Abstract
Free full text
Cellular Mechanisms of Aortic Aneurysm Formation
Abstract
Aortic aneurysms are a common vascular disease in Western populations that can involve virtually any portion of the aorta. Abdominal aortic aneurysms are much more common than thoracic aortic aneurysms and combined they account for over 25,000 deaths in the United States annually. While thoracic and abdominal aortic aneurysms share some common characteristics including the gross anatomic appearance, alterations in extracellular matrix and loss of smooth muscle cells, they are distinct diseases. In recent years, advances in genetic analysis, robust molecular tools, and increased availability of animal models have greatly enhanced our knowledge of the pathophysiology of aortic aneurysms. This review examines the various proposed cellular mechanisms responsible for aortic aneurysm formation and identifies opportunities for future studies.
Introduction
The cellular mechanisms responsible for aortic aneurysm formation constitute a complex, orchestrated series of events that result in dramatic pathological changes in the anatomy and function of the arterial wall. Of fundamental importance is the understanding that there are clear and distinct differences between the mechanisms underlying thoracic and abdominal aortic aneurysms. Thus, while the physical appearance of thoracic and abdominal aortic aneurysms have striking similarities, the pathophysiological change of these two diseases are quite distinct. This review is organized by first focusing on the mechanisms of abdominal aortic aneurysm (AAA) formation followed by similar considerations for thoracic aortic aneurysm (TAA) formation. Aneurysms involving other arterial beds including cerebral and peripheral arterial aneurysms are outside of the scope of the current review. The cellular mechanisms discussed do not include recent advances in the studies of non-coding RNAs or the genetic basis of aortic aneurysms which are specifically discussed in different sections of this compendium.
Abdominal Aortic Aneurysms
Abdominal aortic aneurysms (AAA) are a major cause of morbidity and mortality and it is estimated that the incidence of AAA in men increases by 6% per decade after age 65. A predictive modeling study based on known risk factors suggested that there may be over 1 million people in the US today with AAA.1
The epidemiology of AAA formation appears to be distinct from that of atherosclerotic disease. One of the earlier definitive studies of risks factors for AAA was the Aneurysm Detection and Management (ADAM) Veterans Affairs Cooperative Study Group.2 This retrospective analysis found that a history of cigarette smoking was by far the strongest risk factor for AAA carrying a relative risk of 5.9 when compared to non-smokers. The second most potent risk factor was an age-independent family history of AAA with a relative risk of 1.9. Hypertension, hypercholesterolemia and pre-existing coronary artery disease carried a relative risk of less than 1.5 suggesting that the disease mechanisms of AAA may diverge from those of atherosclerosis.
Contributions of Different Cell Types
Apoptosis of smooth muscle cells and degeneration of the aortic media have long been identified as hallmark of AAA pathology.3 Inflammation, production of reactive oxygen species, and ER stress have all been associated with smooth muscle cell apoptosis in AAA.4 This loss of structural integrity is key to aortic dilation and rupture. Of interest is that the vast majority of AAA occur below the level of the renal arteries which may reflect the differing embryologic origins of vascular smooth muscle cells in the distal abdominal aorta where the mesoderm gives rise to aortic vascular smooth muscle cells whereas the thoracic aorta smooth muscle cells arise from the neural crest.5
While it is obvious that changes in vascular smooth muscle cells of the media are pivotal to the development of AAA, many other cell types are involved in addition to smooth muscle cells including endothelial cells,6 neutrophils,7, 8 monocyte/macrophages,9, 10 lymphocytes,10, 11 adipocytes,12, 13 mast cells,10, 14, 15 and platelets.16 The functional contributions of these cell types is sometimes obvious as in the case of vascular smooth muscle cells. In other cases, studies have utilized depletion strategies in order to define their relative contributions.
The precise role of the endothelium has not yet been fully explored but, it is clear that these cells do play a critical role in that eNOS uncoupling17 and endothelial biomechanical signal transduction18, 19 have roles in AAA formation. In the case of platelets, while von Willebrand factor (VWF) may not be essential for AAA formation,20 the presence of thrombus in AAA portends a worse outcome21, 22 and suggests that there are yet to be determined functional contributions of platelets to AAA development and rupture.
