Volume 221, Issue 1
Invited Review
Free Access

The sour side of neurodegenerative disorders: the effects of protein glycation

Hugo Vicente Miranda

Enzymology group, Departamento de Quimica e Bioquimica, Centro de Quimica e Bioquimica, Faculdade de Ciencias da Universidade de Lisboa, Campo grande, Edificio C8, 1749‐016, Lisboa, Portugal

Cell and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Faculdade de Medicina da Universidade de Lisboa, Av. Prof. Egas Moniz, 1649‐028 Lisboa, Portugal

Search for more papers by this author
Tiago Fleming Outeiro

Corresponding Author

E-mail address: touteiro@fm.ul.pt

Cell and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Faculdade de Medicina da Universidade de Lisboa, Av. Prof. Egas Moniz, 1649‐028 Lisboa, Portugal

Instituto de Fisiologia, Faculdade de Medicina da Universidade de Lisboa, Av. Prof. Egas Moniz, 1649‐028 Lisboa, Portugal

Cell and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Faculdade de Medicina da Universidade de Lisboa, Av. Prof. Egas Moniz, 1649‐028 Lisboa, Portugal.Search for more papers by this author
First published: 01 April 2010
Citations: 84

No conflicts of interest were declared.

Abstract

Neurodegenerative diseases are associated with the misfolding and deposition of specific proteins, either intra‐ or extracellularly in the nervous system. Although familial mutations play an important role in protein misfolding and aggregation, the majority of cases of neurodegenerative diseases are sporadic, suggesting that other factors must contribute to the onset and progression of these disorders. Post‐translational modifications are known to influence protein structure and function. Some of these modifications might affect proteins in detrimental ways and lead to their misfolding and accumulation. Reducing sugars play important roles in modifying proteins, forming advanced glycation end‐products (AGEs) in a non‐enzymatic process named glycation. Several proteins linked to neurodegenerative diseases, such as amyloid β, tau, prions and transthyretin, were found to be glycated in patients, and this is thought to be associated with increased protein stability through the formation of crosslinks that stabilize protein aggregates. Moreover, glycation may be responsible, via the receptor for AGE (RAGE), for an increase in oxidative stress and inflammation through the formation of reactive oxygen species and the induction of NF‐κB. Therefore, it is essential to unravel the molecular mechanisms underlying protein glycation to understand their role in neurodegeneration. Here, we reviewed the role of protein glycation in the major neurodegenerative disorders and highlight the potential value of protein glycation as a biomarker or target for therapeutic intervention. Copyright © 2009 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Introduction

The increase in life expectancy observed over the last century has led to the emergence of a series of age‐related disorders that pose novel challenges to modern societies 1. Neurodegenerative disorders, including Alzheimer's, Parkinson's, amyotrophic lateral sclerosis and the prion diseases, are debilitating and so far incurable disorders that demand intensive research. These diseases are characterized by a slow and progressive loss of neuronal cells, and by the deposition of misfolded and aggregated proteins in the brain, either intra‐ or extracellularly. In Alzheimer's disease (AD), extracellular amyloid β‐peptide (Aβ) deposits (amyloid plaques) and intracellular tau protein deposits (neurofibrillary tangles) are the key pathological hallmarks 2. In Parkinson's disease (PD), cytoplasmic proteinaceous inclusions, mainly composed of the protein α‐synuclein (α‐syn), named Lewy bodies (LBs), are the pathognomonic inclusions 3. While the vast majority of cases of neurodegenerative disorders are sporadic, familial forms are also known where several mutations are associated with the amyloidogenic behaviour of proteins such as Aβ 4, α‐syn 5-7 or transthyretin 8.

Endogenous chemical agents such as oxygen and reducing sugars (oxidative and carbonyl stress) are important players in the damage of biomolecules, leading to the loss of their normal function and, ultimately, to cell dysfunction and death. Some physiological consequences of protein glycation include the development of diabetes mellitus and cardiac dysfunction 9-11, visual disorders 12, nephropathy, vascular disorders and diabetic atherosclerosis 13-15.

