Background: The oxygen induced retinopathy rodent model is widely used, notably for the assessment of developmental dystrophies in preclinical studies of vascular retinal diseases. Typically, the quantification of vessel tufts and avascular regions is computed manually from flat mounted retinas imaged using fluorescent probes that highlight the vascular network. However, such manual measurements are time-consuming and hampered by user variability and bias, thus a rapid and objective alternative is required.

Methods: We employ a machine learning approach to segment and characterize vascular tufts. The proposed quantitative retinal vascular assessment (QuRVA) technique uses quadratic discrimination analysis and morphological techniques to provide reliable measurements of vascular density and pathological vascular tuft regions, devoid of user intervention within seconds. Our algorithms allow also delineating the whole vasculature network, and identifying and analyzing avascular regions.

Results: Our first experiment shows the high degree of error and variability of manual segmentations. In consequence, we developed a set of algorithms to perform this task automatically. We benchmark and validate the results of our analysis pipeline using the consensus of several manually curated segmentations using commonly used computer tools. We describe the method, provide details for reproducing the algorithm, and validate all aspects of the analysis.

Conclusions: Manual and semi-automated procedures for tuft detection present strong fluctuations among users, demonstrating the need for fast and unbiased tools in this highly active research field with tremendous implications for basic research and industry.

Background: The oxygen induced retinopathy rodent model is widely used, notably for the assessment of developmental dystrophies in preclinical studies of vascular retinal diseases. Typically, the quantification of vessel tufts and avascular regions is computed manually from flat mounted retinas imaged using fluorescent probes that highlight the vascular network. However, such manual measurements are time-consuming and hampered by user variability and bias, thus a rapid and objective alternative is required.

Methods: We employ a machine learning approach to segment and characterize vascular tufts. The proposed quantitative retinal vascular assessment (QuRVA) technique uses quadratic discrimination analysis and morphological techniques to provide reliable measurements of vascular density and pathological vascular tuft regions, devoid of user intervention within seconds. Our algorithms allow also delineating the whole vasculature network, and identifying and analyzing avascular regions.

Results: Our first experiment shows the high degree of error and variability of manual segmentations. In consequence, we developed a set of algorithms to perform this task automatically. We benchmark and validate the results of our analysis pipeline using the consensus of several manually curated segmentations using commonly used computer tools. We describe the method, provide details for reproducing the algorithm, and validate all aspects of the analysis.

Conclusions: Manual and semi-automated procedures for tuft detection present strong fluctuations among users, demonstrating the need for fast and unbiased tools in this highly active research field with tremendous implications for basic research and industry.

Background: Retinol dehydrogenase 8 (RDH8) is a 312-amino acid (aa) protein involved in the visual cycle. Bound to the outer segment disk membranes of photoreceptors, it reduces all-trans-retinal to all-trans-retinol1 as one of the rate-limiting steps of the visual cycle2. RDH8 is a member of the short-chain dehydrogenase/reductase family. Its C-terminal segment allows its membrane-anchoring through the postulated presence of an amphipathic α-helix and of 1 to 3 acyl groups at positions 299, 302 and 3043. The secondary structure and membrane binding characteristics of RDH8 and its C-terminal segment have not yet been described.

Methods: To evaluate the membrane binding of RDH8, the full-length protein (aa 1–312), a truncated form (aa 1–296), its C-terminal segment (aa 281–312 and 297–312) as well as different additional variants of this segment were used. The truncated protein binds membranes less efficiently than the full-length form. Thus, the C-terminal segment of RDH8 is essential for the binding and has thus been further examined. The intrinsic fluorescence of tryptophan residues at positions 289 and 310 of the wild-type C-terminal segment of RDH8 and the mutants W289F, W310F and W310R have thus been used to determine their extent of binding to lipid vesicles and to monitor their local environment. Unilamellar lipid vesicles composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) or a mixture of POPC and 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) were used to mimic the phospholipid content of the outer segment disk membranes of photoreceptors.

Results: An increase in fluorescence intensity and in fluorescence lifetime is observed upon increasing the concentration of lipid vesicles. These data allowed calculating values of partition coefficient of the C-terminal segment of RDH8 varying between Kp =1.1 E6 to 1.7 E6. It is noteworthy that the observation of a more intense shift to lower wavelengths upon membrane binding of the mutant W310R and W310F indicates a deeper incorporation of the remaining tryptophan residue at position 289 into the lipid bilayer. The secondary structure of the C-terminal segment of RDH8 observed by circular dichroism and infrared spectroscopy shows a superposition of α-helical, β-turn and unordered structures.

