Auxin promotes susceptibility to Pseudomonas syringae via a mechanism independent of suppression of salicylic acid-mediated defenses

Authors


For correspondence (e-mail kunkel@wustl.edu).

Summary

Auxin is a key plant growth regulator that also impacts plant–pathogen interactions. Several lines of evidence suggest that the bacterial plant pathogen Pseudomonas syringae manipulates auxin physiology in Arabidopsis thaliana to promote pathogenesis. Pseudomonas syringae strategies to alter host auxin biology include synthesis of the auxin indole-3-acetic acid (IAA) and production of virulence factors that alter auxin responses in host cells. The application of exogenous auxin enhances disease caused by P. syringae strain DC3000. This is hypothesized to result from antagonism between auxin and salicylic acid (SA), a major regulator of plant defenses, but this hypothesis has not been tested in the context of infected plants. We further investigated the role of auxin during pathogenesis by examining the interaction of auxin and SA in the context of infection in plants with elevated endogenous levels of auxin. We demonstrated that elevated IAA biosynthesis in transgenic plants overexpressing the YUCCA 1 (YUC1) auxin biosynthesis gene led to enhanced susceptibility to DC3000. Elevated IAA levels did not interfere significantly with host defenses, as effector-triggered immunity was active in YUC1-overexpressing plants, and we observed only minor effects on SA levels and SA-mediated responses. Furthermore, a plant line carrying both the YUC1-overexpression transgene and the salicylic acid induction deficient 2 (sid2) mutation, which impairs SA synthesis, exhibited additive effects of enhanced susceptibility from both elevated auxin levels and impaired SA-mediated defenses. Thus, in IAA overproducing plants, the promotion of pathogen growth occurs independently of suppression of SA-mediated defenses.

Introduction

Pathogens exploit host signaling pathways and modify host physiology to cause disease. Pseudomonas syringae pv. tomato DC3000 is a bacterial pathogen that causes disease in Arabidopsis thaliana and Solanum lycopersicum (tomato). Like other pathogens, DC3000 secretes virulence factors that alter the biology of its host to promote disease (Block and Alfano, 2011). Plant hormone signaling pathways are important targets for DC3000 virulence factors. A well-studied example of this is the virulence factor coronatine, a small molecule secreted by DC3000 that modulates the jasmonic acid (JA) and salicylic acid (SA) pathways in the host to suppress plant defenses (Kloek et al., 2001; Brooks et al., 2005; Laurie-Berry et al., 2006; Spoel and Dong, 2008).

Auxin, a key hormone in plant growth and development, is also involved in plant–pathogen interactions. Long associated with disease caused by gall-forming bacteria, such as Agrobacterium tumefaciens, auxin has more recently been discovered to be important in other plant–pathogen interactions. Many pathogenic microbes, including P. syringae, have been shown to produce auxin (Glickmann et al., 1998; Spaepen and Vanderleyden, 2011) and are reported to alter host auxin biology during infection (Marois et al., 2002; Chen et al., 2007). Levels of indole-3-acetic acid (IAA), the predominant active form of auxin in A. thaliana, increase during infection (O'Donnell et al., 2003; Chen et al., 2007). Furthermore, treating plants with exogenous auxin promotes susceptibility (Navarro et al., 2006; Chen et al., 2007; Wang et al., 2007). Treatment with bacterial flagellin, a microbe-associated molecular pattern that triggers host defense responses, downregulates auxin signaling (Navarro et al., 2006). Taken together, these results indicate that high levels of auxin and enhanced auxin signaling contribute to enhanced disease susceptibility.

The mechanism by which auxin promotes disease is, however, unknown. The favored hypothesis is that auxin suppresses SA-mediated defenses (Robert-Seilaniantz et al., 2011a). Previous studies (Park et al., 2007; Wang et al., 2007) found that plants co-treated with exogenous auxin and SA had reduced expression of the defense marker PATHOGENESIS-RELATED GENE 1 (PR1), compared with plants treated with SA alone. Thus, we sought to further investigate the role of auxin during pathogenesis by examining the interaction of auxin and SA in the context of infection in plants with altered endogenous levels of auxin.

