- Email: [email protected]
WOR Diabetic Rat
Free Radical Biology & Medicine, Vol. 24, No. 1, pp. 111–120, 1998 Copyright q 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/98 $19.00 / .00
PII S0891-5849(97)00202-5
Original Contribution INCREASED NADH OXIDASE ACTIVITY IN THE RETINA OF THE BBZ/WOR DIABETIC RAT E. ANN ELLIS,* MARIA B. GRANT,* FREDERICK T. MURRAY, † MARTHA B. WACHOWSKI, † DENNIS L. GUBERSKI, ‡ PAUL S. KUBILIS,§ and GERARD A. LUTTY \ * Department of Medicine, College of Medicine, University of Florida, Gainesville, FL, USA;† Pharmacia & Upjohn, Kalamazoo, MI, ‡ Department of Pathology, University of Massachusetts School of Medicine, Worcester, MA, USA; § Division of Biostatistics, University of Florida, Gainesville, FL, USA; and \ Department of Ophthalmology, The Johns Hopkins School of Medicine, Baltimore, MD, USA (Received 1 July 1996; Revised 8 April 1997; Accepted 12 May 1997)
Abstract—This morphological study demonstrates a role for endothelial cells in generating reactive oxygen species in early stages of retinopathy in the BBZ / Wor rat, an obese, noninsulin dependent model of diabetes. Hyperglycemia induced pseudohypoxia results in an imbalance in cytosolic NADH / NAD/. In the oxygen-rich environment of the retina, NADH oxidase generates superoxide radical which is dismutated to hydrogen peroxide. Localization of hydrogen peroxide by the cerium NADH oxidase enzyme activity cytochemical localization technique shows a statistically significant increase of peroxide localization in the central retina of diabetic rats as compared to age-matched, nondiabetic controls. Endothelial cell dysfunction, indicated by leakage of endogenous serum albumin, coincided with areas of NADH oxidase activity localization. In diabetic rats there are increased levels of fibronectin in areas of hydrogen peroxide localization. This in vivo , morphological study is the first demonstration of oxidative injury and endothelial cell dysfunction in the retina of a spontaneous, noninsulin dependent model of diabetes. q 1997 Elsevier Science Inc. Keywords—Oxidative injury, Endothelial dysfunction, Vascular leakage, Retinopathy, Diabetes, NADH oxidase
tomy.11 Endothelial cells may produce reactive oxygen species capable of tissue injury.12 Increased flux of glucose via the sorbitol pathway results in increased cytosolic NADH / NAD/ and increased superoxide production described as hyperglycemia induced pseudohypoxia.13 Autoxidation of glucose, a nonenzymatic process, increases with hyperglycemic pseudohypoxia and is a source of reactive oxygen species in diabetes.2 Any and all of the above mechanisms may play a role in oxidative injury. Free radicals or reactive oxygen species in cells and tissues have been demonstrated by a number of methods. Effects of antioxidants such as superoxide dismutase (SOD), glutathione peroxidase, catalase and vitamin E have been detected indirectly.14,15 Direct detection of reactive oxygen species has been difficult because reactive oxygen species are short lived and are catalyzed rapidly by scavenging systems. At the light microscope level, nitroblue tetrazolium has been used for histochemical demonstration of superoxide in the retina.16
INTRODUCTION
Hyperglycemia is the primary initiating factor for all types of diabetic vascular complications including diabetic retinopathy.1 – 3 Increased oxidative stress is observed in diabetes.4,5 Production of free radicals correlates to metabolic control and degree of hyperglycemia.6,7 Free radical mediated retinal vascular dysfunction may occur by several mechanisms. Activated polymorphonuclear leukocytes generate free radical derived oxidants.8 Activation of polymorphonuclear leukocytes occurs in diabetic patients, 9 in an alloxan-induced diabetic rat model, 10 and in a cat model of diabetes induced by partial pancreatecThis study was supported in part by NIH Grant EY07739, the American Heart Association and the Department of Health and Rehabilitative Services of the State of Florida for the University of Florida Diabetes Research, Education and Treatment Center. Address correspondence to: M. B. Grant, Box 100226, Gainesville, FL 32610-0226; E-Mail: [email protected] 111
/ 2b30 5134 Mp 111 Wednesday Nov 12 11:07 AM EL–FRB 5134
112
E. A. ELLIS et al.
The cerium NADH oxidase enzyme activity methodology introduced by Briggs and Karnovsky 17,18 has been used for cytochemical demonstration of hydrogen peroxide. This method has been used to detect reactive oxygen species in vivo in models of IgA damage to lungs, 19 in optic neuritis, 20 ischemia-reperfusion of the heart, 21 and during metastasis of tumor cells.22 In the cerium NADH oxidase enzyme activity localization technique, NADH oxidase enzyme activity, in the presence of oxygen and the substrate NADH, generates the reactive oxygen species superoxide (O2 i 0 ) from which other free radical derived oxidants such as hydrogen peroxide (H2O2 ) and hydroxyl radical (OH i 0 ) are generated. NADH oxidase
NADH / 2O2
2O2 i 0 / NAD / / H /
Superoxide dismutates, either spontaneously or catalyzed by superoxide dismutase, to yield hydrogen peroxide. SOD
2O2 i 0 / 2H/ r H2O2 / O2 In situ subcellular sites of generation of the free radical derived oxidant, hydrogen peroxide, can be shown by using cerium chloride in the reaction medium with the substrate NADH. Cerium ions react with hydrogen peroxide to produce an electron dense precipitate, cerium perhydroxide, Ce(OH)2OOH. H2O2 / CeCl 3 r Ce(OH)2OOH Vascular permeability, disruption of the blood retinal barrier, can be assessed by colloidal gold immunocytochemical localization of extravasation of endogenous serum proteins.23 Serum albumin or other endogenous serum proteins are retained in their normal location by fixation; immunocytochemical localization of these endogenous proteins can be done as a post embedding procedure on tissue sections. This study presents an investigation of oxidative injury and endothelial cell dysfunction in eyes of obese, noninsulin dependent diabetic BBZ/Wor rats.24 Free radical derived oxidant localization is documented by the cerium NADH oxidase enzyme activity cytochemical method and endothelial cell dysfunction is confirmed by colloidal gold immunocytochemical localization of extravasated serum albumin. In addition, fibronectin is elevated in and around vessels where oxidative injury has occurred in diabetic rats.
MATERIALS AND METHODS
Chemicals 3-Amino-1,2,4-triazole, cerium chloride, cold water fish gelatin, dimethylsulfoxide ( DMSO ) , fibronectin, glycine, rat serum albumin, reduced nicotinamide adenine dinucleotide ( NADH ) , maleic acid, periodic acid, Tris base, and Triton X-100 were purchased from Sigma Chemical Co. ( St. Louis, MO ) . Benzyldimethyl amine ( BDMA ) , colloidal gold labeled rabbit antigoat IgG antibodies, dodecenyl succinic anhydride ( DDSA ) , Eponate 12, osmium tetroxide, Quetol 651, and sodium cacodylate were purchased from Ted Pella Inc. (Redding, CA). Polyclonal goat anti-rat serum albumin antibodies, IgG fraction, were purchased from Organon Teknika-Cappel (Durham, NC). Polyclonal rabbit anti-human fibronectin antibodies, IgG fraction, were purchased from Dako Corporation (Carpinteria, CA). Colloidal gold labeled donkey anti-rabbit IgG antibodies were purchased from Jackson Immunoresearch Labs Inc. (West Grove, PA). Isolab Gly Affin GHb test kit was purchased from Isolab (Akron, OH). Sodium pentobarbital was purchased from Abbott Laboratories (Chicago, IL). Methoxyflurane was from Mallinkrodt Veterinary Inc. (Mundelein, IN). Animals Male BBZ/Wor and age matched, nondiabetic BB DR /Wor rats were maintained on Purina 5002 rodent chow and sterilized water ad lib in a viral free facility. All rats were obtained from the NIH facility at the University of Massachusetts and shipped in isolator rat boxes to the viral free facility at Pharmacia and Upjohn in Kalamazoo, MI. All BBZ/Wor rats were housed individually in micro-isolator rat boxes. BB DR /Wor rats were housed two to a micro-isolator rat box. Diabetic and nondiabetic control rats ranged from 7–12 months of age. Diabetes developed at 2–3 months of age and length of diabetes ranged from 4.5–10 months. Blood samples were obtained from tail veins under light methoxyflurane anesthesia. Whole blood glucose was measured using the Glucometer Elite Glucose Meter from Bayer Diagnostic Division (Tarrytown, NY). Total glycosylated hemoglobin was measured from whole blood removed from tail veins using the Isolab Gly Affin Test kit as previously described.25 Cytochemical localization of hydrogen peroxide Rats were euthanized by intraperitoneal injection with sodium pentobarbital at 50 mg/kg body weights. Eyes were enucleated and one eye of each animal was
/ 2b30 5134 Mp 112 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
processed as follows. The globe and attached optic nerve were dissected out and fixed for 1 h in cold 5.0% acrolein in 0.1 M sodium cacodylate-HCl buffers (pH 7.4). Specimens were washed overnight in cold 0.15 M sodium cacodylate-HCl buffers (pH 7.4) plus 5% sucrose and 1% dimethylsulfoxide (DMSO). Specimens were brought to room temperature in the final two buffers washes which contained 0.1 M glycine. Specimens were preincubated for 30 min at 377C with agitation in the following medium: 2.0 mM cerium chloride, 10 mM 3-amino-1,2,4-triazole, 0.1 M Trismaleate buffer (pH 7.5), 7% sucrose, and 0.0002% Triton X-100. They were then incubated for 1 h at 377C with agitation in the following complete reaction medium: 2 mM cerium chloride, 10 mM 3-amino-1,2,4triazole, 0.8 mM NADH, 0.1 M Tris-maleate buffer (pH 7.5), 7% sucrose and 0.0002% Triton-X 100. The reaction was terminated by washing in cold 0.1 M Trismaleate buffer (pH 7.5), 7% sucrose followed by a wash in cold 0.15 M sodium cacodylate-HCl buffer (pH 7.4), 7% sucrose. Control for the specificity of the reaction was done by reacting pieces of tissue with cerium chloride and aminotriazole in the absence of the substrate, NADH. Specimens were postfixed overnight in 1% osmium tetroxide, 0.1 M sodium cacodylate-HCl buffer (pH 7.4), 7% sucrose in the cold, dehydrated through an ethanol series (20%, 40%, 60%, 80%, 95% 1 2, 100% 1 2) to propylene oxide and infiltrated and embedded in epoxy resin (Quetol 651-1.15 g, Eponate 12-1.83 g, DDSA-7.02 g, BDMA-0.14ml). Grids were examined and photographed at 75 kV in the transmission electron microscope (TEM) without post staining.
