Impaired Arteriolar Mechanotransduction in Experimental Diabetes Mellitus

ORIGINAL ARTICLES

Impaired Arteriolar Mechanotransduction in Experimental Diabetes Mellitus Guanglei Yu Hui Zou Russell L. Prewitt Michael A. Hill

ABSTRACT Decreased arteriolar distensibility in diabetes may impair signal transduction mechanisms that are required for converting a pressure stimulus into smooth muscle contraction. These studies aimed to determine if pressure-induced increases in arteriolar intracellular Ca2ⴙ (Ca2ⴙ i ) are altered in diabetes and whether diabetes is associated with alterations in composition of the extracellular matrix. Studies of mechanical properties used single, isolated, and cannulated cremaster arterioles from streptozotocin (60 mg/kg) diabetic rats and age-matched controls. To measure Ca2ⴙ i , arterioles were loaded with Fura 2 (5 ␮M) after which preparations were examined by fluorescence microscopy and image analysis. Matrix protein (type IV collagen, laminin, fibronectin) deposition was studied by immunohistochemistry. Over a range of 30–120 mm Hg control vessels showed a linear relationship (r ⫽ 0.98, p ⬍ 0.01) between intraluminal pressure and Ca2ⴙ i . Vessels from diabetic animals also

showed a linear relationship (r ⫽ 0.99, p ⬍ 0.01), however, the mean slope was significantly (p ⬍ 0.02) less in the diabetic (0.17 ⫾ 0.05, n ⫽ 5) compared to controls (0.51 ⫾ 0.09, n ⫽ 7). Similarly, the slope of the wall tension-Ca2ⴙ i relationship was significantly decreased in vessels from diabetic animals. These differences were ameliorated by treatment of diabetic animals (n ⫽ 5) with aminoguanidine. Increased content of type IV collagen, laminin and fibronectin in vessel media was evident after 2 weeks of diabetes and showed a further increase with duration of diabetes. The data suggest that for a given increase in luminal pressure arterioles from diabetic animals response with an attenuated rise in smooth muscle Ca2ⴙ i . This mechanotransduction defect may relate to alterations in the composition of the extracellular matrix within the arteriolar wall. (Journal of Diabetes and Its Complications 13; 5/6: 235–242, 1999.) 2000 Elsevier Science Inc.

INTRODUCTION

Department of Physiology (G.Y., H.Z., R.L.P., M.A.H.), Eastern Virginia Medical School, Norfolk, Virginia, USA; and Department of Human Biology and Movement Science (M.A.H.), RMIT University, Bundoora, Victoria, Australia Reprint requests to be sent to: Dr. Michael A. Hill, Department of Human Biology and Movement Science, RMIT University, Plenty Road, Bundoora, Victoria 3083, Australia. Journal of Diabetes and Its Complications 1999; 13:235–242 2000 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010

D

espite advances in the treatment and monitoring of diabetes mellitus the secondary microvascular complications remain a major cause of the increased morbidity and mortality associated with this metabolic disorder. While the results of the Diabetes Control and Complications Trial1 has provided evidence linking hyperglycemia with the development and progression of microangiopathy the exact cellular mechanisms remain uncertain. Amongst a number of hypotheses it has 1056-8727/99/$–see front matter PII S1056-8727(99)00050-1

236

YU ET AL.

