Aluminum-Resistant Arabidopsis Mutants That Exhibit Altered Patterns of Aluminum Accumulation and Organic Acid Release from Roots

Abstract

Al-resistant (alr) mutants of Arabidopsis thaliana were isolated and characterized to gain a better understanding of the genetic and physiological mechanisms of Al resistance. alr mutants were identified on the basis of enhanced root growth in the presence of levels of Al that strongly inhibited root growth in wild-type seedlings. Genetic analysis of the alr mutants showed that Al resistance was semidominant, and chromosome mapping of the mutants with microsatellite and random amplified polymorphic DNA markers indicated that the mutants mapped to two sites in the Arabidopsis genome: one locus on chromosome 1 (alr-108, alr-128, alr-131, and alr-139) and another on chromosome 4 (alr-104). Al accumulation in roots of mutant seedlings was studied by staining with the fluorescent Al-indicator dye morin and quantified via inductively coupled argon plasma mass spectrometry. It was found that the alr mutants accumulated lower levels of Al in the root tips compared with wild type. The possibility that the mutants released Al-chelating organic acids was examined. The mutants that mapped together on chromosome 1 released greater amounts of citrate or malate (as well as pyruvate) compared with wild type, suggesting that Al exclusion from roots of these alr mutants results from enhanced organic acid exudation. Roots of alr-104, on the other hand, did not exhibit increased release of malate or citrate, but did alkalinize the rhizosphere to a greater extent than wild-type roots. A detailed examination of Al resistance in this mutant is described in an accompanying paper (J. Degenhardt, P.B. Larsen, S.H. Howell, L.V. Kochian [1998] Plant Physiol 117: 19–27).

Al toxicity is a global problem that limits crop productivity on acidic soils. Al is the most abundant metal in the earth's crust, and in acidic soils (pH < 5.5) the phytotoxic species Al³⁺ is solubilized to levels that inhibit root growth and crop yield (Kochian, 1995). Large areas of the world contain acidic soils (>30% of the arable land), so Al toxicity is a very important worldwide agricultural problem (Von Uexkull and Mutert, 1995). Despite the agronomic importance of this problem, little is known about fundamental mechanisms of Al toxicity and resistance. It has been well documented that many plant species exhibit significant genetic variability in their ability to resist Al toxicity (Delhaize and Ryan, 1995; Kochian, 1995, and refs. therein). Although it is clear that certain plant genotypes have evolved mechanisms that confer Al resistance, the cellular and molecular basis for Al resistance is still poorly understood.

There are two strategies that plants can use to deal with Al toxicity: exclusion from the root apex or development of the ability to tolerate Al once it enters the plant symplasm (Delhaize and Ryan, 1995; Kochian, 1995). Because of the complex interactions between Al and the plant, it is very likely that there are a number of different mechanisms that plants use to confer Al resistance. This is supported by genetic studies of Al resistance, which have shown it to be a dominant, multigenic trait controlled by one or a few major genes and several minor genes (Lafever and Campbell, 1978; Aniol, 1990; Carver and Ownby, 1995).

Recent experimental evidence supports an Al-resistance mechanism that results in the exclusion of Al from the root apex via the release of Al-binding ligands such as organic acids and/or phosphate. When these ligands are released into the rhizosphere, they can effectively chelate Al³⁺ and prevent its entry into the root. In several species, including snapbean, maize, and wheat, increased Al resistance is correlated with Al-dependent organic acid (citrate or malate) release (Miyasaka et al., 1991; Delhaize et al., 1993b; Basu et al., 1994; Pellet et al., 1995). In near-isogenic lines of wheat, increased Al resistance is dependent on Al-inducible release of malate into the rhizosphere (Delhaize et al., 1993a, 1993b). Further examination has shown that in a large number of wheat genotypes, Al resistance is strongly correlated with the level of Al-induced malate that is released from root tips (Ryan et al., 1995b).

A similar mechanism has also been reported in maize; an Al-resistant cultivar was shown to release citrate in response to Al (Pellet et al., 1995). More recently, it was shown that constitutive phosphate release might operate in conjunction with Al-induced organic release to confer a greater degree of Al resistance in certain wheat genotypes (Pellet et al., 1996). Another mechanism of Al exclusion that is frequently proposed is a root-mediated elevation in the pH of the rhizosphere adjacent to the root apex (Kochian, 1995). Because the solubility of Al³⁺ is pH dependent, increases in rhizosphere pH would reduce the activity of Al³⁺ in the rhizosphere, thus decreasing Al entry into the root. However, no direct evidence for this Al-exclusion mechanism has been found to date.

The primary effect of Al toxicity is a rapid inhibition of root growth, which appears to result from complex interactions between Al and the root apex (Ryan et al., 1993). Because this response is rapid (e.g. significant inhibition within 1 h in wheat), it is likely that cell elongation, not cell division, is initially inhibited. However, in the long term, Al-dependent root-growth inhibition is probably dependent on blockage of both cell division and elongation (Kochian, 1995). The molecular and biochemical basis of Al toxicity is not well understood, but it is likely that a number of processes that are associated with both the root apoplast and the symplast are targeted by Al. In Arabidopsis thaliana, mutations affecting at least eight unique loci confer increased Al sensitivity, indicating that Al sensitivity is a complex trait in plants (Larsen et al., 1996).

