Biochemical and Molecular Characterization of a Hydroxyjasmonate Sulfotransferase from Arabidopsis thaliana *

12-Hydroxyjasmonate, also known as tuberonic acid, was first isolated from Solanum tuberosum and was shown to have tuber-inducing properties. It is derived from the ubiquitously occurring jasmonic acid, an important signaling molecule mediating diverse developmental processes and plant defense responses. We report here that the gene AtST2a from Arabidopsis thaliana encodes a hydroxyjasmonate sulfotransferase. The recombinant AtST2a protein was found to exhibit strict specificity for 11- and 12-hydroxyjasmonate with K m values of 50 and 10 μm, respectively. Furthermore, 12-hydroxyjasmonate and its sulfonated derivative are shown to be naturally occurring inA. thaliana. The exogenous application of methyljasmonate to A. thaliana plants led to increased levels of both metabolites, whereas treatment with 12-hydroxyjasmonate led to increased level of 12-hydroxyjasmonate sulfate without affecting the endogenous level of jasmonic acid. AtST2a expression was found to be induced following treatment with methyljasmonate and 12-hydroxyjasmonate. In contrast, the expression of the methyljasmonate-responsive gene Thi2.1, a marker gene in plant defense responses, is not induced upon treatment with 12-hydroxyjasmonate indicating the existence of independent signaling pathways responding to jasmonic acid and 12-hydroxyjasmonic acid. Taken together, the results suggest that the hydroxylation and sulfonation reactions might be components of a pathway that inactivates excess jasmonic acid in plants. Alternatively, the function of AtST2a might be to control the biological activity of 12-hydroxyjasmonic acid.

jasmonate (MeJA) by the development of senescence-like symptoms and by the up-regulation of a large number of genes including genes coding for plant defense proteins such as thionins and proteinase inhibitors or enzymes involved in phytoalexin synthesis (2).
The biosynthesis of JA occurs through the octadecanoid pathway and starts with the oxidation of ␣-linolenic acid to yield 13-hydroperoxylinolenic acid (3). The hydroperoxide is then dehydrated and cyclized to the cyclopentanone 12-oxophytodienoic acid (OPDA). Reduction of OPDA by OPDA reductase 3 followed by three cycles of ␤-oxidation yield JA. JA can undergo several modification reactions, the best studied being methylation at carbon 1 to yield MeJA (4). In addition to the formation of the methyl ester, the biotransformation of JA involves five major reactions (5): (i) hydroxylation at carbon 11 or carbon 12 to form 11-hydroxyjasmonic acid (11-OHJA) or 12-hydroxyjasmonic acid (12-OHJA); (ii) reduction of the keto group at carbon 6 to form cucurbic acid (CA), which can also be hydroxylated on the side chain; (iii) O-glycosylation of the above hydroxylated compounds; (iv) conjugation of JA or its hydroxylated derivatives with amino acids (5), tyramine (6), or adenylate (7); and (v) degradation of the carboxylic acid side chain to form cis-jasmone (8).
Some of these reactions might lead to the irreversible inactivation of JA, whereas others might be reversible and used for transport and/or storage. Alternatively, the metabolism of JA might generate molecules with new properties. For instance, the 12-O-␤-D-glucopyranoside of 12-OHJA was shown to induce tuber formation in vitro (9) and the degradation of JA to cisjasmone was found to affect the plant defense response to herbivores by stimulating the activity of parasitic insects (10). Furthermore, a distinct ratio of several octadecanoid-derived compounds, the so-called "oxylipin signature" may have a biological significance. For instance, such a signature was found to be different among various plant species (11) and attributed to distinct defense status (12), and varied also in the same species during development (13).
A sulfated conjugate of 12-OHJA was found to accumulate in the plant species Tribulus cistoides, a member of the Zygophylaceae family (14). This is the only report on the occurrence of a conjugate of 12-OHJA in a plant species that does not produce tubers. Furthermore, it is the only report of the occurrence of a jasmonate conjugated with a sulfonate group. In mammals, it is well recognized that the sulfonation reaction plays an important role in the modulation of the biological activity of a number of compounds, such as steroids and thyroid hormones, and catecholamine neurotransmitters (15)(16)(17)(18). Sulfonate conjugation not only facilitates transport and excretion of hydrophobic molecules by increasing their water solubility, it abolishes the biological activity of hormones such as estrogens. Recently, the characterization of a brassinosteroid sulfotransferase (ST) from Brassica napus demonstrated that plants like animals use the sulfonation reaction to control the biological activity of hormones (19). The sulfonation reaction can also lead to the production of biologically active molecules. For example, the sulfonation of gallic acid glucoside generates the periodic leaf movement factor 1 involved in the seismonastic response in Mimosa pudica (20).
