Liver-specific adiponectin gene therapy suppresses microglial NLRP3-inflammasome activation for treating Alzheimer's disease

Abstract

Adiponectin (APN) is an adipokine which predominantly expresses in adipocytes with neuroprotective and anti-inflammatory effects. We have recently indicated that circulatory trimeric APN can enter the brain by crossing the blood-brain barrier (BBB) and modulate microglia-mediated neuroinflammation. Here, we found that the microglial NLR family pyrin domain containing 3 (NLRP3)-inflammasome activation was exacerbated in APN^-/-5xFAD mice in age-dependent manner. The focus of this study was to develop a new and tractable therapeutic approach for treating Alzheimer's disease (AD)-related pathology in 5xFAD mice using peripheral APN gene therapy. We have generated and transduced adeno-associated virus (AAV2/8) expressing the mouse mutated APN gene (APN^C39S) into the liver of 5xFAD mice that generated only low-molecular-weight trimeric APN (APN^Tri). Single dose of AAV2/8-APN^C39S in the liver increased circulatory and cerebral APN levels indicating the overexpressed APN^Tri was able to cross the BBB. Overexpression of APN^Tri decreased both the soluble and fibrillar Aβ in the brains of 5xFAD mice. AAV2/8-APN^Tri treatment reduced Aβ-induced IL-1β and IL-18 secretion by suppressing microglial NLRP3-inflammasome activation. The memory functions improved significantly in AAV-APN^Tri-treated 5xFAD mice with reduction of dystrophic neurites. These findings demonstrate that peripheral gene delivery to overexpress trimeric APN can be a potential therapy for AD.

Keywords: Adiponectin; Alzheimer’s Disease; Gene delivery; NLRP3; Neuroinflammation.

Fig. 1.

Adiponectin deficiency exacerbates NLRP3-inflammasome activation in 5xFAD mice. A Representative images of co-immunofluorescent staining of NLRP3 and Iba1 in the hippocampus of 3, 9, and 15-month-old 5xFAD and APN^−/−5xFAD mice. (n = 3 mice for each experimental group) (Magnification: 10 × 20; Magnified: 10 × 40; Scale bar: 50μm). B Relative intensity of NLRP3 immunofluorescent staining in the hippocampus. C Western blotting assay of ASC, Pro-caspase-1, and cleaved caspase-1 in the hippocampus of 9-month-old 5xFAD and APN^−/−5xFAD mice. (n = 3 mice for each experimental group). D Densitometric analysis of ASC, Pro-caspase-1, and cleaved caspase-1 levels. E ELISA assay of hippocampal IL-1β of 3, 9, and 15-month-old 5xFAD and APN^−/−5xFAD mice. (n = 3 mice for each experimental group). Data were presented as the mean ± s.e.m. *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis was performed by unpaired t tests.

Fig. 2

Trimeric adiponectin suppresses Aβ oligomer-induced IL-1β secretion from microglia via NLRP3-inflammsome pathway. A Representative western blotting image indicates pre-treatment of mammalian recombinant trimeric APN protein reduced NLRP3 level in Aβ oligomer-induced microglia. (n = 3 cultures for each group with duplicated experiment). B Densitometric analysis of NLRP3 and α-tubulin levels. C ELISA assay of IL-1β in the microglia BV2 cells. (n = 6 for each group). D ELISA assay of IL-18 in the microglia BV2 cells. (n = 6 for each group). Data were presented as the mean ± s.e.m for at least three independent experiments, and each performed in duplicates (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis was performed by one-way ANOVA followed with Bonferroni’s post hoc comparison tests.

