Plant Physiology Preview. Published on August 26, 2013, as DOI:10.1104/pp.113.223503 1 For Plant Physiology 2 Running head: Phospholipase D in pathogen resistance signaling 3 4 Correspondence to: 5 Mats X. Andersson 6 Department of Biological and Environmental sciences, University of Gothenburg 7 Box 461 8 SE-405 30 Göteborg 9 Sweden 10 Email: [email protected] 11 Phone: +46 31 786 2688 12 13 14 Research area: Signaling and Response 1 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Copyright 2013 by the American Society of Plant Biologists 15 δ Arabidopsis phospholipase D is involved in basal defense and non-host resistance to 16 powdery mildew fungi 17 18 Francesco Pinosa1, Nathalie Buhot1, Mark Kwaaitaal2, Per Fahlberg1, Hans Thordal- 19 Christensen2, Mats Ellerström1, Mats X. Andersson1* 20 1.Department of Biological and Environmental Sciences, University of Gothenburg 21 2.Department of Plant and Environmental Sciences, Faculty of Science University of Copenhagen 22 *Correspondence to [email protected] 23 24 One sentence summary: 25 The Arabidopsis phospholipase D family was probed using reverse genetics for involvement in 26 δ cell wall based defense to non-host powdery mildew and the PLD isoform was identified as a 27 component in the defense reaction. 28 2 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 29 Financial sources 30 31 The financial support of the Swedish Council for Environment, Agricultural Sciences and Spatial 32 Planning (project no. 2007-1563, 2009-888 and 2007-1051), the Carl Tryggers foundation, the 33 Olle Engkvist Byggmästare and The Danish Council for Independent Research , Technology and 34 Production Sciences (project no. 10-082292) foundation is gratefully acknowledged. 35 36 Author addresses 37 38 Francesco Pinosa, Per Fahlberg, Nathalie Buhot, Mats Ellerström and Mats X. Andersson: 39 Department of Biological and Environmental Sciences, University of Gothenburg, Box 461, SE- 40 405 30 Göteborg, Sweden 41 42 Mark Kwaaitaal and Hans Thordal-Christensen: Section for Plant and Soil Science, Department 43 of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 44 Frederiksberg C, Denmark 45 46 Correspondence to Mats X. Andersson, [email protected] 3 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 47 Abstract 48 Plants have evolved a complex array of defensive responses against pathogenic microorganisms. 49 Recognition of microbes initiates signaling cascades that activate plant defenses. The membrane 50 lipid phosphatidic acid (PA) produced by phospholipase D (PLD) has been shown to take part in 51 both abiotic and biotic stress signaling. In this study, the involvement of PLD in the interaction 52 between Arabidopsis thaliana and the barley powdery mildew fungus, Blumeria graminis f. sp. 53 hordei (Bgh) was investigated. This non-adapted pathogen is normally resisted by a cell wall- 54 based defense, which stops the fungal hyphae from penetrating the epidermal cell wall. Chemical 55 inhibition of PA production by PLD increased the penetration rate of Bgh spores on wild-type 56 leaves. The analysis of T-DNA knock-out lines for all Arabidopsis PLD genes revealed that 57 PLDδ is involved in penetration resistance against Bgh, and chemical inhibition of PLDs in 58 plants mutated in PLDδ indicated that this isoform alone is involved in Bgh resistance. In 59 addition, we confirmed the involvement of PLDδ in penetration resistance against another non- 60 adapted pea powdery mildew fungus, Erysiphe pisi. A GFP fusion of PLDδ localized to the 61 plasma membrane at the Bgh attack site, where it surrounded the cell wall reinforcement. 62 δ Furthermore, in the pld mutant transcriptional upregulation of early Microbe Associated 63 Molecular Pattern (MAMP) response genes was delayed after chitin stimulation. In conclusion, 64 we propose that PLD is involved in defense signaling in non-host resistance against powdery 65 mildew fungi and put PLDδ forward as the main isoform participating in this process. 