Anal. Chem. 2003, 75, 2740-2745
Speciation of Nickel in a Hyperaccumulating Plant by High-Performance Liquid Chromatography-Inductively Coupled Plasma Mass Spectrometry and Electrospray MS/MS Assisted by Cloning Using Yeast Complementation Ve´ronique Vacchina,† Ste´phane Mari,‡ Pierre Czernic,‡ Laurence Marque`s,‡ Katia Pianelli,‡ Dirk Schaumlo 1 ffel,† Michel Lebrun,‡ and Ryszard Łobin´ski*,†
Group of Bio-Inorganic Analytical Chemistry, UMR 5034 CNRS, He´ lioparc, 2, Av. Pr. Angot, F-64053 Pau, France, and UMR 5004 (CNRS, Universite´ de Montpellier II, INRA, ENSAM), F-34095 Montpellier Ce´ dex 5, France
A novel analytical approach based on a combination of multidimensional hyphenated techniques and cloning of the Ni-resistance gene using yeast complementation screens was developed for the identification of nickel species in a Thlaspi caerulescens hyperaccumulating plant. The presence of an unknown strong Ni complex was demonstrated by size exclusion HPLC-capillary electrophoresis with ICPMS detection. The Ni-containing peak was characterized by electrospray MS (m/z 360) and shown by collision-induced dissociation MS to be a chelate with a tricarboxylic amino acid ligand. To identify the species and demonstrate its functional character, a cDNA library was constructed from T. caerulescens, expressed in the yeast, and screened on a toxic Ni2+ medium. The extract from the surviving transformant culture gave identical HPLC-ICPMS, CZE-ICPMS, and ES MS/MS data and contained a cDNA insert homologous to the nicotianamine synthase gene. This observation allowed the identification of nicotianamine as the nickel-binding ligand. The presence of the Ni-nicotianamine complex was ultimately demonstrated by comparing tandem mass spectra of the plant and yeast extracts with those of a synthetic standard. The discovery of metal-hyperaccumulating properties in certain plants spurred research toward using them for cleanup of heavymetal-contaminated soils.1-3 The term hyperaccumulator, referring to a plant with a highly abnormal level of metal accumulation, was originally coined to describe a plant with a concentration exceeding 0.1% Ni (dry mass)4 and then extended to other metals, * Corresponding author. Tel: +33-559-80-6884. Fax: +33-559-80-1292. Email:
[email protected]. † UMR 5034 CNRS. ‡ UMR 5004 (CNRS, Universite´ Montpellier II, INRA, ENSAM). (1) Palmer, C. E.; Warwick, S.; Keller, W. Can. Int. J. Phytorem. 2001, 3, 245287. (2) Raskin, I.; Ensley, B. D. Phytoremediation of toxic metals. Using plants to clean up the environment; Wiley: New York, 2000 (3) McGrath, S. P.; Zhao, F. J.; Lombi, E. Adv. Agron. 2002, 75, 1-56. (4) Brooks, R. R.; Lee, J.; Reeves, R. D.; Jaffre´, T. J. Geochem. Explor. 1977, 7, 49-57.
