[1]
16w
Amyloid-b (Ab) deposition in brain is one of the histological hallmarks of Alzheimer disease (AD) [5].
[2]
111w
Nevertheless, the magnitude of Ab deposition may vary considerably from case to case and, in rare instances, deposits may even be missing [48,125]. Conversely, considerable amounts of Ab may be found in the brains of very old subjects who never exhibited clinical signs of dementia before death [71]. Currently, neuro-toxic effects of 0306-9877/$ -see front matter c 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy. 2005.11.016 Ab are conceived to be pivotal factors driving both the mental decline and the neuronal death in AD [4,106]. If so, however, it remains surprising why Ab cyto-toxicity does not yield similar effects in aged subjects with high loads of cerebral Ab deposits but without dementia.
[3]
170w
The 4.2 kDa large, self-aggregating, b-sheated and almost non-digestible Ab molecule has been shown to originate from a larger trans-membrane amyloid precursor (glyco-)protein (APP) most abundantly expressed in brain on the neuronal (synaptic) membrane [5,31]. Ab was shown to be a product of the aberrant cleavage of APP by receptor-regulated proteases such as b-secretase(s) BACE (b amyloid cleaving enzyme) [126] followed by c-secretase(s) processing [83] of the C-terminal fragment. In course of the normal turn-over, APP is processed by an a-secretase within the Ab sequence (amino acids (aa) 597-638 of APP 695 ) at lysine 16 (Lys 16) of Ab [49,142], thus resulting in two N-and C-terminal non-amyloidogenic fragments which can be further processed without problems. It has been widely accepted to date that aberrant cleavage of APP is the main source of Ab and its plaque-like deposition in brain tissue of the aged and even more in AD [5]. However, up to now the reason(s) for the abnormal APP cleavage both in ageing and AD has (have) remained elusive.
[4]
81w
In this particular context, the possible significance for APP processing of non-enzymatic glycation (NEG) or e-glycation, a common molecular ''accident'' that may happen to peptides and proteins in the presence of reducing sugars [13], leading to advanced glycation end-products (AGEs) [17], has but infrequently been addressed scientifically up to now (e.g., [16,22,138]). It could be shown that APP undergoes post-translational Nand O-glycation [104,132] which are enzymatically controlled. However, currently little information is available about the involvement of APP in nonenzymatic e-glycation.
[5]
66w
In view of the emerging knowledge about the widespread occurrence of NEG in various tissues [50,77] including the central nervous system [53,81,119,128] I want to propose the following hypothesis based on a review of recent results and increasing evidence for a putative role of the e-glycation and AGEs in ageing and disease, with special emphasis on APP processing and Ab production in brain ageing and AD:
[6]
36w
Hypothesis. e-Glycation of the amyloid precursor protein (APP) is likely a significant factor that drives Ab production and deposition in normal ageing and AD, being enhanced in the latter by failure of the cellular energy production.
[7]
14w
The non-enzymatic browning (Maillard) reaction results in the formation of advanced glycation end-products (AGE)
[8]
91w
Non-enzymatic e-glycation (NEG) is a kind of a potentially deleterious molecular accident that can happen to amines, i.e., peptides and proteins, through the binding of reducing sugars such as glucose, lactose, fructose, and maltose in e position, which means to lysines, with a preference for N eamino groups of the amino acid (aa) residues [3,14,9,115]. The numerous targets of NEG in vivo include collagen [45,105], low-density lipoproteins (LDL) [115], apolipoproteins (apoE!) [12,24], proteins of the humoral immune defense like IgG [9], and many other proteins [27] such as large cell-surface proteins.
[9]
65w
Via subsequent molecular re-arrangement (Amadori re-arrangement) [35], i.e., complete irreversible oxygen-dependent isomerisation of Nsubstituted aldosylamines to 1-amino-1-deoxy-2ketoses, non-soluble sugar-amine isomeres (ketoamines, Amadori compounds) [26,54,89] are generated. Further processing of the latter in course of the Maillard reaction eventually results in the formation of a multitude of yellow-brown fluorescing products (''brown Maillard products'') named advanced glyc(osyl)ation end products (AGEs) [14,18,35,45,61,69,79,124], which can hardly [23,117] become digested.
