Identification of likely orthologs of tobacco salicylicacid-binding protein 2 and their role in systemic acquiredresistance in Arabidopsis thaliana

Anna Corina Vlot1,†, Po-Pu Liu1, Robin K. Cameron2, Sang-Wook Park1,‡, Yue Yang3,§, Dhirendra Kumar1,–, Fasong Zhou1,**,

Thihan Padukkavidana1,††, Claes Gustafsson4, Eran Pichersky3 and Daniel F. Klessig1,*

1Boyce Thompson Institute for Plant Research, Tower Road, Ithaca, NY 14853, USA,2Department of Biology, McMaster University, 1280 Main St West, Hamilton, ON, Canada L8S 4K1,3Department of Molecular, Cellular and Developmental Biology, University of Michigan, 830 N. University Street,

Ann Arbor, MI 48109, USA, and4DNA2.0, Inc., 1430 O’Brien Drive, Suite E, Menlo Park, CA 94025, USA

Received 7 May 2008; revised 24 June 2008; accepted 25 June 2008; published online 20 August 2008.*For correspondence (fax +1 607 254 6779; e-mail dfk8@cornell.edu).†Present addresses: Max Planck Institute for Plant Breeding Research, 50829 Cologne, Germany.‡Virginia Bioinformatics Institute, Virginia Polytechnic University, Blacksburg, VA 24061, USA.§Department of Plant Biology, Michigan State University, East Lansing, MI 48824, USA.–Department of Biological Sciences, East Tennessee State University, Johnson City, TN 37614, USA.

**Ceres Inc., Thousand Oaks, CA 91320, USA.††Department of Molecular, Cellular, and Developmental Biology, Yale University, New Haven, CN, USA.

Summary

Salicylic acid-binding protein 2 (SABP2) is essential for the establishment of systemic acquired resistance

(SAR) in tobacco; SABP2’s methyl salicylate (MeSA) esterase activity is required in healthy systemic tissues of

infected plants to release the active defense phytohormone SA from MeSA, which serves as a long-distance

signal for SAR. In the current study, we characterize a new gene family from Arabidopsis thaliana encoding 18

potentially active a/b fold hydrolases that share 32–57% identity with SABP2. Of 14 recombinant AtMES (MES

for methyl esterase) proteins tested, five showed preference for MeSA as a substrate and displayed SA

inhibition of MeSA esterase activity in vitro (AtMES1, -2, -4, -7, and -9). The two genes encoding MeSA

esterases with the greatest activity, AtMES1 and -9, as well as AtMES7 were transcriptionally upregulated

during infection of Arabidopsis with avirulent Pseudomonas syringae. In addition, conditional expression of

AtMES1, -7, or -9 complemented SAR deficiency in SABP2-silenced tobacco, suggesting that these three

members of the AtMES family are SABP2 functional homologs (orthologs). Underexpression by knockout

mutation and/or RNAi-mediated silencing of multiple AtMES genes, including AtMES1, -2, -7, and -9,

compromised SAR in Arabidopsis and correlated with enhanced accumulation of MeSA in the systemic tissue

of SAR-induced plants. Together, the data show that several members of the AtMES gene family are

functionally homologous to SABP2 and redundant for MeSA hydrolysis and probably SAR. These data suggest

that MeSA is a conserved SAR signal in Arabidopsis and tobacco.

Keywords: systemic acquired resistance, methyl salicylate, methyl esterase, Arabidopsis, salicylic acid-

binding protein 2, defense signal.

Introduction

In nature, plants are subjected to a wide range of stresses

such as cold and drought, as well as attacks by pathogens

including bacteria, fungi, and viruses. To cope with these

stresses, many different defense mechanisms are activated

by various signal transduction pathways that nevertheless

appear to belong to one complex network revolving around

the phytohormones salicylic acid (SA), jasmonic acid (JA),

and ethylene (ET) (Glazebrook, 2005; Robert-Seilaniantz

et al., 2007). In most experimental systems, SA and JA/ET

signaling antagonize each other, although synergism has

ª 2008 The Authors 445Journal compilation ª 2008 Blackwell Publishing Ltd

The Plant Journal (2008) 56, 445–456 doi: 10.1111/j.1365-313X.2008.03618.x

been reported (e.g. Mur et al., 2006). Salicylic acid is gen-

erally critical for defense against (hemi-)biotrophic patho-

gens, whereas JA and ET synergistically induce defense

against necrotrophic pathogens and insects (Glazebrook,

2005). Two distinct forms of inducible defense are depen-

dent on either SA or JA/ET. Of these, SA-dependent

systemic acquired resistance (SAR) is a long-lasting,

broad-based resistance that is established in uninfected

parts of plants after localized exposure to pathogens

(Durrant and Dong, 2004; Vlot et al., 2008).

Although a hypersensitive response (HR) is not strictly

required to establish SAR (Cameron et al., 1994; Mishina and

Zeier, 2007), SAR signaling generally has been studied in

systemic healthy tissues of plants during or after the

establishment of a localized HR. A HR is induced during a

so-called gene-for-gene interaction between plant and path-

ogen after a plant resistance (R) protein either directly or

indirectly recognizes a cognate pathogen-encoded aviru-

lence (Avr) protein. This recognition event induces a robust

defense response including induction of reactive oxygen

species (ROS), nitric oxide (NO), SA, and expression of

defense genes such as Pathogenesis Related 1 (PR1), and

ultimately results in programmed death of cells in and

around the site of infection (HR; Greenberg and Yao, 2004).

Accumulation of SA in the healthy systemic tissue of

infected plants is essential for SAR. However, grafting

experiments using transgenic tobacco expressing the SA-

degrading salicylate hydroxylase encoded by NahG as

rootstocks for the primary infection showed that SA itself

is most probably not the mobile signal inducing SAR

(Vernooij et al., 1994). Over the past decade a few candidate

mobile signals have been proposed; e.g. several studies

have indicated a role for lipids or lipid-derived molecules in

SAR signaling (Buhot et al., 2004; Chaturvedi et al., 2008;

Maldonado et al., 2002; Nandi et al., 2004; Truman et al.,

2007). Recently, the SA derivative methyl salicylate (MeSA)

was shown to play a pivotal role as a long-distance mobile

SAR signal in tobacco (Park et al., 2007). Salicylic acid-

binding protein 2 (SABP2) mediates hydrolysis of MeSA to

SA, and its enzymatic activity is required in systemic tissues

of tobacco mosaic virus (TMV)-infected tobacco to mount a

successful SAR response. Reduction of MeSA in primary

infected leaves by either overexpression of a mutant SABP2

with unregulated MeSA esterase activity or silencing of a

SAMT (SALICYLIC ACID METHYL TRANSFERASE) gene

abolished SAR in tobacco (Park et al., 2007). Importantly,

MeSA levels increased in petiole exudates of leaves after

TMV infection, and localized treatment of plants with MeSA

resulted in the induction of systemic resistance to sub-

sequent TMV infection. Together, the data strongly argue

that MeSA is an essential part of the SAR signal.

In the current study, we characterize a gene family from

Arabidopsis encoding proteins with significant sequence

similarity with tobacco SABP2. Of the 18 potentially func-

tional members of this family most encode proteins with

esterase activity against one or more biologically relevant

substrates, although substrate preference varies (see also

Yang et al., 2008). Therefore, we have named the genes

AtMES1–18 for methyl esterase 1–18. The biological signif-

icance of genes encoding proteins that prefer MeSA as a

substrate in vitro (this work) or methyl indole-3-acetic acid

(MeIAA) as a substrate (Yang et al., 2008) was assessed in

more detail. The results presented here provide support for

the idea that MeSA is an SAR signal that is conserved

between plant species.