66 67 4 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 68 Introduction 69 70 A wide range of potentially pathogenic microbes are found in the environment of plants. 71 Nevertheless, plants are able to resist the great majority of microbes they encounter and are 72 susceptible to only a small number of specifically adapted pathogens. This capacity hinges 73 primarily on a broad-range form of immunity called “non-host resistance” (NHR), which is 74 defined as the immunity displayed by an entire plant species against all genetic variants of a 75 pathogen (Thordal-Christensen, 2003). NHR against pathogens of distantly related host species 76 is believed to be due to basal defense activated after recognition of microbe-associated molecular 77 patterns (MAMPs) (Schulze-Lefert and Panstruga, 2011). While adapted pathogens have evolved 78 effector proteins that target distinct plant mechanisms for instance to suppress plant defenses 79 (Jones and Dangl, 2006), the effectors of non-adapted pathogens are not able to sufficiently 80 suppress plant defenses. In this way, NHR is provided by multiple genes, which makes it 81 effective and genetically robust (Schulze-Lefert and Panstruga, 2011). This form of resistance 82 has thus gained particular interest for its agro-economic potential. 83 84 A well-studied model system for this type of NHR is the interaction between Arabidopsis 85 thaliana and non-adapted powdery mildew fungi (Micali et al., 2008). After germination on a 86 leaf, fungal spores develop a specialized hyphal structure, called “appressorium”, which attaches 87 to the epidermal cell wall. Subsequently, the fungus attempts to penetrate the cell from the 88 attachment site. Here plants are thought to recognize fungal MAMPs and induce localized 89 defense responses at the site of interaction. Fungal penetration is usually stopped by the 90 formation of a multi-layered cell wall reinforcement known as a papilla. The papilla is rich in 91 callose and other antimicrobial substances, which provides both a physical and a chemical barrier 92 against the invading pathogen (Hardham et al., 2007; Hückelhoven, 2007). When powdery 93 mildew fungi manage to evade this defense, they form a haustorium, a feeding structure that 94 invades the host cell. Successfully penetrating spores are stopped either by a second level of 95 defense: confinement of the haustorium within a callose encasement, or a third level of defense: 96 a hypersensitive cell death response to stop further fungal development (Lipka et al., 2008). 97 Thus, NHR against non-adapted powdery mildews and other biotrophic fungi can be 98 conceptually divided into pre- and post-invasion defenses. 5 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 99 100 It is well established that cell wall-based pathogen recognition in plants is mediated by trans- 101 membrane pattern recognition receptors (PRRs) that recognize highly conserved MAMPs. 102 Examples of fungal MAMPs are xylanase and chitin. (Boller and Felix, 2009; Monaghan and 103 Zipfel, 2012). The corresponding PRRs, Ethylene Inducing-Xylanase (EIX)2 in tomato (Ron and 104 Avni, 2004), Chitin Elicitor Receptor Kinase (CERK)1 in Arabidopsis (Miya et al., 2007) and 105 Chitin elicitor binding protein (CEBiP) in rice (Shimizu et al., 2010) have been identified. 106 Notably, the chitin receptor CERK1 has been found to be essential for resistance against a 107 powdery mildew fungus in Arabidopsis (Wan et al., 2008). Several molecular and physiological 108 changes are initially brought about by MAMP perception. These include fluxes of various ions 109 including Ca2+ over the plasma membrane, protein phosphorylation, generation of reactive 110 oxygen species and nitric oxide, activation of mitogen-activated protein kinase (MAPK), Ca2+- 111 dependent protein kinase (CDPK) cascades and finally transcriptional changes in defense genes 112 (Hückelhoven, 2007; Schwessinger and Zipfel, 2008; Boller and Felix, 2009). These early 113 responses to MAMP detection are well studied and known to depend on one another (Ogasawara 114 et al., 2008; Ma et al., 2009). However, thorough insight into the regulation and the genetic basis 115 of these events remains fragmented. 