Aims Although ventricular septal defects (VSD) are the most common congenital heart lesion, familial clustering has been described only in rare instances. The aim of this study was to identify genetic factors and chromosomal regions contributing to VSD.
Methods and results A unique, large kindred segregating various forms of septal pathologies—including VSD, ventricular septal aneurysms, and atrial septal defects (ASD)—was ascertained and characterized clinically and genetically. Eighteen family members in three generations could be studied, out of whom 10 are affected (2 ASD, 3 septal aneurysm, 4 VSD, and 1 tetralogy of Fallot). Parametric multipoint LOD scores reach significance on chromosome 10p15.3-10p15.2 (max. 3.29). The LOD score support interval is in a gene-poor region where deletions have been reported to associate with septal defects, but that is distinct from the DiGeorge syndrome 2 region on 10p. Multiple linkage analysis scenarios suggest that tetralogy of Fallot is a phenocopy and genetically distinct from the autosomal dominant form of septal pathologies observed in this family.
Conclusion This study maps a rare familial form of VSD/septal aneurysms to chromosome 10p15 and extends the spectrum of the genetic heterogeneity of septal pathologies. Fine mapping, haplotype construction, and resequencing will provide a unique opportunity to study the pathogenesis of septal defects and shed light on molecular mechanisms of septal development.
Tetralogy of Fallot
Septal defects are the most prevalent congenital heart malformation, accounting for 25–50% of all congenital heart defects. The most common septal defects are atrial septal defects of the secundum type (ASD II) as well as muscular ventricular septal defects (VSD).1 The association of septal defects with other congenital heart disease is well recognized, as an example, ASD can occur concomitantly with pulmonary valve stenosis. Despite the high incidence of septal defects, relatively little is known about the genetic origin of these diseases.
In most cases, these defects occur sporadically, i.e. family history is negative for other members with congenital heart disease. However, genetic contributions to these malformations are increasingly recognized.2 Familial recurrence of ASD has been reported more frequently than that of VSD, and recurrence risks in these lesions have been estimated at up to 23%.3 At least four genes have been causally identified in non-syndromic ASD, namely NKX2.5, GATA4, MYH6, and ACTC.4–7 Mutations in causative genes can occur de novo or be transmitted in an autosomal dominant fashion with recurrence risks of 50%.
Recurrence risks in VSD have been estimated at 0.9–4.3%, which is lower than the numbers reported in ASD II.8 Ventricular septal defects with a known genetic aetiology have been mostly described in more complex congenital heart disease, such as tetralogy of Fallot caused by mutations in NKX2.5, JAG1, or FOG2. Likewise, recurrence risks for tetralogy of Fallot within families has been estimated at 2.5–3.0%, with relatively high concordance.9
We previously reported a large pedigree segregating VSD and aneurysms as an autosomal dominant trait with variable expressivity and high penetrance.10 In addition, a single individual with tetralogy of Fallot was identified in this pedigree. We hypothesized that the disease trait in this family is a monogenic trait amenable to a whole-genome scan. Here, we report additional clinical and genetic characterization of this kindred, with significant linkage to a locus on chromosome 10p15 not previously implicated in septal defects.
Samples and genotyping
Results of phenotypical characterization of this family were described previously.10 The family members underwent a cardiac exam including electrocardiogram (ECG) and echocardiography. The study was carried out in keeping with the tenets of Helsinki and approved by the respective Institutional Review Boards at The Prince Charles Hospital, Brisbane, QLD, Australia, and Sainte Justine Hospital, Montréal, QC, Canada. All participants gave informed consent to participate in the study. DNA was extracted from whole blood using the PureGene isolation kit (Gentra, Minneapolis, MN, USA) or the Oragene salive self-sampling kit (Oragene, Kanata, ON, Canada). We used the Illumina HumanLinkage 12 panel to perform a whole-genome SNP scan according to the manufacturer's instructions (Illumina, San Diego, CA, USA).
We used ALOHOMORA_M, GRR, and PedCheck for the preparation of input files.11–13 Linkage analysis was performed using Merlin v1.1.2 on Linux,14 and Genehunter using the easyLINKAGE 5.08 interface.15 We assumed an autosomal dominant model with the following parameters: penetrance 0.90, phenocopy rate 0.01, and disease allele frequency 0.001. Construction of haplotypes was performed using the Viterbi algorithm implemented in Genehunter.16 Family members affected with ventricular septal aneurysm, VSD, or ASD were set to affected. For Individual III-1 with tetralogy of Fallot, three linkage scenarios were considered: a first one in which he was considered to be of unknown phenotype status, and one in which he was considered to be unaffected. The first scenario is justified by the fact that the morphological and genetic relationship between tetralogy of Fallot and the septal pathologies observed in the other family members is unknown. The second scenario is motivated by the stark discrepancy of the observed phenotypes. Moreover, tetralogy of Fallot is relatively more common than familial septal aneurysms; therefore, this individual could constitute a phenocopy. The third linkage scenario was included to test the hypothesis that a monogenic trait is responsible for all cardiac malformations encountered in this family.