Matrix Metalloproteinases
In both AAA and TAA, the amount of elastin, collagen and glycosaminoglycans is reduced compared to normal aortas. An imbalance between the amount of active matrix metalloproteinases (MMPs) and their inhibitors is responsible for most of these changes.23, 24 Thus far, 23 different MMPs have been described in humans and are divided into archetypal, matrilysins, gelatinases and furin-activated MMPs.23 In the normal aorta, endothelial cells, smooth muscle cells, and adventitial fibroblasts are responsible for MMP production. In the setting of AAA, inflammatory cells serve as additional sources of MMPs.23, 24
Increased MMP-1 (collagenase-1) expression has been described in human AAA.25, 26 Along with increased MMP-1 expression, there is a concurrent decrease in the levels of MMP-1 inhibitors.27 However, clinical studies have not been able to correlate genetic polymorphisms of the MMP-1 promoter region with a clinical outcome.28 MMP-1 is produced constitutively by fibroblasts and smooth muscle cells (SMCs) in the aortic wall and MMP-3 or plasmin are required for its activation. Inflammatory cells represent a minor source of MMP-1.29 (18)F-FDG uptake detected by PET in the abdominal aorta is seen in patients with symptomatic AAAs and those at high risk of rupture. Histologically, these sites of uptake correlate with an increase in adventitial inflammatory infiltrates, a reduction in smooth muscle cells and increased in MMP-1 and MMP-13 expression.30
MMP-13 (collagenase-3) expression is increased in human AAAs, especially in symptomatic AAAs and those at high-risk for rupture.25, 31, 32 MMP-13 is produced primarily by SMCs in the aortic wall. A polymorphism (−77A/G) is an independent risk factors for AAA formation, suggesting MMP-13 plays a relevant role in AAAs.27 In mice, nitric oxide (NO) was found to increase the production of CD147 leading to greater levels of MMP-13 and worsening of elastase-induced AAAs. Silencing of CD147 by RNA interference or pharmacological inhibition of iNOS lead to the inhibition of MMP-13 expression and a subsequent decrease in aortic dilation.33
MMP-3 (stromelysin-1) is also expressed at high levels in the wall of AAAs.34 This enzyme, produced by fibroblasts and epithelial cells, appears to be produced by macrophages in the setting of AAA.23 A MMP-3 gene promoter region polymorphism termed 5A/6A (5 adenines vs. 6 adenines at −1612) augments transcriptional activity and serves as an independent risk factor for AAA formation.27
MMP-12 (Metalloelastase) is produced and secreted primarily by macrophages. 35 The role of MMP-12 in the pathogenesis of AAAs is not entirely clear. Experiments in MMP-9 knockout (KO), MMP-12 KO and MMP-9/MMP-12 KO mice using the elastase model of AAA formation revealed that MMP-12 KO mice developed AAAs (similar to wild-type controls). However, MMP-9 KO and MMP-9/MMP-12 were resistant to AAA development.36 Recently, experiments using the CaCl2 model of AAA formation in mice with a genetic inactivation of PI3-kinase delta revealed a significant upregulation of MMP-12 expression and enhanced AAA formation.37 MMP-12 co-localizes with CD68-positive macrophages in the aneurysmal wall. In vitro, genetic inactivation of PI3-kinase delta or treatment with a PI3-kinase delta inhibitor led to enhanced macrophage migration and increased MMP-12 expressio.. 37 While MMP-12 may not be directly involved in the pathogenesis of AAAs, it apparently facilitates other MMPs in extracellular matrix degradation.
MMP-2 (gelatinase A) degrades e;aston and is expressed constitutively in smooth muscle cells, fibroblasts and in macrophages.24, 38 Ang II and CaCl2 augment the activity of MMP-2 in the abdominal and thoracic aorta.38-40 In vivo studies revealed MMP-2 systemic KO had no impact on Ang II-induced AAAs but was protective of CaCl2-induced AAAs.25,27
MMP-9 (Gelatinase B) is constitutively produced by fibroblasts and smooth muscle cells and by infiltrating adventitial macrophages during AAA formation.41 MMP-9 mRNA and protein expression in the aortic wall and MMP-9 plasma levels are significantly greater in patients with AAAs, and higher levels are seen in patients with a luminal thrombus.40, 42, 43 MMP-9 KO mice proved to be resistant to elastase-induced AAAs.44 Infusion of wild-type macrophages into MMP-9 KO mice led to reconstitution of AAA, suggesting macrophage-derived MMP-9 is crucial for AAA formation.40 Recently, a genomic-wide association study of patients with AAA identified an interaction between the MMP-9 gene expression and the expression of genes involved in inflammation/immune function) and cholesterol metabolism.45
MMP-14 (Membrane type 1-MMP) is produced by infiltrating macrophages and smooth muscle cells in the aortic wall. Higher MMP-14 mRNA and protein expression levels have been found in human AAAs.46 In murine CaCl2-induced AAAs, macrophage-derived MMP-14 plays a crucial role in the direct degradation of the extracellular matrix in the tunica media and adventitia leading to AAA formation. Macrophage-derived MMP-14 directly regulates macrophage elastolytic activity even in the setting of MMP-2 deletion.47
Given the pathological process of AAA formation, it is not surprising that an array of MMPs have been implicated involving multiple cell types. It is difficult to assign central causality to a specific MMP and cell type. Indeed, as indicated above, it appears that several MMPs are necessary, but not sufficient to support AAA development as it is most likely that the different cell types act with temporal and spatial specificity during AAA initiation and growth.
