Incipient diabetic retinopathy is characterized by increased vascular permeability and progressive vascular occlusion. Pericyte loss precedes capillary occlusion in the diabetic retina, but its cause remains unclear. One concept proposes that pericyte loss is the result of toxic product accumulation and induction of destructive cellular signals generated within the pericyte. Alternatively, new experimental data indicate that pericyte dropout may result from regulations which induce pericyte elimination as an active process. Pericytes are critical for the development of a proper retinal network, and appear protective for endothelial cells under hyperglycemic conditions. The unifying hypothesis of hyperglycemia-induced microvascular damage centers around hyperglycemia-induced mitochondrial overproduction of reactive oxygen species. The pharmacological prevention of acellular capillaries without the rescue of pericyte loss in experimental diabetic retinopathy suggests that the endothelium is the primary therapeutic target.

Our data indicate that the oxidatively formed CML is present in non-diabetic retinas as a regular constituent but increases in diabetes both in neuroglial and vascular components. Imidazolone-type AGE are restricted to microvessels and spread during later stages over the entire retina, co-localizing with the expression of AGE-receptor.

A linear relation exists between time-averaged blood glucose concentrations and the development and progression of microvascular complications in diabetic patients [1]. Some individual patients, however, with low average blood glucose concentrations develop severe complications, while some with high mean blood glucose concentrations do not. Overall, only one-third of Type I diabetic patients develop clinical nephropathy, usually after 12 to 15 years of Diabetologia (1999) Abstract Aims/hypothesis. We investigated whether either the amount of diabetes-induced intracellular oxidative stress or the concentration of hyperglycaemia-induced advanced glycation endproducts is associated with the risk of diabetic retinopathy. Methods. We measured concentrations of the glycoxidation product N e -(carboxymethyl)lysine and two non-oxidation-dependent advanced glycation endproducts (methylglyoxal-derived and 3-deoxyglucosone-derived) in CD45RA + T-cells from 21 Type I (insulin-dependent) diabetic patients with and without diabetic retinopathy and from age-matched nondiabetic control subjects. Results. Intracellular concentrations of both oxidation-dependent N e -(carboxymethyl)lysine and oxidation-independent advanced glycation endproducts were increased in memory T-cells from diabetic patients. N e -(carboxymethyl)lysine: diabetic median-24 176 arbitrary units/mg protein (95 % confidence interval 18 690±34 099 arbitrary units/mg protein); nondiabetic-9088 arbitrary units/mg protein (confidence interval 6994±10 696 arbitrary units/mg protein; p < 0.0001). Methylglyoxal-derived advanced glycation end products: diabetic-5430 arbitrary units/ mg protein (confidence interval 3458±13 610); nondiabetic-271 arbitrary units/mg protein (confidence interval 61±760 arbitrary units/mg protein; p < 0.0001). 3-Deoxyglucosone-derived advanced glycation end products: diabetic-8070 arbitrary units/mg protein (confidence interval 7049±16 551 arbitrary units/mg protein); nondiabetic-1479 arbitrary units/ mg protein (confidence interval 1169±3170; p < 0.0001). Only N e -(carboxymethyl)lysine concentrations, however, inversely correlated with the duration of retinopathy-free diabetes (r = ±0.51; p < 0.02). Diabetes-dependent N e -(carboxymethyl)lysine accumulation did not correlate with age, diabetes duration, or averaged glycohaemoglobin concentrations. In vitro experiments wih menadione and lymphocytes confirmed that N e -(carboxymethyl)lysine concentrations reflect intracellular oxidative stress. Conclusion/interpretation. Monitoring intracellular concentrations of increased oxidative stress in longlived CD45RA + lymphocytes by markers such as N e -(carboxymethyl)lysine possibly identifies a subgroup of patients at high risk for microvascular complications. [Diabetologia (1999) 42: 603±607]

Proliferative diabetic retinopathy (PDR) and macular degeneration are the most common causes of blindness in industrialised nations. Retinal neovascularization is the main feature of PDR and the angiogenic activity produced by the retina seems to be diffusable because extensive capillary non-perfusion is associated with rubeosis iridis, thus producing angiogenesis as far anterior as the surface of the iris. Accordingly increased vitreal concentrations of angio- Diabetologia (2000) Abstract Aims/hypothesis. Proliferative diabetic retinopathy is a major debilitating disease causing most cases of blindness in humans in the Western world. Photocoagulation is the established therapy of proliferative diabetic retinopathy, although the molecular mechanisms of its effects are still not known. Recently angiostatin has been characterized as a potent inhibitor of neovascularization. Apart from a possible downregulation of angiogenic cytokines, release of angiostatin could initiate the anti-angiogenic effects of retinal photocoagulation. Methods. We investigated the regulation of angiostatin and the angiogenic cytokines vascular endothelial growth factor and basic fibroblast growth factor in vivo by comparing vitreal concentrations of 18 control patients and 34 patients with proliferative diabetic retinopathy with and without previous photocoagulation. Concentrations of basic fibroblast growth factor and angiostatin were additionally measured in serum, while vascular endothelial growth factor is known to be regulated locally in the eye. Cytokines were measured by immunological methods.