Recently, we initiated a systematic investigation of the biochemical functions of the 18 ST-coding genes from Arabidopsis thaliana. We report here the characterization of a ST (AtST2a) from this plant exhibiting strict specificity for hydroxyjasmonates. This is the first report of a ST involved in jasmonate metabolism. We also demonstrate that 12-OHJA and its sulfate derivative are natural constituents in A. thaliana. We propose a role for the hydroxyjasmonate ST in the regulation of the levels of JA and/or 12-OHJA in plants.

Materials
12-OHJA, JA isoleucine conjugate, CA, 7-iso-CA, 6-epi-CA, 6-epi-7iso-CA, and their methyl esters, were from our laboratory collection. MeJA and JA were purchased from Bedoukian Research Inc. and Sigma, respectively. Wild type A. thaliana seeds, ecotype Col-0, were obtained from Lehle Seeds. The enzymes used for cloning were from New England Biolabs and used under conditions recommended by the manufacturer. All reagents were of analytical or molecular biology grade.

Cloning of AtST2a
cDNA clone of AtST2a (GenBank TM accession number T43254) was obtained from the Arabidopsis Biological Resource Center. Oligonucleotide 5Ј-CGGGATCCATGGCTACCTCAAGCATGAAG-3Ј was designed to introduce a BamHI site at the 5Ј-end of the coding sequence. AtST2a was amplified by polymerase chain reaction with Vent DNA polymerase (New England Biolabs) using the above primer and M13-20 primer. The amplified product was digested with BamHI and ligated into the BamHI site of the bacterial expression plasmid pQE30 (Qiagen). Clones containing AtST2a in the proper orientation were determined by restriction enzyme analysis and by sequencing the 5Ј junction.

DNA Sequence Analysis
DNA and protein sequence alignments were performed using the ClustalW program 2 and the similarity/identity values determined from the pairwise comparisons of all the ST genes. Boxshade was used to shade identical residues in the alignments. 3

Expression of Recombinant AtST2a in Escherichia coli
A culture of E. coli, strain XL1-blue harboring AtST2a (A 600 ϭ 0.7) was induced with 1 mM isopropylthio-␤-D-galactopyranoside for 10 h at 22°C. Bacterial cells were collected by centrifugation and resuspended in 50 mM sodium phosphate buffer (pH 8.0) containing 0.3 M NaCl and 14 mM 2-mercaptoethanol. The cells were lysed by sonication and the recombinant proteins were recovered in the soluble fraction by centrifugation at 12,000 ϫ g for 15 min at 4°C. The soluble recombinant AtST2a protein was purified by affinity chromatography onto nickelnitriloacetic acid-agarose matrix as recommended (Qiagen). The Niagarose-purified protein was desalted on a PD-10 (Amersham Biosciences) column equilibrated in 25 mM bis-Tris (pH 6.5) (buffer A). The desalted protein was chromatographed on a PAP-agarose column (0.5 ϫ 10 cm) previously equilibrated in buffer A. The bound protein was eluted with a linear salt gradient of 0 to 1 M NaCl in the same buffer. Affinity chromatography was performed on a Waters 625 LC HPLC system and protein absorbance was monitored at 280 nm. AtST2a eluted at ϳ450 mM NaCl. Protein concentration was estimated using the Bradford reagent (Bio-Rad) and bovine serum albumin as reference protein.

