Receptor for Advanced Glycation End Products Is Involved in Impaired

Original Article Receptor for Advanced Glycation End Products Is Involved in Impaired Angiogenic Response in Diabetes Takuhito Shoji,1 Hidenori Koyama,1 Tomoaki Morioka,1 Shinji Tanaka,1 Akane Kizu,1 Kohka Motoyama,1 Katsuhito Mori,1 Shinya Fukumoto,1 Atsushi Shioi,2 Noriko Shimogaito,3 Masayoshi Takeuchi,3 Yasuhiko Yamamoto,4 Hideto Yonekura,4 Hiroshi Yamamoto,4 and Yoshiki Nishizawa1

Angiogenic response is impaired in diabetes. Here, we examined the involvement of receptor for advanced glycation end products (RAGE) in diabetes-related impairment of angiogenesis in vivo. Angiogenesis was determined in reconstituted basement membrane protein (matrigel) plugs containing vascular endothelial growth factor (VEGF) implanted into nondiabetic or insulin-deficient diabetic wild-type or RAGEⴚ/ⴚ mice. The total, endothelial, and smooth muscle (or pericytes) cells in the matrigel were significantly decreased in diabetes, with the regulation dependent on RAGE. In the matrigel, proangiogenic VEGF expression was decreased, while antiangiogenic thrombospondin-1 was upregulated in diabetic mice, regardless of the presence of RAGE. In wild-type mice, proliferating cell nuclear antigen (PCNA)-positive cells in the matrigel were significantly less in diabetic than in nondiabetic mice, while the numbers of transferase-mediated dUTP nick-end labeling (TUNEL)-positive cells were significantly higher. This alteration in PCNA- and TUNEL-positive cells in diabetes was not observed in RAGEⴚ/ⴚ mice. Similarly, the percentage of nuclear factor ␬B–activated cells is enhanced in diabetes, with the regulation dependent on the presence of RAGE. Importantly, adenovirus-mediated overexpression of endogenous secretory RAGE, a decoy receptor for RAGE, restores diabetes-associated impairment of angiogenic response in vivo. Thus, RAGE appears to be involved in impairment of angiogenesis in diabetes, and blockade of RAGE might be a potential therapeutic target. Diabetes 55: 2245–2255, 2006

T

he morbidity and mortality of diabetes are due to the development of both macrovascular and microvascular complications (1). Progressive vasodegeneration in microvascular beds is the major underlying factor in initiation and progression of diabetic vascular complications (2– 4). In addition, the phenomenon of impaired new vessel growth in the diabetic state is well recognized (5– 8). However, the mechanisms by which diabetes could limit the formation of new blood vessels remain largely undefined. One consequence of long-term hyperglycemia is the formation of advanced glycation end products (AGEs); the accumulation of AGEs in the vessel wall has been implicated in the pathogenesis of diabetes complications (1). Among a variety of AGE receptor or AGE-binding proteins that have been described, the receptor for AGEs (RAGE) is probably the best-characterized molecule. RAGE belongs to the immunoglobulin superfamily of cell surface molecules to which AGEs bind (9,10). Ligation of RAGE in endothelial cells activates the transcription factor nuclear factor-␬B (NF-␬B), subsequently leading to increased expression of proatherogenic mediators, such as monocyte chemoattractant protein-1 or vascular cell adhesion molecule-1 (11,12). Recently, RAGE has been shown to be involved in both microdiabetic (13) and macrodiabetic (14,15) vascular complications. The angiogenic role of AGEs in vitro remains somewhat controversial: antiangiogenic (16) or proangiogenic (17). Some reports showed the involvement of AGEs in impaired angiogenic response in diabetic animals in vivo; inhibition of AGE formation in diabetic mice is shown to restore ischemia-induced angiogenesis in peripheral limbs (18), and AGEs inhibition with soluble RAGE can restore angiogenic potential during wound healing in diabetic mice (19). However, the role of RAGE in altered angiogenic response in diabetes has not been reported. In this article, using RAGE-deficient mice, we examined if RAGE underlays impaired angiogenic response in diabetes. We show that presence of RAGE is essential for impaired angiogenic response in diabetes. We further show that functional blockade of RAGE with a decoy receptor, endogenous secretory RAGE (esRAGE), can restore suppressed angiogenic response in diabetes.

From the 1Department of Metabolism, Endocrinology and Molecular Medicine, Osaka City University Graduate School of Medicine, Osaka, Japan; the 2 Department of Cardiovascular Medicine, Osaka City University Graduate School of Medicine, Osaka, Japan; the 3Department of Pathophysiological Science, Faculty of Pharmaceutical Sciences, Hokuriku University, Kanazawa, Japan; and the 4Department of Biochemistry and Molecular Vascular Biology, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan. Address correspondence and reprint requests to Hidenori Koyama, MD, PhD, Department of Metabolism, Endocrinology and Molecular Medicine, (Second Department of Internal Medicine), Osaka City University Graduate School of Medicine, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan. E-mail: [email protected]. Received for publication 23 October 2005 and accepted in revised form 17 May 2006. ␣SM actin, ␣ smooth muscle actin; Ad-esRAGE, pAdHM15-esRAGE; AGE, advanced glycation end product; CML, N-carboxymethyllysine; esRAGE, endogenous secretory RAGE; MMP, matrix metalloproteinase; NF-␬B, nuclear factor-␬B; PCNA, proliferating cell nuclear antigen; RAGE, receptor for AGE; STZ, streptozotocin; TSP, thrombospondin; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling; VEGF, vascular endothelial growth factor. DOI: 10.2337/db05-1375 © 2006 by the American Diabetes Association.

RESEARCH DESIGN AND METHODS

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.