We were interested in determining the molecular and physiological basis of Al resistance in higher plants, so we identified and characterized Arabidopsis mutants with altered responses to Al. We previously identified and characterized a number of Al-sensitive Arabidopsis mutants, and through analysis of these mutants, are attempting to identify genes encoding targets of Al toxicity (Larsen et al., 1996, 1997). In the present study, we have conducted research aimed at identifying loci that are responsible for Al resistance via screening for Arabidopsis mutants with increased resistance to Al. By isolation and analysis of such mutants, we hope to identify and characterize genes responsible for Al resistance and to gain a better understanding of the genetic and physiological mechanisms resulting in resistance.

MATERIALS AND METHODS

Isolation of Mutants

Mutants were generated in the Columbia (Col-0) ecotype of Arabidopsis thaliana by treating seeds with EMS. Al-resistant (alr) M₂ seedlings were identified by quantifying root growth through gelled-nutrient medium equilibrated with levels of AlCl₃ toxic to wild-type seedlings (Larsen et al., 1996).

Screening was carried out in 100- × 25-mm Petri dishes containing 85 mL of nutrient medium (pH 4.2) plus 0.125% gellan gum (Gell-Gro, ICN) (for composition of nutrient medium, see Larsen et al., 1996). Al was introduced into the gel medium by equilibration for 2 d with a soak solution consisting of a modified nutrient solution, pH 4.2, and 1.5 mm AlCl₃ (Larsen et al., 1996).

Mutagenized seeds were sterilized and cold stratified (4°C) for 2 d in the dark to synchronize germination. Two-hundred-fifty M₂ generation seeds suspended in 0.15% agarose were planted around the periphery of each Petri dish. After incubation for 7 d in a growth chamber (20°C with a day/night cycle of 16/8 h), putative Al-resistant mutant seedlings were identified based on enhanced root growth in the Al-containing environment, and then rescued onto plant nutrient medium without Al and with Suc added (Lincoln et al., 1990). After 2 weeks, putative mutants were transferred to soil and grown in a light room at 20°C, 40% RH, and a light intensity of 50 μE m⁻² s⁻¹. Putative alr mutants were selfed, and M₃ progeny were rescreened using the same screening method. Each mutant was backcrossed four times into the Columbia background and the resulting BC₄ generation was used for subsequent genetic and physiological analyses.

Genetic Analysis

Analysis of inheritance was performed by crossing each alr mutant (male parent) to wild-type (Col-0) plants (female parent) bearing the glabrous-1 mutation (used as a crossing marker).

The chromosome location of each alr mutation was mapped using microsatellite markers (Bell and Ecker, 1994). To generate mapping populations, alr mutants (Col-0 ecotype) were crossed with wild type (Wassilewskija, Ws-0 ecotype). Homozygous wild-type (not resistant to Al) F₂ progeny were identified on gelled-nutrient medium equilibrated with 1.5 mm AlCl₃. Progeny were confirmed as being homozygous by rescreening F₃ families using the same method. Map distances were analyzed using the Mapmaker II program (Lander et al., 1987).

To identify more closely linked markers on chromosome 1, RAPD analysis was performed by comparing pools of homozygous resistant and homozygous wild-type F₂ plants generated from a mapping cross between alr-128 and Ws-0 (Reitter et al., 1992).

Root-Growth Measurements in Solution Culture

Arabidopsis seedlings were grown in solution culture supported on a 250-μm mesh polypropylene screen in 100- × 15-mm Petri dishes containing 40 mL of nutrient solution (Larsen et al., 1996). After 4 d of growth, one-half of the seedlings were removed for root measurement. At this time, the solution was replaced with nutrient solution containing varying concentrations of AlCl₃ or LaCl₃. For growth in LaCl₃-containing solution, KH₂PO₄ was omitted to prevent precipitation of La. The seedlings were grown for an additional 2 d, after which time the roots of the remaining seedlings were measured.

Determination of Al Accumulation in Roots

To determine if any of the Al-resistant mutants were excluding Al from the root tip, two different approaches were used. First, roots were stained with morin (2′,3′,4′,5,7-pentahydroxyflavone), a fluorescent histochemical indicator for Al. Roots of seedlings grown in solution culture were exposed for 1 h to 25 μm AlCl₃ in nutrient medium. The roots were washed in Mes buffer, pH 5.5, for 10 min, stained with morin, and visualized using the methods described by Larsen et al. (1996). A second approach involved the direct quantification of Al in root tissue using ICP-MS. Arabidopsis seedlings were grown hydroponically on 2 × 2 cm squares of 250-μm polypropylene mesh supported by 1.5-cm polycarbonate rods (1.25 cm in diameter) (Small Parts, Inc., Miami, FL). One-hundred seeds were sown on each screen and then placed in a 100- × 25-mm Petri dish containing 85 mL of nutrient medium.

