Modeling Procedure, Validation, and Cervical Artery Anatomy.

(A1-G1) Schematic illustration of unilateral pterygopalatine ophthalmic artery occlusion (UPOAO). (A2-G2) Practical operation of UPOAO. (A1, A2) Blunt separation and left cervical arteries exposure. (B1, B2) Arterial suture ligation. (C1, C2) Silicone wire embolus insertion. The artery was incised a hole and the silicone wire embolus was inserted. (D1, D2) Artery disconnection and movement of the silicone wire embolus. The artery was cut along the hole and the silicone wire embolus was retracted and reinserted. (E1, E2) Removal of the silicone wire embolus and reperfusion. After some time of ischemia, the silicone wire embolus was removed. (F1, F2) Removal of sutures. The sutures at both ends of the disconnected vessel were knotted, and other two sutures were removed. (G1, G2) Anatomic reduction and sutures of the skin. (H and I) Flat-mounted retina after perfusing rhodamine-labeled canavalin A into the heart of a UPOAO mouse. The sham eye was an unpracticed control eye, and the UPOAO lateral eye was the experimental eye. Retinal vessels in the sham eye (H) were filled with fluorescence while retinal vessels in the UPOAO lateral eye (I) were unfilled. Scale bar = 1 mm. (J and K) Fluorescein Fundus Angiography (FFA) of a UPOAO modeling mouse. The retina vessels (J) were perfused while UPOAO lateral retinal perfusion (K) was delayed. (L) The schematic illustration of cervical artery anatomy and ocular blood supply. Embolization of PPA led to ocular ischemia. The red arrow indicates the embolism site of the silicone wire embolus. The silicone wire embolus is the type 602156 wire embolus extended to 7mm and has a diameter of 0.19mm. The blue arrows indicate the molding locations of the HIOP model and the UCCAO model, respectively. CCA: common carotid artery; ICA: internal carotid artery; ECA: external carotid artery; PPA: pterygopalatine artery; MCA: middle cerebral artery; Post. Sup. Alveolar art.: Posterior superior alveolar artery; Infraorbital art.: infraorbital artery; OA: ophthalmic artery; SPCA: short posterior ciliary artery; LPCA: long posterior ciliary artery; CRA: central retinal artery.

Antibodies used in staining of retinal sections

List of primers used in this study.

Staining and quantification of retinal ganglion cells (RGCs) at different ischemia and reperfusion times.

Flat-mounted retina RGCs were labeled with Brn3a staining. (A, B) Representative pictures of peripheral retinal visual field (A), and quantification of surviving RGCs (B) in the 30-minute ischemia and 3-day reperfusion group. N = 3. (C, D) Representative pictures of peripheral retinal visual field (C), and quantification of surviving RGCs (D) in the 30-minute ischemia and 7-day reperfusion group. N = 3. (E, F) Representative pictures of peripheral retinal visual field (E), and quantification of surviving RGCs (F) in the 60-minute ischemia and 3-day reperfusion group. N = 5. (G, H) Representative pictures of peripheral retinal visual field (G), and quantification of surviving RGCs (H) in the 60-minute ischemia and 7-day reperfusion group. N = 5. The results showed the evident RGCs loss after 60-minute ischemia. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, t-test. Scale bar = 100 μm.

Comparison of electroretinographic (ERG) dark-adapted responses at different ischemic and reperfusion times.

Following evaluating survival RGCs, visual function between sham vs. UPOAO experimental eyes at different ischemia and reperfusion times was evaluated using ERG. (A) Representative wave lines in four groups at the stimulus light intensity of 0.1, 1.0, and 10.0 cd.s/m2, respectively. (B) Quantification of amplitudes of a-waves and b-waves in the 30-minute ischemia and 3-day reperfusion group. N = 5. (C) Quantification of amplitudes of a-waves and b-waves in the 30-minute ischemia and 7-day reperfusion group. N = 5. (D) Quantification of amplitudes of a-waves, b-waves in the 60-minute ischemia and 3-day reperfusion group. N = 7. (E) Quantification of amplitudes of a-waves and b-waves in the 60-minute ischemia and 7-day reperfusion group. N = 8. Dark-adapted responses were similar in terms of a-wave amplitude but significantly decreased in b-wave amplitude in the 60-minute ischemia groups. The amplitudes of b-waves declined at 3 days and more intensely at 7 days. (F, G) Representative OPs-waves and quantification of amplitudes in the 60-minute ischemic groups. N = 7 in the 3-day reperfusion group; n = 8 in the 7-day reperfusion group. The OPs-wave amplitudes decreased significantly at 7-day reperfusion. The decline of amplitudes in b-waves, OPs-waves, and the loss of RGCs support the selection of a 60-minute ischemic duration as an appropriate choice. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, two-way ANOVA test in a-waves and b-waves; paired t-test in OPs-waves.

