for regulating the entry of the symbiont and its subsequent colonization are unknown.
Ca 2+ is a universal second messenger, activated very early in signaling cascades upon recognition of both pathogens and symbionts. Rhizobacteria-mediated nodulation and mycorrhiza formation are associated with oscillations in nuclear Ca 2+ in host plants. These oscillations upon perception of rhizobia and mycorrhiza activate induction of the common genes that are important for the establishment of the symbioses (Oldroyd, 2013). Colonization by P. indica in Arabidopsis is independent of these common arbuscular mycorrhizal symbiotic genes (Banhara et al., 2015). However, elevation of Ca 2+ is common with other symbiotic interactions, as P. indica cell-wall extract (PiCWE) elevates root Ca 2+ cyt and is crucial for growth promotion in Arabidopsis . Using the elevation of Ca 2+ as a marker, Johnson et al. (2018) identified cellotriose (CT) as the major elicitor in crude PiCWE. It was further confirmed that CT targets a poly(A)-specific ribonuclease in order to modulate plant responses such as elevation of Ca 2+ cyt , generation of ROS, expression of defense-related genes, phytohormonal signaling, and growth promotion. Elevation of Ca 2+ cyt requires entry of Ca 2+ either across the plasma membrane or from intracellular compartments. In Arabidopsis, ligand-gated channels such as cyclic nucleotide gated channels (CNGCs), glutamate receptor-like channels (GLRs), stretch-activated Ca 2+ channels (OSCAs), and the MID1-complementing activity (MCA) families are the four main plasma membrane Ca 2+ -permeable channels, whilst the slow vacuolar two-pore channel 1 (TPC1) is the key vacuolar channel (Dodd et al., 2010). The Arabidopsis genome encodes 20 members of the CNGC family, with roles in plant development and a functions related to biotic and biotic stresses (Meena and Vadassery, 2015;DeFalco et al., 2016). CNGC2, CNGC4, CNGC11, and CNGC12 have been reported to play crucial roles in defense against bacterial and fungal pathogens (Yoshioka et al., 2001;Ahn, 2007), and we have recently identified a role of the CNGC19 Ca 2+ channel in herbivoryinduced Ca 2+ flux and plant defense against Spodoptera litura (Meena et al., 2019).
CNGC15 has been identified as critical nuclear channel that generates oscillatory Ca 2+ signals during arbuscular mycorrhizal symbiosis with Medicago trancatula roots (Charpentier et al., 2016). In Lotus japonicus, a mutation in the AtCNGC19 homolog BRUSH is reported to result in impaired infection by nitrogen-fixing rhizobia due to a leaky channel (Chiasson et al., 2017). The identity of the channel involved in the elevation of Ca 2+ cyt that is induced by P. indica is not yet known. The P. indica elicitor CT induces expression of GLR Ca 2+ channels in Arabidopsis roots; however, (Johnson et al., 2018) found no functional roles for GLR3.3, GLR2.4, GLR2.5, and TPC1 in this response. Expression levels of CNGCs are altered upon PiCWE treatment in plant roots and these are the only other type of Ca 2+ channel known to be involved in the interaction . We therefore hypothesized that CNGCs might be involved in the generation of elevated Ca 2+ in Arabidopsis roots and in the downstream signaling in response to P. indica mutualism. Our results point to a role of Arabidopsis CNGC19 as an important gatekeeper to regulate P. indica colonization.

Plant and fungal interactions in soil and co-cultivation media
For soil experiments, seeds were sown in pots containing soilrite, Irish peat moss, and exfoliated vermiculite (1:1:1, w:w:w) and kept for 2 d at 4 °C in the dark for stratification. The soil was pre-mixed with 1% P. indica mycelia (w/w), and plants were grown for 6 weeks after stratification. Control plants had no P. indica mycelia in the soil. The plants were grown at 22 °C with a 10/14 h light/dark photoperiod and a light intensity of 150 µmol m −2 s −1 in the growth room. Samples were harvested at 42 d post-inoculation (dpi).
For plate experiments, seeds were surface-sterilized, stratified under the conditions described above, and placed on half-strength MS plates supplemented with 1% sucrose and 0.8% agar, and germinated for 7 d. The seedlings were grown at 22 °C with a 10/14 h light/dark photoperiod and a light intensity of 150 µmol m −2 s −1 in the growth room. They were then transferred to 1× PNM medium for co-cultivation with P. indica discs for 14 d (Johnson et al., 2011) under similar conditions. Samples were harvested at 2, 7, and 14 dpi.

Preparation of P. indica cell-wall extract and application on roots
Piriformospora indica cell-wall extract (PiCWE) was prepared as described by . In brief, the mycelia from 14-d-old liquid cultures were homogenized, filtered using nylon membranes, and washed three times with water, twice with chloroform/methanol (1:1), and finally twice with acetone. The mycelial cell wall material obtained was dried at room temperature, suspended in water, and autoclaved for 30 min at 121 °C. It was then filter-sterilized using a 0.22-μM filter, and 50 µl of the resulting extract was used per seedling root for experiments. The active elicitor of PiCWE was recently identified as cellotriose (CT; Johnson et al., 2018). CT (Sigma, C1167) and cellobiose (Sigma, C7252) were used for some experiments.

