Three Pectin Methylesterase Inhibitors Protect Cell Wall Integrity for Arabidopsis Immunity to Botrytis

Pectin methylesterase inhibitors control PME activity to hinder pectin degradation as part of the plant immune response.

Abstract

Infection by necrotrophs is a complex process that starts with the breakdown of the cell wall (CW) matrix initiated by CW-degrading enzymes and results in an extensive tissue maceration. Plants exploit induced defense mechanisms based on biochemical modification of the CW components to protect themselves from enzymatic degradation. The pectin matrix is the main CW target of Botrytis cinerea, and pectin methylesterification status is strongly altered in response to infection. The methylesterification of pectin is controlled mainly by pectin methylesterases (PMEs), whose activity is posttranscriptionally regulated by endogenous protein inhibitors (PMEIs). Here, AtPMEI10, AtPMEI11, and AtPMEI12 are identified as functional PMEIs induced in Arabidopsis (Arabidopsis thaliana) during B. cinerea infection. AtPMEI expression is strictly regulated by jasmonic acid and ethylene signaling, while only AtPMEI11 expression is controlled by PME-related damage-associated molecular patterns, such as oligogalacturonides and methanol. The decrease of pectin methylesterification during infection is higher and the immunity to B. cinerea is compromised in pmei10, pmei11, and pmei12 mutants with respect to the control plants. A higher stimulation of the fungal oxalic acid biosynthetic pathway also can contribute to the higher susceptibility of pmei mutants. The lack of PMEI expression does not affect hemicellulose strengthening, callose deposition, and the synthesis of structural defense proteins, proposed as CW-remodeling mechanisms exploited by Arabidopsis to resist CW degradation upon B. cinerea infection. We show that PME activity and pectin methylesterification are dynamically modulated by PMEIs during B. cinerea infection. Our findings point to AtPMEI10, AtPMEI11, and AtPMEI12 as mediators of CW integrity maintenance in plant immunity.

The cell wall (CW) is the primary interface where most plant-microbe interactions occur and the main physical and molecular line of defense evolved by plants to restrict pathogen penetration and infection spreading (Malinovsky et al., 2014). Some of the most devastating plant diseases are caused by necrotrophic fungi. Necrotrophic infection is a complex process mediated by numerous extracellular enzymes, proteins, and metabolites. CW-degrading enzymes perturb CW integrity, facilitating penetration into the host surface, while toxins, oxalic acid (OA), and reactive oxygen species may contribute to killing of the host cells (Laluk and Mengiste, 2010; King et al., 2011; Nakajima and Akutsu, 2014). The plant susceptibility to necrothrophs and the efficiency of CW degradation are largely affected by CW composition and structure (Lionetti et al., 2010; Francocci et al., 2013; Bellincampi et al., 2014). CWs of Arabidopsis (Arabidopsis thaliana) leaves consist mainly of cellulose fibers encased in a network of hemicellulose and embedded in a pectin matrix (Zablackis et al., 1995). Hemicelluloses include xyloglucan containing a (1,4)-β-linked glucan backbone substituted with (1,6)-α-linked xylosyl residues or side chains of xylosyl, galactosyl, and fucosyl residues. Low percentages of glucuronoarabinoxylans, mannans, and glucomannans also are present in CW of Arabidopsis leaves. Pectins are a complex group of polysaccharides composed of homogalacturonan (HG), rhamnogalacturonan I (RGI), rhamnogalacturonan II (RGII), and xylogalacturonan. HG, a linear polymer of (1,4)-α-linked GalA residues, is the prevalent component of leaf CW pectins (Zablackis et al., 1995) and is critical for tissue integrity, wall plasticity, and cell adhesion (Lionetti et al., 2014a). RGI consists of a backbone of alternating disaccharides, (1,4)-α-d-GalA-(1,2)-α-l-Rha with (1,4)-β-galactan, branched arabinan, or arabinogalactan side chains and constitute 20% to 25% of pectin of primary CW of dicots (Mohnen, 2008). RGII is a highly substituted galacturonan with side chains containing Ara, Rha, Gal, Xyl, or Fuc residues (Mohnen, 2008; Harholt et al., 2010). Xylogalacturonan is formed by a linear (1,4)-α-GalA backbone substituted with d-Xyl residues. RGII and xylogalacturonan represent a minor component and account of about 10% of leaf CW pectin (Zandleven et al., 2007; Mohnen, 2008).

During pathogen attack, plants sense pathogens and exploit constitutive and/or induced defense mechanisms based on biochemical modification of the CW components (Williamson et al., 2007; Raiola et al., 2011; Bellincampi et al., 2014; Blümke et al., 2015). Upon cuticle breaking, the adhesive pectin matrix is an early target for fungal necrotrophs (van Kan, 2006). Botrytis cinerea is considered one of the most important necrotrophic pathogens, mainly due to its large host range and its ability to produce severe damage, both preharvest and postharvest (Dean et al., 2012). Analysis of the B. cinerea genome indicates the presence of 118 genes associated with plant CW degradation (Amselem et al., 2011), including a large array of pectinases such as polygalacturonases and pectate lyases (Blanco-Ulate et al., 2014). Pectins are synthesized in the Golgi and secreted into the CW in a highly methylesterified form (Harholt et al., 2010; Kim et al., 2015). Methylesterified HG is demethylesterified after biosynthesis by plant pectin methylesterases (PMEs; EC 3.1.1.11; Pfam 01095; CE8; www.cazy.org), which release protons and methanol (MeOH) in the apoplast. PMEs belong to a large multigene family (67 putative isoforms in Arabidopsis) whose members display different modes of action and produce HG with different distribution and degree of methylesters in a locally regulated manner, not yet completely understood (Wang et al., 2013). PME activity can be tightly regulated by PME inhibitors (PMEIs) identified in numerous plant species (Sénéchal et al., 2014; Lionetti et al., 2015c). PMEIs belong to the large multigene family PF04043 (http://pfam.xfam.org/family/PF04043; 69 genes in Arabidopsis), which also includes the structurally related invertase inhibitors (INHs).

The methylesterification status of pectin affects plant resistance to diseases (Lionetti et al., 2012; Bellincampi et al., 2014). In several plant-microbe interactions, a high level of pectin methylesterification correlated with an increased resistance to pathogens. This feature was associated with the low susceptibility of high methylesterified pectin to pectic enzymes, virulence factors with active roles in pathogenesis (Herron et al., 2000; Wydra and Berl, 2006; Lionetti et al., 2012). Biotechnological approaches were used to increase the basal level of pectin methylesterification, aiming at engineering a pectin substrate less prone to degradation by fungal pectinases. These strategies were based mainly on reducing PME activity through the constitutive expression of development-related PMEIs. PMEI overexpression in different plant species reduced their susceptibility to fungal, bacterial, and viral pathogens (Lionetti et al., 2007, 2014b, 2015b; An et al., 2008; Raiola et al., 2011). In particular, plants overexpressing AtPMEI1 or AtPMEI2 (Hothorn et al., 2004; Raiola et al., 2004; Wolf et al., 2009; De Caroli et al., 2011) showed a lower level of PME activity, a higher degree of methylesterification (DME) of pectin, and a reduced susceptibility to B. cinerea and Pectobacterium carotovorum (Lionetti et al., 2007; Raiola et al., 2011). The reduced susceptibility to fungal diseases also was observed in wheat (Triticum aestivum) plants overexpressing AdPMEI from kiwi (Actinidia deliciosa; Camardella et al., 2000; Giovane et al., 2004; Di Matteo et al., 2005; Lionetti et al., 2014b), indicating that the inhibition of PME activity also improves plant resistance in low-pectin-containing monocot species (Vogel, 2008; Volpi et al., 2011).

Despite this evidence, the current knowledge of a possible regulation of pectin methylesterification during disease as well as of the molecular mechanisms involved in this regulation remains scarce. Emerging clues indicate that PME activity and the level of pectin methylesterification are precisely and temporally regulated during the course of a plant disease. A steady increase in plant PME activity resulted in a reduced level of pectin methylesterification, which was observed in Arabidopsis during infection by different pathogens (Bethke et al., 2014; Lionetti, 2015). Genome-wide transcriptional profiling suggests that specific Arabidopsis PMEs and PMEIs are expressed in response to pathogens and strictly dependent on the pathogen lifestyle (Lionetti et al., 2012; Windram et al., 2012). However, very few PME and PMEI isoforms have been identified to date that are involved in disease resistance, and their exact role during disease remains undetermined. AtPME3 is induced in Arabidopsis when challenged with either B. cinerea or P. carotovorum, and the pme3 mutant is more resistant to these necrotrophic pathogens (Raiola et al., 2011). TaPME1 has been considered as a candidate gene involved in wheat susceptibility against the necrotroph Fusarium graminearum (Lionetti et al., 2015a). A recent comprehensive analysis of the PME gene family in wheat indicated the possible involvement of specific TaPMEs in wheat defense against F. graminearum (Zega and D’Ovidio, 2016). Specific Arabidopsis PMEs contribute to immunity against the hemibiotrophic bacterial pathogen Pseudomonas syringae (Bethke et al., 2014). However, there is no evidence that the same genes are important against the necrotrophic fungus Alternaria brassicicola.

