|Home | About | Journals | Submit | Contact Us | Français|
NPR1 is a gene of Arabidopsis thaliana required for the perception of salicylic acid. This perception triggers a defense response and negatively regulates the perception of jasmonates. Surprisingly, the application of methyl jasmonate also induces resistance, and NPR1 is also suspected to be relevant. Since an allelic series of npr1 was recently described, the behavior of these alleles was tested in response to methyl jasmonate.
The response to methyl jasmonate of different npr1s alleles and NPR1 paralogs null mutants was measured by the growth of a pathogen. We have also tested the subcellular localization of some npr1s, along with the protein-protein interactions that can be measured in yeast. The localization of the protein in npr1 alleles does not affect the response to methyl jasmonate. In fact, NPR1 is not required. The genes that are required in a redundant fashion are the BOPs. The BOPs are paralogs of NPR1, and they physically interact with the TGA family of transcription factors.
Some npr1 alleles have a phenotype in this response likely because they are affecting the interaction between BOPs and TGAs, and these two families of proteins are responsible for the resistance induced by methyl jasmonate in wild type plants.
Plants are constantly defending themselves against pathogens by means of a wide array of mechanisms. Some of them are pre-existing (or non inducible) and others are induced in response to the pathogen attack. Salicylic acid (SA, reviewed by ) is a plant hormone which is crucial for the inducible response of Arabidopsis thaliana (Arabidopsis) to biotrophic pathogens like Pseudomonas spp. When a pathogen is perceived, SA is produced and accumulated, producing a proper defense. This SA signaling occurs not only where the attack takes place, since defense is also enhanced in leaves different from the one inoculated. This is called Systemic Acquired Resistance (SAR, ). SA has an intricate crosstalk with other hormones, showing an overall negative crosstalk with auxins, ethylene (ET), and jasmonates (JA, crosstalk of hormones reviewed by ). In the case of JA, it has been shown that the active form in planta is JA-Ile (reviewed by ), while in the laboratory is used exogenously as Methyl Jasmonate (MeJA).
NON-EXPRESSOR OF PATHOGENESIS-RELATED GENES1 (NPR1) is the main gene required for SA perception . There are five paralogs of NPR1 in Arabidopsis , BLADE-ON-PETIOLE1 (BOP1) and BOP2 have an important role in development , NPR3 and NPR4 have a role in defense , probably through SA perception , and no specific function for NPR2 has been described, besides a secondary role in SA perception . There are other genes that are relevant for signal transduction, like the family of TGA transcription factors whose products interact with NPR1 , but they are required in a redundant fashion.
NPR1 has been described as having more than one role in defense, since it is also important in the Induced Systemic Resistance (ISR, ). ISR is defined as the resistance triggered at the leaves by a non pathogenic organism inoculated in the roots, and while SAR requires SA signaling, ISR requires MeJA and ET signaling. As with SA, exogenous applications of MeJA and ET trigger resistance in Arabidopsis towards some biotrophs, like Pseudomonas. It has been proposed that NPR1 is relevant for the resistance-inducing ability of MeJA (, hereafter abbreviated as RIM), although RIM it is not necessarily equivalent to ISR. While the role of NPR1 in SA perception takes place in the nucleus , its function in RIM is not so clearly understood. It has been described a cytosolic function of NPR1 crucial in cross-talk between SA and JA signaling . Furthermore, Arabidopsis transcriptome analysis upon pathogen infection has suggested that such cytosolic function is also involved in the modulation of JA-dependent defenses . The npr1-3 mutant, which produces a truncated cytoplasmatically localized npr1 protein with no nuclear localization signal, has been reported to be affected only in SA-dependent gene expression, not in JA and ET dependent genes. In contrast, the npr1-1 mutant, which has a mutation in a key domain, is affected in the expression of SA, JA and ET-dependent genes . More recent studies support such cytosolic NPR1 function as regulator of JA-dependent defense responses ([18-20]).
BOP1 was first identified by its mutant phenotype of ectopic blades along the petioles, as well as some alterations in the flowers . The first allele identified was a dominant negative, since T-DNA insertions in bop1 did not reproduce the phenotypes of ectopic blades . Once BOP1 was identified as paralog to NPR1, it was shown that other paralog, BOP2, functions redundantly with BOP1. The double mutant bop1 bop2 reproduced all the developmental phenotypes of ectopic blades, but it was wild type when inoculated with Pseudomonas, and it is also wild type for SA perception .
