Review article

Involvement of the renin–angiotensin system in abdominal and thoracic aortic aneurysms

Hong Lu, Debra L. Rateri, Dennis Bruemmer, Lisa A. Cassis, Alan Daugherty


Aortic aneurysms are relatively common maladies that may lead to the devastating consequence of aortic rupture. AAAs (abdominal aortic aneurysms) and TAAs (thoracic aortic aneurysms) are two common forms of aneurysmal diseases in humans that appear to have distinct pathologies and mechanisms. Despite this divergence, there are numerous and consistent demonstrations that overactivation of the RAS (renin–angiotensin system) promotes both AAAs and TAAs in animal models. For example, in mice, both AAAs and TAAs are formed during infusion of AngII (angiotensin II), the major bioactive peptide in the RAS. There are many proposed mechanisms by which the RAS initiates and perpetuates aortic aneurysms, including effects of AngII on a diverse array of cell types and mediators. These experimental findings are complemented in humans by genetic association studies and retrospective analyses of clinical data that generally support a role of the RAS in both AAAs and TAAs. Given the lack of a validated pharmacological therapy for any form of aortic aneurysm, there is a pressing need to determine whether the consistent findings on the role of the RAS in animal models are translatable to humans afflicted with these diseases. The present review compiles the recent literature that has shown the RAS as a critical component in the pathogenesis of aortic aneurysms.

  • abdominal aortic aneurysm
  • angiotensin
  • angiotensin-converting enzyme
  • renin–angiotensin system
  • thoracic aortic aneurysm


Aortic aneurysms are relatively common diseases that are characterized by pathological luminal dilation. Although the presence of an aortic aneurysm usually has limited observable symptoms, it predisposes the risk for rupture that compromises aortic integrity and frequently leads to fatal haemorrhage. The primary parameter of aortic aneurysms used in clinics is aortic diameter, the principal diagnostic index to predict propensity for aortic rupture. The most common aneurysms are present in the infrarenal region of the abdominal aorta [AAAs (abdominal aortic aneurysms)]. Aneurysms also occur in the thoracic region [TAAs (thoracic aortic aneurysms) with the most predominant location restricted to the ascending aorta. For both AAAs and TAAs, surgery is the only therapeutic option to obviate rupture. Hence there is a dire need to understand the aetiology, pathophysiology and mechanisms of these two pathological states in order to facilitate the development and validation of pharmacological approaches.

Although both AAAs and TAAs are pathological dilations, there are major differences between these two forms of aortic aneurysms. For example, AAAs are highly prevalent in the aged population, with a marked proclivity for males, and smoking being a major risk factor [1,2]. In contrast, TAAs frequently occur in young individuals without overt gender propensity and are mainly associated with genetic disorders of connective tissues [3]. There are also major differences in pathological characteristics of the diseased tissues. Despite these differences, evidence from both animal and human studies suggests that the RAS (renin–angiotensin system) may play an important role in the formation and progression of both forms of aortic aneurysms.

We have reviewed previously the role of the RAS in aortic aneurysmal diseases [4]. Therefore the present review will primarily focus on more contemporary literature on the two types of aortic aneurysms, with an emphasis on contrasting mechanisms. In addition, we will highlight findings from mouse models that have provided mechanistic insights into human aortic aneurysms.


As with most areas of biomedical investigation, mice are the dominant species used to develop models of aortic aneurysms. For AAAs, although there are several modes for generating aneurysms in the abdominal aortic region [5], the three most common models are adventitial calcium chloride exposure [6], intraluminal elastase perfusion [7] and chronic subcutaneous AngII (angiotensin II) infusion [8,9]. There are also several mouse models of TAAs [10]. Many have been associated with Marfan's syndrome and involve manipulation of the fibrillin-1 gene, such as a hypomorphic model [11] and transgenic overexpression of a pathognomic mutation [12]. In addition to genetic manipulations, application of calcium chloride to the descending thoracic aorta [13] and chronic subcutaneous infusion of AngII have also been demonstrated to promote TAAs [14].

