Portal de Periódicos da CAPES

Purification, Cloning, and Expression of a Pathogen Inducible UDP-glucose:Salicylic Acid Glucosyltransferase from Tobacco

1999; Elsevier BV; Volume: 274; Issue: 51 Linguagem: Inglês

10.1074/jbc.274.51.36637

1083-351X

Hyung‐il Lee, Ilya Raskin,

Plant Gene Expression Analysis

Salicylic acid (SA) plays an important role in plant disease resistance. Inoculation of tobacco leaves with incompatible pathogens triggers the biosynthesis of SA which accumulates primarily as the SA 2-O-β-d-glucoside (SAG) and glucosyl salicylate (GS). The tobacco UDP-glucose:salicylic acid glucosyltransferase (SA GTase) capable of forming both SAG and GS was purified, characterized, and partially sequenced. It has an apparent molecular mass of 48 kDa, a pH optimum of 7.0, and an isoelectric point at pH 4.4. UDP-glucose was the sole sugar donor for the enzyme. However, SA and several phenolics served as glucose acceptors. The apparent K m values for UDP-glucose and SA were 0.27 and 1–2 mm, respectively. Zn2+ and UDP inhibited its activity. The corresponding cDNA clone which encoded a protein of 459 amino acids was isolated from an SA-induced tobacco cDNA library and overexpressed in Escherichia coli. The recombinant protein catalyzed the formation of SAG and GS, and exhibited a broad specificity to simple phenolics, similar to that of the purified enzyme. Northern blot analysis showed that the SA GTase mRNA was induced both by SA and incompatible pathogens. The rapid induction timing of the mRNA by SA indicates that it belongs to the early SA response genes. Salicylic acid (SA) plays an important role in plant disease resistance. Inoculation of tobacco leaves with incompatible pathogens triggers the biosynthesis of SA which accumulates primarily as the SA 2-O-β-d-glucoside (SAG) and glucosyl salicylate (GS). The tobacco UDP-glucose:salicylic acid glucosyltransferase (SA GTase) capable of forming both SAG and GS was purified, characterized, and partially sequenced. It has an apparent molecular mass of 48 kDa, a pH optimum of 7.0, and an isoelectric point at pH 4.4. UDP-glucose was the sole sugar donor for the enzyme. However, SA and several phenolics served as glucose acceptors. The apparent K m values for UDP-glucose and SA were 0.27 and 1–2 mm, respectively. Zn2+ and UDP inhibited its activity. The corresponding cDNA clone which encoded a protein of 459 amino acids was isolated from an SA-induced tobacco cDNA library and overexpressed in Escherichia coli. The recombinant protein catalyzed the formation of SAG and GS, and exhibited a broad specificity to simple phenolics, similar to that of the purified enzyme. Northern blot analysis showed that the SA GTase mRNA was induced both by SA and incompatible pathogens. The rapid induction timing of the mRNA by SA indicates that it belongs to the early SA response genes. salicylic acid tobacco mosaic virus pathogenesis related SA 1-O-β-d-glucoside glucosyl salicylate UDP glucose:salicylic acid glucosyltransferase cycloheximide triethanolamine 2-[bis(2-hydroxyethyl) amino]-2-(hydroxymethyl)-propane-1,3-diol polyacrylamide gel electrophoresis high performance liquid chromatography fast protein liquid chromatography polymerase chain reaction indole acetic acid Importance of salicylic acid (SA)1 in the signal transduction pathway of plant disease resistance has been well documented in many plants (1Malamy J. Carr J.P. Klessig D.F. Raskin I. Science. 1990; 250: 1002-1004Crossref PubMed Scopus (1075) Google Scholar, 2Métraux J.-P. Signer H. Ryals J. Ward E. Wyss-Benz M. Gaudin J. Raschdorf K. Schmid E. Blum W. Inverardi B. Science. 1990; 250: 1004-1006Crossref PubMed Scopus (802) Google Scholar, 3Delaney T.P. Friedrich L. Ryals J. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6602-6606Crossref PubMed Scopus (588) Google Scholar, 4Hammond-Kosack K.E. Silverman P. Raskin I. Jones J.D.G. Plant Physiol. 