Conclusions: The peptides derived from the C-terminal segment of RDH8 show a strong binding to lipid vesicles. These strength of binding is independent of the type of lipid and the presence of a mutation.

Background: Retinol dehydrogenase 8 (RDH8) is a 312-amino acid (aa) protein involved in the visual cycle. Bound to the outer segment disk membranes of photoreceptors, it reduces all-trans-retinal to all-trans-retinol1 as one of the rate-limiting steps of the visual cycle2. RDH8 is a member of the short-chain dehydrogenase/reductase family. Its C-terminal segment allows its membrane-anchoring through the postulated presence of an amphipathic α-helix and of 1 to 3 acyl groups at positions 299, 302 and 3043. The secondary structure and membrane binding characteristics of RDH8 and its C-terminal segment have not yet been described.

Methods: To evaluate the membrane binding of RDH8, the full-length protein (aa 1–312), a truncated form (aa 1–296), its C-terminal segment (aa 281–312 and 297–312) as well as different additional variants of this segment were used. The truncated protein binds membranes less efficiently than the full-length form. Thus, the C-terminal segment of RDH8 is essential for the binding and has thus been further examined. The intrinsic fluorescence of tryptophan residues at positions 289 and 310 of the wild-type C-terminal segment of RDH8 and the mutants W289F, W310F and W310R have thus been used to determine their extent of binding to lipid vesicles and to monitor their local environment. Unilamellar lipid vesicles composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) or a mixture of POPC and 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) were used to mimic the phospholipid content of the outer segment disk membranes of photoreceptors.

Results: An increase in fluorescence intensity and in fluorescence lifetime is observed upon increasing the concentration of lipid vesicles. These data allowed calculating values of partition coefficient of the C-terminal segment of RDH8 varying between Kp =1.1 E6 to 1.7 E6. It is noteworthy that the observation of a more intense shift to lower wavelengths upon membrane binding of the mutant W310R and W310F indicates a deeper incorporation of the remaining tryptophan residue at position 289 into the lipid bilayer. The secondary structure of the C-terminal segment of RDH8 observed by circular dichroism and infrared spectroscopy shows a superposition of α-helical, β-turn and unordered structures.

Conclusions: The peptides derived from the C-terminal segment of RDH8 show a strong binding to lipid vesicles. These strength of binding is independent of the type of lipid and the presence of a mutation.

Background: The neovascular aged-related macular degeneration (AMD) is the leading cause of legal blindness in the elderly. It is presently treated by anti-VEGF intravitreal injection in order to stop the neovascularization. In seeking of more efficient treatments to prevent retinal damage, it has been proposed that the kinin-kallikrein system (KKS), a key player in inflammation, could be involved in AMD etiology. However, the role of kinin receptors and their interaction with VEGF in AMD is poorly understood.

Methods: In order to address this question, choroidal neovascularization (CNV) was induced in the left eye of Long-Evans rat. After laser induction, anti-VEGF or IgG control were injected into the vitreal cavity. Gene expression was measured by qRT-PCR, retinal adherent leukocytes were labelled with FITC-Concanavalin A lectin, vascular leakage by the method of Evans blue and cellular localisation by immunohistochemistry.

Results: The number of labelled adherent leucocytes was significantly increased in laser-induced CNV compared to the control eye. This was significantly reversed by one single injection of anti-VEGF. Extravasation of Evans blue dye was significantly increased in laser-induced CNV eyes compared to control eyes and partially reversed by one single injection of anti-VEGF or by R954 treatment. The mRNA expression of inflammatory mediators was significantly increased in the retina of CNV rats. Immunodetection of B1R was significantly increased in CNV eyes. B1R immunolabeling was detected on endothelial and ganglion cells.

Conclusions: This study is the first to highlight an effect of the kinin/kallikrein system in a model of CNV that could be reduced by both anti-VEGF therapy and topically administered B1R antagonist R-954.

Background: The neovascular aged-related macular degeneration (AMD) is the leading cause of legal blindness in the elderly. It is presently treated by anti-VEGF intravitreal injection in order to stop the neovascularization. In seeking of more efficient treatments to prevent retinal damage, it has been proposed that the kinin-kallikrein system (KKS), a key player in inflammation, could be involved in AMD etiology. However, the role of kinin receptors and their interaction with VEGF in AMD is poorly understood.

Methods: In order to address this question, choroidal neovascularization (CNV) was induced in the left eye of Long-Evans rat. After laser induction, anti-VEGF or IgG control were injected into the vitreal cavity. Gene expression was measured by qRT-PCR, retinal adherent leukocytes were labelled with FITC-Concanavalin A lectin, vascular leakage by the method of Evans blue and cellular localisation by immunohistochemistry.

