Background

Bicuspid aortic valve (BAV) is the most common congenital cardiac malformation, with a prevalence estimated between 0.5% and 2% [1, 2]. Nearly all patients with BAV develop serious aortic complications, including significant valve regurgitation, severe aortic stenosis, aortic aneurysm and dissection [3, 4]. Therefore, the burden of BAV disease is more significant than any other congenital cardiac abnormalities [4]. Clinical studies have reported high heritability in families of patients with BAV [5–7], suggesting a Mendelian pattern of inheritance and the need for echocardiography screening in first-degree relatives of patients with BAV [8].

A family-based linkage analysis has identified genomic regions on chromosomes 18q, 5q and 13q responsible for BAV and associated cardiac malformations, but the precise genes within these regions were not defined [9].

To date, only mutations in the NOTCH1 gene, a signaling transmembrane receptor (2,556 amino acids) involved in multiple cell aspects of vascular development, have been associated with dominantly inherited BAV in a small number of families [10].

However, the rate of NOTCH1 gene mutations in familial forms of BAV and their association with different clinical complications remains incomplete [11, 12]. Interestingly, McBride et al. demonstrated that mutations in the gene NOTCH1 are found in cases of left ventricular outflow tract (LVOT) malformations including aortic valve stenosis, coarctation of the aorta and hypoplastic left heart syndrome [13]. In particular, a much higher rate of bicuspid aortic valve (BAV) has been found in families with LVOT malformations compared to the general population, suggesting that BAV may be a forme frusta of the more serious LVOT malformations [13].

Recently, a missense mutation in transforming growth factor-beta receptor 2 gene (TGFBR2) was found in a patient with BAV and aortic aneurysm [14], but earlier studies found no mutation in either TGFBR1 or TGFBR2 in patients with familial and sporadic BAV disease [15, 16].

Finally, GATA5 was reported to have an essential role in cardiac morphogenesis and in aortic valve development [17, 18], and targeted deletion of this gene in mice leads to partially penetrant BAV [19]. Accordingly, a very recent study investigated the relationship between GATA5 gene variations and BAV disease identifying rare non-synonymous variations in the functionally important GATA5 transcriptional activation domain [20]. Conversely from NOTCH1 gene, the involvement of GATA5 has not been described in other LVOT malformations.

However, further studies are needed to assess the involvement of NOTCH1, GATA5 and TGFBR1 and TGFBR2 mutations in patients with BAV and its associated aortic complications.

The purpose of our study was to investigate the potential contribution of germline mutations in NOTCH1, GATA5, TGFBR1 and TGFBR2 genes in a cohort of Italian patients with familial BAV.

Methods

Study population

A cohort of 11 unrelated Italian patients (8 males; age = 42 ± 19 years) with familial BAV were recruited for this study. Patients with sporadic BAV and/or with coexisting syndromes, such as Marfan syndrome, Turner syndrome and Di George syndrome or unknown syndromes were excluded from the study. Clinical evaluation consisted of medical history, physical examination, 2-dimensional echocardiography and a complete Magnetic Resonance Imaging (MRI) study. A detailed family history was obtained for each proband. Familial BAV was defined if two or more affected relatives had proven BAV diagnosed by echocardiography or cardiac magnetic resonance. Figure 1 shows the pedigree of BAV patients. Patient’s phenotype is reported in Table 1. Two of the probands (18%) had BAV with unknown morphology. Of the 9 probands with known BAV morphology, 7 (63%) had fusion of the right/left coronary commissure whereas 2 (18%) had a valve with anterior-posterior sinus without raphe.We also included a control cohort comprised of 100 ethnically unrelated individuals (60 male; age = 35 ± 16 years) without BAV as shown by echocardiography. This study was conducted with informed consent of every participant subject and was approved by the local Ethical Research Committee.

Figure 1
figure 1

Pedigrees of BAV families A-K. The proband is indicated by arrow and affected status is indicated by filled symbol. Slanted bars represent deceased individuals.

Full size image
Table 1 Clinical and demographic characteristics of study population
Full size table

Mutational sequencing analysis

A mutational screening was performed by direct gene sequencing of all coding exons including adjacent intronic as well as 5′ and 3′ untranslated sequences of NOTCH1, GATA5 and TGFBR1, TGFBR2 genes. Genomic DNA was extracted from peripheral blood cells using the QIAGEN BioRobot® EZ1 System. Exons and the flanking intronic sequences were amplified by polymerase chain reaction (PCR) using specific primers as described previously for NOTCH1 and TGFBR1 and TGFBR2[11, 21] and using Primer3 (v. 0.4.0) software (http://frodo.wi.mit.edu/primer3/) based on the cDNA sequences available in GenBank for GATA5 screening.

Approximately, PCR amplifications were performed with a volume of 50 microl mixture containing 1.5 mM MgCl2, 10 mMdNTP, 50 ng genomic DNA, 20 micromol of each primer and 2.5 U Taq DNA polymerase. PCR product was used for PCR sequencing reaction with the CEQ DTCS Quick Start Kit. The sequencing reaction products were purified by precipitation with ethanol, re-suspended in a sample loading solution and analyzed with a CEQ 8800 capillary sequence (Beckman Coulter, Germany) according to the manufacturer’s protocol. Resulting sequences were analyzed using CEQ 8800 software packages and aligned against a reference sequence obtained from GenBank.