2740 Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
such as Co 5 and Pb 6 (0.1% threshold value) and Zn and Mn (1% threshold values).7 To date, ∼400 known metal hyperaccumulators have been reported worldwide.8 The most widely referred to have been the mustard Thlaspi caerulescens containing 1-2% Zn,9 Alyssum lesbiacum (>1% of Ni),10 and the New Caledonian tree Sebertia acuminata whose latex was reported to contain above 20% Ni (dry mass).11 The successful use of plants to extract metals from contaminated soils requires a better understanding of the mechanisms of metal uptake, translocation within the plant, and accumulation. It requires the identification of molecules involved, in particular bioligands and metal complexes synthesized by a plant to be used for metal transport and suitable for bioaccumulation. In contrast to the plethora of molecular data available on the biosynthesis of phytochelatins in edible plants (e.g., rice, soybean, maize) exposed to Cd or Cu stress,12,13 remarkably little is known on the molecular forms (speciation14) of metals in hyperaccumulating plants. The notable few exceptions include the correlation of free nickel and histidine levels in A. lesbiacum, suggesting a Ni-histidine complex being responsible for the xylem transport,10 and the demonstration of citrate as one of the ligands binding Ni in the S. acuminata latex by NMR.15 In the hyperaccumulator Thlaspi goesingense, histidine and citrate were found as ligands for nickel in leaves while the majority of the Ni was associated with the cell wall.16 (5) Brooks, R. R.; Reeves, R. D.; Morrison, R. S.; Malaisse, F. Bull. Soc. R. Bot. Belg. 1980, 113, 166-172. (6) Reeves, R. D.; Brooks, R. R. Environ. Pollut., Ser. A 1983, 31, 277-285. (7) Baker, A. J. M.; Brooks, R. R. Biorecovery 1989, 1, 81-126. (8) Reeves, R. D.; Baker, A. J. M.; Borhidi, A.; Berazain, R. Ann. Botany (London) 1999, 83, 29-38. (9) Baker, A. J. M.; Reeves, R. D.; Hajar, A. S. M. New Phytol. 1994, 127, 6168. (10) Kra¨mer, U.; Cotter-Howells, J. D.; Charnock, J. M.; Baker, A. J. M.; Smith, J. A. C. Nature 1996, 379, 635-638 (11) Jaffre´, T.; Brooks, R. R.; Lee, J.; Reeves, R. D. Science 1976, 193, 579-580. (12) Zenk, M. H. Gene 1996, 179, 21-30. (13) Cobbett, C. S. Curr. Opin. Plant Biol. 2000, 3, 211-216. (14) Templeton, D.; Ariese, F.; Cornelis, R.; Danielsson, L. G.; Muntau, H.; van Leeuven, H. P.; Łobin´ski, R. Pure Appl. Chem. 2000, 72, 1453-1470. (15) Sagner, S.; Kneer, R.; Wanner, G.; Cosson, J. P.; Deus-Neumann, B.; Zenk, M. H. Phytochemistry 1998, 47, 339-347. (16) Kra¨mer, U.; Pickering, I. J.; Prince, R. C.; Raskin, I.; Salt, D. E. Plant Physiol. 2000, 122, 1343-1353. 10.1021/ac020704m CCC: $25.00
© 2003 American Chemical Society Published on Web 05/01/2003
Recently, three selenium species, Se-methylselenocysteine, Sehomocysteine, and Se-cystathionine, were detected in the Se accumulator Brassica juncea (Indian mustard).17 Another important class of ligands in plants is the phytosiderophoric family of nonproteinaceous amino acids (mugineic acids) including nicotianamine. The latter was proposed as a chelator of iron in symplastic and phloem transport.18,19 and of copper for its mobilization in xylem transport,20 and was shown to be an extremely strong chelator (log K 16.1) of nickel.21 Nicotianamine has a molecular weight of 303 and is a product of the trimerization of S-adenosylmethionine catalyzed by nicotianamine synthase.22 The latter is coded by a gene identified by Suzuki and Mori in Arabidopsis thaliana (accession number AB021934).23 The identification of element species in plant fluids and tissues encounters two major obstacles on the level of analytical chemistry. They are the difficulty of distinguishing between, potentially numerous, metal complexes involved in the metal translocation and bioaccumulation and the lack of standards together with the underdevelopment of standard-free species-specific methodology (e.g., mass spectrometry, NMR) that would work when applied to complex biological matrixes. These problems can be successfully addressed in many cases by hyphenated techniques based on the combination of a chromatographic separation with the parallel element-specific (ICPMS) and molecule-specific (ES MS/ MS) detection as recently reviewed by Szpunar.24 In particular, the potential of this pair of hyphenated techniques was successfully demonstrated for the identification of Al-citrate complex in human serum25 and plant sap,26 organoselenium compounds in yeast,27,28 garlic,29 and indian mustard,17 organoarsenic compounds in marine biota,30-32 and metal-phytochelatin complexes in plants.33 However, interpretation of mass spectra of unknown metallobiomolecules may be difficult and even impossible. (17) Montes-Bayo´n, M.; LeDuc, D. L.; Terry, N.; Caruso, J. A. J. Anal. At. Spectrom. 2002, 17, 872-879. (18) Schmidke, I.; Scholz, G.; Stephan, U. W.; Physiol. Plant. 1995, 95, 147153. (19) von Wire´n, N.; Klair, S.; Bansal, S.; Briat, J.