After 4 d of growth, the growth solution was replaced with a new medium containing 25 μm AlCl₃, and seedlings were incubated for up to 3 h. Seedlings were then rinsed in nutrient medium for 5 min. For root tissue in which cell wall-bound Al was desorbed, samples were washed in 0.5 mm citrate, pH 4.2, for 30 min at 4°C (Zhang and Taylor, 1990). Root tips were harvested by cutting the terminal portion with a razor blade against the polycarbonate support rod. Collected roots were dried in an oven at 90°C overnight, and dry weights were determined using a microgram balance (model UMT2, Mettler, Greifensee, Switzerland).

Dried tissue samples were ashed in 20 μL of hot, concentrated nitric acid. Samples were resuspended in 10 mL of 1.0% nitric acid and analyzed using an inductively coupled argon plasma mass spectrometer (model 5000, Perkin-Elmer/Sciex). Samples were compared with AlCl₃ standards.

All plasticware used for growth and collection of samples was soaked in 20% HCl before use to minimize Al contamination.

Visualization of Callose in Roots

Five-day-old seedlings were exposed to various concentrations of AlCl₃ for 24 h, after which they were transferred to fixative containing 10% formaldehyde, 5% glacial acetic acid, and 45% ethanol, and vacuum infiltrated for 4 h. Fixed seedlings were stained with 0.1% aniline blue (pH 9.0, 0.1 m K₃PO₄). Callose production was visualized under the same conditions as described for morin staining.

Measurement of Root Organic Acid Content and Exudation

One-hundred seeds of either the wild type or specific mutant lines were sown on 250-μm polypropylene mesh supported in a six-well tissue culture plate (Falcon) containing 18 mL of liquid nutrient medium per well. After 5 d of growth, seedlings were rinsed with 100 μm CaCl₂, pH 4.2, and transferred to 18 mL of 100 μm CaCl₂, pH 4.2, containing either 0 or 2.7 μm AlCl₃. Calculations using the Geochem-PC program (Parker et al., 1987) indicated that a concentration of 2.7 μm AlCl₃ in 100 μm CaCl₂, pH 4.2, yielded the same Al³⁺ activity as 25 μm AlCl₃ in the liquid nutrient medium, pH 4.2. After 24 h of growth in the solutions with and without Al, 15-mL samples of each root-bathing solution were collected and passed through an Ag cartridge (OnGuard, Dionex, Sunnyvale, CA) to remove free Cl⁻ for subsequent HPLC analysis.

Solutions were frozen, lyophilized, and resuspended in 600 μL of deionized water. Samples were analyzed on an ion chromatography system (model DX-300, Dionex) fitted with a 4-mm AG-11 guard column and a 4-mm AG-11 ion-exchange analytical column. Samples were separated using an eluent gradient of NaOH in 17% high-purity methanol, and anions were detected by measurement of electrical conductivity. Identification of organic acids in the exudates was achieved by comparison of retention times with those of organic acid standards. Quantification was based on standard curves generated from peak integration of standards.

Roots used for measurement of internal organic acid concentrations were grown in the absence of Al as previously described. Roots from 100 5-d-old seedlings were rinsed in deionized water, harvested, weighed, and ground in 1 mL of 60% ethanol. Cellular debris were pelleted, the supernatant collected, and the pellet resuspended in 95% ethanol. The debris were repelleted and the supernatant was collected and combined with the previous supernatant. The samples were then dried under vacuum and resuspended in 600 μL of deionized water. Concentrations of organic acids in roots were measured as previously described for measurement of organic acids in root exudates.

RESULTS

Isolation of Mutants with Increased Al Resistance

alr mutants were identified by screening M₂ populations of EMS-mutagenized Arabidopsis for seedlings exhibiting improved root growth in medium containing a phytotoxic level of Al. Screening involved the use of gelled-nutrient medium that was equilibrated with a level of AlCl₃ (1.5 mm) that inhibited root growth in wild-type seedlings by 80%. AlCl₃ was introduced into the gel medium by soaking because we found that Al toxicity was variable when Al-containing medium was autoclaved with the gellan gum (Larsen et al., 1996).

Approximately 2500 M₂ seedlings were screened from each of 40 mutagenized pools, resulting in the isolation of 57 putative mutants from approximately 1 × 10⁵ seedlings. Rescreening of M₃ seedlings confirmed that 6 mutants from different pools were Al resistant (Table I). An “in-gel” AlCl₃ dose-response curve for the alr mutants demonstrated increased Al resistance at concentrations from 0.25 to 1.5 mm AlCl₃ (data not shown). Root growth in the wild type was inhibited in gel medium equilibrated with 0.75 mm or higher concentrations of AlCl₃. In contrast, root growth in alr-128 was inhibited only when grown in gelled-nutrient medium equilibrated with AlCl₃ concentrations ≥ 1.25 mm. It should be noted that root growth in the absence of Al was similar for wild type and all of the alr mutants.

Table I.

