Arch Microbiol(1991) 156:507- 512
Archives of
Micrnbiology
9 Springer-Verlag199t
Aluminium toxicity and binding to Escherichia coli Laura Guida 1, Ziba Saidi 1, Martin N. Hughes 2, and Robert K. Poole 1 1 MicrobialPhysiologyResearch Group, BiosphereSciencesDivision,King's CollegeLondon, Campden Hill Road, London W8 7AH, UK 2 ChemistryDepartment, King's CollegeLondon, London,UK Received March 6, 1991/AcceptedAugust 6, 1991
Abstract. The toxicity and binding of aluminium to Escherichia coli has been studied. Inhibition of growth by aluminium nitrate was markedly dependent on pH; growth in medium buffered to pH 5.4 was more sensitive to 0.9 mM or 2.25 mM aluminium than was growth at pH 6.6-6.8. In medium buffered with 2-(N-morpholino)ethanesulphonic acid (MES), aluminium toxicity was enhanced by omission of iron from the medium or by use of exponential phase starter cultures. Analysis of bound aluminium by atomic absorption spectroscopy showed that aluminium was bound intracellularly at one type of site with a Km of 0.4 mM and a capacity of 0.13 tool (g dry wt)- 1. In contrast, binding of aluminium at the cell surface occurred at two or more sites with evidence of cooperativity. Addition of aluminium nitrate to a weakly buffered cell suspension caused acidification of the medium attributable to displacement of protons from cell surfaces by metal cations. It is concluded that aluminium toxicity is related to pH-dependent speciation [with AI(H20)6a+ probably being the active species] and chelation of aluminium in the medium. Aluminium transport to intracellular binding sites may involve Fe(III) transport pathways. Key words" Aluminium and bacteria - Metal speciation - Iron transport - Biosorption of metals -- Metalmicrobe interactions - Escherichia coli
Aluminium is one of the most abundant elements in the earth's crust, but has no established biological function. Its potential toxicity to biological systems is ameliorated by its existence as oxides and complex aluminosilicates and by its insolubility in aqueous media at neutral pH. However, aluminium may be leached from soils by acid rain, washed into lakes and rivers and reach toxic levels. Toxicity to microbes may be manifest, for example, as Offprint requests to: R. K. Poole
inhibition of microbial soil activity (Firestone et al. 1983; Thompson and Medve 1984) and inhibition of photosynthesis and nitrogen fixation in cyanobacteria (Pettersson et al. 1985a). There is currently grave concern over the toxicity to Man of aluminium in drinking water and the implication of aluminium from various domestic sources as the aetiological agent in Alzheimer's disease, dialysis encephalopathy syndrome and related neurological disorders (for references see Guy et al. 1990). In aqueous solutions, AI(III) forms an octahedral hexahydrate, AI(H20)63+ (A13+) at acid pH. Successive pH-dependent deprotonations of the aluminium-bound water yield AI(I-IzO)5(OH) z +, AI(H20)g(OH) +, AI(OH)3 and AI(OH)g (Martin 1986). At neutral pH, the dominant form is the insoluble precipitate AI(OH)3. In view of this speciation (Hughes and Poole 1991) it is imperative to consider the effects of pH and the choice of buffer on aluminium toxicity. The ionic radius of A13+ (67.5 pm) is close to those of Fe 3+ (78.5 pro) and Ga 3+ (76.0 pm), although formation constants for the reaction of aluminium with the ligands EDTA, transferrin and citrate are notably lower than those for these metals (Hughes and Poole 1989). The similarity between these metals suggests that aluminium might interact with microorganisms by competition with iron. Indeed, certain fungal siderophores transport Fe(III), AI(III) or Ga(III) (Raymond and Carrano 1978) although in the cyanobacterium Anacystis eylindrica, aluminium transport occurs by passive diffusion (Petersson et al. 1986). Unlike iron, however, aluminium is bound exclusively by oxygen ligands. The toxicity of aluminium often appears attributable to slow ligand exchange rates. For example, the toxicity of aluminium often results from inhibition of Mg2+-dependent enzymes because ligand exchange is about 105 times slower for A13+ than for Mg 2+ (Martin 1986). To further our understanding of aluminium toxicity and binding, we have begun a study of aluminium interactions with "model" microbial systems and report here factors affecting the toxicity of aluminium to Escherichia coli, an organism with well-studied metal transport sys-
508 terns, especially for Fe(III). The results suggest c o m p e tition o f A13 + a n d F e 3 + a n d i n d i c a t e c o m p l e x b i n d i n g o f A13 + to cell surfaces.