The Renin-Angiotensin System
The demonstration of induction of AAA by angiotensin II (Ang II) infusion in apoE48 and LDL49 receptor knockout mice is certainly the most direct data in terms of documenting a causal relationship between the renin-angiotensin system and AAA formation. Earlier reports in mice overexpressing angiotensinogen and renin on a high salt diet first suggested this relationship as these animals developed AAA.50 These data are consistent with additional studies showing that angiotensin converting enzyme inhibition limited AAA formation in the elastase model.51 Thus, it appears that angiotensin II may induce AAA but, it may not be sufficient as in most cases, additional factors (e.g., elevated cholesterol) appear to be required to induce AAA. The source of Ang II generation is complex and several studies raise the possibility that, in addition to angiotensin converting enzyme, chymase expressed in mast cells may contribute to local generation of angiotensin II.15, 52 The absolute magnitude of the contribution is less clear due to the additional role of chymase in MMP activation and apoptosis.53, 54
Angiotensin II effects on the cellular components of the aorta have been extensively studied55–59 and include many of the cellular mechanisms described below including production of reactive oxygen species, induction and activation of MMP’s, and infiltration of inflammatory cells. Rather disappointingly, clinical trials have failed to demonstrate a benefit of angiotensin converting enzyme inhibition on AAA growth rate60 and in at least one study, may have had an adverse effect.61 The reason for this discrepancy is unclear and may reflect deficiencies in the available animal models, effects on disease initiation vs. progression, or the endpoint of aortic diameter.
Of note, recent data also suggest a role for the mineralocorticoid receptor in AAA formation as aldosterone administration in non-hyperlipdemic mice was shown to induce AAA, an effect that was age dependent with a more pronounced phenotype in older animals.62
Inflammation: Cells and Pathways
A hallmark of AAA formation is an intense inflammatory response involving essentially all of the classic cellular constituents of inflammation as well as local inflammatory responses in the arterial wall. Neutrophil infiltration occurs very early on in AAAbut is transient. Observations in the elastase model showed that treatment with neutrophil neutralizing antibodies slowed AAA expansion suggesting a functionally relevant role for neutrophils 63, 7. Similarly, resolvins D1 and D2 have been shown to inhibit AAA formation.64, 65 This effect was initially attributed to neutrophil-derived MMP’s. However, more recent data suggests that neutrophil extracellular traps may be an important component of the continued inflammation in AAA.65, 66 Finally, neutrophils are a source of reactive oxygen species as described below that can be generated by several enzymatic sources including the NADPH oxidase and myeloperoxidase.67
Macrophage infiltration of AAA is one of the most consistent pathological findings and certainly not surprising given the biological function of these cells and their production of MMPs, cytokines, and chemokines as well as their ability to remove cellular debris.29, 68, 69 Interestingly, owing to their ability to express different phenotypes encompassing both inflammatory and reparative functions, macrophages participate in both the pathogenesis of AAA as well as the repair response.69 This can occur through classical cytokine pathway or through more novel mechanisms. Macrophages can directly influence smooth muscle cell function via macrophage-derived netrin-1 (a protein classified as a neuronal guidance cue but known to promote macrophage retention in tissues) as evidenced by the finding that deletion of netrin-1 results in a reduction in AAA formation via a MMP3-depedent mechanism.70
Both T cells and B lymphocytes are relatively abundant in aneurysmal tissues and both cell types have been implicated in the disease process. B cell depletion is protective against AAA formation in both the angiotensin II and elastase models.71 The mechanism appears to involve enhancement of a unique population of indole 2,3-dioxygenase–expressing plasmacytoid dendritic cells and regulatory T cells with a subsequent decrease in inflammatory gene exression.71 However, there exist additional layers of complexity in that different B cell subtypes (B1 and B2) may play different and opposing roles in AAA.72 B2 cells appear to be the dominant subtype in AAA which is consistent with their known role in atherosclerosis.71
T cells in general are important in the development of AAA as it has been shown that T cell depletion attenuates AAA formation.73 Regulatory T cells play a role in protection against AAA formation74 likely due to the fact that they secrete IL-10, an anti-inflammatory cytokine75 as well as TGF-β, which can have a stabilizing role for AAA.76 While the precise role of CD8+ T cells in AAA remains unclear,77 CD4+ Th1 and Th2 cells do appear to contribute to the pathogenesis of AAA. In the setting of AAA, there are conflicting data showing that Th1-derrived interferon gamma has a role in inducing AAA78 as well as protecting against AAA formation.79 There are also conflicting studies as regards the ratio of Th1 to Th2 cells.80 These differences potentially reflect different states of disease progression or perhaps differences in animal models.
Cytokine production by both immune cells and cells native to the vessel wall ultimately drive the inflammatory responses leading to AAA formation. As reviewed in detail elsewhere, numerous studies have identified the involvement of multiple cytokines in AAA formation.81 These encompass both inflammatory and anti-inflammatory cytokines. While extremely complex and incompletely understood, suffice it to say that cytokine production by both inflammatory cells and the resident cells of the vascular wall contribute to a pathological feedback loop further enhancing inflammation.