Characteristics of alr mutants

Mutant	Genetic Dominance	Map Location	Genetic Distance
alr-104	Semidominant	chr.a4, BP-1.8	10.7 cMb(n = 48)
alr-108	Semidominant	chr. 1, OPW-10	16.8 cM (n = 14)
alr-128	Semidominant	chr. 1, OPW-10	15.4 cM (n = 87)
alr-131	Semidominant	chr. 1, OPW-10	21.2 cM (n = 26)
alr-139	Semidominant	chr. 1, OPW-10	15.3 cM (n = 19)
alr-142	Semidominant	NDc	ND

^achr., Chromosome. 

^bcM, centimorgan. 

^cND, Not determined. 

Genetic Analysis of alr Mutants

To determine the mode of inheritance of Al resistance, F₂ progeny from a cross of each alr mutant (male) with wild type (gl-1 female) were analyzed on gelled-nutrient medium equilibrated with 1.5 mm AlCl₃. The segregation of Al resistance in each of the crosses was consistent with the expectation that Al resistance is a semidominant trait (data not shown). The major class of F₂ progeny in each cross was intermediate in resistance between the two parents.

Because Al resistance was semidominant, it was not possible to sort the alr mutants unambiguously by complementation analysis. To determine whether the alr mutants represented unique genetic loci, the mutations were mapped on the Arabidopsis genome (Table I). The Ws-0 ecotype was chosen as a mapping partner because it was shown previously that roots of Ws-0 seedlings had Al sensitivity similar to that of Col-0 (Larsen et al., 1996). Homozygous wild-type (non-Al-resistant) F₂ progeny were used to construct a mapping population. The genotypes of the F₂ progeny were confirmed by the analysis of F₃ families. alr-104 mapped near the cleaved amplified polymorphic sequence marker B9–1.8 on chromosome 4 (Table I) (Konieczny and Ausubel, 1993). Four other alr mutants (alr-108, alr-128, alr-131, and alr-139) mapped on chromosome 1 near the microsatellite marker nga280 (Table I). Mapping analysis was not carried out for alr-142 because it has a weak phenotype and is difficult to follow in crosses.

RAPD mapping was performed to identify an anonymous marker that was more closely linked to alr-128 than nga280. Pools of DNA from homozygous Al-resistant and wild-type F₂ plants generated from the cross described above were compared by RAPD analysis. A polymorphic band derived from primer OPW-10 appeared in the resistant pools but not in the wild-type pools and was determined to represent a site tightly linked to the alr-128 locus. Inheritance of the OPW-10 polymorphism was examined for each of the alr mutants that mapped near alr-128 (Table I). The results indicate that these alr mutants map closely together on chromosome 1 and either represent a cluster of genes affecting Al resistance or are alleles of the same gene.

Growth of alr Mutants in Al-Containing Solution Culture

Dose responses were analyzed for seedlings grown hydroponically in liquid nutrient medium rather than on gels because Al interacts with the gel matrix, dramatically lowering its phytotoxicity and presumably the concentration of free Al³⁺ (Larsen et al., 1996). To relate the resistance of the alr mutants to Al³⁺ concentration, it was necessary to examine root growth in hydroponic solution supplemented with physiologically relevant concentrations of AlCl₃. For the studies of Al effects on root growth and root Al accumulation in the alr mutants, alr-128 was chosen to represent the four mutants that map to the same locus on chromosome 1 because it exhibited the strongest phenotype of the four mutants. It was examined in comparison with wild type and alr-104, which was the only mutant to map to chromosome 4. Seedlings of Col-0 wild type, alr-104, and alr-128 were germinated and grown for 4 d in nutrient solution in the absence of added AlCl₃, after which the root length of one-half of the seedlings was measured. The remainder of the seedlings were grown for an additional 2 d in nutrient solution containing varying concentrations of AlCl₃ (0–50 μm). The relative growth increment (percentage increase in root length during 2 d of Al exposure) was determined for each mutant and for the wild-type seedlings (Fig. 1).

Figure 1.

Growth of Arabidopsis roots in hydroponic solution culture containing Al. Al-dependent root growth inhibition was compared for wild type (wt), alr-104, and alr-128. Seedlings were grown for 4 d in nutrient solution with no added AlCl₃, pH 4.2, and then transferred to nutrient solution containing varying concentrations of AlCl₃ and grown for an additional 2 d. Relative root growth increment in the presence of Al was expressed as (RL d 6 − RL d 4)/(RL d 4) × 100, where RL = root length. Error bars represent the se (n = 50).

Root growth of wild-type seedlings was inhibited at concentrations of AlCl₃ as low as 10 μm, and declined sharply at 20 μm. In contrast, at concentrations between 10 and 40 μm AlCl₃, root growth in both alr-104 and alr-128 decreased gradually, with maximal differences in Al resistance between wild type and the alr mutants occurring within this concentration range. In nutrient solution containing 50 μm AlCl₃, root growth of both wild type and the alr mutants was completely inhibited. A concentration of 50 μm AlCl₃ in our nutrient solution represents an Al³⁺ activity of 3.9 μm, as determined by the Geochem-PC program (Parker et al., 1987).