[1]
79w
APP, as a large trans-membraneous cell-suface glyco-protein, can be postulated to be a candidate for NEG, perhaps subsequently being subjected to cross-linking and AGE production. The likelihood of molecular cross-linking is highest in locations with the highest densities of APP molecules. This is true for the neuronal membrane [32] with highest APP densities at the synaptic membrane [103]. The deposition of neuronal amyloid-b in form of non-vascular amyloid plaques was thought to occur in the surrounding of synapses [73].
[2]
95w
The likelihood of NEG and its sequelae further depends on the magnitude of the metabolic capacity and/or the turn-over of a neuron as well as on the concentration of free sugars in the intra-and extra-cellular space. NEG increases with (1) increasing concentration of open-chain structured sugar in sugar-protein systems [54] and (2) decreasing pH to acidic levels because the Amadori rearrangement is acid catalysed [35]. Thus, the increased glucose levels in cortex [72] as a result of reduced neuronal glucose utilization in ageing and AD [38,39] might further favour NEG of cellular proteins including APP.
[3]
69w
Severe damage by NEG to APP might induce proteolytic degradation of the severely functionally hampered molecule. It has been proposed that e-glycated and cross-linked proteins become increasingly affected by catalytic enzymes [9,129]. Inactivation of APP by NEG, cross-linking and proteolysis would activate the APP coding PAD-gene on chromosome 21 [5,6] to regionally over-express for substitution new APP [58]. The latter could again become e-glycated thus inducing a vicious circle.
[4]
133w
Ubiquitin (Ub) marks short-lived proteins for degradation by a labelling mechanism using lysine e-amino groups on the target protein [37,94]. Imagine that in the course of NEG of APP sugar chains would become attached to lysines normally needed for either Ub labelling or a-secretase cleavage or both (e.g., K16). This might hinder the labelling procedure or block the site of action of the asecretase, thus invoceing other secretase(s) (e.g., BACE) to cleave at an alternative lysine which could lead to the formation of abnormal cleavage products with the generation of amyloidogenic aa fragments. Alternatively, a sugar chain becoming linked with a Lys other than K16 might exert a marker effect on the host molecule for degradation at that site. A similar consideration has previously been proposed for e-glycated collagen in diabetes cf. [9].
[5]
117w
Up to now, these and other challenging questions regarding the occurrence of NEG in APP and its putative sequelae have not been adequately addressed in ageing and AD research. However, it could be shown that in prion protein (PrP C ) two lysines at the amino terminal site become involved on generation of the mature molecule in labelling the site of signal cleavage [110, p. 179]. Moreover, glycation in PrP C on asparagines (Asn 181,197) that was demonstrated in plants to be apt to lead to the formation of Amadori compounds [26] similar to eglycation, can significantly alter neuronal targeting and the deposition in prion disease of PrP Sc , probably due to related conformational modification [21].
[6]
104w
Of note, NEG can be prevented by antioxidants and reducing substances like ascorbic acid (vitamin C) [20,27,28,99,100]. The anti-NEG effect of vitamin C becomes already empirically evident by the fact that vitamin C-rich citrus fruits (lemons, oranges, grape fruits), unlike other fruits (apples, pears etc.), never show Maillard's browning. However, in opposite to vitamin C, its degradation products such as L-dehydro-ascorbic acid or formyl threosyl pyrrole [84,96] favour NEG, AGE production and molecular cross-linking [84,96,100]. Beneficial effects were also found for green tea extract. The latter in combination with vitamins C and E, was shown to be most effective in preventing AGE production [99].
[7]
51w
Hence, low-dose aethanol (as mentioned before) and high-dose vitamin C and E intake, together with green tea extract, might be effective means to attenuate detrimental effects of NEG and products of the advanced Maillard reaction, therewith slowing the ageing of protein molecules and tissues and, perhaps, influence the progression of AD.
[8]
82w
All in all, the aim of the present review was to summarize the compelling evidence for a significance of NEG and the Maillard reaction in ageing and AD and, in particular, to emphasize the putative influence of NEG on the processing of APP and the generation and deposition of Ab. The ''NEG hypothesis of APP processing in ageing and AD'' merits further study, because it might be relevant for the development of alternative means or additional preventive and therapeutic strategies in AD.
