RAGE Blockade Stabilizes Established Atherosclerosis in Diabetic Apolipoprotein E–Null Mice
Background— Previous studies suggested that blockade of RAGE in diabetic apolipoprotein (apo) E–null mice suppressed early acceleration of atherosclerosis. A critical test of the potential applicability of RAGE blockade to clinical settings was its ability to impact established vascular disease. In this study, we tested the hypothesis that RAGE contributed to lesion progression in established atherosclerosis in diabetic apoE-null mice.
Methods and Results— Male apoE-null mice, age 6 weeks, were rendered diabetic with streptozotocin or treated with citrate buffer. At age 14 weeks, certain mice were killed or treated with once-daily murine soluble RAGE or albumin; all mice were killed at age 20 weeks. Compared with diabetic mice at age 14 weeks, albumin-treated animals displayed increased atherosclerotic lesion area and complexity. In diabetic mice treated with sRAGE from age 14 to 20 weeks, lesion area and complexity were significantly reduced and not statistically different from those observed in diabetic mice at age 14 weeks. In parallel, decreased parameters of inflammation and mononuclear phagocyte and smooth muscle cell activation were observed.
Conclusions— RAGE contributes not only to accelerated lesion formation in diabetic apoE-null mice but also to lesion progression. Blockade of RAGE may be a novel strategy to stabilize atherosclerosis and vascular inflammation in established diabetes.
Received June 13, 2002; revision received August 23, 2002; accepted August 24, 2002.
Diabetes is associated with aggressive atherosclerosis in human subjects and represents a leading cause of morbidity and mortality.1 Chronic perturbation of diabetic vasculature leads to increased numbers, size, and complexity of atherosclerotic plaques. Furthermore, lesion instability is enhanced in diabetes and mediates increased incidence and severity of clinical events. In diabetes, especially in the setting of hyperglycemia mediated by insulin resistance, an emerging view is that events portending accelerated atherosclerosis are underway even before the formal diagnosis of diabetes.2 These considerations highlight the concept that strategies to intervene in established vascular disease are essential for the optimal management of diabetic macrovascular disease.
See p 2760
Previous studies suggest that ligands of RAGE (receptor for advanced glycation end products [AGEs]) are enriched in diabetic vasculature,3,4⇓ such as AGEs, particularly carboxy(methyl)lysine (CML) adducts of proteins and lipids. AGE deposition occurs in diabetic blood vessels driven by hyperglycemia, oxidant stress, and triggering of proinflammatory mechanisms.5,6⇓ The identification of S100/calgranulins, members of a family of proinflammatory cytokines, as ligands for RAGE underscores the premise that heightened activation of inflammatory mechanisms critically impact diabetic vasculopathy.7 Together with superimposed stresses such as hyperlipidemia, we hypothesize that increased expression of RAGE in a ligand-enriched environment exacerbates proinflammatory mechanisms, thereby accelerating atherosclerotic plaque formation in diabetes. Previously, we demonstrated that administration of the extracellular ligand-binding domain of the receptor, soluble (s) RAGE, to apolipoprotein (apo) E–null mice immediately on the diagnosis of hyperglycemia suppressed accelerated development of atherosclerotic plaques in a lipid- and glycemia-independent manner.3 A critical test of the role of RAGE in macrovascular disease and its potential utility as a target for diabetic vascular disease, however, is the extent to which RAGE contributed to lesion progression in established lesions. In this study, we tested the concept that interruption of ligand-RAGE interaction in established atherosclerosis would impact lesion progression in diabetic apoE-null mice. Our findings support the premise that activation of RAGE is involved not only in accelerating early lesion formation in diabetic apoE-null mice but also in sustaining proinflammatory or prothrombotic mechanisms leading to rapid lesion progression.
Animal studies were performed in accordance with the approval of the Institutional Animal Care and Use Committee of Columbia University. Male apoE-null mice (backcrossed >10 generations in the C57BL/6 background) were purchased from Jackson Laboratories (Bar Harbor, Maine) and maintained on a 12-hour light-dark cycle in a pathogen-free environment with free access to rodent chow and water. At age 6 weeks, mice were rendered diabetic by administration of 5 daily intraperitoneal injections of streptozotocin, 60 mg/kg in citrate buffer (0.05 mol/L; pH 4.5) (Sigma Aldrich). Control mice were treated with citrate buffer.3 Serum glucose was measured from tail vein blood using a glucometer. Beginning at age 14 weeks, certain diabetic and nondiabetic mice were killed. Other animals were randomized to treatment either with murine-soluble RAGE, 100 μg per day by intraperitoneal injection,3,7⇓ or vehicle, murine serum albumin (MSA). These mice were killed at age 20 weeks.