-F.; Khodr, H.; Shiori, T.; Leigh, R. A.; Hider, R. C. Plant Physiol. 1999, 119, 1107-1114. (20) Pich, A.; Scholz, G. J. Exp. Bot. 1996, 47, 41-47. (21) Benes, I.; Schreiber, K.; Ripperger, K.; Kircheiss, A. Experimentia 1983, 39, 261-262. (22) Higuchi, K.; Suzuki, K.; Nakanishi, H.,; Yamaguchi, H.; Nishizawa, N.-K.; Mori, S. Plant Physiol. 1999, 119, 471-479. (23) Suzuki, K.; Higuchi, K.; Nakanishi, H.; Nishizawa, N.-K.; Mori, S. Soil Sci. Plant Nutr. 1999, 45, 993-1002. (24) Szpunar, J. Analyst 2000, 125, 963-988. (25) Bantan, T.; Milacˇicˇ, R.; Mitrovic´, B.; Pihlar, B. J. Anal. At. Spectrom. 1999, 14, 1743-1748. (26) Bantan Polak, T.; Milacˇicˇ, R.; Pihlar, B.; Mitrovic´, B. Phytochemistry 2001, 57, 189-198. (27) Casiot, C.; Vacchina, V.; Chassaigne, H.; Szpunar, J.; Potin-Gautier, M.; Łobin´ski, R. Anal. Commun. 1999, 36, 77-80. (28) Kotrebai, M.; Tyson, J. F.; Block, E.; Uden, P. C. J. Chromatogr., A 2000, 866, 51-63. (29) McSheehy, S.; Yang, W.; Pannier, F.; Szpunar, J.; Łobin´ski, R.; Auger, J.; Potin-Gautier, M. Anal. Chim. Acta 2000, 421, 147-153. (30) Madsen, A.; Goessler, W.; Pedersen, S. N.; Francesconi, K. A. J. Anal. At. Spectrom. 2000, 15, 657-662. (31) McSheehy, S.; Pohl, P.; Ve´lez, D.; Szpunar, J. Anal. Bioanal. Chem. 2002, 372, 457-466. (32) Van Hulle, M.; Zhang, C.; Zhang, X.; Cornelis, R. Analyst 2002, 127, 634640. (33) Vacchina, V.; Łobin´ski, R.; Oven, M.; Zenk, M. H. J. Anal. At. Spectrom. 2000, 15, 529-534.
The metal-binding ligand is usually either a product of gene expression (e.g., metallothionein) or a product of a reaction of an enzyme coded by a gene conferring the metal resistance.34 Hence, the identification of the gene and the enzyme and linking them to the metal species not only aids the identification of the latter but also describes the mechanism of the metal resistance. The most appropriate technique for the gene search is cloning using yeast complementation assays;35 in particular, screens on a toxic metal medium have been used to isolate metal-resistance genes.36 In contrast to other cloning approaches, it does not rely on nucleic acid or amino acid sequence homology, and cDNA isolated in a complementation experiment is directly demonstrated to be functional.35,37 The objective of this research was the development of an analytical approach allowing the detection and identification of the functional ligand playing a role in the process of the Ni hyperaccumulation in T. caerulescens by means of chromatography and capillary electrophoresis with parallel element-specific (ICPMS) and molecule-specific (ES MS/MS) detection. An integral part of the approach was the use of the yeast complementation screen on a toxic Ni2+ medium as an aid for ligand identification and the demonstration of its functionality. EXPERIMENTAL SECTION Procedures. Preparation of Samples. Plants (T. caerulescens) were grown in hydropony and exposed to 100 µmol L-1 Ni2+ (added as NiSO4) during one week. The roots were separated from the rest of the plant and washed twice with water. A yeast strain (Saccharomyces cerevisiae) containing the nicotianamine synthase gene was exposed to 150 µmol L-1 Ni2+ (added as NiSO4) during 24 h. Each of both samples was frozen in liquid nitrogen, ground with pestle and mortar, extracted with 30 mmol L-1 TrisHCl (pH 7.5), and centrifuged (2900 rpm, 30 min). The supernatants were freeze-dried. The dried material obtained from ∼1 g of fresh roots and yeast, respectively, was dissolved in 500 µL of 30 mmol L-1 Tris-HCl (pH 7.5) for analysis by size exclusion chromatography-ICPMS. Chromatography Conditions. A 100-µL sample of either T. caerulescens extract, yeast extract, or Ni-nicotianamine standard solution was eluted with 5 mmol L-1 ammonium acetate buffer (pH 7.0) at a flow rate of 0.75 mL min-1. Nickel was monitored on-line by ICPMS in order to detect the presence of stable Ni complexes in the sample. For the subsequent ES MS/MS experiments, fractions of the eluate were collected at 30-s intervals. A 5-µL aliquot of each fraction was diluted with 2% HNO3 and analyzed for nickel. The Ni concentration values were plotted as a function of the fraction number in order to obtain a chromatogram. Fractions producing a peak were pooled and freeze-dried. The column was conditioned between the injection with a 10 mmol L-1 EDTA solution for 30 min and then for another 30 min with the mobile phase in order to remove the EDTA. Capillary Electrophoresis Conditions. Nickel complexes were separated using a 12 mmol L-1 Tris-HCl buffer (pH 7.5). The voltage applied was 20 kV that corresponded to a current of (34) Cobbett, C.; Goldsbrough, P. Annu. Rev. Plant Biol. 2002, 53, 159-182. (35) Smith, A. G.; Murray, J. A. H. Methods Mol. Cell. Biol. 1996, 5, 278-288. (36) Lasat, M. M.; Pence, N. S.; Letham, D. L. D.; Kochian, L. V. Int. J. Phytorem. 2001, 3, 129-144. (37) Chen, L.; Powers, S. Methods Enzymol. 1995, 255, 465-468.
Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
2741
20 µA. The sample was injected hydrodynamically by a pressure of 3.4 kPa during 3 s. The capillary was conditioned by rinsing with 0.5 mol L-1 NaOH solution for 5 min. Before the new run, the capillary was regenerated by rinsing with 0.1 mol L-1 NaOH solution for 10 min and with the buffer solution for 3 min. ICPMS Conditions. ICPMS measurement conditions (nebulizer gas flow, rf power, and lens voltage) were optimized daily for highest intensity using a standard built-in software procedure. The isotopes monitored were 58Ni, 60Ni, 61Ni, 62Ni, and 64Ni. Electrospray MS/MS Conditions. The pooled and freezedried fractions within a peak in a SEC-ICPMS chromatogram were dissolved in 100 µL of water. ES MS spectra were acquired (10 scans) in the range 150-1000 u using a 5-ms dwell time and a 1 u step size. Zooms around nickel characteristic isotopic pattern were acquired with a 10-ms dwell time and a 0.1 u step size. The orifice potential was 40 V, the electrospray voltage was 5000 V, and the ion detector potential was 2800 V. The positively charged ions were fragmented by collision-induced dissociation and analyzed in the product ion scan mode. The collision gas was nitrogen, and the collision energy was optimized each time to obtain optimum fragmentation. Cloning of the Ni-Resistance Gene by Yeast Complementation Screens. A cDNA library was constructed from the T. caerulescens leaves and expressed in the yeast S. cerevisiae according to the procedure described in detail elsewhere.36,38 The obtained transformants were screened (in the yeast BY4741-his strain) for Ni tolerance on a minimal medium supplemented with a growth-inhibitory dose of 350 µmol L-1 NiSO4. The surviving transformants isolated from the screen were recovered and submitted, on one hand, to the extraction of plasmids and cDNA sequencing and, on the other hand, to SEC-ICPMS and SECES MS/MS analyses as described above. Apparatus. Chromatographic separations were performed using a model 1100 HPLC pump (Agilent, Wilmington, DE) as the delivery system while the eluent was degassed ultrasonically. The size exclusion column was Superdex Peptide HR 10/30 (300 × 10 mm i.d.) (Pharmacia Biotech, Uppsala, Sweden) with an optimum separation range between 100 and 7000 Da. Injections were made using a model 7725 injection valve with a 0.1-cm3 injection loop (Rheodyne, Cotati, CA). All the connections were made of PEEK tubing (0.17-mm i.d.). Extracts and fractions were lyophilized using a model LP3 lyophilizer (Jouan, France). The ICPMS instrument was an ELAN 6000 (Perkin-Elmer Sciex, Thornhill, ON, Canada). The column eluate was introduced into the ICP via a cross-flow nebulizer fitted in a Ryton spray chamber. The ES MS/MS experiments were performed using a Perkin-Elmer Sciex API 300 electrospray triple quadrupole mass spectrometer. The samples were introduced into the ionization source by means of a Harvard Apparatus model 22 syringe pump. Capillary electrophoresis experiments were performed with a Beckman P/ACE 2200 (Beckman Instruments Inc., Fullerto, CA) using uncoated fused-silica capillaries (Beckman) with a length of 100 cm and an inner diameter of 75 µm. For CZE-ICPMS experiments, the ICPMS instrument used was an HP Agilent 7500 (Yokogawa Analytical Systems Inc., Tokyo, Japan). Both instruments were coupled using the CEI-100 interface (CETAC, Omaha, (38) Czernic, P.; Mari, S.; Pianelli, K.; Vacchina, V.; Marque`s, L.; Łobin´ski, R.; Lebrun, M. XIII International conference on Arabidopsis Research, Sevilla, 28.06-02.07, 2002; Abstract 8-40.
