TBC1D15 Inhibits Autophagy of Microglia through Maintaining the Damaged Swelling Lysosome in Alzheimer's Disease

Abstract

Autophagy in microglia is essential for the clearance of amyloid-beta (Aβ) and amyloid plaques in Alzheimer's disease. However, reports regarding the levels of autophagy in microglia have been inconsistent; some studies indicate an early enhancement followed by a subsequent reduction, while others describe a persistently weakened state. Notably, there is a lack of systematic studies documenting the temporal changes in microglial autophagy. TBC1D15, a Rab GTPase, plays a crucial role in lysosomal membrane repair, yet its function in regulating microglial autophagy in Alzheimer's disease remains unexplored. Current research suggests that microglial autophagy is activated in 3-month-old AD mice but gradually decreases by 12 months of age. Furthermore, TBC1D15 levels are significantly elevated in the lysosomes of microglia in Alzheimer's disease. Silencing TBC1D15 markedly inhibits swelling and Aβ phagocytosis in BV2 cells following Aβ treatment while simultaneously promoting autophagy and lysophagy. LIMP II/ATG8-TBC1D15-Dynamin2/RAB7 might participate in lysosome swelling of microglia in AD. These findings indicate that TBC1D15 in microglia is critical for the decline of autophagy in Alzheimer's disease. It is suggested that targeting microglial TBC1D15 may be an important strategy for enhancing autophagy, which facilitates the clearance of amyloid plaques as a therapeutic approach for Alzheimer's disease.

Figure 1.

Dynamics of microglial autophagy in 3xTg AD mice at different ages. (A) The levels of Aβ phagocytosis by microglia in wild-type, 3-month-old, 6-month-old, 9-month-old, and 12-month-old 3xTg AD mice. n=5/group, scale bars: 50 μm. (B, C) Immunofluorescence was used to observe the distribution levels of LC3B and P62 in microglial cells of 3xTg AD mice at different ages. n=3/group, scale bars: 50 μm. (D) Aβ-positive area in microglia at different ages. n=5/group. (E, F) Normalized analysis of the LC3B-positive particle area and the P62-positive particle area in microglia of AD mice at different ages. n=3/group. (G) Western blot analysis of LC3B levels in wild-type and AD mice of different ages. n=5/group. (H, I) Analysis of the relative expression levels of LC3B and P62 in Western blot. n=5/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**), p < 0.001(***) and p < 0.0001(****).

Figure 2.

BV2 cells exhibit time-dependent dynamic changes in autophagy following Aβ treatment. (A) BV2 cells were treated with Aβ oligomers for 6, 9, 12, and 24 h, and controls were treated with PBS, and Western blot was used to detect the levels of LC3B and P62, with β-actin as a loading control. n=3/group. (B, C) Relative quantification of LC3B and P62 levels was performed. n=3/group. (D, E, F) C8D1A cells received corresponding treatments, and the levels of LC3B and P62 were detected with relative quantification analysis. n=3/group. (G, H, I) HT22 cells were subjected to similar treatments to detect LC3B and P62 levels, followed by relative quantification analysis. n=3/group. (J, K, L, M) BV2 cells were treated with Aβ oligomers for 6 and 24 h, and controls were treated with PBS. Immunofluorescence was utilized to observe the levels of LC3B and P62 in BV2 cells, and the relative intensity of LC3B and P62 immunofluorescence was analyzed. n=3/group, scale bars: 50 μm. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**).

Figure 3.

With increasing age, 3xTg mice exhibit enhanced Aβ phagocytosis by microglial cells, along with enlarged lysosomal volume and aggravated damage. (A) Brain slices from wild-type, 3-month-old, and 12-month-old mice were incubated with IBA1 antibody (green), 6E10 antibody (red), and LAMP1 antibody (purple) to observe microglial phagocytosis of plaques and lysosomal morphology. n=3/group. Scale bar=20 μm. (B) The slices were also incubated with IBA1 antibody (green), galectin3 antibody (red), and LAMP1 antibody (purple) to assess lysosomal damage in microglia. n=4/group. Scale bar=25 μm. (C) In IBA1-labeled cells, positive particles co-localizing with 6E10 and LAMP1 were statistically analyzed. n=3/group. (D) The area of LAMP1-positive particles in IBA1-labeled cells was quantified. n=3/group. (E) In IBA1-labeled cells, positive particles co-localizing with galectin3 and LAMP1 were also statistically analyzed. n=4/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**), p < 0.001(***).

Figure 4.

