The tyrosine-sulfated peptides PSKα and PSY1 bind to specific leucine-rich repeat surface receptor kinases and control cell proliferation in plants. In a reverse genetic screen, we identified the phytosulfokine (PSK) receptor PSKR1 as an important component of plant defense. Multiple independent loss-of-function mutants in PSKR1 are more resistant to biotrophic bacteria, show enhanced pathogen-associated molecular pattern responses and less lesion formation after infection with the bacterial pathogen Pseudomonas syringae pv. tomato DC3000. By contrast, pskr1 mutants are more susceptible to necrotrophic fungal infection with Alternaria brassicicola, show more lesion formation and fungal growth which is not observed on wild-type plants. The antagonistic effect on biotrophic and necrotrophic pathogen resistance is reflected by enhanced salicylate and reduced jasmonate responses in the mutants, suggesting that PSKR1 suppresses salicylate-dependent defense responses. Detailed analysis of single and multiple mutations in the three paralogous genes PSKR1, -2 and PSY1-receptor (PSY1R) determined that PSKR1 and PSY1R, but not PSKR2, have a partially redundant effect on plant immunity. In animals and plants, peptide sulfation is catalyzed by a tyrosylprotein sulfotransferase (TPST). Mutants lacking TPST show increased resistance to bacterial infection and increased susceptibility to fungal infection, mimicking the triple receptor mutant phenotypes. Feeding experiments with PSKα in tpst-1 mutants partially restore the defense-related phenotypes, indicating that perception of the PSKα peptide has a direct effect on plant defense. These results suggest that the PSKR subfamily integrates growth-promoting and defense signals mediated by sulfated peptides and modulates cellular plasticity to allow flexible adjustment to environmental changes.
Due to their sessile lifestyle, plants need to adapt to their environment by allocating their resources according to the stresses they are exposed to. Therefore, plants have evolved strategies to counteract the effects of heat, light, wounding and pathogen attack and to integrate adaptive responses to all of these. These responses are interconnected and even defense against biotrophic and necrotrophic pathogen attack is regulated via mutually antagonistic salicylate- (SA) and jasmonate (JA)-dependent signaling pathways (Glazebrook, 2005; Jones and Dangl, 2006; Koornneef and Pieterse, 2008; Spoel and Dong, 2008; Vlot et al., 2009; von Saint Paul et al., 2011).
The prerequisite for an appropriate stress response is its perception. Thus, receptor activation is the first step in the initiation of adaptive processes. The first layer of plant innate immunity to biotrophic pathogens is based on the perception of pathogen-associated molecular patterns (PAMPs) by cell surface receptors such as the leucine-rich repeat receptor-like kinases (LRR-RLKs) FLS2 and EFR, the receptors for bacterial flagellin and Elongation Factor Tu, respectively (Chisholm et al., 2006; Jones and Dangl, 2006; Boller and Felix, 2009; Postel and Kemmerling, 2009; Postel et al., 2010; Segonzac and Zipfel, 2011). These receptors form ligand-dependent complexes with other RLKs such as BRI1-Associated Kinase1 (BAK1) and BAK1-like1 (BKK1) for full activation (Chinchilla et al., 2007, 2009; Heese et al., 2007; Roux et al., 2011). This leads to the initiation of phosphorylation-dependent signaling cascades, which result in the activation of defense responses including the production of reactive oxygen species and salicylate, expression of defense genes such as pathogenesis-related (PR) genes, production of antimicrobial compounds and hypersensitive cell death (Boller and Felix, 2009; Nicaise et al., 2009; Segonzac and Zipfel, 2011). These defense responses are accompanied by a growth arrest of the plants indicating that reprogramming of growth is part of the stress response. In contrast, necrotrophic pathogens are restricted by the activation of JA signaling, resulting in the activation of the plant defensin PDF1.2 (Kazan and Manners, 2008; Browse, 2009). Activation of JA signaling also leads to a growth arrest which can be explained by the influence of JA on auxin biosynthesis and signaling (Kazan and Manners, 2012).
Peptide hormones have multiple functions in cell–cell communication in animals. Recently, several studies have identified a number of new plant peptide hormones (for review see Matsubayashi and Sakagami, 2006; Matsubayashi, 2011). Most of these are involved in developmental processes, although systemin, a peptide produced by Solanaceae, is a messenger involved in wound responses and plant defense to herbivores (Pearce et al., 1991; McGurl et al., 1992; Matsubayashi and Sakagami, 2006; Matsubayashi, 2011). Recently, AtPep1 was described as an endogenous peptide ligand for two LRR-RLKs that is necessary for plant defense against fungi and can induce resistance to bacteria (Huffaker et al., 2006; Yamaguchi et al., 2006, 2010; Krol et al., 2010). Though not proven yet, perception of AtPep1 by the receptors PEPR1 and PEPR2 has been proposed to positively regulate and potentially amplify PAMP-triggered immunity (PTI) in Arabidopsis by using cell-to-cell communication to alert unaffected cells to the attack (Huffaker and Ryan, 2007; Ryan et al., 2007).
