1α,25-dihydroxyvitamin D3 and resolvin D1 retune the balance between amyloid-β phagocytosis and inflammation in Alzheimer's disease patients

Abstract

As immune defects in amyloid-β (Aβ) phagocytosis and degradation underlie Aβ deposition and inflammation in Alzheimer's disease (AD) brain, better understanding of the relation between Aβ phagocytosis and inflammation could lead to promising preventive strategies. We tested two immune modulators in peripheral blood mononuclear cells (PBMCs) of AD patients and controls: 1α,25(OH)2-vitamin D3 (1,25D3) and resolvin D1 (RvD1). Both 1,25D3 and RvD1 improved phagocytosis of FAM-Aβ by AD macrophages and inhibited fibrillar Aβ-induced apoptosis. The action of 1,25D3 depended on the nuclear vitamin D and the protein disulfide isomerase A3 receptors, whereas RvD1 required the chemokine receptor, GPR32. The activities of 1,25D3 and RvD1 commonly required intracellular calcium, MEK1/2, PKA, and PI3K signaling; however, the effect of RvD1 was more sensitive to pertussis toxin. In this case study, the AD patients: a) showed significant transcriptional up regulation of IL1RN, ITGB2, and NFκB; and b) revealed two distinct groups when compared to controls: group 1 decreased and group 2 increased transcription of TLRs, IL-1, IL1R1 and chemokines. In the PBMCs/macrophages of both groups, soluble Aβ (sAβ) increased the transcription/secretion of cytokines (e.g., IL1 and IL6) and chemokines (e.g., CCLs and CXCLs) and 1,25D3/RvD1 reversed most of the sAβ effects. However, they both further increased the expression of IL1 in the group 1, sβ-treated cells. We conclude that in vitro, 1,25D3 and RvD1 rebalance inflammation to promote Aβ phagocytosis, and suggest that low vitamin D3 and docosahexaenoic acid intake and/or poor anabolic production of 1,25D3/RvD1 in PBMCs could contribute to AD onset/pathology.

Figures

Figure 1

1,25 D3 and RvD1 increase Aβ phagocytosis by macrophages of AD patients in a dose-responsive fashion. Replicate AD macrophage cultures were treated with 1,25D3, RvD1, or both, and exposed to FAM-Aβ overnight. The legend below the bar graphs indicates the concentrations (in nM) of 1,25D3 and/or RvD1. Phagocytosis was determined by scanning photographs of cells with FAM-Aβ (Suppl. Fig. 1). Data are mean ± s.e.m integrated optical density (IOD) per macrophage.

Figure 2

Receptors and 1,25D3 and RvD1 signaling mechanisms of phagocytosis in macrophages of AD patients. A) Phagocytosis of Aβ by replicate AD macrophage cultures treated with 1,25D3 (10−8M) or RvD1 (26.0 nM) and inhibitors of relevant receptors or metabolic pathways: VDR (inhibited by MK [10−7M]) and PDIA3 (inhibited by the neutralizing antibody, Ab099) are the known receptors for 1,25D3; Chemokine receptor 32 (inhibited by the neutralizing antibody, Chr32, and pertussis toxin, PTX). B) The effect of increasing concentration of the calcium chelator EGTA (concentrations listed in mM) on RvD1 (26.0 nM) and 1,25D3 (10 nM) promotion of FAM-Aβ (2 μg/ml) phagocytosis. C) The effect of PI3K (inhibited by wortmannin), PKA (inhibited by PKI), and MEK1/2 (inhibited by U0126) on the ability of RvD1 (26.0 nM) and 1,25D3 (10 nM) to promote FAM-Aβ (2 μg/ml) phagocytosis.

