Planta (1998) 205: 531538

Isolation and characterization of wheat aluminum-regulated genes: possible involvement of aluminum as a pathogenesis response elicitor
Francine Hamel, Christian Breton, Mario Houde
Departement des Sciences biologiques, Universite du Quebec a Montreal, C.P. 8888, Succ. Centre-ville, Montreal, Quebec, H3C 3P8, Canada Received: 24 October 1997 / Accepted: 21 January 1998

Abstract. Using dierential screening of a root tip cDNA library prepared from an Al-tolerant wheat cultivar (Triticum aestivum L. cv. Atlas-66) exposed to Al, we have isolated and characterized several wheat aluminum-regulated (War) cDNAs. Sequence comparison revealed that genes up-regulated by Al correspond to peroxidase (war4.2), cysteine proteinase (war5.2), phenylalanine-ammonia lyase (war7.2), and oxalate oxidase (war13.2). Two wheat cultivars that dier in their level of tolerance (cv. Atlas-66: tolerant, and cv. Fredrick: sensitive) were used to evaluate the relationship between the accumulation of War mRNAs and Al toxicity, as measured by root growth inhibition (RGI). The mRNA accumulation was modulated to similar levels in both cultivars compared at equivalent RGIs. This indicates that War mRNA accumulation is associated with the toxicity of Al rather than with the cultivar's tolerance. It appears that most of the genes found to be up-regulated by Al share homologies with genes induced by pathogens. This suggests that Al may act as an elicitor of a pathogenesis-related transduction pathway. The potential functions of the up-regulated war genes in cell wall strengthening and Al trapping are discussed. Key words: Aluminum toxicity Cysteine proteinase Gene regulation Oxalate oxidase Pathogenesis response elicitor Triticum (A1 toxicity)

Introduction Aluminum (Al) is one of the most abundant elements present in the earth's crust. In acidic soils, its toxicity is

Abbreviations: RGI root growth inhibition; War wheat aluminum-regulated Correspondence to: M. Houde; Tel: 1 (514) 987 3000 ext. 3964; Fax: 1 (514) 987 4647; E-mail: houde.mario@uqam.ca

considered to be the most limiting factor for plant productivity. Based on an estimate of the world's potentially arable land resources, it has been estimated that most of the cultivable area (78.4%) is composed of acid soils. The American continent accounts for 40.9% of the world's acidic soils (von Uexkull and Mutert 1995). The combined eect of acid precipitation and some common agricultural practices exacerbates the overall metal toxicity of these soils. The initial and most dramatic symptom of Al toxicity is inhibition of root growth. The abundance of Al products in the environment and the potential of Al as plant growth inhibitor makes it necessary to understand the mechanisms of Al phytotoxicity and tolerance (Kochian 1995). The toxicity of Al to plants has been attributed to alterations in several physiological and biochemical pathways (Rengel 1992). Aluminum accumulates rapidly in the apoplasm, in the plasma membrane, and eventually enters the cytosol (Lazof et al. 1994; Taylor et al. 1996). However, the plasma membrane has been proposed to be the primary site of Al toxicity (Huang et al. 1995; Sasaki et al. 1995; Wagatsuma et al. 1995). Several reports have shown specic eects on membrane potential, Ca2+ uxes, K+ eux, ion channels, and on the phosphoinositide signal transduction pathway (Kochian 1995; Sasaki et al. 1995; Schroeder 1995; Wagatsuma et al. 1995; Jones and Kochian 1995). However, the precise mechanism of Al toxicity is not fully understood. Several genes have been found to be up-regulated upon exposure to Al. These include phenylalanine ammonia-lyase (wali4, Snowden and Gardner 1993), putative proteinase inhibitors (wali3, wali5, and wali6, Snowden and Gardner 1993; Richards et al. 1994), a plant metallothionein-like protein (wali1, Snowden and Gardner 1993), cysteine-rich proteins of unknown function (wali2 and wali7, Snowden and Gardner 1993; Richards et al. 1994), a peroxidase (al201, Ezaki et al. 1996), a b-1,3-glucanase (glc1 Cruz-Ortega et al. 1997), a glutathione S-transferase (al142, Ezaki et al. 1995), and an auxin-regulated gene homolog (al111, Ezaki et al. 1995). Many of these genes were also found to be

532

F. Hamel et al.: Wheat aluminum-regulated genes

32 P-labeled cDNA inserts (Sambrook et al. 1989). The membranes were washed three times for 30 min with 5 SSC (1 SSC 0.15 M NaCl; 0.015 M Na3-citrate, pH 7) at 55 C, then twice for 30 min with 0.5 SSC at 55 C and autoradiographed with intensifying screens at A80 C. A cDNA showing no dierential hybridization during library screening (control probe named W13.3) was used as a load control to correct for minor variations in RNA loads. Densitometry using a Molecular Dynamics personal densitometer SI and ImageQuaNT 4.2 software (Molecular Dynamics, Sunnyvale, Calif., USA) was used to quantitate the accumulation of each transcript under the dierent Al exposure conditions.