Results: The number of labelled adherent leucocytes was significantly increased in laser-induced CNV compared to the control eye. This was significantly reversed by one single injection of anti-VEGF. Extravasation of Evans blue dye was significantly increased in laser-induced CNV eyes compared to control eyes and partially reversed by one single injection of anti-VEGF or by R954 treatment. The mRNA expression of inflammatory mediators was significantly increased in the retina of CNV rats. Immunodetection of B1R was significantly increased in CNV eyes. B1R immunolabeling was detected on endothelial and ganglion cells.

Conclusions: This study is the first to highlight an effect of the kinin/kallikrein system in a model of CNV that could be reduced by both anti-VEGF therapy and topically administered B1R antagonist R-954.

Background: Retinal endothelial cells are very active and contribute to the integrity of the neurovascular unit. Vascular dysfunction has been proposed to contribute to the pathogenesis of glaucoma. Here, we evaluated the hypothesis that ocular hypertension triggers mitochondrial alterations in endothelial cells impairing the integrity of the blood retinal barrier (BRB).

Methods: Ocular hypertension was induced by injection of magnetic microbeads into the anterior chamber of EndoMito-EGFP mice, a strain expressing green fluorescent protein selectively in the mitochondria of endothelial cells. Capillary density, mitochondrial volume, and the number of mitochondrial components were quantified in 3D-reconstructed images from whole-mounted retinas using Imaris software. Dynamin-related protein (DRP-1), mitofusin-2 (MFN-2) and optic atrophy-1 (OPA-1) expression were assessed by western blot analysis of enriched endothelial cells. Mitochondrial structure was evaluated by transmission electron microscopy (TEM) and oxygen consumption rate was monitored by Seahorse analysis. The integrity of the BRB was evaluated by quantifying Evans blue leakage.

Results: Our data demonstrate that two and three weeks after ocular hypertension induction, the total mitochondria volume in endothelial cells decreased from 0.140±0.002 μm³ from non-injured retinas to 0.108±0.005 and 0.093±0.007 μm³, respectively in glaucomatous eyes (mean ± S.E.M, ANOVA, P<0.001; N=6/group). Frequency distribution showed a substantial increase of smaller mitochondria complexes (<0.5 μm³) in endothelial cells from glaucomatous retinas. Significant upregulation of DRP-1 was found in vessels isolated from glaucomatous retinas compared to the intact retinas, while MFN-2 and OPA-1 expression was not affected. Structural alteration in endothelial cell mitochondria was confirmed by TEM, which were accompanied by a 1.93-fold reduction in the oxygen consumption rate as well as 2.6-fold increase in vasculature leakage in glaucomatous retinas (n=3–6/group). In addition, this model did not trigger changes in the density of the vascular network, suggesting that mitochondrial fragmentation was not due to endothelial cell loss.

Conclusions: This study shows that ocular hypertension leads to early alterations in the dynamic of endothelial cell mitochondria, contributing to vascular dysfunction in glaucoma.

Background: Retinal endothelial cells are very active and contribute to the integrity of the neurovascular unit. Vascular dysfunction has been proposed to contribute to the pathogenesis of glaucoma. Here, we evaluated the hypothesis that ocular hypertension triggers mitochondrial alterations in endothelial cells impairing the integrity of the blood retinal barrier (BRB).

Methods: Ocular hypertension was induced by injection of magnetic microbeads into the anterior chamber of EndoMito-EGFP mice, a strain expressing green fluorescent protein selectively in the mitochondria of endothelial cells. Capillary density, mitochondrial volume, and the number of mitochondrial components were quantified in 3D-reconstructed images from whole-mounted retinas using Imaris software. Dynamin-related protein (DRP-1), mitofusin-2 (MFN-2) and optic atrophy-1 (OPA-1) expression were assessed by western blot analysis of enriched endothelial cells. Mitochondrial structure was evaluated by transmission electron microscopy (TEM) and oxygen consumption rate was monitored by Seahorse analysis. The integrity of the BRB was evaluated by quantifying Evans blue leakage.

