Materials and Methods

Patients

In this descriptive study, TBX5 gene mutations in patients with ASD and VSD were investigated in 100 cases of congenital heart disease in two hospitals, the Sanandaj Tohid Hospital and Kermanshah Cardiovascular Center. The target Kurdish population consisted of 50 healthy controls who had been admitted to hospital (with no history of familial disease and who showed no cardiac abnormality by echocardiogram ) and 100 cases (62 females and 38 male) of non-syndromic patients including 57 patients with VSD (42 membranous, 14 muscular, 1 conal septal), 39 patients with ASD (36 ostium secundum, 2 sinus venosus, 1 without information) and four patients possessing AVSD (Atrioventricular) (Table 1).

Extracting Genomic DNA

Genomic DNA was extracted from whole blood by AccuPrep® Genomic DNA Extraction Kit from Bioneer Corporation, South Korea.

  Real Time PCR and HRM Analysis

In order to replicate the coding regions of the TBX5 gene (exons 2-9) specific primers for the reference sequence NM_000192.3 and NG_007373.1 were designed in NCBI using CLC Workbench 5.5 software. Gene Reference Sequence ID is LRG_670 (Table 2). The intron/exon boundary (exons 2 to 8) and exon 9 (containing 2177bp), its 575bp component, which is part of the coding region, was considered for the design of the primer. In this study, the Rotor-Gene ™ 6000 (Corbett Research) and 2x QuantiFast SYBR® Green kits were used. In order to optimize the PCR reaction as well as HRM, the PCR reaction was performed for all primers under the same conditions as follows: 95 °C for 5 minutes and initial denaturation and 40 cycles of 95 °C for 10 seconds and 60 °C for 30 seconds and the HRM step from 72 °C to 95 °C, including an increase in temperature of 0.1 °C per second. At first, 10 healthy individuals were identified as sequencing controls and were used as control (wild type) for HRM.

Sequencing

Patient samples were grouped according to the HRM curve and samples whose melting patterns differed from healthy subjects or controls (Figure 1) were sequenced. Also, 25 samples of patients and 25 controls with normal curves were randomly selected for sequencing. The sequencing of PCR products by ABI 3730XL DNA Analyzers was performed by Bioneer (South Korea) and the results were analyzed using the CLC main workbench.

Predict structural stability of mutation

The prediction of the structural stability of a protein derived from amino acid mutation was performed based on energy changes by using the sequence information online at the MUpro site (http://www.ics.uci.edu/~baldig/mutation.html) using the SVM method and the support vector machine and neural network (Cheng, Randall, & Baldi, 2006; Cheng, Randall, Sweredoski, & Baldi, 2005).

Potential pathogenicity prediction of TBX5 gene sequence variants using PolyPhen-2

Potential Pathogenicity of TBX5 sequence variants was tested based on two standard human datasets, HumDiv and HumVar, which are applied by PolyPhen-2 Online (http://genetics.bwh.harvard.edu/pph2) (Figure 3). The HumDiv dataset (Adzhubei et al., 2010) contains 5564 mutations in the UniProtKB database known to cause Mendelian disease and 7,539 DNA variants between human proteins and their closely related mammalian equivalents assumed to be non-damaging. The HumVar dataset (Capriotti, Calabrese, & Casadio, 2006) consists of 22,196 human disease-causing mutations from UniProtKB and 21,151 neutral mutations that are common human nsSNPs (Non-synonymous single nucleotide polymorphisms) (MAF > 1 %) with no annotated disease involvement. Note that the HumDiv dataset contains annotated mutations directly associated with human diseases, while the HumVar dataset is more noisy as its neutral mutation subset includes many mildly deleterious alleles (Adzhubei et al., 2010; Plekhanova, Nuzhdin, Utkin, & Samsonova, 2019).