Phytosulfokines (PSKs) were first identified as growth-promoting peptides in the supernatant of Asparagus cell cultures (Matsubayashi and Sakagami, 1996). The active 5-amino-acid bisulfated peptide is processed from an ~80-amino-acid secreted precursor protein by post-translational sulfation and proteolytic cleavage (Srivastava et al., 2008; Komori et al., 2009). Five paralogous PSK genes have been identified in the Arabidopsis genome which are predominantly expressed in mature tissues (Yang et al., 2001; Lorbiecke and Sauter, 2002). Addition of PSKα to cell cultures or calli has a growth-promoting effect. PSK also promotes somatic embryogenesis (Hanai et al., 2000; Igasaki et al., 2003), tracheary element differentiation in vitro (Matsubayashi et al., 1999; Motose et al., 2009), pollen germination (Chen et al., 2000) and adventitious root formation (Yamakawa et al., 1998; Amano et al., 2007). A second sulfated peptide involved in cell proliferation is PSY1 (plant peptide containing sulfated tyrosine 1), an 18-amino-acid sulfated and glycosylated secreted peptide (Amano et al., 2007). Phytosulfokines are thought to maintain the cellular potential to proliferate and differentiate.
Perception of PSKα is dependent on the LRR-RLK PSKR1 and, to a lesser extent, its paralog PSKR2 (Matsubayashi et al., 2006). Loss-of-function mutants in PSKR1 have no obvious phenotypes. They are less sensitive to external application of PSKα, which results in decreased callus growth. A weak early senescence phenotype was observed in pskr1 mutants, while conversely PSKR1 overexpression resulted in delayed senescence indicating that PSK perception via PSKR1 is somehow involved in plant senescence (Matsubayashi et al., 2006). On the other hand, wound healing is affected in mutants lacking the PSK receptor, suggesting that loss of these receptors causes pleiotropic phenotypes related to SA- and JA-regulated processes (Matsubayashi et al., 2006). PSY1 is perceived by the PSY1 receptor (PSY1R), also a LRR-RLK related to PSKR1 and PSKR2. Triple mutants in all three RLKs result in moderate dwarfism caused by a reduction in cell number and size (Amano et al., 2007). Taken together, the members of this small LRR-RLK subfamily are partially redundant in controlling cell homeostasis after peptide perception.
Peptide sulfation in plants is mediated by a unique Golgi-membrane localized tyrosylprotein sulfotransferase (TPST) (Moore, 2009). Similar to the triple receptor mutants, tpst-1 mutants exhibit semi-dwarfism, early senescence and reduced cell proliferation. In addition, recombinant TPST has been shown to sulfate PSKα and PSY1 precursor peptides in vitro, indicating that TPST might be the enzyme necessary for post-translational sulfation of PSKα and PSY1 (Komori et al., 2009).
Recently, Igarashi et al. (2012) described the involvement of PSKR1 but not PSKR2 in PAMP responses. Mutants in PSKR1 show enhanced defense gene expression and seedling growth inhibition after PAMP treatment and are more resistant to bacterial infection with Pseudomonas syringae pv. tomato (Pto) DC3000. tpst mutant plants also exhibit enhanced PAMP-trigged seedling growth inhibition and defense gene expression which can be partially rescued by application of PSKα.
Using a reverse genetic approach we studied 49 pathogen-inducible LRR-RLK genes for their function in plant immunity to microbial infection (Postel et al., 2010). Here, we identified PSKR1 and additionally PSY1R as bifunctional receptors that exhibit roles in plant development and immunity. We report new insights into the antagonistic effect of PSKR1 and PSY1R on SA- and JA-mediated defense responses and characterize the receptors involved in this process.
Mutants in PSKR1 are antagonistically affected in their resistance to bacterial infection with Pto DC3000 and necrotrophic fungal infection with Alternaria brassicicola
The analysis of AtGenExpress microarray data (Postel et al., 2010) showed that PSKR1, but not the PSK receptor paralogs PSKR2 and PSY1R, is significantly altered in its expression by bacterial infection and PAMP treatments (Figure S1 in Supporting Information). PSK ligand transcripts are also induced by pathogens and PAMPs, especially PSK2 and PSK4 (Figure S2), as is the PSY1 gene (Figure S3).