Growth of alr Mutants in La-Containing Solution Culture

To determine whether the metal resistance in the alr mutants was specific for Al, the mutants were grown in nutrient solution containing another phytotoxic trivalent cation, La³⁺, which has rhizotoxic effects similar to those of Al. However, it has been previously shown that Al-resistant wheat genotypes were not La resistant (Kinraide et al., 1992; Ryan et al., 1995a). The sensitivity of root growth to La in wild type, alr-104, and alr-128 was assessed by growing seedlings in the same nutrient solution used for determining Al resistance, except that the medium containing LaCl₃ was prepared without KH₂PO₄ to prevent the precipitation of La. During a 2-d exposure to 15 μm LaCl₃, roots of wild type (Col-0) were inhibited by 60 ± 4.3%. Similar La-dependent inhibition of root growth was observed for the alr mutants, with roots of alr-104 inhibited by 55 ± 6% and those of alr-128 inhibited by 58 ± 4.7%.

Callose Accumulation

Callose deposition is an indicator of Al-induced stress because it accumulates in root tips after exposure to toxic levels of Al (Wissemeir et al., 1987; Schreiner et al., 1994; Zhang et al., 1994). Callose accumulation was examined in the alr mutants to determine if increased Al resistance resulted in reduced callose accumulation after exposure to Al. Four-day-old seedlings were transferred for 24 h to nutrient solution containing levels of Al that were more strongly toxic to wild type compared with the alr mutants, after which roots were fixed and stained with aniline blue, pH 9.0. Root tips were examined for callose fluorescence using epifluorescence microscopy (Fig. 2A). Untreated wild-type root tips exhibited little background fluorescence, whereas wild-type root tips exposed to Al fluoresced intensely, indicating callose accumulation. Compared with wild type, callose accumulation in alr-128 in the presence of Al was reduced dramatically to levels slightly greater than that seen for the untreated controls. Callose accumulation was only modestly reduced in roots of alr-104.

Figure 2.

Accumulation of Al and callose in roots after exposure to AlCl₃. A, Patterns of Al-dependent callose accumulation were examined for roots of wild type (wt), alr-104, and alr-128. Roots of 4-d-old seedlings were exposed to nutrient solution containing 75 μm AlCl₃ for 24 h, except for the first panel, in which no Al was added (−Al). Seedlings were then fixed, stained with 0.1% aniline blue, pH 9.0, and observed using epifluorescence microscopy. The top row represents bright-field images of the treated roots, and the bottom row are fluorescence images indicating callose accumulation. B, Patterns of Al accumulation by roots of wild type, alr-104, and alr-128 were observed using the Al-indicator dye morin. Five-day-old seedlings grown in nutrient solution without Al were exposed to nutrient solution containing 25 μm AlCl₃ for 1 h, except for the first panel, in which no Al was added (−Al). Roots were then stained with 100 μm morin, a stain that fluoresces when complexed with Al.

Patterns of Al Accumulation

Al-exclusion mechanisms, including the release of Al-chelating organic acids, reduce Al accumulation in root tips of Al-resistant cultivars. Therefore, we investigated the possibility that increased Al resistance in the alr mutants was associated with a decrease in root apical Al accumulation after short-term exposure to Al.

Roots of intact wild-type, alr-104, and alr-128 seedlings were exposed to a nutrient solution containing 25 μm AlCl₃ for either 0 or 3 h, after which time roots were rinsed for 5 min in nutrient solution without Al. Root tips, representing the terminal 1 to 2 cm of the roots, were analyzed for Al content using ICP-MS. As shown in Table II, analysis of 0-h samples revealed a minimal amount of background Al. After 3 h of Al exposure, wild type accumulated 48% more Al than alr-104 and 61% more than alr-128.

Table II.

Root Al concentrations determined by ICP-MS analysis

Plant Type	No Desorptiona		Desorptionb
	0 h	3 h	0 h	3 h
	nmol Al mg−1 dry wt
Wild type	0.2  ± 0.1	42.3  ± 3.4	0.2  ± 0.1	29.3  ± 0.8
alr-104	0.3  ± 0.1	28.6  ± 1.6	0.4  ± 0.2	31.7  ± 0.9
alr-128	0.1  ± 0.1	26.3  ± 2.2	0.3  ± 0.1	30.7  ± 1.5

Values are means of six replicates ± se.

^aOne-hundred roots were exposed to 25 μm AlCl₃ for 0 or 3 h and then washed for 5 min in nutrient solution containing no Al. 

^bOne-hundred roots were exposed as described above and were then desorbed in 0.5 mm citrate, pH 4.2, at 0°C for 30 min. 

