Role of Adiponectin in Preventing Vascular Stenosis

Obesity is more linked to vascular disease, including atherosclerosis and restenotic change, after balloon angioplasty. The precise mechanism linking obesity and vascular disease is still unclear. Previously we have demonstrated that the plasma levels of adiponectin, an adipose-derived hormone, decreases in obese subjects, and that hypoadiponectinemia is associated to ischemic heart disease. In current the study, we investigated the in vivorole of adiponectin on the neointimal thickening after artery injury using adiponectin-deficient mice and adiponectin-producing adenovirus. Adiponectin-deficient mice showed severe neointimal thickening and increased proliferation of vascular smooth muscle cells in mechanically injured arteries. Adenovirus-mediated supplement of adiponectin attenuated neointimal proliferation. In cultured smooth muscle cells, adiponectin attenuated DNA synthesis induced by growth factors including platelet-derived growth factor, heparin-binding epidermal growth factor (EGF)-like growth factor (HB-EGF), basic fibroblast growth factor, and EGF and cell proliferation and migration induced by HB-EGF. In cultured endothelial cells, adiponectin attenuated HB-EGF expression stimulated by tumor necrosis factor α. The current study suggests an adipo-vascular axis, a direct link between fat and artery. A therapeutic strategy to increase plasma adiponectin should be useful in preventing vascular restenosis after angioplasty.

Adiponectin is an adipocyte-derived factor that was identified by our group in human adipose tissues (8). Acrp30 or AdipoQ, independently cloned by two groups, is the mouse counterpart of adiponectin (9,10). Adiponectin mRNA is expressed exclusively in adipose tissues. Adiponectin is composed of two structurally distinct domains: C-terminal collagen-like fibrous domain and complement C1q-like globular domain. Interestingly, low plasma concentrations of adiponectin are found in obese subjects (11) and patients with coronary artery disease (12). Furthermore, the incidence of cardiovascular death is higher in renal failure patients with low plasma adiponectin compared with those with higher plasma adiponectin levels (13). We have also reported that adiponectin infiltrates rapidly into the subendothelial space of the vascular wall when the endothelial barrier of the arterial wall is injured by balloon angioplasty (14). In tissue cultures, adiponectin attenuates monocyte attachment to endothelial cells by reducing the expression of adhesion molecules on endothelial cells (12,15). Adiponectin also suppresses lipid accumulation in monocytederived macrophages through the suppression of macrophage scavenger receptor expression (16). These in vitro data suggested the anti-atherogenic properties of adiponectin, and hence hypoadiponectinemia might be associated with a higher incidence of vascular diseases in obese subjects.
In the present study, we investigated the role of adiponectin on the vascular wall in vivo using adiponectin knockout (KO) mice and adiponectin-producing adenovirus (17). Our results demonstrate that adiponectin deficiency aggravates neointimal thickening, and adiponectin supplement attenuates neointimal thickening in mechanically injured arteries, presumably through the suppressive effect of adiponectin on the proliferation and migration of vascular smooth muscle cells. Here we show the first in vivo evidence that adiponectin is a fat-derived hormone directly bridging the adipose-vascular axis. * This work was supported in part by "Research for the Future" Program JSPS-RFTF97L00801 from the Japan Society for the Promotion of Science and Grants-in-aid 12307022, 12557090, and 12671084 from the Ministry of Education, Science, Sports and Culture of Japan. 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

EXPERIMENTAL PROCEDURES
Animals-Adiponectin KO male mice (8 -11 weeks old) were generated as described previously (17). Briefly, the targeting vector for the adiponectin KO mice was constructed using a positive selection cassette derived from a vector, pPolIIsneobpA, containing the neo R gene. The 1.4-kb VspI-EcoRI region, located in intron 2 of mouse adiponectin gene, was inserted into the VspI site of pPolIIsneobpA prior to the neo R cassette. The 7-kb SnaBI-AatII region located in intron 1 was inserted into the SnaBI site to the 3Ј site of the neo R cassette. The targeting construct was linearized with NotI and introduced into mouse AB2.2prime embryonic stem cells (Lexicon Genetics, Woodlands, TX) by electroporation (270 V, 500 microfarad, BTX ECM600). Embryonic stem cell clones resistant to G418/gancyclovir were isolated, and 16 positive clones were obtained. Chimeric animals obtained from the microinjections were bred to C57BL/6J mice, and three chimeric males sired offspring that carried the disrupted mouse adiponectin allele through the germ line.