Quantification and Characterization of Atherosclerotic Lesions
On euthanasia, the heart was removed and stored in buffered formalin (10%). Cryostat sections were prepared and embedded in gelatin (25%). Serial sections, 10 μm thick, were cut from the level of the aortic valve leaflets up to ≈480 μm above the leaflets in the aortic sinus; every other section was retrieved and placed onto gelatin-coated slides (5%); 4 sections were placed onto each slide for a total of 5 slides. Sections were stained with Oil Red O and counterstained with hematoxylin/light green. Quantification of atherosclerotic lesion area was performed by a blinded observer, and lesion complexity was determined as described.3,8⇓
Western Blotting and Zymography
For Western blotting, aortic tissue was retrieved and subjected to disruption by a Brinkmann homogenizer in tris-buffered saline containing protease inhibitors (Boehringer Mannheim). For zymography, aortic tissue was subjected to homogenization in tris-buffered saline containing phenylmethylsulfonyl fluoride, 0.001 mol/L. Protein concentration was determined using an assay from Bio-Rad. Equal amounts of protein were subjected to SDS-PAGE (Novex/Invitrogen), and contents of the gels were transferred to nitrocellulose membranes (Novex/Invitrogen) and Western blotting performed as described.7 After completion, membranes were stripped of bound antibodies and reprobed with antibody to β-actin (Sigma). Bands were scanned into a densitometer and quantification was performed after normalization to β-actin. The value 1.0 was arbitrarily assigned to the mean pixel units per control condition. The following antibodies were used: goat anti-human vascular cellular adhesion molecule (VCAM)-1 IgG (0.4 μg/mL) (Santa Cruz Biotechnology, Santa Cruz, Calif); goat anti-murine JE-MCP-1 (0.2 μg/mL) (R&D Systems, Minneapolis, Minn); rabbit anti-murine RAGE (2.2 μg/mL) (0.4)3,7⇓; goat anti-rat tissue factor IgG (21.5 μg/mL) (generously supplied by Dr Vijay Rao, University of Texas Health Sciences Center at Tyler, Tex); rabbit anti-mouse cyclooxygenase 2 igG (2 μg/mL) (Cayman Chemical, Ann Arbor, Mich); rabbit anti-mouse matrix metalloproteinase (MMP) 9 IgG (2 μg/mL) (Chemicon International, Temecula, Calif); and monoclonal mouse anti-mouse β-actin IgG (0.25 μg/mL) (Chemicon). For zymography, aortic extracts were subjected to electrophoresis using gelatin-laden gels (Novex/Invitrogen).
Histology and Immunohistochemistry
Heart tissue was fixed in buffered-formalin (10%) for 1 week, followed by preparation of paraffin-embedded sections (5 μm thick). Certain sections were stained with Direct Red 80 mixed with picric acid (Sigma) to yield Picrosirius Red stain for detection of collagen. Other sections were prepared for immunohistochemistry using rabbit anti-murine RAGE IgG (14 μg/mL), affinity purified anti-CML IgG (0.37 μg/mL),5 rabbit anti-S100 IgG (40 μg/mL) (Sigma), mouse anti-human CD68 IgG (3.6 μg/mL) (DAKO A/S, Glostrup, Denmark), mouse anti-human smooth muscle actin IgG (3.8 μg/mL) (DAKO A/S), goat anti-rat tissue factor IgG (86 μg/mL), rabbit anti-mouse MMP-9 IgG (40 μg/mL) (Triple Point Biologics, Inc, Portland, Ore), and rabbit anti-nitrotyrosine IgG (10 μg/mL) (Upstate, Lake Placid, NY). Control immunostaining was performed using the respective nonimmune IgG; no specific staining was observed. For each mouse and condition, 3 slides from the aortic sinus were randomly chosen, and image analysis using a Zeiss microscope/Media Cybernetics software was used to determine the mean immunoreactive area for macrophages (using anti-CD68 IgG) and smooth muscle cells (using anti-smooth muscle actin IgG).
All data are reported as mean±SD. Where indicated, densitometric analysis of band intensity was performed using ImageQuant and Molecular Dynamics. Data were analyzed by ANOVA and, as indicated, subject to post hoc comparisons using two-tailed t test. Values considered significant were P<0.05.
