Previous studies on BAV (bicuspid aortic valve)-related aortopathy, whose aetiology is still debated, have focused mainly on severe dilatations. In the present study, we aimed to detect earlier signs of aortopathy. Specimens were collected from the ‘concavity’ (lesser curvature) and the ‘convexity’ (greater curvature) of mildly dilated AAs (ascending aortas; diameter ≤4 cm) with stenotic TAV (tricuspid aortic valve) or BAV and from donor normal aortas. Specimens were submitted to morphometry, immunohistochemistry and differential gene-expression analysis, focusing on SMC (smooth muscle cell) phenotype, remodelling, MF (myofibroblast) differentiation and TGFβ (transforming growth factor β) pathway. Smoothelin and myocardin mRNAs decreased in all the samples from patients, with the exception of those from BAV convexity, where a change in orientation of smoothelin-positive SMCs and an increase of α-SMA (α-smooth muscle actin) mRNA occurred. Dilated aortas from BAV and TAV patients showed both shared and distinct alterations concerning the TGFβ pathway, including an increased TGFβ and TGFβR2 (TGFβ receptor 2) expression in both groups and a decreased TGFβR1 expression in BAV samples only. Despite a decrease of the mRNA coding for the ED-A (extra domain-A) isoform of FN (fibronectin) in the BAV convexity, the onset of the expression of the corresponding protein in the media was observed in dilated aortas, whereas the normal media from donors was negative for this isoform. This discrepancy could be related to modifications in the intima, normally expressing ED-A FN and showing an altered structure in mild aortic dilatations in comparison with donor aorta. Our results suggest that changes in SMC phenotype and, likely, MF differentiation, occur early in the aortopathy associated with valve stenosis. The defective expression of TGFβR1 in BAV might be a constitutive feature, while other changes we reported could be influenced by haemodynamics.

CLINICAL PERSPECTIVES

  • BAV (bicuspid aortic valve) stenosis is the most common cardiovascular malformation and is associated with an increased risk of aortic dilatation, aneurysm and dissection. However, an understanding of the early mechanisms of the malformation is lacking.

  • In the present study we have compared the biomolecular and immunohistochemical findings in aortic specimens from tricuspid and BAV patients with severe aortic valve stenosis and mildly dilated aorta, focusing on the possible role of the trans-differentiation of medial SMCs to MFs and the expression of TGFβ and its receptors. The results suggest that a non-specific up-regulation of TGFβ, occurring early in the development of post-stenotic aortic dilatation, may be redirected in BAV patients towards a programme of increased matrix degradation, owing to a disequilibrium in the expression of the two TGFβR subunits.

  • New therapeutic targets or prognostic tools for the aortopathy related to bicuspidy may be identified by expanding the research in this direction.

INTRODUCTION

BAV (bicuspid aortic valve) stenosis is the most common cardiovascular malformation, associated with an increased risk for aortic dilatation, aneurysm and dissection [1]. The historical debate on the pathogenesis of BAV aortopathy, i.e. whether it is determined by genetics or haemodynamics, has lately given way to the consideration that these two causative factors may not necessarily exclude each other [24]. Aortic wall remodelling processes may therefore be concomitantly affected both by congenital defects and by an altered biomechanical environment secondary to long-standing haemodynamic derangements.

The TGFβ (transforming growth factor β) pathway is attracting an increasing attention in studies focusing on AA (ascending aorta) aneurysms, by virtue of its important role in tissue fibrosis and ECM (extracellular matrix) remodelling [5]. Among its multiple functions, TGFβ can induce MF (myofibroblast) differentiation [6], in concert with the ED-A (extra domain-A) splicing variant of FN (fibronectin), whose expression is limited in healthy adult tissues but is up-regulated in pathological settings [7].

MFs, also defined as ‘mesenchyme-like interstitial cells’, are specialized cells arising in pathophysiological conditions and contributing to tissue repair during wound healing. MF differentiation is gaining a growing interest, as these cells could play a key role in aortopathies, thus potentially representing a therapeutic target to prevent disease progression [6]. MF activation in aortic aneurysms has been addressed so far mainly in animal models of disease [8], while specific studies on human aortic specimens are still in their infancy [9]. The univocal identification of vascular MFs can be quite difficult, since these cells, at fully differentiated phenotype, share many of their characteristics with SMCs (smooth muscle cells), including the expression of α-SMA (α-smooth muscle actin) and vimentin.

Recent studies demonstrated that aortic stenosis and regurgitation are associated with distinct AA wall alterations, suggesting that haemodynamic factors may influence their expression [10,11]. On this basis, we previously investigated ECM remodelling and SMC apoptosis in patient groups homogeneous for valve dysfunction, with severe [12] and moderate [13] dilatations. In those studies, for the purpose of distinguishing between constitutive and flow-induced changes, we also introduced a protocol for comparative analysis between two areas of the AA, corresponding to the convexity (greater outer curvature) and the concavity (lesser inner curvature) of its profile, known to be exposed to different haemodynamic stimuli in BAV patients [3,14].

The objective of the present study was to gain information about the earlier phase of aortopathy, comparing aortic wall samples from the convexity and the concavity of mildly dilated AAs in patients with stenotic TAV (tricuspid aortic valve) and BAV, focusing on the expression of genes (i) considered as well-known markers of cell phenotype, (ii) belonging to the TGFβ pathway or (iii) playing a role in vascular remodelling.

MATERIAL AND METHODS

Sample collection

Specimen pairs from the concavity and the convexity of mildly dilated AAs (≤4 cm) were harvested from patients with isolated TAV stenosis (n=13) and BAV stenosis (n=22) undergoing valve replacement (from the ends of the transverse aortotomy, 1–2 cm distal to the sino-tubular junction). None of the patients had positive family history for aneurismal disease of the aorta and, as such, only cases of ‘sporadic BAV’ were included in the present study. The measure of the aortic diameter was taken by transthoracic echocardiography at the ascending tract level, and the aortic dimension was indexed to the patient's height as suggested by Svensson [15]. Within the BAV group, two subgroups were identified, according to valve morphotype, i.e. the cusp fusion pattern (right–left coronary fusion, n=13; right-non-coronary fusion, n=9). To warrant group homogeneity and comparability, we excluded from this study patients with aortic valve stenosis affected by impaired systolic ventricular function.

The study received the Institutional Ethic Committee approval and patients gave their informed consent. Control aortic specimens were obtained during organ harvesting from heart donors with no evidence of aneurismal disease or familiarity (n=12, all TAV, cause of death: trauma in seven cases, spontaneous cerebral haemorrhage in five cases).

Clinical characteristics are summarized in Table 1. To help the interpretation of selected results, specimens were taken also from three aneurismal AAs with TAV (diameter >5 cm), representative of a more advanced stage of the dilative disease. In summary, overall 100 aortic specimens were analysed in the present study.

Table 1
Clinical characteristics in the study groups

Values are means±S.E.M. *P<0.05 compared with donors; P<0.05 compared with TAV; all other differences were not significant. CSA/h, ascending aortic cross-sectional area to patient height ratio; N/A, not available.