Figure 3

Fibrillar Aβ, but not soluble Aβ, induces caspase-3-positive apoptosis in macrophages of AD patients, but not in control macrophages. Replicate macrophage cultures were exposed to 2 μg/ml FAM-fAβ or FAM-sAβ overnight and then fixed and stained for caspase-3, photographed and fluorescent signal determined by scanning. Open symbols highlight data obtained from control macrophages and closed from AD macrophages (n=3). * indicates significance (p

Figure 4

Differential TLR mRNA expression in the Group 1 and Group 2 AD PBMCs at baseline and upon stimulation with exogenous sAβ. Patients 4 and 5 (Group 1, A) showed significant down regulation of TLR2 (*p

Figure 5

Two profiles in the baseline transcription of inflammatory and autoimmune genes observed in PBMCs of AD patients when compared to control subjects. A) A volcano plot demonstrating the fold regulation change and statistics observed when the baseline transcription of eighty four genes (PAHS077G RT2-Profiler array, Qiagen) from group 1 AD (n=3, two patients) and control (n=3, three controls) were compared. For this comparison all five housekeeping genes were used for normalization. In the plot, data points to the left of the center line designate down regulated mRNA expression and those to the right up regulated expression. The vertical lines to the left and right of the center line, highlight a 4-fold regulation change threshold. The horizontal line represents a p-value of 0.05. B) A volcano plot demonstrating the fold regulation change and statistics observed when the baseline transcription of eighty four genes (PAHS077G RT2-Profiler array, Qiagen) from group 2 AD (n=6, three patients) and control (n=3, three controls) were compared. For this comparison all five housekeeping genes were used for normalization. C) A simplified volcano plot showing only the genes whose baseline transcription was statistically different when group 2 AD (n=6) and group 1 (n=3) samples were compared. D) A simplified volcano plot showing only the genes whose baseline transcription was statistically different when all AD samples (n=9) and control (n=3) samples were compared.

Figure 6

Soluble Aβ upregulates the transcription of inflammatory and autoimmune genes more so in AD patients than in controls. A) A volcano plot demonstrating the fold regulation change and statistics observed when controls (n=3, three control subjects) were treated overnight with 2 μg/ml soluble Aβ42 (sAβ42) and the qPCR results compared to the baseline transcription of 84 genes (PAHS077G RT2-Profiler array, Qiagen) from the same PBMC isolation. The vertical lines to the left and right of the center line highlight a 4-fold regulation change threshold. The horizontal line represents a p-value of 0.05. For genes labeled above the line the fold regulation change was significant across the sample population. B) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when group 1 AD patient PBMCs (n=3, two patients) were treated overnight with 2 μg/ml sAβ42 and the qPCR results compared to the baseline transcription from the same PBMC isolation. For this comparison three housekeeping genes were used for normalization. A comparison of group 1 AD and control sAβ42-treated PBMCs is provided in Supplementary Figure 2A. C) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when group 2 AD PBMCs (n=5, three patients) were treated overnight with 2 μg/ml sAβ42 and the qPCR results compared to the baseline transcription from the same PBMC isolation. For this comparison all five housekeeping genes were used for normalization. A comparison of group 2 AD and control sAβ42-treated PBMCs is provided in Supplementary Figure 2B. D) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when group 2 AD (n=5, three patients) and group 1AD (n=3, two patients) sAβ42-treated PBMCs were compared. For this comparison all three housekeeping genes were used for normalization. E) A volcano plot demonstrating the general up regulatory effect of exogenous sAβ42 on the expression of inflammatory and autoimmune genes across all AD patients (n = 6, three from each group) when compared to sAβ42-treated PBMC samples from controls (n=3). For this comparison three housekeeping genes were used for normalization.