J Diab Comp 1999; 13:235–242

been suggested that early diabetes is associated with periods of hyperperfusion2 which through repeated episodes leads to altered vascular structure and ultimately diminished perfusion. Given that diabetes is associated with periods of inappropriate perfusion it is apparent that the normal regulatory mechanisms that control microvascular hemodynamics may be overridden or impaired. Local blood flow and pressure are, in part, controlled through a series of mechanisms that principally aim to match perfusion to local metabolic demand.3,4 Two such mechanisms rely on the ability of vascular cells to respond to physical stimuli through processes of mechanotransduction. For example, arteriolar smooth muscle responds to transmural forces exerted by an increase in intraluminal pressure with vasoconstriction (myogenic response)3,4 while endothelial cells respond to the shear stress imparted by blood flow with production of vasodilators such as nitric oxide.5,6 In the case of myogenic constriction, it is believed that an increase in intraluminal pressure exerts a stretch or distension stimulus to the smooth muscle cells, which respond by altering membrane permeability to Ca2⫹, phosphorylation of the myosin regulatory light chains, and ultimately contraction.7 In previous studies of arteriolar function we have demonstrated that experimental diabetes results in impaired myogenic reactivity8,9 and have suggested that this is a function of decreased distensibility.9 Diabetesinduced alterations is the mechanical properties of the arteriolar wall could conceivably result from alterations to the extracellular matrix through non-enzymatic glycosylation reactions and/or via excess deposition or accumulation of matrix proteins.10–13 The aims of the present studies were to extend our understanding of the effects of diabetes on arteriolar myogenic reactivity and, in particular, to determine (1) whether cannulated arterioles from diabetic rats show impaired intracellular Ca2⫹ (Ca2⫹i) responses to a change in intraluminal pressure and (2) whether such arterioles show alterations in the extracellular matrix component of the vessel wall. METHODS Animal Handling. The studies used male Sprague– Dawley rat, aged 4–5 weeks and weighing 120–150 grams. Diabetes was induced with a single injection of streptozotocin (STZ; 60 mg/kg, IM); induction of diabetes was confirmed by blood glucose levels (⬎ 15 mmol/L) 48 h after administration of STZ. Prior to study rats were housed in a temperature (24⬚C) and humidity (55%) controlled facility with a 12-h light/ dark cycle; during this period all animals were allowed free access to a standard chow diet and tap water. Selected groups of animals were treated with either subcutaneously administered insulin (5 U/kg am and

12 U/kg pm) or subcutaneously administered aminoguanidine (25 mg/kg/day). Treatments were initiated within 48 hrs of streptozotocin administration after confirming hyperglycemia. Insulin treatment was aimed at reducing blood glucose levels while aminoguanidine was administered to inhibit formation of advanced glycosylation reactions. The protocol for use of animals was approved by the Animal Care and Use Committee of Eastern Virginia Medical School. Isolation and Cannulation of Arterioles. Studies of mechanical properties of arterioles were performed on tissues taken from 4 to 6 weeks duration diabetic animals and aged-matched controls. For surgical removal of vascular tissue rats were anesthetized with sodium thiopental (Pentothal, 20 mg/100g body weight). Cremaster muscles were surgically removed from the anesthetized rats and placed into a plexiglass chamber containing cooled (4⬚C) dissection buffer (3 mM 3-Nmorpholino propanesulfonic acid (MOPS); 145 mM NaCl; 5 mM KCl; 2.5 mM CaCl2; 1 mM MgSO4; 1 mM NaH2PO4; 0.02 mM EDTA; 2 mM pyruvate; 5 mM glucose and 1% albumin).14 Segments of first-order arteriole (active diameter approximately 80 ␮m) were cleared from surrounding tissues and cannulated onto glass micropipettes as previously described. The cannulated vessels were positioned in a custom designed superfusion system which was placed on the stage of an inverted microscope which was coupled to a CCD camera and high resolution monitor. The vessel segments were superfused with a physiological salt solution (34⬚C, 4 mL/min), pressurized to 70 mm Hg (in the absence of intraluminal flow) and allowed to develop spontaneous tone. To be considered viable and suitable for study, arterioles were required to be free of pressure leaks, develop spontaneous tone and exhibit myogenic reactivity to acute changes in intraluminal pressure. Intraluminal pressure in the isolated arteriolar segment was controlled by adjusting the height of a fluid reservoir which was connected via polyethylene tubing to one of the cannulation pipettes. To quantitate the response of cremaster muscle arterioles to a given change in transmural pressure the vessel lumen diameter was measured on the video monitor using a calibrated electronic caliper.15 For measurement of changes in intracellular Ca2⫹ arterioles were loaded with the Ca2⫹ sensitive fluorescent indicator fura 2. In brief, vessel segments were exposed, via the albuminal surface to 5 ␮M fura 2-AM in 0.05% DMSO and 0.01% pluronic for 60 minutes followed by a 30-min wash period with physiological salt solution. Fura 2 loaded vessels were exposed to epi-illumination (75W Xenon source) with light of alternating excitation wavelengths (340 and 380 nm) using a computer controlled filter wheel. Fluorescent images were transferred from the microscope (x20 Nikon