E
Methods
c,
Organism and growth conditions
r o e..
Escherichia eoli RPI (= AN2572, prototroph) was grown in a medium that contained (g. 1-1): Tris buffer, 4.85; NH4C1, 1.0; glycerol, 50; KCI, 1.0; CaC12 - 2H20, 0.01 ; K2SO,~, 0.87. A trace element solution (Poole and Haddock 1974; 10ml-l-1), flglycerophosphate (16.5%, w/v; 10 ml. 1-1) and MgCI2 - 6H20 (20% w/v; 1 ml - 1-1) were autoclaved separately and added aseptically to the sterile medium, (pH 7.4). For growth at lower pH values, the Tris buffer was replaced by succinic acid (pH 4.8 to 6) or 2-(Nmorpholino)ethanesulphonicacid (MES; pH 5.4 to 6.8) and the pH adjusted to the required value with NaOH. The buffers used (MES and succinate) were chosen for their appropriate pK values and their low binding affinities for metals in general, in preference to citrate or phosphate, for example (Hughes and Poole 1991). For iron-deficient medium, FeC13 was omitted from the trace element solution. Where the iron content of the medium required supplementation, a 6 mM solution of FeC13 in 12 mM EDTA was adjusted to the same pH as the medium and added as a filtersterilized solution. Aluminium was freshly prepared as a 0.2 M solution of Al(NO3)3 in distilled water which was filter-sterilized or autoclaved with similar results. Cultures were grown in 20 ml medium in 250 ml conical flasks shaken at 200 rpm and 37~ inoculated at 10% (v/v). Growth was monitored by measuring apparent absorbance (optical density) at 600 nm (after appropriate dilution to give OD6oo < 0.5) in cuvettes of 1 cm path length using a Pye Unicam SP6-450 spectrophotometer (Cambridge, UK).
Analysis of aluminium in cells, media and washes Cells were harvested by centrifugation from overnight cultures (50 ml), washed in 100 ml medium, and the pellets suspended in MES buffer (pH 6) to a volume of 10 ml. To 9 ml portions of aluminium solutions prepared in the same buffer or to 9 ml of buffer only (controls) were added, in glass tubes, 1 ml aliquots of the cell suspension. At each aluminium concentration, a control was included without cells to investigate loss of the metal to the tube surface. After incubation at 37~ for 3 - 4 h, cells were pelleted by centrifugation, then washed successively in 4 ml buffer and 4 ml 0.1 M HNO3. All supernatant liquids were kept separately, acidified with HC1 (1% of final vol.) and made up to known volumes with distilled water. For tubes that contained no cells, the entire volume was transferred to 25 ml volumetric flasks, acidified as above and made up to volume. Cell pellets were digested by adding 100 gl conc. H2SO4, heating to 120~ and then adding 50 ~tl aliquots of conc. HNO3 until dissolution was complete. The HNO3 was then boiled off and the digests transferred to 5 ml volumetric flasks. The tubes were washed thorougly with distilled water and the washings pooled with the final volume. Aluminium assays were performed using a Perkin-Elmer 2380 atomic absorption spectrophotometer (PerkinElmer, Beaconsfield, UK) on a N20/acetylene flame at the following settings: 2, 309.3 nm; lamp current, 9 mA; spectral band width, 0.7 nm. The standards were prepared by dilution of a commercial standard. Dry weights of cell suspensions used in aluminiumbinding experiments were determined in duplicate by filtering a diluted suspension through preweighed 50 mm filters (0.45 gm pore size) and drying to constant weight at 105~C.