Blockade of RAGE Stabilizes Atherosclerotic Lesion Area and Complexity in Diabetic ApoE-Null Mice
To test the hypothesis that RAGE contributes to amplification of pathways linked to progression of diabetic atherosclerosis, we studied apoE-null mice treated with soluble RAGE, 100 μg per day, or vehicle, MSA, from age 14 to 20 weeks. Gross examination of the aortae revealed that induction of diabetes was associated with increased atherosclerotic lesion area compared with nondiabetic controls at age 14 and 20 weeks (Figure 1b and 1a and Figure 1d and 1c, respectively). In diabetic mice subjected to blockade of RAGE, atherosclerotic lesion burden was reduced compared with mice treated with MSA (Figure 1e and 1d, respectively). Examination of representative Oil Red O–stained cross-sections of the aorta through the aortic sinus revealed that compared with nondiabetic mice, diabetic mice displayed larger atherosclerotic lesions at 14 weeks (Figure 1f and 1g, respectively) and 20 weeks (Figure 1h and 1i, respectively). In diabetic mice treated with sRAGE, atherosclerotic lesions were smaller than those observed in vehicle-treated diabetic mice at age 20 weeks (Figure 1j and 1i, respectively). At age 14 weeks, after ≈6 weeks of established diabetes, apoE-null mice displayed an ≈2.7-fold increase in mean lesion area compared with nondiabetic mice of the same age, 46 087±4592 versus 16 879±2216 μm2, respectively; P<0.00001 (Figure 1k). At 20 weeks, after ≈12 weeks of established diabetes, diabetic apoE-null mice displayed an ≈2.5-fold increase in mean lesion area compared with nondiabetic mice of the same age, 250 870±13 056 versus 99 758±10 303 μm2, respectively; P<0.00001 (Figure 1k). In diabetic mice treated with sRAGE at 20 weeks, there was an ≈4.2-fold decrease in mean atherosclerotic lesion area compared with MSA-treated diabetic animals, 59 071±6645 versus 250 870±13 056 μm2, respectively; P<0.00001 (Figure 1k). Although mean lesion area increased ≈5.4-fold between diabetic mice at 20 versus 14 weeks; P<0.0001, mean lesion area in sRAGE-treated diabetic mice at 20 weeks was not significantly different than that observed in diabetic mice at 14 weeks, 59 071±6645 versus 46 087±4592 μm2, respectively; P>0.05 (Figure 1k). Furthermore, at age 20 weeks, mean lesion area in sRAGE-treated diabetic mice was reduced ≈1.7-fold compared with lesions in nondiabetic animals of the same age; P<0.05.
Epidemiologic studies in humans and studies in apoE-null mice suggest that a key feature of diabetes is the hastening of lesion complexity.3 Thus, we assessed the effects of RAGE blockade in diabetic apoE-null mice with established atherosclerosis on lesion complexity, defined by the presence of fibrous caps, cholesterol clefts, or necrosis. At age 20 weeks, the mean number of complex lesions per section was increased ≈2.3-fold in vehicle-treated diabetic mice versus nondiabetic animals of the same age; P<0.00001 (Figure 1l). In diabetic mice treated with sRAGE from the age of 14 to 20 weeks, the mean number of complex lesions was reduced ≈3.7-fold compared with diabetic animals treated with MSA; P<0.0001 (Figure 1l). Importantly, lesion complexity was not significantly different in sRAGE-treated diabetic mice at 20 weeks compared with diabetic mice at age 14 weeks; P>0.05 (Figure 1l). Analogous to observations regarding mean atherosclerotic lesion area, a trend toward decreased lesion complexity (≈1.6-fold) was observed in sRAGE-treated diabetic mice at age 20 weeks compared with nondiabetic animals of the same age, although the results did not achieve statistical significance; P=0.099 (Figure 1l).