Since a collection of npr1 alleles has recently been available , we tested the hypothesis that the role of NPR1 in RIM is cytosolic. In this work, we show that NPR1 has no role in RIM in wild type conditions, since the genes responsible for RIM are BOP1 and BOP2, with an important part being played by the TGAs. Therefore, two genes required for the normal development of the leaf, are also required for plant defense.
NPR1 has been characterized as a result of observing the response to SA of the great number of alleles described for it . NPR1 has also been described as essential for RIM, but there are differences between alleles, since npr1-1 and npr1-3 have different RIM ([20,25]). npr1-1 and npr1-3 have other differences in phenotypes related to MeJA. Thus, the SA-JA antagonism is not present in npr1-1, but it is active in npr1-3 . Other difference is the gene expression, whereas npr1-1 was affected in SA, JA, and ET dependent genes upon Pto inoculation, npr1-3 was only affected in SA dependent genes . These different phenotypes have been attributed to the lack of nuclear localization in npr1-3, since the truncated cytosolic protein would be functional to modulated JA-dependent defense response .
In order to determine the precise role of NPR1 in RIM, and to asses the role of the cytosolic vs. nuclear localization, we tested an allelic series of 43 npr1 alleles, defined by their inability to perceive SA . The npr1 alleles produced a mixture of phenotypes in RIM (Figure (Figure1),1), and out of the 43 alleles tested, 11 had a significant RIM (RIM+, Figure Figure1a),1a), meaning that the growth of Pto in MeJA treated plants was significantly lower than in mock treated plants (Student’s test of one tail, with P<0.05). The rest of alleles showed no response (RIM-, 11 out of 32 are shown in Figure Figure1b).1b). These two categories of alleles did not share any obvious feature and had the same proportion of stops and point mutations in each category. In fact, in the RIM+ alleles tested, six point mutations were widely scattered along the protein (Figure (Figure1c,1c, npr1-3 is included as reference).
These alleles could somehow affect the localization of the protein inside the cell, even if the mutation was not in the NLS. To check this possibility, cDNAs of three RIM+ alleles (npr1-22, -35 and −44, Additional file 1), and three RIM- alleles (npr1-1, -40, and −56, Additional file 1), chosen among the point mutations, were cloned along with the wild type NPR1. Then, these seven cDNAs were transiently expressed in N. benthamiana under the control of the 35S promoter and with the GFP marker. The free GFP was detectable in the nucleus and in the cytoplasm. But when the wild type and the six alleles of npr1 were expressed in the same conditions, GFP was detected mainly in the nucleus, with no difference existing between the two classes of alleles (Figure (Figure22a).
As a complementary approach, we took advantage of the transgenic line that overexpresses NPR1 fused to the steroid hormone binding domain of the rat glucocorticoid receptor (HBD, and the transgenic plants are known as NPR1-HBD, ). NPR1-HBD remains exclusively in the cytosol in mock conditions and should be functional in RIM. The original line is in an npr1-3 background (RIM+), and therefore the transgene was transferred to an npr1-1 background (RIM-) to check for complementation. Treatments with BTH and with and without glucocorticoid dexamethasone (DEX) showed that NPR1-HBD was functional (Additional file 1). NPR1-HBD, even under the control of the 35S promoter, did not complement the lack of RIM in npr1-1 (Figure (Figure2b).2b). When DEX was applied, NPR1-HBD moved to the nucleus and npr1-1 was complemented in the RIM phenotype. Note that the presence of cytosolic NPR1-HBD in an npr1-3 background did not enhance RIM in comparison to npr1-3 alone.
The npr1 alleles RIM+ or RIM- did not share any obvious feature, so it would be difficult to assign a precise role to the wild type gene. A critical genetic resource to discern the role of a gene is the null allele. Therefore, the response of two null alleles of npr1 to RIM was measured (Figure (Figure3a).3a). Both npr1-70 and npr1-71 are in Laer-0 background, so an introgression of npr1-1 in Laer-0 was used as control . These two null alleles responded to RIM in all experiments like the wild type. npr1-70 introgressed in Col-0 responded again like the wild type, which ruled out any ecotype effect (Figure (Figure3b).3b). Since the direct role of NPR1 in RIM was in question as a consequence of the aforementioned results, we included an independent RIM- control, coi1-40 (see Methods).