1996; 110: 1381-1394Crossref PubMed Scopus (114) Google Scholar). In tobacco cultivars containing theN gene for resistance, the hypersensitive response occurs after inoculation with incompatible pathogens including tobacco mosaic virus (TMV) or Pseudomonas syringae pv.phaseolicola. The hypersensitive response is characterized by the formation of necrotic lesions which restrict the spread of pathogens and by a dramatic increase in SA levels in the inoculated leaf and, to a lesser extent, throughout the plant. Concomitant with the appearance of SA is an accumulation of pathogenesis-related (PR) proteins and the establishment of systemic acquired resistance throughout the plant. Direct evidence showing the involvement of SA in disease resistance came from a study of transgenic plants that were unable to accumulate SA or to establish systemic acquired resistance (5Gaffney T. Friedrich L. Vernooij B. Negrotto D. Nye G. Ukness S. Ward E. Kesmann H. Ryals J. Science. 1993; 261: 754-756Crossref PubMed Scopus (1421) Google Scholar). Analysis of Arabidopsis mutants showed that mutants with high levels of SA exhibited enhanced disease resistance while plants unable to respond to exogenous SA were impaired in their disease-resistance response (6Durner J. Shah J. Klessig D.F. Trends Plant Sci. 1997; 2: 266-274Abstract Full Text PDF Scopus (499) Google Scholar). Furthermore, SA was shown to be involved in production of active oxygen species and cell death (7Draper J. Trends Plant Sci. 1997; 2: 162-165Abstract Full Text PDF Scopus (229) Google Scholar). Like other plant hormones, SA is present in plants as both a free acid and as conjugated metabolites generated primarily by methylation, hydroxylation, and glucosylation. Among the SA conjugates, the SA 2-O-β-d-glucoside (SAG) has been identified as a predominant and stable metabolite in many plants, including tobacco. In TMV-inoculated tobacco possessing the N gene, newly synthesized SA is converted primarily to glucosylated and methylated conjugates (8Enyedi A.J. Yapani N. Silverman P. Raskin I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2480-2484Crossref PubMed Scopus (385) Google Scholar, 9Shulaev V. Silverman P. Raskin I. Nature. 1997; 385: 718-721Crossref Scopus (586) Google Scholar). The glucosylated forms are localized in the vicinity of the necrotic lesions, while methyl salicylate is volatilized. Glucosyl salicylate (GS), an ester form, was also recently identified as a SA metabolite in virus- or bacteria-inoculated tobacco (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar).In vivo tracer studies with tobacco leaves fed [14C]SA showed that the SAG accumulates over time while GS formation is more transient and occurs rapidly after SA administration. Rapid GS formation may serve as a protective mechanism from the phytotoxic effects of high concentrations of SA, while the more stable SAG may be a slow-release storage form of SA. The precise role of each conjugated form is, however, difficult to document. In tobacco, two SA glucosyltransferase (SA GTase) activities leading to the formation of both SAG and GS were detected (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar, 11Edwards R. J. Plant Physiol. 1994; 143: 609-614Crossref Scopus (45) Google Scholar, 12Enyedi A.J. Raskin I. Plant Physiol. 1993; 101: 1375-1380Crossref PubMed Scopus (88) Google Scholar). UDP-glucose:SA GTase (SAGT) catalyzed the conversion of SA to SAG, while formation of GS was catalyzed by UDP-glucose:SA carboxyl GTase. Both enzyme activities were enhanced in pathogen-inoculated or SA-treated tobacco plants. However, little is known about how closely these two glucosyltransferases are related to each other and whether they are products of one gene or a gene family. For reasons of simplicity, therefore, we will refer to the SA glucosyltransferase simply as SA GTase, realizing that it results from the combined activities of both SAGT and SA carboxyl GTase. In oat, the glucosyltransferase displaying SAGT activity was partially purified (195-fold) and characterized (13Yalpani N. Schulz M. Davis M.P. Balke N.E. Plant Physiol. 1992; 100: 457-463Crossref PubMed Scopus (52) Google Scholar). Here, we purified to near homogeneity, characterized, and partially sequenced the glucosyltransferase that catalyzes the in vitro conversion of SA to GS (an ester form) and to a lesser extent to SAG (a more stable glucoside). Using the peptide sequence information obtained from the purified protein, we were able to isolate and characterize the cDNA clone for the purified SA GTase. We also confirmed that the recombinant protein expressed in Escherichia coli converted SA to both GS and SAG. Tobacco plants (Nicotiana tabacum L. cv Xanthi-nc NN genotype) were grown as described previously (14Yalpani N. Silverman P. Wilson T.M.A. Kleier D.A. Raskin I. Plant Cell. 1991; 3: 809-818Crossref PubMed Scopus (430) Google Scholar). SA infiltration or inoculation of either tobacco mosaic virus (TMV) or Pseudomonas syringae pv. phaseolicola (NPS3121) (15Lindgren P.B. Peet R.C. Panopoulus N.J. J. Bacteriol. 