Inhibitory Effect of Diabetes on Proliferation of Vascular Smooth Muscle After Balloon Injury in Rat Aorta

The effect of streptozotocin-induced diabetes on cell proliferation in rat aortic intima-media, as well as on local gene expression of transforming growth factor-β1 (TGF-β1) was studied. TGF-β1 mRNA was measured by solution hybridization and TGF-β1 protein by ELISA. Proliferation was measured by bromodeoxyuridine incorporation into DNA two days after balloon injury. All BrdU-labelled cells observed were smooth muscle cells. After a diabetes duration of 2 and 4 weeks, labelled cells were significantly fewer compared with controls. Circulating levels of total TGF-β1 were lowered in rats with 2 weeks diabetes. Although the balloon injury procedure by itself stimulated the gene expression of TGF-β1, no significant difference in TGF-β1 mRNA content between diabetic and control rats after injury was found. In conclusion: vascular smooth muscle proliferation in vivo is inhibited by the diabetic state in this model of insulin deficient diabetes and this inhibition is not related to an impaired local expression of TGF-β1.

The effect of streptozotocin-induced diabetes on cell proliferation in rat aortic intima-media, as well as on local gene expression of transforming growth factor-//1 (TGF-//1) was studied. TGF-//1 mRNA was measured by solution hybridization and TGF-//1 protein by ELISA. Proliferation was measured by bromodeoxyuridine incorporation into DNA two days after balloon injury. All BrdU-labelled cells observed were smooth muscle cells. After a diabetes duration of 2 and 4 weeks, labelled cells were significantly fewer compared with controls. Circulating levels of total TGF-jI were lowered in rats with 2 weeks diabetes. Although the balloon injury procedure by itself stimulated the gene expression of TGF-]/1, no significant difference in TGFq/1 mRNA content between diabetic and control rats after injury was found. In conclusion: vascular smooth muscle proliferation in vivo is inhibited by INTRODUCTION In the development of atherosclerosis and in restenosis after balloon angioplasty, proliferation of vascular smooth muscle cells plays a central role. I1"21 Smooth muscle cell growth is regulated by different growth factors that are locally produced in the vessel wall or circulate in the blood. When proliferation of vascular smooth muscle cells is provoked by balloon angioplasty, there is an increased expression of growth factors such as platelet-derived growth factor (PDGF), I31 basic fibroblast growth factor (bFGF), 41 insulin-like growth factor-I (IGF-I)I5j and transforming growth factor-ill (TGF-fll) 6 diabetes associated nephropathy, macrovascu-12 h darkness cycle and had free access to food lar complications 91 and wound healing. [1][2][3][4][5][6][7][8][9][10][11][12][13] and water. All animal experiments were made High glucose has been shown to induce the in agreement with the principles of laboratory expression of TGF-fll in some cell types I91 animal care and were approved by the local and circulating levels of TGF-fll are reported ethics committee for animal experiments. Diato be elevated in diabetic rats [14] and in betes was induced by i.v. injection of streptozo-NIDDM patients. I151 However, in wounds of tocin (65mg/kg body weight) in a tail vein, diabetic rats with retarded wound healing, while rats injected with 0.9% NaC1 served as levels of TGF-fl, bFGF and IGF-I are decreas-controls. Blood samples were taken from tail ed [11"13] and treatment with TGF-fl has been vein and blood glucose was measured using a shown to improve wound healing in diabetic hexokinase method (Gluco-quant(R)) or ONE animals. I1-12 TOUCH(R) test strips for quantitative measure-Bornfeldt et al., studied the effect of experi-ment of glucose in blood. I2l The rats were mental diabetes on the response of arteries to considered diabetic if random blood glucose injury and showed that 3H-thymidine incorpora-concentration exceeded 15  following balloon injury to the aorta. I171 In patients with type II diabetes mellitus, a greater incidence of restenosis after arterial injury caused by coronary stenting has been reported I18 while wound healing in general is impaired. 2,19] The present study was undertaken to investigate the effect of diabetes on early cell proliferation in the aorta after injury. We also investigated if the effect of diabetes was associated with a change in circulating levels of TGF-fll or changes in local TGF-fll gene expression.
De-endothelialization was performed, using a balloon catheter, according to the method of Capron et a/. [21] In short, the rats were kept under light ether anaestesia and a deflated embolectomy catheter (Fogarty, size 2F, Baxter Medical) was introduced into the aorta through the left common carotid artery down to the level of the renal artery. The balloon was inflated with 40-50tl distilled water, and withdrawn to the level of the diaphragm. After the passage of the diaphragm had been felt, additional 20-30 tl distilled water was inflated. This procedure was repeated three times and the carotid artery was then double ligated. Control rats were sham operated: the same procedure was carried out except that no catheter was introduced into the carotid artery.

