Age-Related Changes in the Synaptic Density of Amyloid-β Protein Precursor and Secretases in the Human Cerebral Cortex

INTRODUCTION

The metabolism of amyloid-β protein precursor (AβPP) is mostly processed by two parallel pathways initiated by α-secretases, namely ADAM10 to generate non-amyloidogenic peptides, and byβ-secretases, namely BACE1 to generate amyloid-β peptides (Aβ) [1, 2], which are considered as possible culprits to trigger Alzheimer’s disease (AD) [3]. In accordance with its localization also in synapses [4, 5], AβPP is involved in synapse formation [6, 7] apparently in a manner independent of its metabolism [8]. Therefore, it is believed [2] that AD could result from an imbalanced AβPP metabolism, which would format Aβ production, or could result from changes in AβPP levels, altering synapse maintenance. This is of particular relevance since AD begins by a synaptic dysfunction associated with the emergence of memory dysfunction [9, 10]. Since the greatest risk factor for AD is aging [11], which is also associated with a deteriorated memory performance [12], we now investigated possible age-related alterations in the density of AβPP and of α- and β-secretases selectively in synapses from the frontal cortex collected from autopsies of individuals with ages ranging from 20 to 80 years. Additionally, we also analyzed the sub-synaptic distribution of AβPP and secretases in the human cortex at a single age, to compare it with that described in rodents.

MATERIALS AND METHODS

Human brain samples from the subgenual cingulate cortex were collected at autopsies, performed at the Instituto Nacional de Medicina Legal e Ciências Forenses (INMLCF), of Caucasian Portuguese male subjects who died from natural causes or accidents with ages ranging from 20 to 80 years. All procedures were approved by the INMLCF and followed the rules of the European Consortium of Nervous Tissues: BrainNet Europe II, to protect the identity of individual donors. The samples were rapidly frozen in liquid nitrogen and stored at –80°C until analysis. We selected for analysis only the samples lacking any evident alteration of their morphology (visual inspection of cortical tissue, without specific histological analysis of the eventual presence of plaques and/or tangles) with adequate brain pH and RNA integrity number (RIN) [13]. The subgenual cingulate cortex was selected for analysis because it exhibits hypometabolism without a significant atrophy during the aging process [14] and was available for collection rather than other brain regions used for standard pathology analysis during autopsy. The dissected cingulate cortical samples were homogenized by sonication to prepare total extracts and synaptic terminals membranes (synaptosomes), as previously described (e.g., [15]), or pre-synaptic, post-synaptic and non-active zone fractions, using a previously described fractionation protocol [16]. The samples were then analyzed by Western blot using primary antibodies against AβPP C-terminal (1 : 1,000; Sigma), BACE1 (1 : 1,000; Millipore), ADAM10 (1 : 500; Santa Cruz), SNAP-25 (1 : 10,000; Sigma), PSD-95 (1 : 10,000; Sigma), synaptophysin (1 : 10,000; Millipore), or syntaxin (1 : 20,000; Sigma) and the membranes were re-probed against glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1 : 2,500; Abcam) for normalization (see [5 , 17]). Data are presented as mean±SEM and statistical differences were probed at >95% confidence with a one-way ANOVA followed by a Tukey post hoc test.

RESULTS AND DISCUSSION

We first investigated if the synaptic and subsynaptic distribution in the human subgenual cingulate cortex of AβPP and of the two secretases initiatingits differential catabolism, α- and β-secretases, was similar to that previously reported in rodents [5, 17]. In order to analyze the distribution of AβPP, ADAM10, and BACE1 in synapses, we used synaptosomes, a preparation enriched in synaptic markers such as syntaxin (a marker of the presynaptic active zone), synaptophysin (a synaptic vesicle protein, marker of the non-active zone fraction), and PSD-95 (a marker of the post-synaptic density), when compared to total brain extracts (see Fig. 1A). The levels of AβPP, ADAM10, and BACE1 were significant higher in human subgenual cingulate cortical synaptosomes (of individuals with about 40 years old) than in preparations of total extracts, indicating that these proteins are enriched in synapses (Fig. 1B). We next used a subsynaptic fractionation protocol validated by the segregation of markers characteristic of each fraction (SNAP25 in the pre-synaptic active zone fraction, synaptophysin in the non-active zone fraction and PSD95 in the post-synaptic density fraction; see Fig. 1C), to assess the subsynaptic distribution of AβPP and secretases in the human cortex. We concluded that AβPP was mainly located in the pre-synaptic active zone (41.9±1.4% of total AβPP immunoreactivity in all subsynaptic samples, n = 3), while it was also present in the post-synaptic density (32.3±2.6%, n = 3) and at lower levels in the non-active zone (25.8% ±0.7%, n = 3) (Fig. 1D). The α-secretase ADAM10 was present at higher density in the post-synaptic density (61.3±4.3%, n = 3) than in pre-synaptic active zone (15.3% ±3.1%, n = 3) and in non-active zone (21.6% ±8.9%, n = 3) (Fig. 1E). The β-secretase BACE1 was found at high levels in non-active zone (58.6±10.7%, n = 3) and also present in pre-synaptic active zone (24.5±6.8%, n = 3) and in post-synaptic density (16.9±6.8%, n = 3) (Fig. 1F). This first report of the subsynaptic distribution of AβPP and secretases in the human brain revealed an essentially similar distribution to that reported in the rat hippocampus [5] and mouse cerebral cortex [17], suggesting a highly conserved distribution of AβPP and secretases across brain regions and across species.

