Marked upregulation of cholesterol 25-hydroxylase expression by lipopolysaccharide

Abstract

During screening of genes upregulated by lipopolysaccharide (LPS; endotoxin) treatment of bone marrow-derived mouse macrophages, it was unexpectedly found that cholesterol 25-hydroxylase (Ch25h) was strongly upregulated. Treatment of macrophages with 10 ng/ml of LPS for 2 h resulted in a 35-fold increase in the expression of Ch25h. In contrast, LPS treatment did not increase the expression of Cyp27a1 or Cyp7b1. The increased Ch25h expression was found to be independent of Myeloid differentiation protein 88 signaling but dependent on Toll-like receptor 4 signaling. LPS treatment of macrophages caused a 6- to 7-fold increase in cellular 25-hydroxycholesterol concentration. When macrophages were treated with increasing concentrations of 25-hydroxycholesterol, a dose-dependent release of CCL5 into the culture medium was observed. Intravenous injection of LPS in eight healthy volunteers resulted in an increase in plasma 25-hydroxycholesterol concentration. The possibility is discussed that 25-hydroxycholesterol may have a role in the inflammatory response, in addition to its more established role in the regulation of cholesterol homeostasis.

Fig. 1.

The cholesterol 25-hydroxylase gene is induced by LPS. Bone marrow-derived macrophages were differentiated in vitro with GM-CSF for 8 d and then treated with 10 ng/ml repurified E.coli LPS or media only for 2 h (A). Total RNA was harvested, purified, and labeled, and gene expression analyzed by microarrays (Operon) with oligonucleotide reporters representing approximately 13,000 genes. Cells from four individual mice were used and analyzed in duplicate (n = 8 microarrays total). In B, macrophages were treated with repurified E. coli LPS, live E. coli (one bacterium per macrophage), or media only for 0, 0.5, 1, 2, 4, 6, 12, and 24 h as indicated. Each line represents macrophages taken from one individual mouse. In C, macrophages were derived from Tlr4^−/−, Myd88^−/−, or C57Bl/6 control mouse bone marrow and treated with 10 ng/ml repurified E. coli LPS or media only for 2 h. Cells from four individual mice were used and analyzed in duplicate (n = 8 microarrays total). Fold changes represent expression compared with media controls, i.e., a fold change of 1 indicates no change from media only treated cells. Average data ± SD are presented. Statistics were calculated with a Mann-Whitney test.

Fig. 2.

The cholesterol 25-hydroxylase gene, but not Cyp27a1 or Cyp7b1, was induced by LPS. Bone marrow-derived macrophages were differentiated in vitro with GM-CSF for 6 d and then treated with PBS (vehicle) or 10 ng/ml E. coli LPS for 2 h. Total RNA was isolated, and mRNA expression of Ch25h (A), Cyp27a1 (B), and Cyp7b1 (C) was analyzed by real-time PCR. Expression was normalized to β-actin expression and presented as expression relative to PBS-treated control cells. Average data ± SD are presented. Statistics were calculated with a Mann-Whitney test (n = 4 individual cultures).

Fig. 3.

LPS-treated macrophages produce and release 25-hydroxycholesterol. Bone marrow-derived macrophages were differentiated in vitro with GM-CSF for 6 d and then treated with PBS (vehicle) or 10 ng/ml E. coli LPS for 6 or 24 h. Medium (A) and cells (B) were collected and analyzed for 25-hydroxycholesterol content and normalized to total cellular protein. Average data ± SD are presented. ** P < 0.01 in a one-way ANOVA with Dunnett’s multiple testing correction compared with untreated cells (n = 4 individual cultures).

Fig. 4.

25-Hydroxycholesterol treatment of macrophages induced CCL5 expression. Mouse macrophage-like cells were treated under various conditions and the CCL5 released into the culture media was measured by ELISA. A: The mouse macrophage-like cell line RAW264.7 (in DMEM with 10% FCS) was treated with 25-hydroxycholesterol for 18 h. B: RAW264.7 cells (in DMEM with 10% FCS) were pretreated with 25-hydroxycholesterol for 30 min before treatment with 10 ng/ml LPS for 18 h. C: RAW264.7 cells (in DMEM with 2% FCS) were treated with varying concentrations of 25-hydroxycholesterol for 18 h. Average data ± SD are presented. ** P < 0.01 in a one-way ANOVA with Dunnett’s multiple testing correction compared with untreated cells (n = 4 individual cultures). Culture medium contained <1 ng/ml 25-hydroxycholesterol.

Fig. 5.

LPS infusion in healthy volunteers raises plasma 25-hydroxycholesterol levels after 4 h. Mann-Whitney test compared with time 0 (n = 8 healthy volunteers).