Our analysis of two independent loss-of-function alleles, pskr1-2 and pskr1-3, showed enhanced resistance to bacterial infection with Pto DC3000 as also shown by Igarashi et al. (2012). Bacterial growth was reduced by up to two orders of magnitude, also resulting in reduced symptom development. Wild-type plants exhibited high levels of chlorosis and lesions, while pskr1-2 and pskr1-3 mutant plants remained mostly green (Figure S4a,b). Cell death, which usually occurs 3 days after spray infection with high inocula of Pto DC3000, is almost undetectable even at 5 days after infection (Figure S4c). Potentially due to differences in experimental conditions, the effects on bacterial resistance and in all the following assays are stronger than reported by Igarashi et al. (2012). These data show that loss of PSKR1 has a strong positive impact on resistance to bacterial biotrophic pathogens and that infection-induced symptom development is largely reduced in the mutants.
To further characterize the role of PSKR1 in defense responses to other plant pathogens, pskr1 mutants were inoculated with the necrotrophic fungus Alternaria brassicicola that shows an incompatible interaction with Arabidopsis Col-0 plants (Thomma et al., 1999). This resulted in higher disease indices, greater lesion sizes and stronger disease symptoms in pskr1 mutants compared with wild-type plants (Figure 1a,b). At the microscopic level, trypan blue staining showed that pskr1 mutant plants developed higher levels of cell death than wild-type plants in response to fungal infection (Figure 1c). Furthermore, while fungal growth stopped at the stage of germinated spores at the zone of inoculation in wild-type plants, this was not the case in pskr1-2 and pskr1-3 plants, where the mycelium grew and spread outside this zone, suggesting that loss of PSKR1 has a negative effect on fungal resistance.
Loss of PSY1R has similar effects on plant immunity as loss of PSKR1, and has additive effects in multiple mutant combinations with pskr1 mutants
The small subfamily of LRR-RLKs including PSKR1, PSKR2 and PSY1R was reported to function in a partially redundant manner in PSKα and PSY1-mediated cell proliferation (Amano et al., 2007). Since we found that PSKR1 has a strong impact on the resistance of plants to fungal pathogens, mutants in its close paralogs pskr2 and psy1r were subjected to fungal disease-rating experiments to determine if they have additional functions in defense responses.
While pskr2 mutants were not altered in fungal susceptibility, loss of PSY1R increased symptom formation, disease index, lesion size and fungal growth after infection with A. brassicicola (Figure 2). In combination, double pskr1-3/psy1r mutants cause a very high disease index, severe lesion formation and strong mycelial growth compared with wild-type plants (Figure 3). The pskr2 mutation had little effect on these A. brassicicola phenotypes in all mutant combinations tested (Figure S5). Since the double mutant pskr1-3/psy1r displayed the same level of symptoms as the triple mutant pskr1-3/pskr2/psy1r, the role of pskr2 in defense against A. brassicicola seems to be marginal.
In response to Pto DC3000, pskr2 plants did not show any significant difference in bacterial growth as reported by Igarashi et al. (2012). In contrast, psy1r plants are more resistant with about eight times less bacterial growth than wild-type plants (Figure 4a). Double and triple mutants of all three genes were tested in all combinations to assess the impact of each gene on bacterial resistance. Interestingly, pskr1-3/psy1r double-mutant plants displayed an additive resistance phenotype to Pto DC3000, with just over 100-fold less growth than in wild-type plants (Figure 4b). The triple mutant pskr1-3/pskr2/psy1r did not show any further differences from the pskr1-3/psy1r double mutant, confirming a limited role for PSKR2 in Pto DC3000 resistance (Figure S6). This demonstrates that PSY1R, in addition to PSKR1, has a role in plant defense against bacterial pathogens.
To show that the loss-of-function phenotypes in pskr1 mutants are indeed due to the loss of PSKR1 we expressed the coding region of PSKR1 fused to GFP under its native promoter in the pskr1/pskr2/psy1r mutant background. Expression of PSKR1 in the triple mutant background leads to a partial restoration of both the bacterial resistance and fungal susceptibility phenotypes. The partial character of the complementation might be a result of the construct used, but may also support the impact of the second receptor PSY1R which is not restored in these plants (Figure S7). Taken together, these results show that PSKR1 and PSY1R play a partially redundant role in plant immunity with antagonistic effects on bacterial and fungal resistance.
