Michael R. Wills

A long-term intravenous model of aluminum maltol toxicity in rabbits: Tissue distribution, hepatic, renal, and neuronal cytoskeletal changes associated with systemic exposure☆

We studied the toxicity of an intravenously injected, water-soluble aluminum complex (aluminum maltol) in 20 young adult male New Zealand white rabbits over a period of 8 to 30 weeks. Sixteen rabbits injected with aluminum-free maltol and 15 untreated rabbits served as controls. Rabbits were injected three times per week with 75 mumol of aluminum maltol per injection, or a molar equivalent amount of maltol alone, through an indwelling jugular catheter. Liver contained the highest concentrations of aluminum among the aluminum maltol-treated rabbits, and aluminum accumulation was correlated with the appearance of periportal multinucleated giant cells in 13 of 20 rabbits. These cells stained positively for aluminum when a fluorescent (Morin) stain was applied to tissue from rabbits with a high concentration of aluminum in the liver. Proximal renal tubular necrosis or atrophy was found in 15 of 20 aluminum maltol-treated rabbits but not in maltol-treated and untreated controls. Renal tubules in rabbits with acute proximal renal necrosis stained positively for aluminum. Neurofibrillary tangles, immunoreactive with a monoclonal antibody to the 200-kDa subunit of neurofibrillary protein, were observed in the oculomotor nucleus of 3 aluminum maltol-treated rabbits (treated for 12, 20, and 29 weeks), but in none of the two groups of controls. These tangles were present in 3 of 10 aluminum-treated rabbits in which the nucleus was located. None of the 17 animals in both control groups in which the nucleus was found demonstrated tangles. A slight increase in brain tissue aluminum concentration was confirmed by an electrothermal atomic absorption spectrophotometric method. There were no specific findings in heart or lung tissue from aluminum-treated rabbits, although the aluminum content of these tissues was 10 to 20 times greater than control values. This model should be useful for investigating the effects of systemic exposure to high concentrations of solubilized aluminum.

Tau immunoreactivity associated with aluminum maltolate-induced neurofibrillary degeneration in rabbits

Intracisternal administration of aluminum maltolate to rabbits produces a marked argyrophilic neurofibrillary degeneration (NFD) which is also immunoreactive for both phosphorylated and non-phosphorylated microtubule associated protein tau. Using tissue fixation in PBF, the monoclonal antibodies Tau-2 and AT8 stain the NFD. Dephosphorylation markedly reduces the positivity of AT8. Using PLP-fixed tissue, monoclonal antibody Tau-1 also immunostains aluminum-induced NFD.

Neurofibrillary lesions in experimental aluminum-induced encephalopathy and Alzheimer's disease share immunoreactivity for amyloid precursor protein, Aβ, α1-antichymotrypsin and ubiquitin-protein conjugates

Neurofibrillary tangles of Alzheimers disease contain predominantly tau protein and to a lesser degree amyloid precursor protein (APP), A beta protein, alpha 1-antichymotrypsin (ACT) and ubiquitin. Previously we have demonstrated the presence of phosphorylated tau and neurofilament proteins in neurofibrillary degeneration (NFD) induced by aluminum (Al) maltolate in rabbits [Savory et al., Brain Res. 669 (1995) 325-329; Savory et al., Brain Res. 707 (1996) 272-281]. Using the same animal system we have now detected APP, A beta, ACT and ubiquitin-like immunoreactivities in NFD-bearing neurons, often colocalizing in the NFD. Diffuse cytoplasmic staining for APP, A beta and ubiquitin was also present in neurons without NFD from Al maltolate-treated rabbits. This study provides additional support for immunochemical similarities between Al-induced NFD in rabbits and the neurofibrillary tangles in human subjects with Alzheimers disease.

Neuronal cytoskeletal lesions induced in the CNS by intraventricular and intravenous aluminium maltol in rabbits.

The antigenicity of neuronal cytoskeletal lesions was studied immunohistochemically in adult New Zealand white rabbits after intraventricular (subacute) and intravenous (chronic) administration of a water‐soluble aluminium compound, aluminium (Al) maltol. After short‐term intraventricular administration, rabbits developed widespread neuro‐fibrillary degeneration (NFD) involving pyramidal neurons of the isocortex and allocortex, projection neurons of the diencephalon, and nerve cells of the brain stem and spinal cord. There was a predilection for motor neuron involvement and for the infratentorial portions of the neuraxis. Perikarya and proximal neurites were especially affected. Bundles of 10 nm filaments were frequently present. Three of the animals treated intravenously for 12 weeks or longer displayed NFD in the oculomotor complex and in the pyramidal neurons of the occipital isocortex.

Aluminum neurotoxicity in experimental animals

Neurotoxic effects of aluminum (AI) were recognized >100 years ago, but have only recently been studied in detail. By far, the most dramatic effect of Al is that of producing intraneuronal perikaryal neurofilamentous aggregates, which consist of phosphorylated neurofilaments. Several species have been used to demonstrate this effect, rabbit being most common; the effect also is seen in in vitro systems. Besides its role in producing neurofibrillary pathology, Al appears to modify the blood-brain barrier and exert cholinergic and noradrenergic effects. Possible mechanisms of Al neurotoxicity could be related to cell damage via free radical production, impairment of glucose metabolism, and effects on signal transduction.

