Some Cellular Effects of Al3+
Introduction
As outlined in Aluminium1, it was found that patients on dialysis for kidney failure tended to develop Dialysis Encephalopathy Syndrome (DES), and Dialysis Osteomalacia (DOM). It was further established, unequivocally, that the culprit was an excessive concentration of Al3+ in the dialysis buffer, and hence in the blood. The former syndrome (DES), which in several respects resembles Alzheimer's disease, seemed to result from accumulation of aluminium in the brain; the latter (DOM) to result from competition between Al3+ and Ca2+ at some stage in the bone-forming process.
If excess serum Al3+ can cause either problem, it seems likely that it could, in some individual cases cause both. Searching the medical literature soon found a report of the "Impact of Osteoporosis on the Risk of Dementia in Almost 60,000 Patients Followed in General Practices in Germany" [19] The odd title seems to ascribe the dementia to an effect of the osteoporosis; but the impact is presumably only on the risk of dementia; i.e. it is a merely a correlation.
How, then, does aluminium act in metabolism, and in particular in dementia? Is Al3+ a cause, or a coincidence?
Al3+ ion Leads to Oxidative Damage.
It seems generally accepted that the presence of excess Al3+ can cause oxidative damage to cells, probably by generating the superoxide radical (O2∙-), which itself gives rise to other reactive species such as peroxide, hydroxyl radical, and peroxynitrite. Several mechanism have been suggested: [a] mitochondria can release oxygen radicals, by incomplete reduction of oxygen to water; [b] Al3+ could displace redox-active Fe ions from their complexes; [c] mitochondria are perhaps damaged by sequestration of phosphate as Al3PO4; and [d] by direct formation of superoxide from water molecules co-ordinately bound to the Al3+ atom. Let us consider these in turn.
[a] Mitochondria in ischaemic tissue that cannot get sufficient oxygen can develop an excessively reduced respiratory chain, with components at a sufficiently reducing potential to pass a single electron to an O2 molecule. In those circumstances the highly reactive superoxide anion is produced which can go on to react with nitric oxide, phospholipids and other components of the cell [1]. (The transfer of 1 electron to the O2 molecule generates superoxide radical (O2∙-) while the transfer of one electron to the peroxide molecule generates water and the hydroxyl radical (OH∙).)
[b] The Al3+ and the Fe3+ ions have the same high charge and similar ionic radius (57 and 64 pm respectively). Both Al3+ and Fe3+ ions bind tightly to transferrin in the blood, but the binding of Fe3+ ions is very much tighter (logKeq = 22.5 & 21.4, for Fe3+ binding to C and N sites respectively, while the binding of Al is 10 orders of magnitude less tight [2]. In the presence of Fe3+ ions, transferring does not carry Al3+, but it is not normally saturated with iron. The complexes are carried round the body in the blood plasma, and into iron-requiring cells by endocytosis. The acidic lumen of the endocytotic vesicle releases the trivalent ions. Al3+ can then compete with Fe3+ for the many compounds in the cell that bind Al3+ more tightly than Fe3+ (such as citrate, oxalate, phosphate, acetate)[3]. There are several metabolites (such as ascorbate) that can reduce Fe3+ to Fe2+. The latter, by the Fenton reaction, can generate the destructive hydroxyl radical and tip the cell into the intrinsic (or mitochondrial) pathway of apoptosis. [4].
[c] ATP (and ADP) mostly exits in the cytoplasm complexed with Mg2+ [5]. It is said that Al3+ inhibits hexokinase [6], as the Al3+ salt of ATP is not a substrate. If that were true for transport into the mitochondria of ADP and Pi, the mitochondria in a cell containing Al3+ might enter the hyper-reduced state called state 4., and emit superoxide radicals [7].
[d] The idea, ventured by Exley and Lopez and tested in several theoretical studies [8, 9, 10], that Al3+ can generate bound superoxide directly from bound water, may be correct, but may be irrelevant. It is clear that differential binding of Al3+ to the reduced component of a redox couple would make the bound couple appear more easily reduced than the free couple. But in that case, the bound superoxide would not be as strongly reducing as free superoxide.
Al3+ ion Displaces Mg2+ from ATP, Pi, etc..
As mentioned in [c] above, Al3+ can displace Mg2+ from ATP (and possibly ADP and Pi), under the heading of damage by oxygen radicals. Inhibition of hexokinase, and other reactions involving ATP, could have many wide-reaching effects on the metabolism of the cell including loss of ionic gradients across the cell membrane, and collapse of membrane potential.
Al3+ ion Displaces Ca2+ in the formation of bone.
It is clear, since the discovery of dialysis osteomalacia (DOM), that excess Al3+ in blood causes softening of bones.