Semiquantitative evaluation of peroxide localization by the cerium NADH oxidase enzyme activity method Semiquantitative evaluation of peroxide localization was done by a modification of the method of Briggs et al.26 Eyes were sectioned median longitudinally through the optic nerve head. These sections included the neural retina approximately 1 mm on either side of the optic nerve head. This area was designated as the area of central retina. Sections from the far peripheral retina were taken starting at the ora serrata and extending 1.5-2 mm toward the posterior of the eye and designated as the area of peripheral retina for comparison to the central area. All blood vessels in the neural retina which were not blocked by grid bars were examined at a magnification of 10,000 1 and scored either positive ( / ) or negative ( 0 ) for cerium perhydroxide precipitate. Results were expressed as the percentage of positive vessels relative to the total number of vessels ex-
113
amined for each grid. Seven diabetic eyes and five nondiabetic eyes were examined. Immunocytochemical localization of endogenous serum albumin Sections from eyes in which NADH oxidase enzyme activity had been localized were picked up on nickel grids, oxidized for 30 min with 1% periodic acid followed by 5 1 5 min washes in deionized water. Grids were floated 2 1 5 min on phosphate buffered saline (PBS) (pH 7.2), followed by 30 min on PBS blocker [PBS plus 2% nonfat dry milk and 2% cold water fish gelatin], and then reacted overnight at 47C in a moist chamber with polyclonal goat anti-rat serum albumin antibodies, IgG fraction, diluted 1:200 with PBS blocker. After 2 1 5 min washes with PBS blocker followed by 2 1 5 min washes with PBS, grids were washed with Tris-HCl saline buffer (pH 7.6) (TBS) followed by 30 min on TBS blocker [TBS plus 2% nonfat dry milk and 2% cold water fish gelatin]. Grids were then incubated for 1 h at room temperature in a moist chamber on drops of rabbit anti-goat IgG labeled with 20 nm colloidal gold diluted 1:40 with TBS blocker. Grids were washed with TBS blocker, followed by 2 1 5 min washes with TBS and then 3 1 5 min washes with deionized water. Control grids were incubated with only the gold labeled secondary antibody or with antibodies to rat serum albumin which had been absorbed overnight with an excess of rat serum albumin. Immunocytochemical localization of fibronectin Sections from eyes in which NADH oxidase enzyme activity had been localized were picked up on nickel grids. The immunocytochemical localization procedure was the same as that used for localization of serum albumin. Primary antibody was a polyclonal rabbit anti-human fibronectin diluted 1:100; secondary antibody was a donkey anti-rabbit IgG labeled with 18 nm colloidal gold diluted 1:40. Control grids were incubated with only the gold labeled secondary antibody or with antibodies to fibronectin which had been absorbed overnight with an excess of fibronectin. ADPase histochemical studies One eye from twelve 5-month duration diabetic BBZ/Wor animals and six age-matched nondiabetic BB DR /Wor rats were used for ADPase flat-embedded retina preparations. The retinas from these eyes were teased from the retinal pigment epithelium, fixed in 2% paraformaldehyde, washed, and processed for adeno-
/ 2b30 5134 Mp 113 Wednesday Nov 12 11:07 AM EL–FRB 5134
114
E. A. ELLIS et al.
Table 1. Levels of Blood Sugars and Glycosylated Hemoglobin
Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
Blood Sugars (mg/dl)
Total Glycosylated Hemoglobin (%)
496 { 15 173 { 3
13.0 { 0.4 5.3 { 0.5
sine diphosphatase (ADPase) histochemical staining of the retinal vasculature as previously described.27 Four radial cuts were made in the ADPase incubated retinas so that they could be postfixed flat and embedded in JB-4 plastic (Polysciences, Inc., Warrington, PA) as described in detail previously.27 The lead ADPase reaction product in retinal vessels was visualized in the flat-embedded retinas with dark field illumination. Changes in the vascular pattern and angiopathy were evaluated in the flat perspective and then serial 2.5 mm sections were cut from areas of interest on a dry glass knife. The sections were stained with Multiple Stain (Polysciences, Inc.).
cular formations, nor microaneuysms were observed in any animals with 5 months duration diabetes. However, by TEM occasionally a degenerating pericyte was observed and minimal basement membrane thickening was detected in these 5 month diabetic rats. Hydrogen peroxide localized as an electron dense precipitate, cerium perhydroxide, in cytoplasmic vesicles, tight junctions and plasma membranes of endothelial cells, basement membranes, and vessel lumens in the neural retina of diabetic rat eyes (Fig. 1). There was minimal to no localization of peroxide in comparable sites in eyes from nondiabetic, age-matched controls (Fig. 2). After 10 months of diabetes, there were many endothelial cells with numerous cytoplasmic vesicles which fused with the plasma membrane and appeared to discharge peroxide into the basement membrane and the adjacent pericyte (Fig. 1). Although peroxide localization in the common basement membrane between endothelial cells and pericytes was observed frequently, there was minimal to no localiza-
Statistical methods Data on hydrogen peroxide localization in diabetic and nondiabetic rats was analyzed by random effects logistic regression to determine if diabetes status and retinal area influenced the probability that a sampled vessel would be positive for hydrogen peroxide while accounting for the clustering of sampled vessels within individual animals. Adjusted hydrogen peroxide positive percentages, odds ratios, and corresponding 95% confidence bounds were estimated for each subgroup or comparison of interest using the EGRET epidemiological statistics software package (SERC, Seattle, WA). The independent-sample t-test was used to compare mean fibronectin localization between replicate measurements from a diabetic rat and a nondiabetic rat. RESULTS
The BBZ/Wor rat is an obese, spontaneously diabetic rat with noninsulin dependent diabetes. Blood sugar levels and total glycosylated hemoglobin values, for diabetic and nondiabetic control rats are shown in Table 1. There was significant elevation of blood sugar and total glycosylated hemoglobin levels in the diabetic rats. All 5-month duration diabetic BBZ/Wor animals had a vascular pattern comparable to age-matched nondiabetic BB DR /Wor as determined from the ADPase flat-embedded retinas. No intra-retinal microvascular abnormalities (IRMA formations), preretinal neovas-
FIG. 1. Localization of hydrogen peroxide (arrows) in vessel lumen (L), endothelial cell (EC), and basement membrane (BM) of a capillary in the retina of a BBZ/Wor rat after 5 months of diabetes. Note the localization of hydrogen peroxide in tight junctions (TJ) and in cytoplasmic vesicles (CV) which fuse with the basement membrane. There is no hydrogen peroxide in the pericyte (P) and pericyte basement membrane. Red blood cell (RBC). 1 20,000.
/ 2b30 5134 Mp 114 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
115
FIG. 2. Age matched, nondiabetic control of comparable area to that shown in Fig. 1. There is no localization of hydrogen peroxide. Basement membrane (BM); endothelial cell (EC); pericyte (P); red blood cell (RBC). 1 20,000.
FIG. 3. Control for specificity of NADH oxidase enzyme activity in which the substrate, NADH, was omitted. There is no cerium reaction product in the vessel lumen endothelial cells (EC), basement membranes (BM), or pericytes (P). Red blood cell (RBC). 1 20,000.
tion of peroxide in the external pericyte basement membrane. Endothelial cells with numerous plasma membranes associated vesicles appeared to bulge into and reduce the area of the vessel lumen. Specificity of the NADH oxidase enzyme activity reaction was validated by omission of the substrate NADH in reaction medium which contained both the chromogen, cerium chloride, and the inhibitor of catalase, aminotriazole. There was no cerium perhydroxide precipitate in the vessel lumen, cytoplasm, plasma membrane, tight junction or basement membrane of any blood vessels in sections of retina in which the substrate was omitted (Fig. 3). There was no reaction product in specimens in which aminotriazole was omitted from the reaction medium. Semiquantitation of peroxide localization in the central area of the retina showed significantly more peroxide in retinas from diabetic eyes as compared to nondiabetic eyes (p Å .001) (Table 2). There was a significant difference in the central vs. the peripheral retina in diabetic rats (p Å .001). Peroxide localization in nondiabetic rats did not differ significantly between central retina and peripheral retina (Table 2).
Colloidal gold localization of endogenous serum albumin demonstrated extensive extravasation of albumin around vessels in the retina of diabetic rats (Fig. 4A) as compared to nondiabetic rats (Fig. 4B). Extravasation of serum albumin was increased at sites of oxidative injury as shown by colocalization of serum albumin and peroxide (Fig. 5). Serum albumin localized in cytoplasmic vesicles (Fig. 5) suggesting transcytotic transport of serum albumin. Although serum albumin localized in association with peroxide in tight junctions (Fig. 5), it was not possible to detect any disruption or destruction of tight junctions with the standard voltage (75 kV) TEM studies done here. Fibronectin extravasation, as shown by colloidal gold localization, was similar to the pattern shown for serum albumin leakage in retinas of diabetic eyes as Table 2. Percentage of Blood Vessels Positive for Peroxide
Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
/ 2b30 5134 Mp 115 Wednesday Nov 12 11:07 AM EL–FRB 5134
Central Retina
Peripheral Retina
80.00 { 10.52 38.28 { 9.68
44.24 { 12.79 26.47 { 8.62
116
E. A. ELLIS et al.
FIG. 4. (A) Localization of serum albumin with 20 nm colloidal gold particles (arrowheads) showing leaky vessel with extravasation of serum albumin beyond basement membrane in retina of BBZ/Wor rat after 5 months of diabetes. Note the thickening of the basement membrane (BM) as compared to that in Fig. 4B. Endothelial cell (EC); vessel lumen (L); pericyte (P). 1 20,000; (B) Localization of serum albumin in age matched, nondiabetic BB/WorDR control rat. Serum albumin is confined to the vessel lumen (L). Basement membrane (BM); endothelial cell (EC). Compare the basement membrane to that in Fig. 4A. 1 20,000.
compared to retinas of age matched, nondiabetic control eyes. There was more fibronectin localization associated with areas of peroxide localization in retinas from diabetic eyes (Fig. 6A) than in retinas from nondiabetic eyes (Fig. 6B). Mean fibronectin localization in diabetic rat was significantly greater than in nondiabetic rat (p Å .00001) Table 3. DISCUSSION
This study demonstrates cytochemical and morphological evidence of oxidative injury. NADH oxidase activity is present at sites of albumin and fibronectin leakage. The endothelial cell was documented as a source of free radical derived oxidants in the spontaneously diabetic, noninsulin dependent model of diabetes. Oxidative injury appears to be a very early event in this model of noninsulin dependent diabetes because it was present in animals that were diabetic for 5 months but had no angiopathic changes, as determined by ADPase flat-embedding.
