Omega-3 EPA Supplementation Shapes the Gut Microbiota Composition and Reduces Major Histocompatibility Complex Class II in Aged Wild-Type and APP/PS1 Alzheimer's Mice: A Pilot Experimental Study

Abstract

Background/Objectives: Neuroinflammation, a hallmark of Alzheimer's disease (AD), is characterized by elevated levels of inflammatory signaling molecules, including cytokines and eicosanoids, as well as increased microglial reactivity, and is augmented by gut microbiota dysbiosis via the gut-brain axis. We conducted a pilot experiment to elucidate the anti-inflammatory effects of dietary omega-3 polyunsaturated fatty acid (ω-3 PUFA) eicosapentaenoic acid (EPA) on the gut microbiota and neuroinflammation. Methods: Female APP/PS1 mice (TG) and non-transgenic littermates (WT), 13-14 months old, were fed a diet supplemented with 0.3% EPA or control chow for 3 weeks. The gut microbiota composition, hippocampal and plasma eicosanoids levels, platelet activation, and microglial phagocytosis, as well as the brain and retinal genes and protein expression, were analyzed. Results: EPA supplementation decreased the percentage of Bacteroidetes and increased bacteria of the phylum Firmicutes in APP/PS1 and WT mice. Inflammatory lipid mediators were elevated in the hippocampus of the TG mice, accompanied by a reduction in the endocannabinoid docosahexaenoyl ethanolamide (DHEA). Dietary EPA did not affect hippocampal lipid mediators, but reduced the levels of arachidonic-derived 5-HETE and N-arachidonoylethanolamine (AEA) in WT plasma. Moreover, EPA supplementation decreased major histocompatibility complex class II (MHCII) gene expression in the retina in both genotypes, and MHCII+ cells in the hippocampus of TG mice. Conclusions: This pilot study showed that short-term EPA supplementation shaped the gut microbiota by increasing butyrate-producing bacteria of the Firmicutes phylum and decreasing Gram-negative LPS-producing bacteria of the Bacteroidetes phylum, and downregulated the inflammatory microglial marker MHCII in two distinct regions of the central nervous system (CNS). Further investigation is needed to determine whether EPA-mediated effects on the microbiome and microglial MHCII have beneficial long-term effects on AD pathology and cognition.

Keywords: APP/PS1; Alzheimer’s disease; MHCII; eicosanoids; eicosapentaenoic acid (EPA); gut microbiota; lipid droplets; microglia; omega-3 polyunsaturated fatty acids; phagocytosis.

Figure 1

Microbiome composition. Feces were analyzed for microbiome composition; operational taxonomic units (OTUs) that were identified in all samples and exhibited an average abundance of at least 5% were included in the statistical analysis. The graphs show (A) the alpha diversity of the samples, indicated by the Shannon index; (B,C) the microbiota on the hierarchical taxonomic level of phylum; (D) the ratio of Firmicutes to Bacteroidetes bacteria; (E–G) the microbiota on the hierarchical taxonomic level of class; (H–J) the microbiota on the hierarchical taxonomic level of order; and (K–M) the microbiota on the hierarchical taxonomic level of family. (A–M) A two-way ANOVA for the main diet and genotype effects was performed, followed by Šídák’s multiple comparison test. A 95% confidence interval was used, and p-values less than 0.05 were considered significant. The graphs are presented as the mean ± standard deviation, with all data points shown.

Figure 2

Analysis of platelet activation and plasma lipid mediators. Platelet activation was assessed via flow cytometry. CD9 was used to label platelets, and CD62P was used as an activation marker. The graphs show (A) the percentage of gated CD9+ platelets positive for CD62P, and the quantification of plasma (B) eicosanoids and (C) endocannabinoids and endocannabinoid-like amides. (A–C) A two-way ANOVA for the main diet and genotype effects was performed, followed by Šídák’s multiple comparison test. A 95% confidence interval was used, and p-values less than 0.05 were considered significant. The graphs are presented as the mean ± standard deviation, with all data points shown.

Figure 3

Retinal gene expression and vascular integrity. Retinal gene expression of Trem2 (A) and H2-Aa (B) was analyzed with RT-qPCR. (C) Representative retinal tissue immunohistochemically stained with Collagen IV (ColIV) for blood vessels (in red) and neural/glial antigen 2 (NG2) for pericytes (in white); the scale size is 50 µm. (D) The total length of retinal capillaries per field of view. (E) The pericyte count per vessel length. (A,B,D,E) A two-way ANOVA for main diet and genotype effects was performed. A 95% confidence interval was used, and p-values less than 0.05 were considered significant. The graphs are presented as the mean ± standard deviation, with all data points shown.

Figure 4

Hippocampal lipid mediators. Quantification of hippocampal (A) eicosanoids and (B) endocannabinoids and endocannabinoid-like amides. In the TG+EPA group, one of the samples was unfortunately lost during sample preparation for LC-MS/MS analysis; therefore only 3 data points are available. A two-way ANOVA for the main diet and genotype effects was performed. A 95% confidence interval was used, and p-values less than 0.05 were considered significant. The graphs are presented as the mean ± standard deviation, with all data points shown.

Figure 5

Microglia analysis. (A) Representative images of hippocampal brain sections stained with 4′,6-diamidino-2-phenylindole (DAPI) for cell nuclei (in blue), calcium-binding adapter molecule 1 (Iba1) for microglia (in white), and major histocompatibility complex class II (MHCII) (in green); the scale size is 50 µm. (B) Quantification of MHCII+ cells in TG mice. Student’s t-test was performed with a 95% confidence interval, and a p-value below 0.05 was considered significant. (C,D) Primary microglia of one hemisphere were isolated, and a phagocytosis assay with fluorescent labeled amyloid peptide (Aβ-488) was performed and measured on a flow cytometer. The graphs show the percentage of phagocytic cells (C) and the median fluorescence as a measure of peptide uptake (D). (E) Representative pictures of hippocampal sections stained with Bodipy for lipid droplets (in green) and Iba1 for microglia (in red); the scale size is 25 µm. The lipid droplet count (F), total volume per field of view (G), and mean volume (H) were measured with Imaris (Bitplane) software. (C,D,F–H) A two-way ANOVA for the main diet and genotype effects was performed. A 95% confidence interval was used, and p-values less than 0.05 were considered significant. The graphs are presented as the mean ± standard deviation, with all data points shown.