Results

Arabidopsis encodes at least 18 potential SABP2 orthologs

Basic Local Alignment Search Tool (BLAST) analysis of the

tobacco SABP2 amino acid (aa) sequence against the Ara-

bidopsis protein database (http://www.arabidopsis.org/

Blast/index.jsp, TAIR8 proteins) resulted in the identification

of 20 genes from Arabidopsis, designated AtMES1–20, that

encode potential orthologs (defined here as functional

homologs) of SABP2 (Table 1). Although all 20 proteins dis-

played relatively high homology with SABP2 (score above

120 bites and e-value below 6e-28), AtMES19 and AtMES20

are truncated and lack one or more residues of the highly

Table 1 Salicylic acid-binding analysis of AtMES family membersand tobacco (Nt) SABP2

Protein Genetic locus

SA bindingrelative toAtMES9 (%)a

AtMES1 At2g23620 0AtMES2 At2g23600 0AtMES3 At2g23610 0AtMES4 At2g23580 4

AtMES5 At5g10300 0AtMES6 At2g23550 ndAtMES7 At2g23560 <1

AtMES8 At2g23590 ndAtMES9 At4g37150 100

AtMES10 At3g50440 0AtMES11 At3g29770 ndAtMES12 At4g09900 ndAtMES13 At1g26360 ndAtMES14 At1g33990 0AtMES15 At1g69240 ndAtMES16 At4g16690 0AtMES17 At3g10870 ndAtMES18 At5g58310 ndNtSABP2 250

nd, not determined.aBinding of SA by AtMES9 was arbitrarily set at 100%. Per centbinding activity is based on the SA-binding efficiency measured intriplicate of 15 lg of protein per assay from two independentpreparations of recombinant protein.

446 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

conserved catalytic triad required for enzymatic activity,

suggesting that they are non-functional (Forouhar et al.,

2005; Yang et al., 2008). Therefore, we have confined our

analysis here to AtMES1–18; their encoded proteins share

32–57% aa sequence identity and 46–71% aa sequence

similarity with SABP2. We constructed a phylogenetic tree

including SABP2 and AtMES1–18 using the default settings of

the web-based alignment tool MULTALIN (http://bioinfo.

genopole-toulouse.prd.fr/multalin/multalin.html), and divi-

ded the genes into five subfamilies (Figure 1) (Corpet, 1988):

AtMES1, -2, -5, and -8 cluster with SABP2 forming subfamily I,

AtMES3, -4, -6, -7, and -9 form subfamily II, and AtMES10

forms the single-member subfamily III. AtMES16–18 form

subfamily IV, including the MeIAA esterase AtMES17 (Yang

et al., 2008). Finally, AtMES11–15 contain N-terminal exten-

sions compared with SABP2 and the other AtMES proteins,

and consequently co-segregate as SABP2’s most distantly

related subfamily V. In Yang et al. (2008), we provide a more

detailed phylogenetic analysis of AtMES1–20 with SABP2

and related a/b fold hydrolases. In that analysis, subfamilies I,

II, and III are clustered in one clade of a phylogenetic tree with

SABP2, tomato methyl jasmonate esterase (MJE), and Rau-

volfia serpentina polyneuridine aldehyde esterase (subfam-

ily 1). Subfamilies IV and V as defined here correspond to

subfamilies 3 and 2, respectively, in Yang et al. (2008).

Five AtMES proteins display SA-inhibited MeSA

esterase activity

As an initial screen to delineate which AtMESs might be

functionally homologous to SABP2, 10 recombinant pro-

teins were tested for their ability to bind SA (Table 1).

N-terminal His6-tagged versions of the different AtMES

homologs and SABP2 were expressed in Escherichia coli,

affinity purified, and analyzed for SA binding. We tested all

AtMES members of subfamilies I–III (Figure 1) except

AtMES8, which did not yield a sufficient amount of soluble

protein in E. coli for the SA-binding assay. AtMES6 was also

excluded from this, as well as subsequent assays, because

available cDNA clones did not encode full-length AtMES6

and we could not detect the gene product by RT-PCR fol-

lowed by nested PCR. Only one AtMES protein from each of

subfamilies IV and V was tested for SA binding since these

subfamilies are more distantly related to SABP2. The results

of the SA-binding assay are summarized in Table 1. Three

members of subfamily II, AtMES4, -7, and -9 (Figure 1),

detectably bound SA. AtMES9 reproducibly bound SA

approximately 30–50% as efficiently as SABP2 under the

reaction conditions used, while AtMES4 and -7 bound SA

much less efficiently.

Subsequently, all recombinant proteins including AtMES5

and -8 were tested for their ability to hydrolyze MeSA, MeJA,

or MeIAA (Table 2 and Yang et al., 2008). AtMES8 cleaved

the synthetic ester p-nitrophenyl acetate, but had no activity

on any of the methylated phytohormones tested, while

AtMES5 did not have activity with any of the substrates.

Therefore, neither was characterized further. AtMES1, -2, -4,

-7, and -9 were able to hydrolyze MeSA. AtMES1 and -9

showed the highest hydrolytic activity, followed by AtMES4

and -7, with AtMES2 having substantially less activity

(Table 2). It should be noted that the MeSA esterase activity

of AtMES2 and -7 varied between assays, but generally was

lower than that of AtMES1 or -9, which hydrolyze MeSA with

Km values of 57.3 and 147.1 lM and Kcat values of 0.0124 and

0.0137 sec)1, respectively. Substrate preference for MeSA,

MeJA, or MeIAA was determined, since SABP2 hydrolyzes

these three substrates in vitro with a strong preference for

MeSA with an apparent Km of 8.6 lM and a Kcat of 0.45 sec)1

(Forouhar et al., 2005). A similar substrate preference was

found for AtMES1, -2, -4, and -9, while AtMES7 displayed a

slight preference for MeIAA over MeSA under the reaction

conditions used (Table 2).

Since the MeSA esterase activity of SABP2 is inhibited by

the reaction product SA, we assessed whether these five

AtMESs were similarly controlled by product feedback

inhibition. Although strong SA binding was detected only

for AtMES9 and was much weaker for AtMES4 and -7, all five

enzymes were efficiently inhibited when SA was added to

the esterase reaction mixture in equimolar amounts to

MeSA (Table 2). We compared the SA inhibition kinetics of

the strongest MeSA esterases from subfamilies I and II,

AtMES1 and -9, respectively (Figure S1). The Ki values for SA

of these two enzymes are similar at 8.3 and 8.6 lM for

AtMES1 and -9, respectively, and modestly lower than that

of SABP2 (16.4 lM, SWP and DFK, unpublished result).

Similar to inhibition of the MeSA esterase activity of SABP2,

SA competitively inhibits the MeSA esterase activity of

AtMES14AtMES12AtMES15

AtMES7

AtMES13AtMES11

AtMES4

AtMES9

AtMES16AtMES10

AtMES17AtMES18

AtMES3

AtMES6

AtMES5AtMES1

AtMES8AtMES2

SABP2

10 PAM

V

IV

III

II

I

Figure 1. Phylogenetic tree of AtMES1–18 and NtSABP2.

Methyl salicylate esterases in Arabidopsis 447

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

AtMES1. However, inhibition of AtMES9 by SA is either

irreversible or non-competitive. Possibly, binding of SA to

AtMES1 (and -2; subfamily I) was not detected under the

conditions used for the SA-binding assay due to more

transient SA-binding kinetics compared with that of AtMES9

and possibly the other members of subfamily II.

Expression of AtMES1, -7, and -9 is induced during

pathogen infection

In tobacco, SABP2 gene expression is induced after infection

of leaves with TMV (Kumar and Klessig, 2003). Previously,

expression profiling showed that AtMES9 expression is

induced upon infection of Arabidopsis leaves with Pseudo-

monas syringae (Glazebrook et al., 2003). Therefore, we

assessed AtMES1, -2, -7, and -9 gene expression in Col-0 wild

type (wt) plants before or after mock inoculation or inocula-

tion with P. syringae pv. tomato DC3000 (Pst) AvrRpm1.

