Aims

UCB-J is a recently developed PET-tracer to study synaptic integrity in different neurological disorders. Though several UCB-J in vivo PET studies have been published in AD and other neuropsychological disorders, there have been some discrepancies between the findings. Many of these studies indicated potential correlations between tau deposition, atrophy, cognitive-impairment, and loss of UCB-J binding. Based on these in vivo observations, we found it of utmost relevance to perform in vitro characterization of UCB-J in postmortem AD brains.

Methods

We performed in vitro radioligand binding assays (saturation, competition, and regional distribution) alongside large brain section autoradiography in AD and control brains.

Results

³H-UCB-J saturation studies showed much higher specific binding (above 90%) in synaptosomal membrane P2-fraction than brain homogenates prepared from the same frontal cortex (FC) region of AD and control brains. The specific binding was in nanomolar-range (Kd ~ 5.0 nM) and clearly highlighted the loss of UCB-J binding in FC of AD brain compared to control (33% decrease in Bmax vs. control). This was very interesting as UCB-J showed almost 1.8-fold higher binding in FC of AD brain than control when brain homogenates were used. Competition studies with unlabelled-UCB-J showed one-binding site with nanomolar-affinity in FC of AD and control brains, complementing our saturation-data. In contrast to P2-fraction data, autoradiography with ³H-UCB-J showed unexpectedly higher binding in AD than control brain.

Conclusions

Our in vitro findings indicated off-target binding of UCB-J, if P2-fraction is not used. The ongoing regional distribution and other experiments will provide further understanding of UCB-J binding mechanism.

Aims

To compare the composition of synapses within the entorhinal cortex (EC) between non-pathological controls, Alzheimer’s disease (AD) and primary age-related tauopathy (PART).

Methods

We prepared synaptosomal fractions from the EC of brains without lesions (n=5), with AD (n=8) or PART (n=5) pathology and analyzed the fractions by LC-MS/MS. We normalized peptides using variance stabilizing normalization and to the total intensity of 15 core synaptic proteins. We used MSqRob to compare synaptic protein levels across groups by linear mixed effects regression. Random effects included peptide sequence/modification, run, post-mortem interval and age-at-death. We performed statistical enrichment tests to compare group differences in relative abundance of gene ontologies. All p-values were adjusted for false discovery rate.

Results

We identified 3473 proteins with at least 1 proteotypic peptide. 5 proteins were under-expressed and 44 proteins were over-expressed in AD compared to controls (adj.p<.05). No individual protein was altered between PART and controls (all adj.p>.15). 3 proteins were under-expressed in PART compared to AD (adj.p<.03). Compared to controls, both AD and PART synaptosomes were under-enriched for pathways including vesicle-mediated transport, signal transduction and nervous system development and over-enriched for pathways including mitochondrial transmembrane transport, ATP synthesis and aerobic respiration (adj.p<.05). PART synapses were specifically under-enriched for pathways related to actin filament polymerization and over-enriched for RNA processing. AD synapses were specifically under-enriched for endopeptidase activity and endosomal sorting and over-enriched for mitochondrial gene expression. No ontologies were differentially enriched between PART and AD synaptosomes (adj.p>.05).

Conclusions

AD and PART pathologies affect similar synaptic processes in the EC.

Aims

To investigate the translational potential of CSF Ng for synaptic alterations using the App knock-in AD mice.

Methods

CSF and brain tissue were sampled from 12-months old wild type (WT), App^NL-F/NL-F and App^{NL-G-F/NL-G-F} mice. Ng, Aβ42, and t-tau levels were analyzed using Simoa. UCB-J PET was performed using [18F]SynVesT-1 and nanoPET. Amplitude and frequency of miniature EPSCs were measured by patch clamp recordings of cornu ammonis 1 exitatory neurons.

Results

Ng and tau levels were significantly increased in CSF of App knock-in mice, most pronounced in App^NL-G-F mice. CSF Ng levels negatively correlated with Ng levels in the cortex and Aβ42 levels in the CSF and positively correlated with tau levels both in the CSF and brain tissue. IHC staining revealed a substantial reduction of Ng around Aβ plaques. Frequency of the miniature EPSCs and UCB-J PET signal is similarly increased in App^NL-G-F mice.

Conclusions

App knock-in mice exhibit Aβ, tau and Ng alterations in the CSF and the interdependencies between these pathologies were strikingly similar to the pattern previously observed in AD patients. The changes in the mouse CSF biomarker readout were concurrent with synaptic impairments shown by increased UCB-J PET signal and hyperexcitability as shown by increased synaptic firing frequency. Hence, App knock-in mice and their CSF biomarkers have translational value for investigating synaptic alterations due to AD-related pathologies.

Aims

Physiopathological cleavage of the amyloid precursor protein (APP) allows the secretion of Aβ but also of other peptides, including peptides secreted from the η-secretase pathway, named Aη peptides, described in Willem et al., Nature, 2015. These Aη peptides acutely impair neuron function even at low nanomolar concentrations (Mensch et al. Alzheimers Res Ther., 2021). Here we analysed the consequences of chronic in vivo alterations of Aη peptide levels on hippocampal function.

Methods

To modulate Ah levels in vivo, two new mouse models were created. The first model, named MISEPA2, harbours a transgene expressing a secreted from of Aη-α allowing for a chronic increase of human Aη-α levels in the brain. The second model, named APPΔη, harbours a deletion of a portion of endogenous APP that prevents η-secretase processing. We characterized these models by analysing APP peptides levels by immunoblotting, analysing synaptic plasticity by electrophysiology and analysing spatial memory by submitting the mice to the Morris water maze.

Results

MISEPA2 mice exhibit normal levels of Aβ in the hippocampus. Long-term potentiation (LTP), long-term depression (LTD) and spatial memory are altered in these mice.

APPΔη mice exhibit normal levels of Aβ in the hippocampus. LTP is normal, but LTD is absent and rescued upon acute application of synthetic Aη-α peptide. Spatial memory is impaired in these mice.

Conclusions

These results show that a chronic increase of Aη-α is detrimental to hippocampal function. We also validate the hypothesis that the peptides secreted from the η-secretase pathway are necessary for endogenous modulation of synaptic communication and memory processing.

Aims

Alzheimer Disease (AD) is a neurodegenerative disorder characterized by impaired protein homeostasis. Autophagy is a major clean up system which is impaired in AD leading to an increase of autophagosomes in dystrophic neurites. The aim of this study is to understand the development of autophagy impairment associated with Aβ deposition and the link between autophagy impairment and other AD related pathologies.

Methods

We performed RNA-sequencing analysis of hippocampus from App knock-in mouse models at three different ages to compare and evaluate the early as well as the more advanced stages of neuropathological changes. Autophagy flux was measured in the crude synaptosomal fractionations of hippocampus, and autophagosome accumulation was investigated by immunofluorescence (IF) staining and electron microscopy. We also performed IF staining of tau in autophagy-deficient mouse brain.

Results

We sequenced approximately 46,000 transcripts. Pathway analysis revealed that pathways regulating autophagy were slightly up-regulated before onset of pathology, and it gradually decreased at late stage in App^NL-G-F mice mimicking the processes present in AD brain. We found an accumulation of autophagosomes in 12-months-old App^NL-G-F mice but not in 2-months-old. Interestingly, swollen synapses containing autophagosomes were observed in 12-months-old App^NL-G-Fmice. It may indicate a link between autophagy impairment and synaptic dysfunction. We also revealed phosphorylated tau accumulation in hippocampus of autophagy-deficient mice.

Conclusions

We revealed autophagy impairment in App knock-in mouse models. Autophagy impairment may have a link with other AD related pathologies such as synaptic dysfunction and tau pathology.

Aims

Methods

Results

Conclusions