Partial Purification of Hydroxyjasmonate Sulfotransferase from
A. thaliana 15-Day-old plants grown on MS medium in magenta boxes were treated with 100 M 12-OHJA or MeJA for a period of 8 h. At the end of the treatment, rosette leaves (about 2 g) were harvested, ground in liquid nitrogen, and homogenized in 10 ml of ice-cold 0.5 M bis-Tris (pH 6.5). The homogenate was filtered through nylon mesh and the filtrate was centrifuged at 12,000 ϫ g for 15 min. The supernatant was gently stirred for 10 min with 0.5 ml of PAP-agarose pre-equilibrated in the same buffer. The mixture was centrifuged for 5 min at 1000 ϫ g on a benchtop microcentrifuge. The supernatant was discarded and replaced with 1 ml of 0.5 M bis-Tris (pH 6.5) containing 1.0 M NaCl. The solution was gently stirred for 5 min followed by centrifugation at 1000 ϫ g for 5 min. The supernatant was used for enzyme activity.

Sulfotransferase Assays
Analysis of substrate specificity was performed by testing enzymatic activity with three different concentrations of acceptor substrates: 1, 10, and 100 M. The reaction mixture (50 l) contained 50 pmol of The reactions were allowed to proceed for 10 min at 25°C. Under the assay conditions, the reaction was linear with time from 0 to 40 min and with protein concentrations ranging from 0 to 2.0 g/assay. The AtST2a-sulfated reaction product was extracted with 1-butanol saturated with water and an aliquot was counted for radioactivity in scintillation fluid.

Detection and Quantification of JA and 12-OHJA from A. thaliana
Rosette leaves from fresh plant material (1 g) were homogenized with 10 ml of methanol containing 100 ng of [ 2 H 6 ]JA and 12-[ 2 H 3 ]OAc-JA as internal standards. The filtrate was evaporated and acetylated with pyridine/acetic acid anhydride (2:1) at 20°C overnight. The reaction mixture was evaporated, resuspended in ethyl acetate, and loaded on a silica (SiOH) column (500 mg; Machery-Nagel). The flow-through containing JA and the acetylated form of 12-OHJA was collected and evaporated. The evaporated mixture was resuspended in 5 ml of methanol and loaded on a 3-ml DEAE-Sephadex A25 column (acetylatedform in methanol). The column was washed with 3 ml of methanol followed by 3 ml of 0.1 M acetic acid in methanol. The jasmonates were eluted with 5 ml of 1 M acetic acid in methanol (Fraction A), evaporated, and separated on preparative HPLC for GC-MS analysis.
The SiOH column was washed with methanol and the flow-through (Fraction B) was collected for analysis of 12-OHJA sulfate. Fraction B was evaporated, resuspended in 10% acetonitrile, and chromatographed by reverse phase HPLC (method gradient, 10 -90% acetonitrile in 15 min at a flow rate of 1 ml/min). Fractions were collected from 4.5 to 7 min, evaporated, and resuspended in 50 l of methanol and analyzed by LC-MS.
Preparative HPLC-Fraction A eluted from the DEAE-Sephadex A-25 column was subjected to preparative HPLC column, Eurospher 100-C18 (5 m, 250 ϫ 4 mm). Jasmonates were eluted with methanol, 0.2% acetic acid in H 2 O (1:1) at a flow rate of 1 ml/min and UV detection at 210 nm. Fractions between R t 9.15 and 11 min containing 12-OAc-JA and between 12 and 13.30 min containing JA were collected and evaporated. The samples were dissolved in 200 l of chloroform/N,N-diisopropylethylamine (1:1) and derivatized with 10 l of pentafluorobenzylbromide at 20°C overnight. The evaporated derivatized samples were dissolved in 5 ml of n-hexane and passed through a SiOH column (500 mg; Machery-Nagel). The pentafluorobenzyl esters were eluted with 7 ml of n-hexane/diethyl ether (2:1), evaporated, dissolved in 100 l of acetonitrile, and analyzed by GC-MS.