AtMES4 was not analyzed because we were unable to

reproducibly visualize its expression before or after infection

by PCR followed by nested PCR, suggesting that this gene is

expressed at a very low level. The RT-PCR analysis revealed

that expression of AtMES1 and -9 was induced at both 1 and

3 days post-infection (dpi) with Pst AvrRpm1, similar to that

of the positive control, PR1 (Figure 2). Expression of AtMES7

was induced, but moderately and only at 1 dpi. By contrast,

AtMES2 expression was not induced.

AtMES1, -7, and -9 complement SAR deficiency in

SABP2-silenced tobacco

To establish if one or more of the AtMES genes are true

orthologs of SABP2 in planta, complementation analysis

for SAR development in SAR-deficient SABP2-silenced to-

bacco (Kumar and Klessig, 2003) was done. We selected

the two strongest candidate SABP2 orthologs, AtMES1 and

-9, for complementation analysis, as well as AtMES7.

Previously, we showed that conditional expression of a

10 mM

MgCl2Pst

AvrRPM1

0 1 3 1 3 d.p.i.

AtMES1

AtMES2

AtMES7

AtMES9

PR1

Actin

Figure 2. Expression of AtMES1, -7, and -9 expression is induced following

infection by Pseudomonas syringae pv. tomato DC3000 (Pst) AvrRpm1.

Samples were taken at 0, 1 and 3 days after infiltration (dpi) of 4- to 5-week-old

Col-0 plants with 10 mM MgCl2 or with 105 colony-forming units ml)1 of Pst

AvrRpm1 in 10 mM MgCl2. Transcripts were amplified from total RNA

preparations by RT-PCR using gene-specific primers.

Table 2 Methyl salicylate (MeSA) esteraseactivity, substrate preference, and productinhibition of AtMES1, -2, -4, -7, and -9

ProteinMeSA hydrolysis(nmol min)1 lg)1)a

Substrate preference[conversion per substraterelative to MeSA (%)]b

Inhibition of MeSAesterase by SAcMeSA MeJA MeIAA

AtMES1 28.92 � 4.73 100 8 2 +++AtMES2 1.77 � 4.24 100 14 8 +++AtMES4 16.07 � 4.75 100 5 29 +++AtMES7 12.82 � 12.2 100 0 123 +++AtMES9 25.88 � 5.71 100 30 48 +++

SA, salicylic acid; MeJA, methyl jasmonic acid; MeIAA, methyl indole-3-acetic acid.aHydrolysis of MeSA was measured using 1.5 lg of protein per assay with 10 lM of MeSA. Datafrom three independent experiments (of two or three replicates each) were pooled to obtain theaverage and standard deviation values indicated here.bThe conversion of MeJA and MeIAA is indicated relative to that of MeSA using similar proteinand substrate concentrations per assay. Hydrolysis of MeSA was set at 100% per protein. Thedata from three independent experiments were pooled to obtain the indicated average relativeconversions except for the relative conversions of MeJA and MeIAA by AtMES4, which wereobtained from two independent experiments.cInhibition of MeSA esterase activity was determined in the presence of equimolar amounts of SAand MeSA.

448 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

synthetic version of SABP2 complemented the SAR-defi-

cient phenotype of SABP2-silenced tobacco plants (Fig-

ure 3c; Kumar et al., 2006). SABP2syn2 encodes wt SABP2

protein from a transcript that is modified such that it is no

longer subject to silencing by the SABP2-silencing trans-

gene. To avoid silencing of AtMES1 and -7 by the SABP2-

silencing transgene, the cDNA sequences of both genes

were modified by PCR to eliminate stretches of perfect

identity of 12 nucleotides or longer (Figure S2). The

resulting AtMES1mod and AtMES7mod sequences share

61.6% and 55.9% nucleotide identity with SABP2, respec-

tively, and encode wt AtMES1 and AtMES7 proteins. A

synthetic version of AtMES9 (AtMES9syn) was used that

shares 44.5% nucleotide identity with SABP2, does not

contain stretches of perfect sequence identity with SABP2

of more than five nucleotides, and encodes wt AtMES9

protein (Figure S2).

Transgenes containing AtMES1mod, AtMES7mod,

AtMES9syn, or SABP2syn2 driven by the b-estradiol-induc-

ible XVE promoter were expressed in the systemic leaves of

SABP2-silenced tobacco by estradiol treatment 1 day after

the primary (1�) infection. SABP2-silenced plants carrying

XVE::SABP2nat, encoding native SABP2, were included as a

negative control for complementation since this transcript is

targeted for silencing by the SABP2-silencing transgene

(Figure 3a; Kumar et al., 2006). Although the complement-

ing transgenes are under the control of inducible promoters,

varying degrees of leaky expression were observed at the

RNA level in all lines before estradiol treatment [Figure 3a;

upper panels, 0 days post (d.p.)-estradiol]. With the excep-

tion of SABP2nat, induced expression of all XVE-driven

transgenes was evident after treatment of the leaves with

estradiol (Figure 3a). Importantly, expression of endoge-

nous SABP2 was efficiently silenced in all transgenic plants

generated in the SABP2-silenced background compared

with control plants that carry only the empty silencing vector

(Figure 3a, wild type; Kumar and Klessig, 2003).

Systemic acquired resistance was monitored in control

plants and SABP2-silenced lines expressing the various

transgenes by comparing TMV-induced 1� and secondary

(2�) lesion size. Control (wt) plants mounted a successful

SAR response, as indicated by a reduction in lesion size of

(a)

(b)

(c)

Figure 3. AtMES1, -7 and -9 complement sys-

temic acquired resistance (SAR) deficiency in

SABP2-silenced tobacco.

(a) SABP2nat, AtMES1mod, AtMES7mod and

AtMES9syn were expressed in SABP2-silenced

tobacco from estradiol-inducible transgenes.

Transgene-specific (upper panels), endogenous

SABP2 (middle panels), and EF1a (lower panels;

equal loading control) transcript levels were

assessed by RT-PCR using (trans)gene-specific

primers. Samples were taken from the primary

(1�) inoculated leaves at 0 days post (dp)-estra-

diol treatment of the systemic tissue, and from

the estradiol-treated systemic tissue at 1 and

6 dp estradiol treatment.

(b,c) Lesion sizes were measured 5 days after

either 1� or secondary (2�) TMV infection and the

mean � SD were derived from 25 (b) or 30 (c)

lesions per leaf. Part (b) is derived from the same

experiment as (a). Pictures show representative

leaf areas 5 dp 2� infection.

Methyl salicylate esterases in Arabidopsis 449

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

approximately 50% between the 1� and 2� infection

(Figure 3b). A modest reduction in lesion size was repro-

ducibly observed in SABP2-silenced plants carrying XVE::

SABP2nat (Figure 3b,c), possibly due to residual expression

of the transgene as detected by RT-PCR using transgene-

specific primers (Figure 3a, upper panel). By contrast,

AtMES1mod, AtMES7mod, or AtMES9syn fully comple-

mented SAR deficiency in SABP2-silenced plants, as indi-

cated by a 40–50% reduction in lesion size (Figure 3b). In an

independent experiment, AtMES7mod or AtMES9syn com-

plemented the SAR-deficient phenotype of SABP2-silenced

plants to comparable levels as the synthetic transgene

XVE::SABP2syn2 (Figure 3c). It should be noted that some

variability occurred between experiments. Whereas com-

plementation was always observed, the strength of the SAR

response inversely correlated with the level of leaky trans-

gene expression in the 1� inoculated leaves. It is possible

that heightened levels of MeSA esterase activity in the 1�inoculated leaves prevented the accumulation of sufficient

levels of MeSA for efficient signaling of SAR. Regardless,

these complementation data argue that AtMES1, -7, and -9

can function as SABP2 in planta.

