The AGE-Breaker ALT-711 Restores High Blood Flow–Dependent Remodeling in Mesenteric Resistance Arteries in a Rat Model of Type 2 Diabetes

Twenty adult male ZDF and 20 lean Zucker (LZ) rats, 11–12 weeks old, were purchased from Charles River (L’Arbresles, France). They were anesthetized (isoflurane, 2.5%) and submitted to surgery to modify blood flow, as previously described (19,21,22). In brief, three consecutive first-order mesenteric arteries were used. Ligatures (7–0 silk surgical thread) were applied to second-order branches of the first and third arteries, as shown in Fig. 1A. The artery located between two ligated arteries was designated as an HF artery. Other arteries located further from the ligated arteries were used as control (NF) arteries. Rats were treated with buprenorphine (0.1 mg/kg s.c. temgesic) before and after surgery. One-half of the rats were treated by intraperitoneal injection with ALT-711 (3 mg/kg/day). Treatment started 1 week before surgery. Consequently, the total duration of the treatment was 3 weeks.

Fourteen days after surgery, the animals were anesthetized (isoflurane, 2.5%). The right femoral artery was catheterized for blood pressure measurement (23). The animals then were killed by CO₂ inhalation, the gut excised, and the mesenteric arteries gently dissected. HF and NF arteries from each rat were isolated and divided into several segments used respectively for pressure-diameter relationship measurement, pharmacology, biochemistry, and immunohistological analyses. Before the animals were killed, glycemia was quantified on a sample of arterial blood with a glucometer (9).

The procedure followed in caring for and killing the study animals complied with the European Community Standards on the Care and Use of Laboratory Animals (authorization no. 6422; Ministère de l’Agriculture, France) and the Principles of Laboratory Animal Care (NIH publication no. 85–23, revised 1985; available at http://grants1.nih.gov/grants/olaw/references/phspol.htm). The protocol was approved by the regional ethics committee (protocol CEEA PdL 2008.10).

Segments of NF and HF arteries were cannulated at both ends and mounted on a video-monitored perfusion system (24). Arterial segments were perfused and superfused with a Ca²⁺-free physiological salt solution (PSS) containing EGTA (2 mmol/L) and sodium nitroprusside (SNP) (10 μmol/L) and submitted to a stepwise increase in pressure (10–150 mmHg) in order to determine arterial passive diameter. Data were recorded using a Biopac data acquisition system (La Jolla, CA) and analyzed (Acqknowledge software). Arterial compliance was calculated as previously described (25). In short, during the stepwise increase in pressure (10–150 mmHg), the cross-sectional compliance (CSC) of mesenteric resistance arteries was calculated by the ratio of lumen area (∆A) and pressure (∆P) as CSC = ∆A/∆P.

Other segments of HF and NF mesenteric arteries (2 mm long) were dissected and mounted on a wire myograph (DMT, Aarhus, DK) (26). Two tungsten wires (25 μm in diameter) were inserted into the lumen of the arteries and fixed to a force transducer and a micrometer, respectively. Arteries were bathed in a PSS. Wall tension was applied as previously described (27). The artery’s viability was tested using a potassium-rich solution (KCl, 80 mmol/L). A cumulative concentration-response curve (CRC) to acetylcholine (ACh; 0.001 to 10 μmol/L) was performed after a precontraction induced by phenylephrine to ~50% of the maximal response. Then, a CRC to 5-hydroxy-tryptamine (0.001–10 μmol/L) was completed. Thirty minutes after washout, a second CRC to ACh (0.001–10 μmol/L) was performed in the presence of l-NG-nitro-l-arginine methyl ester (l-NAME) (100 μmol/L) or in the presence of superoxide dismutase (120 units/mL) plus catalase (80 units/mL). Finally, a CRC to SNP (0.001–10 μmol/L) was performed. In another series of experiments, a CRC to l-NAME (0.01–100 μmol/L) was performed after a precontraction induced by phenylephrine to ~10% of the maximal response.