This section will briefly introduce commonly used mouse models of AAAs and TAAs, focusing on the role of the RAS in these models. Similarities and differences between AngII-induced AAAs and TAAs in mice are summarized in Table 1.

View this table:
Table 1 Similarities and differences in characteristics of AngII-induced AAAs and TAAs in mice

RAS in AAA models

Development of mouse models of AAAs has led to a pronounced increase in publications on this disease. A widely used approach is to expose the adventitial surface of infrarenal aortas to a concentrated calcium chloride solution (0.25 M) [6]. No studies have determined a role of the RAS in this model. In another common mouse model, AAAs are formed by transient perfusion of elastase [7]. There is a role for the RAS in this model since AT1a receptor (AngII type 1a receptor) deficiency attenuates aortic expansions [15]. Additionally, pharmacological inhibition of ACE (angiotensin-converting enzyme) or AT1 receptors suppresses AAA formation in rats transiently perfused with intra-aortic elastase [1618].

The most direct verification of a role of the RAS in AAAs is the consistent demonstration of aneurysmal formation induced by subcutaneous infusion of AngII into normal or obese C57BL/6 mice [1922], ApoE (apolipoprotein E)−/− mice [9], and LDL (low-density lipoprotein)-receptor−/− mice [8]. Although high plasma cholesterol concentrations are not required for the generation of AngII-induced AAAs, the frequency of pathology is increased by hypercholesterolaemia. However, the incidence of AAAs is similar in ApoE−/− mice fed on a normal laboratory diet and ApoE−/− and LDL-receptor−/− mice fed on diets enriched in saturated fat, despite strikingly different plasma cholesterol concentrations under these conditions [23]. In addition, AAA development is comparable in male C57BL/6 and LDL-receptor−/− mice fed on saturated fat-enriched diets, although these two strains have highly disparate plasma cholesterol concentrations [21]. Therefore, unlike atherosclerosis, there is not a linear relationship between plasma cholesterol concentrations and the incidence of AngII-infused AAAs.

One similarity of AngII-induced AAAs in mice to the human disease is the sexual dimorphic effect. Male mice are more susceptible to AngII-induced AAA formation than female mice [24,25]. An interesting aspect of this prominent sex difference is that it is restricted to AAAs, since atherosclerosis and SBP [systolic BP (blood pressure)] elevations in AngII-infused mice do not differ between male and female mice [2426]. Another feature that is comparable with the human disease is the relationship with BP. Although hypertension has been frequently detected in patients with AAAs, the ADAM (Aneurysm Detection and Management) study screening programme has provided evidence that hypertension is not an independent risk for AAAs [27]. In agreement with human observations, although AngII can increase SBP in mice, this property is not a major contributor to AAA formation [28].

Similar to the human disease, AAAs formed during AngII infusion in mice have complex pathologies with considerable regional and temporal heterogeneity [29]. The first event detected in the aneurysm prone-region during AngII infusion is the accumulation of macrophages in the aortic media, which has been observed as early as 48 h after initiating AngII infusion [30]. Subsequently, there is a focal and transmural medial rupture resulting in lumen expansion that routinely bulges to the left side of the abdominal aorta [31,32]. At the site of medial rupture, the acutely expanded lumen is encased by a thrombus. The thrombus also dissects the adventitia both proximally and distally to the site of medial rupture. In this region, the thrombus surrounds an intact aortic media that appears grossly normal. Macrophages are highly abundant and may assist in thrombus resolution and replacement with fibrous materials. With persistent AngII infusion, there is continuous lumen expansion and remodelling of the aortic wall [33]. Although the heterogeneous nature of human and experimental AAAs provides impediments to objective comparisons, many elements of human AAAs are recapitulated in mice infused with AngII [30,34].

It is apparent that AAAs generated during AngII infusion into mice do not replicate all facets of the human disease. For example, the acute lumen expansion during AngII infusion is attributed to a transmural medial rupture of the mouse aorta that contains three to five elastin layers. This transmural medial break is unlikely to occur in human abdominal aorta containing nearly 30 elastin layers [35]. It remains to be determined whether restricted disruption of medial integrity affecting a limited number of elastin layers, rather than transmural rupture, contributes to AAA formation in humans.