Reactive Oxygen Species
Reactive Oxygen Species (ROS) play a central role in the development of AAA. Some of the earlier observations in human tissues showed that superoxide levels were elevated in the smooth muscle and inflammatory cells of AAA specimens.82 Many of the known pathological effects of excess ROS including activation of MMPs, induction of pro-inflammatory genes, and apoptosis, are key pathological features of AAA. Animal studies have shown that administration of vitamin E as an antioxidant led to decreases in AAA size and rupture.83 Work from our own laboratory has shown that smooth muscle-specific overexpression of catalase prevented early mechanical changes in the aortic wall after Ang II infusion84 as well as inhibition of AAA formation in the CaCl2 model.85
NADPH Oxidase86–88 is a major source of ROS in AAA as evidenced by increased expression of the NADPH oxidase in segments of human aorta from patients with AAA.82 Specific deletion of the p47 subunit of NADPH oxidase resulted in attenuation of AAA in two different models.17, 89 The NADPH oxidase system is complex and is composed of a family of NOX homologs identified by their catalytic subunits (NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1, DUOX2). The different homologs exhibit varied tissue expression and produce superoxide anion and/or hydrogen peroxide. There exists some controversy as regards the precise role of the various NOX homologs in AAA. Siu et al17 reported that deficiency of NOX1, NOX 2 or NOX 4 as well as deletion of the non-catalytic p47 subunit resulted in complete inhibition of AAA. However, Kigawa et al90 showed that NOX2 deficiency resulted in increased diameter and extent of AAA that was associated with overall decreased ROS production and polarization of macrophages towards aninflammatory phenotype. The reason for this discrepancy may be due to the models used, potential compensatory upregulation of other NOX homologs, or cell-specific effects. The above studies allow one to conclude that, even though the involvement of the specific NOX homologs is unclear, there is an abundance of evidence supporting the role of NADPH oxidase in AAA formation.
Uncoupled endothelial nitric oxide synthase (eNOS) has been proposed to be a contributor to ROS production in a variety of vascular diseases processes in general91 and specifically in AAA.17, 92–94 AAA occur after angiotensin II infusion in hph-1 mice95 which have impaired activity of GTP cyclohydrolase leading to a deficiency in tetrahydrobiopterin which in turn causes uncoupling of eNOS leading to production of superoxide. Importantly, folic acid administration was shown to prevent AAA formation. Ang II infusion in apoE mice leads to decreased tissue and circulating levels of tetrahydrobiopterin,96 an effect likely due to the degradation of tetrahydrobiopterin by oxidative stress. These data suggest that ROS derived from uncoupled eNOS also contribute to AAA formation and that eNOS uncoupling occurs as a result of superoxide production by NADPH oxidases. In essence, uncoupled eNOS can act to amplify the effects of NADPH oxidases.
Myeloperoxidase may also contribute to ROS generation in AAAas neutrophil infiltration of the abdominal aorta occurs early on in experimental models. Neutrophil depletion inhibits AAA formation,7 an effect which likely has multiple mechanistic impacts given the functional contributions of neutrophils. However, neutrophils are the primary source of myeloperoxidase, accounting for up to 5% of the total protein content of neutrophils.97 Genetic deletion of myeloperoxidase or inhibition of myeloperoxidase by administering taurine inhibits AAA formation, confirming the importance of this enzyme in AAA formation.67 Hypochlorous acid (a product of myeloperoxidase) is highly reactive with many biomolecules including lipids. Subsequent lipid oxidation may be one the key steps in oxidative damage in the setting of AAA as upregulation of serum amyloid A, a major target of lipid oxidation, occurs in AAA and genetic depletion of serum amyloid A also results in inhibition of AAA formation.98
Other sources of reactive oxygen species including xanthine oxidase, iNOS, cyclooxygenase, and mitochondrial metabolism may also play a role in the pathogenesis of AAA however, studies from human AAA samples suggest that the contributions of xanthine oxidase and mitochondrial respiration may be less important.99 Deletion of mitochondrial uncoupling protein-2 (UCP-2) resulted in increased incidence of AAA in the Ang II model of AAA though whether this a direct effect on production of ROS by mitochondria or a secondary effect through regulation of other pro- and antioxidant enzyme systems is unclear.100
Dysregulation of antioxidant systems may also determine local and systemic levels of ROS. Catalase expression has been shown to be decreased in aortic tissue in animal models of AAA84 as well as in circulating polymorphonuclear neutrophils and plasma obtained from patients with AAA.101 Studies of the potential involvement of superoxide dismutase (SOD) have shown both increased and decreased activity in AAA samples from patients102,103 Specific analysis of SOD isoform expression showed increases only in MnSOD expression104 with another study reporting a decrease in ecSOD.105 Glutathione peroxidase103 and paraoxinase-1106 also appear to be decreased in human AAA specimens.
