Whole exome sequencing identifies novel mutation in eight Chinese children with isolated tetralogy of Fallot

Background Tetralogy of Fallot is the most common cyanotic congenital heart disease. However, its pathogenesis remains to be clarified. The purpose of this study was to identify the genetic variants in Tetralogy of Fallot by whole exome sequencing. Methods Whole exome sequencing was performed among eight small families with Tetralogy of Fallot. Differential single nucleotide polymorphisms and small InDels were found by alignment within families and between families and then were verified by Sanger sequencing. Tetralogy of Fallot-related genes were determined by analysis using Gene Ontology /pathway, Online Mendelian Inheritance in Man, PubMed and other databases. Results A total of sixteen differential single nucleotide polymorphisms loci and eight differential small InDels were discovered. The sixteen differential single nucleotide polymorphisms loci were located on Chr 1, 2, 4, 5, 11, 12, 15, 22 and X. Among the sixteen single nucleotide polymorphisms loci, six has not been reported. The eight differential small InDels were located on Chr 2, 4, 9, 12, 17, 19 and X, whereas of the eight differential small InDels, two has not been reported. Analysis using Gene Ontology /pathway, Online Mendelian Inheritance in Man, PubMed and other databases revealed that PEX5, NACA, ATXN2, CELA1, PCDHB4 and CTBP1 were associated with Tetralogy of Fallot. Conclusions Our findings identify PEX5, NACA, ATXN2, CELA1, PCDHB4 and CTBP1 mutations as underlying genetic causes of isolated tetralogy of Fallot.


INTRODUCTION
Tetralogy of Fallot (TOF) is the most common cyanotic congenital heart disease (CHD) with an incidence rate of 1/3600 in live births and 10% in CHD [1]. TOF is characterised by cardiac outflow tract malformation caused by non-uniform separation between truncus arteriosus and bulbus arteriosus during embryo stage. TOF pathological features include ventricular septal defect, aortic overriding, right ventricular outflow tract stenosis or pulmonary artery stenosis and right ventricular hypertrophy. In neonatal period, children with TOF may manifest oxygen www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 63), pp: 106976-106988 Research Paper deficiency, pneumonia, intractable congestive heart failure and other complications with high early mortality. Despite TOF treatment in children, some complications will still occur because of poor prognosis. TOF brings heavy burden to the family and society. Therefore, to investigate TOF etiology, possible pathogenesis and risk factors for prenatal diagnosis and counselling and prognostic evaluation is of great significance.
Heart development is a complex and orderly process including cardiac tube formation, loop formation, intracardiac separation and vascular connection. Heart development is related to many genes, which express and interact in different spaces and at different times to form a precise regulatory mechanism. Abnormal expression occurring in any of these genes is likely to affect heart development, leading to heart malformation. For TOF, embryonic developmental mechanism is nearly definite. However, the molecular pathogenesis remains to be clarified. TOF is associated with gene mutation [2]. TOFrelated genes include NKX2-5 [3], GATA4 [3,4], and JAG-1 [5,6]. Approximately 15% of TOF are from 22q11 microdeletion syndrome [7] caused by chromosome 22 long arm 1 zone 1 subzone deletion and manifest extracardiac abnormalities, such as abnormal face, thymic hypoplasia, cleft palate and hypocalcemia. In addition, TOF is also related to chromosomal aneuploidy, and approximately 3% of TOF is from 21-trisomy syndrome [8]. However, the pathogenesis of isolated TOF is still to be elucidated.
Whole exome sequencing (WES) is suitable to high-throughput sequencing for all genomic exon regions. Human exome accounting for approximately 1%-2% of genome contains important information of protein synthesis, which directly reflects gene function. In most diseases, related mutations are located in the exome region. WES may be used to find pathogenic gene and predisposing gene in complex diseases, monogenic diseases and cancer because it can investigate protein encoding information in several individuals; its data are also accurate [9]. The aim of this study was to find TOF-related pathogenic genes through WES technology performed among eight small families with TOF, providing a basis for studying TOF pathogenesis.