Plant treatments and gene expression analysis
Seedlings at 10 d old were treated with 100 μl PiCWE or 10 µM CT by adding to the MS media and were harvested at 0, 15, 30, 45, and 60 min. For the co-cultivation experiment, the seedlings were harvested at 2, 7, and 14 dpi. Each sample consisted of six seedlings and was ground to a fine powder in liquid N 2 , and total RNA was isolated using TRIzol Reagent (Invitrogen) according to the manufacturer's protocol. Four replicate samples were used. An additional DNAse (Turbo DNAse, Ambion) treatment was included to eliminate any contaminating DNA. cDNA synthesis was performed using a High Capacity cDNA kit (Applied Biosystems). Gene-specific primers were designed using the NCBI primer design tool (http://www.ncbi.nlm.nih.gov/tools/primerblast) and are listed in Supplementary Table S1 at JXB online. qRT-PCR was performed in optical 96-well plates on a CFX96 Real-Time PCR Detection System (Bio-Rad) using iTaq universal SYBR green Mix (Bio-Rad). AtActin2 (At3g18780) was used as the endogenous control for normalization of transcripts. The fold-induction values of the target genes were calculated using the ΔΔC T method (Livak and Schmittgen, 2001) and were expressed relative to the mRNA level of the genes in the control seedlings, the values of which were set as 1.

Detection and measurement of P. indica colonization
For tracking of colonization, a green fluorescent protein (GFP)-tagged P. indica strain was utilized (Hilbert et al., 2012). Roots colonized with tagged P. indica were harvested at 2, 7, and 14 dpi, and were cleaned, mounted, and observed using fluorescence microscopy (Nikon 80i).
For confocal microscopy, the roots were treated with propidium iodide and observed under a confocal microscope (Leica TCS M5) at an emission wavelength of 505-530 nm with excitation at 470 nm and digital sectioning of 4-5 µm of root thickness. The relative amount of fungal DNA was determined using real time-qPCR utilizing Arabidopsis Actin2 (At3g18780) and P. indica Tef1 . Relative changes in fungal DNA content were calculated using the C T values of PiTef1, which were normalized by the C T values of AtActin2 using the ΔΔC T equation and setting the P. indica DNA content of the control roots as 1 (Vadassery et al., 2008).

Tissue localization by GUS assays
Transgenics with a fusion of the CNGC19 promoter and β-glucuronidase (GUS) were constructed as previously described by Meena et al., 2019. Arabidopsis seedlings of ProCNGC19::GUS-expressing transgenic plants (T 3 generation) were co-cultivated with P. indica, and were carefully harvested at 2 dpi and 7 dpi. They were vacuum-infiltrated with GUS staining solution and incubated in the dark at 37 °C. Tissues were decolored by treating with 70% ethanol at 65 °C and then observed under a light microscope (Nikon 80i).

Measurements of elevation of cytoplasmic Ca 2+
WT::aeq, cngc19-2::aeq, and pepr1 pepr2::aeq seedlings (Meena et al., 2019) were grown on MS medium and roots of 14-d-old seedlings were transferred to a 96-well white plate (ThermoFisher Scientific) containing 5 μM coelenterazine (PJK, Germany) and left in the dark overnight at 21 °C. Bioluminescence counts in the roots were recorded as relative light units (RLU) per second using a microplate luminometer (Luminoscan Ascent, v. 2.6, ThermoFisher Scientific). After a 1-min background reading, PiCWE (50 µl), cellobiose (100 µM), or CT (10 µM) was added and readings were taken for 15 min. Discharge solution (1 M CaCl 2 and 10% ethanol) was used for calibrations to estimate the aequorin that remained at the end of the experiment (Vadassery et al., 2012;Meena et al., 2019). The luminescence counts obtained were calibrated using the equation presented by Rentel and Knight (2004).

Callose staining, microscopy, and quantification
Seedlings at 2, 7, and 14 dpi were treated with Aniline Blue (0.001%) according to the protocol described by Schenk and Schikora (2015). Colonized and non-colonized seedlings, and control and AtPep1-treated (100 nM) 7 d old seedlings were incubated at room temperature in acetic acid and ethanol (1:3) for decolorization, washed in 150 mM K 2 HPO 4 , and stained with Aniline Blue (0.001%) solution. Slides for observing callose deposition were prepared using 50% glycerol under a Nikon 80i microscope at 358 nm excitation and 460 nm emission. Relative callose intensities were calculated by dividing callose pixels and total pixels using the digital photograph analysis software GIMP (Scalschi et al., 2015).

Phylogenetic analysis
A total of 123 complete CNGC sequences from seven different plants were selected for construction of the phylogenetic tree, namely Arabidopsis thaliana, Glycine max, Medicago truncatula, Solanum lycopersicum, Zea mays, Oryza sativa, and Lotus japonicus. All the amino acid sequences encoding CNGCs were retrieved from previously published reports (Moeder et al., 2011;Nawaz et al., 2014;Charpentier et al., 2016;Chiasson et al., 2017;Moeder and Yoshioka, 2017) and are listed in Supplementary Table  S2. A phylogenetic tree was constructed using the MEGA 10 software (https://www.megasoftware.net/; Kumar et al., 2018), in which the sequences were aligned by MUSCLE with default parameters. These aligned sequences were used to build the phylogenetic tree using the maximum likelihood (ML) method and the evolutionary distances were computed using a Jones-Taylor-Thornton matrix-based method with 1000 bootstrap replications.

Statistical analysis
Statistical differences between treatments were analysed using twotailed Student's t-tests or one-way ANOVA followed by Tukey's test in SigmaPlot 13.0. Figures were generated using Origin 6.0 (www. originlab.com).