In this study, we examine the hypothesis that elevated auxin levels promote pathogen growth by suppressing SA-mediated defenses. Experimental manipulation of auxin levels is challenging in plants because of the hormone's essential role in plant growth and development; however, the availability of transgenic A. thaliana lines overexpressing YUCCA 1 (YUC1; Zhao et al., 2001; Mashiguchi et al., 2011; Won et al., 2011), a key enzyme in the auxin biosynthesis pathway, allowed us to use plants with elevated auxin levels in our studies. We show that plants overexpressing YUC1 have elevated auxin levels and enhanced susceptibility to DC3000, supporting previous findings that auxin promotes pathogenesis; however, our results indicate that this is not primarily the result of suppression of SA-mediated defenses. Rather, elevated auxin levels appear to promote pathogenesis via a mechanism that is independent of the suppression of SA-mediated defenses.

Results

Overexpression of YUC1 leads to enhanced susceptibility to DC3000

To study the effects of elevated levels of endogenous auxin on the susceptibility of A. thaliana to DC3000, we used transgenic plants overexpressing YUC1 under the control of the strong, constitutive cauliflower mosaic virus (CaMV) 35S promoter (35S:YUC1). Seedlings expressing 35S:YUC1 have elevated IAA levels, elongated hypocotyls and epinastic leaves (Zhao et al., 2001). To verify that 35S:YUC1 plants maintain significantly increased levels of IAA as adults, we quantified IAA in 4-week-old plants and found the levels remain elevated in adult leaf tissue (Figure 1a). When infiltrated with a low inoculum of DC3000 (OD600 = 1 × 10−4), 35S:YUC1 plants supported increased bacterial growth compared with wild type Col-0 (Figure 1b). The increased growth was similar to that observed in the salicylic acid induction deficient 2 (sid2) mutant, which has a mutated ISOCHORISMATE SYNTHASE 1 (ICS1) gene, leading to reduced SA biosynthesis and reduced defenses against DC3000 (Wildermuth et al., 2001). The enhanced susceptibility of 35S:YUC1 plants was not observed reproducibly when plants were inoculated via dip inoculation or with a higher dose of bacteria (Figure 2a; A. Mutka, unpublished data), similar to what was previously observed for sid2 (Brooks et al., 2005), presumably because bacterial levels reach the carrying capacity of the leaf. In experiments that enable the observation of enhanced susceptibility, however, the increased bacterial growth on 35S:YUC1 plants suggests that elevated endogenous auxin levels promote susceptibility to DC3000.

Figure 1.

Mature YUC1-overexpressing plants have elevated auxin levels and enhanced susceptibility to DC3000. (a) Measurement of free IAA levels in 4-week-old 35S:YUC1 leaves by LC-MS/MS. Values are an average of three biological replicates. Similar results were obtained in two additional independent experiments; FW, fresh weight. (b) Growth of DC3000 in 35S:YUC1 and sid2-2 following syringe infiltration (OD600 = 1 × 10−4). Values are an average of four biological replicates. Similar results were observed in six out of a total of seven experiments. *< 0.05; **< 0.01, versus Col-0.

Figure 2.

35S:YUC1 plants exhibit strong effector-triggered immunity, but have an attenuated hypersensitive response (HR). (a) Growth of DC3000 (ev) carrying the empty vector pLAFR3 and DC3000 (pLAFR3-avrRpm1) in Col-0 wild-type (WT) and 35S:YUC1 plants following dip inoculation (OD600 = 0.4). Values are an average from four biological replicates; similar results were obtained in two additional experiments; *< 0.05, between samples. (b) Comparison of tissue collapse between Col-0 WT and 35S:YUC1 plants. Leaves were infiltrated with a 10 mm MgCl2 mock solution or bacterial solution (OD600 = 0.05). Strains used were Pseudomonas syringae pv. phaesiolicola (Psp) and Psp (pLAFR3-avrRpm1) to induce the HR. For each treatment, 30–40 leaves were inoculated. Categories H1–H5 indicate the severity of tissue collapse observed in the inoculated leaf area (H1, no evidence of collapse; H2, tissue softening with no color change; H3, tissue softening and turning gray; H4, partial tissue collapse; H5, total tissue collapse). Observations of HR were recorded at 21–23 h post inoculation. Values are an average of three independent experiments. Error bars indicate the standard errors of the means; *< 0.05 compared with Col-0 + Psp (AvrRpm1) for the category indicated.