responsive to other toxic metals, low Ca2+ levels, and physical wounding (Snowden et al. 1995). The goal of this study was to identify new genes modulated by Al exposure, to correlate their level of expression as a function of the degree of toxicity, and to understand their possible function. To achieve that, we isolated several Al-responsive genes from a cDNA library prepared from root tips of the tolerant wheat cultivar Atlas-66 exposed to Al. The accumulation of mRNA corresponding to four genes was evaluated in a sensitive and a tolerant wheat cultivar at equivalent root growth inhibition (RGI) levels (Parker 1995). Under these conditions, the genetic responses were found to be similar in both cultivars, indicating that the response is modulated as a function of the degree of Al toxicity. These results suggest that Al-regulated genes isolated in our study do not confer particular advantages to the tolerant plants and could not be implicated in the dierence observed in the cultivars' tolerance. In addition, we have isolated two new genes involved in the Al-toxicity response: oxalate oxidase and cysteine proteinase. The similarity of several Al-regulated genes to the pathogenesis-related (PR) genes suggests that Al may trigger a PR transduction pathway. Materials and methods
Plant material and growth conditions. Two winter wheat cultivars possessing a high (Triticum aestivum L. cv. Atlas-66) and low (T. aestivum L. cv. Fredrick) tolerance to Al were used in this study. Plants were seeded in moist vermiculite and allowed to germinate for 5 d under a controlled environment (20 C; under continuous light: 25 lmol mA2 sA1). For Al exposure, seedlings were carefully washed with deionized water to remove vermiculite. Flat containers were lled with 5 l of solution at pH 4.15 (adjusted with HCl) containing 1 mM CaCl2, and various concentrations of Al ranging from 0 to 500 lM, as described for each experiment. The AlCl3 stock solution (0.1 M) was prepared in water maintained under pH 3.0 with HCl, while AlCl3 powder was slowly added. Seedlings (35 per container) were exposed (24 h) to Al via their roots under the same temperature and light conditions as used for germination. The root growth inhibition (RGI) was measured in all experiments and is expressed as 100 [1 A (root growth of Altreated seedlings divided by the root growth of control seedlings)]. The pH was checked at the end of each experiment to ensure that it did not vary by more than 0.1 pH unit. Construction and screening of the cDNA library. Polyadenylated RNA was isolated from root tips (5 mm) of Atlas-66 seedlings exposed to 50 lM Al. A cDNA library was constructed in lambda ZAPII (Stratagene, La Jolla, Calif. USA) using EcoRI-NotI linkers from Pharmacia Biotech Inc. (Baie D'Urfe, Quebec, Canada), and transfected into Escherichia coli strain XL-1 Blue. The cDNA library was screened with 32P-labeled cDNA probes prepared from poly(A)+ RNA isolated from the root tips of control and Alexposed wheat plants. The screening of the cDNA library and all the recombinant DNA techniques were performed as described by Sambrook et al. (1989). Northern blot analyses. Ribonucleic acid was isolated from root tips, roots, crown, and leaves of wheat seedlings. The RNA samples (5 lg) were separated on formaldehyde agarose gels with samples containing ethidium bromide to control the loads. The RNA was then transferred to nylon membranes MAGNA MSI (Fisher Scientic, Nepean, Ont., Canada) and hybridized with random

Analysis of DNA sequences. Sequencing (T7 DNA sequencing kit; Pharmacia Biotech Inc.) was performed on both DNA strands using a series of deletions. Sequence analysis and comparisons were carried out with the Genetics Computer Group sequence analysis software Wisconsin package, version 9.0-Unix (Michigan State U., Wis., USA).

Results Eect of Al on root growth. In order to reduce the number of Al species in the exposure medium, 1 mM CaCl2 was used and the pH adjusted to 4.15. Under these conditions, most Al remains in the Al3+ form (minimum of 92%), with small amounts present as Al(OH)2+ (up to 7%) over the range of Al concentrations used (as determined by MINEQL+; Environmental Research Software, Mass., USA). Exposure to dierent Al concentrations showed that the cultivar Fredrick is much more sensitive than the cultivar Atlas66 since a maximal RGI was reached at approximately 50 lM Al for Fredrick compared to 500 lM for Atlas66 (Fig. 1). At the highest concentration used (500 lM), the RGI did not exceed 75%. This indicates that roots can continue to grow during the rst 24 h of exposure. However, there was no further root growth within the next 24 h of exposure (data not shown), indicating that maximal inhibition was achieved. Under our assay