CNGC19 expression is activated by P. indica cell-wall extract
To identify the role of CNGCs in the perception of P. indica by Arabidopsis, we applied either crude cell-wall extract (PiCWE) or its identified active elicitor cellotriose (CT) to Arabidopsis seedlings for 30 min (Fig. 1A). Upon treatment with PiCWE, five CNGCs were found to be induced in Arabidopsis roots, namely CNGC19 (14.3-fold increase), CNGC3 (6.22-fold), CNGC13 (5.9-fold), CNGC10 (5-fold), and CNGC6 (3.7-fold). In response to CT, nine CNGCs were induced, with the highest expression being found for CNGC3 (17.5-fold); however, CNGC19 was not induced by CT. Since it was the highest expressed upon PiCWE treatment and the only transcript specifically induced by PiCWE but not by CT, we hypothesized that CNGC19 was induced by unidentified elicitors. We examined the patterns of CNGC19 expression in wild-type (WT) roots and found that it increased from 10-45 min in response to PiCWE but there was no response to CT (Fig. 1B). We then examined roots of plants co-cultivated with P. indica, and found that expression of CNGC19 was increased by 7-fold at 2 d post-colonization, before returning to the basal level at later time-points (Fig. 1C), thus indicating a potential role in the colonization process. To identify the tissue-specific expression pattern of CNGC19 in Arabidopsis after colonization by P. indica, we used ProCNGC19::GUS. CNGC19 promoter activity was observed in the root primordia and primary vasculature at 2 dpi and 7 dpi (Fig. 1D, E). Interestingly, we also observed a systemic expression of CNGC19 in the leaf vasculature at the same time. The results therefore indicated that CNGC19 is an early-activated gene that is expressed in the vasculature, and it is induced in the roots by unidentified components in PiCWE and systemically in leaves.

The growth of cngc19 mutants is inhibited by P. indica
To identify the functional role of CNGC19 in the promotion of growth induced by P. indica, we utilized the cngc19-2 and cngc19-1 mutant T-DNA lines (Meena et al., 2019). In contrast to the WT plants, we observed no growth promotion in the cngc19 mutants in response to P. indica for parameters such as fresh weight ( Fig. 2A, B) and root length ( Supplementary  Fig. S1) when plants were grown in culture medium. Indeed, the mutants showed growth inhibition at 7 dpi and 14 dpi compared to the WT (Fig. 2B). We found similar effects when we repeated the experiment with plants grown in soil, with the growth of the mutants being strongly inhibited at 42 dpi whereas growth promotion was observed in the WT in response to P. indica ( Supplementary Fig. S2A-C).

Colonization by P. indica is enhanced in cngc19 roots
We tested the hypothesis that the reduced growth in cngc19 mutants upon P. indica inoculation was due to enhanced colonization. The roots of the WT and cngc19-2 were co-cultivated with GFP-tagged P. indica and were used for microscopic analyses. We observed no differences between P. indica colonization in the WT and cngc19-2 at 2 dpi and 7 dpi ( Supplementary Fig. S3); however, at 14 dpi cngc19 roots had increased colonization compared to the WT and clumps of fungal mycelia and spores could be observed (Fig. 2C). This was supported by quantification of the relative fungal DNA content, which increased at 14 dpi (grown in medium) and 42 dpi (grown in soil) in cngc19-2 roots relative to the WT (Fig. 2D). Thus, the loss-of-function of CNGC19 resulted in increased colonization and loss of controlled P. indica growth in the plant-fungal interaction at the post-establishment phase.

The CNGC19 channel is involved in PiCWE-mediated elevation of cytosolic Ca 2+
CNGC19 is a plasma membrane-localized Ca 2+ -permeable channel (Meena et al., 2019) and we therefore hypothesized that it could be involved in generating the elevation in cytosolic Ca 2+ in response to PiCWE. We used the WT::aeq and cngc19::aeq lines to examine intracellular Ca 2+ cyt upon application of PiCWE and CT to the roots. The substrate of CT, cellobiose (CB), was also used as an unrelated elicitor control. Both CT (Fig. 3A) and cellobiose ( Supplementary Fig.  S4) induced elevation of Ca 2+ cyt in the WT::aeq and cngc19 roots at similar levels. When PiCWE was added, the elevation in Ca 2+ cyt was reduced in the cngc19::aeq line relative to the WT::aeq (Fig. 3B), both in the initial peak and for several minutes thereafter. These results suggest that CNGC19 is a crucial channel that is involved in sensing as yet unidentified elicitors in PiCWE and in activating the elevation of Ca 2+ cyt .
Callose deposition in response to P. indica colonization is delayed in cngc19 Increased callose deposition has been reported in Arabidopsis roots colonized by P. indica and indicates the activation of MTI (Jacobs et al., 2011). In the WT plants, callose deposition was increased at 2 dpi and 7 dpi and then remained unchanged at 14 dpi during the established colonization phase (Fig. 4A, C). In contrast, in the cngc19-2 mutant the callose deposition was unaltered compared to the control at both 2 dpi and 7 dpi, indicating a reduced defense in these plants at these early stages. Callose deposition was not induced in cngc19-2 until 14 dpi. Thus, plant defense was lowered in the cngc19 mutant during the initial stages of the plant-fungal interaction, leading to increased colonization.