The role of NPR1 in this response might be indirect. Thus, one scenario would be a reinforcement of the negative crosstalking between SA and MeJA. npr1 alleles produced more SA when infected with Pto and seemed unable to metabolize it . RIM- alleles -defective in terms of SA perception- might have left intact the negative crosstalk between SA and MeJA, and an excess of SA repressed the response to MeJA beyond the wild type levels. Therefore, the RIM+ alleles would be defective in terms of both SA perception and SA-MeJA crosstalking, an explanation that would also be in agreement with the behavior of the null alleles.
To test this hypothesis, the double mutants between npr1-1 and NahG (a transgenic plant that degrades SA, ), eds5 (a mutant in SA transport, ), and sid2 (a mutant in SA biosynthesis, ), were constructed and tested for RIM. npr1-1 did not respond to MeJA even if the levels of SA were low (Figure (Figure3c),3c), so the hypothesis of a reinforcement of the negative crosstalk was not supported.
The experiments with the null alleles showed that NPR1 was not necessary for RIM. Perhaps NPR1 and other gene(s) would be redundantly responsible of RIM, and while null npr1 alleles would have a RIM+ phenotype, some npr1 alleles could be RIM- by interacting with other protein(s) negatively. The most likely candidates for these interactions were the NPR1 paralogs, since their proteins shared the same overall structure. There are five additional paralogs of NPR1 in the Arabidopsis genome , and we analyzed double mutants of npr1-70 with different paralogs (Figure (Figure4).4). There was no proof of NPR1 having a redundant role in RIM. NPR2 did not play a significant role in RIM (Figure (Figure4a)4a) and the same was true for the double npr3 npr4 (Figure (Figure4b,c4b,c and ). Strikingly, the double bop1 bop2 was RIM- (Figure (Figure44d).
Single bop1 and bop2 were also tested for RIM and showed to be wild type (Figure (Figure5a).5a). BOP1 and BOP2 exert their function in part through transcriptional regulation of KNAT6 and physical interaction with PAN (TGA8, ), but T-DNA insertions predicted to disrupt KNAT6 or PAN activity did not have an effect on RIM (Figure (Figure5a).5a). To rule out the possibility that other mutations besides bop1 and bop2 were producing this RIM- phenotype, we constructed an artificial microRNA (amiRNA, ) to deplete the levels of BOP1 and BOP2 at the same time. Eight independent homozygous transgenic lines for amiRNA (BOP1 - BOP2) were analyzed for RIM (Figure (Figure5b).5b). Five out of eight lines were RIM-, and the remaining three responded less than the wild type control. The levels of both genes were partially depleted in the eight lines (Figure (Figure5c);5c); five of the lines had both genes significantly reduced, and all had BOP2 significantly reduced. None of these lines had the characteristic blade-on-petiole macroscopic phenotype, not even as the subtle phenotype of bop1 alone.
The previous experiments had shown that BOP1 and BOP2 were acting redundantly in RIM (Figure (Figure4d).4d). Therefore, increasing the amount of any of them should have an effect on RIM, especially since normal levels of BOP1 and BOP2 are quite low (Additional file 1, ). The overexpression lines of BOP1 and BOP2 described  were analyzed for RIM. 35S:BOP1 had a stronger RIM than Col-0, and 35S:BOP2 had a strong variation in the MeJA treated plants (Figure (Figure6a).6a). At the time of the experiments, each population looked homogeneous, but when these plants were grown to set seeds, two phenotypes could be observed in each transgenic line. Approximately half of the plants showed a wild type phenotype, and the other half reproduced the dwarf plants described . Seeds from both lines and from both phenotypes reproduced the two phenotypes. It seems that it was an issue of silencing, since RNA taken from plants classified by their mentioned phenotype diverged widely in the transgene expression (Figure (Figure66b).