Accumulating evidence suggests that immune hormones regulate pectin methylesterification during plant-pathogen interactions (Nafisi et al., 2015). Jasmonic acid (JA) modulates the pectin DME in potato (Solanum tuberosum) to protect pectin degradation by pectate lyase produced by Dickeya dadantii (Taurino et al., 2014). PME activity is triggered through a JA-dependent pathway when Arabidopsis is challenged with P. syringae or A. brassicicola (Bethke et al., 2014). PME activity was proposed to be involved in the release and perception of defense signals during infection. PME activity is required for the production of demethylesterified oligogalacturonides, damage-associated molecular patterns (DAMPs), released upon partial degradation of HG by fungal pectinases (Ferrari et al., 2013). The ectopic expression of the strawberry (Fragaria × ananassa) gene FaPE1 in Fragaria vesca resulted in the generation of bioactive oligogalacturonides (OGs) responsible for the constitutive activation of defense responses and plant resistance to B. cinerea (Osorio et al., 2008, 2011). OGs are perceived in Arabidopsis by the receptor WALL-ASSOCIATED KINASE1 (WAK1) to activate plant immune responses (Brutus et al., 2010). WAK2 requires AtPME3-induced pectin demethylesterification to activate OG-dependent stress responses in Arabidopsis (Kohorn et al., 2014). Furthermore, PMEs release MeOH, a DAMP-like alarm signal (Hann et al., 2014), which alerts adjacent noninfected tissues or neighboring plants (Dorokhov et al., 2012; Komarova et al., 2014). The molecular mechanisms behind OG/MeOH release and signaling during pathogen infection are still largely unknown.

In this work, we advance the understanding about CW remodeling during necrotrophy. We sought to unravel the mechanisms regulating pectin methylesterification during pathogen infection. We identified AtPMEI10, AtPMEI11, and AtPMEI12 as functional PME inhibitors induced upon B. cinerea attack. AtPMEI10, AtPMEI11, and AtPMEI12 are regulated by immune hormonal pathways and by DAMP_S such as OGs and MeOH. We isolated pmei mutants, which showed an impaired PMEI expression and a compromised fungal resistance. Our findings point to AtPMEI10, AtPMEI11, and AtPMEI12 as mediators of CW integrity maintenance in plant immunity.

RESULTS

The Expression of Specific PMEI Isoforms Is Altered during B. cinerea Infection

We attempted to understand if PMEIs are exploited by plants to regulate pectin methylesterification during disease. Analyzing publicly available microarray data (AbuQamar et al., 2006; Lionetti et al., 2012; Windram et al., 2012), we selected eight Arabidopsis genes, predicted to belong to the plant INH/PMEI family, which showed altered expression during B. cinerea infection. A homology tree, generated with the selected PMEIs and the INH/PMEIs characterized so far in dicotyledonous plants, identified three independent groups (Fig. 1). Group 1 included the functional AdPMEI expressed in kiwi fruits (Balestrieri et al., 1990; Giovane et al., 1995), the pollen-specific AtPMEI1 and AtPMEI2 (Raiola et al., 2004), and At5g46960 (henceforth AtPMEI12). The same group also contained SolyPMEI (Reca et al., 2012) and VvPMEI1 (Lionetti et al., 2015c), two PMEIs from tomato (Solanum lycopersicum) and grapevine (Vitis vinifera), respectively, both involved in fruit development. Group 2 comprised At1g62760 (henceforth AtPMEI10) and At3g47380 (henceforth AtPMEI11), both clustering with CaPMEI1, a pepper (Capsicum annuum) inhibitor involved in plant resistance to biotic and abiotic stresses (An et al., 2008). At5g51520 also was present in the same group with a higher similarity to AtPMEI4, an inhibitor expressed in Arabidopsis roots (Sénéchal et al., 2015). Group 2 also included AtPMEI3, expressed in apical meristems and affecting primordia formation (Peaucelle et al., 2008), and AtPMEI6, involved in seed maturation and germination (Saez-Aguayo et al., 2013). In group 3, At3g17140 clustered with the biochemically characterized INHs, suggesting a possible INH nature for this protein. Also in the same group, although with a low amino acid identity, is AtPMEI5, previously involved in seed germination (Müller et al., 2013). BoPMEI, involved in pollen tube growth of Brassica oleracea (Zhang et al., 2010), At4g15750, and At5g50040 (not yet characterized) showed low levels of amino acid identity with the other proteins analyzed.

Figure 1.

Homology tree of selected PMEI isoforms. The percentage amino acid identity (numbers at branch points) was evaluated between the putative Arabidopsis INH/PMEI isoforms with an altered expression during B. cinerea infection and PMEIs and INHs characterized so far. Multiple sequence alignment was performed using the DNAman software package (Lynnon Biosoft). The defense-related AtPMEI10, AtPMEI11, and AtPMEI12 selected in this study for further characterization are highlighted in boldface.

PMEIs and INHs can be distinguished on the basis of specific features in their amino acid sequences (Di Matteo et al., 2005). With the aim of identifying the nature of the eight selected genes as PMEIs or INHs, we searched for the conserved amino acid domains characterizing the two types of inhibitors (Supplemental Fig. S1). All selected PMEIs presented the four Cys residues typically engaged in the formation of two disulfide bridges and required to stabilize their four-helical-bundle structure. Excluding At3g17140, At5g46960, and At5g50040, the putative PMEIs also showed a conserved Thr residue required to strengthen the PMEI-PME interactions at apoplastic pH. A typical SAA motif, contributing to the PMEI-PME complex formation (Di Matteo et al., 2005), also was shared by putative PMEIs, except for At3g17140 and At4g15750. In At3g17140, the presence of a typical PKFAE motif, critical for invertase-INH interaction (Hothorn et al., 2010), suggests that the protein is an INH; therefore, it was not investigated further.

The level of expression of the selected genes was analyzed by reverse transcription (RT)-PCR at 0, 24, 48, and 72 h post B. cinerea infection. Different from what was reported in the microarray data, our analysis did not reveal an altered expression for At5g51520, At5g50040, and At4g15750, which were not further investigated. An altered pattern of expression of AtPMEI3 (Peaucelle et al., 2008), AtPMEI10, AtPMEI11, and AtPMEI12 was detected (Fig. 2). In particular, AtPMEI3 showed an early down-regulation at 24 h post inoculation (hpi), which proceeded during the course of infection. AtPMEI11 was rapidly and significantly repressed at 24 hpi, subsequently induced at 48 hpi, and then declined at 72 hpi. AtPMEI10 and AtPMEI12 were significantly induced at 48 hpi and further increased at 72 hpi. These results indicate that PMEI members have different levels and timings of expression in response to B. cinerea. Therefore, we decided to further study only AtPMEI10, AtPMEI11, and AtPMEI12, which were induced during B. cinerea infection.

Figure 2.

Levels and kinetics of AtPMEI3, AtPMEI10, AtPMEI11, and AtPMEI12 expression in Arabidopsis leaves during B. cinerea infection. The expression of different PMEI genes in infected leaves of 6-week-old Arabidopsis wild-type plants was analyzed by quantitative PCR (qPCR) at 24, 48, and 72 hpi. The expression levels were normalized to UBIQUITIN5 (UBQ5) expression. Values are means ± sd (n = 3). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.05). The experiments were repeated three times with similar results. B, B. cinerea-inoculated leaves; M, mock-inoculated leaves.

AtPMEI10 and AtPMEI11 share about 32% identity at the amino acid level with each other and about 13% identity with AtPMEI12 (Supplemental Fig. S2). All three PMEIs showed the presence of an N-terminal signal peptide for targeting to the secretory pathway. Interestingly, AtPMEI10 stands out for the presence, after the signal peptide and upstream of the PMEI domain, of a 115-amino acid region enriched in Ser/Pro-rich repeat (SPRR) modular domains, to our knowledge never described previously in PMEIs (Supplemental Figs. S2 and S3). The SPRR region is delimited by two SSSS motifs and is formed mainly by contiguous and repeated SL(S)SPS(L)-SP(A)P regions including two interspersed repetitions of four Pro residues. Prediction analysis indicated an extracellular localization for the three PMEIs (http://abi.inf.uni-tuebingen.de/Services/MultiLoc2). We attempted to investigate the in vivo subcellular localization of the three new inhibitors. None of the three inhibitors were predicted to show the C-terminal ω site common to glycosylphosphatidylinositol-anchored proteins (predGPI software; http://gpcr2.biocomp.unibo.it/predgpi/) and found previously for AtPMEI1 (De Caroli et al., 2011). Consequently, we decided to link GFP to the C terminus of the proteins. AtPMEI11-GFP and AtPMEI12-GFP constructs under the control of the 35S promoter were expressed transiently in Arabidopsis cotyledon epidermal cells using the FAST technique (Li et al., 2009). The Phaseolus vulgaris polygalacturonase-inhibiting protein2 fused to red fluorescent protein (PGIP2-RFP) was coexpressed as a CW marker (Fig. 3, B and F) and the plasma membrane aquaporin fused to mCherry (pm-rk) as a marker of the plasma membrane (Fig. 3, D and H; Nelson et al., 2007; De Caroli et al., 2011, 2015). At 36 h post transformation, the CW was fluorescently labeled by both AtPMEI11-GFP and AtPMEI12-GFP protein fusions (Fig. 3), as indicated by (1) the colocalization of both PMEI11-GFP (Fig. 3A) or PMEI12-GFP (Fig. 3E) with PGIP2-RFP (Fig. 3, C and G), and (2) the absence of colocalization of PMEI11-GFP or PMEI12-GFP chimeras with pm-rk in plasmolyzed cells (Fig. 3, D and H). These results confirm that AtPMEI11 and AtPMEI12 are correctly secreted into the apoplast. The consistent efforts spent on the construction of the AtPMEI10-GFP fusion were unsuccessful.