The RAGE targeting construct and the generation of RAGE–/– mice is described elsewhere (20). Animals used in the described experiments were

DIABETES, VOL. 55, AUGUST 2006

2245

RAGE AND IMPAIRED ANGIOGENESIS IN DIABETES

TABLE 1 Body weights and blood glucose levels of the mice

Body weight (g) Blood glucose (mg/dl) Glucose-derived AGEs (units/ml) CML (units/ml)

Nondiabetic wild-type

Diabetic wild-type

Nondiabetic RAGE⫺/⫺

Diabetic RAGE⫺/⫺

23.7 ⫾ 3.2 161.1 ⫾ 37.8 1.11 ⫾ 0.59 0.20 ⫾ 0.14

18.9 ⫾ 4.3* 463.1 ⫾ 65.3* 2.68 ⫾ 0.38* 0.56 ⫾ 0.32

22.5 ⫾ 2.9 139.9 ⫾ 42.8 0.70 ⫾ 0.38 0.25 ⫾ 0.10

19.0 ⫾ 2.2* 457.5 ⫾ 81.0* 2.20 ⫾ 1.09* 0.42 ⫾ 0.30

Data are means ⫾ SD. Body weight and postprandial blood glucose were measured just before the implantation of the matrigel. Glucose-derived AGEs and CML were measured when the mice were killed. *P ⬍ 0.05 vs. control, Student’s t test. from F2 or F5 backcrosses onto the C57BL/6J genetic background. C57BL/6J wild-type and RAGE⫺/⫺ mice were produced by mating C57BL/6J RAGE⫹/⫺ mice. Mice were weaned at age 4 –5 weeks and housed in conventional cages. Mice were fed a pelleted standard rodent diet. Mice were maintained in a temperature-controlled (24°C) facility with a strict 12-h light/dark cycle and given free access to food and water. Procedures in this study were approved by the animal care and use committee at the Osaka City University Graduate School of Medicine, Osaka, Japan. Experimental protocols. For the induction of insulin-deficient diabetes, mice were given 60 mg/kg of streptozotocin (STZ; Sigma, St. Louis, MO) via intraperitoneal injection at age 6 weeks for 6 days as previously described (21). For nondiabetic controls, mice were injected intraperitoneally with solvent alone (0.05 mol/l sodium citrate buffer, pH 4.5). Wild-type or RAGE⫺/⫺ mice were randomly divided into diabetic and nondiabetic groups to produce four groups: a nondiabetic wild-type, a diabetic wild-type, a nondiabetic RAGE⫺/⫺, and a diabetic RAGE⫺/⫺ group. In this experimental protocol, plasma glucose was promptly increased at age 7 weeks after STZ injection (data not shown), and serum AGE level was increased when the mice were killed (Table 1). Determination of AGE concentration. Serum N-carboxymethyllysine (CML) and glucose-derived AGEs were differentially determined using a competitive enzyme-linked immunosorbent assay as described (22). One unit per milliliter of CML or glucose-derived AGEs corresponded to a protein concentration of 1 ␮g/ml CML-BSA or glucose-derived AGE-BSA, respectively. In vivo matrigel assay. Angiogenic response in nondiabetic and STZ-induced diabetic wild-type or RAGE⫺/⫺ mice was evaluated through the matrigel assay as described previously (23). Briefly, reconstituted basement membrane proteins, matrigel solution (BD Biosciences), was supplemented with 300 ng/ml vascular endothelial growth factor (VEGF; Genzyme Techne, Minneapolis, MN) and injected subcutaneously into the abdominal midline of wild-type or RAGE⫺/⫺ mice at age 13 weeks, where it polymerized to form a plug. The plug was removed 2 weeks later and processed for histology analysis or for total RNA isolation. For histochemical analyses, it was fixed in a zinc fixative (BD Biosciences) for 24 – 48 h at room temperature and sequentially in 4% paraformaldehyde overnight at 4°C. It was then embedded in paraffin. All tissues were sectioned (5-␮m thickness) and mounted onto slides. The angiogenic response in the matrigel was evaluated with hematoxylin and eosin staining using standard techniques or immunohistochemistry with cell type– specific antibodies. Total, CD31-positive, ␣ smooth muscle actin (␣SM actin)positive, or desmin-positive cell numbers were counted in the five sequential 200 ⫻ 300 –␮m fields adjacent to the fascia. Immunohistochemical staining and TUNEL in situ labeling. The following antibodies were used as the first antibodies for immunohistochemistry: rat anti-CD31 antibody (550274, dilution 1:100; BD Pharmingen, San Diego, CA) for staining vascular endothelial cells, mouse monoclonal anti–␣SM actin antibody (clone 1A4, dilution 1:800; Sigma) or anti-desmim antibody (dilution 1:1,000; DakoCytomation, Carpinteria, CA) to identify smooth muscle cell or pericyte, goat anti-mouse RAGE antibody (1:20 dilution; R&D systems), anti-VEGF antibody (clone sc-7269, dilution 1:100; Santa Cruz Biotechnology, Santa Cruz, CA), anti–thrombospondin (TSP)-1 antibody (Ab4, dilution 1:50; NeoMarkers, Fremont, CA), anti–angiopoietin-1 antibody (dilution 1:200; Alpha Diagnostic International, San Antonio, TX), anti–angiopoietin-2 antibody (dilution 1:200; Alpha Diagnostic International), anti–Mac-3 antibody (dilution 1:10; BD Pharmingen), the rabbit polyclonal antibody against the COOH-terminal 16 amino acids of human esRAGE (dilution 1:1,000) not detecting full-length RAGE (24), and anti-proliferating cell nuclear antigen (PCNA) antibody (clone PC10, dilution 1:40; Oncogene Research Products, San Diego, CA). Paraffin-embedded implants were sectioned at 5 ␮m thickness. Following deparaffinization and rehydration of sections, endogenous peroxidase activity was blocked using 0.3% hydrogen peroxide before application of blocking serum. Sections were then incubated with primary antibody. Immunohistochemical staining was performed by the aridin-biotin complex method (ABC Kit; Vector Laboratories, Burlingame, CA). Color was 2246