1986; 168: 512-522Crossref PubMed Scopus (299) Google Scholar) was described previously (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar). For SA incubation, leaf disks were floated on a solution containing SA at the indicated concentration. Distilled water was used as a control. For inhibition of protein synthesis, the leaf disks were preincubated in 20 μg/ml cycloheximide (CHX) solution for 1 h and then floated in solutions containing either 20 μg/ml CHX alone or 20 μg/ml CHX and 1 mm SA. All of the samples harvested at the indicated time points were frozen in liquid nitrogen, and stored at −80 °C for RNA extraction. Frozen tobacco leaves were pulverized in liquid nitrogen using a mortar and pestle, and resuspended in 20 mm triethanolamine (TEA) buffer (pH 7.3) containing 14 mm β-mercaptoethanol, leupeptin (1 μg/ml), 1 mm phenylmethysulfonyl fluoride, and 2% (w/v) polyvinylpyrrolidone. The tissue slurry was removed by centrifugation at 15,000 × g for 20 min and supernatant filtered through Miracloth. The filtrate was used for further purification. The protein fractions which precipitated between 40 and 75% of ammonium sulfate saturation were isolated by centrifugation. The pellet was resuspended in 20 mm TEA buffer containing 14 mmβ-mercaptoethanol (TEAM) and desalted in a PD-10 column (Amersham Pharmacia Biotech) which was equilibrated with TEAM buffer. After desalting, the protein samples were subject to Mono Q FPLC column (Amersham Pharmacia Biotech) which was equilibrated with TEAM buffer. The proteins were eluted with a 0–2 m KCl gradient in TEAM buffer (pH 7.3) at a flow rate of 1.5 ml/min. Fractions were collected and assayed for enzyme activity and protein concentration. The fractions containing enzyme activity were desalted on a PD-10 column and concentrated using a Centricon 30 microconcentrator (Amicon Inc., Beverly, MA). Concentrated active fractions from the Mono Q column were applied to a UDP-glucuronic acid affinity column (Sigma). The eluent and washed fractions that contained the enzyme activity were collected and used for further purification. The glucosyltransferase activity was not detected in the fractions eluted with salts or UDP-glucose. The active fractions were further purified with a Hitrap Blue affinity column (Amersham Pharmacia Biotech) which was linked to the Ranin FPLC/HPLC system. After washing the column with 20 mm TEAM buffer, the active fractions were eluted from the column with 10 mm UDP and concentrated with a Centricon 30. After removal of UDP, the fractions were assayed for the enzyme activity. Active fractions from the dye affinity column were applied to the Superose 12 FPLC column. The column was eluted at a flow rate of 0.5 ml/min with 20 mm TEAM. Fractions were assayed for enzyme activity and protein concentration. The enzyme extract in 25 mmbis-Tris (pH 6.3) was applied to a Mono P chromatofocusing column. Fractions were eluted with 100 ml of diluted Polybuffer 74 (Amersham Pharmacia Biotech) in a linear pH range of 7.0–4.0. Fractions (1 ml) were collected in tubes containing 1 ml of TEA buffer in order to reduce the inhibitory effect of Polybuffer 74 on the enzyme activity. Polybuffer from each fraction was removed after centrifugation using the Centricon 30. Thereafter, each fraction was assayed for enzyme activity or analyzed by SDS-PAGE. Samples from each step of protein purification were analyzed by SDS-PAGE (16Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (207227) Google Scholar). Protein bands were visualized with a silver-staining procedure. Protein concentrations were measured using Coomassie Blue dye reagent (Bio-Rad) with bovine serum albumin as a standard (17Bradford M.M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (216428) Google Scholar). The purified protein fraction containing SA GTase was separated on SDS gel and electroblotted on polyvinylidene difluoride membranes. The protein band was stained with Coomassie Blue