The Novel Omega-6 Fatty Acid Docosapentaenoic Acid Positively Modulates Brain Innate Immune Response for Resolving Neuroinflammation at Early and Late Stages of Humanized APOE-Based Alzheimer's Disease Models

Abstract

Neuroinflammation plays a crucial role in the development and progression of Alzheimer's disease (AD), in which activated microglia are found to be associated with neurodegeneration. However, there is limited evidence showing how neuroinflammation and activated microglia are directly linked to neurodegeneration in vivo. Besides, there are currently no effective anti-inflammatory drugs for AD. In this study, we report on an effective anti-inflammatory lipid, linoleic acid (LA) metabolite docosapentaenoic acid (DPAn-6) treatment of aged humanized EFAD mice with advanced AD pathology. We also report the associations of neuroinflammatory and/or activated microglial markers with neurodegeneration in vivo. First, we found that dietary LA reduced proinflammatory cytokines of IL1-β, IL-6, as well as mRNA expression of COX2 toward resolving neuroinflammation with an increase of IL-10 in adult AD models E3FAD and E4FAD mice. Brain fatty acid assays showed a five to six-fold increase in DPAn-6 by dietary LA, especially more in E4FAD mice, when compared to standard diet. Thus, we tested DPAn-6 in aged E4FAD mice. After DPAn-6 was administered to the E4FAD mice by oral gavage for three weeks, we found that DPAn-6 reduced microgliosis and mRNA expressions of inflammatory, microglial, and caspase markers. Further, DPAn-6 increased mRNA expressions of ADCYAP1, VGF, and neuronal pentraxin 2 in parallel, all of which were inversely correlated with inflammatory and microglial markers. Finally, both LA and DPAn-6 directly reduced mRNA expression of COX2 in amyloid-beta42 oligomer-challenged BV2 microglial cells. Together, these data indicated that DPAn-6 modulated neuroinflammatory responses toward resolution and improvement of neurodegeneration in the late stages of AD models.

Keywords: APOE; Alzheimer's disease; EFAD; docosapentaenoic acid; fatty acid; linoleic acid; neuroinflammation.

Figure 1

The high n-6 linoleic acid diet reduced brain hippocampal tissue pro-inflammatory cytokines in EFAD mice. (A) High linoleic acid (n-6 or LA) diet reduced brain pro-inflammatory cytokine IL-1β in E3FAD (p < 0.0001) and a trend in E4FAD mice (p = 0.1) compared to the EFAD mice on standard (ST) diet. (B) High n-6 diet reduced pro-inflammatory cytokine IL-6 in E3FAD (p < 0.0001) and E4FAD mice (p < 0.5). (C) High n-6 diet did not change pro-inflammatory cytokine TNF-α. (D) High n-6 diet increased the inflammation-resolving cytokine IL-10 in both E3FAD (p < 0.05) and E4FAD mice (p < 0.001). Data represent the mean ± SEM. Significance was determined by one-way ANOVA, *p < 0.05, ***p < 0.001, and ****p < 0.0001.

Figure 2

DPAn-6 reduced mRNA expression of proinflammatory cytokine and cytokine receptors in the brain of aged E4FAD mice. In RNA-Seq data, (A) DPAn-6 reduced interleukin-1 receptor-like 2 (ILRL2, p < 0.01). (B) DPAn-6 reduced interleukin 6 receptor alpha (IL6RA, p = 0.01). (C) DPAn-6 reduced IL6 signal transducer (IL6ST, p = 0.01). (D) DPAn-6 reduced tumor necrosis factor receptor 2 (TNFR2, p = 0.01). (E) DPAn-6 reduced tumor necrosis factor-related apoptosis-inducing ligand (TRAIL, also known as tumor necrosis factor superfamily, member 10 (TNFSF10), p < 0.05)). (F) DPAn-6 decreased IL-10 receptor beta (IL-10Rß, p < 0.05). The RNA-Seq data were normalized by Count Per Million (CPM) and presented as Log2 expression. Data represent the mean ± SEM. Significance was determined using an unpaired, two-tailed Student's t-test, *p < 0.05, and **p < 0.01.