The precise mechanisms through which ROS elicit cellular responses has been reviewed previously107–111 and remains an active area of study that is well beyond the scope of this review. It is clear that dysregulation of ROS leads to widespread changes in the arterial wall that are central in the process of AAA development. The result being that increased levels of reactive oxygen species lead to increased expression of pro-inflammatory gene products, smooth muscle apoptosis, increased expression and activation of MMPs, and other key events in the pathogenesis of AAA.
Risk factors and mechanisms
The strongest risk factors for AAA are male sex, family history, and cigarette smoking. How these risk factors impact the known cellular mechanisms of AAA is key to our understanding of the disease. The overall rate at which AAA affects men is higher than women at ratio of 5:1.112 In both the elastase113 and the angiotensin II models of AAA114, male mice exhibit a higher incidence and larger AAAs when compared to females. A very striking gender-based difference in the disease is its progression over time. A population based study examining the incidence of AAA in men and women as a function of age showed that in the 40–60 year age group (pre- and early post-menopausal women), the rate of AAA was eleven times higher in men than women. In the 60–90 year group this difference falls to three times, and by age 90 men and women have AAA at an equivalent rate.115
There is not a single, causal mechanism responsible for the sex-dependent differences in the incidence of AAA. As recently reviewed116 there are many clinical and pre-clinical studies examining the role of sex hormones on AAA demonstrating complex and sometimes paradoxical effects. While endogenous estradiol is likely protective, exogenous estrogen replacement has shown conflicting results in human studies with both increases117 and decreases118 in AAA events in women receiving hormone replacement therapy. Conversely, androgen signaling increases production of reactive oxygen species, expression of angiotensin II receptors,119 inflammation in the arterial wall116 and counteracts the effects of exogenous estrogen.120 Very interestingly, specific loci on the Y chromosome have been linked to activation of the renin angiotensin system and subsequent AAA formation.114, 121 These data suggest that androgen-mediated events are the primary drivers of sex differences in AAA. However, in human studies, low testosterone levels were shown to be independently associated with the presence of AAA.122 Whether this discrepancy is due to differences in disease state or is related to lack of translation of animal models to the human disease state is unclear.
Work from our laboratory has suggested that local difference in vascular hemodynamics may also contribute to the sex differences in AAA in that local shear stress patterns differ between males and females as a result of differences in abdominal blood flow patterns to the reproductive organs.19 In addition, mechanical properties of the aorta exhibit sex-dependent differences that are also age-dependent.123–126 This may be due to altered extracellular matrix content126 and cross linking of the extracellular matrix which ultimately reflect androgen-mediated physiology as described above. The downstream effectors of differential androgen and estrogen signaling involve essentially all of the same cellular mechanisms of AAA as described above with reported differences in production of ROS,127 inflammatory mediators, extracellular matrix composition,128 and inflammatory cells.
The mechanisms of other risk factors is less clear. Genetic factors are discussed in a separate section of this compendium. Cigarette smoking is a complex stimulus and there are likely multiple components in cigarette smoke that promote vascular disease. It has been shown that cigarette smoke extract induces expression of MMP-2 as well as MMP-9 and that nicotine, through increased expression of MMP-2 induces AAA.129, 130 The mechanisms of other, protective risk factors including diabetes and race remain unknown.
Thoracic Aortic Aneurysms
The prevalence of TAA is less than that of abdominal aortic aneurysms but they still occur in up to 6–10 people per 100,000. Approximately 20% of TAA can be attributed to a specific genetic variant. The differences between TAA and AAA reflects several fundamental differences in their pathophysiology as exemplified by the fact that multiple genetic variants have been identified for TAA which do not confer increased risk for AAA. While there is clear evidence for increased risk for AAA among family members of individuals with AAA and clusters of AAA within families, there is surprisingly little overlap between AAA and TAA.
TAAs can be broadly categorized as syndromic, familial non-syndromic, and sporadic with the latter category being quite diverse. This heterogeneity of TAA complicates mechanistic studies using human samples and animal models as TAA is a multi-factorial disease with multiple unique mechanisms resulting in similar clinical presentations. The classic syndromic TAAs include Marfan, Loeys Dietz, Ehlers-Danlos, etc., while the familial non-syndromic TAAs encompass a wide variety of genetic variants (see the section of this compendium on the genetics of aneurysms). Overlapping this construct is TAA associated with a bicuspid aortic valve as a bicuspid valve occurs at higher rates in some specific syndromes (e.g., Marfan Syndrome) and also in the general population where a single genetic variant is not present. 131
Much of the fundamental science of TAA stems from genetic studies and observations made in mouse models that focused on alterations in extracellular matrix components, smooth muscle function, and related proteins. The volume of data related to mechanisms of TAA is less than that for AAA, likely reflecting differences in the ease and applicability of small animal models for AAA vs. TAA. However, significant insights into the mechanisms of this disease have been obtained over the past several decades and we now more fully appreciate that TAAs represent a spectrum of disease pathologies that are the result of complex changes in the cellular and extracellular environment and not a simple degenerative process.
