Elastin and collagen fibre microstructure of the human aorta in ageing and disease: a review

Abstract

Aortic disease is a significant cause of death in developed countries. The most common forms of aortic disease are aneurysm, dissection, atherosclerotic occlusion and ageing-induced stiffening. The microstructure of the aortic tissue has been studied with great interest, because alteration of the quantity and/or architecture of the connective fibres (elastin and collagen) within the aortic wall, which directly imparts elasticity and strength, can lead to the mechanical and functional changes associated with these conditions. This review article summarizes the state of the art with respect to characterization of connective fibre microstructure in the wall of the human aorta in ageing and disease, with emphasis on the ascending thoracic aorta and abdominal aorta where the most common forms of aortic disease tend to occur.

1. Introduction

Aortic disease is currently a large health concern because it is both common and can lead to fatal outcomes. These consist of a variety of conditions targeting the aorta, with the most common forms being aneurysm [1,2], dissection [3,4], occlusion owing to atherosclerosis [5,6] and a general stiffening of the normally elastic aorta that is thought to be a natural consequence of ageing ([7–11]; table 1). There are many co-morbid abnormalities that can lead to or be associated with one or more of these conditions, including hypertension [12–14], genetic mutations (such as Marfan syndrome (MFS) [15,16]), developmental defects (such as bicuspid aortic valve (BAV) [17–19]), connective tissue disorders (such as Ehler–Danlos disorder [20,21], scleroderma [22,23], osteogenesis imperfecta [24,25], polycystic kidney disease [26,27] and Turner syndrome [28,29]), as well as injury. The co-morbid abnormalities that will be discussed below are shown in table 1. All aortic diseases are associated with microstructural changes, either to the content or architecture of the connective fibres elastin or collagen.

Table 1.

Summary of aortic diseases and co-morbid conditions addressed in this review article. BAV, bicuspid aortic valve; MFS, Marfan syndrome; SVAS, supravalvular aortic stenosis; WBS, Williams–Beuren syndrome; AAE, annuloaortic ectasia; CMD, cystic medial degeneration; AVPS, aortic valve pure stenosis; AVR, aortic valve regurgitation; ATA, ascending thoracic aorta; AA, abdominal aorta; IRAA, infrarenal abdominal aorta.

disease	description	associations	common locations
aortic diseases
aneurysm	widening or balloon-like formation of an artery owing to local weakness of the wall	BAV, MFS, AVPS, AVR, atherosclerosis	ATA, AA, IRAA
secondary bleb in aneurysm	focal out-pouching within aneurysm		IRAA
dissection	tear in the inner wall of the aorta which causes blood to flow between wall layers and forces the layers apart	hypertension, MFS, atherosclerosis, CMD, medionecrosis, scarring, aortitis	ATA
atherosclerosis	a condition in which the arterial wall thickens owing to development of a fatty plaque	aneurysm, dissection	ATA, AA, IRAA
stiffening	progressive increase in wall stiffness	hypertension, ageing	ATA
co-morbid abnormalities
pressure-related
hypertension	chronic medical condition in which the arterial blood pressure is elevated causing stiffening in arteries	dissection	ATA
genetic defects
MFS	autosomal dominant genetic disorder causing an alteration in the protein fibrillin-1 which is an important structural component of the aortic wall	aneurysm, AAE, AVR, dissection	ATA
WBS	people with WBS are missing genetic material from chromosome 7, including the gene elastin. Because they lack elastin, individuals with WBS have circulatory system disorders and heart defects. This lack of elastin may cause stenosis in small and large vessels	SVAS	ATA
developmental defects
BAV disease	a congenital heart abnormality. It is a defect of the aortic valve that results in the formation of two leaflets instead of the normal three	aneurysm, AVPS, AVR	ATA
SVAS	aortic defect that develops before birth. It is a narrowing of the aorta just above the aortic valve	WBS	ATA
fibrotic and necrotic
CMD	characterized by elastin fragmentation and loss of smooth muscle cells in the aortic medial layer	dissection, medionecrosis	ATA
medionecrosis	pools of necrotic tissue within the aortic media. May predispose the formation of aneurysm	CMD, dissection	ATA
aortic scarring	the development of dense fibrous tissue which is usually accompanied by calcification in aortic wall	dissection	ATA
inflammatory
aortitis	inflammation of the aorta	dissection	ATA
combined effects
AAE	dilatation or enlargement of ATA and aortic annulus	MFS	ATA
AVPS	disease of the aortic valve characterized by narrowing of the valve opening	aneurysm, BAV	ATA
AVR	the aortic valve does not close completely. As a result, some blood leaks back (regurgitates) through the aortic valve into the left ventricle	aneurysm, MFS, BAV	ATA

The aorta, the blood vessel responsible for delivering blood from the heart to the systemic circulation, normally possesses a high degree of elasticity, which aids in the propulsion of blood downstream to the systemic vasculature [30], and a microstructure that supports this function ([31,32]; figure 1). It is the connective fibres within this microstructure, elastin and collagen, which impart the elastic properties and strength of the aorta, respectively. Often, it is alteration of the quantity and/or architecture of these fibres that leads to the mechanical, and hence functional, changes associated with aortic disease [9,14,34–42]. For example, structural alterations in the walls of large arteries with progressing age causes a decrease in the total arterial compliance [9,10,43–46], which in turn leads to both a decreased distal blood flow and an increase in aortic pulse pressure [30]. This increased pulse pressure has been shown to be the strongest predictor of cardiovascular mortality, because it increases the mechanical load on the left ventricle [47].

Figure 1.

Idealized architecture of a healthy human artery. Arteries possess a three layer structure consisting of an inner layer (intima (I)), middle layer (media (M)) and outer layer (adventitia (A)). The intima is composed mainly of a single layer of endothelial cells, a thin basal membrane and a subendothelial layer of collagen fibrils. The media is composed of smooth muscle cells, a network of elastic and collagen fibrils, and elastic laminae which separate M into a number of transversely isotropic fibre-reinforced units. The adventitia is the outermost layer surrounded by loose connective tissue. The primary constituents of the adventitia are thick bundles of collagen fibrils arranged in helical structures. (Adapted from Gasser et al. [33].)

The aorta also possesses a non-uniform structure displaying distinct regions or segments (figure 2a) that are more susceptible to certain types of disease than others [48,49]. It stands to reason that each segment can undergo different types and degrees of remodelling during ageing and disease, and structurally responds distinctly to the various loading conditions seen throughout the length of the aorta. For instance, it has been suggested that the aorta ages ‘from the bottom up’, i.e. biomechanical changes are manifested at an earlier age in the infrarenal abdominal aorta (IRAA) than in the descending thoracic aorta (DTA) [50]. It would be prudent to believe that this is a direct result of variable age-related architectural changes of the aortic wall, which occur earlier in the distal segments of the aorta than in the proximal ones.

Figure 2.

(a) Illustration of the different regions of the human aorta. (i) ascending thoracic, (ii) descending thoracic, (iii) supraceliac abdominal, (iv) suprarenal abdominal and (v) mid-infrarenal abdominal aorta. (Adapted from Halloran et al. [31].) (b) Schematic of ascending thoracic aorta and arch. R, root; P, proximal ascending thoracic aorta; D, distal ascending thoracic aorta; and A, arch. (Online version in colour.)