The ability of RAGE blockade to reduce atherosclerotic lesion area in diabetic mice beyond that seen in nondiabetic animals suggested that RAGE-associated mechanisms accounted, at least in part, for atherosclerosis in the absence of diabetes. To study this, nondiabetic mice at age 14 weeks received either MSA or sRAGE from age 14 to 20 weeks. At 20 weeks, MSA-treated mice displayed an ≈5.9-fold increase in mean atherosclerotic lesion area compared with nondiabetic mice at age 14 weeks; P<0.00001 (Figure 1m). In sRAGE-treated nondiabetic mice at age 20 weeks, mean lesion area, ≈62 771±3392 μm2, was significantly lower than that observed in MSA-treated nondiabetic mice of the same age (99 758±10 303 μm2); P=0.03 (Figure 1m). Mean lesion area in sRAGE-treated nondiabetic mice at 20 weeks was, however, significantly higher than that observed in nondiabetic mice at age 14 weeks; P<0.00001 (Figure 1m). In addition, we examined the impact of RAGE blockade on lesion complexity. At age 20 weeks, the mean number of complex lesions per section was increased ≈16.2-fold in nondiabetic mice versus animals at age 14 weeks; P<0.01 (Figure 1n). However, in nondiabetic mice treated with sRAGE from the age of 14 to 20 weeks, a trend toward decreased mean number of complex lesions was noted (≈2.4-fold lower) compared with nondiabetic animals treated with MSA (Figure 1n). Analogous to observations regarding mean atherosclerotic lesion area, the differences between mean number of complex lesions in nondiabetic mice at age 14 weeks versus sRAGE-treated nondiabetic mice remained statistically significant, suggesting that factors complementary to RAGE activation importantly contributed to progression of atherosclerosis in the absence of diabetes (Figure 1n).
Blockade of RAGE Stabilizes Vascular Inflammation in Diabetic ApoE-Null Mice
To dissect the molecular mechanisms underlying RAGE-dependent acceleration of diabetes-associated atherosclerosis, we focused on inflammatory and prothrombotic mediators in diabetic aortae retrieved from mice at age 14 or 20 weeks, in the presence or absence of RAGE blockade. We retrieved the aorta from a point just distal to the aortic sinus to the bifurcation of the iliac arteries. First, we assessed levels of RAGE antigen. Compared with diabetic mice at 14 weeks, aortae from diabetic animals at age 20 weeks revealed an ≈2.1-fold increase in RAGE expression by Western blotting; P<0.05 (Figure 2a through 2c). In the presence of RAGE blockade, at age 20 weeks, levels of RAGE in the aorta were ≈1.9-fold lower in diabetic animals treated with sRAGE versus MSA; P<0.05 (Figure 2a, 2c, and 2d). Levels of RAGE antigen in aortae retrieved from sRAGE-treated diabetic mice at age 20 weeks were not significantly different from levels observed in diabetic mice at age 14 weeks; P>0.05 (Figure 2a, 2b, and 2d). In addition, enhanced expression of signal-transducing ligands of RAGE, S100/calgranulins, and CML-AGEs overlapped with that of RAGE in the atherosclerotic plaques of diabetic apoE-null mice (Figure 2e through 2g and Figure 2h through 2j, respectively).
We next examined the impact of diabetes on lesion characteristics, specifically with respect to numbers of mononuclear phagocytes (MPs) and smooth muscle cells (SMCs), and how migration or proliferation of these cell types modulated diabetes and RAGE blockade. Quantitative immunohistochemistry using an image analysis program was performed to determine the mean area/lesion occupied by either MP or SMC, because these are the principal cells comprising atherosclerotic plaques in the apoE-null mouse. At age 20 weeks, an ≈2.6-fold decrease in mean area per lesion occupied by CD68-expressing MP was observed in diabetic apoE-null mice treated with sRAGE compared with diabetic animals treated with MSA; P<0.05 (Figure 3c and 3b, respectively, and Figure 3d). In addition, mean MP area and lesion in sRAGE-treated mice at 20 weeks was not significantly different than that observed in diabetic mice at age 14 weeks; P>0.05 (Figure 3c and 3a, respectively, and Figure 3d).
The mean area occupied by SMC within atherosclerotic plaques was also enhanced in diabetes. At 20 weeks, an ≈4.5-fold decrease in mean area per lesion occupied by SMC was observed in diabetic apoE-null mice treated with sRAGE compared with diabetic mice treated with MSA; P<0.01 (Figure 4c and 4b, respectively, and Figure 4d). Blockade of RAGE instituted at age 14 weeks effectively halted additional migration or proliferation of SMC into atherosclerotic lesions, because mean area occupied by SMC in diabetic mice treated with sRAGE at 20 weeks did not differ from that observed in diabetic mice at age 14 weeks; P>0.05 (Figure 4c and 4a, respectively, and Figure 4d). These data suggest that diabetes was associated with increased numbers of MP and SMC in atherosclerotic plaques in apoE-null mice; however, in the presence of RAGE blockade, additional influx or proliferation of these key cell types critical to progression of atherosclerotic lesions was halted.
