Layer-specific Nos3 expression and genotypic distribution in bicuspid aortic valve aortopathy

Abstract

OBJECTIVES

We hypothesized that expression and activity of nitric oxide synthase-3 enzyme (Nos3) in bicuspid aortic valve (BAV) aortopathy are related to tissue layer and Nos3 genotype.

METHODS

Gene expression of Nos3 and platelet and endothelial cell adhesion molecule-1 (Pecam1) and NOS activity were measured in intima-containing media and adventitial specimens of ascending aortic tissue. The presence of 2 Nos3 single-nucleotide polymorphisms (SNPs; −786T/C and 894G/T) was determined for non-aneurysmal (NA) and aneurysmal patients with BAV (n = 40, 89, respectively); patients with tricuspid aortic valve (TAV) and aneurysm (n = 151); and NA patients with TAV (n = 100).

RESULTS

Elevated Nos3 relative to Pecam1 and reduced Pecam1 relative to a housekeeping gene were observed within intima-containing aortic specimens from BAV patients when compared with TAV patients. Lower Nos3 in the adventitia of aneurysmal specimens was noted when compared with specimens of NA aorta, independent of valve morphology. NOS activity was similar among cohorts in media/intima and decreased in the diseased adventitia, relative to control patients. Aneurysmal BAV patients exhibited an under-representation of the wild-type genotype for −786 SNP. No differences in genotype distribution were noted for 894 SNP. Primary intimal endothelial cells from patients with at least 1 C allele at −786 SNP exhibited lower Nos3 when compared with wild-type cells.

CONCLUSIONS

These findings of differential Nos3 in media/intima versus adventitia depending on valve morphology or aneurysm reveal new information regarding aneurysmal pathophysiology and support our ongoing assertion that there are distinct mechanisms giving rise to ascending aortopathy in BAV and TAV patients.

It remains incompletely understood how and why aneurysm in the proximal ascending aorta frequently arises amidst a backdrop of bicuspid aortic valve (BAV).

INTRODUCTION

It remains incompletely understood how and why aneurysm in the proximal ascending aorta frequently arises amidst a backdrop of bicuspid aortic valve (BAV). The prevalence of BAV within first degree relatives of an affected patient has been reported to be 5% and 9%, with a BAV heritability ratio reported as high as 0.89 [1–3]. BAV morphotypes’ heterogeneity seen in clinical practices of large surgical centres [4, 5] and the existence of different BAV morphotypes within families [1] exemplify that family history is not the only risk factor. The array of genes identified in patients affected with thoracic aortic aneurysm (TAA) and dissection [6–8] highlights the complexity of the disease and complicates interpretations from genetically modified models.

Murine models have revealed that insufficiency of several genes [9–11], including nitric oxide synthase 3 (Nos3), results in partially penetrant BAV [12]. Murine Nos3 knockout models reflect a different spectrum of BAV morphotypes than what is observed in humans, primarily arising as the type 1 R/N morphotype [4, 5]. Unfortunately, these animal models imperfectly recapitulate human BAV-associated aortopathy and have not yet uncovered a putative genetic basis for both the BAV and the associated aneurysm in the proximal ascending aorta. Therefore, studies that apply human tissue and cell-based strategies to establish the functional relevance of genetic polymorphisms in BAV could help to identify biological and/or biomechanical mechanisms that contribute to the pathophysiology of the associated aortopathy.

There have been discordant findings regarding Nos3 expression in aortic specimens from BAV patients [13–17] including findings by our group of increased Nos3 expression in aortic media/intima specimens from aneurysmal patients with BAV [18]. Prior studies implicated Nos3 genotype in BAV-associated aortopathy by demonstrating an over-representation of the wild-type single-nucleotide polymorphism (SNP) in the Nos3 promoter (−786T/C) [19] and a related influence of this SNP on Nos3 expression in vitro [20]. We hypothesized that a genetic alteration in Nos3 alters its expression in BAV-associated aortopathy.

METHODS

Ethics statement

Ascending aortic specimens were procured with approval of the University of Pittsburgh Institutional Review Board (protocol # PRO07020120, first approved 10 July 2007) and with formal written patient consent obtained.