The obese BBZ/Wor rat, a new animal model for noninsulin dependent diabetes mellitus, has been developed by crossing the BBDP/Wor rat, a lean, lymphopenic animal with autoimmune, insulin dependent diabetes mellitus, with the Zucker fatty rat, an obese animal with insulin resistance and glucose intolerance. Obese BBZ/Wor males become diabetic more frequently and at an earlier age than obese females, suggesting a role for gender in the pathogenesis of this type of diabetes. The obese BBZ/Wor rat is lymphopenic, hyperinsulinemic with peripheral insulin-resistance and develops spontaneous autoimmune noninsulin dependent diabetes mellitus at a mean age of 70 days. Lymphocytic insulitis is mild and beta-cell destruction is incomplete with hyperplasia of existing beta cells allowing prolonged survival without insulin therapy.24 In the diabetic rats, urinary protein and albumin were significantly elevated over age-matched diabetes resistant controls after one month of diabetes and further increased with the duration of diabetes.28
/ 2b30 5134 Mp 116 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
117
FIG. 5. Co-localization of serum albumin (arrowheads) and hydrogen peroxide (arrows) in retina of BBZ/Wor diabetic rat which had been diabetic for 10 months. Serum albumin localizes in cytoplasmic vesicles and tight junction (TJ) in association with hydrogen peroxide. Basement membrane (BM); endothelial cell (EC); red blood cell (RBC). 1 40,000.
Previous studies of oxidative injury in diabetes have demonstrated serum levels of free radicals, conjugated dienes, and lipid peroxidation products, all of which are measures of free radical mediated processes. The biochemical studies of Murata et al. demonstrated the involvement of NADH in lipid peroxidation in the retina of rats with alloxan-induced diabetes.29 Nishida et al. showed that superoxide reacted with unsaturated fatty acids in the retina of rats with streptozotocin-induced diabetes resulting in increased levels of lipid peroxidation products and a decrease in levels of the free radical scavenger, SOD.30 Armstrong and Al1–Wadi observed a time-dependent increase in lipid hydroperoxides which correlated with focal damage to the photoreceptor layer of the retina of rats with streptozotocininduced diabetes.31 The above studies involved an aggressive, chemically induced, insulin dependent diabetes. Diabetic retinopathy is a disease process which affects predominantly the central retina. The semiquantitative data for peroxide localization indicate that the changes observed in this animal model are consistent with the pattern observed in patients with diabetic retinopathy. The cerium NADH oxidase enzyme activity technique shows actual sites of peroxide generation.
Semiquantitation of amounts of cerium reaction product, a direct indication of oxidase activity, shows a significant increase of NADH oxidase enzyme activity in diabetic animals as compared to nondiabetic controls. Elevated ROS can occur as a consequence of cellular damage and inflammation, which may be contributing to the increased NADH oxidase activity observed in the retinas of the diabetic rats. The studies of Murata et al.29 indicated that NADH plays an important role in lipid peroxidation in retina with a 40% increase in the rate of peroxidation in diabetic retina as compared to nondiabetic retina. There is increased cytosolic free NADH/NAD/ in tissues exposed to elevated glucose levels at normal tissue pO2 . This redox imbalance has been linked to increased metabolism of glucose via the sorbitol pathway in retina and other tissues subject to vascular injuries of diabetic hyperglycemia.13 Retina is one of the most highly oxygenated tissues in the body. In the presence of oxygen and an abundance of the substrate NADH, NADH oxidase enzyme activity increases superoxide production in the retina. More recent studies have validated that NADH oxidase enzyme activity is the major generator of superoxide in vascular endothelial cells.32,33
/ 2b30 5134 Mp 117 Wednesday Nov 12 11:07 AM EL–FRB 5134
118
E. A. ELLIS et al.
FIG. 6. (A) Co-localization of fibronectin with 18 nm colloidal gold particles (arrowheads) with hydrogen peroxide (arrows) in retina of BBZ/Wor diabetic rat which had been diabetic for 10 months. Basement membrane (BM); endothelial cell (EC); vessel lumen (L); red blood cell (RBC). There are increased levels of fibronectin as compared to the age matched, nondiabetic control in Fig. 6B. 1 40,000; (B) Localization of fibronectin in retina of an age matched, nondiabetic control rat. Fibronectin is confined to the vessel lumen. Basement membrane (BM); endothelial cell (EC); vessel lumen (L). 1 40,000.
Recent studies indicate that disruption of the blood-retinal barrier is mediated by an increase in transendothelial transport of serum proteins through transcytotic vesicles.34,35 Studies of binding and uptake of native and glycosylated albumin-gold complexes in perfused rat lungs showed enhanced binding, internalization and transport of glycosylated albumin.36 Morphometric studies of capillary abnormalities in sural nerves of patients with diabetic neuropathy showed increased endothelial cell area and decreased lumen area in capillaries where the endothelial cell appeared to bulge into the lumen. A reduction in lumen size would be expected to contribute to capillary closure, reduced blood flow, and increased hypoxia.37 Increased vascular permeability is an early complication of diabetic retinopathy. Most studies of retinal dysfunction in diabetic retinopathy have utilized fluorescein angiography. The post embedding procedure presented here offers several advantages in
that there is no need to inject or infuse nonphysiological tracers and there is the improved resolution of TEM. In addition, this post embedding procedure on sections allowed correlative studies after cytochemical localization procedures. Studies by Bendayan et al. demonstrated leakage of endogenous serum albumin in the glomerular wall of streptozotocin-induced diabetes.38 A similar immunocytochemical technique has been used to study bloodretinal barrier disruption in cadaveric eyes from patients with diabetic retinopathy.39 Treatment of confluent monolayers of human umbilical vein endothelial cells with sublethal
Table 3. Immunocytochemical Localization of Fibronectin Colloidal Gold Particles/um2 Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
/ 2b30 5134 Mp 118 Wednesday Nov 12 11:07 AM EL–FRB 5134
4.45 { 1.09 1.56 { 0.79
NADH oxidase activity in the rat
concentrations of hydrogen peroxide resulted in partial redistribution of actin filaments from the cell periphery and cell retraction that opened gaps between cells.40 There was also redistribution of endothelial cell adhesion molecules. These in vitro changes were interpreted as manifestations of sublethal hydrogen peroxide injury which could cause endothelial dysfunction.40 Hyperglycemic activation of protein kinase C is associated with increased vascular permeability in diabetic retinopathy.41 Phorbol myristate acetate which activates protein kinase C induced endocytosis and exocytosis ( transcytosis ) in toad bladder with cellular profiles of plasma membrane associated cytoplasmic vesicles similar to those shown here.42 More recent studies showed oxygen radical activation of protein kinase C and increased transendothelial permeability in lung inflammation.43 In addition to extravasation of serum albumin, our immunocytochemical studies of disruption of the blood-retinal barrier (serum albumin leakage) showed leakage which correlated with hydrogen peroxide localization. Our morphological findings are consistent with a possible oxygen radical activation of protein kinase C and increased vascular permeability through transcytotic vesicles as demonstrated by other studies of disruption of the blood-retinal barrier 39 and bloodnerve barrier 37 in diabetes. Fibronectin synthesis and release into the circulation has been documented as a result of oxidantinduced vascular injury in lung.44 Increased synthesis of fibronectin in retinal vessels from diabetic patients has been documented.45 The studies shown here show elevation of total fibronectin and extravasation of fibronectin in diabetic rat eyes at sites of oxidative injury ( peroxide localization ) as compared to age matched controls. Previous studies of oxidative injury in diabetic retinopathy involved insulin dependent models and focused on the role of activated leukocytes in initiating endothelial cell dysfunction. In this study, increased numbers of leukocytes were not observed in the retinas of the BBZ/Wor diabetic rats compared to control rats. This may be due to the fact that this noninsulin dependent model has a milder diabetes with an attenuated disease coarse as compared to insulin dependent diabetes. Porta proposed that since the endothelium is strategically situated in the vessel wall, it plays a major role in remodeling the retinal microvasculature in diabetes.46 This report is the first in vivo demonstration that the endothelial cell in diabetic retinopathy may play a role in the initiation of oxidative injury. This oxidative injury may contribute to the observed dysfunction of the blood-retinal barrier. Cytochemical lo-