Biosynthesis of indole glucosinolates (IGs), a class of defense compounds, occurs through a branch of metabolism originating with tryptophan, similar to the auxin biosynthesis pathway (Sugawara et al., 2009; Mashiguchi et al., 2011; Won et al., 2011). The YUC1 enzyme catalyzes the conversion of IPA to IAA in the final step of IAA biosynthesis. Because YUC1-overexpressing plants exhibit increased conversion of tryptophan to IAA, in theory this could lead to reduced levels of IGs and subsequently enhanced disease susceptibility. Although IGs are thought to play a role in pathogen interactions (Clay et al., 2009), it is not clear to what extent they impact DC3000 growth in planta. To investigate whether reduced IG levels affect DC3000 growth, we infected the cyp79B2 cyp79B3 double mutant, in which the production of IGs is eliminated (Zhao et al., 2002). Growth of DC3000 in cyp79B2 cyp79B3 was indistinguishable from that in the wild type (Figure S1). Thus, the elimination of IGs does not have a detectable effect on DC3000 growth in planta, and it is unlikely that reduced IG levels cause enhanced susceptibility in 35S:YUC1 plants. Rather, it is more likely that elevated auxin levels cause enhanced susceptibility when YUC1 is overexpressed.

AvrRpm1-triggered immunity occurs in YUC1-overexpressing plants

A favored hypothesis is that auxin promotes susceptibility in virulent interactions by suppressing host defenses. To investigate plant defense responses in the context of elevated auxin levels, we inoculated wild-type and 35S:YUC1 plants with strains of DC3000 expressing the type-III effector protein AvrRpm1 and monitored resistance to pathogen growth and the hypersensitive response (HR). Wild-type Col-0 plants infected with DC3000 expressing AvrRpm1 initiate a strong defense response that restricts pathogen growth (Debener et al., 1991), compared with wild-type plants infected with the empty vector control (Figure 2a). Similarly, 35S:YUC1 plants inoculated with DC3000 (AvrRpm1) restrict bacterial growth compared with 35S:YUC1 plants inoculated with the empty vector control. Thus, AvrRpm1-triggered immunity occurs in 35S:YUC1 plants; however, growth was higher in 35S:YUC1 plants treated with the avirulent strain compared with wild-type plants treated with the avirulent strain (Figure 2a), suggesting that elevated auxin levels can promote pathogen growth, even in the context of a strong resistance response. Alternatively, auxin may partially suppress the defense response.

Another hallmark of defense is the HR, a programmed cell death response triggered upon recognition of pathogen effectors. We examined whether elevated auxin levels in 35S:YUC1 plants impact the HR. When inoculated with the non-pathogen strain P. syringae pv. phaseolicola (Psp), expressing AvrRpm1, wild-type plants showed a strong HR, with about 70% of the inoculated leaves exhibiting partial or total tissue collapse (Figure 2b). When 35S:YUC1 plants were inoculated with Psp (AvrRpm1), only about 23% of inoculated leaves exhibited partial or total tissue collapse. Thus, elevated auxin levels may suppress the progression of cell death associated with the HR, consistent with what has been previously reported by others (Gopalan, 2008).

SA levels and SA-dependent gene expression are not suppressed in infected YUC1-overexpressing plants

Previous studies found that treatment with exogenous auxin reduced expression of PR1, an SA-dependent defense gene (Park et al., 2007; Wang et al., 2007). We investigated whether YUC1 overexpression suppresses SA levels and SA-mediated defenses in the context of DC3000 infection. In infected wild-type plants, SA accumulated to high levels compared with the mock-treated control, but no SA accumulation occurred in infected sid2-2 plants, as expected (Figure 3a). In mock-treated plants, SA levels were higher in YUC1-overexpressing plants than in the wild type in three out of four experiments (Figure 3a), suggesting that there is a positive relationship between auxin and SA in uninfected 35S:YUC1 plants. For infected 35S:YUC1 plants, we observed that SA accumulated to levels similar to those observed in infected wild-type plants; however, this accumulation was variable between experiments. In two out of four experiments infected 35S:YUC1 plants accumulated SA to levels that were not significantly different from those in infected wild-type plants, but in the other two, SA accumulation was lower in 35S:YUC1 plants compared with the wild type (Figure S2). Given that 35S:YUC1 plants exhibited enhanced susceptibility in all experiments, it is interesting to note that increased pathogen growth occurs despite high SA accumulation, and that 35S:YUC1 and sid2-2 have similar elevated susceptibility, despite their differences in SA accumulation (Figures 1b and 3a). Furthermore, our observation that auxin levels remain elevated in infected 35S:YUC1 plants (Figure S3) indicates that the effect of elevated auxin on pathogen growth occurs despite SA accumulation. In summary, the elevated auxin levels in 35S:YUC1 plants do not have a strong or reproducible effect on SA levels, suggesting that auxin impacts plant defense responses downstream or independently of SA accumulation.