Figure 3

DPAn-6 attenuated microgliosis and reduced mRNA expression of microglial markers in aged E4FAD mice. (A,B) Immunostaining for the microglial marker Iba1 showed that DPAn-6 altered microglia morphology from an overactive hypertrophied shape to small ramified cell shape. (C) Quantification showed reduction of numbers of hypertrophied microglia and cellular size of activated microglia by DPAn-6 (p < 0.05). (D–F) DPAn-6 suppressed microglial gene expressions of TMEM119 (p < 0.01, D), CD68 (p = 0.01, E), and TREM2 (p < 0.05, F) compared to controls (CTRL) on standard diet. (G–I) TREM2 was positively correlated with Iba1 (Aif1, p = 0.0002, R² = 0.78, G), TMEM119 (p = 0.0002, R² = 0.77, H), and CD68 (p = 0.0148, R² = 0.46, I). (J–N) Activated microglial marker CD68 was positively corelated with IL1RL2 (p < 0.0001, R² = 0.799, J), IL6RA (p = 0.0012, R² = 0.667, K), IL6ST (p = 0.0001, R² = 0.788, L), TNFR1 (p = 0.0028, R² = 0.608, M), IL10Rß (p = 0.0011, R² = 0.673, N). The RNA-Seq data were normalized by Count Per Million (CPM) and presented as Log2 expression. Data represents the mean ± SEM. Significance was determined using an unpaired, two-tailed Student's t-test, *p < 0.05, and **p < 0.01.

Figure 4

DPAn-6 reduced apoptosis in aged E4FAD mice. In RNA-Seq data, (A–C) DPAn-6 reduced gene expression of caspase 2 (CASP2, a trend p = 0.064, (A)), caspase 6 (CASP6, p < 0.05, (B)), and caspase 8 (CASP8, a trend p = 0.057, (C)) compared to controls (CTRL) on standard diet (ST). (D,E) Immunofluorescence staining revealed that DPAn-6 reduced caspase-cleaved fragment of actin (Fractin (red), D; p < 0.001, E). Nuclei were stained by DAPI (blue). (F,G) Activated microglial marker CD68 was positively correlated with CASP2 (p = 0.0274, R² = 0.40, (F)) and CASP8 (p = 0.0138, R² = 0.47, (G)). The RNA-Seq data were normalized by Counts Per Million (CPM) and presented as Log2 expression. Data represent the mean ± SEM. Significance was determined using an unpaired, two-tailed Student's t-test, *p < 0.05, and ***p < 0.001.

Figure 5

DPAn-6 increased mRNA expression of ADCYAP1, VGF, and NPTX2, which were inversely correlated with activated microglia in aged E4FAD mice. In RNA-Seq data, (A-C) DPAn-6 significantly increased gene expressions of ADCYAP1 (p < 0.0001, A), VGF (p < 0.01, B), and NPTX2 (p < 0.001, C). (D–F) These neuronal markers were positively correlated with each other. ADCYAP1 was positively correlated with VGF (p = 0.0019, R² = 0.635, D) and NPTX2 (p = 0.0029, R² = 0.605, E). NPTX2 was positively correlated with VGF (p = 0.0003, R² = 0.748, F). (G–I) The microglial activation marker CD68 is was inversely correlated with ADCYAP1 (p = 0.0003, R² = 0.745, G), VGF (p = 0.0003, R² = 0.742, H), and NPTX2 (p = 0.0084, R² = 0.517, I). The RNA-Seq data were normalized by Count Per Million (CPM) and presented as Log2 expression. Data represent the mean ± SEM. Significance was determined using an unpaired, two-tailed Student's t-test, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure 6

Correlation analysis of ADCYAP1, VGF, and NPTX2 with proinflammatory cytokines, cytokine receptors and caspases. In RNA-Seq data, (A–C). IL1RL2 was inversely correlated with ADCYAP1 (p = 0.0003, R² = 0.738, A), VGF (p = 0.0015, R² = 0.653, B), and NPTX2 (p = 0.0019, R² = 0.635, C). (D,E) IL6RA was inversely correlated with ADCYAP1 (p = 0.0052, R² = 0.559, D) and VGF (p < 0.0001, R² = 0.817, E). (F–H) IL6ST was inversely correlated with ADCYAP1 (p = 0.0053, R² = 0.557, F), VGF (p < 0.0001, R² = 0.817, G), and NPTX2 (p = 0.0026, R² = 0.613, H). (I–K) TNFAF10 was inversely correlated with ADCYAP1 (p = 0.0082, R² = 0.519, I), VGF (p = 0.016, R² = 0.456, J), and NPTX2 (p = 0.0109, R² = 0.493, K). (L,M) IL10Rß was inversely correlated with ADCYAP1 (p = 0.0093, R² = 0.507, L) and VGF (p = 0.0153, R² = 0.461, M). (N,O) CASP2 was inversely correlated with VGF (p = 0.0016, R² = 0.646 N) and NPTX2 (p = 0.0032, R² = 0.598, O).

Figure 7

Linoleic acid and DPAn-6 reduced cyclooxygenases (COXs) in EFAD mice and in Aβ42 oligomers-stimulated microglial BV2 cells. (A) A high LA diet reduced mRNA expression of COX2 in both E3FAD and E4FAD mice compared to standard (ST) diet (p < 0.0001) measured by RT-qPCR. (B) DPAn-6 reduced COX1 gene expression in RNA-Seq data (p < 0.01) in DPAn-6 treatment of aged E4FAD. (C,D) DPA n-6 inhibited Aβ42 oligomer-stimulated elevation of COX2 mRNA measured by RT-qPCR at 1 h and 4 h compared to the Aβ42 oligomer-stimulated cells without DPAn-6 treatment (p < 0.01, C,D). LA showed a trend for short-term protective effects at 1 h (p = 0.056, C) but not at 4 h (p > 0.05, D). Data represent the mean ± SEM. Significance was analyzed by one-way ANOVA (A,C,D) and an unpaired, two-tailed Student's t-test (B), **p < 0.01 and ****p < 0.0001.

Figure 8

Schematic representation of omega-6 fatty acid docosapentaenoic acid (DPAn-6) positively resolving neuroinflammation in early and late stages of Alzheimer's disease (AD) models with the background of humanized APOE isoforms. In aging, MCI and early-stage AD, Aβ and other cellular debris can induce innate immune responses to activate microglia for phagocytosis (74). During this process, activated microglia release proinflammatory cytokines that can damage neurons if the response is not resolved adequately. Many studies have indicated that NSAIDs and other anti-inflammatory drugs, including COX2 inhibitors, can inhibit neuroinflammation in animal models (8, 9). However, clinical trials with these drugs have failed to date (12). Here we found that DPAn-6 resolved neuroinflammation in early-stage AD models, suggesting DPAn-6 may function similarly to NSAIDs. Further, with the progression of AD, sustainably activated microglia (overactivated microglia) may lose their phagocytotic function. These overactivated microglia may reduce expression of some known proinflammatory cytokines but sustain increases in cytokine receptors of IL1RL2, IL6RA, IL6ST, TNFR2, TRAIL, and IL10Rß (–43), which may further damage neurons. Here we found that DPAn-6 suppressed the expression of these selected cytokines and cytokine receptors at late stages in the E4FAD model. It also improved toxicity evidenced by reductions of caspases and caspase-cleavage fragments. DPAn-6, which is normally produced in the liver and can be taken up by the brain in response to high n-6 diets, can resolve neuroinflammation at the late stage of AD models with advanced AD pathology in the background of APOE4 isoform, a strong genetic risk factor for AD.