Introduction

Atherosclerosis is a major complication of diabetes, representing the predominant cause of morbidity and mortality in both type 1 and type 2 diabetic patients [1]. The renin–angiotensin system (RAS), and in particular its effector molecule, angiotensin II (AII), participates in the development of atherosclerosis, through the regulation of at least two key processes, inflammation and fibrosis [2]. As a result, pharmacological inhibition of the RAS has been proposed as a strategy for reducing atherosclerosis, beyond BP reduction. Pharmacological agents include ACE inhibitors (ACEi), which block the conversion of the pro-hormone angiotensin I to the active hormone AII, and selective AII subtype 1 receptor blockers (ARBs), which competitively inhibit the action of AII at the AT1 receptor subtype (AT1R) [3–6]. Both classes of compounds are widely used for reducing BP in patients with type 1 and type 2 diabetes. Furthermore, it has been suggested that there is a potentially important BP-independent protective effect on the risk of CHD for ACEi that has not been demonstrated for ARBs [7], although this possibility remains controversial [8].

AII binds with similar affinity to two major receptor subtypes, AT1R and the type 2 (AT2R) receptor [9]. While most of the well-known effects of AII are mediated via AT1R, there is growing interest in effects mediated via AT2R. The AT2R is ubiquitously produced in fetal tissues, but its production declines after birth [10]. In adults, AT2R production is detectable in the pancreas, heart, adrenals, brain, kidney and vasculature [9, 11, 12]. AT2R is overproduced in pathological situations involving tissue remodelling or inflammation, including kidney damage [13–15] and atherosclerosis [16, 17]. However, the role of AT2R in diabetes-associated atherosclerosis is not completely understood. It has been postulated that AT1R and AT2R have opposing actions on proliferation and apoptosis. The proliferative properties of AII have generally been considered to be associated with AT1R, whereas AT2R is viewed as promoting apoptosis and inhibiting cell growth [18]. However, the effects of AT2R activation are not always uniform and reproducible. Indeed, there is currently no consensus on the role of AT2R in atherosclerosis, with many studies, albeit in the non-diabetic setting, suggesting either a neutral or anti-atherosclerotic effect of this AII receptor subtype [17, 19–23]. In addition, recent data suggest that AT2R may promote cell growth and inflammation [24–26].

The present study examined the role of AT2R in the context of long-term diabetes in a well-characterised and widely used model of diabetic atherosclerosis, the streptozotocin-induced diabetic apolipoprotein E knockout (Apoe-KO) mouse [27–30]. The validity of this model has recently been confirmed as appropriate for the study of diabetes-associated atherosclerosis by the National Institutes of Health (NIH)/Juvenile Diabetes Research Foundation (JDRF)-supported Animal Models of Diabetic Complications Consortium [29]. Our aim was to determine whether disruption of the AT2R would protect these animals from developing vascular lesions in the presence of diabetes. Thus, two different approaches were employed: first, a pharmacological strategy using an AT2R antagonist, PD123319, and second, a genetic strategy involving the use of At 2 r-deficient rodents.

Methods

Experimental model

Six-week-old male Apoe-KO mice (backcrossed 20 times onto a C57BL6/J background; Animal Resource Centre, Canning Vale, WA, Australia) and At 2 r (also known as Agtr2) double-KO (DKO) mice, confirmed by PCR genotyping and generated by backcrossing At 2 r-KO mice [31] on the C57BL6/J background (had ≥98% C57BL6/J background, genome scan by Jackson Laboratory; www.jax.org) into Apoe-KO mice for ten generations were housed at the Precinct Animal Centre (Baker IDI Heart and Diabetes Institute, Melbourne, VIC, Australia). The protocols followed for animal handling and experimentation were in accordance with ethical guidelines of the Alfred Medical Research and Education Precinct animal ethics committee and the National Health and Medical Research Council of Australia guidelines.

