Cholesterol 25-hydroxylase production by dendritic cells and macrophages is regulated by type I interferons

Mice

BALB/cJ, C57BL/6, myd88^−/−, trif^−/−, and stat6^−/− were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). ifnar^−/− mice on a BALB/cJ background [24, 25] were obtained with permission from Michael Aguet (Swiss Institute for Experimental Cancer Research, Lausanne, Switzerland). Stat1^−/− mice on a 129S6/SvEvTac background and the WT 129S6/SvEvTac controls were purchased from Taconic (Rockville, MD, USA). Male mice at 6–8 weeks of age were used for the experiments. All animal procedures described in this study were performed under the approval of the Johns Hopkins Animal Care and Use Committee (Baltimore, MD, USA) in accord with the guidelines set by the National Research Council's Guide for Care and Use of Laboratory Animals.

Isolation of lung macrophages

Lungs were harvested and processed for the isolation of CD11c⁺ lung macrophages as outlined previously [20]. Briefly, finely chopped lungs were suspended in RPMI (Invitrogen, Carlsbad, CA, USA) and digested for 30 min with 1 mg/ml Collagenase II (Gibco, Carlsbad, CA, USA) and 30 μg/ml DNase I (Roche, Indianapolis, IN, USA). After passing through a sterile strainer (352360, BD Falcon, Franklin Lakes, NJ, USA), RBCs were removed, and CD11c⁺ cells were isolated by the IMag™ cell separation system protocol (BD PharMingen, San Diego, CA, USA). The purity of the cells was checked by flow cytometry (FACSCalibur, BD Biosciences, San Jose, CA, USA) and was typically >90% CD11c⁺ cells.

Preparation of BMDCs and BMMs

Marrow cells were collected from long bones using a syringe fitted with a 25-G needle, filtered through strainers (BD Falcon), and treated with ACK Lysing Buffer (Quality Biologicals, Gaithersburg, MD, USA) to remove RBCs. For BMDCs, cells were cultured with IMDM (Invitrogen) and supplemented with 10% FBS, 1× antibiotic-antimycotic (Invitrogen), 55 μM 2-ME (Sigma-Aldrich, St. Louis, MO, USA), and 20 ng/ml GM-CSF (Invitrogen). Half of the culture medium was changed at Days 3, 5, and 7. On Day 8 or 9 of culture, CD11c⁺ cells were isolated by the IMag™ cell separation system protocol (BD PharMingen) using the manufacturer's protocol. For BMM differentiation, cells were cultured in supplemented IMDM containing 30% L929 cell culture supernatant. Half of the medium was changed at Days 3 and 5 of culture. At Day 6, floating cells were removed, and the culture medium was changed to IMDM, supplemented 10% FBS without L929 culture supernatant. On Day 7, fully differentiated macrophage cultures were used for stimulation.

Stimulation of cells with TLR ligands and IFNs

In vitro stimulations of the BMDC, BMM, and CD11c⁺ lung macrophages were carried out with 1 × 10⁶ cells/well in a 24-well plate (BD Falcon). The cells were exposed to a panel of TLR agonists (Invivogen) that included 1 μg/ml Pam3CSK4 (TLR1/2 agonist), 1 × 10⁸/ml HKLM (TLR2 agonist), 25 μg/ml poly I:C (TLR3 agonist), 1 μg/ml of ultra-pure LPS (TLR4 agonist) from Escherichia coli K12, 1 μg/ml flagellin from Salmonella typhimurium (TLR5 agonist), 1 μg/ml FSL-1 (TLR6/2 agonist), 1 μg/ml ssRNA40 (TLR7 agonist), and 2.5 μM CpG ODN1826 (TLR9 agonist). Cells were also stimulated with IFN-α (1000 units/ml; 12105-1, PBL InterferonSource, Piscataway, NJ, USA), IFN-β (1000 units/ml; I9032, Sigma-Aldrich), or IFN-γ (1 μg/ml; I4777, Sigma-Aldrich) for 6 h. Cells were harvested for RNA isolation as described below.

