Regulation of versican expression in macrophages is mediated by canonical type I interferon signaling via ISGF3

Abstract

Growing evidence supports a role for versican as an important component of the inflammatory response, with both pro- and anti-inflammatory roles depending on the specific context of the system or disease under investigation. Our goal is to understand the regulation of macrophage-derived versican and the role it plays in innate immunity. In previous work, we showed that LPS triggers a signaling cascade involving Toll-like receptor (TLR)4, the Trif adaptor, type I interferons, and the type I interferon receptor, leading to increased versican expression by macrophages. In the present study, we used a combination of chromatin immunoprecipitation, siRNA, chemical inhibitors, and mouse model approaches to investigate the regulatory events downstream of the type I interferon receptor to better define the mechanism controlling versican expression. Results indicate that transcriptional regulation by canonical type I interferon signaling via interferon-stimulated gene factor 3 (ISGF3), the heterotrimeric transcription factor complex of Irf9, Stat1, and Stat2, controls versican expression in macrophages exposed to LPS. This pathway is not dependent on MAPK signaling, which has been shown to regulate versican expression in other cell types. The stability of versican mRNA may also contribute to prolonged versican expression in macrophages. These findings strongly support a role for macrophage-derived versican as a type I interferon-stimulated gene and further our understanding of versican's role in regulating inflammation.NEW & NOTEWORTHY We report the novel finding that versican expression is regulated by the interferon-stimulated gene factor 3 (ISGF3) arm of canonical type I Ifn signaling in LPS-stimulated macrophages. This pathway is distinct from mechanisms that control versican expression in other cell types. This suggests that macrophage-derived versican may play a role in limiting a potentially excessive inflammatory response. The detailed understanding of how versican expression is regulated in different cells could lead to unique approaches for enhancing its anti-inflammatory properties.

Keywords: ISGF3; macrophage; type I interferon; versican.

Graphical abstract

Figure 1.

Evidence for transcriptional and posttranscriptional regulation of versican mRNA in macrophages. A: bone marrow-derived macrophages (BMDMs) from wild-type mice were exposed to PBS (vehicle control) or LPS (10 ng/mL) for up to 24 h. Versican mRNA expression was evaluated by real-time qPCR. Data represent means ± SE for cells from n = 7 mice. B: sheared cross-linked chromatin were assayed using an antibody to the C-terminal domain of RNA polymerase II. Chromatin immunoprecipitation (ChIP) DNA was analyzed at the versican promoter (Prom 400) and first exon (Exon 1) by real-time qPCR. Data are expressed as % Input and represent means ± SE, n = 3 mice. A schematic of the versican gene (V0 variant) is shown: translated and untranslated exons are represented as solid and open rectangles, respectively; lines represent introns; arrows show locations of the amplicons. C and D: BMDMs were treated with LPS for 4 h and then exposed to vehicle (d²H₂O) or actinomycin D (5 ug/mL). Cells were harvested at the indicated time points. qPCR data represent means ± SE for n = 3 mice. Statistics were analyzed by one-way ANOVA: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2.

Roles of Ifn-β and Irf3/7 in the induction of versican mRNA and protein. BMDMs from wild-type or Irf3/7 double knockout mice were exposed to PBS (vehicle control), LPS (10 ng/mL), or Ifn-β (100 U/mL) for up to 24 h. Ifn-β (A) and versican mRNA (B) expression were evaluated by real-time qPCR. Data represent means ± SE for n = 5–7 mice. C: representative Western blot analyses for versican are shown for secreted proteins from wild-type BMDMs treated with PBS (vehicle control), LPS, or Ifn-β for 24 h. Statistics were analyzed by two-way ANOVA with multiple comparisons: ^aP < 0.0001 for WT LPS, 4 h vs. 0 h control; ^bp < 0.05 for WT Ifn-β, 4 h vs. 0 h control; ^cP < 0.001 for WT Ifn-β, 8 h vs. 0 h control; ^dP < 0.0001 for WT LPS vs. Irf3/7 double knockout (dKO) LPS at 4 h; ^eP < 0.05 for WT Ifn-β vs. Irf3/7 dKO Ifn-β at 8 h; ^fP < 0.0001 for WT LPS, 4 or 8 h vs. 0 h control; ^gP < 0.0001 for WT Ifn-β, 4 or 8 h vs. 0 h control; ^hP < 0.0001 for Irf3/7 dKO Ifn-β, 4 or 8 h vs. 0 h control; ⁱP < 0.05 for Irf3/7 dKO LPS, 8 h vs. 0 h control; ^jP < 0.001 for WT LPS vs. Irf3/7 dKO LPS at 4 or 8 h; ^kP < 0.05 for WT LPS vs. WT Ifn-β at 4 h.

Figure 3.