MATERIALS AND METHODS
DNA synthesis was measured two days after injury. The rats were injected intraperitoneally Animals with 5-bromo-2'-deoxyuridine (BrdU) 0.1 mg/g body weight. Two hours after injection, the Male Sprague-Dawley rats (ALAB, Stockholm, animals were killed and the aortic segment Sweden) weighing about 200g were used in between the left subclavian artery and the celiac artey was removed and immersed in ice-cold saline. Within 30 minutes, cross sections of the central part of the aortic segment were cut out and quickly frozen at -70C. With a setting at 4 tm, cryostat sections were cut out and brought to chrome-gelatinized glass slides. The slides were fixed in 70% ethanol for I h and were allowed to dry. DNA was denatured with 1.5 mol/1 NaOH for 90 sec followed by neutralisation in 0.1 mol/1 Na2B407 (pH 8.5) for 10sec and rinsing twice in a phosphate buffer (PBS).
After incubation with an anti-BrdU monoclonal antibody and a peroxidase-conjugated secondary antibody developed with diaminobenzidine as a chromogen, cells in the S-phase were easily visualized. DNA synthesis was expressed as number of labelled cells in a whole cross section of aortic media layer counted under an ordinary light microscope (Fig. 1). One cross section from each aorta was counted.

Measurement of mRNA by a Solution Hybridization Assay
The intima-media was separated from adventitia according to the method of Wolinsky and Daly. [22] The intima-media of the aortas were homogenized in a glass-glass homogenizer for 30 s. Nucleic acids were extracted essentially as described by Durnam and Palmiter. I231 Samples were digested with proteinase K and extracted with phenol and chloroform. Nucleic acids were precipitated with ethanol. Total nucleic acids were measured by spectrophotometry. DNA content was measured by fluorimetry according to the method of Labarca and Paigen. [24] The mRNA level of TGF-fll was determined by a solution hybridization assay 231 using a [35S]-UTP labelled probe per incubation. The samples were exposed to RNases and the hybrids precipitated with 100tl TCA (6mol/1), collected on glass microfibre filters and radioactivity measured in a liquid scintillation counter (1214 Rackbeta; LKB). The radioactivity FIGURE 1 Photomicrograph of the media layer of aorta in a control rat, two days after balloon angioplasty to the aorta. No reendothelialization of the injured aorta can be seen at this time point. The structure of the artery wall can be discerned and the nuclei of two smooth muscle cells in S-phase are strongly labelled by BrdU incorporation. (See Color Plate VI). of each sample was then compared with a RESULTS standard curve constructed from a sample with known amount of in vitro synthesized TGF-fll Induction of diabetes was monitored by measursense RNA complementary to the probe. A ing body weight and blood glucose. Blood standard curve was included in each assay and glucose was markedly elevated and body weight samples were analysed in triplicate. Tubes of streptozotocin injected rats was lower than including only hybridization buffer and probe in control rats (Tab. I). served as blanks.
Two days after balloon angioplasty the aortic intima-media was examined with regard to the presence of BrdU-labelled cells. At this time Immunoassay for TGF-//1 point no re-endothelialization of the injured aorta was seen and of the aortic intima-media Serum levels of TGF-fll were measured by a quantitative immunoassay kit for TGF-fll (R&D only the media layer was present. All cells reacting with antibodies against BrdU were Systems). Before analysis TGF-fll protein was activated by acidification of serum samples. A standard curve made from recombinant TGF-fll was included in each assay. Chemicals smooth muscle cells (Fig. 1). The number of labelled cells was lower in diabetic rats as compared with controls (Fig. 2). Proliferating cells in aortic media remained low in diabetic rats, while in nondiabetic rats, BrdU-labelled cells decreased along with the increased age. Consequently, the difference between control Streptozotocin was obtained from Sigma Cheand diabetic rats was more pronounced in mical Co. (St Louis, MO, USA) and ONE TOrats with a diabetes duration of 2 and 4 weeks UCH(R) test strips were from Lifescan (Milpitas, than in rats with 8 weeks diabetes. California, USA). Gluco-quant(R), 5-bromo-Total TGF-fll was measured in serum from rats 2'deoxyuridine, RNases, Proteinase K and herwith a diabetes duration of 2 weeks (Fig. 3). ring sperm DNA were from Roche Diagnostics Levels of TGF-fll were slightly but significantly (Mannheim, Germany). Anti-BrdU monoclonal lower in diabetic animals as compared with antibody and peroxidase-conjugated secondcontrols. No difference was seen between balloon ary antibody was purchased from Dakopatts injured rats and sham operated rats in either AB (H/igersten, Sweden). Immunoassay kit control animals or diabetic animals. Levels of for human TGF-fll was obtained from R&D active TGF-fll in serum was measured. A tend-Systems (Abingdon, UK). [35S]-UTP was from ency to lower values of active serum TGF-fll in Amersham International (Amersham, Bucks, diabetic rats as compared to controls was seen UK), chemicals for probe synthesis were received from Promega (Madison, WI, USA) and