To elucidate the mechanisms by which activation of RAGE mediated progression of atherosclerosis in diabetic vasculature, we assessed indices of vascular inflammation and expression of prothrombotic mediators. Because diabetes was associated with a striking increase in numbers of MP infiltrating atherosclerotic plaques, we assessed levels of a key chemokine, JE-MCP-1,9,10⇓ and adhesion molecule vascular cell adhesion molecule-1 (VCAM-1)11 in contributing to this phenomenon. By Western blotting, compared with diabetic mice at age 14 weeks, diabetic animals at age 20 weeks displayed an ≈2.3-fold increase in vascular expression of JE-MCP-1; P<0.05 (Figure 5a). An ≈3-fold decrease in JE-MCP-1 antigen was observed in sRAGE-treated diabetic apoE-null mice versus MSA-treated diabetic animals at age 20 weeks; P<0.05 (Figure 5a). Levels of JE-MCP-1 in the aortae of sRAGE-treated diabetic mice at age 20 weeks were not significantly different from those observed in diabetic animals at age 14 weeks; P>0.05 (Figure 5a). In addition to RAGE-driven enhanced expression of chemokines linked to atherosclerosis, the aortae of diabetic animals at age 20 weeks displayed an ≈2.7-fold increase in VCAM-1 antigen compared with animals at age 14 weeks; P<0.05 (Figure 5b). In diabetic mice treated with sRAGE at age 20 weeks, levels of VCAM-1 were reduced by ≈2.2-fold compared with vehicle-treated diabetic mice; P<0.05 (Figure 5b). Vascular expression of VCAM-1 was not significantly different in sRAGE-treated diabetic mice at age 20 weeks versus diabetic animals at age 14 weeks; P>0.05 (Figure 5b).
We next studied vascular levels of cyclooxygenase-2 (cox-2), because this molecule is associated with proinflammatory mechanisms and oxidant stress in atherosclerosis.12 Levels of cox-2 antigen in the aorta were enhanced ≈3.8-fold in diabetic mice at age 20 weeks compared with 14 weeks; P<0.01 (Figure 5c). Activation of RAGE in diabetic aorta importantly contributed to this phenomenon; compared with MSA-treated diabetic mice at age 20 weeks, levels of cox-2 antigen in the aortae of mice treated with sRAGE were decreased ≈2.6-fold; P<0.05 (Figure 5c). Levels of cox-2 in sRAGE-treated diabetic aortae at 20 weeks were not significantly different from those observed in diabetic aortae of apoE-null mice, age 14 weeks; P>0.05 (Figure 5c). In parallel with increased duration of diabetes, the larger and more complex lesions of MSA-treated diabetic mice displayed increased epitopes for 3-nitrotyrosine, a marker of oxidative stress,13 compared with either diabetic lesions at age 14 weeks or lesions observed in sRAGE-treated diabetic mice at age 20 weeks (Figure 5d through 5f).
In addition to increased expression of proinflammatory mediators in diabetic vasculature, enhanced expression and activity of MMPs has been linked to multiple mechanisms germane to the progression of atherosclerosis lesions, such as SMC migration and lesion instability.14–16⇓⇓ By Western blotting and zymography, levels of MMP-9 antigen and activity in the aorta were increased ≈2.1-fold and ≈5.1-fold in the aortae of diabetic mice at age 20 weeks compared with 14 weeks (Figure 5g through 5k); P<0.05. In the presence of sRAGE, at 20 weeks, levels of MMP-9 antigen/activity were reduced ≈1.9- and ≈3-fold, respectively, in aortae compared with vehicle-treated diabetic animals; P<0.05 (Figure 5g through 5k). Blockade of RAGE effectively suppressed the exaggerated effects of diabetes on enhancing MMP-9 expression and activity, as these levels were not significantly different in sRAGE-treated diabetic mice at age 20 weeks versus 14 weeks (Figure 5g, 5h, 5j, and 5k); P>0.05.