Femoral Artery Injury-The femoral artery injury procedure in mice was conducted as described previously (18,19). Briefly, wild-type (WT) and adiponectin KO male mice underwent bilateral femoral artery injury by a straight spring wire (0.36-mm diameter, no. SKI 175 FLP 14-S, Invatec, Concesio (BS), Italy), denuding vascular endothelium and inducing neointimal hyperplasia. At 2-3 weeks after vascular injury, the mice were anesthetized, and both femoral arteries were harvested after perfusion fixation with 10% formalin and embedded in paraffin. Following embedding in paraffin, parallel sections were stained with hematoxylin and eosin. Smooth muscle cells were identified by immunostaining for ␣-smooth muscle actin using clone 1A4 from Sigma as the primary antibody. Intimal and medial area were measured using the image analysis software MacSCOPE.
BrdUrd Staining-Following vascular injury, 100 g/g BrdUrd was administered intraperitoneally every 24 h until harvesting the femoral arteries. On the 14th day after vascular injury, the femoral arteries were perfusion-fixed in 10% formalin, harvested, and embedded in paraffin. After deparaffinization, parallel sections were immunostained with BrdUrd using a BrdUrd staining kit (Oncogene Research Products, Boston, MA). BrdUrd-labeled and -unlabeled smooth muscle cells in neointima were counted for each section. The proliferation index was calculated by dividing the number of BrdUrd-labeled cells by the number of unlabeled cells as described previously (20).
Preparation and Administration of Adenovirus-Adenovirus producing the full-length mouse adiponectin was prepared by using the Adenovirus Expression Vector kit (Takara, Kyoto, Japan). 2 ϫ 10 8 plaque-forming units of adenovirus-adiponectin (Ad-APN) or adenovirus-␤-galactosidase (Ad-␤gal) was injected into the jugular vein of mice 3 days prior to the femoral artery injury. On the 14th day after the virus injection (11th day after the injury), the femoral arteries were harvested for analysis.
Cell Culture-Human aortic smooth muscle cells (HASMCs) (Clonetics) were maintained and used for experiments at passage 4 or 5 as previously described (21). Human aortic endothelial cells (HAECs) (Clonetics) were maintained in plastic plates precoated with type I collagen (BD PharMingen) as described previously (15). Human recombinant adiponectin was prepared as reported previously (11).
Cell Proliferation Assays-HASMCs were treated for 18 h in Dulbecco's modified Eagle's medium containing 2% fetal calf serum (Invitrogen) with or without 10 ng/ml human recombinant platelet-derived growth factor (PDGF)-BB, HB-EGF, basic fibroblast growth factor (FGF), and EGF (R&D Systems) in the presence or absence of 30 g/ml human recombinant adiponectin. The cells were exposed to [ 3 H]thymidine (Amersham Biosciences) at 20 Ci/ml for 6, then trypsinized, and retrieved onto glass fiber filters using an automatic cell harvester.
[ 3 H]Thymidine uptake was measured in a direct ␤ counter. Cell number was counted with the hemocytometer method as described previously (22).
Cell Migration Assay-Migration assays were performed using a Boyden chamber. HASMCs (5 ϫ 10 4 cells/ml) were added to the Transwell inserts (Costar, 12-mm diameter, 12.0-m pore size) precoated with collagen type I. Migration was induced by HB-EGF (10 ng/ml) with or without adiponectin (30 g/ml) added to the lower chamber beneath the insert membrane. The Transwell chambers were then incubated for 4 h under culture condition. Migrated HASMCs on the lower surface of the membrane were fixed with ethanol and stained with hematoxylin. Migration activity was evaluated microscopically by counting the number of stained nuclei per high power field (ϫ400). All assays were performed in triplicate, and each sample was counted randomly in 10 different areas in the center of the membrane.