Underexpression of AtMESs enhances systemic MeSA

accumulation during SAR

To assess the biological significance of AtMES-encoded

MeSA esterases in Arabidopsis, we obtained T-DNA inser-

tion knockout (KO) lines of AtMES1 and -9 from the SALK

library. M1/9 double KO lines were generated, and RT-PCR

with gene-specific primers confirmed the absence of

AtMES1 or AtMES9 transcripts due to the M1KO or M9KO

mutation, respectively (Figure 4b). Expression of the other

three MeSA esterase-encoding AtMESs, which cluster with

AtMES1 on chromosome 2, was downregulated in the M1/9

KO or Col-0 wt backgrounds by using RNAi technology. First,

the full-length AtMES7 sequence was targeted for silencing

using construct RNAi::M7 in the M1/9 KO background

(Figure 4a); a reduction of the AtMES7 expression level was

observed in this methyl esterase deficient 1-1 line (med1-1;

Figure 4b). Second, AtMES2 and AtMES7 were targeted for

silencing in both the M1/9 KO and the Col-0 wt backgrounds

with a single RNAi construct, RNAi::M2_7 (Figure 4a). Again,

expression of AtMES7 was reduced but not abolished in

med2-2 and med3-1, whereas silencing of AtMES2 was most

efficient in med3-1 and variable, but generally less effective,

in med2-2 (Figure 4b). Due to significant sequence homol-

ogy in the 3¢ half of the AtMES3 transcript and the M7 frag-

ment of the RNAi::M2_7 construct, AtMES3 was co-silenced

in the Col-0 wt background (med3-1; Figure 4b). In addition,

AtMES8, which is a close homolog of AtMES2, was effi-

ciently silenced in all RNAi::M2_7 lines, and AtMES1, which

shares significant sequence homology with the M2 fragment

of the RNAi::M2_7 construct, was efficiently silenced in the

med3-1 mutant (Figure 4b).

Any phenotypes observed in the above-described trans-

genic lines (except med3-1) would be determined by three

genetic loci (two T-DNA insertions and the RNAi transgene).

In an attempt to construct a genetic tool to study the MeSA–

SA equilibrium in Arabidopsis, nine AtMES genes were

targeted with a single RNAi construct, RNAi::combo_9, in the

Col-0 wt background (Figure 4a). The RNAi::combo_9 silenc-

ing sequence contains ten 100-nucleotide (nt) fragments that

target all of the AtMES genes encoding esterases displaying

discernible hydrolytic activity on a high concentration of

MeSA in vitro (5 mM; AtMES1, -2, -3, -4, -7, -9, and -16). In

(a)

(c)

(b) Figure 4. Underexpression of AtMES genes

enhances accumulation of methyl salicylate

(MeSA) in Arabidopsis.

(a) Schematic representation of silencing con-

structs. Numbers indicate the nucleotides that

were used from each open reading frame.

(b) Silencing of target genes was analyzed by RT-

PCR using gene-specific primers. Samples were

taken from 5-week-old plants in the homozygous

T3 generation.

(c) The MeSA was quantified in systemic tissue

of 4-week-old plants at 0, 12, and 24 h after a

primary infection with coronatine-deficient Pseu-

domonas syringae pv. maculicola DC3000 (Psm)

AvrRpt2. Analyses were done in triplicate, and

the mean � SD values are presented.

450 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

addition, fragments were included to silence genes that

display significant sequence homology with active MeSA

esterases despite encoding proteins without detectable

MeSA esterase activity in vitro (AtMES5 and -8). In general,

silencing of AtMES2, -3, and -8 was efficient, whereas

AtMES7 was variably silenced in two independent trans-

genic lines (Figure 4b, med4-1 and med4-2). The expression

levels of AtMES1 and AtMES9 varied between individual

plants of the same mutant. In the samples used to generate

Figure 4(b), AtMES9 was silenced fairly efficiently, whereas

expression of AtMES1 in the med4-1 line resembled that in

Col-0 wt plants and expression in the med4-2 line was highly

elevated. Despite this variability, the overall expression of

AtMES genes encoding or resembling MeSA esterases was

reduced in med4-1 and med4-2 as compared to wt.

Since MeSA in tobacco is hydrolyzed systemically by the

MeSA esterase activity of SABP2 to trigger SAR, we quan-

tified MeSA levels in uninoculated systemic tissue of plants

that had been SAR-induced by a 1� infection with coronatine-

deficient P. syringae pv. maculicola (Psm) AvrRpt2 (Cui

et al., 2005). The level of MeSA increased slightly in systemic

tissue of Col-0 wt and M1/9 KO plants during SAR induction,

with levels being significantly higher at 24 h after infection

(hpi) as compared with 0 hpi (Figure 4c and Figure S3).

Levels of MeSA were reproducibly elevated compared with

those in Col-0 wt plants in systemic tissues of med2-2, med3-

1, and med4-1 before and after SAR induction (Figure 4c and

Figure S3), indicating that conversion of MeSA to SA was

compromised in these mutants in spite of residual expres-

sion of either AtMES2, AtMES9, or AtMES1 in med2-2,

med3-1, and the med4 lines, respectively. Taken together,

the data show that the AtMES genes are functionally

redundant and involved in MeSA hydrolysis in Arabidopsis.

Underexpression of AtMESs compromises SAR

in Arabidopsis

Since SABP2-silenced tobacco plants appear somewhat

more susceptible than wt to TMV infection (Kumar and

Klessig, 2003) and virulent P. syringae pv. tabaci (unpub-

lished result, E. Kaimoyo [Boyce Thompson Institute for

Plant Research, Ithaca, NY], DK, and DFK), R gene-mediated

resistance and basal resistance were analyzed in the mutant

lines described above. R gene-mediated resistance in inoc-

ulated leaves was not affected in the various lines, as similar

titers of Pst AvrRpm1 were detected in inoculated leaves of

Col-0 wt, med2-2, and med4-1 plants (Figure 5a and data not

shown). No difference was observed in growth of virulent

Pst in any of the lines, indicating that underexpression of

AtMES genes does not affect basal resistance to pathogens

(Figure 5b,c; compare Pst growth in uninduced plants,

i.e. black bar).

Since SAR is completely abolished in SABP2-silenced

tobacco, we analyzed the ability of the various mutants to

1

2

3

4

5

6

Pst

Avr

Rp

m1

(lo

g c

fu c

m–2

)

5

6

7

2º in

ocu

lum

gro

wth

(lo

g c

fu p

er le

af d

isc)

40 dpi 2 dpi M1/9KOwt M1KO M9KO

med4-1

wt

(a)

(c)

(b)

med2-2

3

4

5

6

7

2nd

ino

culu

m g

row

th(l

og

cfu

cm

–2 o

f le

af)

wt med1-1 med2-2 med3-1 med4-1 med4-2P = 0.014 0.015 0.647 0.859 0.081< 0.001

Figure 5. Underexpression of AtMES genes

compromises systemic acquired resistance

(SAR), but not local resistance.

(a) In planta Pseudomonas syringae pv. tomato

DC3000 (Pst) AvrRpm1 titers were determined in

triplicate 0 and 2 days after infection of 4- to 5-

week-old plants, and the mean and SD values are

represented on a logarithmic scale.

(b, c) SAR was analyzed in untreated (b, black

filled bars) or 10 mM MgCl2-treated (c, black filled

bars) versus coronatine-deficient Psm AvrRpt2-

infected (b, c, striped bars) plants. The Pst titers

in the systemic challenged tissue were deter-

mined in triplicate at 3 days post-inoculation

(dpi), and the mean and SD values are presented

on logarithmic scales. P, Student’s t-test proba-

bility value of equal Pst growth in mock-treated

and SAR-induced plants.