A central role of the Extracellular Matrix
As is the case for AAA, TAA is characterized in part by abnormalities in the extracellular matrix that compromise the structural integrity of the aorta. Studies have implicated several genetic variants in proteins that directly impact the mechanical characteristics of the aorta leading to TAA. Among these is lysl oxidase, which is responsible for cross linking collagen and elastin. Genetic data are supported by experimental studies in which administration of beta-aminopropionitrile (a lysl oxidase inhibitor) to mice in the setting of angiotensin infusion leads to aneurysmal dilation of the aorta at multiple locations.132 Interestingly, beta-aminopropionitrile is naturally occurring in sweet peas133 and either feeding experimental animals sweet peas133 or farm animals eating sweet peas results in aortic aneurysms among other manifestations related to loss of collagen and elastin structure. It is important to note that the effects of beta-aminopropionitrile are not specific to the thoracic aorta suggesting that other factors are likely involved. In addition, it is unknown if genetic variants in other members of the lysl oxidase family contribute to TAA formation.
Some forms of Ehlers-Danlos Syndrome are associated with TAA and aortic dissection. Ehlers-Danlos Syndrome encompasses a group of diseases due to mutations in collagen genes. The vascular phenotype variant of Ehlers-Danlos (type IV) is due to one of several mutations in the type III pro-collagen gene and is associated with TAA and other vascular abnormalities.134 The formation of aneurysm and dissection in affected individuals is presumably due to the compromise of the mechanical integrity of the wall.
One of the initial, major contributions to our understanding of TAA came from observations related to Marfan Syndrome with the finding there was a mutation in the fibrillin 1 gene.135 Subsequently, additional mutations in fibrillin 1 have been identified. Fibrillin 1 is an extracellular matrix glycoprotein that serves as the major structural component of microfibrils in the aortic wall and while the structural contributions of this protein to the arterial wall may be important, Neptune and colleagues proposed that the mutations in fibrillin 1 actually functioned through dysregulation of TGF-β signalling.136 The model that emerged was that fibrillin-1 binds to the large latent TGF binding protein that sequesters TGF-β in a complex and prevents it from becoming active.137–141 Subsequent work has also implicated variants in the large latent TGF binding protein.142 In Marfan syndrome, it was proposed that the mutations in fibrillin 1 led to increased availability of TGF-β and subsequent activation of TGF-β mediated signaling pathways.141 With this model, transcriptional regulation occurs through pathways that are both Smad-dependent and -independent.143 Animal studies performed with either mutated or knocked out fibrillin-1 mimicked Marfan Syndrome. While there are reports of increased TGF-β in aortic tissue of Marfan patients, other reports demonstrated increased total TGF-β but no change in active TGF-β despite increased expression of total TGF-β and activation of Smad2.144 Furthermore, while Smad2 is a downstream effector of TGF-β, other signaling pathways also converge on Smad2. Recent publications using a Marfan mouse model with mutated fibrillin-1 demonstrated that genetic disruption of TGF-β receptor increased aortic dilatation suggesting a protective effect of TGF-β signaling as opposed to a causal role.145, 146 Thus, the precise role of TGF-β signaling in Marfan Syndrome remains unclear and is a source of significant controversy.143, 145, 147 Fibrillin-1 is clearly important and may have alternative mechanisms of action related to its structural contributions, integrin binding capacity, involvement in sequestering other growth factors, or other, as yet undetermined functions.
Additional variants in the structural components of the arterial wall have been linked to TAA are reviewed in detail in the section of this compendium devoted to the genetic basis of aneurysmal diseases. There is clearly a significant contribution of variants in the structural proteins of the aortic wall to the pathogenesis of TAA in both syndromic and sporadic disease.
Contributions from Smooth Muscle Cells
Smooth muscle cells are the major cellular constituent of the aorta and their loss through apoptosis or necroptosis is a major defining feature of both AAA and TAA. Genetic studies have shown that several mutations in the contractile proteins of smooth muscle cells predispose individuals to TAA. This suggests that smooth muscle contractile function plays an important role in TAA though the mechanism is unclear. Possibilities include both a structural, load bearing function or a signal mechanical signal transduction function as proposed in the section of this compendium devoted to the genetic basis of aortic aneurysms.