Effects of silencing Jak1 or Tyk2 on the induction of versican mRNA. BMDMs were transfected with 0, 30, or 300 nM of siRNA for Jak1 (A–D), or Tyk2 (E–J), for 24 h prior to exposure to PBS (vehicle control), LPS (10 ng/mL), or Ifn-β (100 U/mL) for 4 h. Jak1 (A), Tyk2 (E), Ifn-β (C and G), versican (D and H), IL1-β (I), and IL-18 (J) mRNA expression were evaluated by real-time qPCR. Representative Western blots are shown for Jak1 (B) or Tyk2 (F) proteins with normalization to β-actin at 24 h after transfection, with no stimulation. qPCR data represent means ± SE for cells from n = 3–6 wild-type mice. K: BMDMS were transfected with negative control siRNA and evaluated for versican mRNA expression. Statistics were analyzed by two-way ANOVA with multiple comparisons: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4.

Effects of silencing Irf9, Stat1, or Stat2 on the induction of versican mRNA. BMDMs were transfected with 0, 30, or 300 nM of siRNA for Irf9 (A–D), Stat1 (E–H), or Stat2 (I–L) for 24 h prior to exposure to PBS (vehicle control), LPS (10 ng/mL, or Ifn-β (100 U/mL) for 4 h. Irf9 (A), Stat1 (E), Stat2 (I), Ifn-β (C, G, and K), and versican (D, H, and L) mRNA expression were evaluated by real-time qPCR. Representative Western blots are shown for Irf9 (B), Stat1 (F), or Stat2 (J) proteins with normalization to β-actin at 24 h after transfection, with no stimulation. qPCR data represent means ± SE for n = 3–9 wild-type mice. Statistics were analyzed by two-way ANOVA with multiple comparisons: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5.

Role of Stat1 phosphorylation in the induction of versican mRNA. A: BMDMs were transfected with 0, 30, or 300 nM of Stat1 siRNA for 24 h prior to exposure to PBS (vehicle control), LPS (10 ng/mL), or Ifn-β (100 U/mL) for 4 h. Representative Western blots are shown for Stat1, p-Stat1, and β-actin; digital quantification data reflect means ± SD for cells from n = 2 wild-type mice. B: BMDMs were exposed to 0 (DMSO control), 33.3 or 100 mM fludarabine for 24 h prior to exposure to PBS (vehicle control), LPS, or Ifn-β (10 or 100 U/mL, as indicated) for 4 h. Representative Western blots are shown for pStat1 (Tyr701), Stat1, and β-actin with digital quantification data for n = 3 wild-type mice. Versican (C), Ifn-β (D), TNF-α (E), and Has1 (F) mRNA expression were evaluated by real-time qPCR. Data represent means ± SE for cells from n = 3–6 wild-type mice. Statistics were analyzed by two-way ANOVA with multiple comparisons vs. control (no differences between DMSO vs. PBS): *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6.

Effect of Stat1 deletion on the induction of Vcan. A: BMDMs from wild-type or Stat1 knockout (KO) mice were evaluated by Western blot analysis for Stat1 protein with normalization to β-actin. Representative blots with digital quantification of cells from replicate mice are shown. B–E: cells were exposed to PBS (vehicle control), LPS, or Ifn-β (100 U/mL) for 4 h and evaluated by real-time qPCR for versican (B), Ifn-β (C), Ifit2 (D), or Has1 (E) mRNA expression. Data represent means ± SE for cells from n = 6 mice per strain. Statistics were analyzed by two-way ANOVA with multiple comparisons: *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7.

Role of Mek1/2 in the induction of Vcan. A: BMDMs from Mek1^fl or Mek1^flLysM^Cre mice were evaluated by Western blot analysis for Mek1 protein with normalization to β-actin. Representative blots with digital quantification of cells from replicate mice are shown. B and C: cells were exposed to PBS (vehicle control), LPS (10 ng/mL), or Ifn-β (10 U/mL) for 4 h and evaluated by real-time qPCR for versican (B) or Nos2 (C) mRNA expression. D: BMDMs from wild-type (WT) mice were treated with PD032901 (0.5 mM in DMSO) and evaluated by Western blot analyses for phospho-ERK1/2, total ERK1/2, and β-actin. Representative blots with digital quantification of cells from replicate mice are shown. E: cells were then exposed to PBS, LPS, or Ifn-β for 4 h and evaluated by real-time qPCR for versican mRNA expression. Data represent means ± SE for cells from n = 3 mice per strain. Statistics were analyzed by ANOVA with multiple comparisons: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 8.

Schematic of regulatory events that control versican expression in macrophages. Previous work has shown that engagement of macrophage Toll-like receptors TLR4 and TLR3 by LPS and polyinosinic:polycytidylic acid [poly(I:C)], respectively, result in enhanced versican expression. Subsequent to activation of TLR4 and TLR3, engagement of the TRIF adaptor molecule is known to activate transcription factors Irf3 and Irf7, leading to the production of Type I Ifns (Ifn-α/β) and recognition by type I Ifn receptors (Ifnar1/2) (29). We now identify Jak1, Irf9, Stat1, and Stat2 as essential signaling molecules downstream of Ifnar1/2 that mediate the induction of versican expression in macrophages.