The role of the blood-brain barrier in a mouse model of Alzheimer's disease

The blood-brain barrier (BBB) protects the brain microenvironment from external damage. It is formed by endothelial cells (ECs) lining the brain vessels, expressing tight junctions and having reduced transcytosis, resulting in a very low paracellular and transcellular passage of substances, respectively (low permeability). The specific BBB phenotype is maintained by Wnt molecules secreted by astrocytes (ACs) that bind to receptors in ECs, and start a molecular cascade that leads to β-catenin translocating to the nucleus and activating the transcription of BBB genes. An increasing number of studies report BBB dysfunction in Alzheimer’s disease (AD), although the topic is currently under debate. AD is a neurodegenerative condition characterized by brain depositions of Aβ aggregates and Tau neurofibrillary tangles. The aetiology of AD is unknown, although round 5% of all AD cases have a genetic origin. Mutations in APP or PSEN1/2 can lead to Aβ over-production and accumulation, causing familiar AD. There is no cure for AD, as all clinical trials failed during the past years. Consequently, I studied the role of the BBB in AD, aiming to investigate if a BBB dysfunction occurs in AD, and to identify by transcriptomic analysis novel gene regulations happening at the BBB in AD. The final objective was to evaluate the potential of identified BBB genes as therapeutical target. I used transgenic mice expressing the human APP mutations Swiss, Dutch and Iowa under the control of the neuronal promoter Thy1 (Thy1-APPSwDI) as AD model. In this AD mouse model, I could detect Aβ deposits and memory loss by immunofluorescence (IF) and behavioural tests. Importantly, I identified an increase of BBB permeability to 3-4 kDa dextrans in 6 months, 9-12 months, and 18 months or older AD mice compared to age-matched control wild types (WT), indicating BBB dysfunction in AD mice. In order to study the BBB transcriptional changes in AD, I sequenced the RNA from 6 and 18 months old AD and WT mouse brain microvessels (MBMVs), as well as of FACS-sorted ECs, mural cells (MuCs), ACs, and microglia (MG) in collaboration with GenXPro, a company specialized in 3’ RNA sequencing. Currently, no transcriptomic datasets of ECs and MuCs are publicly available, suggesting that this is the first study sequencing those cell types in the context of AD. The analysis of sequencing data from MBMVs and ECs revealed a Wnt/β-catenin repression, and an increase of inflammatory genes like Ccl3 in ECs, that could explain the BBB dysfunction observed in AD mice. Furthermore, the sequencing data from MuCs identified a set of 11 genes strongly regulated in both 6 and 18 month AD groups. Three of those 11 genes are known to be involved in inflammatory processes, demonstrating that inflammation affects and plays an important role in MuCs and ECs during AD. Thanks to published sequencing data, some up-regulated MG genes in AD are well known and recognized, such as Trem2 and Apoe. Those genes were found in the FACS-sorted MG data as well, validating the AD model and with it, the other novel sequenced datasets. Importantly, one of the most strongly AD-regulated genes in MBMV and MG samples was Dkk2, a member of the Dickkopf family of secreted proteins known to be involved in Wnt signalling modulation. Importantly, a dual luciferase reporter assay proved that Dkk2 is a Wnt inhibitor. A preliminary immunohistochemistry examination of DKK2 in human brain autopsy tissue from an AD patient and age-matched control revealed a stronger DKK2 immunoreactivity in the AD brain. In order to answer the question whether a rescue of BBB function would ameliorate AD symptoms, I made use of a tamoxifen-inducible transgenic mouse line to activate the Wnt/β-catenin pathway specifically in ECs, leading to a gain of function (GOF) condition (Cdh5-CreERT2+/–/Ctnnb1(Ex3)fl/fl). This mouse line was then crossed with the AD line, creating AD/GOF and AD/control groups. AD/GOF mice performed better in a Y-Maze memory test than AD/controls when the Wnt/β-catenin pathway was induced before AD onset, indicating a protective effect. Moreover, the finding implies that shielding BBB functioning in AD further protects the brain from AD toxic effects, suggesting an important role of brain vasculature in AD and its potential as therapeutic target