J Diab Comp 1999; 13:235–242

ARTERIOLAR MECHANOTRANSDUCTION IN DIABETES

237

Fluar objective lens, N.A. 0.75) to the imaging system using an image intensifier (Videoscope International, Washington, D.C.) and a charge-coupled device (CCD; Hamematsu, Bridgewater, NJ). Fluorescent image intensities were expressed as the 340/380 nm ratio to allow quantitative estimates of changes in arteriolar wall intracellular Ca2⫹. Details of these methods are described in previous publications.7,16 Tissue Preparation for Immunohistochemical Staining. Segments of femoral artery, small mesenteric arteries and cremaster first order arterioles were removed from anesthetized rats and fixed in freshly prepared 4% paraformaldehyde in PBS. Segments were then processed through a graded series of ethanol, xylene and embedded in paraffin. Tissue sections (5 ␮m) from matched diabetic, treated and control animals were mounted on the same glass slide and processed simultaneously. This procedure was performed to minimize variation in staining between immunohistochemical determinations. For staining, tissues were deparaffined and rehydrated using standard procedures. To increase antigenicity, tissues were incubated (3 min, 22⬚C) with 0.1% pepsin diluted in 0.01N HCl. To quench endogenous peroxidase sections were treated with 3% hydrogen peroxide in methanol. To reduce non-specific binding 5% non-immune goat sera and 2% bovine serum albumin were applied to each section for 20 min prior to the application of a given primary antibody. For identification of specific matrix proteins tissue sections were exposed to anti-collagen type IV (1:100 dilution; ICN Laboratories), anti-laminin (1:200; E-Y Laboratories) or antifibronectin (1:400; ICN or Gibco) antibodies for 30 min. For negative controls, non-immune rabbit IgG was used at concentrations matched to that of the primary antibody. After washing, tissue sections were then incubated with biotinylated goat anti-rabbit secondary antibody (1:800) followed by extravidinhorseradish peroxidase. Color was then developed utilizing the substrate 3-amino-9-ethylcarbazole (25 mg) in 2.5 mL dimethylformamide, 25 ␮L 30% hydrogen peroxide and 47.5 mL 50mM acetate buffer. Segments were counterstained with hematoxylin. The relative staining of sections was evaluated using a video-based analysis image system (Universal Imaging, PA) coupled to an upright Nikon microscope. Degree of staining was assigned using a scale of 0 to 4⫹. As stated above, to minimize variation between samples, sections from control and diabetic tissues were processed simultaneously and analyzed under identical conditions. Data Analysis. For simple comparisons of two means the Student t-test was used to calculate levels of significance. For comparisons of multiple groups one way analysis of variance (ANOVA) was used. Pairwise

FIGURE 1 [Ca2⫹]i-pressure relationships for isolated arterioles from control, streptozotocin diabetic and aminoguanidine-treated streptozotocin diabetic animals. The upper panel illustrates the data in terms of percent change in the fluorescent ratio (relative to that at 30 mm Hg) while the lower panel shows the same data in terms of percent change in calculated intracellular Ca2⫹. Results are expressed as mean ⫾ SEM.

comparison were then made with Bonferroni corrected t-tests. Linear regression analysis was applied to the intraluminal pressure—Ca2i ⫹ relationships for isolated arterioles. Significance was considered to occur when p ⬍ 0.05. RESULTS Intraluminal Pressure-Ca2ⴙ i Relationships for Arterioles Isolated from Control and STZ Diabetic Animals. Over a range of 30–120 mm Hg control vessels showed a linear relationship (r ⫽ 0.98, p ⬍ 0.01) between intraluminal pressure and Ca2i ⫹ (Figure 1). Vessels from diabetic animals similarly showed a linear relationship (r ⫽ 0.99, p ⬍ 0.01), however, the mean slope was significantly (p ⬍ 0.02) less in the diabetic group (0.17 ⫾ 0.05, n ⫽ 5) compared to that of controls (0.51 ⫾ 0.09, n ⫽ 7) (Figure 1). These data indicate that for a given change in intraluminal pressure, arterioles from diabetic animals showed an impaired Ca2i ⫹ response compared to that of controls. The defect is partially prevented by treating animals with aminoguani-

238

YU ET AL.

