Objective Glucosamine, an intermetabolite of the hexosamine biosynthesis pathway (HBP), is a widely used nutritional supplement in osteoarthritis patients, a subset of whom also suffer from diabetes. HBP is activated in diabetic retinopathy (DR). The aim of this study is to investigate the yet unclear effects of glucosamine on DR. Methods In this study, we tested the effect of glucosamine on vascular and neuronal pathology in a mouse model of streptozotocin-induced DR in vivo and on cultured endothelial and Müller cells to elucidate the underlying mechanisms of action in vitro . Results Glucosamine did not alter the blood glucose or HbA 1c levels in the animals, but induced body weight gain in the non-diabetic animals. Interestingly, the impaired neuronal function in diabetic animals could be prevented by glucosamine treatment. Correspondingly, the activation of Müller cells was prevented in the retina as well as in cell culture. Conversely, glucosamine administration in the normal retina damaged the retinal vasculature by increasing pericyte loss and acellular capillary formation, likely by interfering with endothelial survival signals as seen in vitro in cultured endothelial cells. Nevertheless, under diabetic conditions, no further increase in the detrimental effects were observed. Conclusions In conclusion, the effects of glucosamine supplementation in the retina appear to be a double-edged sword: neuronal protection in the diabetic retina and vascular damage in the normal retina. Thus, glucosamine supplementation in osteoarthritis patients with or without diabetes should be taken with care.

C. elegans are used to study molecular pathways, linking high glucose levels (HG) to diabetic complications. Persistent exposure of C. elegans to a HG environment induces the mitochondrial formation of reactive oxygen species (ROS) and advanced glycation endproducts (AGEs), leading to neuronal damage and decreased lifespan. Studies suggest that transient high glucose exposure (TGE) exerts different effects than persistent exposure. Thus, the effects of TGE on ROS, AGE-formation and life span were studied in C. elegans. Four-day TGE (400 mM) as compared to controls (0mM) showed a persistent increase of ROS (4-days 286 ± 40 RLUs vs. control 187 ± 23 RLUs) without increased formation of AGEs. TGE increased body motility (1-day 0.14 ± 0.02; 4-days 0.15 ± 0.01; 6-days 0.16 ± 0.02 vs. control 0.10 ± 0.02 in mm/s), and bending angle (1-day 17.7 ± 1.55; 3-days 18.7 ± 1.39; 6-days 20.3 ± 0.61 vs. control 15.3 ± 1.63 in degree/s) as signs of neuronal damage. Lifespan was increased by 27% (21 ± 2.4 days) after one-day TGE, 34% (22 ± 1.2 days) after four-days TGE, and 26% (21 ± 1.4 days) after six-days TGE vs. control (16 ± 1.3 days). These experiments suggest that TGE in C. elegans has positive effects on life span and neuronal function, associated with mildly increased ROS-formation. From the perspective of metabolic memory, hormetic effects outweighed the detrimental effects of a HG environment.