1 0.6
s o r t~
0.3
~~ ~
t~
> 108.8 (e.g. 10 is at 1 m M A1, pH 6), toxicity always increased at lower p H values, even though Al13 is more than an order of magnitude more toxic than A13 + (Parker et al. 1989). We thus identify a mononuclear species,
% Recovery
Aluminium in control tubes (lag)
% Recovery compared to controls
44 51 48 54 55 52
1800 1350 425 155 55 ~20
88 92 82 85 73 84
probably A1a+, as the toxic species. Pettersson et al. (1985 a) have claimed that aluminium toxicity to another prokaryote, Anabaena cylindrica, at high p H was due to mononuclear hydroxy-A1 species (AI[OH]~] -")+, where n = 1 to 4 and coordinated water has been omitted), but the solutions were likely to have contained A113. Succinate buffer appears more protective than MES, suggesting that the former has a higher affinity for A13 § than does MES; indeed, succinic (as well as citric and malic) acid are effective in protection against acute aluminium toxicity in Man (Llobet et al. 1987; Domingo et al. 1988). The proposal that the A13§ ion is the major toxic species is supported by experiments that suggest interaction of the transport systems for A13+ and Fe 3+. Iron(III), AI(III) [and Ga(III)] have similar effective ionic radii (see Hughes and Poole 1989) and each forms complexes with siderophores (Tait 1975; Theil et al. 1983), transferrin and citrate. Although the small size of the aluminium cation prevents effective binding of large ligands, aluminium is known to be transported by certain iron uptake pathways. For example A13+- (or G a 3+-) substituted ferrichrome is transported by the fungus Ustilago (Emery and Hoffer 1980). In the present study, aluminium toxicity was exacerbated by iron deficiency, perhaps allowing more effective competition of aluminium for iron-binding sites.
512
References Davis WB, McCauley MJ, Byers BR (1971) Iron requirements and aluminium sensitivity of an hydroxamic acid-requiring strain of Bacillus megaterium. J Bacteriol 105 : 589 - 594 Domingo JL, Gomez M, Llobet JM, Corbella JM (1988) Comparative effects of several chelating agents on the toxicity, distribution and excretion of aluminium. Hum Toxicol 7 : 2 5 9 - 262 Emery T, Hoffer PB (1980) Siderophore-mediated mechanism of gallium uptake demonstrated in the microorganism Ustilago sphaerogena. J Nucl Med 21:935--939 Firestone MK, Killham R, McColl JG (1983) Fungal toxicity of mobilized soil aluminium and manganese. Appl Environ Microbiol 46:758 - 761 Guy SP, Seabright PJ, Day JP, Itzhak RF (1990) Uptake of aluminium by human neuroblastoma cells. Biochem Soc Trans 18 : 3 9 2 - 393 Hughes MN, Poole RK (1989) Metals and microorganisms. Chapman and Hall, London Hughes MN, Poole RK (1991) Metal speciation and microbial growth - the hard (and soft) facts. J Gen Mierobio1137:725734 Kinraide TB, Parker DR (1989) Assessing the phytotoxicity of mononuclear hydroxy-aluminium.Plant Cell Environ 12: 4 7 9 487 Lawford HG, Garland PB (1972) Proton translocation coupled to quinone reduction by reduced nicotinamide-adenine dinucleotide in rat liver and ox heart mitochondria. Biochem J 130:1029 -- 1044 Llobet JM, Domingo JL, Gomez M, Tomas JM, Corbella J (1987) Acute toxicity studies of aluminium compounds: antidotal efficacy of several chelating agents. Pharmacol Toxicol 60:280-283 Macdonald TL, Martin RB (1988) Aluminium ion in biological systems. Trends Biochem Sci 13 : 15-19 Martin RB (1986) The chemistry of aluminium as related to biology and medicine. Clin Chem 32:1797-1806
