Characterising the astrocyte responses to oligomeric and fibrillary Aβ1-42 in Alzheimer's disease

Background: Alzheimer’s disease is a devastating disease, and the most common cause of dementia. Alzheimer’s disease is characterised by the formation of neurofibrillary tangles (NFTs) composed of tau protein, and senile plaques composed mainly of Aβ fibrils. Small and soluble Aβ oligomers have also been implicated in disease, however the exact mechanism of how Aβ oligomers and fibrils modulate Alzheimer’s disease pathology remains elusive. Astrocytes are the most abundant glial cells, which function to provide essential metabolic and trophic support for neurons. Astrocytes are implicated in Alzheimer’s disease and change early in disease. The mechanisms of the interplay of astrocytes and Aβ fibrils and oligomers in Alzheimer’s disease are unclear. Hypothesis: Different forms of Aβ1-42 (oligomers vs fibrils) elicit a varying response in astrocytes, driving astrocytes to a neurodegenerative or neuroprotective phenotypes in Alzheimer’s disease. Results: Stable monomeric Aβ1-42 preparations were generated in vitro and were confirmed using asymmetric flow field flow fractionation and transmission electron microscopy. Aβ1-42 was aggregated into oligomers and fibrils in vitro. The morphology of the preparations was confirmed using transmission electron microscopy, size exclusion chromatography with multi-angle light scattering, Thioflavin T assay, and immunoblotting. The highest physiological concentration of Aβ1-42 oligomers and fibrils (1 μM) does not cause any significant cell death, DNA damage, or cell morphology changes in human fetal astrocytes or iAstrocytes. However, the individual iAstrocyte cell lines could have a varying response to different Aβ aggregation species, suggesting a small degree astrocyte heterogeneity. As a confirmation of the model, and to characterise the astrocyte responses to disease-related stressors, human fetal astrocytes were treated with 100 μM of hydrogen peroxide to induce oxidative stress. Acute oxidative stress caused a rapid upregulation of DNA damage response markers and formation of γH2AX-positive DNA foci, which indicate double stranded DNA breaks. Furthermore, acute oxidative stress causes a significant cell death in human fetal astrocytes after 24h of treatment. Finally, RNAseq analysis, followed by differential gene expression pathway analyses showed that Aβ can elicit heterogeneous responses in iAstrocytes, resulting in differential expression of many cellular pathways, and corresponding genes. Furthermore, the astrocytes showed a heterogenic response to different types of Aβ (oligomers, fibrils, extracts). The gene changes identified imply Aβ-mediated astrocyte changes in inflammation (MIRLET7I, MMP9, IL1A, IL32, CXCL8, SAA1), increased Aβ production (MIRLET7I), breakdown of the extracellular matrix (MMP1, MMP9, MMP13), loss of tight junctions (CLDN2, CLDN18, MARVELD3), and impairment in neurotransmitter signalling (EPHB2, EPHB6, GRIN2C, GABRA1, GABRAP, SHISA7). Conclusions: Well-characterised Aβ1-42 preparations were generated and optimised, which allowed for reliable and reproducible treatment of astrocytes. Astrocytes show a rapid DNA damage response to oxidative stress, which is not seen as a response to Aβ treatments in fetal astrocytes or iAstrocytes. This implies that astrocytes could display a heterogenic response to different types of stress and injury in disease (Aβ vs oxidative stress). Both younger and aged astrocytes appear resistant to various forms of Aβ, which differs from neuronal responses in disease. Treatment of iAstrocytes with Aβ shows a differential expression of genes, suggesting that Aβ-modulated astrocytes may play a role in neuroinflammation, breakdown of BBB, oxidative stress, and impairment in neurotransmitter signalling. However, different types of Aβ (oligomers, fibrils, extracts) can elicit varying gene expression responses in astrocytes, implying that Aβ can induce heterogenic changes to astrocytes based on their final conformation. This may alter astrocyte function, which could be relevant to Alzheimer’s disease progression. Dysregulation of these astrocytic functions could show potential mechanisms behind Alzheimer’s disease pathology and should be investigated further