developed with 3,3⬘-diaminobenzidine tetrahydrochloride. Sections stained by omitting the primary antibodies were used as negative controls. NF-␬B activation in cells in the matrigel was analyzed by the direct fluorescent immunohistochemistry. The sections were incubated with rhodamine-conjugated anti–NF-␬B p65 antibody (F6, dilution 1:50; Santa Cruz Biotechnology), and the nuclei were counterstained with Haechist 33258 dye. The cells with the nuclei stained with p65 were considered NF-␬B–activated cells. The in situ apoptosis detection kit (Takara, Tokyo, Japan) was used to detect apoptotic cells in the matrigel plugs by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay. The sections were also couterstained with 3% methyl green. Gene expression analyses with real-time RT-PCR. Total RNA was extracted from matrigel plugs using Trizol. cDNA was synthesized by TaqMan reverse transcription reagents, and expressions of mouse VEGF, basic fibroblast growth factor, hepatocyte growth factor, angiopoietin-1, angiopoietin-2, TSP-1, pigment epithelium– derived factor, tumor necrosis factor ␣, plateletderived growth factor, tissue factor, interleukin-6, and matrix metalloproteinase (MMP) 2 and 9 mRNA were measured quantitatively by the TaqMan real-time RT-PCR technique (Applied Biosystems, Foster City, CA). Ready-touse primers and fluorescence probes for each of the mouse gene and 18S ribosomal RNA were purchased from Applied Biosystems (Assay on Demand) and used according to the manufacturer’s protocol. The Ct value for every sample was measured, and mRNA expression levels of each mouse gene were determined by a comparative Ct method using 18S ribosomal RNA as endogenous reference (Applied Biosystems). Construction and generation of recombinant adenovirus vectors. Human esRAGE-expressing pAdHM15-esRAGE (Ad-esRAGE) was constructed using the in vitro ligation method as previously described (25,26). In brief, a full-length human esRAGE cDNA (24) was first inserted into the shuttle plasmid, pHMCMV6, and was subcloned into the pAdHM15, using unique I-CeuI and PI-SceI sites in the E1 deletion region. Ad-esRAGE vector was transfected to human embryonic kidney 293 cells using Superfect transfection reagent (Qiagen) according to the manufacturer’s protocol. Recombinant adenovirus expressing ␤-gal (Ad-LacZ) was used as a control. Adenovirus-mediated esRAGE gene transfer in vivo. Control or insulindeficient diabetic C57BL/6J mice (The Jackson Laboratories, Bar Harbor, ME) were used to examine the effect of esRAGE overexpression in vivo. VEGF (300 ng/ml) and either Ad-esRAGE or Ad-LacZ were suspended in the matrigel (5 ⫻ 108 pfu/ml), and the matrigel were implanted subcutaneously into the abdominal midline of mice at age 13 weeks. The plug was removed 2 weeks later and processed for histochemical analysis. Statistical analysis. Data are presented as means ⫾ SD. Statistical analysis was done by using the Student’s t test or ANOVA. These analyses were carried out using Stat View V software. Differences were considered significant when the P value was ⬍0.05.

RESULTS

RAGE deficiency restores impaired angiogenic response in diabetic mouse. Table 1 shows body weights and postprandial blood glucose levels of four groups of mice (nondiabetic wild-type, diabetic wild-type, nondiabetic RAGE⫺/⫺, and diabetic RAGE⫺/⫺). In both the wildtype and RAGE⫺/⫺ mice, the diabetic group showed significantly lower body weight and higher blood glucose level than the nondiabetic control group. There were no significant differences in body weight and blood glucose between wild-type and RAGE⫺/⫺ mice both in nondiabetic and diabetic conditions. To determine the levels of circulating AGEs, serum was differentially measured by competitive enzyme-linked immunosorbent assay for CML DIABETES, VOL. 55, AUGUST 2006

T. SHOJI AND ASSOCIATES

FIG. 1. RAGE immunostaining of the implanted matrigel or adjacent fascia in RAGEⴙ/ⴙ or RAGEⴚ/ⴚ mice. CD31 or ␣SM actin was also immunostained in adjacent sections to identify endothelial cells or smooth muscle–like cells. Bars represent 50 ␮m.

(derivative of AGEs mainly formed by peroxidation) and for glucose-derived AGEs (22). Serum glucose-derived AGE levels were significantly higher in diabetic than nondiabetic mice, regardless of the presence or absence of RAGE (Table 1). There was a tendency of higher serum CML levels in the diabetic groups than in the nondiabetic groups but without significant differences (Table 1). Expression of RAGE in wild-type or RAGE⫺/⫺ mice was confirmed by immunohistochemistry (Fig. 1). In wild-type mice, RAGE was expressed in the vessels outside of the matrigel and in cells migrated into matrigel. Staining of CD31-positive endothelial cells and ␣SM actin–positive smooth muscle–like cells in adjacent sections revealed that RAGE was expressed in both of these cell types. As expected, these RAGE-positive cells were not observed in RAGE⫺/⫺ mice. To examine the angiogenic response, we analyzed the numbers of total, endothelial (CD31 positive), and smooth muscle or pericytic cells in the matrigel (Fig. 2). Because of the heterogeneity of the marker expression for smooth muscle cell or pericyte (27), we used two representative markers in immuonohistochemistry: ␣SM actin and desmin. In wild-type mice, the total cell number in the matrigel of the STZ-treated diabetic mice was significantly decreased compared with that of the nondiabetic mice. In sharp contrast in RAGE⫺/⫺ mice, the total cell number of diabetic mice was not significantly different from that of the nondiabetic mice. The CD31-positive, ␣SM actin– positive, or desmin-positive cell number in the matrigel of the diabetic wild-type mice was also significantly less than that of the nondiabetic wild-type mice. Whereas, neither the number of CD31-positive, ␣SM actin–positive, nor desmin-positive cells was significantly different compared with nondiabetic RAGE⫺/⫺ mice. The numbers of ␣SM actin–positive cells and desmin-positive cells showed DIABETES, VOL. 55, AUGUST 2006

weak, but significant, correlation (r ⫽ 0.421, P ⫽ 0.0085) even though the adjacent sections were used for the analyses. This result is consistent with the idea that expression of the markers for smooth muscle cell/pericyte could be heterogeneous. Taken together, the angiogenic response to VEGF is suppressed in diabetic mice, with the impairment dependent on the presence of RAGE. RAGE is not principally involved in diabetes modulation of the mRNA or protein expression of genes in the matrigel involved in the angiogenic response. Next, we investigated whether RAGE was involved in diabetes-mediated regulation of genes in the matrigels involved in angiogenic response (Table 2). Expression of proangiogenic VEGF mRNA of the STZ-treated diabetic wild-type mice was significantly decreased compared with that of the nondiabetic wild-type mice. Presence of diabetes also significantly suppressed the expression of VEGF mRNA in RAGE⫺/⫺ mice. Other proangiogenic factors, including basic fibroblast growth factor, hepatocyte growth factor, angiopoietin-1, and angiopoietin-2, did not show significant changes by diabetes, regardless of the presence or absence of RAGE. In sharp contrast to VEGF, the levels of antiangiogenic TSP-1 mRNA of the diabetic wild-type mice was significantly higher than those of the nondiabetic RAGE⫹/⫹ mice. RAGE deficiency, however, did not affect diabetes-mediated increase in the expression of TSP-1. mRNA levels of pigment epithelium growth factor were significantly lower in the diabetic than nondiabetic group regardless of the presence of RAGE. Tissue factor, a primary cellular initiator of blood coagulation, is recently shown to contribute to a variety of biological processes, including inflammation, angiogenesis, metastasis, and cell migration (28). In the present system, tissue factor mRNA was also suppressed by diabetes in both wild-type and RAGE⫺/⫺ mice. Interestingly, platelet-de2247