The imbalance in defense responses in pskr1 and pskr1/pskr2/psy1r mutants is reflected by elevated salicylate responses after infection with Pto DC3000
To further characterize the effect of pskr1 and pskr1/pskr2/psy1r mutations on defense responses at a molecular level, plants were treated with Pto DC3000 and SA levels were analyzed. In pskr1-3 mutant plants, SA levels are about 1.6-fold higher than in wild-type plants at 12 and 24 h post-inoculation. In the triple mutant pskr1/pskr2/psy1r plants, SA accumulation was delayed, but by 24 h post-inoculation the plants contained about two times more SA than wild-type plants (Figure 5a).
Real-time PCR-based gene expression analysis was performed after Pto DC3000 treatment to assess whether PR-gene expression was altered in the mutants. The single pskr1-3 and the triple pskr1/pskr2/psy1r mutant exhibited significantly higher induction of PR1 than in wild-type plants (three- and eight-fold, respectively) (Figure 5b). In the case of PR2, the induction was about three- and five-fold higher in pskr1-3 and pskr1/pskr2/psy1r, respectively (Figure 5c). Expression levels of FRK1, a marker for PAMP-induced gene expression, peaked at 12 h post-inoculation with an increased induction of about seven- and twelve-fold in pskr1-3 and pskr1/pskr2/psy1r mutants, respectively (Figure 5d). On the other hand, upon Pto DC3000 treatment, both pskr1-3 and pskr1/pskr2/psy1r plants expressed lower levels of PDF1.2, a marker gene for JA-dependent defense responses, as compared with the wild-type. In both mutant plants, expression of PDF1.2 was lower in uninoculated mutant plants (about 30- and 100-fold, respectively) and remained lower during Pto DC3000 infection (Figure 5e). In the case of OPR3, untreated pskr1-3 and pskr1/pskr2/psy1r plants also had fewer transcripts than the wild-type. Induction of OPR3 never reached wild-type levels in either mutant plant (about ten- and two-fold less, respectively) (Figure 5f). Taken together SA- and PAMP-responsive genes are transcriptionally up-regulated in pskr1 and more strongly in pskr1/pskr2/psy1r mutants, while JA-responsive genes are down-regulated: this observation may reflect a shift in the antagonistic hormone signaling pathways to the advantage of SA signaling.
Pathogen-associated molecular pattern responses are enhanced in pskr1, psy1r and pskr1/pskr2/psy1r triple mutants
Igarashi et al. (2012) reported enhanced defense gene activation and seedling growth inhibition in pskr1 and tpst mutants upon elf18 treatment and to a lesser extent after flg22 treatment. Based on microarray data, not only PSKR1 and PSK but also PSY1 transcription is specifically up-regulated by PAMPs (Figure S1–S3). To address whether mutations in PSY1R also alter PAMP responses, plants were treated with 100 nm flg22 and callose deposition was visualized by aniline blue staining (Figure 6a,b). Microscopically, it is obvious that the receptor mutants produce an increased amount of callose after PAMP treatment compared with wild-type plants. Quantification reveals that psy1r plants have the same enhanced amount of callose stained as pskr1 mutants (both about 1.7-fold) compared with wild-type plants indicating that both PSKR1 and PSY1R receptors suppress PAMP-induced callose formation. The triple mutant pskr1/pskr2/psy1r reflects the additive effects of both receptors with 2.4-fold more callose being detected. To further characterize alterations in PAMP responses, we studied seedling growth inhibition, which usually accompanies PAMP responses and FRK1 expression. After flg22 treatment, all three mutants showed a strongly enhanced response compared with wild-type at the 10 and 100 nm levels, with slightly less responsiveness in psy1r mutants reflected by only insignificant changes at the 1 nm level (Figure 6c). Adult leaves were treated with 100 nm flg22 for 4 h and the induction of FRK1 was measured by quantitative (q) RT-PCR. Mutants in pskr1-3, psy1r and pskr1/pskr2/psy1r all displayed an enhanced induction of FRK1 compared with the wild-type upon treatment, suggesting that PSK and PSY1 signaling is required for down-regulating PAMP responses (Figure 6d). Cumulatively, our data show that PAMP responses are suppressed by PSKR1 and PSY1R signaling: this suggests that PSKα and probably PSY1 signaling act in a negative regulatory loop to prevent over-responsiveness to PAMPs.