The remaining NF and HF arteries of each mesenteric vascular bed were pooled and then homogenized. Proteins (25 μg total protein from each sample) were separated by SDS-PAGE using a 4% stacking gel, followed by a 10% running gel. Proteins were detected with specific antibodies (eNOS 1:1,000, phospho-eNOS 1:500, AGE 1:100, receptor for advanced end products [RAGE] 1:500, and β-actin 1:1,000 in bovine serum albumin in Tris-buffered saline with Tween [T-TBS-BSA] 5%; Transduction Laboratories). Protein expression was visualized using the ECL Plus chemiluminescence kit (Amersham) (28).

As previously described (24), segments of mesenteric resistance arteries were mounted in embedding medium (Tissu-Tek; Miles), frozen in isopentane precooled in liquid nitrogen, and stored at −80°C. Transverse cross-sections (7 μm thick) were used for cyclooxygenase (COX)-2 (24), advanced end products (AGEs) (29), and methylglyoxal-derived hydroimidazolene (MG-¹H) (30). Fluorescence staining was visualized using confocal microscopy and image analysis (Histolab; Microvision, Paris, France) (9).

MMP activity was determined by in situ zymography, as previously described (31). Arterial segments isolated from LZ or ZDF rats were incubated for 15 min with angiotensin II (100 nmol/L) in PSS. Arteries then were quickly embedded vertically in Tissue-Tek and frozen. Frozen sections (7 μm thick) were incubated overnight (37°C) with a fluorogenic gelatin substrate (Molecular Probes) dissolved to 25 μg/mL in zymography buffer (50 mmol/L Tris-HCl, 10 mmol/L CaCl₂, and protease inhibitor cocktail, pH 7.4). The gelatin with a fluorescent tag remains caged until the gelatin is cleaved by gelatinase activity. In situ gelatinolysis was revealed by the appearance of fluorescence visualized and quantified using confocal microscopy. In control experiments, sections were incubated with MMP inhibitors (1,10-phenanthroline or EDTA).

Segments of mesenteric arteries were bathed in Ca²⁺-free PSS containing 10 μmol/L SNP. Pressure was set at 75 mmHg, and the artery was fixed in a 10% buffered formaldehyde solution, as previously described (32). Sections (7 μm thick) were stained with orcein. External diameter, lumen diameter, and media thickness were determined after image acquisition (Olympus T100 microscope, Sony camera) and analyzed using Histolab software (Microvision) for cross-sectional area calculation, as previously described (6).

Results were expressed as a means ± SEM. Significance of the variances between groups was determined by a two-way ANOVA for repeated measurements, in pressure-diameter or CRCs. For the other experiments, we used a one-way ANOVA, followed by the Bonferroni post hoc test or a t test. Values of P < 0.05 or lower were considered to be statistically significant.

Rat body weight was slightly, but not significantly, higher in ZDF rats than in LZ rats (363 ± 17 g vs. 336 ± 21 g; P < 0.05) and was not significantly affected by ongoing treatment with ALT-711 in either ZDF (353 ± 10 g) or LZ (347 ± 19 g) rats. Blood glucose was significantly enhanced in ZDF rats compared with LZ rats (432 ± 55 mg/dL vs. 157 ± 18 mg/dL). Blood glucose was not modified by continued treatment with ALT-711 in either ZDF (447 ± 32 mg/dL) or LZ (166 ± 15 mg/dL) rats. Type 2 diabetes significantly increased mean blood pressure (92 ± 3 mmHg in ZDF vs. 80 ± 3 mmHg in LZ rats; P < 0.05). However, mean blood pressure was not affected by ongoing treatment with ALT-711, in either the ZDF (97 ± 4 mmHg) or LZ (83 ± 3 mmHg) rats.

In LZ rats, HF artery diameter was significantly larger than in NF arteries (Fig. 1A). In contrast, HF artery diameter in ZDF rats did not increase significantly compared with NF arteries (Fig. 1B). Thus, arterial outward remodeling did not occur in ZDF rats. In LZ rats treated with ALT-711, remodeling was similar to that in control rats (Fig. 1C and E). However, in ZDF rats, ALT-711 restored HF-induced remodeling (Fig. 1D and F) without changing arterial diameter in mesenteric arteries submitted to NF, compared with control rats.