Figure 3.

Subcellular localization of AtPMEI11-GFP and AtPMEI12-GFP proteins. Confocal microscopy images of cotyledon epidermal cells of Arabidopsis 7-d-old-seedlings show the coexpression of AtPMEI11-GFP (A) or AtPMEI12-GFP (E) with PGIP2-RFP (B and F) and their colocalization in the apoplast (C and G) at 36 h post transformation. Plasmolyzed cells (1 m NaCl) expressing AtPMEI11-GFP (D) or AtPMEI12-GFP (H) and the plasma membrane marker pm-rk show the green fluorescent CW and the retracted red fluorescent plasma membrane. Bars = 20 μm (A–C and E–G) and 5 μm (D and H).

We attempted to explore possible PMEs expressed during B. cinerea infection and possibly recognized by the selected PMEIs. Twelve Arabidopsis PMEs exhibiting an altered expression upon B. cinerea infection were selected by exploiting publicly available microarray data (AbuQamar et al., 2006; Lionetti et al., 2012; Windram et al., 2012). We challenged Arabidopsis leaves with B. cinerea and quantified the expression level of the selected PMEs at 0, 24, 48, and 72 hpi. METHYLESTERASE PCR A (PMEPCRA), PME1, PME3, PME17, PME20, PME21, PME34, and PME41 members exhibited significantly altered timing and level of expression during B. cinerea infection, while the expression of PME4, PME6, PME7, and PME8 was not altered under our experimental conditions (Fig. 4). The higher level of expression was detected for PME17 and PME20. AtPMEPCRA, PME17, PME20, and PME21 expression started at 24 hpi, showed peaks at 48 hpi, and declined at 72 hpi. A steady increase of PME3 expression was observed during infection, consistent with previous results (Raiola et al., 2011). PME41 showed a peak of expression at 24 hpi and declined at 48 hpi. In contrast, PME1 and PME34 were repressed during the course of infection.

Figure 4.

Specific Arabidopsis PMEs show altered expression during B. cinerea infection. The expression of different PME isoforms in infected leaves of 6-week-old Arabidopsis wild-type plants was analyzed by qPCR at 0, 24, 48, and 72 h post infection. The expression levels were normalized to UBQ5 expression. The values are means ± sd (n = 3). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.01). The experiments were repeated three times with similar results. B, B. cinerea-inoculated leaves; M, mock-inoculated leaves.

AtPMEI10, AtPMEI11, and AtPMEI12 Expression Is Controlled by Plant Immune Signaling

We evaluated the possibility that PMEI expression is exploited by Arabidopsis as part of the immune response against B. cinerea. The expression of AtPMEI10, AtPMEI11, and AtPMEI12 was determined in mutants defective in defense hormone signaling or biosynthetic pathways after infection with B. cinerea (Table I; Fig. 5). The induction of AtPMEI10 was reduced significantly in jar1-1 plants with respect to wild-type plants and almost completely abolished in ein2-5 plants. The same mutant also showed a lower induction of AtPMEI11 expression. Instead, AtPMEI12 induction was not affected in jar1-1 while it was significantly higher in ein2-5 mutants. These results indicate that JA and ethylene (ET) positively contribute to AtPMEI10 and AtPMEI11 induction against B. cinerea and that ET could exert a negative regulation of AtPMEI12. No differences were observed in sid2-2 and bak1-4 mutants with respect to the wild type, indicating that salicylic acid (SA) and BAK1 are not required for the induction of the inhibitors during B. cinerea attack. Interestingly, the induction of expression of all inhibitors also was higher in infected rlp44 plants with respect to the wild type, indicating that the signaling triggered by this receptor could negatively regulate PMEI expression. However, the high PMEI induction in rlp44 and the high expression of AtPMEI12 in ein2-5 also could be influenced by the high susceptibility of these mutants to the pathogen (Supplemental Fig. S4).

Table I. Arabidopsis mutants defective in immune hormone signaling used in this study.

Name	Description	Gene Function	Literature
jar1-1	Jasmonate-resistant1	Defective in JA signaling	Staswick et al. (1992)
sid2-2	SA induction-deficient mutant	Defective in SA biosynthesis	Nawrath and Métraux (1999)
bak1-4	BRI1-associated receptor kinase	Defective in brassinosteroid perception	Kemmerling et al. (2007)
rlp44	Receptor-like protein44	Defective in the activation of brassinosteroid signaling triggered by alteration of methylesterification	Wolf et al. (2014)
ein2-5	ET-insensitive	Defective in ET signaling	Alonso et al. (1999)

Figure 5.

Plant immunity hormones regulate the expression of disease-related PMEIs. AtPMEI10, AtPMEI11, and AtPMEI12 expression was analyzed by qPCR at 0 and 48 hpi in B. cinerea- or mock-inoculated leaves of 6-week-old Arabidopsis wild-type (WT) plants. The expression levels were normalized to UBQ5 expression. The results represent means ± sd (n = 3). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.01). The experiments were repeated three times with similar results. B, B. cinerea-inoculated leaves; M, mock-inoculated leaves.

The 1,000 bp upstream of the transcriptional start site of the identified PMEIs was analyzed using the PLACE database (http://www.dna.affrc.go.jp/PLACE/) for the presence of putative cis-acting DNA elements with regulatory functions in plant immunity (Supplemental Fig. S5). All PMEIs hold the AWTTCAAA motif, an ethylene-responsive element (Tapia et al., 2005), and AACGTG, a T/G box element involved in JA gene induction (Boter et al., 2004). AtPMEI10 and AtPMEI12 promoters contain CGTG(T/C), a brassinosteroid response element (He et al., 2005). In addition, the AtPMEI11 and AtPMEI12 promoters incorporate TGTCA, a binding site of different transcription factors whose expression is associated with the resistance response in rice (Oryza sativa; Boyle and Brisson, 2001; Luo et al., 2005). Highly enclosed in all PMEI promoters is the W box motif, (T)TGAC(C/T), that is also found in the NPR1 promoter of Arabidopsis and was shown previously as an elicitor-responsive element recognized by different WRKY transcription factors (Eulgem et al., 1999; Chen and Chen, 2002). The AtPMEI11 promoter contains the Myb1 element, GTTAGTT, recognized by MYB transcription factors and found in well-characterized hypersensitive response-related genes (Kranz et al., 1998; Daniel et al., 1999; Pontier et al., 2001). We next investigated whether PME-related DAMPs can impact the expression of disease-related PMEIs. Leaves from wild-type plants were infiltrated with OGs, MeOH, or water, and the expression of AtPMEI10, AtPMEI11, and AtPMEI12 was evaluated at 1.5 h post infiltration. AtPMEI11 expression was induced after treatment with OGs and repressed by MeOH (Fig. 6). The expression of AtPMEI10 and AtPMEI12 was not altered by DAMP treatment. These results indicate that a posttranscriptional regulation of PME activity by AtPMEI11 can be mediated by PME-related DAMPs. Collectively, our findings point to a regulation of PME activity by PMEIs as part of the Arabidopsis immune response to B. cinerea.

Figure 6.

DAMPs regulate AtPMEI11 expression. The expression of AtPMEI10, AtPMEI11, and AtPMEI12 was analyzed by qPCR at 1.5 h post infiltration of leaves of 6-week-old Arabidopsis wild-type plants with OGs (50 μg mL⁻¹), MeOH (0.1%, v/v), or water as a control. The expression levels were normalized to UBQ5 expression. Results represent means ± sd (n = 3). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.01). The experiments were repeated three times with similar results.

AtPMEI10, AtPMEI11, and AtPMEI12 Contribute to Arabidopsis Resistance against B. cinerea

A reverse genetic approach was used to elucidate the role of the PMEIs against B. cinerea. Arabidopsis T-DNA and transposon insertional PMEI mutants were selected (Fig. 7A). Two independent mutant lines for each AtPMEI were characterized: SALK_007859C (henceforth pmei10-1 plants) and SALK_072421 (henceforth pmei10-2 plants), both with a T-DNA insertion located in the 5′ UTR of AtPMEI10; SALK_015169c (henceforth pmei11-1 plants) and GT_5_108240 (henceforth pmei11-2 plants), with a T-DNA/transposon insertion located in the starting codon and in the exon of AtPMEI11, respectively; and SALK_108076c (henceforth pmei12-1 plants) and GT_5_108791 (henceforth pmei12-2 plants), with a T-DNA/transposon insertion located in the 5′ UTR and in the exon of AtPMEI12, respectively. We quantified the AtPMEI10, AtPMEI11, and AtPMEI12 transcript abundance by RT-PCR in wild-type and mutant plants challenged with B. cinerea. The induction of the three inhibitors was significantly reduced in pmei10, pmei11, and pmei12 mutants during disease (Fig. 7B).

Figure 7.