The bop1 bop2 double mutant lacked a RIM response, but this failure in MeJA signaling might occur at different points of the signal transduction. For example, the defect could target a general signaling component affecting all MeJA responses (e.g. coi1,), or a specialized part of the pathway, affecting a subset of MeJA responses (e.g. jin1,). When bop1 bop2 plants were grown in plates containing MeJA, the growth of the roots was similar to the wild type controls (Figure (Figure7a).7a). Other phenotypes of bop1 bop2 plants growing in MeJA plates were similar to the wild type controls (carotenoids production, size of aerial part, number of trichomes, etc.; data not shown). Another effect of MeJA is the increase of senescence in detached leaves, measured as chlorophyll production . bop1 bop2 responded as the wild type in this particular system (Figure (Figure7b).7b). Coronatine is a virulence factor of several Pseudomonas isolates with structural and functional similarities to JA-Ile (the functional form of Jasmonate in planta, ), therefore a mutant Pto that lacks coronatine  grows less in Col-0 than the wild type Pto. bop1 bop2 was also wild type in response to Pto with and without coronatine (Figure (Figure7c).7c). Inoculations with Plectospharaella cucumerina, a fungus that causes more disease in MeJA mutants than in wild type plants , did not cause any more disease in bop1 bop2 than in Col-0 (data not shown). If bop1 bop2 was not a MeJA signaling mutant, but specifically in RIM, it could be defective in the other signaling required for ISR; ET. It was not; when 1 mM of 1-aminocyclopropane-1-carboxylic acid (ACC, an ET precursor) was sprayed to bop1 bop2, the resistance triggered was similar to that triggered in the wild type controls (Figure (Figure7d).7d). etr1 (ETHYLENE RESPONSE 1,) was included as a negative control of resistance induced by ethylene.
Once it was clear that both BOP genes are required for RIM, we tested the model that npr1 RIM- alleles could have a dominant negative effect on BOP activity, either directly or indirectly. To first test whether NPR1 had an effect in the interaction between BOP proteins, we used a yeast two-hybrid assay. As reported, BOP1 and BOP2 interacted with each other . Next, we introduced in a third plasmid containing wt NPR1 or various mutant npr1 alleles presented in Figure Figure2a.2a. If the effect of npr1 on RIM were a direct interaction between NPR1 and the two BOPs, the alleles that diverge in their RIM phenotype would diverge in their ability to interfere in the interaction of BOP1 and BOP2. The two classes of npr1 alleles did not have a distinct behavior (Figure (Figure8a,8a, the first three npr1 alleles are RIM-, and the last three are RIM+), therefore the dominant negative effect did not seem to be direct.
All the NPR1 paralogs tested interact with members of the TGA family in a different degree [7,21]. Therefore, the TGAs would be a reasonable candidate for being the third component, and their interaction with RIM- alleles would indirectly affect the function of BOP1 and BOP2. As a control, single mutants in TGA1 and TGA7 produced a significant RIM (Figure (Figure8b),8b), but when three specific tgas are knocked out at the same time (a triple which phenocopies an npr1 mutant in SA response, ), there is no RIM (Figure (Figure88b).
We reasoned that one or several of these three TGAs (TGA2, 5, and 6) might have a functional interaction with the BOPs, which might be affected by the RIM- alleles. To test this hypothesis BOP1 and each of the mentioned TGAs were introduced in the yeast two-hybrid system with the npr1 alleles mentioned above in a third plasmid. TGA2 and TGA6 interact differentially with BOP1 depending on the npr1 protein present (Figure (Figure8c).8c). There was an enhancement of the interaction in two out of the three RIM- alleles, and no interference in two out of the three RIM+ alleles. The interaction TGA5-BOP1 was not affected by the presence of npr1 proteins (data not shown). The experiments were repeated with BOP2 producing similar results (data not shown). In sum, the data indicated that BOPs and TGA2, TGA5 and TGA6 are required for RIM, that BOPs interact with these (and other) TGAs, and that NPR1 may modulate the affinity or stability of the interactions.
NPR1 is an essential gene for SAR and SA perception . npr1-1, the most widely used allele, is also impaired in RIM . We speculated that since npr1-3 is wild type for RIM ([20,25]), and it has been reported that the difference of some phenotypes between npr1-1 and npr1-3 was due to the lack of NLS in npr1-3 ([17,19]), the same could be true for RIM. However, we show here that the nuclear localization of the alleles makes no difference. This conclusion is supported by multiple lines of evidence. First, the npr1 alleles with RIM+ are not structurally similar to npr1-3, since not all of them are affected in the NLS (Figure (Figure1c).1c). Even an allele with a point mutation in the NLS (npr1-22, Additional file 1) should be partially localized in the nucleus . Second, three RIM- and three RIM+ alleles do not differ in their nuclear localization or stability when transiently expressed in N. benthamiana (Figure (Figure2a,2a, Additional file 1). While these proteins are no longer functional, they respond to the signals of a wild type background by localizing in the nucleus. Third, when a functional NPR1 is anchored in the cytoplasm there is no complementation of the RIM- phenotype in an npr1-1 background (Figure (Figure2b),2b), nor there is an increase in RIM phenotype in an npr1-3 background. In fact, the application of DEX triggered an increased RIM in both backgrounds (discussed below).