Al bound to the root cell wall was desorbed with an ice-cold 0.5 mm citrate solution following the methods of Zhang and Taylor (1990) to determine if the differences in root Al accumulation between wild type and the alr mutants was attributable to symplastic or apoplastic Al. It should be noted that it is likely that this desorption regime does not remove all of the cell wall Al (see Rengel, 1996), but primarily removes Al in the free space and loosely bound to the cell wall. Under these conditions, after desorption the Al left in the root would be either tightly bound to the cell wall or sequestered in the symplasm. Citrate desorption reduced the amount of Al that accumulated in wild-type roots, but had no effect on Al concentrations in roots of alr-104 and alr-128 (Table II), suggesting that Al accumulated in alr roots represents Al that cannot easily be desorbed from the root cell wall. Therefore, the fraction of Al desorbed from wild-type roots may represent a free pool of apoplastic Al that was not present in the alr roots.

Wild-type seedlings and the alr mutants were further examined for their capability to exclude Al by using the Al-indicator dye morin. The use of morin, which fluoresces when complexed with Al, allows for better spatial resolution of Al accumulation in the root apex, the primary site of Al toxicity. Seedlings were exposed to 25 μm AlCl₃ for 1 h, washed, and then stained with morin. As shown in Figure 2B, morin staining revealed that there were significant differences between wild type and the alr mutants in terms of Al accumulation within the root apex. Roots not exposed to Al exhibited a very slight fluorescence. In contrast, intense morin fluorescence was observed in wild-type root tips exposed to 25 μm AlCl₃. Morin fluorescence in Al-treated root tips of both alr-104 and alr-128 was reduced compared with wild type, with alr-128 exhibiting the lowest levels of fluorescence after the 1-h exposure to Al. This was especially true right at the root apex of alr-128, where staining intensity approximated that seen in untreated wild-type roots.

Organic Acid Release

In wheat and maize, increased Al resistance is associated with Al-induced exudation of Al-chelating organic acids into the rhizosphere (Delhaize et al., 1993a, 1993b; Pellet et al., 1995). To determine if a similar mechanism operates in Arabidopsis, the alr mutants were examined for altered patterns of organic acid release in comparison with wild-type seedlings.

Seedlings of wild type and each alr mutant were grown in sterile culture for 5 d in nutrient solution and subsequently transferred to a sterile solution consisting of 100 μm CaCl₂, pH 4.2, containing either 0 or 2.7 μm AlCl₃, and grown for an additional 24 h. Subsequently, the root-bathing solution was collected and analyzed for organic acids. These root-exudate experiments were carried out in simple salt (CaCl₂) solutions to eliminate interference from the high levels of inorganic anions in the nutrient solution (i.e. NO₃⁻, SO₄²⁻, and PO₄³⁻) during HPLC analysis. It was determined with the Geochem-PC program that the addition of 2.7 μm AlCl₃ to the low-salt solution (100 μm CaCl₂) would yield the same Al³⁺ activity as 25 μm AlCl₃ in the full nutrient solution.

When roots of wild-type seedlings were incubated in the root-bathing solution without added Al, the profile of organic acid release during the 24-h time period was complex (Fig. 3). Citrate was the predominant organic acid released by wild-type roots, along with smaller amounts of pyruvate, succinate, and malate. Exposure to 2.7 μm Al³⁺, which inhibited root growth in wild-type seedlings but not in the alr mutants, did not significantly change the profile or magnitude of organic acid release by wild-type roots.

Figure 3.

Organic acid exudation by roots of wild-type (wt) seedlings and alr mutants in the presence or absence of Al. One-hundred seedlings of wild type (Col-0), alr-104, alr-108, alr-128, and alr-131 were grown for 5 d in nutrient solution and then transferred to a solution consisting of 100 μm CaCl₂, pH 4.2, containing either 0 or 2.7 μm Al³⁺ for 24 h. Root exudates were collected, free Cl⁻ was removed, and samples were analyzed using an ion-chromatography system. Each panel represents the total picomoles of organic acid released by 100 roots during the 24-h period in the absence (−Al) or presence (+Al) of added Al. Error bars represent the se (n = 6).

The profiles of released organic acids indicate that the group of mutants mapping on chromosome 1 (alr-108, alr-128, and alr-131) are phenotypically similar but not identical (Fig. 3). These mutants constitutively exhibited an increased rate of malate and/or citrate exudation. alr-108, alr-128, and alr-139 each exhibited stimulated malate release, particularly alr-128, which had a 2-fold increase in malate exudation compared with wild type. In the absence of Al, each mutant exhibited a slight to moderate increase in citrate release. This was most evident for alr-131, a mutant that did not exhibit enhanced malate exudation. Increased rates of citrate exudation were not sustained when the mutants were exposed to Al. Each of the four chromosome 1 mutants exhibited a 2- to 3-fold stimulation in pyruvate exudation (which does not effectively chelate Al). alr-104, the lone mutation on chromosome 4, did not exhibit an enhanced citrate or malate exudation compared with wild type (Fig. 3). Thus, the mechanism of Al resistance in alr-104 does not appear to involve stimulated organic acid exudation.