Figure 7

RvD1 retunes the transcription of inflammatory and autoimmune genes whose transcription was altered by sAβ42. A) A scatter plot showing the fold regulation change in the transcription of eighty four genes (PAHS077G RT2-Profiler array, Qiagen) following overnight treatment of control PBMCs with sAβ42 and sAβ42 + 26.0 nM RvD1 (n=2 for both groups). Genes up or down regulated more than 2.6-fold are labeled and the average fold regulation change provided in the figure panel. B) A scatter plot showing the fold regulation change in the transcription of genes following overnight treatment of group 1 PBMCs with sAβ42 and sAβ42 + 26.0 nM RvD1 (n=2 for both groups, one sample from two AD patients). Genes up or down regulated more than 4.0-fold are labeled and the average fold regulation change provided in the figure panel. C) A scatter plot showing the fold regulation change in the transcription of genes following overnight treatment of group PBMCs with sAβ42 and sAβ42 + 26.0 nM RvD1 (n=5 for both treatments, two group 2 patients). Genes up or down regulated more than 2.6-fold are labeled and the average fold regulation change provided in the figure panel. A volcano plot demonstrating the effect of RvD1 observed in one group 2 AD patient is provided in Supplementary Figure 3A. D) A scatter plot showing the fold regulation change in the transcription of the eighty four genes following overnight treatment of all AD patient PBMCs with sAβ42 or sAβ42 + 26.0 nM RvD1 (n=6 for both groups). Genes up or down regulated more than 2.6-fold are labeled and the average fold regulation change provided in the figure panel.

Figure 8

1,25D3 retunes the transcription of inflammatory and autoimmune genes whose transcription was altered by sAβ42. A) A scatter plot showing the fold regulation change in the transcription of eighty four genes (PAHS077G RT2-Profiler array, Qiagen) following overnight treatment of control PBMCs with sAβ42 and sAβ42 + 10.0 nM 1,25D3 (n=2 for both groups). Genes up or down regulated more than 4-fold are labeled in the figure panel. B) A scatter plot showing the fold regulation change in the transcription of genes following overnight treatment of group 1 PBMCs with sAβ42 and sAβ42 + 10.0 nM 1,25D3 (n=2 for both groups, one sample from two AD patients). Genes up or down regulated more than 4.0-fold are labeled in the figure panel. C) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when group 2 AD PBMCs (n=4, two patients) were treated overnight with 2 μg/ml sAβ42 + 10.0 nM 1,25D3 and the qPCR results compared to those following sAβ42-treatment of PBMCs isolated from the same blood draw. D) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when all AD PBMCs (n=6, four patients) were treated overnight with 2 μg/ml sAβ42 + 10.0 nM 1,25D3 and the qPCR results compared to transcription of eighty four genes (PAHS077G RT2-Profiler array, Qiagen) following sAβ42-treatment of PBMCs isolated from the same blood draw. E) A volcano plot demonstrating the fold regulation change (≥ 4-fold) and statistics (p<0.05) observed when all AD PBMCs (two group 1 and two group 2 AD patients) were treated overnight with 2 μg/ml sAβ42 + 10.0 nM 1,25D3 (n=6) or + 26.0 nM RvD1 (n=6) and the qPCR results compared. The plot shows how 1,25D3 differs from RvD1 (e.g., IL1R1 is down regulated more by 1,25D3 across the AD patient population then it is by RvD1).

Figure 9

The cytokine IL1β inhibits the action of 1,25D3 and RvD1 on Aβ phagocytosis. Replicate macrophage cultures were pre-treated with 1,25D3 (10−8M) or RvD1 (26.0 nM) with or without IL1β (2.5 μg/ml) and then exposed to 2 μg/ml FAM-Aβ overnight. Data are presented as mean ± STD IOD per macrophage (n=4; *p<0.01 compared to sAβ-treated group; **p<0.001 compared to sAβ-treated group).

Figure 10

RvD1 and 1,25D3 inhibit the secretion of inflammatory cytokines from PBMCs of a group 2 patient stimulated by sAβ. Three million PBMC’s were treated overnight with sAβ42 ± 10.0 nM 1,25D3 or 26.0 nM RvD1. The supernatants were harvested and the cytokines were tested by ELISA. * indicates significance (p<0.02) when compared to sham (AD); ** indicates significance when compared to fAβ control and sham (p<0.0001); and *** indicates significance when compared to AD fAβ-treated macrophages (p<0.0001) according to a two-tailed t-test (n=3).