Arabidopsis pmei10, pmei11, and pmei12 mutants are more susceptible to B. cinerea. A, Schematic representation of PMEI gene structures. The localizations of T-DNA and transposon insertions in the genomic DNA sequences are shown (black arrows). The 5′ untranslated region (UTR) and 3′ UTR are represented in white, and exons are represented in gray. The SPRR domain is represented as a black block in PMEI10. B, The expression of PMEI genes was analyzed by qPCR using cDNA from leaves of 6-week-old wild-type (WT) and mutant plants at 48 hpi with B. cinerea. The expression levels were normalized to UBQ5 expression. The relative mRNA levels are represented as the ratio between gene expression in the mutants and that in the wild type. Results represent means ± sd of three independent experiments (n = 3). C, Quantification of lesion areas produced by the spreading of the fungus at 48 hpi. The values are means ± sd of three independent experiments (n = 30). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.05).

We next evaluated the susceptibility of pmei mutants to B. cinerea. Compared with the wild type, the local symptoms of the fungus were significantly higher in pmei10-1, pmei10-2, pmei11-1, pmei11-2, pmei12-1, and pmei12-2 plants, indicating that each inhibitor contributes to the resistance against the necrotroph (Fig. 7C). Since the two mutant lines for each gene showed a similar susceptibility phenotype against B. cinerea, we used pmei10-1, pmei11-1, and pmei12-1 mutants for further analysis. A greater development of B. cinerea mycelium around the inoculation sites was detected in leaf tissue of pmei mutants with respect to wild-type plants using Trypan Blue staining (Fig. 8A). We next tested if the increased susceptibility observed in pmei mutants could be due to a higher production of fungal toxic compounds, such as OA and botcinic acid. We analyzed the expression of BcOX1, encoding an oxaloacetate hydrolase, and of BcBOA6, encoding a polychetide synthase, key enzymes involved in the biosynthesis of OA and botcinic acid, respectively (Han et al., 2007; Dalmais et al., 2011; Schumacher et al., 2012). The induction of BcOX1 is higher, while the induction of BcBOA6 is significantly lower, in all pmei mutants with respect to the wild type (Fig. 8, B and C). These results indicate that a stimulated biosynthesis of the fungal OA can contribute to the higher susceptibility of pmei mutants.

Figure 8.

AtPMEI10, AtPMEI11, and AtPMEI12 expression alters B. cinerea colonization and mycotoxin synthesis. A, Photographs showing lesion areas produced by B. cinerea on leaves of 6-week-old Arabidopsis wild-type (WT) and pmei mutant plants (left) and microphotographs showing B. cinerea colonization revealed by Trypan Blue staining (right). B and C, The expression of BcOX1 (B) and BcBOA6 (C) genes was analyzed by qPCR at 48 hpi. Expression levels were normalized to BcTUBULIN expression. The results represent means ± sd (n = 3). Different letters indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.05). The experiments were repeated three times with similar results.

To determine if the increased susceptibility of pmei mutants could be due to an impaired ability to induce defense responses, we measured callose deposition and hydrogen peroxide (H₂O₂) accumulation after B. cinerea infection. Both defense responses were not compromised in leaves of pmei mutants with respect to the wild type (Supplemental Fig. S6). pmei mutants did not show any evident defects in rosette leaves, floral stems, flowers, seeds, and all developmental stages in which the inhibitors are significantly expressed (Supplemental Figs. S7 and S8).

AtPMEI10, AtPMEI11, and AtPMEI12 Control PME Activity and Pectin Methylesterification during B. cinerea Infection

After establishing that AtPMEI10, AtPMEI11, and AtPMEI12 play a role in immunity, we evaluated their possible contribution to the control of PME activity and pectin methylesterification during B. cinerea infection. The level of PME activity was monitored in wild-type, pmei10-1, pmei11-1, and pmei12-1 plants after B. cinerea infection using histochemical and biochemical assays (Lionetti, 2015). A higher increase of pathogen-induced PME activity was observed in mutant plants with respect to the wild type (Fig. 9A; Supplemental Figs. S9 and S10). This indicates that AtPMEI10, AtPMEI11, and AtPMEI12 are functional PMEIs and are able to control the induction of PME activity observed during disease. No significant differences were observed in PME activity of untreated wild-type, pmei10-1, and pmei12-1 plants. Instead, a higher basal level of PME activity was detected in the pmei11-1 mutant compared with the wild type (Fig. 9A). Most likely, AtPMEI11 also modulates the PME activity resident in leaf tissue. B. cinerea is known to produce PMEs during infection (Valette-Collet et al., 2003; Kars et al., 2005). We attempted to understand if the PME activity controlled by the identified PMEIs was derived by plant and/or by fungus. To this purpose, the recombinant AtPMEI1, purified to homogeneity and unable to inhibit B. cinerea PMEs (Lionetti, 2015), was added to protein extracts isolated from infected leaves of wild-type and pmei plants at 48 hpi. The strong inhibition of PME activity by AtPMEI1 in both the wild type and pmei mutants indicated that the enzymatic activity targeted by defense-related PMEIs was mainly plant derived (Supplemental Fig. S10).

Figure 9.

AtPMEI10, AtPMEI11, and AtPMEI12 regulate PME activity and the level of pectin methylesterification during disease. Leaves of 6-week-old Arabidopsis wild-type (WT) and pmei mutant plants were inoculated with B. cinerea and mock treatment, and PME activity (A) and the DME of pectin (B) were quantified at 0 and 48 hpi. The results represent means ± sd (n = 3). Different letters on the bars indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.05). The experiment was repeated three times with similar results. B, B. cinerea-inoculated leaves; M, mock-inoculated leaves.

In order to understand the role of the identified PMEIs in the modulation of pectin methylesterification after infection, we started to monitor the DME in wild-type leaves at 24, 48, and 72 h post B. cinerea infection. The DME was reduced significantly up to 48 hpi, reaching a level of 80% reduction compared with a mock control (Supplemental Fig. S11), and intriguingly, no further reduction of DME was found at a later stage of infection. The possible contribution of PMEI expression on the level of pectin methylesterification was explored 48 h after B. cinerea infection by comparing the DME in the wild type and pmei10-1, pmei11-1, and pmei12-1 mutants. A significantly higher reduction of DME was observed in pmei10-1, pmei11-1, and pmei12-1 mutants compared with the wild type (Fig. 9B). The DME of uninfected pmei11-1 was lower with respect to the other genotypes, consistent with its higher PME activity in the leaves.

AtPMEI10, AtPMEI11, and AtPMEI12 Control CW Integrity against B. cinerea

To study the role of the selected PMEIs in CW remodeling, it was useful first to clarify the dynamics of CW changes in wild-type plants at different times during B. cinerea infection (Supplemental Fig. S12). The amount of cellulose-derived Glc was reduced significantly at 24 hpi and decreased to a higher extent at 48 hpi, indicating a massive degradation of the polymer by B. cinerea cellulases. At the latter stage of infection (72 hpi), the level of Glc was not significantly different from that observed at 48 hpi (Supplemental Fig. S12A). The monosaccharide composition indicates a progressive reduction of GalA content at 24 and 48 hpi, confirming the pectinolytic nature of B. cinerea (Supplemental Fig. S12B). Interestingly, no further reduction of GalA content was detected later at 72 hpi. The lack of reduction of cellulose-derived Glc and GalA content at the latter stage of infection could be related to the lower increment of the lesion area detected between 48 and 72 hpi (0.4-fold change) with respect to that observed between 24 and 48 hpi (2.3-fold change; Supplemental Fig. S13). A significant increase in the relative amounts of all other monosaccharides was observed (Supplemental Fig. S12B), indicating a possible CW biosynthetic compensatory effect in response to cellulose and pectin degradation. No significant differences were observed in mock-inoculated leaves at the different time points (Supplemental Figs. S12A and S14). The contribution of PMEIs in the degradation and remodeling of CW polysaccharides was investigated by analyzing the monosaccharide composition of pmei10, pmei11, and pmei12 mutants in comparison with the wild type at 48 h after B. cinerea infection. A greater reduction of GalA level was detected in the CW of all pmei mutants with respect to the wild type, with a minor decrease in pmei12 plants (Fig. 10). This result indicates that a higher pectin degradation occurs if AtPMEI10, AtPMEI11, or AtPMEI12 expression is reduced. A higher induction in the relative amounts of Ara, Gal, Glu, and Xyl was observed in all pmei mutants with respect to the wild type, while increases of Rha and GlcA were found only in pmei10 and pmei11 plants. The monosaccharide composition of the noncellulosic polysaccharides extracted from mock-inoculated leaves of wild-type and pmei10 and pmei12 plants was similar (Supplemental Fig. S15). Instead, pmei11 CW had a reduced GalA content and an increased content of Ara, Gal, Glu, and Xyl. Since pmei11 has a basal higher PME activity and a basal low DME with respect to the wild type (Fig. 9), its pectin could be particularly sensitive to endogenous pectinases.

Figure 10.

pmei10, pmei11, and pmei12 mutants showed a higher pectin degradation with respect to the wild type (WT) during B. cinerea infection. The monosaccharide composition of matrix CW polysaccharides was monitored in leaves of 6-week-old Arabidopsis wild-type and pmei mutant plants challenged with B. cinerea at 48 hpi. The molar percentages of Fuc, Rha, Ara, Gal, Glc (Glu), Xyl, Man, GalA, and GlcA released after 2 m trifluoracetic acid (TFA) hydrolysis were quantified by high-performance anion-exchange chromatography with pulsed amperometric detection. Results represent means ± sd (n = 3). Different letters for each monosaccharide indicate data sets significantly different according to ANOVA followed by Tukey’s test (P < 0.05). The experiments were repeated three times with similar results.

