Fig 2: Regulation of inflammatory cytokine expression by ADAMTS13 in human retinal microvascular endothelial cells. Endothelial cultures were left untreated (control) or were treated with 25 mM of high glucose (HG), 300 µM of cobalt chloride (CoCl2), or 1 ng/mL of tumor necrosis factor-α (TNF-α) for 24 h with or without a 1 h preincubation with 100 ng/mL of ADAMTS13. A total of 25 mM of mannitol was used as control for cell damage induced by high glucose levels. In panel (A), signaling events were probed by measurement of phospho-ERK1/2 levels in cell lysates. In panel (B), cell culture medium levels of monocyte chemotactic protein-1 (MCP-1) and, in panel (C), cell culture medium levels of fractalkine were quantified with the use of specific ELISAs. Data are expressed as means ± standard deviation from independent (n = 3) experiments with triplicates per experiment. Statistical comparisons were performed as described in Section 2.12. * p < 0.05 indicates comparisons with values obtained from control cells. # p < 0.05 indicates comparisons with values obtained from stimulated cells.

Fig 3: Regulation of proangiogenic and inflammatory molecule expression by ADAMTS13 in human retinal Müller glial cells. Human glial cells were left untreated or treated with high glucose (HG) (25 mM), tumor necrosis factor-α (TNF-α) (1 ng/mL), or cobalt chloride (CoCl2) (300 µM) for 24 h or ADAMTS13 (100 ng/mL) for 1h followed by HG, CoCl2, or TNF-α. A total of 25 mM of mannitol was used as an inert control for osmotic effects by HG treatment. Levels of monocyte chemotactic protein-1 (MCP-1), matrix metalloproteinase-9 (MMP-9), and vascular endothelial growth factor (VEGF) were quantified in the culture media by ELISA. The present data were generated from three different experiments, each performed in triplicates, and the results are provided as means ± standard deviation; statistical comparisons were performed as described in Section 2.12. * p < 0.05 indicates the comparisons with values obtained from control cells. # p < 0.05 documents the differences with values obtained from cells treated with HG, TNF-α, or CoCl2.

Fig 4: Intravitreal administration of ADAMTS13 reduces retinal inflammation in diabetic rats. The relative protein amounts of phospho-ERK1/2 (A), high-mobility group box-1 (HMGB1) (B), vascular cell adhesion molecule-1 (VCAM-1) (C), and intercellular adhesion molecule-1 (ICAM-1) (D) were determined in rat retinas with the use of western blots. The animals were made diabetic with the use of a single streptozotocin bolus, ADAMTS13 was injected in the vitreous, and its effects on inflammation markers were evaluated by comparison of ADAMTS13-injected with the contralateral PBS-injected eyes in single animals. Statistical comparisons (mean standard deviation of 8–10 rats in each group) were performed as described in Section 2.12. * p < 0.05 compared with nondiabetic controls. # p < 0.05 compared with diabetic rats.

Fig 5: Hemostasis biomarkers in vitreous fluid of diabetes patients. Detection of thrombotic markers in vitreous fluid of patients with proliferative diabetic retinopathy (PDR). Determination of the von Willebrand factor (VWF) (A), the platelet marker CD41 (B), and ADAMTS13 (C) levels in vitreous fluid samples. A total of 15 µL of vitreous fluid samples from 12 patients with PDR and from 12 nondiabetic patients with rhegmatogenous retinal detachment (RD) was subjected to gel electrophoresis and the presence of VWF, CD41, and ADAMTS13 (5C11 monoclonal antibody) was illustrated by representative western blots and the levels of the antigens compared between the RD and PDR cohorts. Results are expressed as medians (interquartile range). (* p < 0.05; Mann-Whitney test).