[1]
78w
AGE production is likely enhanced by Ca 2+ , which is thought to play a pivotal role in the pathogenesis of age-dependent and AD-related neural changes [55,74,102]. Ca 2+ antagonists were found to slow down significantly the formation of AGE [112,134]. By enhancing AGE production, Ca 2+ may, for instance, cause increased signalling through AGE specific receptors (see below), thus favouring oxidative stress, which is thought to foster neuronal pathology and cognitive decline in ageing and AD [33,43,120,134].
[2]
49w
Moreover, oxygen free radicals are also produced during the Maillard reaction, e.g., in course of the decomposition of Amadori compounds that leads to the generation of highly reactive intermediates like 3-deoxyglucosone [26], and oxygen free radicals [61], the former acting as crosslinkers [45], the latter inducing oxidative stress [111].
[3]
64w
NEG and its sequelae may induce changes at the cellular and nuclear level, favouring ageing and disease NEG and its sequelae (Maillard intermediates and AGEs) are apt to induce a number of unfavourable intra-and extra-cellular changes, including crosslinking [8,100], oxidative stress (as mentioned before), conformational protein change [20], aberrant intra-cellular [133,141] and nuclear signalling, as well as mutagenesis [63] and change of mitogenesis [40].
[4]
145w
These changes may form the basis of or play a key-role in various somatic diseases such as diabetes [119,135] where peripheral nerve proteins are highly glycated [128], and nervous system disorders related to ageing [18,20,27,29,30,52,79,124] with and without mental decline [24,81,93,107,111,119,138,139]. NEG at critical sites may change the function of cell-surface receptors such as the agonist binding capacity of ion channels (e.g., through glycation aside the aspartate D152 of the extra-cellular domain of the nAChR a-subunit [116]) or affect nucleic acids. [62]. Oxygen free radicals as well as possibly signalling of AGEs to the nucleus through their specific receptor(s) (RAGE) (see below) may induce DNA strand scission [51,80] or cause altered protein transcription, e.g., by stimulation of nuclear factor NF-jB [137,59,66,42,113,121]. DNA replication may become inhibited by intermediate products of the Maillard reaction such as 3-deoxyglucosone which may lead to impairment of cell proliferation [108].
[5]
34w
NEG, Maillard intermediates and AGEs favour the cross-linkingofproteinmoleculesthroughe(lysine-lysine)groups [9,41,45,81,97,105].Cross-linkingofcell surfacereceptorproteinswasshowntobeapttoinitiate apoptoticcell death [19,43].Theinteractionofcross-linkingandaccumulatedAGEs [85]mayalsocause(neuronal)celldeath [120],changeofthebio-chemicaland bio-mechanical properties of cell membranes and long-livedtissueproteins [11,97],createvasculardisease [15,86,101], perturb the blood-brain barrier (BBB) [15,95],promoteendothelialtrans-cytosisofcir-culatingAb proteinintobraintissue [68]andtriggerinflammatoryreactions [130,56].Allthesechanges occurinageingand,particularly,inAD.
[6]
46w
Receptors reacting specifically with AGEs (RAGE) have been detected in various mammalian including human [70] cell types such as monocyte/macrophages and microglia [76,130], vascular endothelia [7,10,87] and smooth muscle cells [10,131], fibroblasts [40,92] as well as various neural cells, including neurons and white matter glia [70].
[7]
43w
After identification of a bovine form of RAGE, a cell surface polypeptide of 35 kDa molecular mass [87], a secreted 90% human homologue (hRAGEsec) of the same molecular weight was cloned, sequenced [70] and identified as a member of the immuno-globulin super-family [87].
[8]
23w
In addition to RAGE, further receptors for AGE such as AGE-R2, AGE-R3, macrophage scavenger receptors, and oligosaccharyl transferase 48 were subsequently identified [123].