Figure 3.

Overexpression of YUC1 impacts pathogen growth independently of SA accumulation and SA-related defense responses. (a) Quantification of SA levels in 35S:YUC1 and sid2-2 mock-treated and infected plants (OD600 = 1 × 10−4) by LC-MS/MS. Plant tissue was isolated at 24 h post-inoculation, and values are an average of three replicates. Similar results were seen in two out of a total of four experiments. Letter differences indicate < 0.05. (b, c) ICS1 expression (b) and PR1 expression (c) in plants infected by syringe infiltration (OD600 = 1 × 10−2). Values are an average of three biological replicates. Similar results were obtained in a second independent experiment. (d) Growth of DC3000 in 35S:YUC1, sid2-2 and 35S:YUC1 sid2-2 following syringe infiltration (OD600 = 1 × 10−5). Values are an average of seven biological replicates. Similar results were obtained in two additional experiments. Letter differences indicate < 0.10 between samples.

To investigate whether YUC1 overexpression impacts SA biosynthesis, we monitored the expression of ICS1, which encodes the enzyme responsible for most SA synthesis during infection (Wildermuth et al., 2001). We investigated ICS1 expression in mock- and DC3000-treated plants. In both infected wild-type and 35S:YUC1 plants, ICS1 was induced by 24 h post-inoculation (Figure 3b). ICS1 induction was somewhat lower in 35S:YUC1 than in the wild type, but this difference was not significant. These results, indicating that expression of the ICS1 SA biosynthesis gene is not dramatically impacted by high auxin levels, correlate with our finding that SA levels are not significantly altered in infected 35S:YUC1 plants relative to the wild type.

We also monitored the expression of PR1, a commonly used marker gene of SA-mediated defense. PR1 expression was upregulated to similar levels in infected wild-type and 35S:YUC1 plants by 24 h post-inoculation (Figure 3c). Thus, elevated auxin levels in 35S:YUC1 do not promote susceptibility simply by suppressing SA-mediated defenses. This provides further evidence that auxin promotes pathogenesis independently or downstream of SA-mediated defense signaling.

YUC1 overexpression promotes pathogenesis in an ICS1-independent manner

Our studies above suggest that auxin promotes susceptibility to DC3000 by acting independently of SA accumulation. If so, we predict that overexpression of YUC1 in the sid2-2 background would exhibit the additive effects of enhanced susceptibility from both elevated auxin levels and impaired SA-mediated defenses. Thus, 35S:YUC1 sid2-2 plants would support even higher levels of pathogen growth than a line with either 35S:YUC1 or sid2-2 alone. To test this, we crossed 35S:YUC1 into the sid2-2 mutant background, infected the plants with a low inoculum of bacteria (OD600 = 1 × 10−5) by syringe infiltration, and monitored bacterial growth. The 35S:YUC1 sid2-2 line supported higher bacterial growth than either 35S:YUC1 or sid2-2 (Figure 3d). Thus, YUC1 overexpression promotes bacterial growth independently of ICS1, again supporting the hypothesis that auxin promotes pathogen growth independently of the suppression of SA-mediated defenses.

The ability of 35S:YUC1 to promote pathogen growth does not depend on GH3.2

The results above leave unanswered the question of how elevated auxin levels promote pathogenesis. One possibility is that when IAA is overproduced, either free IAA or a modified form of IAA could have direct effects on pathogen virulence. Much of the auxin in plants occurs in the form of IAA-amino acid conjugates. One such conjugate, IAA-Asp, was reported to promote DC3000 pathogenesis during infection (Gonzalez-Lamothe et al., 2012). The A. thaliana auxin-responsive GH3.2 gene, which encodes an enzyme believed to be responsible for synthesizing IAA-Asp, was induced during pathogen infection. When GH3.2 was mutated, IAA-Asp accumulation was reduced, and the gh3.2 mutant was less susceptible to infection by DC3000 (Gonzalez-Lamothe et al., 2012).