DISCUSSION
It is known that proliferation of aortic cells induced by de-endothelialization peaks two days after injury. I21] We therefore determined BrdU-labelling two days after balloon catheter injury, a time point when the aorta was devoid of intima and there was no re-endothelialization. All BrdU-labelled cells were smooth muscle cells and the number of BrdU-labelled cells were lowered in diabetic aorta in comparison to aorta from nondiabetic control rats. Bornfeldt and coworkers [51 showed that 3H-thymidine incorporation into DNA was inhibited in aorta of diabetic rats after balloon injury and that the responded with lower proliferation than DNA content was lower in diabetic aorta 2 expected. weeks after injury. In this study and in the It might be argued that streptozotocin by itstudy of Bornfeldt et al., untreated streptozo-self causes the impaired proliferation of smooth tocin diabetic Sprague-Dawley rats, an insulin muscle cells in diabetic rat aorta. However, deficient model of diabetes, was used. Another earlier studies in our laboratory showed that study using the same rat model reported an effects of streptozotocin induced diabetes on inhibited neointima formation in the carotid Hg-thymidine incorporation 5 and C14-1eucine artery after balloon injury. I161 Our results incorporation [29] in rat aorta, could be reversed show that inhibition of smooth muscle cell by insulin treatment. This suggests that the proliferation is a feature of the inhibited arterimpaired metabolism and growth seen in ial response to injury in aorta of streptozotocin streptozotocin diabetic rat aortas, is caused by diabetic rats. In line with these results it was the diabetic state and not by a toxic effect of recently reported that the hypercellularity of streptozotocin. the intima is reduced in restenotic tissue from TGF-fll is an important regulator of vascular diabetic patients while collagen-rich sclerotic smooth muscle cell growth and is reported to content is increased. 27 It was suggested that stimulate smooth muscle cell proliferation after an accelerated fibrotic rather than a prolife-arterial balloon injury. 6" 7] It has, however, also rative response is the main feature of restenosis been stated that TGF-fll might inhibit smooth in diabetic patients, muscle cell proliferation in vivo. [31 Expression In order to see whether diabetes duration of TGF-fll is correlated with increased extrahad an effect on proliferation of smooth muscle cellular matrix production and neointima for+/cells, BrdU incorporation was measured at mation. [6"31"321 In wounds, low levels of TGF-fl 2, 4 and 8 weeks after induction of diabetes, are considered to contribute to the impaired The proliferative response of smooth muscle healing seen in diabetes. I1-131 The role of cells in the intima-media remained low in TGF-fll in diabetic vascular diseases is, howrats with different diabetes duration without ever, still mainly unknown. We investigated the any significant change with time. Because, in possibility that the low proliferation of smooth our study, the proliferation of smooth muscle muscle cells in diabetic rats two days after incells in normal rats decreased along with injury might be associated with altered levels of creased age, the difference between controls circulating TGF-fll and/or a changed expression and diabetic rats was most pronounced 2 and 4 of TGF-fll as compared with injured controls weeks after induction of diabetes. It has been locally in the vessel wall. Levels of total cirreported that proliferation of vascular smooth culating TGF-fll was slightly but significantly muscle cells is enhanced by increased age. I281 lower in balloon injured diabetic rats as com-This does not explain the observed decline pared to balloon injured control rats. The same in smooth muscle cell proliferation in normal pattern was found in sham operated rats. Active rats in this study. The proliferative response TGF-fll in serum was measured but levels were of smooth muscle cells has been shown to below the lowest concentration in the standard be dependent on the severity of the arterial curve. However, a tendency to lower levels of injury. 71 It is possible that the control rats active TGF-fll was seen in diabetic rats as at 8 weeks diabetes duration, which were compared to controls in agreement with results considerably bigger than the control rats at on total TGF-fll. In contrast to these results 4 weeks of diabetes duration, were less injured Bollinenni and colleagues found elevated levels by the balloon angioplasty procedure and thus of circulating active TGF-fll in diabetic rats 6 weeks after induction by streptozotocin and levels were markedly higher than we found in this study. [141 The reason for such a discrepancy is not evident but might be partly due to the different methods used for measurement of TGF-fll, enzyme-linked immunosorbent assay in our study and a mink-lung epithelial cell bioassay in the study of Bollinenni and colleagues. Also the diabetes duration was different, 2 weeks and 6 weeks respectively.
A previous study, 61 reported that TGF-fll mRNA was elevated after balloon injury to the aorta of non-diabetic rats. We found a significant increase in TGF-fll mRNA by balloon injury in diabetic rats. This increased gene expression of TGF-fll might contribute to the proliferation of smooth muscle cells triggered by the balloon injury procedure. 6'71 TGF-fll mRNA in diabetic rats after injury was not, however, decreased as compared to injured controls. Thus, the low proliferation of smooth muscle cells in diabetic rats after injury was not related to a significant impairment of TGF-fll gene expression locally. However, other growth factors, as previously described for IGF-I, Isl might be of importance for regulating smooth muscle cell proliferation in diabetic rats after balloon injury.
In conclusion, balloon injury to rat aorta induced proliferation of vascular smooth muscle cells. The cell proliferation was impaired in streptozotocin diabetic rats. Although the injury procedure by itself increased TGF-fll mRNA levels, no significant difference in TGF-fll mRNA levels between injured diabetic rats and injured controls could be seen, suggesting that the inhibited smooth muscle cell proliferation in diabetic rats was not caused by an impaired local TGF-fll mRNA expression.