Methyl salicylate esterases in Arabidopsis 451

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

mount a SAR response. First, SAR was analyzed in M1KO,

M9KO, and M1/9 KO plants by monitoring growth of Pst in

systemic leaves of induced plants, which had received a 1�infection with coronatine-deficient Psm AvrRpt2, as com-

pared with untreated plants (Figure 5b). All three mutants

mounted a successful SAR response consistent with wt-like

levels of MeSA accumulation in the systemic tissue of M1/9

KO plants (Figure 4c). Second, Pst growth was significantly

reduced in SAR-induced versus mock-treated med1-1 plants,

indicating that SAR was activated and that a reduction of the

expression of AtMES7 in the M1/9 KO background is not

sufficient to compromise SAR (Figure 5c). However, in more

than 10 experiments in three different locations/countries,

SAR was compromised approximately 50% of times in

med2-2, med3-1, med4-1, and med4-2. A representative

experiment is shown in Figure 5(c). In this experiment,

inhibition of SAR correlated with the absence of PR1

transcripts in systemic tissue of SAR-induced med3-1 and

med4-1 (Figure S4).

A weaker SAR response in Col-0 wt was associated with

SAR deficiency in med2, med3, and med4 lines (Figure S5).

When a very robust SAR response was observed (100-fold

reduction in Pst growth), incomplete AtMES gene silencing

(e.g. med2-2 in Figure 5c and data not shown) was not

sufficient to inhibit SAR. In numerous experiments SAR was

less compromised in med4-2 than in med4-1 (Figure 5c); this

correlates with the expression level of AtMES1 (Figure 4b).

Similarly, the silencing efficiency of AtMES2 may help

determine the SAR status of med2-2. AtMES2 has a very

low MeSA esterase activity in vitro and its expression is

not induced by pathogen challenge, but it is nonetheless

the most abundantly expressed MeSA esterase-encoding

AtMES gene in planta, as detected by northern analysis

(G. Wang [University of Michigan, Ann Arbor, MI] and EP,

unpublished result) and by the relatively low number of PCR

cycles required to visualize the transcript by RT-PCR (Fig-

ure 4b). Taken together, SAR is at least partially impaired by

inhibition of MeSA hydrolysis upon silencing of AtMESs,

particularly those encoding MeSA esterases including

AtMES1, -2, -7, and -9.

Discussion

In this study, we have characterized a new gene/protein

family from Arabidopsis based on homology with SABP2,

the tobacco gene encoding a protein with MeSA esterase

activity. At least five AtMES proteins, AtMES1, -2, -4, -7, and

-9, displayed SA-inhibited MeSA esterase activity in vitro

(Table 2) suggesting that they could perform a biochemical

function similar to that of SABP2 in the perception of the

MeSA defense signal in Arabidopsis (Forouhar et al., 2005;

Park et al., 2007). Expression of AtMES1, AtMES9, and to a

lesser degree AtMES7, was also induced by pathogen

infection (Figure 2). Possible functional homology between

SABP2 and AtMES1, -7, and -9 was confirmed by the

complementation of SAR deficiency in SABP2-silenced to-

bacco (Figure 3). Moreover, underexpression of AtMES

genes, including those that encode functional MeSA ester-

ases, compromised SAR (Figure 5 and Figure S5); this

correlated reasonably well with elevated MeSA levels in the

systemic tissue of SAR-induced plants (Figure 4 and Fig-

ure S3). Silencing of AtMES family members was increas-

ingly less efficient when more MeSA esterase-encoding

genes were targeted (Figure 4 and data not shown). In

many independent med4 lines, for instance, expression of

one or more MeSA esterases was elevated above wt levels

when silencing of others was highly efficient (as illustrated

in this study for AtMES1 in med4-2). Also, silencing of At-

MES2 was much less efficient in the med2 lines, in which

AtMES1 and -9 are KO, compared to that in the med3 lines,

in which AtMES1 and -9 are wt. These results underscore

the difficulty in assessing gene function when there is a lot

of redundancy as well as a possible requirement for plant

viability of the activity under study (in this case MeSA

esterase activity).

Several AtMES proteins other than the five MeSA

esterases characterized in this study display esterase

activity in vitro, but their substrate preference is for other

methylated phytohormones (Yang et al., 2008; YY, ACV,

EP, and DFK, unpublished data). For instance, AtMES3, -10,

-16, -17, and -18 preferred MeIAA as a substrate over MeSA

(Yang et al., 2008). Several AtMES proteins were able to

hydrolyze MeJA in vitro even though this compound was

not their preferred substrate, with the possible exception of

AtMES10. Additionally, some AtMES proteins, including

the MeSA esterases AtMES7 and -9, displayed significant

hydrolytic activity towards two or all three of the substrates

tested (Table 2). Although other methylated substrates

remain to be tested, it is already apparent that members

of the AtMES gene family are involved in multiple biolog-

ical processes. Perhaps differential expression of the

different AtMES genes combined with their varying sub-

strate preferences helps modulate a balance among

methylated plant hormones to regulate plant growth,

development, and responses to stresses including patho-

gen attack.

Previous studies have indicated a role for methylation

and/or demethylation of MeSA in plant defense. The

AtBSMT1 gene, which encodes a protein capable of methy-

lating both benzoic acid and SA, was induced upon wound-

ing, uprooting, MeJA treatment, or insect herbivory (Chen

et al., 2003; Koo et al., 2007). A similar finding was reported

in tomato, where JA-induced expression of LeSAMT pre-

ceded emission of MeSA vapor into the plant headspace

(Ament et al., 2004). In the latter study, JA-induced produc-

tion of MeSA and other volatile compounds positively

correlated with the attraction of predatory mites to spider

mite-infected plants, thus affecting indirect defense of

452 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

tomato against insects. Interestingly, elevated MeSA levels

in the headspace of OsBSMT1-overexpressing Arabidopsis

induced SA-dependent defense gene expression in neigh-

boring non-transgenic plants (Koo et al., 2007). This

confirms earlier observations that MeSA can serve as an

airborne defense signal in plants (Shulaev et al., 1997). Since

enhanced conversion of SA to MeSA in OsBSMT1-express-

ing Arabidopsis virtually abolished SA-dependent defenses

and since MeSA itself does not induce resistance or defense

gene expression in SA-degrading NahG tobacco, it is highly

likely that MeSA itself is biologically inactive and needs to be

converted to SA for biological activity (Koo et al., 2007;

Seskar et al., 1998). We propose that SABP2, its Arabidopsis

orthologs AtMES1, -2, -4, -7, and -9, and as yet unidentified

orthologs in other plant species, mediate hydrolysis of both

airborne and phloem-transported MeSA to initiate defense

signaling.

Growth of virulent or avirulent Pst was comparable in the

primary inoculated leaves of wt plants and all of the mutant

lines studied (Figure 5 and data not shown). This result

suggests that MeSA esterases in Arabidopsis may play a

lesser role in basal or R gene-mediated local resistance to P.

syringae than in SAR. However, basal resistance is only

moderately compromised in SABP2-silenced tobacco, in

which SAR is more efficiently inhibited than in the med

mutants (E. Kaimoyo, DK, and DFK, unpublished result).

Suppression of PR1 gene expression and, to a lesser extent,

N gene-mediated resistance was only detected in tobacco

plants strongly silenced for SABP2, while the SAR-deficient

phenotype was more robust and appeared less affected by

modest changes in SABP2-silencing effectiveness (Kumar

and Klessig, 2003). Thus, different threshold levels of MeSA

esterase activity may be required for different defense

responses, with SAR requiring very high levels, R gene-

mediated resistance requiring intermediate levels, and PR1

expression requiring only moderate levels.