We hypothesized that elevated levels of auxin resulting from the overexpression of YUC1 could result in the elevated expression of GH3.2 and in the accumulation of IAA-Asp, which could promote DC3000 virulence. To test this hypothesis, we first examined GH3.2 expression in mock-treated and DC3000-infected 35S:YUC1 plants. The gene was induced in infected wild-type and 35S:YUC1 plants, as previously reported. However, counter to our prediction, GH3.2 expression was not elevated in mock-treated 35S:YUC1 plants (Figure 4a). Furthermore, in preliminary experiments, we did not detect elevated levels of IAA-Asp in either uninfected or infected 35S:YUC1 (A. Mutka, unpublished data). Thus, the induction of GH3.2 seems to be specific to infected plants and is not simply the result of elevated IAA.

Figure 4.

The ability of 35S:YUC1 to promote pathogen growth does not depend on GH3.2. (a) GH3.2 expression in plants infected by syringe infiltration (OD600 = 1 × 10−2). Values are an average of three biological replicates. Similar results were obtained in 1 additional experiment. (b) Growth of DC3000 in gh3.2, 35S:YUC1 and 35S:YUC1 gh3.2 plants following syringe infiltration (OD600 = 1 × 10−4). Values are an average of seven biological replicates. Similar results were obtained in one additional experiment. *< 0.05, versus Col-0.

To investigate whether increased susceptibility in 35S:YUC1 depends on GH3.2, we crossed the gh3.2 mutation into the 35S:YUC1 background and monitored pathogen growth following inoculation. We were not able to replicate the previous results (Gonzalez-Lamothe et al., 2012) that gh3.2 mutants have reduced susceptibility to DC3000 (Figure 4b). Furthermore, 35S:YUC1 gh3.2 plants supported elevated pathogen growth, similar to that observed in 35S:YUC1 alone, indicating that GH3.2 is not required for the elevated growth of DC3000 in 35S:YUC1 plants.

Discussion

In summary, we showed that 35S:YUC1 plants accumulate elevated auxin levels as adults, and have an enhanced susceptibility to DC3000 (Figure 1). Counter to expectations, enhanced auxin biosynthesis did not have significant impacts on plant defense. 35S:YUC1 plants mounted a resistance response when inoculated with avirulent bacterial strains (Figure 2). These plants also did not exhibit significant suppression of either SA accumulation or SA-responsive gene expression (Figure 3). Furthermore, a plant line carrying both 35S:YUC1 and the sid2-2 mutation exhibited additive effects of enhanced susceptibility from elevated auxin levels and impaired SA synthesis (Figure 3d). Together, our findings suggest that auxin promotes pathogen growth through a mechanism that is independent of the suppression of SA-mediated defenses.

Elevated auxin biosynthesis enhances susceptibility

Several studies have found that elevated auxin levels can promote disease development (Navarro et al., 2006; Chen et al., 2007; Wang et al., 2007), but the mechanism behind this activity is not well understood. Despite the fact that previous work has shown that exogenous auxin treatment can suppress SA-mediated defense gene expression (Park et al., 2007; Wang et al., 2007), our results do not support the hypothesis that the primary role of auxin during pathogenesis is to suppress SA-mediated defenses. This discrepancy between our findings and published results could be explained by different experimental conditions. Treating plants exogenously with high concentrations of hormones may have different effects than if auxin levels are elevated as a result of increased biosynthesis of the hormone in plant tissues. Likewise, plants may have very different responses to hormones in the absence of infection, compared with those during infection. Pathogen infection modulates a variety of hormone signaling pathways that involve many regulatory interactions (Robert-Seilaniantz et al., 2011a). Thus, our experiments with infected plants potentially provide a more biologically relevant look at the interaction between auxin and SA in a host–pathogen system.

A caveat of using 35S:YUC1 plants in our experiments is that they are somewhat abnormal developmentally and morphologically, raising the possibility that it is not elevated auxin levels that directly promote pathogen growth in these plants, but other pleiotropic effects. However, other transgenic lines or mutants that exhibit significant changes in auxin levels or auxin signaling exhibit very severe developmental phenotypes (Woodward and Bartel, 2005; Cheng et al., 2006), making the plants impractical for pathogen infection experiments. Thus, 35S:YUC1 plants are a useful tool for studying the effects of elevated auxin on pathogenesis.

Multiple roles for auxin during infection

It is important to emphasize that any SA-independent mechanism through which auxin acts during pathogenesis is likely to function in combination with other mechanisms that do suppress SA signaling. Altered auxin signaling has been shown to impact SA-mediated defenses. For example, the overexpression of AFB1, an auxin receptor, led to the suppression of SA levels and signaling, and to a higher susceptibility to biotrophic pathogens (Robert-Seilaniantz et al., 2011b). Our experiments, however, used plants with altered auxin biosynthesis, which could have additional or different effects during infection. Pathogens, for instance, may sense and respond directly to auxin levels rather than being impacted by changes in the host caused by increased auxin signaling. This leads us to hypothesize that auxin plays multiple roles in promoting pathogenesis, an idea that is consistent with the wide variety of processes that auxin regulates in plant and microbial systems.