Specimen and blood collection

Inclusion criteria for the study were patients undergoing elective aortic valve and/or ascending aortic replacement due to aortic stenosis or insufficiency and/or aneurysm or heart transplantation. Aortic specimens were also procured from organ donors via the Center for Organ Recovery and Education. Specimens were classified as non-aneurysmal (NA) by the attending surgeon author (Thomas G. Gleason) according to the maximal orthogonal diameter of the ascending aorta measuring ≤42 mm from computed tomographic angiography. Patients with known connective tissue disorders (e.g. Marfan, Ehlers–Danlos, Loeys–Dietz syndromes) or undocumented aortic valve morphotype were excluded. Blood specimens collected from patients without aortopathy (n = 90) and aortic media specimens from heart transplant donors (n = 10) and exhibiting tricuspid aortic valve (TAV) served as controls for bioassays and genotype analysis. Demographics for all patients are displayed in Supplementary Material, Table S1. Specimens were either preserved in RNAlater solution (Invitrogen Thermo Fisher Scientific, Waltham, MA, USA) at −20°C or snap frozen in liquid nitrogen and stored at −80°C.

Blood samples were collected immediately prior to surgical intervention. Upon retrieval, the ‘buffy coat’ layer containing peripheral blood mononuclear cells was collected via centrifugation with Ficoll (Fisher Scientific) and stored long term in liquid nitrogen.

Endothelial cell isolation and culture

Human intimal endothelial cells (n = 20) were isolated by placing the aortic specimen intima-side down in a Petri dish with Dubeccos Modified Essential Medium (DMEM) supplemented with 2.5 mg/ml collagenase type IV and then incubating for 15 min in at 37°C. The intimal surface was then scraped to remove detached endothelial cells, pelleted and plated in Endothelial Cell Growth Medium (Cell Applications, San Diego, CA, USA) on bovine type I collagen-coated tissue culture flasks and then expanded for 2–3 passages. Cell populations were enriched for CD31⁺ cells using a magnetic bead separation technique during initial propagation and expansion, and prior to experimentation. Confluent cells were serum-derived for 24 hours and resuspended in Buffer RA1 (Takara Bio USA, Mountain View, CA, USA) with 0.01% β-mercaptoethanol (MilliporeSigma, Burlington, MA, USA) for storage in −80°C for later gene expression analysis.

Gene expression

RNA from aortic media/intima (N = 71) was isolated as previously described [18]. Adventitial specimens (N = 76) were incubated in excess nuclease-free water for 1–4 h on ice to remove red blood cells before RNA was extracted using the Nucleospin RNA kit with manufacturer’s recommended fibrous tissue modifications (Takara Bio USA). cDNA synthesis was completed using the High-Capacity cDNA Reverse Transcription kit and quantified with the Qubit 2.0. qPCR was performed on 100 ng of media/intima cDNA input or 10 ng of adventitial cDNA using Taqman™ technology (Thermo Fisher Scientific) for Nos3 and normalized to either Pecam1 or Ppia gene expression (Supplementary Material, Table S2). In addition, Pecam1 expression was independently evaluated relative to Ppia gene expression for each layer of the aortic wall.

RNA was isolated from human primary aortic intimal endothelial cells (N = 11) using the Nucleospin RNA kit (Takara Bio USA). cDNA synthesis and quantiative real-time polymerase chain reaction (qPCR) were performed as described above using 10 ng cDNA input for Nos3 and normalized to Ppia. qPCR for all samples was performed on an ABStepONEPlus™ Real-Time PCR System (Thermo Fisher Scientific). Data analysis was performed on StepOnePlus™ software v2.2.2, and relative target gene expression was determined using the 2−^ΔCt method.

DNA isolation and genotyping

The Nos3 SNP genotypes were determined for a subset of patients listed in Supplementary Material, Table S1 who self-identified as Caucasian (N = 267). DNA was isolated from peripheral blood mononuclear cells (control patients) or snap-frozen aortic specimens using the NucleoSpin Tissue kit (Takara Bio USA). Genotyping of each SNP was completed using Taqman™ Genotyping technology (Thermo Fisher Scientific) with SNP assay primer/probe sets [# rs1799983 (894G/T), rs2070744(−786T/C)]. Data analysis was performed using StepOnePlus™ software v2.2.2.

