Abstract Body

Animal models of late-onset Alzheimer’s Disease are valuable resources to understand the pathogenesis of Alzheimer’s Disease (AD). For the past few decades, mouse models that harbor autosomal dominant mutations found in familial AD and other dementias have been the gold standard for the field. These models have been proven invaluable for understanding key features of AD pathology and for helping to evaluate and discover new therapeutic targets and disease-modifying strategies. As the field advances, so too must our development of animal models, with particular emphasis being focused on generating models of late-onset AD. The goal of the MODEL-AD consortium is to produce, characterize and make available the next generation of animal models for AD through comprehensive and systematic analysis (deep-phenotyping), and to standardize a phenotyping pipeline to understand the strength of each model. The Consortium is led by NIA, and our ongoing collaborative efforts take place at different locations across the US (University of California, Irvine, Jackson laboratory, University of Pittsburgh, Indiana University and Sage Bionetworks). Analyses include functional phenotyping, gene expression and network analysis, pathology, and biochemical analysis to obtain comprehensive results, which will be openly accessible to the scientific community through the AD Knowledge Portal.

Abstract Body

We have tested several human tau mutation variants and the common variant (wild type) for their capacity to produce tauopathy in middle aged mice. Twelve month old C57BL/6Nia mice were injected with AAV serotype 9 vectors expressing either wild type human tau, P301L tau, R406W tau or GFP (control) into the anterior cortices and hippocampus bilaterally. Mice injected with tau variants also were injected with a 1/10^th dose of the GFP virus to control for injection variability. Both P301L and wild type injected mice exhibited memory deficits in radial arm water maze performance 4 months later. Detergent soluble total tau was lower in the R406W mice than the other tau variant injected groups, yet GFP expression was elevated in this group, suggesting this variant impaired tau expression. Insoluble total tau was greatest in the P301L injected group. Soluble p-199-tau and p-396-tau were highest in the wild type mice, while insoluble p-tau and Gallyas staining were highest in the P301L mice. Remarkably, the only group demonstrating hippocampal atrophy was the wild type injected mice, with a mean 40% reduction in hippocampal volume. These data suggest that even though P301L increased insoluble, silver positive tau deposits, wild type tau with large amounts of soluble tau isoforms was more toxic to the hippocampus. Supported by AG 051500 to DM and AG062217 to MNG

Abstract Body

Objective: Complement C3 is an innate immune molecule that plays an important role in synapse elimination. Previously, we reported that germline deletion of complement C3 (C3KO) protected against hippocampal decline in aged C57BL/6 and APP/PS1dE9 mice. Here, we generated global and cell-specific C3 inducible conditional knockout mice.

Methods: We generated and crossed C3 floxed (C3fl/fl) mice to Rosa26-Cre-ERT2+/- mice to create a novel global C3 inducible conditional knockout (C3iKO) mouse model. Adult mice were treated with 5 daily i.p. injections of 75 mg/kg tamoxifen (TAM) or corn oil (control). C3 protein levels were quantified in serum (ELISA) and brain (Western blot) at various timepoints. Complement C3 mRNA expression was quantified by qPCR in the liver and brain post-treatment. Currently, we are crossing the C3fl/fl mice with Cx3cr1-Cre-ERT2 (YFP) mice to lower C3 in microglia (C3-mg-iKO) and Aldh1l1-Cre-ERT2 (BAC) mice to lower C3 in astrocytes (C3-as-iKO).

Results: Tamoxifen treatment of C3iKO mice resulted in reduced C3 levels in serum (~85-90%) and brain (~60%) at day 150. C3 mRNA was significantly reduced in liver and brain of C3iKO+TAM mice at 60 and 150 days. The cell-specific models are underway and thus far, show recombination in brain.

Conclusions: Our novel C3iKO mouse model shows long-term C3 lowering. We will cross these mice to amyloid and tau mouse models to determine if C3 lowering is protective during early-stage AD pathogenesis. Our novel cell-specific models will allow us to further delineate C3 signaling in brain.

Funding: NIH R21 AG044713 (CAL, MC); NIH RF1 AG060057 (CAL)

Aims

Despite the assumed key role of Aβ for Alzheimer´s pathogenesis, clinical trials directed at reducing Aβ failed to substantially modify the course of the disease. Thus, it now appears crucial to understand AD pathogenesis in the context of APP physiological functions. However, the early postnatal lethality of germline triple knockout mice has so far precluded the analysis of APP family physiological functions in the developing and adult brain.

Methods

Here, we generated conditional APP/APLP1/APLP2 triple KO (cTKO) mice lacking the APP family in excitatory forebrain neurons from E12 onwards using NexCre mice.

Results

Conditional cTKO mice showed altered brain morphology with agenesis of the corpus callosum and impaired lamination of the hippocampus. In addition, electrophysiological recordings in the hippocampus of cTKO mice revealed impaired basal synaptic transmission and severely reduced long-term potentiation that was associated with reduced spine density of hippocampal neurons. At the behavioral level cTKO mice were not only severely impaired in several tests for learning and memory, but also exhibited autism-like behaviors including stereotypic repetitive behaviors, reduced social communication and impaired social interaction. Further, in vivo two-photon imaging of spontaneous Ca²⁺ transients revealed that behavioral impairments were associated with impaired network activity in the cortex.

Conclusions

Together, our study identifies essential functions of the APP family during brain development and for hippocampal and cortical networks in the adult brain, important to consider for future therapeutic strategies.

Aims

Anesthesia in the elderly has been associated with subsequent cognitive decline. Although the mechanisms for this phenomenon are unclear, acute anesthesia exposure has been shown to increase tau hyperphosphorylation in rodent models. In order to further interrogate this relationship, we examined the long-term effects of anesthesia exposure on mouse models of tauopathy.

Methods

We exposed P301L, COMTKO/P301L, and hTau mice to isoflurane for two-hours without temperature control. In addition to ELISA and immunocytochemistry, we assessed gene expression changes in the brain using mRNA-seq in order to evaluate their association with tau hyperphosphorylation and pathology. In the hTau mouse, we examined spatial learning and memory of mice using the Barnes maze at baseline and 1-month and 7-months post-anesthesia.

Results

Across all three mouse models, the levels of phosphorylation at Thr231, Ser202, and Ser396/404 in the soluble fractions increased in the anesthesia group. However, 24-hours post-anesthesia, we did not observe a difference. At one-month post-anesthesia, we found elevated Thr231 hyperphosphorylation in the insoluble fraction in all three models. According to differential expression analysis, short- and long-term effects of acute hyperphosphorylation were different between the hTau mice and the mice with mutant human tau. We did not observe any differences in spatial learning and memory 1- or 7-months post-anesthesia.

Conclusions

Anesthesia was associated with acute hyperphosphorylation of tau in each model of tauopathy. Additionally, anesthesia induced short- and long-term changes in gene expression that may be associated with tau hyperphosphorylation.

Aims

Alzheimer’s disease (AD) is characterized by the accumulation of aggregated Tau and Amyloid beta in the brain. While autophagy is a crucial mechanism for the clearance of aggregated proteins in healthy cells, this process is impaired in AD. Therefore, pharmacological modulation of autophagy is of high interest as a potential disease-modifying therapy for AD. Due to its central role in autophagosome biogenesis, the protein LC3B is an excellent biomarker candidate to monitor autophagy.

Methods

The presence of different LC3B forms in human CSF was investigated by Western Blot. Quantification of total LC3B in CSF from an AD cohort with attached clinical and biomarker data and from Thy1-hTau.P301S mice was performed with an immunoassay developed and validated in house using the Singulex platform.

Results

We show that the cytoplasmic form LC3B-I is dominating in human CSF, whereas the membrane-bound form LC3B-II is likely present in low amounts. Quantification of total LC3B revealed significantly higher levels in the CSF of AD patients (n=18) compared to healthy controls (n=22) and Abeta-negative patients with cognitive impairment (n=30). Interestingly, LC3B strongly correlates with total tau in the CSF. Transgenic Thy1-hTau.P301S mice also showed an increase of LC3B in the CSF that is strongly correlating with pathological Tau inclusions in the brain.

Conclusions

Our data suggests a link between elevated levels of LC3B in the CSF and tau pathology. Due to the central role of LC3B in autophagosome biogenesis, its elevation in the CSF has the potential to indicate impaired autophagic processes in AD patients.

Aims

Mouse models of Alzheimer’s disease (AD), carrying rare variants in genes such as APP and PS1 (APP/PS1), have usually been created on the C57BL/6J (B6J) genetic background. While these strains often exhibit amyloid accumulation and neuroinflammation, many additional molecular alterations present in human AD are absent. To broaden the phenotypes of mouse models, we introduced genetic diversity by incorporating Collaborative Cross (CC) lines, a recombinant inbred mouse panel created from eight highly diverse founder strains.

Methods

Five CC strains were selected for maximal genetic and gene expression variation at twelve late-onset GWAS loci, including TREM2, BIN1, and CLU. Transgenic APPand PS1alleles with a humanized APOE4 allele on a B6J background were crossed with each CC line. Brain hemisphere transcriptomes and neuropathology were assessed at 8-months. Neuropathology focused on amyloid deposition, glial cell activation and neuronal health.

Results

The effect of humanized APOE4 demonstrated differences across CC lines and in the presence of mutant APP and PS1 transgenes. RNA-Seq data revealed allele-specific gene expression profiles associated with neuropathological differences. We mapped strain specific transcriptional signatures to Late-Onset AD subtypes identified in the study cohorts from AMP-AD consortium and observed correlations with subtypes specific to APP/PS1 and APOE4 alleles.

Conclusions

The findings provide new insights into the role of APOE4 in amyloid pathogenesis. Diverse genetic backgrounds of CC lines exhibit a unique resource to assess genome-wide allele-specific gene expression connecting AD risk variants to molecular and neuropathological profiles.This study suggests use of CC lines mouse models to better represent the genetic variation in AD.