Microglial TLR4-Lyn kinase is a critical regulator of neuroinflammation, Aβ phagocytosis, neuronal damage, and cell survival in Alzheimer's disease

Abstract

Disease-Associated Microglia (DAM) are a focus in Alzheimer's disease (AD) research due to their central involvement in the response to amyloid-beta plaques. Microglial Toll-like receptor 4 (TLR4) is instrumental in the binding of fibrillary amyloid proteins, while Lyn kinase (Lyn) is a member of the Src family of non-receptor tyrosine kinases involved in immune signaling. Lyn is a novel, non-canonical, intracellular adaptor with diverse roles in cell-specific signaling which directly binds to TLR4 to modify its function. Lyn can be activated in response to TLR4 stimulation, leading to phosphorylation of various substrates and modulation of inflammatory and phagocytosis signaling pathways. Here, we investigated the TLR4-Lyn interaction in neuroinflammation using WT, 5XFAD, and 5XFAD x Lyn^-/- mouse models by western blotting (WB), co-immunoprecipitation (co-IP), immunohistochemistry (IHC) and flow cytometric (FC) analysis. A spatial transcriptomic analysis of microglia in WT, 5XFAD, and 5XFAD x Lyn^-/- mice revealed essential genes involved in neuroinflammation, Aβ phagocytosis, and neuronal damage. Finally, we explored the effects of a synthetic, TLR4-Lyn modulator protein (TLIM) through an in vitro AD model using primary murine microglia. Our WB, co-IP, IHC, and FC data show an increased, novel, direct protein-protein interaction between TLR4 and Lyn kinase in the brains of 5XFAD mice compared to WT. Furthermore, in the absence of Lyn (5XFAD x Lyn^-/- mice); increased expression of protective Syk kinase was observed, enhanced microglial Aβ phagocytosis, increased astrocyte activity, decreased neuronal dystrophy, and a further increase in the cell survival signaling and protective DAM population was noted. The DAM population in 5XFAD mice which produce more inflammatory cytokines and phagocytose more Aβ were observed to express greater levels of TLR4 and Lyn. Pathway analysis comparison between WT, 5XFAD, and 5XFAD x Lyn^-/- mice supported these findings via our microglial spatial transcriptomic analysis. Finally, we created an in vitro co-culture system with primary murine microglial and primary murine hippocampal cells exposed to Aβ as a model of AD. When these co-cultures were treated with our TLR4-Lyn Interaction Modulators (TLIMs), an increase in Aβ phagocytosis and a decrease in neuronal dystrophy was seen. Lyn kinase has a central role in modulating TLR4-induced inflammation and Syk-induced protection in a 5XFAD mouse model. Our TLIMs ameliorate AD sequalae in an in vitro model of AD and could be a promising therapeutic strategy to treat AD.

Keywords: 5XFAD; Alzheimer’s disease; Disease-Associated Microglia; Lyn kinase; TLR4; TLR4-Lyn modulator protein.

Fig. 1

Macrophage TLR4-Lyn co-expressions in Alzheimer’s mouse brain. a) IHC imaging of WT and 5xFAD mouse brains for Lyn (1:350; green) and TLR4 (1:500; red). All scale bars = 50 µm. b) Surface plots of Lyn-TLR4 colocalization for each group was quantified using ImageJ software, yellow color represents highest possible colocalization. c) Co-immunoprecipitation of Lyn kinase and TLR4 to establish direct interaction between the two proteins in AD vs WT mice brains. d) DAMs are characterized by CD11c⁺Clec7a⁺ population plotted on microglia (CD11b⁺CD45^med) in mouse brains. Bottom plots show Lyn and TLR4 double positive DAMs. e) Quantification of DAMs and the DAM subset that is pLyn and TLR4 double positive. (n = 4; students t-test *P < 0.05, **P < 0.001, ***P < 0.0001 between groups).

Fig. 2

Cell signaling pathway analysis. a) WB of TLR4, pLyn, Lyn, pSyk, Syk, PI3K, pAkt, Akt, pGSK3β, and GSK3β in WT, 5xFAD and 5xFAD X Lyn^-/-mice, GAPDH was used as loading control. b) Quantification of band area of the proteins analyzed by western blot, n = 3; F = 70.34, 99.16, 74.45, 93.51, 76.57, 90.32, 71.08, 52.35, 51.50, 57.97 for all proteins, respectively. F for GAPDH was 1.03 and not significant. One-Way ANOVA, *P < 0.05, **P < 0.001, ***P < 0.0001 between groups. c) Immunohistochemical imaging of 5xFAD and 5xFAD X Lyn^-/-mice for Lyn (1:350; green) and TLR4 (1:500; red). All scale bars = 50 µm. Surface plots of Lyn-TLR4 colocalization for each group was quantified using ImageJ software, yellow color represents highest possible colocalization. d) Flow cytometric analysis of Syk, TLR4, Lyn expression on the total brain cells of WT, 5XFAD and 5XFAD X Lyn^-/- mice. e) Quantification of Syk (F = 75.29), TLR4 (F = 43.56), Lyn (F = 48.15) expression by the brain cells (n = 4; One-way ANOVA *P < 0.05, **P < 0.001, ***P < 0.0001 between groups).