Donors (n=12)TAV (n=13)BAV (n=22)
Male gender (%) 50 46 59 
Age (years) 39.7±13.2 (range 20–61) 71.3±2.53 (range 55–84)* 62.1±1.87 (range 42–79)* 
Aortic diameter (cm) 3.0±0.09 3.6±0.18* 3.9±0.16* 
Aortic CSA/h (cm2/m) 4.5±0.2 6.6±0.6* 7.6±0.6* 
Peak valve gradient (mmHg) N/A 83±9 86±11 
Mean valve gradient (mmHg) N/A 55±7 59±9 
Aortic valve area index (cm2/m2N/A 0.39±0.02 0.38±0.02 
Hypertension (%) 25 84* 81* 
Medication    
 Angiotensin receptor blockers (%) 38* 18 
 Statins (%) 41* 
 Aspirin (%) 31 14 
Donors (n=12)TAV (n=13)BAV (n=22)
Male gender (%) 50 46 59 
Age (years) 39.7±13.2 (range 20–61) 71.3±2.53 (range 55–84)* 62.1±1.87 (range 42–79)* 
Aortic diameter (cm) 3.0±0.09 3.6±0.18* 3.9±0.16* 
Aortic CSA/h (cm2/m) 4.5±0.2 6.6±0.6* 7.6±0.6* 
Peak valve gradient (mmHg) N/A 83±9 86±11 
Mean valve gradient (mmHg) N/A 55±7 59±9 
Aortic valve area index (cm2/m2N/A 0.39±0.02 0.38±0.02 
Hypertension (%) 25 84* 81* 
Medication    
 Angiotensin receptor blockers (%) 38* 18 
 Statins (%) 41* 
 Aspirin (%) 31 14 

Histological analysis

Harvested samples were fixed in 4% buffered formaldehyde, dehydrated and embedded in paraffin. Consecutive 5 μm cross-sections were stained with haematoxylin or with haematoxylin/orcein for nucleus and elastic fibre staining respectively. Image screening and photography were performed using the Leica IM1000 software. The same software was also used to measure, in five consecutive cross-sections for each AA sample, the wall thickness (intima+media) and the cell density in the media, expressed as the number of haematoxylin-stained cells/mm2 in the central region of the media. Alternatively, 5 μm cross-sections were stained with fluorescent Hoechst 33258 (Sigma–Aldrich) for nucleus staining and detection of natural autofluorescence emitted by elastic laminae. Image screening and photography were then performed using the Leica 4000F software. The intima and the media were defined as the layers enclosed between the lumen and the internal elastic lamina and between the internal and the external elastic laminae respectively. The adventitia was not included in measurements, as this layer can be partially damaged or removed during AA sample harvesting, which makes its measurement less reliable. All the measurements were performed by the same person, blinded to the patient phenotype.

RNA extraction and RT–PCR (reverse transcription–PCR) analysis

Total RNA was extracted from whole AA samples stored in RNALater (Qiagen) at −80°C immediately after harvesting, using the RNAeasy minikit (Qiagen). RNA was treated with DNase (Qiagen) to remove DNA contamination, its concentration was measured using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies) and its integrity was verified by electrophoresis on denaturing 1% agarose gel. The absence of residual DNA was verified by PCR on total RNA without reverse transcription. GeneBank® sequences for human mRNAs and the Primer Express software (Applied Biosystem) were used to design primer pairs for the target genes (Table 2). RT–PCR amplification experiments were performed as published previously, using both semi-quantitative RT–PCR, both real-time RT–PCR for selected targets [16]. GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was chosen as reference housekeeping gene. Relative quantitative RT–PCR was used to determine the fold difference for genes. The Opticon II machine (Bio-Rad) was used for real-time RT–PCR. The PCR efficiency was determined for each primer pair and was calculated using a dilution series and the Opticon II analysis software. The Verity thermal cycler (Applied Biosystems) and the ChemiDoc and associated software Quantity One (Bio-Rad) were used for semi-quantitative densitometric analysis of PCR products after electrophoresis.

Table 2
Summary of the RT–PCR primer sequences, position, annealing temperature and PCR product length for each target gene analysed
Gene namePrimer positionPrimer sequenceAnnealing tempeature (°C)PCR product length (bp)
GAPDH 472 5′-GCATCCTGCACCACCAACTG-3′ 55 327 
 799 5′-GCCTGCTTCACCACCTTCTT-3′   
Smoothelin 2319 5′-AGCACCATGATGCAAACCAAG-3′ 62 143 
 2461 5′-TCTTTCTTCTTCTCGGCCTGC-3′   
ED-A FN 5875 5′-ACAACCACGGATGAGCTG-3′ 54.8 185 
 5690 5′-CCAGTGCACAGCTATTCCTG-3′   
Myocardin 501 5′-TCGAAGTGAAGACCCCCAAA-3′ 61 108 
 609 5′-TGAGCCAGAAGCAAGATCCTG-3′   
α-SMA 1049 5′-TCCGGAGCGCAAATACTCTGT-3′ 61 102 
 1150 5′-CCGGCTTCATCGTATTCCTGT-3′   
TGFβ 1716 5′-CTATTGCTTCAGCTCCACGGA-3′ 61 114 
 1829 5′-AAGTTGGCATGGTAGCCCTTG-3′   
TGFβRI 421 5′-AAAGTCATCACCTGGCCTTGG-3′ 60 113 
 533 5′-CAGTGCGGTTGTGGCAGATAT-3′   
TGFβRII 3237 5′-TCAAGCACAGTTTGGCCTGAT-3′ 56 110 
 3346 5′-ACGCTGTGCTGACCAAACAA-3′   
LTBP-1 2203 5′-TGGATCTTACCAGTGCGTTCC-3′ 60 101 
 2303 5′-CATTTGCGCAGACGTTTGG-3′   
SFSWAP 487 5′-TAGCCTTGCATACGGACTTGC-3′ 59 110 
 597 5′-AAGTGAACCCCACGGCATT-3′   
TIMP-1 626 5′-TTGGCTGTGAGGAATGCACA-3′ 60 106 
 731 5′-TTTTCAGAGCCTTGGAGGAGC-3′   
TIMP-2 2295 5′-TGTGGTTTCCTGAAGCCAGTG-3′ 57 164 
 2458 5′-CGCCAAGCTTCGGTTTCAT-3′   
MMP-2 493 5′-GAGCTGCAACCTGTTTGTGCT-3′ 59 105 
 597 5′-CGCATGGTCTCGATGGTATTC-3′   
MMP-9 1146 5′-CCTCGAACTTTGACAGCGACA-3′ 61 111 
 1256 5′-AGGAATGATCTAAGCCCAGCG-3′   
Gene namePrimer positionPrimer sequenceAnnealing tempeature (°C)PCR product length (bp)
GAPDH 472 5′-GCATCCTGCACCACCAACTG-3′ 55 327 
 799 5′-GCCTGCTTCACCACCTTCTT-3′   
Smoothelin 2319 5′-AGCACCATGATGCAAACCAAG-3′ 62 143 
 2461 5′-TCTTTCTTCTTCTCGGCCTGC-3′   
ED-A FN 5875 5′-ACAACCACGGATGAGCTG-3′ 54.8 185 
 5690 5′-CCAGTGCACAGCTATTCCTG-3′   
Myocardin 501 5′-TCGAAGTGAAGACCCCCAAA-3′ 61 108 
 609 5′-TGAGCCAGAAGCAAGATCCTG-3′   
α-SMA 1049 5′-TCCGGAGCGCAAATACTCTGT-3′ 61 102 
 1150 5′-CCGGCTTCATCGTATTCCTGT-3′   
TGFβ 1716 5′-CTATTGCTTCAGCTCCACGGA-3′ 61 114 
 1829 5′-AAGTTGGCATGGTAGCCCTTG-3′   
TGFβRI 421 5′-AAAGTCATCACCTGGCCTTGG-3′ 60 113 
 533 5′-CAGTGCGGTTGTGGCAGATAT-3′   
TGFβRII 3237 5′-TCAAGCACAGTTTGGCCTGAT-3′ 56 110 
 3346 5′-ACGCTGTGCTGACCAAACAA-3′   
LTBP-1 2203 5′-TGGATCTTACCAGTGCGTTCC-3′ 60 101 
 2303 5′-CATTTGCGCAGACGTTTGG-3′   
SFSWAP 487 5′-TAGCCTTGCATACGGACTTGC-3′ 59 110 
 597 5′-AAGTGAACCCCACGGCATT-3′   
TIMP-1 626 5′-TTGGCTGTGAGGAATGCACA-3′ 60 106 
 731 5′-TTTTCAGAGCCTTGGAGGAGC-3′   
TIMP-2 2295 5′-TGTGGTTTCCTGAAGCCAGTG-3′ 57 164 
 2458 5′-CGCCAAGCTTCGGTTTCAT-3′   
MMP-2 493 5′-GAGCTGCAACCTGTTTGTGCT-3′ 59 105 
 597 5′-CGCATGGTCTCGATGGTATTC-3′   
MMP-9 1146 5′-CCTCGAACTTTGACAGCGACA-3′ 61 111 
 1256 5′-AGGAATGATCTAAGCCCAGCG-3′   

The panel of genes analysed in the present study was selected on the basis of the current literature as informative for cellular phenotype transformation, with particular focus on SMC switch and MF differentiation in order to highlight the possible onset of these phenomena in specimens from patients with BAV or TAV with only mild aortic dilatation (see the Discussion for further details). Moreover, we selected for RT–PCR analysis a group of genes known to be regulated mainly at transcriptional level, in order to obtain reliable data.

