Absence of CX3CR1 impairs the internalization of Tau by microglia

Abstract

Background: Extracellular Tau is toxic for neighboring cells, and it contributes to the progression of AD. The CX3CL1/CX3CR1 axis is an important neuron/microglia communication mechanism.

Methods: We studied Tau clearance by microglia both in vitro (microglia primary cultures treated with Cy5-Tau, affinity chromatography to study the binding of Tau to CX3CR1, and Tau-CX3CL1 competition assays) and in vivo (stereotaxic injection of Cy5-Tau into WT and CX3CR1^-/- mice). The expression of CX3CR1, CX3CL1 and the microglial phagocytic phenotype were studied in brain tissue samples from AD patients.

Results: Tau binding to CX3CR1 triggers the internalization of the former by microglia, whereas S396 Tau phosphorylation decreases the binding affinity of this protein to CX3CR1. Of note, the progressive increase in the levels of phosho-Tau occurred in parallel with an increase in CX3CR1. In addition, our studies suggest that the phagocytic capacity of microglia in brain tissue samples from AD patients is decreased. Furthermore, the CX3CR1/CX3CL1 axis may be impaired in late stages of the disease.

Conclusions: Our data suggest that the CX3CR1/CX3CL1 axis plays a key role in the phagocytosis of Tau by microglia in vitro and in vivo and that it is affected as AD progresses. Taken together, our results reveal CX3CR1 as a novel target for the clearance of extracellular Tau.

Keywords: Alzheimer’s disease; CX3CR1; Microglia; Phagocytosis; Tau; Tauopathies.

Fig. 1

The absence of CX3CR1 leads to a decrease in extracellular Tau internalization by microglia in vitro. Representative immunofluorescence images of primary cultures of microglia derived from WT mice (a) and their high-power magnification (b), and from KO animals (c) and their magnification image (d). Cells were incubated with 1 μM Tau-Cy5. Immunofluorescence images for Iba1 (white) and Cy5 (red) are shown separately. (e) Quantification of Cy5⁺ area in individual microglial cells treated with either Tau-Cy5, or BSA-Cy5 (f) for 0, 5, 30 and 60 min. g-l Representative images of primary microglia cultures from WT mice treated with PBS-Cy5 (g) or Tau-Cy5 (h) and from KO animals treated with PBS-Cy5 (i) or Tau-Cy5 (j) showing CD68 (white) and Iba1 (green) staining. k Quantification of CD68⁺ area in individual WT and KO microglia after treatment with PBS or Tau at 0 and 30 min. Bars show means ± S.E. ∗∗∗p ≤ 0.001. White scale bar: 100 μm, purple scale bar: 50 μm

Fig. 2

The absence of CX3CR1 leads to a decrease in Tau internalization by microglia in vivo. Representative images of the dentate gyrus (DG) of WT (a-b) and KO mice (c-d) stereotaxically injected with PBS-Cy5 or Tau-Cy5 and sacrificed one week later. Immunofluorescence images for Iba1 (green) and Cy5 (red) and high-power magnification are shown separately. e Colocalization between Iba1 and Cy5 in mice injected with PBS-Cy5 or Tau-Cy5 in the DG. Reduced colocalization between Cy5 and Iba1 can be observed in KO animals injected with Tau-Cy5 in comparison to WT mice that received the same injection. f Quantification of Cy5 fluorescence intensity in mice injected with PBS-Cy5 or Tau-Cy5. An increased level of Cy5 fluorescence intensity can be observed in KO mice injected with Tau-Cy5 in comparison to WT mice that received the same injection. Bars show means ± S.E. ∗p ≤ 0.05; ∗∗p ≤ 0.01; + 0.05 < p ≤ 0.1. Purple scale bar: 100 μm, Blue scale bar: 50 μm. GL, granule layer; H, hilus; ML, molecular layer; CA3, cornu ammonis region 3; DG, dentate gyrus