Afterward, we explored the impact of AtPMEI10, AtPMEI11, and AtPMEI12 expression on the dynamic of CW polysaccharide structures and proteins during infection. We performed a carbohydrate microarray polymer profiling (CoMPP; Moller et al., 2007, 2008) on pectin- and hemicellulose-enriched fractions isolated from CW of B. cinerea-infected and mock-inoculated wild-type and pmei leaves at 48 hpi. A set of monoclonal antibodies (mAbs) specific for CW glycan epitopes was used (Table II). A heat map showing the mean CoMPP signals obtained for all the mAbs is shown in Figure 11A. Specific modifications of CW structure and components occurred in Arabidopsis wild-type leaves when challenged with B. cinerea. The analysis indicates pectin as the substrate most significantly altered during B. cinerea infection. All epitopes relative to different degrees of HG methylesterification were reduced in the pectin fractions of infected wild-type and pmei plants, although to different extents. Interestingly, pmei mutants, in comparison with wild-type plants, showed a greater reduction in JIM7 signal, specific for high methylesterified pectins with respect to JIM5, recognizing pectin with a low DME (Fig. 11B). In contrast, a similar increase of unesterified pectin recognized by LM19 was detected in hemicellulose fractions of infected CWs in all plants analyzed. In pectin fractions of wild-type infected leaves, RGI backbone epitopes (recognized by mAbs INRA-RU1 and INRA-RU2) decreased at a higher extent with respect to epitopes relative to the RGI side chain (recognized by mAbs LM5 and LM6). These reductions were slightly higher in pmei mutants with respect to the wild type. In hemicellulose fractions, all RGI epitopes increased to the same extent in wild-type and pmei plants. The increase of different pectic epitopes in the hemicellulose fraction could reflect the induction of linkage between pectin and hemicelluloses as potential mechanisms to defend CW against the action of the degradative fungal enzymes. Signals of LM15 and BS-400-4, specific for nongalactosylated xyloglucan and galactomannans, respectively, were reduced in the hemicellulose of infected wild-type and mutant leaves. Instead, the signals of LM24, relative to galactosylated xyloglucan epitopes, and LM21, specific for galactomannans/glucomannans, seem not to be affected during infection. These results suggest that the addition of specific substitutions could represent a strengthening mechanism to protect xyloglucan and mannans from enzymatic degradation during disease, as already proposed during development (Peña et al., 2004). The appearance of signals of BS-400-2, specific for β-(1,3)-d-glucan, in infected CWs indicates that both the wild type and mutants accumulated callose against B. cinerea. The signals of JIM11 and JIM20, both recognizing extensin epitopes, and of JIM13, specific for arabinogalactan proteins, increased to similar extents in all infected genotypes. The increase of extensins and arabinogalactan proteins could be exploited by plants to increase CW polysaccharide assembly and to reduce the accessibility of fungal CW-degrading enzymes (Rashid, 2016).

Table II. mAbs used for CoMPP.

AGP, Arabinogalactan protein; XXXG, oligosaccharide motif consisting of three Xyl-substituted (X) residues and one unsubstituted (G) glucosyl residue. L in XLLG indicates Gal linkage to Xyl.

mAb	Detected Epitope
JIM5	Partially methylesterified/unesterified HG
JIM7	Partially methylesterified/methylesterified HG
LM18	Partially methylesterified HG
LM19	Unesterified HG
LM20	Methylesterified HG
INRA-RU1	Backbone of RGI
INRA-RU2	Backbone of RGI
LM5	(1→4)-β-d-Galactan
LM6	(1→5)-α-l-Arabinan
BS-400-4	(1→4)-β-d-(Galacto)mannan
LM21	(1→4)-β-d-(Galacto)(gluco)mannan
BS-400-2	(1→3)-β-d-Glucan
LM15	Xyloglucan (XXXG motif)
LM24	Xyloglucan (XLLG motif)
LM10	(1→4)-β-d-Xylan
LM11	(1→4)-β-d-Xylan/arabinoxylan
JIM11	Extensin
JIM20	Extensin
JIM4	AGP
JIM13	AGP

Figure 11.

Changes in Arabidopsis CW structures of wild-type (WT) and pmei mutant plants during B. cinerea infection analyzed by CoMPP. A, Heat map showing the relative abundance of CW glycans recognized by different monoclonal antibodies (top) on pectin-enriched (diaminocyclohexanetetraacetic acid [CDTA]) and hemicellulose-enriched (NaOH) fractions extracted from CW prepared from B. cinerea-infected and mock-inoculated leaves of 6-week-old Arabidopsis wild-type and pmei10-1, pmei11-1 and pmei12-1 plants at 48 h post infection. The distribution of CW polysaccharide epitopes is represented as a heat map where mean CoMPP spot signals (numbered values) are correlated to color intensity. The highest mean signal value in the data set was set to 100, and all other values were adjusted accordingly. B, Heat map reporting percentages of the reduction of CoMPP signals related to antibodies recognizing HG methylesterification in the CDTA fraction of B. cinerea-infected leaves with respect to mock-inoculated leaves.

DISCUSSION

In this work, we argued about the possibility that plants can modulate PME activity by expressing specific defense-related PMEIs as a physiological response to pathogens. We selected AtPMEI10, AtPMEI11, and AtPMEI12 as possible pathogenesis-related PMEIs. These inhibitors showed a sequence homology and the presence of structural motifs typical of functional PMEIs. Interestingly, AtPMEI10 shows an SPRR/PMEI structure not described previously in this class of CW proteins. Similar domains are present in CW structural proteins such as extensins (Lamport et al., 2011). Interestingly, chimeric Leu-rich repeat/extensin CW proteins with roles in plant development and defense were reported (Jones and Jones, 1997; Baumberger et al., 2001; Draeger et al., 2015). As described for Leu-rich repeat/extensin CW protein chimeras (Ringli, 2010), the SPRR region in SPRR/PMEI could be required to anchor PMEI10 within the CW during infection. We demonstrated that AtPMEI11 and AtPMEI12 are correctly secreted into the apoplast, as also reported for other characterized PMEIs (Röckel et al., 2008; Hong et al., 2010; Zhang et al., 2010; De Caroli et al., 2011; Vandevenne et al., 2011; Reca et al., 2012). Further experiments are needed to confirm the predicted extracellular localization of AtPMEI10. The different levels and timing of expression in response to B. cinerea suggest that AtPMEI10, AtPMEI11, and AtPMEI12 are not functionally redundant and could play different roles during disease. AtPMEI11 is expressed in uninfected Arabidopsis leaves and repressed early during infection. At early stages of pathogen attack, the AtPMEI11 down-regulation as well as the absence of induction of the other inhibitors could favor the effects mediated by pathogen-induced PMEs, such as the possible release of OGs and MeOH. Later during infection, AtPMEI10, AtPMEI11, and AtPMEI12 are significantly induced. Microarray experiments showed that AtPMEI10, AtPMEI11, and AtPMEI12 also are induced by P. syringae, an hemibiotrophic bacterial pathogen, but not by Erysiphe orontii, a fungal biotroph (Supplemental Fig. S16; www.geneinvestigator.ethz.ch/at/). Overall, this evidence suggests that the new AtPMEIs are involved in immunity against pathogens, including the necrotrophy in their lifestyle. Intriguingly, the three PMEI members are highly expressed in senescing leaves, a developmental stage in which CW degradation occurs (Supplemental Fig. S8; Patro et al., 2014), suggesting that the expression of these PMEIs is activated in response to the loss of CW integrity.

Meta-analyses indicated that plant PMEs are differentially expressed in Arabidopsis in response to B. cinerea (Lionetti et al., 2012; Windram et al., 2012); however, information about the kinetics of expression of these genes remains scarce. We found that six PMEs were induced significantly while two isoforms were repressed during infection. The rapid and efficient activation of several PME isoforms (PME17, PME20, PME21, and, in particular PME41) indicates that plants employ an expensive metabolic response early against B. cinerea, most likely to favor the release of OGs and MeOH and trigger immunity against the pathogen. On the other hand, the steady down-regulation of PME members (PME1 and PME34) may represent, together with PMEI induction, a defense mechanism to limit pectin demethylesterification and degradation during infection. Consistently, TaPME1 expression is repressed in a resistant wheat genotype but induced in a susceptible accession during F. graminearum infection (Lionetti et al., 2015a). On the basis of PME/PMEI coexpression, we suggest PME3, PMEPCRA, PME17, PME20, and PME21 as putative ligands of AtPMEI10, AtPMEI11, and AtPMEI12. Future efforts will be devoted to study the molecular bases of PME-PMEI interactions during pathogenesis.