Another discrepancy of AngII-induced AAAs in mice is the suprarenal location that is in contrast with the most prevalent location of human AAAs in the infrarenal aortic region. The suprarenal location has been a reproducible finding in studies of AngII-induced AAAs and other genetic mouse models. These include aged hypercholesterolaemic mice (ApoE−/− and LDL-receptor−/−) [36], eNOS (endothelial NO synthase)-deficient mice [37], and smooth-muscle-cell-specific LRP1 (LDL-receptor-related protein 1)-deficient mice [38]. The basis for the suprarenal location of AAAs in mouse models compared with the infrarenal aortic location in humans is undefined. A possible explanation of the difference in AAA locations may be the distinct blood flow characteristics between mice and humans in aneurysm-prone areas of the aorta [32,39,40].

RAS in TAA models

There are many naturally occurring and genetically engineered mice that develop TAAs. The most frequently used models have manipulations of fibrillin-1, the genetic determinant of Marfan's syndrome. Complete deficiency of fibrillin-1 results in embryonic lethality. In contrast, mice that are hypomorphic for fibrillin-1 are viable, but are afflicted with progressive aneurysmal expansion and dissection of the ascending aorta [11]. There is evidence that losartan attenuates aortic pathology in fibrillin-1 hypomorphic mice [41]. More recently, mice have been developed that express a mutation of fibrillin-1, C1039G, that is commonly found in patients with Marfan's syndrome [12]. The proposed mechanistic basis for fibrillin-1 mutations promoting TAAs that are predominant in the ascending aorta is the inability to effectively sequester latent TGFβ (transforming growth factor β). Ascending aortic dilation that occurs in fibrillin-1 C1039G-expressing mice is ablated by the AT1 receptor antagonist losartan [12]. This finding infers that TAAs generated in this mouse model are a consequence of an interaction, either direct or indirect, between TGFβ and AngII.

Infusion of AngII also promotes pathologies that are localized to the thoracic aorta, predominantly in the ascending portion. This was first detected during AngII infusion into normal C57BL/6 mice [42]. Dissections of the ascending aorta occurred within 6–10 days after initiating AngII infusion. It was subsequently observed that AngII infusion into LDL-receptor−/− mice also led to lumenal expansion that was restricted to the ascending aorta. Lumen expansion progresses with prolonged AngII infusion [33]. Unlike AngII-induced AAA pathology, changes in the aortic media of the ascending aorta include concentric medial thickening in which the distance between elastin layers increases progressively from the lumen to the adventitial side of the aortic wall. Furthermore, elastin fragmentation is more frequent and distributed throughout the media laminae [14,43].


Mechanisms of RAS in AAAs

Multiple arrays of mechanisms have been proposed for the contribution of the RAS to AAAs. These comprise leucocyte infiltration and the consequent inflammatory responses, extracellular matrix protein degradation through activation of a broad range of proteases, and vascular oxidative stress [44,45].

Recent transcriptional profiling and network analyses of mouse AngII-induced AAAs have provided comprehensive information regarding this aneurysmal disease [46,47]. These analyses have demonstrated that AngII increases the abundance of a wide spectrum of genes for inflammation, immunity, matrix degeneration, apoptosis, cell cycling, angiogenesis and signalling pathways [46,47]. Another noteworthy attribute is that these analyses reveal that the pattern of gene abundance differs widely during progressive intervals of AngII infusion. These findings imply that this heterogeneous disease is propelled by complex molecular and cellular mechanisms during progression of AAAs in AngII-infused mice [30,33]. Although a single study has demonstrated that AT2 receptor (AngII type 2 receptor) antagonism by PD123319 administration augmented AngII-induced AAAs [48], there is more extensive evidence that AngII exerts its effects predominantly through the AT1a receptor in AngII-infused mice [48,49]. Table 2 represents a summary of the approaches that have been used to modulate AngII-induced AAAs, and the following text highlights some salient features.