Glycation has pivotal clinical relevance, since it may be used as a specific biomarker for several disorders. In diabetes, for example, AGEs may be used as markers of tissue damage and may predict long‐term complications 16, 17. Non‐invasive techniques have been developed to assess the levels of AGE in the skin, such as the auto‐fluorescence reader, which rapidly measures skin auto‐fluorescence and, therefore, AGE accumulation 17. Here we review the contribution of protein glycation to protein aggregation and toxicity in the context of neurodegenerative disorders, and propose that the process may open novel avenues for therapeutic intervention.

Protein glycation—a sour reaction with sweets

In 1912, Louis Camille Maillard studied the reaction between reducing sugars and amine‐containing compounds 18, first described in vivo in either diabetic or healthy individuals 19. This post‐translational modification involves multi‐step non‐enzymatic reactions between carbonyl‐containing groups and amino groups, leading to the formation of irreversible modifications called advanced glycation end products (AGEs) 20. The main targets are biomolecules with free amine groups, such as proteins, nucleotides and also some phospholipids. Side‐chains of arginine and lysine residues, the N‐terminal amino group of proteins, and the thiol groups of cysteine residues, are the main targets of protein glycation. This process depends on several conditions, such as the concentration and reactivity of the glycation agent 21-23, the presence of catalytic factors (metals, buffer ions and oxygen), the physiological pH and temperature 24-26 and the half‐life of each protein 27.

Since the discovery of glycated haemoglobin by D‐glucose 19, which is currently used as a biomarker of diabetes disease, glycation has been vastly investigated. However, D‐glucose is the least reactive of all reducing sugars and its intracellular concentration is negligible, while reactive dicarbonyl compounds are far more reactive (Figure 1). This observation has focused attention on methylglyoxal, present in all cells and considered to be the most reactive glycation agent 28-30. Methylglyoxal is mainly formed by the non‐enzymatic β‐elimination of the phosphate group from the triose phosphates derived from glycolysis 31. An increase in dicarbonyl compounds (carbonyl stress) is often achieved under hyperglycaemia, oxidative stress or a diminished activity of their catabolic pathways. AGEs may directly affect protein structure and function and, in addition, AGEs‐modified proteins also exert cellular effects mediated by specific receptors, including the macrophage scavenger receptor, MSR type II, OST‐48, 80K‐H, galectin‐3, CD36 and the receptor for AGEs (RAGE) 32-36. RAGE belongs to the immunoglobulin superfamily of receptors, with a broad repertoire of ligands that can be generated endogenously, such as AGEs, Aβ fibrils, transthyretin amyloid fibrils, and amphoterin and proinflammatory cytokine‐like mediators of the S100/calgranulin family 37. The interaction of Aβ with RAGE induces neuronal stress and the activation of different signalling pathways, which mediate the effects of Aβ in microglia, the blood–brain barrier and neurons 38. Moreover, the engagement of RAGE by AGEs results in VCAM‐1, tissue factor and IL‐6 expression and their endothelial permeability to macromolecules 39-41.

image

General chemical structure of dicarbonyls, including the major glycation agents, methylglyoxal, glyoxal and 3‐deoxyglucosone. The electrophilic carbonyl group of an aldehyde or ketone (dicarbonyls) suffers a nucleophilic attack by the nitrogen atom of the amino‐group, forming an unstable Schiff's base. After undergoing intramolecular rearrangements, oxidative and non‐oxidative fragmentation and dehydratation reactions, advanced glycation end‐products are formed.