Immunohistochemistry

Target proteins were smoothelin, ED-A FN, total FN, α-SMA and vWF (von Willebrand factor). Primary antibodies used are listed in Table 3. Consecutive 4% formaldehyde-fixed 5 μm cross-sections were deparaffinized and rehydrated. Experiments were performed as previously published [16]. Primary antibodies were omitted in the negative controls of the reactions (see Supplementary Figure S1 at http://www.clinsci.org/cs/124/cs1240097add.htm). Nuclei were counterstained with Mayer's haematoxylin (Sigma–Aldrich). Image screening and photography were performed using a Leica IM1000 System.

Table 3
Monoclonal primary antibodies used for immunohistochemical analysis

See the text for abbreviations

Target proteinAntibody typeCloneDilutionCompany
Smoothelin Mouse monoclonal R4A 1:100 Abcam 
ED-A FN Mouse monoclonal IST-9 1:100 Santa Cruz Biotechnology 
FN Mouse monoclonal IST-4 1:100 Sigma 
α-SMA Mouse monoclonal 1A4 1:200 Sigma 
vWF Rabbit polyclonal − 1:500 Abcam 
Target proteinAntibody typeCloneDilutionCompany
Smoothelin Mouse monoclonal R4A 1:100 Abcam 
ED-A FN Mouse monoclonal IST-9 1:100 Santa Cruz Biotechnology 
FN Mouse monoclonal IST-4 1:100 Sigma 
α-SMA Mouse monoclonal 1A4 1:200 Sigma 
vWF Rabbit polyclonal − 1:500 Abcam 

Statistical analysis

Statistical analysis was performed using the GraphPad software (Prism 4.0) and SPSS ver13.0. Significant differences in RT–PCR were identified by a cut-off 2-fold change in gene expression or detected by a Student's t test comparing target gene-fold expression in BAV or TAV aortic samples compared with baseline in normal aortas, in agreement with others [17,18].

Summaries for continuous variables are presented as means±S.E.M., for categorical variables as numbers and percentages. Normality of data distribution was assessed for each variable in order to choose between parametric and non-parametric tests. Comparisons between the convexity and the concavity were performed through paired Student's t test or Wilcoxon signed-rank test. Comparisons across study groups (TAV, BAV, donors) were done through the ANOVA test, with Bonferroni correction, or through Kruskall–Wallis test with Exact correction, for normally or asymmetrically distributed continuous variables respectively. Within the BAV group, comparisons between the two valve morphotypes were performed by unpaired Student's t test or Mann–Whitney test. All comparisons for categorical variables were done through the χ2 test. Correlations were analysed by Spearman test. Values of P<0.05 were considered significant.

RESULTS

Mildly dilated BAV and TAV aortic specimens show morphological and morphometric alterations in comparison with donors

Morphometric analysis results are summarized in Table 4. The overall intima+media thickness significantly decreased in all samples from dilated AAs in comparison with donor aortas (P<0.05). Conversely, cell density showed a trend for increase in both groups and in both curvatures compared with donor AAs, with a trend for greater increase in cellularity in BAV than in TAV samples (Table 4). The correlation between aortic diameter and cell density was significant in the convexity of BAV AAs (r=0.93, P<0.05). Cell density in the convexity showed significant inverse correlations with age in TAV patients (r=−0.90, P<0.05) and with intima+media thickness in donors and TAV patients (r=−0.90, P<0.05 and r=0.70, P<0.05 respectively). No significant correlation was found in mildly dilated AAs between aortic diameter and age, neither between aortic diameter and intima+media thickness.

Table 4
Morphometry results

Values are means±S.E.M. *P<0.05 compared with the same curvature in control samples from donors.

DonorsTAVBAV
ParameterConcavityConvexityConcavityConvexityConcavityConvexity
Medial cell density (cells/mm2767±32 731±55 833±152 788±94 1045±186 944±192 
Intima+media thickness (mm) 2.07±0.15 2.03±0.15 1.29±0.12* 1.23±0.18* 1.36±0.14* 1.46±0.09* 
DonorsTAVBAV
ParameterConcavityConvexityConcavityConvexityConcavityConvexity
Medial cell density (cells/mm2767±32 731±55 833±152 788±94 1045±186 944±192 
Intima+media thickness (mm) 2.07±0.15 2.03±0.15 1.29±0.12* 1.23±0.18* 1.36±0.14* 1.46±0.09* 

Mildly dilated BAV and TAV specimens show structural alterations in the intima

Since by definition the presence of the elastic laminae characterizes the arterial media, the aortic layers can be delimited by techniques that highlight these laminae, such as orcein staining and autofluorescence. On this basis, a marked alteration of the intima, losing the well-organized structure observed in control AAs (Figures 1A–1D) was observed in all the mildly-dilated AAs. In more detail, while the normal intima was negative to orcein staining (Figure 1A) and to elastic lamina autofluorescence (results not shown), we quite consistently observed a homogeneous orcein staining in samples from mildly dilated AAs with BAV and TAV (Figures 1B and 1C), without a clear distinction between the intima and the media. Conversely, the intima in TAV aneurysmal aortas showed increased thickness, altered structure and cellularity but no orcein staining (Figure 1D).

Initial changes in the mild dilatation of the AA with TAV and BAV stenosis and in aneurysms of the AA

Figure 1
Initial changes in the mild dilatation of the AA with TAV and BAV stenosis and in aneurysms of the AA

Upper panel: RT–PCR data (normalized to the expression level of the endogenous GAPDH housekeeping gene) for ED-A FN expression in AA samples from donors, TAV and BAV patients with mild AA dilatation and from AA aneurysms. Results are expressed as means±S.E.M. P<0.05 compared with the same curvature in donor AAs. Lower panel: orcein/haematoxylin staining (AD) and immunohistochemical analysis of the expression of the ED-A FN isoform (EH) and of vWF (IN) in representative cross-sections of samples from the convexity of donor AAs, TAV and BAV mildly dilated AAs and AA aneurysm. Brown staining corresponds to target protein. Haematoxylin nuclei counterstaining, (AH): ×40 magnification; (IN): ×100 magnification; scale bar represents 20 μm. Solid dark lines in (A) and (D) indicate the thickness of the intima in donor and in aneurysm AAs, as indicated by the absence of orcein-stained elastic laminae and by positivity to ED-A FN (E, H). In TAV and BAV samples the boundaries of the intima cannot be defined on the basis of these parameters (B, C, F, G). Arrows in (IN) indicate representative vWF-positive ECs on the lumen surface.

Figure 1
Initial changes in the mild dilatation of the AA with TAV and BAV stenosis and in aneurysms of the AA

Upper panel: RT–PCR data (normalized to the expression level of the endogenous GAPDH housekeeping gene) for ED-A FN expression in AA samples from donors, TAV and BAV patients with mild AA dilatation and from AA aneurysms. Results are expressed as means±S.E.M. P<0.05 compared with the same curvature in donor AAs. Lower panel: orcein/haematoxylin staining (AD) and immunohistochemical analysis of the expression of the ED-A FN isoform (EH) and of vWF (IN) in representative cross-sections of samples from the convexity of donor AAs, TAV and BAV mildly dilated AAs and AA aneurysm. Brown staining corresponds to target protein. Haematoxylin nuclei counterstaining, (AH): ×40 magnification; (IN): ×100 magnification; scale bar represents 20 μm. Solid dark lines in (A) and (D) indicate the thickness of the intima in donor and in aneurysm AAs, as indicated by the absence of orcein-stained elastic laminae and by positivity to ED-A FN (E, H). In TAV and BAV samples the boundaries of the intima cannot be defined on the basis of these parameters (B, C, F, G). Arrows in (IN) indicate representative vWF-positive ECs on the lumen surface.

