Apolipoprotein E induces antiinflammatory phenotype in macrophages

Abstract

Objective: Apolipoprotein E (apoE) exerts potent antiinflammatory effects. Here, we investigated the effect of apoE on the functional phenotype of macrophages.

Methods and results: Human apoE receptors very-low-density lipoprotein receptor (VLDL-R) and apoE receptor-2 (apoER2) were stably expressed in RAW264.7 mouse macrophages. In these cells, apoE downregulated markers of the proinflammatory M1 phenotype (inducible nitric oxide synthase, interleukin [IL]-12, macrophage inflammatory protein-1α) but upregulated markers of the antiinflammatory M2 phenotype (arginase I, SOCS3, IL-1 receptor antagonist [IL-1RA]). In addition, M1 macrophage responses (migration, generation of reactive oxygen species, antibody-dependent cell cytotoxicity, phagocytosis), as well as poly(I:C)- or interferon-γ-induced production of proinflammatory cytokines; cyclooxygenase-2 expression; and activation of nuclear factor-κB, IκB, and STAT1, were suppressed in VLDL-R- or apoER2-expressing cells. Conversely, the suppression of the M2 phenotype and the enhanced response to poly(I:C) were observed in apoE-producing bone marrow macrophages derived from VLDL-R-deficient mice but not wild-type or low-density lipoprotein receptor-deficient mice. The modulatory effects of apoE on macrophage polarization were inhibited in apoE receptor-expressing RAW264.7 cells exposed to SB220025, a p38 mitogen-activated protein kinase inhibitor, and PP1, a tyrosine kinase inhibitor. Accordingly, apoE induced tyrosine kinase-dependent activation of p38 mitogen-activated protein kinase in VLDL-R- or apoER2-expressing macrophages. Under in vivo conditions, apoE-/- mice transplanted with apoE-producing wild-type bone marrow showed increased plasma IL-1RA levels, and peritoneal macrophages of transplanted animals were shifted to the M2 phenotype (increased IL-1RA production and CD206 expression).

Conclusions: ApoE signaling via VLDL-R or apoER2 promotes macrophage conversion from the proinflammatory M1 to the antiinflammatory M2 phenotype. This effect may represent a novel antiinflammatory activity of apoE.

Figure 1. ApoE induces macrophage M2 polarization markers in VLDL-R- or apoER2-expressing RAW264.7 cells

VLDL-R, apoER2 or wild-type RAW264.7 cells were cultured for 24h with or without apoE (5μg/mL) or with increasing concentrations of apoE. A. Total cell lysates were separated by SDS-PAGE and immunoblotted with antibodies against iNOS, arginase-1, FIZZ1/RELM, and SOCS3. Data are representative for 3 independent experiments. B. Cell culture media were collected and concentrations of nitrite and nitrate (products of iNOS) and urea (product of arginase-1) were determined by fluorimetric or photometric assays, respectively. Shown are results from 3 independent experiments. C and D. Concentrations of M1 (MIP-1α, IL-12) and M2 (IL-1RA, G-CSF, IL-4, IL-13) cytokines were determined by ELISA. IL-13Rα2 was co-incubated with apoE for 24h at concentration of 0.1 μg/mL. Shown are results from 3 to 5 independent experiments. * - p<0.05; § - p<0.01; # - p<0.001, −apoE vs. +apoE).

Figure 2. ApoE induces M2 functional phenotype in VLDL-R- or apoER2-expressing RAW264.7 cells

VLDL-R, apoER2 or WT RAW264.7 macrophages were incubated with apoE (5 μg/mL) for 24h. A. Cell suspension (2.0 × 10⁵/mL) was exposed to chemoattractans fMLP (0.1 μmol/L) or M-CSF (0.1 μmol/L) for 2h. Cells emigrated through the insert membrane were fixed, stained and quantified as described in Methods. Shown are results from 3 independent experiments. B. Adhering cells were incubated for 2h with opsonized erythrocytes and hemoglobin released to the media was quantified as described in Methods. Shown are results from 3 independent experiments. C. Cell suspensions (5.0 × 10⁵/mL) were loaded with H₂DCF-DA and DCF fluorescence reflecting H₂O₂ concentration was monitored fluorimetrically for 20 min. Original tracing curves from one representative experiment out of three were superimposed for comparison (left panels). Right panel: comparison of ROS production rate in cells treated or not with apoE. D. Adhering cells were co-incubated with apoE and FITC-coupled latex beads for 18h. The single cell fluorescence was analyzed by flow cytometry. Note fluorescence peaks corresponding to macrophage populations ingesting one, two or more beads. Original tracing curves from one representative experiment out of three were superimposed for comparison (left panels). Right panel: comparison of latex beads ingestion stimulated by apoE in macrophages. * - p<0.05; § - p<0.01; # - p<0.001, −apoE vs. +apoE.