View this table:
Table 2 Recent studies on potential mediators of AngII-induced AAAs in mice

HDAC, histone deacetylase; PPARγ, peroxisome-proliferator-activated receptor γ; RAP, receptor-associated protein; VEGF, vascular endothelial growth factor.

Cell types

Whole-body deficiency of AT1a receptor ablates the development of AngII-induced AAAs [49]. The absolute requirement of AT1a receptor stimulation for AAA formation has been used to determine its role on specific cell type(s). The initial approach was to perform bone marrow transplantation in which chimaeric mice were generated by irradiation of AT1a receptor wild-type and deficient mice and repopulating them with bone-marrow-derived stem cells from mice with either AT1a receptor genotype. This study failed to demonstrate an effect of AT1a receptor on bone-marrow-derived stem cells in the development of AngII-induced AAAs [49]. This result infers that AngII does not directly stimulate AT1a receptors on any leucocyte population to promote aneurysmal formation. Recently, the AT1a receptor has been floxed in mice to permit its cell-specific deletion [43]. The availability of these mice (C57BL/6N-Agtr1a tm1Uky/J, stock number 016211; The Jackson Laboratory) will greatly assist in determining the cell types that are directly stimulated by AngII to induce aneurysmal formation.

Although we have been unable to detect a direct effect of the AngII–AT1a receptor interaction on macrophages during AAA formation, macrophages are the predominant infiltrating cell type present in AngII-induced AAA tissues [9,30,33,49]. Macrophage accumulation in the aortic media occurs during the initiation of AngII-induced AAAs and is persistently present in each stage of AAA progression [30,33]. Macrophage accumulation is also detected in periaortic adipose tissues surrounding the abdominal aorta during AngII infusion [21,22]. Furthermore, macrophage presence is increased in periaortic adipose tissues in AngII-infused mice fed on a diet enriched in saturated fat and is associated with augmented AAA formation in C57BL/6 mice [21]. Conversely, exercise of mice fed previously on a saturated fat-enriched diet not only leads to a loss of body weight, but also ameliorates periaortic macrophage accumulation and reduces aortic expansion [22]. Therefore, although there has not been a direct role for the AT1a receptor on macrophages in AngII-induced AAA formation, the associative evidence in aneurysmal tissues is consistent with a primary role for this cell type in the disease aetiology.

A strategy for determining whether macrophages contribute to AngII-induced AAAs is the use of M-CSF (macrophage colony-stimulating factor)-deficient mice. However, these osteopetrotic mice have many defects that compromise data interpretation [50]. Although depletion of macrophages in osteopetrotic mice was unable to provide insights into AngII-induced AAAs, the impaired function of macrophages has been implicated in a reduction in AngII-induced AAAs using mouse models with a deficiency of either CCR2 (CC chemokine receptor 2), a receptor for MCP-1 (monocyte chemoattractant protein-1) [14,51], or MyD88 (myeloid differentiation factor 88), a component that is important for macrophage-mediated immune responses [52]. Both CCR2 and MyD88 are present on multiple cell types. However, for both proteins, bone marrow transplantation studies have demonstrated that the absence of these proteins on bone-marrow-derived cells diminishes AngII-induced AAAs [51,52]. This evidence is in agreement with a macrophage-function-based mechanism. Other components that have been identified as crucial mechanisms in AngII-induced AAAs include TERT (telomerase reverse transcriptase), a critical element to stabilize telomeres, in macrophages [53].

In addition to macrophages, both B- and T-lymphocytes have been detected in AngII-induced AAAs [9,30]. Nevertheless, the absence of both classes of lymphocytes in Rag-1−/− mice did not affect the incidence or the maximal diameter of AngII-induced AAAs [54].

Several other cell types have been implicated in AAA formation, including mast cells, neutrophils, SMCs (smooth muscle cells) and endothelial cells. However, AngII-related functions of these cell types in AAAs are undefined.