Detection of 12-OHJA Sulfate by LC MS/MS-The negative ion electrospray (ES) mass spectra were obtained from a Finnigan MAT TSQ 7000 instrument (electrospray voltage, 4 kV; heated capillary temperature, 220°C; sheath gas, nitrogen) coupled with a Micro-Tech Ultra-Plus MicroLC system equipped with a RP18 column (4 m, 1 ϫ 100 mm, Ultrasep). For HPLC, a gradient was used starting from H 2 O:acetonitrile (90:10; containing 0.2% acetic acid) to 10:90 in 15 min followed by a 10-min isocratic period at a flow rate of 70 l/min. The collisioninduced dissociation (CID) mass spectra during the HPLC run were performed with a collision energy of 30 eV (collision gas, argon; collision pressure, 1.8 ϫ 10 Ϫ3 torr). All mass spectra are averaged and background subtracted. The following results were obtained: 12-OHJA sulfate, R t (LC-MS), 12.32 min, negative electrospray MS m/z (relative intensity), 305 ([M Ϫ H] Ϫ , 100); CID spectrum, 225 (93), 147 (9), 97 (100), 59 (58). For the determination of 12-OHJA sulfate in plant material, the daughter ions at m/z 225, 97, and 59 were measured in the selected ion monitoring mode. The CID spectrum of 12-OHJA sulfate displays significant ions at m/z 225, 97, and 59 reflecting the typical structural features of the compound. The ion at m/z 97 represents a key ion in the negative CID mass spectra of sulfated compounds (21).
Northern Blot Analysis-Rosette leaves of A. thaliana plants were pulverized in liquid nitrogen and total RNA was extracted in buffer/ phenol/chloroform as described (22). 10 g of total RNA was used for agarose gel electrophoresis after denaturing with glyoxal and Me 2 SO (23). Northern blot analysis of total RNA was achieved under high stringency conditions according to standard procedures (24) using the 32 P-labeled coding sequence of AtST2a as a probe. SDS-PAGE Electrophoresis-To verify the solubility and evaluate the level of purity of the recombinant protein after chromatography on nickel-agarose and PAP-agarose, aliquots of the recombinant enzyme were subjected to 12% polyacrylamide gel electrophoresis according to the method of Laemmli (25). The proteins were visualized by Coomassie Blue staining.
Reverse Transcriptase-Polymerase Chain Reaction-2.5 g of total RNA was treated with 20 units of DNase I (Roche Diagnostics) in 50 l of 0.1 M sodium acetate, 5 mM MgSO 4 (pH 5.0) for 10 min at 37°C. DNase I was heat inactivated at 95°C for 5 min and RNA was ethanol precipitated. cDNA was synthesized using Moloney murine leukemia virus-reverse transcriptase (New England Biolabs) in a 25-l reaction volume as recommended by the manufacturer. 1 l of room temperature reaction product was used for PCR with Ex Taq DNA polymerase (Takara Biochemicals).

RESULTS
Cloning of AtST2a-The AtST2a cDNA (GenBank TM accession number T43254) was obtained from the Arabidopsis Biological Resource Center. AtST2a is localized on chromosome V and can be retrieved from the GenBank TM data base under accession number AB010697 from nucleotides 53,936 to 55,015. It is flanked by an open reading frame encoding another putative ST, AtST2b from nucleotides 50,627 to 51,670. AtST2a and AtST2b encode proteins of 359 and 347 amino acids corresponding to molecular masses of 41.3 and 39.6 kDa, respectively. The AtST2a and AtST2b deduced protein sequences contain all the domains known to be involved in the binding of the sulfonate donor PAPS and in catalysis (Fig. 1) (26). Amino acid sequence alignment indicates that they share 85% amino acid sequence identity and 92% similarity, suggesting that they might be isozymes with similar substrate specificities. This hypothesis is further supported by the fact that the amino acids known to interact with the sulfonate acceptor substrate are identical between the two sequences (26). The AtST2a and -b proteins share 45% sequence identity with the A. thaliana and B. napus 24-epibrassinosteroid STs, 40% sequence identity with the flavonol STs of Flaveria species, and ϳ25% sequence identity with the mammalian STs.