Figure 3. ApoE suppresses poly(I:C)- and IFN-γ-induced inflammatory response in VLDL-R- or apoER2-expressing RAW264.7 cells

VLDL-R, apoER2 or WT RAW264.7 macrophages were incubated with apoE (5 μg/mL) for 24h. A and B. Cells were stimulated for further 24h with poly(I:C) (10 ng/mL) and IFN-γ (50 ng/mL), supernatants were collected and analyzed by ELISA for cyto/chemokine content (C) or by EIA for PGE₂ concentration (D, lower panel). Shown are results from stimulated cells representative for 3 to 5 independent experiments. * - p<0.05; § - p<0.01; # - p<0.001, stimulated vs. unstimulated cells. ApoE alone did not affect cyto/chemokine or PGE₂ concentrations. Cell lysates were immunoblotted and probed with anti-COX-2 antibody (B, upper panel). Shown is one representative result out of three. Lower panel: densitometric analysis of blots. C. Cells were transfected with p(κB)₅-Luc or p(Gas)-Luc reporter plasmids, stimulated for 24h with, respectively, poly(I:C) (10 ng/mL) or IFN-γ (50 ng/mL) with or without apoE, and analyzed for luciferase. Luminescence level in un-stimulated cells was set as 1. ApoE alone did not change luminescence levels. Shown are results from 3 to 4 independent experiments. * - p<0.05; § - p<0.01; −apoE vs. +apoE. D. Cells were exposed to poly(I:C) or IFN-γ for indicated times and lysates were probed with antibodies against total (t) and phosphorylated (p) isoforms of IκB and STAT1. Shown are blots representative for 3 to 5 independent experiments.

Figure 4. Bone marrow (BM) macrophages derived from VLDL-R^−/− mice show enhanced pro-inflammatory phenotype

BM macrophages were incubated for 24h in serum-free medium containing 0.1% (v/v) albumin. A. mRNA levels were assessed by RT-PCR. Shown are representative agarose gels of amplified VLDL-R, apoER2 and GAPDH DNA fragments. Cerebellum (Cer) mRNA was taken as positive control. B. Concentrated cell culture media and cell lysates were subjected to Western blot using anti-apoE, anti-arginase-1 and anti-SOCS-3 antibodies. Urea was determined by photometric assay C. Cytokine concentrations were determined by ELISA. D and E. Cells were stimulated for further 24h with poly(I:C) (10 ng/mL), supernatants were collected and analyzed by ELISA for cyto/chemokine content (D) or by EIA for PGE₂ concentration (E, right panel). Cell lysates were immunoblotted with anti-COX-2 antibody and blots were analyzed by densitometry (E, right panel). Data represent mean ± SD from 3 independent experiments, each in duplicate. * - p<0.05; § - p<0.01, WT vs. VLDL-R^−/− cells.

Figure 5. ApoE induces tyrosine kinase-dependent p38^MAPK activation in VLDL-R- or apoER2-expressing RAW264.7 cells

A. VLDL-R, apoER2 or WT RAW264.7 macrophages were incubated with apoE (5 μg/mL) for indicated times. Cell lysates were immunobloted and probed with antibodies against total (t) and phosphorylated (p) isoforms of Akt and p38^MAPK. Results are representative for 3 experiments. B. Cells expressing apoE receptors were pretreated for 0.5h min with PP1 (10 μmol/L) and exposed to apoE (5 μg/mL) for 1h. p38^MAPK phosphorylation was analyzed as described above. Shown are results representative for 2 experiments C. VLDL-R, apoER2 or wild-type RAW264.7 macrophages were stimulated with apoE (5 μg/mL) for 24 h in the absence or presence of PP1 (10 μmol/L) or SB203580 (20 μmol/L). Incubation media were analyzed for IL1-RA concentrations. Shown are results from 4 independent experiments. § - p<0.01; # - p<0.001, apoE−inhibitor vs. apoE+inhibitor.

Figure 6. ApoE promotes macrophage M2 polarization in vivo

ApoE-deficient mice were transplanted with bone marrow obtained from apoE^−/− (BMT-E^−/−, n=6) or WT (BMT-E^+/+, n=6) mice as described in Methods. A. Verification of reconstitution with donor hematopoietic cells by PCR amplification of murine apoE using genomic DNA from blood cells (BC; upper left panel), immunoblot of apoE in cell media from peritoneal macrophages (PM) (lower left panel), and cholesterol distribution among serum lipoprotein fractions (right panel). Note normalization of lipoprotein profile in apoE^−/− mice transplanted with WT bone marrow. B. IL-1RA concentrations in sera from C57Bl6 control mice, or apoE^−/− or WT bone marrow transplanted C57Bl6 mice were determined by ELISA. C. PM supernatants from C57Bl6 control mice, C57Bl6 or mice transplanted with apoE^−/− or WT bone marrow were analyzed for IL-12 and IL-1RA concentrations by ELISA. Results were normalized for cell protein content. * - p<0.05; § - p<0.01; apoE^−/− vs. WT. D. MHC-II and CD206 cell surface expression was analyzed by flow cytometry in peritoneal macrophages from mice transplanted with apoE^−/− or WT bone marrow. Representative histograms were superimposed for comparison. Isotype-matched irrelevant immunoglobulins were used as negative controls (Ctrl).