Mice were randomised to have diabetes induced via five daily i.p. injections of streptozotocin at 55 mg kg−1 day−1 (Boehringer, Mannheim, Germany) [2, 28] or sham injected with vehicle (citrate buffer, pH 4.5). On the sixth day, blood glucose was tested using Accu-Chek Advantage II test strips (Roche Diagnostics, Mannheim, Germany), and only mice with blood glucose >15 mmol/l were included as diabetic in the study (>90% of injected mice).

Control and diabetic Apoe-KO mice were then randomised to treatment with PD123319 (5 mg kg−1 day−1; Sigma-Aldrich, St Louis, MO, USA) [13, 32] or to no treatment for 20 weeks. The PD123319 drug and the vehicle (sterile Milli-Q water) were administrated s.c. via osmotic minipumps (Model 2004; Alzet, Cupertino, CA, USA), and these minipumps were replaced every 28 days. Animals were allowed access to standard mouse chow and water ad libitum. After 20 weeks, excised aortas (n = 15–20 per group) were placed in 10% neutral buffered formalin (10%, vol./vol.) and quantified for lesion area (n = 7–8 per group) before being embedded in paraffin for immunohistochemical analysis. In the remaining mice (n = 8–12 per group) aortas were snap-frozen in liquid nitrogen and stored at −80°C for subsequent RNA extraction [2, 28].

Metabolic variables and BP

After 20 weeks, blood was collected from the left ventricle for the measurement of glycated haemoglobin (HbA1c) [33], fasting glucose, total cholesterol and triacylglycerol [34].

Systolic BP was measured at week 19 of the study period by a computerised, non-invasive tail cuff system in conscious prewarmed mice [35].

Plaque area quantification

Aortas removed from mice were cleaned of excess fat under a dissecting microscope and subsequently stained with Sudan IV-Herxheimer’s solution (0.5% wt/vol.) (Gurr; BDH, Poole, UK) as described previously [2, 28]. The aortas were divided into arch, thoracic and abdominal sections and then cut longitudinally. After pinning en face onto wax, images were acquired with a dissecting microscope (Olympus SZX10, Olympus Optical, Tokyo, Japan) with a videocamera (Q-capture Pro Version 5.1, Burnaby, BC, Canada). Total plaque area was quantified as a percentage area of the whole aorta stained by Sudan IV (Adobe Photoshop version 6.0). Aortas were subsequently embedded in paraffin and sections cut for cross-sectional analysis.

Real-time RT-PCR

Total RNA was extracted from the whole aorta by homogenising and treated with DNAse as described previously [2, 28]. cDNA was synthesised by reverse transcription (Pierce, Rockford, IL, USA). Quantitative real-time RT-PCR was performed using the Taqman System on an ABI Prism 7500 Sequence Detector (Applied Biosystems, Foster City, CA, USA) and analysed using Sequence Detection Software (SDS version 1.9; Applied Biosystems, Foster City, CA, USA). Gene expression was normalised to 18S rRNA (Applied Biosystems). Detailed information on probes is provided in the Electronic supplementary material (ESM) Table 1.

Immunostaining

Serial 4 µm paraffin aortic cross-sections were stained with haematoxylin and eosin to evaluate the atherosclerotic lesion complexity, or with Picrosirius-Red with polarisation microscopy to detect collagen content [36, 37]. Serial 4 µm paraffin aortic cross-sections were immunostained for alpha-smooth muscle actin (α-SMA) (1:500; Serotec, Oxford, UK), vascular cell adhesion molecule-1 (VCAM-1) (1:50; Pharmingen, San Diego, CA, USA), monocyte chemotactic protein-1 (MCP-1) (1:500; R&D Systems, Minneapolis, MN, USA) and the macrophage marker F4/80 (1:50; Serotec) as previously described [34]. Results were quantified as per cent of positively stained tissue. Image analysis was performed using ImageJ (http://rsb.info.nih.gov/ij, accessed 3 March 2008) and ImagePro 6.0 software (Media Cybernetics, Inc., Bethesda, MD, USA).

Statistical analysis

Data were analysed by ANOVA using Statview (version 5.0) and post hoc analysis of group means was performed by Fisher’s least significant difference method. Data are expressed as means±SEM unless otherwise specified. p < 0.05 was considered to be statistically significant.