FIG. 2. Decrease in angiogenic response in diabetes is dependent on the presence of RAGE. A: Angiogenic response in nondiabetic and STZ-induced diabetic wild-type (WT) or RAGEⴚ/ⴚ mice was evaluated through the matrigel assay, which was supplemented with 300 ng/ml VEGF. The plug was removed 2 weeks after the implantation and processed for histology analysis. The angiogenic response in the matrigel was evaluated by quantifying neovascularization in sections stained with hematoxylin and eosin using standard techniques or immunostained with cell type–specific antibodies. Bars represent 50 ␮m. B: Summary of the quantitative analyses of angiogenic response (nondiabetic wild-type [N-WT]: n ⴝ 14, diabetic wild-type [D-WT]: n ⴝ 12, nondiabetic RAGEⴚ/ⴚ [N-RAGEⴚ/ⴚ]: n ⴝ 7, and diabetic RAGEⴚ/ⴚ [D-RAGEⴚ/ⴚ]: n ⴝ 7). Total, CD31-positive, ␣SM actin–positive, and desmin-positive cell numbers were counted in the five sequential 200 ⴛ 300 –␮m fields adjacent to the fascia. Data are shown as means ⴞ SD. *P < 0.05 vs. control, ANOVA with multiple comparison (Scheffe’s type).

rived growth factor gene, an essential factor involved in pericyte proliferation and migration (29), was also significantly downregulated in diabetic mice, with its suppression again being independent on the presence of RAGE. Inflammatory cytokines including tumor necrosis factor ␣ and interleukin-6 were not significantly altered in diabetic 2248

mouse. MMP-2 gene expression was slightly lower in diabetic mice, while in RAGE⫺/⫺ mice, no significant differences were observed. MMP-9 mRNA expression was not significantly altered by diabetes in both wild-type and RAGE⫺/⫺ mice. To further examine the altered gene regulation in diaDIABETES, VOL. 55, AUGUST 2006


TABLE 2 Gene expression of the factors involved in angiogenesis in the matrigel

VEGF Basic fibroblast growth factor Hepatocyte growth factor Angiotensin-1 Angiotensin-2 TSP-1 Pigment epithelium–derived factor Tumor necrosis factor-␣ Platelet-derived growth factor-B Tissue factor Interleukin-6 MMP2 MMP9

Nondiabetic wild-type

Diabetic wild-type

Nondiabetic RAGE⫺/⫺

Diabetic RAGE⫺/⫺

100.0 ⫾ 49.6 100.0 ⫾ 21.5 100.0 ⫾ 12.6 100.0 ⫾ 35.2 100.0 ⫾ 14.6 100.0 ⫾ 42.5 100.0 ⫾ 24.9 100.0 ⫾ 72.0 100.0 ⫾ 20.7 100.0 ⫾ 24.1 100.0 ⫾ 28.4 100.0 ⫾ 21.9 100.0 ⫾ 59.6

57.6 ⫾ 25.3* 89.2 ⫾ 18.5 91.2 ⫾ 22.9 80.4 ⫾ 34.9 78.8 ⫾ 23.3 282.6 ⫾ 109.4* 63.3 ⫾ 24.2* 124.3 ⫾ 66.5 77.3 ⫾ 21.8* 60.1 ⫾ 24.3* 87.6 ⫾ 48.5 74.4 ⫾ 32.4* 110.3 ⫾ 57.9

75.1 ⫾ 40.1 91.8 ⫾ 28.0 95.5 ⫾ 25.5 88.9 ⫾ 40.6 79.5 ⫾ 30.4 109.5 ⫾ 42.2 102.5 ⫾ 14.9 67.1 ⫾ 30.3 86.5 ⫾ 22.8 104.3 ⫾ 36.7 91.6 ⫾ 46.4 97.7 ⫾ 19.9 87.4 ⫾ 41.9

38.7 ⫾ 11.7* 74.0 ⫾ 9.0 95.8 ⫾ 20.8 67.8 ⫾ 28.8 83.8 ⫾ 25.9 261.4 ⫾ 83.9* 66.6 ⫾ 28.6* 64.5 ⫾ 24.8 60.7 ⫾ 23.3* 56.3 ⫾ 23.0* 86.6 ⫾ 66.4 92.2 ⫾ 32.2 107.1 ⫾ 36.0

Data are means ⫾ SD. Total RNA was isolated from the matrigel of nondiabetic wild-type (n ⫽ 11), diabetic wild-type (n ⫽ 12), nondiabetic RAGE⫺/⫺ (n ⫽ 9), and diabetic RAGE⫺/⫺ (n ⫽ 9) mice. Each mouse was implanted with two matrigel plugs, and these two plugs were pooled to isolate single preparation of total RNA. Expression levels of the genes were quantitated by real-time RT-PCR. Each value was expressed as the percentage of the average value of nondiabetic wild-type samples. *P ⬍ 0.05 vs. control, Student’s t test .

betes, we performed immunohistochemistry to clarify their protein expression. Neither the localization of VEGF, TSP-1, angiopoietin-1, nor angiopoietin-2 expression was markedly different between nondiabetic and diabetic or between wild-type and RAGE⫺/⫺ mice (Fig. 3A). The intensity of immunostaining for VEGF and angiopoietin-2 inside the matrigel was rather decreased in the diabetic mouse, regardless of the presence of RAGE. In accordance with the gene regulation, abundance of TSP-1 protein expression tends to be elevated inside the matrigel and in the fascia around the matrigel in diabetic mice. However, diabetic wild-type and diabetic RAGE⫺/⫺ mice showed similar TSP-1 staining. Since monocyte or macrophage is the major source or angiogenic growth factors, we also immunostained macrophage Mac-3 in these tissue sections (Fig. 3B). We found scattered localization of Mac-3 both inside the matrigel and in the fascia around the matrigel, with the numbers of cells not being significantly different between nondiabetic and diabetic or wild-type and RAGE⫺/⫺ mice. Taken altogether, regulation of multiple genes involved in angiogenesis appears to be altered in diabetes, with the majorities of regulation being independent of the presence of RAGE. Imbalance between proand antiangiogenic factors in diabetes was not strictly associated with local inflammation in this experimental system. Suppression of angiogenic response in diabetes is associated with decreased cell proliferation and increased cell death, which are restored by RAGE deficiency. To examine the potential mechanism underlying impaired angiogenic response in diabetes, we evaluated proliferating or dying cells by PCNA or TUNEL staining. As shown in Fig. 4, diabetic wild-type mice had significantly less PCNA-positive and higher TUNEL-positive cells than nondiabetic wild-type mice. However, in RAGE⫺/⫺ mice, there were no significant differences in PCNA- and TUNEL-positive cells between nondiabetic and diabetic mice. Thus, altered balance of cell proliferation and cell death could be involved in impaired angiogenic response in diabetes, with the alteration dependent on the presence of RAGE. NF-␬B is a major signal transduction pathway activated by RAGE and is involved in cellular activation and changes DIABETES, VOL. 55, AUGUST 2006