Matrix Metalloproteinases
As is the case with AAA, matrix metalloproteinases play a pivotal role in the remodeling of TAA but they are far less studied. MMPs are potentially involved in both the maladaptive degradation of the extracellular matrix as well as the adaptive remodeling of extracellular matrix to stabilize an enlarging aorta. The largest body of data related to TAA involve MMP-2 where the results are somewhat paradoxical. In contrast to AAA, MMP-2 deletion increased Ang II-induced TAA formation but was also protective of CaCl2-induced TAAs.36 A potential explanation for these findings is that there is a higher content of elastin in the thoracic aorta, making it more susceptible to extracellular matrix synthesis alterations whereas the abdominal aorta more susceptible to excess extracellular matrix degradation. Systemic deletion of the MMP-2 gene in a mouse model of Marfan Syndrome delayed TAA rupture, inhibited the activation of TGF-β and phosphorylation of ERK1/2 and SMAD2, preserved aortic lamellar integrity, and prolonged mice survival. MMP-9 protein and activity levels in the aortic wall are increased in the setting of TAA.148, 149 An ex vivo study revealed that after Ang II administration, there is SMAD2 activation, increased MMP-9 expression via ERK, and subsequent extracellular matrix degradation.148 However, in TAAs associated with bicuspid aortic valves, MMP-9 expression was not augmented.149 In addition, MMP-3 may play a role in the development of human ascending TAAs (4.0- to 5.9-cm size range), especially in older patients.150
Inflammation
Inflammation is a cardinal feature of AAA but the data supporting a role for inflammation in TAA are less extensive. There are intriguing data suggesting that there is some commonality between AAA and TAA and that the structural changes described above may lead to increased production of reactive oxygen species and inflammatory responses. Both T cells and macrophages are present in the media of TAA151 a finding that is supported by genomic analysis of human TAA samples which revealed upregulation of multiple inflammatory pathways.152 T cells may be a source of FAS ligand that results in smooth muscle apoptosis, a cardinal feature of TAA. Specific mediators of inflammation in TAA are not well defined. One study reported elevated IL-6 levels in TAA relative to both normal tissue and tissue from AAA.153 Furthermore, inhibition of IL-6 signaling inhibited aneurysm growth in the murine elastase model of TAA.153 Similar data from the same group were obtained with regard to IL-1β.154 Thus, while there is evidence of inflammation in TAA and plausible mechanisms for the induction of inflammation in TAA, the data set currently available is somewhat limited.
Reactive Oxygen Species
The role of ROS in TAA has received less attention as compared to studies of AAA. Aortic tissue from patients with Marfan syndrome have increased levels of oxidative stress, decreased expression of antioxidant enzymes, and increased expression of iNOS, xanthine oxidase, and NADPH oxidase subunits.155 Similarly superoxide dismutase 3 expression in the ascending aorta is downregulated in patients with bicuspid aortic valves. Smooth muscle cells from the aortas of patients with Marfan Syndrome have increased expression of NOX4 and enhanced production of H2O2.156 More mechanistic data have been obtained in models of Marfan and Loeys-Dietz syndrome where altered mitochondrial respiration and increased production of ROS have been reported.157
Risk factors and mechanisms
The familial risk of TAA is discussed in a separate section of this compendium. For sporadic TAA, hypertension, cigarette smoking, and atherosclerotic disease are the strongest risk factors. The mechanisms behind these risk factors is unclear is generally unclear except in the case of atherosclerosis which shares many common disease mechanisms. Hypertension can impact the vessel wall through direct mechanical effects on inflammation or through humoral stimuli. Female sex and diabetes are also protective in TAA but the mechanisms are unknown.
Discussion
Perhaps the major conclusion to be derived from the evolving literature defining the cellular and molecular mechanisms of thoracic and abdominal aortic aneurysms is that while thoracic and abdominal aortic aneurysms have striking similarities at the gross anatomical level, the underlying pathophysiology has quite distinct differences. AAA and TAA are best considered as different disease processes with inflammation as hallmark of AAA formation and distinct alterations in extracellular matrix formation perhaps being a common feature of TAA. However, we must be cautious about such sweeping generalizations as they may be driven by the experimental approaches used to study these diseases and in essence, be self-fulfilling prophesies because of these approaches.
The vast majority of the studies of AAA have been fueled by the emergence of several relatively simple models. While informative that angiotensin II can induce AAA in the appropriate settings, the extensive use of this model raises the question; Does the fact that angiotensin II is a potent pro-inflammatory stimulus inform us about the fundamental pathophysiology of AAA or does it constrain us to an exclusive focus on inflammatory mechanisms? Similarly, in the case of TAA, much of the mechanistic work has been driven by genetic models that are often derived from human disease where unique genetic mutations have been identified. While extraordinarily informative, this approach is also somewhat limited in terms of the generalizability to the greater population of individuals with sporadic TAA. Indeed, the relatively limited number of more “generic” animal models of TAA (i.e., those that are not based on a specific genetic mutation) may be a limiting factor in the study of TAA disease mechanisms. As discussed in other sections of this compendium, clinical trials in humans using pharmacological approaches sometimes fail to produce the predicted effects which is likely a consequence of the limitations of available animal models.