It is known that JA and ET signaling plays a strong role in the immunity of Arabidopsis to the necrotrophs (Thomma et al., 2001; Laluk and Mengiste, 2010). Our findings uncover the involvement of JA and ET immune signaling pathways in the activation of AtPMEI10 and AtPMEI11 in response to B. cinerea. A similar regulation also was proposed for the defense proteins PDF1.2, β-Chi, and Thi2.2 (Epple et al., 1995; Penninckx et al., 1998; Bari and Jones, 2009). It is also known that JA- and ET-dependent defense responses are triggered when the integrity of the plant CW is perturbed (Ellis and Turner, 2001; Ellis et al., 2002; Zhong et al., 2002). AtPMEI10 and AtPMEI11 could be part of a plant defense machinery against necrotrophs activated by both JA and ET aimed to maintain CW integrity, preventing excessive damage to host tissues (Overmyer et al., 2003). In contrast, the expression of AtPMEI12 is not affected by the alteration of JA signaling while the gene is strongly induced in the ein2-5 mutant, suggesting that ET signaling could contribute to the repression of the inhibitor during infection. The higher susceptibility of bak1-4 mutants to B. cinerea (Supplemental Fig. S4) indicates that, although BAK1 is involved in disease resistance against B. cinerea, as demonstrated previously for hemibiotrophs and biotrophs (Roux et al., 2011), it seems not to be required for AtPMEI activation. Our results exclude a possible role of the SA-mediated pathway in AtPMEI activation during defense, consistent with the major role of SA in Arabidopsis-biotroph interactions (Thomma et al., 2001; Nafisi et al., 2015). The ability of Arabidopsis to orchestrate the activation of specific AtPMEI isoforms during disease indicates the requirement of a fine modulation of PME activity during the course of infection.

PMEs release MeOH, a DAMP-like alarm signal, to alert adjacent noninfected tissues or neighboring plants (Dorokhov et al., 2012; Komarova et al., 2014). MeOH treatment can alter plant signaling responses to different DAMPs and microbe-associated molecular patterns (Hann et al., 2014). We revealed that AtPMEI11 expression is repressed by MeOH. This effect could be related to the down-regulation of AtPMEI11 observed at early stages of infection. This evidence, together with previous observations indicating that specific PMEs are MeOH-induced genes, support a role of MeOH as a danger signal to amplify PME-related immunity (Downie et al., 2004; Dorokhov et al., 2012; Komarova et al., 2014). RLP44 senses the alteration of pectin methylesterification during growth and abiotic stresses (Wolf et al., 2014). However, a possible role of RLP44 in response to pathogens has not yet been explored. We found that rlp44 mutants showed a higher induction of AtPMEI10, AtPMEI11, and AtPMEI12 and was more susceptible to B. cinerea with respect to the wild type (Supplemental Fig. S4). It is possible that RLP44 perceives the alterations of pectin methylesterification induced during infection and, by controlling PMEI expression, could boost PME-related immunity in concert with MeOH. OGs are able to elicit defense responses, including the accumulation of reactive oxygen species and pathogenesis-related proteins, and to protect plants against pathogen infection (Ferrari et al., 2013). We demonstrated that OG specifically induced AtPMEI11 expression. AtPMEI11 expression also is triggered by a number of elicitors/effectors (Supplemental Figs. S16 and S17). For instance, AtPMEI11 is affected by another DAMP, Pep2 (Huffaker et al., 2006; Supplemental Fig. S17), by different pathogen-associated molecular patterns (PAMPs), such as elf18 (Kunze et al., 2004), flg22 (Felix et al., 1999), LPS (Zeidler et al., 2004), and GST-NPP1 (Fellbrich et al., 2002), and by the effector HrpZ (Alfano et al., 1996; Supplemental Figs. S16 and S17). This evidence indicates that AtPMEI11, and consequently PME activity, can be regulated precisely by DAMPs, PAMPs, and effector-triggered immunity.

We isolated pmei10-1, pmei10-2, pmei11-1, pmei11-2, pmei12-1, and pmei12-2 mutants that, when infected with B. cinerea, exhibited a compromised resistance to the fungus. The higher susceptibility observed in pmei mutants was related to higher mycelium growth. We found a higher induction of BcOX1 in all pmei mutants with respect to wild-type plants. OA may be a cofactor in pathogenesis, acting in synergy with endopolygalacturonases during tissue maceration (ten Have et al., 2002). OA can cause hydration and swelling of pectin, favoring its degradation by endopolygalacturonases. Moreover, by sequestering the Ca²⁺ ions, OA may perturb the integrity of the pectic structure (Mansfield and Richardson, 1981). Through the expression of PMEIs, plants could engineer a substrate less enriched in Ca²⁺ ions and, thus, less susceptible to OA-mediated CW damage. BcBOA6 is expressed at lower levels in pmei mutants with respect to the wild type. These results indicate that pectin methylesterification can affect the production of fungal toxic compounds and suggest that B. cinerea selects specific toxins based on the different CW substrate to be degraded during infection. pmei mutants showed a higher increase of PME activity upon infection, indicating that AtPMEI10, AtPMEI11, and AtPMEI12 are able to control plant PME activity induced during disease. The low level of BcPME1 and BcPME2 expression at 48 hpi (Supplemental Fig. S18), the strong inhibition of B. cinerea-induced PME activity by AtPMEI1 (Supplemental Fig. S10), together with the evidence that PMEIs typically recognize plant PMEs (Raiola et al., 2004; Di Matteo et al., 2005; Reca et al., 2012; Lionetti et al., 2015c) support our conclusion. However, further research is needed to evaluate if the identified PMEIs also are able to inhibit fungal enzymatic activity.

The compositional profiling of CW, extracted from infected Arabidopsis wild-type leaves, revealed an extensive degradation of cellulose and HG and a reduction of DME, which intriguingly stops at the latter stages of infection. This physiological dynamics suggests that, when pectin demethylesterification reaches a dangerous level, it could be limited to protect CW by the fungus-degrading machinery later during infection. It is noteworthy that AtPMEI induction fits well with the block of pectin demethylesterification and CW degradation observed at late stages of infection. All pmei mutants exhibited a higher reduction of HG methylesterification and a greater degradation of HG with respect to the wild type. CoMPP indicated that pmei mutants showed a greater reduction of highly methylesterified pectins with respect to pectin with a low DME and a higher degradation of HG and RGI with respect to wild-type plants. The lack of PMEI expression does not affect hemicellulose strengthening, callose deposition, and the synthesis of structural defense proteins, here detected as CW-remodeling mechanisms induced against B. cinerea. Overall, our evidence indicates that AtPMEI10, AtPMEI11, and AtPMEI12 are exploited by Arabidopsis to limit demethylesterification and to protect pectin degradation and CW integrity in response to the pathogen. During infection, the increased amount of some monosaccharides observed in the wild type, and to a greater extent in all pmei mutants, could reflect the activation of a CW biosynthetic compensatory effect dependent on the extent of CW degradation. Intriguingly, this can be a CW-remodeling mechanism induced as a plant defense response against fungal OA production. The evidence that the toxin DON from the necrotrophic pathogenic fungus F. graminearum was able to alter the monosaccharide composition of Brachypodium distachyon CW supports this last hypothesis (Blümke et al., 2015).

The PMEI family appeared evolutionarily later than the PME genes (Wang et al., 2013). It is possible that plants have evolved specific PMEIs to control PME activity as a result of plant-pathogen coevolution (Jones and Dangl, 2006). In Figure 12, a dynamic model shows the possible roles of PMEIs in plant immunity against necrotrophs. In early phases of pathogen infection, plants induce PMEs to release MeOH and OGs, danger signals sensed by specific receptors to activate immune responses. By a feedback loop, MeOH and RLP44 receptor down-regulate resident PMEIs, possibly to favor a rapid and efficient local production of PME-related DAMPs. However, necrotrophic pathogens have evolved the ability to efficiently degrade low methylesterified pectin. At later stages of infection, plants defend CW integrity by triggering a signaling mediated by OG, JA, and ET. This leads to the induction of specific PMEIs able to lock the decrease of pectin methylesterification, protecting CW by further enzymatic degradations and limiting pathogen diffusion. This study brings new insights to molecular mechanisms of the regulation of PME activity during infection and the dynamic role of pectin methylesterification in plant immunity against necrotrophs. We highlight pectin methylesterification as a biochemical determinant of plant immunity and indicate AtPMEI10, AtPMEI11, and AtPMEI12 as new targets for crop breeding approaches to improve plant resistance to disease.

Figure 12.

Dynamic model showing the possible roles of PMEIs in plant immunity against necrotrophs. Step 1, In the early phases of pathogen infection, PME is required to release MeOH and OG danger signals, sensed by specific receptors to activate immunity. Step 2, In a feedback loop, MeOH and RLP44 down-regulate the resident pathogen-related AtPMEI expression to favor a rapid and efficient elicitation of defense responses by PME-related DAMPs. Pathogens have evolved the ability to efficiently degrade low-methylesterified pectin produced by PME activity. Step 3, At later stages of infection, plants defend CW integrity (CWI) by triggering a signaling mediated by OGs, JA, and ET. This leads to the induction of specific PMEIs able to lock the decrease of pectin methylesterification, protecting CW by further enzymatic degradation and limiting fungal diffusion. fPGs, Fungal polygalacturonases; fPMEs, fungal PMEs; PM, plasma membrane; pPMEs, plant PMEs.