and excised after destaining. After elution of the protein from the membrane, the sample was digested with Lys-C endopeptidase. The peptide fragments were then separated by HPLC, followed by peptide sequence analysis. Three of the peptide fragments, SGP1, SGP2, and SGP3, were microsequenced by Harvard Microchemistry Facility (Cambridge, MA). The enzyme assay for SA GTase activity was described previously (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar). One unit of enzyme activity was defined as formation of 1 μmol of either GS or SAG min−1 from SA. From the two internal peptide sequences (SGP1 and SGP3) of the purified SA GTase, two degenerate oligonucleotides, LP1S and LP3A, were designed and used as polymerase chain reaction (PCR) primers. LP1S is 5′-GARGCNGARCARATGGRAGA-3′ and LP3S is 5′-CCCATYTCCCANACRTCYTC-3′. PCR was performed using Taq DNA polymerase as suggested by manufacturer (Life Technologies, Inc.). The PCR product was analyzed by gel electrophoresis and visualized after ethidium bromide staining. The PCR product was recovered from the agarose gel with a gel extraction kit (CLONTECH, Palo Alto, CA), according to the manufacturer's recommendation. The purified PCR fragment was then labeled with [32P]dCTP by random primering and used as a probe to screen 100,000 plaque forming units from an SA-induced λ TriplEx cDNA library constructed byCLONTECH Laboratories, Inc. Forty positive plaques were selected and evaluated on the basis of their insert sizes by PCR. One positive plaque containing the longest insert in λ TriplEx was further purified and converted to the pTriplEx vector viaCre-lox recombination. The insert was sequenced by an automatic sequencer (Applied Biosystems) as recommended by the manufacturer. Total RNA was prepared using Tri reagents (Molecular Research Center, Inc.) as described in the manufacturer's instructions. For RNA gel blotting, total RNA was separated by electrophoresis in 1.3% agarose, 2% formaldehyde gels and blotted on nylon membranes (Zeta probe, Bio-Rad). Cloned DNA probes were labeled with [α-32P]dCTP by a random priming method (18Feinberg A.P. Vogelstein B. Anal. Biochem. 1983; 137: 266-267Google Scholar). Prehybridization, hybridization, and washes were performed according to the manufacturer's protocol (Bio-Rad). The coding region of the SA GTase cDNA was subcloned into the pET30a vector (Novagen, Madison, WI). The recombinant plasmid was transformed into E. coli BL-21 (DE3) (Novagen, Madison, WI). The recombinant protein was then expressed in BL-21 (DE3) after isopropyl-β-d-thiogalactopyranoside induction according to manufacturer's recommendation. The glucosyltransferase activity leading to the formation of either SAG or GS was detected in tobacco plants treated with SA or inoculated with TMV or P. syringae pv.phaseolicola where induction of both enzyme activities was localized to the infected areas (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar, 11Edwards R. J. Plant Physiol. 1994; 143: 609-614Crossref Scopus (45) Google Scholar, 12Enyedi A.J. Raskin I. Plant Physiol. 1993; 101: 1375-1380Crossref PubMed Scopus (88) Google Scholar). Tobacco plants inoculated with TMV and incubated at 32 °C for 4 days showed no change in SA content or glucosyltransferase activity. However, a temperature shift to 24 °C resulted in similar induction of both glucosyltransferase activities in the whole infected leaves (data not shown). Since protein purification requires large quantities of plant material, tissue samples were prepared either from tobacco leaves 6 h after SA infiltration (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar) or from TMV-inoculated tobacco leaves 12 h after temperature shift from 32 to 24 °C, when the SA glucosyltransferase activity was maximum. The enzyme in crude extracts was stable for months when stored at −80 °C. However, the enzyme activity was sensitive to oxidation (data not shown). Therefore, β-mercaptoethanol was always included in the extraction buffer as a reducing reagent. Comparison of relative enzyme activities in different buffer systems showed that the optimum pH for this enzyme is around pH 7.0. SA GTase activity was also relatively high in TEAM buffer at pH 7.3 which is a recommended buffer for Mono Q column. Therefore, TEAM buffer at pH 7.3 was used as the