Response time for a-waves and b-waves in ERG under different light intensities.

(A, B) Response times of a-waves and b-waves under different light intensities at 3 and 7 days post-UPOAO in the 30-minute ischemia group. (C, D) Response times of a-waves and b-waves under different light intensities at 3 and 7 days post-UPOAO in the 60-minute ischemia group. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, two-way ANOVA test.

Changes in retina morphology of the UPOAO animals.

(A) Representative OCT images of mouse retina at 3 days. Green lines indicated the OCT scan area from the optic disc. IMRL: inner middle retina layer, including RNFL, GCL and IPL layers. MRL: middle retina layer, including INL and OPL layers. ORL: outer retina layer, including IS, OS and RPE layers. TRT: total retina thickness, including all layers of the retina. (B and C) Quantification of IMRL and TRT thickness at 3 days. The thickness of IMRL and TRT in OCT was measured and compared at distances of 1.5 PD, 3.0 PD and 4.5 PD from the optic disc, respectively. N = 5. (D) Representative OCT images of mouse retina at 7 days. (E and F) Quantification of IMRL and TRT thickness at 7 days. N = 5. (G) Representative HE images of mouse retina at 3 days. (H, I and J) Quantification of NFL + GCL, IPL and INL thickness at 3 days. Retinal thickness in HE was measured near the optic nerve head and compared. N = 3. (K) Representative HE images of mouse retina at 7 days. (L, M and N) Quantification of NFL + GCL, IPL and INL thickness at 7 days. N = 3. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, two-way ANOVA test in OCT and paired t-test in HE. PD: papillary diameters; RNFL: retinal nerve fiber layer; GCL: ganglion cell layer; IPL: inner plexiform layer; INL: inner nuclear layer; OPL: outer plexiform layer; IS: inner segment; OS: outer segment; RPE: retinal pigment epithelium. Scale bar = 50 μm.

Quantification of MRL and ORL thickness in OCT at 3-d and 7-d reperfusion of the UPOAO animals.

MRL and ORL thickness in OCT were measured and compared at distances of 1.5 PD, 3.0 PD and 4.5 PD from the optic disc, respectively. (A and B) Quantification of MRL (A) and ORL (B) thickness at 3 days. n = 5. No significant differences. (C and D) Quantification of MRL (C) and ORL (D) thickness at 7 days. N = 5. No significant differences. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, two-way ANOVA test.

Changes of bipolar cells, Horizontal cells and cholinergic amacrine cells in the UPOAO mice.

(A) Representative images of DAPI and PKCα co-staining of mouse retina at 3 days. (B) Representative images of DAPI and PKCα co-staining of mouse retina at 7 days. (C) Quantification of PKCα fluorescence density at 3 days. (D) Quantification of PKCα fluorescence density at 7 days. (E) Representative images of DAPI and ChAT co-staining of mouse retina at 3 days. (F) Representative images of DAPI and ChAT co-staining of mouse retina at 7 days. (G) Quantification of ChAT fluorescence density at 3 days. (H) Quantification of ChAT fluorescence density at 7 days. Data points were from sections of four animals. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, t- test. Scale bar = 50 μm.

Changes of photoreceptor cells in the UPOAO mice.

(A) Representative images of DAPI and Recoverin co-staining of mouse retina at 3 days. (B) Representative images of DAPI and Reconerin co-staining of mouse retina at 7 days. Data points were from sections of four animals. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, t- test. Scale bar = 50 μm.

Transcriptomic features at different times of reperfusion.

RNA-seq evaluation of 0, 3 and 7 days in UPOAO and mostly enriched in the pathways of immune cells migration, oxidative stress and immune inflammation. (A, D, G) Volcano plots showing differential expression genes (DEGs) in the UPOAO and sham eyes in the no-perfusion group (A), 3-day group (D) and 7-day group (G), respectively. Red dots: significant upregulated genes, green and blue dots: significant downregulated genes, grey dots: stable expressed genes, adjusted P < 0.05. log2FC = 1. (B, E, H) Gene ontology (GO) analysis of differential genes in the no-perfusion group (B), 3-day group (E) and 7-day group (H). The overlapping DEGs between UPOAO and Sham eyes were enriched in the pathways of immune cells migration (0d), oxidative stress (3d) and immune inflammation (7d), respectively. (C, F, I) Heatmap showing TOP 150 overlapping DEGs related with pathways of immune cells migration (C), oxidative stress (F) and immune inflammation (I) based on the DEGs between UPOAO and Sham at no-perfusion, 3-day and 7-day reperfusion, respectively. Ranking was determined by the magnitude of fold change. In each heatmap, upper box showed the top 10 up-regulated genes, and the below one showed the top 10 down-regulated genes.