2742
Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
NE) described in detail elsewhere.39 In brief, the makeup liquid, grounded by a Pt electrode, was mixed with the CZE buffer at the end of the CZE capillary. It was transported to the nebulizer by self-aspiration at 6 µL min-1. Reagents, Solutions, and Materials. Analytical reagent grade chemicals purchased from Sigma-Aldrich (Saint-Quentin Fallavier, France) and water (18 MΩ cm) obtained with a Milli-Q system (Millipore, Bedford, MA) were used in this study throughout unless stated otherwise. The T. caerulescens plants were cultivated from seeds collected at the Saint-Laurent Le Minier site (ancient mine) close to Montpellier (France). Total RNA was extracted from leaves using TRIzol R reagent (GIBCOL BRL, Life Technologies, Grand Island, NY). Messenger RNA was purified using the PolyAtract system (Promega, Madison, WI). Complementary DNA synthesis was performed using the cDNA Synthesis Kit (Stratagene, La Rolla, CA) and XhoI/EcoRI cloned in pYX212 (R&D Systems, Minneapolis, MN). Screening were realized in the yeast S. cerevisiae BY4741-his strain (histidine autotrophe derivative of BY4741, ATCC). Pure nicotianamine (2(S),3′S),3′′S)-N-[N-(3-amino-3-carboxypropyl)-3-amino-3-carboxypropyl]azetidine-2-carboxylic acid) was a generous gift of Prof. Satoshi Mori, Laboratory of Plant Molecular Physiology, University of Tokyo. The Ni-nicotianamine (Ni-NA) complex solution was prepared by mixing a 100-µL volume of 15 µmol L-1 nicotianamine solution in water with an equimolar amount of neutral (pH 7.0) solution of Ni2+. RESULTS AND DISCUSSION Detection and Fractionation of Ni Complexes in T. caerulescens by Size Exclusion LC-ICPMS. Size exclusion chromatography allows the separation of metallobiomolecules with dilute buffers close to the physiological pH whereas ICPMS is a convenient element-specific detection technique.24 Recently, some examples of its use for nickel speciation analysis have been reported.40,41 Because the presence of relatively low molecular weight metabolites was expected, a packing (Superdex Peptide) with a narrow separation range within low Mr values (below 10 000) was chosen. Ammonium acetate buffer with a pH value (7.0) close to the intracellular pH was selected in order to avoid the dissociation of the metal chelates. Another important reason for the choice of ammonium acetate was its being well tolerated by the electrospray ionization source in the subsequent analytical step. Ammonium acetate concentrations between 5 and 50 mmol L-1 did not affect the morphology of the chromatogram, so the smallest concentration (5 mmol L-1) was chosen. The elution in the absence of buffer (water as the mobile phase) negatively affected peak height and shape. Other chromatographic separation mechanisms (cation exchange, reversed phase, and anionexchange) were tested but eluted the Ni either in the void or the shape of the peaks was poor. Figure 1a shows SEC-ICPMS chromatograms of an extract of T. caerulescens sample exposed to Ni in hydropony, of a control sample, and of a reagent blank. The morphology of unexposed (39) Schaumlo ¨ffel, D.; Prange, A. Fresenius’ J. Anal.Chem. 1999, 364, 452-456. (40) Mihucz, V. G.; Tatar, E.; Varga, A.; Zaray, G.; Cseh, E. Spectrochim. Acta 2001, 56B, 2235-2246. (41) Koplik, R.; Borkova, M.; Mestek, O.; Kominkova, J.; Suchanek, M. J. Chromatogr., B 2002, 775, 179-187.