Measurement of HB-EGF mRNA-HAECs in a confluent state were preincubated for 18 h in medium M199 (Invitrogen) containing 0.5% fetal calf serum and 3% bovine serum albumin with or without 30 g/ml recombinant adiponectin and then exposed to human recombinant TNF␣ (R&D Systems) or vehicle at a final concentration of 10 ng/ml for 2 h. Cells were harvested, and total RNA was prepared with an RNA STAT-60 kit (Tel-Test, Friendswood, TX). cDNA was produced using Taqman reverse transcription kits (PerkinElmer Life Sciences). Realtime PCR was performed on an ABI-Prism 7700 using the Master Mix SYBR Green kit (PE-Applied Biosystems, Norwalk, CT) according to the manufacturer's instructions. Primers were: 5Ј-TCCTCCAAGCCA-CAAGCACT-3Ј and 5Ј-GCCCATGACACCTCTCTCCA-3Ј for HB-EGF and 5Ј-ACCACAGTCCATGCCATCAC-3Ј and 5Ј-CACCACCTTCTT-GATGTCATC-3Ј for glyceraldehyde-3-phosphate dehydrogenase.
Statistical Analysis and Ethical Considerations-Results were expressed as mean Ϯ S.E. Differences between groups were examined for statistical significance using the Student's t test or analysis of variance with Fisher's protected least significant difference test. A p value less than 0.05 denoted the presence of a statistically significant difference. The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of Osaka University School of Medicine.

RESULTS
Basal Profile of Adiponectin Knockout Mice-Adipose mRNA and plasma protein of adiponectin were deficient in KO mice studied in the current analysis (data not shown). Table I describes the phenotypic comparison in WT and adiponectin KO mice under non-fasted and 12-h fasted conditions. No significant differences were observed in the weights of body and various tissues including epididymal white fat, brown fat, liver, gastrocnemius muscle, and heart. Plasma concentration of glucose, insulin, cholesterol, triglyceride, and free fatty acid were not altered significantly in the adiponectin KO mice. Adiponectin Deficiency Increases Neointimal Thickening in Injured Arteries-In the present study, we denuded the vascular endothelium of the femoral artery as we described previously (19,20) and compared the neointimal thickening of the arteries at 3 weeks after the injury between WT and KO mice. Hematoxylin-eosin staining (Fig. 1A, upper panel) demonstrated that the neointimal hyperplasia in the injured artery was worse in KO mice than in WT mice. Immunohistochemical staining revealed that the neointima in both WT and KO mice was composed of ␣-smooth muscle actin-positive smooth muscle cells (Fig. 1A, lower panel). We also quantitatively measured the intimal and medial area by computerized morphometry (Fig. 1B). The I/M ratio (the ratio of intimal area/medial area) was significantly greater in KO mice compared with WT mice (p Ͻ 0.01). Notably, two of seven injured femoral arteries were occluded only in KO mice at 3 weeks after injury.
Adiponectin Deficiency Increases Proliferation of Vascular Smooth Muscle Cells in Injured Arteries-Next we assessed proliferation of vascular smooth muscle cells (VSMCs) by immunohistochemical detection of BrdUrd-labeled VSMCs in the sections from each femoral artery at 2 weeks after injury ( Fig.  2A). Quantitative data of proliferation index revealed that intimal VSMC proliferation induced by vascular injury was ϳ2fold greater in KO mice than in WT mice (p ϭ 0.001) (Fig. 2B). In non-injured arteries of both WT and KO mice, BrdUrdlabeled VSMCs were barely detectable (data not shown). These data demonstrate that deficiency of adiponectin in KO mice caused severe neointimal hyperplasia after artery injury, suggesting an inhibitory effect of adiponectin on the proliferation of VSMCs in injured arteries.