But, most importantly, NPR1 is not required for RIM, since the null npr1 alleles are RIM+ regardless of the background (Figure (Figure3a,b).3a,b). We also discarded that NPR1 could be a part of RIM in a redundant fashion with its paralogs (Figure (Figure44).
An interesting alternative for the role of NPR1 in RIM would be an effect on the crosstalking between SA and MeJA. NPR1 has been described as a key point in the negative regulation between SA and MeJA. Thus, the RIM+ alleles could be defective in both SA perception and in SA-MeJA crosstalk, while the RIM- alleles would be defective only in SA perception but not in SA-MeJA crosstalk. The inoculation with Pto triggers an increase in the levels of SA, and in the case of the npr1 alleles, there is more SA than in the wild type . Although this hypothesis would explain the phenotype of the null alleles, it was rejected after the experiment of Figure Figure3c,3c, where a severe reduction of SA levels in a RIM- allele did not have any effect on the phenotype.
The redundant functions of BOP1 and BOP2 are essential for normal development. Previous work has shown that the double mutant has numerous defects in plant architecture including altered leaf morphology , changes in floral patterning , defects in the conversion of shoots to flowers  and loss of floral-organ abscission . The double mutant was tested for basal defense  and SA perception (Additional file 1) but no difference from wild type was found. We show here that both genes are also redundantly required in defense against pathogens triggered by MeJA. Interestingly, whereas significant loss of BOP activity is required to exert changes in development , RIM is abolished in plants that are only partially silenced for the BOP genes (Figure (Figure5b,c).5b,c). Thus, the levels of gene expression required for RIM are higher than those required for normal development. Compatible with this idea, BOPs expression in plants is highly localized, restricted to young organ primordia, leaf petioles, and lateral organ boundaries, which may make systemic responses to MeJA sensitive relatively minor changes in BOP transcript abundance. Both NPR1 and the BOPs localize to the cytoplasm as well as nucleus and interact with members of the TGA family of bZIP transcription factors, albeit with different affinities (e.g., ). In development, BOP1 and BOP2 form a nuclear complex with TGA8/PERIANTHIA (PAN) to regulate number of sepals and petals in flowers and potentially to promote floral meristem fate . Given that pan loss-of-function did not reproduce the RIM- phenotype (Figure (Figure5a)5a) other genes, perhaps TGAs, are involved in this phenotype, as shown for SA perception . Given that BOPs play both positive and negative roles in transcriptional regulation of the KNOX (Knotted1-like homeobox) gene KNAT6, we also tested if RIM was affected by knat6 loss-of-function, but again, no difference was observed (Figure (Figure5a).5a). This may reflect redundancy with other KNOX genes, or more likely, that BOP regulation of RIM is independent of KNAT6.
Whether bop1 bop2 recapitulates or not all the phenotypes of the RIM- npr1 alleles (e.g. ISR, ; Verticilium resistance, ; resistance induced by Piriformospora indica,; etc.) remains to be assessed. We did check that there were similar phenotypes in the specificity of response to MeJA as well as the fact that bop1 bop2 was wild type for the rest of MeJA phenotypes (Figure (Figure7).7). But there were strong differences, since npr1-3 is affected in basal defense and SA perception while bop1 bop2 is wild type for both phenotypes . Regarding ET, the other hormone relevant for ISR, it has been proposed that applications of this hormone could render the crosstalk between SA and MeJA independent of NPR1 . It seems plausible that the ET induced resistance works as the MeJA induced resistance and other proteins -perhaps NPR1 paralogs, but not the BOPs (Figure (Figure7d)7d) - might also be affected by some alleles of npr1.