Parker DR, Kinraide TB, Zelazny LW (1988) Aluminium speciation and phytotoxicity in dilute hydroxy-aluminium solutions. Soil Sci Soc Am J 52:438-444 Parker DR, Kinraide TB, Zelazny LW (1989) On the phytotoxicity of polynuclear hydroxy-aluminium complexes. Soil Sci Am J 53 : 789-- 796 Pettersson A, H/illbom L, Bergman B (1985a) Physiological and structural responses of the cyanobacterium Anabaena cylindrica to aluminium. Physiol Plant 63 : 153 -- 158 Pettersson A, Kunst L, Bergman B, Roomans GM (1985 b) Accumulation of aluminium by Anabaena cylindrica into polyphosphate granules and cell walls: an X-ray energy-dispersive microanalysis study. J Gen Microbiol 131:2545-2548 Pettersson A, Hallbom L, Bergman B (1986) Aluminium uptake by Anabaena cylindrica. J Gen Microbiol 132:1771 -- 1774 Poole RK, Haddock BA (1974) Energy-linkedreduction ofnicotinamide-adenine dinucleotide in membranes derived from normal and various respiratory-deficient mutant strains of Escherichia coil Biochem J 144: 7 7 - 85 Raymond KN, Carrano CJ (1978) Coordination chemistry and microbial iron transport. Acc Chem Res 12:183 - 190 Schwyn B, Neilands JB (1987) Universal chemical assay for the detection and determination of siderophores. Anal Biochem 160:47-56 Tait HG (1975) The identificationand biosynthesis of siderochromes formed by Micrococcus denitrificans. Biochem J 146:191 - 2 0 4 Theil EC, Eichhorn GL, Marzilli LG (1983) Iron-binding proteins without cofactors or sulfur clusters. Advances in Inorganic Biochemistry, vol. 5. Elsevier, New York Thompson GW, Medve RJ (1984) Effects of aluminiumand manganese on the growth of etomycorrhizal fungi. Appl Environ Microbiol 48 : 556-- 560 Van Haecht JL, Bolipombo M, Rouxhet PG (1985) Immobilization of Saccharomyces cerevisiae by adhesion: treatment of the cells by A1 ions. Biotechnol Bioeng 27:217--224
Archives of
Micrnbiology
9 Springer-Verlag199t
Aluminium toxicity and binding to Escherichia coli Laura Guida 1, Ziba Saidi 1, Martin N. Hughes 2, and Robert K. Poole 1 1 MicrobialPhysiologyResearch Group, BiosphereSciencesDivision,King's CollegeLondon, Campden Hill Road, London W8 7AH, UK 2 ChemistryDepartment, King's CollegeLondon, London,UK Received March 6, 1991/AcceptedAugust 6, 1991
Abstract. The toxicity and binding of aluminium to Escherichia coli has been studied. Inhibition of growth by aluminium nitrate was markedly dependent on pH; growth in medium buffered to pH 5.4 was more sensitive to 0.9 mM or 2.25 mM aluminium than was growth at pH 6.6-6.8. In medium buffered with 2-(N-morpholino)ethanesulphonic acid (MES), aluminium toxicity was enhanced by omission of iron from the medium or by use of exponential phase starter cultures. Analysis of bound aluminium by atomic absorption spectroscopy showed that aluminium was bound intracellularly at one type of site with a Km of 0.4 mM and a capacity of 0.13 tool (g dry wt)- 1. In contrast, binding of aluminium at the cell surface occurred at two or more sites with evidence of cooperativity. Addition of aluminium nitrate to a weakly buffered cell suspension caused acidification of the medium attributable to displacement of protons from cell surfaces by metal cations. It is concluded that aluminium toxicity is related to pH-dependent speciation [with AI(H20)6a+ probably being the active species] and chelation of aluminium in the medium. Aluminium transport to intracellular binding sites may involve Fe(III) transport pathways. Key words" Aluminium and bacteria - Metal speciation - Iron transport - Biosorption of metals -- Metalmicrobe interactions - Escherichia coli
Aluminium is one of the most abundant elements in the earth's crust, but has no established biological function. Its potential toxicity to biological systems is ameliorated by its existence as oxides and complex aluminosilicates and by its insolubility in aqueous media at neutral pH. However, aluminium may be leached from soils by acid rain, washed into lakes and rivers and reach toxic levels. Toxicity to microbes may be manifest, for example, as Offprint requests to: R. K. Poole