The purpose of this review article is to provide an overview of the literature associated with the characterization and quantification of elastin and collagen, and changes in their fibrous architecture within the wall of the human aorta during ageing and disease. We first present various reports that describe generally the connective fibre content and architecture in unspecified aortic locations as a function of ageing and disease. Then, the state of the art with respect to variations in connective fibre content and architecture in the ascending thoracic aorta (ATA) and abdominal aorta (AA), locations in which the most common forms of aortic disease tend to occur, in ageing and disease is presented. The pertinent information presented in these sections is summarized in tables 2–5, which allow the reader to identify trends regarding fibre content, concentration and architecture in different sections of the aorta as a function of age and disease, as well as note inconsistencies and gaps in the current knowledge.

Table 2.

Summary of studies published to date regarding elastin content, concentration and architecture in human ascending thoracic aorta (ATA): effects of age, disease and location. Tilde symbol (∼), no change; up arrow (↑), increase; down arrow (↓), decrease; SVAS, supravalvular aortic stenosis; WBS, Williams–Beuren syndrome; CMD, cystic medial degeneration; AAE, annuloaortic ectasia; BAV, bicuspid aortic valve; w, with; wrt, with respect to.

additional co-morbidity	elastin content	elastin concentration	elastin cross-linking	no. elastin lamellar units	elastin microstructure	references
ascending thoracic aorta: elastin
age
	∼	↓				[51–54]
aneurysm
	overall: ↓ layer-specific: ↑ in adventitia, ↓ in media, intima location-specific: lowest in the CIRC direction in the media of right lateral region				overall: fragmentation, disrupted, irregular arrangement, location-specific: orthotropic alignment in CIRC–LONG directions in media of anterior, left lateral, and posterior regions. aligned with LONG axis in inner media of right lateral region	[55,56]
aneurysm convexity w BAV	overall: ↑ wrt concavity				fragmentation, shorter fibres wrt concavity	[57]
MFS	overall: ↓				anisotropy ↓	[58]
dissection
	overall: ↓ location-specific: ∼ CIRC regions age-specific: ∼	overall: ↓ location-specific: ∼ CIRC regions	overall: ↓		overall: fragmentation, disrupted, irregular arrangement, decrease in interlaminar fibres	[49,54,56,59–61]
hypertension	overall: ↓				overall: decrease in interlaminar fibres	[60,61]
MFS	overall: ↓				overall: cystic and laminar medionecrosis	[60]
atherosclerosis					overall: mild or no fragmentation	[59]
CMD + medionecrosis					overall: fragmentation	[59,62]
total scarring					overall: only remnants of medial elastin fibres	[59]
co-morbidities	elastin content	elastin concentration	elastin cross-linking	no. elastin lamellar units	elastin microstructure	references
other co-morbidities
hypertension	layer-specific: ↑ in inner layers, ↓ in middle and outer layers					[63]
BAV	overall: ∼					[64]
SVAS		overall: ∼		overall: ↑	overall: thin, fragmented	[65]
SVAS + WBS		overall: ↓		overall: ↑	overall: thin, dispersed, fragmented	[65]
AAE		overall: ↓			overall: fragmentation, accellularity	[52]
AAE + MFS		overall: varied drastically (total lack of to normal)				[52]

Table 5.

Summary of studies published to date regarding collagen content, concentration and architecture in human abdominal aorta (AA): effects of age and disease. Tilde symbol (∼), no change; up arrow (↑), increase; down arrow (↓), decrease; IRAA, infrarenal abdominal aorta; AAA, abdominal aortic aneurysm; ATA, ascending thoracic aorta; SRAA, suprarenal abdominal aorta; s-PIIINP, serum carboxyterminal propeptide of type III collagen; wrt, with respect to; ctrl, control.

additional co-morbidity	collagen content	collagen concentration	collagen cross-linking	collagen undulation	collagen microstructure	references
abdominal aorta: collagen
location: IRAA
IRAA	location-specific: ↓ wrt ATA, ∼ among SCAA, SRAA and mid-IRAA, ↓ less than elastin between SRAA and IRAA, ∼ type I wrt SRAA					[31,71]
aneurysm: IRAA
	overall: ∼ /↑ layer-specific: ↑ type III in media and adventitia, ↑ procollagen I in the intima, and ↑ procollagen III in the media	overall: ∼ /↑, ∼ type I, ∼ type III	overall: ↑	overall: ↓	overall: loss of normal architecture and distinction between adventitial, medial and intimal collagen, parallel sheets, not coherent network, disordered, no layered directional organization location-specific: widely dispersed fibre orientation in anterior region	[31,48,72,74–76,81]
secondary bleb	layer-specific: media consisted almost entirely of collagen fibres					[77]
atherosclerosis
atherosclerosis	overall: ↓ layer-specific: mature collagen abundant in media and periphery of plaque. New collagen mostly in fibrous cap. sex-specific: higher in females than males		overall: ↑			[73]
age and location: AA
	age-specific: ↑ overall, ↓ type III location-specific: ↓ type III wrt ATA	overall: ↑			overall: 2 distinct counter-rotating fibre families found in intima, media and adventitia. Fibres organized in separate layers. Almost no interlaminar fibres.	[70,82,83]
aneurysm: AA
	overall: ↑ s-PIIINP				overall: degraded type III	[79,84]
atherosclerosis	overall: ↑, type I the most abundant of I, III and V in both AAA and ctrl	overall: ∼ types I, III and V				[80]
atherosclerosis: AA
early atherosclerosis	overall: ↑					[82]

It is important to note that both the terms content and concentration are used to quantitate the results below but these terms do not always imply the same thing. If a change in content is seen, the actual amount of the fibre of interest is altered, but if a change in concentration is seen, it could mean that the amount of the fibre of interest is altered or the amount of surrounding components (collagen, elastin and sugars) are altered while the fibre of interest is maintained constant.

2. Human aorta: trends in unspecified locations

In many literature reports, the elastin and collagen content and architecture were reported for unspecified aortic locations. As a result they cannot necessarily be considered representative of how the connective fibres vary in specific aortic sections. While this information has some value, and is summarized in this section, having knowledge of regional differences in fibre orientation is of critical importance because certain portions of the aorta are more susceptible to certain types of disease than others. For instance, as will be described below, the ATA is susceptible to aneurysm and dissection, potentially owing to haemodynamic effects [55], whereas the AA presents with an aneurysm. The sections that follow this one will focus on literature reports that specifically characterize and quantify the variation in amount and architecture of elastin and collagen specifically in the ATA and AA in ageing and disease.

2.1. Elastin

2.1.1. Effects of age

One of the earliest reports on the effect of age on elastin showed that its concentration within the aorta decreased with age, but the elastin content remained unchanged [85]. This trend was confirmed in various subsequent studies [86,87], which also suggested that the decrease in elastin concentration was in part due to increases of other components, such as collagen [88–90], while maintaining total elastin content [91].

It has also been shown that age could have altering effects on the amino acid structure of elastin. John & Thomas [92] found that the content of the elastin cross-linking amino acids desmosine and isodesmosine remained unchanged at 3/1000 amino acid residues, and started to decrease after the age of 63 reaching a value of 2/1000 amino acid residues in subjects aged 86 years. In another study, the contents of these amino acids decreased slightly with age throughout adult life and the content of histidinoalanine, which is considered to be one of the senescent elastin and collagen cross-linking residues, increased markedly with age during adult life [93]. Histidinoalanine amino acids can form cross-links between neighbouring acidic proteins and elastin, between acidic proteins and collagen, and between the acidic protein molecules themselves, which might be responsible for the ageing-associated changes seen in the human aorta such as loss in elasticity.