Candidate gene sequencing
The following candidate genes from the critical interval were sequenced bidirectionally in at least one affected and one unaffected individual from the pedigree: IDI1, IDI2, WDR37, and PFKP. Primers were designed using the Exon Primer interface integrated into the UCSC Genome Browser (Institut fuer Humangenetik, Munich, Germany, http://ihg2.helmholtz-muenchen.de/cgi-bin/primer/ExonPrimerUCSC.pl?db=hg18&acc=uc010jdh.1). Furthermore, we sequenced the known cardiac-specific candidate genes GATA4 and NKX2.5 in two individuals (III-1 and IV-1; all primer sequences available upon request).
Recruitment of additional family members
In addition to our previous report,10 more detailed phenotypic information became available for Individual II-8, previously reported to have hypertrophic obstructive cardiomyopathy. A detailed echocardiographic study now shows a marked apical septal aneurysm and dyskinesis. A similar finding was made in Individual II-2, in whom marked thinning of the apical septum was found, with slight bulging towards the right ventricular cavity, but without clear aneurysm formation (Figure 1). These defects are similar to those reported in the other family members, with the notable exception of Individual III-1, who carries a diagnosis of tetralogy of Fallot. Supplementary material online, Table S2 provides a synopsis resuming all clinical findings encountered in this family. A more detailed histological analysis of a sample of tissue obtained from Patient IV-1 in the course of surgical resection of the septal aneurysms demonstrated a thin ventricular septum with fibroelastotic thickening of the endocardium and fibrosis within the myocardium (Figure 2A). The fibrosis within the myocardium extended between individual myocytes. The cardiac myocytes have a variable morphology with some appearing atrophic and others showing changes of hypertrophy with nuclear enlargement. An elastin stain (Verhoff Van Gieson) confirmed the increased elastic fibres within the endocardium (Figure 2B). Electron microscopy did not reveal any structural changes of the sarcomeres (data not shown).
Two-dimensional echocardiography in family members recruited in addition to those included in our previous report.10 Arrowheads depict septal thinning/aneurysm. (A) Parasternal short-axis view in Individual II-2 showing marked septal thinning at mid-septal/apical level, with slight bulging of the thinned interventricular septum into the right ventricular cavity. (B) Apical four-chamber view in Individual II-8 of a marked apical aneurysm.
Histology of excised tissue of ventricular septal aneurysm of Patient IV-1. (A) Haematoxylin/eosin stain of surgical specimen at ×100 magnification. Arrows depict the thickened endocardium and arrowheads the invasion of myocardium with fibroblastic material. (B) Verhoeff Van Gieson staining for elastic fibres, ×100 magnification (blue-black). Arrows indicate the thickened endocardium and arrowheads the increase in elastic fibres and fibrotic material.
Linkage analysis and construction of haplotypes
As described above, we considered three distinct scenarios for linkage analysis.
In the first scenario, in which Individual III-1 with tetralogy of Fallot was considered unknown, parametric two-point analysis yielded a maximum LOD score of 2.32 on chromosome 12 at marker rs1995257 (117 857 059 bp). Multipoint analysis did not confirm any further evidence for linkage of this region. In parametric multipoint analysis, LOD scores above 2 were obtained only on chromosome 10 (marker rs4328141 (2 888 143 bp) to marker rs3814595 (3 201 429 bp)) with a maximum LOD score of 3.03 from marker rs7902082 (2 785 366 bp) to marker rs4328141 (2 877 893 bp) (see Supplementary material online, Table S1). Construction of the haplotype pedigree shows recombination events in Individuals III-8 and IV-1, yielding a telomeric border at marker rs1057304 (852 356 bp) and a centromeric border at marker rs3814595 (3 201 678 bp) (Figures 3 and 4).
Genome-wide LOD scores of for all three linkage scenarios. X-axis: concatenated chromosomes, drawn to scale. Y-axis: parametric multipoint LOD scores. Top, middle, and bottom: linkage scenarios 1, 2, and 3, respectively. Asterisk indicates the maximal LOD score in models 1 and 2.