Substrate Specificity-To characterize the biochemical function of AtST2a, we studied the activity of purified recombinant AtST2a expressed in E. coli. The histidine-tagged AtST2a enzyme could be purified to apparent homogeneity by a combination of nickel-and PAP-agarose chromatography (Fig. 2) (27). The purified recombinant AtST2a enzyme was used to test a variety of acceptor molecules of plant and mammalian origin. The substrate library comprises more than 100 compounds and includes phenolic acids, desulfoglucosinolates, flavonoids, steroids, gibberellic acids, phenylpropanoids, hydroxyjasmonates, and coumarins, to mention a few. AtST2a was found to exhibit strict specificity for 11-and 12-OHJA (Fig. 3). No activity could be detected with the remaining compounds of the library. The enzyme did not accept structurally related jasmonates such as 12-OHJA methyl ester, CA, 7-iso-CA, 6-epi-CA, 6-epi-7-iso-CA, and their methyl esters. Furthermore, the AtST2a enzyme did not sulfonate the structurally related mammalian prostaglandin E 2 , arachidonyl alcohol, and 11-eicosenol. AtST2b was also expressed in E. coli and could be recovered in a soluble form. No activity was observed with all the substrates tested including 11-and 12-OHJA. This result is surprising when we consider the high level of sequence identity with AtST2a and the presence of all the amino acids known to be required for cosubstrate binding and catalysis (28). Other substrates related in structure to 11-and 12-OHJA will have to be tested to determine the biochemical function of this enzyme.
Enzyme Properties-To determine the pH optimum of AtST2a enzyme activity, the sulfonation of 12-OHJA was carried out at fixed substrate (50 M) and PAPS (5 M) concentrations in buffers ranging from pH 6 to 10. Highest enzyme activity was observed at pH 7.5 in Tris buffer. However, the enzyme is active in a broad pH range and retains ϳ80% activity at pH 9.5. The kinetic parameters of AtST2a were determined from the results of substrate interaction kinetic experiments. AtST2a exhibited preference for 12-OHJA over 11-OHJA with K m values of 10 and 50 M, respectively. Double reciprocal plots with 12-OHJA as the variable substrate at several fixed concentrations of PAPS resulted in an intersecting pattern (data not shown). A similar pattern was obtained with PAPS as the variable substrate and the secondary slope, and intercept replots were linear. As calculated from the replots, the V max for the conversion of 12-OHJA to 12-OHJA sulfate was 37.5 pkatal/mg of protein. The K m value for PAPS was found to be 1 M. Because JA is structurally similar to 12-OHJA, we tested if it could act as a competitive inhibitor in enzyme assays. The assays were performed with a fixed concentration of 12-OHJA (50 M) and variable concentrations of JA (0.1-100 M). Addition of JA had no effect on the amount of product formed (data not shown), indicating that JA does not interfere with the binding of 12-OHJA at the enzyme active site.
Identification and Quantification of Jasmonates in A. thaliana-To date, the presence of 12-OHJA has been reported to occur only in tuber-producing plant species (29) and in the fungus Botryodiplodia theobromae (30). To validate the results of the in vitro studies, GC-MS was used to analyze A. thaliana extracts for the presence of 12-OHJA. To minimize the effects of biological variability, pools consisting of 30 -50 plants were used for metabolite extraction. Internal standards were used to compensate for differences in extraction efficiencies. The technical variability of the GC-MS and LC-MS/MS upon repeated injections was found to be smaller than 5%. Table I shows that 12-OHJA is present at a level ϳ4-fold lower than JA. Preliminary results indicate that 12-OHJA is also present in barley, tomato, and tobacco leaves suggesting that it might have a wide distribution in the plant kingdom (data not shown).
Identification of the Enzyme Reaction Product Formed in Vitro and in Vivo-To identify the reaction product formed by recombinant AtST2a using LC-MS/MS, the enzyme reaction was carried out using unlabeled PAPS. The product of the enzymatic reaction was found to elute at the same retention time and gave the same fragmentation pattern as authentic 12-OHJA sulfate (Fig. 3) (14).
12-OHJA sulfate has only been reported to occur in T. cistoides, a member of the Zygophylaceae family (14). To determine whether it is present in A. thaliana, methanolic extracts were subjected to LC-MS/MS. 12-OHJA sulfate was identified as an endogenous compound in A. thaliana (Table I) (Table  I). Treatment with MeJA led to 20-and 4-fold increases of the endogenous levels of 12-OHJA and 12-OHJA sulfate, respectively. Treatment with 12-OHJA has no effect on the accumulation of JA. However, a 9-fold increase in 12-OHJA sulfate level is observed following this treatment.