In ZDF rats, in vitro CSC was significantly lower in both HF and NF arteries, compared with LZ animals. The change in vessel CSC during the stepwise increase of intraluminal pressure from 10 to 50 mmHg was greater in the ZDF group of ALT-711–treated animals, whereas changes in CSC in the ALT-711–treated LZ group showed no difference. Hence, the CSC of the mesenteric resistance arteries was significantly increased with the AGE-breaker treatment (Fig. 2A and B).

Chronic increase in blood flow significantly increased the media cross-sectional area of HF arteries in LZ rats. No difference in media cross-sectional areas was observed between HF and NF arteries from ZDF rats. Nevertheless, type 2 diabetes induced an increase in the media cross-sectional area of the NF artery, compared with LZ rats. The treatment with ALT-771 significantly decreased the media cross-sectional area of NF arteries in ZDF rats (Fig. 2C).

In both HF and NF arteries from ZDF rats, ACh-induced relaxation was lower than in LZ animals. In rats treated with ALT-711, ACh-induced relaxation in HF and NF arteries was improved compared with corresponding arteries from untreated rats (Fig. 3A and C).

To determine the role of NO in endothelium-dependent dilation, CRCs to ACh were performed in the presence of the NOS inhibitor, l-NAME. In LZ rats, l-NAME abolished ACh-induced relaxation in HF and NF arteries. However in ZDF rats, inhibition of ACh-dependent relaxation with l-NAME was greater under ALT-711 treatment in both NF and HF arteries when compared with vessels from untreated animals (Fig. 3B and D). The further addition of cyclooxygenase inhibitor indomethacin did not significantly affect dilation (data not shown).

l-NAME–mediated contraction was significantly lower in arteries (NF and HF) from ZDF than from LZ rats, and l-NAME–mediated contraction was higher in ZDF rats treated with ALT-711 than in nontreated rats (Fig. 3E and F).

The association of SOD plus catalase improved ACh-mediated dilation in both NF and HF arteries from ZDF rats, whereas it did not affect the dilation in arteries isolated from LZ or ZDF rats treated with ALT-711 (Fig. 4). Serotonin (5-hydroxy-tryptamine)-induced contraction and SNP-induced relaxation were not affected by chronic changes in blood flow, by type 2 diabetes or by ALT-711 treatment (Fig. 5).

As previously shown by our group (9), eNOS expression levels increased in HF arteries compared with NF vessels in LZ rats (Supplementary Fig. 1). Endothelium dysfunction in ZDF rats was associated with a decreased eNOS expression in HF and NF arteries compared with LZ animals. Of interest, ALT-711 significantly increased eNOS expression in arteries from ZDF rats. Nevertheless, this change in eNOS expression may not totally explain the changes in NO-dependent dilation as ACh-mediated phosphorylation of eNOS increased in both LZ and ZDF rats, treated or not with ALT-711 (Supplementary Fig. 2).

Mesenteric resistance arteries revealed an increase of AGE formation and RAGE expression in ZDF rats compared with LZ rats. ALT-711 significantly reduced AGE formation and RAGE expression in both HF and NF arteries from diabetic rats (Supplementary Fig. 1). Immunolocalization of AGEs showed the presence of AGEs and MG-¹H in mesenteric arteries from ZDF rats and a reduction in AGEs and MG-¹H level in ALT-711–treated rats (Fig. 6). Nevertheless, ALT-711 reduced AGE and MG-¹H levels in ZDF rats without normalizing their level.

Because MMP activation plays a key role in high flow–mediated remodeling (31), we measured MMP activity (33) in mesenteric arteries (Fig. 7). In arteries from LZ rats, angiotensin II significantly increased MMP activity as assessed by an in situ zymographic gelatinase activity assay. On the other hand, in ZDF rats, angiotensin II induced no significant increase in MMP activity, whereas in ALT-711–treated ZDF rats, angiotensin II significantly increased MMP activity (Fig. 7).

In the current study, we found that type 2 diabetes impaired the ability of mesenteric arteries to remodel and improve NO-dependent dilation in response to a chronic increase in blood flow. Indeed, AGEs and RAGEs were overexpressed, whereas eNOS level and MMP activity were reduced in arteries from ZDF rats. Ongoing treatment with the AGE-breaker ALT-711 restored the ability of mesenteric arteries from ZDF rats to increase their diameter and improved endothelium-dependent dilation in response to a chronic rise in blood flow.