Alterations in the intima of dilated AA specimens were paralleled by the expression of the ED-A FN isoform, a component of the ECM of major interest as a player in MF differentiation, showing a marked expression in the intima of donor AAs and aneurysms (Figures 1E and 1H), while only a very thin periluminal layer expressed this isoform in mildly dilated BAV and TAV AAs (Figures 1F and 1G).

The expression of ED-A FN was evaluated also at mRNA level by differential RT–PCR, revealing a decrease only in the convexity of BAV AAs and a significant increase in the concavity of aneurismal aortas in comparison with healthy aortic samples (Figure 1, histogram). This trend was consistent with the expression of SFSWAP (SRp40), a splicing factor facilitating the inclusion of the ED-A exon in FN. The peculiar expression profile of ED-A FN mRNA should be interpreted in the light of its canonical localization within the AA wall (normally confined to the intima) and related to the above-mentioned intima alterations in mildly dilated BAV and TAV AAs, as depicted by the immunohistochemical analysis (Figures 1E–1H).

The integrity of the endothelial layer was similarly assessed in the same series of AA specimens through the immunohistochemical detection of vWF, a well-established marker of ECs (endothelial cells). The endothelial integrity was clearly visible in all the AA specimens we analysed (Figure 1I–1N), allowing us to exclude the possibility that the intima alterations we observed in mildly dilated specimens were related to artefacts occurring during sample harvesting and/or manipulation.

Mildly dilated BAV and TAV AA specimens show distinct expression profiles of members of the TGFβ pathway

Unlike donor AAs, BAV and TAV AAs showed distinct signatures of mRNA alterations between the concavity and the convexity, suggesting asymmetric patterns of gene expression (Figures 2 and 3). No significant difference in gene expression was found between BAV subgroups of different cusp fusion pattern, probably owing to the low numbers.

Smoothelin, myocardin, α-SMA, TGFβ, TGFβR1 and TGFβR2 expression as determined by RT–PCR

Figure 2
Smoothelin, myocardin, α-SMA, TGFβ, TGFβR1 and TGFβR2 expression as determined by RT–PCR

Results were normalized to the expression level of the endogenous GAPDH housekeeping gene. *P<0.05 as determined using a Bonferroni post-hoc test.

Figure 2
Smoothelin, myocardin, α-SMA, TGFβ, TGFβR1 and TGFβR2 expression as determined by RT–PCR

Results were normalized to the expression level of the endogenous GAPDH housekeeping gene. *P<0.05 as determined using a Bonferroni post-hoc test.

LTBP-1, ED-A FN, SFSWAP, MMP-2, TIMP-1 and TIMP-2 expression as determined by RT–PCR

Figure 3
LTBP-1, ED-A FN, SFSWAP, MMP-2, TIMP-1 and TIMP-2 expression as determined by RT–PCR

Results were normalized to the expression level of the endogenous GAPDH housekeeping gene. *P<0.05 as determined using a Bonferroni post-hoc test.

Figure 3
LTBP-1, ED-A FN, SFSWAP, MMP-2, TIMP-1 and TIMP-2 expression as determined by RT–PCR

Results were normalized to the expression level of the endogenous GAPDH housekeeping gene. *P<0.05 as determined using a Bonferroni post-hoc test.

The mRNAs coding for TGFβ and for its binding factor, the LTBP-1 (latent TGFβ binding protein 1), showed a significant increase in BAV convexity (Figure 3). TGFβ mRNA increased in TAV AAs as well, however, with greater variability, causing the lack of significant differences compared with donor AAs (P=0.055). As far as the expression of the two TGFβR (TGFβ receptor) subunits is concerned, we observed a significant 1.85-fold decrease of the mRNA coding for the R1 subunit (TGFβR1) in the BAV group in both the AA curvatures and a 1.38-fold increase of the mRNA coding for the R2 subunit (TGFβR2), reaching the significance in the AA concavity of both BAV and TAV groups (Figure 2). Of interest, the expression of TGFβR1 in BAV convexity showed a significant positive correlation with aortic diameter (r=0.78, P<0.05).

Mildly dilated BAV and TAV specimens show an altered expression of ECM components

As to the differential expression of genes involved in ECM remodelling, the MMP-2 (matrix metalloproteinase 2) mRNA showed a significant increase in the convexity of both BAV and TAV AAs, counteracted by an increase of TIMP-2 (tissue inhibitor of metalloproteinases-2) only in the convexity of TAV AAs (Figure 3). The MMP-9 mRNA was undetectable in all the aortic samples we analysed (Figure 3).

Immunohistochemical analysis revealed that total FN, an adhesive glycoprotein relevant in many tissue matrices, was expressed in all the aortic layers, both in dilated aortas and in control samples (Figure 4). Conversely, although the ED-A FN isoform was observed exclusively in the intima of AAs from donors (Figure 1E), its expression was activated in the media of mildly dilated BAV and TAV AAs (Figure 4 and Supplementary Figure S1), with the exclusion of TAV concavity (Figure 4D). ED-A FN in BAV and TAV specimens was localized mainly around degraded ECM and appeared as a blurry staining in the proximity of cells showing different dimension and morphology, ranging from spindle-shaped to round-shaped cells (Figures 1E–1H and 4, and Supplementary Figure S1), possibly related to their heterogeneous origin.

Immunohistochemical analysis of total FN and of the ED-A FN isoform in all sample groups

Figure 4
Immunohistochemical analysis of total FN and of the ED-A FN isoform in all sample groups

Representative images of AA cross-sections hybridized with primary antibody for total FN (first and third column) and for ED-A FN (second and fourth column). Brown staining corresponds to target protein. All images refer to the central region of the media. ED-A FN is not expressed in the media of donor AAs and is activated in dilated aortas, with the exception of TAV AA concavity. Haematoxylin nuclei counterstaining, ×40 magnification, bar represents 20 μm.

Figure 4
Immunohistochemical analysis of total FN and of the ED-A FN isoform in all sample groups

Representative images of AA cross-sections hybridized with primary antibody for total FN (first and third column) and for ED-A FN (second and fourth column). Brown staining corresponds to target protein. All images refer to the central region of the media. ED-A FN is not expressed in the media of donor AAs and is activated in dilated aortas, with the exception of TAV AA concavity. Haematoxylin nuclei counterstaining, ×40 magnification, bar represents 20 μm.

SMCs switch their phenotype in mildly dilated BAV and TAV specimens

Differential RT–PCR gene expression analysis revealed that both the mRNAs coding for smoothelin, an univocal marker of SMC contractile phenotype, and myocardin, a transcriptional co-activator able to modulate the expression of smooth muscle-specific serum response factor target genes, showed a 0.62–0.51-fold decrease in the concavity of BAV and TAV AAs, while only in TAV AAs the myocardin mRNA showed a 0.45-fold decrease also in the convexity (Figure 2). Of note, the smoothelin and myocardin expression at the concavity of TAV AAs showed a significant inverse correlation with aortic diameter (r=−0.74, P<0.05). The immunohistochemical analysis confirmed the RT–PCR data, indicating a decrease of smoothelin-positive cells in dilated AAs (Figure 5). Moreover, both smoothelin-positive and smoothelin-negative SMCs were found to have altered orientation in BAV samples, in both aortic curvatures (Figures 5C and 5F) and, to a lesser extent, also in TAV samples (Figures 5B and 5E).

Immunohistochemical analysis of smoothelin in the concavity and in the convexity of normal and mildly dilated AA

Figure 5
Immunohistochemical analysis of smoothelin in the concavity and in the convexity of normal and mildly dilated AA

Brown staining corresponds to target protein. All images refer to the central region of the media. ×100 Magnification, scale bar represents 20 μm. The inset in the right upper corner of each image represent a ×40 magnification of the central region of the media detailing where the ×100 magnification is taken from. Arrows indicate representative smoothelin-positive SMCs shifting their orientation from a circumferential to a longitudinal direction in BAV and TAV AA samples.

Figure 5
Immunohistochemical analysis of smoothelin in the concavity and in the convexity of normal and mildly dilated AA

Brown staining corresponds to target protein. All images refer to the central region of the media. ×100 Magnification, scale bar represents 20 μm. The inset in the right upper corner of each image represent a ×40 magnification of the central region of the media detailing where the ×100 magnification is taken from. Arrows indicate representative smoothelin-positive SMCs shifting their orientation from a circumferential to a longitudinal direction in BAV and TAV AA samples.