CNGC19-mediated basal immunity upon perception of P. indica is dependent on AtPep-PEPR
Plants roots encounter damage-associated molecular patterns (DAMPs) upon microbial invasion (Boller and Felix, 2009; Albert,   2013). In Arabidopsis, a family of endogenous elicitor peptides referred to as AtPeps acts as DAMPs (Huffaker and Ryan, 2007;Bartels et al., 2013) and the plasma membrane Pep-receptors PEPR1 and PEPR2 perceive them (Yamaguchi et al., 2006(Yamaguchi et al., , 2010Krol et al., 2010). We tested the possibility that P. indica may suppress AtPep-induced defense, and the role of CNGC19 in such a process. We found that callose deposition induced by application of AtPep1 was suppressed in WT seedlings inoculated with P. indica at 2 dpi (Fig. 4B, D). For the cngc19 mutant, the AtPep1induced deposition of callose was constitutively lower than that of the WT for all the treatments, and it was not suppressed by colonization by P. indica (Fig. 4D). Thus, the results indicated that CNGC19 was also crucial for P. indica-mediated suppression of DAMP-triggered immunity. The Pep-receptors PEPR1 and PEPR2 have a putative guanyl cyclase domain that generates cyclic nucleotides and they are upstream of CNGC2 (Ma et al., 2012). In order to determine the role of PEPR1 and PEPR2 in the Arabidopsis-P. indica association, we examined their expression. In WT seedlings, expression of PEPR1 and PEPR1 was found to be induced by PiCWE treatment (up to 3.5-fold; Fig. 5A) and by P. indica colonization (up to 6-fold; Supplementary Fig.  S5A). To examine the dependency of CNGC19 activation on PEPR, we measured the expression of CNGC19 in the pepr1 pepr2 background in response to treatment with PiCWE, and Representative images of callose deposition in roots of non-inoculated controls and in roots colonized with P. indica (+Pi) at 2-14 d post-inoculation (dpi) in the wild-type (WT) and the cngc19-2 mutant. (B) Representative images of callose deposition in roots of non-inoculated controls and in roots colonized with P. indica for the WT and cngc19-2 plants with or without AtPep1 treatment (100 nM) applied for 24 h at 2 dpi. (C) Relative callolose intensity of the roots shown in (A) as determined from staining for non-inoculated controls and for roots colonized by P. indica for WT and cngc19-2 seedlings at 2-14 dpi. The intensity was expressed as the number of fluorescent callose-corresponding pixels relative to that of the total number of pixels. Data are is means (±SE) of n=20 seedlings. (D) Relative callose intensity in roots shown in (B). Data are means (±SE) of n=15 seedlings. In (C, D), different letters indicate significant differences among the different treatments, as determined by one-way ANOVA and a post hoc Tukey Test (P≤0.001).
found that it was reduced at 45 min and 60 min relative to the WT (Fig. 5B). We then examined the role of PEPRs in P.indicainduced growth promotion. Colonization by P. indica was relatively high in pepr1 pepr2, and instead of growth promotion, inhibition was observed both on plates (Fig. 5C, D) and in soil ( Supplementary Fig. S6A, B). However, elevation of Ca 2+ cyt was unaltered when compared to the WT in pepr1 pepr2::aeq upon treatment with PiCWE (Fig. 5E) and CT (Supplementary Fig.  S5B). PEPR signaling contributes to the JA signaling pathway upon herbivory (Klauser et al., 2015;Meena et al., 2019). The level of the JA marker VSP2 was found to be reduced in pepr1 pepr2 upon both treatment with PiCWE and colonization by P. indica (Fig. 5E, Supplementary S5C), indicating its function downstream of CNGC19 via jasmonate signaling. Fig. 5. The roles of PEPR1 and PEPR2 in the Arabidopsis-P. indica interaction. Expression of (A) PEPR1 and PEPR2 in wild-type (WT) and (B) expression of CNGC19 in the pepr1 pepr2 double-mutant in response to treatment with P. indica cell-wall extract (PiCWE) in 10-d-old seedlings. Transcripts levels were normalized to AtActin2 mRNA and the fold-change in expression is relative to the value at time zero, which was set as 1. Data are means (±SE) of four replicates, each of which consisted of six seedlings. Significant differences were determined using two-tailed Student's t-test (*P<0.05). (C) Effects of P. indica colonization on the fresh weight of the WT and the pepr1 pepr2 double-mutant in non-inoculated controls in response to colonization by P. indica at 14 d post-inoculation (dpi). Data are means (±SE), n=30. Different letters indicate significant differences among the different treatments, as determined using one-way ANOVA and a post hoc Tukey Test (P≤0.001). (D) Effect of P. indica colonization on fungal colonization in the WT and the pepr1 pepr2 double-mutant. Plants were co-cultivated with or without fungal discs on 1× PNM agar plates and the roots were harvested at 14 dpi. The relative fungal colonization was calculated by subtracting the C T values of P. indica Tef1 from the C T values of Arabidopsis Actin2. Data are means (±SE) of four replicates, with six seedlings per replicate. The significant difference was determined using a two-tailed Student's t-test (**P≤0.005). (E) Response of cytosolic calcium (Ca 2+ cyt ) to treatment with P. indica cell-wall extract (PiCWE, 50 µl) in roots of transformed 10-d-old seedlings of the WT and pepr1 pepr2 expressing cytosolic apoaequorin. Data are means (±SE), n=5. The experiment was repeated three times with similar results and the data from one experiment are shown. Water was used as the control and gave background readings in the WT and pepr1 pepr2. The arrow indicates the time of treatment with the elicitor. (F) Expression of the defense-related gene VSP2 in response to treatment with PiCWE in roots of transformed 10-d-old seedlings of the WT and pepr1 pepr2 double-mutant expressing cytosolic apoaequorin. Transcripts levels were normalized to AtActin2 mRNA and the fold-change in expression is relative to the value at time zero, which was set as 1. Data are means (±SE), n=4.) Significant differences were determined using two-tailed Student's t-test (*P≤0.05; **P≤0.005).

Loss-of-function of CNGC19 down-regulates jasmonate biosynthesis upon P. indica colonization
Piriformospora indica activates the JA signaling pathway in Arabidopsis, and this is crucial for colonization and for balancing the beneficial interaction between the two organisms (Stein et al., 2008;Vahabi et al., 2015). To identify the role of various phytohormones in P. indica colonization in the cngc19 mutant, we measured their levels at 2 dpi and 14 dpi. The levels of JA and jasmonoyl isoleucine (JA-Ile) were increased in WT plants in response to P. indica at 2 dpi and 14 dpi, and cis-(+)-12-oxo-phytodienoic acid (cis-OPDA) was also strongly increased at 2 dpi ( Fig. 6A-C). In contrast, no significant effects of colonization were observed for cngc19. The lack of an effect on the levels of cis-OPDA, JA, and JA-Ile at 2 dpi may have contributed to uncontrolled colonization in cngc19 roots. JA-Ile-OH also showed similar trends ( Supplementary Fig.  S7A), but no changes in ABA and salicylic acid (SA) were observed in response to P. indica colonization in either genotype at either time-point ( Supplementary Fig. S7B, C).