FIGURE 2 Relationships between arteriolar [Ca2⫹]i and calculated wall tension (pressure ⫻ radius) for isolated arterioles from control (upper panel), streptozotocin diabetic (middle panel) and aminoguanidine-treated streptozotocin diabetic (lower panel) animals. The individual data points were calculated from the group data shown in Figure 1.

dine; a compound known to prevent the formation of advanced glycosylation endproducts (Figure 1). As our previous studies7 in arterioles from nondiabetic animals have shown that Ca2i ⫹ is more closely related to wall tension (tension ⫽ pressure ⫻ radius), than either pressure or vessel radius alone, the individual Ca2⫹ data points in Figure 1 have been plotted in terms of calculated wall tension (Figure 2). Similar to the data above, both control vessels (Ca2i ⫹ ⫽ 0.7 ⫻ wall tension ⫹ 87.3) and arterioles from diabetic animals (Ca2i ⫹ ⫽ 0.2 ⫻ wall tension ⫹ 99.2) showed linear relationships between Ca2i ⫹ and calculated wall tension (Figure 2). The slope of this relationship was significantly decreased in arterioles from diabetic rats (Figure 2). Treatment with aminoguanidine normalized the Ca2i ⫹-tension relationship (Ca2i ⫹ ⫽ 0.7 ⫻ wall tension ⫹ 89.0) without having an effect on blood glucose levels.

J Diab Comp 1999; 13:235–242

FIGURE 3 The figure shows typical immunohistochemical staining patterns for matrix proteins within the walls of selected arteries. Panels A and B show fibronectin staining for control and 2-week diabetic femoral arteries, respectively. Panels C and D illustrates fibronectin staining for control and 4-week diabetic femoral arteries respectively. Note that intensity of staining increases with duration of experimental diabetes. Panels E and F shows representative laminin staining for control and 4 week diabetic femoral arteries respectively. See methods for immunohistochemical procedures.

Effect of Experimental Diabetes on Arterial Wall Matrix Protein Content. Matrix proteins were identified immunohistochemically and the extent of staining was graded on a relative scale of 0–4. Sections of arteries from diabetic and matched controls were simultaneously processed on the same slide to minimize variability. A generalized accumulation of type IV collagen, fibronectin and laminin in the medial layer of femoral, mesenteric and cremaster arteries was apparent as early as 2 weeks after induction of STZ diabetes. The intensity of matrix protein staining appeared to increase with duration of diabetes. Examples of immunohistochemical identification of matrix proteins are shown in Figure 3 and group data for the staining index are presented in Figures 4 and 5. Aminoguanidine treatment of STZ-diabetic animals

J Diab Comp 1999; 13:235–242

ARTERIOLAR MECHANOTRANSDUCTION IN DIABETES

FIGURE 4 Group data showing the effect of 2 weeks of experimental diabetes on arterial wall matrix protein content. Results are shown for cremaster arteriole (CA), femoral artery (FA) and small mesenteric artery (MA). Results are presented as mean ⫾ SEM for the staining index; *p ⬍ 0.05.

(4–6 weeks) resulted in a significant decrease in type IV collagen staining in the media of cremaster, mesenteric and femoral arteries relative to that of untreated diabetic animals (Figure 6a). Although there appeared to be a trend towards decreased staining for fibronectin and laminin in aminoguanidine treated diabetic animals (relative to untreated diabetic rats) this only reached statistical significance in the case of laminin in the media of femoral arteries (p ⬍ 0.05). To demonstrate whether or not the observed accumulation of arterial wall matrix proteins was related to the diabetic state an additional group of streptozotocin-treated animals were given exogenous insulin.

239

FIGURE 5 Group data showing the effect of 4 weeks of experimental diabetes on arterial wall matrix protein content. Results are shown for cremaster arteriole (CA), femoral artery (FA) and small mesenteric artery (MA). Results are presented as mean ⫾ SEM for the staining index; *p ⬍ 0.05. Note that durations of 2 and 4 weeks were not quantitatively compared to each other.

Insulin treatment (4 weeks) of diabetic animals caused a partial normalization of matrix protein staining and significantly reduced matrix accumulation relative to that of untreated diabetic animals. As an example, data for the effect of insulin treatment on type IV collagen in the media of cremaster, mesenteric and femoral arteries are shown in Figure 6b. DISCUSSION The results of these studies demonstrate impaired mechanotransduction in arterioles isolated from rats with streptozotocin-induced diabetes. In particular it was shown that in arteriolar smooth muscle the increase in Ca2i ⫹, which follows a rise in intraluminal

240

YU ET AL.