119
calization of NADH oxidase enzyme activity as presented here supports a role for hyperglycemia in development of free radical-associated vascular complications in diabetes. REFERENCES 1. Pugliese, G.; Tilton, R. G.; Williamson, J. R. Glucose-induced metabolic imbalances in the pathogenesis of diabetic vascular disease. Diabetes Metab. Rev. 7:35–39; 1991. 2. Ruderman, N. B.; Williamson, J. R.; Brownlee, M. Glucose and diabetic vascular disease. FASEB J. 6:2905–2914; 1992. 3. Collier, A.; Rumley, A.; Rumley, A. G.; Paterson, J. R.; Leach, J. P.; Lowe, G. D. O.; Small, M. Free radical activity and hemostatic factors in NIDDM patients with and without microalbuminuria. Diabetes 41:909–913; 1992. 4. Oberly, L. W. Free radicals and diabetes. Free Rad. Biol. Med. 5:113–124; 1988. 5. Baynes, J. W. Role of oxidative stress in development of complications in diabetes. Diabetes 40:405–412; 1991. 6. Ceriello, A.; Giugliano, D.; Quatraro, A.; Dello Russo, P.; Lefebvre, P. J. Metabolic control may influence the increased superoxide generation in diabetic serum. Diab. Med. 8:540–542; 1991. 7. Tesfamariam, B. Free radicals in diabetic endothelial cell dysfunction. Free Rad. Biol. Med. 16:383–391; 1994. 8. Weiss, S. J.; Young, J.; LoBuglio, A. F.; Slivka, A.; Mineh, N. F. Role of hydrogen peroxide in neutrophil-mediated destruction of cultured endothelial cells. J. Clin. Invest. 68:714–721; 1981. 9. Wieruz–Wysocka, B.; Wysocky, H.; Siekirka, H.; Wykretowicz, A.; Szczepanik, A.; Klimas, R. Evidence of polymorphonuclear neutrophil activation in patients with insulin-dependent diabetes mellitus. J. Leuk. Biol. 42:519–523; 1987. 10. Schroder, S.; Palinski, W.; Schmid–Schonbein, G. W. Activated monocytes and granulocytes, capillary nonperfusion, and neovascularization in diabetic retinopathy. Am. J. Path. 139:81– 100; 1991. 11. Freedman, S. F.; Hatchell, D. L. Enhanced superoxide radical production by stimulated polymorphonuclear leukocytes in a cat model of diabetes. Exp. Eye Res. 55:767–773; 1992. 12. Rosen, G. M.; Freeman, S. A. Detection of superoxide generated by endothelial cells. Proc, Natl. Acad. Sci. USA 81:7269–72732; 1984. 13. Williamson, J. R.; Chang, K.; Frangos, M.; Hasan, K. S.; Ido, Y.; Kawamura, T.; Nyengaard, J. R.; van den Enden, M.; Kilo, C.; Tilton, R. G. Hyperglycemic pseudohypoxia and diabetic complications. Diabetes 42:801–813; 1993. 14. Bravenboer, B.; Kappelle, A. C.; Hamers, F. P. T.; van Buren, T.; Erkelens, D. W.; Gispen, W. H. Potential use of glutathione for the prevention and treatment of diabetic neuropathy in the streptozotocin-induced diabetic rat. Diabetologia 35:813–817; 1992. 15. Cameron, N. E.; Cotter, M. A.; Maxfield, E. K. Anti-oxidant treatment prevents the development of peripheral nerve dysfunction in streptozotocin-diabetic rats. Diabetologia 36:299–304; 1993. 16. Zhang,H.; Agardh, E.; Agardh, C-D. Nitro blue tetrazolium staining — A morphological demonstration of superoxide in the rat retina. Graefe’s Arch. Clin. Exp. Ophthalmol. 231:178–183; 1993. 17. Briggs, R. T.; Karnovsky, M. L.; Karnovsky, M. J. Localization of NADH oxidase enzyme activity on the surface of human polymorphonuclear leukocytes by a new cytochemical method. J. Cell Biol. 67:566–586; 1975. 18. Karnovsky, M. J. Cytochemistry and reactive oxygen species. Histochemistry 102:15–27; 1994. 19. Warren, J. S.; Kunkel, R. G.; Simon, R. H.; Johnson, K. J.; Ward, P. A. Ultrastructural cytochemical analysis of oxygen radicalmediated immunoglobulin A immune complex induced lung injury in the rat. Lab. Invest. 60:641–658; 1989.
/ 2b30 5134 Mp 119 Wednesday Nov 12 11:07 AM EL–FRB 5134
120
E. A. ELLIS et al.
20. Guy, J.; Ellis, E. A.; Rao, N. A. Hydrogen peroxide localization in experimental optic neuritis. Arch. Ophthalmol. 108:1614– 1621; 1990. 21. Shlafer, M.; Brosamer, K.; Forder, J. R.; Simon, R. H.; Ward, P. A.; Grum, C. M. Cerium chloride as a histochemical marker of hydrogen peroxide in reperfused ischemic hearts. J. Mol. Cardiol. 22:83–97; 1990. 22. Soares, F. A.; Shaughnessy, S. G.; MacLarkey, W. R.; Orr, F. W. Quantification and morphological demonstration of reactive oxygen species produced by Walker 256 tumor cells in vitro and during metastasis in vivo . Lab. Invest. 71:480–489; 1994. 23. Bendayan, M. Use of the protein A-gold technique for the morphological study of vascular permeability. J. Histochem. Cytochem. 28:1251–1254; 1980. 24. Guberski,D.L.; Butler, L.; Manzi, S. M.; Stubbs, M.; Like, A. A. The BBZ/Wor rat: Clinical characteristics of the diabetic syndrome. Diabetologia 36:912–919; 1993. 25. Murray, F. T.; Beyer–Mears, A.; Johnson, R. D.; Sima, A. A. F.; Cameron, D. F.; Sninsky, C. A.; Selawry, H. Assessment of proteinuria and neuropathy in the nonimmunosupressed BB diabetic rat after abdominal intratesticular islet transplantation. Transplantation 56:680–686; 1993. 26. Briggs, R. T.; Karnovsky, M. L.; Karnovsky, M. J. Hydrogen peroxide in chronic granulomatous disease: A cytochemical study of reduced pyridine nucleotide oxidases. J. Clin. Invest. 59:1088–1098; 1977. 27. Lutty, F. A.; McLeod, D. S. A new technique for visualization of the human retinal vasculature. Arch Ophthalmol. 110:2:267– 276; 1992. 28. Murray, F. T.; Wachowski, M. B.; Dani, A.; Ellis, E. A.; Grant, M. B. Intermediate and long term diabetic (Type II) complications in the spontaneously diabetic BBZ/Wor rat. Diabetes 45 suppl. 2:272A abstr. (1996). 29. Murata, T.; Nishida, T.; Eto, S.; Mukai, N. Lipid peroxidation in diabetic rat retina. Metab. Ped. Ophthalmol. 5:83–87; 1981. 30. Nishida, T.; Nakagawa, S.; Manabe, R. Superoxide dismutase activity in diabetic rat retina. Jpn. J. Ophthalmol. 28:377–382; 1984. 31. Armstrong, D.; Al–Awadi, F. Lipid peroxidation and retinopathy in streptozocin-induced diabetes. Free Rad. Biol. Med. 11:433–436; 1991. 32. Mohazzab, K. M.; Kaminski, P. M.; Wolin, M. S. NADH oxidoreductase is a major source of superpoxide anion in bovine coronary artery endothelium. Am. J. Physiol. 266:H2568– H2572; 1994. 33. Rajagopalan, S.; Kurz, S.; Munzel, T.; Tarpey, M.; Freeman, B. A.; Griendling, K. K.; Harrison, D. G. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation. J. Clin. Invest. 97:1916–1923; 1996.
34. Lin, S.-J.; Hong., C.-Y.; Chang, M.-S.; Chiang, B. N.; Chien, S. Increased aortic endothelial cell death and enhanced transendothelial macromolecular transport in streptozotocin-diabetic rats. Diabetologia 36:926–930; 1993. 35. Derevjanik, N. L.; Wilson, C. A.; Berkowitz, B. A.; Malow, J.; Vinores, S. A. Blood-retinal barrier (BRB) breakdown in diabetic rabbits assessed by electron microscopic immunocytochemistry and magnetic resonance imaging (MRI). Invest. Ophthalmol. Vis. Sci. 37(suppl.): S968 abstr. (1996). 36. Villaschi, S.; Johns, L.; Cirigliano, M.; Pietra, G. G. Binding and uptake of native and glycosylated albumin-gold complexes in perfused rat lung. Microvascular Res. 32:190–199; 1986. 37. Malik, R. A.; Tesfaye, S.; Thompson, S. D.; Veves, A.; Hunter, A.; Sharma, A. K.; Ward, J. D.; Boulton, A. J. M. Transperineurial capillary abnormalities in the sural nerve of patients with diabetic neuropathy. Microvascular Res. 48:236–245; 1994. 38. Bendayan, M.; Gingras, D.; Charest, P. Distribution of endogenous albumin in the glomerular wall of streptozotocin-induced diabetic rats as revealed by high resolution immunocytochemistry. Diabetologia 29:868–875; 1986. 39. Vinores, S. A.; Gadegbeku, C.; Campochiaro, P. A.; Green, W. R. Immunohistochemical localization of blood-retinal barrier breakdown in human diabetics. Am. J. Pathol. 134:231–235; 1989. 40. Bradley, J. R.; Thiru, S.; Pober, J. S. Hydrogen peroxide-induced endothelial retraction is accompanied by a loss of the normal spatial organization of endothelial cell adhesion molecules. Am. J. Pathol. 147:627–641; 1995. 41. Lee, T-S.; MacGregor, L. C.; Fluharty, s. J.; King, G. L. Differential regulation of protein kinase C and (Na,K)-adenosine triphosphatase activities by elevated glucose levels in retinal capillary endothelial cells. J. Clin. Invest. 83:90–94; 1989. 42. Masur, S. K.; Sapirstein, V.; Rivero, D. Phorbol myristate acetate induces endocytosis as well as exocytosis and hydroosmosis in toad urinary bladder. Biochim. Biophys. Acta 821:286–296; 1985. 43. Lynch, J. J.; Ferro, T. J.; Blumenstock, F. A.; Brockenauer, A. M.; Malik, A. B. Increased endothelial albumin permeability mediated by protein kinase C activation. J. Clin. Invest. 85:1991–1998; 1990. 44. Peters, J. H.; Ginsberg, M. H.; Bohl, B. P.; Sklar, L. A.; Cochrane, C. G. Intravascular release of intact cellular fibronectin during oxidant-induced injury of the in vitro perfused rabbit lung. J. Clin. Invest. 78:1596–1603; 1986. 45. Roy, S.; Cagliero, E.; Lorenzi, M. Fibronectin over expression in retinal microvessels of patients with diabetes. Invest. Ophthalmol. Vis. Sci. 37:258–266; 1996. 46. Porta, M. Endothelium: The main actor in the remodeling of the retinal microvasculature in diabetes. Diabetologia 39:739–744; 1996.