α-SMA mRNA expression showed a significant 1.5-fold increase in BAV convexity (Figure 2), while the immunohistochemical analysis of the corresponding protein, a major component of the cytoskeleton expressed in both SMCs and MFs, showed a non-homogeneous mild increase in the media of dilated AAs (Figure 6), with areas in samples from the aortic convexity featuring a marked cell rarefaction and absence of α-SMA expression (Figures 6F and 6G). α-SMA was also expressed in the intima and in the adventitia in correspondence of large vasa vasorum (results not shown).

Immunohistochemical analysis of α-SMA in the concavity and in the convexity of the three study groups

Figure 6
Immunohistochemical analysis of α-SMA in the concavity and in the convexity of the three study groups

Brown staining corresponds to target protein expression. All images refer to the central region of the media. Haematoxylin nuclei counterstaining, ×20 magnification, scale bar represents 20 μm.

Figure 6
Immunohistochemical analysis of α-SMA in the concavity and in the convexity of the three study groups

Brown staining corresponds to target protein expression. All images refer to the central region of the media. Haematoxylin nuclei counterstaining, ×20 magnification, scale bar represents 20 μm.

DISCUSSION

The original features of the present study include: (i) patient groups homogeneous for the type of haemodynamic disturbance (aortic valve stenosis) and healthy donors rather than surgical patients as reference group [16]; (ii) a comparison between the concavity and the convexity of each aorta, to account for regional variations possibly related to haemodynamics [19]; (iii) focus on mild aortic dilatation, with the aim of highlighting some of the early molecular events underlying its development; (iv) the combined analysis of the expression of a set of molecules potentially informative for SMC and MF phenotypic alterations [20].

Early intimal and medial alterations occur in mild AA dilatations with BAV and TAV stenosis

We highlighted a significantly decreased thickness of the aortic wall in aortic specimens from TAV and BAV patients representative of the early stage of AA dilatation (Table 4), as previously observed also in more advanced stages [11,21]. A parallel increase in cell density suggested that the aortic remodelling phenomena associated with aortic valve stenosis already occur when vessel dilatation is just mild. Of note, the cell rarefaction in subintimal areas of the convexity media has already been reported in aortic aneurysms as well [11].

To our knowledge, the intima alterations we highlighted in mildly dilated BAV and TAV AAs (Figure 1) have not so far been reported in the literature. Conversely, a thicker intima in aneurysms compared with normal aortas has already been described [22], in agreement with our sub-analysis on TAV aneurysms (Figures 1A and 1D).

The restriction of ED-A FN expression to the intima of healthy aortas (Figures 1 and 2) is in agreement with previous studies [23], whereas the reduced ED-A FN and SFSWAP mRNA levels in mildly dilated samples is consistent with the loss of the ED-A FN protein isoform in the altered intima revealed by the immunohistochemical analysis. The previously unreported activation of ED-A FN synthesis in the media of mildly dilated AAs, observed in the present study, is probably not fully compensatory, at mRNA level, for the loss in the altered intima (Figure 4). This observation confirms the utility of a combined molecular and immunohistochemical approach to catch all the phenomena involved in the early phase of this pathophysiological process, also considering that ECM proteins accumulate in tissues, and consequently mRNA levels may not correspond to the actual protein levels.

On an observational basis, we can speculate that the intima markedly reduces its thickness during early dilatation; nonetheless, we cannot exclude a massive synthesis of elastin in the intima by resident cells, leading to the loss of a clear distinction between the intima and the media. In either case, these results underscore the importance of the previously neglected involvement of the intima in the early phases of the aortopathy associated with TAV and BAV stenosis.

Imbalance in TGFβ receptor expression and activation of FN alternative splicing occur in mild post-stenotic dilatations

TGFβ is a pleiotropic cytokine synthesized as a precursor and complexed with LTBP-1 in the ECM. TGFβ is activated when LTBP-1 is removed extracellularly via proteolytic cleavage or through a mechanical stimulus [24]. Activated TGFβ binds to a heterodimeric receptor complex comprising two subunits (TGFβR1 and TGFβR2), endowed with a tyrosine kinase activity. TGFβ drives the alternative splicing of FN in response to external stimuli and, in turn, ED-A FN and TGFβ, in the presence of a mechanical stress, promote the up-regulation of α-SMA and the differentiation of MFs [6,25]. Others reported that the rate of ED-A exon inclusion in FN increased with increasing aneurysm diameter in TAV patients, while this correlation was impaired in BAV patients [9]. The evidence of activation of ED-A FN expression in the media of our BAV AA specimens argues against those authors’ hypothesis of a constitutive defect of FN splicing in BAV aortopathy. However, in their studies BAV and TAV patient groups were not comparable for aortic valve functional status [9], thus possibly explaining the discrepancy, as stenosis and regurgitation are known to be associated with diverging patterns of aortic remodelling during aneurismal dilatation [10,11]. The significant difference in ED-A FN mRNA levels between convexity and concavity in BAV AAs (with preserved expression in the concavity) could be related to the typical asymmetric BAV-related flow and wall stress patterns, as already demonstrated through biomechanical studies both by our group and by others [3,14,26].

We observed an increase of TGFβ mRNA in mild AA dilatations (Figure 2), while others showed an increase of TGFβ protein in frankly aneurismal aortas with no increase of its mRNA [27], possibly implying different mechanisms of regulation depending on the stage of the aortopathy.

The increased TGFβR2 mRNA expression observed in this study seems not specific to the pathogenesis of BAV aortopathy, as it has been reported in disparate settings [28,29], possibly representing an epiphenomenon of the aortic dilative remodelling.

Conversely, the decreased expression of TGFβR1 mRNA in both the curvatures of BAV AAs only, and the positive correlation of its mRNA levels with aortic diameter, could together suggest that TGFβR1 down-regulation could be a very early event in BAV-related aortopathy, and that patients with BAV stenosis who develop post-stenotic dilatation of the AA may have an intrinsic defect of TGFβR1 expression. Most importantly, the altered imbalance between the TGFβR1 and the TGFβR2 subunits, particularly evident in BAV AAs, are in agreement with data obtained in a murine model of aortic aneurysm and could trigger non-canonical TGFβ-mediated signalling pathways, leading to ECM degradation and aneurysm [6,28]. A relationship between the expression levels of TGFβR1 and TGFβR2 and mutations in the corresponding genes cannot be ruled out, but it is improbable, as it has been demonstrated that mutations in TGFβR1 and TGFβR2 are rarely present in sporadic BAV patients [30].

Early SMC phenotypic alteration and differentiation of a MF-like cell lineage likely occur in AA dilatations with BAV and TAV stenosis

Alterations of SMC phenotype have been described previously both in experimentally induced aneurysms [31] and in patients with AA disease of different aetiologies [32], thus appearing as a common pathogenetic step for distinct diseases. Smoothelin is a univocal marker of SMC contractile phenotype [33] that allows distinguishing contractile SMCs from other cells with smooth muscle-like characteristics, including MFs. Its deficiency results in a considerable loss of contractile potential and hence in impaired smooth muscle function. To our knowledge, smoothelin differential expression in AA samples from BAV and TAV patients has never been evaluated before. Despite the above-mentioned increased cell density, the smoothelin decrease occurred in both BAV and TAV AAs, but it reached the significance only at the concavity, possibly indicating that the greater haemodynamic burden acting on the AA convexity [3,14] could at least in part preserve the contractile phenotype of SMCs [34]. We also observed a change in SMC orientation, from circumferential to longitudinal direction, already known to occur in vessels submitted to altered tensile strain and stress [35] and in the more mature phases of aneurysm development [21]. The greater change in SMC orientation in BAV compared with TAV AA samples is in agreement with the notion that, for equal extents of stenosis, the BAV induces more marked flow derangement and shear stress anomalies than TAV [3,36]. It remains to be clarified whether SMC orientation changes contribute to alter the mechanical properties of the dilated aortic wall.

The increased α-SMA expression in mild BAV AA dilatations (Figures 2 and 6) could be causally related to the parallel increase of TGFβ expression and be possibly suggestive of MF differentiation, given the concomitant decrease of smoothelin expression.

The convexity of mild post-stenotic AA dilatations shows altered ECM homoeostasis

Our data showed an early increase of MMP-2 expression in both the TAV and BAV AA convexity (Figure 3), where the haemodynamic stress is known to be greater [3]. This is in agreement with the previously reported increased expression of MMP-2 in SMCs submitted to cyclic mechanical stretch [37] as well as with the activation of the MMP-2 promoter as a function of wall tension [38]. TIMP-1 mRNA level did not differ between BAV and control aorta, in agreement with others [39], while TIMP-2 mRNA showed a significant increase only in TAV patients (Figure 3), suggesting a possible attempt to preserve ECM homoeostasis. These findings are only partially in agreement with those obtained in unselected BAV patients with larger aortic diameters [40].

The absence of MMP-9 expression was probably related to the low-inflammatory infiltrates at the early stage of aortic dilatation, as this enzyme is known to be a major product of macrophages, neutrophils and SMCs with a pro-inflammatory phenotype [41].

Study limitations

We are aware that, because of the characteristics of the material we had at our disposal, we were limited to the use of histological and expression analysis techniques. Isolation and culturing of vascular cells from diseased aortic specimens would allow a broader range of analytical experiments, but results can be affected by changes in biochemical, physical and haemodynamic cell environment. Our choice of retrieving the control aortic specimens from healthy donors rather than from surgical patients with other cardiac diseases has evident advantages on the reliability of results but affected the mean age in the control group that was inevitably younger than in patients. Consequently, it cannot be excluded that some of the changes observed in mild aortic dilatations may reflect aging rather than dilative aortopathy. Similarly, the different mean age in BAV and TAV groups was related to the epidemiology of aortic valve stenosis. However, the normal aging process of the aorta is known to be associated with increase in TGFβ signalling, increased expression of the receptor subunit 2 of TGFβ, increased FN and MMP-2 expression [29]. We found all these changes being more marked and more constant in the younger BAV group than in the older TAV group; therefore the differences observed between patient groups might have been rather reduced than emphasized by a possible interference of age.

The present study did not include the investigation of the Smad pathway (e.g. phosphorylation of Smad2) for two reasons: (i) the main driver in aortic diseases implying TGFβ signalling dysregulation has been authoritatively demonstrated to be the non-canonical, Smad-independent, pathway [28,42]; (ii) it has been shown that an aspecific phosphorylation of Smad2 occurs in aortic dilative disease, which is independent of TGFβ signalling and irrespective of the underlying aetiology [27,43], and could be secondary to the flow changes that are entailed by the increase in vessel radius [44].

Conclusions

We have observed significant mRNA expression changes in five out of the 13 genes we analysed in TAV AAs, compared with ten out of 13 the genes significantly altered in BAV AAs. These data could be related both to intrinsic differences between TAV and BAV AAs, both to a greater variability of the expression of some genes in TAV in comparison with BAV AAs, thus suggesting a greater heterogeneity of the underlying abnormalities in TAV AA specimens. Among the results we obtained, the defect of TGFβR1 mRNA expression in both concavity and convexity was a peculiarity of patients with BAV stenosis, and likely independent of different local stress levels. Additional studies should address the post-receptorial TGFβ signalling pathways specifically involved in BAV aortopathy to evaluate the consequences of this likely intrinsic defect of TGFβR1 expression. It could be hypothesized that this defect, along with the up-regulation of TGFβ and TGFβR2 expression observed in both BAV and TAV stenosis, may determine a redirection of the TGFβ signalling towards non-canonical effects.

Overall data suggest that the alterations of the intimal structure, the cell loss in the subintimal media and the SMC phenotype switch possibly leading to the emergence of differentiating MFs, are concomitant events occurring early in the development of both TAV and BAV AA dilatations, with some events more evident in BAV convexity, possibly suggesting an intraluminal factor (e.g. shear stress) triggering a maladaptive remodelling. Whether a putative concomitant genetic defect leading to disturbed expression of TGFβ receptors predisposes BAV patients to a greater susceptibility to the triggering factor should be object of further investigation.

FUNDING

This work was supported in part by PRIN 2007 from the Italian Ministry of University and Research [grant no. 2007KA2HSZ (to M.C.)] and in part by the Italian Ministry of Health [grant number GR-2009-1580434 (‘Ricerca Sanitaria–GR2009’ grant) (to A.D.C)]. C.B. was supported by the Ph.D. Program ‘Medical and Surgical Physiopathology of the Cardio-respiratory System and Associated Biotechnologies’, Second University of Naples.

AUTHOR CONTRIBUTION

Amalia Forte conceived and designed the study, carried out the experiments, interpreted the results and wrote the paper; Alessandro Della Corte conceived and designed the study, collected samples during surgery, interpreted the results and contributed to writing the paper; Mario Grossi and Mauro Finicelli carried out the experiments and collected literature; Ciro Bancone and Rafaela Provenzano collected the data and literature and contributed to patient enrolment; Marisa De Feo and Luca De Santo contributed to patient/control enrolment, sample collection at surgery and figure generation; Gianantonio Nappi and Maurizio Cotrufo collected samples at surgery and contributed to study design and data interpretation, Umberto Galderisi and Marilena Cipollaro contributed to study design, experiments and writing.

We thank Dr Ciro Maiello (Department of Cardiovascular Surgery and Transplant at the V. Monaldi Hospital) for providing the control AA specimens, and M. Rosaria Cipollaro for her assistance.

Abbreviations

     
  • AA

    ascending aorta

  •  
  • BAV

    bicuspid aortic valve

  •  
  • EC

    endothelial cell

  •  
  • ECM

    extracellular matrix

  •  
  • ED-A

    extra domain-A

  •  
  • FN

    fibronectin

  •  
  • GAPDH

    glyceraldehyde-3-phosphate dehydrogenase

  •  
  • LTBP-1

    latent TGFβ-binding protein 1

  •  
  • MF

    myofibroblast

  •  
  • MMP

    matrix metalloproteinase

  •  
  • RT–PCR

    reverse transcription–PCR

  •  
  • SMC

    smooth muscle cell

  •  
  • TAV

    tricuspid aortic valve

  •  
  • TGFβ

    transforming growth factor β

  •  
  • TGFβR

    TGFβ receptor

  •  
  • TIMP

    tissue inhibitor of metalloproteinases

  •  
  • vWF

    von Willebrand factor

  •  
  • α-SMA

    α-smooth muscle actin

References

References
1
Michelena
 
H. I.
Khanna
 
A. D.
Mahoney
 
D.
Margaryan
 
E.
Topilsky
 
Y.
Suri
 
R. M.
Eidem
 
B.
Edwards
 
W. D.
Sundt
 
T. M.
Enriquez-Sarano
 
M.
 
Incidence of aortic complications in patients with bicuspid aortic valves
JAMA, J. Am. Med. Assoc.
2011
, vol. 
306
 (pg. 
1104
-
1112
)
2
Della Corte
 
A.
Bancone
 
C.
Quarto
 
C.
Dialetto
 
G.
Covino
 
F. E.
Scardone
 
M.
Caianiello
 
G.
Cotrufo
 
M.
 
Predictors of ascending aortic dilatation with bicuspid aortic valve: a wide spectrum of disease expression
Eur. J. Cardiothorac. Surg.
2007
, vol. 
31
 (pg. 
397
-
404
)
3
Della Corte
 
A.
Bancone
 
C.
Conti
 
C. A.
Votta
 
E.
Redaelli
 
A.
Del Viscovo
 
L.
Cotrufo
 
M.
 
Restricted cusp motion in right-left type of bicuspid aortic valves: a new risk marker for aortopathy
J. Thorac. Cardiovasc. Surg.
2012
, vol. 
144
 (pg. 
360
-
369.e1
)
4
Girdauskas
 
E.
Borger
 
M. A.
Secknus
 
M. A.
Girdauskas
 
G.
Kuntze
 
T.
 
