PDF, English Restricted access: Repository staff only until 13 November 2026. Login+Download (20MB) | Terms of use

Abstract

Neuroinflammation is a key hallmark of neurodegenerative diseases, driven mostly by microglia, the innate immune cells of the brain. This is particularly evident in Alzheimer’s Disease (AD), where an increasing number of genetic risk variants in immune-related genes have been associated with disease onset and development. Unsurprisingly, many of these variants are expressed in microglia, highlighting their importance in AD pathology. Among known genetic risk factors for AD, the APOE4 variant of APOE is the strongest genetic determinant, increasing the risk of developing AD in comparison with the neutral APOE3 variant and the protective APOE2 variant. In recent years, APOE4 has been associated with increased inflammation, but the precise consequences and underlying mechanisms remain unexplored. In addition, the increasing evidence highlighting differences in the impact of APOE4 depending on its cellular source demonstrates the value of studying APOE in a cell-type specific manner. The aim of this study was to characterize the temporal transcriptomic and underlying regulatory response of neurons and microglia upon inflammation and to understand how APOE4 may modulate this response. For that, a hiPSC-derived microglia and neuron co-culture model system was combined with single-cell RNA expression and chromatin accessibility sequencing readout. To disentangle the specific impact of APOE4 in each of the cell types, they were co-cultured in four different combinations: both expressing APOE4, both expressing APOE3, neurons expressing APOE4 with APOE3 microglia and neurons expressing APOE3 with APOE4 microglia. These co-cultures were then stimulated with pro-inflammatory molecules like LPS, TNF or IL-1 over a time course, resulting in 48 conditions. However, available single-cell RNA and ATAC methods were not compatible with the profiling of a large number of samples. To overcome this limitation, a novel method called SUM-seq was developed for parallel profiling of chromatin accessibility and mRNA expression from single nuclei at ultra-high throughput, enabling analysis of many samples in a single experiment. The SUM-seq method was benchmarked in a human-mouse species mixing experiment and its ability to capture relevant biological insights was validated using a well-established macrophage polarization model comparable to the microglia model. SUM-seq achieved data quality that surpasses existing ultra-high-throughput methods, whilst being the first to introduce sample multiplexing to single-cell multiomic profiling. Employing SUM-seq for profiling neuron-microglia co-cultures, revealed temporal gene expression and regulatory factor response dynamics for each cell type upon inflammation. In microglia, APOE4 expression altered the early inflammatory response, inducing an upregulation of several cytokine and NF-κB related genes. This effect was independent of neuronal APOE4. Conversely, in later stages of the response, an upregulation of a set of genes including AD-associated genes and a downregulation of interferon-related genes arose in an APOE4 dose dependent manner. In neurons, a synergistic effect of APOE4 being expressed in both neurons and microglia was observed, triggering the downregulation of synapse-related genes in later stages of inflammation. Overall, this study presents SUM-seq as a powerful novel single-cell method to study gene expression and chromatin accessibility from single nuclei in a multiplexed high throughput manner, and its application to multiple hiPSC-derived culture systems. Moreover, it delivers the first comprehensive overview of the temporal responses of microglia and neurons to inflammation and highlights how APOE4 impacts these responses, offering insights to how APOE4 may contribute to AD onset and progression.