It has also been reported that the remodelling of the human thoracic aorta that occurs with ageing was associated with fragmentation of elastin fibres [94]. This fragmentation has also been reported in old, highly calcified thoracic aortas accompanied by atherosclerotic lesions [95]. It should be noted, however, that calcification of the aorta does not always require atherosclerosis and occurs in a process known as medial elastocalcinosis [96]. While the exact mechanism for which these elastin fibres are fragmented with age is not known, it could be a consequence of mechanical fatigue failure caused by the pulsatile wall strain experienced by the aorta over the number of cardiac cycles experienced in a lifetime [11,88,97]. It might also be due to chemical degradation caused by the upregulation of matrix metalloproteases (MMPs) with age owing to an imbalance with their inhibitors [98]. Additionally, there is evidence that the increased expression of MMPs could accompany the aforementioned calcification of elastin [99].

2.1.2. Effects of disease

In aortas from patients with MFS, elastin was deficient in desmosine cross-linking residues and the content of elastin was decreased by almost 50 per cent [100]. This was further validated in a report also showing elastin content was decreased by 50 per cent, but specifically in the media of MFS aorta with structural alterations of elastin fibres being characterized by enlarged interlaminar spaces (between elastin laminae) and loss of interlaminar elastin fibrils [101]. This loss of elastin content and decrease in cross-linking could explain the higher prevalence of MFS patients to aneurysm, because the degradation of elastin could cause a release of significant compressive pre-stresses within the wall and subsequently lead to diameter enlargement [102].

Elastin has also been analysed in the media of dissecting aortic aneurysms with one study showing its content remained unchanged as compared with normal [101], but a later one showing contrary results with the elastin concentration decreasing at site of dissection in dissected aneurysms of the thoracic aorta [103].

2.2. Collagen

2.2.1. Effects of age

An increase in the content of collagen within the aortic wall with increased age has been well characterized [88–90]. One study showed collagen concentration increasing from 20 to 30.5 per cent [86], and another showed an increase from 19.9 to 25.5 per cent [104]. This trend was corroborated by a study showing the content of hydroxyproline, which is proportional to the content of collagen, to be increased with age [105]. However, it might be possible that this increase in collagen content was layer-specific, as it remained unchanged with age in the intimal layer in another study [106]. This increase in collagen content did not occur linearly with age for the thoracic aorta, as it was approximately constant for ages younger than 45 and increased slowly thereafter [87]. The structure of the collagen fibres themselves was also changed with advanced age, showing an increase in irregularly arranged fibres in the media of the human thoracic aorta [94].

The amount of cross-links between collagen fibres has also been found to increase with age because of two different mechanisms. The first is due to a marked increase in the content of cross-linking amino acids histidinoalanine [93] and pentosidine [107] in collagen itself. The second is due to a marked increase in the accumulation of advanced glycation end-products with age, which are produced by the glycation and oxidizing reactions between sugars and the amino groups in protein molecules, and form bridges between collagen fibres [108].

2.2.2. Effects of disease

The collagen content in aortas of MFS, fusiform aortic aneurysm and dissecting aortic aneurysm patients was similar to the normal aorta, and exhibited no structural alterations either [100,101,109]. However, while this trend was consistent for the thoracic aorta as a whole in dissecting aneurysms, specifically at sites of dissection, collagen content was increased and collagen concentration was decreased [103].

3. Human ascending thoracic aorta

3.1. Elastin

3.1.1. Effects of age

The elastin concentration (% per mg of sample, dry weight) in the wall of ATA was reported to gradually decrease by 36 per cent in subjects aged from newborn to 81 years old [110]. A later study evaluated a broader population and also showed a decrease with age by 33 per cent between the second and ninth decade of life [51]. In another study, elastin concentration was evaluated specifically in intimal–medial samples of ATA wall, and it was found that the elastin concentration was highest in the ATA wall of children and decreased with age [52]. Despite the previous studies showing a decrease in concentration, the elastin content in the ATA wall was found to remain unchanged with age [53].

3.1.2. Effects of disease

In dissected ATAs, the elastin content is generally decreased compared with control (non-dissected) ATAs. One study showed that elastin content was slightly decreased in dissected ATA, whereas the concentration was significantly decreased compared with control ATA [49]. There was also a slight decrease in the concentration of elastin cross-links (amount of desmosine and isodesmosine) compared with control. In another study, the elastin content in the wall of proximal (ascending or transverse arch) thoracic aortic dissection (figure 2b) was not different between acute (less than 14 days) and chronic (greater than 14 days) dissections, but it was decreased compared with control [54]. This decreased amount of elastin correlated strongly with decreased expression of fibulin-5, which is known to be involved in elastogenesis [111]. While one potential mechanism for aortic dissection can be the elastolysis by elastase and MMPs [112–119], the decreased expression of fibulin-5 may suggest that the loss of elastin turnover also has an effect on the presence of ATA dissection. With respect to microstructure, severe medionecrosis, a decreased number of smooth muscle cells (SMCs), and severe fragmentation of elastin lamellae was seen in the media, which correlated strongly with decreased expression of fibulin-5. It was suggested that the decreased expression of fibulin-5 might impair the assembly of mature elastin in the ATA tissue, which could render the wall susceptible to dissection. This trend of decreased number of SMCs in the aortic wall is also common in ageing, hypertension and in presence of atheromatous plaque [88–90,94].

Disrupted and irregular elastin lamellae have been observed in the medial layer of dissecting ATAs with and without accompanying aneurysm, with some areas either being devoid of an elastin fibre framework, or having localized elastin fragmentation [56,59]. This was opposed to that seen in the control ATA, where the layers of SMCs were separated by prominent elastin lamellae, which were interconnected by a network of small elastin fibres and collagen fibres. Fragmented elastin fibres were also often found in the basement membrane in cases of ATA dissection in aortas accompanied by cystic medial degeneration (CMD) and medionecrosis [59,62]. One theory as to why this fragmentation happens is that the elastin fibres could be the target for enzymes such as elastases in the aortic wall, where increased levels have been shown in cases of dissecting aortas [120] as well as aneurysms [121]. This process of enzymatic destruction of elastin might also be highly selective to specific portions of elastin within the wall, as substantial features of medial elastin architecture could be preserved in abdominal aortic aneurysm (AAA; [122]). While fragmentation of elastin is often found in ATA dissection and also when accompanied by aneurysm, this may not be the case when accompanied by atherosclerosis [59]. Samples with fatty streaks or fibrotic plaques maintained an intact media with only mild or no elastin fragmentation seen. It was only in severely atherosclerotic samples that fragmentation could be seen. Only remnants of medial elastin fibres were found in ATA dissection when total scarring occurred in samples with advanced atherosclerotic lesions.