Tyrosine-sulfated peptides are necessary for balancing defense signaling
In animals and plants, peptide tyrosine sulfation is mediated by a unique membrane localized TPST (Moore, 2009). Loss-of-function mutants in TPST mimic the triple receptor mutant phenotypes and have additional phenotypes comparable to plants lacking sulfated peptides or their corresponding receptors (Komori et al., 2009; Zhou et al., 2010; Stuhrwohldt et al., 2011). Mutants in TPST show a strong bacterial resistance phenotype after infection with Pto DC3000, phenocopying the effects observed in the triple receptor mutants (Figure 7a). Enhanced PAMP-induced callose formation is found in tpst-1 mutants as well as seedling growth inhibition and FRK1 gene expression as observed by Igarashi et al. (2012) (Figure S8). Furthermore, susceptibility to A. brassicicola was strongly enhanced and indistinguishable from that of the triple receptor mutant (Figure 7b–d). The effect on fungal susceptibility indicates that the defense-related phenotypes of tpst mutants are not solely due to the effect on PAMP responses reported by Igarashi et al. (2012). Taken together, these data show that peptide tyrosine sulfation is necessary for balancing bacterial susceptibility and fungal resistance, two antagonistically regulated defense pathways.
Ectopic expression of PSK2, PSK4 and PSKR1 enhances susceptibility to Pto DC3000 and increases resistance to Alternaria brassicicola
The previous results indicate that PSKR1 and PSY1R receptors are necessary for balancing defense responses, as is peptide tyrosine sulfation. To prove that these phenomena are indeed due to PSK signaling we overexpressed PSK2, PSK4 and PSKR1-encoding regions in Arabidopsis and tested these lines in bacterial and fungal infection assays. Indeed, expression of all three proteins leads to opposite effects in infection assays than observed in pskr1 and psy1r mutants. Upon treatment with Pto DC3000, PSK2-, PSK4- and PSKR1-overexpressing plants display a susceptibility phenotype, allowing approximately 10-fold more bacterial growth than in wild-type plants (Figure 8a). On the other hand, treatment with A. brassicicola revealed an enhanced resistance phenotype in the overexpressing plants (Figure 8b,c). Taken together, these gain-of-function experiments support that PSK signaling has a strong effect on both defense responses to biotrophic and necrotrophic pathogens in addition to the reported effect of PSK signaling on PAMP responses (Igarashi et al., 2012).
PSKα has an immediate effect on plant immunity
To determine if secondary effects such as growth or morphological differences in the mutants might be the reason for the immunity-related phenotypes, we tested if exogenous PSKα application can be directly tied to defense responses. Application of PSKα to wild-type plants slightly promotes root growth (Figure S9) but has only weak effects on bacterial resistance (Figure 9) and on seedling growth inhibition after elf18 treatment (Igarashi et al., 2012). In contrast, tpst-1 plants were partially rescued from the bacterial resistance phenotype by PSKαtreatment (Figure 9). tpst-1 plants treated with inactive non-sulfated PSK allowed about 25-fold less bacterial growth than control- treated wild-type plants. However, upon treatment with active PSKα, bacteria grew about 30-fold more in tpst-1 plants and only nine-fold less compared with wild-type plants. The pskr1/pskr2/psy1r triple mutants were completely insensitive to PSKα treatment, as expected. These data show that the defense-related phenotypes are due to direct effects of PSKα signaling rather than due to secondary effects from developmental differences in the mutants.
Here, we have shown that PSKR1 influences defense responses to biotrophic and necrotrophic pathogens in an antagonistic manner. Bacterial resistance is strongly increased in pskr1 mutants while non-host resistance to the necrotrophic fungus Alternaria brassicicola is impaired. This indicates that the regulation of defense responses is differentially altered in pskr1 mutants. Depending on the type of invader, a particular subset of defense responses is activated, such as SA- and JA-mediated signaling pathways, to specifically fend off specific classes of pathogens. Although the concerted actions of both of these pathways have been reported, a number of cases have illustrated that impairment in defense responses to biotrophs and/or SA signaling leads to increased necrotrophic resistance, and vice versa (Glazebrook, 2005; Jones and Dangl, 2006; Koornneef and Pieterse, 2008; Spoel and Dong, 2008; Vlot et al., 2009; von Saint Paul et al., 2011). Biotrophic pathogens are restricted by SA-mediated responses resulting in enhanced PR gene expression and hypersensitive cell death.
