SMAD1 Loss-of-Function Variant Responsible for Congenital Heart Disease

As the most common form of developmental malformation affecting the heart and endothoracic great vessels, congenital heart disease (CHD) confers substantial morbidity and mortality as well as socioeconomic burden on humans globally. Aggregating convincing evidence highlights the genetic origin of CHD, and damaging variations in over 100 genes have been implicated with CHD. Nevertheless, the genetic basis underpinning CHD remains largely elusive. In this study, via whole-exosome sequencing analysis of a four-generation family inflicted with autosomal-dominant CHD, a heterozygous SMAD1 variation, NM_005900.3: c.264C > A; p.(Tyr88∗), was detected and validated by Sanger sequencing analysis to be in cosegregation with CHD in the whole family. The truncating variation was not observed in 362 unrelated healthy volunteers employed as control persons. Dual-luciferase reporter gene assay in cultured COS7 cells demonstrated that Tyr88∗-mutant SMAD1 failed to transactivate the genes TBX20 and NKX2.5, two already well-established CHD-causative genes. Additionally, the variation nullified the synergistic transcriptional activation between SMAD1 and MYOCD, another recognized CHD-causative gene. These data indicate SMAD1 as a new gene responsible for CHD, which provides new insight into the genetic mechanism underlying CHD, suggesting certain significance for genetic risk assessment and precise antenatal prevention of the family members inflicted with CHD.


Introduction
Congenital heart disease (CHD) constitutes the most frequent type of birth deformity in humans, inflicting~1% of live births globally [1,2]. Based on cardiac anatomic abnormalities, CHD is clinically categorized into >20 distinct types, including pulmonary stenosis (PS), ventricular septal defect (VSD), atrial septal defect, patent ductus arteriosus (PDA), and tetralogy of Fallot [1]. CHD may result in neurodevelopmental abnormality, thromboembolic complications, infective endocarditis, pulmonary hypertension, heart failure, arrhythmias, and death [3,4]. Tremendous advancement has been achieved in pediatric cardiovascular surgery during recent decades, which enables the overwhelming majority (up to 97%) of neonates with CHD to survive childhood and reach adulthood, and now, CHD adults outnumber CHD children [5,6]. Unfortunately, prolonged life span has led to an increasing number of adult CHD survivors, who are prone to suffering from miscellaneous late complications, with heart failure and cardiac arrhythmias being the most prominent [7,8]. Consequently, CHD has conferred substantial morbidity and mortality as well as socioeconomic burden on humans [1]. Despite clinical importance, the etiologies accountable for CHD remain largely obscure.
It has been demonstrated that cardiac morphogenesis is a sophisticated biological process and both genetic defects and nongenetic precipitating risk factors may disturb this complex process, giving rise to CHD [2,[9][10][11][12][13][14]. Welldocumented nongenetic risk factors predisposing to CHD include maternal diabetes, folate deficiency, viral infections, autoimmune disorder, and environmental exposures to air pollutants and medications [9]. However, aggregating evidence indicates that genetic components exert a key effect  III-9 -III-10 III-11 -III-12 +   IV-6  IV-7 +   III-3 III-4   II-2 -II-3 +  II-5 +  II-4 -I-1  I-2   Family 1   II-6 -II-8 -II-7 III-13 III-14 on the occurrence of CHD, and in addition to chromosomal alterations (aneuploidies) and copy number variations, pathogenic mutations in >100 genes, including TBX20, NKX2.5, and MYOCD, have been involved in the occurrence of CHD [2,[10][11][12][13][14]. Nevertheless, the genetic culprit components for CHD in up to 55% of cases remain unveiled [12]. Hence, there is still much research work to be fulfilled to show a complete picture of genetic causes for CHD. Approval of the research protocol was achieved from the local institutional medical ethics committee, with an ethical approval number of LL(H)-09-07. Written informed consent was provided by the study participants or their parents. For this research, a four-generation pedigree with high incidence of autosomal-dominant CHD was enlisted. A total of 362 unrelated ethnicity-matched volunteers without CHD were recruited as control individuals. All research participants experienced a comprehensive clinical investigation, encompassing review of personal and medical histories as well as familial histories, careful physical examination, transthoracic echocardiogram, and electrocardiogram. Diagnosis of CHD was made as previously described [15]. Peripheral blood specimen was collected from each study subject, and genomic DNA was prepared from blood leucocytes of each study subject.