buffer system for the entire purification protocol. Most of the SA GTase activity (>95%) was recovered from the 40–75% fractionation with ammonium sulfate. This step allowed the protein sample to be purified approximately 2-fold and to be concentrated for anion exchange chromatography. The enzyme activity was detected as a single peak after FPLC separation on a Mono Q column. Although the SA GTase had no affinity for UDP-glucuronic acid, some of the UDP-glucuronic acid-binding proteins with molecular masses of 40–50 kDa were eliminated from the extract using a UDP-glucuronic acid affinity column. However, SA GTase was able to bind to a blue dye-ligand affinity column, providing a considerable purification. Glucosyltransferase substrates, SA or UDP-glucose, were unable to elute the glucosyltransferase from the dye affinity column (data not shown). Only UDP, the end product of the reaction, eluted the enzyme with relatively high specificity. Because UDP inhibited the enzyme activity in the eluted fraction, UDP had to be removed from the sample by Centricon filtration. The SA GTase could also be eluted from the blue dye-ligand affinity column with a salt gradient, along with many other proteins (data not shown). Therefore, gel filtration chromatography was used as the final purification step to eliminate proteins with high molecular mass. A summary of the efficiency of the purification process is provided in Table I. The protein patterns at various steps of this purification are shown in Fig. 1. After the final purification step, SA GTase protein appeared as a dominant single peptide on a silver-stained gel.Table IPurification of SA GtasePurification stepTotal proteinTotal activityaOne unit is defined as the formation of 1 μmol · min−1 of GS from SA.Specific activityEnrichmentRecoveryμgunitsunits mg −1-fold%Crude extract50,0001,85031.0100(NH4)2SO440–75%31,1001,775571.595Ion-exchange2102171,03228.011UDPGA affinity2102421,15031.511Dye affinity17.51377,850195.07Gel filtration0.43479,5002,000.02a One unit is defined as the formation of 1 μmol · min−1 of GS from SA. Open table in a new tab In order to determine whether the SA GTase is a monomer or a multimer, the molecular masses of denatured and native forms of this protein were estimated by SDS-PAGE and gel filtration. The molecular mass of the denatured protein was around 48 kDa in SDS gel when compared with protein standards (data not shown). Similarly, the native form of the enzyme had an apparent molecular mass of 47 kDa estimated by gel filtration on Superose 12, indicating that the glucosyltransferase is a monomeric protein. The SA GTase activity is induced by SA (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar). However, it was never determined whether the enhanced glucosyltransferase activity is due to activation of a pre-existing enzyme or due to an increase in the de novo biosynthesis of the enzyme. In order to distinguish between these possibilities, the glucosyltransferase activity was partially purified from the leaves of both the TMV-resistant tobacco cultivar (Xanthi-nc NN genotype) and the TMV susceptible tobacco cultivar (Xanthi-nc nn genotype) after SA treatment (Fig. 2, lanes 2 and4) by a combination of ammonium sulfate fractionation, anion exchange, and affinity chromatographic steps. SDS gel analysis showed that the 48-kDa protein band attributed to SA GTase was strongly induced in both tobacco plants treated with SA (Fig. 2, lanes 2 and 4), compared with that in healthy control plants (Fig. 2, lanes 1 and 3). The staining intensity of the putative SA GTase protein in the gel also correlated with the enzyme activity in respective fractions. The purified enzyme was tested for substrate specificity using various nucleotide sugars as glucose donors and phenolic compounds as glucose acceptors. Only UDP-glucose served as the glucose donor for the SA GTase. None of the other tested nucleotide diphosphate glucose compounds or UDP-glucose analogues functioned as sugar donors for the enzyme (Table II). In contrast, the SA GTase possessed a range of substrate specificities toward phenolic compounds forming glucose esters with the carboxyl groups. All the tested phenolics were glucosylated, while plant hormones such as IAA, GA3, and zeatin were not glucosylated. Interestingly, the enzyme