116 117 Three major components of pre-invasion defense were identified in forward genetic screens for 118 Arabidopsis mutants with enhanced penetration frequencies of the non-adapted powdery mildew 119 fungus, Blumeria graminis f. sp. hordei (Bgh) (Collins et al., 2003; Lipka et al., 2005; Stein et 120 al., 2006). These correspond to the PENETRATION1 (PEN1), PEN2 and PEN3 genes, which 121 encode a plasma membrane-localized syntaxin, a glycoside hydrolase and a pleiotropic drug 122 resistance ATP binding cassette transporter, respectively. The current model predicts existence 123 of a PEN1-defined pathway where this syntaxin is involved in polar secretion of material at sites 124 of attempted fungal ingress, together with an ADP ribosylation factor (ARF) GTPase and the 125 corresponding, GTP exchange factor, GNOM (Böhlenius et al., 2010; Nielsen et al., 2012). In a 126 second pathway, PEN2 is required for production of an antifungal glucosinolate product, 127 secreted into the apoplast by PEN3 (Bednarek et al., 2009). The three PEN proteins accumulate 128 at the incipient fungal entry site: PEN1 in exosomes in the papilla body (Meyer et al., 2009; 6 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 129 Nielsen et al., 2012), PEN2 in peroxisomes in the cytoplasmic area surrounding the papilla 130 (Lipka et al., 2005) and PEN3 at the plasma membrane (Stein et al., 2006). 131 132 Several lipids and lipid metabolites have been shown to function in the signal transduction 133 pathway leading to activation of plant defense responses (Laxalt and Munnik, 2002; Kachroo and 134 Kachroo, 2009). A prominent role has been postulated for phosphatidic acid (PA), whose levels 135 increase within minutes in plant cells recognizing different MAMPs (van der Luit et al., 2000) or 136 upon recognition of pathogen effector proteins (De Jong et al., 2004; Andersson et al., 2006; 137 Kirik and Mudgett, 2009). PA has been implicated in the modulation of MAPK activity, Ca2+ 138 influx from the apoplast, and oxidative burst during abiotic and biotic stresses (Testerink and 139 Munnik, 2011). In plants, PA can be formed via two main pathways: the direct hydrolysis of 140 phospholipids by phospholipase D (PLD) or the combined action of phospholipase C (PLC) and 141 diacylglycerol kinase (Wang, 2004). Both pathways are involved in the early signaling events of 142 plant-pathogen interactions and contribute to the activation of plant defenses. Specifically, PLD 143 activity has been found to be required for the hypersensitive response induced by recognition of 144 bacterial effector proteins (Andersson et al., 2006; Kirik and Mudgett, 2009). 145 146 The twelve different PLD isoforms encoded in the Arabidopsis genome are classified into six 147 groups (α, β, γ, δ, ε, ζ), based on sequence similarity and in vitro activity. Two HxKxxxxD 148 (HKD) motifs that interact with each other are essential for lipase activity in all eukaryotic PLD 149 isoforms (Bargmann and Munnik, 2006). Although an increase in PLD activity upon treatment 150 with various MAMPs has been reported for different cell culture systems (van der Luit et al., 151 2000; Yamaguchi et al., 2005; Suzuki et al., 2007), genetic evidence for the involvement of 152 specific PLD isoforms in pathogen defense is so far missing. Here, we reveal a role for PLD 153 activity in Arabidopsis cell wall-based defense against powdery mildew fungi and genetically 154 δ identify PLD as the responsible PLD isoform. We also report that this PLD is involved in 155 MAMP signaling at a more general level. 156 157 Results 158 159 Inhibition of PLD-generated PA causes increased Bgh penetration in Arabidopsis 7 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 160 The development of Bgh spores on Arabidopsis leaves is synchronous and penetration attempts 161 occur in a narrow time frame around 12 hours-post-inoculation (hpi) (Assaad et al., 2004). It is 162 thus possible to score and compare fungal penetration on different plants at a certain time point. 