in cell phenotype. We therefore examined NF-␬B activation in cells in the matrigels. As shown in Fig. 4C, the percentage of cells with the nuclei stained with NF-␬B was significantly higher in diabetic wild-type mice than nondiabetic wild-type mice. Importantly, the percentage of cells with NF-␬B activation was not affected by diabetes in RAGE⫺/⫺ mice. Thus, altered cellular function in diabetes might be regulated through RAGE-mediated NF-␬B activation. Adenovirus-mediated overexpression of esRAGE restores diabetes-mediated impairment of angiogenic response in vivo. Recently, an esRAGE has been identified as a novel splice variant that directs the synthesis of RAGE proteins carrying all of the extracellular domains but devoid of the transmembrane and intracytoplasmic domains (24). esRAGE was found to be released outside from the cells, to bind AGEs, and to be capable of neutralizing AGE actions on endothelial cells in culture (24). We have recently reported that esRAGE is also detected in human plasma and is inversely associated with carotid atherosclerosis (30). Thus, we next examined if adenoviral overexpression esRAGE restores diabetes impairment of angiogenic response in vivo. Our adenoviral esRAGE construct (Ad-esRAGE) successfully and dose dependently increased amounts of the protein secreted into the culture media when infected to human smooth muscle cells in vitro (Fig. 5A). The major 50-kDa and minor 46-kDa bands correspond to the difference of Nglycosylation as previously described (24). When the AdesRAGE was included in the matrigel plug assay, abundant esRAGE expression was detected in the muscles and vessels adjacent to the matrigel compared with Ad-LacZ– infected mice (Fig. 5B). Similar to Fig. 2, in Ad-LacZ– infected mice, the total cell number in the matrigel was significantly less in diabetic than that of the nondiabetic mice (Fig. 6). In contrast in Ad-esRAGE–infected mice, the diabetes-mediated decrease in total cell number in the matrigel was significantly less. In accordance with the total cell number, decrease in both CD31- and ␣SM actin– positive cells in the matrigel in diabetic mice were significantly recovered by overexpression of esRAGE (Fig. 6), suggesting that esRAGE restores diabetes impairment of angiogenic response in vivo. 2249


FIG. 3. A: Immunohistochemistry of pro- and antiangiogenic factors in and outside of the matrigel. The matrigel was implanted to nondiabetic and diabetic wild-type or RAGEⴚ/ⴚ mice at 13 weeks of age. The matrigel was removed 2 weeks later and processed for histochemical analyses using antibodies against indicated pro- or antiangiogenic factors. Bars represent 200 ␮m. B: Immunohistochemical analyses for macrophage Mac-3. Tissue sections, prepared as described above, were stained with anti–Mac-3 antibody. Bars represent 100 ␮m. Right panel shows the summary of the quantitative analyses (n ⴝ 12, 14, 7, and 5 for nondiabetic wild-type, diabetic wild-type, nondiabetic RAGEⴚ/ⴚ, and diabetic RAGEⴚ/ⴚ mice, respectively). Data are shown as means ⴞ SD. No significant differences were observed among the groups.

2250



FIG. 4. Decrease in PCNA-positive cells, increased TUNEL-positive cells, and increased cells with activated NF-␬B in diabetes are dependent on the presence of RAGE. A and B: PCNA- and TUNEL-positive cells were determined by immunohistochemistry. Arrow head indicates TUNEL-positive cells. Bars represent 100 ␮m. C: Immunofluorescence analyses of the nuclear localization of NF-␬B. NF-␬B was stained with rhodamine-conjugated anti–NF-␬B p65 antibody (red) with the nuclear staining with Haechist 33258 dye (blue). Bars represent 50 ␮m. Cells with merged color (white) were counted and expressed as the percentage of total cells. Lower right side is the summary of the quantitative analyses (control RAGEⴙ/ⴙ: n ⴝ 9, diabetic RAGEⴙ/ⴙ: n ⴝ 9, control RAGEⴚ/ⴚ: n ⴝ 7, and diabetic RAGEⴚ/ⴚ: n ⴝ 7). Data are shown as means ⴞ SD. *P < 0.05 vs. control, ANOVA with multiple comparison (Scheffe’s type).

FIG. 5. Adenovirus-mediated overexpression of esRAGE. A: Confluent human arterial smooth muscle cells were infected with 0, 1.0, 3.0, and 10 MOI of Ad-esRAGE and cultured in 2% FCS/ DMEM for indicated days. Media were collected, and esRAGE secreted into the media was determined by immunoblotting. B: VEGF (300 ng/ml) and either Ad-esRAGE or Ad-LacZ were suspended in matrigel (5 ⴛ 108 pfu/ml), and the matrigel were implanted subcutaneously into the abdominal midline of C57BL/6J mice at age 13 weeks. The plug was removed 2 weeks later and processed for immunohistochemical analysis. Arrow heads indicate esRAGE-expressing cells. Bars represent 200 ␮m. DIABETES, VOL. 55, AUGUST 2006

2251


FIG. 6. Adenovirus-mediated overexpression of esRAGE restores diabetes impairment of angiogenic response. Control or insulin-deficient diabetic C57BL/6J mice were used to examine the effect of esRAGE overexpression in vivo. VEGF (300 ng/ml) and either Ad-esRAGE or Ad-LacZ were suspended in matrigel (5 ⴛ 108 pfu/ml), and the matrigel were implanted subcutaneously into the abdominal midline of mice at age 13 weeks. The plug was removed 2 weeks later and processed for immunohistochemical analysis. 〉ars represent 50 ␮m. B: Summary of the quantitative analyses (n ⴝ 5 for each group). Total, CD31-positive, and ␣SM actin–positive cell numbers were counted in the five sequential 200 ⴛ 300 –␮m fields adjacent to the fascia. Data are shown as means ⴞ SD. *P < 0.05 vs. Ad-LacZ–infected control group; **P < 0.05 vs. Ad-LacZ–infected group, ANOVA with multiple comparison (Scheffe’s type). 2252