FIGURE 6 Effect of aminoguanidine (upper panel; n ⫽ 7 per group) and insulin (lower panel: n ⫽ 5 per group) treatment on type IV collagen accumulation in the media of cremaster, mesenteric and femoral arteries. Results are presented as mean ⫾ SEM of the staining index; *p ⬍ 0.05.

pressure, is significantly attenuated. Importantly the diabetes-induced effect occurred over a physiologically relevant pressure range. The physiological significance of this observation is that such an alteration in Ca2⫹ handling would be expected to lead to impaired myogenic autoregulatory processes during periods of altered systemic arterial pressure. A second observa-

J Diab Comp 1999; 13:235–242

tion was that this defect in mechanotransduction appears to be related to advanced non-enzymatic glycosylation reactions as the defect in Ca2⫹ handling was prevented by treatment of diabetic animals with aminoguanidine. This latter finding is consistent with previous studies of ours9 and others17 showing that aminoguanidine treatment prevents diabetes-induced stiffening of the arterial wall. An additional factor possibly impacting on the mechanical properties of the arteriolar wall was provided by the observation that experimental diabetes resulted in accumulation of matrix proteins (type IV collagen, laminin, fibronectin) within the vessel media. Most tissues exhibit local microvascular autoregulatory mechanisms that are aimed in providing a constancy of hemodynamic conditions while also matching perfusion with metabolic demands. One such mechanism is the myogenic response that is manifested in the ability of arterioles to constrict in response to an increase in intraluminal pressure or dilate following a decrease in pressure.3 Thus the change in vessel caliber acts, at least in part, to offset the effects of an alteration in perfusion pressure on local blood flow. Myogenic responsiveness of arteriolar smooth muscle is also considered to set the basal level of vascular tone upon which neurohumoral factors can act to alter vascular resistance.3 The stimulus for this vasoactive response is thought to be the vascular smooth muscle stretch or transmural wall tension exerted by the blood pressure.3,7 Given this local property of the microcirculation we have speculated that if early diabetes is associated with disturbances in microvascular hemodynamics then such control mechanisms must be overridden or fundamentally altered. While the exact cellular mechanisms that transduce an intraluminal pressure stimulus across arteriolar smooth muscle membranes and ultimately into an active mechanical response are uncertain, previous studies have demonstrated relationships between Ca2i ⫹ and the degree of wall tension or cell stretch.7,18 As such, alterations in the distensibility of the arteriolar wall may be expected to impact on signaling mechanisms underlying myogenic reactivity. Consistent with this we have previously reported that arterioles from diabetic animals are less distensible, compared to those of control animals, and that diabetes results in a stiffening of blood vessels indicated by a rightward shift in their stress-strain curve (9). The present study extends these findings to show that, for a given change in intraluminal pressure, arterioles from diabetic animals show an impaired ability to increase Ca2i ⫹ relative to that seen in control vessels. This impaired ability to increase Ca2i ⫹ would be expected to lead to a decreased vasoconstriction as Ca2⫹ is critical to development of a normal constrictor response. Consistent with previous studies, treatment of ani-