showed the highest activity with benzoic acid, a biosynthetic precursor of SA, among tested phenolics.Table IISpecificity of the purified enzyme for glucose acceptors and donorsSubstrateRelative activityGlucose acceptorsSalicylic acid1.0trans-Cinnamic acid1.1Benzoic acid5.73-Hydroxybenzoic acid1.14-Hydroxybenzoic acid3.5IAA0GA30Zeatin0Glucose donorsUDP-Glc1.0UDP-GA0UDP-Gal0CDP-Glc0GDP-Glc0TDP-Glc0 Open table in a new tab The pI of the SA GTase was determined by Mono P chromatofocusing chromatography. The enzyme activity was detected between pH 4.3 and 4.4. In addition, the glucosyltransferase activity was slightly increased by the addition of divalent cations such as 10 mm Ca2+, Mg2+, and Mn2+ (data not shown). EDTA (5 mm) appeared to have no effect on either the stability or activity of this enzyme, while Zn2+ (10 mm) had a strong inhibitory effect on the enzyme activity (data not shown). Consistent with UDP-glucose being the sole glucose donor, the SA GTase was inhibited by UDP, an end product of the enzyme reaction (data not shown). The apparent K m for UDP-glucose was 0.27 mm which is similar to the K m values of other known glucosyltransferases. In contrast, the apparentK m for SA was 1–2 mm. The SA GTase was partially purified from tobacco plants treated with 1 mm SA by the combination of ammonium sulfate fractionation, ion-exchange chromatography on a Mono Q column, and dye-affinity chromatography (see Table I). After Lys-C endopeptidase digestion and HPLC separation of the purified protein samples prepared as described under "Experimental Procedures," the amino acid sequences of the three longest peptide fragments were determined by Edman degradation. A 320-base pair PCR fragment was obtained by polymerase chain reaction (PCR) using the degenerate oligonucleotides designed from the amino acid sequences of the two peptide fragments (see "Experimental Procedures"). The identity of the PCR product was confirmed by sequencing. The PCR product was then used to screen a SA-induced cDNA library which was constructed in λ TriplEx with mRNA from tobacco leaves induced with SA for 6 h. Among the 40 positive clones, the longest clone was isolated and sequenced. The nucleotide and deduced amino acid sequence of the full-length clone is shown in Fig. 3. The nucleotide sequence of the cDNA clone contains an open reading frame of 1,377 base pairs that encodes a protein of 459 amino acid residues. The calculated molecular mass (51.3 kDa) of the cDNA encoded protein is close to the molecular mass (∼48 kDa) of the purified glucosyltransferase determined by gel filtration on Superose 12. The cDNA-encoded protein contained domains that perfectly matched the sequences of all three internal peptide fragments of the purified protein. The locations of amino acid sequences corresponding to all peptide sequences of the purified protein are indicated in Fig. 3. The degree of homology between the cDNA-encoded protein and other glucosyltransferases is shown in Fig. 4. The highly homologous proteins are plant glucosyltransferases such as indole acetic acid (IAA) glucosyltransferase (19Szerszen J.B. Szczyglowski K. Bandurski R.S. Science. 1994; 265: 1699-1701Crossref PubMed Scopus (173) Google Scholar) and flavonol glucosyltransferase (20Ralston E.J. English J.J. Dooner H.K. Genetics. 1988; 119: 185-197PubMed Google Scholar). In order to determine if the isolated putative glucosyltransferase gene encodes an active SA GTase enzyme, the coding region of the corresponding cDNA was introduced into an overexpression vector, pET30a, and the recombinant protein was expressed in E. coli. Following the addition of UDP-glucose and [14C]SA, two 14C-glucosylated metabolites of SA were formed only in the reaction mixture containing the isopropyl-β-d-thiogalactopyranoside-induced bacterial lysates expressing the recombinant protein (Fig. 5). The identity of the labeled products as SAG and GS was confirmed by their R F values and following treatment with β-glucosidase and esterase. Treatment with glucosidase resulted in conversion of both metabolites back to SA, while esterase treatment, as expected, was only able to hydrolyze GS back to SA (data not shown). Furthermore, the recombinant protein converted [14C]UDP-glucose to radioactive-labeled SAG and GS when non-radioactive SA was used as