Real-time RT-PCR

Lung tissues were homogenized in Trizol (Invitrogen), and RNA cleanup was performed using RNeasy mini columns from Qiagen (Valencia, CA, USA) with on-column DNase I treatment. Total RNA was isolated from 10⁶ cultured cells using the RNeasy mini kit (Qiagen) with on-column DNase I treatment. Total RNA (1.0 μg) was reverse-transcribed with SuperScript II (Invitrogen) using random hexamers following the manufacturer's protocols. The following primer-probe sets were obtained as Taqman® gene expression assays-on-demand (Applied Biosystems, Foster City, CA, USA): Ch25h (Mm00515483_s1), ifnβ (Mm004395496_s1), Taqman® rRNA control reagent. A standard curve was constructed from the expression levels in serial dilutions of pooled cDNAs from known positive samples. The data obtained using the ABI 7500 were normalized to the expression of the gene encoding the 18S rRNA.

JAK inhibition

BMDCs and BMMs were exposed to 0.2, 1.0, or 5.0 μM InSolution™ JAK inhibitor 1 (Calbiochem, San Diego, CA, USA) for 1 h prior to stimulation with 25 μg/ml poly I:C, 1 μg/ml LPS, or 1000 units/ml IFN-β. After 6 h, the cells were harvested for RNA isolation. The inhibition of STAT1-phosphorylation by JAK inhibitor 1 was confirmed using immunostained Western blots of cell extracts harvested after 3 h of stimulation using anti-STAT1 (BD PharMingen), antiphospho-STAT1 (Cell Signaling Technologies, Danvers, MA, USA), and anti-actin (Abcam, Cambridge, MA, USA).

Statistical analysis

The Student's t test and ANOVA were used to calculate statistical significance using Prism 4 (GraphPad Software, La Jolla, CA, USA).

TLR ligands induce Ch25h in BMMs and BMDCs

Recent studies have reported that TLR4 and TLR3 agonists selectively induce Ch25h expression in macrophages [14, 18]. To determine if DCs also express Ch25h in response to TLR ligands, transcription levels of Ch25h were measured in BMDCs stimulated with a panel of eight TLR ligands for 1, 3, 6, or 24 h. As reported for macrophages, the ligands for TLR3 (poly I:C) and TLR4 (LPS) robustly increased Ch25h expression in BMDC (Fig. 1A). By comparison, the other six TLR agonists induced small but detectable levels of Ch25h transcription. Under these in vitro conditions, the highest poly I:C- and LPS-induced transcription levels of Ch25h by BMDCs were measured at 6 h (Fig. 1A). The same TLR ligands were tested for their ability to induce Ch25h in BMMs after 6 h of stimulation. As observed for BMDCs, ligands for TLR3 and TLR4 induced robust Ch25h expression in BMMs, and the ligands for other TLRs had only a modest effect (Fig. 1B).

Signal transduction events initiated by TLRs rely on two adaptor molecules: MyD88 and TRIF. TLR3 signals solely through a TRIF-dependent pathway, and TLR4 uses the MyD88 and TRIF pathways [26]. To test the role of these adaptor proteins in TLR-mediated expression of Ch25h, BMDCs were generated from myd88^−/− and trif^−/− animals. In the absence of MyD88, poly I:C-induced expression of Ch25h in BMDCs was not different from WT controls (Fig. 1C). LPS treatment of myd88^−/− BMDCs resulted in a modest but statistically significant reduction, indicating that MyD88 has a role to play in the regulation of this enzyme. In contrast, there was little to no TLR3- or TLR4-mediated Ch25h transcription in trif^−/− BMDCs, suggesting that TLR-mediated expression of this enzyme functions primarily through a TRIF-dependent mechanism. The importance of TRIF in the mechanism of Ch25h expression is underscored by the near-total absence of transcription levels in the TRIF-deficient BMDCs even under control conditions (Fig. 1C).