inhibition of microbial soil activity (Firestone et al. 1983; Thompson and Medve 1984) and inhibition of photosynthesis and nitrogen fixation in cyanobacteria (Pettersson et al. 1985a). There is currently grave concern over the toxicity to Man of aluminium in drinking water and the implication of aluminium from various domestic sources as the aetiological agent in Alzheimer's disease, dialysis encephalopathy syndrome and related neurological disorders (for references see Guy et al. 1990). In aqueous solutions, AI(III) forms an octahedral hexahydrate, AI(H20)63+ (A13+) at acid pH. Successive pH-dependent deprotonations of the aluminium-bound water yield AI(I-IzO)5(OH) z +, AI(H20)g(OH) +, AI(OH)3 and AI(OH)g (Martin 1986). At neutral pH, the dominant form is the insoluble precipitate AI(OH)3. In view of this speciation (Hughes and Poole 1991) it is imperative to consider the effects of pH and the choice of buffer on aluminium toxicity. The ionic radius of A13+ (67.5 pm) is close to those of Fe 3+ (78.5 pro) and Ga 3+ (76.0 pm), although formation constants for the reaction of aluminium with the ligands EDTA, transferrin and citrate are notably lower than those for these metals (Hughes and Poole 1989). The similarity between these metals suggests that aluminium might interact with microorganisms by competition with iron. Indeed, certain fungal siderophores transport Fe(III), AI(III) or Ga(III) (Raymond and Carrano 1978) although in the cyanobacterium Anacystis eylindrica, aluminium transport occurs by passive diffusion (Petersson et al. 1986). Unlike iron, however, aluminium is bound exclusively by oxygen ligands. The toxicity of aluminium often appears attributable to slow ligand exchange rates. For example, the toxicity of aluminium often results from inhibition of Mg2+-dependent enzymes because ligand exchange is about 105 times slower for A13+ than for Mg 2+ (Martin 1986). To further our understanding of aluminium toxicity and binding, we have begun a study of aluminium interactions with "model" microbial systems and report here factors affecting the toxicity of aluminium to Escherichia coli, an organism with well-studied metal transport sys-
508 terns, especially for Fe(III). The results suggest c o m p e tition o f A13 + a n d F e 3 + a n d i n d i c a t e c o m p l e x b i n d i n g o f A13 + to cell surfaces.
E
Methods
c,
Organism and growth conditions
r o e..
Escherichia eoli RPI (= AN2572, prototroph) was grown in a medium that contained (g. 1-1): Tris buffer, 4.85; NH4C1, 1.0; glycerol, 50; KCI, 1.0; CaC12 - 2H20, 0.01 ; K2SO,~, 0.87. A trace element solution (Poole and Haddock 1974; 10ml-l-1), flglycerophosphate (16.5%, w/v; 10 ml. 1-1) and MgCI2 - 6H20 (20% w/v; 1 ml - 1-1) were autoclaved separately and added aseptically to the sterile medium, (pH 7.4). For growth at lower pH values, the Tris buffer was replaced by succinic acid (pH 4.8 to 6) or 2-(Nmorpholino)ethanesulphonicacid (MES; pH 5.4 to 6.8) and the pH adjusted to the required value with NaOH. The buffers used (MES and succinate) were chosen for their appropriate pK values and their low binding affinities for metals in general, in preference to citrate or phosphate, for example (Hughes and Poole 1991). For iron-deficient medium, FeC13 was omitted from the trace element solution. Where the iron content of the medium required supplementation, a 6 mM solution of FeC13 in 12 mM EDTA was adjusted to the same pH as the medium and added as a filtersterilized solution. Aluminium was freshly prepared as a 0.2 M solution of Al(NO3)3 in distilled water which was filter-sterilized or autoclaved with similar results. Cultures were grown in 20 ml medium in 250 ml conical flasks shaken at 200 rpm and 37~ inoculated at 10% (v/v). Growth was monitored by measuring apparent absorbance (optical density) at 600 nm (after appropriate dilution to give OD6oo < 0.5) in cuvettes of 1 cm path length using a Pye Unicam SP6-450 spectrophotometer (Cambridge, UK).