the glucose acceptor (data not shown). The E. coli expression data indicate that the cDNA clone encodes functionally active SA GTase. To further confirm the functional identity of the recombinant enzyme, the substrate specificity of the recombinant SA GTase was tested with various phenolics as glucose acceptors. Similar to the SA GTase purified from tobacco, the recombinant protein from E. colishowed a broad range of glucose acceptor specificity (Table III). Consistent with the benzoic acid being the most favored substrate of the purified enzyme, the recombinant protein showed high substrate specificity toward benzoic acid.Table IIISpecificity of the recombinant SA GTase for glucose acceptorGlucose AcceptorGlucosylation siteRelative Activity-foldSalicylic acid−COOH1.0−OH0.8trans-Cinnamic acid−COOH1.3Benzoic acid−COOH6.03-Hydroxybenzoic acid−COOH7.74-Hydroxybenzoic acid−COOH5.5IAA−COOH0.3GA3−COOH0.0Zeatin−COOH0.0 Open table in a new tab It was shown that the SA GTase activities involved in the formation of SAG and GS were induced during pathogen inoculation (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar). We wanted to determine if the enhanced enzyme activity was due to transcriptional activation. In tobacco inoculated with TMV and incubated at 32 °C for 4 days, SA GTase mRNA began to accumulate between 6 and 8 h after temperature shift (32 → 24 °C) and was maintained at high levels for at least 4 h (Fig. 6 A). In P. syringaepv. phaseolicola inoculated tobacco, levels of the corresponding mRNA increased 12 h postinoculation and were maintained 24 h postinoculation when the infected tobacco leaves began to collapse (Fig. 6 B). In TMV- or Pseudomonas-inoculated tobacco, the timing of induction of the SA GTase mRNA correlated with the accumulation of SA (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar), suggesting that SA is involved in the transcriptional activation of the SA GTase mRNA. The SA GTase activity was shown earlier to be rapidly induced in response to SA (10Lee H.-I. Raskin I. Phytopathology. 1998; 88: 692-697Crossref PubMed Scopus (68) Google Scholar). We wanted to perform a more detailed analysis of induction pattern of the SA GTase mRNA following SA infiltration (Fig. 6 C). The levels of SA GTase mRNA began to increase within 1 h with a maximum reached by 2 h, and then decreased rapidly. SA GTase mRNA returned to background expression levels between 6 and 24 h. The rapid inducibility of SA GTase mRNA by SA suggested that this gene might be classified as an early SA response gene. In order to study the effect of CHX on the SA GTase expression, a time course experiment on the SA-treated leaves was carried out in the presence or absence of CHX (Fig. 7). In control tobacco treated with H2O, there were no significant changes in the mRNA levels during the time course, while SA alone temporarily induced the SA GTase mRNA. However, following treatment with CHX alone or a combination of SA and CHX, the mRNA accumulated steadily and remained high for 24 h. As expected, PR1agene, one of the late SA inducible genes, was induced by SA treatment alone (Fig. 7). After treatment with CHX and SA, however, expression of the PR1a mRNA was dramatically reduced, confirming thatPR1a gene expression is blocked in the presence of CHX (21Horvath D.M. Chua N-H. Plant Mol. Biol. 1996; 31: 1061-1072Crossref PubMed Scopus (102) Google Scholar). Salicylic acid glucosyltransferase activity has been observed in many plants including oat and tobacco (11Edwards R. J. Plant Physiol. 1994; 143: 609-614Crossref Scopus (45) Google Scholar, 12Enyedi A.J. Raskin I. Plant Physiol. 1993; 101: 1375-1380Crossref PubMed Scopus (88) Google Scholar). We have purified a SA GTase which forms primarily GS in vitro. The peptide sequences of the protease-digested fragments of the purified protein were homologous to UDP-glucose-binding domains of other known glucosyltransferases, suggesting that the purified protein is a glucosyltransferase. Also, the molecular mass of the purified enzyme was similar to other known glucosyltransferases that range from 40 to 60 kDa. However, Western blot analysis showed that SA GTase was not recognized by antibodies against flavonol GTase from tobacco (a gift of Dr. C.