In cases of ATA dissecting aneurysms (DeBakey's type III, figure 3, type I–III), the presence of MFS was reported in 11 out of 111 cases, and a high degree hypertension was present in 95 out of 111 cases, 70 of which had no other obvious risk factors, suggesting that hypertension might have an effect on the presence of dissection in ATA aneurysm [60]. In these cases of ATA dissections, men outnumbered women three to two, and the age range was generally between 60 and 80 years old, though patients with MFS were typically younger (under 40). In ATA dissections, the microstructure was characterized by a loss of elastin content and a decrease in interlaminar elastin fibres [60]. Furthermore, these fibres were fragmented and irregularly arranged [60], which, as previously mentioned, is also common in ATA dissections with and without accompanying aneurysm [54,56,59], in ATA dissections accompanied by CMD and medionecrosis [59,62], or with severe atherosclerosis [59]. In the control tissue, the structure of elastin was continuous and formed elastin laminae, which were interconnected through interlaminar fibres. We hypothesize that the interlaminar elastin fibres and particularly those elastin fibres that run radially could contribute to the bonding forces that hold the aortic wall layers together under applied haemodynamic loading conditions [124]. Therefore, in the degenerated ATA aneurysm wall, the haemodynamic loads may exceed the bonding forces holding the mural layers together leaving the weakened aorta prone to spontaneous tears [125]. Also, the occurrence of MFS appeared to be related to the presence of cystic medionecrosis and laminar medionecrosis [60]. The latter was considered to be a secondary ischaemic change following dissection [60]. The results showing a loss and fragmentation of elastin and a decrease in interlaminar elastin fibres in ATA dissecting aneurysms of MFS or hypertensive cases were similarly shown for ATA dissecting aneurysms of type A (figure 3, type A,B), hypertensive ATAs and control ATAs [61]. Also it was suggested that hypertension could be related to a decrease of interlaminar fibres of medial elastin in ATA dissections. The latter is consistent with the results of a recent study on hypertensive ATAs using multiphoton microscopy [63], whereby it was reported that elastin content was increased in the inner wall layers, and slightly decreased in the middle and outer wall layers, compared with normotensive ATA.

Figure 3.

DeBakey classification (types I, II, III) and Stanford classification (types A, B) of aortic dissection. Type I refers to dissections that propagate from the ascending aorta, extend to the aortic arch, and commonly, beyond the arch distally. Type II refers to dissections that are confined to the ascending portion of the aorta. Type III dissections are limited to the descending aorta. Type III also includes the dissections that start in the descending aorta that extend proximally to the arch and ascending aorta. Stanford type A includes dissections that involve the ascending aorta, arch and descending thoracic aorta. Stanford type B refers to dissections that involve only the descending aorta. Rarely, dissections that start in the descending aorta can also extend proximally into the aortic arch and the ascending aorta. This is a special case of type B dissection, and should be called retro-A dissection. (Adapted from Tran & Khoynezhad [123].)

The elastin concentration was studied in the ATA wall from patients with annuloaortic ectasia (AAE), some of which also possessed MFS [52]. In some of the ATAs, the concentration of elastin was less than control, but in others there was less marked difference. Also, the specimens in which elastin concentration was decreased displayed severe elastin fragmentation and acellularity. In the ATA of AAE + MFS patients, elastin concentration varied drastically, from a total absence to normal values. It is important to note that, although the results of this study add to our knowledge base regarding the effects of AAE and MFS on the architecture of elastin in ATA, further studies will need to be performed to better characterize the trends in these cases.

The media layer in specimens of MFS ATA aneurysms exhibited a profound decrease in the amount of elastin and reduced anisotropy (directionality), as compared with control where the elastin fibres were well-formed and arranged in a lamellar pattern [58]. The elastin content in the media layer of ATA with BAV and in control ATA (those with tricuspid aortic valve (TAV)) has also been quantified [64], and it was shown to be unchanged.

Elastin content, elastin cross-linking and the number of elastin lamellar units were analysed in the media layer of the walls of a small number (n = 2) healthy control ATA, proximal ATA (figure 2b) with supravalvular aortic stenosis (SVAS), and ATA with SVAS and Williams–Beuren syndrome (WBS) [65]. The elastin content was greater in SVAS and control ATA compared with SVAS + WBS ATA. The elastin lamellae were thin, numerous (approx. 120 units), and fragmented in the wall of ATA with SVAS as compared with control (approx. 43 units). In the wall of ATA with SVAS + WBS, the elastin lamellae (approx. 120) also appeared to be highly fragmented and thinner than control. The stenotic region contained thin and dispersed elastin components, and the elastin material was observed as small puddles.

3.1.3. Effects of location

The elastin content and concentration was analysed in three circumferential regions (figure 4, sample sites S1, S2, S3) in dissected ATA and in the healthy control ATA [49], and was found to not vary significantly with respect to circumferential location in either the dissected ATA or control ATA.

Figure 4.

Diagrammatic representation of the sample sites in the thoracic aorta. (a) Band position with respect to anatomical features: B1, B2, B3 and B4, bands 1–4. (b) Transverse section of the vessel wall indicating orientation of the sample sites with respect to each other; arrow denotes blood flow into the page. S1 samples were taken from the outer curvature of the aortic wall. (c) Relative sample positions illustrating sites within bands. Design allows for longitudinal and transverse comparisons to be made between sites. (Adapted from Cattell et al. [49].)

The amount of elastin and its architecture was evaluated in four circumferential regions (anterior, posterior, left lateral and right lateral) and also in the three mural layers (intima, media and adventitia) in ATA aneurysms and in control ATA. It was shown that the amount of elastin was lower in all regions of aneurysmal samples compared with control [55], and higher in the adventitia, but lower in the intima and media layers compared with control. The alignment of elastin fibres revealed fibres running in both the CIRC and LONG directions in the media of anterior, left lateral and posterior regions. However, in the right lateral region, the elastin fibres of the media were aligned with the LONG axis in the inner layers, but primarily oriented in a CIRC direction in the outer layers. It was suggested that this deviation might reflect regional heterogeneity in the biomechanical properties of ATA aneurysm. Also, the orientation-dependent differences might associate with CIRC tears of the inner wall that have been observed in dissection [126]. Supporting this hypothesis, the right lateral region where these structural differences occurred is the most frequent site of dissection [126].

The elastin integrity was investigated in the media of proximal ATA aneurysm (figure 2b) with BAV from 27 patients, 12 of which had aortic valve pure stenosis (AVPS) and the rest had isolated aortic valve regurgitation (AVR; [57]). The results were compared with healthy control ATAs from six normotensive heart donors with TAV. A more severe elastin fragmentation with a higher number of shorter elastin fibres was observed in the media of the convexity versus concavity of proximally dilated ATA, in both the regurgitant and stenotic BAV groups (figure 4). No differences in elastin integrity between convexity and concavity were reported for the control ATA.

3.1.4. Summary table

Table 2 summarizes the current knowledge of elastin amount and architecture in the wall of human ATA in ageing and disease. The empty positions in the table indicate the information that is missing from the literature.

3.2. Collagen

3.2.1. Effects of age

Studies examining collagen in the human aorta across various ages showed a significant positive correlation for collagen content with age [53] and a significant increase in the number of hydroxyproline residues (indicative of amount of collagen) after age 50 [52]. However, a separate report found only a slight, non-statistically significant increase in collagen content with age [51].

3.2.2. Effects of disease

In control ATA and in histologically normal samples of ATA dissection (in absence of CMD, medionecrosis, atherosclerosis and inflammation), type IV collagen [127–129] was observed as longitudinal sheets between SMCs in the subintimal basement membrane and in the media [59]. Type IV collagen was also seen in the basement membrane of the adventitial layer and in the vasa vasorum. In mild or moderate atherosclerosis, type IV collagen was observed in these longitudinal sheets similar to control but in advanced atherosclerosis with medial scarring, type IV collagen was missing above the intimal plaque (ablumenal side), but not elsewhere [59]. In the case of ATA dissection with CMD or medionecrosis, there often were areas missing collagen, leading to a discontinuous and irregular structure, and these areas were smaller in the media than in the basement membrane. Type IV collagen was more continuously seen in samples with mild or moderate medionecrosis compared with severe medionecrosis, and was normally seen in the subendothelial basement membrane of the vasa vasorum [59]. In the case of ATA dissection with aortitis, in areas with inflammatory cells, type IV collagen was irregular and fragmented but this was a small, local defect [59]. In the case of ATA dissection with total scarring of the aortic wall, type IV collagen was not expressed around the medial cells, but in the subendothelial basement membrane [59].