Control group validation

The distribution of genotypes for the control group was compared with European allele frequencies obtained by the 1000 Genomes Project Phase 3 [21], where −786T/C was reported to have allele frequencies of T = 56% and C = 44% and 894G/T was reported to have allele frequencies of G = 66% and T = 34%. The control group was determined to be within Hardy–Weinberg equilibrium (HWE), as described below in the ‘Statistical Analysis’ section for both Nos3 SNPs (−786T/C χ² = 2.923, 894 G/T χ² = 0.243), and therefore was deemed suitable for comparison with pathological groups.

NOS activity

Snap-frozen specimens of aortic media/intima (N = 56) or adventitia (N = 71) were homogenized, and 50 µg samples were utilized for a NOS activity assay using a commercial kit (Cayman Chemical, Ann Arbor, MI, USA) per the manufacturer’s instructions. Briefly, conversion of radiolabeled [¹⁴C] arginine to citrulline was measured in the presence of requisite NOS substrates as previously reported [22]. NOS activity was quantified as counts per minute (CPM) per milligram of protein. Purified inducible NOS was utilized as a positive control for the NOS activity assay.

Statistical analyses

Analyses were completed using SigmaPlot 14 (SysStat Software Inc, San Jose, CA, USA). Distribution of genotypes in pathological groups was compared to our validated control group using HWE when all genotypes had n ≥ 5 (using 1 degree of freedom, a chi-square value of χ² ≥ 3.84 was considered significant) or Fisher’s test where any 1 genotype had n < 5. All remaining data were assessed using Kruskal–Wallis to compare among genotypes and Mann–Whitney for pairwise comparisons. Pearson correlation was utilized for the determination of diameter or age correlation with gene expression or NOS activity. A P-value of ≤0.05 was considered significant for all analyses.

RESULTS

Distinct differences in Nos3 expression in the media/intima and adventitia by aortic valve morphology and dilatation

Nos3 expression normalized to Pecam1 was elevated in media/intima specimens from BAV patients when compared with TAV patients (median ± standard error of the mean, 0.169 ± 0.030 vs 0.073 ± 0.008, P < 0.001, Fig. 1A). Media/intima specimens exhibited no differences among patient cohorts in Nos3 gene expression when normalized to the housekeeping gene Ppia (Fig. 1B). A decrease in Pecam1 expression normalized to Ppia was observed in media/intima specimens from BAV patients when compared with TAV patients (0.031 ± 0.009 vs 0.012 ± 0.006, P = 0.027, Fig. 1C). We did not identify any correlation for Pecam1-normalized Nos3 in TAV or BAV patient-derived specimens for either diameter (Supplementary Material, Fig. S1A) or patient age (Supplementary Material, Fig. S1B).  Similarly, Ppia-normalized Pecam1 expression did not correlate for either TAV or BAV patient-derived specimens for diameter (Supplementary Material, Fig. S1C) or age (Supplementary Material, Fig. S1D).

Figure 1:

Nos3 gene expression in human ascending aorta media/intima specimens (A–C) and adventitia (D–F). Nos3 expression normalized to (A) Pecam1 and (B) Ppia expression. (C) Pecam1 expression in media/intima specimens normalized to Ppia. Nos3 expression normalized to (D) Pecam1 and (E) Ppia expression. (F) Pecam1 expression in adventitial specimens normalized to Ppia. BAV: bicuspid aortic valve; NA: non-aneurysmal; TAA: thoracic aortic aneurysm; TAV: tricuspid aortic valve. In (D) and (E), P-value denotes the comparison when non-aneurysmal from bicuspid aortic valve and tricuspid aortic valve are pooled* versus thoracic aortic aneurysm from bicuspid aortic valve and tricuspid aortic valve are pooled^#.

Diminished Nos3 gene expression in thoracic aortic aneurysm adventitia

Examination of gene expression in adventitial specimens revealed reduced Nos3 expression in TAA specimens when compared with NA specimens when normalized to Pecam1 (0.045 ± 0.007 vs 0.082 ± 0.015, P < 0.01, Fig. 1D) or Ppia (0.053 ± 0.009 vs 0.077 ± 0.013, P = 0.03, Fig. 1E). No differences were observed in Ppia-normalized Pecam1 expression among patient cohorts (Fig. 1F). For TAV adventitial specimens, Pecam1-normalized Nos3 expression negatively correlated with increasing diameter (R² = 0.1554, P = 0.016, Supplementary Material, Fig. S2A) and age (R² = 0.0967, P = 0.04, Supplementary Material, Fig. S2B). The correlation of Nos3 expression with diameter or age was not observed in BAV patient-derived specimens. Gene expression of Ppia-normalized Pecam1 expression did not correlate with diameter (Supplementary Material, Fig. S2C) or age (Supplementary Material, Fig. S2D) for either TAV or BAV patient-derived specimens.