The RIM- npr1 alleles were the majority of the alleles found (32 RIM- vs. 11 RIM+). How is this compatible with the fact that the null npr1 alleles are RIM+? A possible explanation was the selection used in the screening. Since the selection was made for complete loss of SA perception, perhaps most of the RIM+ alleles had a phenotype of partial SA perception, as the null alleles. Then, the prediction would be that a good number of random alleles of npr1 would be RIM+ and partially receptive to SA. We previously showed that for SA perception, there are genetic interactions between the npr1 alleles and the NPR1 paralogs . The work reported herein points to a genetic interaction too, this being between npr1 alleles on one side and the BOPs on the other. Thus, the RIM- alleles were a phenocopy of the bop1 bop2 mutant in defense but not in development. This discrimination was a consequence of the different thresholds for the phenotype in development and defense (Figure (Figure55b,c).
Mechanistically, the levels of expression of the BOPs were low in comparison to NPR1 (Additional file 1), so a direct or indirect negative interference of NPR1 with the BOPs would be favored stoichiometrically. Once the pathogen was inoculated, the levels of SA would rise and in a wild type plant NPR1 is degraded as part of the signaling process . In an npr1 background, this signaling would not be transmitted and perhaps the npr1 proteins would be able to interfere longer in RIM. This would explain the behavior of NPR1-HBD in npr1-1 (v 2b); NPR1-HBD in the cytoplasm did not complement npr1-1 in the RIM phenotype, but when DEX was applied there was complementation of the phenotype. Likely, when no DEX was present npr1-1 would somehow interfere with the function of the BOPs. When DEX was present, the presence of NPR1-HBD in the nucleus would trigger the degradation of both NPR1-HBD and npr1-1. If npr1-1 was degraded, the BOPs would function normally.
There was no evidence for a direct interaction in yeast, since the presence of NPR1 or mutated versions of this protein did not interfere in the interaction between BOP1 and BOP2 in a consistent manner with the phenotype (Figure (Figure8a).8a). A first alternative was that the interference of the RIM- alleles would occur with the BOPs without affecting the interaction between the BOPs. A second alternative would be that the RIM- alleles would interfere with other proteins that normally interact with the BOPs. In both cases there is a family of proteins that interacts with both NPR1 and the BOPs, the transcription factors TGAs , with -again- functional redundancy (Figure (Figure8b).8b). Two out of three RIM- alleles enhanced or stabilized the BOPs-TGAs interaction, while two out of three RIM+ alleles did not (Figure (Figure8c8c and data not shown). It was clear that the npr1 mutated proteins had an unpredicted effect on the BOPs-TGAs interaction, but the yeast experiments did not produce an absolute answer about the role of npr1 proteins in RIM. We speculate that in planta, all the RIM- alleles enhance the interaction between the BOPs and the TGAs, titering out the TGAs and thus rendering them unable to fulfill their function of triggering defense. On the other hand, the RIM+ alleles (including the null alleles), and NPR1 would not affect the interaction either way. Since in the yeast assays two out of three alleles worked as proposed in either way, it may be possible that a factor(s) is present in the plant that is not in yeast, or perhaps the fact that there are ten TGAs, and that NPR1 is expressed between 3 and 18 times more than BOP1 + BOP2 (Additional file 1) could explain this difference. If this hypothesis were to be true, it will definitively explain the role of npr1 in RIM.
In sum, we have shown that, in wild type conditions, the BOPs and the TGAs (but not NPR1) are required for the resistance triggered by methyl jasmonate against Pto. We propose that the phenotype of the npr1 RIM- alleles is caused by their interference between BOPs and TGAs.
Arabidopsis thaliana (L.) Heynh. was sown and grown as described  in controlled environment rooms with days of 8 h at 21°C, 150 μmol m-2 s-1, and nights of 16 h at 19°C. Treatments, inoculations, and sampling started 30 minutes after the initiation of the artificial day to ensure reproducibility. The following genotypes were used: npr1-1 and npr1-3 ; npr2, npr1-20 to npr1-71, and combinations of npr1-70 with other genotypes ; 35SCaMVp:NPR1HBD; sid2; eds5; NahG; npr3 and npr4; bop1-3 and bop2-1 ; coi1-40 (Dobón, Wulff, Canet and Tornero, to be published elsewhere); kant6, pan1-1 to pan1-3, tga1, and tga7; 35S:BOP1 and 35S:BOP2; etr1-3 ; tga2 tga5 tga6. Pseudomonas syringae pv. tomato DC3000 (Pto) was grown, inoculated and measured as described . Briefly, plants of 14 days were inoculated by spray with Pto at OD600=0.1 with 0.02% Silwet L-77 (Crompton Europe Ltd, Evesham, UK). Three days later, the amount of colony forming units (cfu) per plant was quantified and represented in a logarithmic scale. When indicated, a strain of Pto lacking coronatine was used (Pto(cfa–),). For all the experiments, at least three independent treatments were performed (three independent sets of plants sown and treated on different dates).