Figure 1. SEC-ICPMS chromatograms of the T. caerulescens sample. (a) Thick line, exposed root extract; medium line, control sample extract; thin line, blank (Tris solution). (b) Preparative separation of the exposed root extract (the dots correspond to the individual fractions analyzed).
and exposed plant chromatograms is similar: a major peak with a front shoulder followed by a much less intense second peak. The peak intensity is distinctly higher in the exposed plant, which indicates the bioaccumulation of Ni from the hydroponical solution. No new ligand seems to be synthesized in the plant as result of increasing Ni supply. The metal is accumulated by complexation with an already existing ligand of which the concentration, however, increases as a result of exposure to nickel stress. The shoulder peak and the second peak are present in the chromatogram of the experimental blank (Tris solution) that suggest there being due to the Ni2+ and Ni-Tris complexes, respectively. Indeed, the intensity of these peaks was found to increase when Ni2+ was added to the experimental blank solution. Size exclusion chromatography was also run with an off-line detection by ICPMS that allowed collection of the Ni-containing fraction for the purpose of further experiments. The 30-s fraction collection interval resulted in 325-µL fractions. A 5-µL aliquot of each fraction was sufficient (upon dilution) for the determination of the Ni concentration by ICPMS that allowed the plotting of a chromatogram shown in Figure 1b. Despite the 10-fold more concentrated extract injected, the resolution and the morphology of the chromatogram was still acceptable. Verification of the SEC Peak Purity by Capillary Electrophoresis-ICPMS. To check the chromatographic purity of the major Ni-containing fraction, capillary electrophoresis-ICPMS was applied as a high-resolution technique with an independent separation mechanism. The CZE-ICPMS electropherogram (Figure 2) shows two peaks. The broad peak at a migration time of 400 s corresponds, under these experimental conditions, to the elution of Ni2+, which was weakly bound by carboxylic or amino acids in the original plant extract and dissociated off prior to capillary electrophoresis. The sharp peak at 650 s corresponds to an unknown nickel complex, highly likely to be the only stable Ni complex present in the T. caerulescens extract. The complex was further characterized by electrospray MS/MS. Characterization of the Ni Species by Electrospray MS/ MS. An electrospray mass spectrum (Figure 3a) corresponding
Figure 2. CZE-ICPMS electropherogram of the major Ni-containing fraction in Figure 1b.
to the major peak in the SEC-ICPMS chromatogram (Figure 1) was screened for the presence of the Ni characteristic isotopic pattern (shown in the inset). Two putative Ni-containing ions, at m/z 360 and at m/z 458, could be detected as shown in Figure 3b and c, respectively. The ions, when submitted to CID mass spectrometry, gave product ion mass spectra shown in Figure 3d and e. The CID mass spectrum of the m/z 360 ion shows Nicontaining fragments corresponding to losses of CO2 (m/z 316), CO2 and NH3 (m/z 299), and CO2 and HCOOH (m/z 270) characteristic of a multicarboxylic amino acid. The CID mass spectrum of the m/z 458 ion shows a prominent m/z 360 fragment that suggests this ion to be an adduct of the former species and a molecule with an Mr of 98, possibly sulfate (present in the hydroponic culture medium). In the fraction corresponding to the second peak in the SECICPMS chromatogram (cf. Figure 1b), a number of Ni complexes Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
2743
Figure 3. ES MS/MS analysis of the major Ni-containing fraction in Figure 1b. (a) ES MS spectrum; inset, theoretical Ni isotope distribution profile; (b) zoomed part of the ES MS spectrum around m/z 360 containing the Ni pattern; (c) zoomed part of the ES MS spectrum around m/z 458 containing the Ni pattern; (d) CID MS spectra of the m/z 360 and 362 ions; (e) CID MS spectra of the m/z 458 and 460 ions.