Ch25h expression is controlled primarily by type I IFN signaling

TRIF-dependent signaling culminates in the activation of NF-κB and IRF3, subsequently leading to the induction of IFN-β [26]. Indeed, the same BMDCs and BMMs that demonstrated a significant up-regulation of Ch25h expression after poly I:C or LPS treatment also exhibited an induction in ifnβ transcription (Fig. 2A). Under the conditions used here, none of the other six TLR ligands induced significant levels of ifnβ expression. Using BMDCs from WT, myd88^−/−, and trif^−/− animals, it was demonstrated that TRIF signaling had a dominant effect on the transcription levels of type I IFN in poly I:C- and LPS-treated cells (Fig. 2B) and that the levels of ifnβ expression paralleled that of Ch25h for each treatment (Fig. 1C).

These results suggested that induction of Ch25h by poly I:C/TLR3 and LPS/TLR4 is indirectly controlled through IFNβR–IFNαR interactions. To determine the role of type I IFN signaling in the regulation of Ch25h expression, BMMs and BMDCs from type 1 ifnar^−/− and WT mice were stimulated with IFN-α or IFN-β. After 6 h, the cells were harvested, and the transcription level of Ch25h was measured by real-time RT-PCR. IFN-α and IFN-β treatment resulted in a significant induction of Ch25h transcription in the BMMs and BMDCs from WT mice but not in cells from ifnar^−/− mice (Fig. 3). IFN-α and IFN-β appeared to be equally potent in inducing Ch25h expression in BMDCs and BMMs.

BMDCs and BMMs from ifnar^−/− mice were also stimulated with the panel of TLR agonists. In the absence of the IFNαR, neither the BMDC nor the BMM responded to TLR activation with Ch25h expression (Fig. 3). These results support the hypothesis that TLR-mediated induction of Ch25h expression is controlled indirectly through type I IFN signaling.

As a control experiment, the type II IFN, IFN-γ, was included. Interestingly, IFN-γ also induced significant levels of Ch25h transcription in BMDCs and BMMs from WT animals that were modulated but not fully suppressed in the cells from ifnar^−/− animals (Fig. 3). Thus, it appears that IFN-γ-mediated signaling also has a role in the regulation of Ch25h expression.

Induction of Ch25h is JAK- and STAT1-dependent

The IFNαR-dependent nature of Ch25h expression is also supported by the results of experiments in which the activity of JAKs was blocked. Using a JAK inhibitor, the levels of Ch25h transcription induced by IFN-β, poly I:C, or LPS were reduced in a dose-dependent manner in BMMs and BMDCs from WT mice (Fig. 4A and B). To verify the action of the JAK inhibitor, immunostained Western blots of BMDC extracts were used to demonstrate that the activation via phosphorylation of STAT1, a key transcription factor in the IFN-β (and IFN-γ) signaling pathway, was blocked (Fig. 4C).

To determine further if the induction of Ch25h transcription was STAT1-dependent, BMMs and BMDCs were generated from stat1^−/− and WT animals and stimulated with TLR ligands, IFN-α, IFN-β, or IFN-γ for 6 h. Induction of Ch25h was abolished effectively in BMDCs and BMMs from stat1^−/− in all of the treatment groups (Fig. 5). Thus, in BMMs and BMDCs, Ch25h expression is regulated directly by the actions of type I or type II IFNs that signal in a JAK- and STAT1-dependent manner.

Regulation of Ch25h in vivo

To determine if TLR-mediated induction of Ch25h was also dependent on signaling though IFNRα and STAT1 in vivo, we administered poly I:C intratracheally to ifnar^−/−, stat1^−/−, and strain-matched control mice and processed the lungs to determine Ch25h levels in the RNA isolated from whole lung tissue and from an enriched population of CD11c⁺ DCs and macrophages (lung macrophages and DCs express elevated levels of CD11c [27]). In WT animals, administration of poly I:C resulted in a significant up-regulation of Ch25h in the whole lung with a majority of the response derived from the CD11c⁺ cell population (Fig. 6A and B). The Ch25h levels were diminished significantly in the whole lungs and CD11c⁺ cells isolated from ifnar^−/− (Fig. 6A) and stat1^−/− (Fig. 6B) mice. The results indicate that similar to the responses observed in vitro, TLR-mediated induction of Ch25h in DCs and macrophages in vivo is type I IFN- and STAT1-dependent.