Involvement of peroxisomes at the onset and during the progression of Alzheimer's disease in a transgenic mouse model

Abstract

Alzheimer’s Disease (AD) is the most common form of dementia, characterized by progressive neurodegeneration. Histopathological changes, including formation of beta-amyloid (Aβ) plaques, neurofibrillary tangles, followed by inflammation and neuronal loss, are especially prominent in the hippocampus and the neocortex. The classical hypothesis to explain the pathogenic events occurring in AD, the so-called amyloid cascade, involves a critical and early role of A peptide, involving production of reactive oxygen species (ROS) and consequent oxidative stress (Sayre et al., 2008). Even though Atoxicity is still widely accepted, this hypothesis has been recently challenged by Nunomura (2010) who proposed oxidative stress as the primary culprit in AD pathogenesis. In this scenario, peroxisomes may play an important role, as they are involved in a wide variety of metabolic processes, including ROS and lipid metabolism (Schrader and Fahimi, 2008). It known that loss of peroxisomes makes neuronal cells vulnerable to oxidative stress and leads to degeneration (Stamer et al., 2002). On the other hand, peroxisomal proliferation attenuates Aβ-dependent toxicity in hippocampal neurons (Santos et al., 2005). This PhD project aims to investigating the possible involvement of peroxisomes in AD onset and progression, with special reference to their role in oxidative stress. To this purpose we examined the transgenic mouse strain Tg2576, which recapitulates human pathological phenotype and is especially suited to the study of early stages, for it is characterized by slow progression (Jacobsen et al., 2006, D’Amelio et al., 2011). Our study was performed on early, advanced, and late stages of the disease (3, 6, 9, 12, 18 months of age), focusing on the hippocampus and neocortex, i.e., the areas where primary neuronal injury is known to occur. First, we analyzed the distribution of neuronal and glial markers, in WT and Tg2576 ageing/diseased mice, in order to detect possible alterations in brain organization. Our data show that the overall cytoarchitecture is conserved during normal aging, while in the pathological genotype, neuronal layering gradually changes starting from 9 months and appears dramatically altered at 18 months, when hyperproliferated and hypertrophic astrocytes and microglia cells are detected in both the neocortex and hippocampus. Consistently with the observed cytoarchitectural alterations, Congo Red staining demonstrates that small amyloid plaques are already present in the neocortex at 9 months, while their first appearance in the hippocampal formation occurs at 18 months. Main objective of this PhD project was to investigate age-dependent variations of several peroxisome-related proteins in Tg2576 and WT hippocampus and neocortex in the course of the disease. Specifically, we characterized the peroxisomal population from a molecular and morphological point of view, examining the expression of biogenesis markers, membrane proteins, and matrix enzymes. Our biochemical, morphological and ultrastructural data concur to demonstrate a significant peroxisomal induction in 3-month-old diseased hippocampus, as assessed by the peroxin Pex5p, the peroxisomal membrane protein PMP70, and acyl-CoA-oxidase (AOX, the first enzyme of the fatty acid peroxisomal -oxidation pathway), all of which show higher levels in Tg2576, compared to controls. This increase, paralleled by detectable oxidative modifications to biomolecules and by PPAR activation in Tg2576 hippocampus, suggests an early response to redox imbalance, mediated by PPAR, which is known to be induced by oxidized lipids. Strikingly, the other crucial enzyme for the peroxisomal -oxidation pathway, i.e., thiolase (THL), is unchanged at 3 months, indicating possible inefficiency of the -oxidation pathway in the Tg2576 than in the WT. This would imply accumulation of VLCFA-derived intermediates in the brain, in line with the recent findings on the brain of AD patients (Kou et al., 2011). Even more remarkably, at this early stage, catalase (CAT) levels show no genotype-based differences, suggesting that increased H2O2 production by AOX is not accompanied by its efficient removal. These data, together with the low expression levels shown by another major H2O2-scavenging enzyme, glutathione peroxidase (GPX1), strongly favour the idea that oxidative stress occurs early in AD. To this respect, it is also worth noting that mitochondrial superoxide dismutase (SOD2) is increased in the hippocampus of 3-month-old Tg2576 mice. Despite its commonly referred role as a ROS scavenger, this enzyme could even act as a pro-oxidant in this context. Indeed, the augmentation of SOD2, converting superoxide ion to H2O2, in