These observations suggested that RAGE-driven proinflammatory mechanisms accounted, in large part, for the sustained increases in vascular expression of JE-MCP-1,VCAM-1, cox-2, nitrotyrosine epitopes, and MMP-9 antigen/activity. To identify likely intracellular signaling pathways triggered by ligand activation of RAGE in diabetes, we studied p38 mitogen-activated protein (MAP) kinase, because its activation has been linked to regulation of proinflammatory genes.17 On normalization to the level of total p38 MAP kinase, an ≈9.6-fold increase in phospho-p38 MAP kinase was noted in diabetic aortae at 20 versus 14 weeks; P<0.05 (Figure 5l). In the presence of RAGE blockade at 20 weeks, a significant decrease in levels of phospho-p38 MAP kinase was observed in diabetic aortae compared with MSA; P<0.05 (Figure 5l). Levels of phospho-p38 MAP kinase were not significantly different between sRAGE-treated diabetic mice at age 20 weeks versus diabetic animals at age 14 weeks; P>0.05 (Figure 5l).
In addition to the striking RAGE-driven enhancement of proinflammatory mediators in diabetic vasculature, we tested the role of the ligand-RAGE axis in expression of prothrombotic tissue factor (TF).18 Compared with diabetic mice at age 14 weeks, the aortae of diabetic animals at age 20 weeks revealed an ≈3.5-fold increase in expression of TF antigen; P<0.05 (Figure 5m through 5o). Administration of sRAGE to diabetic mice resulted in an ≈4.8-fold reduction in levels of tissue factor compared with MSA-treated animals at age 20 weeks; P<0.05 (Figure 5m, 5o, and 5p). RAGE blockade rendered vascular expression of TF in sRAGE-treated diabetic mice not significantly different from that observed in diabetic vasculature at age 14 weeks; P>0.05 (Figure 5m, 5n, and 5p).
Blockade of RAGE Stabilizes Collagen Generation in Diabetic ApoE-Null Mice
Lastly, multiple studies have suggested that extracellular matrix molecules such as collagen are produced in atherosclerotic plaques, largely by activated SMCs that migrate into and proliferate within the lesions.19–21⇓⇓ By Picrosirius Red staining and polarized light microscopy, diabetes was associated with a significant increase, ≈1.8-fold, in extent of collagen per lesion in diabetic mice versus nondiabetic controls at 20 weeks of age; P<0.05 (Figure 6d and 6c, respectively, and Figure 6f). In the presence of RAGE blockade, an ≈2.1-fold decrease in collagen content per lesion was noted compared with that seen in diabetic mice treated with MSA; P<0.05 (Figure 6e and 6d, respectively, and Figure 6f). The extent of collagen deposition in diabetic sRAGE-treated mice was not significantly different from that observed in nondiabetic animals of the same age; P>0.05 (Figure 6e and 6c, respectively, and Figure 6f). Furthermore, collagen deposition in sRAGE-treated diabetic mice at age 20 weeks was not different from that observed in diabetic mice at age 14 weeks; P>0.05 (Figure 6e and 6b, respectively).
Mean levels of blood glucose did not differ between diabetic mice treated with MSA versus sRAGE, 378±16 and 398±22 mg/dL, respectively; P>0.05. Levels of total plasma cholesterol were not significantly different between diabetic mice treated with sRAGE versus MSA, 702±153 and 714±155 mg/dL, respectively; P>0.05. In addition, separation of plasma cholesterol components by FPLC failed to reveal differences (data not shown).
The dramatic regression of atherosclerosis in euglycemic apoE-null nude mice subsequent to infection with adenovirus encoding human apoE cDNA suggested that the formation or progression of atherosclerotic lesions is a dynamic process.22–26⇓⇓⇓⇓ Recent evidence suggests that the beneficial effects of statins in modulation of atherosclerosis extend beyond their capacity to lower serum lipid levels and, in addition, include their potent effects on diminishing vascular inflammation and oxidant stress in animals and human subjects.20,27–29⇓⇓⇓ In diabetes, evidence is mounting that an exaggerated inflammatory response within the blood vessel wall contributes to acceleration of vascular lesion area and complexity. Previously, we addressed the impact of RAGE blockade in newly diabetic mice and demonstrated that RAGE blockade impacted importantly on accelerated lesion initiation in these animals.3 In this study, we show that activation of RAGE-dependent mechanisms is linked to accelerated lesion progression in apoE-null mice. Indeed, in certain settings, it is possible that individual molecular mediators may impact distinct phases of atherosclerosis. In this context, experiments using blockade of cox-2 using pharmacologic inhibitors or mice genetically deficient in cox-2 suggest that this inflammatory mediator is more likely to contribute to lesion initiation, and not progression, because in advanced lesions, expression of cox-2 was noted to decline.12
In contrast, in diabetic apoE-null mice, accumulation of RAGE ligands, AGEs and S100/calgranulins, continues to increase in atherosclerotic plaques. Support for the concept that enhanced RAGE-driven inflammatory responses within the diabetic blood vessel account for accelerated progression of diabetic atherosclerosis is the observation that multiple mediators of inflammation and plaque instability, such as inflammatory cell chemoattractants (JE-MCP-1 and VCAM-1), cox-2, tissue factor, and MMP-9 (antigen and activity), are additionally enhanced in diabetic vasculature at age 20 versus 14 weeks. In each case, blockade of RAGE, commencing at age 14 weeks, effectively halted progression of vascular inflammation. Consistent with the concept that RAGE-driven vascular inflammation is especially heightened in diabetic vasculature was the observation that although RAGE blockade significantly reduced mean atherosclerotic lesion area in nondiabetic mice at age 20 weeks versus that observed in sRAGE-treated mice at the same age, a significant increase in both lesion area and complexity was observed in sRAGE-treated nondiabetic mice at age 20 weeks versus nondiabetic mice at age 14 weeks. These observations strongly suggest that factors beyond RAGE activation importantly contribute to progression of atherosclerosis in the absence of diabetes, such as vascular disease triggered by hyperlipidemia.
