INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plants are exposed to various environmental influences during their life cycle, which may present different forms of stress, e.g. attacks by herbivores that lead to mechanical injuries but also mechanical stress by wind or touch. Invasion of pathogens has to be prevented and stress by high light intensities or UV light has to be met. Most plants react by activating the transcription of a set of appropriate genes whose gene products either have a defensive role per se or help to repair the effects resulting from stress. The particular reaction is dependent on the kind of stress and the developmental stage of the plant and can involve the strengthening of the cell wall (Hause et al., 1994; Kaiser et al., 1994) or the production of secondary plant products, such as phytoalexins or alkaloids (for overview, see Harborne, 1988).

Many of these induced plant responses are mediated by octadecanoids, of which jasmonic acid (JA) and its precursor 12-oxo-phytodienoic acid (OPDA) were identified as the major signaling compounds (Farmer and Ryan, 1990; Weiler et al., 1994; Parchmann et al., 1997). Upon wounding, the endogenous level of JA increases (Bowles, 1991; Berger et al., 1996; Laudert et al., 1996; Laudert and Weiler, 1998), a phenomenon that can be observed also in plant tissue cultures upon elicitation (Gundlach et al., 1992), whereas tendril coiling in Bryonia dioica is accompanied by an increase of OPDA (Blechert et al., 1999). Although the biosynthesis of jasmonates from linolenic acid is well understood (Vick and Zimmerman, 1984) and many of the enzymes of the pathway have been cloned and characterized (for compilation, see Weiler et al., 1999), the mechanisms of regulating octadecanoid production and the octadecanoid signaling pathway are less well understood.

The non-host-specific phytotoxin coronatine is produced by several cultivars of Pseudomonas syringae and induces leaf chlorosis, inhibits root growth, and induces the production of ethylene and of several proteins, e.g. proteinase inhibitors (Pautot et al., 1991; Bohlmann et al., 1998). Coronatine has been shown to be an essential factor in the early stages of infection of Arabidopsis and presumably acts by suppressing defense-related genes (Dong et al., 1991; Mittal and Davis, 1995). On the other hand, coronatine seems to mimic several effects of octadecanoids in plants and is thought to be an octadecanoid analog mediating and inducing defense reactions (Weiler et al., 1994). Mutants of Arabidopsis (coi1) that are insensitive to coronatine are also insensitive to methyl jasmonate (MeJA) and resistant to infection with P. syringae. The MeJA-insensitive mutants jin1 and jin4 show reduced responses to wounding (Feys et al., 1994; Benedetti et al., 1995; Benedetti et al., 1998). Thus, there seems to be a common site of action for coronatine and octadecanoids. An identification of genes induced or suppressed by coronatine therefore might yield new genes that are regulated by octadecanoids as well and further enhance our understanding of the physiological role of the octadecanoids in the plant.

In this paper, we report the isolation of a Tyr aminotransferase (TAT) from Arabidopsis obtained by differential RNA display from coronatine-treated plants, the first characterization of the heterologous overexpressed protein and the regulation of the enzyme's activity in the plant.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Identification of Coronatine-Regulated Genes

The differential RNA display originally identified 28 differential bands from coronatine-treated plants. Eight clones, designated CORI2 to CORI9 (CORI = coronatine induced), turned out to represent mRNAs, which show different levels of expression after treatment of plants, one of which, CORI3, is characterized in detail in this paper.

Isolation of a cDNA Clone Homologous to TAT

The original CORI3 fragment had a size of 250 bp and thus was relatively short. Screening of the cDNA library from Arabidopsis yielded the clone pZL1pCORI3 (1.1 kb) which lacked the 5' end. The full-length clone pZL1CORI3 was generated via PCR and had a size of 1,460 bp, encoding a putative protein of 422 amino acids (AtTAT1; Fig. 1; gene bank accession no. AF268090).

View larger version (77K): [in this window] [in a new window]   Figure 1.   Alignment of the deduced amino acid sequence for the TAT from Arabidopsis with those from different species using CLUSTALW multiple sequence alignment (Higgins et al., 1996). Regions identical to all TATs are boxed in black. AtTAT1, TAT from Arabidopsis, this work; ratTAT, rat TAT (Hargrove et al., 1989); TrypTAT, TAT from Trypanosoma cruzi (Botempi et al., 1993); RhizTAT, TAT from Rhizobium meliloti (Rastogi and Watson, 1991); HumTAT, human TAT (Rettenmeier et al., 1990). The putative transit peptide of AtTAT1 is underlined.

BLAST (Altschul et al., 1990) database searches (National Center for Biotechnology Information, Bethesda, MD) revealed that the amino acid sequence deduced from the nucleotide sequence of pZL1CORI3 exhibited a 35% identity to TATs from humans and rats (Fig. 1; Rettenmeier et al., 1990).

In addition, the Arabidopsis genome sequencing project released a gene sequence (accession no. CAA 23026) of a TAT-like gene which encodes the CORI3 sequence. The gene has seven exons and six introns and the nucleotide sequence of the gene and the cDNA is identical except for four nucleotides. These changes do not lead to different amino acids. The PSORT program (Nakai and Horton, 1999) predicted a putative transit peptide of the CORI3 protein (AtTAT1) with a cleavage site between amino acid 36 (C) and 37 (D; Fig. 1) and suggested a localization of the protein in the chloroplast.

DNA-Blot Analysis

DNA-blot analysis was conducted to examine the number of copies of the TAT gene in Arabidopsis. Total DNA was digested with restriction enzymes XhoI and BglI, which do not cut, with XbaI, which cuts once and with HindIII, which cuts twice within the cDNA of the CORI3 cDNA. Hybridization of the blot with ³²P-labeled pZL1CORI3 cDNA was done at different stringencies with the following results: As expected, restriction with XhoI and BglI resulted in only one band each and XbaI and HindIII gave two and three bands, respectively. However, as can be seen in Figure 2, there was one additional weak band after restriction with XhoI and two weak bands with XbaI, possibly from other TAT-like genes.

View larger version (58K): [in this window] [in a new window]   Figure 2.   DNA blot of Arabidopsis genomic DNA. DNA (20 µg) was digested with XbaI (X), HindIII (H), XhoI (Xh), or BglI (B) and resolved on 0.6% (w/v) agarose gel. Hybridization was with the 32P-labeled pZL1CORI3 cDNA clone at 65°C. The positions of size markers (in kilobase pairs) are indicated at the left.

A comparison of TAT-like genes published from the Arabidopsis genomic sequencing project shows that another gene with 78% similarity of the amino acid sequence of AtTAT1 could be detected (accession no. CAA 23025, chromosome 4). Four other TAT-like genes (accession nos. CAA 16881, chromosome 4; AAD 23027, chromosome 2; AAD 21706, chromosome 2; and BAA 96891, chromosome 5) have been identified with 55% to 44% homologies to the AtTAT1 amino acid sequence. For the coding regions of CAA 23025 and AAD 23027, the PSORT program predicts putative transit peptides and plastidic localization.

Expression of the CORI3 Gene by Treatment with Octadecanoids and Hormones

Accumulation of CORI3-specific mRNA was found not only for plants sprayed with coronatine but also with methyl-OPDA (OPDAMe) and MeJA and was dosage dependent. Whereas untreated control plants gave only a weak signal of a size of 1.7 kb, the treated plants were induced already at concentrations as low as 0.5 µM, with maximum induction at 5 µM for coronatine, 20 µM for OPDAMe, and 50 µM for MeJA (Fig. 3).

View larger version (67K): [in this window] [in a new window]   Figure 3.   CORI3 mRNA accumulation after treatment with octadecanoids. RNA-blot analysis of total RNA (20 µg per lane), isolated from Arabidopsis plants 2 h after treatment with different concentrations of coronatine (A), MeJA (B), and OPDAMe (C). Concentrations are given in micromoles on top of the blots. Hybridizations were carried out with the 32P-labeled pZL1CORI3 cDNA. Uniform loading was checked by comparing RNA intensities after ethidium bromide staining (not shown).

When other hormones were applied, only abscisic acid gave a slight induction, whereas gibberellic acid, cytokinin, indolylbutyric acid, and ethephon did not cause any reaction at the given concentrations (Fig. 4). When Arabidopsis leaves were wounded a strong accumulation of the CORI3 mRNA was observed.

View larger version (64K): [in this window] [in a new window]   Figure 4.   CORI3 mRNA accumulation. Conditions are given in Figure 3. Results are from two different blots. C, Control plants; Co, 5 µM coronatine; J, 50 µM MeJA; O, 20 µM 12-OPDAMe; A, 100 µM ABA; E, 7 mM ethephon; W, wounded plants; S, 1 mM salicylic acid; Y, 25 µg mL1 yeast elicitor; I, 50 µM IBA; Cy, 50 µM cytokinin; G, 50 µM gibberellic acid.

Heterologous Expression of TAT

CORI3 protein was expressed in Escherichia coli to conduct a functional characterization of the putative TAT. Cloning of the full-length CORI3 cDNA from pZL1CORI3 into the expression vector pET21a(+) with a C-terminal His-tag led to the construct pET21a(+)CORI3. Another plasmid [pET21a(+)pCORI3] was made, in which the potential transit peptide sequence in CORI3 had been removed. Both constructs were transformed into BL21p-Lys E. coli cells. Coomassie staining of SDS-PAGE gels containing total protein extract from induced bacterial cells did not detect overproduction of the protein (Fig. 5A). However, immunoblot analysis using the AtCORI3 antiserum clearly showed a band of the expected size of 47 kD for pET21a(+)CORI3 and a band of approximately 43 kD for pET21a(+)pCORI3 (Fig. 5B).

View larger version (17K): [in this window] [in a new window]   Figure 5.   Expression of Arabidopsis TAT in E. coli cells. Soluble protein extracts were obtained from E. coli cells transformed with the plasmids pET21a(+) as control (lane 1), pET21a(+)pCORI3 (without potential transit peptide; lane 2), and pET21a(+)CORI3 (full-length TAT; lane 3). Crude protein extracts were from cells harvested 40 h after induction with 1 mM isopropyl -D-thiogalactopyranoside (IPTG) and grown at 6°C. They were subjected to SDS-PAGE (10% [w/v]; 10 µg protein/lane) and stained with Coomassie Blue (A) or analyzed by immunoblot with AtCORI3 antiserum (B). Expressed, His-tagged proteins from these extracts were purified by Ni-NTA agarose chromatography, eluted with imidazole, separated by SDS-PAGE (10% [w/v]), and were either Coomassie stained (C, 5 µg of protein applied per lane) or analyzed by immunoblotting using AtCORI3 antiserum (D, 5 µg of protein, lanes 1 and 2; 2 µg of protein, lane 3). Molecular mass markers are given in kilodaltons.

As the proteins possessed a His-tag, they were purified on nickel-nitriloacetic acid (Ni-NTA) agarose columns under native conditions. Figure 5, C and D, show that the 47-kD full-length CORI3 protein was 75% to 80% pure, whereas the 43-kD protein was enriched to a lesser extent after purification. Activity measurements for TAT of the purified proteins from both constructs showed that, when expression of the protein after induction with IPTG was done at 32°C for 3 to 4 h, activity was relatively low with 5 to 12 nanokatals (nkat) mg¹ protein (Table I). Longer incubations at this temperature caused a decrease of the activity, probably due to proteolysis of the protein and lysis of the cells. When temperatures during induction were lowered to 6°C and induction times prolonged, the amount of soluble, full-length protein increased considerably, with a specific activity up to 55 nkat mg¹ protein. The amount of protein without the putative transit peptide remained lower (Table I).

Native gel electrophoresis followed by immunodetection revealed that at 32°C only a small amount of the protein was present as a dimer. The majority was detected as higher M_r multimers or aggregates (data not shown).

Regulation of TAT in Arabidopsis Plants

Immunodetection of proteins after SDS-PAGE of plant extracts from Arabidopsis leaves with the AtCORI3 antiserum revealed a positive signal of a protein with an apparent molecular mass of 47 kD (Fig. 6A), which would correspond to the size of the protein deduced from the full-length CORI3 cDNA. The plant protein had the same size as the recombinant, full-length protein produced in E. coli (data not shown). This result suggests that in plants the putative transit peptide is not processed. Under native conditions a band of a molecular mass of approximately 98 kD was detected, suggesting that in plants TAT is active as a dimer (Fig. 6B). The monomer was not detectable (data not shown).

View larger version (19K): [in this window] [in a new window]   Figure 6.   Molecular mass determination of TAT in Arabidopsis. On immunoblots of plant protein extracts (20 µg) after SDS-PAGE, the antiserum AtCORI3 detected a single band of 47 kD (A). Blue native gel electrophoresis showed only one immunoreactive band of 98 kD (B). The enhanced chemiluminescence (ECL) kit from Amersham was used. Molecular mass markers are given in kilodaltons.

When expression of TAT was checked at the protein level, in non-treated leaves from Arabidopsis rosettes, a weak, basal level could be detected. Also, low basal levels of activity in the range of 0.2 to 0.3 nkat mg¹ protein could be measured (Fig. 7). As had been shown earlier (Fig. 3), levels of TAT mRNA accumulation increased after treatment of plants with coronatine. Activity levels for TAT were enhanced 3 h after application of coronatine and showed a maximum, with a 6- to 7-fold increase, at 9 h to decline slowly afterward (Fig. 7A). Immunodetection of the protein also showed an increase. In this case, the level of the protein was maximal at 12 h and remained constant for about 48 h (Fig. 7C). Activity kinetics and protein level measurements were quite similar when the octadecanoid MeJA was applied, whereas for OPDAMe maximal activity was observed around 7.5 h after application, followed by a dramatic decrease of activity (Fig. 7B). Also, protein accumulation was lower.

View larger version (14K): [in this window] [in a new window]   Figure 7.   Regulation of TAT in Arabidopsis. TAT activity was measured at different time points after treatment of the plants with: A, 50 µM coronatine (white squares) or 100 µM MeJA (black squares); and B, 100 µM OPDAMe (white circles) or after wounding (black circles). Acitivity in untreated control plants is given for comparison in each graph (diamonds). C, Immunodetection of TAT in total protein extracts (20 µg/lane) of plants treated as given above after separation on SDS-PAGE. C, Coronatine; J, MeJA; O, OPDAMe; W, wounded. The time after treatment is given in hours. Detection was with the ECL system. Equal loading was checked by staining the blots with Ponceau (not shown).

Wounding, which leads to an increase of endogenous OPDA and jasmonate levels, also led to induction of TAT activity with a maximum of a 5-fold increase and a time course similar to that shown after treatment with OPDAMe. Again, maximal activity was reached after 7.5 h and then dropped rapidly down to basal levels. TAT protein levels increased, although not as strong as after coronatine treatment, reaching maximal expression between 7 and 12 h and then remained either constant or in some experiments decreased slightly (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plants respond to mechanical injuries and pest attack by inducing genes activated by different signal transduction pathways, which can be jasmonate dependent or independent. Although the role of octadecanoids in herbivore attack seems indisputable, in plant-pathogen interactions their role is less well understood (McConn et al., 1997; Dong, 1998; Vijayan et al., 1998). With the aid of the non-host-specific phytotoxin coronatine, which mimics octadecanoids in several plant processes, it was possible to isolate and characterize several coronatine-inducible genes. The one reported in this study is CORI3, which encodes a TAT. This is the first full-length cDNA from plants for TAT reported. TAT activity of the clone could be verified when the cDNA was heterologously expressed and the protein purified over Ni-NTA agarose columns. The amount of soluble full-length protein increased considerably when induction with IPTG and growth prolonged for longer time at 6°C (Table I). In the plant, native TAT obviously forms a dimer of 98 kD (Fig. 6B) and folding of the protein and formation of the complex might be better at lower temperatures. Expression of proteins forming homo- or heterodimers often requires lower growth temperatures; e.g. geranyl diphosphate synthase, a heterodimer, is expressed in bacteria at 15°C (Burke et al., 1999).

CORI3 mRNA levels were increased by different octadecanoids of the jasmonate cascade (OPDAMe and MeJA) as well as by wounding (Figs. 3 and 4). TAT activity in Arabidopsis was also induced with a maximum at 7.5 or 9 h after treatment. TAT induction kinetics and the strength of induction are very similar for coronatine and MeJA with a maximum at 9 h and a relatively slow decrease. Protein levels of TAT increase as well, but then remain constant for at least 24 to 48 h (Fig. 7A). Wounding and treatment with OPDAMe result in very similar kinetics of activity with a maximum at 7.5 h and following rapid decrease to control levels. At the protein level, the accumulation is less strong when compared with coronatine and MeJA, with slight decreases in the amount over time (Fig. 7B). So in this case, coronatine seems to act more like MeJA, whereas wounding and OPDAMe display similar time courses. This is different from induction patterns of allene oxide synthase, an enzyme of the jasmonate biosynthesis (Laudert and Weiler, 1998). Also, allene oxide synthase mRNA accumulates in response to salicylate and ethephon, which is not the case for TAT (Fig. 4). Thus, although both enzymes are induced by treatment with octadecanoids and coronatine, there seem to be differences in induction patterns and separate defense responses because one enzyme belongs to the biosynthetic pathway of jasmonates and the other one is involved in a pathway leading from Tyr to potential defense compounds of the plant.

It is interesting that Titarenko et al. (1997), when using the differential display method to obtain jasmonate- and wound-inducible genes from Arabidopsis, were also able to isolate a cDNA fragment of 489 bp that showed homology to TAT from humans and rat. However, no sequence was published.

In plants, the product formed from Tyr by the reaction of TAT is 4-hydroxyphenyl-pyruvate (Fig. 8). Few reports on this plant enzyme have been published. De-Eknamkul and Ellis (1987) purified TAT from Anchusa officinalis. They could distinguish three different forms as homotetramers with native molecular weights of 180 to 220 kD and relatively high pH optima of 8.8 to 9.6. In A. officinalis, rosmarinic acid is formed via TAT, one of several phenolic compounds protecting the plant against peroxidative damage by scavenging free radicals (Liu et al., 1992). A similar function might well be mediated by TAT in Arabidopsis. It is well known that 4-hydroxyphenylpyruvate is converted to homogentisate by a dioxygenase in Arabidopsis and other plants (Löffelhardt and Kindl, 1979; Fiedler et al., 1982). The dioxygenase has been cloned from several plants (Garcia et al., 1997, 1999; Kleber-Janke and Krupinska, 1997; Norris et al., 1998). Homogentisate is the precursor of both tocopherols and plastoquinones in plants (Fig. 8). The synthesis of the prenylquinones has been shown to be localized at the inner envelope membrane of the chloroplast and both quinones can be found at the outer and inner membrane (Soll et al., 1985). -Tocopherol, the main component of vitamin E and the reduced form of plastoquinone, are known as scavengers for reactive oxygen species (ROS; Hundal et al., 1995; Shintani and DellaPenna, 1998). Upon pathogen attack and stress, chlorophyll breakdown can occur and ROS can be formed that might have to be neutralized by prenylquinones. Higher demand for prenylquinones might lead to an increased activity of TAT, the first enzyme of the pathway.

View larger version (15K): [in this window] [in a new window]   Figure 8.   The biosynthetic pathway of tocopherols and plastoquinones.

In an alternate manner, a more direct role of TAT upon pathogen attack could be that phenolics formed from 4-hydroxyphenylpyruvate might serve as cross-linkers in the cell wall, thereby strenghtening this physical barrier (Hause et al., 1994; Schmidt et al., 1998). Future studies on localization and regulation of TAT under different stress conditions and the analysis of metabolites in transgenic plants exhibiting altered levels of TAT will help to elucidate the role of this enzyme in plant defense and its control by the octadecanoid pathway.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material and Growth Condition

Seeds of Arabidopsis, ecotype C24, were surface sterilized and grown on one-half-strength Murashige and Skoog medium with 0.8% (w/v) agar (Murashige and Skoog, 1962) in a growth chamber at 24°C and a 9-h-light/15-h-dark cycle with a light intensity of 80 µE m² s¹_. Non-sterile Arabidopsis plants were grown in the greenhouse in trays containing standard soil. Four- to 6-week-old plants were used in both cases for further experiments.

For induction experiments, plants from the greenhouse were sprayed with different hormones and octadecanoids at the given concentrations in 0.1% (w/v) Tween 20. Control plants were sprayed with 0.1% (w/v) Tween 20 only. Immediately after harvesting, the plant material was frozen in liquid nitrogen and stored at 75°C. For wounding experiments, 50% to 80% of the leaf area was crushed with a hemostat. Plant material was harvested, frozen in liquid nitrogen, and stored at 75°C. For enzymatic assays, fresh plant material was used. All experiments were done at least three times in independent experiments.

RNA Isolation and Differential RNA Display

For differential display, total RNA from control Arabidopsis plants and plants sprayed with 5 µM coronatine and incubated for 2 h was extracted as described by Barkan (1989). Poly(A⁺) RNA was isolated using the Oligotex mRNA Kit (Qiagen, Hilden, Germany). Differential RNA display was performed as described by Liang and Pardee (1992). Purified poly(A⁺) RNA (0.2 µg) was reverse transcribed in a 20-µL reaction containing 1 µM of the D collection of primers and 20 µM deoxyribonucleotide triphosphates (GeneExScreen Primer Kit, Biometra, Göttingen, Germany). The reaction was carried out as follows: The mixture was heated for 5 min at 65°C and then for 10 min at 37°C, after which 10 units of reverse transcriptase (Promega, Mannheim, Germany) were added. The reaction proceeded for another 60 min and the mixture was then heated at 95°C for 5 min and stored at 20°C until used for PCR.

Two-microliter aliquots were used as templates for PCR with the D primers in combination with an arbitrary primer U (GeneExScreen Primer Kit) in the presence of ³²S-labeled dATP (Amersham-Pharmacia, Freiburg, Germany). Reaction conditions for PCR were as follows: 94°C for 30 s, 40°C for 2 min, and 72°C for 30 s for 40 cycles, followed by extension at 72°C for 5 min. Aliquots (4.5 µL) of the PCR reaction were run on a 6% (w/v) non-denaturing polyacrylamide gel to separate the amplified cDNAs. Duplicates of each reaction were subjected to electrophoresis and only bands that were amplified in a consistent manner were used for further analysis. Bands containing the differentially expressed cDNAs were excised from the gel, eluted in 100 µL of distilled, sterile water for 10 min, and then boiled for 15 min. The supernatant was ethanol precipitated. Each pellet was dissolved in 10 µL of distilled water and an aliquot (4 µL) was used for reamplification with the appropriate pair of primers. The primers that subsequently led to the isolation of the cDNA were D8 and U8 from the GeneExScreen Primer Kit. The PCR products were purified on 6% to 8% (w/v) non-denaturing polyacrylamide gels and subcloned into the pBluescript SK+ vector (Stratagene, Amsterdam). The partial cDNA fragment was further used for library screening and DNA- and RNA-blot analysis.

RNA-Blot Analysis

Total RNA (20 µg) was mixed with an equal volume of loading buffer {50% [v/v] formamide, 10% [v/v] formaldehyde, 0.1 M MOPS [3-(N-morpholino)-propanesulfonic acid], pH 8.0, and 50 µg mL¹ ethidium bromide}, separated by gel electrophoresis on a 1.3% (w/v) agarose-formaldehyde gel (10% [v/v]) and then transferred to Hybond-N membranes (Amersham-Pharmacia) using 20× SSPE (1× SSPE: 0.22 M NaCl, 10 mM NaH₂PO₄, and 1 mM EDTA, pH 7.4) (Sambrook et al., 1989). Hybridization was performed at 65°C with ³²P-random-labeled cDNA probes following standard procedures (Sambrook et al., 1989). For all blots, equal loading was confirmed by ethidium bromide staining of ribosomal RNAs (25S and 18S).

DNA-Blot Analysis

Genomic DNA was isolated from Arabidopsis leaves following the procedure of Pich and Schubert (1993). DNA (20-30 µg) was digested with the appropriate restriction enzymes, separated by gel electrophoresis on a 0.6% (w/v) agarose gel, and blotted onto Hybond-N membranes. Prehybridization and hybridization with ³²P-random-labeled cDNA probes were performed at 65°C according to standard procedures (Sambrook et al., 1989).

Isolation and Cloning of the Full-Length cDNA Clone

The cDNA library (vector, Lambda ZIP LOX; BRL, Carlsbad, CA) from Arabidopsis, kindly provided by Dr. Stefan Trentmann (Ruhr-Universität Bochum, Germany) was screened by colony hybridization using the ³²P-labeled CORI3 cDNA fragment from the differential RNA display as a probe. Approximately 150,000 plaques from the library were screened. Two positive clones were isolated and subcloned in vivo. The sequences of the two clones were determined and they turned out to be identical. The clone was named pZL1pCORI3.

To obtain the missing 5' end of the clone, a Hybri-ZAP-cDNA library (Stratagene) was used and PCR was performed essentially as described in Chiang et al. (1995). The PCR was performed in a 50-µL reaction volume, containing 2.5 µL of the amplified cDNA Hybri-ZAP library with 50 pmol of each primer (primer A, 5'-end of the LambdaZAP-clone; 5'-AGGGATGTTTAATACCACTAC-3', primer CORI3 R, specific for the partial CORI3cDNA; and 5'- TCAGAAACCACCACCATAATCTTGA-3'). The conditions for PCR were as follows: 94°C for 5 min, then 94°C for 1 min, 55°C for 2 min, and 72°C for 2 min for 35 cycles, followed by incubation at 72°C for 10 min. The obtained PCR products were subcloned into the pBluescript vector and religated with pZL1pCORI3. The resulting full-length clone was named pZL1CORI 3.

Preparation of Polyclonal Antiserum

The immunization of a rabbit was done according to Weiler (1986). For overexpression in Escherichia coli, the cDNA from the partial clone pZL1pCORI3 was subcloned into the pQE31 vector (Qiagen). Overexpression and purification under denaturing conditions using Ni-NTA agarose was done as described in the QIAexpressionist handbook (Qiagen). For pre-immunization, the rabbit was injected once per week subcutaneously with 150 µg of purified protein in two volumes of Freund's incomplete adjuvant for 4 weeks, followed by intramuscular injections of 150 µg of protein in Freund's incomplete adjuvant every 3 weeks. Serum was obtained 10 d after each boost and the antiserum AtCORI3, which detected TAT in heterologous bacterial systems and in Arabidopsis, was stored at 20°C.

Cloning of the CORI3 cDNA into the pET 21 Vector

For overexpression of the CORI3 protein in the pET 21a(+) vector (Novagen, Madison, WI), two constructs with and without a potential transit peptide (see "Results") were made. To generate the NdeI and XhoI restriction sites necessary for subcloning into the vector, the full-length pZl1CORI3 clone was amplified using the primers pET21A and pET21R for the full-length construct (pET21A, 5'-GGGATTTCCATATGGCAACCCTTAAGTGCATTGAT-3'; and pET21R, 5'-CCGCTCGAGCTTAACACCATTGACGTCTTT-3') and primers pET21B and pET21R for the construct without the potential transit peptide (pET21B, 5'- GGGATTTCCATATGGATCCTCATGGCAAACCCATT-3'). Both PCR products were restricted with NdeI and XhoI, and cloned using corresponding restriction sites in the multicloning site of the vector pET21a (+). The resulting clones were named pET21a(+)CORI3 (primers pET21A and pET21R, full length), and pET21a(+)pCORI3 (primers pET21B and pET21R, lacking the potential transit peptide). Both constructs contained a His₆-tag at the C terminus.

TAT Activity

For expression of the CORI3 protein (AtTAT1), clones pET21a(+)CORI3 and pET21a(+)pCORI3 were transformed into BL21(DE3) pLysS bacteria (Promega). Induction was with 1 mM IPTG. Cells then were grown for different times and different temperatures until harvest. Harvest and lysis followed the instructions of the manufacturer (pET manual, Novagen) and purification of the protein was done over a Ni-NTA-agarose column (Qiagen) according to the manufacturer's instructions under non-denaturing conditions. Activity of TAT was measured after Granner et al. (1970) at 35°C immediately after purification and protein determination as the protein proved to be unstable. Protein determination was done according to Bradford (1976), using bovine serum albumin as a standard.

For measurements of TAT activity, plants (5 g fresh weight) were homogenized in 2 volumes of ice-cold buffer, containing 100 mM potassium phosphate buffer (pH 7.5), 0.1 mM EDTA, 8.0 mM -ketoglutarate, 0.2 mM pyridoxal-5-phosphate, and 1 mM dithiothreitol. After filtration through cheesecloth and centrifugation at 6,000g, the supernatant was made 0.1% (v/v) with tergitol and a 15-min incubation period on ice followed. The supernatant of the following 30,000g centrifugation was used as crude enzyme extract. Immediately after protein determination, measurement of enzyme activity followed at 35°C, using 50 µg of total protein in an assay mixture containing 0.125 M potassium phosphate buffer (pH 7.5), 0.75 mM EDTA, 5.5 mM Tyr, 10 mM -ketoglutaric acid, and 0.1 mM pyridoxal 5-phosphate. Addition of KOH converted 4-hydroxyphenylpyruvate, the product of TAT activity, to 4-hydroxybenzaldehyde, which was determined spectrophotometrically at 331 nm.

PAGE and Protein Analysis

Determination of protein concentrations was done according to Bradford (1976) with bovine serum albumin as a protein standard. For SDS-PAGE, proteins were separated on a 10% (w/v) separating gel (Laemmli, 1970).

Blue native gel electrophoresis was done according to Schägger and von Jagow (1991).

Transfer of proteins separated by SDS-PAGE onto nitrocellulose was done according to Towbin et al. (1979) and transfer of proteins after blue native PAGE according to Schägger and von Jagow (1991).

Immunodetection was performed after blotting using either the ECL western blotting analysis system (Amersham-Pharmacia) or the alkaline phosphatase detection system (Promega). For ECL, the AtCORI3 antiserum was used at a 1:15,000 dilution and the anti-rabbit secondary antibody was used at a dilution of 1: 5,000 (v/v). With the alkaline phosphatase system, the AtCORI3 antiserum was used at a dilution of 1:20,000 (v/v) for proteins expressed in bacteria and 1:7,500 (v/v) for protein extract from plants. The second antibody, goat-anti-rabbit IgG (Promega) was used at a dilution of 1:7,500 (v/v). Detection was with the substrates nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate.