Receptors for pro-resolving mediators are increased in Alzheimer's disease brain

Abstract

Neuroinflammation is a key element of AD pathology and conceivably a result of a disturbed resolution. Resolution of inflammation is an active process which is strictly orchestrated following the acute inflammatory response after removal of the inflammatory stimuli. Acute inflammation is actively terminated by specialized pro-resolving mediators (SPMs) thereby promoting healing and return to homeostasis. Failed resolution may contribute to persistent neuroinflammation and aggravate AD pathology. BLT1 (leukotriene B₄ receptor) and ChemR23 (chemerin receptor 23) are receptors for the SPM resolvin (Rv) E1 and are important clinical targets for ending inflammation. In AD, the levels of SPMs are decreased, and pro-inflammatory mediators are increased. In the current study, the distribution of BLT1 and ChemR23 receptors in control brains and in AD as well as correlations with AD pathology was examined for the first time. BLT1 and ChemR23 were analyzed in different regions of post-mortem human brain from cases with AD, early-onset AD and mild cognitive impairment (MCI) and healthy controls, using western blotting and immunohistochemistry. BLT1 and ChemR23 were detected in neurons and glial cells in all examined regions of the human brain, with markedly higher levels in AD than in controls. The receptor levels correlated with the density of staining for the inflammation markers HLA-DR and YKL-40 for microglia and astrocytes, respectively, and elevated staining coincided with high Braak stages in AD. The relative staining densities of these receptors were higher in the basal forebrain, cingulate gyrus and hippocampal regions compared to the cerebellum and frontal cortex (BA46). In conclusion, alterations in the expression of the resolution receptor BLT1 in AD have not been reported previously and the changes in both BLT1 and ChemR23 suggest a disturbed resolution pathway in several regions of the AD brain that may play a role in disease pathology.

Keywords: Alzheimer's disease; Braak; amyloid; immunohistochemistry; resolution of inflammation; specialized pro-resolving mediators.

Figure 1

BLT1—receptor for the SPM resolvin E1 (RvE1) in different regions of Alzheimer's disease (AD) and control brain. A–C. Sections from the human hippocampus including CA1, CA2, CA3 and the dentate gyrus (DG) from AD and control cases were incubated with antibodies to BLT1. The signal for BLT1 is clearly stronger in AD in all of these areas, with the largest difference in CA2 and CA3. B. The granular appearance of the staining in pyramidal neurons is seen in both control and AD (black and white arrow heads in CA2 and CA3). A similarly strong level of BLT1 immunoreactivity in AD can be seen in neurons in the hilus region, also with a granular appearance. The BLT1 signal is also seen in axons close to the neuronal cell somata (short black arrows in CA3 and hilus). A strong signal for BLT1 is seen in glial cells, also with a granular appearance (black arrow heads in CA2, CA3, hilus and DG (C)). C. Granular cells in the DG are also positive for BLT1 (long black arrows), with markedly stronger staining in AD. Bars = 50 µm. The high magnification micrographs are part of the corresponding low magnification micrographs. BLT1 = leukotriene B4 receptor, CA = cornu Ammonis, SPM = specialized pro‐resolving mediator.

Figure 2

BLT1—receptor for the SPM resolvin E1 (RvE1) in different regions of Alzheimer's disease (AD) and control brain. A–B. Sections from the human entorhinal cortex (ENT), basal forebrain (BF), Brodmann area (BA) 46, cingulate gyrus (CG), cerebellum (CB) and corpus callosum (CC) from AD and control cases were incubated with antibodies to BLT1. The signal for BLT1 in neurons is clearly stronger in AD in all of these areas, but with a minor difference in the Purkinje cells of the CB (short arrows). B. A granular staining is evident in the neurons of both control and AD brains, for example, seen here in the large neurons of BF in AD (arrowhead), extending into the axon (long arrows), and in BA46 in a control case (arrowhead). Varicose fibers with BLT1 immunoreactivity can be seen in higher magnification of the BA46 cortex (long arrows). Numerous glial cells with a strong signal for BLT1 are seen in the CC in AD, whereas only a few weakly labeled glia can be seen in control brain. Bars = 50 µm. BLT1 = leukotriene B4 receptor, SPM = specialized pro‐resolving mediator.