Limitations to this model should be noted. Induction of diabetes in apoE-null mice results in sustained increases in levels of blood glucose. To assess the impact of hyperglycemia within a reasonable period of time in the mouse, we have not treated the animals with agents to reduce hyperglycemia. Thus, in contrast to the setting in human subjects, where treatment with insulin or insulin-sensitizing agents or typical postprandial fluctuations leads to intermittent hyperglycemia, here we have studied largely sustained elevation of blood glucose. In both settings, however, processes of advanced glycation will occur. We speculate that formation of these signal transduction ligands of RAGE is a key factor accelerating vascular inflammation. Furthermore, a limitation of the present studies is that diabetic animals were examined at one time point after intervention. It is conceivable that at later times, frank lesion regression might have been observed. Multiple studies in both animal models and human clinical trials suggest that lesion size and the degree of impingement on lumen patency do not necessarily predict lesion vulnerability or the likelihood of rupture. As the studies of Libby and colleagues have shown, lesion stability, at least in the face of lipid-lowering strategies, most likely ensues from suppression of exaggerated proinflammatory mechanisms, as well as reduction in levels of prothrombotic tissue factor expressed within the vascular plaques.23,30⇓ Thus, despite lack of regression by RAGE blockade in established diabetic atherosclerosis in apoE-null mice, we speculate that this intervention, nevertheless, modulates key features linked to accelerated cardiovascular events. In this context, because RAGE blockade was not associated with alteration in lipid number or profile or levels of blood glucose, we propose that concurrent administration of agents to block RAGE, together with optimal management of these complementary risk factors, has the potential to profoundly impact the course of atherosclerosis in diabetes.
Lastly, it is important to note that in euglycemic animals and human subjects, ruptured human atherosclerotic lesions frequently displayed thin fibrous caps with few SMCs and, in parallel, decreased collagen,31 thereby suggesting that increased numbers of vascular SMC and enhanced deposition of extracellular matrix molecules, in particular, collagen, enhanced lesion stability.18,19⇓ Here, we found that at age 20 weeks, atherosclerotic lesions in diabetic apoE-null mice were enriched both in SMC and collagen compared with nondiabetic controls and compared with lesions in diabetic mice at age 14 weeks. In diabetic mice treated with sRAGE at 20 weeks, collagen content was reduced. Based on birefringence patterns of Picrosirius Red–stained sections, diabetic lesions displayed markedly compact collagen, as suggested by its red appearance on polarization microscopy. In contrast, nondiabetic and sRAGE-treated diabetic lesions displayed evidence of less dense collagen generation or deposition, as indicated by the green to yellow color.32 Thus, we propose that in diabetes, this apparent paradox in content and density of collagen within atherosclerotic plaques may be considered as follows. Even in the face of increased SMC and collagen in diabetic atherosclerotic plaques, the significant increase in numbers of MP,33 along with the striking enhancement of proinflammatory mechanisms, particularly increased MMP expression/activity and generation of tissue factor, is likely to tip a delicate balance between structural integrity and lesion stability.
Taken together, our findings delineate a central role for RAGE in sustained MP and SMC activation in the diabetic atherosclerotic plaque and extend the potential application of RAGE blockade in diabetic atherosclerosis. In addition, for the first time, these findings demonstrate that RAGE blockade impacts atherosclerosis progression in euglycemia. In established vascular disease, interruption of ligand-RAGE interaction is likely to impact favorably on lesion progression and, potentially, clinical outcome.