/ 2b30 5134 Mp 120 Wednesday Nov 12 11:07 AM EL–FRB 5134
PII S0891-5849(97)00202-5
Original Contribution INCREASED NADH OXIDASE ACTIVITY IN THE RETINA OF THE BBZ/WOR DIABETIC RAT E. ANN ELLIS,* MARIA B. GRANT,* FREDERICK T. MURRAY, † MARTHA B. WACHOWSKI, † DENNIS L. GUBERSKI, ‡ PAUL S. KUBILIS,§ and GERARD A. LUTTY \ * Department of Medicine, College of Medicine, University of Florida, Gainesville, FL, USA;† Pharmacia & Upjohn, Kalamazoo, MI, ‡ Department of Pathology, University of Massachusetts School of Medicine, Worcester, MA, USA; § Division of Biostatistics, University of Florida, Gainesville, FL, USA; and \ Department of Ophthalmology, The Johns Hopkins School of Medicine, Baltimore, MD, USA (Received 1 July 1996; Revised 8 April 1997; Accepted 12 May 1997)
Abstract—This morphological study demonstrates a role for endothelial cells in generating reactive oxygen species in early stages of retinopathy in the BBZ / Wor rat, an obese, noninsulin dependent model of diabetes. Hyperglycemia induced pseudohypoxia results in an imbalance in cytosolic NADH / NAD/. In the oxygen-rich environment of the retina, NADH oxidase generates superoxide radical which is dismutated to hydrogen peroxide. Localization of hydrogen peroxide by the cerium NADH oxidase enzyme activity cytochemical localization technique shows a statistically significant increase of peroxide localization in the central retina of diabetic rats as compared to age-matched, nondiabetic controls. Endothelial cell dysfunction, indicated by leakage of endogenous serum albumin, coincided with areas of NADH oxidase activity localization. In diabetic rats there are increased levels of fibronectin in areas of hydrogen peroxide localization. This in vivo , morphological study is the first demonstration of oxidative injury and endothelial cell dysfunction in the retina of a spontaneous, noninsulin dependent model of diabetes. q 1997 Elsevier Science Inc. Keywords—Oxidative injury, Endothelial dysfunction, Vascular leakage, Retinopathy, Diabetes, NADH oxidase
tomy.11 Endothelial cells may produce reactive oxygen species capable of tissue injury.12 Increased flux of glucose via the sorbitol pathway results in increased cytosolic NADH / NAD/ and increased superoxide production described as hyperglycemia induced pseudohypoxia.13 Autoxidation of glucose, a nonenzymatic process, increases with hyperglycemic pseudohypoxia and is a source of reactive oxygen species in diabetes.2 Any and all of the above mechanisms may play a role in oxidative injury. Free radicals or reactive oxygen species in cells and tissues have been demonstrated by a number of methods. Effects of antioxidants such as superoxide dismutase (SOD), glutathione peroxidase, catalase and vitamin E have been detected indirectly.14,15 Direct detection of reactive oxygen species has been difficult because reactive oxygen species are short lived and are catalyzed rapidly by scavenging systems. At the light microscope level, nitroblue tetrazolium has been used for histochemical demonstration of superoxide in the retina.16
INTRODUCTION
Hyperglycemia is the primary initiating factor for all types of diabetic vascular complications including diabetic retinopathy.1 – 3 Increased oxidative stress is observed in diabetes.4,5 Production of free radicals correlates to metabolic control and degree of hyperglycemia.6,7 Free radical mediated retinal vascular dysfunction may occur by several mechanisms. Activated polymorphonuclear leukocytes generate free radical derived oxidants.8 Activation of polymorphonuclear leukocytes occurs in diabetic patients, 9 in an alloxan-induced diabetic rat model, 10 and in a cat model of diabetes induced by partial pancreatecThis study was supported in part by NIH Grant EY07739, the American Heart Association and the Department of Health and Rehabilitative Services of the State of Florida for the University of Florida Diabetes Research, Education and Treatment Center. Address correspondence to: M. B. Grant, Box 100226, Gainesville, FL 32610-0226; E-Mail: [email protected] 111
/ 2b30 5134 Mp 111 Wednesday Nov 12 11:07 AM EL–FRB 5134
112
E. A. ELLIS et al.
The cerium NADH oxidase enzyme activity methodology introduced by Briggs and Karnovsky 17,18 has been used for cytochemical demonstration of hydrogen peroxide. This method has been used to detect reactive oxygen species in vivo in models of IgA damage to lungs, 19 in optic neuritis, 20 ischemia-reperfusion of the heart, 21 and during metastasis of tumor cells.22 In the cerium NADH oxidase enzyme activity localization technique, NADH oxidase enzyme activity, in the presence of oxygen and the substrate NADH, generates the reactive oxygen species superoxide (O2 i 0 ) from which other free radical derived oxidants such as hydrogen peroxide (H2O2 ) and hydroxyl radical (OH i 0 ) are generated. NADH oxidase
NADH / 2O2
2O2 i 0 / NAD / / H /
Superoxide dismutates, either spontaneously or catalyzed by superoxide dismutase, to yield hydrogen peroxide. SOD
2O2 i 0 / 2H/ r H2O2 / O2 In situ subcellular sites of generation of the free radical derived oxidant, hydrogen peroxide, can be shown by using cerium chloride in the reaction medium with the substrate NADH. Cerium ions react with hydrogen peroxide to produce an electron dense precipitate, cerium perhydroxide, Ce(OH)2OOH. H2O2 / CeCl 3 r Ce(OH)2OOH Vascular permeability, disruption of the blood retinal barrier, can be assessed by colloidal gold immunocytochemical localization of extravasation of endogenous serum proteins.23 Serum albumin or other endogenous serum proteins are retained in their normal location by fixation; immunocytochemical localization of these endogenous proteins can be done as a post embedding procedure on tissue sections. This study presents an investigation of oxidative injury and endothelial cell dysfunction in eyes of obese, noninsulin dependent diabetic BBZ/Wor rats.24 Free radical derived oxidant localization is documented by the cerium NADH oxidase enzyme activity cytochemical method and endothelial cell dysfunction is confirmed by colloidal gold immunocytochemical localization of extravasated serum albumin. In addition, fibronectin is elevated in and around vessels where oxidative injury has occurred in diabetic rats.
MATERIALS AND METHODS
Chemicals 3-Amino-1,2,4-triazole, cerium chloride, cold water fish gelatin, dimethylsulfoxide ( DMSO ) , fibronectin, glycine, rat serum albumin, reduced nicotinamide adenine dinucleotide ( NADH ) , maleic acid, periodic acid, Tris base, and Triton X-100 were purchased from Sigma Chemical Co. ( St. Louis, MO ) . Benzyldimethyl amine ( BDMA ) , colloidal gold labeled rabbit antigoat IgG antibodies, dodecenyl succinic anhydride ( DDSA ) , Eponate 12, osmium tetroxide, Quetol 651, and sodium cacodylate were purchased from Ted Pella Inc. (Redding, CA). Polyclonal goat anti-rat serum albumin antibodies, IgG fraction, were purchased from Organon Teknika-Cappel (Durham, NC). Polyclonal rabbit anti-human fibronectin antibodies, IgG fraction, were purchased from Dako Corporation (Carpinteria, CA). Colloidal gold labeled donkey anti-rabbit IgG antibodies were purchased from Jackson Immunoresearch Labs Inc. (West Grove, PA). Isolab Gly Affin GHb test kit was purchased from Isolab (Akron, OH). Sodium pentobarbital was purchased from Abbott Laboratories (Chicago, IL). Methoxyflurane was from Mallinkrodt Veterinary Inc. (Mundelein, IN). Animals Male BBZ/Wor and age matched, nondiabetic BB DR /Wor rats were maintained on Purina 5002 rodent chow and sterilized water ad lib in a viral free facility. All rats were obtained from the NIH facility at the University of Massachusetts and shipped in isolator rat boxes to the viral free facility at Pharmacia and Upjohn in Kalamazoo, MI. All BBZ/Wor rats were housed individually in micro-isolator rat boxes. BB DR /Wor rats were housed two to a micro-isolator rat box. Diabetic and nondiabetic control rats ranged from 7–12 months of age. Diabetes developed at 2–3 months of age and length of diabetes ranged from 4.5–10 months. Blood samples were obtained from tail veins under light methoxyflurane anesthesia. Whole blood glucose was measured using the Glucometer Elite Glucose Meter from Bayer Diagnostic Division (Tarrytown, NY). Total glycosylated hemoglobin was measured from whole blood removed from tail veins using the Isolab Gly Affin Test kit as previously described.25 Cytochemical localization of hydrogen peroxide Rats were euthanized by intraperitoneal injection with sodium pentobarbital at 50 mg/kg body weights. Eyes were enucleated and one eye of each animal was
/ 2b30 5134 Mp 112 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
processed as follows. The globe and attached optic nerve were dissected out and fixed for 1 h in cold 5.0% acrolein in 0.1 M sodium cacodylate-HCl buffers (pH 7.4). Specimens were washed overnight in cold 0.15 M sodium cacodylate-HCl buffers (pH 7.4) plus 5% sucrose and 1% dimethylsulfoxide (DMSO). Specimens were brought to room temperature in the final two buffers washes which contained 0.1 M glycine. Specimens were preincubated for 30 min at 377C with agitation in the following medium: 2.0 mM cerium chloride, 10 mM 3-amino-1,2,4-triazole, 0.1 M Trismaleate buffer (pH 7.5), 7% sucrose, and 0.0002% Triton X-100. They were then incubated for 1 h at 377C with agitation in the following complete reaction medium: 2 mM cerium chloride, 10 mM 3-amino-1,2,4triazole, 0.8 mM NADH, 0.1 M Tris-maleate buffer (pH 7.5), 7% sucrose and 0.0002% Triton-X 100. The reaction was terminated by washing in cold 0.1 M Trismaleate buffer (pH 7.5), 7% sucrose followed by a wash in cold 0.15 M sodium cacodylate-HCl buffer (pH 7.4), 7% sucrose. Control for the specificity of the reaction was done by reacting pieces of tissue with cerium chloride and aminotriazole in the absence of the substrate, NADH. Specimens were postfixed overnight in 1% osmium tetroxide, 0.1 M sodium cacodylate-HCl buffer (pH 7.4), 7% sucrose in the cold, dehydrated through an ethanol series (20%, 40%, 60%, 80%, 95% 1 2, 100% 1 2) to propylene oxide and infiltrated and embedded in epoxy resin (Quetol 651-1.15 g, Eponate 12-1.83 g, DDSA-7.02 g, BDMA-0.14ml). Grids were examined and photographed at 75 kV in the transmission electron microscope (TEM) without post staining.