Fig. 3

AAV2/8-mediated gene delivery of APN^C39S transgene to 5xFAD exhibits dose-dependent transduction efficiency, liver-specific expression of transgene and increase of circulatory adiponectin. A AAV plasmids design of control GFP and APN^C39S. B Fluorescence of GFP in the liver and duodenum of AAV2/8-GFP-treated 5xFAD 1 month after injection. (n = 3 mice). C Vector copy number of APN transgene in the liver, duodenum and muscle of AAV2/8-APN^Tri-treated 5xFAD at different doses 1 month after injection. (n = 3 mice for each experimental group). D ELISA analysis of plasma APN in AAV2/8-APN^Tri-treated 5xFAD at different doses 1 month after injection. (n = 6 mice for each experimental group). E Fold change analysis of plasma APN in AAV2/8-APN^Tri-treated 5xFAD at different doses 1 month after injection (normalized by wildtype plasma APN). Data were presented as the mean ± s.e.m. *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis was performed by one-way ANOVA followed with Bonferroni’s post hoc comparison tests.

Fig. 4

Single dose of AAV2/8-APN^Tri increases trimeric adiponectin in circulation and cerebral adiponectin levels in 5xFAD mice 4 months after injection. A Preclinical study scheme of single dose AAV injection at 4-month-old 5xFAD. B ELISA analysis of plasma APN in wildtype, AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD 4 month after injection. (n = 6 mice for each experimental group). C Non-denaturing SDS PAGE analysis of plasma APN by anti-flag and anti-APN antibodies indicated the levels of trimeric APN in the circulation 4 months after injecting AAV2/8-GFP and AAV2/8-APN^Tri. (n = 4 mice for each experimental group). D Densitometric analysis of APN^Tri in AAV2/8-GFP-treated and AAV2/8-APN^Tri-treated 5xFAD. E Immunofluorescent staining of APN using anti-flag and anti-APN antibodies in the cerebral cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD 4 month after injection. (n = 4 mice for each experimental group) (Magnification: 10 × 20; Scale bar: 50μm). Data were presented as the mean ± s.e.m.; *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis was performed by one-way ANOVA followed with Bonferroni’s post hoc comparison tests and unpaired t test.

Fig. 5

AAV2/8-APN^Tri treatment improve memory functions and reduces dystrophic neurites in the 5xFAD mice. A The escape latency of different treatment group in the visible platform of Morris water maze (MWM) test. (n = 9 mice for wildtype and n = 11 for both AAV-treated groups). B The escape latency in 5-day sessions performed by different treatment group in MWM test. (n = 9 mice for wildtype and n = 11 for both AAV-treated groups). C Representative tracks of different treatment groups in the hidden sessions (H1, H3 & H5). D Probe test indicates the percentage of time spent in the target quadrant (platform location) by different mouse treatment groups. (n = 9 mice for wildtype and n = 11 for both AAV-treated groups). E Representative images of co-staining of Aβ and Lamp1 in the cortex and hippocampus of AAV2/8-GFP-treated (n = 6), and AAV2/8-APN^Tri-treated 5xFAD (n = 6) (Total 780 plaques were analyzed; Magnification: 10 × 40; Scale bar: 10μm). F, G Quantification of total dystrophic volume represented by Lamp1 staining surrounding the compact Aβ and filamentous Aβ. Data were presented as the mean ± s.e.m. *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis of MWM tests (A, B) was analyzed with two-way ANOVA followed by Bonferroni’s post hoc test. Statistical analysis in (D) was performed by one-way ANOVA followed with Bonferroni’s post hoc comparison tests and in (F, G) were performed by unpaired t test.

Fig. 6

AAV2/8-APN^Tri treatment reduces amyloid pathology and microglia activation in the 5xFAD mice. A Immunohistochemistry analysis of Aβ (4G8; brown) in the hippocampus and cerebral cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (n = 6 mice for each experimental group) (Magnification: 10 × 10; Scale bar: 50μm). B Quantitative analysis of the Aβ loading and number of Aβ deposits in the hippocampus and cortex. C, D ELISA analysis of Aβ₄₂ and Aβ₄₀ levels in the hippocampus and the frontal cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (n = 5-6 mice for each experimental group). E Iba1 (red) and thioflavin S (green) double-immunofluorescent staining in the hippocampus of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (n = 6 mice for each experimental group) (Magnification: 10 × 10; Scale bar: 50μm). F Quantitative analysis of the Aβ loading in the hippocampus of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. G Relative intensity of Iba1 in the hippocampus of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. Data were presented as the mean ± s.e.m. n.s. not significant; *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analysis was performed by unpaired t tests.

Fig. 7

Recombinant APN increases microglial phagocytic activities induced by Aβ peptides and oligomers in vitro. A Representative images of FAM-Aβ_1-42 phagocytosis in BV2 microglia cells. BV2 cells were pre-incubated with APN for 2 h followed by incubation with FAM-Aβ_1-42 (green) for 24 h before staining Iba1 (red). Scale bar = 100 μm. B The bar graphs showed percentage of phagocytic cells Data were presented as the mean ± SEM for at least three independent experiments (n = 3). C Representative images of latex bead phagocytosis in BV2 microglia cells. BV2 cells were pre-incubated with APN for 2 h followed by incubation with AβO for 24 h and then were loaded with fluorescent beads (green) for 1 h at 37 °C before staining Iba1 (red). Scale bar = 100 μm. D The bar graphs show percentage of phagocytic cells Data were presented as the mean ± SEM for at least three independent experiments (n = 3). One-way ANOVA with Tukey’s multiple comparison test revealed difference between groups. *p < 0.05

Fig. 8

AAV2/8-APN^Tri treatment reduces NLRP3-inflammasome activation in the 5xFAD mice. A Iba1 (red) and NLRP3 (green) double-immunofluorescent staining in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (n = 6 mice for each experimental group) (Magnification: 10 × 20; Scale bar: 50μm). B Relative intensity of microglial NLRP3 in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (C) Relative intensity of NLRP3 in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (D) Iba1 (green) and ASC (red) double-immunofluorescent staining in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (n = 6 mice for each experimental group) (Magnification: 10 × 20; Scale bar: 50μm). E Percentage of Iba1 + ASC + cells in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. F Relative intensity of microglial ASC in the hippocampus and cortex of AAV2/8-GFP-treated, and AAV2/8-APN^Tri-treated 5xFAD. (G) Western blotting assay of ASC, Pro-caspase-1, and cleaved caspase-1 in the hippocampus of AAV2/8-GFP-treated and AAV2/8-APN^Tri-treated 5xFAD. (n = 5 mice for each experimental group). H Densitometric analysis of ASC, pro-caspase-1, and cleaved caspase 1. (I) ELISA assay of hippocampal IL-1β of AAV2/8-GFP-treated and AAV2/8-APN^Tri-treated 5xFAD. (n = 6 mice for each experimental group). (J) ELISA assay of hippocampal IL-18 of AAV2/8-GFP-treated and AAV2/8-APN^Tri-treated 5xFAD. (n = 5 mice for each experimental group). Data were presented as the mean ± s.e.m. n.s. not significant; *p < 0.05; **p < 0.01; ***p < 0.001. Statistical analyses in (B, C, E, F, G, and H) were performed by unpaired t tests and (I, J) were performed by one-way ANOVA followed with Bonferroni’s post hoc comparison tests.

Fig. 9

Schematic summary of this study. A Strategy of liver-specific adiponectin gene delivery using recombinant AAV as a treatment for Alzheimer’s disease. Intravenous injection of AAV2/8-APN^Tri which carries APN mutant (C39S) under the ApoE enhancer and hAAT promoter overexpresses trimeric adiponectin in the liver. Increasing circulatory trimeric APN which crosses the blood–brain barrier to inhibits microglia-mediated neuroinflammation. B Molecular mechanism of the model. AβO stimulates nuclear translocation of NF-κB through binding of toll-like receptors (TLRs) or CD36. Trimeric adiponectin binds to adiponectin receptor 1 (AdipoR1) and activates AMPK to suppress NF-κB nuclear translocation. This decreases the mRNA transcriptions and the levels of NLRP3, ASC, pro-caspase 1, IL-1β, and IL-18