NPR1 and six alleles of this gene were cloned in pDONR222 or pDONR221 (Invitrogen, Barcelona, Spain) and then transferred to pMDC43  for expression in planta with GFP and to pARC352  for expression in yeast. Similarly, BOP1, BOP2, TGA2, TGA5, and TGA6, were cloned and then transferred to pDEST22 and pDEST32 (Invitrogen) for expression in yeast. Yeast n-hybrid analyses were done as described , and the interactions were quantified as described . N. benthamiana leaf tissue was mounted in water under a coverslip 4 days after infiltration with Agrobacterium tumefaciens containing the constructs. All imaging was conducted with a Leica TCS SL confocal laser scanning microscope (Leica, Barcelona, Spain) using an HCX PL APO CS 40X/1.25 oil objective to study the subcellular localization of the fluorescence-tagged proteins. Green fluorescent protein was visualized by 488-nm excitation with an Ar laser, and its emissions were examined with a band-pass filter for 500 to 530 nm. The primers used are included as Additional file 1. Primers and chemical products were purchased from SIGMA (St. Louis, MO, USA) unless otherwise is stated. For the construction of amiRNA(BOP1-BOP2), the plasmid pRS300 was modified , cloned in pGW14 , and plants were transformed as described .
To measure the effect in Pto growth 100 μM methyl jasmonate (MeJA) in 0.1% DMSO and 0.02% Silwet L-77 (Crompton Europe Ltd) was applied by spray one day before the pathogen inoculation . Dexamethasone was applied at 2 μM diluted in water from a stock of 20 mM in EtOH. 1-Aminocyclo- propane-1-carboxylic acid (ACC) was sprayed at 1 mM in water with 0.02% Silwet L-77.
For in vitro culture, plants were grown in Johnson’s media  with 1 mM KH2PO4. When indicated, the plates were supplemented with 50 μM MeJA. The length of the roots was measured with ImageJ software . Senescence induced by MeJA was measured as described .
Total RNA from 3-week-old (Figure (Figure5c)5c) or 6-week-old plants (Figure (Figure6b)6b) was extracted with Trizol (Invitrogen), following the manufacturer’s instructions. cDNA was synthesized with RevertAid™ First Strand cDNA Synthesis Kit (Fermentas, Madrid, Spain), and the quantitative PCR performed with LuminoCt Sybr Green qPCR Ready Mix (SIGMA) in a 7000 RT-PCR Systems machine (Applied Biosystems, Madrid, Spain), following the manufacturer’s instructions. For each measurement three biological replicates were done. The obtained values were referred to the geometric average of three reference genes (At3G18780, At1G49240, and At5G60390), as described , and normalized, being the value of Col-0 in mock equal to one. The list of primers used is provided in Additional file 1.
The authors declare that they do not have any competing interest.
JVC, AD, and JF performed the experiments, analyzed the data and revised the paper. PT designed the research, analyzed the data and wrote the paper. All authors read and approved the final manuscript.
Figure S1. Localization of cloned npr1 alleles mentioned in the text. Figure S2. NPR1HBD treated with DEX is more sensitive to BTH. Figure S3. Expression levels of NPR1 paralogs. Figure S4. Response of bop1 bop2 to SA and BTH. Table S1. List of primers used.
This work was supported by the “Ministerio de Economía y Competitividad” (MINECO) of Spain (grant BIO201018896 to PT, a JAE-CSIC Fellowship to JVC and a FPI-MINECO to AD) and “Generalitat Valenciana” of Spain (grant ACOMP/2012/105 to PT). Thanks to Dr. Xinnian Dong for NPR1 overexpression lines and to Dr. Ove Nilsson for BOPs overexpression lines. We appreciate the opinions and generous help of Drs. Vicente Ramirez, Pablo Vera, and Shelley Hepworth about the manuscript.