Phytohormone-and defense-related genes are downregulated in cngc19 during colonization
Since colonization by P. indica was enhanced in the cngc19 lines, we examined the expression of marker genes of different defense pathways such as those of JA, ROS, and phytoalexin. The JA markers VSP2, PDF1.2, and LOX1 were found to be induced by P. indica in WT plants at both 2 dpi and 14 dpi compared to non-inoculated controls, whereas these genes were found to be down-regulated in cngc19 except LOX1 at 2 dpi (Fig. 7A). The ROS markers RBOHD, RRTF1, and OXI1 were selected based on their known functional roles in the Arabidopsis-P. indica interaction (Camehl et al., 2011;Matsuo et al., 2015;Johnson et al., 2018). RRTF1 was found to be up-regulated in the WT but not in cngc19 (Fig. 7B), RBOHD was up-regulated in both the WT and cngc19 and did not differ between the two, and OXI1 was strongly up-regulated at 2 dpi in the WT but showed no change in cngc19 compared to non-inoculated controls. In contrast, at 14 dpi OXI1 was observed to be strongly up-regulated in cngc19, which may have been due to increased colonization. In addition, the ROSrelated genes SOD1, GSTF8, and APX1 were found to be up-regulated in the WT and down-regulated in cngc19 at 2 dpi ( Supplementary Fig. S8), and at 14 dpi GR1 and CAT2 were found to be up-regulated in cngc19 but unaltered in the WT. WRKY33 and PAD3, which are related to phytoalexin biosynthesis, were also found to be up-regulated in the WT but not in cngc19 at 2 dpi (Fig. 7C). At 14 dpi, WRKY33 was up-regulated in both the WT and cngc19 and did not differ between the two, whilst PAD3 was also up-regulated in both but had a greater increase in the WT. Thus, the loss-of-function of CNGC19 affected JA-responsive genes, and genes involved in ROS signaling and defense, and was associated with overcolonization by P. indica in cngc19 plants.

CNGC19-mediated defense signaling in roots acts via indole glucosinolates
CNGC19 loss-of-function results in constitutively reduced levels of aliphatic GSs in Arabidopsis rosettes, which are crucial for defense against herbivory (Meena et al., 2019). Upon plant-microbe interactions, accumulation of antimicrobial indole glucosinolates (iGSs) and camalexin trigger immunity (Bednarek et al., 2009;Clay et al., 2009;Böhm et al., 2014). The iGS pathway is critical for mutualistic P. indica colonization and for uncompromised plant immunity (Nongbri et al., 2012;Lahrmann et al., 2015). We therefore decided to test the effects of mutation in CNGC19 on iGS levels. We found that the constitutive levels of both iGSs and aliphatic GSs in non-inoculated (control) seedlings were the same in the WT and cngc19 at 14 dpi (Fig. 8A, B). At the rosette stage in noninoculated plants grown in soil (42 dpi), the level of iGSs did not differ between the WT and cngc19 whilst aliphatic GSs were significantly lower in cngc19, as also reported by Meena et al., 2019. We then looked at the effects of P. indica colonization and found that iGSs were increased in the WT at 14 dpi but were not affected in cngc19 (Fig. 8A). The levels of aliphatic GSs also showed a similar trend (Fig. 8B). After a prolonged period of colonization by P. indica (42 dpi, grown in soil), iGS levels were reduced significantly in WT plants but were unaltered in cngc19 (Fig. 8A). Aliphatic GS levels did not change at 42 dpi in either of the genotypes (Fig. 8B). We examined the relative expression of key genes related to the iGS biosynthesis pathway and found that they were generally up-regulated in the WT but not in cngc19 at 2 dpi and 14 dpi (Fig. 8C). The results therefore indicate that CNGC19 plays a crucial role in modulating the content of indole GSs during colonization by P. indica.

Phylogenetic analysis of CNGC19 indicates it has a distinct role in microbial interactions
Unlike rhizobial nodulation and mycorrhizal symbiosis, P. indica has a broad host range and is a primitive symbiont . We constructed a phylogenetic tree in order to understand the relationship between AtCNGC19 and its orthologs in other host plants, many of which form symbiotic interactions. Genome-wide analyses in seven different species have identified distinct CNGCs (Mäser et al., 2001;Nawaz et al., 2014;Saand et al., 2015;Charpentier et al., 2016). In our unrooted phylogenetic Fig. 7. Effects of P. indica colonization on expression of genes related to phytohormones, reactive oxygen species (ROS), and defense in seedlings of the Arabidopsis wild-type (WT) and the cngc19 mutant. Expression of marker genes for (A) jasmonates, (B) ROS, and (C) defense early 2 d post-inoculation (dpi) and 14 dpi. Transcripts levels were normalized to AtActin2 mRNA and the fold-change in expression is relative to that of the corresponding noninoculated control, which was set as 1. Data are means (±SE) of four replicates, each of which consisted of six seedlings. Significant differences were determined using two-tailed Student's t-test: *P≤0.05, ***P<0.0001. tree, 123 CNGCs clustered into four different groups (Groups I-IV; Fig. 9). Among these, Group IV was further subdivided into IVA and IVB. AtCNGC19 and AtCNGC20 were clustered into Group IVA with orthologs from different legume plants that form symbiotic interactions, such as G. max, M. truncatula, and L. japonicas. AtCNGC19 was clustered with BRUSH from L. japonicas, which plays a crucial role in rhizobial symbiosis by regulating Ca 2+ fluxes (Chiasson et al., 2017). AtCNGC19 also grouped with OsCNGC13 (its ortholog in O. sativa; Moeder and Yoshioka, 2017), which has been shown to be up-regulated by a bacterial pathogen (Nawaz et al., 2014). Nuclear-localized MtCNGC15 has been shown to be critical in generating oscillations in nuclear Ca 2+ during mycorrhizal symbiosis (Charpentier et al., 2016); however, it was found to be clustered into Group III, which suggests it is evolutionarily divergent from AtCNGC19.