Physiologically, a chronic rise in blood flow in resistance arteries enlarges vascular diameter and improves endothelium-dependent dilation (9,19,34). This remodeling is essential to adjust organ perfusion during physiological processes, such as development (35), pregnancy (36), or exercise training (37), as well as during pathological processes (mainly ischemic diseases). A similar remodeling also occurs in response to vasodilator treatments (38,39). This remodeling plays a key role in revascularization after occlusion of a large artery, in addition to arteriogenesis and angiogenesis (40). The advantage of the model used in the current study is that it involved resistance arteries and allowed the study of the effects of blood flow on the arterial diameter and wall thickness, independent of blood pressure or metabolic changes and without ischemia.

In type 2 diabetes, the endothelium is less capable of inducing vasodilatation, especially in resistance arteries, which control blood supply to end organs (41). To improve endothelium-dependent dilation, and consequently local blood flow, vasodilator treatments, therapies improving insulin sensitivity, and exercise are commonly used. These treatments are associated with a higher eNOS expression, which is at least partly the consequence of a chronic rise in blood flow (9,34). Here, we show that despite no change in blood glucose level, ALT-711 increased eNOS protein expression in HF and NF arteries from ZDF rats (Fig. 5A). Nevertheless, this may not necessarily explain the improvement in ACh-dependent relaxation observed in mesenteric resistance arteries (Fig. 3A and C) because ACh-mediated phosphorylation of eNOS was equivalent in the different experimental groups. It is most likely that an excessive ROS production reduces NO bioavailablity in arteries of ZDF rats and consequently endothelium-mediated relaxation. Indeed, both inflammation, evidenced by an increased MCP-1, CD68, COX-2, and inducible NOS expression mRNA level (Supplementary Fig. 3) and ROS production, evidenced by a higher 3-nitrotyrosine expression (Supplementary Fig. 4), were higher in arteries from ZDF rats, compared with LZ rats. An immunohistological detection of inducible NOS and COX-2 confirmed the occurrence of inflammation in ZDF rat arteries (Supplementary Figs. 5 and 6). A treatment with ALT-711 reduced both inflammation and oxidative stress.

Our findings are in agreement with a work performed by Su et al. (17), providing a link between AGE formation, oxidative stress, and resistance artery endothelial dysfunction in type 2 diabetic mice. In addition, we have previously shown that the effect of a chronic rise in blood flow on vascular function and remodeling in resistance arteries in type 2 diabetes is impaired and associated with excessive ROS production. However, ROS reduction with catalase and superoxide dismutase (present study) or with tempol (6) restored endothelium-mediated relaxation. On the other hand, a chronic treatment with tempol (6) did not fully restore flow-mediated remodeling or diameter enlargement (6) compared with ALT-711 (present study). As a result, besides superoxide production affecting the endothelium, another mechanism prevents diameter enlargement because antioxidant treatment did not restore flow-mediated remodeling despite a reduction in superoxide level (6).

A possible explanation for the altered HF-dependent remodeling in ZDF rats is that type 2 diabetes already has induced an outward hypertrophic remodeling (42–45) and that a further increase in diameter is no longer possible. Nevertheless, after chronic treatment with ALT-711, an increase in diameter was observed and found to be equivalent to that in LZ rats. The initial flow-sensing process was probably not affected in ZDF rats, because the response to the chronic rise in flow was “normal” with a rise in media cross-sectional area in HF arteries equivalent to that observed in LZ rats. Thus, the reduction in HF-induced diameter enlargement might be caused by the presence of AGEs, which have been shown to reduce the activity of several processes in type 2 diabetes. AGEs have been reported to alter the matrix proteins collagen, vitronectin, and laminin through AGE-AGE intermolecular covalent bonds or cross-linking (45,46). Furthermore, AGE cross-linking on type I collagen and elastin causes an increase in the extracellular matrix area, resulting in increased stiffness of the vasculature (46)_. The most probable explanation for this defect is that MMP activity is reduced in arteries from ZDF rats, as evidenced by our experiments (Fig. 7). Indeed, in situ gelatinase activity was reduced in ZDF rats and increased by ALT-711. This effect was most probably mediated by restoration of matrix degradation processes, which are reduced as a result of AGE accumulation. This is in agreement with a previous study showing that preventing AGE formation in streptozotocin-induced diabetic mice restored MMP activity and postocclusion revascularization (29).