Analysis of aluminium in cells, media and washes Cells were harvested by centrifugation from overnight cultures (50 ml), washed in 100 ml medium, and the pellets suspended in MES buffer (pH 6) to a volume of 10 ml. To 9 ml portions of aluminium solutions prepared in the same buffer or to 9 ml of buffer only (controls) were added, in glass tubes, 1 ml aliquots of the cell suspension. At each aluminium concentration, a control was included without cells to investigate loss of the metal to the tube surface. After incubation at 37~ for 3 - 4 h, cells were pelleted by centrifugation, then washed successively in 4 ml buffer and 4 ml 0.1 M HNO3. All supernatant liquids were kept separately, acidified with HC1 (1% of final vol.) and made up to known volumes with distilled water. For tubes that contained no cells, the entire volume was transferred to 25 ml volumetric flasks, acidified as above and made up to volume. Cell pellets were digested by adding 100 gl conc. H2SO4, heating to 120~ and then adding 50 ~tl aliquots of conc. HNO3 until dissolution was complete. The HNO3 was then boiled off and the digests transferred to 5 ml volumetric flasks. The tubes were washed thorougly with distilled water and the washings pooled with the final volume. Aluminium assays were performed using a Perkin-Elmer 2380 atomic absorption spectrophotometer (PerkinElmer, Beaconsfield, UK) on a N20/acetylene flame at the following settings: 2, 309.3 nm; lamp current, 9 mA; spectral band width, 0.7 nm. The standards were prepared by dilution of a commercial standard. Dry weights of cell suspensions used in aluminiumbinding experiments were determined in duplicate by filtering a diluted suspension through preweighed 50 mm filters (0.45 gm pore size) and drying to constant weight at 105~C.
1 0.6
s o r t~
0.3
~~ ~
t~
> 108.8 (e.g. 10 is at 1 m M A1, pH 6), toxicity always increased at lower p H values, even though Al13 is more than an order of magnitude more toxic than A13 + (Parker et al. 1989). We thus identify a mononuclear species,
% Recovery
Aluminium in control tubes (lag)
% Recovery compared to controls
44 51 48 54 55 52
1800 1350 425 155 55 ~20
88 92 82 85 73 84
probably A1a+, as the toxic species. Pettersson et al. (1985 a) have claimed that aluminium toxicity to another prokaryote, Anabaena cylindrica, at high p H was due to mononuclear hydroxy-A1 species (AI[OH]~] -")+, where n = 1 to 4 and coordinated water has been omitted), but the solutions were likely to have contained A113. Succinate buffer appears more protective than MES, suggesting that the former has a higher affinity for A13 § than does MES; indeed, succinic (as well as citric and malic) acid are effective in protection against acute aluminium toxicity in Man (Llobet et al. 1987; Domingo et al. 1988). The proposal that the A13§ ion is the major toxic species is supported by experiments that suggest interaction of the transport systems for A13+ and Fe 3+. Iron(III), AI(III) [and Ga(III)] have similar effective ionic radii (see Hughes and Poole 1989) and each forms complexes with siderophores (Tait 1975; Theil et al. 1983), transferrin and citrate. Although the small size of the aluminium cation prevents effective binding of large ligands, aluminium is known to be transported by certain iron uptake pathways. For example A13+- (or G a 3+-) substituted ferrichrome is transported by the fungus Ustilago (Emery and Hoffer 1980). In the present study, aluminium toxicity was exacerbated by iron deficiency, perhaps allowing more effective competition of aluminium for iron-binding sites.