For prediction of the functional consequences of mutations on protein sequence, we used software programs available on the internet, namely PolyPhen-2(http://genetics.bwh.harvard.edu/pph/) SIFT (http://sift.jcvi.org/) and Mutation Taster (http://www.mutationtaster.org/ ). PolyPhen-2 predicts the effect of an amino acid substitution on the structure and function of a protein, using sequence homology. The PolyPhen-2 score represents the probability that a substitution is damaging, so values nearer 1 are more confidently predicted to be deleterious (note that this is the opposite of SIFT).

SIFT uses sequence homology to predict whether an amino acid substitution will affect the protein function. SIFT score <0.05 indicates the amino acid substitution is damaging while scores ≥ 0.05 are predicted to be tolerant. Mutation Taster is a free, web-based application for rapid evaluation of the disease-causing potential of DNA sequence alterations. Mutation Taster integrates information from different biomedical databases and uses established analysis tools. Analyses comprise evolutionary conservation, splice-site changes, loss of protein features and changes that might affect the amount of mRNA.

Detected mutations in probands were confirmed by sequencing in the opposite direction from another two independent PCR products.

The presence of a novel mutation was assessed in 200 ethnically reference alleles derived from volunteers by PCR-based restriction fragment length polymorphism (PCR-RFLP) or by direct gene sequencing.

Results

Clinical and demographic characteristics of each patient are summarized in Table 1. The group comprised 73% males, and coarctation of the aorta (CoA) as coexistent cardiovascular anomalies occurred in 18% of the cohort.

Genetic analysis of NOTCH1gene

A mutational screening was performed by direct gene sequencing of all 34 coding exons including adjacent intronic as well as 5′ and 3′ untranslated sequences of NOTCH1 gene. Seventeen separate NOTCH1 sequence variants listed previously in public dbSNPs or in the literature were found in our patients, while five variants were novel. These included one synonymous variant (Exon 21, p.G1166G), two intronic (IVS11 + 63; IVS32-41) and two novel mutations (Exon 5,p.P284L; Exon 26, p.Y1619X). Details of the specific sequence variants and number of positive patients are summarized in Table 2.

Table 2 Genetic variants within the human NOTCH1 gene identified in familial BAV patients
Full size table

The novel C > T heterozygous substitution in exon 5, resulting in the substitution of a Proline with a Leucine (P284L) at position 284, was found in the sequencing electropherograms of a 40-year-old male patient (Proband A; Figure 2). The identified missense mutation was absent in 200 ethnically reference alleles derived from volunteers. PolyPhen-2 analysis predicted the p.P284L mutation as probably damaging with a score of 0.993. This result was confirmed by SIFT (score: 0.04).

Figure 2
figure 2

DNA sequencing electropherograms demonstrating the NOTCH1 heterozygous sequence mutations (Exon 5, p.P284L and Exon 26, p.Y1619X).

Full size image

The missense mutation (p.P284L) identified in our proband, tested by Mutation Taster, induced the loss of an important residue with putative significance for calcium binding to epidermal growth factor (EGF)-like 7 domain.

The proband, at the age of 39 years, presented severe calcified aortic valve and normal diameter of ascending aorta (39 mm). One year later, the patient underwent elective aortic valve and ascending aorta repair (48 mm) replacement. Family history and echocardiography analysis revealed the presence of BAV in his 7-year-old son. Genetic screening of the proband’s child confirmed the presence of p.P284L mutation in exon 5 of NOTCH1 gene.

Moreover, we detected a novel heterozygous C-to-G transition, in exon 26, of nucleotide 4856 leading to a premature stop codon at position 1619 (p.Y1619X) instead of tyrosine in the sequencing electropherograms of a 38-year-old male patient (Proband I; Figure 2). The identified mutation was absent in 200 ethnically reference alleles derived from volunteers. At age 6 months the proband had surgery to correct the CoA, while at age 25 years he underwent elective aortic valve replacement. Family history and echocardiography analysis revealed the presence of BAV in his 5-year-old daughter. Genetic screening of the proband’s child confirmed the presence of p.Y1619X mutation in exon 26 of NOTCH1 gene.

Genetic analysis of GATA5gene

Sequencing of all 6 exons and splice signal sequences of GATA5 in our cohort revealed five synonymous, one non-synonymous and three intronic variants. Details of the specific sequence variants and number of positive patients are summarized in Table 3. We identified one novel intronic variant (C > T; IVS 5–15) in one patient, but it was also present in our control group.

Table 3 Genetic variants within the human GATA5 gene identified in familial BAV patients
Full size table

Genetic analysis of TGFBR1 and TGFBR2genes

TGFBR1 and TGFBR2 genes were analyzed by direct gene sequencing of all coding exons including adjacent intronic as well as 5′ and 3′ untranslated sequences. No pathogenetic mutation was identified; we found only known common genetic polymorphisms. Details of the specific sequence variants in TGFBR1 and TGFBR2 genes are summarized in Table 4.