Fig. 1. Eect of Al on root growth in the winter wheat cultivars Atlas-66 and Fredrick. Seedlings of 5-d-old Atlas-66 (s) and Fredrick (d) were exposed to varying concentrations of Al for 24 h. RGI was calculated as 100 [1 A (root growth of Al-treated seedlings divided by the root growth of control seedlings)]. Values are the means SE of 30 root measurements from 3 to 10 independent experiments

F. Hamel et al.: Wheat aluminum-regulated genes

533

conditions, half of the maximal inhibition observed (37.5% RGI) would be reached at Al concentrations of 5 and 45 lM for Fredrick and Atlas-66, respectively. Isolation and characterization of Al-responsive genes. Screening of the Atlas-66 cDNA library allowed us to isolate four wheat aluminum-regulated (War) cDNAs that do not cross-hybridize (War4.2, War5.2, War7.2, and War13.2). The eect of four Al concentrations (0, 5, 50, and 500 lM) on the accumulation of War mRNAs in the tolerant cultivar Atlas-66 and the sensitive cultivar Fredrick is shown in Fig. 2. The War4.2, War5.2, War7.2, and War13.2 mRNAs accumulated in response to Al in both Atlas-66 (Fig. 2A) and Fredrick (Fig. 2B). In the tolerant cultivar Atlas-66, the expression of these genes increased at an intermediate Al concentration (50 lM) and their mRNA accumulated to a high level at 500 lM of Al. These mRNAs accumulated at a lower concentration of Al in the sensitive cultivar Fredrick compared to the tolerant cultivar Atlas-66. However, the maximal level of accumulation of these mRNAs appears to be comparable between the sensitive and the tolerant cultivars (Fig. 2). If we compare the level of mRNA accumulation between Atlas-66 exposed to 50 lM Al (Fig. 2A) and Fredrick exposed to 5 lM Al (Fig. 2B), we nd that the transcript level for the dierent genes is very similar. At these Al concentrations, both cultivars show a similar RGI (see Fig. 1). These results suggest
Fig. 3AD. Eect of comparable Al toxicity levels on War mRNA accumulation in the winter wheat cultivars Atlas-66 and Fredrick. Comparable degrees of Al toxicity measured as RGI were obtained by treating Atlas-66 (h) and Fredrick (j) with 0500 lM Al and 0 50 lM Al, respectively. The accumulation of War4.2 (A), War5.2 (B), War7.2 (C), and War13.2 (D) was quantitated by densitometry of northern hybridizations, and normalized using the W13.3 control probe. Results are shown for the control plants (0% RGI), plants exposed to an intermediate degree of toxicity (3560% RGI), and to a high degree of toxicity (>60% RGI). The relative level of mRNA accumulation for each RGI was calculated as a percentage of the maximal level of accumulation for Atlas-66, which was set to 100% in all experiments. Relative expression values are the means + SE of 3 to 12 independent measurements. The asterisk indicates a Student's t test giving a signicant dierence between cultivars at P < 0.05; otherwise, there was no signicant dierence between cultivars

Fig. 2A,B. Eect of Al on the accumulation of War mRNAs in the winter wheat cultivars Atlas-66 and Fredrick. Northern blot hybridizations of 5 lg of total RNA extracted from the root tips of wheat seedlings Atlas-66 (A) and Fredrick (B) exposed for 24 h to 0500 lM Al. Each membrane was successively hybridized with the specied 32P-labeled cDNA, dehybridized, and rehybridized with the W13.3 control probe (shown under each War cDNA hybridization)

that the transcripts of the up-regulated genes accumulate in response to Al toxicity, and is not implicated in the dierential tolerance observed between cultivars. Quantitative analyses of gene expression. The accumulation of the dierent mRNAs was characterized in detail in the sensitive and tolerant cultivars exposed to similar degrees of toxicity. We used dierent ranges of Al concentrations in the sensitive (050 lM) and the tolerant cultivar (0500 lM) in order to achieve comparable RGIs. Aluminum toxicity was thus expressed as RGI and divided into three groups: 0%, 3560%, and >60% RGI (Fig. 3). In most cases, there was no signicant dierence in the accumulation of War4.2, War5.2, War7.2, and War13.2 between the tolerant and the sensitive cultivars compared at the same RGI