Clinical features
All the eight TOF children manifested cyanosis and rough systolic-ejection murmurs at the left sternal border between the second and fourth ribs. Echocardiography showed right ventricular enlargement, right ventricular anterior wall thickening, right ventricular outflow tract and pulmonary artery stenosis, ventricular septal defect and aortic overriding ( Figure 1). The echocardiographic results of eight TOF children in the right ventricle, right ventricular anterior wall, right ventricular outflow tract, main pulmonary artery, left pulmonary artery, right pulmonary artery, ventricular septal defect and left ventricular ejection fraction ( Table 1). The eight TOF children displayed no other malformations, and their parents showed no abnormalities.

Mutation detection
A total of sixteen differential single nucleotide polymorphisms (SNPs) loci and eight differentially small InDels were discovered by sequencing all exons in the eight small families and alignment of SNPs and small InDels within families and between families ( Table 2 and  Table 3).
We discovered fourteen SNPs-and Small InDels-

Results verified by Sanger sequencing
We designed twenty-four pairs of primers to verify sixteen differential SNPs loci and eight differentially small InDels. The results of Sanger sequencing are consistent with that of WES ( Figure 2).
The verification of sixteen differential SNPs sites included NACA (rs745387506, T>G),

Gene ontology (GO)/pathway analysis
GO analysis indicated the percentages of 22 genes enriched in GO term ( Figure 3). Growth was related to CHD, and it was associated with some genes including PEX5, NACA, ATXN2 and CELA1 (Table 5). PEX5, ATXN2 and CELA1 are responsible for multicellular organism growth (GO:0035264). PEX5, ATXN2, NACA and CELA1 are responsible for developmental growth (GO:0048589). PEX5, ATXN2 and NACA are responsible for regulation of developmental growth (GO:0048638) and growth regulation (GO:0040008). PEX5 and ATXN2 are responsible for regulation of multicellular organism growth (GO:0040014). PEX5 and NACA are responsible for positive regulation of developmental growth (GO:0048639) and positive growth regulation (GO:0045927). PEX5 is responsible for positive regulation of multicellular organism growth (GO:0040018). ATXN2 is responsible for negative regulation of multicellular organism growth (GO:0040015), epidermal growth-factor receptor binding (GO:0005154), negative regulation of developmental growth (GO:0048640), growth-factor receptor binding (GO:0070851) and negative growth regulation (GO:0045926).  Repression of Wnt target genes and Notch signalling were related to CHD, and was associated with PCDHB4 and CTBP1 genes ( Table 6). Pathway analysis showed that twenty-two genes were enriched at the top fifty of pathway term (

Variant analysis
ChrX 300502 a nuclear-encoded mitochondrial multienzyme complex that catalyzes the overall conversion of pyruvate to acetyl-CoA and CO (2), and provides the primary link between glycolysis and the tricarboxylic acid cycle Pyruvate dehydrogenase E1-alpha deficiency of PCDHB4 was highly evolutionarily conserved in diverse species including Homo sapiens, Pan paniscus, Pongo abelii and gorilla (Figure 5a). In addition, Y5S of CELA1 was highly evolutionarily conserved in diverse species including H. sapiens, gorilla, Saimiri boliviensis, Papio anubis, Peromyscus maniculatus, Tapirus bairdii, Eptesicus fuscus and Sus scrofa (Figure 5b). The highly evolutionarily conserved protein sequences were not found in PEX5, NACA, ATXN2 and CTBP1 in other species.