Defense against necrotrophic organisms is regulated by activation of a JA-dependent signaling pathway which is accompanied by the activation of marker genes such as the plant defensin PDF1.2 (Kazan and Manners, 2008). Wounding responses are also regulated by this pathway (Reymond et al., 2000) while senescence processes share some molecular characteristics with SA-dependent defense programs (Robatzek and Somssich, 2002; Buchanan-Wollaston et al., 2005; Spoel and Dong, 2008). In pskr1 mutants the homeostasis between SA- and JA-dependent defense pathways is shifted toward SA responses. Levels of SA are enhanced in pskr1 single and pskr1/pskr2/psy1r triple mutants after infection with Pto DC3000. Consequently, SA-responsive PR gene expression is also enhanced in the single and triple mutants, confirming that the SA signaling pathway is more strongly activated in the mutants than in wild-type plants. In Zinnia, addition of PSKα represses defense-related genes such as PR1, chitinases and several genes involved in SA biosynthesis (Motose et al., 2009). The authors conclude that PSK signaling might attenuate stress responses by suppression of SA signaling in Zinnia.
In this study we show that alterations in SA responses correlate with an enhanced resistance to the biotrophic pathogen Pto DC3000 and, antagonistically, loss of resistance to the necrotrophic pathogen A. brassicicola, and suggest that the PSKα receptor is needed for negative regulation of SA-dependent defense responses. The JA content is not significantly altered in pskr1 single or pskr1/pskr2/psy1r triple mutants. This is consistent with other mutants showing altered JA responses, but an unchanged hormone content (Cui et al., 2010). However, JA-responsive gene expression (PDF1.2 and OPR3) is down-regulated in pskr1 and the triple receptor mutants after Pto DC3000 treatment, supporting a model where PSKα perception causes a shift in defense homeostasis toward JA responses. This is further supported by reports of impaired wound healing in pskr1 mutants (Matsubayashi et al., 2006; Loivamaki et al., 2010). Several publications have reported wound responsiveness of PSKR1 and, depending on the assay conditions, for PSK3, -4 and -5 (Matsubayashi et al., 2006; Kilian et al., 2007; Loivamaki et al., 2010). In addition, early senescence was described for pskr1 mutants and confirmed by delayed senescence in PSKR1-overexpressing plants. Senescent leaves express high levels of the SA-regulated PR1 gene (Robatzek and Somssich, 2001). Consistent with this, developmental senescence is at least partially regulated by SA-dependent signaling pathways (Buchanan-Wollaston et al., 2005) supporting the idea that the reported wound- and senescence-associated phenotypes correlate with a shift in SA–JA homeostasis caused by the loss of PSKα perception and function.
Non-host and non-pathogenic bacteria which are recognized by plants, depending on their repertoire of PAMPs, as well as direct exposure to PAMPs, induce the expression of PSKR1 and some of the PSK peptide-encoding genes (PSK1, -2 and -4). In pskr1, psy1r, pskr1/pskr2/psy1r and tpst-1 mutant plants, responses to flg22 such as callose deposition, FRK1 expression and seedling growth inhibition are enhanced. Similar results were recently obtained by Igarashi et al. (2012). The differences observed on early flg22 responses in pskr1 mutants are most likely a matter of the different tissue used (seedlings versus adult). In our hands generally stronger responses were found in adult plants and in the triple mutants. FRK1 expression and callose deposition upon flg22 treatment is strongly affected in these tissues.
A tight negative regulation of PAMP responses might be necessary to prevent over-induction of PAMP responses. The negative effects of overly strong PAMP-triggered immunity are evident from mammals, where overstimulation of the innate immune system leads to fatal sepsis (Underhill and Ozinsky, 2002; Hotchkiss and Karl, 2003). Interestingly, PAMP treatment has been shown to be capable of causing accumulation of SA (Tsuda et al., 2008). However, PAMP responses are also influenced by SA signaling (Tsuda et al., 2009). Therefore, experiments addressing whether the enhanced PAMP responses in PSK receptor mutants are the cause or the consequence of the enhanced SA levels will be required in the future. Balancing the antagonistic SA- and JA-dependent defense pathways might be an explanation for the SA-repressive effect of PSK perception. A consequence of the antagonism in defense responses is the fact that enhancing resistance to biotrophic pathogens renders the plant more susceptible to necrotrophic attack. An over-amplification of one pathway leaves the plant unprotected against pathogens defended by the other (Spoel and Dong, 2008). Therefore, activation of the PSK signaling pathway by PAMPs might be necessary for maintaining a balanced defense arsenal against all pathogens after an initial boost of PTI by PAMPs. Recently, other hormone influences were discussed in the context of plant defense (Robert-Seilaniantz et al., 2007, 2011a). Auxin was reported to suppress SA-dependent defense responses such as PR1 expression and SA biosynthesis, while gibberellic acid has antagonistic effects (Park et al., 2007; Navarro et al., 2008; Robert-Seilaniantz et al., 2011b). Consistent with the antagonistic relationship between SA and auxin signaling, auxin signaling mutants are more susceptible to the necrotrophic fungi Plectosphaerella cucumerina and Botrytis cinerea (Llorente et al., 2008). Auxin is involved in all areas of plant growth and development (Benjamins and Scheres, 2008), it can induce TPST expression (Zhou et al., 2010) and seems to be required for PSK-induced growth promotion in carrot cells (Eun et al., 2003). Therefore, the effects of auxin might be the cause of the growth-promoting effects of PSKs and the here described SA/JA hormone imbalance in pskr mutants. Alternatively, Motose et al. (2009) report that PSKα suppresses expression of stress-responsive genes even in the absence of auxin or cytokinin in Zinnia and conclude that both hormones are not necessary for PSK-dependent stress signaling. Further analysis of hormone contents and the interplay of their signaling pathways is needed to fully understand the influence of PSKs on hormone homeostasis.