Quantitative study of aluminum binding to human serum albumin and transferrin by a chelex competitive binding assay

Binding of aluminum to human serum albumin and transferrin was investigated using a competitive binding assay incorporating a cation exchange resin, chelex. Both albumin and transferrin were found to produce linear Scatchard plots of aluminum binding data over the aluminum and protein concentration ranges found in humans. Binding constants measured for albumin and transferrin were 1.96 and 0.515 microM, respectively.

Quantitative image analysis of temporal changes in tau and neurofilament proteins during the course of acute experimental neurofibrillary degeneration; non-phosphorylated epitopes precede phosphorylation

Perturbation of the neuronal cytoskeleton represents an integral feature of neurofibrillary tangles which are characteristic neuropathological findings seen in Alzheimers disease. Microtubule associated protein tau (tau) is considered to be the major component of these lesions although neurofilament proteins also are present. The present study explores the formation of intraneuronal tau and neurofilament protein aggregates using intracisternal administration of aluminum maltolate to rabbits. The time course of the formation of these aggregates and subsequent phosphorylation have been investigated by immunohistochemical methods using a panel of monoclonal antibodies, with quantitation of the staining by image analysis. Neurofilament proteins begin to aggregate by day 1 following aluminum maltolate injection on day 0. Increases in non-phosphorylated neurofilament proteins are observed first, with phosphorylated epitopes being recognized by day 3. Tau follows a similar pattern in that non-phosphorylated epitopes appear to precede phosphorylation. The monoclonal antibody Alz-50 which recognizes a phosphorylation-independent epitope of tau in Alzheimers disease paired helical filaments, demonstrates positivity in the aluminum maltolate-treated rabbits by day 3. Other tau monoclonal antibodies which recognize phosphorylated tau in paired helical filaments (AT8 and PHF-1) show positive immunostaining on days 6-8. These results indicate that intraneuronal aggregation of cytoskeletal proteins can be initiated by factors other than phosphorylation. However, phosphorylation occurring as a secondary event probably contributes to stabilization of the aggregates.

Aluminium Toxicity in Chronic Renal Insufficiency

Summary Aluminium is a ubiquitous element in the environment and has been demonstrated to be toxic, especially in individuals with impaired renal function. Not much is known about the biochemistry of aluminium and the mechanisms of its toxic effects. Most of the interest in aluminium has been in the clinical setting of the haemodialysis unit. Here aluminium toxicity occurs due to contamination of dialysis solutions, and treatment of the patients with aluminium-containing phosphate binding gels. Aluminium has been shown to be the major contributor to the dialysis encephalopathy syndrome and an osteomalacic component of dialysis osteodystrophy. Other clinical disturbances associated with aluminium toxicity are a microcytic anaemia and metastatic extraskeletal calcification. Aluminium overload can be treated effectively by chelation therapy with desferrioxamine and haemodialysis. Aluminium is readily transferred from the dialysate to the patients bloodstream during haemodialysis. Once transferred, the aluminium is tightly bound to non-dialysable plasma constituents. Very low concentrations of dialysate aluminium in the range of 10–15 μg/l are recommended to guard against toxic effects. Very few studies have been directed towards the separation of the various plasma species which bind aluminium. Gel filtration chromatography has been used to identify five major fractions, one of which is of low molecular weight and the others appear to be protein-aluminium complexes. Recommendations on aluminium monitoring have been published and provide ‘safe’ and toxic concentrations. Also, the frequency of monitoring has been addressed. Major problems exist with the analytical methods for measuring aluminium which result from inaccurate techniques and contamination difficulties. The most widely used analytical technique is electrothermal atomic absorption spectrometry which can provide reliable measurements in the hands of a careful analyst.

Partial reversal of aluminium-induced neurofibrillary degeneration by desferrioxamine in adult male rabbits.

J. Savory, M. M. Herman, R. T. Erasmus, J. C. Boyd and M. R. Wills (1994) Neuropathology and Applied Neurobiology 20, 31–37

Aluminium maltol–induced neurocytoskeletal changes in fetal rabbit midbrain in matrix culture

We have developed a neuronal culture system to evaluate the neurotoxic effects of aluminium maltol on fetal rabbit midbrain sections containing the oculomotor nucleus. Cultures were treated with 5, 7, 9, 11, 13 and 15 μ.mol/1 aluminium maltol or 39 and 45 umol/1 maltol (molal equivalents to 13 and 15 umol/1 aluminium maltol). Control cultures were maintained in nutrient medium alone. Silver–positive neuritic swellings and occasional perikaryal neurofibrillary tangles were observed in cultures treated with 11, 13 and 15 umol/1 aluminium maltol. The number of tangles (involved neurons) produced in aluminium maltol treated cultures were counted and compared to (untreated) controls. We observed a total of 3, 7 and 7% of involved neurons following treatment with 11, 13 and 15 μmol/1 aluminium maltol respectively, and none in the control group. By immunohistochemistry, neurofibrillary tangles were immunoreactive with MAbs to phosphorylated (SMI–31), non–phosphorylated, phosphorylation dependent (SMI–32) and phosphorylation independent (SMI–33) epitopes of the high (–H) and middle (–M) molecular weight neurofilament subunits (NF–H/M). By contrast these lesions were nonreactive with MAbs recognizing tau, MAP2 or different β–tubulin isotypes. The perikaryal tangles consisted of focal accumulations of 10 nm straight filaments by electron microscopy.