Abstract Body

Objectives:

The beta-site APP cleaving enzyme (BACE1) is known for its “amyloidogenic” activity which leads to the production of Abeta40 and Abeta42 peptides. We recently discovered a novel “amyloidolytic” activity for BACE1 whereby it degrades longer Abeta peptides into a common, non-toxic Abeta34 intermediate (Liebsch et al., 2019). Our current work examined whether altered BACE1 activity shifts the equilibrium between Abeta production (Abeta40, Abeta42) and clearance (Abeta34), and how the generation of certain N-terminally truncated Abeta peptides is influenced.

Methods:

Pericytes are vascular mural cells within the neurovascular unit, essential for the blood brain barrier and able to metabolize amyloid peptides (Kirabali et al., 2019). Thus, we compared pericytes with neuroblastoma cells (SH-SY5Y) and organotypic brain slice cultures to analyze the production and clearance of Abeta species by immunoprecipitation – mass spectrometry (IP-MS) and our ultra-sensitive multiplexing assay (custom MSD).

Results:

In the presence of surplus BACE1, longer Abeta peptides were processed via the “amyloidolytic” pathway, yielding elevated levels of Abeta34. N-terminally truncated peptides (Abeta5-x) were detected when BACE1 activity was reduced; the truncated species can aggregate and gain toxic properties in combination with other Abeta peptides.

Conclusions:

Our new findings caution that clinical trials that have utilized BACE1 inhibitors may have inadvertently attenuated (i) the favourable degradation of toxic Abeta species into non-toxic Abeta34 and (ii) increased the generation of N-terminally truncated Aβ peptides. The influence of truncated Abeta5-x species on amyloid seeding, aggregation, and toxicity is not well understood and represents an exciting new area of APP biology.

Aims

Several phase III clinical trials with BACE1 (beta-secretase) inhibitors were terminated due to occurrence of side-effects, most notably cognitive decline and psychiatric symptoms. Side effects are assumed to result from loss-of-cleavage of BACE1 substrates with essential CSN functions, but it remains largely unclear which of the many BACE1 substrates are particularly relevant in primates/humans, whether their cleavage-inhibition correlates with side-effects and how their function may be altered upon BACE1 inhibition. Here, we identified BACE1 substrates from primary neurons and from CSF of humans and non-human primates treated with a BACE1 inhibitor.

Methods

Substrate levels upon BACE1 inhibition were determined by nLC-MS/MS. Functional consequences of BACE1 substrate cleavage were analyzed using primary neurons.

Results

Across species, BACE1 inhibition reduced cleavage of several new and known BACE1 substrates in a time- and dose-dependent manner, including SEZ6, SEZ6L, CACHD1, CD200, VCAM1 and a cytokine receptor. Some substrates, e.g. SEZ6, were exclusively cleaved by BACE1 and paralleled the reductions of Abeta and sAPPbeta. We further validated CACHD1, CD200 and the cytokine receptor as new BACE1 substrates in vitro and demonstrate that cytokine receptor signaling in primary neurons is altered upon BACE1 inhibition or deficiency.

Conclusions

This study identifies and mechanistically characterizes new and known in vivo-relevant BACE1 substrates and demonstrates that SEZ6 and additional BACE1 substrates are suitable pharmacodynamic CSF biomarkers to measure BACE1 activity in vivo. Application of these biomarkers may help to understand and prevent the occurrence of side-effects observed in the phase III studies with BACE1 inhibitors.

Aims

Amyloid precursor protein (APP) is associated with both familial and sporadic forms of Alzheimer’s disease. Despite its importance, the role of APP family in neuronal function and survival remains unclear due to perinatal lethality exhibited by knockout mice lacking all three APP family members. We aim to determine whether APP family is required for cortical neuronal survival during aging and whether APP family plays an essential role in the regulation of neuronal function.

Abstract Body

Aims

Amyloid precursor protein (APP) is associated with both familial and sporadic forms of Alzheimer’s disease. Despite its importance, the role of APP family in neuronal function and survival remains unclear due to perinatal lethality exhibited by knockout mice lacking all three APP family members. We aim to determine whether APP family is required for cortical neuronal survival during aging and whether APP family plays an essential role in the regulation of neuronal function.

Methods

We developed triple conditional knockout mice, in which APP, APLP1 and APLP2 are selectively inactivated in excitatory neurons of the postnatal forebrain. We then used molecular, biochemical, histological, behavioral and electrophysiological analyses to determine the consequences of APP family inactivation in the cerebral cortex.