[9]
100w
The AGE-RAGE signalling cascade involves p21(ras) [59], cdc/Rac [46] and likely MAP kinases such as p38MAPK, p44/42MAPK, ERK-1 and ERK-2 [46,59,141] facilitating both trophic and toxic effects [43,44,119]. The latter may include neuronal cell death [120], while the former promote neuritic outgrowth (in cultured neuroblastoma cells) [36,42], enhance the activity of nuclear transcription factor NF-jB (see above) [42,57], activate nuclear factor CREB (cAMP response element-binding protein) [44] and facilitate pro-inflammatory cytokine secretion [141]. The latter pertains to macrophages in which interaction of AGEs with RAGE was shown to stimulate the secretion of TNF-a, IL-1 and platelet-derived growth factor (PDGF) [56,75,129,130].
[10]
66w
NEG with production of AGEs has been regarded as a process of molecular [13] and cellular [65] ageing. Factors such as oxidant stress, molecular cross-linking, the sequelae of RAGE signalling and changes of DNA integrity (see above) extrapolate this process from the molecular to the biological (cellular and tissue) level [42][43][44]51,63,130]. This pertains to ageing (''cross-linkage'' or ''Maillard reaction theory of aging'') [8,79] and AD [34,69,88,105,111,122].
[11]
82w
NEG of cellular and tissue proteins inside and outside the CNS and the deposition of AGEs were shown to be increased in the aged brain [17] and AD [81]. In the latter, accumulation of AGEs that likely activate astro-and microglia [82,118,138,139] was demonstrated particularly in pyramidal neurons [65] and their surrounding [118], in NFT [60] and in senile plaques as well as vascular amyloid deposits [127]. AGEs were thought to contribute to the aggregation and deposition of Ab [69,127] (see also below).
[12]
97w
It is not surprising that AGEs may affect especially neurons since the latter almost exclusively use glucose as substrate for energy production so that abundant amounts of glucose are permanently available for NEG of neuronal proteins. This is, in particular, true for diabetes mellitus [77]. The formation and tissue deposition of AGEs has been shown to be particularly increased in type 2 diabetes mellitus [78], in which age-related decline of cognitive functions appears to be accelerated [93] and the risk of dementia seems to be increased [64,90,91,114], presumably (but not unequivocally) including the dementia of Alzheimer-type [34,88,122].
[13]
100w
Cerebral glucose utilization decreases with ageing and in AD, probably due to a decrease in the activity of enzymes of the pyruvat-dehydrogenase complex that facilitate the transfer of the products of anaerobic glycolysis (pyruvate/lactate) to the citrate cycle for oxidative processing [39]. It was hypothesized that the increased use of non-oxidative pathways for energy supply with accumulation of lactate (acidification) might favour NEG with AGE production [35], and cross-linking [8]. Conversely, no information is currently available in how far the ageing-and AD-related attenuation of enzymes that catalyze the cellular energy metabolism [38] might be due to or supported by NEG.
[14]
101w
Ample evidence has been provided that, in particular molecular cross-linking as favoured by NEG may drive amyloid formation. The occurrence of NEG-related cross-linkage of b-sheeted proteins was rendered likely in various amyloidoses [8,130,75,76,50,113], while AGEs were supposed to enhance the aggregation of amyloid peptides [127]. Further evidence has been provided that the receptor for AGE, RAGE, which is over-expressed in AD neurons [82,138], potentiates Ab-induced perturbation of neuronal function [2]. RAGE can act as an acceptor for Ab peptide in neural cells [42,139] and vascular endothelia [68], stimulate (pro-) inflammatory pathways [25] and induce early cyto-toxic effects [68,70,[138][139][140] (see, however, [67]).
[15]
58w
AGE immuno-reactivity has been demonstrated in NFT [111], tomb stone tangles staining much more intensely for AGE than intracellular NFT [24]. In vivo, NEG of hyper-phosphorylated s can result in the formation of s AGEs [136,137] that may favour s aggregation into PHF [60] and, like Ab, induce oxidant stress in neurons as well as cytokine release [136,137].
[16]
97w
Alcohol appears to have a protective effect against the formation of AGE in that binding of the aethanol-derived metabolite acetaldehyde to Amadori products seems to prevent progression of the latter to AGE [1]. This seems to correspond with the epidemiologic finding that low-dose alcohol intake is likely to decrease the AD risk [98]. That is because the prevention of AGE production might be, at least, one factor to attenuate the AD risk, in addition to alcohol-induced increase of network inhibition by enhancement of the ligand sensitivity of GABA A receptors [109] and other (e.g., nutritional) factors [47]).