DISCUSSION

In this study, we showed for the first time that RAGE is involved in impaired angiogenic response in diabetes. Moreover, our results implicate esRAGE as a therapeutic factor to protect impaired angiogenic response in diabetes. In addition to the vasodegenerative changes (2– 4), several observations indicated that angiogenic response or development of new vessels in response to local ischemia/ inflammation is significantly reduced in diabetic patients and animals (5– 8). However, only few reports showed the involvement of AGEs/RAGE system in diabetic-related impaired angiogenesis. Goova et al. (19) demonstrated that blockade of RAGE restores effective wound healing in diabetic mice. Tamarat et al. (18) demonstrated that blockade of AGE formation restored ischemia-induced angiogenesis in diabetic mice. Both of these reports implicate the functional role of AGEs in impaired angiogenic response but do not necessarily delineate the role of RAGE. Goova et al.’s (19) observation is made by using sRAGE, a decoy by binding RAGE ligands, which has the capacity to prevent ligands from interacting with receptors other than RAGE. Tamarat et al. (18) used aminoguanidine to prevent AGE formation. Recently, ablation of galectin-3, another receptor for AGEs, has been shown to abolish the AGE-mediated increase in retinal ischemia and restored the neovascular response to that seen in controls (8). Here, by using RAGE deficient mice, we showed for the first time that RAGE is indeed involved in diabetic-related impairment of angiogenesis. We also demonstrated that, by using adenovirus-mediated gene transfer, overexpression of esRAGE restored diabetes-related impaired angiogenesis. The mechanism underlying AGEs/RAGE-mediated impaired angiogenic response is not clear at present. VEGF is a potent agonist of angiogenesis that activates both endothelial cell proliferation and migration. By contract, TSP-1 suppresses angiogenesis by inhibiting endothelial cell proliferation (31,32). Some in vitro analyses showed that AGE can elicit angiogenesis through the induction of autocrine vascular VEGF, thereby playing a potential role in the development and progression of diabetic retinopaty (17,33,34). However, it has not been clear how the AGEs/ RAGE system regulates VEGF or TSP-1 expression in diabetic peripheral tissue. Our present quantitative analyses for mRNA expression revealed that VEGF is decreased while TSP-1 is upregulated in the matrigel implanted in diabetic mouse. These alterations may be related to impaired angiogenesis in diabetes but was not recovered by RAGE deficiency. Moreover, since our angiogenesis assay contains exogenous VEGF, sufficient VEGF appears not enough to overcome impairment of angiogenesis in diabetes. Moreover, our data showed that genes involved in cell migration, such as tissue factor (28) and platelet-derived growth factor (29), are also downregulated in diabetic mouse. However, these changes in gene regulation again are not dependent on the presence of RAGE. MMP-2 mRNA expression is slightly suppressed in diabetes only in wild-type, not in RAGE⫺/⫺, mice, suggesting that MMP-2 regulation may be at least partly involved in RAGEdependent diabetes suppression of angiogenic response. Thus, aberrant gene regulation involved in angiogenesis appears not to be the primary mechanism underlying RAGE-mediated impaired angiogenic response in diabetes. In these sets of gene expression analyses, a limitation for the interpretation of the data are that there is obviously a DIABETES, VOL. 55, AUGUST 2006

distinct conformational changes in the number, complexity, and differentiation state of the vessels in the plugs between nondiabetic and diabetic mice. Thus, there is a danger that the changes in gene expression profiles may reflect these changes rather than a diabetes-mediated response. One mechanism through which AGEs may affect pathologic processes is by enhanced apoptosis as supported by in vitro cultured cells (35,36). The mechanisms by which AGEs lead to apoptosis could be through increasing oxidative stress, cytoplasmic, or mitochondrial caspase pathways or via induced expression of proapoptotic cytokines (35,36). In our experimental system, TUNEL-positive cells in the matrigel are indeed increased in diabetic mice, and the increase is dependent on the presence of RAGE. However, quantitative analyses reveal that the average TUNEL-positive cells in diabetes is ⬍1.5%, suggesting only partial contribution to the less angiogenic response in diabetic mice. On the other hand, decrease in PCNA-positive cells in diabetes is well correlated with the suppressed angiogenic response in diabetes. Our data also show that the decrease in PCNA-positive cells in diabetes is also dependent on the presence of RAGE. Although the effect of diabetes on endothelial cell proliferation is often presented on the basis of what is observed in the late phase of diabetic retinopathy, recent observations suggest that hyperglycemia or AGEs suppresses angiogenesis even in retina in vivo (8,37). Larger et al. (37) directly measured endothelial cell proliferation in the chicken chorioallantoic membrane assay, which is suppressed by hyperglycemia. Regarding smooth muscle cells in major arteries, such as mouse arterial injury model, RAGE appears to be involved in increased smooth muscle cell proliferation in vivo (38,39). However, little is known about the effect of AGEs or hyperglycemia on the microvascular mural cell proliferation in vivo, although pericyte loss is the hallmarks of early changes of diabetic retinopathy (40,41). Although our current study does not delineate which types of cells are associated with the decrease in cell proliferation in the matrigel of diabetic mice, our data do suggest that diabetic condition suppresses microvascular cell proliferation through RAGE in vivo. Ligation of RAGE leads to activation of the transcription factor NF-␬B, which has been proposed as a factor implicated in diabetes complications (11). Various recent reports have described increased AGE accumulation, RAGE expression, and NF-␬B activation in the organs of diabetic animals and have postulated the pathophysiological relevance of these cellular events in diabetes complications such as diabetic retinopathy (42,43). Our data also show that the percentage of cells with NF-␬B activation in the matrigel is higher in diabetic than control mice. Moreover, the NF-␬B activation in diabetes is not observed in RAGEdeficient mice, implicating that RAGE is involved in diabetes-mediated NF-␬B activation. It is not clear at present, however, whether the RAGE–NF-␬B system is involved in altered cell proliferation and apoptosis and leads to impaired angiogenic response in diabetes. Apparently, further studies will be warranted to clarify this question. The genetically engineered soluble form of RAGE has been successfully used for prevention of diabetic vascular complications (14,19). In recent human studies, circulating soluble RAGE, including esRAGE, has been shown to be inversely associated with atherosclerotic diseases (30,44, 45). Our current results also indicate that esRAGE might 2253


be a useful therapeutic tool for the prevention of impaired angiogenic response in diabetes. In the present study, we could successfully overexpress esRAGE inside and outside of the matrigel by the adenoviral system and could restore impaired angiogenic response in diabetes. In conclusion, the AGEs/RAGE system appears to be involved in impaired angiogenesis in diabetes. esRAGE might be a potential therapeutic candidate for the treatment of this disorder. ACKNOWLEDGMENTS

