Julie De Backer

Hartcentrum, Ghent University Hospital

There are different possibilities for the classification of genetic defects in congenital heart disease (CHD). From a genetic point of view one would classify the different entities according to the size, the location and the type of the genetic defect, going from chromosomal defects (e.g. Trisomy 21) over multiple gene deletions (e.g. Velocardiofacial syndrome, del22q11) to monogenetic diseases (e.g. Marfan syndrome, 15q21 mutations). (Pediatric) Cardiologists are used to classifications according to the anatomic abnormalities. Embryologists distinguish abnormalities according to their effect on cardiac morphogenesis.

It is clear that limitation to the left ventricular outflow tract is only possible in the case of an anatomical classification. Both genetics and morphogenesis are by definition not limited to these boundaries and it would do more harm than good to try to restrict this overview to this area.

We will give a brief overview of possible approaches in the identification of genes causing CHD, as well as some basic principles of cardiac morphogenesis, after which we will focuss on the current knowledge of some important genetic defects with LVOT abnormalities. This way we hope to provide useful information for the clinician involved in congenital heart disease.

1. Identification of causal genes

A first step in the search for causal genes is the identification of candidate genes. This can be achieved through linkage analysis of large kindreds. Alternatively, a balanced translocation in a patient with a congenital heart defect whose breakpoint disrupts a gene can also be used in the identification of a candidate gene.

Proving that a candidate gene causes a specific heart defect remains a formidable task. Two complementary approaches are possible.

The first one consists of mutation screening in patients with a specific defect. To test the causality of a suspect mutation, either exclusion of the mutation in a large sample of normal controls or cosegrargation of the mutant allele with the disease in large families is required. At present mutation screening of candidate genes is restricted to research laboratories, but with the development of high throughput DNA sequencing techniques, mutation screening will become more widely available.

The second approach to prove that a candidate gene causes a specific heart defect involves the use of genetically engineered animal models. Both knock-out and knock-in models are possible.

2. Morphogenesis of the cardiovascular system

As early as the third week of gestation, cardiac progenitor cells become committed to a cardiogenic fate. The specific responsible signalling molecules are largely unknown at present. Both the TGFβ family and the Wnt signalling pathway play an important role in this process. TGFβ2 is required for normal development of the truncal and conal septa. BMP’s which are part of the TGFβ superfamily also play a role in conotruncal development.

 The heart tube is then looped to a left-right polarity. Ventricular chambers arise from the outer curvature, whereas the inner curvature gives rise to the outflow portions. Further septation and remodelling eventually leads to a four chambered heart.

The aortic sac arises from another major cell type, known as the neural crest. They will give rise to the aorta, pulmonary artery, semilunar valves and the superior part of the ventricular septum.

It is easy to understand that mutations in the signalling molecules as well as mutations in genes responsible for the formation of the neural crest may give rise to defects in the LVOT development.

3. Specific genetic defects

3.1 Genes involved in morphogenesis

3.1.1 22q11 deletion syndrome (Velocardiofacial syndrome, Di George syndrome)

This is the most common human gene deletion syndrome (1 in 4000 live births) and the second most common genetic cause of CHD after trisomy 21. Most of the derivatives of the pharyngeal arches and pouches are affected. Of individuals with 22q11, 75% have defects of the conotruncus and/or aortic arch (derived from the neural crest) in addition to pharyngeal arch defects including cleft palate, dysmorphic facial features, thymic hypoplasia and hypoparathyroidism.

The genetic defect consists of a monoallelic microdeletion of chromosome 22q11 spanning approximately 3Mb, containing nearly 30 genes. Whether the cardiac defects are caused by the lack of one single gene is not known at present. Based on mouse models, Tbx1, a transcription factor expressed in the pharyngeal arches, was studied as a possible candidate. Confirmation came from the observation in humans of Tbx1 mutations in subjects with phenotypic features of VCF in whom no 22q11 deletion could be demonstrated. Tbx1 mutations are responsible for 5 major phenotypes of the 22q11 syndrome, namely, abnormal facies, cardiac defects, thymic hypoplasia, velopharyngeal insufficiency of the cleft palate, and parathyroid dysfunction with hypocalcemia; these mutations did not appear to be responsible for typical mental retardation that is commonly seen in patients with the deletion form of 22q11 syndrome. Another gene, Crkol has been identified in the del22q11 region; Crkol is involved in growth factor and focal adhesion signalling. Crkol mutations cause aortic arch defects in the mouse. Mutations in humans have not been identified until now.

3.1.2 Alagille syndrome(AS)

This is an autosomal dominant condition, caused by mutations in Jagged-1, a ligand for the transmembrane receptor Notch, which is involved in embryonic patterning and cellular differentiation. CHD is present in 95% of AS patients. TOF is the most common finding. Other common findings in AS are a characteristic facies and biliary atresia.

Jagged-1 mutations have also been found in isolated Pulmonary Stenosis or TOF.

3.1.3 Aortic coarctation

Aortic coarctation is known to have a high familial recurrence rate. Several families with autosomal dominant inheritance have been described, but until now no causative gene in humans is identified. A zebrafish model suggests mutations in HRT2/Hey2 gene as possible candidates. These genes have an essential role in controlling the decision of an early vascular progenitor to adopt an arterial fate in a specific vascular segment.

Aortic coarctation is found in association with several syndromes such as Turner syndrome (X0) and Noonan syndrome (PTPN11 mutations). Up to 85% of patients with coarctation also have bicuspid aortic valves.

3.1.4 Char syndrome

Char syndrome is a rare disorder characterized by patent ductus arteriosus and craniofacial abnormalities.  Mutations in TFAP2B have been identified in patients with Char syndrome, suggesting an essential role for TFAP2B in governing closure of this vessel. TFAP2B encodes a transcription factor expressed in the neural crest.