163 To test a possible contribution of PLD-generated PA to plant resistance against Bgh penetration, 164 we exploited the transphosphatidylation activity of PLD, which uses primary alcohols as 165 substrates to form an artificial phosphatidyl alcohol. The preferential formation of this compound 166 impairs PA production (Munnik et al., 1995). Thus, we employed n-butanol to inhibit PA 167 formation by PLD. Leaves were infiltrated with different n-butanol or tert-butanol concentrations 168 3 hours before inoculation and penetration was scored at 3 days-post-inoculation (dpi). 169 Increasing n-butanol concentrations caused progressively higher Bgh penetration rates, from 170 16% (SD 1.0) for a water control to 38% (SD 7.5) for leaves treated with 0.6% n-butanol (Fig. 171 1A). In contrast, the penetration rate for leaves infiltrated with tert-butanol, an alcohol with no 172 inhibitory effects on PA formation, was not significantly different compared to that of the water 173 control. The results thus indicate the involvement of PLD and PA in cell wall-based defense 174 against Bgh. We next attempted to measure PLD activity and PA accumulation during the 175 defense response. To this end, leaf explants were labeled with 33PO over-night before 4 176 inoculation with Bgh spores. However, no increase in PA label could be measured at 4, 8, 12 and 177 24 hpi (Figure S1). 178 179 PLDδ is the only PLD isoform involved in penetration resistance against Bgh 180 We set out to identify which, if any, of the twelve known Arabidopsis PLD isoforms contributes 181 to penetration resistance against Bgh. For this purpose, we identified at least one T-DNA 182 insertion line for each PLD gene. The insertion site of each T-DNA line was verified by 183 sequencing (Table S I and Fig. S2). Homozygous plants for T-DNA insertion lines not 184 previously published were tested by RT-PCR for the presence of the corresponding PLD 185 transcript using primers flanking the insertion site (Fig. S3 and Table S II). Absence of 186 transcripts confirmed that the T-DNA lines indeed are knock-out (KO) mutants of the PLD 187 genes. Absence of transcript and/or protein in the pldα1, pldα3, pldζ1 and pldζ2 mutants was 188 described in previous publications (Zhang et al., 2004; Li et al., 2006; Hong et al., 2008). The 189 different PLD single KO lines and two PLD double KO lines (pldβ1-2 pldβ2 and pldζ1 pldζ2) 190 were scored for Bgh penetration resistance by counting interaction events in leaves stained by 8 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 191 trypan blue two days after inoculation with Bgh spores. Among all, only the PLDδ KO line 192 δ (pld ) displayed a significantly higher Bgh penetration rate compared to wild-type (Fig. 1B). The 193 penetration rate in this mutant was approximately doubled compared to wild-type, representing 194 an increase from 12% (SD 0.8) penetration in wild-type to over 23% (SD 2.3) in pldδ. None of 195 the PLD mutants investigated was affected in the cell death response to successfully penetrating 196 spores (Fig. S4). Furthermore, none of the lines displayed any visible growth or other phenotype 197 under our cultivation regime (Fig. S5). 198 199 Absence of the PLDδ−specific transcript in pldδ was verified with a primer pair annealing 200 downstream of the insertion site in addition to the T-DNA flanking pair (Fig. 2A-B). To confirm 201 that the penetration phenotype of pldδ is caused by the specific loss of the PLDδ protein, the 202 mutant was complemented with PLDδ coding sequence fused to GFP under the transcriptional 203 regulation of a 1.4 kb PLDδ promoter region. This restored wild-type levels of Bgh penetration 204 (Fig. 2C). In addition, pldδ plants were transformed with a mutated version of the pPLDδ:PLDδ- 205 GFP transgene, which carried a point mutation turning a histidine into aspartic acid residue 206 (Η707D) in one of the catalytically active HKD sites of PLDδ. This H to D mutation was 207 previously shown to abolish the enzymatic activity of a PLD isoform in cabbage (Brassica 208 oleracea) (Lerchner et al., 2006). Transgenic pldδ plants expressing the pPLDδ:PLDδΗ707D- 209 GFP construct displayed Bgh penetration levels similar to the untransformed pldδ (Fig. 2B-C), 210 further supporting that PLDδ activity is essential for normal penetration resistance. 