Results

Metabolic variables and BP

The induction of diabetes in both Apoe-KO and At 2 r/Apoe-DKO mice resulted in increased plasma glucose and HbA1c concentrations, which were comparable between the diabetic groups. Plasma total cholesterol concentrations were also significantly increased in both groups following the induction of diabetes. BP was not altered after 20 weeks of diabetes in either Apoe-KO or At 2 r/Apoe-DKO groups compared with the non-diabetic groups. PD123319 treatment in the diabetic Apoe-KO mice did not alter any of these variables (Table 1).

Table 1 Characteristics of mice at the conclusion of the 20 week study
Full size table

Atherosclerotic plaque area

The induction of diabetes in Apoe-KO mice led to an approximately sixfold increase in atherosclerotic plaque area of the whole aorta compared with non-diabetic Apoe-KO mice (Fig. 1). In these diabetic animals lesions were predominantly complex fibrous plaques (Fig. 2). AT2R blocker (AT2RB) treatment significantly attenuated the extent and complexity of plaques in diabetic Apoe-KO mice. Plaque area was also significantly reduced in diabetic At 2 r/Apoe-DKO mice (Fig. 1). By contrast, no effect was detected in control Apoe-KO mice treated with AT2RB, nor did control At 2 r/Apoe-DKO mice show any difference in plaque area compared with control Apoe-KO mice alone.

Fig. 1
figure 1

a–f Representative en face photomicrographs of aortas with each box divided into aortic arch (left), thoracic (centre) and abdominal aorta (right) segments. g Total aortic plaque area expressed as the percentage of aortic area staining for Sudan IV Red (black, arch; white, thoracic; horizontal lines, abdominal). C, control Apoe-KO; C+PD, AT2RB-treated control Apoe-KO; C DKO, control At 2 r/Apoe-DKO; D, diabetic Apoe-KO; D+Veh, vehicle-treated diabetic Apoe-KO; D+PD, AT2RB-treated diabetic Apoe-KO; D DKO, diabetic At 2 r/Apoe-DKO. Means ± SEM. *p < 0.05 vs control Apoe-KO; † p < 0.05 vs diabetic Apoe-KO (n = 7–8 per group)

Full size image
Fig. 2
figure 2

Representative thoracic aortic sections stained with Picrosirius Red with polarisation microscopy to evaluate the presence of collagen I and III in control Apoe-KO (C), AT2RB-treated control Apoe-KO (C+PD), control At 2 r/Apoe-DKO (C DKO), diabetic Apoe-KO (D), AT2RB-treated diabetic Apoe-KO (D+PD) and diabetic At 2 r/Apoe-DKO (D DKO). Magnification ×200

Full size image

At2r expression

In diabetic Apoe-KO mice, aortic mRNA levels of the At 2 r gene were increased compared with control Apoe-KO mice and this increase in gene expression was decreased by AT2RB treatment (Fig. 3).

Fig. 3
figure 3

At 2 r gene expression quantified by real-time RT-PCR at the conclusion of the 20 week study in the whole aorta from the six experimental groups. C, control Apoe-KO; C+PD, AT2RB-treated control Apoe-KO; C DKO, control At 2 r/Apoe-DKO; D, diabetic Apoe-KO; D+PD, AT2RB-treated diabetic Apoe-KO; D DKO, diabetic At 2 r/Apoe-DKO. Means±SEM. *p < 0.05 vs control Apoe-KO; † p < 0.05 vs diabetic Apoe-KO. a.u., arbitrary units; ND, not detected