Hub genes in PPI analysis and GSEA analysis at no-reperfusion in UPOAO.

(A) Hub genes in PPI analysis of DEGs at no-reperfusion group. (B-D) GSEA analysis of the pathway of the cell killing (B), leukocyte mediated-immunity (C) and lymphocyte mediated-immunity (D) at no-reperfusion group.

Hub genes, GSEA analysis at 3-day reperfusion in UPOAO and relation with mitochondrial genes.

(A) Hub genes in PPI analysis of DEGs at 3-day reperfusion group. (B) GSEA analysis of the pathway of plasma membrane protein complex (top left), bounding membrane of organelle (top right) and cell surface (bottom). (C) Venn diagram of mitochondrial genes and DEGs at 3-day reperfusion (44 genes). (D) GO analysis of 44 overlapping DEGs.

Hub genes, GSEA analysis at 7-day reperfusion in UPOAO and relation with immune genes.

(A) Hub genes in PPI analysis of DEGs at 7-day reperfusion group. (B) GSEA analysis of the pathway of reactome innate immune system (top left), reactome immunoregulatory interactions between a lymphoid and a non-lymphoid cell (top right), reactome immune system (bottom left) and reactome adaptive immune system (bottom right). (C) Venn diagram of immune genes and DEGs at 7-day reperfusion (112 genes). (D) GO analysis of 112 overlapping DEGs.

The co-expressed genes for 3-day and 7-day reperfusion.

(A) Venn diagram of DEGs at 3-day and 7-day reperfusion and list of 17 overlapping DEGs. (B) GO analysis of 17 overlapping DEGs.

Immune inflammation-related gene expression was upregulated in the UPOAO-7d group.

ABCA1, CD86, CD48, TLR4, TLR6, and TSPO genes were significantly upregulated in the UPOAO-7d group, while other immune-inflammatory and chemotaxis-related genes showed no significant differences. The data points are from retina of four animals. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, Paired t-test.

Peripheral leukocyte infiltration and retinal resident microglial activation.

Rhodamine-labeled canavalin A was immediately used for cardiac perfusion to mark blood vessels, followed by CD45 immunofluorescent staining to observe the relationship between blood vessels and CD45+ cells in Sham (A), UPOAO-1d (B), 3d (C), and 7d (D). The white arrows indicate the presence of CD45+ cells in the blood vessels. UPOAO-1d (E), 3d (F), 7d (G) whole-retinal CD45+ cell counts were performed. The cellular morphology and distribution of microglial cells in the superficial retina of UPOAO-3d (H) and 7d (I). And cell counting of microglial cells in the superficial retina of UPOAO-3d (J) and 7d (K). Data points are from retinal sections of four animals. Data were presented as means ± s.e.m, *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001, t- test. Scale bar = 50 μm.

Transcriptomic results comparison between UPOAO model, HIOP model and UCCAO model.

(A) Schematic illustration of HIOP model. RNA-seq for HIOP retina samples at 7 days. (B) GO analysis of differential genes in the HIOP model. (C) Venn diagram indicating the overlapping DEGs (146 genes) between HIOP and UPOAO and their respective remaining DEGs (both 593 genes in each group). (D) GO analysis of the overlapping DEGs between HIOP and UPOAO. (E) GO analysis of remaining DEGs in the HIOP_7d model except for overlapping DEGs. (F) GO analysis of remaining DEGs in the UPOAO_7d model. (G) Heatmap showing the inter-sample distribution of lipid and steroid-related DEGs from the analysis of remaining DEGs in the UPOAO model. (H) Schematic illustration of UCCAO model. (I) GO analysis of DEGs in the UCCAO model. (J) Hub genes in PPI network analysis of the DEGs in the UCCAO model were shown. (K) Venn diagram indicating the overlapping DEGs between UCCAO and UPOAO (15 genes). (L) Venn diagram indicated there were two overlapping DEGs between UCCAO DEGs and m6A related genes.

The characteristic features of UPOAO model in ischemia-reperfusion periods.