Semiquantitative evaluation of peroxide localization by the cerium NADH oxidase enzyme activity method Semiquantitative evaluation of peroxide localization was done by a modification of the method of Briggs et al.26 Eyes were sectioned median longitudinally through the optic nerve head. These sections included the neural retina approximately 1 mm on either side of the optic nerve head. This area was designated as the area of central retina. Sections from the far peripheral retina were taken starting at the ora serrata and extending 1.5-2 mm toward the posterior of the eye and designated as the area of peripheral retina for comparison to the central area. All blood vessels in the neural retina which were not blocked by grid bars were examined at a magnification of 10,000 1 and scored either positive ( / ) or negative ( 0 ) for cerium perhydroxide precipitate. Results were expressed as the percentage of positive vessels relative to the total number of vessels ex-
113
amined for each grid. Seven diabetic eyes and five nondiabetic eyes were examined. Immunocytochemical localization of endogenous serum albumin Sections from eyes in which NADH oxidase enzyme activity had been localized were picked up on nickel grids, oxidized for 30 min with 1% periodic acid followed by 5 1 5 min washes in deionized water. Grids were floated 2 1 5 min on phosphate buffered saline (PBS) (pH 7.2), followed by 30 min on PBS blocker [PBS plus 2% nonfat dry milk and 2% cold water fish gelatin], and then reacted overnight at 47C in a moist chamber with polyclonal goat anti-rat serum albumin antibodies, IgG fraction, diluted 1:200 with PBS blocker. After 2 1 5 min washes with PBS blocker followed by 2 1 5 min washes with PBS, grids were washed with Tris-HCl saline buffer (pH 7.6) (TBS) followed by 30 min on TBS blocker [TBS plus 2% nonfat dry milk and 2% cold water fish gelatin]. Grids were then incubated for 1 h at room temperature in a moist chamber on drops of rabbit anti-goat IgG labeled with 20 nm colloidal gold diluted 1:40 with TBS blocker. Grids were washed with TBS blocker, followed by 2 1 5 min washes with TBS and then 3 1 5 min washes with deionized water. Control grids were incubated with only the gold labeled secondary antibody or with antibodies to rat serum albumin which had been absorbed overnight with an excess of rat serum albumin. Immunocytochemical localization of fibronectin Sections from eyes in which NADH oxidase enzyme activity had been localized were picked up on nickel grids. The immunocytochemical localization procedure was the same as that used for localization of serum albumin. Primary antibody was a polyclonal rabbit anti-human fibronectin diluted 1:100; secondary antibody was a donkey anti-rabbit IgG labeled with 18 nm colloidal gold diluted 1:40. Control grids were incubated with only the gold labeled secondary antibody or with antibodies to fibronectin which had been absorbed overnight with an excess of fibronectin. ADPase histochemical studies One eye from twelve 5-month duration diabetic BBZ/Wor animals and six age-matched nondiabetic BB DR /Wor rats were used for ADPase flat-embedded retina preparations. The retinas from these eyes were teased from the retinal pigment epithelium, fixed in 2% paraformaldehyde, washed, and processed for adeno-
/ 2b30 5134 Mp 113 Wednesday Nov 12 11:07 AM EL–FRB 5134
114
E. A. ELLIS et al.
Table 1. Levels of Blood Sugars and Glycosylated Hemoglobin
Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
Blood Sugars (mg/dl)
Total Glycosylated Hemoglobin (%)
496 { 15 173 { 3
13.0 { 0.4 5.3 { 0.5
sine diphosphatase (ADPase) histochemical staining of the retinal vasculature as previously described.27 Four radial cuts were made in the ADPase incubated retinas so that they could be postfixed flat and embedded in JB-4 plastic (Polysciences, Inc., Warrington, PA) as described in detail previously.27 The lead ADPase reaction product in retinal vessels was visualized in the flat-embedded retinas with dark field illumination. Changes in the vascular pattern and angiopathy were evaluated in the flat perspective and then serial 2.5 mm sections were cut from areas of interest on a dry glass knife. The sections were stained with Multiple Stain (Polysciences, Inc.).
cular formations, nor microaneuysms were observed in any animals with 5 months duration diabetes. However, by TEM occasionally a degenerating pericyte was observed and minimal basement membrane thickening was detected in these 5 month diabetic rats. Hydrogen peroxide localized as an electron dense precipitate, cerium perhydroxide, in cytoplasmic vesicles, tight junctions and plasma membranes of endothelial cells, basement membranes, and vessel lumens in the neural retina of diabetic rat eyes (Fig. 1). There was minimal to no localization of peroxide in comparable sites in eyes from nondiabetic, age-matched controls (Fig. 2). After 10 months of diabetes, there were many endothelial cells with numerous cytoplasmic vesicles which fused with the plasma membrane and appeared to discharge peroxide into the basement membrane and the adjacent pericyte (Fig. 1). Although peroxide localization in the common basement membrane between endothelial cells and pericytes was observed frequently, there was minimal to no localiza-
Statistical methods Data on hydrogen peroxide localization in diabetic and nondiabetic rats was analyzed by random effects logistic regression to determine if diabetes status and retinal area influenced the probability that a sampled vessel would be positive for hydrogen peroxide while accounting for the clustering of sampled vessels within individual animals. Adjusted hydrogen peroxide positive percentages, odds ratios, and corresponding 95% confidence bounds were estimated for each subgroup or comparison of interest using the EGRET epidemiological statistics software package (SERC, Seattle, WA). The independent-sample t-test was used to compare mean fibronectin localization between replicate measurements from a diabetic rat and a nondiabetic rat. RESULTS
The BBZ/Wor rat is an obese, spontaneously diabetic rat with noninsulin dependent diabetes. Blood sugar levels and total glycosylated hemoglobin values, for diabetic and nondiabetic control rats are shown in Table 1. There was significant elevation of blood sugar and total glycosylated hemoglobin levels in the diabetic rats. All 5-month duration diabetic BBZ/Wor animals had a vascular pattern comparable to age-matched nondiabetic BB DR /Wor as determined from the ADPase flat-embedded retinas. No intra-retinal microvascular abnormalities (IRMA formations), preretinal neovas-
FIG. 1. Localization of hydrogen peroxide (arrows) in vessel lumen (L), endothelial cell (EC), and basement membrane (BM) of a capillary in the retina of a BBZ/Wor rat after 5 months of diabetes. Note the localization of hydrogen peroxide in tight junctions (TJ) and in cytoplasmic vesicles (CV) which fuse with the basement membrane. There is no hydrogen peroxide in the pericyte (P) and pericyte basement membrane. Red blood cell (RBC). 1 20,000.
/ 2b30 5134 Mp 114 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
115
FIG. 2. Age matched, nondiabetic control of comparable area to that shown in Fig. 1. There is no localization of hydrogen peroxide. Basement membrane (BM); endothelial cell (EC); pericyte (P); red blood cell (RBC). 1 20,000.
FIG. 3. Control for specificity of NADH oxidase enzyme activity in which the substrate, NADH, was omitted. There is no cerium reaction product in the vessel lumen endothelial cells (EC), basement membranes (BM), or pericytes (P). Red blood cell (RBC). 1 20,000.
tion of peroxide in the external pericyte basement membrane. Endothelial cells with numerous plasma membranes associated vesicles appeared to bulge into and reduce the area of the vessel lumen. Specificity of the NADH oxidase enzyme activity reaction was validated by omission of the substrate NADH in reaction medium which contained both the chromogen, cerium chloride, and the inhibitor of catalase, aminotriazole. There was no cerium perhydroxide precipitate in the vessel lumen, cytoplasm, plasma membrane, tight junction or basement membrane of any blood vessels in sections of retina in which the substrate was omitted (Fig. 3). There was no reaction product in specimens in which aminotriazole was omitted from the reaction medium. Semiquantitation of peroxide localization in the central area of the retina showed significantly more peroxide in retinas from diabetic eyes as compared to nondiabetic eyes (p Å .001) (Table 2). There was a significant difference in the central vs. the peripheral retina in diabetic rats (p Å .001). Peroxide localization in nondiabetic rats did not differ significantly between central retina and peripheral retina (Table 2).
Colloidal gold localization of endogenous serum albumin demonstrated extensive extravasation of albumin around vessels in the retina of diabetic rats (Fig. 4A) as compared to nondiabetic rats (Fig. 4B). Extravasation of serum albumin was increased at sites of oxidative injury as shown by colocalization of serum albumin and peroxide (Fig. 5). Serum albumin localized in cytoplasmic vesicles (Fig. 5) suggesting transcytotic transport of serum albumin. Although serum albumin localized in association with peroxide in tight junctions (Fig. 5), it was not possible to detect any disruption or destruction of tight junctions with the standard voltage (75 kV) TEM studies done here. Fibronectin extravasation, as shown by colloidal gold localization, was similar to the pattern shown for serum albumin leakage in retinas of diabetic eyes as Table 2. Percentage of Blood Vessels Positive for Peroxide
Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
/ 2b30 5134 Mp 115 Wednesday Nov 12 11:07 AM EL–FRB 5134
Central Retina
Peripheral Retina
80.00 { 10.52 38.28 { 9.68
44.24 { 12.79 26.47 { 8.62
116
E. A. ELLIS et al.
FIG. 4. (A) Localization of serum albumin with 20 nm colloidal gold particles (arrowheads) showing leaky vessel with extravasation of serum albumin beyond basement membrane in retina of BBZ/Wor rat after 5 months of diabetes. Note the thickening of the basement membrane (BM) as compared to that in Fig. 4B. Endothelial cell (EC); vessel lumen (L); pericyte (P). 1 20,000; (B) Localization of serum albumin in age matched, nondiabetic BB/WorDR control rat. Serum albumin is confined to the vessel lumen (L). Basement membrane (BM); endothelial cell (EC). Compare the basement membrane to that in Fig. 4A. 1 20,000.
compared to retinas of age matched, nondiabetic control eyes. There was more fibronectin localization associated with areas of peroxide localization in retinas from diabetic eyes (Fig. 6A) than in retinas from nondiabetic eyes (Fig. 6B). Mean fibronectin localization in diabetic rat was significantly greater than in nondiabetic rat (p Å .00001) Table 3. DISCUSSION
This study demonstrates cytochemical and morphological evidence of oxidative injury. NADH oxidase activity is present at sites of albumin and fibronectin leakage. The endothelial cell was documented as a source of free radical derived oxidants in the spontaneously diabetic, noninsulin dependent model of diabetes. Oxidative injury appears to be a very early event in this model of noninsulin dependent diabetes because it was present in animals that were diabetic for 5 months but had no angiopathic changes, as determined by ADPase flat-embedding.