There are only a few studies in which the effects of cytokine deficiencies have been determined on AngII-induced AAAs. Abundance of mRNA for IFN (interferon) γ and CXCL10 (CXC chemokine ligand 10), an IFNγ-inducible T-cell chemoattractant, were increased in aneurysmal tissues of AngII-infused mice. Although deficiency of either cytokine reduced atherosclerosis, the incidence and severity of AngII-induced AAAs was greatly increased [55]. Consistent with these findings, deficiency in a critical molecule in IFNγ signalling, STAT1 (signal transducer and activator of transcription 1), also augmented the formation of AngII-induced AAAs [56]. In mice with CXCL10 deficiency, a concomitant reduction in T-lymphocytes and IFNγ production was observed in aneurysmal tissues with enhanced TGFβ activation. In concert with these results, administration of a neutralizing anti-TGFβ antibody reduced AngII-induced AAAs in CXCL10-deficient mice [55], inferring a beneficial effect of blocking TGFβ signalling on AAA prevention. However, there is a contrasting report that inhibition of TGFβ activity augments the progression and aortic rupture of AAAs in AngII-infused mice [57].

Of the other cytokines examined, no effects on AngII-induced AAAs have been reported in mice administered IFNβ or with p55 TNF (tumour necrosis factor) receptor deficiency [58,59].

Proteases and extracellular matrix proteins

A wide spectrum of proteases have been detected in AngII-induced AAA tissues, including aspartic proteases, serine proteases and metalloproteinases. Among the proteases, MMPs (matrix metalloproteinases) are the most frequently studied, with MMP-2 and MMP-9 being commonly detected in AAA tissues [6062]. A functional role of MMPs in AngII-induced AAAs has been implied by the effect of doxycycline, an MMP inhibitor with broad selectivity, in attenuating AngII-induced AAAs [63,64]. However, there is no direct evidence to implicate the effects of specific MMPs in AngII-induced AAAs.

A serine protease, uPA (urokinase-type plasminogen activator), increases in aneurysmal tissue of AngII-infused mice [19,20,60]. uPA cleaves plasminogen to plasmin that acts on several proteins, including MMPs. Local overexpression of PAI-1 (plasminogen activator inhibitor-1), an inhibitor of uPA, prevents AngII-induced AAAs in mice [65]. Despite the consistent demonstration of uPA activation via AngII stimulation, deficiency of uPA has revealed conflicting results on AngII-induced AAA in mice [19,20]. In hypercholesterolaemic mice, uPA deficiency reduced AngII-induced AAA incidence in ApoE−/− mice, whereas it had no effect in LDL-receptor−/− mice [19,20]. Discrepant results of uPA deficiency were also reported in AngII-infused C57BL/6 mice [19,20]. Although the basis of these contradictory findings is unclear, these results infer further the complexity of AAA mechanisms.

Cysteine proteases and serine proteases have also been proposed to contribute to AngII-induced AAAs. For example, caspase inhibition [66], calpain inhibition [67] or deficiency of granzyme B [68] attenuates AngII-induced AAAs in mice. Cathepsin K deficiency has no effect on AngII-induced AAAs [69]. However, deficiency of cystatin C, an endogenous inhibitor of all cysteinyl cathepsins, augments AAAs in AngII-infused mice [70]. Mechanisms proposed in these reported studies involve extracellular matrix protein degradation and subsequent loss of vessel wall integrity, as well as inflammatory responses.

In addition to activating multiple proteases, AngII also enhances the abundance of osteopontin, a secreted extracellular matrix protein, in the arterial wall [71]. Its presence appears to have functional consequences, since osteopontin deficiency reduces AngII-induced AAAs in mice that are predominantly ascribed to its absence in bone-marrow-derived leucocytes [72].