Since underexpression of AtMES genes encoding MeSA

esterases did not compromise local resistance in the

primary infected leaves, which generate the SAR signal,

the SAR-deficient phenotypes shown in Figure 5 and

Figure S5 are probably a consequence of compromised

reception of the signal in the distal systemic tissues. The

documented role of SABP2 in perception of the SAR signal

in systemic tissues (Park et al., 2007) is further confirmed

here by complementation of SAR deficiency in SABP2-

silenced plants by expression of AtMES1, -7, or -9 in the

distal, SAR signal-perceiving leaves of infected plants.

Moreover, leaky uninduced expression of these estradiol-

inducible AtMES transgenes in the 1� infected tissue of

SABP2-silenced tobacco appeared to inversely correlate

with SAR development. Together, the data argue that,

similar to SABP2, at least one function of the AtMES MeSA

esterases is perception of a long-distance SAR signal in

systemic tissues of infected plants. However, because the

extent of functional redundancy is very large and the

correlation between elevated MeSA levels in systemic tissue

and loss of SAR was not complete, we cannot rule out the

possibility that the roles of these five AtMES proteins and

MeSA in SAR may be somewhat different in Arabidopsis

compared with tobacco.

Truman et al. (2007) proposed an early function for JA in

SAR signaling. Since SABP2 is capable of hydrolyzing MeJA

in vitro, although with low specific activity (Forouhar et al.,

2005), one might argue that SABP2-mediated hydrolysis of

MeJA could account for the involvement of both JA and

SABP2 enzymatic activity in SAR. However, this possibility

seems unlikely since expression of AtMES7, which is

incapable of hydrolyzing MeJA in vitro, complements SAR

deficiency in SABP2-silenced tobacco. The possible involve-

ment of both JA and MeSA in SAR probably does not involve

an antagonistic relationship between JA and SA during

perception of the SAR signal(s). Truman et al. (2007)

suggested that the two signals could be temporally sepa-

rated. Another possibility is that the relative concentrations

of JA and SA during SAR signal perception and relay are

within the range supporting cooperativity (Mur et al., 2006).

The data presented here and in Park et al. (2007) argue that

MeSA is an essential part of SAR signaling in plants.

However, many more factors could be directly or indirectly

involved in the generation or perception of the signal(s) as

illustrated by the increasing number of studies implicating

lipids, including JA, proteins, NO, and other small mole-

cules, in systemic defense signaling in plants (reviewed in

Vlot et al., 2008).

In summary, we have identified a new gene family in

Arabidopsis some of whose members perform a similar

biochemical function to SABP2 in the hydrolysis of MeSA

and probably also in SAR. The data argue that MeSA may

function as a signal for SAR in Arabidopsis as well as in

tobacco, and therefore that this SAR signal is conserved

between members of the solanaceous and cruciferous plant

families. Since the phenotypes of med3-1, med4-1, and

med4-2 depend on a single genetic locus, these mutant lines

may prove to be useful genetic tools to further dissect the

biological role(s) of MeSA and its hydrolysis in planta.

Experimental procedures

Plant material and growth

Nicotiana tabacum (tobacco) empty vector control line C3 wasgrown as described (Kumar and Klessig, 2003). SABP2-silencedtobacco lines carrying XVE::SABP2nat or XVE::SABP2syn2 weredescribed by Kumar et al. (2006).

Arabidopsis thaliana ecotype Colombia-0 (Col-0) was used forall experiments and grown in a 10-h photoperiod (140 lE m)2

sec)1) at 22�C in 60–80% relative humidity. Knockout mutants,obtained from the Arabidopsis Biological Resource Center (ABRC,Ohio State University, OH, USA), contained transfer-DNA inser-

Methyl salicylate esterases in Arabidopsis 453

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

tions in At2g23620 (SALK_006044; M1KO) and At4g37150(SALK_030442; M9KO). Homozygous KO plants were selected byPCR amplification using the T-DNA-specific pLB primer and gene-specific primers p044-LP and p044-RP for the M1KO, and p442-LPand p442-RP for the M9KO (Table S1). Knockout of gene expres-sion in the KO lines was confirmed by RT-PCR using pMES1-5¢ andpMES1-3¢ for the M1KO and pMES9-5¢ and pMES9-3¢ for the M9KO(Table S1). M1/9 KO plants were generated by crossing M1KOand M9KO plants and selecting double homozygous plants in theF2 generation.

DNA constructs

For expression of N-terminally 6-histidine (His6-)tagged recom-binant proteins, AtMES-encoding DNAs (cDNAs) were cloned intothe E. coli expression vector pET28a (Novagen, http://splash.emdbiosciences.com/). AtMES DNAs were PCR-amplified as out-lined in Table S1. AtMES1 was amplified from cDNA clone 194A22(ABRC), and cloned into pET28a using BamHI. AtMES9 was ampli-fied from cDNA clone SQ089h02F (Kazusa DNA Research Institute,Kisarazu, Japan), and cloned into pET28a using EcoRI and XhoI. Allother cDNAs for expression were obtained via RT-PCR on Col-0 totalRNA isolated 3 days after infection of the leaves with Pst AvrRpm1.AtMES2, AtMES4, AtMES5, AtMES7, AtMES8, AtMES10, AtMES14,AtMES16, and AtMES17 were each cloned into pET28a using EcoRI.AtMES3 was cloned into pET28a using SacI. All PCR products weresequenced after cloning. Construction of pET28a-SABP2 has beendescribed (Kumar and Klessig, 2003).

For expression of AtMES1, -7, or -9 in SABP2-silenced tobacco,modified or synthetic transcripts were cloned in the binary vectorpER8 (Zuo et al., 2000). The sequences of all PCR primers used areshown in Table S1. To generate AtMES1mod, AtMES1 DNA frag-ments were amplified from pET28a-AtMES1 with pMES1mod.11and .4 (PCR 114), pMES1mod.5 and .6 (PCR 56), pMES1mod.7 and .8(PCR 78), and pMES1mod.9 and .10 (PCR 910). Gel-purified PCRfragments were annealed and amplified by PCR as follows: PCR 114and 56 with pMES1mod.1 and .6 (PCR 16), and PCR 78 and 910 withpMES1mod.7 and .10 (PCR 710). Gel-purified PCR products 16 and710 were annealed and amplified by PCR with pMES1mod.1 and .10to yield AtMES1mod. To generate AtMES7mod, two AtMES7 DNAfragments were amplified from pET28a-AtMES7 using pMES7-mod.1 and .2 and pMES7mod.3 and .4, respectively. Thegel-purified PCR fragments were annealed and amplified by PCRwith pMES7mod.1 and .4 to yield AtMES7mod. Both PCR frag-ments were cloned into pGEMT-easy (Promega, http://www.promega.com/) and sequenced. A cDNA encoding AtMES9(AtMES9syn) was designed and synthesized by DNA2.0 Inc.(http://www.dna20.com/) as described (Kumar et al., 2006).AtMES1mod and AtMES7mod were cloned into pER8 using ApaI,and AtMES9syn was cloned into pER8 using XhoI.

The RNAi constructs were made in pHANNIBAL (Wesley et al.,2001). Full-length AtMES7 was amplified from pET28a-AtMES7 withpMES7HAN-fwd and pMES7HAN-rev (M7). A 5¢ fragment of AtMES2was amplified from pET28a-AtMES2 with pMES2_MES7_1 and _2. A3¢ fragment of AtMES7 was amplified from pET28a-AtMES7 withpMES2_MES7_3 and _4, and the two fragments were annealed andamplified by PCR with pMES2_MES7_1 and _4 (M2_7). M7 and M2_7were each cloned into pGEMT-easy, sequenced, and cloned intopHANNIBAL in sense and antisense orientations using EcoRI andClaI, respectively. The RNAi cassettes were excised from pHANNI-BAL with NotI, blunt ended with T4 DNA polymerase, and insertedinto SmaI and blunt ended KpnI sites of the binary vector pGreenII-61 (Hellens et al., 2000). The silencing sequence combo_9 wasgenerated using partially complementary primers to merge 100

nucleotide fragments from nine AtMES genes. The fragments weregenerated by PCR as outlined in Table S1. Gel-purified PCRfragments were annealed and amplified by two or three insubsequent PCR reactions using the complementarity betweenprimers. All 10 fragments were fused in a total of four rounds of PCRamplification. The resulting combo_9 silencing sequence wascloned into pGEMT-easy, sequenced, and cloned into pHANNIBALin sense and antisense orientations using EcoRI and XbaI, respec-tively. The combo_9 RNAi cassette from pHANNIBAL was trans-ferred to the binary vector pART27 using NotI.