Underrepresentation of −786TT Nos3 single-nucleotide polymorphism genotype (wild type) in bicuspid aortic valve-thoracic aortic aneurysm patients

The BAV-TAA population showed a significant deviation from HWE (χ² = 4.56, P = 0.033) characterized by an underrepresentation of individuals with the TT wild-type genotype (35 expected, 25 observed) and a modest overrepresentation of individuals with the heterozygous TC genotype (42 expected, 50 observed, Fig. 2A). The TAV-TAA and BAV-NA patient populations exhibited similar distributions of the −786T/C Nos3 polymorphism as the control population.

Figure 2:

Nos3 genotype distribution by patient cohort. (A) Population distribution of Nos3 −786 single-nucleotide polymorphisms within each cohort. (B) Population distribution of Nos3 894 single-nucleotide polymorphism genotype within each cohort. Nos3 expression normalized to Ppia expression for each cohort by Nos3 −786 single-nucleotide polymorphism genotype for media/intima specimens (C) and adventitial specimens (D). Nos3 expression normalized to Pecam1 expression by Nos3 −786 single-nucleotide polymorphism genotype for media/intima specimens (E) and adventitial specimens (F). BAV: bicuspid aortic valve; NA: non-aneurysmal; TAA: thoracic aortic aneurysm; TAV: tricuspid aortic valve.

894G/T Nos3 single-nucleotide polymorphism exhibited normal distribution for all patient cohorts

Analysis of 894G/T SNP genotype revealed that the distribution was within HWE for aneurysmal patient populations and NA BAV patients (Fig. 2B). Aneurysmal patients in both the BAV and TAV cohorts did not show any deviation in genotype distribution when stratified by the degree of aortic insufficiency (AI) (Supplementary Material, Table S3) or degree of aortic stenosis (AS). In addition, no differences were observed among genotypes for either aneurysmal group when examined by age or aortic diameter. Genotype distribution for BAV-NA patients did not differ among degrees of AI or AS (Supplementary Material, Table S2), nor did they correlate with patient age.

Differences in Nos3 expression associated with −786CC SNP

TAV-TAA patients with the −786 SNP CC genotype exhibited increased Nos3 expression in media/intimal specimens when compared with specimens from wild-type (TT) patients (0.007 ± 0.003 vs 0.001 ± 0.001, P = 0.04). There were no differences in Nos3 expression identified based on −786 SNP genotype for media/intimal specimens from control (TAV-NA) patients or for either NA or aneurysmal BAV patients when normalized to either Ppia (Fig. 2C) or in any patient cohorts when normalized to Pecam1 (Fig. 2E). In adventitial specimens normalized to Ppia expression, a decrease in Nos3 expression was noted in TAV-TAA specimens collected from patients exhibiting the −786CC genotype when compared with expression levels in specimens collected from patients with −786TC genotype (0.022 ± 0.019 vs 0.092 ± 0.026, P = 0.026). Nos3 expression was found to be uniformly decreased in the adventitia irrespective of −786T/C SNP genotype for NA or aneurysmal BAV patients, as well as control patients (Fig. 2D). Nos3 gene expression did not differ in either tissue layer for 894 SNP genotype (Media/Intima: Supplementary Material, Fig. S3A; Adventitia: Supplementary Material, Fig. S3B), nor when normalized to Pecam1 gene expression (Fig. 2F).

NOS enzymatic activity differed in media/intima versus adventitia

No differences were identified for NOS activity in media/intima specimens (Fig. 3A). Differences in NOS activity among genotypes were observed only for the control group where media/intima specimens from patients exhibiting the −786CC genotype were found to express higher NOS activity when compared with specimens from −786TC patients (70 500 ± 4070 vs 38 600 ± 3740 CPM/mg protein, P = 0.003). There was no change among genotypes for any pathological group (Fig. 3B), nor was there an effect of 894 SNP genotype on NOS activity (Supplementary Material, Fig. S3C). NOS activity in the adventitia was lower in all pathological groups when compared with specimens from control patients [48 100 ± 4630 (P < 0.001), 47 300 ± 9920 (P = 0.006), 42 200 ± 4600 (P < 0.001) vs 72 600 ± 7000 CPM/mg protein, Fig. 3C]. NOS activity did not appear to be influenced by −786TC SNP genotype (Fig. 3D) or by 894GT SNP genotype (Supplementary Material, Fig. S3D) within any cohort.