534

F. Hamel et al.: Wheat aluminum-regulated genes Table 1. Characteristics of the War cDNA clones. Sizes (in bases) of mRNAs were estimated from northern hybridizations. Identities were obtained from the FASTA search program of the GenBank release 101.0 Clone War4.2 War5.2 War7.2 War13.2
a

mRNA (bases) 1450 1700 2500 1000

Gene Peroxidase Cysteine proteinase Phenylalanine ammonia-lyase Oxalate oxidase

Identitya (%) 60.6 78.5 89.8 79.0

Fig. 4. Relative accumulation of War13.2 mRNA in dierent tissues from wheat Atlas-66 seedlings. Total RNA was isolated from root tips, roots, crown, and leaves of 5-d-old Atlas-66 seedlings exposed for 24 h to 0500 lM Al, to give RGIs of 0% (h), 3560% ( ) or >60% (j). The accumulation was quantitated by densitometry of northern hybridizations, and normalized to those obtained from the W13.3 control probe. The relative level of mRNA accumulation for each RGI was calculated as a percentage of the maximal level of accumulation observed in root tissue, which was set to 100% in all experiments. Relative expression values are the means + SE of at least three independent measurements

Nucleotide identity was compared between War4.2 (AF005087), War5.2 (AF005088), War7.2 (AF005089), War13.2 (AF005084) and a peroxidase (X58396), a cysteine proteinase (D45402), a phenylalanine ammonia-lyase (X16099), and an oxalate oxidase (X93171) cDNA

War4.2 and War7.2 show 60.6% identity to a peroxidase and 89.8% identity to a phenylalanine ammonia-lyase cDNA respectively (Table 1). These genes have already been shown to be up-regulated by Al (Ezaki et al. 1996; Snowden and Gardner 1993) and were not characterized in detail.

(Fig. 3). On the other hand, mRNA accumulation was signicantly increased in the two wheat cultivars exposed to Al concentrations resulting in a moderate (3560% RGI; P < 0.005) or high (>60% RGI; P < 0.005) degree of toxicity compared to the control (0% RGI). The mRNA accumulation of up-regulated genes is thus mostly regulated by the degree of Al toxicity (as measured by RGI) and is independent of the cultivars' tolerance. However, the absolute Al concentration needed to achieve similar RGIs, is cultivar dependent. Tissue-specicity of up-regulated war genes. To examine the tissue-specicity of up-regulated genes, we isolated mRNAs from dierent parts of the plants. At dierent RGIs, the level of mRNA accumulation of War13.2 increased in roots as well as in root tips (Fig. 4), suggesting that the function of this gene is closely associated with the eect of Al on root physiology. The relative mRNA accumulation of the other up-regulated genes (war4.2, war5.2, and war7.2) was found to be higher in roots than in root tips, but was unaected by varying RGIs in roots (data not shown). All four genes were expressed at low levels in crown and leaf tissues and increased slightly (but non signicantly) in leaf at high RGI. This result may suggest that a low amount of Al is translocated between roots and other tissues during the 24 h of exposure or that these genes are not responsive to Al in other tissues. Analysis of DNA sequences. We determined the complete sequences of War4.2, War5.2, War7.2, and War13.2 cDNAs. Details of mRNA size, and identity with other cDNAs are summarized in Table 1; DNA sequence comparisons have shown that their identity with known genes is greater than 60% in all cases.

Fig. 5. Comparison of cysteine proteinase protein sequences. Alignment of the deduced amino acid sequences of the wheat (Triticum aestivum) and corn (Zea mays) cysteine proteinase proteins. Identity of 82.3% was obtained using the TFASTA search program of the GenBank release 101.0. Identities, and conservative changes are indicated by vertical lines, and colons, respectively. Dots within the sequences indicate gaps that were introduced to maximize homology between the two sequences. An arrow indicates the putative cleavage site of the signal peptide. An arrowhead indicates a possible cleavage site in post-translational processing. Also indicated are the conserved cysteine and histidine residues of the active sites (h) as well as the conserved cysteines involved in disulphide bridge formation (j). The protein sequences were deduced from the following cDNA sequences: wheat (War5.2, acc. no. AF005088) and corn (Ccp1, acc. no. D45402)

