Brain EV holds promise to assist in the treatment of AD-type neurodegeneration and its associated cognitive signs [
40,
41]. This intercellular communication mechanism has been implicated in the spreading of neurodegenerative culprits from cell-to-cell in the brain, and suppression of EV biogenesis has been recently hypothesized as a viable strategy for slowing AD neurodegeneration [
6]. However, to advance the achievement of novel therapeutic applications and to better understand the contributions of these vesicles to AD neuropathology, their divergent compositions throughout the progressive course of AD and potential alterations in their biogenesis need to be uncovered, as recently noted [
42]. Enrichment of EVs from brain tissues can be considered a challenging task and the presence of contaminants from brain tissues in the obtained EV fractions cannot be overlooked. Thus, the methodology employed here previously shown that whole organelle contamination in brain EV preparations can be kept below 10% (particles larger than 700 nm in diameter) [
14], a result consistent with the NTA data obtained in this study. However, we still consider that whole brain contamination in brain EV preparation is a challenging limitation that requires of further research, by complementary methodologies, and of continuous improvement.
In this study, our systems biology approach revealed that EV biogenesis becomes altered during the preclinical stage of AD linked to an increase in the genesis of a specific population of EV that express the MHC class-type markers. These findings are partially consistent with previous reports in other brain diseases [
15]. EV containing MHC class antigens have been vastly attributed to bone marrow dendritic cells (DCs) and brain microglia cells [
43,
44]. Several studies indicate that DCs alter their functional properties and release proinflammatory molecules in the presence of amyloid-β toxicity [
45,
46]. Based on the findings reported here, the interaction between DCs and amyloid-β is mediated by EV. Additionally, these data do not support the hypothesis that an increase in brain EV biogenesis would be caused by interactions of the Aβ peptide and hyperphosphorylated Tau with the endocytic pathway, as previously hypothesized [
6]. An increase in EV biogenesis was only found to be significantly evident in preclinical AD in this study; thus, our findings align better with previous reports that show mild-to-moderate hypoxia as the core cause of the boost in EV biogenesis in the diseased brain [
47]. This is especially manifested when we consider that brain EV manifest the influence of the AD vascular component during the preclinical stage of the disease [
17]. Similarly, we found by advanced proteome clustering that common patterns between mixed dementias and all analyzed AD stages could only be identified in brain EV from preclinical AD, a finding that reaffirms our recently suggested hypothesis indicating that the influence of the vascular component in brain EV may decay over the course of AD [
17]. Although we found that the amyloid cascade seems to not interfere with EV biogenesis, especially throughout the disease course, an abrupt increase in the levels of amyloid precursor protein APP in brain EV during preclinical AD was encountered. We also observed that this increase in APP in brain EV was maintained throughout the whole AD neuropathology. The presence of APP in brain EV from experimental AD models was initially reported in 2012 by Perez-Gonzalez [
48], although very recently, it has also been shown that APP is present in brain EV from subjects with AD at higher levels compared to age-matched controls, as we have also observed here. These authors also found that these EV loaded with APP show the ability to spread Aβ toxicity cell-to-cell [
49]. Similarly, previous in vitro and in vivo studies have linked the upregulated presence of APP in EV to the upregulated presence of the prion protein PrP in these vesicles [
49]. However, this co-upregulation was not confirmed in clinical subjects with AD until here, when we found significant co-upregulation of the prion protein PrP and APP in brain EV from subjects with AD compared to age-matched controls. According to these data, PrP was at its highest level in brain EV during the preclinical stage of the disease. The interaction between APP and PrP in brain cells has been largely described [
50,
51], although the purpose of that interaction in brain EV remains unknown [
42]. The available data about the interactions between these proteins come mainly from Falker and colleagues, who proposed that PrP in EV may contribute to sequestering toxic amyloid oligomers [
52]. Our findings throughout AD progression support this hypothesis; hence, a significant increase in PrP in brain EV was uniquely manifested during the preclinical stage of AD, and the presence of PrP in brain EV was reduced to similar levels as in age-matched controls at the initial manifestation of clinical signs. Thus, based on our data, we consider that PrP may work as a switch activating the sequestration of APP in brain EV, a mechanism that abruptly decays in the temporal brain after passing through the preclinical stage of the disease. Furthermore, our results do not support the hypothesis that brain EV enriched with toxic amyloid oligomers and PrP contribute to increased levels of aggregates in AD brains; hence, co-upregulation of both proteins in brain EV is highly active only during the preclinical stage of the disease. In line with these findings, an extensive review analysis of the CNS-derived EV literature in the context of AD is included in Table
2 and indicates that circulating neuronal EV contain Aβ1-42 up to 10 years prior to the clinical onset of AD [
53]. Similarly, several studies demonstrate that circulating EV exert potential diagnostic/prognostic abilities in AD dementias [
54‐
58]. Furthermore, the upregulated presence of phosphorylated Tau residues in circulating brain EV from cerebrospinal fluid was encountered in preclinical subjects with AD (Braak stage 3) [
59]. We also found that brain EV are significantly loaded with higher levels of essential brain enzymes during the preclinical stage of the disease, which included protein-
l-isoaspartate (
d-aspartate) o-methyltransferase (PCMT1), neuroendocrine convertase 2 (PCSK2), dihydropyrimidinase-like 2 (DPYLS2), and biliverdin reductase-A (BLVRA). All these enzymes have previously been implicated in AD [
60‐
63]. We also observed that the levels of these enzymes in brain EV, albeit upregulated in preclinical AD, were significantly downregulated throughout the entire disease progression. Similarly, we found that the levels of the immunoglobulin J chain in brain EV gradually and significantly increased during the preclinical and early symptomatic stages of AD, whereas these levels were downregulated in the terminal stages of the disease. The observed increase in immunoglobulin J chain can be associated with enhanced permeability of the blood brain barrier and an increased supply of outer EV to the diseased CNS [
64].
Table 2
Extensive review analysis of CNS-derived EVs findings in the context of AD
P-S396-tau, P-T181-tau and Aβ1-42 | Upregulation in AD | NDE | Immunoassay | Predictor of AD development prior to clinical onset | |
SNAP-25 | Downregulation in AD | NDE | Immunoassay | Negative correlation between the levels of SNAP-25 and cognitive status | |
Gelsolin | Downregulation in DLB compared to AD | Plasma-derived EVs | Proteomics | Potential biomarker able to differentiate DLB and AD | |
Growth-associated protein 43 and synapsin 1 | Downregulation in AD but not in dementia patients | NDE | Immunoassay | Possible early differential diagnosis marker to differentiate AD and dementia | |
β/γ-secretase and sAPPβ | Upregulation in AD | NDE | Immunoassay | Astrocyte-derived exosomes of AD patients show up to 20-fold upregulation than neuron-derived exosomes | |
pS396 tau and Aβ | Upregulation in AD | Cortical grey matter EVs | Immunoassay | – | |
ANXA5, VGF, GPM6A and ACTZ | Presence | Cortical grey matter EVs | Quantitative proteomics and machine learning | EVs signature panel of proteins in AD | |
Total and phosphorylated tau | Upregulation in AD | CSF from AD Braak stage 3 | Immunoassay | Considered patients with mild AD | |