Transcriptomic Profiling Reveals Discrete Poststroke Dementia Neuronal and Gliovascular Signatures

• Published in: Translational stroke research Vol. 14; no. 3; pp. 383 - 396
• Main Authors: Waller, Rachel, Hase, Yoshiki, Simpson, Julie E., Heath, Paul R., Wyles, Matthew, Kalaria, Rajesh N., Wharton, Stephen B.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2023; Springer Nature B.V
• Subjects: Age; Animals; Biomedical and Life Sciences; Biomedicine; Brain; Brain - pathology; Cardiac arrhythmia; Cardiology; Cognitive ability; Dementia; Disease; Endothelial Cells - pathology; Executive function; Heart failure; Humans; Ischemia; Metastasis; Mice; Neurology; Neurons - pathology; Neurosciences; Neurosurgery; Original; Original Article; Pathology; Peritonitis; Pneumonia; Proteins; Stroke; Stroke - complications; Stroke - genetics; Stroke - pathology; Transcriptome; Vascular Surgery; United Kingdom > UK; Vascular dementia; Neurons; Gliovascular unit; Poststroke dementia; Transcriptome
• ISSN: 1868-4483; 1868-601X; 1868-601X
• DOI: 10.1007/s12975-022-01038-z

Poststroke dementia (PSD) is associated with pathology in frontal brain regions, in particular dorsolateral prefrontal cortex (DLPFC) neurons and white matter, remote from the infarct. We hypothesised that PSD results from progressive DLPFC neuronal damage, associated with frontal white matter gliovascular unit (GVU) alterations. We investigated the transcriptomic profile of the neurons and white matter GVU cells previously implicated in pathology. Laser-capture microdissected neurons, astrocytes and endothelial cells were obtained from the Cognitive Function After Stroke cohort of control, PSD and poststroke non-dementia (PSND) human subjects. Gene expression was assessed using microarrays and pathway analysis to compare changes in PSD with controls and PSND. Neuronal findings were validated using NanoString technology and compared with those in the bilateral common carotid artery stenosis (BCAS) mouse model. Comparing changes in PSD compared to controls with changes in PSND compared to controls identified transcriptomic changes associated specifically with dementia. DLPFC neurons showed defects in energy production (tricarboxylic acid (TCA) cycle, adenosine triphosphate (ATP) binding and mitochondria), signalling and communication (MAPK signalling, Toll-like receptor signalling, endocytosis). Similar changes were identified in neurons isolated from BCAS mice. Neuronal findings accompanied by altered astrocyte communication and endothelium immune changes in the frontal white matter, suggesting GVU dysfunction. We propose a pathogenic model in PSD whereby neuronal changes are associated with frontal white matter GVU dysfunction leading to astrocyte failure in supporting neuronal circuits resulting in delayed cognitive decline associated with PSD. Therefore, targeting these processes could potentially ameliorate the dementia seen in PSD.