Is aortopathy in bicuspid aortic valve disease a congenital defect or a result of abnormal hemodynamics? A critical reappraisal of a one-sided argument
Eur. J. Cardiothorac. Surg.
2011
, vol. 
39
 (pg. 
809
-
814
)
5
Olivieri
 
J.
Smaldone
 
S.
Ramirez
 
F.
 
Fibrillin assemblies: extracellular determinants of tissue formation and fibrosis
Fibrogenesis Tissue Repair
2010
, vol. 
3
 pg. 
24
 
6
Forte
 
A.
Della Corte
 
A.
De Feo
 
M.
Cerasuolo
 
F.
Cipollaro
 
M.
 
Role of myofibroblasts in vascular remodelling: focus on restenosis and aneurysm
Cardiovasc. Res.
2010
, vol. 
88
 (pg. 
395
-
405
)
7
Muro
 
A. F.
Chauhan
 
A. K.
Gajovic
 
S.
Iaconcig
 
A.
Porro
 
F.
Stanta
 
G.
Baralle
 
F. E.
 
Regulated splicing of the fibronectin EDA exon is essential for proper skin wound healing and normal lifespan
J. Cell. Biol.
2003
, vol. 
162
 (pg. 
149
-
160
)
8
Jones
 
J. A.
Beck
 
C.
Barbour
 
J. R.
Zavadzkas
 
J. A.
Mukherjee
 
R.
Spinale
 
F. G.
Ikonomidis
 
J. S.
 
Alterations in aortic cellular constituents during thoracic aortic aneurysm development: myofibroblast-mediated vascular remodeling
Am. J. Pathol.
2009
, vol. 
175
 (pg. 
1746
-
1756
)
9
Paloschi
 
V.
Kurtovic
 
S.
Folkersen
 
L.
Gomez
 
D.
Wagsater
 
D.
Roy
 
J.
Petrini
 
J.
Eriksson
 
M. J.
Caidahl
 
K.
Hamsten
 
A.
, et al 
Impaired splicing of fibronectin is associated with thoracic aortic aneurysm formation in patients with bicuspid aortic valve
Arterioscler. Thromb. Vasc. Biol.
2011
, vol. 
31
 (pg. 
691
-
697
)
10
Roberts
 
W. C.
Vowels
 
T. J.
Ko
 
J. M.
Filardo
 
G.
Hebeler
 
R. F.
Henry
 
A. C.
Matter
 
G. J.
Hamman
 
B. L.
 
Comparison of the structure of the aortic valve and ascending aorta in adults having aortic valve replacement for aortic stenosis versus for pure aortic regurgitation and resection of the ascending aorta for aneurysm
Circulation
2011
, vol. 
123
 (pg. 
896
-
903
)
11
Cotrufo
 
M.
Della Corte
 
A.
De Santo
 
L. S.
Quarto
 
C.
De Feo
 
M.
Romano
 
G.
Amarelli
 
C.
Scardone
 
M.
Di Meglio
 
F.
Guerra
 
G.
, et al 
Different patterns of extracellular matrix protein expression in the convexity and the concavity of the dilated aorta with bicuspid aortic valve: preliminary results
J. Thorac. Cardiovasc. Surg.
2005
, vol. 
130
 (pg. 
504
-
511
)
12
Della Corte
 
A.
De Santo
 
L. S.
Montagnani
 
S.
Quarto
 
C.
Romano
 
G.
Amarelli
 
C.
Scardone
 
M.
De Feo
 
M.
Cotrufo
 
M.
Caianiello
 
G.
 
Spatial patterns of matrix protein expression in dilated ascending aorta with aortic regurgitation: congenital bicuspid valve versus Marfan's syndrome
J. Heart Valve Dis.
2006
, vol. 
15
 (pg. 
20
-
27
)
13
Della Corte
 
A.
Quarto
 
C.
Bancone
 
C.
Castaldo
 
C.
Di Meglio
 
F.
Nurzynska
 
D.
De Santo
 
L. S.
De Feo
 
M.
Scardone
 
M.
Montagnani
 
S.
Cotrufo
 
M.
 
Spatiotemporal patterns of smooth muscle cell changes in ascending aortic dilatation with bicuspid and tricuspid aortic valve stenosis: focus on cell-matrix signaling
J. Thorac. Cardiovasc. Surg.
2008
, vol. 
135
 (pg. 
8
-
18
)
14
Conti
 
C. A.
Della Corte
 
A.
Votta
 
E.
Del Viscovo
 
L.
Bancone
 
C.
De Santo
 
L. S.
Redaelli
 
A.
 
Biomechanical implications of the congenital bicuspid aortic valve: a finite element study of aortic root function from in vivo data
J. Thorac. Cardiovasc. Surg.
2010
, vol. 
140
 (pg. 
890
-
896
)
15
Svensson
 
L. G.
Kim
 
K. H.
Lytle
 
B. W.
Cosgrove
 
D. M.
 
Relationship of aortic cross-sectional area to height ratio and the risk of aortic dissection in patients with bicuspid aortic valves
J. Thorac. Cardiovasc. Surg.
2003
, vol. 
126
 (pg. 
892
-
893
)
16
Forte
 
A.
Della Corte
 
A.
Grossi
 
M.
Finicelli
 
M.
Bancone
 
C.
Provenzano
 
R.
Pepino
 
P.
Nappi
 
G. A.
De Feo
 
M.
Galderisi
 
U.
, et al 
Morphological and molecular characterization of healthy human ascending aorta
Histol. Histopathol.
2012
, vol. 
27
 (pg. 
103
-
112
)
17
Jones
 
J. A.
Ruddy
 
J. M.
Bouges
 
S.
Zavadzkas
 
J. A.
Brinsa
 
T. A.
Stroud
 
R. E.
Mukherjee
 
R.
Spinale
 
F. G.
Ikonomidis
 
J. S.
 
Alterations in membrane type-1 matrix metalloproteinase abundance after the induction of thoracic aortic aneurysm in a murine model
Am. J. Physiol. Heart Circ. Physiol.
2010
, vol. 
299
 (pg. 
H114
-
H124
)
18
Jones
 
J. A.
Zavadzkas
 
J. A.
Chang
 
E. I.
Sheats
 
N.
Koval
 
C.
Stroud
 
R. E.
Spinale
 
F. G.
Ikonomidis
 
J. S.
 
Cellular phenotype transformation occurs during thoracic aortic aneurysm development
J. Thorac. Cardiovasc. Surg.
2010
, vol. 
140
 (pg. 
653
-
659
)
19
Della Corte
 
A.
Romano
 
G.
Tizzano
 
F.
Amarelli
 
C.
De Santo
 
L. S.
De Feo
 
M.
Scardone
 
M.
Dialetto
 
G.
Covino
 
F. E.
Cotrufo
 
M.
 
Echocardiographic anatomy of ascending aorta dilatation: correlations with aortic valve morphology and function
Int. J. Cardiol.
2006
, vol. 
113
 (pg. 
320
-
326
)
20
Hao
 
H.
Gabbiani
 
G.
Camenzind
 
E.
Bacchetta
 
M.
Virmani
 
R.
Bochaton-Piallat
 
M. L.
 
Phenotypic modulation of intima and media smooth muscle cells in fatal cases of coronary artery lesion
Arterioscler. Thromb. Vasc. Biol.
2006
, vol. 
26
 (pg. 
326
-
332
)
21
Kirsch
 
E. W.
Radu
 
N. C.
Gervais
 
M.
Allaire
 
E.
Loisance
 
D. Y.
 
Heterogeneity in the remodeling of aneurysms of the ascending aorta with tricuspid aortic valves
J. Thorac. Cardiovasc. Surg.
2006
, vol. 
132
 (pg. 
1010
-
1016
)
22
Tang
 
P. C.
Coady
 
M. A.
Lovoulos
 
C.
Dardik
 
A.
Aslan
 
M.
Elefteriades
 
J. A.
Tellides
 
G.
 
Hyperplastic cellular remodeling of the media in ascending thoracic aortic aneurysms
Circulation.
2005
, vol. 
112
 (pg. 
1098
-
1105
)
23
Glukhova
 
M. A.
Frid
 
M. G.
Koteliansky
 
V. E.
 
Phenotypic changes of human aortic smooth muscle cells during development and in the adult vessel
Am. J. Physiol.
1991
, vol. 
261
 (pg. 
78
-
80
)
24
Hinz
 
B.
 
