The transcription factor STAT-1 couples macrophage synthesis of 25-hydroxycholesterol to the interferon antiviral response

Abstract

Recent studies suggest that the sterol metabolic network participates in the interferon (IFN) antiviral response. However, the molecular mechanisms linking IFN with the sterol network and the identity of sterol mediators remain unknown. Here we report a cellular antiviral role for macrophage production of 25-hydroxycholesterol (cholest-5-en-3β,25-diol, 25HC) as a component of the sterol metabolic network linked to the IFN response via Stat1. By utilizing quantitative metabolome profiling of all naturally occurring oxysterols upon infection or IFN-stimulation, we reveal 25HC as the only macrophage-synthesized and -secreted oxysterol. We show that 25HC can act at multiple levels as a potent paracrine inhibitor of viral infection for a broad range of viruses. We also demonstrate, using transcriptional regulatory-network analyses, genetic interventions and chromatin immunoprecipitation experiments that Stat1 directly coupled Ch25h regulation to IFN in macrophages. Our studies describe a physiological role for 25HC as a sterol-lipid effector of an innate immune pathway.

Graphical abstract

Figure 1

25HC Synthesis and Secretion by Mϕ after Interferon Treatment or MCMV Infection (A) Work-flow for the LC-MS(MSⁿ) analysis of oxysterols. Quantification was performed from reconstructed ion chromatograms by reference to deuterated internal standards. (B) Structure of cholesterol, the scaffold on which oxysterols are built. (C) Intracellular levels of oxysterols in BMDM following MCMV infection (MOI = 1), Ifn-γ (25 U/ml) or Ifn-β (25 U/ml) stimulation for 12 hr. A mock sample of BMDM without stimulation or infection was similarly analyzed. (D) Analysis of oxysterols secreted by cells into medium when treated as in (C). n = 3, data are mean ± SEM.

Figure 2

Temporal Analyses of Ch25h Synthesis and 25HC Production (A) Levels of secreted (A) and cellular (B) 25HC and cellular cholesterol (C) with Ifn-β (25 U/ml) or Ifn-γ (25 U/ml) in delipidized media. n = 3, mean ± SEM. (D) Expression level of Ch25h, Cyp27a1, and Cyp7a1 from microarray time course experiment in WT BMDM following MCMV infection (MOI = 1) or Ifn-γ (10 U/ml) stimulation (Cyp46a and Cyp7b1 did not reach the threshold of expression). (E) Ch25h expression BMDM following MCMV infection (MOI = 1) or Ifn-γ (10 U/ml) stimulation. (F) 25HC level in BMDM 24 hr following MCMV infection, Ifn-β (10 U/ml), Ifn-γ (10 U/ml) or PolyI:C (10 ng/ml) with the mean ± SEM for data obtained from three independent biological replicates.

Figure 3

Analysis of 25HC Anti-Viral Effects, See Also Figure S1 Virus replication was monitored (A, B, D, E, and F) as a function of kinetic GFP fluorescence. Replication slopes over the linear phase were calculated and normalized to vehicle treated cells, and the mean replication slope from independent experiments calculated. For (C) Influenza virus, supernatant was collected at 16 hpi for analysis of viral titer by plaque assay. Data are mean ± SEM (three replicates).