We also found that endothelium-dependent dilation was not improved but instead was further reduced in arteries from ZDF rats submitted to a chronic rise in blood flow, in agreement with our previous work (6). Furthermore, HF arteries from ZDF rats exhibited a greater reduction in endothelium-dependent dilation than NF arteries from ZDF rats (Fig. 3A and C), showing that a further endothelial dysfunction occurred in response to the chronic increase in blood flow. The defect observed in ZDF rats was not a result of a change in smooth muscle response to NO, because the relaxation induced by the NO donor SNP was not affected. On the other hand, the involvement of NO in the dilation was severely reduced, as evidenced by the reduction of l-NAME’s effect on ACh-induced relaxation. In LZ rats, l-NAME strongly reduced ACh-induced relaxation, in agreement with previous studies showing the key role of NO in the mesenteric arteries (9,12). Of interest, in ZDF rats ALT-711 significantly improved NO-dependent relaxation in both NF and HF vessels (Fig. 3). This lower capacity of NO to induce relaxation also was evidenced by a reduced ability of l-NAME to induce contraction following precontraction (Fig. 3E and F). These findings agree with previous studies showing a deleterious effect of AGEs on endothelium-dependent dilation (17,47), although our study provides direct evidence that AGE breaking might reverse vascular damage, especially in response to a chronic increase in blood flow. The reduction in NO-dependent dilation is probably the result of a reduced NO bioavailability, as suggested by two experiments: 1) the ratio of phospho-eNOS to eNOS, reflecting eNOS activity, was not different in ZDF rats, LZ rats, and ZDF rats treated with ALT-711 (Supplementary Fig. 2) and 2) ROS reduction with superoxide dismutase and catalase restored ACh-dependent dilation in ZDF rats to the level found in LZ rats (Fig. 4).

Thus, AGEs not only block diameter enlargement in response to a chronic rise in blood flow, they also increase the endothelial dysfunction already present in ZDF rats. We can draw multiple conclusions from the current study. First, increasing flow, which is expected to increase arterial diameter and improve endothelium-dependent dilation, is not only inefficient in ZDF rats but it accelerates endothelial dysfunction. This finding might be of importance when recommending exercise or prescribing vasodilators to diabetic patients. Indeed, both aim to induce flow-mediated outward remodeling (38,48). However, breaking AGEs not only improved endothelial function, but our study suggests that it also might improve local blood flow supply and hence prevent damage to end organs. Thus, breaking AGEs and/or preventing their formation certainly is an important way of reducing ischemic disorders associated with diabetes.

Nevertheless, because ALT-711 also reduced early AGEs such as MG-¹H, we cannot exclude that ALT-711 also could reduce AGE formation, at least in part. Furthermore, some beneficial effects of ALT-711, such as improved endothelium-dependent dilation, might result, at least in part, from reduced RAGE stimulation. This is suggested by our results showing that ALT-711 reduced inflammation and oxidative stress in arteries of ZDF rats. Thus, a more in-depth analysis of the effect of AGEs, per se, and of those of RAGE would be necessary in the remodeling process.

In summary, AGE formation in type 2 diabetes altered the ability of resistance arteries to adapt their structure and function in response to a chronic increase in blood flow (summary of the main results of the study in Fig. 8). This impairment was reversed by an AGE-breaking treatment that reduced both AGE level and RAGE expression. Thus, breaking AGE cross-links provides a therapeutic potential for overcoming the microvascular complications of ischemic diseases in long-term diabetes.

This work was supported in part by the Foundation for Medical Research (Fondation pour la Recherche Médicale, FRM), Paris, France. M.L.F. was supported by a grant from Intervet/Schering-Plough, Angers, France.

No potential conflicts of interest relevant to this article were reported.

M.L.F. wrote the manuscript and researched data. K.T. and B.T. researched data. C.F. and L.L. contributed to the discussion and reviewed and edited the manuscript. D.H. designed the protocol, obtained grants, and wrote the manuscript. D.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

The authors thank Jérôme Roux, Pierre Legras, and Dominique Gilbert (Animal Care Facility, University of Angers) for their kind help in treating the rats.