the absence of parallel increase in any of the H2O2 scavengers, likely contributes itself to redox imbalance. Therefore, our data prefigures an early oxidative stress condition occurring in the hippocampus, strongly supporting the idea that this is the primary culprit in AD pathogenesis (Nunomura et al., 2010). Indeed, our data on the oxidative damage markers 8-OHG/OHdG and acrolein confirm this view. During the progression of AD pathology, peroxisomal population undergoes dramatic changes, in that several proteins, including PMP70, Pex14p, and AOX, are decreased in 6-month-old Tg2576 hippocampus. This may reflect an impaired peroxisomal biogenesis or autophagic removal of excess organelles. Interestingly, in the hippocampus relatively high levels of CAT are detected at this age in the pathological genotype, suggesting a late response to peroxisomal induction, in agreement with what described in the liver following peroxisomal agonists administration (Reddy et al., 1986). At 9 and 12 months, hippocampal levels of PMP70, AOX, and THL are relatively stable in both conditions. Other antioxidant enzymes show interesting changes at more advanced stages of the disease. Indeed, GPX1 and SOD1 increase from 6 to 9 months of age in the Tg2576 hippocampus, suggesting a late response, not involving the peroxisomes. However, at 12 months, when amyloid plaques start to form, GPX1, SOD1 and SOD2 show significantly lower levels in the Tg2576 mice, with respect to WT. Accordingly, decreased activities of SOD1 and GPX have been reported in patients with symptomatic AD by Casado et al. (2008). The situation observed in 12-month-old Tg2576 hippocampus, closely resembling that of SOD1 and of GPX1, may reflect a decreased ability of neurons of this region to respond to A-mediated insult, possibly involving ROS generation. A striking result regarding peroxisomes was obtained on 18-month-old mouse hippocampus. In fact, a peak of PMP70 expression is observed irrespective of the genotype, allowing to hypothesize a change in peroxisomal population related to the ageing process. Consistently, its main regulator, PPAR, shows intense immunoreactivity in the aged hippocampus, in both WT and Tg2576 animals. This late increase of peroxisomal number, not accompanied by induction of any of the peroxisomal matrix enzymes, may indicate the occurrence in the hippocampus of an attempt to counteract cellular damage, which however fails to result in improved efficiency of the organelles, either as ROS scavengers, or as lipid metabolizing sites. Indeed, AOX levels remain stably low at 18 months, while THL is dramatically decreased at the same age, allowing to hypothesize that inefficient -oxidation is associated with normal and pathological aging. Antioxidant enzymes show protein-specific variations in the aged hippocampus. Overall, it is conceivable that a pro-oxidant environment is still present in the cytosol, rather than in peroxisomes, since decreased levels of GPX1 and SOD1 are observed in Tg2576 mice, while CAT is unchanged. Concerning the characterization of the peroxisomal population in the neocortex, our data show a delayed response in this brain area, with respect to the hippocampus, as well as substantial region-based differences in the susceptibility to A-mediated damage. The reason for this behaviour may relate to the relatively high anti-oxidant enzyme levels, which likely make the neocortex less susceptible to oxidative stress than the hippocampal formation (Cimini et al., 2009). In fact, at 3 months of age we failed to detect any variations in the expression levels of peroxisomal proteins. Nevertheless, catalase showed a delocalization, as assessed by immunoelectron microscopy, being found at extraperoxisomal sites, including the nucleus, cytosol, and post-synaptic densities. This evidence leads us to hypothesize that the presence of catalase to sensitive sites may contribute to enhancing protection against oxidative damage. The most relevant quantitative change in the neocortex are observed in 6-month-old Tg2576 mice, where significantly lower levels of peroxisomal proteins (PMP70, CAT) are detected, with respect to WT. Molecular data concerning peroxisomal -oxidation enzymes show no statistically significant variations during aging or disease progression, indicating that no perturbation of this pathway occur in the neocortex. In conclusion, while 3 months of age is an especially promising time point for Tg2576 mice for devising therapies aimed at delaying or even preventing AD onset (Cimini et al., 2009; D’Amelio et al., 2011), the 6-month-old time point seems a more critical period, in which different approaches could be designed to counteract AD-like neurodegeneration. Based on our results, we suggest that potential therapies using PPAR agonists may be beneficial around 6 months of age.