Figure 3

Colocalization of BLT1 with HLA‐DR and TM119 in microglia. Z‐stack images were created by merging serial confocal microscopy scans of 5 µm sections of double immunofluorescence labeling for BLT1 and HLA‐DR or TM119. Co‐localization of the signal for BLT1 (green) can be seen with HLA‐DR (red) (A–C) or with TM119 (red) (G–I) in the gray matter of CG of an Alzheimer's disease (AD) case. The graphs show the fluorescence intensity profile from a line crossing over the microglia shown in E and K, and intensity peaks for both red and green display the co‐localization of BLT1 with HLA‐DR (D, E) and TM119 (J, K). The ortho‐view of the z‐stack images shows that BLT1 is expressed in the cell soma of microglia (F, L). Bar = 20 µm. BLT1 = leukotriene B₄ receptor, CG = cingulate gyrus, HLA‐DR = human leukocyte antigen‐D‐related, TM119 = transmembrane protein 119.

Figure 4

Colocalization of BLT1 with GFAP and S100β in astrocytes. Z‐stack images were created by merging serial confocal microscopy scans of 5 µm sections of double immunofluorescence labeling for BLT1 and GFAP or S100β. Co‐localization of the signal for BLT1 (green) can be seen with GFAP (red) (A–C) or s100β (red) (G–I) in the gray matter of CG of an Alzheimer's disease (AD) case. The graphs show the fluorescence intensity profile from a line crossing over the astrocytes shown in E and K, and intensity peaks for both red and green display the co‐localization of BLT1 with GFAP (D, E) and S100β (J, K). The ortho‐view of the z‐stack images shows that BLT1 is expressed in the cell soma of the astrocytes (F, L). Bar = 20 µm. BLT1 = leukotriene B₄ receptor, CG = cingulate gyrus, GFAP = glial fibrillary acidic protein, s100β = S100 calcium binding protein β.

Figure 5

Analysis of BLT1 immunoreactivity in different regions of Alzheimer's disease (AD) and control brain. A. Densitometric analysis of BLT1 immunoreactivity in CA1, CA2, entorhinal cortex (ENT), dentate gyrus (DG), basal forebrain (BF), Brodmann area (BA) 46, cingulate gyrus (CG) and cerebellum (CB), as well as visual scoring of the BLT1 staining in the same regions. The visual scoring is performed on glia (G) and neurons (N). B. Western blot (WB) analysis of BLT1 receptor protein of 55‐kDa molecular weight (MW). Representative images of bands from each region of AD and control cases. Horizontal bars indicate median. C. Expression of BLT1 during disease progression. Increased staining intensity of BLT1 in all investigated brain regions is associated with disease pathology based on Braak stage. BLT1 = leukotriene B4 receptor, CA = cornu Ammonis, eAD = early onset AD, HPC = hippocampus, MCI = mild cognitive impairment, N/A = not available.

Figure 6

ChemR23—receptor for the SPM resolvin E1 (RvE1) in different regions of the Alzheimer's disease (AD) and control brain. A–B. Sections from the hippocampus including CA1, CA2 and the dentate gyrus (DG) from AD and control cases, and a section of the hilus region from AD, were incubated with antibodies to ChemR23. The signal for ChemR23 is clearly stronger in AD in all of these areas, with a particularly strong signal in pyramidal cells in the CA regions and hilus. B. The signal for ChemR23 has a granular appearance (black and white arrow heads in CA1 and CA2). In DG, a few granular cells have a weak ChemR23 labeling (black arrow heads). Glial cells with a strong signal in AD are seen in all of the regions (white arrows in CA1, CA2, DG and hilus). Bars = 50 µm. ChemR23 = chemerin‐like receptor‐1, CA = cornu Ammonis, SPM = specialized pro‐resolving mediator.

Figure 7

ChemR23—receptor for the SPM resolvin E1 (RvE1) in different regions of the Alzheimer's disease (AD) and control brain. Sections from the entorhinal cortex (ENT), basal forebrain (BF), Brodmann area (BA) 46, cingulate gyrus (CG), cerebellum (CB) and corpus callosum (CC) from AD and control cases were incubated with antibodies to ChemR23. The signal for ChemR23 in neurons is clearly stronger in AD in all of these areas, but with a smaller difference with regard to CB, including the Purkinje cells (arrows). Particularly strong signal for ChemR23 is observed in the large neurons of BF. Note also neurons with distorted morphology in AD (arrow heads in BF and BA46). Many glial cells with a strong signal for BLT1 are seen in the CC in AD, whereas only few weakly labeled glia can be seen in control. Bar = 50 µm. ChemR23 = chemerin‐like receptor‐1, SPM = specialized pro‐resolving mediator.

Figure 8

Analysis of ChemR23 immunoreactivity in different regions of Alzheimer's disease (AD) and control brain. A. Densitometric analysis of ChemR23 immunoreactivity in CA1, CA2, entorhinal cortex (ENT), dentate gyrus (DG), basal forebrain (BF), Brodmann area (BA) 46, cingulate gyrus (CG) and cerebellum (CB), as well as visual scoring of the BLT1 staining in the same regions. The visual scoring is performed on glia (G) and neurons (N). B. Western blot (WB) analysis for ChemR23 receptor protein of 42‐kDa molecular weight (MW). Representative images of bands from AD and control cases. Horizontal bars indicate median. C. Staining intensity of ChemR23 using densitometry analysis is summed for all brain regions investigated to show the relationship between ChemR23 expression and AD pathology (NFTs and Aβ plaques). Aβ = β amyloid, CA = cornu Ammonis, ChemR23 = chemerin‐like receptor‐1, eAD = early onset AD, HPC = hippocampus, MCI = mild cognitive impairment, N/A = not available, NFTs = neurofibrillary tangles.

Figure 9

Microglial activation in different regions of the Alzheimer's disease (AD) and control brain. A. Sections from the CA1, CA2, dentate gyrus (DG) and hilus region from AD and control cases were incubated with antibodies to human leukocyte antigen–D‐related (HLA‐DR). A clear increase in labeling intensity for HLA‐DR and density of labeled microglia is seen in AD compared to control in all of these regions. B. High magnification of CA2 (part of CA2 (AD) in A) shows unlabeled neurons (arrow heads) surrounded and contacted by microglial processes. A cluster of microglia is observed in the hilus region of an AD brain. Bars = 50 µm. CA = cornu Ammonis.

Figure 10

Microglia in different regions of the Alzheimer's disease (AD) and control brain. A. Sections from the entorhinal cortex (ENT), basal forebrain (BF), Brodmann area (BA) 46, cingulate gyrus (CG) and cerebellum (CB) from AD and control cases were incubated with antibodies to human leukocyte antigen–D‐related (HLA‐DR). A clear increase in labeling intensity for HLA‐DR and density of labeled microglia is seen in AD compared to control in all of these regions. B. Unlabeled neurons (arrow heads) in BF of an AD brain are surrounded and contacted by microglial processes. The low density of labeled microglia in control is exemplified by the high‐magnification micrograph of BA46. Bars = 50 µm.

Figure 11

Analysis of the inflammatory marker YKL‐40 in different regions of Alzheimer's disease (AD) and control brain. A. Western blot (WB) analysis of the hippocampus (HPC), and basal forebrain (BF), show higher levels of YKL‐40 in AD patients compared to controls, whereas no difference was seen in Brodmann area (BA) 46 or cingulate gyrus (CG). B. Relationship between YKL‐40 levels and the presence of the E4 allele of apolipoprotein (Apo). Horizontal bars indicate median. YKL40 = Chitinase‐3‐Like Protein 1.

Figure 12

Multivariate analysis (MVA) of data on the SPM‐receptors BLT1 and ChemR23 and the number of HLA‐DR‐positive microglia in different regions of the Alzheimer's disease (AD) and control brain. The orthogonal projections to latent structures (OPLS) analysis was performed on A densitometric measurements of BLT1 and ChemR23 immunohistochemical staining, and B cell counts of HLA‐DR immunoreactive microglia. The data were obtained from analysis of BA46, basal forebrain (BF), CA1, CA2, cerebellum (CB), cingulate gyrus (CG) and dentate gyrus (DG) in post‐mortem brain tissue from individuals diagnosed with AD (black symbols) and cognitively normal controls (gray symbols). For HLA‐DR (B), estimated number of positive cells, as well as the number of positive cells with activated morphology (indicated by a following the region), in the regions were included in the analysis. Furthermore, the analysis of HLA‐DR included data from the white matter (WM) regions of BA46 and CG. The OPLS‐DA models for BLT1 and ChemR23, as well as HLA‐DR, showed good ability to discriminate individuals with AD from cognitively normal controls (R2Y(cum) = 0.804 (A), and R2Y(cum) = 0.693 (B), respectively). * indicates variables with a significant impact on the model. The symbols are coded for the presence of the E2, E3 and E4 alleles of apolipoprotein (Apo) (see legend in figure). # indicates individuals with early‐onset AD (eAD). BA46 = Brodmann area 46, BLT1 = leukotriene B4 receptor 1, ChemR23 = chemokine‐like receptor 23, HLA‐DR = human leukocyte antigen–D‐related, SPM = specialized pro‐resolving mediator.