J Diab Comp 1999; 13:235–242

ARTERIOLAR MECHANOTRANSDUCTION IN DIABETES

mals for the duration of the study with aminoguanidine prevented diabetes-induced alterations in arteriolar wall mechanical properties.9,17 In particular administration of aminoguanidine normalized the relationship between arteriolar wall tension and Ca2i ⫹. As aminoguanidine is known to inhibit the formation of advanced glycosylation endproducts and the cross linking of long lived proteins19,20 it is conceivable that the alteration in arteriolar wall mechanical properties results from structural changes in the extracellular matrix. An additional possibility is that the presence of advanced glycosylation endproducts within the vascular wall leads to stimulation of matrix synthesis or a decrease in degradation which results in a net accumulation of such matrix proteins.21,22 Such an effect could be mediated for example through interaction of glycated species with cell surface receptors which have been shown to be linked to the production of growth factors (e.g., TGF␤)22 or as a result of glycation reactions altering the activity or accessibility of matrix modifying enzymes.23 Consistent with this suggestion it was of interest to note that in the present study aminoguanidine treatment not only improved mechanical properties of the arteriolar wall but also appeared to decrease the amount of immunoreactive type IV collagen within the vessel media. Similarly, Rumble et al.22 reported that aminoguanidine treatment of diabetic rats decreased both the mRNA for type IV collagen and the immunoreactive protein in the media of small mesenteric arteries. Interestingly these authors also found that such effects were evident early (3 weeks) in the timecourse of experimental diabetes. The exact mechanisms by which diabetes-induced changes in matrix protein composition alter vessel function are not fully understood. As suggested above simple alterations in arteriolar distensibility could conceivably impair mechanotransduction via decreasing the extent of a stretch stimulus that results from a given change in intravascular pressure. An additional possibility arises, however, from recent studies demonstrating that matrix proteins interacting with cell surface proteins such as integrins can modulate intracellular signaling.24–27 For example, using in vitro electrophysiological techniques it has been shown that fibronectin fragments acting via ␣5␤1 integrins can modulate voltage gated Ca2⫹ entry into vascular smooth muscle cells.27 This observation together with studies showing that cyclic RGD peptides decrease vascular smooth muscle intracellular Ca2⫹, with subsequent arteriolar relaxation,28 indicate that integrin binding may modulate vascular tone. Thus a change in matrix proteins could more directly affect smooth muscle cell signaling and therefore the functional properties of arterioles. Further, given that there is specificity in the binding of matrix proteins to a variety of cell surface receptors the increased deposition of matrix proteins,

241

in general, could result in multiple vascular abnormalities. Diabetes-induced alterations in matrix synthesis could conceivably result in impaired vascular function via effects on cells other than smooth muscle. For example, alterations to the basal lamina underlying endothelial cells may result in altered shear stress-induced release of nitric oxide5,6 which would impair normal mechanisms of flow-dependent dilatation. While such dilation is a further example of mechanotransduction it is unlikely that an endothelial impairment contributed to the current results as the myogenic response was studied in the absence of intraluminal flow. Under these conditions the endothelium does not exert a direct effect on arteriolar tone.29 In summary, the present study provides further evidence for impaired arteriolar function early in the course of experimental diabetes. The results demonstrate impaired mechanotransduction as shown by an impaired ability of arteriolar vascular smooth muscle to raise intracellular Ca2⫹ in response to an increase in intraluminal pressure. Such a defect would result in a decreased ability to respond with vasoconstriction during an increase in arterial pressure thus exposing the distal vasculature to periods of hyperperfusion and increased capillary pressure. A possible mechanism underlying impaired mechanotransduction is an alteration to the vessel wall through deposition of matrix proteins and/or advanced non-enzymatic glycosylation reactions. ACKNOWLEDGMENT Studies described in this manuscript have been supported by NIH (NHLBI), Juvenile Diabetes Foundation International and the National Health and Medical Research Council of Australia.

REFERENCES 1. Diabetes Control and Complications Trial Study Group. Effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New Eng J Med 329:977–968, 1993. 2. Tooke JE: Microvascular physiology in diabetes: A physiologic perspective. Diabetes 44:721–726, 1995. 3. Davis MJ, Hill MA: Signaling mechanisms underlying the vascular myogenic response. Physiological Rev 79:387–423, 1999. 4. Johnson PC: Autoregulation of blood flow. Circ Res 59:483–495, 1986. 5. Davies PF: Flow-mediated endothelial mechanotransduction. Physiological Rev 75:519–560, 1995. 6. Busse R, Fleming I: Pulsatile stretch and shear stress: physiological stimuli determining the production of endothelium derived relaxation factors. J Vasc Res 35:73– 84, 1998. 7. Zou H, Ratz PH, Hill MA: Role of myosin phosphoryla-

242

YU ET AL.

J Diab Comp 1999; 13:235–242

tion and [Ca2⫹)i in myogenic reactivity and arteriolar tone. Am J Physiol 269:H1590–H1596, 1995.

mechanotransduction of the myogenic response. Am J Physiol 273:H175–H182, 1997.

8. Hill MA, Meininger GA: Impaired arteriolar myogenic reactivity in early experimental diabetes. Diabetes 42:1226–1232, 1993.