could be detected (mass spectra not shown). The CID fragmentation patterns always contained peaks at m/z 122 and 243 that are characteristic of the Tris molecule and its dimer. This confirms the above evoked hypothesis that this peak is due to a Ni complex with Tris. In summary, ES MS/MS data supported by SEC-ICPMS and CZE-ICPMS prove the presence of a single, previously unreported, nickel compound produced by T. caerulescens as a response to Ni stress. An insight into the identity of this species could be obtained in an alternative set of experiments using molecular genetics techniques. Cloning of the Ni-Resistance Gene by Yeast Complementation Screens. The production of a metal-complexing metabolite in plants is usually a reaction of an enzyme, coded by a specific gene and activated by the metal. Hence, the identification of the gene gives information on the identity of the enzyme and its metabolite. The principle of the approach toward the identification of the gene conferring resistance to nickel is based on the cloning by the complementation approach35,36 and followed by the isolation of the Ni-resistant clone by screening on a toxic Ni medium in which it will be the only one to survive. For the purpose of the complementation experiment, a leaf cDNA library was expressed in the yeast S. cerevisiae using a standard molecular genetics approach.35,36 Out of 320 000 clones obtained, 8 have survived the screening experiment using a culture 2744 Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
Figure 4. Analysis of the Ni-resistant yeast strain. (a) Size exclusion LC-ICPMS chromatograms: thick line, exposed Ni-resistant yeast extract; thin line, control sample extract. (b) Zoomed part of the ES MS spectrum around m/z 360 containing the Ni pattern. (c) CID MS spectra of the m/z 360 ion.
medium supplemented with a growth-inhibitory dose of 350 µmol L-1 NiSO4; one of the transformants turned out to be particularly resistant. The plasmid of this clone was extracted, and its cDNA was sequenced. It turned out to contain a 1.6-kb insert homologous to the A. thaliana nicotianamine synthase gene described elsewhere by Suzuki et al.23 in the context of iron metabolism in barley plants. To verify this identity hypothesis and to correlate the presence of nicotiamine with the resistance to nickel, the nickel-resistant yeast strain was further analyzed by the techniques described above for T. caerulescens. Analysis of the Ni-Resistant Yeast Strain by SEC-ICPMS and ES MS/MS. Size exclusion LC-ICPMS chromatograms of the Ni-resistant yeast strain and that of yeast containing an empty vector as the control samples are shown in Figure 4a. The elution time (18.2 min) of the Ni peak (absent in the control sample) matches that of the compound found in the T. caerulescens extract (Figure 1a). The identity of the complex is further confirmed by mass spectrometry. Indeed, both the mass spectrum zoomed around the m/z 360 (Figure 4b) and the CID product ion mass spectrum (Figure 4c) are identical to those obtained for the plant extract. This proves that the nicotianamine synthase gene is responsible for triggering the synthesis of the nickel-binding ligand, nicotianamine, in T. caerulescens. The DNA of the other seven Ni-resistant transformants was also sequenced showing other genes, which could not, however, be correlated with mass
Figure 5. ES MS/MS analysis of the reconstructed nickel-nicotianamine complex. (a) Zoomed part of the ES MS spectrum around m/z 360 containing the Ni pattern. (b) CID MS spectra of the m/z 360 ion; inset, structure of the ligand nicotianamine.
spectrometric data, notably in the context of the presence of nicotianamine. The ultimate proof of the identification was obtained by the synthesis of a Ni-nicotianamine standard and analyzing it by electrospray MS/MS. The mass spectra (Figure 5) are identical to those obtained earlier for the plant and Ni-resistant yeast strain extract. CONCLUSIONS The combination of hyphenated techniques and molecular biology techniques offers a new way of investigating the processes of metal accumulation and metal tolerance in hyperaccumulating plants. Metal complexes can be detected by HPLC-ICPMS assisted by CZE-ICPMS for the verification of the peak purity and characterized by electrospray MS/MS. Their functional importance, however, can be demonstrated only by the identification of the gene coding for the enzyme responsible for the synthesis of the metal-binding ligand, followed by the reproduction of the biochemical process by isolating and expressing the gene in bacteria or yeast. Cloning using yeast complementation screens on toxic metal medium followed by cDNA analysis of the surviving transformant can provide valuable information allowing an easier
interpretation of HPLC-ICPMS chromatograms and ES MS/MS spectra and is a unique tool to demonstrate the functional importance of the detected metabolites. On the other hand, hyphenated techniques offer a valuable tool for metabolite screening in original plant and genetically modified yeast. A chemical synthesis of the putative ligand followed by the verification of similarity of tandem mass spectra with those from the original plant and the metal-resistant yeast transformant remains the ultimate tool for the validation of the approach. ACKNOWLEDGMENT This research has been supported by a Marie Curie Fellowship for D.S. of the European Community program IHP-MCFI-99-1 under Contract HPMF-CT-2001-01196.
Received for review November 12, 2002. Accepted March 11, 2003. AC020704M
Analytical Chemistry, Vol. 75, No. 11, June 1, 2003
2745