F. Hamel et al.: Wheat aluminum-regulated genes

535

The deduced amino acid sequences of War5.2 and War13.2 show 78.5% identity to a cysteine proteinase and 79.0% identity to an oxalate oxidase cDNA, respectively (Table 1). The results of the search indicate that War5.2, encoding a 41-kDa protein with a calculated pI of 6.14, is homologous to Ccp1, a cysteine proteinase of corn (Zea mays) encoding a 40.3-kDa peptide with an isoelectric point (pI) of 6.32 (Domoto et al. 1995). Figure 5 shows a comparison of the predicted protein sequences of the wheat War5.2 with the corn Ccp1 cDNAs. It was previously noted that the CCP1 protein contains a putative peptide signal of 23 amino acids. A very similar signal was present in the wheat polypeptide. WAR5.2, as well as CCP1, contains a prepro-sequence, probably from 1M to 141G, which was assigned based on sequence similarity to the mature sequences of other known cysteine proteinases. The wheat protein (Fig. 5) has conserved the relative position of the catalytic cysteine (C-166) and histidine (H-308) residues, and the cysteine residues involved in disulphide-bridge formation (C-163/C-209, C-197/C246, C-302/C-359) compared to other mature cysteine proteinases (Linthorst et al. 1993). In the case of War13.2, the deduced amino acid sequence contains 221 residues with a molecular weight of 24.3 kDa and a pI of 6.67. It was found that this cDNA is highly homologous to the Bh6-903 cDNA of barley (Hordeum vulgare) with 79.0% identity at the DNA level and 72.4% identity at the protein level (Table 1, Fig. 6). Bh6-903 encodes an oxalate oxidaselike protein of 229 residues with a molecular weight of 24.7 kDa and a pI of 6.5 (Thordal-Christensen et al. 1997). Figure 6 shows a comparison of the predicted protein sequences of the wheat War13.2 with the barley Bh6-903 cDNAs. The proteins contain the rare HI/ THPRATEI sequence that is found in oxalate oxidases and in the spherulins 1a/b. It has been postulated that a membrane transit domain may be common to these functionally disparate proteins, which are destined for the extracellular matrix (Lane 1994).

Discussion Wheat aluminum-regulated (war) genes show similarities with known PR genes. We have identied four genes (war4.2, war5.2, war7.2, and war13.2) that are upregulated by Al exposure in two wheat cultivars diering in their level of Al tolerance. Expression of RNA is traditionally compared at similar Al concentrations (Snowden and Gardner 1993). Using this paradigm, we observed that tolerant cultivars express these genes at a much lower level than the sensitive cultivar (Fig. 2). This dierential sensitivity to Al toxicity could be explained by the higher capacity of tolerant plants to exclude Al (Samuels et al. 1997). In order to compare the level of gene expression under similar conditions, we used RGI as a measure of Al toxicity, as suggested by Parker (1995). Our results show that the dierent war genes are up-regulated proportionately to Al toxicity (Fig. 3). This suggests that these Al-regulated genes do not confer particular advantages to the tolerant plants and could not be implicated in the dierential Al-tolerance mechanisms of wheat cultivars. The dierent war genes upregulated upon Al exposure are homologous to a peroxidase (war4.2), a cysteine proteinase (war5.2), a phenylalanine ammonia-lyase (war7.2), and an oxalate oxidase (war13.2) gene. Peroxidases (such as war4.2) were previously found to be up-regulated upon Al exposure (al201, Ezaki et al. 1996). An increase in the activity of both soluble and cell wall peroxidases was observed by these authors either using phosphate starvation or a combination of Al exposure and Pi starvation. Peroxidase activities occur mostly at the cell wall, where these enzymes have been suggested to modulate cell wall rigidity and extensibility by insolubilizing extensin monomers via a progressive increase in inter- and intramolecular cross-linking involving isodityrosine bridges (Welinger 1992), thus reducing the rate of Al diusion through the cell wall. Reinforcement of the cell wall has also been postulated to be important to slow down pathogen invasion. Furthermore, since extensins are negatively charged, they could act as a ``y paper'' through electrostatic interaction with Al3+ (Showalter 1993). Similarly, pectins which contain a large number of carboxyl groups could also bind or chelate Al3+ ions. Analysis of cell wall components in squash root seedlings inhibited by Al exposure has shown an increase in pectin, hemicellulose, and cellulose content after a few hours of exposure (Le Van et al. 1994). Thickening of the cell wall may lead to Al detoxication but, unfortunately, it may also result in the arrest of root growth and cellular elongation (Le Van et al. 1994). We have isolated a cysteine proteinase (war5.2) gene which, for the rst time, has been shown to be inducible by Al. This enzyme is involved in diverse metabolic events such as protein degradation or post-translational processing of protein precursors to a mature form (Hara-Nishimura et al. 1993). It may participate in the defense against pathogen infection since it was recently shown that benzothiazole, a novel class of inducer of systemic acquired resistance, activates the induction of a

Fig. 6. Comparison of oxalate oxidase protein sequences. Alignment of the deduced amino acid sequences of the wheat (Triticum aestivum) and barley (Hordeum vulgare) oxalate oxidase-like proteins. Identity of 72.4% was obtained using the TFASTA search program of the GenBank database release 101.0. The box indicates a rare consensus sequence. For other symbols, see Fig. 5. The protein sequences were deduced from the following cDNA sequences: wheat (War13.2, acc. no. AF005084) and barley (Bh6-903, acc. no. X93171)