Full size image

Inflammation

Inflammation plays a central role in the development and progression of atherosclerotic lesions. Thus, we assessed the regulation of nuclear factor-kappa B (NF-κB), a key mediator of inflammation, by measuring aortic expression of the gene encoding NF-κB subunit p65. Diabetes induced a greater than threefold increase in the expression in Apoe-KO aorta. This upregulation was significantly attenuated by both AT2RB treatment and At 2 r gene deletion (Table 2). MCP-1, an NF-κB-dependent chemokine, is considered a key mediator in AII-induced progression of atherosclerosis [38]. Indeed, in this study, diabetes was associated with a significant increase in levels of this chemokine. AT2RB treatment and At 2 r gene deletion significantly decreased the aortic production of MCP-1 (Table 2, Figs 4 and 5) as well as suppressing the induction of pro-inflammatory cytokines such as TNF-α (Table 2) and cell adhesion molecules such as VCAM-1 (Table 2, Figs 4 and 5). Furthermore, macrophage accumulation, as detected by staining with the F4/80 marker, was increased in diabetic Apoe-KO mice compared with control Apoe-KO mice and also reduced by both AT2RB treatment and At 2 r gene deletion (Fig. 5). Macrophage staining was detected in both the media and the fibrous cap of the atherosclerotic lesions, particularly in diabetic Apoe-KO mice (Fig. 5).

Table 2 Aortic gene expression quantified by real time RT-PCR (arbitrary units) at the conclusion of the 20 week study
Full size table
Fig. 4
figure 4

Aortic protein abundance quantified by immunohistochemistry (% stained area) for MCP-1 (a), VCAM-1 (b), collagen (c) and α-SMA (d) at the conclusion of the 20 week study in control Apoe-KO (C), AT2RB-treated control Apoe-KO (C+PD), control At 2 r/Apoe-DKO (C DKO), diabetic Apoe-KO (D), AT2RB-treated diabetic Apoe-KO (D+PD) and diabetic At 2 r/Apoe-DKO (D DKO). Means±SEM. *p < 0.05 vs control Apoe-KO; †p < 0.05 vs diabetic Apoe-KO

Full size image
Fig. 5
figure 5

Representative aortic sections stained for MCP-1, VCAM-1 and F4/80 in control Apoe-KO (C), diabetic Apoe-KO (D), AT2RB-treated diabetic Apoe-KO (D+PD) and diabetic At 2 r/Apoe-DKO (D DKO). Magnification ×200 for MCP-1 and VCAM-1; ×400 for F4/80

Full size image

Smooth muscle cell recruitment and extracellular matrix accumulation

Atherosclerosis in diabetic Apoe-KO mice involves the proliferation and migration of medial vascular smooth muscle cells into the vessel intima, as demonstrated by increased gene and protein expression of α-SMA compared with control Apoe-KO mice. The atherosclerotic plaque in AT2RB-treated diabetic Apoe-KO mice and diabetic At 2 r/Apoe-DKO mice was significantly less complex, with reduced α-SMA production compared with diabetic Apoe-KO mice (Table 2, Fig. 4).

Extracellular matrix accumulation is a dominant feature of the diabetic plaque. Aortic expression of genes encoding fibronectin (Fn1), collagen I (Col1a1) and collagen III (Col3a1) were significantly increased in diabetic Apoe-KO mice compared with non-diabetic Apoe-KO mice. mRNA levels for these variables were decreased in AT2RB-treated diabetic Apoe-KO mice and in diabetic At 2 /Apoe-DKO mice compared with diabetic Apoe-KO mice (Table 2). Aortic fibrillar collagen content was also assessed by summing the percentage of orange-red (type I collagen) and yellow-green (type III collagen) fibres with polarisation microscopy. Collagen content was significantly increased in diabetic Apoe-KO mice compared with non-diabetic Apoe-KO mice. AT2RB treatment and At 2 r gene deletion significantly reduced collagen content compared with untreated diabetic Apoe-KO mice (Fig. 2).

Discussion

The current study has demonstrated in vivo that both AT2R blockade using a pharmacological approach and At 2 r deficiency induced by gene deletion attenuated to a similar degree diabetes-associated atherosclerosis. This occurred despite similar HbA1c and BP levels between diabetic mice with and without interruption of AT2R. Furthermore, this attenuation of diabetes-associated atherosclerosis, as a result of modulation of AT2R, was associated with reduced levels of pro-inflammatory and profibrotic molecules.