Arachidonic acid pathway

Enzymes related to release of arachidonic acid and molecules associated with its subsequent modification have been implicated in development of AngII-induced AAAs. Increased arachidonic acid release occurs through the actions of phospholipases. Non-selective pharmacological inhibition of phospholipases reduces AngII-induced AAAs [73]. Deficiency of group X sPLA2 [secretory PLA2 (phospholipase A2)] in mice attenuates AAAs induced by either AngII or calcium chloride [74,75]. PGE2 (prostaglandin E2), a product of the COX (cyclo-oxygenase)/PGE synthase pathway, suppresses the production of inflammatory cytokines, such as TNFα, IL (interleukin)-12 and IFNγ, through binding to its receptors EP (prostanoid receptor)1–EP4. The absence of EP4 in bone-marrow-derived cells increases AngII-induced AAAs [76]. In contrast, mPGES-1 (microsomal PGE synthase-1) deletion attenuates AngII-induced AAAs [77]. LTB4 (leukotriene B4), generated by the activation of 5-LO (lipoxygenase), exerts its effects through the G-protein-coupled receptor BLT1 (LTB4 receptor 1). Genetic deficiency [78] or pharmacological inhibition [79] of BLT1 reduces the incidence of AngII-induced AAAs. Despite the consistent findings that activation of the arachidonic acid cascade contributes to AngII-induced AAAs, studies have yielded inconsistent outcomes regarding the role of 5-LO on AngII-induced AAAs [80,81].

Other potential mechanisms

In addition to the mechanisms described above, many other components have been demonstrated to contribute to AngII-induced AAAs directly or indirectly via inflammation and impaired extracellular matrix protein homoeostasis. AngII promotes oxidative stress. Consequently, administration of the antioxidant vitamin E or deficiency of p47phox, a cytosolic subunit of NADPH oxidase, reduces AngII-induced AAAs [61,82].

Changes in transcription factors also play important roles in AngII-induced AAAs. For example, the loss of the regulator KLF15 (Kruppel-like factor 15) increases AAA incidence and aortic rupture rate [83]. MicroRNAs, as universal post-transcriptional regulators, have also been implicated in AAA formation. miR-29b is up-regulated in AngII-induced AAAs [84], and inhibition of miR-29b reduces AngII-induced AAAs [85].

There is evidence that testosterone promotes site-specific increases in AT1a receptor expression in abdominal aortas of male and female mice to promote AngII-induced AAAs [25]. Administration of testosterone to neonatal female mice imposed long-lasting increases in abdominal aortic AT1a receptor expression and increased AAA susceptibility [26].

Although the cellular and molecular mechanisms have been studied extensively in AngII-induced AAAs, following advancement of the state-of-the-art imaging techniques, disturbed haemodynamics have also been postulated as an initiatory and enhancing factor for AngII-induced AAAs through influencing inflammation and accelerating the degeneration of the arterial wall [86].

In summary, although there is growing evidence that the RAS plays a crucial role in the development of AAAs, definitive mechanisms for AngII-induced AAA formation remain undefined. The pathogenesis of AAAs involves abnormalities in the homoeostasis of a variety of components and intricate signalling regulatory networks. Validation of consistent findings and exploration of the origins of the conflicting results would be important to fully understand this complex pathology.

Mechanisms of RAS in TAAs

Cell types

It is evident that AngII infusion promotes TAAs via binding to the AT1a receptor in mice since whole-body AT1a receptor deficiency ablates the formation of AngII-induced TAAs [43]. Our subsequent study has shown that the AT1a receptor on endothelial cells, but not on SMCs or bone-marrow-derived leucocytes, plays a modulating role in the development of AngII-induced TAAs [43]. Although this result demonstrates the importance of the AngII–AT1a receptor interaction on endothelial cells in the development of AngII-induced TAAs, it does not negate the roles of other cell types. One pronounced feature of the AngII-induced TAAs is that its location is restricted to the ascending aortic portion. During development, there are diverse SMC lineages populating the ascending aorta, which may potentially influence the specific location of AngII-induced TAAs [87].