Results: Our data demonstrate that two and three weeks after ocular hypertension induction, the total mitochondria volume in endothelial cells decreased from 0.140±0.002 μm³ from non-injured retinas to 0.108±0.005 and 0.093±0.007 μm³, respectively in glaucomatous eyes (mean ± S.E.M, ANOVA, P<0.001; N=6/group). Frequency distribution showed a substantial increase of smaller mitochondria complexes (<0.5 μm³) in endothelial cells from glaucomatous retinas. Significant upregulation of DRP-1 was found in vessels isolated from glaucomatous retinas compared to the intact retinas, while MFN-2 and OPA-1 expression was not affected. Structural alteration in endothelial cell mitochondria was confirmed by TEM, which were accompanied by a 1.93-fold reduction in the oxygen consumption rate as well as 2.6-fold increase in vasculature leakage in glaucomatous retinas (n=3–6/group). In addition, this model did not trigger changes in the density of the vascular network, suggesting that mitochondrial fragmentation was not due to endothelial cell loss.

Conclusions: This study shows that ocular hypertension leads to early alterations in the dynamic of endothelial cell mitochondria, contributing to vascular dysfunction in glaucoma.

Background: Retinopathy of prematurity (ROP) is the major cause of blindness in children, mainly caused by the retinal neovascularization (NV). Mounting of evidences shown that macrophage plays a pivotal role in the regulation of angiogenesis in ROP. Numerous studies confirmed that the deletion of macrophage significantly reduce the neovascularized areas in the oxygen-induced retinopathy (OIR) model. We have been studied the effect of lymphocyte derived-microparticles (LMPs) over ten years. LMPs are extracellular vesicles derived from apoptotic human CEM T lymphocytes. Our previous studies demonstrated that LMPs possess strong anti-angiogenic effect. Recently we observed that LMPs are capable to switch the phenotype of macrophage, thus to suppress the choroidal neovascularization (CNV). However, the role of LMPs on macrophage in ROP has not been clarified. Thus, my project is to disclose the relationship between LMPs and macrophage in ROP using the OIR model. Hypothesis: LMPs may inhibit retinal NV in the OIR model through targeting at macrophage by affecting the migration of macrophage, thus to inhibit pathological angiogenesis in ROP.

Methods: Cell culture [RAW 264.7 and bone marrow-derived macrophage (BMDM)] for cell migration and viability assay. Generate the OIR model for in vivo detection of macrophage recruitment. Quantification of retinal NV, immunohistostaining of the macrophage in vivo, ex vivo retinal explants for cell migration and qPCR.

Results: LMPs do not affect RAW 264.7 and BMDM cell viability (P>0.05). LMPs significantly decrease the BMDM cell migration indirectly (P<0.05). I successfully generate the OIR model and confirm that more macrophages infiltrate during retinal angiogenesis with counting the F4/80 immunostaining in the retinal flat mount. LMPs exert inhibiting effect on retinal angiogenesis through decreasing the migration of macrophages in vivo.

Conclusions: LMPs have the negative effect on retinal angiogenesis via reducing the infiltrated macrophages to the neovascularized areas in the OIR model.

Background: Retinopathy of prematurity (ROP) is the major cause of blindness in children, mainly caused by the retinal neovascularization (NV). Mounting of evidences shown that macrophage plays a pivotal role in the regulation of angiogenesis in ROP. Numerous studies confirmed that the deletion of macrophage significantly reduce the neovascularized areas in the oxygen-induced retinopathy (OIR) model. We have been studied the effect of lymphocyte derived-microparticles (LMPs) over ten years. LMPs are extracellular vesicles derived from apoptotic human CEM T lymphocytes. Our previous studies demonstrated that LMPs possess strong anti-angiogenic effect. Recently we observed that LMPs are capable to switch the phenotype of macrophage, thus to suppress the choroidal neovascularization (CNV). However, the role of LMPs on macrophage in ROP has not been clarified. Thus, my project is to disclose the relationship between LMPs and macrophage in ROP using the OIR model. Hypothesis: LMPs may inhibit retinal NV in the OIR model through targeting at macrophage by affecting the migration of macrophage, thus to inhibit pathological angiogenesis in ROP.

Methods: Cell culture [RAW 264.7 and bone marrow-derived macrophage (BMDM)] for cell migration and viability assay. Generate the OIR model for in vivo detection of macrophage recruitment. Quantification of retinal NV, immunohistostaining of the macrophage in vivo, ex vivo retinal explants for cell migration and qPCR.

Results: LMPs do not affect RAW 264.7 and BMDM cell viability (P>0.05). LMPs significantly decrease the BMDM cell migration indirectly (P<0.05). I successfully generate the OIR model and confirm that more macrophages infiltrate during retinal angiogenesis with counting the F4/80 immunostaining in the retinal flat mount. LMPs exert inhibiting effect on retinal angiogenesis through decreasing the migration of macrophages in vivo.

Conclusions: LMPs have the negative effect on retinal angiogenesis via reducing the infiltrated macrophages to the neovascularized areas in the OIR model.