Regulation Studies-Rosette leaves from 15-day-old A. thaliana plants expressed very low levels of AtST2a mRNAs when grown under normal conditions (Fig. 4, lane C). This result is consistent with the absence of detectable hydroxyjasmonate sulfotransferase activity in protein extracts from plants growing under the same conditions. However, treatment with the jasmonates 12-OPDA, JA, and JA-Ile resulted in strong induction of AtST2a expression (Fig. 4). Interestingly, AtST2a expression is also induced in a dose-dependent manner by 12-OHJA (Fig. 4). The steady state AtST2a mRNA levels increased within 30 min following 12-OHJA treatment (Fig.  5A). Similar kinetics of induction were obtained when plants were treated with MeJA (Fig. 5B). The specific up-regulation of AtST2a by 12-OHJA and MeJA was confirmed by RT-PCR using AtST2a-specific primers (Fig. 6). Furthermore, the induction of AtST2a expression by 12-OHJA and MeJA was found to correlate with increased hydroxyjasmonate ST activity in partially purified protein preparations from plants treated for 8 h with 12-OHJA (about 9 pkatal/mg of protein) or MeJA (about 14 pkatal/mg of protein). To test if the accumulation of AtST2a transcript is specific to jasmonates or a general stress response, Arabidopsis plants were subjected to a number of physical and chemical treatments. AtST2a transcript level is not altered in rosette leaves from plants subjected to cold, heat, hypoxia, salt stress, or treatment with salicylic acid. In addition, treatments with phytohormones like gibberellins, abscisic acid, zeatin, and indole 3-acetic acid did not affect the AtST2a transcript level (data not shown).
Thi2.1 is a marker gene for MeJA treatment or wounding (31,32). As expected, Thi2.1 expression is induced when plants are treated with MeJA (Fig. 6). However, the expression of Thi2.1 is not induced in rosette leaves from plants treated with 12-OHJA. In contrast, AtST2a expression is induced by both MeJA and 12-OHJA.  a 18-Day-old plants grown in magenta boxes were treated by the addition to the MS medium of MeJA or 12-OHJA to a final concentration of 20 M. Rosette leaves were harvested 36 h after the treatments and processed for jasmonate quantification according to the method described under "Experimental Procedures." The values represent the results of an experiment in which a pool of 30 -50 plants was analyzed.
b The large amount of free JA indicates in vivo cleavage of the methyl ester.

DISCUSSION
A cDNA clone encoding an enzyme catalyzing the sulfonation of 12-OHJA has been characterized in this study. The Arabidopsis gene AtST2a is a member of a small family comprising 18 members. It shares 85% amino acid identity with AtST2b and between 40 and 45% identity with other members of the family. To characterize the biochemical function of AtST2a, a large number of substrates were tested. The substrate library comprises the desulfo-derivatives of most of the known plantsulfated metabolites as well as a collection of metabolites for which no sulfated derivatives have been reported in the literature. AtST2a was found to exhibit strict specificity for 11-and 12-OHJA with K m values of 50 and 10 M, respectively. However, these values are probably not reflecting the real affinity of the enzyme for its substrates, because the assay mixture contained four steroisomers as a result of the presence of two chiral carbons at C-3 and C-7 of the cyclopentanone ring. Substrate-dependent stereoselectivity has been reported for the rat hydroxysteroid and phenol STs and more recently for the B. napus brassinosteroid ST (19,33). Although the stereochemistry of naturally occurring 12-OHJA has not been established unambiguously, naturally occurring 12-OHJA seems to have a cis conformation, and it was shown that cis-12-OHJA has a much stronger tuber-inducing activity than the trans isomer (34). It will be interesting to see whether AtST2a functions in a stereoselective manner, because such a property has been shown for other cyclopentenone-generating (allene oxide cyclase) and converting (OPDA reductase 3) enzymes (35,36).
The fact that the enzyme accepts 12-OHJA and not its methyl ester derivative indicates a strict requirement for a carboxyl function at carbon 1. Furthermore, the fact that JA is not a competitive inhibitor of the reaction indicates that the presence of a hydroxyl group at positions 11 or 12 on the side chain is required for binding at the active site of the enzyme. The strict structural specificity exhibited by AtST2a is a common feature of plant STs and has been observed with the flavonoid, gallic acid glucoside, and brassinosteroid STs (19,37).