This study was supported in part by Grant-in-Aid for scientific research (15590953 and 17590946 to H.K and H.Y.) from the Japan Society for the Promotion of Science and by a grant from the Osaka Medical Research Foundation for Incurable Diseases (to Y.N.). REFERENCES 1. Brownlee M: Biochemistry and molecular cell biology of diabetic complications. Nature 414:813– 820, 2001 2. Mizutani M, Kern TS, Lorenzi M: Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy. J Clin Invest 97:2883–2890, 1996 3. Stitt A, Gardiner TA, Anderson NL, Canning P, Frizzell N, Duffy N, Boyle C, Januszewski AS, Chachich M, Baynes JW, Thorpe SR: The AGE inhibitor pyridoxamine inhibits development of retinopathy in experimental diabetes. Diabetes 51:2826 –2832, 2002 4. Hammes HP, Du X, Edelstein D, Taguchi T, Matsumura T, Ju Q, Lin J, Bierhaus A, Nawroth P, Hannak D, Neumaier M, Bergfeld R, Giardino I, Brownlee M: Benfotiamine blocks three major pathways of hyperglycemic damage and prevents experimental diabetic retinopathy. Nat Med 9:294 – 299, 2003 5. Abaci A, Oguzhan A, Kahraman S, Eryol NK, Unal S, Arinc H, Ergin A: Effect of diabetes mellitus on formation of coronary collateral vessels. Circulation 99:2239 –2242, 1999 6. Rivard A, Silver M, Chen D, Kearney M, Magner M, Annex B, Peters K, Isner JM: Rescue of diabetes-related impairment of angiogenesis by intramuscular gene therapy with adeno-VEGF. Am J Pathol 154:355–363, 1999 7. Waltenberger J: Impaired collateral vessel development in diabetes: potential cellular mechanisms and therapeutic implications. Cardiovasc Res 49:554 –560, 2001 8. Stitt AW, McGoldrick C, Rice-McCaldin A, McCance DR, Glenn JV, Hsu DK, Liu F-T, Thorpe SR, Gardiner TA: Impaired retinal angiogenesis in diabetes: role of advanced glycation end products and galectin-3. Diabetes 54:785–794, 2005 9. Neeper M, Schmidt AM, Brett J, Yan SD, Wang F, Pan YC, Elliston K, Stern D, Shaw A: Cloning and expression of a cell surface receptor for advanced glycosylation end products of proteins. J Biol Chem 267:14998 –15004, 1992 10. Schmidt AM, Yan SD, Yan SF, Stern DM: The biology of the receptor for advanced glycation end products and its ligands. Biochim Biophys Acta 1498:99 –111, 2000 11. Schmidt AM, Hori O, Chen JX, Li JF, Crandall J, Zhang J, Cao R, Yan SD, Brett J, Stern D: Advanced glycation endproducts interacting with their endothelial receptor induce expression of vascular cell adhesion molecule-1 (VCAM-1) in cultured human endothelial cells and in mice: a potential mechanism for the accelerated vasculopathy of diabetes. J Clin Invest 96:1395–1403, 1995 12. Basta G, Lazzerini G, Massaro M, Simoncini T, Tanganelli P, Fu C, Kislinger T, Stern DM, Schmidt AM, De Caterina R: Advanced glycation end products activate endothelium through signal-transduction receptor RAGE: a mechanism for amplification of inflammatory responses. Circulation 105:816 – 822, 2002 13. Yamamoto Y, Kato I, Doi T, Yonekura H, Ohashi S, Takeuchi M, Watanabe T, Yamagishi S, Sakurai S, Takasawa S, Okamoto H, Yamamoto H: Development and prevention of advanced diabetic nephropathy in RAGEoverexpressing mice. J Clin Invest 108:261–268, 2001 14. Yan SF, Ramasamy R, Naka Y, Schmidt AM: Glycation, inflammation, and RAGE: a scaffold for the macrovascular complications of diabetes and beyond. Circ Res 93:1159 –1169, 2003 15. Nawroth P, Bierhaus A, Marrero M, Yamamoto H, Stern DM: Atherosclerosis and restenosis: is there a role for RAGE? Curr Diab Rep 5:11–16, 2005 2254