Internal Organic Acid Concentrations

To determine if the increased release of organic acids by roots of the alr mutants that map to chromosome 1 was driven primarily by increased organic acid synthesis, organic acid concentrations were quantified in wild-type and alr-128 roots using HPLC. For comparison purposes, we also quantified organic acid levels in roots of alr-104, even though it did not exhibit enhanced organic acid release. As shown in Table III, there were only very modest changes in root organic acid concentrations between wild type and the two alr mutants. Roots of wild-type and alr-128 seedlings were similar in pyruvate concentration, whereas roots of alr-104 contained no detectable pyruvate. Concentrations of malate were similar in roots of wild type and alr-104, with alr-128 roots exhibiting a modest (<10%) increase in malate levels. Finally, root citrate levels were similar in wild type and alr-128, whereas alr-104 exhibited a significantly greater concentration of citrate in the roots (an approximately 30% increase). It does not appear that these changes in root organic acid content could account for the increased organic acid release in roots of alr-128 or the other alr mutants that map to the same locus on chromosome 1.

Table III.

Concentrations of organic acids in Arabidopsis roots

Plant Type	Pyruvate	Malate	Citrate
	nmol g−1 fresh wt
Wild type	83.15  ± 10.30a	203.86  ± 8.91	167.53  ± 8.45
alr-104	NDb	191.39  ± 8.80	204.19  ± 8.97
alr-128	87.19  ± 9.92	236.17  ± 11.2	150.48  ± 10.16

^an = 8. 

^bND, Not detectable. 

DISCUSSION

Al Resistance in Arabidopsis

Al resistance in other plant species is usually a dominant trait and in some cases is considered to be a gain-of-function character (Lafever and Campbell, 1978; Delhaize et al., 1993a; Carver and Ownby, 1995). Because of this, the recovery of alr mutants from a standard Arabidopsis EMS-mutagenesis procedure was somewhat unexpected. Initial screening resulted in the identification of six unique mutants from a mutagenized pool of approximately 1 × 10⁵ seeds, indicating that Al resistance appears at a low frequency in mutagenized populations of Arabidopsis. It is not clear whether the Al resistance in the alr mutants results from gain-of-function or loss-of-function mutations. Enhanced organic acid release was found to be correlated with increased Al resistance and it is possible that mechanisms involved in organic acid synthesis or transport may be altered in such a way as to increase organic acid production or transport out of the root. Alternatively, increased Al resistance could result from a loss-of-function mutation such as a defect in a repressor, resulting in either ectopic or elevated expression of a gene responsible for Al resistance.

We have identified at least two loci in Arabidopsis that confer Al resistance, which does not appear to be as genetically complex in Arabidopsis as Al sensitivity, for which eight different loci have been found (Larsen et al., 1996). All eight Al-sensitive loci have not been mapped yet, but so far we have no evidence that the mutations affecting Al sensitivity and resistance involve common genes. Certainly, the Al-sensitive mutations may affect the same mechanisms affected by alr mutations, but they probably would do so in different ways. It is likely that some of the Al-sensitive mutations are just as important in understanding Al resistance as the alr mutations.

To varying degrees, both alr-104 and alr-128 appear to exclude Al from the root tip. This was shown both by staining root tips with Al-indicator dyes and by quantifying Al accumulation with ICP-MS. Differences in Al accumulation between wild type and the alr mutants appeared to be much greater when analyzed by morin staining than by ICP-MS. This is probably a result of technical limitations attributable to the difficulty of isolating only root tips for analysis via ICP-MS. In both wheat and maize, Al exclusion and organic acid release were specifically localized to the root apex, and there were no differences in Al accumulation for mature root regions between Al-resistant and Al-sensitive genotypes (see Rincon and Gonzales, 1992). We also found little difference in Al accumulation between the alr mutants and wild-type seedlings in the mature root regions based on results from morin staining (data not shown). The smallest root tip segments we could obtain by harvesting root tips en masse were about 5 mm in length. Because the root apex in Arabidopsis is only about 1 mm long, it would be expected that the differences in Al concentration between root tips of the alr mutants and wild type as determined by ICP-MS analysis are an underestimation of the actual differences within the root apex.

In addition to the fact that root growth in alr-104 and alr-128 is resistant to Al, each accumulated less callose, which is a good indicator of Al toxicity in roots (Wissemeir et al., 1987; Schreiner et al., 1994; Zhang et al., 1994). After Al exposure, callose accumulation in the roots of both alr mutants was reduced compared with wild type. Unlike Al-sensitive Arabidopsis mutants, in which no correlation was found between Al sensitivity and Al-dependent callose accumulation (Larsen et al., 1996), a reduction in callose accumulation in the alr mutants appears to be a good marker of increased Al resistance.

Organic Acid Exudation in Relation to Al Resistance

alr-108, alr-128, alr-131, and alr-139 were found to be closely linked to the microsatellite marker nga280 on chromosome 1. Further mapping with RAPD markers demonstrated that each mapped within a few centimorgans of a single RAPD marker, suggesting that they are tightly clustered or represent alleles of the same gene. Mutants in this group are phenotypically similar in that they have nearly identical profiles of organic acid release (except alr-131, which does not exhibit enhanced rates of malate release but does exhibit a moderate [40%] increase in citrate release). The differences between alr-131 and the other mutants may be the result of allelic variation or may indicate that alr-131 actually represents another closely linked gene.