We next tested if there were age-related changes of the density of AβPP and secretases in human cortical synapses. The analysis of subgenual cingulate cortical samples representative of four distinct age groups (average age of 20, 40, 60, and 80 years old) showed that although the levels of AβPP, ADAM10, and BACE1 in total extracts did not change significantly with aging (Fig. 2E-G), the synaptic density of AβPP tended to decrease with aging (Fig. 2B). In fact, in synaptosomes from the subgenual cingulate cortex, the AβPP levels in aged individuals (80 years old, n = 4) were lower (p < 0.05) than in young adult individuals (20 years old, n = 3). These alterations in synaptic AβPP levels were not accompanied byconsistent changes of synaptic markers (synaptophysin, SNAP-25 and PSD-95) with aging (Fig. 2A). Surprisingly, the description of age-related changes of AβPP density is scarce, and there is only information about the preservation of the AβPP mRNA levels in aged individuals [18], as well as in rats [19, 20]. This suggests that changes in the synaptic turnover rather that in the synthesis of AβPP might account for this reported age-related decrease of the synaptic density AβPP in the subgenual cingulate cortex. However, we observed that the synaptic levels of ADAM10, responsible for the non-amyloidogenic cleavage of AβPP, displayed a non-significant (n = 4-5 per age group; p > 0.05) tendency to decrease with aging (Fig. 2C); similarly BACE1 levels did not change significantly with increased age (Fig. 2D). This rules out the possibility of excessive local AβPP degradation in synapses as a cause of the decrease of the synaptic levels of AβPP with increasing age. Curiously, previous studies reported an increase of BACE1 activity, but not in BACE protein levels [21] and increased intra-neuronal ADAM10 immunoreactivity in the aged human cortex [22], which, together with the present results, suggest a particular metabolism of AβPP in synapses, differing from the extra-synaptic AβPP metabolism. Future studies attempting to quantify the levels of AβPP C terminal fragments (CTF83 and CTF99) might bring additional confirmation to this working hypothesis. It would also be of interest to further detail for possible age-related alterations of the sub-synaptic localization of AβPP and secretases, which was not possible since the number of donor was not sufficient to obtain enough tissue for sub-synaptic fraction in all age groups.

This decreased density of AβPP in cortical synapses upon increased age is unlikely to be a reflection of the loss of cortical synapses with aging [23] since we did not observe significant alterations in synaptic markers with age progression (Fig. 2A) and besides that we normalized the evaluation of AβPP density by the amount of GAPDH to ensure that similar amounts of nerve terminals were analyzed across the age groups. Instead, the decreased density of synaptic AβPP might be associated with the decreased ability to maintain synapses and synaptic dynamic with aging [24], in view of the role of AβPP in the control of the formation and function of synapses [8 , 26].

In conclusion, this study provides the first description of the subsynaptic distribution of AβPP and secretases in human brain cortical synapses and shows for the first time that the synaptic levels of AβPP decrease with increased age in the human cerebral cortex. This suggests a possible participation of the decreased levels of AβPP in the lower allostasis of synapses in the aged brain and their greater susceptibility to dysfunction characteristic of the early phases of neurodegenerative disorders. It is important to stress that the present study only focused on synaptic variations, leaving open the possibility that more global changes of AβPP and secretases might influence other processes critical for the enhanced susceptibility of the aged brain to disease, such as neurogenesis [27] also controlled by AβPP [28].