Molecular Genetic Studies.
For a study participant, a whole-exome library was prepared using 2 μg of genomic DNA and captured with the SureSelect Human All Exon V6 Kit (Agilent Technologies), as per the manufacturer's manual. The exome library was enriched and then sequenced on the Illumina HiSeq 2000 Genome Analyzer (Illumina) using the HiSeq Sequencing Kit (Illumina), fol-lowing the protocol. Bioinformatics assays of the data produced by whole-exome sequencing (WES) were performed as previously described [16][17][18][19]. The candidate variants identified by WES and bioinformatical analyses of the DNA samples from the CHD family underwent Sanger sequencing analysis in the whole family with CHD. For a verified genetic variation, the entire coding region and splicing donors/acceptors of the gene were sequenced in all available family members of the family with CHD and 362 unrelated control persons and such population genetics database as the Single Nucleotide Polymorphism (SNP) Database (https://www.ncbi.nlm.nih.gov/) and the Genome Aggregation Database (gnomAD; https://gnomad .broadinstitute.org/) were consulted to check its novelty.

Statistical
Analysis. The activity of a promoter was given as a ratio of firefly luciferase to Renilla luciferase. Values for promoter activity were expressed as mean ± standard deviation (SD) of the results from three transfection experiments in triplicate. Student's t-test was applied to statistical analysis. A two-sided p < 0:05 denoted statistical difference.

Clinical Characteristic Information of the Study Pedigree.
In the present investigation, a large family with CHD spanning four generations (Figure 1(a)) was enrolled from the Chinese Han-race population. In this Chinese family, there were 32 family members, including 30 living members (15 male members and 15 female members, with ages ranging from 3 to 55 years) and all the nine affected members had echocardiogram-documented PDA. In addition, four members also suffered from VSD and three members also suffered from PS. Genetic analysis of this four-generation pedigree (Figure 1(a)) revealed that PDA was inherited in an autosomal-dominant fashion, with 100% penetrance. The proband, a three-year-old boy, underwent catheterbased cardiac repairment due to PDA and VSD. The proband's affected relatives also underwent interventional procedures for correction of CHD, except for his mother's grandfather (I-1), who died of congestive heart failure at the age of 69 years. No recognized noninherited risk factors contributing to CHD were ascertained in all the family

3.2.
Detection of a CHD-Causing SMAD1 Mutation. WES was carried out in four CHD-inflicted family members (including III-5, III-12, IV-4, and IV-7) and three unaffected family members (including III-6, III-11, and IV-3) of the family (Figure 1(a)   Three cellular transfection experiments were repeated in triplicate for every expression plasmid. All values are given as the mean ± standard deviation of the results from three independent reporter gene analyses. Both * and * * indicate that p < 0:005, in comparison with their corresponding wild-type counterparts.