MATERIALS AND METHODS

Plant Material and Growth Conditions

Arabidopsis (Arabidopsis thaliana) plants were grown in a controlled-environment chamber maintained at 22°C and 70% relative humidity with a 12-h/12-h day/night cycle (photosynthetically active radiation level of 100 µmol m⁻² s⁻¹). Germplasm used includes jar1-1 (Staswick et al., 1992), bak1-4 (Kemmerling et al., 2007), rlp44 (Wolf et al., 2014), ein2-5 (Alonso et al., 1999), and sid2-2 (Nawrath and Métraux, 1999). Mutants characterized in this study are as follows: SALK_007859C (pmei10-1 plants), SALK_072421 (pmei10-2 plants), SALK_015169c (pmei11-1 plants), GT_5_108240 (pmei11-2 plants), SALK_108076c (pmei12-1 plants), and GT_5_108791 (pmei12-2 plants). All SALK_c lines belong to the SALK Homozygote T-DNA Collection (http://signal.salk.edu/cgi-bin/homozygotes.cgi; Alonso et al., 2003) and were obtained from the Nottingham Arabidopsis Stock Centre. The homozygosity of the mutants was confirmed by PCR-based genotyping.

Cloning of AtPMEI11 and AtPMEI12 and Plasmid Construction and Agrobacterium tumefaciens Transient Transformation of Arabidopsis Leaf Cotyledons

The intronless regions of AtPMEI11 and AtPMEI12, encoding the predicted protein, were amplified by PCR from genomic DNA (50 ng) with ExTaq polymerase (Takara) and then amplified to introduce the restriction sites BamHI and NheI in each gene using primers reported in Supplemental Table S3. The BamHI/NheI fragments were inserted into a GFP-containing vector (De Caroli et al., 2015). All constructs were inserted as BamHI/SacI fragments into the plant binary vector (Haseloff et al., 1997) for A. tumefaciens-mediated expression in Arabidopsis seedlings. The cloned genes and the GFP constructs were checked by sequencing (Eurofins Genomics). The GFP constructs AtPMEI11-GFP, AtPMEI12-GFP, pm-rk, and PGIP2-RFP (De Caroli et al., 2011, 2015) were introduced into A. tumefaciens (strain LBA4404) and used for transient gene expression and coexpression of the chimeras in Arabidopsis seedling cotyledons (Li et al., 2009).

Confocal Laser Scanning Microscopy

Observations were made using a confocal laser microscope (LSM 710; Zeiss) and Zen Software. To detect GFP fluorescence, a 488-nm argon ion laser line was used, and the emission was recorded with a 505- to 530-nm filter set; RFP was detected with a 560- to 615-nm filter set after helium-neon laser excitation at 543 nm, while chlorophyll epifluorescence was detected with the filter set for tetramethylrhodamine isothiocyanate (greater than 650 nm) and eliminated. The power of each laser line, the gain, and the offset were identical for each experiment so that the images were comparable. Appropriate controls were performed to exclude the possibility of cross talk between the two fluorochromes before image acquisition.

Arabidopsis Infection with Botrytis cinerea

B. cinerea strain SF1 (Lionetti et al., 2007) was grown for 15 d on potato dextrose agar at 39 g L^–1 and 23°C with a 16-h-light/8-h-dark photoperiod before conidia collection. Conidia were harvested by washing the surface of the mycelium with 10 mL of sterile distilled water. Conidia suspensions were filtered to remove residual mycelium, and the spore concentration was determined using a Thoma chamber. To synchronize the germination, 2 × 10⁵ mL of conidia was incubated in potato dextrose broth at 24 g L⁻¹ at room temperature for 3 h. Fully developed leaves were detached from 6-week-old Arabidopsis plants. The detached leaves were placed in square petri dishes with petioles embedded in 0.8% (w/v) agar. Six droplets of conidia suspension (5 µL each) were placed on the surface of each leaf. Mock inoculation was performed using potato dextrose broth. Leaves were incubated at 24°C with a 16-h-light/8-h-dark photoperiod. The lesion size produced by B. cinerea was evaluated as an indicator of susceptibility to the fungus (Mengiste et al., 2003; Denby et al., 2004).

Determination of PME Activity

The quantification of PME activity was performed using the PECTOPLATE assay as described previously (Lionetti, 2015). Extractions of total proteins were obtained by homogenizing uninfected and infected leaves of 6-week-old Arabidopsis plants in the presence of 1 m NaCl, 12.5 mm citric acid, 50 mm Na₂HPO₄, 0.02% (w/v) sodium azide, and 1:100 (v/v) protease inhibitor (P8849; Sigma-Aldrich), pH 6.5 (2 mL of extraction buffer per g of tissue). The homogenates were shaken for 1.5 h at 4°C and centrifuged at 14,000g for 15 min, and the supernatant was collected. Protein concentration was determined in supernatants using the Bradford protein assay method (Bradford reagent; Sigma-Aldrich) and bovine serum albumin as a standard (Bradford, 1976). Equal amounts of protein samples (2 µg of total proteins in 20 μL) were loaded in each well of the PECTOPLATE. Plates were incubated at 30°C for 16 h and stained with 0.05% (w/v) Ruthenium Red (R2751; Sigma-Aldrich) for 30 min. The plates were destained by several washes with water, and the area of the fuchsia-stained haloes, resulting from the demethylesterification of pectin, was measured with ImageJ software (Abramoff et al., 2004). Known amounts of commercially available PME from orange (Citrus spp.) peel (P5400; Sigma-Aldrich) was used in the PECTOPLATE to generate a standard curve used to calculate PME activity in the protein extracts.

Elicitor Treatments, Gene Expression Analysis, and Mutant Genotyping

Leaves of 6-week-old Arabidopsis plants were infiltrated with OGs (50 μg mL⁻¹), MeOH (0.1%, v/v), or water as a control using a needleless syringe. At 1.5 hpi, leaves were collected and processed for RNA extraction. In all experiments, leaves were frozen in liquid nitrogen and homogenized using a mixer mill (MM301; Retsch) and inox beads (5 mm diameter) for about 1 min at 30 Hz, and total RNA was extracted with Isol-RNA Lysis Reagent (5′-Prime) according to the manufacturer’s instructions. RNA was treated with RQ1 DNase (Promega), and first-strand cDNA was synthesized using ImProm-II reverse transcriptase (Promega). qPCR analysis was performed using the CFX96 Real-Time System (Bio-Rad). One microliter of cDNA (corresponding to 50 ng of total RNA) was amplified in 30 μL of reaction mix containing 1× Go Taq qPCR Master Mix (Promega) and 0.4 μm of each primer. Primer sequences are shown in Supplemental Table S1. Expression levels of each gene, relative to the UBQ5 gene, were determined using a modification of the method of Pfaffl (2001). Meta-analysis of publicly available microarray data was performed using the Genevestigator database (https://www.genevestigator.com/gv/plant.jsp; Zimmermann et al., 2004).

Genomic DNA was extracted from rosette leaves of 6-week-old Arabidopsis wild-type plants as described previously (Edwards et al., 1991) and subjected to a PCR-based screening using primer pairs described in Supplemental Table S2. Go-Taq DNA polymerase (Promega) was used at 0.5 units per 30-µL assay. Amplification was carried out in the presence of deoxyribonucleotide triphosphates (0.5 mm), and specific primers (300 nm) in the buffer were provided by the supplier. Conditions for amplification were as follows: 94°C for 3 min; 30 cycles of amplification at 94°C for 30 s, 60°C for 30 s, and 72°C for 60 s; and a final extension of 72°C for 1 min. PCR products were separated by agarose gel electrophoresis and visualized by ethidium bromide staining.

Determination of H₂O₂ Accumulation, Callose Deposition, and Mycelium Growth

Arabidopsis leaves were assayed for H₂O₂ accumulation employing 3,3′-diaminobenzidine staining as described (Reem et al., 2016). For the determination of callose deposition in B. cinerea-infected and mock-inoculated leaves from 6-week-old Arabidopsis, plants were cleared with 100% (v/v) ethanol. Leaves were fixed in an acetic acid:ethanol (1:3) solution for 2 h, sequentially incubated for 15 min in 75% (v/v) ethanol, in 50% (v/v) ethanol, and in 150 mm phosphate buffer, pH 8, and then stained for 2 h at 25°C in 150 mm phosphate buffer, pH 8, containing 0.01% (w/v) Aniline Blue. After staining, leaves were mounted in 50% (v/v) glycerol and examined with a UV2A filter (excitation wavelength = 330–380 nm) using the epifluorescence microscope Nikon Eclipse E200. Photographs were taken with a Nikon Digital Sight DS-Fi1c camera. For the detection of B. cinerea hyphae in infected plant tissues, detached Arabidopsis leaves were washed for 1 h with gentle agitation in absolute ethanol at 60°C to remove chlorophyll. Thereafter, leaves were incubated for 30 min in lactophenol Trypan Blue solution (water:glycerol:lactic acid [1:1:1] + 10 µL of Trypan Blue solution [25 mg mL⁻¹; Sigma-Aldrich]). Finally, stained leaves were incubated for 20 min with gentle agitation in a destained solution (water:glycerol:lactic acid [1:1:1]) and transferred into 50% (v/v) glycerol solution for microscopy.