It remains controversial as to what is the best model to study diabetes-associated atherosclerosis. As recently reviewed [29], there are strengths and limitations of the various models. Multiple injections of low-dose streptozotocin has been used to induce diabetes-associated atherosclerosis in Apoe-KO mouse, and this is one of the more popular approaches to be employed by investigators [27, 28, 30]. One of the potentially confounding factors of this model is the increase in lipids seen in the diabetic mice. However, importantly in this study, despite diabetic Apoe-KO mice having elevated cholesterol levels, the diabetic At 2 r/Apoe-DKO mice and AT2RB-treated diabetic Apoe-KO mice had similar elevations in serum cholesterol, yet demonstrated reduced plaque accumulation.

AT2R is produced at very low levels in the cardiovascular system of the adult. Its rate of production changes according to age [39], vessel type and the presence of pathophysiological states associated with tissue remodelling or inflammation, including diabetes. In several models of renal and vascular damage characterised by an inflammatory response, AT2R overproduction has been described [13–15]. In this study, we have reported that At 2 r gene expression is indeed increased in the aorta from diabetic Apoe-KO mice.

The role of AT2R in the vasculature is not well defined, but it has been conventionally considered to act in an opposite manner with respect to the trophic responses mediated by the AT1R subtype. The vasodilator, antigrowth and apoptotic actions of AT2R are postulated to act in a counter-regulatory manner to those conferred by AT1R. Thus, the effects of stimulation of AT2R on the cardiovascular system have been viewed as beneficial and thus no harm has been assumed as a result of increased activation of these receptors. However, there is recent evidence to suggest that activation of AT2R could, in certain contexts, exert growth stimulatory and pro-inflammatory effects that result in complementary rather than opposite effects to AT1R activation [22]. In the context of diabetes, our findings suggest that AT2R may have deleterious effects in the development and progression of atherosclerosis. We have used two disparate approaches to interrupt AT2R-dependent pathways, first a pharmacological blocker and second the use of At 2 r/Apoe-DKO diabetic mice. While no clear anti-atherosclerotic effects of AT2RB treatment or At 2 r deficiency were observed in control animals, as reported previously by other groups [17, 20], interrupting AT2R showed clear anti-atherosclerotic effects in diabetic mice. This is consistent with the view that AT2R activation may enhance vascular injury in certain contexts, particularly in pathological conditions such as diabetes.

Our group has previously shown that the atherosclerotic plaques observed in these animals resemble the complex morphology seen in patients with diabetes, including enhanced accumulation of macrophages, foam cells and cholesterol clefts within the fatty atheroma in association with increased levels of inflammatory markers, pro-inflammatory cytokines and chemokines [2, 40, 41]. In line with these findings, Levy et al. observed that chronic AT2R inhibition by PD123319 in AII-induced hypertensive rats was associated with a reduction in aortic collagen accumulation, hypertrophy and fibrosis [42]. Although the findings from the present study suggest that AT2R plays a pro-atherogenic role in the diabetic setting, this has not been a universal finding. For example, in AII-infused or non-diabetic high-fat fed mice, two different models of atherosclerosis, blockade of AT2R action, either using an antagonist such as PD123319 [20, 43] or deletion of this receptor subtype [17, 19], showed no effect on atherosclerosis, or possibly even an increase.

It remains unexplained why in the context of diabetes we observed a potential pro-atherogenic role for AT2R in this study. Other studies in the atherosclerotic setting were performed in different contexts such as AII infusion [43] and high-fat diet [17] and therefore it is possible that AT2R plays a different role in the various experimental settings that have been examined. However, in keeping with these previously reported findings, the atherosclerosis seen in non-diabetic Apoe-KO mice was not significantly influenced by AT2R blockade or deletion in this study. Indeed, the mechanisms examined in this study that may be responsible for the development of atherosclerosis in the context of high-fat diet and diabetes may be different, although this has not been extensively investigated. This issue clearly needs to be fully clarified with the advent of AT2R agonists such as C21, which has been shown to improve cardiac function in post-myocardial infarction [44]. However, the role of such an agonist has not been assessed in models of atherosclerosis, particularly with concomitant diabetes.