It may also deploy other cellular responses, including monocyte chemotaxis stimulation followed by mononuclear infiltration 42, increased angiogenesis through the production of vascular endothelial growth factor (VEGF) 43, 44 and cell proliferation 42, 45, 46. RAGE also presents a central role in inducing inflammatory responses 47, 48 via two mechanisms: activating the nuclear transcription factor‐κB (NF‐κB) by interaction with leukocytes or endothelial cells 49, 50; or directly interacting with leukocyte β2‐integrins on endothelial cells, recruiting inflammatory cells 51, 52. Carbonyl stress may then lead to an extensive non‐programmed glycation, resulting in a deleterious effect on protein structure and function and ultimately causing cellular and tissue damage, features observed in several pathological conditions and ageing.

Glycation in Alzheimer's disease

AD is the most prevalent neurodegenerative disease, affecting ∼5% of people aged 65–75 and almost 50% of people over the age of 85 53. One of the most recognized symptoms is gradual difficulty in memorizing new information, followed by confusion, disorientation and disordered thinking. Neurodegeneration occurs in several areas, such as the hippocampus, amygdala, nucleus basalis and entorhinal cortex. Most AD cases are sporadic, whereas only 6% have a known genetic origin 54. Genetic cases are, in the great majority, related to the presence of the ε4 allele of the apolipoprotein E (APOEε4) and also to mutations in the amyloid β precursor protein (APP) 55. APOEε4 belongs to the family of apolipoproteins responsible for the regulation of lipoprotein metabolism 56, whereas APP is a membrane‐spanning protein involved in synapse formation and function and neural plasticity 57, 58.

The diagnosis of AD can only be confirmed upon autopsy of the brain, where the presence of amyloid β (Aβ) plaques and neurofibrillary tangles (NFTs) are the hallmark pathological signs 59. Over‐expression of tau protein (a microtubule‐associated protein), hyperphosphorylation and mutations are known to contribute to its aggregation 60-62. Also, the accumulation of Aβ in mitochondria is associated with the neuronal toxicity observed in this disorder 63-65. Nevertheless, glycation is also believed to play an important role in NFTs formation as well as in the development of senile plaques.

In AD, there is an increased content of AGEs in the Aβ plaques and NFTs 66, 67. Tau glycation enhances the formation of paired helical filaments in AD and reduces its ability to bind microtubules in vitro 68. In addition, modification of tau by AGEs is responsible for increased fibrillization of the protein, although it does not affect the lag time of the equilibrium. These results suggest that glycation, together with pseudophosphorylation, do not promote fibrillization per se but, instead, promote the stabilization of the filaments once they nucleate 69. In addition, AGE‐modified tau leads to an increase in the production and secretion of Aβ, followed by the formation of reactive oxygen species (ROS) 70. Interestingly, glycation agents such as methylglyoxal are able to activate p38 MAP kinase, which is able to phosphorylate tau 71, 72, a process believed to occur in neurons in AD 73. These results suggest that glycation may also play a role on tau phosphorylation.

In Aβ, glycation by methylglyoxal promotes the formation of β‐sheets, oligomers and protofibrils and also increases the size of the aggregates 74. Interestingly, the expression levels of glyoxalase I in the brains of AD patients is lower in later stages of the disease, as confirmed at the level of both mRNA and protein 75. Since this is the main methylglyoxal catabolic pathway, a higher carbonyl stress is expected and an increase in AGE content, oxidative stress, sustained inflammation, plaques and tangles formation and, ultimately, apoptosis are likely to occur.

Hyperglycaemia is a major player in carbonyl stress, which is observed in AD, and there is also a link between diabetes mellitus and AD. Diabetic patients have a two‐ to five‐fold higher susceptibility for developing AD compared to non‐diabetic control patients 76. Another factor eliciting carbonyl stress is the AGEs derived from glyceraldehydes, which are present in the cytosol of neurons in the hippocampus and parahippocampal gyrus 77. These may inhibit the enzyme glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH), causing an increase in glyceraldehydes and, consequently, methylglyoxal, forming a vicious cycle leading to neurodegeneration 78-80. Remarkably, glycation levels in AD are three‐fold higher than normal subjects 66.