Results

We found that selective inactivation of APP family members in excitatory neurons of the postnatal forebrain results in neither cortical neurodegeneration nor increases in apoptosis and gliosis up to ~2 years of age. However, hippocampal synaptic plasticity, learning and memory are impaired in these mutant mice. Furthermore, hippocampal neurons lacking APP family exhibit hyperexcitability, as evidenced by increased neuronal spiking in response to depolarizing current injections, whereas blockade of Kv7 channels mimics and largely occludes the effects of APP family inactivation.

Conclusions

These findings demonstrate that APP family is not required for neuronal survival, and suggest that APP family may regulate neuronal excitability through Kv7 channels. I will also present our latest findings on pathogenic mechanisms underlying FAD mutations in Presenilin-1.

Methods

We developed triple conditional knockout mice, in which APP, APLP1 and APLP2 are selectively inactivated in excitatory neurons of the postnatal forebrain. We then used molecular, biochemical, histological, behavioral and electrophysiological analyses to determine the consequences of APP family inactivation in the cerebral cortex.

Results

We found that selective inactivation of APP family members in excitatory neurons of the postnatal forebrain results in neither cortical neurodegeneration nor increases in apoptosis and gliosis up to ~2 years of age. However, hippocampal synaptic plasticity, learning and memory are impaired in these mutant mice. Furthermore, hippocampal neurons lacking APP family exhibit hyperexcitability, as evidenced by increased neuronal spiking in response to depolarizing current injections, whereas blockade of Kv7 channels mimics and largely occludes the effects of APP family inactivation.

Conclusions

These findings demonstrate that APP family is not required for neuronal survival, and suggest that APP family may regulate neuronal excitability through Kv7 channels. I will also present our latest findings on pathogenic mechanisms underlying FAD mutations in Presenilin-1.

Abstract Body

Aims

The majority of the early cases of AD genetic cases relate to the presenilin 1 (PS1). PS1 is considered important determinants of the γ-secretase catalytic site that process amyloid precursor protein (APP) to Aβ toxic isoforms and is essential for activation of signaling important for cell activity such as NOTCH. Previous work suggested that mutation in PS1 results in reduction in gamma secretase activity. Microglia activation in AD is consider double sword suggesting that in one case impair in clearing amyloid load and another case increase in neuronal pruoining might lead to neurodegeneration. We discovered that deficiency in PS1 reduce microglia activity in migration and uptake of Abeta1-42 (Farfara et al ANN 2011). We further aim to assess the effect of pathological changes in PS1 on microglia activity.

Methods

We target the role of PS1 in microglia activity both in cell culture and in animal model.

Results

We found that Alzheimer's mutated PS1 in microglia increase secretion of pro-inflammatory cytokines and production of reactive oxidative species (ROS). Furthermore, mutated PS1 microglia show significantly reduction in uptake of Abeta 1-42 that is affiliated with impair in expressing of scavenger receptor and neurotoxicity.

Conclusions

Further research on the pathways that pathological changes within PS1 affect microglia activity may lead to identify new candidate targets for therapeutic intervention in Alzheimer's disease and related neurodegenrative diseases.

Aims

Axonal transport of diverse cargoes is essential for neuronal function and viability. Defects in transport have been implicated as an early event in many neurodegenerative diseases including Alzheimer’s disease (AD). Axonal transport is likely to be highly coordinated and regulated since many cargoes must be selected, packaged, and associated with specific motors (kinesin-1 for anterograde and dynein for retrograde) for efficient simultaneous transport on microtubules (MTs). One possible mechanism by which transport defects could occur is by improper regulation of motors.

Methods

Indeed, excess GSK3β causes axonal blockages by increasing the binding of motors to cargo. Presenilin (PS), the protein implicated in early-onset familial AD, also influences motor function, with functional PS and the catalytic loop region of PS being essential for the rescue of GSK3β-mediated transport defects.

Results

Intriguingly, while active GSK3β phosphorylates the kinesin-1 motor domain, the phospho-defective Crispr-Cas-9 motor domain mutant decreases membrane binding and impairs coordinated motility by reducing ATP activity, but did not abolish MT binding. In contrast, the phospho-mimicking kinesin-1 mutant completely abolished motility. Strikingly, active GSK3β, kinesin-1, and PS form a complex.

Conclusions

Together our observations propose a scaffolding mechanism for PS where the loop region sequesters GSK3β away from motors for the proper regulation of motor function by differential phosphorylation, demonstrating how PS together with GSK3β fine-tunes kinesin-1 motor activity in vivo. These findings highlight the complex regulatory mechanisms that likely exist in vivo, and reveals how disruption of these processes can lead to neuronal dysfunction and disease.

Aims

Innate immunity is associated with Alzheimer’s disease, but the influence of immune activation on the production of amyloid-beta is unknown.

Methods

We used photolableing using a GSM followed by mass spectrometry to identify gamma-secretase modulatory proteins.We used aging mouse brains, 5xFAD alzheimer transgenic mice, primary neurons, astrocytes, and late-onset human brains to investigate the effect of the innate immunity protein IFITM3 on gamma-secretase and its activity.

Results

We identify interferon-induced transmembrane protein 3 (IFITM3) as a gamma-secretase modulatory protein, and establish a mechanism by which inflammation affects the generation of amyloid-beta. Inflammatory cytokines induce the expression of IFITM3 in neurons and astrocytes, which binds to gamma-secretase and upregulates its activity, thereby increasing the production of amyloid-beta. The expression of IFITM3 is increased with ageing and in mouse models that express familial Alzheimer’s disease genes. Furthermore, knockout of IFITM3 reduces gamma-secretase activity and the formation of amyloid plaques in a transgenic mouse model (5xFAD) of early amyloid deposition. IFITM3 protein is upregulated in tissue samples from a subset of patients with late-onset Alzheimer’s disease that exhibit higher gamma-secretase activity. The amount of IFITM3 in the γ-secretase complex has a strong and positive correlation with gamma-secretase activity in samples from patients with late-onset Alzheimer’s disease.

Conclusions

These findings reveal a mechanism in which gamma-secretase is modulated by neuroinflammation via IFITM3 and the risk of Alzheimer’s disease is thereby increased.

Aims

Production of the amyloid-beta peptide (Aβ) of Alzheimer's disease is carried out by the membrane-embedded γ-secretase complex. Mutations in the transmembrane domain of the amyloid precursor protein (APP) associated with autosomal-dominant early-onset familial Alzheimer's disease (FAD) alter the ratio of aggregation-prone 42-residue Aβ (Aβ42) to 40-residue Aβ (Aβ40). However, APP is processively processed by γ-secretase along two pathways: Aβ49→Aβ46→Aβ43→Aβ40 and Aβ48→Aβ45→Aβ42→Aβ38. The effects of FAD mutations on each proteolytic step are unknown, largely due to the difficulty in detecting the longer Aβ peptides. To address this problem, we carried out a systematic and quantitative analysis of all the tri- and tetra-peptide products from wild-type and 14 different FAD-mutant APP substrates by γ-secretase (8 X 15 = 120 cleavage events) in a purified biochemical system.

Methods

Small peptide products were detected by LC/MS/MS, establishing standard curves for each, including all FAD-mutant peptides. APP intracellular domain (AICD) products were quantified by western blot, and the ratio of AICD products corresponding to Aβ48 and Aβ49 were determined by MALDI-TOF MS. Levels of individual Aβ peptides were determined by subtracting small peptides associated with degradation from those associated with production. This method was validated for Aβ40 and Aβ42 by specific ELISAs, which gave remarkably consistent results.

Results

Not all FAD-mutant APP substrates led to increased Aβ42/40. However, every disease-causing mutation led to inefficient processing of intermediate forms of Aβ ≥ 45 residues.

Conclusions

These findings point to deficient trimming of elusive long Aβ peptides—dubbed here “dark amyloid”—as potentially pathogenic in FAD.

Aims

Long amyloid beta peptides act as both substrates and products in gamma-secretase-mediated proteolysis. Our studies tested the hypothesis that these peptides can compete and displace the amyloid precursor protein C-terminal fragment (APP-CTF) from binding to gamma-secretase. In particular, we hypothesized that long amyloid beta peptides, when present at the high concentrations in Alzheimer's disease (AD) brain, inhibit gamma-secretase, leading to the accumulation of unprocessed gamma-secretase substrates. The accumulation of unprocessed substrates may have pathophysiological implications for AD progression.

Methods

We performed rigorous gamma-secretase kinetic analysis in detergent-based (purified gamma-secretase) or native-like cell-free conditions (detergent resistant membranes) in the presence of 3xFLAG-tagged gamma-secretase substrates and amyloid beta peptides. The enzyme activity was estimated by western blotting and mass spectrometry. In addition, analysis of APP-CTFs in neuronal cell lines exposed to amyloid beta peptides was performed.

Results

Kinetic evaluation of gamma-secretase activity demonstrated an amyloid beta-mediated feedback inhibition of the processing of multiple substrates by gamma-secretase. Interestingly, human amyloid beta 1-42, but not the analogous human (17-42) p3 peptide, exerted inhibition. The pathophysiological relevance of these observations is supported by the accumulation of APP-CTFs in neuronal cell lines exposed to amyloid beta 1-42 and not to p3 (17-42) peptides.

Conclusions

Feedback inhibition is exerted by pathologically relevant amyloid beta peptides on gamma-secretase activity, giving evidence for a novel mechanism for amyloid beta 42-mediated toxicity in AD that also helps to explain the accumulation of unprocessed APP-CTFs in AD-linked conditions.