211 212 To verify that PLDδ is the only PLD isoform that plays a role in penetration resistance against 213 Bgh, the pldδ mutant was treated with n-butanol before inoculation with Bgh. This treatment 214 caused no further increase in Bgh penetration rate (Fig. 3A). The transcripts of PLDα1 and PLDδ 215 are the most abundant among the PLD genes in green tissues (Li et al., 2006), while PLDβ1 and 216 PLDβ2 are the two isoforms closest related to PLDδ (Qin and Wang, 2002). Bgh penetration was 217 scored in pldα1 pldδ double- and pldβ1-2 pldβ2 pldδ triple-mutants, but none of these displayed 218 higher levels than the pldδ single mutant (Fig. 3B and C). This, together with the n-butanol 219 treatment of pldδ, verified that PLDδ is the sole PLD isoform involved in penetration resistance. 9 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 220 Double mutants combining the pldδ mutation with the pen1-1 or the pen3-1 mutations, did not 221 display higher Bgh penetration frequencies than the respective parent pen mutants (Fig. 3D). 222 223 PLDδ contributes to penetration resistance against the non-host pea powdery mildew 224 To investigate the conservation of the involvement of PLDδ in penetration resistance against 225 powdery mildew fungi, we used the Arabidopsis non-host fungal pathogen, Erysiphe pisi (Ep), 226 which is an adapted powdery mildew fungus on pea plants. Ep penetration rate on wild-type 227 Arabidopsis leaves was higher than that of Bgh, suggesting a higher level of adaptation to 228 Arabidopsis. The pldδ mutant displayed a significantly higher penetration rate compared to wild- 229 type for Ep (Fig. 4). Penetration success rate increased from 18% (SD 3.1) of the germinated 230 spores on wild-type leaves to 35% (SD 4.0) on the pldδ mutant. This, thus clearly demonstrates 231 that the PLDδ gene is involved in penetration resistance against this fungus as well. 232 233 PLDδ in chitin-induced MAMP-response gene activation 234 Since knockout of the PLDδ gene caused decreased penetration resistance against two different 235 non-host powdery mildew fungi, and since the gene appears not specifically linked to either 236 PEN1- or PEN2/PEN3-defined penetration resistance pathways, we hypothesized that the 237 activity of the PLDδ isoform might have a more general role in plant pathogen defense. In 238 particular, it seems reasonable to suggest that the gene might be involved in signaling after 239 recognition of MAMPs. It would be interesting to test the response of the pldδ mutant to 240 MAMPs, such as chitin, since a chitin receptor mutant has been previously found to have 241 reduced powdery mildew resistance (Wan et al., 2008). For this purpose, wild-type and 242 pldδ mutant seedlings were treated with a chitin extract and transcriptional changes of the early 243 MAMP-induced genes, FRK1, NHL10 and PHI-1 (Boudsocq et al., 2010; Kwaaitaal et al., 2011), 244 were evaluated by QRT-PCR (Fig. 5). The transcript levels of all three genes rapidly increased 245 following treatment with chitin in both the wild-type and mutant plants. PHI1 and NHL10 both 246 reached a peak of transcript accumulation 30 min after treatment, whereas FRK1 transcript 247 increased throughout the experiment. However, the pldδ mutant showed a trend for an 248 approximately 50% reduction in transcript accumulation after 30 min for PHI1 and NHL10. For 249 all three genes, a trend for a delayed chitin-induced accumulation of transcripts was observed. 10 Downloaded from on April 13, 2019 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved.