Reactive Astrocytes Contribute to Alzheimer's Disease-Related Neurotoxicity and Synaptotoxicity in a Neuron-Astrocyte Co-culture Assay.

Abstract

Pathological hallmarks of Alzheimer’s disease (AD) include deposition and accumulation of amyloid- β (Aβ), neurofibrillary tangle formation, and neuronal loss. Pathogenesis of presymptomatic disease stages remains elusive, although studies suggest that the early structural and functional alterations likely occur at neuronal dendritic spines. Presymptomatic alterations may also affect different CNS cell types. However, specific contributions of these cell types as cause or consequence of pathology are difficult to study in vivo. There is a shortage of relatively simple, well-defined, and validated in vitro models that allow a straightforward interpretation of results and recapitulate aspects of pathophysiology. For instance, dissecting the AD-related processes (e.g., neurotoxicity vs. synaptotoxicity) may be difficult with the common cell-based systems such as neuronal cell lines or primary neurons. To investigate and characterize the impact of reactive astrocytes on neuronal morphology in the context of AD-related cues, we modified an in vitro co-culture assay of primary mouse neurons and primary mouse astrocytes based on the so-called Banker “sandwich” co-culture assay. Here, we provide a simple and modular assay with fully differentiated primary mouse neurons to study the paracrine interactions between the neurons and the astrocytes in the co-culture setting. Readouts were obtained from both cell types in our assay. Astrocyte feeder cells were pre-exposed to neuroinflammatory conditions by means of Aβ42, Aβ40, or lipopolysaccharide (LPS). Non-cell autonomous toxic effects of reactive astrocytes on neurons were assessed using the Sholl analysis to evaluate the dendritic complexity, whereas synaptic puncta served as a readout of synaptotoxicity. Here, we show that astrocytes actively contribute to the phenotype of the primary neurons in an AD-specific context, emphasizing the role of different cell types in AD pathology. The cytokine expression pattern was significantly altered in the treated astrocytes. Of note, the impact of reactive astrocytes on neurons was highly dependent on the defined cell ratios. Our co-culture system is modular, of low cost, and allows us to probe aspects of neurodegeneration and neuroinflammation between the two major CNS cell types, neurons, and astrocytes, under well-defined experimental conditions. Our easy-to-follow protocol, including work-flow figures, may also provide a methodological outline to study the interactions of astrocytes and neurons in the context of other diseases in the future.