Novel mutations of TCTN3/LTBP2 with cellular function changes in congenital heart disease associated with polydactyly

Abstract Congenital heart disease (CHD) associated with polydactyly involves various genes. We aimed to identify variations from genes related to complex CHD with polydactyly and to investigate the cellular functions related to the mutations. Blood was collected from a complex CHD case with polydactyly, and whole exome sequencing (WES) was performed. The CRISPR/Cas9 system was used to generate human pluripotent stem cell with mutations (hPSCs‐Mut) that were differentiated into cardiomyocytes (hPSC‐CMs‐Mut) and analysed by transcriptomics on day 0, 9 and 13. Two heterozygous mutations, LTBP2 (c.2206G>A, p.Asp736Asn, RefSeq NM_000428.2) and TCTN3 (c.1268G>A, p.Gly423Glu, RefSeq NM_015631.5), were identified via WES but no TBX5 mutations were found. The stable cell lines of hPSCs‐LTBP2mu/TCTN3mu were constructed and differentiated into hPSC‐CMs‐LTBP2mu/TCTN3mu. Compared to the wild type, LTBP2 mutation delayed the development of CMs. The TCTN3 mutation consistently presented lower rate and weaker force of the contraction of CMs. For gene expression pattern of persistent up‐regulation, pathways in cardiac development and congenital heart disease were enriched in hPSCs‐CM‐LTBP2mu, compared with hPSCs‐CM‐WT. Thus, the heterozygous mutations in TCTN3 and LTBP2 affect contractility (rate and force) of cardiac myocytes and may affect the development of the heart. These findings provide new insights into the pathogenesis of complex CHD with polydactyly.


| INTRODUC TI ON
Congenital heart disease (CHD) is the most common congenital anomaly with a worldwide occurring in about 9 per 1000 live births and accounts for nearly one-third of all major congenital anomalies, 1 and the prevalence may be up to 2~3%, if minor cardiac abnormalities such as bicuspid aortic valve are included. 2 The advances in surgical management and longevity of patients have been improved.
However, there is approximately 20% early mortality for the most complex cardiac defects and late mortality is a relatively common occurrence. 3 The genetic factors associated with CHD remain challenging. We recently identified that mutations of TBX5, 4 PLGL1 5 and HOMEZ 6 are associated with ventricular septal defect. Further, we also identified the prevalence of 22q11.2 variations in CHD. 7 CHD associated with malformation of dactyl has a complex genetic basis. Among various syndromes and complexes, Holt-Oram syndrome (HOS, OMIM 142900) is a recognized rare condition (1/100 000 live births) characterized by anterior pre-axial limb and cardiac malformations including atrial septal defects (ASD), ventricular septal defect (VSD), or multiple and complex malformations. 8 Mutations in TBX5 have been reported to cause autosomal dominant HOS. 9 Studies have shown that the genetic heterogeneity in HOS may be caused by the mutations of different genes. Several genes, such as homeobox genes, peptide growth factors and retinoic acid receptors, have been proposed to contribute to cardiac and limb development and regarded as important candidate genes for HOS. 10 To identify the genetic characteristics of complex CHD and polydactyly with emphasis on whether TBX5 was involved, we performed whole exome sequencing (WES) on blood sample of a patient with complex CHD and polydactyly who had successful surgical repair. On the basis of the findings of the mutations in a number of genes, we subsequently performed related cellular function studies.

| Whole Exome Sequencing (WES) study
Blood sample was obtained from the patient who was one-anda-half-year-old girl with complex CHD including complete atrioventricular septal defect (AVSD), patent ductus arteriosus (PDA), secondary atrial septal defect, pulmonary hypertension, and polydactyly ( Figure 1A). Genomic DNA was extracted. WES was performed F I G U R E 1 Typical phenotypes of the patient and potential pathogenic mutations. (A) Clinical features of the patient with both congenital heart defects and polydactyly. Atrioventricular septal defect (AVSD) was diagnosed by echocardiography; The X-ray shows the enlarged heart shadow and the polydactyly. (B) Chromatograms of the two heterozygous mutations in the study. Top panels show wild type, and bottom panels show heterozygous mutations. Mutations are marked with arrows by BGI (Beijing Genomic Institute, Shenzhen, china), and the SNV/ Indel was detected and analysed, using the UCSC hg19 human genome. The frequency of the variants in our study was filtered by population databases of dbSNP, 1000Genomes, ESP6500,gnomAD, and HGVD (BGI owned).
The study protocol was approved by the Institutional Review Board and Ethics Committee of TEDA international cardiovascular hospital.
Primers (Table S1) were designed to amplify, and the PCR products were sequenced by 3500 Genetic Analyzer (Applied Biosystems) to verify the results of WES.

