Eicosapentaenoic acid prevents the progression of intracranial aneurysms in rats

Abstract

Background: As subarachnoid hemorrhage due to rupture of an intracranial aneurysm (IA) has quite a poor outcome despite of an intensive medical care, development of a novel treatment targeting unruptured IAs based on the correct understanding of pathogenesis is mandatory for social health.

Methods: Using previously obtained gene expression profile data from surgically resected unruptured human IA lesions, we selected G-protein coupled receptor 120 (GPR120) as a gene whose expression is significantly higher in lesions than that in control arterial walls. To corroborate a contribution of GPR120 signaling to the pathophysiology, we used an animal model of IAs and examine the effect of a GPR120 agonist on the progression of the disease. IA lesion was induced in rats through an increase of hemodynamic stress achieved by a one-sided carotid ligation and induced hypervolemia. Eicosapentaenoic acid (EPA) was used as an agonist for GPR120 in this study and its effect on the size of IAs, the thinning of media, and infiltration of macrophages in lesions were examined.

Result: EPA administered significantly suppressed the size of IAs and the degenerative changes in the media in rats. EPA treatment also inhibited infiltration of macrophages, a hallmark of inflammatory responses in lesions. In in vitro experiments using RAW264.7 cells, pre-treatment of EPA partially suppressed lipopolysaccharide-induced activation of nuclear factor-kappa B and also the transcriptional induction of monocyte chemoattractant protein 1 (MCP-1), a major chemoattractant for macrophages to accumulate in lesions. As a selective agonist of GPR120, TUG-891, could reproduce the effect of EPA in RAW264.7 cells, EPA presumably acted on this receptor to suppress inflammatory responses. Consistently, EPA remarkably suppressed MCP-1 expression in lesions, suggesting the in vivo relevance of in vitro studies.

Conclusions: These results combined together suggest the potential of the medical therapy targeting GPR120 or using EPA to prevent the progression of IAs.

Keywords: Eicosapentaenoic acid; GPR120; Intracranial aneurysm.

Fig. 1

Expression of GPR120 in human IA lesion and the control arterial wall. The representative images of immunostaining of specimens from human IA lesion (the upper panels) and the control arterial wall (middle meningeal artery, the lower panels) for GPR120 are shown. The magnified images corresponding to the squares in the left panels are also shown on a right. Bar; 200 μm. The arrow head indicates the endothelial cells where GPR120 expression is induced

Fig. 2

Induction of GPR120 expression in endothelial cells and macrophages of human IA lesions. The representative images of immunostaining of specimens from human IA lesion (the upper panels) and the control arterial wall (middle meningeal artery, the lower panels) for GPR120 (green), CD31, a marker for endothelial cells (red in a), CD68, a marker for macrophages (red in b) or smooth muscle α-actin (SMA), a marker for smooth muscle cells (red in c), nuclear staining by DAPI (blue) and merged images are shown. The magnified images corresponding to the squares in the left panels are also shown on a right. In d, the image of immunostaining without a primary antibody is shown as a negative control study. Bar; 10 μm. Note the induction of GPR120 in CD31-positive endothelial cells and CD68-positive macrophages while expression in smooth muscle cells is constitutive

Fig. 3

Inhibitory effect of EPA on the size of intracranial aneurysms in a rat model. a Plasma concentration of EPA, docosahexaenoic acid (DHA), and arachidonic acid (AA) in rats orally administered EPA. Rats were orally administered EPA (vehicle, n = 9, 100 mg/kg/day, n = 8, 1000 mg/kg/day, n = 9) once a day and, on the 5th day, plasma concentration of EPA, DHA, and AA was measured and the ratio of EPA over AA was also calculated. Data represents box-and-whisker plots. Statistical analysis was done by a Kruskal-Wallis test. *p < 0.05. b, c Effect of EPA on the size of induced aneurysms. EPA was administered in a rat model subjected to aneurysm induction once a day for 12 days and the size of induced aneurysms at right anterior cerebral-olfactory artery bifurcation was measured after Elastica van Gieson staining (vehicle, n = 8, 100 mg/kg/day, n = 10, 1000 mg/kg/day, n = 9). The representative microscopic images of induced aneurysms with Elastica van Gieson staining in each group are shown in b. Bar; 50 μm. Data represents box-and-whisker plots (c). Statistical analysis was done by a Kruskal-Wallis test. *p < 0.05. d Plasma concentration of Resolvin E1 in rats administered EPA. Rats were orally administered EPA once a day and, after 12 days, plasma concentration of Resolvin E1 was measured by ELISA (vehicle, n = 8, 1000 mg/kg/day, n = 9). Data represents box-and-whisker plots. Statistical analysis was done by a Mann-Whitney test. *p < 0.05