Protein expression and assays

The pET28a-AtMES clones were transformed into E. coli strain BL21for protein expression. His6-tagged recombinant proteins wereaffinity purified using Ni2+-NTA resin (Novagen) according to themanufacturer’s instructions except that protein expression wasinduced overnight at 18�C using 0.1 mM isopropyl b-D-thiogalacto-side. The SA binding was measured as described using 15 lg ofprotein per assay (Slaymaker et al., 2002). The MeSA, MeJA, andMeIAA esterase assays were performed as described using 1.5 lg ofprotein and 10 lM of substrate per assay (Forouhar et al., 2005).

Plant transformation

pER8 constructs were transformed into A. tumefaciens strainLBA4404 for transformation of SABP2-silenced tobacco line 1-2(Kumar and Klessig, 2003). Transformation and regeneration oftobacco was performed as described (Shah and Klessig, 1996),and transgenic plants were maintained in culture on MS medium(Gibco, http://www.invitrogen.com/) containing 10 lg ml)1 hygro-mycin (Sigma, http://www.sigmaaldrich.com/) to select for thepER8-derived transgenes.

pGreenII-61 constructs were transformed into A. tumefaciensstrain GV3101 carrying pSOUP, and pART-RNAi::combo_9 wastransformed into A. tumefaciens strain GV3101 carrying pMP90. TheM7 silencing cassette was transformed as described (Clough andBent, 1998) into M1/9 KO plants, and the M2_7 silencing cassetteinto Col-0 and M1/9 KO plants. Transgenic T1 plants were selectedby spraying vernalized seeds with 50 lM of phosphinotricin (Finale)before germination. The combo_9 silencing cassette was trans-formed into Col-0 wt plants, and transgenic T1 plants were selectedon MS medium containing 50 lg ml)1 kanamycin. Homozygous T3

transgenic plants of selected lines were used for the experimentsdescribed here.

Plant pathogen experiments

Complementation of SAR in tobacco was analyzed essentially asdescribed (Kumar et al., 2006). Systemic acquired resistance wasinduced in tobacco plants 6 weeks after sowing (empty vectorcontrol line) or transfer from MS medium to soil (all other transgenicplants) by infection of three lower leaves per plant with TMV. Oneday after infection, two to three upper leaves per plant were sprayedwith 30 lM b-estradiol in 0.01% Tween-20 for induction of transgeneexpression. Seven days after the 1� TMV infection, the estradiol-treated leaves were challenged by a 2� TMV infection. Lesion sizeswere measured with a vernier caliper 5 days after 1� or 2� infection.

All P. syringae strains used in this study have been previouslydescribed (Aarts et al., 1998; Cui et al., 2005). Pst carrying AvrRpm1or the corresponding empty vector pVSP61 were grown on King’s B(KB) medium (King et al., 1954; Tornero and Dangl, 2001), and

454 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

infiltrated with a syringe into leaves of 4- to 5-week-old plants at aconcentration of 105 colony forming units (cfu) ml)1 in 10 mM

MgCl2. Mock inoculation was performed using 10 mM MgCl2.In planta bacterial titers were determined by shaking leaf dics frominfected leaves in 10 mM MgCl2 supplemented with 0.01% Silwet-L77 at 28�C for 1 h (Tornero and Dangl, 2001). The resulting bacterialsuspensions were serially diluted and spots of 20 ll per dilutionwere grown on KB medium and counted.

Pathogen propagation and pathogen infections for SAR assayswere performed as described (Maldonado et al., 2002) except that106 cfu ml)1 coronatine-deficient Psm AvrRpt2 was used for the 1�inoculation (Cui et al., 2005) and 105 cfu ml)1 Pst was used for the 2�inoculation. The leaf discs used for quantification of secondarypathogen growth in Figure 5(b) and Figure S5 were 4 mm indiameter.

MeSA quantification

Methyl SA was extracted and quantified as described from 110 mgof tissue per sample (Park et al., 2007).

RT-PCR

Total RNA from leaves was extracted using TRIZOL reagent (Invi-trogen, http://www.invitrogen.com/) according to the manufac-turer’s instructions. After treatment with RNase-free DNase (RQ1;Promega), the RNAs were again extracted with TRIZOL and used forRT reactions (SuperscriptII; Invitrogen). The PCRs were performedusing primers outlined in Table S1.

Acknowledgements

We thank Drs Naomi E. Pierce and Georg Jander for the coronatine-deficient Psm AvrRpt2 strain, Drs Melanie A. Sacco and PeterMoffett for pGreenII-61 and GV3101 carrying pSOUP, Dr D’MarisDempsey for critically reading the manuscript, and Drs Hong GuKang, Melanie A. Sacco, and Peter Moffett for helpful discussions.This work was funded in part by the Netherlands Organization forScientific Research (NWO; fellowship to ACV), and by NationalScience Foundation grant IOS-0525360 to DFK.

Supporting Information

Additional Supporting Information may be found in the onlineversion of this article:Figure S1. Determination of enzymatic parameters of recombinantAtMES1 and AtMES9.Figure S2. DNA sequence alignment of SABP2 with (a) AtMES1modand AtMES1, (b) AtMES7mod and AtMES7, and (c) AtMES9syn andAtMES9.Figure S3. Accumulation of MeSA in the systemic tissue of systemicacquired resistance-induced AtMES underexpressing plants ascompared with wild-type plants.Figure S4. PR1 expression in the systemic tissue of wild-type andmed mutant plants upon induction of systemic acquired resistance.Figure S5. Development of systemic acquired resistance in medmutants compared with wild-type plants.Table S1. Oligonucleotides used in this study.Please note: Wiley-Blackwell are not responsible for the content orfunctionality of any supporting materials supplied by the authors.Any queries (other than missing material) should be directed to thecorresponding author for the article.

References

Aarts, N., Metz, M., Holub, E., Staskawicz, B.J., Daniels, M.J. and

Parker, J.E. (1998) Different requirements for EDS1 and NDR1 bydisease resistance genes define at least two R gene-mediatedsignaling pathways in Arabidopsis. Proc. Natl Acad. Sci. USA, 95,10306–10311.

Ament, K., Kant, M.R., Sabelis, M.W., Haring, M.A. and Schuurink,

R.C. (2004) Jasmonic acid is a key regulator of spider mite-induced volatile terpenoid and methyl salicylate emission intomato. Plant Physiol. 135, 2025–2037.

Buhot, N., Gomes, E., Milat, M.-L., Ponchet, M., Marion, D., Lequeu,

J., Delrot, S., Coutos-Thevenot, P. and Blein, J.-P. (2004) Modu-lation of the biological activity of a tobacco LTP1 by lipid com-plexation. Mol. Biol. Cell, 15, 5047–5052.

Cameron, R.K., Dixon, R.A. and Lamb, C.J. (1994) Biologicallyinduced systemic acquired resistance in Arabidopsis thaliana.Plant J. 5, 715–725.

