Mara Lorenzi

Pathogenesis of Diabetic Retinopathy

Diabetic retinopathy is the most common sequela of diabetes, developing in more than 90% of individuals who have had the disease for fifteen or more years. Sixty percent of such individuals will eventually develop the sight-threatening stage of the disease. The process that makes diabetic retinopathy sight-threatening is the progressive damage to the small vessels that feed the retina. The retinal capillaries become leaky, permitting the extravasation of fluid into the retinal tissue (retinal edema), and eventually become nonperfused and obliterated leading to retinal ischemia and proliferation of new vessels (proliferative retinopathy). We are interested in reconstructing the cellular and molecular events by which diabetes leads to retinal capillary damage, with the goal of identifying rational targets for adjunct therapy. We previously observed that the retinal capillary cells--endothelial cells and pericytes--of individuals or experimental animals with diabetes die at an accelerated rate by a form of suicidal death termed apoptosis. Because such accelerated death can readily explain most of the capillary lesions occurring in diabetic retinopathy, our work targeted the mechanisms responsible for apoptosis and, more generally, for capillary dysfunction and obliteration.

Pericytes are cells that provide support to the delicate capillaries of the retina. They die quite early in diabetes, and we identified several players in the sequence of events that leads to their early death. In both diabetic patients and experimentally diabetic rats, we found that retinal pericytes activated a transcription factor named NF-kB, which is triggered by cellular stress and controls the expression of several proteins. Among these, are proteins of the Bcl-2 family that regulate apoptosis. The pro-survival family members can inhibit apoptosis induced by a variety of cytotoxic insults, and the pro-apoptotic members mostly work by antagonizing the effects of the pro-survival members although the Bax group can also kill directly by damaging mitochondria. In diabetes, retinal pericytes with activated NF-kB showed intense expression of proapoptotic Bax and the nuclear changes indicative of apoptosis. Using pericytes cultured in vitro we determined that the high glucose levels characteristic of diabetes were sufficient to induce the sequence of events that are observed in the diabetic retina in vivo, and that inhibition of NF-kB activation prevented the increase in Bax and pericyte death. Thus, one of the pathways to the early demise of retinal pericytes in diabetes begins with the stress caused by high glucose which activates NF-kB, in turn leading to increased expression of Bax, and apoptosis. Studies are in progress to ascertain whether preventing the apoptotic action of Bax, and thus the death of pericytes, will have a favorable influence on the other abnormalities of retinal vessels in diabetes.

It was surprising and intriguing that endothelial cells, the other type of cells present in retinal vessels, did not show NF-kB activation or Bax overexpression in diabetes. This suggested that the mechanisms leading to accelerated death of endothelial cells in diabetes may differ from those operative in the death of pericytes. On the basis of the clinical observation that the lack of capillary perfusion in the diabetic retina is occasionally transient, we examined the possibility that the apoptotic death of endothelial cells may be secondary to reversible occlusive phenomena. Drs. Daria Boeri and Michele Maiello, Visiting Scientists from Italy, set up in our laboratory methods to detect the presence of even minuscule thrombi in the retinal capillaries, obtain an accurate count of such microthrombi, measure their extension, and image their topographical relationship with apoptotic endothelial cells. The occurrence of retinal microthrombosis in human diabetes has long been suspected but never systematically demonstrated. We documented that diabetic individuals manifested an increased prevalence of platelet-fibrin thrombi in retinal capillaries, and that the microthrombi were often associated with apoptotic endothelial cells. Because platelet-fibrin microthrombosis can be a rational target of antiplatelet drugs such as aspirin, the findings have immediate clinical implications. They have received widespread attention, and commentaries have appeared in the lay press, from The New York Times, to the Washington Post as well as in ophthalmology- and diabetes-related publications. We now need to learn how early after the onset of diabetes retinal microthromboses begin to occur, and we are thus planning a clinical trial in which we will test antiplatelet drugs on early functional abnormalities of retinal vessels in young patients with recent onset of diabetes.

The mechanisms for increased retinal microthrombosis in diabetes are almost certainly multiple, and some are likely to stem from altered properties of the endothelial cells. We recently began to investigate in the diabetic retina the regulation of complement, because the cascade of events involved in complement activation can alter the properties of endothelial cells causing both increased permeability and microthrombosis, the vascular pathology that we encounter in the diabetic retina. Complement is normally activated as a defense against microbes and other pathogenic invaders, but there are circumstances in which activation occurs because of decreased availability of inhibitors of the activation process. We found that diabetes reduces the availability of complement inhibitors on the endothelial cells of retinal vessels, and that in these vessels there is complement activation. The goal is now to determine the relationship between complement activation, microthrombosis, and endothelial cell apoptosis in diabetic retinal vessels.