Table 4 Genetic variants within the human TGFBR1 and TGFBR2 genes identified in familial BAV patients
Full size table

Discussion

In our cohort of 11 familial patients with BAV two novel mutations were found in the NOTCH1 gene in two unrelated families. The mutations segregated with the disease in these families, and they were not found in 200 chromosomes from ethnically matched controls.

On the contrary, analysis of GATA5, TGFBR1 and TGFBR2 genes identified only known genetic variants. Our findings confirm the important role of NOTCH1 as conserved intracellular regulator in the pathogenesis of congenital valve disease and in its complications. Indeed, the finding of two independent de novo mutations in this gene in a small number of affected subjects is compelling evidence of disease causation.

NOTCH1 encodes for a transmembrane protein that activates a signaling pathway with a major role in cardiac embryogenesis, including aortic and pulmonary valve development as well as the development and maintenance of the aorta and other great vessels [22]. Murine model studies indicate that Notch signaling promotes the epithelial-to mesenchyme transition that gives rise to the primordial cardiac valve, which later is sculpted into mature valves [23]. Moreover, a recent study showed that heterozygous Notch1-null (Notch1(+/)(−) mice had greater than fivefold more aortic valve calcification than age- and sex-matched wild type littermates [24].

Indeed, in vitro studies showed that Notch1 signaling repressed valvular Bmp2 expression, and de-repression of Bmp2 was involved in calcification induced by Notch1 inhibition. Thus, Notch1 signaling appears to prevent aortic valve calcification in part by repressing Bmp2 expression within the valve [24].

Defective NOTCH signaling might also contribute to aortic valve calcification by repressing the activity of the hairy-related family of transcription factors (Hrt), and thereby also repressing the Runx2/Cbfa1 pathway, a central transcriptional regulator of osteoblast cell fate [24, 25].

In addition, there is increasing evidence of interaction between the Notch1 pathway and transforming growth factor beta (TGFβ) [26]. This is particularly important in light of evidence showing that TGFβ plays an important role in vascular remodeling and alterations in TGFβ signaling activity might enhance matrix degradation [27, 28].

In 2005, the discovery of a nonsense (p.R1108X) and a frameshift mutation (p.H1505del) in two families represented the first demonstration of NOTCH1 germline mutations as a cause of BAV and severe valve calcification in non-syndromic autosomal-dominant human pedigrees [10].

Subsequently, two novel mutations (p.T596M and p.P1797H) located in highly conserved regions of the protein were identified in patients with sporadic BAV, providing evidence that mutations in NOTCH1 do not only play a role in familial BAV, but can also be observed in approximately 4% of sporadic cases [11]. Furthermore, a targeted mutational analysis of NOTCH1 demonstrated the presence of two novel mutations (p.A1343V and p.P1390T) among patients with BAV with sporadic and concomitant thoracic aortic aneurysms, providing evidence that mutations in NOTCH1 may also confer susceptibility to aortic aneurysm formation [12].

In a very recent paper, Kent et al. suggested a NOTCH1-dependent mechanism that produces stenotic, insufficient and/or calcified aortic valves with a rare aneurysm, and a NOTCH1-independent mechanism that produces highly penetrant ascending aortic aneurysm (AscAA) in the presence of a non-calcified and often normally functioning BAV [29]. Our data appear to agree with and fit well into the proposal made in this article by Kent et al., since we found two novel mutations in these patients with valve malformation, calcification, and dysfunction as predominant phenotype caused by altered NOTCH1 signaling.

Conversely, we did not identify any pathogenetic mutations in GATA5. The non-synonymous variant Thr67Pro located in the transcriptional activator domain encoded by exon 1, detected with a frequency of 45% in our population, is a common genetic polymorphism previously described [20, 30].

Regarding to TGFBR1 and TGFBR2 genes, in contrast with a recent paper showing a missense mutation in TGFBR2 gene in a patient with BAV and aortic aneurysm [14], our findings support the studies suggesting that mutations in these genes are rarely identified in patients with familial BAV [15, 16].

Conclusion

Although the small number of patients (11) might be an important study limitation, we have identified two novel NOTCH1 mutations in two Italian families with BAV. Our findings confirm the critical role of NOCTH1 in the development of aortic valve and aortic complications. Genetic screening for NOTCH1 mutation may be considered for diagnosis and family screening as part of the standard management of patients with family history of BAV. However, the etiology of this common valve disease is largely unknown. Future studies based on new genetic screening strategies, the so-called next-generation sequencing, are needed in order to unravel the still largely unknown genetics of BAV.

Abbreviations: BAV = Bicuspid Aortic Valve, TGFBR1 = transforming growth factor-beta receptorI gene; TGFBR2 = transforming growth factor-beta receptorII gene; PCR-RFLP = PCR-based restriction fragment length polymorphism; MAF = minor allele frequency; CoA = coarctation of the aorta; TGFβ = transforming growth factor beta; MRI = Magnetic Resonance Imaging; AscAA = ascending aortic aneurysm; LVOT = left ventricular outflow tract.