| Establishment of human pluripotent stem cells (hPSCs) with mutations
To verify the correlation between TCTN3 (RefSeq NM_015631. 5

| Detection of pluripotency markers in hPSCs
The hPSCs in experimental and control groups were collected and resuspended in cell lysis buffer. Total protein was extracted and measured by BCA Protein Assay Kit (23227, Thermo Fisher Scientific). Protein samples were electrophoresed on SDS-PAGE gel, blotted onto PVDF membrane and probed with primary antibodies, including anti-OCT4 (sc5279, Santa Cruz), anti-Nanog (sc293121, Santa Cruz) and anti-β-actin (P30002, Abmart). Secondary antibodies conjugated with horseradish peroxidase (HRP) were used to detect the expression of the target protein.
Total RNA was extracted using TRIzol (15596018, Thermo Fisher Scientific) according to the manufacturer's protocol. The first-strand cDNA was synthesized using random primers and M-MLV Reverse Transcriptase (Invitrogen). Real-time PCR was set up in duplicate with the Faststart Universal SYBR Green Master Mix (Roche) and carried out on an iCycler MyiQ2 Detection System (BIO-RAD). Each sample was repeated 3 times and normalized using GAPDH as internal control. Relative quantitative evaluation of target gene was determined by comparing the threshold cycles. Primers (listed in Table S1) were confirmed for the specificity with dissociation curves.

| Differentiated hPSCs into cardiomyocytes (CMs)
The PSC Cardiomyocyte Differentiation Kit (A2921201, Thermo Fisher Scientific) was used for differentiation of hPSCs into CMs according to the manufacturer's protocol. Cell samples were collected on day 0, 9 and 13 during differentiation.

| RNA-sequencing and bioinformatics analysis
The CMs differentiated from hPSCs (hPSCs-CMs) were cultured and harvested. A total amount of 3 mg of RNA per sample was extracted. The mRNA was enriched and fragmented into short fragments and was reverse transcripted into cDNA with random primers.
Sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, United States) following manufacturer's recommendations and index codes were added to attribute sequences to each sample. The sequencing was performed by Gene Denovo Biotechnology Company (Guangzhou, China).
After filtering, clean reads were mapped to the human reference genome (GRCh38.p10). Gene abundances were quantified by RSEM.

| Statistical analysis
Data were analysed by ANOVA or t tests using StatView software from SAS Institute Inc (Cary, NC). The P value was calculated, and statistical significance was defined as P < 0.05(*), P < 0.01(**) or P < 0.001(***).
The more detailed methods are described in the Supplementary Method.

| TCTN3 and LTBP2 mutations identified by WES
WES was performed on the patient with complex CHD, and the mean read depth of the target regions of each sample ranged from 80×.  (Table S2). In the 137 variants, four genes were found to be involved in both heart and dactyly. Those were LTBP2, TCTN3, CHD7 and TWIST1 (Table S2). PolyPhen2 (score: 1) predicted TCTN3 c.1268G>A to be deleterious; MutationTaster predicted that the mutant was damaging; Human Splicing Finder showed that TCTN3 c.1268G>A broke an exon splicing enhancer and created an exon splicing silencer, thus potentially alter the splicing.
The allele frequency of LTBP2 c.2206G>A was 0.0009, 0.0035 and 0.0008 in population databases of dbSNP, 1000Genomes, and HGVD (BGI). The charge was changed by the mutation from negative-charged to uncharged. PhyloP (score: 3.278) predicted that the site was highly conserved in vertebrates; SIFT (score: 0.02) and PolyPhen2 (score: 0.966) predicted the mutation to be deleterious; MutationTaster predicted that the mutant was damaging; Human Splicing Finder showed that the variant broke an exon splicing enhancer thus potentially alter the splicing. Sanger sequencing was used to validate these two mutations. Figure 1B shows the chromatograms of the mutations. Interestingly, there were no mutations found in the TBX5 gene in this patient.

| The LTBP2/TCTN3 mutation was introduced separately into hPSCs by CRISPR/Cas9 system
To study the relationship between gene mutations and clinical phenotype of the patient, LTBP2/TCTN3 with point mutation (LTBP2 mu /TCTN3 mu ) was introduced separately into hPSCs by CRISPR/ Cas9 technology to observe the changes of cells.
Double-strand DNA with LTBP2 mu or TCTN3 mu was generated as template to repair DNA double-strand break. Synonymous mutations within protospacer-adjacent motif (PAM) were introduced to avoid breaking the templates again (Figure 2A and B). Templates together with sgRNAs were co-transfected into hPSCs WA26 (wild-type). Sanger sequencing was used to validate homozygous mutations LTBP2 (c.2206G>A) and TCTN3 (c.1268G>A) in clones ( Figure 2C and D). We successfully established two cell lines, hP-SC-LTBP2 mu and hPSC-TCTN3 mu . The cell morphologic changes were observed using inverted microscope and cells grew well in three groups ( Figure 2E).