Figure 3:

NOS activity in human ascending aortic specimens. NOS activity is displayed according to the patient cohort for media/intima specimens (A) and by Nos3 −786 single-nucleotide polymorphism genotype (B). NOS activity in adventitial specimens by patient cohort (C) and by Nos3 −786 single-nucleotide polymorphism genotype (D). BAV: bicuspid aortic valve; NA: non-aneurysmal; TAA: thoracic aortic aneurysm; TAV: tricuspid aortic valve.

Intimal cells from patients with a −786C allele exhibit decreased Nos3 expression

We did not observe any differences in Nos3 expression in primary human aortic endothelial cells isolated from TAV versus BAV cohorts comprising all genotypes and all aortic diameters (Fig. 4A). However, intimal endothelial cells isolated from patients exhibiting at least 1 C allele for −786 SNP (TC and CC) showed reduced Nos3 in vitro when compared with wild-type (TT) cells (0.005 ± 0.001 vs 0.013 ± 0.005, P = 0.011, Fig. 4B). There was no difference in Nos3 expression among 894 G/T genotypes (Fig. 4C).

Figure 4:

Nos3 gene expression in cultured human primary aortic intimal endothelial cells. (A) Expression in TAV versus BAV patients. (B) Expression in cells exhibiting different −786 single-nucleotide polymorphism genotype and (C) by 894 single-nucleotide polymorphism genotype. BAV: bicuspid aortic valve; TAV: tricuspid aortic valve.

DISCUSSION

A role for Nos3 in BAV-associated aortopathy has been considered by several groups, including our own with varying results. We previously reported elevated Nos3 gene and protein expression in intima-containing medial specimens of aneurysmal aorta from BAV patients [18], findings consistent with others, despite use of different internal reference genes [17]. The same group more recently described decreased Nos3 mRNA and protein in NA aorta from BAV patients when compared with normal and aneurysmal specimens from BAV patients [14]. Indeed, others also reported decreased Nos3 at the mRNA level [15] and protein levels [16], which may be related to the choice of internal reference genes. For example, the reduced level of Pecam1 within intima/medial specimens from BAV patients may reflect a decrease in endothelial cells of the intima or medial vasa vasorum. Thus, endothelial cell produced Nos3 was higher in specimens derived from BAV patients relative to specimens from TAV patients.

Here, we considered the adventitia separately and found decreased Nos3 in the adventitia of aneurysmal patients compared to NA specimens, regardless of valve morphology. Others reported decreased Nos3 in the aorta of BAV patients, which we presumed are full-thickness aortic specimens that include the intima, media and adventitia, absent of details to the contrary [14–16]. That NOS activity was decreased in all pathologic adventitial specimens and gene expression in specimens from BAV-NA patients remained unchanged when compared with the control group suggests an earlier stage of BAV-associated aortopathy than their aneurysmal counterparts. Our previous finding of elevated NOS protein in BAV-TAA absent of evident downstream NO signalling via phosphorylation of vasodilatory stimulated phosphoprotein [18] tells us that other factors may regulate NOS activity, hence our choice to measure enzymatic activity rather than protein levels. Thus, progression of aortic dilatation may involve post-translational regulation affecting NOS activity in the adventitia while Nos3 down-regulation occurs during a different or later stage of aneurysmal disease.