Mutations in MHC11 (myosin heavy chain 11) [88,89] or α-SMA (smooth muscle α-actin) [90] are associated with inherited TAAs. Furthermore, fibulin-4 deficiency in mice leads to TAAs that are associated with the suppression of SMC-specific contractile proteins and degenerative changes that impair the integrity of the aortic structure [91]. A recent study has also provided evidence that AT1 receptor blockade with losartan attenuates aortic wall abnormality in fibulin-4-deficient mice [92]. This finding indicates that the AngII–AT1 receptor activation contributes to the structural destruction of the aortic wall induced by depletion of fibulin-4. In addition to endothelial cells and SMCs, aortic fibroblasts have been implicated in TAA formation [93]. In thoracic aortic fibroblasts isolated from TAA tissues in mice, the abundance of selected genes for MMPs, collagen/elastic architecture and transcription factors are up-regulated by AngII. These results imply that aortic fibroblasts contribute to progression of TAAs via the AngII-mediated enhancement of inflammation and extracellular matrix proteolysis [93]. Studies have also revealed that leucocyte–fibroblast interactions in aortic adventitia contribute to AngII-induced aneurysmal formation via magnifying vascular inflammation, extracellular matrix remodelling and disruption of aortic structure [42,94]. Altogether, the development and progression of TAAs involve intricate mechanisms and complex pathological changes in the vascular wall structure and interactions of multiple cell types, including leucocytes and resident cell types, in the aortic wall.


A mouse model of Marfan's syndrome with transgenic expression of a mutated form of fibrillin-1 exhibits augmented TGFβ signalling [12]. TAA tissues from patients with Marfan's syndrome also have an increased abundance of TGFβ [95]. Loeys–Dietz syndrome, another inherited disease with a predisposition to TAAs due to mutations in the genes encoding TGFβ receptors (receptors 1 or 2), confirms the importance of TGFβ signalling in TAA pathogenesis [10]. Studies have also provided mechanistic insights that the RAS contributes to TAAs through a convoluted network of cross-talk between AngII and TGFβ [12,9699].

In mice expressing the C1039G mutation of the fibrillin-1 gene, AT1 receptor antagonism substantially diminishes the expansion of ascending aortic diameter [12,100]. In addition to direct effects of AT1 receptor antagonism, there has also been an inference that ascending aortic expansion is associated with AT2 receptor signalling [98]. It has been demonstrated that AT2 receptors are necessary to inhibit TGFβ-mediated activation of cell signalling that contributes to TAA formation and progression.

AngII receptors have also been inferred in other models of ascending aortic aneurysms. In agreement with findings from the Marfan's syndrome mouse model, AT1a receptor deletion ablates AngII-induced TAAs [43]. There are currently no reports describing a role of AT2 receptors in AngII-induced TAAs. In vitro experiments have documented that AngII activates NF-κB (nuclear factor κB) via both AT1 and AT2 receptor activation in rat thoracic aortic SMCs. However, NF-κB-mediated transcription, for example, MCP-1, increases exclusively through AT1 receptors [101]. In contrast with the findings that TGFβ activation promotes TAAs in a mouse model of Marfan's syndrome [12,98], a recent study has reported that TGFβ activity protects against the development of aortic aneurysms and aortic rupture in AngII-infused mice [57]. These conflicting findings highlight the complex mechanisms of TAAs that require further studies to clarify.


RAS in human AAAs

The major approach used to define the relevance of the RAS to human AAAs has been the identification of SNPs (single nucleotide polymorphisms) of specific RAS components. These findings have been discussed in detail in our previous review [4]. In brief, genetic association studies of ACE polymorphisms have generated conflicting outcomes. Many of these studies have used small population sample sizes. A positive association has been observed between the A1166C polymorphism in the 3′-UTR (untranslated region) of human AT1 receptor and AAAs [102]. That study analysed data from large populations and was replicated in three distinct groups of multiple countries.

Retrospective analysis has demonstrated potential benefits of ACE inhibition in preventing aortic rupture [103]. In addition, administration of AT1 receptor antagonists was associated with reduced aortic expansion [104]. In contrast with the findings from these two studies, a recent retrospective analysis reported that ACE inhibition was associated with increased aortic expansion [105].

