Therapies for Alzheimer’s disease: a metabolic perspective

Highlights

• In Alzheimer’s disease, cerebral glucose transporters (mainly GLUT1 and GLUT3) seem to be downregulated.

• Glucose hypometabolism destabilizes the normal flow of the hexosamine biosynthesis pathway, reducing O-GlcNAcylation rates.

• An unbalanced relation between glycation and O-GlcNAcylation can promote Alzheimer’s disease pathological features.

• Impaired insulin signalling is often observed in AD patients and associated with its more prominent features.

• Metabolic therapies have shown to be promising therapeutic approaches for AD.

Abstract

Alzheimer’s disease (AD) is one of the most common forms of dementia in the elderly. Currently, there are over 50 million cases of dementia worldwide and it is expected that it will reach 136 million by 2050. AD is described as a neurodegenerative disease that gradually compromises memory and learning capacity. Patients often exhibit brain glucose hypometabolism and are more susceptible to develop type 2 diabetes or insulin resistance in comparison with age-matched controls. This suggests that there is a link between both pathologies. Glucose metabolism and the tricarboxylic acid cycle are tightly related to mitochondrial performance and energy production. Impairment of both these pathways can evoke oxidative damage on mitochondria and key proteins linked to several hallmarks of AD. Glycation is also another type of post-translational modification often reported in AD, which might impair the function of proteins that participate in metabolic pathways thought to be involved in this illness. Despite needing further research, therapies based on insulin treatment, usage of anti-diabetes drugs or some form of dietary intervention, have shown to be promising therapeutic approaches for AD in its early stages of progression and will be unveiled in this paper.

Introduction

Alzheimer’s disease (AD) is a progressive neurological condition that gradually deteriorates memory and learning capabilities through neuronal degeneration. It is the most common form of dementia, currently affecting 20 to 30 million people worldwide [1]. By 2050, it is estimated that around 136 million people will be afflicted by some form of dementia, making it a global health and economical challenge [2].

Major hallmarks of this illness are formation and accumulation of senile plaques, mostly made of amyloid-β (Aβ) and dystrophic neurites, and neurofibrillary tangles (composed primarily by hyperphosphorylated tau protein). Aβ plaques are primarily comprised of Aβ with 40 or 42 (Aβ42) amino acids, which are two by-products of amyloid precursor protein (APP) metabolism [3]. The Aβ42 peptide is the most abundant isoform due to its higher fibrillization rate and insolubility [4]. An imbalance between Aβ production and its clearance is a trigger for AD’s pathological process. Tau is a major microtubule-associated protein (MAP), and it interacts with tubulin to promote microtubules assembly and network stabilization. In AD, hyperphosphorylated tau loses the ability to bind tubulin and thus disrupts microtubule assembly. This toxic property involves the sequestration of normal tau protein and other MAPs, which impairs their normal function [5].

One of the major genetic risk factors associated with sporadic AD is the allele ε4 pertaining to the apolipoprotein E (APOE) gene. Its major roles are related to lipid transport and cholesterol homeostasis [6]. In addition, APOE protein also mediates other processes like synapse formation, modulation of neurite outgrowth, synaptic plasticity, destabilization of microtubules, and Aβ clearance [7]. APOE has three isoforms: APOE2, APOE3 and APOE4. Carriers of APOE4 are shown to be at higher risk of developing AD. Among the three isoforms, APOE4 is primarily responsible for increasing cholesterol levels in the brain. This isoform was shown to be hypolipidated and less effective than APOE3 in inducing cholesterol efflux, thus implying that the pathological effects of APOE4 could be related to lipid metabolism. APOE2 carriers on the other hand tend to exhibit lower cholesterol levels due to the reduced binding affinity for low-density lipoprotein [6].

Another important feature of this illness is neuroinflammation. It entails the activation of different types of brain immune cells like microglia and astrocytes [8,9]. It is largely mediated by redox events both at molecular (e.g., nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), inflammasome) and cellular level (e.g., inflammatory signals transmitted to neurons by astrocytes, via oxidants like H₂O₂) [10]. It is assumed that the inflammatory responses are driven by a positive feedback-loop relationship between astrocytes and microglia. Both cells types can amplify each other’s inflammatory responses, worsening the neurotoxic environment and promoting neurodegeneration [10]. Aβ is able to activate both microglia and astrocytes, inducing the production of H₂O₂, nitric oxide, prostaglandins, pro-inflammatory cytokines and chemokines, which can lead to neuronal death [10,11]. However, despite this, it is important to emphasize that a pro-inflammatory environment is required for the activation of microglial cells that are able to clear Aβ and even support neurogenesis after neuronal damage [12]. During an inflammatory process, active microglia exhibit a shift in their cellular metabolism from oxidative phosphorylation to aerobic glycolysis. This is known as the Warburg effect [11]. Despite being less effective mechanism to generate adenosine triphosphate (ATP) than mitochondrial respiration, glycolysis is much faster than oxidative phosphorylation (OXPHOS), allowing high energy processes like proliferation, migration, cytokine secretion and phagocytosis to occur. A study conducted by Baik et al. [13] showed that Aβ induced microglial activation by shifting their metabolism toward aerobic glycolysis. This activation was dependent on glycolysis and the mammalian target of rapamycin (mTOR)- hypoxia-inducible factor (HIF)-1α pathway. However, after prolonged exposure to Aβ, microglia became tolerant and exhibited defects in their cellular metabolism, involving both glycolysis and OXPHOS, as well as diminished inflammatory responses. They also reported that treatment with interferon-γ was able to restore phagocytic activity to tolerant microglia, as well as reduce Aβ plaque accumulation and neuronal loss. This alludes to the conclusion that metabolic modulation of microglia phenotypes could reduce the extent of neuronal damage.

Mitochondrial dysfunction, endoplasmic reticulum (ER) stress and abnormal calcium influxes are also very prominent events observed in this pathology and they are very closely related [14]. Oxidative damage to DNA can interfere with gene transcription and proper promoter function, hindering the transcription of essential genes and generating mutations. Mitochondrial DNA can also be affected, which contributes to energy production impairment and mitochondrial dysfunction [15]. In turn, this leads to a decrease in ATP production and dysfunctional ion gradient maintenance, hindering neurotransmission [16]. Loss of ion gradients also leads to higher intracellular calcium levels, which reduces the buffering ability of the ER and mitochondria. This causes stress to both organelles and further aggravates synaptic dysfunction, causing neuronal death. High levels of intraneuronal free calcium also lead to loss of fidelity of microtubule assembly, consequently, reducing anterograde and retrograde transport of mitochondria and neurotransmitter vesicles [17].

Currently, AD treatment is based on symptom management using cholinesterase inhibitors, N-methyl-D-aspartate antagonists and combination therapy. However, these drugs exert weak beneficial effects on cognitive function, only providing some relief of the behavioural and psychological symptoms of dementia [18]. Several biomarkers are used for AD diagnosis, namely Aβ42, normal and phosphorylated tau in cerebrospinal fluid, APOE genotype and brain imaging techniques [19,20]. It is known that individuals who will eventually develop AD, exhibit pathological signs almost two decades before the onset of clinical symptoms. It follows that current therapy may begin too late for the reversal of AD’s hallmarks. Further research on how to achieve an earlier diagnosis is required to develop new approaches in order to slow, or preferably halt, AD progression [16]. In this paper we sought to review the potential therapeutic effect of metabolic strategies.