First Scientific Session - Cerebrovascular Disease, Brain Tumours and Traumatic Brain Injury

Background: Stroke doubles the risk of dementia in the over 65s and around a quarter of stroke survivors develop progressive cognitive decline more than 3 months after the stroke, eventually leading to vascular dementia. Impaired executive function is a characteristic feature of vascular dementia, including post stroke dementia (PSD). PSD is associated with pathology of the anterior cognitive circuits, but the mechanisms leading to PSD are unclear. This study aims to find novel pathways by which cell damage in the brain can give rise to PSD.

Methods: Dorsal lateral prefrontal cortex (DLPFC) of 10 control, 10 PSD, and 10 post-stroke non-demented (PSND) cases from the Cognitive Function After STROKE (CogFAST) study were subjected to laser capture microdissection to obtain RNA enriched for; astrocytes, endothelium and neurons. Microarray analysis was performed to identify differentially expressed genes in cortical neurons and the frontal white matter cells of the gliovascular unit.

Results: We identified altered gene expression associated with PSD. Pathway analysis demonstrated a decrease in communication and signalling pathways in astrocytes, whilst endothelium showed an increase in inflammatory pathways and altered metabolism. Neurons showed reduction in pathways for metabolism and endocytosis. These changes will be confirmed at both the cellular (immunohistochemistry) and molecular (NanoString) level.

Conclusions: Within the anterior cognitive circuit, cortical neurons and cells of the gliovascular unit showed alterations in their ability to function normally and communicate with each other. These findings provide new information about how stroke can lead to dementia, and may identify novel targets for future studies.

Type 2 diabetes mellitus (T2D) is a risk factor for dementia. It may act partly through vascular and Alzheimer’s pathologies. However, whether diabetes may directly damage cells of the brain is not well-understood. We sought to identify transcriptomic changes in cortical neurones to discover alterations in neuronal cell pathways. We also examined the transcriptomes of astrocytes and endothelial cells. Neurones, astrocytes, and endothelial-rich samples were isolated using laser capture microdissection from temporal cortex of six cases with self-reported T2D and six without from the Cognitive Function and Ageing Study neuropathology cohort. Differentially expressed genes (DEGs) were assessed by microarray analysis and altered cellular pathways determined using Integrated Molecular Pathway Level Analysis. Gene expression changes in neurones were validated by a second method, the NanoString nCounter platform. Neurones showed 912 DEGs between samples from diabetic and non-diabetic individuals. Cell pathway changes included alterations in insulin signalling, cell cycle, cellular senescence, inflammatory mediators, and components of the mitochondrial respiratory electron transport chain. There were 2,202 and 1,227 DEGs in astrocytes and endothelial cells respectively. There was little overlap in individual DEGs in different cell types, but pathway analysis demonstrated that impaired insulin signalling was shared by neurovascular unit cells. Additionally, apoptotic pathway changes in astrocytes and dysregulation of advanced glycation end-product signalling in endothelial cells were identified. Transcriptomic analysis identified changes in key cellular pathways associated with T2D that may contribute to neuronal damage and dysfunction. These alterations potentially contribute to a diabetic dementia, and may provide novel avenues for investigation.

Glioblastoma (GBM) is the commonest primary brain tumour in adults.  Approximately a third of the cells in GBM are microglia/macrophages rather than neoplastic cells, with evidence supporting a role for immune cells in GBM development.   Here, we aim to characterise the complexity of the inflammatory environment.

Tissue-microarrays were prepared from 60 GBM, IDH wildtype (WHO grade IV) biopsies from adult patients, representing the tumour core, infiltrating zone and relatively normal grey matter (leading edge of the tumour).  Immunostaining for homeostatic microglia (Iba1, P2Y12), phagocytic microglia (CD68), microglial antigen-presentation (HLA-DR), Fcg receptors (CD64, CD16, CD32a), macrophages (CD163, CD206, TREM2), T cells (CD3, CD4, CD8), and NK cells (NKp46) has been performed.  The analysis is currently ongoing and will include relevant clinical information. 

Iba1, HLA-DR and CD68 showed a significantly higher expression in the core vs. leading edge (p<0.001) and in the infiltrating zone vs the leading edge (p<0.003).  CD163 expression was highest in the core, followed by the infiltrating zone (p<0.002). CD206 was significantly higher in the core vs leading edge (p=0.008).  Within the different areas, correlations were observed between the microglial markers and CD163; while CD206 was only associated with CD163.  

Our initial findings confirm the presence of a substantial immune response to the neoplastic cells of a GBM both by resident microglia and potentially by macrophages with a haematogenous origin.  CD206 identified a subpopulation of CD163+ macrophages within the tumour implying a different role of microglia and macrophages in the response to GBM. 

The accumulation of βAPP caused by axonal injury is an active energy dependant process thought to require blood circulation; therefore, is closely related to the post-injury survival time. Currently, the earliest reported time at which axonal injury can be detected in post mortem traumatic brain injury tissue by βAPP immunohistochemistry is 35 minutes. In this study we investigate whether βAPP staining for axonal injury can be detected in patients with near instantaneous/rapid death.

Forty-nine deaths following road traffic collision (RTC) where we have reliable information concerning the time between the incident and death, were retrospectively studied. Thirty seven patients (group 1) died virtually instantaneously or very rapidly at the scene. A further three died between 30 minutes to 11 hours (group 2) and eight between 2 and 31 days (group 3). Brains from four patients who died instantaneously due to sudden unexpected death in epilepsy were used as a non-traumatic control group. The brains were comprehensively examined, and βAPP immunohistochemistry carried out on sections from a number of brain areas

βAPP immunoreactivity was demonstrated in in 35/37 brains in group 1, albeit with a low frequency and in a variable pattern and with more intensity and frequency in all brains of group 2 and 7/8 brains from group 3, compared with no βAPP immuno reactivity in any of the control group.

The results suggest axonal injury can be detected even in those who died near instantaneously following fatal road traffic collision.