ApoM-bound S1P acts via endothelial S1PR1 to suppress choroidal neovascularization and vascular leakage

Abstract

Neovascular age-related macular degeneration (nAMD) is a major cause of vision loss worldwide. Current standard of care is repetitive intraocular injections of vascular endothelial growth factor (VEGF) inhibitors, although responses may be partial and non-durable. We report that circulating sphingosine 1-phosphate (S1P) carried by apolipoprotein M (ApoM) acts through the endothelial S1P receptor 1 (S1PR1) to suppress choroidal neovascularization (CNV) in mouse laser-induced CNV, modeling nAMD. In humans, low plasma ApoM levels were associated with increased choroidal and retinal pathology. Additionally, endothelial S1pr1 knockout and overexpressing transgenic mice showed increased and reduced CNV lesion size, respectively. Systemic administration of ApoM-Fc, an engineered S1P chaperone protein, not only attenuated CNV to an equivalent degree as anti-VEGF antibody treatment but also suppressed pathological vascular leakage. We suggest that modulating circulating ApoM-bound S1P action on endothelial S1PR1 provides a novel therapeutic strategy to treat nAMD.

Keywords: Age-related macular degeneration; Angiogenesis; Apolipoprotein M; High-density lipoprotein; Sphingosine 1-phosphate; Vascular leak.

Fig. 1

ApoM-bound S1P action via endothelial S1PR1 attenuates choroidal neovascularization (CNV) in mice. a Representative images of the retina from Apom^−/−, Apom^+/− and wild type (wt) control mice at day (D)14. b Quantification of IB4-positive CNV area (wt (n = 13 eyes/ 7 mice), Apom^+/− (n = 10 eyes/ 5 mice) and Apom^−/− (n = 12 eyes/6 mice), One-way ANOVA***, P = 0.0001, **, P = 0.0037). c Representative images of the CNV lesion from Apom^TG and littermate control at D14. d Quantification of CNV area (wt (n = 12 eyes/ 7 mice) and Apom^TG (n = 13 eyes/ 8 mice), unpaired, two-tailed students t-test (**, P = 0.0013)). e Representative images of the retina from S1pr1 ECKO and littermate control at D14. f Quantification of CNV area (wt (n = 14 eyes/ 8 mice) and S1pr1 ECKO (n = 13 eyes/ 8 mice), unpaired, two-tailed Student's t-test (**, P = 0.0056)). g Representative images of the retina from S1pr1 ECTG and wt control at D14. h Quantification of CNV area (wt (n = 14 eyes/ 8 mice) and S1pr1 ECTG (n = 14 eyes/ 7 mice), (unpaired, two-tailed students t-test ***, P = 0.0003)). Data show means ± SD. Scale bar: 500 μm

Fig. 2

Intraperitoneal ApoM-Fc administration suppresses CNV lesions. a Representative western blot showing ApoM-Fc and endogenous ApoM level in plasma 72 h after systemic ApoM-Fc (4 mg/kg) administration in mice. b, c Quantification of the bands shown in a. (Vehicle (n = 3 mice) and ApoM-Fc (n = 4 mice), unpaired, two-tailed Student’s t-test (*, P = 0.0102, ns, P > 0.9999). d ApoM-Fc protein levels in mouse plasma at various time points after a single i.p. injection of ApoM-Fc (n = 4 mice/ time point, one-way ANOVA, ***, P < 0.0001, **, P < 0.001). e S1P levels in plasma at various time points after ApoM-Fc administration (n = 4/ time point, one-way ANOVA, *, P < 0.01). f Experimental scheme for ApoM-Fc regimen. Mice were given ApoM-Fc (4 mg/kg, i.p.) or vehicle at the time indicated. g Representative images of CNV lesion from vehicle- and ApoM-Fc treated mice at D10. h Quantification of CNV area (Vehicle (n = 14 eyes/ 8 mice) and ApoM-Fc (n = 18 eyes/ 10 mice), unpaired, two-tailed Student’s t-test **, P = 0.0022). Data are means ± SD. Scale bar: 100 μm