PSKα was first described as a secreted conditioning factor in Asparagus cell cultures with growth-promoting properties (Matsubayashi and Sakagami, 1996). In Arabidopsis roots, the growth-promoting effect of PSKα leads to enhanced cell numbers and sizes (Amano et al., 2007). In Zinnia, PSKs were described to be involved in the attenuation of stress responses during the transdifferentiation of mesophyll cells into tracheary elements (Motose et al., 2009). The PSK precursor genes are preferentially expressed in differentiated tissue and therefore most likely have a role in repair processes rather than in developmental or differentiation programs (Matsubayashi et al., 2006). Cellular plasticity is extremely important in plants, and PSKs may contribute to developmental stability under different environmental conditions. After pathogen attack, PSK1, -2 and -4, PSY1 and PSKR1 gene expression is activated, and the respective gene products might influence both growth and responses to pathogens and wounding. Therefore, PSK signaling might be necessary for restricting energy-consuming defense responses for the benefit of senescence prevention and growth.
The PSK receptor PSKR1 has a paralog PSKR2 and a close relative PSY1R that perceives the tyrosine-sulfated peptide PSY1. Triple mutants in all three RLKs exhibit a semi-dwarf growth phenotype and are completely insensitive to PSKα (Amano et al., 2007). In this study we show that in addition to PSKR1, PSY1R has an additional function in plant defense. Double mutant analyses reveal an additive effect of PSY1R mutation on bacterial resistance and fungal susceptibility. Mutants lacking PSKR2 still respond to PSKα treatment in root growth assays, indicating that PSKR2 has only a marginal impact on PSKα-mediated growth promotion (Amano et al., 2007) and therefore its contribution might not be sufficiently detectable in plant defense responses. By contrast, PSY1R has a strong impact on plant immunity. Amano et al. (2007) report no involvement of PSY1R in PSKα perception, no binding of PSKα to PSY1R and that both peptides redundantly contribute to similar developmental processes such as cellular proliferation, expansion and wound repair (Amano et al., 2007). Deduced from the mutant analysis and the partial character of the triple mutant and tpst complementation experiments, both peptides most likely also contribute redundantly to the plant defense phenotypes. Such a redundancy might be caused by duplication events within the large receptor kinase family and also within the ligand gene family which contains five PSK and three PSY1 paralogous genes (Yang et al., 2001; Lorbiecke and Sauter, 2002; Amano et al., 2007). Diversification after duplication, which is evident from the distinct expression patterns of these genes, might explain the varying involvement of the individual proteins and is the basis of stability of the plant species when exposed to constantly changing environmental conditions.
From the phenotypes of receptor mutants, it can be deduced that loss of PSKR1 and PSY1R has an impact on plant defense homeostasis. We asked if this is indeed a result of lost PSK perception or a defect in the receptor mutants independent of PSK signaling. The tpst-1 mutants, which are impaired in the sulfotransferase that can sulfate PSKα and PSY1 (Komori et al., 2009), phenocopy the defense responses associated with the receptor triple mutant, suggesting that probably no further sulfated peptides are involved in this effect. Exogenous application of PSKα has only weak effects on defense responses in wild-type plants as described for other phenotypes (Motose et al., 2009; Stuhrwohldt et al., 2011; Igarashi et al., 2012). Delivery of exogenous peptides might be suboptimal as PSK2 and PSK4 overexpressing plants show a strong impact on bacterial and fungal resistance. In tpst-1 mutants, which are likely to be deficient in active PSKα and PSY1, a clear but partial complementation of the bacterial defense phenotype can be detected upon exogenous application of PSKα. The triple receptor mutant is insensitive to PSKα and is not affected by PSKα application, ruling out that PSKα has a direct effect on bacterial growth. This shows that PSKα indeed has a direct effect on plant defense responses. The partial character of this complementation leaves room for the speculation that the PSY1 peptide has additional redundant functions as proposed from the receptor analysis. It also shows that the defense-related tpst-1 and receptor mutant phenotypes are not due to developmental differences compared with wild-type but are caused by the direct and acute influence of the peptide on plant defense.