The staining of interlaminar collagens types I (mature) and III (newly synthesized) was more intense in cases of ATA dissection with CMD, medionecrosis and atherosclerosis, than in control ATA. In dissected samples with CMD and medionecrosis, collagen bundles (types I and III) were not as regular, but were characterized by thick longitudinal sheets or bundles in the media, which were generally thicker than bundles of collagen IV [59]. This increase in collagen was also confirmed by a study showing that the total collagen content within the wall of dissected ATA tissue was increased compared with control [49]. In atherosclerosis, collagen types I and III were intensely expressed in the intimal plaque and in severe cases of atherosclerosis, collagen types I and III were distorted in the media. Also, in severely atherosclerotic samples with ATA dissection, the demarcation between the intima and media was not clear owing to fragmentation of the internal elastic lamina, and fibrosis (excessive collagen formation [130]) continued into the medial layer [59]. In aortitis, there was normal expression of collagens I and III in the media, and in total scarring of the aortic wall, collagen types I and III were intensely expressed throughout the media [59].

With regards to cross-linking in ATA dissection, it was reported that there were no significant differences in the amount or concentration of pyridinoline cross-links compared with control ATA [67].

Collagen content and fibrosis were histologically examined in the wall proximal to the thoracic aortic dissection and in control ATA [54]. It was reported that in this area the collagen content increased in the medial layer of dissected samples compared with control. Also increased fibrosis was seen in the wall of dissected tissue.

It has also been observed that the collagen proportion (% cross-sectional area) in the media layer specifically of the ATA varied with disease conditions [56]. In ATA dissection, the collagen proportion was decreased from 33 ± 12% in the inner half of the media to 19 ± 12% in the outer half (p < 0.01). In the wall of ATA aneurysm, collagen proportion did not differ significantly (p = 0.71) between inner (20 ± 10%) and outer halves (18 ± 12%). In control ATA, the collagen proportion was decreased from 50 ± 13% in the inner half of the media to 40 ± 8% in the outer half (p = 0.04). The collagen proportion in the wall of ATA dissection and in the wall of ATA aneurysm was less than control (p < 0.01). The homogeneous proportion of collagen seen in the media layer of ATA aneurysm could be associated with an overall weakening of the wall that would lead to aneurysmal dilation. In the media layer of ATA aneurysm and dissection, dramatic morphological changes in collagen bundles were observed with collagen fibres being thin and having more scattered fibres. In contrast, in the control ATA, thick collagen fibres and bundles were observed with a parallel arrangement in the media as well as a few thin collagen fibres disposed perpendicularly.

In the ATA of AAE patients, some samples contained more collagen and a marked fibrosis than control but others had a less marked difference [52]. In the ATA of AAE + MFS patients, collagen concentration was increased in 3/6 samples. Changes were much smaller in the other three samples.

In MFS patients with ATA aneurysm, the collagen concentration was unaltered compared with control ATA, and the collagen content also displayed a similar trend [58]. However, microstructural alterations were seen with an increase in the collagen cross-linking (number of hydroxylysyl pyridinoline and lysyl pyridinoline cross-links per triple helix), and the structure of collagen in the medial layer was more disorganized compared with control ATA [48]. Also, by using three-dimensional confocal imaging, a decrease was seen in the collagen fibre waviness, and defects were detected in the collagen microstructure (figure 5c,d), compared with control aorta (figure 5a,b) in MFS ATA aneurysms. Atomic force microscopy was also used to conclude that these collagen fibres did not behave as a coherent network as they do in control tissue. Additionally, there was a complete absence of the normal collagen fibril organization, with only deposition of thin parallel collagen fibrils in the adventitia layer of MFS ATA aneurysms, as opposed to the adventitia layer of control ATA tissue, where the fibres are loosely knit, highly organized ribbon-like collagen bands.

Figure 5.

(a,b) Three-dimensional reconstructions (confocal microscopy) of collagen networks in the normal media and adventitial layer. (c,d) Aneurysms in patients with MFS. (e) Aortic abdominal aneurysm. (Adapted from Lindeman et al. [48].)

The collagen content in the media layer of ATA with BAV was approximately the same compared with control (TAV) ATA [64]. Given that patients with BAV are prone to aneurysm or dissection in the ATA [125], one might expect that more localized, as-yet unknown perturbations in the connective fibre architecture of ATA might be implicated in this.

In the medial layer of proximal ATA aneurysms (figure 2b) with congenital BAV from patients with either AVPS or isolated AVR [57], collagen types I and III decreased in concentration with a stronger decrease in the regurgitant (AVR) BAV group, and collagen type IV increased with a stronger increase in the stenotic (AVPS) BAV group. Also by comparing dilated ATAs with AVR with either BAV, MFS or controls, the MFS group had a decrease in type I collagen and an increase in type III collagen [66].

3.2.3. Effects of location

In both dissected aneurysm and control tissue, the number of pyridinoline cross-links per molecule of collagen in the ATA and in the arch approached the theoretical maximum for lysyl derivatives [67]. It was suggested that pyridinoline represents the major stabilizing cross-link of collagen in these regions of the aorta. On the other hand, in both dissected aneurysm and control tissue, the number of pyridinoline cross-links per molecule of collagen decreased by approximately 10-fold from the arch to the proximal DTA.

In cases of proximal ATA aneurysm (figure 2b) with congenital BAV from patients with either AVPS or isolated AVR, collagen types I and III were decreased, and collagen type IV was increased in the medial layer compared with healthy control ATAs as a whole. No significant differences in the contents of collagen types I, III and IV were found between the region of convexity and concavity of the aortic arch (figure 4) in the aneurysmal and control aortic wall [57]. However, another group showed that type I and III collagens were more markedly reduced in the convexity than in the concavity of the media layer of BAV-dilated ATAs with AVR [66]. In the MFS-dilated ATAs with AVR, the contents of collagen types I and III were not significantly different in the convexity and concavity of ATA.

The collagen content and architecture was analysed in four circumferential regions (anterior, posterior, left lateral and right lateral) of TAV ATA aneurysm and control ATA, and found to be unchanged compared with control [55]. However, the amount of collagen was decreased in the intima and media layers of ATA aneurysm, and increased in the adventitia compared with control. In the right lateral region of ATAA wall, the subintimal collagen fibres formed bundles that had a LONG orientation. In the media of anterior, posterior and left lateral regions, the collagen bundles had a CIRC arrangement. In the adventitia layer, collagen bundles were aligned with both axes. Similar to elastin, it was suggested that this deviation in the orientation of collagen in the right lateral region of ATAA wall might reflect regional heterogeneity in the biomechanical properties of ATAA associated with CIRC tears of the inner wall that have been observed in dissection [126].

3.2.4. Summary table

Table 3 summarizes the current knowledge regarding collagen amount and architecture in the wall of human ATA in ageing and disease. The empty positions in the table indicate the information that is missing from the literature.

Table 3.