536

F. Hamel et al.: Wheat aluminum-regulated genes

cysteine proteinase gene as well as disease resistance in wheat (Gorlach et al. 1996). Accumulation of phenylalanine ammonia-lyase (PAL; War7.2) homologs in the presence of Al was previously reported in wheat (Wali4, Snowden and Gardner 1993). This enzyme could play a benecial role in detoxifying Al that has entered the symplasm since PAL was shown to catalyze the rst step of the multibranched phenylpropanoid metabolism in higher plants. The synthesis of anthocyans and avonoids, which were also referred to as catechins, is performed via PAL. Catechins are very abundant (up to 1030% dry weight) in tea, which is known to accumulate Al in amounts reaching 30 000 ppm (Nagata et al. 1992). The phenylpropanoid pathway and PAL are also involved in the biosynthesis of numerous compounds (Bowles 1990). In particular, they participate in the synthesis of precursors such as ferulate groups (derived from p-coumaric acid) which are used in pectin to increase the gelling of polysaccharides and thus in strengthening the cell wall (Bowles 1990). Phenylalanine ammonialyase is a well known defense protein that has been shown to accumulate in several dierent incompatible plant-pathogen combinations and in response to elicitors (Ebel and Cosio 1994). This is the rst report indicating that an mRNA homologous to oxalate oxidase (War13.2) can accumulate upon Al exposure. The enzyme was shown to accumulate upon fungal infection of barley (Zhang et al. 1995). Oxalate oxidase was also implicated in the resistance of transgenic rapeseed to an oxalate-secreting fungus (Lane 1994). The enzyme is involved in the degradation of oxalate, accumulated in plant cells as the calcium salt, to produce Ca2+, CO2, and H2O2. The four genes identied (war4.2, war5.2, war7.2, and war13.2) are up-regulated proportionately to the Al toxicity level (as measured by the RGI) in both cultivars, suggesting that a common signal is involved in their regulation. An intriguing aspect of these genes is their homologies to genes induced in the plant defense responses against wounding or pathogen invasion. In both stresses, these genes may provide protection by increasing cell wall thickness. To this end, oxalate oxidase would provide the H2O2 needed by the peroxidase to perform cross-linking reactions; PAL would provide precursor molecules and cysteine proteinase would be involved in the processing of new enzyme molecules, such as peroxidase and oxalate oxidase. Atlas-66 wheat plants become much more sensitive to Al when the mucilage is removed from their root cap (Puthota et al. 1991). Furthermore several Arabidopsis mutants with increased sensitivity to Al have been isolated (Larsen et al. 1996), indicating that many genes are involved in the protection against Al toxicity. The four war genes isolated in our study may participate in the strengthening of the root cell wall and in the exclusion of Al in both sensitive and tolerant plants. This inducible detoxication mechanism may involve Al trapping through an increase in the cell wall negative charges and through the increased cross-linking. However, this hypothesis needs to be conrmed by direct

enzyme activity measurements in relationship with cytoplasmic Al-entry rates. Relationships between Al and pathogen stresses. The plant defense mechanisms that have been associated with the response to pathogens include: hypersensitive cell death, lignication, callose deposition, the synthesis of cell wall extensins, the induction of pathogenesis-related proteins such as b-endoglucanases and chitinases, the accumulation of phytoalexins, proteins that induce cell lysis in sporangia, proteinase inhibitors, and other toxic substances referred to as thionins which are toxic cysteinerich peptides produced during pathogen invasion (Hammond-Kosack and Jones 1996; Ebel and Cosio 1994). Many of these responses are also associated with symptoms of Al toxicity. The production of proteinase inhibitors has been described by Richards et al. (1994) and several cDNAs encoding cysteine-rich peptides (similar to thionins) have been isolated (Richards et al. 1994; Ezaki et al. 1995). The induction of b-1,3glucanases by Al toxicity was recently shown in wheat roots (Cruz-Ortega et al. 1997). During pathogen infection, the function of the inducible b-1,3-glucanase would be to degrade fungal pathogen cell walls (Bowles 1990). However, since pathogens do not appear to be involved in Al toxicity (unless Al increases sensitivity to pathogens), the function of this enzyme may be to allow the turnover of callose (b-1,3-glucan polymer) to proceed. On the other hand, this enzyme may not be useful and could be induced serendipitously through the elicitation of a defense-related signalling pathway. One of the most interesting similarities between the response to Al exposure and the infection by pathogens is the production of active oxygen species, that participate in the growth and development of plant cells. These molecules in concert with antioxidant metabolism have been recently suggested as important signalling molecules in response to pathogens or other stresses (Mehdy et al. 1996). Ezaki et al. (1995) have isolated a gene induced by Al that is homologous to glutathione S-transferase. The induction of this enzyme could limit lipid peroxidation which was previously observed in soybean root tips exposed to Al (Cakmak and Horst 1991). These last authors have also shown an induction of superoxide dismutase, indicating an increased metabolism of oxygen free radicals. Such free radicals have been associated with signalling pathways involving pathogens (Mehdy et al. 1996; Hammond-Kozack and Jones 1996); they may also be involved in the response to Al toxicity since several oxidative stress genes are induced by this metal (Richards et al. 1998). This may explain why many of the genes induced by Al are similar to those induced by pathogens or elicitors. Recent studies support the possibility that various components of signalling pathways are aected by Al toxicity, since it was shown that Al could aect anion channels (Schroeder 1995) or phosphoinositide transduction pathways (Jones and Kochian 1995). Aluminum was also shown to block Ca2+ channels in the wheat root plasma membrane (Huang et al. 1996) and, thus, regulatory pathways involving Ca2+ could aect the