Fig. 3

Subcutaneous ApoM-Fc administration attenuates CNV lesions. a Representative western blot showing ApoM-Fc and endogenous ApoM level in plasma 72 h after subcutaneous ApoM-Fc (4 mg/kg) administration in mice. b, c Quantification of bands shown in a. (Vehicle (n = 4 mice) and ApoM-Fc (n = 4 mice), unpaired, two-tailed Student's t-test (***, P = 0.0008, ns, P > 0.9999). d ApoM-Fc protein levels in mouse plasma at various time points after a single subcutaneous injection of ApoM-Fc (n = 4 mice/ time point, one-way ANOVA, *, P < 0.01, **, P < 0.001). e S1P levels in plasma at various time points after ApoM-Fc administration (n = 4/ time point, one-way ANOVA, **, P < 0.001). f Experimental scheme for ApoM-Fc regimen. Mice were given ApoM-Fc (4 mg/kg) or vehicle subcutaneous at the time indicated. g Representative images of IB4 staining-positive CNV lesion from vehicle- and ApoM-Fc treated mice at D14. h Quantification of CNV area (Vehicle (n = 9 eyes/ 5 mice) and ApoM-Fc (n = 10 eyes/ 6 mice), unpaired, two-tailed Student's t-test *, P = 0.0188). Data are means ± SD. Scale bar: 500 μm

Fig. 4

ApoM-Fc and anti-VEGF antibody reduces CNV lesion size and vascular leakage. a Experimental scheme for ApoM-Fc regimen and α-VEGF treatment. Mice were given either ApoM-Fc (4 mg/kg, i.p.) or vehicle or anti-VEGF (0.5 mg/kg, i.p.) at D0 and D5. b Representative images of IB4 staining-positive CNV lesions from treatment group at D10. Scale bar: 500 μm. c Quantification of CNV area (control (n = 17 eyes/ 10 mice), ApoM-Fc (n = 16 eyes/ 10 mice), anti-VEGF (n = 11 eyes/ 6 mice) and ApoM-Fc + anti-VEGF (n = 16 eyes/ 10 mice), One-way ANOVA ****, P < 0.0001, ***, P = 0.0004, **, P = 0.0020, ns, P > 0.9999). d Plasma S1P measurement at D10 (control (n = 7 mice), ApoM-Fc (n = 8 mice), α-VEGF (n = 7 mice) and ApoM-Fc + anti-VEGF (n = 7 mice), One-way ANOVA *, P < 0.01, ns, P > 0.9999). e Experimental scheme for ApoM-Fc regimen and anti-VEGF treatment and fundus fluorescein angiography (FFA) at D10. f Representative images of fluorescein angiogram of IgG control-, ApoM-Fc-, anti-VEGF-, and both ApoM-Fc and anti-VEGF-treated mice 10 days after CNV induction. g Corrected fluorescence intensity normalized by optic nerve head area (control (n = 6 eyes/ 4 mice), ApoM-Fc (n = 7 eyes/ 4 mice), anti-VEGF (n = 7 eyes/ 4 mice) and ApoM-Fc + anti-VEGF (n = 8 eyes/ 5 mice), One-way ANOVA **, P = 0.0075, *, P < 0.01, ns, P > 0.9999). h Mice were given ApoM-Fc (4 mg/kg) or ApoM-TM (4 mg/kg) i.p. 24 h prior to intraocular injection of VEGF (0.1 µg/ eye), followed by FITC-albumin. i Representative images of vehicle-, ApoM-Fc- and ApoM-TM-treated retina. Scale bar: 20 μm. j Quantification of VEGF-induced vascular leakage (vehicle (n = 8 mice), ApoM-Fc (n = 7 mice), ApoM-TM (n = 8 mice), One-way ANOVA *, P < 0.01). Data show means ± SD