Acute Communication Between Microglia and Nonparenchymal Immune Cells in the Anti-Aβ Antibody-Injected Cortex

Abstract

Anti-Aβ immunotherapy use to treat Alzheimer's disease is on the rise. While anti-Aβ antibodies provide hope in targeting Aβ plaques in the brain, there still remains a lack of understanding regarding the cellular responses to these antibodies in the brain. In this study, we sought to identify the acute effects of anti-Aβ antibodies on immune responses. To determine cellular changes due to anti-Aβ antibody exposure, we intracranially injected 14 mo APP male and female mice with anti-Aβ IgG1 (6E10) or control IgG1 into the cortex. After 24 h or 3 d, we harvested the cortex and performed a glial cell-enriched preparation for single-cell sequencing. Cell types, proportions, and cell-to-cell signaling were evaluated between the two injection conditions and two acute timepoints. We identified 23 unique cell clusters including microglia, astrocytes, endothelial cells, neurons, oligos/OPCs, immune cells, and unknown. The anti-Aβ antibody-injected cortices revealed more ligand-receptor (L-R) communications between cell types, as well as stronger communications at only 24 h. At 3 d, while there were more L-R communications for the anti-Aβ antibody condition, the strength of these connections was stronger in the control IgG condition. We also found evidence of an initial and strong communication emphasis in microglia-to-nonparenchymal immune cells at 24 h, specifically in the TGFβ signaling pathway. We identify several pathways that are specific to anti-Aβ antibody exposure at acute timepoints. These data lay the groundwork for understanding the brain's unique response to anti-Aβ antibodies.

Keywords: Alzheimers disease; amyloid; antiamyloid antibody; inflammation; microglia; single-cell sequencing.

Figure 1.

Anti-Aβ antibody alters cell–cell communication networks. A, A schematic of study design and analysis. B, The UMAP depicting 23 unique clusters found from glial-enriched single-cell sequencing prep and analysis. UMAPs for sample, condition, sex, and timepoint can be found in Extended Data Figure 1-1. There were no preparation batch or sex-specific differences noted (Extended Data Table 1-1). C, The number of cells and percentage of total cell population for each cluster. D, Marker genes of various cell types to aid in identification of cluster cell type identity (Exteded Data Table 1-2). E, The difference in the number of significant ligand–receptor pairing counts (Aβ-IgG) at 24 h postinjection. Sending (ligand) cell types are along the y-axis/left, while receiving (receptor) cell types are along the x-axis/top. Red indicates more interactions in the Aβ injection condition, and blue indicates more interactions in the IgG injection condition. F, The difference in weight/strength of the significant ligand–receptor interactions (Aβ-IgG) at 24 h postinjection. G, The difference in the number of significant ligand–receptor communication counts (Aβ-IgG) at 3 d postinjection. H, The difference in weight/strength of the significant ligand–receptor interactions (Aβ-IgG) at 3 d postinjection.

Figure 2.

Microglia show a temporal change in communication due to anti-Aβ antibody exposure. A, Marker genes to identify subtypes of various microglia states. B, Percent change in microglia cells per cluster from 24 h to 3 d (Extended Data Fig. 2-1). C, Signaling pathways enriched in micro (sending):micro (receiving) communication at 24 h in anti-Aβ (pink) and IgG (teal). D, Signaling pathways enriched in micro (sending):micro (receiving) communication at 3 d in anti-Aβ (pink) and IgG (teal). E, A chord diagram of significant CD48 signaling at 24 h in Aβ injected (left) and IgG injected (right). F, The sum of sending and receiving communication probability per cluster for CD48 signaling at 24 h. G, A chord diagram of significant PD-L1 signaling at 3 d in Aβ injected (left) and IgG injected (right). H, The sum of sending and receiving communication probability per cluster for PD-L1 signaling at 3 d. Extended Data Figure 2-2 shows CCL signaling and SPP1 signaling results. Microglia communications are colored according to cluster sending (ligand), and nonmicroglia communications are in gray. Statistically significant communication terms are noted with their respective colors (Wilcoxon test, p < 0.05). Mg, microglia; homeo, homeostatic; pre-DAM, pre-disease-associated microglia; DAM,¸disease-associated microglia; prolif, proliferative; ast-L, astrocyte-like.

Figure 3.

Microglia signal nonmicroglial immune cells through TGFβ. A, Immune cell cluster identification through multiple marker genes. Cluster 14 expresses multiple perivascular macrophage genes, cluster 16 expresses some PVM and high response to interferon response genes suggesting a macrophage cell type, and cluster 19 expresses markers for T-cells and NK cells. B, The PanglaoDB annotation of immune cell clusters graphed by odds ratio for the top five likely cell types (Extended Data Fig. 3-1). C, Percent changes in the number of immune cells in anti-Aβ and IgG injected conditions between 24 h and 3 d. D, The overall outgoing and incoming communication interaction strength for microglia and nonparenchymal cell clusters in anti-Aβ-injected and IgG-injected mice at 24 h (3 d; Extended Data Fig. 3-1). The black outline differentiates nonmicroglia immune cell clusters, and no outline indicates a microglia cluster. E, The overall outgoing and incoming communication interaction strength for microglia and nonparenchymal cell clusters in anti-Aβ- and IgG-injected mice at 3 d. The black outline differentiates nonmicroglia immune cell clusters, and no outline indicates a microglia cluster. F, The signaling pathways enriched in micro (sending):immune (receiving) communication at 24 h in anti-Aβ (pink) and IgG (teal). G, The signaling pathways enriched in micro (sending):immune (receiving) communication at 3 d in anti-Aβ (pink) and IgG (teal). H, A chord diagram of significant TGFβ signaling at 24 h in Aβ-injected (left) and IgG-injected (right). I, The sum of sending and receiving communication probability per immune cluster for TGFβ signaling at 24 .

Figure 4.

Nonmicroglial immune cells produce dynamic communication patterns at 24 h and 3 d. A, A chart categorizing significant signaling pathways and their status at 24 h and 3 d for cluster 14 perivascular macrophages, cluster 16 macrophages, and cluster 19 T-cells and NK cells (Extended Data Fig. 4-1). B, Significant ligand–receptor TNF signaling for immune clusters to all other clusters at 24 h. C, Significant ligand–receptor TNF signaling for immune clusters to all other clusters at 3 d. D, Significantly increased ligand–receptor signals for cluster 14 PVMs to endothelial cells clusters (10, 22) in the anti-Aβ antibody condition at 24 h. E, Significantly increased ligand–receptor signals for cluster 14 PVMs to endothelial cells (clusters 10, 22) in the anti-Aβ antibody condition at 3 d.