Tissue stiffness, latent TGF-β1 activation and mechanical signal transduction: implications for the pathogenesis and treatment of fibrosis
Curr. Rheumatol. Rep.
2009
, vol. 
11
 (pg. 
120
-
126
)
25
Hinz
 
B.
 
Formation and function of the myofibroblast during tissue repair
J. Invest. Dermatol.
2007
, vol. 
127
 (pg. 
526
-
537
)
26
Viscardi
 
F.
Vergara
 
C.
Antiga
 
L.
Merelli
 
S.
Veneziani
 
A.
Puppini
 
G.
Faggian
 
G.
Mazzucco
 
A.
Luciani
 
G. B.
 
Comparative finite element model analysis of ascending aortic flow in bicuspid and tricuspid aortic valve
Artif. Organs.
, vol. 
34
 (pg. 
1114
-
1120
)
27
Gomez
 
D.
Al Haj Zen
 
A.
Borges
 
L. F.
Philippe
 
M.
Gutierrez
 
P. S.
Jondeau
 
G.
Michel
 
J. B.
Vranckx
 
R.
 
Syndromic and non-syndromic aneurysms of the human ascending aorta share activation of the Smad2 pathway
J. Pathol.
2009
, vol. 
218
 (pg. 
131
-
142
)
28
Jones
 
J. A.
Barbour
 
J. R.
Stroud
 
R. E.
Bouges
 
S.
Stephens
 
S. L.
Spinale
 
F. G.
Ikonomidis
 
J. S.
 
Altered transforming growth factor-β signaling in a murine model of thoracic aortic aneurysm
J. Vasc. Res.
2008
, vol. 
45
 (pg. 
457
-
468
)
29
Wang
 
M.
Zhao
 
D.
Spinetti
 
G.
Zhang
 
J.
Jiang
 
L. Q.
Pintus
 
G.
Monticone
 
R.
Lakatta
 
E. G.
 
Matrix metalloproteinase 2 activation of transforming growth factor-β1 (TGF-β1) and TGF-β1-type II receptor signaling within the aged arterial wall
Arterioscler. Thromb. Vasc. Biol.
2006
, vol. 
26
 (pg. 
1503
-
1509
)
30
Arrington
 
C. B.
Sower
 
C. T.
Chuckwuk
 
N.
Stevens
 
J.
Leppert
 
M. F.
Yetman
 
A. T.
Bowles
 
N. E.
 
Absence of TGFβR1 and TGFβR2 mutations in patients with bicuspid aortic valve and aortic dilation
Am. J. Cardiol.
2008
, vol. 
102
 (pg. 
629
-
631
)
31
Ailawadi
 
G.
Moehle
 
C. W.
Pei
 
H.
Walton
 
S. P.
Yang
 
Z.
Kron
 
I. L.
Lau
 
C. L.
Owens
 
G. K.
 
Smooth muscle phenotypic modulation is an early event in aortic aneurysms
J. Thorac. Cardiovasc. Surg.
2009
, vol. 
138
 (pg. 
1392
-
1399
)
32
Lesauskaite
 
V.
Tanganelli
 
P.
Sassi
 
C.
Neri
 
E.
Diciolla
 
F.
Ivanoviene
 
L.
Epistolato
 
M. C.
Lalinga
 
A. V.
Alessandrini
 
C.
Spina
 
D.
 
Smooth muscle cells of the media in the dilatative pathology of ascending thoracic aorta: morphology, immunoreactivity for osteopontin, matrix metalloproteinases and their inhibitors
Hum. Pathol.
2001
, vol. 
32
 (pg. 
1003
-
1011
)
33
van Eys
 
G. J.
Niessen
 
P. M.
Rensen
 
S. S.
 
Smoothelin in vascular smooth muscle cells
Trends Cardiovasc. Med.
2007
, vol. 
17
 (pg. 
26
-
30
)
34
Stegemann
 
J. P.
Hong
 
H.
Nerem
 
R. M.
 
Mechanical, biochemical and extracellular matrix effects on vascular smooth muscle cell phenotype
J. Appl. Physiol.
2005
, vol. 
98
 (pg. 
2321
-
2327
)
35
Liu
 
S. Q.
 
Influence of tensile strain on smooth muscle cell orientation in rat blood vessels
J. Biomech. Eng.
1998
, vol. 
120
 (pg. 
313
-
320
)
36
Richards
 
K. E.
Deserranno
 
D.
Donal
 
E.
Greenberg
 
N. L.
Thomas
 
J. D.
Garcia
 
M. J.
 
Influence of structural geometry on the severity of bicuspid aortic stenosis
Am. J. Physiol. Heart Circ. Physiol.
2004
, vol. 
287
 (pg. 
H1410
-
416
)
37
Grote
 
K.
Flach
 
I.
Luchtefeld
 
M.
Akin
 
E.
Holland
 
S. M.
Drexler
 
H.
Schieffer
 
B.
 
Mechanical stretch enhances mRNA expression and proenzyme release of matrix metalloproteinase-2 (MMP-2) via NAD(P)H oxidase-derived reactive oxygen species
Circ. Res.
2003
, vol. 
92
 (pg. 
e80
-
e86
)
38
Ruddy
 
J. M.
Jones
 
J. A.
Stroud
 
R. E.
Mukherjee
 
R.
Spinale
 
F. G.
Ikonomidis
 
J. S.
 
Differential effect of wall tension on matrix metalloproteinase promoter activation in the thoracic aorta
J. Surg. Res.
2010
, vol. 
160
 (pg. 
333
-
339
)
39
LeMaire
 
S. A.
Wang
 
X.
Wilks
 
J. A.
Carter
 
S. A.
Wen
 
S.
Won
 
T.
Leonardelli
 
D.
Anand
 
G.
Conklin
 
L. D.
Wang
 
X. L.
, et al 
Matrix metalloproteinases in ascending aortic aneurysms: bicuspid versus trileaflet aortic valves
J. Surg. Res.
2005
, vol. 
123
 (pg. 
40
-
48
)
40
Ikonomidis
 
J. S.
Jones
 
J. A.
Barbour
 
J. R.
Stroud
 
R. E.
Clark
 
L. L.
Kaplan
 
B. S.
Zeeshan
 
A.
Bavaria
 
J. E.
Gorman
 
J. H.
Spinale
 
F. G.
Gorman
 
R. C.
 
Expression of matrix metalloproteinases and endogenous inhibitors within ascending aortic aneurysms of patients with bicuspid or tricuspid aortic valves
J. Thorac. Cardiovasc. Surg.
2007
, vol. 
133
 (pg. 
1028
-
1036
)
41
Li
 
H.
Xu
 
H.
Liu
 
S.
 
Toll-like receptors 4 induces expression of matrix metalloproteinase-9 in human aortic smooth muscle cells
Mol. Biol. Rep.
2011
, vol. 
38
 (pg. 
1419
-
1423
)
42
Holm
 
T. M.
Habashi
 
J. P.
Doyle
 
J. J.
Bedja
 
D.
Chen
 
Y.
van Erp
 
C.
Lindsay
 
M. E.
Kim
 
D.
Schoenhoff
 
F.
Cohn
 
R. D.
, et al 
Noncanonical TGFbeta signaling contributes to aortic aneurysm progression in Marfan syndrome mice
Science
2011
, vol. 
332
 (pg. 
358
-
361
)
43
Gomez
 
D.
Coyet
 
A.
Ollivier
 
V.
Jeunemaitre
 
X.
Jondeau
 
G.
Michel
 
J. B.
Vranckx
 
R.
 
Epigenetic control of vascular smooth muscle cells in Marfan and non-Marfan thoracic aortic aneurysms
Cardiovasc. Res.
2011
, vol. 
89
 (pg. 
446
-
456
)
44
Shepherd
 
R. D.
Kos
 
S. M.
Rinker
 
K. D.
 
Flow-dependent Smad2 phosphorylation and TGIF nuclear localization in human aortic endothelial cells
Am. J. Physiol. Heart Circ. Physiol.
2011
, vol. 
301
 (pg. 
H98
-
H107
)

Supplementary data