Potential SA-independent mechanisms that promote pathogenesis

As auxin is involved in many processes, there are numerous potential mechanisms through which it may promote pathogenesis. Elevated auxin may have physiological effects on plant tissues and cells that may render the host more suitable to pathogen growth or promote the progression of disease (Kazan and Manners, 2009). For example, as auxin is known to promote cell wall loosening (Cosgrove, 2005), high levels of auxin may cause changes in the properties of plant cell walls that give pathogens easier access to host cells for the delivery of type III secreted effectors and other virulence factors. Changes in cell wall properties may also facilitate the release of water or nutrients into the apoplast, allowing the pathogen to further proliferate. Additionally, auxin may interfere with programmed cell death associated with the HR (Gopalan, 2008). Our observation that 35S:YUC1 plants exhibit a weaker HR (Figure 2b) is consistent with this idea.

Direct effects on pathogen virulence

Auxin may directly affect pathogen biology, for example by enabling it to sense the plant host and upregulate virulence genes or alter its metabolism in ways that allow it to better grow within the host and cause disease. Previous studies suggest that IAA and related indolic compounds serve as signaling molecules for bacteria (Mazzola and White, 1994; Bianco et al., 2006; Hirakawa et al., 2009). Further studies examining the effects of auxin on pathogen gene expression would be useful.

The distinction between auxin biosynthesis and signaling implied by our data suggests that different forms of auxin may be involved in different aspects of plant–pathogen interactions. IAA-Asp has been reported to promote pathogenesis, possibly by impacting the pathogen directly (Gonzalez-Lamothe et al., 2012), but we were not able to replicate the previously published results that support this finding. Although it is possible that these discrepancies stem from differences in growth conditions or inoculation procedures, our results call into question the importance of GH3.2, and possibly IAA-Asp, in promoting pathogen growth in planta. Given that IAA-amino acid conjugates are thought to be located in the cytosol and endoplasmic reticulum of plant cells (Ludwig-Muller, 2011), IAA-Asp is unlikely to directly impact the growth of DC3000 in the apoplast. Nonetheless, other auxin conjugates produced by the plant may have other significant roles in mediating the interactions between pathogens and plants. Additionally, IAA or other auxinic compounds produced by pathogens may promote disease development. Thus, further studies on the form, source and mode of action for auxin are necessary.

Figure 5 illustrates our working model regarding the roles of auxin during pathogenesis. Our findings with YUC1-overexpressing plants indicate that the primary role of increased auxin levels in promoting disease susceptibility to DC3000 occurs through a mechanism that is independent of the suppression of SA-mediated defense. This mechanism may involve impacts on host physiology, pathogen virulence mechanisms or a combination of both, and is likely to act in addition to other mechanisms by which auxin signaling suppresses SA-mediated defenses (Park et al., 2007; Wang et al., 2007; Robert-Seilaniantz et al., 2011b).

Figure 5.

Working model for the roles of auxin during DC3000 pathogenesis. Upon pathogen attack, expression of the salicylic acid (SA) biosynthesis gene ICS1 is upregulated, SA levels increase, and plant defenses (e.g. PR1 expression) are induced via an NPR1-dependent signaling process to restrict pathogen growth and disease in the host. In plants overexpressing the YUCCA1 (YUC1) auxin biosynthesis gene, auxin (IAA) levels are elevated, which promotes pathogen growth and disease development via a mechanism that does not involve the suppression of SA-mediated defense. This may be mediated through physiological changes in the host or through direct effects of IAA on pathogen virulence. Additionally, IAA or stimulation of auxin signaling may also act to promote pathogenesis by inhibiting SA-mediated defense (Park et al., 2007; Wang et al., 2007; Robert-Seilaniantz et al., 2011b). IAA, indole-3-acetic acid; ICS1, ISOCHORISMATE SYNTHASE 1; IPA, indole-3-pyruvic acid; NPR1, NONEXPRESSER OF PR GENES 1; PR1, PATHOGENESIS-RELATED GENE 1; SA, salicylic acid; TRP, tryptophan.