Discussion
Plant roots interact with both pathogenic and symbiotic microbes and recognize them as a potential threat by activation of basal MAMP-triggered immunity (MTI). Activation of MTI prevents the establishment of pathogens and regulates colonization by symbionts, and thus the process acts as a gatekeeper (Yu et al., 2019a). Root MTI efficiently restricts penetration and colonization of the mutualist P. indica and prevents overcolonization, with the result that a symbiotic interaction is established (Jacobs et al., 2011). Recognition of P. indica cell-wall extract (PiCWE) by Arabidopsis roots induces elevation of cytosolic Ca 2+ (Ca 2+ cyt ) and is crucial in activating the symbiotic interaction and in promoting growth Johnson et al., 2018). The identity of the ion channels responsible for the influx of Ca 2+ and the activation of signaling is currently unknown. In our present study, we identified CNGC19 as an early-activated gene (Fig. 1A) and found that it was induced by unidentified components in PiCWE and not by treatment with cellotriose (CT) (Fig. 1A, B). Cyclic nucleotide gated channels (CNGCs) are altered by treatment with PiCWE and glutamate receptor-like channels (GLRs) are altered by treatment with CT in plant roots Johnson et al., 2018). PiCWE (which contains many elicitors including CT) and CT also differ in their activation of other pathways. Importantly, CT induces a defense Fig. 8. Effects of P. indica colonization on glucosinolates in seedlings of the Arabidopsis wild-type (WT) and the cngc19 mutant. Levels of (A) indole glucosinolates (iGS) and (B) aliphatic glucosinolates in non-inoculated controls and plants colonized by P. indica grown either on plates (14 d post-inoculation, dpi) or in soil (42 dpi). Data are means (±SE) of three replicates each consisting of 40 seedlings (14 dpi) or five replicates each consisting of one seedling (42 dpi). Different letters indicate significant differences among the different treatments as determined using one-way ANOVA and a post hoc Tukey's test (P≤0.001). C. Relative expression of iGS biosynthesis pathway genes in seedlings grown on plates at 2 dpi and 14 dpi. Transcripts levels were normalized to AtActin2 mRNA and the fold-change in expression is relative to that of the corresponding non-inoculated control, which was set as 1. Data are means (±SE) of four replicates each consisting of six seedlings. Significant differences were determined using two-tailed Student's t-test (*P≤0.05, **P≤0.005).
pathway comprising ROS accumulation and the expression of its marker gene RBOHD (Johnson et al., 2018), whereas PiCWE does not activate this pathway . We found that CNGC19 was crucial for P. indica-induced promotion of growth, as its loss-of-function resulted in increased colonization and the complete loss of growth-promotion phenotype ( Fig. 2A, B). Importantly, CNGC19 was found to be critical for the generation of the PiCWE-induced elevation of Ca 2+ cyt . No other Ca 2+ channels have yet been implicated in the growth promotion and elevation of Ca 2+ cyt that is induced by P. indica. Since PiCWE-induced Ca 2+ signals were not completely abolished in the cngc19 mutant (Fig. 3B), they might be controlled by additional genetic interactions between CNGCs and other unknown channels. A leucine-rich repeat protein mutant, Piriformospora indica-insensitive12 (pii12), has previously been reported to show no promotion of growth when it is associated with P. indica (Shahollari et al., 2007). Overall, our results indicate that the PiCWE-activated CNGC19 is a critical Ca 2+ channel for growth-promotion signaling.
Expression of CNGC19 in Arabidopsis is associated with salinity stress (Kugler et al., 2009;Oh et al., 2010). We have previously reported that CNGC19 expressed in the leaf vasculature is crucial for Arabidopsis defense against herbivory by Spodoptera moths by regulating the spread of Ca 2+ signals and the levels of jasmonate and aliphatic glucosinolates (Meena et al., 2019). The AtCNGC19 homolog SlCNGC15 in tomato is induced by both salinity stress and P. indica colonization (Ghorbani et al., 2019). SlCNGC15 is also associated with disease resistance against the necrotropic fungus Sclerotinia sclerotiorum (Saand et al., 2015). cngc19 and cngc20 mutants are also more susceptible to infection by Botrytis cinerea (Moeder et al., 2011). The receptor kinase BAK1/SERK4 phosphorylates the Ca 2+ -channel complex CNGC20/CNGC19 and has crucial role in pathogen-induced cell death (Yu et al., 2019b). The roles of other CNGCs in plant defense have been demonstrated in many studies. AtCNGC2 and AtCNGC4 are known to regulate Ca 2+ -induced PAMP-triggered immunity (Chin et al., 2013;Tian et al., 2019). They act as a heterotetrameric Ca 2+ channel and are phosphorylated and activated by the kinase BIK1 of the pattern-recognition receptor complex, triggering an increase in the concentration of Ca 2+ cyt (Tian et al., 2019). CNGC2 and CNGC4, which are also known as dnd1 (defense no death1) and dnd2, respectively, are involved in the hypersensitivity response and DAMP perception during bacterial infection in Arabidopsis (Ahn, 2007). cngc11 and cngc12 are also hypersusceptible to fungal infection (Yoshioka et al., 2001). In apple, overexpression of MdCNGC1 results in increased susceptibility to fungal infection and in reduced callose deposition when plants are treated with flg22 and chitosan . Interestingly, in our study CNGC2, CNGC4, CNGC11, and CNGC12 were not significantly induced by P. indica (Fig. 1A), indicating that a different set of channels is involved in the interaction. The brush mutant has been isolated in a screen of an ethyl-methanesulfonate-mutated population Fig. 9. Phylogenetic relationships of AtCNGC19 with homologs from other species. Multiple sequence analysis was performed using the MUSCLE software. A maximum-likelihood unrooted tree was constructed using 123 complete amino acid sequences of CNGCs in the MEGA10 software. The evolutionary relationships were analysed with 1000 bootstrap replicates. Each node is labelled with the gene ID and its previously reported name. Genes examined in the current study are highlighted in blue.
for plants defective in symbiotic cell development (Maekawa-Yoshikawa et al., 2009). At 26 °C, brush roots are stunted and infection threads in root hairs do not progress into the cortex, resulting in the formation of non-infected nodules. This has been mapped as a gain-of-function CNGC.IVA mutation and it is orthologous to AtCNGC19 and AtCNGC20, resulting in a leaky tetrameric channel (Chiasson et al., 2017). Similarly, nuclear-localized CNGC15 in Medicago forms a complex with the potassium-permeable channel DMI1, and is responsible for nuclear Ca 2+ release upon mycorrhizal symbiosis (Charpentier et al., 2016). Our phylogenetic analysis also indicated that CNGCs of diverse plants that clustered in Groups IV and III are involved in both symbiotic and pathogenic interactions (Fig. 9). Thus, activation of Ca 2+ channels belonging to the CNGC family seems to be a conserved element in symbiotic interactions.