Fig. 3

Tau binds to CX3CR1. Synthetic CX3CR1 was coupled to CNBr-activated Sepharose4B. Tau detected with anti-Tau5 (a), the positive control CX3CL1 (b), the negative control c-terminal Tau detected with anti-Tau46 (c), and Tau phosphorylated by GSK-3β detected with anti-pTau396 (d) were loaded on the column, as described in the Methods section. Each protein sample was used at a concentration of 1 mg/ml. Exclusion volume (excl. Vol.), aliquots corresponding to the column washed with buffer A, bound proteins eluted by addition of 0.5 M NaCl (marked with an arrow), and collected fractions (Fract. N°) are shown. Note that the profiles indicate that Tau, CX3CL1 and phospho-Tau were bound to the column, whereas the C-terminal fragment of Tau was not

Fig. 4

Tau competes with CX3CL1 for binding to CX3CR1. a-b Representative images of Tau-Cy5 attached to the surface of WT (a) and KO (b) microglia after treatment with Tau-Cy5 for 5 min at 4 °C. c Quantification of Cy5 fluorescence intensity on the microglial surface. In the absence of CX3CR1, the levels of Tau-Cy5 on the membrane of microglial cells are markedly reduced. d Tau profile in the competition assay between Tau and CX3CL1 detected with anti-Tau5. Note that the profiles indicate that the binding of Tau to the column was weaker in the presence of CX3CL1. e-g In vitro competition assay performed in primary microglial cell culture. Representative images and quantification of Cy5 fluorescence intensity (red) and Iba1 (green) in cells treated with Tau-Cy5 in the absence (c) or presence (d) of CX3CL1. As shown, Tau internalization is reduced in the presence of CX3CL1. Bars show means ± S.E. ∗∗∗p ≤ 0.001. White scale bar: 100 μm, purple scale bar: 50 μm

Fig. 5

Tau phosphorylation, the number of microglial cells, colocalization between phosphorylated Tau and CX3CR1, and CX3CR1 expression increase in advanced stages of AD. Representative images of hippocampal tissue derived from AD patients showing microglia (Iba1, green), phospho-Tau (p-S396, white, red arrow) and DAPI (blue) staining at different stages of the disease: Braak-Tau I (a), Braak-Tau III (b) and Braak-Tau VI (c). Immunofluorescence images for Iba1- (green) and Tau-labeled structures (white, red arrow) are shown separately. d Representative high-power magnification image of microglia (green) and phospho-Tau (red) in Braak-Tau stage VI are shown. Orthogonal views are provided to highlight the colocalization between microglia and phospho-Tau. e Quantification of phospho-Tau fluorescence intensity present in the human hippocampus in AD. f Quantification of the number of microglia. g Colocalization between Iba1 and phospho-Tau at different stages of the disease. h Quantification of fluorescence intensity of CX3CR1 in the human hippocampus. (i-j) Representative images and quantification of the Cy5⁺ area relative to the internalization of Tau-Cy5 or phospho-Tau-Cy5 by microglia in vitro. Cell density was calculated as number of cells/μm³. Bars show means ± S.E. + p ≤ 0.05; ∗p ≤ 0.05; ∗∗ p ≤ 0.01 and ∗∗∗p ≤ 0.001. Red scale bar: 40 μm. White scale bar: 50 μm

Fig. 6

The impaired phagocytic capacity of microglia increases in advanced stages of AD. a Representative immunofluorescence images of phagocytic pouche (yellow arrow) in a microglial cell (Iba1, green) of human AD tissue. b Quantification of the number of phagocytic pouches per microglial cell. c Quantification of nucleus area of microglia in human tissue. d Quantification of CX3CL1 fluorescence intensity present in the human hippocampus in AD. Bars show means ± S.E. + p ≤ 0.05; ∗p ≤ 0.05; ∗∗ p ≤ 0.01 and ∗∗∗p ≤ 0.001. Purple scale bar: 40 μm. N, microglia nucleus (DAPI)