| Detection of the developmental pluripotency of hPSCs
The expression levels of important pluripotent markers Oct4, Nanog and SSEA4 were detected by immunofluorescence (IF) microscopy, Western bolt and Q-PCR ( Figure 3). The expression level of Nanog was WT<LTBP2 mu < TCTN3 mu with P value > 0.05 ( Figure 3B and C); the other markers had no differences between the study groups ( Figure 3A-C). These results indicated that there were no significant differences between hPSCs-LTBP2 mu /TCTN3 mu and the wild type; these two mutations may not affect the developmental pluripotency of hPSCs.

| Gene mutations changed the contraction frequency and cell differentiation of cardiomyocytes
To simulate the development of CMs, hPSCs were directly differentiated into CMs (hPSC-CMs) in vitro. As CMs marker, TNNT2 was detected by immunofluorescence, indicating that cell differentiation was successful ( Figure 3D).
During the process of differentiation, the contraction rate of cells was calculated ( Figure 3E). The rate of group hPSC-CMs-LTB-P2 mu was significantly slower than hPSC-CMs-WT on day 9 (Video S1, S2) and the rate increased on days 11-13 (Video S3), suggesting that LTBP2 mu resulted in delayed development of hPSC-CMs.
Interestingly, the contraction of hPSC-CMs-TCTN3 mu was affected with significantly lower beating rate and weaker force (Video S4).
These observations suggest that LTBP2 mu and TCTN3 mu affect cardiac rhythm and contraction of the heart.

| Transcriptomics analysis of hPSC-CMs
Transcriptomics analysis of group hPSC-CMs-LTBP2 mu , hPSC-CMs-TCTN3 mu , and hPSC-CMs-WT were performed using RNAsequencing (RNA-Seq) on day 0, day 9 and day 13. The gene transcripts with the absolute value of log2FC>1 (minimal 2-fold difference in expression) and FDR<0.05 (false discovery rate, adjusted P-value threshold) were considered as significant DEGs.

| Bioinformatics analysis of DEGs: expression pattern, function, and pathway
The DEGs of each group on different days were clustered based on gene expression pattern using STEM software. Figure

| D ISCUSS I ON
The present study demonstrates that in complex CHD associated with polydactyly the mutations of LTBP2 and TCTN3 may be potential pathological cause since these mutations are associated with changes of cellular functions that may affect the development of the heart. Further, this study again demonstrates that TBX5 mutations may not present in complex CHD associated with polydactyly.
As an important transcription factor, TBX5 regulates a wide variety of developmental processes. Particularly, TBX5 interacts with NKX2-5 and GATA4 and co-regulates cardiac gene expressions in heart development. 4 In human, more than one hundred of TBX5 mutations identified are associated with HOS. 11 Although it is well known that causal genes in HOS involve mutations in regulatory parts of TBX5, this syndrome may also involve other genes. 12 Darwich and associate reported that KLF13 interacts physically and functionally with TBX5 to synergistically activate transcription of cardiac genes and therefore KLF13 may be a genetic modifier of the HOS gene TBX5. 13 Another study identified  Thus, these two genes are related to both congenital heart defects and digital abnormalities and therefore are likely to be associated with phenotypes of the patient in this study.

| Functional changes of cardiomyocytes associated with mutations of LTBP2 and TCTN3
During the process of differentiation, we calculated the contraction rate of hPSCs-CMs with mutation and find that the group hPSCs-

| Translational significance in cardiovascular medicine
The present study demonstrated that in clinically seen CHD associated with polydactyly, mutations of genes other than TBX5 may be involved. Owing to the complexity of the CHD in this study, for personalized diagnosis and treatment in CHD patients associated with polydactyly, WES may be necessary to identify the pathological mutations in the future.

| Study limitation
This study has some limitations. Owing to the fact that the patient is an orphan who underwent repair surgery through our charity programme, no family history and data were available.
The present study demonstrates that in complex CHD associated with polydactyly, the mutations of LTBP2 and TCTN3 may be potential pathological cause since these mutations are associated with changes of cellular functions that may affect the development of the heart.
Further, this study again demonstrates that TBX5 mutations may not present in complex CHD associated with polydactyly.
In summary, two heterozygous mutations in LTBP2 and TCTN3 without TBX5 mutations were identified in this study. Cellular functional analysis suggested that these two mutations affect contractility (rate and force) of cardiac myocytes and may affect the development of the heart. These findings provide new insights into the pathogenesis of complex CHD associated with polydactyly.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available upon request from the corresponding author.