Adenovirus plasia induced by vascular injury, we constructed recombinant adenovirus producing mouse adiponectin. WT and KO mice were infected with Ad-␤gal or Ad-APN prior to vascular injury. Ad-APN infection resulted in a 2-3-fold increase in plasma levels of adiponectin on the 4th day after adenoviral injection in both WT and KO mice compared with those in Ad-␤gal-infected WT mice (Fig. 3A). Hematoxylin-eosin-stained sections of the injured femoral arteries of adenovirus-treated KO mice showed that injection of Ad-APN resulted in the suppression of neointimal formation induced by vascular injury (Fig. 3B). Quantitative analysis of these sections revealed that the I/M ratio of femoral arteries of Ad-␤gal-treated KO mice was significantly greater than that of Ad-␤gal-treated WT mice (p ϭ 0.002) (Fig.  3C) as was shown in mice without adenoviral infection in Fig.  1. In the KO mice, the adenovirus-mediated production of adiponectin in plasma attenuated the increase of I/M ratio to the levels of WT mice (p ϭ 0.001) (Fig. 3C). These results demonstrate that adiponectin supplement could reverse neointimal hyperplasia in KO mice.
Adiponectin Suppresses Growth Factor-induced Proliferation and Migration of Cultured VSMCs-In vitro experiments provided strong evidence that adiponectin exerts suppressive effects on VSMC proliferation. PDGF, HB-EGF, basic FGF, and EGF have potent mitogenic activities on HASMCs. Adiponectin treatment attenuated growth factor-induced DNA synthesis in HASMCs (Fig. 4A). The inhibitory effect of adiponectin on HASMC proliferation induced by HB-EGF was directly shown by counting the cell number (Fig. 4B). In addition, adiponectin also suppressed HB-EGF-induced migration of HASMCs (Fig. 4C).
Adiponectin Attenuates the Expression of HB-EGF mRNA in Cultured Endothelial Cells-Next we investigated whether adiponectin could suppress the production of HB-EGF in endothelial cells. Adiponectin treatment completely blocked the TNF␣-mediated increase of HB-EGF mRNA in HAECs (Fig. 5). DISCUSSION In the current study, we demonstrated that adiponectin-null mice exhibited augmented intimal proliferation in mechanically injured vascular walls. Adenovirus-mediated supplement of adiponectin improved the intimal thickening in KO mice to the WT level. How does adiponectin suppress intimal thickening? Fig. 6  injury-induced intimal thickening.
Plasminogen activator inhibitor-1 and HB-EGF are vasoactive substances produced by adipose tissue, although these substances are not adipose-specific. Both factors are considered to promote the development of vascular diseases in obesity (6,7). Contrary to these factors, the plasma concentration of adipose-specific adiponectin is lower in obese subjects and patients with coronary artery disease (11,12). The present study demonstrated in vivo and in vitro that adiponectin suppressed VSMC proliferation. Taken together, adipose tissue secretes both the offense molecules (plasminogen activator inhibitor-1 and HB-EGF) and the defense molecule (adiponectin) into the blood stream, reaching the vascular wall. Then, in obesity, both the increase of offense molecules and decrease of defense molecule(s) in plasma should aggravate vascular diseases. Considering the adipose specificity, adiponectin should play a major role in the adipo-vascular axis.
Recent studies have identified the role of various molecules derived from adipose tissue in the development of insulin resistance. These include TNF␣, leptin, and resistin (3)(4)(5). More recently adiponectin treatment has been shown to improve fatty acid oxidation and insulin resistance in diabetic animals (23,24). Adiponectin-null mice show normal insulin sensitivity under a regular diet but severe insulin resistance under a high fat/high sucrose diet (17). Interestingly, subjects carrying a missense mutation in the adiponectin gene associated with hypoadiponectinemia exhibit the phenotype of the metabolic syndrome, including insulin resistance and coronary artery disease (25). These findings suggest that hypoadiponectinemia associated with obesity is located upstream of metabolic syndrome in the pathophysiology. In the present study, adiponectin-null mice showed profound neointimal hyperplasia despite normal glucose and lipid metabolism. Our results indicate that injury-induced neointimal formation does not accelerate as a result of abnormalities of glucose and lipid metabolism but is directly caused by adiponectin deficiency. Therapeutic approaches that increase plasma adiponectin concentration could be useful in preventing restenosis after vascular intervention.