F. Hamel et al.: Wheat aluminum-regulated genes

537 Huang JW, Pellet DM, Papernik LA, Kochian LV (1996) Aluminum interactions with voltage-dependent calcium transport in plasma membrane vesicles isolated from roots of aluminumsensitive and -resistant wheat cultivars. Plant Physiol 110: 561569 Jones DL, Kochian LV (1995) Aluminum inhibition of the inositol 1,4,5-triphosphate signal transduction pathway in wheat roots: a role in Al toxicity? Plant Cell 7: 19131922 Kochian LV (1995) Cellular mechanisms of aluminum toxicity and resistance in plants. Annu Rev Plant Physiol Plant Mol Biol 46: 237260 Lane BG (1994) Oxalate, germin, and the extracellular matrix of higher plants. FASEB J 8: 294301 Larsen PB, Tai CY, Kochian LV, Howell SH (1996) Arabidopsis mutants with increased sensitivity to aluminium. Plant Physiol 110: 743751 Lazof DB, Goldsmith JG, Rufty TW, Linton RW (1994) Rapid uptake of aluminum into cells of intact soybean root tips. Plant Physiol 106: 11071114 Le Van H, Kuraishi S, Sakurai N (1994) Aluminum-induced rapid root inhibition and changes in cell-wall components of squash seedlings. Plant Physiol 106: 971976 Linthorst SJM, van der Does C, Brederode FT, Bol JF (1993) Circadian expression and induction by wounding of tobacco genes for cysteine proteinase. Plant Mol Biol 21: 685694 Mehdy MC, Sharma YK, Sathasivan K, Bays NW (1996) The role of activated oxygen species in plant disease resistance. Physiol Plant 98: 365374 Nagata T, Hayatsu M, Kosuge N (1992) Identication of aluminium forms in tea leaves by 27Al NMR. Phytochemistry 31: 12151218 Parker DR (1995) Root growth analysis: an underutilised approach to understanding aluminium rhizotoxicity. Plant Soil 171: 151 157 Puthota V, Cruz-Ortega R, Johnson J, Ownby J (1991) An ultrastructural study of the inhibition of mucilage secretion in the wheat root cap by aluminium. In: Wright RJ, Baligar VC, Murrmann RP (eds) Plant-soil interactions at low pH. Kluwer Academic Publishers, Dordrecht, pp. 779787 Rengel Z (1992) Role of calcium in aluminium toxicity. New Phytol 121: 499513 Richards KD, Schott EJ, Sharma YK, Davis KR, Gardner RC (1998) Aluminum induces oxidative stress genes in Arabidopsis thaliana. Plant Physiol 116: 409418 Richards KD, Snowden KC, Gardner RC (1994) wali6 and wali7: genes induced by aluminum in wheat (Triticum aestivum L.) roots. Plant Physiol 105: 14551456 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Samuels TD, Kucukakyuz K, Rincon-Zachary M (1997) Al partitioning patterns and root growth as related to Al sensitivity and Al tolerance in wheat. Plant Physiol 113: 527534 Sasaki M, Kasai M, Yamamoto Y, Matsumoto H (1995) Involvement of plasma membrane potential in the tolerance mechanism of plant roots to aluminium toxicity. Plant Soil 171: 119124 Schroeder JI (1995) Anion channels as central mechanisms for signal transduction in guard cells and putative functions in roots for plant-soil interactions. Plant Mol Biol 28: 353361 Showalter AM (1993) Structure and function of plant cell wall proteins. Plant Cell 5: 923 Snowden KC, Gardner RC (1993) Five genes induced by aluminum in wheat (Triticum aestivum L.) roots. Plant Physiol 103: 855 861 Snowden KC, Richards KD, Gardner RC (1995) Aluminuminduced genes: induction by toxic metals, low calcium, and wounding and pattern of expression in root tips. Plant Physiol 107: 341348 Taylor GJ, Hunter DB, Stephens J, Bertsch PM, Elmore D, Rengel Z, Reid RJ (1996) Direct measurement of Al transport across the plasma membrane of Chara corallina. Proceedings of the 4th