3.1.5 Defects in valve development

Thickened valve leaflets that result in stenotic valves are a common form of CHD. The Smad proteins are intracellular transcriptional mediators of signaling initiated by TGFβ ligands. Smad 6 is specifically expressed in the AV cushions and outflow tract during cardiogenesis and is a negative regulator of TGFβ signaling. In mice, Smad 6 targeting results in thickened and gelatinous AV and semilunar valves similar to those observed in human disease. In addition to Smad 6 there are likely other genes in the TGFβ signaling pathway that, when mutated result in the formation of stenotic and hyperplastic valves.

3.2 Genes with a structural function

3.3.1 Williams Beuren syndrome (WBS) and Supravalvular aortic stenosis (SVAS)

WBS is a rare autosomal dominant dysmorphic syndrome characterized by distinctive facial features (elfin facies), “cocktail-party behaviour”, short stature, hypercalcemia in infancy, and vascular stenosis localized at the aorta, pulmonary artery or other muscular arteries.

WBS is a so-called “contiguous gene disroder” caused by a heterozygous microdeletion at chromosome band 7q11.23 spanning approximately 1.6Mb and comprising 19 identified genes, including genes that code for structural proteins, transcription factors and transmembrane receptors. The vascular lesions are caused by mutations in the elastin gene (ELN). Elastin is produced as tropoelastin, which is then modified by lysyl-oxydase. Elastin is deposited on a framework of fibrillin and other microfibril associated proteins to form elastic fibres. ELN mutations lead to reduced amounts of elastin and compensatory increases in elastin lamellae and smooth muscle cells.

Familial SVAS is an autosomal dominant syndrome with the identical vascular phenotype of WBS, but without mental or growth retardation and without hypercalcemia. SVAS is not limited to the aorta, but is rather a systemic arteriopathy with pulmonary and systemic arterial stenoses. More than 50 different mutations in the elastin gene (ELN) have been reported. Functional haploinsufficiency resulting in reduced synthesis and secretion of tropoelastin seem to be the pathomechanism underlying most cases of nonsyndromic SVAS. 

Marfan syndrome (MFS)

Marfan syndrome is an autosomal dominant connective tissue disorder characterized by manifestations in different organ systems, including the skeletal, cardiovascular and ocular system. The diagnosis is based on the identification of major and minor clinical criteria in each organ system, according to the Gent Nosology. Cardiovascular findings include aortic dilatation and/or dissection, mitral valve prolapse and dilatation of the main pulmonary artery.

Marfan syndrome is caused by mutations in the Fibrillin 1 gene (FBN1). More than 500 different mutations are currently identified. Nearly each family has its own private mutation. There is strong evidence for locus homogeneity in MFS. With the current available techniques, the mutation detection rate in subjects fulfilling the diagnostic criteria achieves 90%. MFS is characterized by a large inter- and intrafamilial variability, which cannot be explained by the genotype. Other genetic and/or epigenetic factors must be involved.

Fibrillin 1 has been considered as a structural protein, providing a scaffold for the deposition of tropo elastin. Recent observations however, show evidence that FBN1 is more than a structural protein. From observations in mouse models, it appears that FBN1 participates in the TGFβ related signalling. FBN1 plays a role as a reservoir for TGFβ and at least part of the phenotypic features in MFS could be explained by an excess in TGFβ. These findings open the way to new therapeutic approaches in MFS.

A new aortic aneurysm syndrome (Loeys-Dietz syndrome)

In a very recent publication, ten families were described with a novel aortic aneurysm syndrome characterized by the triad of widely spaced eyes (hypertelorism), bifid uvula and/or cleft palate, and generalized arterial tortuosity with ascending aortic aneurysm/dissection. This syndrome shows autosomal dominant inheritance and variable clinical expression. Other findings were present in multiple systems and include: craniosynostosis, structural brain abnormalities, mental retardation, aneurysms with dissection throughout the arterial tree and congenital heart disease, including patent ductus arteriosus and ASD.

TGFBR2 was considered as a candidate gene because mouse models have shown that TGFβ signaling plays a prominent role in vascular and craniofacial development and that conditional knockout of TGFBR2 in neural crest cells causes cleft palate and calvaria defects. Furthermore, another recent publication identified TGFBR2 mutations in a large family with a Marfan like condition. TGFBR2 mutations were identified in 6 families. A TGFBR2 mutation was not found in four additional families with a clinically indistinguishable phenotype. All exons of TGFBR1 were sequenced and a unique missense mutation was found in each family that substituted evolutionarily conserved residues in TβRI. Histological examination of the aorta from individuals with TGFBR2 mutations showed complete loss of elastic fiber organization. These findings were observed in young children and in the absence of inflammation, suggesting a severe defect in elastogenesis rather than secondary elastic fiber destruction. While some individuals with mutations in TGFBR1 or TGFBR2 show some overlap with MFS (variable evidence for bone overgrowth and aortic root dilatation and/or dissection), none satisfied diagnostic criteria for MFS. More importantly, all affected individuals showed manifestations in multiple organ systems that are not associated with MFS. Aneurysms tend to be particularly aggressive, with rupture at an early age or at a size not associated with high risk in MFS.

References

1. Chien. Molecular Basis of Cardiovascular Disease: a companion to Braunwald’s heart disease.Second edition, 2004

2. Deepak Srivastava, Eric Nolsen. A Genteic Blueprint of Cardiac Development. Nature, 14 Sept 2000 Vol 407, p 221-226

3. Bart Loeys et al. A syndrome of altered cardiovascular, craniofacial, neurocognitive and skeletal development caused by mutations in TGFBR1 or TGFBR2. Nature Genetics 2005, in press

<terug>