A delicate balance between the SA- and JA-mediated defense signaling pathways exists in plants, as a large number of studies have demonstrated (Glazebrook, 2005; Jones and Dangl, 2006; Koornneef and Pieterse, 2008; Spoel and Dong, 2008; Vlot et al., 2009; von Saint Paul et al., 2011). This fact is emphasized by the numerous genes that have been implicated in the interplay between these two important signaling cascades. In this report we have demonstrated that both PSKR1 and PSY1R are not only necessary for cell proliferation but also for modulation of SA- and JA-dependent defense responses. These results suggest a mechanism whereby endogenous sulfated peptides are required to keep a healthy balance between growth and two (or more) divergent modes of defense. Further studies of PSKR1 and PSY1R signaling as modulators of growth, defense responses and hormone homeostasis will expand our understanding of mechanisms used by plants to adapt to constantly changing environmental challenges.
Plant material and growth conditions
The T-DNA insertion mutants used are pskr1-2 (SAIL_673_H07), pskr1-3 (SALK_008585), pskr1-5 (FLAG_407D02), pskr2 (SALK_024464), psy1r (SALK_072802C) and tpst-1 (SALK_009847), the latter kindly provided by Y. Matsubayashi (National Institute for Basic Biology, Okazaki, Japan). Double and triple mutants were crossed and selected for homozygous offspring by PCR. Plants were grown for 5 to 6 weeks on soil in a growth chamber (8 h light, 16 h dark, 22°C, 110 mE m−2 sec−1).
A 2.4-kb promoter region of PSKR1 plus the coding region was amplified from Col-0 genomic DNA, cloned into pCR8-/GW-TOPO and recombined into pBIB-BASTA-GFP (Gou et al., 2010). To generate transgenic Arabidopsis lines harboring the constructs p35S-AtPSK2, p35S-AtPSK4 (in pK2GW7) and p35S-AtPSKR1 (in pH2GW7), entire coding sequences were amplified by PCR using sequence-specific primers (Table S1) and recombined into the respective Gateway vectors described by Karimi et al. (2002). Constructs were transformed into wild-type or pskr1-3/pskr2/psy1r mutants via Agrobacterium tumefaciens strain GV3101-mediated transformation (Clough and Bent, 1998).
Pseudomonas syringae pv. tomato DC3000 was grown for 16 h in King's B medium at 28°C, and placed in 10 mm MgCl2 at a concentration of 108 colony-forming units (cfu) ml−1 for transcript and SA determination and 104 cfu ml−1 for measuring bacterial growth as described by Zipfel et al. (2004).
Alternaria brassicicola infection assays were carried out as described by Kemmerling et al. (2007).
Dead cells and fungal mycelium in infected tissue were stained with trypan blue in lactophenol and ethanol as described by Kemmerling et al. (2007). Leaves were analyzed by light microscopy. To visualize callose deposition, plants were stained with aniline blue (Gomez-Gomez et al., 1999) and visualized using UV-epifluorescence microscopy. Signal intensities were measured using ImageJ software by measuring the mean signal intensity from 10 replicate fields of 2500 square pixels per data point.
Seedling growth inhibition assays
Surface sterilized seedlings were sown on 1/2 MS medium containing 1% agar, incubated for 5 days in the dark at 4°C and allowed to germinate for 5 days at 22°C. Seedlings were transferred to 1/2 MS liquid medium containing 1% sucrose in sterile 24-well plates. Seedlings were treated with 0.1 m NaCl and 0.1% bovine serum albumin with and without flg22.
To analyze transcript levels, total RNA was extracted from leaves using TRIzol followed by DNase I treatment and cDNA synthesis using 1μg of RNA, oligo dT primers and Moloney murine leukemia virus reverse transcriptase. Quantitative real-time PCR reactions were set up using SYBR green and a Bio-Rad iQ5-cycler (http://www.bio-rad.com/). Primer sequences are listed in Table S1. Gene expression values were normalized to EF1α and presented as a ratio to wild-type expression levels using the mathematical model described by Pfaffl (2001).
Salicylate and jasmonate contents were measured as described by Lenz et al. (2011).
We thank Natalia Rodiuc for transgenic lines and Delphine Chinchilla and Thorsten Nürnberger for critical discussions and comments on the manuscript. We are grateful for funding by the DFG (NU70/7-1 and KE1485/1-1).