19. Brownlee M, Vlassara H, Kooney A, Ulrich P, Cerami A: Aminoguanidine prevents diabetes-induced arterial wall protein cross-linking. Science 232:1629–1632, 1986.

9. Hill MA, Ege EA: Active and passive mechanical properties of isolated arterioles from STZ-induced diabetic rats: Effect of aminoguanidine treatment. Diabetes 43:1450–1456, 1994.

20. Booth AA, Khalifah RG, Todd P, Hudson BG: In vitro kinetic studies of formation of antigenic advanced glycation endproducts (AGEs). J Biol Chem 272:5403–5437, 1997.

10. Rasmussen LM, Ledet T: Aortic collagen alterations in human diabetes mellitus. Changes in basement membrane collagen content and in the susceptibility of total collagen to cyanogen bromide solubilization. Diabetologia 36:445–453, 1993.

21. Cohen MP, Hud E, Wu VY, Ziyadeh FN: Albumin modified by Amadori adducts activates mesangial cell type IV collagen gene transcriptions. Mol Cell Biochem 151:61– 67, 1995.

11. Tomizawa H, Yamazaki M, Kunika K, Itakura M, Yamashita K: Association of elasting glycation and calcium deposit in diabetic rat aorta. Diabetes Res Clin Pract 19:1–8, 1993. 12. Sims TJ, Rasmussen LM, Oxlund H, Bailey AJ: The role of glycation cross-links in diabetic vascular stiffening. Diabetologia 39:946–951, 1996. 13. Vranes D, Cooper ME, Dilley RJ: Cellular mechanisms of diabetic vascular hypertrophy. Microvasc Res 57:8–18, 1999. 14. Duling BR, Gore RW, Dacey Jr RG, Damon DN: Methods of isolation, cannulation, and in vitro study of single microvessels. Am J Physiol 241:H108–H116, 1991. 15. Goodman AH: Un calibreur video simple pour l’utilisation en microscopie video. Innov Tech Biol Med 9:350– 356, 1988. 16. Meininger GA, Zawieja DC, Falcone JC, Hill MA, Davey JP: Calcium measurement in isolated arterioles during myogenic and agonist stimulation. Am J Physiol 261:H950–H959, 1991. 17. Huijberts MSP, Wolffenbuttel BHR, Struijker Boudier HAJ, Crijns FRL, Nieuwenhuijzen Kruseman AC, Poitevin P, Levy BI: Aminoguanidine treatment increases elasticity and decreases fluid filtration of large arteries from diabetic rats. J Clin Invest 92:1407–1411, 1993. 18. D’Angelo G, Davis MJ, Meininger GA: Calcium and

22. Rumble JR, Cooper ME, Soulis T, Cox A, Wu L, Youssef S, Jasik M, Jerums G, Gilbert RE: Vascular hypertrophy in experimental diabetes. Role of advanced glycation endproducts. J Clin Invest 99:1016–1027, 1997. 23. Mott JD, Khalifah RG, Nagase H, Shield CF, Hudson JK, Hudson BG: Nonenzymatic glycation of type IV collagen and matrix metalloproteinase susceptibility. Kidney Int 52:1302–1312, 1997. 24. Clark EA, Brugge JS: Integrins and signal transduction pathways: The road taken. Science 268:233–239, 1995. 25. Schoenwaelder SM, Burridge K: Bidirectional signaling between the cytoskeleton and integrins. Curr Opin Cell Biol 11:274–286, 1999. 26. Clark S, Muggli E, La Greca N, Dunlop ME: Increased phosphorylation of focal adhesion kinase in diabetic rat kidney glomeruli. Diabetologia 38:1131–1137, 1995. 27. Wu X, Mogford JE, Platts SH, Davis GE, Meininger GA, Davis MJ: Modulation of calcium current in arteriolar smooth muscle by integrin receptor ligands. J Cell Biol 143:241–252, 1998. 28. D’Angelo G, Mogford JE, Davis GE, Davis MJ, Meininger GA: Integrin-mediated reduction in vascular smooth muscle [Ca2⫹]i induced by RGD-containing peptide. Am J Physiol 272:H2065–H2070, 1997. 29. Falcone JC, Davis MJ, Meininger GA: Endothelial independence of the myogenic response in isolated skeletal muscle arterioles. Am J Physiol 260:H130–H135, 1991.