Chaturvedi, R., Krothapalli, K., Makandar, R., Nandi, A., Sparks,

A.A., Roth, M.R., Welti, R. and Shah, J. (2008) Plastid x3-fatty aciddesaturase-dependent accumulation of a systemic acquiredresistance inducing activity in petiole exudates of Arabidopsisthaliana is independent of jasmonic acid. Plant J. 54, 106–117.

Chen, F., D’Auria, J.C., Tholl, D., Ross, J.R., Gershenzon, J., Noel,

J.P. and Pichersky, E. (2003) An Arabidopsis thaliana genefor methylsalicylate biosynthesis, identified by a biochemicalgenomics approach, has a role in defense. Plant J. 36, 577–588.

Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method forAgrobacterium-mediated transformation of Arabidopsis thaliana.Plant J. 16, 735–743.

Corpet, F. (1988) Multiple sequence alignment with hierarchicalclustering. Nucl. Acids Res. 16, 10881–10890.

Cui, J., Bahrami, A.K., Pringle, E.G., Hernandez-Guzman, G., Bend-

er, C.L., Pierce, N.E. and Ausubel, F.M. (2005) Pseudomonassyringae manipulates systemic plant defenses against pathogensand herbivores. Proc. Natl Acad. Sci. USA, 102, 1791–1796.

Durrant, W.E. and Dong, X. (2004) Systemic acquired resistance.Annu. Rev. Phytopathol. 42, 185–209.

Forouhar, F., Yang, Y., Kumar, D. et al. (2005) Structural and bio-chemical studies identify tobacco SABP2 as a methyl salicylateesterase and implicate it in plant innate immunity. Proc. NatlAcad. Sci. USA, 102, 1773–1778.

Glazebrook, J. (2005) Contrasting mechanisms of defense againstbiotrophic and necrotrophic pathogens. Annu. Rev. Phytopathol.43, 205–227.

Glazebrook, J., Chen, W., Estes, B., Chang, H.-S., Nawrath, C.,

Metraux, J.-P., Zhu, T. and Katagiri, F. (2003) Topology of thenetwork integrating salicylate and jasmonate signal transductionderived from global expression phenotyping. Plant J. 34, 217–228.

Greenberg, J.T. and Yao, N. (2004) The role and regulation of pro-grammed cell death in plant-pathogen interactions. Cell. Micro-biol. 6, 201–211.

Hellens, R.P., Edwards, E.A., Leyland, N.R., Bean, S. and Mullineaux,

P.M. (2000) pGreen: a versatile and flexible binary Ti vector forAgrobacterium-mediated plant transformation. Plant Mol. Biol.42, 819–832.

King, E.O., Ward, M.K. and Raney, D.E. (1954) Two simple media forthe demonstration of phycocyanin and fluorescin. J. Lab. Clin.Med. 44, 301–307.

Koo, Y.J., Kim, M.A., Kim, E.H. et al. (2007) Overexpression ofsalicylic acid carboxyl methyltransferase reduces salicylicacid-mediated pathogen resistance in Arabidopsis thaliana. PlantMol. Biol. 64, 1–15.

Kumar, D. and Klessig, D.F. (2003) High-affinity salicylic acid-binding protein 2 is required for plant innate immunity and has

Methyl salicylate esterases in Arabidopsis 455

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456

salicylic acid-stimulated lipase activity. Proc. Natl Acad. Sci. USA,100, 16101–16106.

Kumar, D., Gustafsson, C. and Klessig, D.F. (2006) Validationof RNAi silencing specificity using synthetic genes: salicylicacid-binding protein 2 is required for innate immunity in plants.Plant J. 45, 863–868.

Maldonado, A.M., Doerner, P., Dixon, R.A., Lamb, C.J. and

Cameron, R.K. (2002) A putative lipid transfer protein involvedin systemic resistance signalling in Arabidopsis. Nature, 419,399–403.

Mishina, T.E. and Zeier, J. (2007) Pathogen-associated molecularpattern recognition rather than development of tissue necrosiscontributes to bacterial induction of systemic acquired resistancein Arabidopsis. Plant J. 50, 500–513.

Mur, L.A.J., Kenton, P., Atzorn, R., Miersch, O. and Wasternack, C.

(2006) The outcomes of concentration-specific interactionsbetween salicylate and jasmonate signaling include synergy,antagonism, and oxidative stress leading to cell death. PlantPhysiol. 140, 249–262.

Nandi, A., Welti, R. and Shah, J. (2004) The Arabidopsis thalianadihydroxyacetone phosphate reductase gene SUPPRESSOR OFFATTY ACID DESATURASE DEFICIENCY 1 is required for glycer-olipid metabolism and for the activation of systemic acquiredresistance. Plant Cell, 16, 465–477.

Park, S.-W., Kaimoyo, E., Kumar, D., Mosher, S. and Klessig, D.F.

(2007) Methyl salicylate is a critical mobile signal for plant sys-temic acquired resistance. Science, 318, 113–116.

Robert-Seilaniantz, A., Navarro, L., Bari, R. and Jones, J.D.G. (2007)Pathological hormone imbalances. Curr. Opin. Plant Biol. 10, 372–379.

Seskar, M., Shulaev, V. and Raskin, I. (1998) Endogenous methylsalicylate in pathogen-inoculated tobacco plants. Plant Physiol.116, 387–392.

Shah, J. and Klessig, D.F. (1996) Identification of a salicylicacid-responsive element in the promoter of the tobacco patho-

genesis-related b-1,3-glucanase gene, PR-2d. Plant J. 10, 1089–1101.

Shulaev, V., Silverman, P. and Raskin, I. (1997) Airborne signalingby methyl salicylate in plant pathogen resistance. Nature, 385,718–721.

Slaymaker, D.H., Navarre, D.A., Clark, D., Del Pozo, O., Martin, G.B.

and Klessig, D.F. (2002) The tobacco salicylic acid-binding protein3 (SABP3) is the chloroplast carbonic anhydrase, which exhibitsantioxidant activity and plays a role in the hypersensitiveresponse. Proc. Natl Acad. Sci. USA, 99, 11640–11645.

Tornero, P. and Dangl, J.L. (2001) A high-throughput method forquantifying growth of phytopathogenic bacteria in Arabidopsisthaliana. Plant J. 28, 475–481.

Truman, W., Bennett, M.H., Kubigsteltig, I., Turnbull, C. and Grant,

M. (2007) Arabidopsis systemic immunity uses conserveddefense signaling pathways and is mediated by jasmonates.Proc. Natl Acad. Sci. USA, 104, 1075–1080.

Vernooij, B., Friedrich, L., Morse, A., Reist, R., Kolditz-Jawhar, R.,

Ward, E., Uknes, S., Kessmann, H. and Ryals, J. (1994) Salicylicacid is not the translocated signal responsible for inducing sys-temic acquired resistance but is required in signal transduction.Plant Cell, 6, 959–965.

Vlot, A.C., Klessig, D.F. and Park, S.-W. (2008) Systemic acquiredresistance: the elusive signal(s). Curr. Opin. Plant Biol. (in press).

Wesley, S.V., Helliwell, C.A., Smith, N.A. et al. (2001) Constructdesign for efficient, effective and high-throughput gene silencingin plants. Plant J. 27, 581–590.

Yang, Y., Xu, R., Ma, C.J., Vlot, A.C., Klessig, D.F. and Pichersky, E.

(2008) Inactive methyl indole-3-acetic acid ester can be hydro-lyzed and activated by several esterases belonging to the AtMESesterase family of Arabidopsis thaliana. Plant Physiol. 147, 1034–1045.

Zuo, J., Niu, Q.-W. and Chua, N.-H. (2000) An estrogen receptor-based transactivator XVE mediates highly inducible geneexpression in transgenic plants. Plant J. 24, 265–273.

456 A. Corina Vlot et al.

ª 2008 The AuthorsJournal compilation ª 2008 Blackwell Publishing Ltd, The Plant Journal, (2008), 56, 445–456