Much remains to be discovered about how pathogens infect plants and cause disease. The roles of classical plant defense pathways, such as the SA pathway, are important for understanding plant–pathogen interactions. Justifiably, many studies of pathogenic mechanisms have focused on how pathogens modulate these defense pathways to cause disease (Block and Alfano, 2011). Our results, however, suggest that a broader range of pathogenic mechanisms relating to auxin should be considered, including those resulting in the modulation of other aspects of host biology. Further understanding of the roles of auxin during pathogenesis may provide valuable insight into these processes and elucidate disease-causing mechanisms that impact both the host and the pathogen.

Experimental procedures

Plant material and growth conditions

All A. thaliana transgenic lines and mutants used in this study were in the Col-0 ecotype background. 35S:YUC1 (Zhao et al., 2001) was obtained from Yunde Zhao. sid2-2 (Wildermuth et al., 2001) was obtained from Mary Wildermuth. cyp79B2 cyp79B3 (Zhao et al., 2002; Clay et al., 2009) was obtained from Georg Jander.

Plants were grown on soil in a growth chamber with a short-day photoperiod (8-h light/16-h dark) at 21°C and 75% relative humidity, with a light intensity of approximately 130 μEinsteins sec−1 m−1. The gh3.2 mutant (SALK_037520C) was obtained from the Arabidopsis Biological Resource Center (Columbus, OH, USA).

Generation of the 35S:YUC1 sid2-2 line

A homozygous 35S:YUC1 line was crossed with sid2-2, and the F1 progeny were allowed to self-pollinate. Approximately 75% of the F2 progeny exhibited the typical yucca morphology, which includes a long hypocotyl, and epinastic cotyledons and leaves, with elongated petioles (Zhao et al., 2001). Several of these plants were genotyped using polymerase chain reaction (PCR) to identify plants homozygous for the sid2-2 mutation or for the wild-type SID2/ICS1 allele. The sid2-2 mutation (also known as eds16-1) is a deletion of approximately 6500 bp spanning the ICS1 (At1 g74710) gene (M. Wildermuth, personal communication). Genotyping was carried out using the following three primers: sm108F (5′-TTCTTCATGCAGGGGAGGAG-3′), sm30F (5′-CAACCACCTGGTGCACCAGC-3′) and L1849R (5′-AAGCAAAATGTTTGAGTCAGCA-3′). Amplification with the sm108F and L1849R primers gives rise to a product of approximately 7300 bp for the wild-type SID2/ICS1 allele present in Col-0, and a product of approximately 580 bp for the sid2-2 mutant. Amplification with sm30F and L1849R primers yields a product of approximately 880 bp for the wild-type SID2/ICS1 allele, and no product for the sid2-2 mutation. The sid2-2 and SID2/ICS1 homozygous plants were grown to maturity, and their F3 self progeny were tested to identify lines that were homozygous for the 35S:YUC1 expression construct.

Generation of the 35S:YUC1 gh3.2 line

A homozygous 35S:YUC1 line was crossed with gh3.2 (SALK_037520C), which carries a T-DNA insertion that disrupts the GH3.2 gene (Gonzalez-Lamothe et al., 2012). F1 progeny from the cross were allowed to self-pollinate, and F2 plants were genotyped by PCR using the following primers: LBb1.3 (5′-ATTTTGCCGATTTCGGAA-3′), GH3.2 F1 (5′-TACGTAACCACCGGAACTTTG-3′) and GH3.2 R1 (5′-AGAGCGGATGATTGTTGATTG-3′). The GH3.2 gene-specific primers are the same as those described by Gonzalez-Lamothe et al. (2012). Amplification with GH3.2 F1 and GH3.2 R1 yields a product of approximately 1000 bp from a wild-type copy of GH3.2 and no product from a gh3.2 mutant copy. Amplification with LBb1.3 and GH3.2 R1 yields a product of approximately 600 bp from the gh3.2 mutant and no product from the wild-type copy. A heterozygous plant yields a product with both sets of primers. The presence of 35S:YUC1 was determined from the visual yucca phenotype (Zhao et al., 2001).

Bacterial strains and culture conditions

Pseudomonas syringae pv. tomato DC3000 wild type, DC3000(pLAFR3) and DC3000(p48), which carries AvrRpm1 on a pLAFR plasmid (Debener et al., 1991; Kunkel et al., 1993), were used in pathogen growth assays. Pseudomonas syringae pv. phaseolicola (Psp) strain 3121 wild type (Lindgren et al., 1986) and Psp3121(pK48) were used for hypersensitive response assays. Bacterial cultures were grown on NYG agar media with 100 μg ml−1 rifampicin at 28–30°C, plus other antibiotics as needed (25 μg ml−1 kanamycin, 16 μg ml−1 tetracycline).