16. Kuzuya M, Satake S, Ai S, Asai T, Kanda S, Ramos MA, Miura H, Ueda M, Iguchi A: Inhibition of angiogenesis on glycated collagen lattices. Diabetologia 41:491– 499, 1998 17. Yamagishi S, Yonekura H, Yamamoto Y, Katsuno K, Sato F, Mita I, Ooka H, Satozawa N, Kawakami T, Nomura M, Yamamoto H: Advanced glycation end products-driven angiogenesis in vitro: induction of the growth and tube formation of human microvascular endothelial cells through autocrine vascular endothelial growth factor. J Biol Chem 272:8723– 8730, 1997 18. Tamarat R, Silvestre JS, Huijberts M, Benessiano J, Ebrahimian TG, Duriez M, Wautier MP, Wautier JL, Levy BI: Blockade of advanced glycation end-product formation restores ischemia-induced angiogenesis in diabetic mice. Proc Natl Acad Sci U S A 100:8555– 8560, 2003 19. Goova MT, Li J, Kislinger T, Qu W, Lu Y, Bucciarelli LG, Nowygrod S, Wolf BM, Caliste X, Yan SF, Stern DM, Schmidt AM: Blockade of receptor for advanced glycation end-products restores effective wound healing in diabetic mice. Am J Pathol 159:513–525, 2001 20. Yamamoto Y, Doi T, Kato I, Shinohara H, Sakurai S, Yonekura H, Watanabe T, Myint KM, Harashima A, Takeuchi M, Takasawa S, Okamoto H, Hashimoto N, Asano M, Yamamoto H: Receptor for advanced glycation end products is a promising target of diabetic nephropathy. Ann N Y Acad Sci 1043:562–566, 2005 21. Kunjathoor VV, Wilson DL, LeBoeuf RC: Increased atherosclerosis in streptozotocin-induced diabetic mice. J Clin Invest 97:1767–1773, 1996 22. Takeuchi M, Makita Z, Yanagisawa K, Kameda Y, Koike T: Detection of noncarboxymethyllysine and carboxymethyllysine advanced glycation end products (AGE) in serum of diabetic patients. Mol Med 5:393– 405, 1999 23. Tamarat R, Silvestre JS, Durie M, Levy BI: Angiotensin II angiogenic effect in vivo involves vascular endothelial growth factor- and inflammationrelated pathways. Lab Invest 82:747–756, 2002 24. Yonekura H, Yamamoto Y, Sakurai S, Petrova RG, Abedin MJ, Li H, Yasui K, Takeuchi M, Makita Z, Takasawa S, Okamoto H, Watanabe T, Yamamoto H: Novel splice variants of the receptor for advanced glycation endproducts expressed in human vascular endothelial cells and pericytes, and their putative roles in diabetes-induced vascular injury. Biochem J 370: 1097–1109, 2003 25. Mizuguchi H, Koizumi N, Hosono T, Utoguchi N, Watanabe Y, Kay MA, Hayakawa T: A simplified system for constructing recombinant adenoviral vectors containing heterologous peptides in the HI loop of their fiber knob. Gene Ther 8:730 –735, 2001 26. Morioka T, Koyama H, Yamamura H, Tanaka S, Fukumoto S, Emoto M, Mizuguchi H, Hayakawa T, Kojima I, Takahashi K, Nishizawa Y: Role of H1-calponin in pancreatic AR42J cell differentiation into insulin-producing cells. Diabetes 52:760 –766, 2003 27. Armulik A, Abramsson A, Betsholtz C: Endothelial/pericyte interactions. Circ Res 97:512–523, 2005 28. Mackman N: Role of tissue factor in hemostasis, thrombosis, and vascular development arterioscler. Arterioscler Thromb Vasc Biol 24:1015–1022, 2004 29. Lindahl P, Johansson BR, Leveen P, Betsholtz C: Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 277:242–245, 1997 30. Koyama H, Shoji T, Yokoyama H, Motoyama K, Mori K, Fukumoto S, Emoto M, Shoji T, Tamei H, Matsuki H, Sakurai S, Yamamoto Y, Yonekura H, Watanabe T, Yamamoto H, Nishizawa Y: Plasma level of endogenous secretory RAGE is associated with components of the metabolic syndrome and atherosclerosis. Arterioscler Thromb Vasc Biol 25:2587–2593, 2005 31. Jimenez B, Volpert OV, Crawford SE, Febbraio M, Silverstein RL, Bouck N: Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat Med 6:41– 48, 2000 32. Sargiannidou I, Zhou J, Tuszynski GP: The role of thrombospondin-1 in tumor progression. Exp Biol Med (Maywood) 226:726 –733, 2001 33. Yamagishi S, Kobayashi K, Yamamoto H: Vascular pericytes not only regulate growth, but also preserve prostacyclin-producing ability and protect against lipid peroxide-induced injury of co-cultured endothelial cells. Biochem Biophys Res Commun 190:418 – 425, 1993 34. Yamagishi S, Hsu CC, Taniguchi M, Harada S, Yamamoto Y, Ohsawa K, Kobayashi K, Yamamoto H: Receptor-mediated toxicity to pericytes of advanced glycosylation end products: a possible mechanism of pericyte loss in diabetic microangiopathy. Biochem Biophys Res Commun 213:681– 687, 1995 35. Yamagishi S, Inagaki Y, Amano S, Okamoto T, Takeuchi M, Makita Z: Pigment epithelium-derived factor protects cultured retinal pericytes from advanced glycation end product-induced injury through its antioxidative properties. Biochem Biophys Res Commun 296:877– 882, 2002 36. Alikhani Z, Alikhani M, Boyd CM, Nagao K, Trackman PC, Graves DT: Advanced glycation end products enhance expression of pro-apoptotic DIABETES, VOL. 55, AUGUST 2006


genes and stimulate fibroblast apoptosis through cytoplasmic and mitochondrial pathways. J Biol Chem 280:12087–12095, 2005 37. Larger E, Marre M, Corvol P, Gasc JM: Hyperglycemia-induced defects in angiogenesis in the chicken chorioallantoic membrane model. Diabetes 53:752–761, 2004 38. Sakaguchi T, Yan SF, Yan SD, Belov D, Rong LL, Sousa M, Andrassy M, Marso SP, Duda S, Arnold B, Liliensiek B, Nawroth PP, Stern DM, Schmidt AM, Naka Y: Central role of RAGE-dependent neointimal expansion in arterial restenosis. J Clin Invest 111:959 –972, 2003 39. Zhou Z, Wang K, Penn MS, Marso SP, Lauer MA, Forudi F, Zhou X, Qu W, Lu Y, Stern DM, Schmidt AM, Lincoff AM, Topol EJ: Receptor for AGE (RAGE) mediates neointimal formation in response to arterial injury. Circulation 107:2238 –2243, 2003 40. Cogan D, Toussaint D, Kuwabara T: Retinal vascular pattern. IV. Diabetic retinopathy. Arch Ophthalmol 66:366 –378, 1961 41. Hammes HP, Lin J, Renner O, Shani M, Lundqvist A, Betsholtz C, Brownlee M, Deutsch U: Pericytes and the pathogenesis of diabetic retinopathy. Diabetes 51:3107–3112, 2002 42. Bierhaus A, Schiekofer S, Schwaninger M, Andrassy M, Humpert PM, Chen


J, Hong M, Luther T, Henle T, Kloting I, Morcos M, Hofmann M, Tritschler H, Weigle B, Kasper M, Smith M, Perry G, Schmidt AM, Stern DM, Haring HU, Schleicher E, Nawroth PP: Diabetes-associated sustained activation of the transcription factor nuclear factor-␬B. Diabetes 50: 2792–2808, 2001 43. Alves M, Calegari VC, Cunha DA, Saad MJ, Velloso LA, Rocha EM: Increased expression of advanced glycation end-products and their receptor, and activation of nuclear factor kappa-B in lacrimal glands of diabetic rats. Diabetologia 48:2675–2681, 2005 44. Falcone C, Emanuele E, D’Angelo A, Buzzi MP, Belvito C, Cuccia M, Geroldi D: Plasma levels of soluble receptor for advanced glycation end products and coronary artery disease in nondiabetic men. Arterioscler Thromb Vasc Biol 25:1032–1037, 2005 45. Katakami N, Matsuhisa M, Kaneto H, Matsuoka TA, Sakamoto K, Nakatani Y, Ohtoshi K, Hayaishi-Okano R, Kosugi K, Hori M, Yamasaki Y: Decreased endogenous secretory advanced glycation end product receptor in type 1 diabetic patients: its possible association with diabetic vascular complications. Diabetes Care 28:2716 –2721, 2005

2255