Whether glycation is a trigger or a late stage event remains controversial. Some data suggest that AGE formation is a late secondary event in AD, since Aβ alone induces free radical generation that can promote cross‐linking between peptides and sugars 81. On the other hand, glycation of paired helical filaments of A68 proteins, a known precursor of NFTs, suggests that it could be an early event 82. Although AGE levels increase with age, in AD the increase is much greater (37.5% and 72.6%, respectively) 83. Oxidative stress and carbonyl stress were also shown to precede NFTs and plaque formation in AD 83. ApoE also seems to be modified by AGEs, since the staining patterns of AGEs and ApoE are similar in AD 84.

RAGE is an important player in AD, since Aβ is also recognized by this receptor 85. RAGE is expressed in astrocytes and recruits monocytes across the blood‐brain barrier activating NF‐κB and, consequently, increasing oxidative stress 78, 86, 87. In fact, AGEs, Aβ and RAGE colocalize in astrocytes, suggesting that Aβ may be taken via RAGE to be degraded by the lysosomal pathway in these cells 88. AGEs are also found in the cerebrospinal fluid (CSF) of AD patients, suggesting that this may be explored as a biomarker for AD.

Glycation in Parkinson's disease

PD is the second most common progressive neurodegenerative disorder, with a prevalence of around 2% in people over age 60 89. Muscle rigidity, resting tremor, bradykinesia, rigidity and postural instability are the most common symptoms of the disease 90, 91. Loss of nigrostriatal dopaminergic neurons and the accumulation of proteinaceous intraneuronal cytoplasmic inclusions called LBs are the main pathological hallmarks 92, 93. The majority of PD cases are sporadic, although several genetic loci have been associated to familial cases of the disease. Three mutant forms of PARK1, the gene encoding for α‐syn, are associated with familial cases 94-96. In addition, over‐expression of wild‐type α‐syn is sufficient to cause disease 97. While the function of α‐syn is still poorly understood, it is thought to be associated with vesicular trafficking, modulation of synaptic function and plasticity modulation, and the regulation of dopamine neurotransmission 98. α‐Syn is the target of several post‐translational modifications. In LBs, oxidized and phosphorylated forms of the protein have been reported 99-101. Under physiological conditions, α‐syn exhibits a natively unfolded conformation. If associated with membranes, it can adopt different folded conformations rich in α‐helical content 102. In other situations, it may also form aggregates or oligomers, with higher β‐sheet content, which are believed to represent the most cytotoxic forms 103-106. Nevertheless, the process leading to aggregation and LBs formation is unclear and, in addition to oxidation and phosphorylation, glycation might constitute another factor affecting the aggregation process.

Glycation was first reported in the substantia nigra and locus coeruleus, displaying higher immunoreactivity at the periphery of LBs in PD patients 107. These results suggest that glycation may be involved in the chemical crosslinking and proteolytic resistance of the protein deposits. While glycation was also detected in the cerebral cortex, amygdala and substantia nigra of older control individuals, the numbers and levels of glycated proteins were significantly higher in PD patients 108. RAGEs were also highly expressed in PD patients when compared to age‐matched controls 108, suggesting a role for AGEs in the disease.

α‐Syn is a lysine‐rich protein with 15 residues which are putative glycation targets (Figure 2). Thus, it is likely that glycation occurs, influencing the nucleation of pathological aggregates. One important feature of PD is an acute decrease in the levels of cellular reduced glutathione (GSH) in early stages, which results in a lower activity of the glyoxalase system, an important catabolic pathway of the most important glycation agents in vivo, such as methylglyoxal 109. The carbonyl stress will then be responsible for an increase in AGEs concentrations that will raise oxidative stress, which further induces AGEs formation. This deleterious cycle might contribute to the cell damage reported in dopaminergic neurons where, moreover, dopamine auto‐oxidation and its degradation by monoamine oxidase is likely to contribute to an increase in oxidative stress. Strikingly, methylglyoxal and glyoxal are able to induce oligomerization of α‐syn in vitro 110. The membrane‐binding ability of glycated α‐syn was also decreased 110. Thus, we hypothesize that in a glycation‐prone environment, more cytotoxic α‐syn aggregates or oligomers are present in the cytoplasm, contributing to the development of PD.