Plant inoculations and quantification of bacterial growth

Arabidopsis thaliana plants were infected at 4–5 weeks of age. For dip inoculations, plants were dipped in a solution containing bacteria at approximately 4 × 108 cells ml−1 (OD600 = 0.4), 10 mm MgCl2 and 0.02% Silwet L-77. To quantify bacterial growth in the plant, whole leaves were sampled at various time points after inoculation, ground in 10 mm MgCl2 and then plated in serial dilutions on NYG media with rifampicin. Between four and seven leaves were sampled per treatment, depending on the experiment. Following incubation at 28°C for 48 h, colonies were counted to determine the number of bacteria in the leaves. On the day of inoculation, leaves were sampled at 2 h after inoculation, surface sterilized with 15% H2O2 and then washed twice with water before grinding to remove bacteria from the surface of the leaf. For syringe infiltrations, a solution containing 104–105 cells ml−1 (OD600 = 10−5–10−4) and 10 mm MgCl2 was injected into leaves using a 1-ml needleless syringe. Bacterial growth was monitored as described for dip inoculations.

Hypersensitive response assays

Arabidopsis thaliana plants were inoculated with Psp strains to assess tissue collapse associated with the HR. Solutions containing 5 × 107 cells ml−1 in 10 mm MgCl2 (OD600 = 5 × 10−2) were injected into leaves using a needleless syringe. Mock-treated leaves were injected with 10 mm MgCl2. Around 30–40 leaves were inoculated per treatment. The severity of tissue collapse was recorded at 21–23 h after inoculation.

Hormone quantification

Approximately 100 mg of plant leaves were sampled for each of three replicates per treatment, frozen in liquid nitrogen, and stored at −80°C prior to being assayed. IAA and SA were quantified by LC-MS/MS as described previously (Chen et al., 2009).

Quantitative RT-PCR

For the analysis of defense-related gene expression in infected plants we infiltrated entire rosette leaves of 4-week-old plants with DC3000 (OD600 = 1 × 10−2) or a mock solution of 10 mm MgCl2. Approximately 100 mg of leaf tissue was isolated at 6 and 24 h post infiltration, and frozen immediately in liquid nitrogen. Three biological replicates per treatment were harvested and analyzed for each time point. Total RNA was isolated with the RNeasy Plant Mini Kit (Qiagen, http://www.qiagen.com), and genomic DNA was removed with DNase I (Life Technologies, http://www.lifetechnologies.com). cDNA synthesis was performed using the RNA, SuperScript III Reverse Transcriptase (Life Technologies) and oligo(dT)20 primers. Negative control reactions lacking reverse transcriptase were run in parallel to verify that there was no contamination from genomic DNA. Quantitative RT-PCR reactions were then set up with the cDNA and SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich, http://www.sigmaaldrich.com), using a final reaction volume of 25 μl. See Table S1 for the gene-specific PCR primers used. Three technical replicates were run per sample. Reactions were run on the Applied Biosystems 7500 Real-Time PCR System. Relative expression was determined using a sigmoidal model (Chervoneva et al., 2007; Rutledge and Stewart, 2008). Data analysis calculations were performed using the Java program lre analyzer (Rutledge, 2011), and expression was normalized to the reference gene PP2AA3 (At1g13320; Czechowski et al., 2005). Sequences for the qPCR primers used can be found in Table S1.

Statistical analysis

The Student's t-test was used for all statistical analysis. Error bars on all figures represent the standard error of the mean.

Acknowledgements

We thank Liz Haswell and Lucia Strader for comments on the article, and Sheri McClerklin, Ram Dixit, Corey Westfall, Joe Jez and Libo Shan for helpful discussion. We thank Mary Wildermuth for sharing unpublished information about the molecular nature of the sid2-2 allele. We also thank the Proteomics and Mass Spectrometry Facility of the Donald Danforth Plant Science Center (St Louis, MO, USA) for hormone analysis. Funding for the 4000 QTRAP mass spectrometer was provided through a National Science Foundation Major Research Instrumentation Program (NSF-MRI) grant (DBI-0521250). This work was funded by NSF grant IOS-1030250. A.M. was funded by an NSF graduate research fellowship.

Ancillary