Resolution of inflammation is altered in Alzheimer's disease

Abstract

Background: Resolution is the final stage of the inflammatory response, when restoration of tissue occurs. Failure may lead to chronic inflammation, which is known as part of the pathology in the brain of individuals with Alzheimer's disease (AD).

Methods: Specialized pro-resolving mediators (SPMs), receptors, biosynthetic enzyme, and downstream effectors involved in resolution were analyzed in postmortem hippocampal tissue from AD patients and non-AD subjects. SPMs were analyzed in cerebrospinal fluid (CSF).

Results: SPMs and SPM receptors were detected in the human brain. Levels of the SPM lipoxin A4 (LXA4) were reduced in AD, both in the CSF and hippocampus. An enzyme involved in LXA4 synthesis and two SPM receptors were elevated in AD brains. LXA4 and RvD1 levels in CSF correlated with Mini-Mental State Examination (MMSE) scores.

Conclusions: A resolution pathway exists in the brain and the alterations described herein strongly suggest a dysfunction of this pathway in AD. MMSE correlations suggest a connection with cognitive function in AD.

Keywords: 15-Lipoxygenase-2; ALX/FPR2; ChemR23; ELISA; FPRL1; Human; Immunohistochemistry; Lipoxin A(4); Mild cognitive impairment; Resolvin D1; Specialized pro-resolving mediators; Tau.

Fig. 1. LXA₄ and RvD1 in human CSF and hippocampus

(A and B) Lower CSF levels of LXA₄ in AD (n = 15) as compared to MCI (n = 20; ANCOVA corrected for age, Bonferroni post hoc, p = 0.006) and SCI (n = 21; age as covariate by ANCOVA corrected to age, Bonferroni post hoc, p < 0.001) groups, whereas RvD1 levels do not differ between the groups. (C) Levels of LXA₄ and RvD1 in the CSF are positively correlated (Spearman’s rho test, r = 0.843, p < 1 × 10⁻⁶). (D and E) Levels of LXA₄ and RvD1 in the CSF are correlated to MMSE score in a positive manner (Spearman’s rho test, r = 0.475, p < 0.0005, and r = 0.343, p < 0.05 respectively). (F) Lower tissue levels of LXA₄ in the hippocampus of AD-patients (n = 7) than in control subjects (n = 7, Mann-Whitney U test, z = −2.236, p = 0.026). (G) No difference found between AD-patients (n = 7) and control subjects with regard to hippocampal levels of RvD1 (n = 7, Mann-Whitney U test, z = −0.575, p = 0.62). Horizontal bar indicates mean (A and B) or median (F). *p< 0.05, **p < 0.01.

Fig. 2. LXA₄R in human hippocampus

(A) Immunohistochemistry of LXA₄R in AD and control shows neuronal (arrow) and glial (arrowhead) staining; no labeling is observed in CA1 pyramidal neurons, neither in AD nor control; neuronal staining in CA2 is similar in AD and control; glial labeling is stronger in AD, in CA1–CA4 as well as in the dentate gyrus; bar = 20 µm. (B) Strongly labeled glial cells containing multiple large vesicles can be seen in AD. The glia in controls have weaker staining and fewer vesicles; bar = 8 µm. (C) Double labeling of cells with LXA₄R (red) and GFAP (blue) or HLA-DR (blue), shows localization of LXA₄R in both astrocytes and microglia; bar = 8 µm. (D) WB data show no statistically significant difference in LXA₄R levels between AD (n = 7) and control (n = 9) cases (Mann-Whitney U test, z = −0.053, p = 0.958; Horizontal bar indicates median).

Fig. 3. ChemR23 in human hippocampus

(A) Immunohistochemistry of ChemR23 in AD and control shows staining in pyramidal cells (solid arrows) and glia (arrow heads), and in dentate gyrus (DG) granular cells (unfilled arrows); in AD ChemR23-labeled neurons in CA1 with strong staining are more numerous than in control; ChemR23 labeling in CA2 neurons and glial cells is stronger in AD than in control. Similarly, the staining of granular cells and glia in the DG is stronger in AD; bar = 20 µm. (B) Strongly labeled glial cells containing multiple large vesicles can be seen in AD. The glia in controls have weaker staining and fewer vesicles; bar = 8 µm. (C) Double labeling of cells with ChemR23 (red) and GFAP (blue) or HLA-DR (blue) shows localization of ChemR23 in both astrocytes and microglia; bar = 8 µm. (D) WB data show higher levels of ChemR23 in AD (n = 7) than in controls (n = 9) (Mann-Whitney U test, z = −2.593, p = 0.008; Horizontal bar indicates median; **p < 0.01).

Fig. 4. PPAR-γ and IL-10 in human hippocampus

(A) The levels of PPAR-γ are higher in AD (n = 7) than in controls (n = 9) (Mann-Whitney U test, z = −3.228, p < 0.001). (B) Levels of IL-10 are lower in AD (n = 8) than in controls (n = 9) (Mann-Whitney U test, z = −2.117, p = 0.036). Horizontal bar indicates median (A, D); *p < 0.05, **p < 0.01.

Fig. 5. 15-LOX-2 in human hippocampus

(A) Immunohistochemistry of AD and control shows glial staining (arrow) with higher intensity and higher numbers of labeled cells in AD; bar = 20 µm. (B) Close-up view of 15-LOX-2; bar = 8 µm. (C) Double labeling of cells with 15-LOX-2 (red) and GFAP (blue) or HLA-DR (blue) show localization of 15-LOX-2 in both astrocytes and microglia; bar = 8 µm. (D) Western blot data show higher levels of 15-LOX-2 in AD (n = 7) than in controls (n = 9) (Mann-Whitney U test, z = −2.382, p = 0.016; Horizontal bar indicates median; *p < 0.05).