Summary of studies published to date regarding collagen content, concentration and architecture in human ascending thoracic aorta (ATA): effects of age, disease, and location. Tilde symbol (∼), no change; up arrow (↑), increase; down arrow (↓), decrease; CMD, cystic medial degeneration; AAE, annuloaortic ectasia; BAV, bicuspid aortic valve; AVPS, aortic valve pure stenosis; AVR, aortic valve regurgitation; ctrl, control.

additional co-morbidity	collagen content	collagen concentration	collagen cross-linking	collagen undulation	collagen microstructure	references
ascending thoracic aorta: collagen
age
	overall: ∼/↑					[52–54,59]
aneurysm
	overall: ∼ layer-specific: ↓ in intima and media, ↑ in adventitia location-specific: lowest in CIRC direction in media of right lateral region	overall: ↓ location-specific: ∼ between inner and outer half			overall: thin scattered fibres layer- and location-specific: bundles having LONG orientation subintimally in right lateral region, CIRC orientation in media of anterior, posterior, and left lateral regions, aligned with both axes in adventitia	[55,56]
MFS	overall: ∼	overall: ∼	overall: ↑	overall: ↓	overall: disorganized, not coherent network layer-specific: thin parallel fibrils in the adventitia	[48,58]
MFS + AVR	overall: ↑ type III, ↓ type I location-specific:∼between convexity and concavity					[66]
BAV + AVPS	overall: ↓ type I and III, ↑ type IV location-specific: types I, III and IV∼between convexity and concavity	overall: (especially in AVPS) ↑ type IV, ↓ types I and III, location-specific: types I, III and IV ∼ between convexity and concavity				[57]
BAV + AVR	overall and location-specific: ↑ type IV, ↓ type I and III, more in convexity than concavity	overall: ↑ type IV, ↓ (especially in AVR) types I and III				[57,66]
dissection
	overall: ↑	overall: ↓ location-specific: ↓ from inner to outer half	overall: ∼ location-specific: max in ATA and arch in both aneurysm and ctrl, ↓ from arch to DTA		overall: thin scattered fibres, fibrosis	[49,54,56,59,67]
mild/moderate atherosclerosis	overall: ↑ type I and III				overall: increased interlaminar types I and III, defected type IV, longitudinal sheets layer-specific: types I and III intensely expressed in intimal plaque	[59]
advanced atherosclerosis with medial scarring	overall: ↑ type I and III				overall: fibrosis layer-specific: types I and III were distorted in the media, type IV missing above intimal plaque, not clear demarcation between intima and media	[59]
CMD + medionecrosis	overall: ↑ type I and III	overall: ↑ type I and III			overall: types I and III longitudinal bundles thicker than longitudinal sheets of type IV, interlaminar types I and III, areas missing collagen, discontinuous irregular structure	[59]
total scarring	overall: ↑ type I and III				layer-specific: type IV collagen not expressed around medial cells, but at subendothelial basement membrane	[59]
aortitis	overall: ∼ type I and III				overall: type IV irregular and fragmented	[59]
co-morbidities	collagen content	collagen concentration	collagen cross-linking	collagen undulation	collagen microstructure	references
other co-morbidities
BAV	overall: ∼					[64]
AAE	overall: ∼/↑				overall: fibrosis	[52]
AAE + MFS		overall: ↑				[52]

4. Human abdominal aorta

4.1. Elastin

4.1.1. Effects of age

Evaluating the amount of insoluble elastin and elastin cross-links in the IRAA wall from elderly patients with and without aneurysm, the concentration of elastin decreased with age [68]. However, with the limited age group in this study, one cannot feasibly conclude that the concentration of elastin decreased with age as a general trend.

4.1.2. Effects of disease

It was found that elastin concentration was 8.1 ± 3.2% dry defatted weight in the media of the human AAA, as opposed to 35.0 ± 3.2% dry weight in the media of the control AA [78]. The elastin fibres were observed to be disrupted in the media of AAA and the elastin cross-linking (ratio desmosines/isodesmosines) was higher in the media of the AAA compared with the media of the control AA. However, the authors cautioned that there was difficulty in discriminating the two amino acid peaks.

Evaluating the degree of elastin degradation (elastolysis) in the wall of AAA from older (median 68 years) male patients, Wilson et al. [79] measured serum elastin peptides (SEPs) and plasma elastin-alpha1-antitrypsin complex (E-AT), which are considered markers of elastin degradation. The content of SEP was increased in the wall of AAA, indicating a higher degree of degradation, and both E-AT and SEP were related inversely to the pressure–strain elastic modulus and stiffness. The decrease in the effective stiffness of elastin in AAA development could favour diameter enlargement under applied haemodynamic loads [131]. Furthermore, the elastin degradation correlated with increased aortic wall distensibility and aneurysm formation. Within the wall of atherosclerotic AAA (12 men and one woman, aged 52–74 years) the elastin content was three times lower than within the wall of normal AA (seven men , aged 23–44 years; [80]).

Another study suggested that the incidence of AAA formation occurred less often in females and this could be attributed to favourable local haemodynamic conditions in female aortas [132]. More specifically, their data suggested that a low peripheral vascular resistance of the female internal iliac arteries results in decreased overall oscillatory shear stress burden on the endothelium of the female infrarenal aorta, preventing local pro-inflammatory changes.

Elastin content was also analysed specifically in the wall of IRAA aneurysms showing less and fragmented elastin [72]. This trend also held true specifically looking at the medial layer [73] where elastin was virtually absent and completely disrupted with fewer cross-links. The elastin concentration in the wall of IRAA aneurysms was about 10 times less than the value in the wall of control IRAA [72]. This was further validated by two other studies that showed that the amount of total elastin in IRAA aneurysm was decreased compared with control [68,74]. These studies also characterized the number of elastin cross-links showing a similar amount to control IRAA, but this could vary depending on the method used. The elastin content and concentration was found to be decreased by 2.5-fold in mid-IRAA aneurysm compared with control tissue [75].

In a study that evaluated the architecture of elastin lamellar units in the wall of IRAA aneurysm with secondary blebs, in the media without a bleb, multiple layers of fragmented elastin lamellae were formed, whereas in the wall of the secondary bleb, the number of medial elastin tissue elements along the circumference of the bleb progressively decreased with no identifiable elastin lamellae [77]. Only a few scattered isolated fragments of elastin tissue were present at the apex of the blebs. It was suggested that the marked attenuation of elastin at the apex of the blebs could provide the focal site for rupture.

The elastin content and microstructure in the media layer of the wall of IRAA aneurysm and normal IRAA was studied by Lopez-Candales et al. [76]. They reported a pronounced decrease in medial elastin fibres and elastin-associated microfibrillar material in the wall of IRAA aneurysm as opposed to intact medial elastin fibres in the wall of control IRAA.

4.1.3. Effects of location

It was shown that the wall of distal TA contained about 60 elastin lamellar units, whereas the number of elastin lamellar units was about 28 in the wall of IRAA [69]. Other studies have also shown this trend of decreased elastin content as you move down the aortic tree (figure 2b) [31,70]. Specifically, elastin content was decreased sharply by approximately 50 per cent from DTA to the supraceliac AA (SCAA), and was decreased by 58 per cent from suprarenal AA (SRAA) to mid-IRAA. This trend was further validated by another study showing that elastin content decreased by 49 per cent from the SRAA to the IRAA [71]. The number of elastin lamellar units was highest in the ATA wall (80 ± 4) and lowest in the mid-IRAA wall (32±4) [31]. Also, besides differences in amount of elastin, the thickness of various layers of the wall differed between locations. The intimal and adventitial thickness was lowest in the ATA wall (0.08 ± 0.03 and 0.32 ± 0.06 mm, respectively) and was increased along the aorta to become highest in the mid-IRAA (0.64 ± 0.24 and 0.45 ± 0.01 mm, respectively). On the other hand, the thickness of the media layer was highest in the ATA wall (1.41 ± 0.09 mm) and was decreased along the aorta to become lowest in the mid-IRAA wall (0.64 ± 0.18 mm).