Figure 4

Investigation of Specificity and Selectivity of 25HC Antiviral Activity, See Also Figure S2 (A) NIH 3T3 cells were infected with MCMV-GFP virus (MOI = 0.05) treated with 25HC in the presence or absence of Ly295427 (20 μM) or Ly306039 (20 μM). The level of infection was measured by determining the GFP fluorescence level after 72 hr infection. n = 3, data are mean ± SEM (three replicates per experiment). (B) NIH 3T3 cells infected with MCMV-GFP virus (MOI = 0.05) received conditioned medium from BMDM infected, or treated with IFN-β (25 U/ml) or IFN-γ (10 U/ml) ± 20 μM Ly295427 or 20 μM Ly306039. n = 3, and data are mean ± SEM (with three biological replicates per experiment). (C) Expression analysis of Gpr183 (relative to t = 0) in WT, Ifnb^−/− or Ifnar1^−/− BMDM infected with MCMV or MCMVdie3 (MOI = 1) or treated with PolyI:C (10 ng/ml). Each column represents one time point and row represents one gene. Gene expression is shown as a pseudocolor – blue = decrease, yellow = increase. Log fold change values were calculated by subtracting the t = 0 signal value from the signal value at time points after infection or treatment. (D) MEF were pretreated with 25HC or ent-25HC in delipidized medium for 24 hr and infected with MCMV-GFP (MOI = 1). n = 2, data are mean ± SEM (with three biological replicates per experiment). (E) BMDM treated with 10 U/ml Ifn-γ or mock-treated (F). Heatmap showing temporal alterations in abundance of sterol-related transcripts in Ifn-γ-treated BMDM (E, relative to t = 0) or MCMV infected BMDM (F, relative to t = 0) as in (C). (G) NIH 3T3 cells were pretreated with GW3965 (1 μM), T0901317 (1 μM) or 25HC (10 μM) in normal or delipidized medium for 24 hr and then infected with MCMV-GFP. After infection, all wells received fresh medium containing normal or delipidized serum and GW3975, T0901317, or 25HC. Data are normalized to negative controls treated with vehicle alone and are the mean of six independent experiments ± SEM (three biological replicates per experiment). (H) MEFs were pretreated with GW3965 (1 μM), T0901317 (1 μM) or 25HC (10 μM) in normal or delipidized medium for 24 hr and then infected with MCMV-GFP as described in (G). Data are normalized to negative controls treated with vehicle alone and are the mean of six independent experiments ± SEM (three replicates per experiment). (I) Left shows MEFs infected with MCMV (MOI = 0.1) incubated with 25HC (1 μM) and/or GGOH (20 μM) or MEV (20 μM) and MCMV titer determined by plaque assay. Data are mean of three independent experiments ± SEM. ^∗p ≤ 0.05. Right shows that MEFs infected with MCMV (MOI = 0.1) were incubated with vehicle, 25HC – (1 μM), and/ or Farnesol (FOH – 20 μM) or Squalene (Sqle – 20 μM) and MCMV titer was determined by plaque assay. Data are the mean of two independent experiments. (J) RISC-Free, Fnta, Fntb, Pggt1b, Rabggta, Rabggtb, or MCMV M54 siRNA (25nM) were reverse-transfected into NIH/3T3 cells. Forty-eight hr after transfection, cells were infected with MCMV-GFP (MOI = 0.05) and virus replication monitored. Data are normalized to RISC-free wells and are mean of six independent experiments ± SEM.^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗ < 0.001. (K) Left shows that MEF were pretreated with 25HC (5 μM) or psoromic acid (20 μM) for 24 hr, infected with MCMV-GFP (MOI = 0.05), and then incubated with EMEM with the same concentrations of 25HC or psoromic acid. Data are normalized to vehicle treated wells and are mean of two independent experiments. Right shows that MEFs were pretreated with vehicle, FTI-277 (10 μM), or GGTI-2133 (10 μM) for 24 hr, infected with MCMV-GFP (MOI = 0.05), and incubated with the same concentrations of FTI-277 or GGTI-2133 as prior to infection. Data are normalized to vehicle treated wells and are mean of two independent experiments with three biological replicas (error bars represent range).