-H. Nam, Rockefeller University) or IAA GTase from maize (supplied by Dr. Bandurski, Michigan State University) or thiohydroximate GTase fromBrassica (a gift of Dr. D. W. Reed, Plant Biotechnology Institute, Canada). In addition, none of those antibodies were able to immunoprecipitate the SA GTase or to inhibit its activity. This immunological data suggested that the SA GTase was different from other isolated glucosyltransferases. The purified enzyme preparation preferentially glucosylated phenolics with the carboxy group, suggesting that a carboxy group and the structure of the aglycone (benzene ring) are necessary for glucosylation. The broad substrate specificity of the enzyme for SA as well as its precursors, benzoic acid and trans-cinnamic acid, suggests that SA GTase may have an additional role in regulating the metabolism of SA precursors. The identity of the clone was confirmed by the following observations. (i) The deduced amino acid sequence of the cDNA clone contained domains that perfectly matched the amino acid sequences of all three peptide fragments from the purified protein. (ii) The sequence of SA GTase gene was homologous to other sequenced glucosyltransferases and contained the characteristic highly conserved regions. (iii) Recombinant proteins expressed in E. coli glucosylated SA and showed a comparable substrate specificity of the purified native protein. The purified enzyme fraction displayed a major GS forming activity and lesser SAG forming activity. However, we could not exclude the possibility that SAG formation was due to another co-purified glucosyltransferase. The isolation of the cDNA clone followed by its overexpression in E. coli clearly confirmed that the enzyme possesses both activities. However, it remains to be determined how the dual enzyme activity is regulated in vivo, since, unlike the purified native enzyme, the recombinant enzyme formed GS and SAG in almost equal proportions. One possibility is that the ratio of both activities is regulated by cellular factor(s) which are present only in tobacco plants. Alternatively, the enzyme activities might be influenced by post-translational modification, which became altered inE. coli. As expected, in tobacco plants carrying the N gene, SA GTase mRNA was induced by bacterial or viral pathogens. The induction of SA GTase mRNA paralleled the increase in endogenous SA in the inoculated tissue. Indeed, SA GTase mRNA was activated by SA treatment, indicating that the SA GTase gene is induced through the SA-mediated signal transduction pathway. Furthermore, the correlation between the increased quantity of the SA GTase mRNA and the enzyme activity suggests that the induction of SA GTase activity is primarily due to the de novo synthesis of this enzyme. The relative lack of CHX effect on SA GTase transcription suggests that the pre-existing transcriptional factors are involved in activation of this gene in vivo. This phenomenon has been found in many other early response genes in plants and animals (23Almendral J.M. Sommer D. MacDonald-Bravo H. Burckhardt J. Perera J. Bravo R. Mol. Cell. Biol. 1988; 8: 2140-2148Crossref PubMed Scopus (408) Google Scholar, 24Abel S. Theologis A. Plant Physiol. 1996; 111: 9-17Crossref PubMed Scopus (595) Google Scholar). Recently, some of the early SA responsive genes were cloned from tobacco and characterized on the basis of the timing of mRNA induction in the presence of SA and/or CHX (25Horvath D.M. Huang D.J. Chua N.-H. Mol. Plant-Microbe Interactions. 1998; 11: 895-905Crossref PubMed Scopus (67) Google Scholar). However, their functions were not determined. The role of SA in disease resistance has been the subject of intensive investigation for decades but still requires further elucidation. The successful cloning of the SA GTase gene will provide means to define how the process of glucosylation is involved in the regulation of SA metabolism and in the general pathogen defense responses as well as to give insights into basic mechanisms of disease resistance. Only with a thorough understanding of the mechanisms of plant pathogen interactions together with usage of SA GTase cDNA clone as a tool will it be possible to regulate plant disease resistance in field-grown crops. We thank Dr. David Ribnicky for critical reading the manuscript and Dr. Alex Poulev for technical support during the preparation of the manuscript.