Systemic delivery of proresolving lipid mediators resolvin D2 and maresin 1 attenuates intimal hyperplasia in mice

Abstract

Vascular injury induces a potent inflammatory response that influences vessel remodeling and patency, limiting long-term benefits of cardiovascular interventions such as angioplasty. Specialized proresolving lipid mediators (SPMs) derived from ω-3 polyunsaturated fatty acids [eicosapentaenoic acid and docosahexaenoic acid (DHA)] orchestrate resolution in diverse settings of acute inflammation. We hypothesized that systemic administration of DHA-derived SPMs [resolvin D2 (RvD2) and maresin 1 (MaR1)] would influence vessel remodeling in a mouse model of arterial neointima formation (carotid ligation). In vitro, SPM treatment inhibited mouse aortic smooth muscle cell migration (IC₅₀ ≅ 1 nM) to a PDGF gradient and reduced TNF-α-stimulated p65 translocation, superoxide production, and proinflammatory gene expression (MCP-1). In vivo, adult FVB mice underwent unilateral carotid artery ligation with administration of RvD2, MaR1, or vehicle (100 ng by intraperitoneal injection at 0, 1, 3, 5, and 7 d after ligation). In ligated carotid arteries at 4 d, SPM treatment was associated with reduced cell proliferation and neutrophil and macrophage recruitment and increased polarization of M2 macrophages in the arterial wall. Neointimal hyperplasia (at 14 d) was notably attenuated in RvD2 (62%)- and MaR1 (67%)-treated mice, respectively. Modulation of resolution pathways may offer new opportunities to regulate the vascular injury response and promote vascular homeostasis.

Keywords: fatty acid; inflammation; neointimal hyperplasia; vascular remodeling.

Figure 1.

RvD2 and MaR1 reduce chemotaxis of mouse ASMC to PDGF-BB in vitro. A) ASMC migration response to PDGF-BB (50 ng/ml, 4 h) with RvD2 (1 nM) or MaR1 (1 nM) pretreatment, using a transwell assay with Boyden chambers. Representative images stained with DAPI. B) Results of the transwell assay are expressed as percentage change in migration from unstimulated control (no PDGF). Inhibition of chemotaxis is shown for both RvD2 and MaR1 in dose-dependent manner (n = 3). Inhibition is PTX sensitive. C) In the ASMC scratch assay, wound closure was significantly reduced by RvD2 and MaR1 treatments. *P < 0.05, **P < 0.01 vs. control; Dunnett post hoc test. Unpaired t test between RvD2 or Mar1 100 nM with PTX vs. without PTX in B.

Figure 2.

Cytokine activation of mouse ASMCs is altered by RvD2 and MaR1 in vitro. A, B) RvD2 or MaR1 treatment reduces TNF-α–induced superoxide production in cultured ASMCs. Superoxide production was evaluated by DHE staining. ASMCs were stimulated with TNF-α (10 ng/ml, 18 h) in the presence or absence of RvD2 or MaR1. A) Representative merged images of DHE staining, counterstained with DAPI. Negative control (no treatment), positive control (TNF-α only), TNF-α with RvD2 (100 nM), and TNF-α with MaR1 (100 nM). B) Quantitative comparison of DHE staining intensity normalized to control. *P < 0.05, **P < 0.01 vs. positive control (TNF-α); Dunnett post hoc test. Results shown are means ± se. C, D) RvD2 and MaR1 modulate p65 translocation in ASMCs. Representative merged images of p65 translocation assay, counterstained with DAPI. C) Positive control (TNF-α 10 ng/ml for 4 h), negative control (no treatment), TNF-α with 100 nM RvD2, and TNF-α with 100 nM MaR1. D) Quantitative comparison represented by the ratio of p65 fluorescence intensity of nucleus to cytoplasm. *P < 0.05, **P < 0.01 vs. control; Dunnett post hoc test. Results are means ± se. E) Proinflammatory gene expression in ASMCs was modulated by RvD2 and MaR1. ASMCs were stimulated with TNF-α (5 ng/ml) for 18 h in the presence or absence of RvD2 or MaR1 (100 or 10 nM, respectively), and quantitative PCR was performed. Proinflammatory gene expression of IL-6, ICAM-1, MCP-1, and VCAM-1 were measured. Shown is a significant attenuation in the expression of MCP-1 at 10 and 100 nM RvD2 and at 100 nM MaR1, with a nonsignificant trend of decrease in IL-6, ICAM-1, and VCAM-1 (n = 5). *P < 0.05 vs. control; Dunnett post hoc test. Results are means ± se.