Increased P. indica colonization in the cngc19 mutants (Fig. 2) indicated that the normal functioning of CNGC19 is crucial for maintaining controlled colonization, and suggests that it has a role in MTI. Plants deposit callose (β1, 3-glucan) into cell walls upon microbial invasion as a part of MTI (Thordal-Christensen, 2003;Nürnberger and Lipka, 2005). Colonization by P. indica is known to induce callose deposition and additional exposure to the elicitor flg22 does not increase callose because of P. indicamediated suppression of late MTI (Jacobs et al., 2011). Thus, the suppression of callose deposition is required for progression of P. indica colonization. We found that callose deposition was initially unaltered in cngc19 (Fig. 4A, C), indicating that the plant defense was lowered and hence led to increased colonization and pathogen-like growth. Plants activate robust MAMP perception and subsequent MTI by the action of phytohormones such as salicylic acid, ethylene, and jasmonate to regulate P. indica colonization. It is known that colonization increases JA/JA-Ile levels in co-cultivated plants with the result that plant defense responses are altered and tolerances to pathogenic microbes and root herbivory are improved (Vahabi et al., 2013(Vahabi et al., , 2015Lahrmann et al., 2015;Cosme et al., 2016). It has also been reported that JA signaling is required to suppress late MTI in order to facilitate the progression of P. indica colonization at the late biotrophic stage (Jacobs et al., 2011). It is known that the jasmonate insensi-tive1-1 (jin1-1) and jasmonate resistant1-1 (jar1-1) mutants show no promotion of growth upon P. indica colonization (Jacobs et al., 2011). Loss-of-function of CNGC19 down-regulated jasmonate biosynthesis and JA-responsive genes upon P. indica colonization ( Fig. 6 and 7A). This suggests that CNGC19mediated signaling leads to the activation of JA/JA-Ile signaling (Fig. 10), as has also been observed for defense against Spodoptera herbivory in Arabidopsis leaves (Meena et al., 2019). In our study, the cngc19 mutants also displayed reduced expression of early and late MTI-related genes involved in phytohormone, ROS, and secondary metabolite pathways (Fig. 7A-C), which contributed to unbalancing the mutualistic relationship between the fungus and host plant (Fig. 10). We also found that H 2 O 2induced Oxidative Signal Inducible1 (OXI1) kinase was reduced in cngc19 plants at an early stage of colonization (Fig. 7B). The oxi1 mutant together with the agc2-2 (OXI1 kinase homolog) and pdk1.1 pdk1.2 (3-PHOSPHOINOSITIDE-DEPENDENT PROTEIN KINASE1) mutants also have reduced growth upon P. indica colonization (Camehl et al., 2011). In addition, we also found that genes related to the antioxidant system were altered at early and late stages of P. indica colonization, namely SOD1, GSTF8, GR1, CAT2, and APX1 ( Supplementary Fig. S8). Thus, CNGC19 is critical for MTI responses in Arabidopsis roots (Fig. 10).
Damage-associated molecular patterns (DAMPs) are recognized by leucine-rich repeat (LRR)-like receptors, which activate downstream signaling. AtPep1 is perceived as a stronger danger signal by Arabidopsis roots than the MAMP-like bacterial flg22 or chitin, and hence AtPep-PEPR signaling is a major component of surveillance in the roots (Poncini et al., 2017). We have previously identified that CNGC19 is involved in AtPep1-induced elevation of Ca 2+ cyt (Meena et al., 2019). Pep-induced PEPR signaling further intensifies the plant defense response together with MTI (Ross et al., 2014). Upon addition of AtPep1, P. indica-mediated suppression of callose occurred in the WT but not in the cngc19 mutant (Fig. 4). pepr1 pepr2 double-mutants also showed a growth inhibition phenotype (Fig. 5C), but these genes are not involved in the PiCWE-induced elevation of Ca 2+ cyt and instead might be acting via the JA pathway (Fig. 5E, F). PEPR signaling maintains basal immunity by regulating JA and SA signaling, locally and systemically (Ross et al., 2014;Yamada et al., 2016), which agrees with the down-regulation of the JA-responsive gene VSP2 that we observed upon P. indica colonization in the pepr1pepr2 mutant (Fig. 5). Thus, PEPR signaling works downstream of CNGC19, contributes to the JA pathway, and may interact with unknown receptors and kinases for modulating downstream targets.
It is known that cngc19 mutants are constitutively deficient in aliphatic glucosinolate accumulation and that they hyperaccumulate its precursor, methionine (Meena et al., 2019). CNGC19 modulates aliphatic glucosinolate biosynthesis in tandem with BRANCHED-CHAIN AMINO ACID TRANSAMINASE4 (BCAT4), which is involved in the chain elongation pathway of metionine-derived glucosinolates (Meena et al., 2019). However, this phenotype appeared to be age-dependent as it was absent in the seedlings that we studied, and upon P. indica colonization we found that it was the activation of iGS that was important (Fig. 8A). Thus, regulation of glucosinolates by CNGC19 is age-and stimuli-dependent. The indolic glucosinolate pathway plays a major role in the growth restriction of P. indica (Lahrmann et al., 2015). The lossof-function mutants cyp79b2/3 and cyp81f2, which are genes involved in iGS biosynthesis, exhibit growth inhibition upon P. indica colonization similar to what we observed in cngc19 (Nongbri et al., 2012;Lahrmann et al., 2015). Cytochrome P450 enzymes (CYP79B2, CYP79B3, CYP81F2) and transcription regulators of iGS (MYB51, MYB122, WRKY33) and other iGS-related genes (IGMT1, IGMT2, PAD3, PEN2) have previously been observed to be stimulated during the interaction with P. indica (Jacobs et al., 2011;Nongbri et al., 2012;Lahrmann et al., 2015;Peskan-Berghöfer et al., 2015). It has also been shown that genes related to the iGS biosynthesis pathway are essential for callose deposition (Clay et al., 2009). In our study, such genes were found to be down-regulated upon P. indica colonization in cngc19 (Fig. 8), and the levels of iGS were increased at 14 dpi and reduced at 42 dpi in colonized WT plants, but were unaltered in cngc19. The activation of iGS biosynthesis might occur via activation of CNGC19 in the colonized plants. Some other iGS-related mutants (myb34/51/122, pen2-1) have also been shown to have higher levels of P. indica colonization (Jacobs et al., 2011;Nongbri et al., 2012;Lahrmann et al., 2015). In addition, the mutant of a β-glucosidase (ΔPYK10), which is involved in hydrolysing iGS, has also been observed to have no growth promotion upon P. indica colonization (Sherameti et al., 2008;Nakano et al., 2017). All these findings place CNGC19 as an upstream element in iGS activation upon P.indica colonization.
In conclusion, our results indicate that CNGC19 is activated by as yet unidentified elicitors in the cell-wall extract of P.indica. The CNGC19-mediated pathway affects the basal immunity and the levels of phytohormones and glucosinolates in Arabidopsis upon colonization by P.indica, and subsequently affects the growth of the plants. These events do not occur in the cngc19 mutants, and this leads to over-colonization and detrimental effects on plant health and growth (Fig. 10). Thus, CNGC19 plays a central role as a gatekeeper during colonization by P. indica, maintaining a robust innate immunity that ensures that the interaction is mutualistic.