DISCUSSION
TOF as the most common cyanotic CHD is characterised by right ventricular outflow tract obstruction, pulmonary stenosis, ventricular septal defect and aortic overriding caused by non-uniform separation between the truncus arteriosus and bulbus arteriosus at embryo stage. Currently, surgical repair is usually performed in the first month after birth to relieve the stenosis of the right outflow tract and close ventricular septal defect, enabling an exclusive ejection of oxygenated blood via the left ventricle. Cumulative survival and event-free survival were 72% and 25%, respectively, 40 years after follow-up [12].
TOF etiology still needs to be completely clarified because of TOF complexity attributed by both genetic and nongenetic effectors. Next-generation sequencing technology, such as WES with rapid, high throughput and cost-effective features, can meet the requirements for medical studies. WES has been used to obtain information about genetic alterations and potential predispositions possibly associated with TOF occurrence.
Eight small families were tested through WES, and differential SNP loci and InDels were found. From the comparison within each family and among the families, sixteen differential SNPs loci and eight differential Small InDels were screened out. Corresponding genes exist in the database, wherein ten of the sixteen differential SNPs have been reported and six of the sixteen differential SNPs have not been reported. Among the eight Small InDels, seven small Indels have corresponding genes in the database and one Small InDels does not have. Six of the eight differential Small InDels have been reported, whereas two of them have not. Among them, no gene in the database corresponds to the mutation of chr4, position: 13632182. It might be located among the genes or among noncoding regions, and the function requires further study. Go/Pathway analysis of the sixteen differential SNPs loci and eight differential Small InDels were conducted and PEX5, NACA, ATXN2, CELA1, PCDHB4 and CTBP1 were finally screened out and were associated with TOF.
Similar to some genes found to be related to TOF by detection of chromosome karyotype, gene copy number variation and de novo mutation (Table 7), Greenway et al. [13] found four genes (PRKAB2, CHD1L, BCL9 and GJA5) with the highest expression in the right ventricular outflow tract, which is malformed in TOF. Bittel et al. [14] included both idiopathic TOF patients and three patients, which harbour the syndromic 22q11.2 deletion. The expression of the genes A2BP1, VEGF, NPPA and NPPB located in the 22q11.2 region was half-reduced among syndromic patients as expected, whereas none of these genes was differentially expressed in any of the idiopathic TOF subjects. Notch pathway was also   Our study aimed to identify other monogenic causes of TOF. GO analysis indicated that PEX5 was related to multicellular organism growth (GO:0035264), developmental growth (GO:0048589), regulation of developmental growth (GO:0048638), regulation of multicellular organism growth (GO:0040014), positive regulation of developmental growth (GO:0048639), positive regulation of growth (GO:0045927), growth (GO:0040007) and positive regulation of multicellular organism growth. In the 26 cases with TOF, Grunert M et al. [20] found deleterious SNPs in 16 genes, including  BARX1, BCCIP, DAG1, EDN1, FANCL, FANCM, FMR1,  FOXK1, HCN2, MYOM2, PEX6, ROCK1, TCEB3,  TP53BP2, TTN and WBSCR16 which resulted in the imbalance of function networks in TOF. Huffnagel IC et al. [21] found that two of the 21 genetically proven Rhizomelic chondrodysplasia punctata patients suffered from TOF and showed PEX7 mutations. Plasmalogens were not detected in cardiac tissue of PEX7 knock-out mice, which is a model for RCDP type 1. The abovementioned PEX5, PEX6 and PEX7 are all PEX family gene named Peroxins (PEXs) which are proteins essential for the assembly of functional peroxisomes. Therefore, PEX5 might participate in TOF growth and development process. PCDHB4 is located on Chr5q31.3. This gene is a member of the protocadherin beta gene cluster and is one of the three related and tandemly linked gene clusters at chromosome 5. Their specific functions are unknown but they most likely play a critical role in the establishment and function of specific cell-cell neural  connections. Alazami AM et al. [22] performed WES on 143 multiplex consanguineous families, in which known disease genes had been excluded by autozygosity mapping. Patients, whose clinical phenotype is featured by global developmental delay and brain atrophy, have mutation in PCDHB4 (NM_021908:c.489T>G:p.Y163X) based on the candidate gene analysis, and the function of which participated in the growth and development process. Base on our research, Go analysis indicated that PCDHB4 (P00057) was related to WNT signaling pathway. Thus, PCDHB4 (P00057) may be a potential TOF candidate gene. NACA is located on Chr12q13.3. The protein encoded by NACA is associated with basic transcription factor 3, which forms the nascent polypeptide-associated complex. In this study, GO analysis indicated that NACA was related to positive regulation of developmental growth (GO:0048639) and positive regulation of growth (GO:0045927). Thus, NACA may be a potential TOF candidate gene. CELA1 is located on Chr12q13.13. This gene forms a subfamily of serine proteases that hydrolyze many proteins in addition to elastin. Humans possess six elastase genes, which encode the structurally similar elastases 1, 2, 2A, 2B, 3A and 3B. Unlike other elastases, pancreatic elastase 1 is not expressed in the pancreas. GO analysis indicated that CELA1 was related to multicellular organism growth (GO:0035264), developmental growth (GO:0048589) and growth (GO:0040007). Thus, CELA1 may have a potential relevance to TOF. ATXN2 is located on Chr12q24.12 with a transcript number of ENST00000608853.5. This gene belongs to a group of This study has some limitations, and the sample size is small. In the future, the number of samples will be further increased and an in-depth research will be conducted. The identified mutation sites have not been verified based on large TOF sample (n>50). Thus, whether the identified mutation sites may occur in a large group of TOF people has not been confirmed. Moreover, further study is necessary to verify relevant function genes and accordingly illustrate the pathogenesis of TOF.
In summary, TOF pathogenesis is complex because TOF is associated with more than one genetic variant. Eight small families were observed through WES and found that TOF-related genes were PEX5, NACA, ATXN2, CELA1, PCDHB4 and CTBP1, whose potential function might participate in TOF growth and development process. Although none of the three variants were predicted to be highly influential, they are in genes belonging to networks involved in relevant developmental pathways. Further study on the potential developmental mechanisms will be conducted based on these results.