Extraction of Alcohol-Insoluble Residue

Uninfected, mock-inoculated, and B. cinerea-infected leaves of 6-week-old Arabidopsis plants were collected in screw-cap tubes, frozen in liquid nitrogen, and homogenized for 1 min at 30 Hz using a mixer mill (MM301; Retsch) and inox beads (5 mm diameter). Milled tissue was washed twice in prewarmed (70°C) 70% (v/v) ethanol, vortexed, and pelleted by centrifugation at 14,000g for 10 min. The pellet was suspended twice with a chloroform:MeOH mixture (1:1, v/v), shaken for 30 min at room temperature, and centrifuged at 14,000g for 10 min. Samples were pelleted by centrifugation at 14,000g for 10 min. Pellets were resuspended twice in 1 mL of 80% (v/v) acetone and spun at 14,000g for 5 min. Supernatants were discarded, and the pellet containing the alcohol-insoluble residue (AIR) was dried overnight at room temperature under a chemical hood. Starch was removed by treating the AIR with porcine type IA α-amylase (100 units g⁻¹ AIR; product no. A4268; Sigma-Aldrich) in 100 mm potassium phosphate buffer, pH 7.5, 5 mm NaCl, and 0.02% (w/v) NaN₃ for 24 h at 37°C. The suspension was centrifuged at 14,000g for 20 min, and the pellet was then washed with distilled water and 80% (v/v) acetone.

Determination of DME and Monosaccharide Composition of CW

Destarched AIR (2 mg) was saponified by suspending it in 30 μL of water and 10 μL of 1 m NaOH. The solution was incubated at room temperature for 1 h and neutralized afterward with 10 μL of HCl. After centrifugation at 14,000g for 10 min, aliquots of the supernatant (10 and 20 μL) were loaded on 96-well microtiter plates (cod.9018; Costar), and the volume was adjusted to 50 μL. Alcohol oxidase (50 μL) was added to each well (0.03 units in 0.1 m sodium phosphate, pH 7.5; Sigma-Aldrich), and this mixture was incubated at room temperature for 15 min on a shaker. Thereafter, 100 µL of a mixture containing 0.02 m 2,4-pentanedione in 2 m ammonium acetate and 0.05 m acetic acid was added to each well. After 10 min of incubation at 68°C, samples were cooled on ice, and the absorbance was measured at 412 nm in a microplate reader (ETI-System reader; Sorin Biomedica Cardio). The MeOH amount was estimated as described previously (Klavons and Bennett, 1986). The DME was expressed as MeOH-to-uronic acid molar ratio (%).

Two milligrams of saponified AIR samples was incubated in 200 µL of 2 m TFA at 121°C. After 1 h, 200 µL of isopropanol was added, and the mixtures were evaporated at 40°C with a stream of N₂ gas. This step was repeated twice, and the samples were dried at room temperature overnight. The TFA-hydrolyzed monosaccharides were suspended in 200 µL of water. The monosaccharide composition of destarched and TFA-hydrolyzed AIR was determined by high-performance anion-exchange chromatography with pulsed amperometric detection using a PA20 column (Dionex). Peaks were identified and quantified by comparison with a standard mixture of Rha, Ara, Fuc, Gal, Xyl, Man, GalA, and GlcA (Sigma-Aldrich). The crystalline cellulose was determined as described previously (Updegraff, 1969). The cellulose-derived Glc content in destarched AIR was determined by an anthrone colorimetric assay (Scott and Melvin, 1953) with Glc (G8270; Sigma-Aldrich) as a standard.

CoMPP Analysis of CW Material

For the CoMPP analysis, a pool of 100 B. cinerea-infected and mock-inoculated leaves isolated from at least 10 different Arabidopsis plants was used for AIR extraction for each experiment. The chemical fractionation process of AIR material (10 mg) was performed in two extraction steps: first, with 50 mm diaminocyclohexanetetraacetic acid for predominant pectin extraction; then, with 4 m NaOH in 0.1% (w/v) NaBH₄, and the samples were spotted accordingly (Moller et al., 2007, 2008) using a microarray robot (Sprint; Arrayjet). Once printed, arrays were blocked with phosphate-buffered saline (PBS) containing 5% (w/v) low-fat milk powder (MPBS). Arrays were washed with PBS and probed with antibodies (PlantProbes; Leeds University) in 5% (w/v) MPBS. Subsequently, arrays were washed in PBS and incubated with anti-rat secondary antibody conjugated to alkaline phosphatase (Sigma-Aldrich) in 5% (w/v) MPBS (1:5,000). Arrays were developed in a solution containing 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium in alkaline phosphatase buffer (100 mm NaCl, 5 mm MgCl₂, and 100 mm diethanolamine, pH 9.5). Developed microarrays were scanned at 2,400 dpi (CanoScan 8800 F; Canon) and converted to TIFFs. Antibody signals were measured using appropriate software (Array-Pro Analyzer 6.3; Media Cybernetics). The highest mean spot signal in the data set was assigned a value of 100%, and all other signals were adjusted accordingly (Moller et al., 2007, 2008).

Accession Numbers

Details regarding the sequences of the genes analyzed in this study can be obtained with the following accession numbers: AtPMEI1 (At1g48020), AtPMEI2 (At3g17220), AtPMEI3 (At5g20740), AtPMEI4 (At4g25250), AtPMEI5 (At2g31430), AtPMEI6 (At2g47670) AtPMEI10 (At1g62760), AtPMEI11 (At3g47380), AtPMEI12 (At5g46960), PMEPCRA (At1g11580), PME1 (At1g53840), PME3 (At3g14310), PME4 (At3g14300), PME6 (At4g33230), PME7 (At5g07420), PME8 (At3g60730), PME17 (At2g45220), PME20 (At2g47550), PME21 (At3g05610), PME34 (At3g49220), PME41 (At4g02330), JAR1 (At2g46370), ABA2 (At1g52340), BAK1 (At4g33430), RLP44 (AT3G49750), EIN2 (At5g03280), SID2 (At1g74710), UBQ5 (At3g62250), CaPMEI (ABG47806), AdPMEI (P83326), BoPMEI (AAZ20131), SolyPMEI (SGN-U601352), VvPMEI1 (XP_010660323), AtC/VIF1 (AT1G47960), NtCIF (CAA73333), NtVIF (CAA73334), SolyCIF (SGN-U317539), and SolyVIF (SGN-U567308)

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Alignment of amino acid sequences of putative Arabidopsis invertase/PMEI isoforms with an altered expression during B. cinerea infection.

• Supplemental Figure S2. Alignment of amino acidic sequences of AtPMEI10, AtPMEI11, and AtPMEI12.

• Supplemental Figure S3. Nucleotide and amino acid sequences of AtPMEI10.

• Supplemental Figure S4. Disease symptoms produced by B. cinerea infection in immune signaling mutants.

• Supplemental Figure S5. Analysis of cis-acting DNA elements with regulatory functions in plant immunity in the 5′ flanking regions of AtPMEI10, AtPMEI11, and AtPMEI12.

• Supplemental Figure S6. Arabidopsis pmei10-1, pmei11-1, and pmei12-1 mutants are not defective in H₂O₂ and callose accumulation against B. cinerea.

• Supplemental Figure S7. Representative images illustrating the morphology of different tissues of pmei10-1, pmei11-1, and pmei12-1 mutants and wild-type plants.

• Supplemental Figure S8. AtPMEI10, AtPMEI11, and AtPMEI12 gene expression in different tissues and developmental stages of Arabidopsis.

• Supplemental Figure S9. Histochemical evidence indicating the control of PME activity by AtPMEI10, AtPMEI11, and AtPMEI12 during B. cinerea infection.

• Supplemental Figure S10. Effect of exogenous addition of recombinant AtPMEI1 on PME activity at 48 hpi.

• Supplemental Figure S11. Dynamic modifications of DME in Arabidopsis leaves during B. cinerea infection.

• Supplemental Figure S12. Dynamic modifications of CW in Arabidopsis leaves during B. cinerea infection.

• Supplemental Figure S13. Evaluation of Arabidopsis susceptibility during B. cinerea infection.

• Supplemental Figure S14. Monosaccharide composition at different time points of CW of mock-inoculated leaves of Arabidopsis wild-type plants.

• Supplemental Figure S15. Monosaccharide composition of CW of mock-inoculated leaves of Arabidopsis wild-type and pmei mutant plants.

• Supplemental Figure S16. AtPMEI10, AtPMEI11, and AtPMEI12 gene expression during infection of Arabidopsis with different pathogens and following treatment with several elicitors.

• Supplemental Figure S17. Expression patterns of AtPMEI10, AtPMEI11, and AtPMEI12 mRNA in Arabidopsis in response to different PAMPs, effectors, and DAMPs.

• Supplemental Figure S18. Levels and kinetics of expression of B. cinerea PMEs during infection.

• Supplemental Table S1. Primers used for RT-PCR.

• Supplemental Table S2. Primers used for mutant genotyping.

• Supplemental Table S3. Primers used for AtPMEI11 and AtPMEI12 cloning and plasmid construction.

Glossary

CW

cell wall

HG

homogalacturonan

RGI

rhamnogalacturonan I

RGII

rhamnogalacturonan II

MeOH

methanol

DME

degree of methylesterification

JA

jasmonic acid

DAMP

damage-associated molecular pattern

OG

oligogalacturonide

RT

reverse transcription

hpi

hours post inoculation

ET

ethylene

SA

salicylic acid

UTR

untranslated region

OA

oxalic acid

CoMPP

carbohydrate microarray polymer profiling

mAb

monoclonal antibody

PAMPs

pathogen-associated molecular patterns

qPCR

quantitative PCR

AIR

alcohol-insoluble residue

TFA

trifluoracetic acid

PBS

phosphate-buffered saline

MPBS

phosphate-buffered saline containing 5% (w/v) low-fat milk powder