BV2 cells, rather than C8D1A and HT22 cells, exhibit an increase in phagocytosis that is dependent on time and Aβ concentration. (A) BV2 cells were treated with Aβ oligomers at concentrations of 2 μM, 5 μM,10 μM, and 15 μM, and the control group was treated with PBS, and then BV2 phagocytosis was measured with FAM-labeled Aβ oligomers. n=3/group, scale bar=50 μm. (B) BV2 cells were treated with 10 μM Aβ oligomers and incubated for 4 h, 6 h, 12 h, and 24 h. Controls were treated with PBS, and their phagocytosis of Aβ oligomers was observed with FAM-labeled Aβ oligomers. n=3/group, scale bar=50 μm. (C, D) C8D1A cells and HT22 cells were treated with 10 μM Aβ oligomers and incubated for 4 h, 6 h, 12 h, and 24 h. Controls were treated with PBS, and their phagocytosis of Aβ oligomers was observed with FAMlabeled Aβ oligomers. n=3/group, scale bar=10 μm. (E) A relative quantitative analysis of Aβ phagocytosis levels at different concentrations was performed for the three cell types. n=3/group. (F) A relative quantitative analysis of Aβ phagocytosis levels at the different incubation times was conducted for the three cell types. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*).

Figure 5.

BV2 cells, rather than C8D1A cells and HT22 cells, show lysosomal swelling and damage after Aβ phagocytosis. (A) BV2 cells, C8D1A, and HT22 cells were incubated with 10 μM Aβ, and lysotracker staining was used to observe lysosomal volume. n=3/group, scale bar=10 μm. (B) BV2 cells were treated with 10 μM Aβ oligomers and incubated for 6 and 24 h. Controls were treated with PBS. An ATP-sensitive AT 1.03-LAMP1 protein was used to detect ATP levels within lysosomes. n=3/group, scale bar=10 μm. (C) BV2 cells were treated with 10 μmol/L Aβ oligomers and incubated for 6 h, 12 h, and 24 h. Controls were treated with PBS. The GFP-mCherry-LC3 protein was employed to assess the pH levels inside lysosomes. n=3/group, scale bar=10 μm. (D) After Aβ oligomer treatment of BV2 cells for 6 hours and 24 hours, the area of lysotracker-positive particles was normalized for analysis. n=3/group. (E) After Aβ oligomer treatment of BV2 cells for 6 h and 24 h, the fluorescence intensity of AT1.03-LAMP1 protein was normalized for analysis. n=3/group. (F) Aβ oligomer treatment of BV2 cells for 6, 12, and 24 h were analyzed by normalizing the ratio of GFP-positive particle area to mCherry-positive particle area. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**).

Figure 6.

Aβ treatment significantly impairs the autophagy flux and phagocytic ability of BV2 cells. (A) BV2 cells were treated with Aβ or Aβ combined with chloroquine for 6h, and controls were treated with PBS. And the levels of LC3B and P62 were detected using western blot analysis. n=3/group. (B, C) A relative quantitative analysis of the levels of LC3B and P62 was conducted. n=3/group. (D) BV2 cells were treated with Aβ or Aβ combined with rapamycin for 24h. Controls were treated with PBS. And the levels of LC3B and P62 were detected using western blot analysis. n=3/group. (E, F) A relative quantitative analysis of the levels of LC3B and P62 was performed. n=3/group. (G, H) BV2 cells were treated with FAM-labeled Aβ or with chloroquine and rapamycin, to observe phagocytosis of Aβ. n=3/group, scale bar=25 μm. (I, J) A relative quantitative analysis of Aβ levels within BV2 cells was conducted. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**).

Figure 7.

In the microglial lysosomes of 3xTg AD mice, the expression level of TBC1D15 was significantly increased. (A) Brain slices from wild-type, 3-month-old, and 12-month-old 3xTg AD mice were stained with antibodies against IBA1 (green), TBC1D15 (red), and LAMP1 (purple). n=3, scale bar=25 μm. (B) After treating BV2 cells with Aβ oligomers, controls were treated with PBS. The cells were stained for TBC1D15 (green) and LAMP1 (red). n=3/group, scale bar=25 μm. (C) The area of co-localized particles of TBC1D15 and LAMP1 in IBA1-positive cells was quantitatively analyzed. n=3/group (D) Similarly, the area of colocalized particles of TBC1D15 and LAMP1 in BV2 cells was also analyzed. n=3/group. BV2 cells were incubated separately with LLoMe (which can specifically damage lysosomal membranes), Aβ, and microbeads. (E, F, G) Cells were treated with LLoMe, Aβ and microbeads. The control group was treated with PBS. Western blot was employed to detect TBC1D15, galectin3 (a marker of lysosomal damage), TREM2 (a receptor mediating Aβ phagocytosis), and β-actin as a loading control. n=3/group. (H, I, J) TBC1D15, galectin3, and TREM2 were relatively quantified for LLoMe treatment. n=3/group. K, L, M. TBC1D15, galectin3, and TREM2 were relatively quantified for Aβ treatment. n=3/group. (N, O, P) TBC1D15, galectin3, and TREM2 were relatively quantified for microbeads treatment. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**).

Figure 8.

