INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

There are certain plants, such as Thlaspi goesingense, that have the ability to accumulate concentrations of Ni in their shoots far exceeding those observed in the soil, without suffering the detrimental effects of Ni toxicity (Reeves and Brooks, 1983). Brooks et al. (1977) first used the term "hyperaccumulator" to describe plants that contain >1,000 µg g¹ (0.1%) Ni in their dried leaves, a concentration at least an order of magnitude higher than Ni levels in nonaccumulator species. Progress is being made in understanding the mechanisms involved in metal hyperaccumulation, but the molecular basis of this intriguing phenomenon remains elusive (Salt and Krämer, 1999).

Recent work on the mechanism of Ni hyperaccumulation in T. goesingense has established that Ni tolerance is a primary determinant of the hyperaccumulation phenotype in hydroponically cultured plants (Krämer et al., 1997). An important component of this Ni tolerance mechanism appears to be based on the efficient intracellular compartmentalization of the Ni into the vacuole (U. Krämer and D.E. Salt, unpublished data).

Recently, Krämer et al. (1996) observed a 36-fold increase in the concentration of free His in the xylem exudate of the Ni hyperaccumulator Alyssum lesbiacum after exposure to Ni. However, no significant change was observed in the nonaccumulator Alyssum montanum. The authors also observed a significant linear correlation in the xylem exudate concentrations of free His and Ni in several Ni hyperaccumulators in the genus Alyssum (Krämer et al., 1996). Because His is an effective chelator of Ni at cytoplasmic pH (Dawson et al., 1986), the authors suggested that His may be involved in chelating Ni during the transport and or storage of Ni in the Alyssum Ni hyperaccumulators. This was supported by the use of x-ray absorption spectroscopy, which identified putative Ni-His complexes in the xylem sap, root, and shoot tissue of A. lesbiacum (Krämer et al., 1996). However, the mechanism by which Ni hyperaccumulation is achieved through the action of His has not been established.

Both Alyssum and Thlaspi Ni hyperaccumulator species are members of the Brassicaceae family, suggesting that free His may also play a role in the mechanism of Ni hyperaccumulation in T. goesingense. This is supported by the recent identification of putative Zn-His complexes in the roots of the closely related Zn hyperaccumulator Thlaspi caerulescens (Salt at al., 1999b). Therefore, to determine if free His is involved in Ni hyperaccumulation in T. goesingense, we investigated the regulation of His biosynthesis at both the molecular and biochemical levels in T. goesingense.

To determine if Ni regulates the expression of genes involved in His biosynthesis in T. goesingense, we cloned genes encoding ATP phosphoribosyltransferase (THG1), imidazoleglycerol phosphate dehydratase (THB1), and histidinol dehydrogenase (THD1), enzymes that catalyze potentially rate-limiting steps in His biosynthesis.

Previously, several authors have published the sequences of various His biosynthetic genes, including HisD (encoding histidinol dehydrogenase [HDH]) from Arabidopsis (Bevan et al., 1998) and Brassica oleracea (Nagai et al., 1991); HisB (encoding imidazoleglycerol phosphate dehydratase [IGPD]) from Pisum sativum (Kim and Theologis, 1996), Triticum aestivum, and Arabidopsis (Tada et al., 1994); HisC (encoding L-histidinol phosphate aminotransferase [HPA]) from Nicotiana tabacum (El Malki et al., 1998); and HisIE (encoding phosphoribosyl-ATP pyrophosphohydrolase/phosphoribosyl-AMP cyclohydrolase [PR-ATP/PR-AMP]) from Arabidopsis (Fujimori and Ohta, 1998). However, to our knowledge, this is the first study to isolate ATP phosphoribosyltransferase (ATP-PRT), the enzyme that catalyzes the first committed step in His biosynthesis, from plants.

To examine the effects of Ni exposure on His biosynthesis, we also measured the concentration of free His in the roots, shoots, and xylem sap in both the Ni hyperaccumulator T. goesingense and the nonaccumulator Thlaspi arvense. Additionally, we quantified putative Ni-His complexes in both T. goesingense and T. arvense using x-ray absorption spectroscopy.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant Growth Conditions and Ni Treatment

For cDNA library construction, Thlaspi goesingense seeds were germinated and grown hydroponically according to the method of Krämer et al. (1997). The plants were exposed to 25 µM Ni(NO₃)₂ for 5 weeks with two exchanges of hydroponic solution per week. After 5 weeks, whole plants were harvested, immediately frozen in liquid nitrogen, and stored at 80°C.

Plants for RNA and genomic DNA studies were grown as follows. Seeds were germinated on filter papers moistened with double-distilled water for 7 d and subsequently transferred to hydroponic culture solution according to the method of Krämer et al. (1997). Plants were maintained on a 12-h d period and day/night temperature of 25°C/20°C, with weekly changes of hydroponic culture solution. Both fluorescent and incandescent lights were used to provide 170 µmol m² s¹ of photosynthetic photon flux (PPF) at the level of the plants. After 6 weeks of growth from the date of transfer into the hydroponic culture solution, plants were exposed to 50 µM Ni for 24 or 48 h by the addition of Ni(NO₃)₂ to the culture solution. Both roots and shoots were harvested separately, immediately frozen in liquid nitrogen, and stored at 80°C. For the efficient isolation of RNA from shoot material, it was necessary to deplete the plants of starch. This was achieved by harvesting treated plants just before the onset of the light period.

His Levels in T. goesingense

For analysis of the free His concentration, root and shoot material was extracted as follows. Approximately 2 g of tissue was frozen in liquid nitrogen and ground to a fine powder in a pestle and mortar. To the frozen powder 6 mL of 3% (w/v) sulfosalicylic acid was added and the slurry ground until the plant tissue had completely thawed. The slurry was then centrifuged at 1,550g for 15 min at room temperature. The supernatant was filtered through a 0.45-µm filter to remove suspended particulate material. Phenylthiocarbamyl derivatization was carried out following the method of Fierabracci et al. (1991). The non-biological amino acid Met sulfoxide was added to 10 µL of tissue extract or xylem exudate to a final concentration of 40 µM and used during analysis as an internal standard. Samples were vacuum-dried (model SVC200, Savant Instruments, Holbrook, NY), redissolved in 20 µL of ethanol:water:TEA (2:2:1, v/v), and vacuum-dried again. To the dried samples 20 µL of ethanol:triethylamine (TEA):water:phenyl isothiocyanate (7:1:1:1, v/v) was added, and the samples incubated at room temperature for 20 min. Samples were vacuum-dried to remove excess reagent and reconstituted in 250 µL of phosphate buffer (pH 7.4). Separation of the phenylthiocarbamyl amino acid derivatives was performed on a Nucleosil C₁₈ 5-µm HPLC column (Sigma, St Louis) (250 × 4.6 mm) at 38°C. The elution solvent consisted of 0.14 M sodium acetate in water plus 0.5 mL/L TEA, titrated to pH 6.4 with glacial acetic acid, with the addition of 60 mL/L acetonitrile (solvent A); and 60% (v/v) acetonitrile in water (solvent B). Phenylthiocarbamyl amino acid derivatives were detected at 254 nm using an UV spectrophotometer (Spectroflow 783, Kratos Analytical, Ramsey, NJ).

Determination of Ni Speciation in Thlaspi

Plant samples were shipped to the Stanford Synchrotron Radiation Laboratory (Stanford University, Stanford, CA) on dry ice. To minimize breakdown and mixing of cellular components within the plant material, care was taken to keep the tissue frozen at all times prior to measurement. To this end, frozen plant tissues were carefully ground under liquid nitrogen and compacted into liquid-nitrogen-cooled 1-mm pathlength lucite sample holders with mylar windows. Aqueous model compounds were diluted by 30% to 50% (v/v) with glycerol (to avoid ice crystal formation) before being pipetted into holders and rapidly frozen in liquid nitrogen. During data collection, samples were held at approximately 15 K using a flowing liquid helium cryostat.

X-ray absorption spectroscopy was carried out on beamline 7-3 of the Stanford Synchrotron Radiation Laboratory using a Si(220) double crystal monochromator, 1-mm upstream vertical aperture, and no focusing optics. Incident intensity was measured using a nitrogen-filled ion chamber, and the absorption spectrum was collected in fluorescence using a 13-element germanium detector (Cramer et al., 1988) by monitoring the Ni K_a fluorescence line at 7,472 eV. Spectra were energy calibrated with respect to a spectrum of Ni foil, and collected simultaneously with the spectrum of each sample, the first energy inflection of which was assumed to be 8,333 eV.

X-ray absorption spectroscopy data reduction was carried out using the EXAFSPAK suite of programs (George, 1998) according to standard methods (Koningsberger and Prins, 1988). Quantitative edge-fitting analysis was performed using the program DATFIT (George et al., 1991). Here, the near-edge spectrum of the plant material is fit using a least-squares algorithm to a linear combination of edge spectra from a library of Ni model compounds. The fractional contribution of each model spectrum to the fit is then directly proportional to the percentage of Ni present in that form in the plant material. By analyzing the total Ni content of the tissue samples, the percentage fractional contribution can then be simply converted into the absolute amount of complex present.

Total RNA Isolation and TriplEx cDNA Library Construction

Total RNA was isolated according to the method of Murphy and Taiz (1995) as follows. Frozen Ni-treated plant material (3-4 g) was ground in liquid nitrogen to a fine powder using a pre-chilled mortar and pestle. The ground tissue was added to 3.5 mL of extraction buffer (1% [w/v] triisopropylnapthylenesulfonic acid, 6% [w/v] p-aminosalicylic acid, 1% [w/v] NaCl, 3% [w/v] polyvinylpyrrolidine, 5% [v/v] -mercaptoethanol, 4 M guanidine thiocyanate, and 25 mM sodium citrate), shaken gently to mix, and incubated for 5 min at room temperature after the tissue had thawed. To the samples, 3.5 mL of dilution buffer (6× SSC, 10 mM Tris, 1 mM EDTA, and 0.25% [w/v] SDS) was added and vortexed for 1 min, then incubated at 65°C for 5 min.

Phenol (6 mL) was added and the mixture was vortexed for 1 min and incubated at room temperature for 5 min. Chloroform (6 mL) was added and the mixture was vortexed for 1 min. Samples were then centrifuged at 5,000g for 10 min at 4°C to separate the phases. The upper aqueous phase was removed, 7 mL of phenol and 10 mL of chloroform were added, and the mixture was vortexed for 1 min, incubated at room temperature for 5 min, and vortexed again for 1 min. The phases were then separated by centrifugation at 3,000g for 10 min at 4°C. The upper aqueous layer was removed and 7 mL of isopropanol was added and the sample incubated for 4 h at 20°C to precipitate the RNA. RNA was collected by centrifugation at 13,500g for 15 min, and the pellet was resuspended in 400 µL of diethyl pyrocarbonate (DEPC)-treated water. The resuspended pellet was then re-extracted in 1 mL of 5 M LiCl by vortexing for 1 min, and RNA was reprecipitated by incubation at 20°C for 45 min. The RNA was pelleted by centrifugation at 27,000g for 25 min at 4°C. The RNA pellet was washed by the addition of 1 mL of 70% (v/v) ethanol and collected by centrifugation at 27,000g for 25 min at 4°C. The final purified RNA was resuspended in 100 µL of DEPC-treated water and stored at 80°C.

The total T. goesingense RNA was sent to CLONTECH Laboratories (Palo Alto, CA), where mRNA isolation, cDNA synthesis, and construction of recombinant Triplex were performed. The cDNA synthesis resulted in 1.6 × 10⁶ independent clones with an average insert size of 2.0 kb (range 0.7-4.0 kb). All T. goesingense cDNAs were cloned into the EcoRI and XhoI sites within the TriplEx multiple cloning site and converted to the phagmid pTriplEx using a cre-recombinase Escherichia coli expressing strain (BM25.8).

The choice of the TriplEx vector was made because of several of its unique features. Each vector has separate initiation codons in two differing frames, both followed by a single poly-dT tract (slip site) that allows either the RNA polymerase or the ribosomes to skip nucleotides, thereby allowing reading in all three frames. Furthermore, the vector incorporates the 5'-untranslated region (UTR) of the ompA gene from E. coli to increase mRNA stability. These features allow every recombinant vector to express some protein, regardless of the initial frame of the insert.

Cell Transformation and Functional Complementation in E. coli

T. goesingense cDNAs encoding functional homologs of the E. coli His biosynthetic enzymes ATP-PRT, IGPD, and HDH were isolated by screening for cDNAs that could complement the His requirement of various His auxotrophic E. coli mutants. E. coli lacking a functional copy of the gene encoding ATP-PRT (strain KL738), IGPD (strain SB3930), or HDH (strain UTH4758) were transformed with the T. goesingense pTriplEx cDNA library by electroporation (model EC100 Electroporator, EC Apparatus, St. Petersburg, FL), per the manufacturer's protocol. Transformed cells were selected on Luria-Bertani medium and ampicillin (100 µg mL¹) agar plates, resulting in 10⁶ to 10⁷ transformants/µg pTriplEx plasmid DNA. The resulting colonies were replica plated onto M-9 minimal media supplemented with 19 amino acids (excluding His), each at a concentration of 25 µg mL¹, 0.1 mM isopropylthiogalactoside, and ampicillin (100 µg mL¹). Colonies that were able to grow were retested for growth on M-9 minimal medium lacking His. Plasmid DNA was isolated from colonies that grew after replating, and the ability of these plasmids to complement the His requirement of the E. coli auxotrophic mutants was confirmed by retransformation of the appropriate His auxotrophic E. coli mutant.

Sequencing and Analysis

Double-stranded DNA from all three genes was completely sequenced using pTriplEx plasmid primers and primers based on previous sequence data using a DNA sequencer (model 373, Applied Biosystems, Foster City, CA) and a dye terminator cycle sequencing ready reaction kit (catalog no. P/N 402078, ABI PRISM, Perkin-Elmer, Foster City, CA).

Predicted translations of the THG1, THD1, and THB1 genes were generated and a search was performed to identify any homologous sequences in GenBank. The THG1 predicted amino acid sequence was aligned with existing ATP-PRT sequences acquired from GenBank using the CLUSTAL_W algorithm (Thompson et al., 1994). The predicted amino acid sequence of THB1 and THD1 was aligned with existing plant sequences using ALIGN (Myers and Miller, 1989). A phylogenetic construction of the resulting aligned sequences for THG1 (ATP-PRT) was also performed using the PHYLIP algorithm (Phylogeny Inference Package, version 3.57c, Department of Genetics, University of Washington, Seattle) (Felsenstein, 1993). Both the alignment and phylogenetic analysis was performed at the University of Illinois Biology Workbench (http://biology.ncsa.uiuc.edu/). Analysis for the presence and type of protein targeting sequence was performed using the PSORT algorithm (Nakai and Kanehisa, 1992) found at http://psort.nibb.ac.jp:8800/. The targeting sequence cleavage site was predicted using the SignalP (version 1.1) algorithm (Nielsen et al., 1997) found at http://www.cbs.dtu.dk/services/SignalP/.

Nucleotide Probe Preparation

One-hour restriction digests of pTriplEx-(THG1) and pTriplEx-(THD1) with XhoI, pTriplEx-(THB1) with EcoRI and XhoI, and an Arabidopsis actin gene (GenBank accession no. U37281) with BamHI and EcoRI were performed at 37°C. The resulting fragments were run on a 1.5% (w/v) agarose gel and the appropriate size fragment was excised for the gel and recovered by electroelution. For THG1 and THD1, the resulting fragments were approximately 400 and 500 bp in size, respectively, and both contained the 3'-UTR of the gene. For THB1, the entire cDNA was recovered. The resulting actin probe was approximately 950 bp and did not contain either 5'- or 3'-UTRs, only the protein coding sequence.

In advance of blot hybridization, 75 to 150 ng of DNA (approximately 5 µL) was denatured at 100°C for 10 min in 35 µL of double-distilled water, and the sample was snap-cooled on ice for 30 s. Then, 2 µL of bovine serum albumin (BSA) (10 mg/mL), 10 µL of 5× OLB buffer (250 mM Tris-HCl [pH 8.0], 25 mM MgCl₂, 0.35% [v/v] -mercaptoethanol, 100 µM each dGTP, dCTP, and dTTP, 1 M HEPES [pH 6.0], and 0.54 µg/µL pdN₆ random hexamers), 2 to 3 µL of [-³²P]dATP (10 µCi/µL), and 5 units of DNA polymerase I Klenow fragment was added to the denatured DNA. The labeling reaction was incubated for at least 5 h at room temperature. The labeled DNA probe was boiled for 10 min, snap-cooled on ice for 30 s, centrifuged for 10 s at 14,000g, and added directly to the hybridization buffer.

Southern Analysis

To obtain genomic DNA, 0.5 to 1.0 g of frozen T. goesingense shoot tissue was placed in a 15-mL centrifuge tube (Falcon 2059, Becton Dickinson, Lincoln Park, NJ), frozen in liquid nitrogen, and ground to a fine powder with a glass rod. Urea extraction buffer (700 µL; 7 M urea, 312 mM NaCl, 20 mM EDTA, 1% [w/v] N-lauroyl sarkosine, and 50 mM Tris-HCl [pH 8.0) was added, and the sample was thawed to room temperature with frequent gentle mixing. Phenol/chloroform (1:1, 500 µL) was added and the sample was incubated for 15 min at 37°C in a rotary shaker. The sample was then transferred to a 1.5-mL microfuge tube and the aqueous phase was separated by centrifugation at 14,000g for 10 min. The upper aqueous phase (approximately 500 µL) was removed and placed in a fresh 1.5-mL microfuge tube. To the aqueous phase 50 µL of 4.4 M ammonium acetate and 700 µL of isopropanol were added, the sample was mixed well, and the genomic DNA pelleted by centrifugation at 14,000g for 1 min. The genomic DNA pellet was resuspended in 500 µL of sterile water and reprecipitated as above. The final DNA pellet was washed once with 70% (v/v) ethanol, spun at 14,000g for 3 min, air-dried for 10 to 15 min inverted on a paper towel, and resuspended in 50 to 100 µL of sterile water.

For genomic Southern analysis, genomic DNA was digested with 60 units of each of the following restriction enzymes: BamHI, EcoRI, HindIII, XbaI, XhoI, or PstI, for 6 to 8 h at 37°C. The resulting fragments were electrophoresed in a 0.8% (w/v) agarose Tris-boric acid-EDTA gel, and capillary blotted onto nylon membranes (catalog no. 80-6221-93, Pharmacia Biotech, Piscataway, NJ) overnight using 10× SSC. The genomic DNA was UV-crosslinked (model FB-UVXL-1000, Fisher Scientific, Loughborough, Leicestershire, UK) to the membrane. The blots were pre-hybridized at 65°C for at least 2 h in 10 mL of a pre-hybridization solution containing 50 mM Tris-HCl (pH 8.0), 10 mM EDTA (pH 8.0), 5× SSC, 5× Denhardt's solution, 0.2% (w/v) SDS, 7.5% (w/v) dextran sulfate, and 100 µg mL¹ sheared salmon-sperm DNA. The blots were then probed with a denatured ³²P-labeled probe added directly to the hybridization solution and the blots incubated at 65°C for 12 to 16 h. After hybridization the blots were rinsed with 50 mL of 2× SSC and 0.1% (w/v) SDS for 5 min at room temperature and then washed two to four times for 15 min at 65°C with 50 mL of 0.1× SSC and 0.1% (w/v) SDS. After washing, the membranes were placed on x-ray film and exposed at 80°C for 10 to 16 h.

Northern Analysis

Total RNA was isolated according to the method of Puissant and Houdebine (1990). Approximately 5 to 10 g of 6-week-old T. goesingense shoots or roots were frozen in liquid nitrogen and ground to a fine powder in a chilled mortar and pestle. The ground tissue was placed in four to six centrifuge tubes each containing 5 mL of GuISCN extraction buffer (4 M guanidinium isothycyanate, 25 mM sodium citrate [pH 7.0], 0.5% [w/v] N-lauroyl sarkosine, and 0.1 M -mercaptoethanol), and the tubes were mixed by inversion.

The samples were mixed with 0.1 volume of 2 M sodium acetate (pH 4.0) and 5 mL of phenol:chloroform (5:1) was added. The samples were mixed and centrifuged at 5,000g for 15 min at 4°C. The aqueous phase (approximately 7 mL) was removed and placed in a fresh 15-mL centrifuge tube, and RNA was precipitated by adding an equal volume of isopropanol at 4°C. The RNA was collected by centrifugation at 4,000g for 10 min at 4°C, and each pellet was resuspended in 2 mL of 4 M LiCl, mixed well, and re-centrifuged at 4,000g for 10 min at 4°C. Each pellet was resuspended in 2 mL of Tris-EDTA buffer containing 0.5% (w/v) SDS, and an equal volume of chloroform was added. After mixing and centrifugation at 4,000g for 10 min at 4°C, the upper aqueous phase was removed and the total RNA precipitated after adding 0.1 volume of 2 M sodium acetate (pH 5.0) and an equal volume of isopropanol. The total RNA was collected by centrifugation at 4,000g for 15 min at 4°C and washed with 70% (v/v) ethanol and 100% ethanol. The samples were air-dried for 15 min, resuspended in 300 µL of sterile, DEPC-treated water, and stored at 80°C.

For northern analysis, 30 µg of total RNA was electrophoresed on 1.2% (w/v) agarose-formaldehyde gels and capillary blotted overnight onto nylon membranes (catalog no. 80-6221-93, Pharmacia Biotech) using 10× SSC. The RNA was UV-crosslinked to the membrane and the blot prehybridized in 10 mL of a pre-hybridization solution containing 200 mM Na₂PO₄ (pH 7.2), 5% (w/v) SDS, 1 mM EDTA, 10 mg/mL BSA, and 0.1 mg/mL sheared salmon-sperm DNA for at least 2 h at 65°C. The blots were probed with denatured -³²P-labeled probes added directly to the hybridization solution and incubated at 65°C for 12 to 16 h. The blots were washed twice for 15 to 20 min at 65°C with 50 mL of a solution containing 40 mM Na₂PO₄ (pH 7.2), 5% (w/v) SDS, 1 mM EDTA, and 5 mg/mL BSA. The blots were then washed for a second time in 50 mL of a solution containing 40 mM Na₂PO₄ (pH 7.2), 1% (w/v) SDS, and 1 mM EDTA for 30 to 35 min at 60°C to 65°C. After washing, blots were placed on x-ray film for 1 to 4 d at 80°C.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Functional Complementation of E. coli His Mutants

T. goesingense cDNAs that encode ATP-PRT, IGPD, and HDH were isolated. These genes were designated THG1, THB1 (Persans et al., 1998), and THD1 (Persans et al., 1998), and their sequences were submitted to GenBank. When expressed in E. coli HisG, HisB, and HisD mutant strains, these cDNAs were able to complement the E. coli mutants' inability to grow in the absence of His (Fig. 1). This suggests that the T. goesingense THG1, THB1, and THD1 cDNAs encode ATP-PRT, IGPD, and HDH, respectively. THG1 is the first cDNA sequence that encodes ATP-PRT to be isolated from plants.

View larger version (63K): [in this window] [in a new window]   Figure 1.   Functional complementation of HisG, HisB, and HisD His auxotropic E. coli mutants with T. goesingense cDNAs. Three T. goesingense cDNAs were isolated by screening a pTriplEx cDNA library for complementation of HisG (A), HisB (B), and HisD (C) mutations in E. coli. The three clones represent ATP phosphoribosyltransferase (THG1, accession no. AF003347), imidazolglycerol phosphate dehydratase (THB1, accession no. AF023140), and histidinol dehydrogenase (THD1, accession no. AF023141). Each plate contained the following: a mutant unable to grow in the absence of His (1); a mutant transformed with complementing T. goesingense cDNA (2); and a wild-type TOPP10F' E. coli strain (3).

The predicted amino acid sequences derived from the cDNA all had high identity to several protein sequences in GenBank. The THG1 amino acid sequence had 83% identity with a putative Arabidopsis ATP-PRT expressed sequence tag (EST) (GenBank accession no. Z31670) amino acid sequence. Also, the THG1 amino acid sequence had 29% and 26% identity to the ATP-PRT from E. coli (HisG) (GenBank accession no. X13462) and Saccharomyces cerevisiae (His1) (GenBank accession no. V01306) amino acid sequences, respectively (Fig. 2A). The THB1 amino acid sequence had 86%, 87%, and 84% identity to an Arabidopsis IGPD (GenBank accession no. 2244848), a Triticum aestivum IGPD (GenBank accession no. 551331), and a Pisum sativum IGPD (GenBank accession no. 2495230) amino acid sequence, respectively. (Fig. 2B). The THD1 amino acid sequence had 89% and 86% identity with Brassica oleracea HDH (GenBank accession no. 60466) (Fig. 2C) and a putative Arabidopsis EST HDH (GenBank accession no. T42850) sequence.

View larger version (3925K): [in this window] [in a new window]   Figure 2.   Alignment of the predicted amino acid sequences of THG1, THB1, and THD1. Sequences used are as follows. A, T. goesingense ATP-PRT (THG1; accession no. AF003347), Arabidopsis EST sequence (At-HisG; accession no. Z31670), E. coli ATP-PRT (Ec-HisG; accession no. X13462), S. cerevisiae ATP-PRT (Sc-His1; accession no. V01306). Green, Amino acids conserved in all aligned sequences; yellow, amino acids conserved in at least two of the aligned sequences; blue, conservative amino acid substitutions. B, T. goesingense (THB1; accession no. AF023140), and Arabidopsis IGPD (At-HisB; accession no. U02689). C, T. goesingense (THD1; accession no. AF023141), and B. oleracea HDH (Bo-HisD; accession no. M60466). Colons, Identity; periods, conservative replacements.

An analysis of the protein leader sequences for all three T. goesingense His biosynthetic genes predicts that both THG1 and THD1 proteins are targeted to the chloroplast, and the THB1 protein appears to be targeted to the mitochondria (Table I).

A phylogenetic tree was constructed to determine the evolutionary placement of the THG1 protein sequence in relation to existing protein sequences (Fig. 3). The THG1 amino acid sequence was closely grouped with an Arabidopsis EST encoding a putative ATP-PRT. The THG1 amino acid sequence is more distantly related to nine archaebacteria, eubacteria, and unicellular eukaryotic ATP-PRT sequences, and falls within its own unique group.

View larger version (14K): [in this window] [in a new window]   Figure 3.   Unrooted phylogenetic tree of T. goesingense ATP-PRT and nine other ATP-PRT sequences. T. goesingense ATP-PRT (THG1; accession no. AF003347), Arabidopsis EST sequence (At-HisG; accession no. Z31670), S. cerevisiae ATP-PRT (Sc-His1; accession no. V01306), Yarrowia lipoytica (Yl-His1; accession no. U40563), Schizosaccharomyces pombe (Sp-His1; accession no. Z70691), Hemophilus influenzae (Hi-HisG; accession no. U32729), Salmonella typhimurium (St-HisG; accession no. X13464), E. coli ATP-PRT (Ec-HisG; accession no. X13462), Archaeoglobus fulgidus (His1-Arcfu; accession no. AE001064), Methanococcus jannaschii (His1-Metja; accession no. U67562), and Methanobacterium thermoautotrophicum (His1-Metth; accession no. AE000911).

Southern Blot of THG1

To examine the number of THG1 genes present in the T. goesingense genome, genomic DNA was digested with various restriction enzymes, run on an agarose gel, and Southern blotted. The genomic DNA was probed with an internal 400-bp XhoI fragment containing the 3'-UTR of the THG1 cDNA (Fig. 4.) Of the restriction enzymes used, BamHI, EcoRI, HindIII, PstI, and XbaI do not cut within the THG1 cDNA sequence and XhoI cuts once. As expected, BamHI, PstI, and HindIII digestion resulted in a single band, while digestion with XhoI resulted in two bands. Interestingly, EcoRI and XbaI digestion produced two bands. This is inconsistent with the restriction map of the cloned THG1 cDNA. This result implies that there may be more than one THG1 gene present in the genome. However, these results are consistent with the assumption that two or perhaps a small family of THG1 genes is present in the genome.

View larger version (120K): [in this window] [in a new window]   Figure 4.   Southern-blot analysis of T. goesingense THG1. Genomic DNA was cut with BamHI (lane A), EcoRI (lane B), HindIII (lane C), XbaI (lane D), XhoI (lane E), and PstI (lane F), and probed with a 32P-labeled THG1 cDNA fragment.

Ni Regulation of THG1, THD1, and THB1 Gene Expression

Northern analysis of total RNA isolated from T. goesingense exposed to 50 µM Ni for 0, 24, and 48 h showed that mRNA levels of THG1, THB1, or THD1 in both the roots and shoots are not affected by exposure to Ni in the hydroponic culture solution (Fig. 5).

View larger version (131K): [in this window] [in a new window]   Figure 5.   Northern analysis of total RNA isolated from T. goesingense exposed to 50 µM Ni for 0, 24, and 48 h showing the expression of THG1, THD1, and THB1. Total RNA was isolated from T. goesingense shoot (A) and root (B) tissue and probed with 32P-labeled THG1 (a), THB1 (b), and THD1 (c) cDNAs probes. Bottom rows in both A and B show blots from total RNA being probed with an Arabidopsis 32P-labeled actin cDNA fragment as a loading control.

Free His Concentrations and Ni Speciation

Free His concentrations in both shoots and xylem sap of the Ni hyperaccumulator T. goesingense did not differ significantly from those observed in the nonaccumulator Thlaspi arvense (Table II). However, His concentrations in the roots of T. goesingense were significantly higher than those observed in T. arvense. Interestingly, after exposure to 50 µM Ni for 7 d, His concentrations in both the shoot and xylem exudate of T. goesingense remained unchanged. However, His concentrations in the roots dropped to levels observed in unexposed T. arvense (Table II). Acid hydrolysis of selected samples showed that there were no increases in the amount of His associated with peptides and proteins (data not shown).

X-ray absorption spectroscopy clearly demonstrated that the amount of Ni coordinated by His in both roots and shoots of T. arvense always exceeded that found in T. goesingense during 1 to 7 d of exposure to 10 µM Ni (Table III). The x-ray absorption edge spectra for Ni-His were significantly different from Ni-imidazole (data not shown); therefore, the x-ray absorption spectroscopy data presented suggest that the Ni-His complex observed in Thlaspi tissues represents Ni coordinated with free His and not His residues in proteins.

View this table: [in this window] [in a new window]   Table III.   Nickel coordination by His ligands in T. goesingense and T. arvense measured by x-ray absorption spectroscopy Values in parentheses represent the total Ni content of the tissue mmol kg1 dry biomass.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

To test the hypothesis that free His plays a role in the mechanism of Ni hyperaccumulation in T. goesingense, we investigated the regulation of His biosynthesis at the molecular and biochemical levels, and studied the role of His in Ni coordination in planta. By functionally complementing His auxotrophic E. coli mutants with T. goesingense cDNAs, we isolated genes (THG1, THB1, and THD1) encoding enzymes that catalyze three steps in the His biosynthetic pathway. The first of these, THG1, encodes ATP-PRT, a functional homolog of an enzyme that catalyzes the production of N-(5'-phosphoribosyl)-ATP (PR-ATP) from ATP and phosphoribosyl pyrophosphate, the first committed step in His biosynthesis in E. coli. This is the first conclusive evidence that ATP-PRT exists in plants and confirms the earlier observation of ATP-PRT-like enzymatic activity in plant tissue extracts (Waiter et al., 1971).

The presence of ATP-PRT in T. goesingense also supports the growing body of evidence suggesting that His biosynthesis in plants follows a very similar pathway to that observed in E. coli (Nagai et al., 1991; Tada et al., 1993, 1994; Kim and Theologis, 1996; Bevan et al., 1998; El Malki et al., 1998; Fujimori and Ohta, 1998). Because THG1 is the first example of a gene encoding ATP-PRT in plants, we performed a genomic Southern analysis to determine how many copies of this gene occur in the T. goesingense genome. This analysis suggested that there are only a small number of THG1 genes present in T. goesingense, similar to the low copy number of genes encoding HDH and IGPD in B. oleracea and Arabidopsis (Nagai et al., 1991; Tada et al., 1994). An investigation of the evolutionary relationships between the T. goesingense ATP-PRT and other archaebacteria, eubacteria, and unicellular eukaryotic ATP-PRT sequences placed the plant ATP-PRT on a separate evolutionary branch from the three other groups (Fig. 3). The T. goesingense ATP-PRT did cluster with an Arabidopsis EST clone (Z31670) that showed 81% amino acid identity to the T. goesingense ATP-PRT amino acid sequence.

Expression of the T. goesingense THG1 in E. coli generated a protein with an apparent molecular mass of 49,000 D, based on SDS-PAGE (data not shown). This corresponds closely to a predicted molecular mass of 49,335 D based on the protein translation product of the fusion of the THG1 cDNA sequence and the expression vector sequence. The predicted molecular mass of the T. goesingense ATP-PRT after cleavage of the putative chloroplast target sequence, at the predicted cleavage site between amino acids residues 45 and 46, was calculated to be 39,056 D. This is similar to the molecular masses of other ATP-PRTs of approximately 32,500 D (Alifano et al., 1996).

In E. coli ATP-PRT is an important control point for His biosynthesis, being regulated at the level of transcription, translation, and allosteric activation/inhibition (Alifano et al., 1996). The highly regulated nature of ATP-PRT in E. coli and its role in the catalysis of the first committed step in His biosynthesis suggests that it might also play a key regulatory role in plants, making it a likely target for the regulation of His biosynthesis by Ni. However, northern analysis of the THG1 RNA message clearly demonstrated that Ni does not induce or suppress transcription of the THG1 mRNA in either the roots or shoots of T. goesingense (Fig. 5). Because it is possible that ATP-PRT is not the key regulated step in His biosynthesis in plants, we also analyzed expression of two other genes in the His biosynthetic pathway, THB1 and THD1, which encode the enzymes IGPD and HDH, respectively. The predicted amino acid sequence of both THB1 and THD1 show very high homology to other known plant homologs, suggesting that the His biosynthetic pathway is highly conserved in plants. The IGPD enzyme catalyzes the conversion of imidazolglycerol phosphate to imidazoleacetol phosphate, the first step after the branch point that feeds into the purine recycling pathway. This enzyme also catalyzes the conversion of L-histidinol phosphate to L-histidinol. Because of its key position at a branch point in the His biosynthetic pathway this enzyme may also be regulated. HDH catalyzes the oxidation of L-histidinol to L-hisitidine, the final step in His biosynthesis. Because this catalytic step uses NAD⁺ as an oxidant, it is possible that it is also regulated in plants.

Northern analysis of the mRNA levels for both THB1 and THD1 clearly showed that expression of these mRNAs is not induced or repressed by Ni treatment in either the roots or the shoots of T. goesingense (Fig. 5). Because THG1, THB1, and THD1 mRNA expression levels were not changed by Ni treatment, it is unlikely that control of the His biosynthetic pathway at the transcriptional level by Ni is involved in Ni hyperaccumulation in T. goesingense.

To determine if Ni modifies His biosynthesis at the post-transcriptional level in T. goesingense, we also analyzed the concentration of free His in root, xylem sap, and shoot tissue. It is clear from this data (Table II) that His concentrations remain basically unchanged after Ni exposure in both the xylem sap and the shoots. These biochemical data strongly support the molecular evidence that free His concentrations in T. goesingense are not increased by Ni exposure. It is possible, however, that the constitutive concentration of free His observed in T. goesingense is sufficient to fulfill its theorized role in Ni hyperaccumulation. To test this hypothesis, we compared the His concentration in T. goesingense and the nonaccumulator T. arvense. This comparison revealed that the nonaccumulator T. arvense contained equal concentrations of His in roots, shoots, and xylem sap as that found in T. goesingense during Ni exposure.

The His concentration in the xylem sap of T. goesingense after Ni exposure was also similar to that measured in the nonaccumulators Vitis rotundifolia and Lagerstroemia indica (Anderson and Brodbeck, 1989; Anderson et al., 1993). However, we did observe that roots of T. goesingense before Ni exposure had a significantly higher concentration of free His compared with T. arvense. An interesting possibility is that His is overproduced in the roots of T. goesingense as a Ni-scavenging mechanism to enhance Ni acquisition under low external Ni conditions. Once plants were exposed to a higher Ni concentration, the free His concentration in the roots of the hyperaccumulator were observed to significantly decrease (Table II), and this may reflect the fact that Ni scavenging is no longer required. However, this reduction in His concentration in the roots of the hyperaccumulator was not reflected in reduced expression of THG1, THB1, or THD1. Also, this His loss was not accounted for by increased His concentrations in xylem sap or shoots. It is possible that reduced His concentrations may reflect increased catabolism or efflux of His into roots. However, recent analysis of T. goesingense root exudate showed no increases in the rates of His exudation from roots after exposure to Ni (Salt et al., 1999a).

If free His is involved in the hyperaccumulation of Ni, as has been suggested to occur in Alyssum species (Krämer et al., 1996), we would predict that His binds Ni within the plant. To directly address this hypothesis we used x-ray absorption spectroscopy to determine the in planta coordination environment of the Ni in both the hyperaccumulator and nonaccumulator Thlaspi species (Table III). From these data it was clear that free His or a free-His-like molecule is involved in coordinating Ni in both the roots and shoots of T. goesingense and T. arvense. However, the concentration of the Ni-His complex in the shoots of the nonaccumulator T. arvense appears to be approximately 5- to 10-fold higher than in T. goesingense, and equal in the roots, suggesting that increased Ni coordination by His is not a primary determinant of the Ni hyperaccumulation mechanism in Thlaspi. Furthermore, the addition of L-His to the culture solution of hydroponically grown T. arvense had no effect on the accumulation of Ni in either the root or shoot tissue (data not shown).

Our data suggest that Ni hyperaccumulation in T. goesingense is not simply related to an enhanced ability of the hyperaccumulator to accumulate more free His in response to Ni. We would also caution that the role of free His in Ni hyperaccumulation in Alyssum remains speculative and will remain so until more detailed mechanistic data are available. For example, using Arabidopsis probes for northern and western analysis in the Ni hyperaccumulator Alyssum pintodasilvae (Baker and Brooks, 1989), expression levels of the His biosynthetic enzymes ATP-PRT, IGDH, and HDH were found not to be regulated by Ni (data not shown).

However, there is certain limited evidence suggesting that His may be involved in Ni transport in Thlaspi species in general. For example, in this study a significant proportion of root Ni was found to be coordinated by free His ligands in both hyperaccumulator and nonaccumulator Thlaspi species. Also, exposure of the nonaccumulator T. arvense to D-His was observed to reduce the accumulation of Ni in shoots, but was found to have no effect on root Ni concentrations. This suggests that the D-His-Ni complex may compete with an endogenous L-His-Ni complex for transport to the shoot. However, we would again like to stress that the primary determinant of the Ni hyperaccumulation phenotype in T. goesingense is not governed by the overproduction of free His, as has been suggested for Alyssum Ni hyperaccumulators.