Fig. 3

Microglial morphology analysis and inflammatory cytokine (IFNɣ) production. a) The ratio of cell body area to pseudopodia area was calculated and termed microglial morphology analysis index (MMAI), scale bars = 10 µm. Area measurements were done using ImageJ software. n = 4 ; F = 39.82, statistical analysis was done by one-way ANOVA, *P < 0.05, **P < 0.001, ***P < 0.0001 between groups. b) IHC imaging of 5xFAD and 5xFAD X Lyn^-/-mice for GFAP (1:500; red), counterstained with DAPI (blue), scale bars = 50 µm. c) Quantification of reactive astrocytes (GFAP⁺) per field (n = 4; Student’s t-test *P < 0.05 between groups). Flow analysis of TNF⍺ and IFNɣ production by microglia, characterized by CD11b⁺CD45^med populations of WT, 5XFAD and 5XFAD X Lyn^-/- mouse brains. e) Quantification of TNF⍺ (F = 25.48) and IFNɣ (F = 114) production by the MG (n = 4; One-way ANOVA *P < 0.05, **P < 0.001, ***P < 0.0001 between groups).

Fig. 4

Spatial transcriptomics analysis of microglia specific genes and functional pathways. Heatmaps generated by gene set enrichment analysis (GSEA) of set of signature genes expression in the microglia of WT, 5XFAD and 5XFAD X Lyn^-/- mice involved in a) inflammatory cytokine signaling pathway and b) microglial phagocytosis pathways c) cell signaling pathway d) primary component analysis (PCA) plot on the genes involved in phagocytosis in 5XFAD and 5XFAD X Lyn^-/- e) volcano plot compares the microglial gene upregulation in 5XFAD and 5XFAD X Lyn^-/- mouse.

Fig. 5

In vivo phagocytosis of Aβ in WT, 5xFAD and 5xFAD X Lyn^-/- mouse brain cells. a) Immunostaining of Iba-1 (green), Aβ oligomers were bound to the IP administered methoxy-X04 (Me-X04) (red), all scale bars = 50 µm b) Surface plots for Iba1-Me-X04 colocalization. c) in vivo Aβ phagocytosis measured by Tmem119⁺Me-X04⁺ population gated on DAMs (CD11c⁺Clec7a⁺), plotted on microglia (CD11b⁺CD45^med). Data analyzed by FlowJo; d) Quantification of DAMs (F = 37.70) and DAM Aβ phagocytosis (F = 31.32). (n = 4; One-way ANOVA *P < 0.05, **P < 0.001, ***P < 0.0001 between groups).

Fig. 6

Lyn controls neuronal dystrophy and cognitive dysfunction in AD. a) Immunostaining to determine dystrophic neurites around the Aβ plaque. WT, 5xFAD and 5xFAD X Lyn^-/- mouse brain cortical sections were stained with AT8 (green) for phosphorylated tau (p-tau) and Aβ (red). All scale bars = 50 µm. b) Surface plots for AT8-Aβ colocalization. c) Cognitive function test by Barnes maze (F = 54.80, 2-way ANOVA where the 2 variables are time and genotype to assess for differences amongst all groups and Bonferroni post hoc test to assess for differences between pairs of groups). d) Discrimination index measured by Novel object recognition test (F = 37.23, One way ANOVA). n = 6; *P < 0.05, **P < 0.001, ***P < 0.0001 were considered statistically significant. e) Flow analysis to determine the neuronal population in the brain cells using anti-mouse NeuN, a neuronal marker. f) Quantification of neuronal cell percentage in the mouse brain by measuring Neun⁺ cells, n = 3; F = 52.10, One-way ANOVA to assess for differences amongst all groups and Bonferroni post hoc test to assess for differences between pairs of groups. *P < 0.05, **P < 0.001, ***P < 0.0001 was considered statistically significant.

Fig. 7

Effects of TLIM on TLR4-Lyn interaction to improve microglial phagocytosis and neuronal dystrophy. a) TLR4 domains, region of interaction (ROI) defining TLR4-TLIM and Dose–response curve for TLR4-TLIM IC50 = 17.9 nM. b) Microglial phagocytosis assay; primary microglia were incubated with pHrodo tagged Aβ_1-42 oligomers and GST-liposomes or TLIM-liposomes. Data analyzed by FlowJo; n = 4 per group. Quantifications of c) DAMs, d) Lyn⁺TLR4⁺ positive DAMs and e) DAM-Aβ phagocytosis in GST or TLIM liposome treated AD-microglia cultures (n = 4; Student’s t-test *P < 0.05, **P < 0.001 between groups). f) Microglia-Neuron Trans-well Assay. Primary microglia (MG) plated on the top chamber and neuronal cell line (hippocampal) plated in the bottom chamber. Cells were incubated with Aβ oligomers for 2 h, followed by 1-h incubation with 20 nM of either GST-liposomes or TLIM-liposomes. g) Immunohistochemistry staining of AT8 (green, 1:500) and DAPI (blue) in Aβ treated trans-well cultures. h) Quantification of number of dystrophic neurites per filed stained with AT8 (green), scale bars = 25 µm n = 4; Student’s t-test *P < 0.05, **P < 0.001 between groups.