Knockdown of TBC1D15 can improve the autophagy decline and lysosomal damage in BV2 cells treated with Aβ oligomers. (A) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. The control group was treated with PBS. LC3B and P62 protein levels were detected using Western blot analysis. n=3/group. (B, C) LC3, and P62 proteins were relatively quantified. D. BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. TBC1D15, TREM2, and Galectin3 protein levels were detected using Western blot analysis. n=3/group. (E, F, G) TBC1D15, TREM2, and Galectin3 proteins were relatively quantified. H. BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. Lysotracker was used to label lysosomes to observe lysosomal volume. N=3/group. Scale bars: 25 μm. I. Normalized analysis of lysosomal volume. n=3/group. (J) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. The pCMV-mLAMP1-AT1.03 plasmid was transfected to assess lysosomal ATP levels. n=3/group. Scale bars: 25μm. (K) Normalized analysis of the area of pCMV-mLAMP1-AT 1.03-positive particles. n=3/group. (L) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. The GFP-mCherry-LC3B plasmid was transfected to evaluate lysosomal pH levels. n=3/group. Scale bars: 25μm. (M) The ratio of fluorescence intensity GFP-positive particle to mCherry-positive particle was normalized for analysis. n=3/group. (N) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. Phagocytosis experiments were conducted using FAM-labeled Aβ oligomers. n=3/group. Scale bars: 50 μm. (O) The area of Aβ oligomers in BV2 cells was normalized for analysis. n=3/group. (P) BV2 cells were treated with Aβ oligomers, rapamycin, or MG132. The control group was treated with PBS. Western blot was used to detect the protein levels of TBC1D15 and galectin3. n=3/group. (Q, R) Relative quantification analysis of TBC1D15 and galectin3 protein levels. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*), p < 0.01(**).

Figure 9.

Knockdown of TBC1D15 can promote autophagy in microglial cells by enhancing lysophagy in 3xTg AD mice. (A, B) Sections from wild-type and 12-month-old 3xTg AD mice, as well as 3xTg AD mice expressing TBC1D15 shRNA virus driven by the IBA1 promoter, were incubated with antibodies against IBA1 (green), LC3B (red), and P62 (red). n=3/group. Scale bars: 25 μm. (C, D) A comparative quantitative analysis was conducted to determine the area size of P62 and LC3B-positive particles in IBA1-labeled cells. n=3/group. (E) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15. The levels of galectin3 (green) and LAMP1 (purple) were observed using immunofluorescence. n=3/group. Scale bars: 10 μm. (F) A relative quantitative analysis was conducted to assess the colocalization levels of galectin3 and LAMP1 proteins. n=3/group. (G) BV2 cells were treated with Aβ oligomers or Aβ oligomers combined with siRNA TBC1D15, and immunofluorescence was used to observe the levels of ubiquitin protein (purple), galectin3 protein(green) LAMP1 protein (red). n=3/group. Scale bars: 20 μm. (H) A relative quantitative analysis was performed to assess the colocalization levels of ubiquitin, galectin3, and LAMP1.n=3/group. (I) BV2 was treated with siRNA control or siRNA TBC1D15. After 24h transfection, all of them were treated with Aβ for 12h. After that, Western blot was performed to detect RNF13. n=3/group. (J) The analysis of normalized relative expression of RNF13. n=3/group. (K) BV2 was treated with siRNA control or siRNA TBC1D15. After 24h transfection, all of them were treated with Aβ for 12h. After that, immunofluorescence was performed to detect the colocalization area of RNF13, LAMP1, and ubiquitin. n=3/group, scale bar=20 µm. (L) The analysis of normalized colocalization area of LAMP1, RNF13, and ubiquitin. n=3/group. Data from all experiments are expressed as mean ± standard error of the mean (SEM). Non-parametric tests were used for data that did not follow a normal distribution. Nonparametric tests for two independent samples were performed with the Wilcoxon rank-sum test. For nonparametric tests between multiple groups, the Kruskal-Wallis test followed by Dunn’s multiple comparison test was used. Significance levels are indicated by p-values < 0.05(*), and p < 0.01(**).

Figure 10.

LIMP II/ATG8-TBC1D15-Dynamin2/RAB7 pathway might involve microglia lysosome membrane swelling in AD. (A) BV2 was treated with Aβ and controlled with PBS, and the Western blot showed the expression level of LIMP II, ATG8, TBC1D15, Dynamin2, and RAB7. n=3/group. (B) BV2 was treated with Aβ or Aβ combined with siRNA LIMP II, controlled with PBS. Western blot showed the expression level of LIMP II, ATG8, TBC1D15, Dynamin2, and RAB7. n=3/group. (C) BV2 was treated with Aβ or Aβ combined with siRNA ATG8 and controlled with PBS. Western blot showed the expression level of LIMP II, ATG8, TBC1D15, Dynamin2, and RAB7.n=3/group. (D) BV2 was treated with Aβ or Aβ combined with siRNA LIMP II, controlled with PBS. Immunofluorescence showed the co-localization of LAMP1, LIMP II, and TBC1D15.n=3/group. Scale bar=20μm. (E) BV2 was treated with Aβ or Aβ combined with siRNA ATG8 and controlled with PBS. Immunofluorescence showed the co-localization of LAMP1, ATG8, and TBC1D15.n=3/group. Scale bar=20 μm. F. Brain slices from these animals were subjected to 6E10 antibody staining (red). n=3/group. Scale bars: 50 μm. (Statistical analysis graphs are available in Figure S3 and Figure S4.)

Figure 11.

Schematic diagram of the mechanism of this study. (Created with BioRender.com)