This study was supported by Surgical Research Fund of the College of Physicians & Surgeons, Columbia University, and grants from the United States Public Health Service to Drs Stern and Schmidt. Dr Bucciarelli is a postdoctoral research fellows of the Juvenile Diabetes Research Foundation. Dr Schmidt is a recipient of a Burroughs Wellcome Fund Clinical Scientist Award in Translational Research.
Dr Schmidt receives research aid and is a paid consultant of TransTech Pharma, Inc.
- ↵King H, Aubert R, Herman W. Global burden of diabetes, 1995–2005: prevalence, numerical estimates and projections. Diabetes Care. 1998; 21: 1414–1431.
- ↵Haffner SM, Mykkänen L, Festa A, et al. Insulin resistant prediabetic subjects have more atherogenic risk factors than insulin sensitive prediabetic subjects: implications for preventing coronary heart disease during the prediabetic state. Circulation. 2000; 101: 975–980.
- ↵Kislinger T, Tanji N, Wendt T, et al. RAGE mediates inflammation and enhanced expression of tissue factor in the vasculature of diabetic apolipoprotein E null mice. Arterioscler Thromb Vasc Biol. 2001; 21: 905–910.
- ↵Kislinger T, Fu C, Huber B, et al. Nε (carboxymethyl)lysine modifications of proteins are ligands for RAGE that activate cell signalling pathways and modulate gene expression. J Biol Chem. 1999; 274: 31740–31749.
- ↵Burleigh ME, Babaev VR, Oates JA, et al. Cyclooxygenase-2 promotes early atherosclerotic lesion formation in LDL receptor-deficient mice. Circulation. 2002; 105: 1816–1823.
- ↵Aikawa M, Rabkin E, Okada Y, et al. Lipid lowering by diet reduces matrix metalloproteinase activity and increases collagen content of rabbit atheroma: a potential mechanism of lesion stabilization. Circulation. 1998; 23: 2433–2444.
- ↵Mach F, Schonbeck U, Bonnefoy JY, et al. Activation of monocyte/macrophage functions related to acute atheroma complication by ligation of CD40: induction of collagenase, stromelysin, and tissue factor. Circulation. 1997; 15: 396–399.
- ↵Bendeck MP, Irvin C, Reidy MA. Inhibition of matrix metalloproteinase activity inhibits smooth muscle cell migration but not neointimal thickening after arterial injury. Circ Res. 1996; 78: 38–43.
- ↵Aikawa M, Rabkin E, Okada Y, et al. Lipid lowering by diet reduces matrix metalloproteinase activity and increases collagen content of rabbit atheroma: a potential mechanism of lesion stabilization. Circulation. 1998; 97: 2433–2444.
- ↵Fukumoto Y, Libby P, Rabkin E, et al. Statins alter smooth muscle cell accumulation and collagen content in established atheroma of Watanabe heritable hyperlipidemic rabbits. Circulation. 2001; 103: 993–999.
- ↵Desurmont C, Caillaud JM, Emmanuel F, et al. Complete atherosclerosis regression after human ApoE gene transfer in ApoE-deficient/nude mice. Arterioscler Thromb Vasc Biol. 2000; 20: 435–442.
- ↵Hathaway CA, Heistad DD, Piergors DJ, et al. Regression of atherosclerosis in monkeys reduces vascular superoxide levels. Circ Res. 2002; 90: 277–283.
- ↵Corti R, Fayad ZA, Fuster V, et al. Effects of lipid-lowering by simvastatin on human atherosclerotic lesions. Circulation. 2001; 104: 249–252.
- ↵Aikawa M, Rabkin E, Okada Y, et al. Cerivastatin, an HMG-CoA reductase inhibitor, suppresses growth of macrophage expressing matrix metalloproteinases and tissue factor in vivo and in vitro. Circulation. 2001; 103: 276–283.
- ↵Brown BG, Zhao XQ, Sacco DE, et al. Lipid lowering and plaque regression: new insights into prevention of plaque disruption and clinical events in coronary disease. Circulation. 1993; 87: 1781–1791.
- ↵Davies MJ, Richardson PD, Woolf N, et al. Risk of thrombosis in human atherosclerotic plaques: role of extracellular lipid, macrophage, and smooth muscle cell content. Br Heart J. 1993; 69: 377–381.