Figure 3.

Systemic RvD2 and MaR1 treatment altered acute leukocyte recruitment and macrophage phenotype following carotid ligation in mice. A) Neutrophils were detected by immunostaining of NIMP-R14. Pictures of nonligated arteries (sham operation) were shown as negative controls. Ligated carotid arteries treated with vehicle served as positive controls. All samples were counterstained with DAPI (scale bars, 40 μm). B) Total monocytes/macrophages were detected by MOMA-2 and M2 macrophages were also detected by liver Arg-1 (bar, 40 μm). C) Quantification of neutrophil infiltration was shown as ratios of NIMP-R14–positive cells to total cells (DAPI) numbers. The extent of neutrophil infiltration was reduced by either RvD2 (48% reduction) or MaR1 (56% reduction) treatments. D) Quantification of total monocyte/macrophage infiltration or M2 macrophage was shown as ratios of positive cells of MOMA-2 or Arg-1 to total cells (DAPI) numbers. The ratio of MOMA-2 positive cells was reduced by RvD2 (48% reduction) or MaR1 (43% reduction) treatments. This overall reduction in monocyte/macrophage number included the Arg-1–positive subset. E) The ratio of M2 macrophages was calculated as the ratio of Arg-1–positive to MOMA-2–positive cells. The relative proportion of M2 in ligated carotid arteries was higher in RvD2- (53% higher) or MaR1- (19% higher) treated animals than in controls. Values were measured as average of 3 sections for 1 artery (n = 4). *P < 0.05, **P < 0.01 vs. control; Dunnett post hoc test. Results are means ± se. M, media; A, adventitia.

Figure 4.

Systemic RvD2 and MaR1 treatment reduces the early proliferative response in ligated carotid arteries at 4 d. A) Proliferating cells are shown by immunostaining of Ki67. All samples were counterstained with DAPI. Images of nonligated arteries (sham operation) are shown as negative controls. Vehicle treatment is shown as positive control (scale bars, 100 μm). B) Quantification of Ki67-positive cells to total cells (DAPI) numbers. The Ki67 proliferation indices were reduced by RvD2 (77% reduction) or MaR1 (52% reduction) treatments. Values were measured as an average of 3 sections for each artery (n = 4). *P < 0.05, **P < 0.01 vs. control; Dunnett post hoc test. Results are means ± se.

Figure 5.

Vessel remodeling in the ligated carotid artery was modulated by systemic RvD2 and MaR1 treatment at 14 d. Controls were ligated arteries treated with vehicle. A) Representative photomicrographs of elastin staining are shown (bar, 100 μm). B) Representative photomicrographs of immunostaining of α-smooth muscle cell actin are shown (scale bars, 100 μm). Smooth muscle cells were mainly observed in neointima in all of 3 groups.

Figure 6.

Systemic RvD2 and MaR1 treatment attenuates low flow-induced neointima formation in the mouse carotid artery at 14 d. Controls were ligated arteries treated with vehicle. A) Neointima/media ratio was significantly reduced by treatment by RvD2 (67%) or MaR1 (71%). B) Summary of morphometrical analysis shown as areas of each layer of injured arteries. Neointimal formation was attenuated by RvD2 (62%) or MaR1 (67%) treatment, although the areas of media, adventitia, lumen, and whole vessel were not changed by treatments. Values were measured as average of 3 sections for 1 artery (n = 10). **P < 0.01 vs. control; Dunnett post hoc test. Results are means ± se.