Subjects
A total of eight small families whose Chinese children received surgical TOF treatment at the Children's Heart Center of Henan Provincial People's Hospital between January 2016 and June 2016 were enrolled in this study. The eight children ageing 3-11 months included four male babies and four female babies. All manifested isolated TOF. Their parents were normal. All the eight children were diagnosed with TOF by echocardiogram, clinical symptoms and signs and intraoperative findings.

DNA extraction
In each member of the eight small families including 24 individuals, 600 μl peripheral blood was collected to extract genomic DNA using genomic DNA Extraction Kit (Qiage, Germany). The remaining blood samples were stored at −80°C.

WES and mutation screening
WES and subsequent variant annotation were performed on genomic DNA derived from the eight TOF children and their parents. Paired-end libraries were prepared according to the manufacturer's protocols (Agilent). Exome libraries for the eight families were constructed using Agilent SureSelectXT Target Enrichment System according to Illumina Paired End Sequencing Protocol (Agilent Technologies, CA, USA). Capturing of whole exon was carried out according to the protocol of Agilent's Sure Select Human All 1 UTRs 71 MB v4 kit. The flow cells were sequenced as paired-end 150 base pair reads on an Illumina HiSeq ×10 platform to a minimum depth of 50× targeted region coverage using TruSeq SBS sequencing kit version 3 and HiSeq data collection version 2.0.12.0 software (Illumina, Inc., San Diego, CA, USA).
The raw sequence reads we obtained were aligned to the human genome reference sequence (hg19) using Burrows-Wheeler Aligner (BWA) with standard parameters [10]. The BWA-aligned reads were statistically calculated using PICARD software to exclude polymerase chain reaction duplicates. Regional realignment and quality score recalibration were carried out using Genome Analysis Toolkit [11] with recommended parameters, which included local realignment of the sequences around InDels, base quality score recalibration, variant calling and variant quality score recalibration.

Variant detection in the eight TOF children by Sanger sequencing
BigDye ® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) and ABI 3130 Genetic Analyzer were used to detect variants among the eight TOF children.

Ethics statement
Approval was obtained from local ethics committees of Henan Provincial People's Hospital, Zhengzhou www.impactjournals.com/oncotarget University People's Hospital. Informed consents were provided to all the participants of this study.

Author contributions
Lin Liu designed the experiments, recruited patients, collected samples, analysed the data, performed statistics and wrote the manuscript. Hong-dan Wang designed the experiments, recruited the patients, collected samples, analysed the data, performed statistics and wrote the manuscript. Cun-ying Cui recruited the patients, collected the samples, analysed the data and wrote the manuscript. Yun-yun Qin collected samples, analysed the data and wrote the manuscript. Tai-bing Fan collected the samples. Bang-tian Peng collected the samples. Lian-zhong Zhang performed the statistics. Cheng-zeng Wang designed the experiments, recruited patients and critically reviewed the manuscript.