Minocycline Reduces Proinflammatory Cytokine Expression, Microglial Activation, and Caspase-3 Activation in a Rodent Model of Diabetic Retinopathy J. Kyle Krady, 1 Anirban Basu, 1 Colleen M. Allen, 1 Yuping Xu, 2 Kathryn F. LaNoue, 2 Thomas W. Gardner, 2 and Steven W. Levison 1,3 Diabetes leads to vascular leakage, glial dysfunction, and neuronal apoptosis within the retina. The goal of the studies reported here was to determine the role that retinal microglial cells play in diabetic retinopathy and assess whether minocycline can decrease microglial ac- tivation and alleviate retinal complications. Immunohis- tochemical analyses showed that retinal microglia are activated early in diabetes. Furthermore, mRNAs for interleukin-1 and tumor necrosis factor-, proinflam- matory mediators known to be released from microglia, are also increased in the retina early in the course of diabetes. Using an in vitro bioassay, we demonstrated that cytokine-activated microglia release cytotoxins that kill retinal neurons. Furthermore, we showed that neu- ronal apoptosis is increased in the diabetic retina, as measured by caspase-3 activity. Minocycline represses diabetes-induced inflammatory cytokine production, re- duces the release of cytotoxins from activated microg- lia, and significantly reduces measurable caspase-3 activity within the retina. These results indicate that inhibiting microglial activity may be an important strat- egy in the treatment of diabetic retinopathy and that drugs such as minocycline hold promise in delaying or preventing the loss of vision associated with this dis- ease. Diabetes 54:1559 –1565, 2005 D iabetic retinopathy is a progressive neurologi- cal disease characterized by degeneration of neurons and macroglia and accompanied by profound vascular changes that eventually lead to legal blindness. Despite progress in understanding the pathogenesis of diabetic retinopathy, our knowledge of the mechanisms leading to neuronal cell loss, glial dys- function, and vascular remodeling is incomplete. Whereas the retina has traditionally been viewed as an immune privileged tissue, evidence is accumulating to support a role for local inflammation in the pathogenesis of diabetic retinopathy. For example, alterations in the level of adhe- sion molecules, which facilitate the trafficking of leuko- cytes into the retina, accompany changes in retinal blood vessel permeability (1). Joussen et al. (2) identified tumor necrosis factor- (TNF-) and cyclo-oxygenase-2 (COX-2), two proinflammatory mediators, as operative in the early signature pathologies of diabetic retinopathy. In addition, Funatsu et al. (3) documented local production of inter- leukin (IL)-6 in the vitreous humor of diabetic patients and showed that levels of IL-6 correlate with the severity of retinal pathophysiology. Elevated levels of cytokines can activate the endoge- nous immune cells of the central nervous system (CNS), cells known as microglia. Microglia are related to dendritic cells in other tissues and, although similar to macro- phages, are a distinct cell type (4). In response to an activating stimulus, quiescent microglia undergo a series of stereotyped morphological, phenotypical, and func- tional changes (5). In response to signals from dying cells, activated microglia evolve into phagocytes capable of clearing debris (6). It is yet to be determined what signals during diabetes activate the retinal microglia, whether microglial activation precedes the major histopathologic changes that occur in the retina, and what affect these microglia may play in the pathophysiology of the disease. In this study, we examined the putative role of microglia and inflammation in the progression of early diabetic retinopathy and tested the ability of the drug, minocycline, to abrogate some of the negative consequences of microg- lial activation. Minocycline is a second-generation, semi- synthetic tetracycline that exerts anti-inflammatory effects that are completely separate from its antimicrobial ac- tions. Studies suggest that it affords neuroprotection because it crosses the blood-brain barrier, inhibits the proliferation and activation of microglia, and inhibits apo- ptosis (7). It has proven effective in a number of animal models of neurodegeneration in which microglia and in- flammation have been implicated, but it has yet to be tested in animal models of diabetes. Because minocycline is well tolerated, it represents a potential new therapeutic From the 1 Department of Neural and Behavioral Sciences, The Pennsylvania State University College of Medicine, Hershey, Pennsylvania; the 2 Department of Cellular and Molecular Physiology, The Pennsylvania State University College of Medicine, Hershey, Pennsylvania; and the 3 Department of Neurol- ogy and Neuroscience, University of Medicine and Dentistry in New Jersey, New Jersey Medical School, Newark, New Jersey. Address correspondence and reprint requests to J. Kyle Krady, PhD, Dept. of Neural and Behavioral Sciences, H109, The Pennsylvania State University College of Medicine, Hershey, PA 17033. E-mail: jkk7@psu.edu. Received for publication 3 November 2004 and accepted in revised form 27 January 2005. AMC, 7-amino-4-methyl coumarin; CNS, central nervous system; COX-2, cyclo-oxygenase-2; DAPI, 4',6-diamidino-2-phenylindole; IL, interleukin; LDH, lactate dehydrogenase; MEM, minimum essential medium; MMP, matrix metalloproteinases; NCS, newborn calf serum; ROS, reactive oxygen species; STZ, streptozotocin; TNF-, tumor necrosis factor-; TUNEL, transferase-medi- ated dUTP nick-end labeling; VEGF, vascular endothelial growth factor. © 2005 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. DIABETES, VOL. 54, MAY 2005 1559