image

Schematic representation of the primary sequence of α‐syn, whose sequence may be subdivided into three domains. The N‐terminal domain (residues 1–65) includes six imperfect repeats of an unusual sequence of 11 amino acid residues with the consensus sequence KTKEGV (shadowed in the figure). The central hydrophobic domain of α‐syn is known as NAC (non‐Aβ component of AD amyloid; residues 66–95). The C‐terminal domain is negatively charged and is composed by a Glu‐rich sequence (residues 96–140). The 15 putative protein glycation targets are shown in yellow and the three known mutations A30P, E46K and A53T, are also shown above the amino acid sequence.

Glycation in amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease affecting 4–6 per 100 000 people, and shares with other disorders of the ageing nervous system a polygenic, multifactorial origin. 111. ALS is characterized by the degeneration of motor neurons in the brain and spinal cord, leading to muscle weakness, which in turn affects voluntary movements and also the respiratory muscles. Ultimately, this results in muscle atrophy, paralysis and breathing complications 112. The causes of ALS are still unknown and 90% of the cases are sporadic 113. Several inherited cases of the disease are related to SOD1 copper–zinc superoxide dismutase (Cu,Zn‐SOD1) missense mutations 113, 114. This protein is one of the three isozymes that catabolizes superoxide radicals, suggesting that oxidative stress may play an important role in the disease 115. The intracellular and perikaryal accumulation of protein inclusions of neurofilaments as well as LB‐like cytoplasmic inclusions, together with TAR DNA binding protein 43 (TDP‐43) cytoplasmic aggregation, are the pathological hallmarks of the disease 116. TDP‐43 is involved in the regulation of neuronal plasticity via transcriptional regulation 117, 118. Both familial and sporadic cases display extensive neurofilament conglomeration in motor neurons and, in extreme conditions, the formation of axonal spheroids 119. Thus, as in other neurodegenerative disorders, it is also essential to understand the role of protein aggregation in ALS.

Glycation was first detected in both sporadic and inherited forms of ALS, in the spinal cord and brain samples from patients. Initially it was postulated that glycation could be involved in the time‐dependent cross‐linking of neurofilament protein 120. Neurofilament protein subunits contain multiple Lys–Ser–Pro sequences. If the lysine residue is glycated, the assembly process may be impaired 120. Further studies revealed that AGEs levels were higher in patients carrying SOD1 mutations and also in mutant SOD1 transgenic mice, while control cases did not display AGEs immunoreactivity 121. The presence of AGEs in astrocytes in the spinal cord suggests that carbonyl stress in SOD1 mutant subjects may induce the potentially deleterious effects of AGEs. Strikingly, levels of soluble RAGE (sRAGE), a C‐terminal truncated isoform of RAGE, are significantly lower in the serum of ALS patients 122. sRAGE, lacking the transmembrane‐anchoring domain, is found to ameliorate the deleterious effects of RAGE by scavenging its ligands without further activating RAGE mediated‐processes 123, 124. Thus, sRAGE may function as an endogenous protection factor in ALS, indicating that the low sRAGE levels may pose a risk factor in the disease.

One interesting finding is that glycation, which is thought to be a random process, modifies SOD1, specifically at Lys3, Lys9, Lys39, Lys36m, Lys122 and Lys128 residues 125, although 15 possible lysine and arginine residue glycation targets are present. Diminished activity is present when Lys122 and Lys128 are modified 125 and a higher susceptibility of mutated SOD1 to glycation was observed 126. These results suggest that glycation might be responsible for the observed oxidative stress in ALS. Recently, it was established that the concentration of N‐ε‐(carboxymethyl)lysine (CML, an AGE derived from the reaction between glyoxal and the side chain of lysine residues) significantly increases in serum and cerebrospinal fluid of ALS patients, possibly representing a novel biomarker for diagnosis, further highlighting the relevance of glycation in ALS 127.

