Fig 2: Effects of PGF and VEGF inhibition on microgliosis in retinal flat mounts. a Representative images of Iba1-stained microglia/macrophages in single laser spots of retinal flat mounts 3 and 7 days after laser coagulation. Scale bar: 100 μm. b Quantification of mononuclear phagocytes per laser spot in retinal flat mounts 3 days after laser-induced damage (n = 16 laser spots for IgG, n = 14 laser spots for aVEGF, n = 14 laser spots for aPGF, n = 15 laser spots for aVEGF/aPGF, n = 14 laser spots for aflibercept). c Quantification of mononuclear phagocytes per laser spot in retinal flat mounts 7 days after laser-induced damage (n = 14 laser spots for IgG, n = 14 laser spots for aVEGF, n = 15 laser spots for aPGF, n = 27 laser spots for aVEGF/aPGF, n = 28 laser spots for aflibercept). d Quantification of Iba1 signals 3 days after laser coagulation in retinal flat mounts by counting the mean of colored pixels per image (n = 16 laser spots for IgG, n = 14 laser spots for aVEGF, n = 14 laser spots for aPGF, n = 15 laser spots for aVEGF/aPGF, n = 14 laser spots for aflibercept). e Quantification of Iba1 signals 7 days after laser coagulation in retinal flat mounts by counting the mean of colored pixels per image (n = 14 laser spots for IgG, n = 14 laser spots for aVEGF, n = 15 laser spots for aPGF, n = 27 laser spots for aVEGF/aPGF, n = 28 laser spots for aflibercept). f Interleukin 6 (IL-6) levels in retinal flat mounts 6 h after laser damage quantified by ELISA (n = 8 flat mounts per group). Naive (not lasered) animals were used as controls. g Interleukin 1β (IL-1β) levels in retinal flat mounts 6 h after laser damage quantified by ELISA (n = 8 flat mounts per group). h Tumor necrosis factor (TNF) levels in retinal flat mounts 6 h after laser damage quantified by ELISA (n = 8 flat mounts per group). Data are shown as mean ± SEM (*P < 0.05, **P < 0.01, ***P < 0.001)

Fig 3: The abundances of PlGF and VEGF are increased in HG-treated hRECs. hRECs were treated with high glucose (HG, 25 mM) for the indicated time periods. Non-treated cell (CTL) or mannitol (MN) treated cell was used as the controls. (a). Quantitative real PCR was performed to assess the mRNA levels of VEGF and PlGF. (b). Culture medium was collected and ELISA was performed to quantitate the amount of secreted PlGf and VEGF, respectively. (c). Cells were lysed and total protein was extracted for immunoblotting with anti-PlGF and anti-VEGF antibodies. Full length blots are shown for PlGF and VEGF. MN-treated hRECs were used as the treatment control (Supplementary Fig. 9). Data are presented mean ± SD. n = 3. b: *p < 0.05, **p < 0.01, ***p < 0.001. n.s. non-significance. a and c: *p < 0.01 vs 0 h.

Fig 4: Mononuclear phagocytes co-express PGF and VEGF mRNAs in laser lesions. Scale bar: 100 μm. a Representative images of in situ hybridization of single laser spots in RPE/choroidal flat mounts of IgG-treated animals at days 3 and 7. Probes targeted both Iba1 and VEGF. b Representative images of in situ hybridization of laser spots in RPE/choroidal flat mounts of IgG-treated animals at days 3 and 7. Probes targeted both Iba1 and PGF. c Representative images of in situ hybridization of laser spots in RPE/choroidal flat mounts of IgG-treated animals at days 3 and 7. Probes targeted both PGF and VEGF. The frames show higher magnification areas

Fig 5: VNS-Induced PlGF Release and Immune Response Initiation Are Mediated by α-Adrenergic Receptor Signaling(A–C) Vehicle-, phentolamine-, or propranolol-treated mice were subjected to either sham or VNS procedure. (A) SSNA (vehicle, sham versus VNS: q(17) = 4.936, ∗p < 0.05; phentolamine, sham versus VNS: q(17) = 4.753, ∗p < 0.05; propranolol, sham versus VNS: q(17) = 4.729, ∗p < 0.05) and (B and C) TH-positive staining (vehicle, sham versus VNS: q(17) = 7.480, ∗∗∗p < 0.001; phentolamine, sham versus VNS: q(17) = 7.118, ∗∗p < 0.01; propranolol, sham versus VNS: q(17) = 4.599, ∗p < 0.05) (scale bar, 100 μm) were increased in all VNS groups.(D and E) Although vehicle- and propranolol-treated mice shown a similar activation of PlGF, phentolamine treatment significantly reduced it (vehicle, sham versus VNS: q(17) = 5.856, ∗∗p < 0.01; propranolol, sham versus VNS: q(17) = 7.861, ∗∗∗p < 0.001; VNS, phentolamine versus vehicle: q(17) = 4.991, ∗p < 0.05; VNS, phentolamine versus propranolol: q(17) = 7.361, ∗∗∗p < 0.001) (scale bar, 50 μm).(F and G) A reduced area of CD3+ cells (red), delimited by B220+ cells (green), is observed in both vehicle- and propranolol-treated VNS mice. Conversely, phentolamine-treated VNS mice display a CD3+ area comparable with sham mice (vehicle, sham versus VNS: q(17) = 8.337, ∗∗∗p < 0.001; propranolol, sham versus VNS: q(17) = 4.955, ∗p < 0.05; VNS, phentolamine versus vehicle: q(17) = 6.983, ∗∗p < 0.01; VNS, phentolamine versus propranolol: q(17) = 6.688, ∗∗p < 0.01) (scale bar, 100μm).For all panels, data were obtained from n = 4 sham vehicle, n = 4 VNS vehicle, n = 4 sham phentolamine, n = 4 VNS phentolamine, n = 3 sham propranolol, and n = 4 VNS propranolol mice and are represented as mean ± SEM.