512
References Davis WB, McCauley MJ, Byers BR (1971) Iron requirements and aluminium sensitivity of an hydroxamic acid-requiring strain of Bacillus megaterium. J Bacteriol 105 : 589 - 594 Domingo JL, Gomez M, Llobet JM, Corbella JM (1988) Comparative effects of several chelating agents on the toxicity, distribution and excretion of aluminium. Hum Toxicol 7 : 2 5 9 - 262 Emery T, Hoffer PB (1980) Siderophore-mediated mechanism of gallium uptake demonstrated in the microorganism Ustilago sphaerogena. J Nucl Med 21:935--939 Firestone MK, Killham R, McColl JG (1983) Fungal toxicity of mobilized soil aluminium and manganese. Appl Environ Microbiol 46:758 - 761 Guy SP, Seabright PJ, Day JP, Itzhak RF (1990) Uptake of aluminium by human neuroblastoma cells. Biochem Soc Trans 18 : 3 9 2 - 393 Hughes MN, Poole RK (1989) Metals and microorganisms. Chapman and Hall, London Hughes MN, Poole RK (1991) Metal speciation and microbial growth - the hard (and soft) facts. J Gen Mierobio1137:725734 Kinraide TB, Parker DR (1989) Assessing the phytotoxicity of mononuclear hydroxy-aluminium.Plant Cell Environ 12: 4 7 9 487 Lawford HG, Garland PB (1972) Proton translocation coupled to quinone reduction by reduced nicotinamide-adenine dinucleotide in rat liver and ox heart mitochondria. Biochem J 130:1029 -- 1044 Llobet JM, Domingo JL, Gomez M, Tomas JM, Corbella J (1987) Acute toxicity studies of aluminium compounds: antidotal efficacy of several chelating agents. Pharmacol Toxicol 60:280-283 Macdonald TL, Martin RB (1988) Aluminium ion in biological systems. Trends Biochem Sci 13 : 15-19 Martin RB (1986) The chemistry of aluminium as related to biology and medicine. Clin Chem 32:1797-1806
Parker DR, Kinraide TB, Zelazny LW (1988) Aluminium speciation and phytotoxicity in dilute hydroxy-aluminium solutions. Soil Sci Soc Am J 52:438-444 Parker DR, Kinraide TB, Zelazny LW (1989) On the phytotoxicity of polynuclear hydroxy-aluminium complexes. Soil Sci Am J 53 : 789-- 796 Pettersson A, H/illbom L, Bergman B (1985a) Physiological and structural responses of the cyanobacterium Anabaena cylindrica to aluminium. Physiol Plant 63 : 153 -- 158 Pettersson A, Kunst L, Bergman B, Roomans GM (1985 b) Accumulation of aluminium by Anabaena cylindrica into polyphosphate granules and cell walls: an X-ray energy-dispersive microanalysis study. J Gen Microbiol 131:2545-2548 Pettersson A, Hallbom L, Bergman B (1986) Aluminium uptake by Anabaena cylindrica. J Gen Microbiol 132:1771 -- 1774 Poole RK, Haddock BA (1974) Energy-linkedreduction ofnicotinamide-adenine dinucleotide in membranes derived from normal and various respiratory-deficient mutant strains of Escherichia coil Biochem J 144: 7 7 - 85 Raymond KN, Carrano CJ (1978) Coordination chemistry and microbial iron transport. Acc Chem Res 12:183 - 190 Schwyn B, Neilands JB (1987) Universal chemical assay for the detection and determination of siderophores. Anal Biochem 160:47-56 Tait HG (1975) The identificationand biosynthesis of siderochromes formed by Micrococcus denitrificans. Biochem J 146:191 - 2 0 4 Theil EC, Eichhorn GL, Marzilli LG (1983) Iron-binding proteins without cofactors or sulfur clusters. Advances in Inorganic Biochemistry, vol. 5. Elsevier, New York Thompson GW, Medve RJ (1984) Effects of aluminiumand manganese on the growth of etomycorrhizal fungi. Appl Environ Microbiol 48 : 556-- 560 Van Haecht JL, Bolipombo M, Rouxhet PG (1985) Immobilization of Saccharomyces cerevisiae by adhesion: treatment of the cells by A1 ions. Biotechnol Bioeng 27:217--224