4.1.4. Summary table

Table 4 summarizes the current knowledge regarding elastin amount and architecture in the wall of human AA in ageing and disease. The empty positions in the table indicate information that is missing from the literature.

Table 4.

Summary of studies published to date regarding elastin content, concentration and architecture in human abdominal aorta: effects of age and disease. Tilde symbol (∼), no change; up arrow (↑), increase; down arrow (↓), decrease; IRAA, infrarenal abdominal aorta; AAA, abdominal aortic aneurysm; ATA, ascending thoracic aorta; SRAA, suprarenal abdominal aorta; TA, thoracic aorta; w/o, without; wrt, with respect to. Asterisks (**) to be treated with caution owing to difficulty in discriminating the two amino acid peaks.

additional co-morbidity	elastin content	elastin concentration	elastin cross-linking	no. of elastin lamellar units	elastin microstructure	references
abdominal aorta: elastin
age and location: IRAA
	location-specific: ↓ wrt ATA and SRAA	age-specific: ↓ (limited age group)		location-specific: ↓ wrt distal TA		[31,68–71]
aneurysm: IRAA
	overall: ↑/↓	age-specific: ↓ (limited age group) overall: ↓	overall: ∼/↓		overall: microfibrillar, fragmented layer-specific: virtually absent and completely disrupted in media	[68,72–76]
secondary bleb	location-specific: ↓ progressively along the circumference of the bleb				layer-specific: multiple layers of fragmented elastin lamellae in media w/o bleb, no identifiable elastin lamellae in media of bleb location-specific: scattered at bleb apex	[77]
aneurysm: AAA
		overall: ↓	overall: ↑, **		overall: degraded, disrupted	[78,79]
atherosclerosis	overall: ↓					[80]

4.2. Collagen

4.2.1. Effects of age

Measuring the content of collagen types I and III in four segments of the human aorta equally spaced from proximal end to distal end, it was found that collagen overall was non-uniformly distributed among the different segments of the aorta [70]. In particular, with ageing, the content of collagen type III was decreased from proximal to distal aorta, with collagen type I always being the major type present in all the segments of the human aorta. It has also been shown that collagen content as well as concentration were increased with age in the AA [82].

4.2.2. Effects of disease

It has been reported that there was a significant direct correlation between the increase in the collagen content in AA and the levels of esterified cholesterol in males [82]. The latter is associated with lipid deposition in the genesis of atherosclerosis, and it was suggested that this increase of the collagen content accompanied early atherosclerotic changes.

Increased turnover of collagen was observed in AAA compared with control, which was associated with an increase in the metabolism of type III collagen in the wall of AAA resulting in an overall degradation of type III collagen in the AAA wall [84]. Increased collagen neosynthesis was observed in AAA compared with control indicating an increased collagen turnover, and related positively to pressure–strain elastic modulus and stiffness in the wall of AAA [79]. These results were in agreement with previous similar studies that were performed specifically in the wall of IRAA aneurysms and control IRAAs. The content of collagen was found to be increased in aneurysmal IRAA compared with control IRAA [75]. Also, maturely cross-linked type III collagen fibrils were detected in the media and adventitia, and newly synthesized type III procollagen was concentrated in the media, of the IRAA aneurysm [133]. The staining for mature type III collagen and type III procollagen was weak in the intima. Newly synthesized type I procollagen was localized in the intima of the IRAA aneurysm, and its staining was slight or completely absent in media/adventitia. In the wall of control IRAA, no new collagen was synthesized and the turnover of collagen was in equilibrium. The maturely cross-linked type III collagen was varied transmurally in both IRAA aneurysm and control IRAA, but was more abundant in the media and adventitia of IRAA aneurysm. Specifically in the wall of atherosclerotic AAA, the content of collagen was found to be 0.41 mg per mg dry tissue compared with 0.37 mg per mg in normal AA [80]. Here, type I collagen was found to be the most abundant type in both the AA and AAA, constituting about 60 per cent of the total collagen. Type III and V collagens constituted about 22 per cent and 17 per cent of total collagen, respectively, and when comparing between AA and AAA, there were no significant differences between any of the collagen types. However, it has been shown in another study that collagen content decreased in the medial layer of atherosclerotic IRAA aneurysms by 48 per cent compared with control [73]. It was suggested that the decrease in collagen content in IRAA aneurysm might be due to low biosynthesis rate of type I collagen, which was probably associated with atherosclerosis in AAAs [133]. It is also important to note that the intimal layer had marked atherosclerosis and old collagen was abundant in the media and in the periphery of the plaque, whereas new collagen was mostly in the fibrous cap [73]. Another report also exists that showed that the collagen content remained unchanged in the wall of non-atherosclerotic IRAA aneurysm compared with control IRAA [74].

Total collagen concentration and concentration of collagen types I and III have also been evaluated specifically in the wall of IRAA aneurysm and in control IRAA [72]. The total collagen concentration was higher in the wall of IRAA aneurysms (37 ± 16%) compared with control IRAA (24 ± 5%), but no differences were seen in the concentrations of type I collagen (74 ± 4% and 73 ± 4%, respectively) or type III (26 ± 4% and 27 ± 4%, respectively). Rizzo et al. [72] also reported a higher concentration of collagen in the wall of IRAA aneurysms compared with control IRAA. In another study, however, the collagen concentration was not statistically different between IRAA aneurysm and control IRAA [75]. The reason for this inconsistency could be attributed to differences in the protocols or methods, different demographic data or location within the IRAA where the analysis was performed.

Gandhi et al. showed that there was a decrease in the collagen fibre waviness and a defect was detected in the collagen microstructure in the wall of IRAA aneurysm (figure 5e), compared with control aorta (figure 5a,b). They concluded that collagen fibres did not behave as a coherent network in the wall of IRAA aneurysm [74]. According to another study, the collagen in the medial layer of IRAA aneurysms was disordered compared with the well-arranged fibres seen in the control tissue [76]. This was further validated by Lindeman et al. [48] who showed that both collagen content and cross-linking were increased in IRAA aneurysms. In the media and adventitia layers of IRAA aneurysm, a complete loss of the normal architecture was seen. There was a loss of the distinction between media and adventitia collagen organization, and there was deposition of aggregated parallel collagen sheets that appeared rigid [48]. This was opposed to the media layer of control IRAA tissue, where there were small, interdispersed collagen fibrils that ran mainly perpendicular to the circumferential elastin sheets. In the adventitia layer of control IRAA tissue, there was loose knitting of highly organized ribbon-like collagen bands that braced the medial and intimal layers of the vascular wall.

It was reported that the medial layer of the residual wall of secondary blebs in IRAA aneurysms consists mostly of collagen fibres [77].

4.2.3. Effects of location

Collagen content has been shown to decrease by 54 per cent in the mid-IRAA compared with the ATA [31]. Collagen content decreased more gradually from the SCAA to the SRAA, but was not significantly different among SCAA, SRAA and mid-IRAA. The decrease of collagen content from ATA to SRAA was similar to the aforementioned decrease of elastin content from ATA to mid-IRAA. The elastin/collagen ratio was approximately constant from the ATA to the SRAA, and was decreased in the mid-IRAA. The content of type I collagen did not differ between the SRAA and the IRAA [71].