It is also clear that Al3+ ion gets into bone, as half of the total body content of Al in healthy humans is found in the skeleton (and 25% in the lungs) [6]. It therefore seems clear that Al3+ can compete with and replace Ca2+ at some stage (or stages) in the bone-making process, summarised in the equation below:
10Ca2+ + 6H2PO4- + 2H2O ↔︎ Ca10(H2PO4)6(OH)2 + 14H+
Al3+ ion (probably) binds tightly to nucleophilic oxygen atoms in Nucleic acids and phosphorylated proteins.
The leather-tanning process exploits the tendency of Al3+ to form crosslinks between collagen molecules: RCOO- ...Al3+.....OH-.... Al3+.....O-OCR. And the very tight binding of Al3+ to nucleophilic oxygen atoms in ligands such as orthophosphate, acetate, citrate, and hydroxyl ions make it highly credible that Al3+ would bind to such groups in proteins and nucleic acids. Ganrot's comprehensive (but early) review of "Metabolism and possible health effects of aluminum" [6] suggested that proteins (in general) bind Al3+. He believed that the tight binding of Al3+ to proteins and nucleic acids would be irreversible and that the ageing human would gradually fill up with aluminium. That has proved very hard to demonstrate. Brain tissue from donors with neurodegenerative disease contain significantly more Al than control brains; but these control brains show no correlation between Al content and age [11].
Famously, the amyloid beta protein (Aβ) of extracellular presenile plaques in Alzheimer's Disease, was said to bind Al3+, though how, where, how tightly and how many are all question that still need clarifying.
Amyloid, so called as it takes up iodine stain 'like starch', has nothing to do with starch. It represents a misfolded form of any of some 20 different proteins, is characteristic of ageing, is insoluble, and may cause organ failure in a rather passive way [4]. The amyloid of dementia is the result of misfolded peptides (Aβ40, Aβ42 ) derived from the Amyloid Precursor Protein, APP. More about APP and Aβ in a subsequent post.
The other characteristic feature of neurodegenerative dementia is the presence of hyper-phosphorylated Tau-protein in 'neurofibrillary tangles' (NTFs). It is suggested that the phosphorylation of tau causes it to dissociate from intracellular microtubules and migrate to form these intracellular tangles [12]. Both the fibrillar amyloid plaques of aggregated Aβ42 peptide, and the neurofibrillary tangles of hyper-phosphorylated Tau-protein are primarily β-pleated-sheet structures [13].
However, it is not known if Al3+ binding has a causative rôle in plaque formation; for example, by favouring β-sheet structure over 𝛼-helix, though there have been a number of attempts to establish such an effect [14]. Miller et al. detected in amyloid plaques the amide-1 absorbance of β-sheet and deposits of Cu and Zn, but no Al [15]. Indeed, two recent, deep and thorough reviews on the structure of "Amyloid beta" [16, 17] make no mention of Al3+, which seems a strange omission after 40 years of the aluminium hypothesis of dementia.
A recent paper by the Exley group reported Al3+ deposits in the same cell as neurofibrillary tangles, but in a different location [18].
References.
[1] Indo et al. (2015) J Clin Biochem Nutr.; 56: 1–7.)
[2] Ott, D.B. et al. (2019) Metallomics 11(5):968-981; doi: 10.1039/c8mt00308d.
[3] Vekeman et al. (2021) Nanomaterials (Basel); 11: 1763; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8308151/
[4] Wikipedia
[5] Kiss, T.; Sovago, I.; Martin, R. B. Al3+ Binding by Adenosine 5′-Phosphates: AMP, ADP, and ATP. Inorg. Chem. 1991, 30, 2130− 2132.
[6] Ganrot, P.O., (1986) "Metabolism and possible health effects of aluminium", Environmental Health Perspectives 65, 363-441
[7] J Bioenerg Biomembr. (1999);31(4):347-66. doi: 10.1023/a:1005427919188.
[8] Exley, C., (1992) Free Radic. Biol. Med. 13: 79–81
[9] Mujika et al. (2014) https://doi.org/10.5936/csbj.201403002
[10] Lopez (2022) Free Radical Biology and Medicine, Volume 179, 1 February 2022, Pages 200-207.
[11] Exley C, Clarkson E. (2020) Sci Rep. 10(1):7770. doi: 10.1038/s41598-020-64734-6.
[12] Siddhartha Mondragón-Rodríguez et al. (2020) Hindi Neural Plasticity, Volume 2020 | Article ID 2960343 |
[13] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5589967/
[14] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7582331/ Proc Natl Acad Sci U S A. 1994 Nov 8; 91(23): 11232–11235
[15] Miller et al., 2006, J. Struct. Biol., 155 (2006), pp. 30-37.
https://www.sciencedirect.com/science/article/pii/S1878535215001914
[16] Ow & Dunstan (2014) Protein Sci. 2014 Oct; 23(10): 1315–1331.
[17] Chen et al. (2017) PMC5589967]
[18] Mold and Exley (2021) J Alzheimers Dis Rep v.5(1); 2021 PMC8150251
[19] Kostev, Hadji & Jacob (2018) J. Alzheimers Dis. 65(2):401-407.