The obese BBZ/Wor rat, a new animal model for noninsulin dependent diabetes mellitus, has been developed by crossing the BBDP/Wor rat, a lean, lymphopenic animal with autoimmune, insulin dependent diabetes mellitus, with the Zucker fatty rat, an obese animal with insulin resistance and glucose intolerance. Obese BBZ/Wor males become diabetic more frequently and at an earlier age than obese females, suggesting a role for gender in the pathogenesis of this type of diabetes. The obese BBZ/Wor rat is lymphopenic, hyperinsulinemic with peripheral insulin-resistance and develops spontaneous autoimmune noninsulin dependent diabetes mellitus at a mean age of 70 days. Lymphocytic insulitis is mild and beta-cell destruction is incomplete with hyperplasia of existing beta cells allowing prolonged survival without insulin therapy.24 In the diabetic rats, urinary protein and albumin were significantly elevated over age-matched diabetes resistant controls after one month of diabetes and further increased with the duration of diabetes.28
/ 2b30 5134 Mp 116 Wednesday Nov 12 11:07 AM EL–FRB 5134
NADH oxidase activity in the rat
117
FIG. 5. Co-localization of serum albumin (arrowheads) and hydrogen peroxide (arrows) in retina of BBZ/Wor diabetic rat which had been diabetic for 10 months. Serum albumin localizes in cytoplasmic vesicles and tight junction (TJ) in association with hydrogen peroxide. Basement membrane (BM); endothelial cell (EC); red blood cell (RBC). 1 40,000.
Previous studies of oxidative injury in diabetes have demonstrated serum levels of free radicals, conjugated dienes, and lipid peroxidation products, all of which are measures of free radical mediated processes. The biochemical studies of Murata et al. demonstrated the involvement of NADH in lipid peroxidation in the retina of rats with alloxan-induced diabetes.29 Nishida et al. showed that superoxide reacted with unsaturated fatty acids in the retina of rats with streptozotocin-induced diabetes resulting in increased levels of lipid peroxidation products and a decrease in levels of the free radical scavenger, SOD.30 Armstrong and Al1–Wadi observed a time-dependent increase in lipid hydroperoxides which correlated with focal damage to the photoreceptor layer of the retina of rats with streptozotocininduced diabetes.31 The above studies involved an aggressive, chemically induced, insulin dependent diabetes. Diabetic retinopathy is a disease process which affects predominantly the central retina. The semiquantitative data for peroxide localization indicate that the changes observed in this animal model are consistent with the pattern observed in patients with diabetic retinopathy. The cerium NADH oxidase enzyme activity technique shows actual sites of peroxide generation.
Semiquantitation of amounts of cerium reaction product, a direct indication of oxidase activity, shows a significant increase of NADH oxidase enzyme activity in diabetic animals as compared to nondiabetic controls. Elevated ROS can occur as a consequence of cellular damage and inflammation, which may be contributing to the increased NADH oxidase activity observed in the retinas of the diabetic rats. The studies of Murata et al.29 indicated that NADH plays an important role in lipid peroxidation in retina with a 40% increase in the rate of peroxidation in diabetic retina as compared to nondiabetic retina. There is increased cytosolic free NADH/NAD/ in tissues exposed to elevated glucose levels at normal tissue pO2 . This redox imbalance has been linked to increased metabolism of glucose via the sorbitol pathway in retina and other tissues subject to vascular injuries of diabetic hyperglycemia.13 Retina is one of the most highly oxygenated tissues in the body. In the presence of oxygen and an abundance of the substrate NADH, NADH oxidase enzyme activity increases superoxide production in the retina. More recent studies have validated that NADH oxidase enzyme activity is the major generator of superoxide in vascular endothelial cells.32,33
/ 2b30 5134 Mp 117 Wednesday Nov 12 11:07 AM EL–FRB 5134
118
E. A. ELLIS et al.
FIG. 6. (A) Co-localization of fibronectin with 18 nm colloidal gold particles (arrowheads) with hydrogen peroxide (arrows) in retina of BBZ/Wor diabetic rat which had been diabetic for 10 months. Basement membrane (BM); endothelial cell (EC); vessel lumen (L); red blood cell (RBC). There are increased levels of fibronectin as compared to the age matched, nondiabetic control in Fig. 6B. 1 40,000; (B) Localization of fibronectin in retina of an age matched, nondiabetic control rat. Fibronectin is confined to the vessel lumen. Basement membrane (BM); endothelial cell (EC); vessel lumen (L). 1 40,000.
Recent studies indicate that disruption of the blood-retinal barrier is mediated by an increase in transendothelial transport of serum proteins through transcytotic vesicles.34,35 Studies of binding and uptake of native and glycosylated albumin-gold complexes in perfused rat lungs showed enhanced binding, internalization and transport of glycosylated albumin.36 Morphometric studies of capillary abnormalities in sural nerves of patients with diabetic neuropathy showed increased endothelial cell area and decreased lumen area in capillaries where the endothelial cell appeared to bulge into the lumen. A reduction in lumen size would be expected to contribute to capillary closure, reduced blood flow, and increased hypoxia.37 Increased vascular permeability is an early complication of diabetic retinopathy. Most studies of retinal dysfunction in diabetic retinopathy have utilized fluorescein angiography. The post embedding procedure presented here offers several advantages in
that there is no need to inject or infuse nonphysiological tracers and there is the improved resolution of TEM. In addition, this post embedding procedure on sections allowed correlative studies after cytochemical localization procedures. Studies by Bendayan et al. demonstrated leakage of endogenous serum albumin in the glomerular wall of streptozotocin-induced diabetes.38 A similar immunocytochemical technique has been used to study bloodretinal barrier disruption in cadaveric eyes from patients with diabetic retinopathy.39 Treatment of confluent monolayers of human umbilical vein endothelial cells with sublethal
Table 3. Immunocytochemical Localization of Fibronectin Colloidal Gold Particles/um2 Diabetic (BBZ/Wor) Non-Diabetic (BBDR/Wor)
/ 2b30 5134 Mp 118 Wednesday Nov 12 11:07 AM EL–FRB 5134
4.45 { 1.09 1.56 { 0.79
NADH oxidase activity in the rat
concentrations of hydrogen peroxide resulted in partial redistribution of actin filaments from the cell periphery and cell retraction that opened gaps between cells.40 There was also redistribution of endothelial cell adhesion molecules. These in vitro changes were interpreted as manifestations of sublethal hydrogen peroxide injury which could cause endothelial dysfunction.40 Hyperglycemic activation of protein kinase C is associated with increased vascular permeability in diabetic retinopathy.41 Phorbol myristate acetate which activates protein kinase C induced endocytosis and exocytosis ( transcytosis ) in toad bladder with cellular profiles of plasma membrane associated cytoplasmic vesicles similar to those shown here.42 More recent studies showed oxygen radical activation of protein kinase C and increased transendothelial permeability in lung inflammation.43 In addition to extravasation of serum albumin, our immunocytochemical studies of disruption of the blood-retinal barrier (serum albumin leakage) showed leakage which correlated with hydrogen peroxide localization. Our morphological findings are consistent with a possible oxygen radical activation of protein kinase C and increased vascular permeability through transcytotic vesicles as demonstrated by other studies of disruption of the blood-retinal barrier 39 and bloodnerve barrier 37 in diabetes. Fibronectin synthesis and release into the circulation has been documented as a result of oxidantinduced vascular injury in lung.44 Increased synthesis of fibronectin in retinal vessels from diabetic patients has been documented.45 The studies shown here show elevation of total fibronectin and extravasation of fibronectin in diabetic rat eyes at sites of oxidative injury ( peroxide localization ) as compared to age matched controls. Previous studies of oxidative injury in diabetic retinopathy involved insulin dependent models and focused on the role of activated leukocytes in initiating endothelial cell dysfunction. In this study, increased numbers of leukocytes were not observed in the retinas of the BBZ/Wor diabetic rats compared to control rats. This may be due to the fact that this noninsulin dependent model has a milder diabetes with an attenuated disease coarse as compared to insulin dependent diabetes. Porta proposed that since the endothelium is strategically situated in the vessel wall, it plays a major role in remodeling the retinal microvasculature in diabetes.46 This report is the first in vivo demonstration that the endothelial cell in diabetic retinopathy may play a role in the initiation of oxidative injury. This oxidative injury may contribute to the observed dysfunction of the blood-retinal barrier. Cytochemical lo-