To explore the definitive effects of RAS inhibition on human AAAs will require randomized double-blind clinical trials. Two ongoing trials are currently evaluating the effects of pharmacological inhibition of the RAS on AAAs. In one study, effects of ACE inhibition (perindopril) on aortic growth rates of small AAAs will be determined ( accession number NCT01118520). In the other study, the effects of a renin inhibitor, aliskiren, on abdominal aortic expansion rate will be assessed in patients with AAAs ( accession number NCT01425242). Both studies have been designed as randomized double-blinded prospective trials.

RAS in human TAAs

There is only limited information regarding genetic associations of the RAS with TAAs. Two studies have investigated the association of the I/D (insertion/deletion) polymorphism of the ACE gene with TAAs in patients with either non-Marfan's syndrome-related diseases [106] or an aortic valve abnormality [107], revealing that the D allele of the ACE gene conferred a risk of TAAs.

Findings in the Marfan's syndrome mouse model expressing the C1039G mutant of fibrillin-1 has stimulated the retrospective analysis of a small cohort consisting of 18 paediatric patients with Marfan's syndrome [96]. Although standard therapy with β-adrenergic blockade (usually atenolol) failed to attenuate aortic dilation, administration of AT1 receptor antagonists (losartan in 17 patients and irbesartan in one patient) for 12–47 months significantly slowed the rate of progressive aortic root dilation [96]. Several prospective double-blinded trials are now ongoing to test the efficacy of angiotensin receptor antagonists on ascending aortic dilation, primarily in patients with Marfan's syndrome [99]. These include a randomized clinical trial designed by the U.S.A. Pediatric Heart Network to compare aortic root growth in subjects with Marfan's syndrome receiving either atenolol or losartan [108]. Completion of these trials will provide information on a number of variables, including different angiotensin receptor antagonists, their use in combination with standard care and different age groups.


Overall, there is strong evidence that the RAS is important in the development of both AAAs and TAAs. AngII promotion of AAAs has been a highly consistent finding, although there are many undefined issues regarding mechanisms. Genetic association studies and pharmacological inhibition of the RAS in human AAAs have yielded variable results that may be attributable to multiple compound factors. Albeit sparse, information from polymorphism association studies and pharmacological inhibition has provided insights into the relevance of the RAS to TAAs. The commonly used mouse models will continue to serve as a feasible approach to understanding complex pathological changes and the intricate molecular mechanisms in order to explore non-invasive therapeutic strategies for these two aortic pathologies. However, there remains a dire need to validate the findings from pre-translational research to determine whether RAS inhibition attenuates human aortic aneurysms and whether different modes of RAS inhibition have differential effects on aortic aneurysms.


Our own work was supported by the National Institutes of Health [grant numbers HL80100, HL062846, HL073085 HL107319 and HL107326].


We appreciate all the efforts from members of our laboratories at the University of Kentucky for generating the data for the many manuscripts described in the present review.

Abbreviations: AAA, abdominal aortic aneurysm; ACE, angiotensin-converting enzyme; AngII, angiotensin II; ApoE, apolipoprotein E; AT1a receptor, AngII type 1a receptor; BP, blood pressure; CCR2, CC chemokine receptor 2; COX, cyclo-oxygenase; CXCL10, CXC chemokine ligand 10; EP, prostanoid receptor; IFN, interferon; IL, interleukin; KLF15, Kruppel-like factor 15; LDL, low-density lipoprotein; LO, lipoxygenase; LRP1, LDL-receptor-related protein 1; LTB4, leukotriene B4; BLT1, LTB4 receptor 1; MCP-1, monocyte chemoattractant protein-1; M-CSF, macrophage colony-stimulating factor; MMP, matrix metalloproteinase; MyD88, myeloid differentiation factor 88; NF-κB, nuclear factor κB; PGE, prostaglandin E; mPGES-1, microsomal PGE synthase-1; PLA2, phospholipase A2; RAS, renin–angiotensin system; SBP, systolic BP; SMC, smooth muscle cell; sPLA2, secretory PLA2; STAT1, signal transducer and activator of transcription 1; TAA, thoracic aortic aneurysm; TERT, telomerase reverse transcriptase; TGFβ, transforming growth factor β; TNF, tumour necrosis factor; uPA, urokinase-type plasminogen activator


View Abstract