Wall shear stress (WSS) has been raised as an important parameter associated with haemodynamic disturbances in BAV patients [23]. Other studies using 4D flow magnetic resonance imaging revealed regional changes in WSS on the basis of BAV morphotype with aortic stenosis severity affecting the magnitude and variability of elevated WSS independent of aortic valve morphology [24]. While elevated Nos3 gene expression was most pronounced in media/intima specimens of the ascending aorta directly proximal to the right coronary sinus [18], others examining full thickness aortic specimens of the greater and lesser curvature reported no regional differences in Nos3 expression for BAV patients [14]. The Schäfers group recently examined NA aortic specimens from patients with ‘unicuspid’ aortic valves (i.e. BAV type 2 morphotype) and reported decreased intimal Nos3 protein expression despite elevated phos-NOS3 when compared with normal specimens from patients with TAV [13]. This further suggests that Nos3 expression is differentially regulated at the transcriptional, translational and post-translational levels in a layer-specific manner, not only when aneurysm or BAV are present, but also with potential influence of BAV morphotype absent of aortic dilatation. Vasa vasorum-associated endothelial cells could be influenced by increased nutritional demands due to aneurysm. Endothelial cells residing within the adventitia may be less influenced by aortic valve morphology-associated haemodynamics. Other patient comorbidities and demographics may be associated with endothelial dysfunction. However, because our pathological groups include a similar number of specimens from male and female patients, less smokers and diabetics, and in some groups, less hypertension and/or statin use than that of our control group and marked differences in NOS enzymatic activity persist, stratification by sex or comorbidity is unlikely to yield different results with sufficient experimental power. We reason that Nos3 gene expression and enzymatic activity down-regulation in the adventitia may be related to the paucity of vasa vasorum in the adventitia of patients with aneurysm [25]. We believe that discrepancies in Nos3 expression among the discussed reports may be due to differences in specimen sampling, reference gene selection, the layer(s) analysed, variations in genotypic distribution, ethnicity and ethnic diversity of the patient population, BAV morphotype or degree of aortic insufficiency or stenosis. Thus, development of novel imaging modalities that detect endothelial dysfunction at the microvascular level may offer additional clues to developing or progressing aortopathy.

This genotype distribution of Nos3 SNPs is in contrast to a prior study of a smaller Sicilian population determined to have an overrepresentation of the −786T allele in patients with BAV and zero occurrences of the homozygous mutant −786CC [19]. Conversely, we found an underrepresentation of the wild-type allele that reflects a slightly increased proportion of individuals with at least 1 C allele at −786. The 894 SNP exhibited the expected population distribution in all patient groups examined, indicating lack of involvement of this SNP in TAA for the patients included in this study. In human umbilical vein endothelial cells, decreased luciferase activity was noted when driven by a Nos3 promoter construct with a C allele at the −786 position when compared with an analogous construct with a T allele and was inducible under hypoxia [26]. We found this notion of oxygen-dependent differential genotypic effects on Nos3 expression to be intriguing in light of our recent report revealing evidence of chronic hypoxia in the medial layer of human aneurysmal aorta associated with decreased density of adventitial vasa vasorum [25]. Similarly, when all cohorts were considered together, human aortic intimal endothelial cells exhibiting at least 1 C allele at the −786 SNP showed lower levels of basal Nos3 when compared with the wild-type SNP (TT) suggesting that the C allele in −786 SNP may reduce the Nos3 promoter activity. Since there are more individuals in the BAV-TAA population with at least 1 C allele at −786, our finding of medial hypoxia [25] could explain our observations of increased Nos3 in BAV-associated aneurysm when compared with degenerative aneurysm [18].

Our finding of altered representation of Nos3 SNPs in aneurysmal BAV patients infuses additional complexity to our current understanding of how genetic variants affect biological pathways governing vascular (dys)function in the setting of BAV-related aortopathy and aneurysmal disease. Nos3 expression appears to be differentially regulated in the media/intima versus adventitia depending on valve morphology or the presence of aneurysm. These observations are consistent with other distinct mechanisms contributing to TAA in BAV and TAV patients, the unique matrix microarchitecture [27], wall tensile stress [28] and strength [29], distinct profiles of oxidants and anti-oxidants [30, 31] and local hypoxia in the media [25] that when taken together yields a more nuanced picture of aortic integrity for BAV-associated and degenerative aneurysms. These findings suggest that tissue-level bioactivity of gene products is likely more informative than Nos3 genotype in pursuit of a more complete understanding of BAV-associated aortopathy. We envision that improved risk stratification for BAV patients will be made more impactful through multi-layer interrogation of disease manifestation and via advances in tissue and cellular level detection of the biological and biomechanical contributors to ascending aortic disease onset, progression, and severity.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Glossary

ABBREVIATIONS

BAV

Bicuspid aortic valve

CPM

Counts per minute

HWE

Hardy–Weinberg equilibrium

NA

Non-aneurysmal

SNPs

Single-nucleotide polymorphisms

TAA

Thoracic aortic aneurysm

TAV

Tricuspid aortic valve

WSS

Wall shear stress