Glycation in familial amyloid polyneuropathy

Familial amyloid polyneuropathy (FAP) is a systemic amyloid disease, characterized by the extracellular deposition of fibrillar transthyretin (TTR), particularly in the peripheral nervous system. The disease was first described in 1952 as a sensorimotor peripheral polyneuropathy also affecting the autonomic nervous system 128. TTR is a protein responsible for the transport of thyroxine hormone (T4) and retinol in the blood. It is a homotetrameric protein, present in the plasma and cerebrospinal fluid 129. TTR is the main constituent of the typical amyloid fibrils present in the disease, but its misfolding and aggregation are far from being understood 130. The best established model for TTR fibril formation postulates that, under certain unknown conditions, TTR undergoes structural changes and dissociates into partially unfolded monomers. These may then associate into high molecular mass soluble aggregates that might further progress to insoluble amyloid deposits 131, 132. In fact, mutant forms of TTR responsible for familial FAP forms display a higher propensity to form partially unfolded monomers and soluble aggregates 133. Although mutations in TTR are responsible for the development of familial FAP, wild‐type TTR also forms amyloid fibrils in patients with senile systemic amyloidosis, a disease that affects around 25% of the population aged over 80 134.

The intrinsic amyloidogenic behaviour of TTR indicates that non‐genetic factors may also contribute to the disease progression. Since glycation promotes cross‐linking and formation of insoluble aggregates, it might do so by impairing the reversibility in the formation of soluble aggregates. This hypothesis is reinforced by the detection of AGEs in TTR fibrils a few years after the first report of glycation in the amyloid deposits of FAP patients 135, 136. Glycated TTR may contribute to cytotoxicity via oxidative stress directly induced by AGEs, or by interaction with RAGE 137-141. The interaction of TTR fibrils with RAGE results in the translocation of NF‐κB to the nucleus and the concomitant induction of tumor necrosis factor‐α (TNFα) and interleukin‐1β (IL‐1β) 140, 141. Thus, altogether these data suggest that blocking RAGE may constitute a good target for therapeutic intervention in FAP.

Glycation in prion diseases

Prion diseases are characterized by the accumulation and aggregation of abnormally folded cellular prion protein (PrPequation image). This protein is believed to play a role in long‐term memory formation and in haematopoietic stem cell self‐renewal 142, 143. Neuronal loss, vacuolation, microgliosis, astrogliosis and spongiform alterations are hallmarks of this group of diseases, which includes Creutzfeld‐Jakob disease (CJD) in men, bovine spongiform encephalopathy (BSE) in cattle and scrapie in sheep 144, 145. CJD may occur either sporadically or through inherited genetic mutations within the prion protein 146. The ‘protein only’ hypothesis states that these diseases are caused by the conversion of a normally folded prion protein (PrPequation image) into an abnormal isoform that is resistant to degradation by proteinase K (PrPequation image). Nevertheless, the exact mechanism is still controversial and there are arguments against this hypothesis 147. PrPequation image is a glycoprotein that exhibits SOD‐like activity 148, 149 and its conversion into the PrPequation image results in the intra‐ or extracellular accumulation of the protein, one of the pathological hallmarks of the disease 150.