It is also important to consider the potential differences between aneurysm initiation, growth, and rupture. Much of the animal work has focused on the mechanisms responsible for the initiation of aneurysms and while these studies have provided extremely valuable insights into the potential pathogenesis of aortic aneurysms, they cannot help to define the entire disease process and in some instances, they can diverge from the normal progression of human disease. There is clear opportunity for additional studies to help us better understand the mechanisms of aortic rupture and dissection as they relate to both AAA and TAA.Human studies must also be placed into the correct context both in terms of disease state and anatomic location. As the majority of human samples are obtained at the time of surgery, they represent more advanced disease. For TAA, the heterogeneity of the disease is also an important consideration as while syndromic, non-syndromic and sporadic TAA share some common features, it would be naïve to think that they share a single, common cellular disease mechanism.
The past two decades have seen an enormous growth in our understanding of the pathology of both thoracic and abdominal aortic aneurysm formation. While the entirety of the field cannot be summarized in a single review, we have attempted to define some of the key themes that have emerged from this growing body of work. Much of the experimental work to date has focused on cellular and molecular events that occur early in the disease process leaving much to be learned about the more chronic events in aneurysm pathology. In addition, we do not yet fully understand how various risk factors like cigarette smoking, diabetes, and race impact these disease mechanisms. The mere fact that there are no specific, mechanistically based curative therapies for patients with aortic aneurysms tells us that we have much to learn about the cellular mechanisms of both thoracic and abdominal aortic aneurysms.
Acknowledgments
This work was supported by a grant from the National Institutes of Health (PO1 HL095070).
Non-standard Abbreviations and Acronyms
AAA | Abdominal Aortic Aneurysm |
Ang II | Angiotensin II |
eNOS | Endothelial Nitric Oxide Synthase |
FDG | Fluorodeoxyglucose |
IL | Interleukin |
IFN | Interferon |
LDLR | Low Density Lipoprotein Receptor |
MMP | Matrix Metalloproteinase |
PI3-kinase | Phosphatidylinositol-4,5-bisphosphate 3-kinase delta |
PET | Positron Emission Tomography |
ROS | Reactive Oxygen Species |
SMC | Smooth Muscle Cell |
TAA | Thoracic Abdominal Aortic Aneurysm |
TGF-β | Transforming Growth Factor Beta |
Footnotes
The authors have no conflicts of interest to disclose.
Citations
Full text links
Read article at publisher's site: https://doi.org/10.1161/circresaha.118.313187
Read article for free, from open access legal sources, via Unpaywall: https://www.ahajournals.org/doi/pdf/10.1161/CIRCRESAHA.118.313187
Citations & impact
Impact metrics
Citations of article over time
Alternative metrics
Discover the attention surrounding your research
https://www.altmetric.com/details/127302323
Article citations
The potential of a nomogram risk assessment model for the diagnosis of abdominal aortic aneurysm: a multicenter retrospective study.
Sci Rep, 14(1):21536, 15 Sep 2024
Cited by: 0 articles | PMID: 39278952 | PMCID: PMC11402964
The Predictive Role of Indices Indirectly Indicating Hyperlipidemia and Inflammation in Ascending Aortic Aneurysm.
Cureus, 16(9):e68531, 03 Sep 2024
Cited by: 0 articles | PMID: 39364508 | PMCID: PMC11448741
Vinpocetine alleviates the abdominal aortic aneurysm progression via VSMCs SIRT1-p21 signaling pathway.
Acta Pharmacol Sin, 23 Aug 2024
Cited by: 0 articles | PMID: 39179867
Structural reconstruction of mouse acute aortic dissection by intravenously administered human Muse cells without immunosuppression.
Commun Med (Lond), 4(1):174, 09 Sep 2024
Cited by: 0 articles | PMID: 39251746 | PMCID: PMC11384757
Tea polyphenol nanoparticles enable targeted siRNA delivery and multi-bioactive therapy for abdominal aortic aneurysms.
J Nanobiotechnology, 22(1):471, 08 Aug 2024
Cited by: 0 articles | PMID: 39118143 | PMCID: PMC11308685
Go to all (179) article citations
Similar Articles
To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.
Hyperplastic cellular remodeling of the media in ascending thoracic aortic aneurysms.
Circulation, 112(8):1098-1105, 01 Aug 2005
Cited by: 90 articles | PMID: 16116068
Smooth muscle cell deletion of low-density lipoprotein receptor-related protein 1 augments angiotensin II-induced superior mesenteric arterial and ascending aortic aneurysms.
Arterioscler Thromb Vasc Biol, 35(1):155-162, 13 Nov 2014
Cited by: 41 articles | PMID: 25395615 | PMCID: PMC4332619
Relevance of angiotensin II-induced aortic pathologies in mice to human aortic aneurysms.
Ann N Y Acad Sci, 1245:7-10, 01 Dec 2011
Cited by: 41 articles | PMID: 22211965
Review
Matricellular proteins: Potential biomarkers and mechanistic factors in aortic aneurysms.
J Mol Cell Cardiol, 169:41-56, 13 May 2022
Cited by: 5 articles | PMID: 35576729
Review