Figure 5

Analysis of 25HC Effects on Cell Death, Virus Entry and Cell-to-Cell Spread, See Also Figure S3 and Movie S1 (A) Left shows NIH 3T3 cells treated with 25HC in delipidized medium and MCMV infected (MOI = 2). Three hpi DNA levels were determined by qRT-PCR. Middle shows RAW 264.7 cells treated with 25HC (5 μM) pre- or postinfection with MCMV (MOI = 1) in normal medium. At 3 hpi, amounts of internalized viral DNA were determined by qRT-PCR. Right shows MEFs treated with 25HC (5 μM) 24 hr prior to and after infection with MCMV (MOI = 2.5) in delipidized media and infected virus titers determined by plaque assay 4 dpi. n = 4, with mean ± SEM. ^∗∗∗p ≤ 0.001. (B) NIH 3T3 cells were treated with 10 μM 25HC or 10 μM staurosporine in normal media and cell death measured. Data are mean ± SEM for three replicates. (C) Left shows images of NIH 3T3 cells with and without 25HC 2.5 μM infected with MCMV (MOI 0.01). Middle shows cells treated with 2.5 μM 25HC in normal media. Images of NIH 3T3 cells infected with MHV-68 39 hpi. Right shows HeLa cells infected with HSV-1 VP26-YFP (MOI = 3), treated with 25HC (20 μM) in normal medium and images taken 24 hpi. (D) NIH 3T3 cells infected with MCMV (MOI = 0.005) and treated with 0.1, 1, or 5 μM 25HC in normal media. Images were taken at 72 hpi. (E) p53^−/− MEFs were infected with MCMV-GFP (MOI = 0.001) and overlaid in normal medium containing 2.5% agarose and 0.1, 1, 2.5, or 5 μM 25HC. Plaque diameter was measured 3 dpi. Data are mean 170 and 220 plaques counted ± SEM (three biological replicates). ^∗p ≤ 0.05, ^∗∗p ≤ 0.01. (F) H1N1: Plaque reduction assay of Influenza A/WSN/33 (H1N1) virus infection of MDCK cells in the presence or absence of 25HC (2.5 μM) and a 0.5% agarose overlay. Forty-eight hpi, cells were stained with 0.1% toluidine blue and number and size of plaques determined. Left upper (vehicle) and lower (25HC) panels are representative images showing differences in plaque size and morphology. Right panels show mean number of plaques per well (upper) and mean plaque area (lower, mm²) per well. HSV-1-eGFP (C12) infection of A549 cells pretreated with vehicle or 25HC (5 μM) with a 0.5% agarose overlay. Plaque numbers and diameters were determined 72 hpi. cell. Representative fluorescent microscopic images of plaques from vehicle- (top left) or 25HC- (bottom left) treated cells are shown. Plaque numbers (top right) and mean plaque diameter (bottom right) from HSV-1 infected cells are shown. n = 6, data are mean ± SEM. ^∗p ≤ 0.05, ^∗∗p ≤ 0.01. (G) Left shows that NIH 3T3 cells were infected with MCMV (MOI = 2.5) in delipidized medium containing vehicle or 25HC (5 μM). Two dpi supernatants (extracellular virus) and intracellular virus quantitated by plaque assay. n = 4, mean ± SEM. ^∗∗∗p ≤ 0.001. Right shows MDCK cells infected with A/WSN/33 (H1N1) influenza virus (MOI = 1), serum-free medium containing vehicle, or 25HC (2.5μM), and 0.5% agarose was added to the monolayers. After 16 hpi, monolayers were fixed and stained. n = 4 mean ± SEM. ^∗∗∗p ≤ 0.001. (H) Left shows plaque diameters observed in experiment and predicted by modeling when cell permissivity and viral growth both mediate 25HC antiviral effect. Right shows that plaque diameters observed in experiment and predicted by modeling when only cell permissivity contributes to the infectious transmission between cells in a dose-dependent manner. (I) MEFs pretreated with 25HC (5 μM) for 24 hr, infected with MCMV (MOI = 0.3), and DNA levels quantitated at indicated times. (J) Left shows that BMDM were pretreated with Ifn-γ (10 U/ml) or 25HC (5 μM) for 24 hr, infected with GLucMCMV and at 2 or 4 hpi and Gaussia luciferase activity measured using supernatants. n = 6 ± SEM. Right shows that RAW264.7 cells were pretreated for 24 hr with Ifn-γ (10 U/ml) or 25HC (2 μM), infected with MCMV (MOI = 2), in normal medium containing vehicle, IFN-γ (10 U/ml) or 25HC (2 μM) for 24 hr, relative abundance of the MCMV ie1 assayed by qRT-PCR. n = 3; mean ± SEM. ^∗p ≤ 0.05.

Figure 6

25HC Synthesis Is Interferon-Dependent and Regulated by Stat1; See Also Figure S4 and Tables S1 and S2 (A and B) Ch25h gene expression (relative to mock infected cells) in WT, Ifnb^−/− or Ifnar1^−/− BMDM infected with MCMV or MCMVdIE3 (MOI = 1). (C) Changes in intracellular 25HC in Ifnar1 KO BMDM following MCMV infection, Ifn-β (10 U/ml) or Ifn-γ (10 U/ml) stimulation for 24 hr. n = 3; data are mean ± SD. (D) qRT-PCR analysis of Ch25h mRNA expression at 6 and 24 hpi with MCMV (MOI = 1) or treatment with Ifn-γ (10 U/ml) or Ifn-β (25 U/ml) in WT, Ifnar1^−/− or Stat1^−/− BMDM. n = 3, data are mean ± SD. (E) Differential ntRNA synthesis between 60 and 90 min in Ifn-γ treated (F) or MCMV-infected BMDM (H) relative to mock were analyzed and fold-change values in treated or infected cells calculated relative to the control. (G) WT BMDM treated with Ifn-γ (10 U/ml). Microarray analysis of newly transcribed RNA isolated from 16 successive 30 min windows after treatment. (H) UCSC genome browser image is shown for Ch25h locus indicating Stat1 binding site (labeled ch25h_8) in promoter region. Normalized tag counts for H3K4me2-MNase ChIP-Seq, H4K5Ac-ChIP-Seq, and H3K4me3-ChIP-Seq are shown for untreated BMDM. (I) ChIP analysis of Stat1 binding at Igk (negative control), Tap1, and Nos2 (positive controls), and Ch25h loci (as indicated in Figure 5I) in mock- or Ifn-γ− (20 U/ml, 1 hr) treated BMDMs. Fold enrichment for each locus is the ratio of percentage input of Ifn-γ-treated versus mock-treated BMDMs. IgK (negative control) refers to genomic region lacking Stat1 enrichment. Data are representative of two independent experiments performed in duplicate.