The formation of AGEs has been studied in this group of disorders because glycation promotes the formation of protein cross‐links originating protease‐resistant proteins, and may induce oxidative stress. AGEs and RAGE were found to be present in the occipital lobe of CJD patients in higher amounts than in control individuals 151. The hypothesis was that glycation would advance over time and, as with Aβ, PrP would be degraded through the lysosomal pathway in a RAGE‐mediated process. It was reported that this post‐translational modification occurs in the N‐terminus of the protein, since upon PK digestion, which cleaves ∼90 amino acid residues from the PrPequation image N‐terminal, glycation was no longer detected 152. Although human PrP contains 21 arginine and lysine residues (in the cleaved and processed form of the protein), glycation does not appear to be a random process. Only lysines 23, 24 and 27 and the arginine 37 of PrPequation image can be glycated 152. Glycation was reported in several mouse models infected with 139A, ME7, 22L and 87V scrapie strains, in hamster‐adapted 263K and 139H scrapie strains, and in both sporadic and genetic CJD. Glycated PrP was found, using western blot analysis, to form oligomeric species, where imunoreactivity was observed for monomeric and dimeric forms of PrP 152. Therefore, although the role of glycation in the formation of PrPequation image cannot be excluded, some questions still need to be addressed to better understand the role of glycation in prion aggregation and disease development.

Anti‐AGE systems

AGE‐defence systems must be able to cope with pre‐existent forms of AGEs or to inhibit their formation by scavenging or degrading glycation agents. Although the generation of glycating compounds is unavoidable in living systems, several protective enzymatic mechanisms are involved in their degradation. Methylglyoxal and glyoxal are the most reactive species, damaging proteins and other biomolecules through the Maillard reaction, which can culminate in cell death 153, 154. The glyoxalase system is the most extensively studied methylglyoxal catabolic route, but other oxide‐reductases and dehydrogenases are also able to convert methylglyoxal into its oxidized or reduced form 153. These include α‐oxoaldehyde dehydrogenase 155, aldehyde dehydrogenase 156, aldose reductase 157, methylglyoxal reductase 158, and pyruvate dehydrogenase 159. The real contribution of each of these enzymes in vivo is still controversial; nevertheless, it is commonly agreed that the glyoxalase system and aldose reductase are the most relevant methylglyoxal catabolic routes. Indeed, glyoxalase I, as already mentioned, is clearly decreased in AD 75, resulting in the accumulation of α‐dicarbonyls and subsequent deleterious effects.

Some specific inhibitors may also prevent glycation‐induced cross‐linking of proteins. Aminoguanidine, via its fast reaction with α‐dicarbonyl compounds, forms 3‐amino‐1,2,4‐triazine derivatives 160, whereas tenilsetam is a known AGEs cross‐linking inhibitor 161. Interestingly, both compounds were found to protect against the neurotoxic effects of methylglyoxal 162. Some studies also suggest the presence of specific enzymes able to degrade glycated proteins (deglycases or amadoriases). These include fructosamine 3‐kinase, one enzyme that phosphorylates protein‐bound fructosamines and spontaneously breaks down the Amadori rearrangement into inorganic phosphate, along with 3‐deoxyglucosone and the amino compound 163, 164. This process was observed in vivo, leading to deglycation of haemoglobin 165 and protein‐bound psicosamines and ribulosamines in erythrocytes 166. Fructoselysine oxidase 167 and fructoselysine 6‐kinase 168 are also believed to play a role in protein deglycation. The importance of deglycases is still controversial, since the resulting degraded compounds are able to react once more with amino groups of proteins, lipids and nucleic acids. Nevertheless, the inhibition of AGEs formation may represent a therapeutic approach also to be explored in neurodegenerative diseases.