Haplotype pedigree of the family described. Disease haplotypes are boxed. Filled symbols indicate affected clinical status; open symbols, unaffected clinical status. Circles indicate females; squares, males; slash, deceased. III-5, IV-1, and IV-3 are recombinant for disease allele. No DNA samples were available in individuals without haplotypes. Haplotypes were inferred for Individuals I-1, I-2, I-5, I-6, II-9, and III-7 (grey symbols).
In the second scenario, in which Individual III-1 with tetralogy of Fallot was considered to be unaffected, parametric two-point analysis yielded a maximum LOD score of 2.02 on chromosome 11 at marker rs1365406 (12 562 720 bp). Multipoint analysis did not confirm any further evidence for linkage of this region. In parametric multipoint analysis, LOD scores above 2 were obtained only on chromosome 10 between marker rs1029182 (1 965 937 bp) to rs3814595 (3 201 679 bp) with a maximum LOD of 3.29 from marker rs7902082 (2 785 366 bp) to rs4328141 (2 888 143 bp) (see Supplementary material online, Table S1; Figures 3 and 4). Construction of the haplotype pedigree yielded an identical result when compared with the first linkage scenario. Common to both scenarios was the exclusion of 84.8% of the genome at LOD scores of −2 and lower.
In the third scenario, in which Individual III-1 with tetralogy of Fallot was set to affected and therefore an obligate carrier of the disease haplotype, parametric multipoint analysis yielded a maximum LOD score of 1.88 on chromosome 11 at rs1050068 (11 769 948 bp).
In the first two scenarios, the disease haplotype segregating with ventricular septal aneurysm, VSD, and ASD is not shared by Individual III-1, who is affected with tetralogy of Fallot. A more detailed comparison shows no overlap of the phenotypes in Individual III-1 and the remainder of the family, i.e. he does not exhibit any septal thinning, and none of the other affected family members have heart disease reminiscent of tetralogy of Fallot such as seen in Individual III-1.
The critical interval spans a relatively gene-poor region of ∼2.3 Mb on chromosome 10p15, with a total of 12 genes (4 non-coding) and 30 different transcripts in this region according to the ENSEMBL database (see Supplementary material online, Table S3). The recombination rate in this interval varies between 2.0 and 5.2 cM/Mb. Cross-species comparisons reveal high conservation of long stretches of non-coding DNA among mammals. Several genomic copy number variants have been mapped in this region according to publicly available databases.17
We selected several candidate genes from this region based on their expression pattern as determined from publicly available databases for bidirectional sequencing. This mutational survey of IDI1, IDI2, WDR37, and PFKP did not reveal any variants thought to cause disease (see Supplementary material online, Table S4). Sequencing of the known candidate genes GATA4 and NKX2.5 in two affected individuals, the one with tetralogy of Fallot (III-1) and the index case individual (IV-I), was unremarkable (data not shown).
In this study, we provide further evidence for the genetic heterogeneity of septal defects by mapping a novel autosomal dominant trait of VSD/aneurysms and ASD to chromosome 10p15. The phenotypes observed in this family are all in septal locations, i.e. ventricular aneurysms, VSD, and ASD. The divergence of the phenotype in Individual III-1 with a tetralogy of Fallot is notable in this context. Interestingly, the most plausible linkage scenarios strongly suggest that this case of tetralogy of Fallot is genetically distinct from the septal defect phenotype observed in all other family members. Close examination of the linkage results (Figure 3) reveals that only the first two scenarios, in which Individual III-1 is set as either unknown or unaffected, are compatible with a statistically significant monogenic factor linked to disease. This is further corroborated by the haplotype analysis and hints to genetic heterogeneity of septal defects and a conotruncal defect within a single extended family. Despite the mapping of the disease-linked interval to the telomeric end of chromosome 10, we are confident that the recombinants observed on both sides of the critical interval avoid a misspecification of the linkage model with a false-positive signal, in particular due to the recombination between markers rs1057304 and rs1029182 in Individual III-8 (Figure 4).
Interestingly, despite the rarity of the clinical entity of congenital septal aneurysms, several familial cases are reported in the literature.18–21 Inheritance patterns in these families are consistent with autosomal dominance, and the anatomic characteristics of interventricular septal thinning and myocardial disease are similar to those reported in this current pedigree. In four of the five reported familial cases, ASD have been described as well. Tetralogy of Fallot has not been reported in those familial cases, compatible with Individual III-1 being a phenocopy in our family.