Supplementary data
Supplementary data are available at JXB online. Table S1. Primers used in this study. Table S2. Amino acid sequences of CNGCs used for constructing the phylogenetic tree. Fig. S1. Root lengths of WT and cngc19-2 seedlings co-cultivated with P. indica. Fig. S2. Growth of WT and cngc19 seedlings co-cultivated with P. indica in soil. Fig. S3. Colonization patterns of P. indica in WT and cngc19 roots. Fig. S4. Cytosolic Ca 2+ levels in response to treatment with P. indica elicitors. Fig. S5. Role of PEPR signaling in the Arabidopsis-P. indica interaction. Fig. S6. Growth effects of P. indica colonization in the pepr1 pepr2 double-mutant. Fig. S7. Phytohormone levels in WT and cngc19-2 seedlings during P. indica colonization. Fig. S8. Relative gene expression of several ROS-related genes in WT and cngc19 seedlings after P. indica inoculation. Fig. 10. Schematic model of the role of CNGC19 in the interaction between Arabidopsis and P. indica. Upon perception of P. indica cell wall-associated elicitor(s), as yet unknown receptors are activated in the root and lead to activation of CNGC19, which elevates the level of cytosolic calcium. This in turn modulates downstream systemic and defense-related pathways of MAMP-triggered immunity (MTI), such as callose deposition, glucosinolate biosynthesis, and jasmonic acid signaling. AtPep-PEPR signaling functions downstream of CNGC19 via jasmonate signaling. The activation of these pathways synergistically leads to maintenance of the P. indica colonization at a level that provides a mutualistic association between the plant and the fungus. central instrumentation, phytotron facility and DBT-eLibrary Consortium (DeLCON) for providing access to e-resources. This work was funded by the Department of Biotechnology (DBT), India through a NIPGR core grant and the MPG-India partner group program of the Max Planck Society (Germany) and Department of Science and Technology (DST), India.

Authors contributions
JV designed the experiments; AJ performed the experiments; MKM performed the luminometer experiments and generated the transgenic lines; AK performed the HPLC and LC-MS experiments; MV performed the phylogenetic tree analysis; AJ and JV analysed the results and wrote the manuscript.