Fig. 4

Inhibitory effect of EPA on the thinning of medial smooth muscle cell layer in lesions. a–c Inhibitory effect of EPA on the thinning of medial smooth muscle cell layer. Medial smooth muscle cells in IA lesions from rats treated with vehicle or EPA (100 or 1000 mg/kg/day) for 12 days were visualized by immunostaining for smooth muscle α-actin (SMA), a marker for smooth muscle cells. The representative images from immunohistochemistry for SMA (red), nuclear staining by DAPI (blue) and merged images are shown (a). The magnified images corresponding to the squares are shown on a right (a). Bar; 50 μm. The thickness of the thinnest portion in the media of IA lesions (b) and the relative thickness over that of the distal normal part (c) were calculated (vehicle, n = 8, 100 mg/kg/day, n = 10, 1000 mg/kg/day, n = 9). Data represents box-and-whisker plots. Statistical analysis was done by a Kruskal-Wallis test. *p < 0.05

Fig. 5

Inhibitory effect of EPA on the infiltration of macrophages in lesions. a, b Inhibitory effect of EPA on the infiltration of macrophages in lesions. Macrophages infiltrating in IA lesions from rats treated with vehicle or EPA (100 or 1000 mg/kg/day) for 12 days were visualized by immunostaining for CD68, a marker for macrophages. The representative images from immunohistochemistry for CD68 (green) or SMA, a marker for smooth muscle cells (red), nuclear staining by DAPI (blue) and merged images are shown (a). The magnified images corresponding to the squares are also shown in the lower panels (a). Bar; 10 μm. The number of infiltrating macrophages in IA lesions was calculated (vehicle, n = 8, 100 mg/kg/day, n = 10, 1000 mg/kg/day, n = 9). Data represents box-and-whisker plots (b). Statistical analysis was done by a Kruskal-Wallis test. *p < 0.05

Fig. 6

Suppression of NF-κB activation and CCL2 expression by EPA treatment in vitro. a Suppression of LPS-induced NF-κB activation by the pre-treatment with EPA. RAW264.7 cells were treated with EPA (300 μM) for 60 min and then stimulated with LPS (3.3 μg/ml) for additional 10 min. NF-κB activation was then assessed by Western blot analysis using the whole cell lysate. The representative images of Western blot analyses from two independent experiments for phosphorylated form of NF-κB p65 subunit (Ser536), NF-κB p65 subunit or α-tubulin served as an internal control are shown. b Suppression of LPS-induced Ccl2 expression by the pre-treatment with EPA. RAW264.7 cells were treated with EPA (300 μM, n = 6) or an agonist for GPR120, TUG-891 (3 μM, n = 7), for 60 min and then stimulated with LPS (3.3 μg/ml) for additional 60 min. Ccl2 expression was then assess by RT-PCR analyses. Data represents mean ± SD. Statistical analysis was done by a Kruskal-Wallis test. *p < 0.05. c Suppression of CCL2 expression in IA lesions by EPA in rats. EPA (1000 mg/kg/day) was administered in a rat model subjected to aneurysm induction once a day for 12 days and expression of CCL2 in IA lesions was then assessed by immunohistochemistry. The representative images from immunohistochemistry for CCL2 (green), nuclear staining by DAPI (blue), and merged images are shown. The image from immunostaining without a primary antibody for CCL2 is also shown as a negative control study in the lowest panel. Bar; 20 μm