Prior linkage and candidate gene studies have assigned a role for several key developmental molecules in human septal defects, and genes identified through linkage analysis in extended families or in animal models have been used successfully as candidates in sporadic cases of congenital heart disease (for review, see Andelfinger2). Some genotype–phenotype correlations emerge from these studies, such as the association of atrioventricular block with NKX2.5 mutations and that of pulmonary stenosis with GATA4 mutations.22 However, the prototypical clinical phenotype in the family described here does not fit any of the known genetic causes of septal defects reported to date, nor does the critical interval harbour any obvious candidates previously implicated in heart development. Of note, no cardiac-specific transcription factor or signalling molecule was found nor is any of the biological functions associated with the positional candidates particularly suggestive for being involved in cardiovascular phenotypes. According to database searches, several of the positional candidates are expressed in the heart; however, detailed expression studies during the development of the heart have not yet been undertaken. Unfortunately, the changes found upon histological examination of heart tissue did not provide any additional clues as to a possible prioritization of candidate genes.
Also, the critical interval clearly lies in a recombination hotspot as indicated through the regional recombination rate.23 The paucity of clear functional candidates in the critical interval as defined by haplotype analysis as well as its 2.3 Mb size raise the possibility that the disease-causing sequence change may not affect known coding, but rather non-coding regulatory sequence.24–27 A comparative genomic analysis shows that this interval harbours long conserved segments of non-coding sequence. In fact, recent studies suggest that long-range genetic effects in non-coding sequences can control gene expression and be involved in disease pathogenesis.28 As such, KLF6, a Kruppel-like factor strongly expressed in the heart and the developing vasculature, is in the direct proximity of the critical interval and may underlie control through elements within it.29
Previous studies have identified the 300 kb interval between D10S547 and D10S585 on chromosome 10p14 as critical for DiGeorge syndrome type 2.30,31 Interestingly, the critical interval mapped in the family described here does not overlap with this interval. However, genotype–phenotype correlation of patients with 10p syndromes has revealed that the most telomeric deletions involving 10p15 are associated with septal defects (mostly ASD II and VSD) or other congenital heart disease in at least 15 of 21 patients.32 Yet, no ventricular aneurysms have been reported in those patients. A recent study assessing copy-number variation at a genome-wide level in tetralogy of Fallot has identified several novel loci associated with septal defects, but not the novel locus described here.33 Taken together, we speculate that the critical interval mapped in our study harbours novel candidate sequences which may be different from those that are causative in DiGeorge syndrome type 2 or that the mutation exerts its effects through mechanisms other than haploinsufficiency. Future studies of this unique family should therefore include exhaustive resequencing of the entire critical interval, such as made possible now through next-generation sequencing techniques.
The finding of a novel locus for septation defects extends the concept that congenital heart disease can be influenced by strong genetic determinants. The additional finding of septal aneurysms provides a unique opportunity to tease apart disease processes which lead to these overlapping phenotypes. The genetic heterogeneity of these conditions is further highlighted by the fact that within a single extended family, a single member with the more common phenotype of tetralogy of Fallot does not contribute to the linkage signal. All molecular cascades of chamber septation may not lead to disease in the same fashion when disturbed, and divergent pathogenetic events rather than a final common pathway may have been triggered. Further molecular investigation of the rare entity in this family will provide a unique opportunity to deepen our understanding of the mechanisms governing septal formation.
This work was supported by the Canadian Institutes of Health Research, the Heart and Stroke Foundation of Canada, GMHD79045 (to G.A.), and the André Nussia Aisenstadt Foundation. G.A. is a Clinician-Scientist of the Canadian Institutes of Health Research. N.T. is a scholar of the Fondation des Étoiles/Fondation CHU Sainte-Justine and the University of Montréal/Bourse d'excellence de la Faculté des Études supérieures. M.-P.H. is supported by a scholarship from the German Heart Foundation.
Conflict of interest: none declared.
We gratefully acknowledge the contribution of all family members and supporting staff to this project.
. Mutation in the alpha-cardiac actin gene associated with apical hypertrophic cardiomyopathy, left ventricular non-compaction, and septal defects. Eur Heart J 2007;16:1953-1961. doi:10.1093/eurheartj/ehm239.
. Autosomal dominant myocardial disease diagnosed by fetal presentation of proband with an aneurysm of the muscular interventricular septum. Pediatr Cardiol 2002;23:27-31. doi:10.1007/s00246-001-0007-1.
. Disease-causing 7.4 kb cis-regulatory deletion disrupting conserved non-coding sequences and their interaction with the FOXL2 promotor: implications for mutation screening. PLoS Genet 2009;5:e1000522. doi:10.1371/journal.pgen.1000522.
. Expression and mutation analysis of BRUNOL3, a candidate gene for heart and thymus developmental defects associated with partial monosomy 10p. J Mol Med 2002;80:431-442. doi:10.1007/s00109-002-0331-9.