119
calization of NADH oxidase enzyme activity as presented here supports a role for hyperglycemia in development of free radical-associated vascular complications in diabetes. REFERENCES 1. Pugliese, G.; Tilton, R. G.; Williamson, J. R. Glucose-induced metabolic imbalances in the pathogenesis of diabetic vascular disease. Diabetes Metab. Rev. 7:35–39; 1991. 2. Ruderman, N. B.; Williamson, J. R.; Brownlee, M. Glucose and diabetic vascular disease. FASEB J. 6:2905–2914; 1992. 3. Collier, A.; Rumley, A.; Rumley, A. G.; Paterson, J. R.; Leach, J. P.; Lowe, G. D. O.; Small, M. Free radical activity and hemostatic factors in NIDDM patients with and without microalbuminuria. Diabetes 41:909–913; 1992. 4. Oberly, L. W. Free radicals and diabetes. Free Rad. Biol. Med. 5:113–124; 1988. 5. Baynes, J. W. Role of oxidative stress in development of complications in diabetes. Diabetes 40:405–412; 1991. 6. Ceriello, A.; Giugliano, D.; Quatraro, A.; Dello Russo, P.; Lefebvre, P. J. Metabolic control may influence the increased superoxide generation in diabetic serum. Diab. Med. 8:540–542; 1991. 7. Tesfamariam, B. Free radicals in diabetic endothelial cell dysfunction. Free Rad. Biol. Med. 16:383–391; 1994. 8. Weiss, S. J.; Young, J.; LoBuglio, A. F.; Slivka, A.; Mineh, N. F. Role of hydrogen peroxide in neutrophil-mediated destruction of cultured endothelial cells. J. Clin. Invest. 68:714–721; 1981. 9. Wieruz–Wysocka, B.; Wysocky, H.; Siekirka, H.; Wykretowicz, A.; Szczepanik, A.; Klimas, R. Evidence of polymorphonuclear neutrophil activation in patients with insulin-dependent diabetes mellitus. J. Leuk. Biol. 42:519–523; 1987. 10. Schroder, S.; Palinski, W.; Schmid–Schonbein, G. W. Activated monocytes and granulocytes, capillary nonperfusion, and neovascularization in diabetic retinopathy. Am. J. Path. 139:81– 100; 1991. 11. Freedman, S. F.; Hatchell, D. L. Enhanced superoxide radical production by stimulated polymorphonuclear leukocytes in a cat model of diabetes. Exp. Eye Res. 55:767–773; 1992. 12. Rosen, G. M.; Freeman, S. A. Detection of superoxide generated by endothelial cells. Proc, Natl. Acad. Sci. USA 81:7269–72732; 1984. 13. Williamson, J. R.; Chang, K.; Frangos, M.; Hasan, K. S.; Ido, Y.; Kawamura, T.; Nyengaard, J. R.; van den Enden, M.; Kilo, C.; Tilton, R. G. Hyperglycemic pseudohypoxia and diabetic complications. Diabetes 42:801–813; 1993. 14. Bravenboer, B.; Kappelle, A. C.; Hamers, F. P. T.; van Buren, T.; Erkelens, D. W.; Gispen, W. H. Potential use of glutathione for the prevention and treatment of diabetic neuropathy in the streptozotocin-induced diabetic rat. Diabetologia 35:813–817; 1992. 15. Cameron, N. E.; Cotter, M. A.; Maxfield, E. K. Anti-oxidant treatment prevents the development of peripheral nerve dysfunction in streptozotocin-diabetic rats. Diabetologia 36:299–304; 1993. 16. Zhang,H.; Agardh, E.; Agardh, C-D. Nitro blue tetrazolium staining — A morphological demonstration of superoxide in the rat retina. Graefe’s Arch. Clin. Exp. Ophthalmol. 231:178–183; 1993. 17. Briggs, R. T.; Karnovsky, M. L.; Karnovsky, M. J. Localization of NADH oxidase enzyme activity on the surface of human polymorphonuclear leukocytes by a new cytochemical method. J. Cell Biol. 67:566–586; 1975. 18. Karnovsky, M. J. Cytochemistry and reactive oxygen species. Histochemistry 102:15–27; 1994. 19. Warren, J. S.; Kunkel, R. G.; Simon, R. H.; Johnson, K. J.; Ward, P. A. Ultrastructural cytochemical analysis of oxygen radicalmediated immunoglobulin A immune complex induced lung injury in the rat. Lab. Invest. 60:641–658; 1989.
/ 2b30 5134 Mp 119 Wednesday Nov 12 11:07 AM EL–FRB 5134
120
E. A. ELLIS et al.
20. Guy, J.; Ellis, E. A.; Rao, N. A. Hydrogen peroxide localization in experimental optic neuritis. Arch. Ophthalmol. 108:1614– 1621; 1990. 21. Shlafer, M.; Brosamer, K.; Forder, J. R.; Simon, R. H.; Ward, P. A.; Grum, C. M. Cerium chloride as a histochemical marker of hydrogen peroxide in reperfused ischemic hearts. J. Mol. Cardiol. 22:83–97; 1990. 22. Soares, F. A.; Shaughnessy, S. G.; MacLarkey, W. R.; Orr, F. W. Quantification and morphological demonstration of reactive oxygen species produced by Walker 256 tumor cells in vitro and during metastasis in vivo . Lab. Invest. 71:480–489; 1994. 23. Bendayan, M. Use of the protein A-gold technique for the morphological study of vascular permeability. J. Histochem. Cytochem. 28:1251–1254; 1980. 24. Guberski,D.L.; Butler, L.; Manzi, S. M.; Stubbs, M.; Like, A. A. The BBZ/Wor rat: Clinical characteristics of the diabetic syndrome. Diabetologia 36:912–919; 1993. 25. Murray, F. T.; Beyer–Mears, A.; Johnson, R. D.; Sima, A. A. F.; Cameron, D. F.; Sninsky, C. A.; Selawry, H. Assessment of proteinuria and neuropathy in the nonimmunosupressed BB diabetic rat after abdominal intratesticular islet transplantation. Transplantation 56:680–686; 1993. 26. Briggs, R. T.; Karnovsky, M. L.; Karnovsky, M. J. Hydrogen peroxide in chronic granulomatous disease: A cytochemical study of reduced pyridine nucleotide oxidases. J. Clin. Invest. 59:1088–1098; 1977. 27. Lutty, F. A.; McLeod, D. S. A new technique for visualization of the human retinal vasculature. Arch Ophthalmol. 110:2:267– 276; 1992. 28. Murray, F. T.; Wachowski, M. B.; Dani, A.; Ellis, E. A.; Grant, M. B. Intermediate and long term diabetic (Type II) complications in the spontaneously diabetic BBZ/Wor rat. Diabetes 45 suppl. 2:272A abstr. (1996). 29. Murata, T.; Nishida, T.; Eto, S.; Mukai, N. Lipid peroxidation in diabetic rat retina. Metab. Ped. Ophthalmol. 5:83–87; 1981. 30. Nishida, T.; Nakagawa, S.; Manabe, R. Superoxide dismutase activity in diabetic rat retina. Jpn. J. Ophthalmol. 28:377–382; 1984. 31. Armstrong, D.; Al–Awadi, F. Lipid peroxidation and retinopathy in streptozocin-induced diabetes. Free Rad. Biol. Med. 11:433–436; 1991. 32. Mohazzab, K. M.; Kaminski, P. M.; Wolin, M. S. NADH oxidoreductase is a major source of superpoxide anion in bovine coronary artery endothelium. Am. J. Physiol. 266:H2568– H2572; 1994. 33. Rajagopalan, S.; Kurz, S.; Munzel, T.; Tarpey, M.; Freeman, B. A.; Griendling, K. K.; Harrison, D. G. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation. J. Clin. Invest. 97:1916–1923; 1996.
34. Lin, S.-J.; Hong., C.-Y.; Chang, M.-S.; Chiang, B. N.; Chien, S. Increased aortic endothelial cell death and enhanced transendothelial macromolecular transport in streptozotocin-diabetic rats. Diabetologia 36:926–930; 1993. 35. Derevjanik, N. L.; Wilson, C. A.; Berkowitz, B. A.; Malow, J.; Vinores, S. A. Blood-retinal barrier (BRB) breakdown in diabetic rabbits assessed by electron microscopic immunocytochemistry and magnetic resonance imaging (MRI). Invest. Ophthalmol. Vis. Sci. 37(suppl.): S968 abstr. (1996). 36. Villaschi, S.; Johns, L.; Cirigliano, M.; Pietra, G. G. Binding and uptake of native and glycosylated albumin-gold complexes in perfused rat lung. Microvascular Res. 32:190–199; 1986. 37. Malik, R. A.; Tesfaye, S.; Thompson, S. D.; Veves, A.; Hunter, A.; Sharma, A. K.; Ward, J. D.; Boulton, A. J. M. Transperineurial capillary abnormalities in the sural nerve of patients with diabetic neuropathy. Microvascular Res. 48:236–245; 1994. 38. Bendayan, M.; Gingras, D.; Charest, P. Distribution of endogenous albumin in the glomerular wall of streptozotocin-induced diabetic rats as revealed by high resolution immunocytochemistry. Diabetologia 29:868–875; 1986. 39. Vinores, S. A.; Gadegbeku, C.; Campochiaro, P. A.; Green, W. R. Immunohistochemical localization of blood-retinal barrier breakdown in human diabetics. Am. J. Pathol. 134:231–235; 1989. 40. Bradley, J. R.; Thiru, S.; Pober, J. S. Hydrogen peroxide-induced endothelial retraction is accompanied by a loss of the normal spatial organization of endothelial cell adhesion molecules. Am. J. Pathol. 147:627–641; 1995. 41. Lee, T-S.; MacGregor, L. C.; Fluharty, s. J.; King, G. L. Differential regulation of protein kinase C and (Na,K)-adenosine triphosphatase activities by elevated glucose levels in retinal capillary endothelial cells. J. Clin. Invest. 83:90–94; 1989. 42. Masur, S. K.; Sapirstein, V.; Rivero, D. Phorbol myristate acetate induces endocytosis as well as exocytosis and hydroosmosis in toad urinary bladder. Biochim. Biophys. Acta 821:286–296; 1985. 43. Lynch, J. J.; Ferro, T. J.; Blumenstock, F. A.; Brockenauer, A. M.; Malik, A. B. Increased endothelial albumin permeability mediated by protein kinase C activation. J. Clin. Invest. 85:1991–1998; 1990. 44. Peters, J. H.; Ginsberg, M. H.; Bohl, B. P.; Sklar, L. A.; Cochrane, C. G. Intravascular release of intact cellular fibronectin during oxidant-induced injury of the in vitro perfused rabbit lung. J. Clin. Invest. 78:1596–1603; 1986. 45. Roy, S.; Cagliero, E.; Lorenzi, M. Fibronectin over expression in retinal microvessels of patients with diabetes. Invest. Ophthalmol. Vis. Sci. 37:258–266; 1996. 46. Porta, M. Endothelium: The main actor in the remodeling of the retinal microvasculature in diabetes. Diabetologia 39:739–744; 1996.
/ 2b30 5134 Mp 120 Wednesday Nov 12 11:07 AM EL–FRB 5134