Increasingly, it is appreciated that the biological actions of AT2R involve the participation of AT2R-interacting proteins and promyelocytic zinc finger proteins, but the status of these proteins in the various contexts have not yet been fully examined [45]. Other possibilities for the different results among the various studies include the differences in mouse strains from which the At 2 r/Apoe-DKO mice were generated [31, 46]. For example, in a recent study of At 2 r/Apoe-DKO mice, the At 2 r-KO mice used were on a predominantly FVB background, which is a different background from that of the Apoe-KO mouse [17]. By contrast, in the present study the At 2 r-KO mice were on a C57BL6/J background, the same background as the Apoe-KO mice that were used to generate the At 2 r/Apoe-DKO mice. Finally, in previous reports PD123319 was employed at the lower dose of 3 mg kg−1 day−1 rather than the dose of 5 mg kg−1 day−1 used in the present study and this lower dose could explain the neutral effects seen with the AT2R antagonist in these other studies [20, 43]. The dose chosen in the present study was based on pilot studies and previous experiments using in vitro autoradiography to define a dose effectively inhibiting AT2R without influencing AT1R [13, 32].

AT2R activation is involved not only in the acceleration of atherosclerotic lesion formation but also in promoting pro-inflammatory pathways considered to play an important role in diabetes-accelerated atherosclerosis. Indeed, the diabetic plaque is characterised by increased production of inflammatory cytokines, chemokines and adhesion molecules that promote leucocyte infiltration and foam cell accumulation. Gene and protein expression of VCAM-1 were upregulated in aortas from diabetic Apoe-KO mice and these variables were attenuated in both diabetic AT2RB-treated and diabetic At 2 r/Apoe-DKO mice. In addition, although activation of NF-κB by AII has been reported to be AT1R dependent, there is also a significant body of literature implicating the AT2R subtype in AII-induced NF-κB activation, as reported by Ruiz-Ortega et al. in vascular smooth muscle cells [25]. In the present study, there was upregulation of expression of genes encoding the pivotal NF-κB subunit p65 in association with upregulation of the well-characterised NF-κB-dependent chemokine, MCP-1. Production of both of these proteins was attenuated in diabetic Apoe-KO mice with either gene deletion of At 2 r or with treatment with PD123319. Furthermore, other proteins such as TNF-α that are also pro-inflammatory and linked to NF-κB activation were also modified by AT2R blockade. This prominent inflammatory phenotype previously reported by several groups in diabetic Apoe-KO mice [34] remains to be fully investigated, but it appears that these changes, including macrophage infiltration, may be more prominent in the diabetic setting than in various other pro-atherosclerotic contexts such as high-fat feeding. It is hoped that ultimately more comprehensive elucidation of diabetes-specific mechanisms of atherosclerosis with a particular emphasis on inflammation may lead to a more targeted approach to reduce cardiovascular burden in diabetes.

In this study there was a reduction in extracellular matrix accumulation with blockade or deletion of At 2 r. It remains to be determined if this is a truly beneficial effect with the possibility, albeit unproven in experimental models, that a reduction in collagen content may lead to enhanced plaque rupture. Various other approaches have been reported to reduce matrix accumulation, including blockade of the RAS [2, 28] and molecular and pharmacological strategies to interrupt the advanced glycation pathway [30, 34, 41]. Indeed, some of these strategies that reduce vascular matrix accumulation, when adapted to the clinical context, have been shown to reduce cardiovascular events [47].

The potential clinical significance of AT2R as a target for reducing atherosclerosis in diabetic patients currently remains controversial. One must be very cautious in extrapolating these positive findings linking AT2R to diabetes-associated atherosclerosis to the clinical context. Nevertheless, a recent meta-analysis has suggested the possibility of superiority of an ACEi over an AT1R blocker in reducing cardiovascular events [7]. However, the recent ONTARGET study [8] suggested no advantage of one approach to interrupt the RAS over the other, albeit that that study was performed predominantly in non-diabetic individuals who already had established cardiovascular disease at the time of commencement of the study. Thus, given that currently no AT2RBs are available for clinical use, it remains to be fully clarified whether blockade of AT2R could be a beneficial therapeutic strategy in diabetic patients, particularly since they are at a high risk of premature cardiovascular disease as a result of accelerated atherosclerosis.