Concluding remarks

Glycation has been reported in several neurodegenerative disorders, but several questions remain to be answered. Whether glycation of susceptible proteins is a triggering event or just a result of its reactivity towards low‐turnover aggregated species, which are highly insoluble and protease‐resistant, remains to be demonstrated. Several results suggest that glycation may be an early event promoting or accelerating abnormal protein deposition, followed by increased protease resistance and insolubility (Figure 3). For example, in AD, glycation of diffused NFTs 82, along with AGEs in diffuse Aβ plaques 169, are observed; in PD, AGEs are present in very early LB; in prion diseases, AGEs occur in astrocytes prior to the formation of PrPequation image 152.

image

Potential role of protein glycation in protein misfolding and aggregation. After synthesis, proteins may acquire different misfolded conformations and may form dimers, oligomers and, in later stages, amyloid fibrils, which may be sequestered into inclusion bodies (such as LBs for the case of α‐syn in PD). Reducing sugars, increased in hyperglycaemia conditions or when AGE defence mechanisms are overwhelmed, may affect the biology of proteins in early or later stages of the aggregation process (glycating protein monomers, dimers, oligomers and amyloid fibrils). This modification is responsible for an increase in both protein stability by the formation of crosslinks formation, and in protease resistance. AGEs may lead to cell damage by interacting with RAGE, inducing oxidative stress and inflammatory cell responses.

Regardless of the chronology of AGEs formation, it is known that its accumulation is related to sustained inflammatory responses and oxidative stress, which is a common feature in many neurodegenerative disorders. Glycation may then be understood as a dynamic contributor to these multifactorial diseases by promoting, accelerating or stabilizing pathological protein aggregation and inducing responses leading to cell dysfunction, damage and death. Thus, it will be important to further investigate the biochemical effects of glycation on protein function and dysfunction. For example, the activity of enzymes that have central roles in the cell, such as cysteine proteases 170, Cu,Zn‐superoxide dismutase 171, GAPDH, glutathione reductase, lactate dehydrogenase 172, aspartate aminotrasferase 173, catalase 174 and enolase 2 175, are known to be inhibited by AGEs. However, the reported effects are not always deleterious, as observed in haemoglobin and myoglobin, whose esterase activity is increased 176. Likewise, the chaperone activity of HSP27 and αB‐crystallin is also increased upon AGE modification 177, 178.

The question of whether glycation promotes protein fibrilization remains under intense debate. Although it seems to promote Aβ fibrilization 74 and α‐syn 110 and PrP 152 oligomerization, a direct relationship between glycation and β‐sheet formation in vivo is still missing. Since glycation is understood as a non‐enzymatic process, the modifications are predicted to occur at random. However, as already mentioned, SOD1 activity is only affected when specific residues are found to be glycated 125. Moreover, only four of the 21 PrPequation image putative glycation sites are modified 152. These results suggest that glycation may modify specific targets and be understood as a post‐translational modification that alters protein function, similarly to phosphorylation or acetylation. If this is true, a novel avenue for therapeutic intervention for several diseases may emerge by studying the specific glycation sites in the proteins involved in neurodegenerative disorders (eg α‐syn, Aβ, TTR). By developing specific inhibitors or enhancers of glycation, it may be possible to modulate the oligomerization of those proteins to promote the formation of species with lower cytotoxicity.

It will also be important to evaluate the effect of glycation on protein–protein interactions that may compromise several cellular pathways. For example, it is described that, upon RAGE interaction, Aβ may be taken up and degraded via the lysosomal degradation pathway 88. If Aβ is glycated at the site of RAGE recognition, its proteolysis may be compromised, leading to an increase in the concentration of Aβ.

In the future it will be important to evaluate the potential of using specific AGEs as biomarkers of disease. In the context of diabetes mellitus, for example, AGE fluorescence is currently being used as a marker of these complications 16, 17. It has been described that several neurodegenerative disorder‐related proteins are present in extracellular fluids 104, 179-182. It would be of great importance to probe the presence of AGEs in these proteins and establish a relationship between the extent of glycation and the pathological condition to develop new early diagnostic tools.

Acknowledgements

HVM was supported by fellowship SFRH/BD/23035/2005 from Fundação para a Ciência e Tecnologia, Portugal. TFO was supported by an EMBO Installation Grant and by a Marie Curie IRG.

      The full text of this article hosted at iucr.org is unavailable due to technical difficulties.