Discussion
In the present study, a four-generation Chinese family suffering from CHD transmitted as an autosomal dominant trait was enrolled. Via WES analysis of the DNA samples from the family members, a novel variation in the SMAD1 gene, NM_005900.3: c.264C > A; p.(Tyr88 * ), was identified, which was confirmed by Sanger sequencing analysis to be in cosegregation with CHD in the whole family. The heterozygous variation was absent from 724 control chromosomes nor reported in the databases of SNP and gnomAD. Reporter gene assays demonstrated that Tyr88 * -mutant SMAD1 failed to transcriptionally activate the promoters of TBX20 and NKX2.5, two well-established CHD-causing genes [22,23]. Moreover, the mutation nullified the synergistic transcriptional activation between SMAD1 and MYOCD, another CHD-causative gene [11,[24][25][26]. These observational results indicate that genetically compromised SMAD1 predisposes to CHD.
SMAD1 maps on human chromosome 4q31.21, coding for a protein comprising 465 amino acids, a member of the SMAD superfamily of proteins like the products of the Sma gene from Caenorhabditis elegans and the Mad (Mothers against decapentaplegic) gene from Drosophila [27,28]. As a transcription factor and signal transducer that regulates multiple signal pathways, the SMAD1 protein possesses two evolutionarily conserved structural domains, MAD homology 1 (MH1) and MAD homology 2 (MH2), which are separated by Linker [29]. MH1 functions mainly to bind to the DNA consensus sequence of GNCN in target gene promoters and to transcriptionally activate the expression of target genes, in addition to mediating nuclear accumulation of SMAD1 and interaction of SMAD1 with other transcription factors. MH2 is responsible for the interaction of SMAD1 with a wide variety of proteins, provides selectivity and specificity to SMAD1 function, contributes to the binding affinity, and is involved in nuclear accumulation of SMAD1. Linker connecting MH1 and MH2 contains multiple critical peptide motifs, encompassing a nuclear export signal and several potential phosphorylation sites, and hence has a role in transcriptional activation [29]. SMAD1 is highly expressed in the heart throughout embryogenesis, playing a crucial role in transactivating the expression of target genes essential for cardiovascular morphogenesis, including TBX20, NKX2.5, ACTC1, and MYH6, alone or synergistically with MYOCD and TBX20 [30][31][32]. Furthermore, loss-of-function variations in TBX20, NKX2.5, ACTC1, MYH6, and MYOCD have been involved in the molecular pathogenesis of CHD [11,22,23,[33][34][35][36]. In the current study, the variation discovered in cases with familial CHD was predicted to generate a truncating SMAD1 protein lacking MH2 and Linker as well as a part of MH1 and functional data demonstrated that Tyr88 * -mutant SMAD1 failed to transactivate its downstream target genes. These findings support that haploinsufficiency of SMAD1 is the genetical mechanism of CHD which occurred in this family.
It may be ascribed to abnormal cardiovascular development that SMAD1 mutation gives rise to CHD. In most animal species, including xenopus, zebrafish, mice, rats, and humans, SMAD1 is amply expressed in the cardiovascular system during embryonic development, playing a pivotal role in cardiovascular morphogenesis via mediating the cellular proliferation, growth, apoptosis, differentiation, and morphogenesis in the heart and vessels [29,37]. In mice, homozygous knockout of Smad1 led to embryonic demise because the embryos failed to seed the placenta and the Smad1-null embryos showed markedly impaired allantois formation and drastically reduced primordial germ cells; while the embryos with heterozygous deletion of Smad1 developed remarkably normally, probably due to functional compensation by Smad5 and Smad8, which shared common 6 BioMed Research International expression profiles and functional characteristics with Smad1 [38,39]. Although the mice with heterozygous knockout of either Smad1 (Smad1 +/− ) or Smad5 (Smad5 +/− ) developed properly, the murine embryos with double heterozygous knockout of Smad1 and Smad5 (Smad1 +/− /Smad5 +/− ) died by E10.5 and the double heterozygous embryos (Smad1 +/− /Smad5 +/− ) presented defects of heart looping and laterality [39]. Furthermore, the mice with conditional knockout of the Smad1 gene by disrupting Smad1 either in endothelial cells or in smooth muscle cells displayed increased pulmonary pressure, right ventricular hypertrophy, and thickened pulmonary arterioles [40]. Collectively, these results from experimental animals suggest that genetically compromised SMAD1 predisposes to CHD in human beings.

Conclusions
The current investigation indicates SMAD1 as a novel causative gene responsible for CHD, which provides a new potential target for antenatal prophylaxis and personalized treatment of CHD patients.

Data Availability
The data supporting the findings of this investigation are available upon reasonable request.

Conflicts of Interest
The authors declare that they have no conflicts of interest.