expression of specic genes. Furthermore, a low level of extracellular Ca2+ was previously shown to up-regulate Al-responsive genes (Snowden et al. 1995). The induction of these genes by a low level of Ca2+ is interesting since a Ca2+ deciency was previously shown to inhibit root growth and to enhance lipid peroxidation (Cakmak and Horst 1991). An increase in oxalate oxidase (WAR13.2) activity would favor the release of Ca2+ which may compensate, at least partially, for the eect of Al on calcium channels. It is plausible that accumulation of this mRNA caused by Al may be regulated by a signalling pathway involving Ca2+. These dierent results indicate that Al may not only act through interactions with Ca2+ signalling pathways but it may also trigger more than one pathway through the generation of several active molecules. This may thus explain the complexity of responses observed in relation to Al toxicity.
The authors thank Dr B.F. Carver (Oklahoma State U., OK, USA) who provided the Atlas-66 seeds to initiate this project, and Dr N. Chevrier (UQAM) for producing sucient seeds. This work was supported by an NSERC grant (OGP0138557) to M.H. The salary of M.H. was in part provided by the Canadian Network of Toxicology Centers via the ``Centre de Toxicologie de l'Environnement'' of UQAM.

References
Bowles DJ (1990) Defense-related proteins in higher plants. Annu Rev Biochem 59: 873907 Cakmak I, Horst WJ (1991) Eect of aluminium on lipid peroxidation, superoxide dismutase, catalase, and peroxidase activities in root tips of soybean (Glycine max). Physiol Plant 83: 463468 Cruz-Ortega R, Cushman JC, Ownby JD (1997) cDNA clones encoding 1,3-b-glucanase and a mbrin-like cytoskeletal protein are induced by Al toxicity in wheat roots. Plant Physiol 114: 14531460 Domoto C, Watanabe H, Abe M, Abe K, Arai S (1995) Isolation and characterization of two cDNA clones encoding corn seed cysteine proteinases. Biochim Biophys Acta 1263: 241244 Ebel J, Cosio EG (1994) Elicitors of plant defense responses. Int Rev Cytol 148: 136 Ezaki B, Yamamoto Y, Matsumoto H (1995) Cloning and sequencing of the cDNAs induced by aluminium treatment and Pi starvation in cultured tobacco cells. Physiol Plant 93: 1118 Ezaki B, Tsugita S, Matsumoto H (1996) Expression of a moderately anionic peroxidase is induced by aluminium treatment in tobacco cells: possible involvement of peroxidase isozymes in aluminum ion stress. Physiol Plant 96: 2128 Gorlach J, Volrath S, Knauf-Beiter G, Hengy G, Beckhove U, Kogel K-H, Oostendorp M, Staub T, Ward E, Kessmann H, Ryals J (1996) Benzothiadiazole, a novel class of inducers of systemic acquired resistance, activates gene expression and disease resistance in wheat. Plant Cell 8: 629643 Hammond-Kosack KE, Jones JDG (1996) Resistance gene-dependent plant defense responses. Plant Cell 8: 17731791 Hara-Nishimura I, Takeuchi Y, Nishimura M (1993) Molecular characterization of a vacuolar processing enzyme related to a putative cysteine proteinase of Schistosoma mansoni. Plant Cell 5: 16511659 Huang JW, Grunes DL, Kochian LV (1995) Aluminium and calcium transport interactions in intact roots and root plasmalemma vesicles from aluminium-sensitive and tolerant wheat cultivars. Plant Soil 171: 131135

538 International Symposium on Plant-Soil Interactions at Low pH. Belo Horizonte, Brazil, O15 Thordal-Christensen H, Zhang Z, Wei Y, Collinge DB (1997) Subcellular localization of H2O2 in plants: H2O2 accumulation in papillae and hypersensitive response during the barleypowdery mildew interaction. Plant J 11: 11871194 von Uexkull HR, Mutert E (1995) Global extent, development and economic impact of acid soils. Plant Soil 171: 115 Wagatsuma T, Ishikawa S, Obata H, Tawaraya K, Katohda S (1995) Plasma membrane of younger and outer cells is the

F. Hamel et al.: Wheat aluminum-regulated genes primary specic site for aluminium toxicity in roots. Plant Soil 171: 105112 Welinger KG (1992) Plant peroxidases: structure-function relationships. In: Penel C, Gaspar Th, Greppin H (eds) Plant peroxidases 19801990, topics and detailed literature on molecular, biochemical, and physiological aspects. University of Geneva, Switzerland, pp. 124 Zhang Z, Collinge D, Thordal-Christensen H (1995) Germin-like oxalate oxidase, a H2O2-producing enzyme, accumulates in barley attacked by the powdery mildew fungus. Plant J 8: 139145