Both malate and citrate have high affinities for Al³⁺ (Hue et al., 1986; Delhaize et al., 1993b). Recent studies have shown that Al-resistant, near-isogenic lines of wheat bearing the Alt1 gene exhibit a severalfold stimulation in the rate of root malate exudation compared with the Al-sensitive line. The function of Alt1, which is responsible for differential Al resistance, has not been ascertained. When Ryan et al. (1995b) compared a number of different wheat cultivars exhibiting varying degrees of Al resistance, a good correlation was found between Al-inducible malate release and Al resistance. A similar mechanism has been described in maize, in which Al exclusion from the root tip in Al-tolerant lines is correlated with an Al-inducible exudation of citrate localized to the root apex (Pellet et al., 1995).

Malate (and Cl⁻) efflux across the plasma membrane coupled to K⁺ efflux has been studied extensively in relation to stomatal closure. Work by several groups has shown that anion channels are responsible for the observed release of Cl⁻ and malate during stomatal closing (Linder and Raschke, 1992; Hedrich and Becker; 1994; Hedrich et al., 1994; Schroeder, 1995). Similar anion channels have been studied in the plasma membrane of epidermal cells of Arabidopsis hypocotyls and are assumed to exist throughout the plant, including the root (Thomine et al., 1995). Therefore, it is presumed that organic acid release from root cells is mediated by an outward-rectifying, plasma membrane-localized anion channel.

Because plant anion channels can be regulated by a number of external and internal factors (Ebel and Cosio, 1994; Schroeder, 1995; Thomine et al., 1995; Ward et al., 1995), it is possible that at least in wheat and maize, Al may play a role in anion-channel gating (Delhaize et al., 1993b; Pellet et al., 1995, 1996; Ryan et al., 1995a, 1995b). Support for this possibility comes from the results of a recent study by Ryan et al. (1997), which were based on an electrophysiological (patch-clamp) analysis of protoplasts isolated from the root apex of the Al-resistant wheat line used to initially demonstrate Al-inducible malate release. They presented evidence for an Al-activated anion channel in the plasma membrane of these protoplasts that could play a key role in the Al-exclusion response previously studied in intact roots. The physiological characteristics of organic acid release observed in the Arabidopsis alr mutants may be somewhat different than that seen in Al-resistant lines of maize and wheat. Unlike maize and wheat, the altered pattern of organic acid release observed, for example, in alr-128 does not appear to be Al induced, suggesting that the mutation responsible for this phenotype does not result in a change in the Al-dependent regulation of the transporter.

However, it is possible that the increased organic acid release of alr-128 results from a change in the Al-independent gating of an organic acid transporter, thus allowing constitutive increases in citrate, malate, and pyruvate release. The anion channel involved in guard-cell closure has been reported to have low ion selectivity, with the capability to transport a broad range of anions out of the guard cell (Schmidt and Schroeder, 1994; Schroeder, 1995). The low selectivity of such a transporter involved in anion efflux suggests that organic acids in roots may be transported via a similar system.

The observed increase in the rate of organic acid exudation by roots of alr-128 could also represent some change in the compartmentalization of organic acids. Citrate, malate, and pyruvate are all organic acid species that accumulate in the cytoplasm after export of citrate from mitochondria. It is possible that the increased release of organic acids out of the root may result from an alteration in citrate export from mitochondria. Based on the data shown in Figure 3, which indicate that in alr-128 citrate efflux is increased slightly, whereas malate efflux is doubled and pyruvate release is tripled, it can be speculated that a mutation that leads to either enhanced transport of citrate from the mitochondria or reduced citrate transport into the vacuole results in the progressive increase in the cytoplasm of derivatives of cytoplasmic citrate. Inherent in this speculative model is the concept that increased levels of cytoplasmic organic acids lead to increased transport of these species out of the cell. Support for this comes from the recent work by de la Fuente et al. (1997), in which transgenic tobacco and papaya seedlings expressing bacterial citrate synthase resulted in increased citrate accumulation in the root cytoplasm, enhanced citrate exudation from the root, and a significant increase in Al resistance.

Apparently, Al resistance in alr-104, a locus that maps to a different chromosomal location than alr-128, arises from a different mechanism than enhanced organic acid release. Degenhardt et al. (1998) describe a mechanism by which roots of alr-104 exhibit an Al-inducible increase in rhizosphere pH, resulting in a decrease in the activity of the rhizotoxic Al³⁺ species in the rhizosphere.

This study represents the first report to our knowledge of Arabidopsis mutants selected for increased Al resistance in roots. Isolation of such mutants will allow for further characterization of Al-resistance mechanisms and may provide an opportunity to identify other Al-resistance mechanisms that have not been previously described. In addition, such mutants provide the opportunity to isolate genes responsible for these mechanisms. This is significant because previous work to characterize Al resistance mechanisms has been performed in plants with complex genomes, such as wheat and maize, in which it is more difficult to isolate the genes involved in Al resistance.

Abbreviations:

EMS

ethyl methylsulfonate

ICP-MS

inductively coupled argon plasma MS

RAPD

random amplified polymorphic DNA