Regarding distribution of collagen fibre orientations down the aortic tree, two very distinct counter-rotating collagen fibre families were found in the intima, media and adventitia layers of the DTA and AA, often with a third and sometimes fourth family of fibres in the intima [83]. The two distinct families of collagen fibres were almost symmetrically arranged with respect to the cylinder axis, and organized in separate layers in the media and adventitia, with one preferred fibre direction per layer. Very small dispersion of the interlaminar components was reported through the thickness for both DTA and AA.

Gasser et al. [81] calculated the distribution of collagen fibre orientations within the anterior wall of IRAA aneurysm. It was found that, in the anterior wall of IRAA aneurysm, the collagen fibre orientation was widely dispersed, with the dispersion being higher in the tangential than in the cross-sectional plane. Also, there was no significant difference between the medial and adventitial layers, and the layered directional organization of collagen in control AA was not evident in the IRAA aneurysm.

4.2.4. Summary table

Table 5 summarizes the current knowledge regarding collagen amount and architecture in the wall of human AA in ageing and disease. The empty positions in the table indicate information that is missing from the literature.

5. Conclusions

The content of elastin in ATA remains unaltered with age; however, its concentration is decreased. This imparts a loss in the effective stiffness of elastin and the recoil capacity of the wall which favour diameter enlargement [9,131]. Evidently, it is seen that the amount of elastin is decreased in ATA aneurysms, and interestingly, there exist region-specific alterations specifically in those accompanied by BAV where a decrease is seen more so in the concavity than in the convexity of the ATA. In ATA dissection, the content, concentration and amount of cross-links of elastin are decreased. This change in the micro-architecture of elastin can compromise the elastic integrity of the vascular tissue potentially predisposing the initiation of dissection especially when considering that the amount of interlaminar fibres is decreased. A decrease in those fibres which are radially running reduces the adhesive strength between the aortic wall layers [125] leading to spontaneous tissue tears when the pressure-induced wall stresses exceed this strength. When the diseases of dissection and aneurysm are combined in the ATA, elastin appears fragmented, disrupted, irregularly arranged and with decreased anisotropy. Interestingly though, in dissection accompanied by atherosclerosis, the fragmentation of elastin is mild or non-existent, implying that the atherosclerotic plaque itself might play a role in reducing the strength of the wall and leading to dissection.

Currently, it remains unclear if the content of collagen in the ATA increases with age or is unchanged; however, when dissection is present the concentration of collagen decreases with age. A decrease in collagen concentration is also seen in aneurysmal ATA compared with non-aneurysmal which, similar to elastin in ATA aneurysms, results in a reduction of the effective stiffness. This reduction in effective stiffness of both collagen and elastin weakens the vascular wall predisposing it to dilation. Interestingly, in MFS patients with ATA aneurysm the concentration of collagen remains unchanged, but their collagen possesses a higher amount of cross-links and a decreased fibre undulation. This decreased undulation is not only associated with diameter enlargement, potentially explaining the aneurysm formation of MFS ATA with unchanged collagen concentration, but also can have a negative effect on the delamination strength of the ATA tissue providing insight on why MFS ATA aneurysms often dissect. Additional co-morbidities in the presence of ATA aneurysm also have an effect on content and concentration of collagen types: the content and concentration of type IV are increased, and the content and concentration of types I and III are decreased (except in MFS ATA aneurysm with AVR where the content of type III is increased). Similarly, additional co-morbidities have an effect on the content and concentration of collagen types in ATA dissection: in all cases of co-morbidities except aortitis, the content of types I and III is increased; in aortitis they remain unchanged. ATA dissection with atherosclerosis also follows this trend of increased collagen type I and III, but the type IV appears to be defected particularly above the intimal plaque. Overall, collagen appears disorganized, fibrotic and disrupted in most forms of aortic disease of the ATA. It takes the form of thin scattered fibres in ATA dissection and aneurysm, with collagen bundles having longitudinal orientation subintimally in the right lateral region of ATA aneurysms. The latter could increase the stiffness of the wall in the longitudinal direction, and could predispose the tissue to failure in the other direction (circumferential in the subintima).

It appears that the elastin concentration in IRAA is decreased with age, and the elastin content as well as the number of elastin lamellar units are decreased in IRAA when compared with more proximal segments of the aorta. From a mechanics point of view, the reduction in elastin with age as well as regionally makes the IRAA a prime location for aneurysmal dilation owing to a loss in effective stiffness and recoil capacity of elastin. Evidently, it is seen that in cases of both IRAA aneurysm and AAA the concentration of elastin is decreased in the vascular wall. Elastin concentration also decreases with age in IRAA aneurysms and if atherosclerosis coexists. However, despite this change in concentration in IRAA aneurysm, it is unclear if the content of elastin or the amount of elastin cross-links are different from control. Interestingly, the amount of elastin cross-links seems to be increased in the wall of AAA, which is something that could enhance the integrity of the wall and prevent it from dilating further or rupturing. Overall, in the wall of AAA and specifically in the wall of IRAA, elastin is disrupted, fragmented and microfibrillar.

Looking regionally, the overall content of collagen decreases from ATA to IRAA, although type I does not change significantly from the SRAA to the IRAA, and this decrease is less severe than that of elastin between SRAA and IRAA. Adding to this, the ratio of elastin to collagen content in the wall of IRAA is less compared with more proximal segments of the aorta, indicating a dominating role of collagen in the mechanical strength of IRAA. Despite this dominant role in mechanical strength, it is unclear if the content and concentration of collagen increases or remains unaltered overall in the wall of IRAA aneurysm highlighting the importance of elastin loss in initial aneurysmal dilation. However, type III collagen content and number of collagen cross-links are shown to increase, whereas the undulation of collagen fibres is decreased. Together these augment the effective stiffness of the wall and can protect it from further dilation and rupturing, highlighting the importance of collagen in the later stages of aneurysm. Also supporting this, an increase in collagen III turnover in AAA is seen, which resists wall stresses. However, the microstructure of collagen appears to be degraded in the AAA wall, most likely newly laid down collagen is not formed into coherent sheets, and the normal architecture is not seen in IRAA aneurysms. Interestingly, in IRAA with atherosclerosis there is a marked decrease in collagen content and an increase in the amount of collagen cross-links, but some controversy has been seen when collagen was reported to be raised in AA with early atherosclerosis. In AAA with atherosclerosis, the amount of collagen, but not the concentration, is increased with type I collagen being the most abundant of types I, III and V both in the wall of AAA and in the wall of non-aneurysmal AA.

The aim of this review article was to synthesize the literature regarding the interrelation of location, age and disease impact with the microstructure of elastin and collagen in the human aorta. The incomplete information exemplified in tables 2–5 indicates that, at present, the regional variation in aortic wall fibre content and architecture in ageing and disease—including co-morbid conditions such as MFS and BAV disease—is still not fully understood. This information and the lack of information in the literature should motivate future experimental and theoretical studies combined with state-of-the-art imaging tools. For example, information from the regional variation in the content and distribution of the connective fibres in different locations of the human aorta, in ageing and disease, could be incorporated into a computational model of the human aorta to simulate ageing-induced arteriosclerosis, aneurysm formation and dissection, which are very common forms of aortic disease. An expected long-term outcome of such work would be a subject-specific simulation tool that would enable the precise prediction of aortic remodelling in health and disease, and could possibly guide the decision-making in clinical diagnosis and treatment.