Previous to this work, there were no reports on the occurrence of 12-OHJA or its sulfated derivative in A. thaliana or any other members of the Brassicaceae family. 12-OHJA was initially isolated as a tuber-inducing compound from Solanum tuberosum (9). To date, its presence has also been reported in Solanum demissum (38) and in the fungus B. theobromae (30). The glucoside of 12-OHJA has also been detected in Helianthus tuberosus (39) and Astragalus complanatus (40). The presence of 12-OHJA in A. thaliana indicates that its distribution can no longer be considered to be restricted to tuber-producing plants and it raises interesting questions about its function in plants that do not produce tubers. This is the first report on the characterization of an enzyme that sulfonate hydroxylated jasmonates. Recently a mouse cDNA clone encoding an ST that sulfonates compounds structurally similar to jasmonates has been characterized (41). This enzyme, which is predominantly expressed in the kidneys and the uterus, accepts eicosanoids (prostaglandins, thromboxanes, and leukotrienes) as substrates. However, the biological function and the regulation of this enzyme have not been studied yet.
Treatments of Arabidopsis plants with JA, 12-OPDA, MeJA, or JA-Ile lead to increased accumulation of the AtST2a transcript (Fig. 4). It has been shown previously that 12-OPDA can alter the expression of specific genes (42) and was found to exhibit a strong activity in several JA bioassays (43,44). However, we cannot rule out the possibility that the effect of 12-OPDA on AtST2a expression is caused by its metabolism to JA or 12-OHJA. JA-Ile was also shown to induce expression of specific genes when applied exogenously to barley leaves (45). In this case, induction of gene expression could be attributed directly to the conjugate, because very little conversion to JA was observed. The significance of AtST2a induction by the isoleucine conjugate of JA remains to be determined. Treatments with MeJA led to an increase in the amount of AtST2a mRNA, 12-OHJA, and 12-OHJA sulfate. The metabolism of JA to 12-OHJA sulfate might be a route leading to its inactivation. The widespread occurrence of STs and sulfatases in mammals has led to the hypothesis that the concerted action of both groups of enzymes might regulate the levels of bioactive molecules (15). However, the sulfonation of 12-OHJA might be irreversible considering the lack of evidence for the presence of genes encoding sulfatases in the genome of A. thaliana.
Alternatively, the function of AtST2a might be to control directly the biological activity of 12-OHJA. Several lines of evidence suggest that 12-OHJA might have a biological activity different from JA or MeJA. Unlike JA or MeJA, 12-OHJA does not exhibit inhibitory effects on plant growth, such as promotion of senescence of oat leaves, inhibition of soybean callus growth, and inhibition of lettuce seedling growth (34). Furthermore, 12-OHJA was shown previously to have tuber-inducing properties (9). It is possible that 12-OHJA controls developmental processes other than tuberization in plants such as A. thaliana.
One of the most striking features of AtST2a regulation is its induction following treatment with 12-OHJA (Fig. 5). Prior to this work, the tomato gene encoding allene oxide cyclase was the only gene known to be up-regulated by 12-OHJA (46). It is interesting to note that in both Arabidopsis and tomato, treatments with 12-OHJA did not result in increased endogenous JA levels ruling out the possibility that the latter might be the mediator of the response (Table I) (46). The up-regulation of AtST2a expression in response to 12-OHJA indicates the presence of a feed-forward mechanism that tightly controls the in vivo concentration of the free acid by sulfonation. Similar feedforward mechanisms have been shown to regulate the level of the enzymes abscisic acid 8-hydroxylase and gibberellin-2 oxidase involved in the catabolism of abscisic acid and gibberellins, respectively (47,48).
In contrast with AtST2a and allene oxide cyclase (AOC), Thi2.1 expression is not induced by 12-OHJA (Fig. 6). Thi2.1 encodes a thionin that is specifically induced by wounding, pathogen infection, and following treatment with Me-JA (31). Similarly, treatments of barley leaves with 12-OHJA did not induce the expression of a multigene family of barley-encoding thionins (49). The lack of induction of Thi2.1 following 12-OHJA treatment suggests that Me-JA and 12-OHJA mediate their response via two independent pathways. Taken together, the results suggest that 12-OHJA might have a biological function different from JA and MeJA and that the sulfonation reaction might be required to control its activity. In the future, the construction of transgenic plants with altered levels of AtST2a will allow to address the role of 12-OHJA in plant growth, development, and defense responses.