A loss-of-function mutation p.T52S in RIPPLY3 is a potential predisposing genetic risk factor for Chinese Han conotruncal heart defect patients without the 22q11.2 deletion/duplication

All assessments were done with the approval of the Medical Ethics Committee of Xinhua hospital and Shanghai Children’s Medical Center. And all experiments were carried out in accordance with the approved guidelines. Fully written informed consent was obtained from all participants (or their parents if children were too young to consent by themselves).

From November 2011 to January 2014, a cohort of 615 unrelated patients of Han ethnicity diagnosed with CTD were recruited from Shanghai Children’s Medical Center affiliated to Shanghai Jiao Tong University School of Medicine, a cardiologist confirmed the CHD diagnosis for all patients by reviewing and evaluating patient history, physical examinations, and medical records (Table 1) [12]. All subjects were unrelated, and the median age at the time of diagnosis was 10 months with a range of 3 days to 18 years. A total of 391 unrelated healthy individuals (median age was 3.5 years), free of CHD, of Han ethnicity were also enrolled as controls. Peripheral blood samples of all cases were obtained, and genomic DNA was extracted using the QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany) with standard protocols.

22q11.2 loci in all CTD patients were explored by CNVplex^® (a technique for high-throughput detection of sub-chromosomal copynumber aberrations) as described before [12]. Finally, 577 patients without 22q11.2 deletion/duplication were recruited.

For the patients without 22q11.2 deletion/duplication, fragments covering the promoter region, 5′UTR, coding sequence and splicing site of RIPPLY3 were amplified using the EasyTarget^® amplification kit (Genesky Biotechnologies Inc, Shanghai, China) which was developed according to the cycled primer extension and ligation-dependent amplification (CPELA) method. The CPELA was a new method developed by Genesky Biotechnologies for fast and simple enrichment of multiple gene regions for massively parallel sequencing [24].

All candidate variants of RIPPLY3 were confirmed by Sanger sequencing. The primer pairs used to amplify the coding regions contain candidate variants were designed using Primer Premier 5 in Additional file 1: Table S1. Samples were amplified in a volume of 10 μl containing 1× MyTaq™Mix (Bioline USA Inc.), 200 ng of DNA and 0.4 μM of each primer following the suggested PCR protocol. PCR products were sequencing on an ABI 3730 sequencer (Applied Biosystems, USA). The PCR program was as follows: 98 °C for 1 min; 30 cycles of 98 °C for 20 s, 62 °C for 30 s and 72 °C for 30 s; then 72 °C for 10 min. Amplified products were analyzed on 1% agarose gels stained with GelRed (Biotium, USA). The sequence traces were aligned with the reference sequence for RIPPLY3 (RefSeq NM_018962.2) using the GenBank BLAST program (http://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi).

Furthermore, because mutations in cardiac kernel members, such as GATA4, NKX2-5, and TBX5, underlie a range of CHDs [25, 26], and TBX1 plays a vital role in the development of the cardiac outflow tract [7‐11]. To validate the potential function of identified RIPPLY3 variant, we also screened coding sequences and splicing sites of these known CHD pathogenic genes (TBX1, GATA4, NKX2.5 and TBX5) in CTD patients and controls using the EasyTarget^® amplification kit (Genesky Biotechnologies Inc, Shanghai, China). All candidate variants were also validated by Sanger sequencing. The primer pairs used to amplify the coding regions contain candidate variants were designed using Primer Premier 5 in Additional file 1: Table S2. The verified sequence variants were queried in the SNP database at National Center for Biotechnology Information (NCBI; http://​www.​ncbi.​nlm.​nih.​gov), 1000 Genomes database (http://​www.​1000genomes.​org), Exome Aggregation Consortium database (ExAC, http://​exac.​broadinstitute.​org) and Ensembl database (http://​asia.​ensembl.​org).

To predict the potential pathogenic impact of missense variants, three different pathogenicity prediction tools (PPTs), namely PolyPhen2, SIFT and Mutation Taster were used.

Multiple RIPPLY3 protein sequences across various species were aligned using the clustalx 2.1 software.

RIPPLY3 expression vector pCMV6-Entry-RIPPLY3 (Myc-DDK-tagged) containing the cDNA of human RIPPLY3 (RefSeq NM_018962.2) was purchased from Origene (Rockville, MD, USA). The identified variants were introduced into wild type pCMV6-Entry-RIPPLY3 using a QuickChange II Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA, USA). Both wild type and variant inserts were sequenced by Sanger sequencing to confirm the variant sequence and exclude any other sequence variations. TBX1C expression vector pCMV6-XL6-TBX1 containing the cDNA of human TBX1C (RefSeq NM_080647.1) was purchased from Origene (Rockville, MD, USA). TBX1C cDNA was digested and inserting into the plasmid pcDNA3.1(+) at the KpnI and XhoI sites. The luciferase reporter Wnt5a-luc, was constructed by inserting the conserved T-box binding sites (TBEs) in the 3′-UTR of human WNT5A, a target of TBX1 [27], into the pGL3-promoter plasmid (Promega, USA). An FGF10 luc reporter was also constructed according to the described by Agarwal et al. [28]. FGF10 genomic sequence was obtained from the GenBank (http://​www.​ncbi.​nlm.​nih.​gov). A fragment comprising 4.5 kb upstream of the coding region (− 4.3 kb/+ 176 bp) of FGF10 was amplified from DNA of a healthy individual and inserted into the MluI/SmaI sites of the pGL3-Basic vector (Promega, Madison, Wisconsin, USA) to generate the FGF10-luc reporter construct. The schemes of these two luciferase reporter plasmids were showed in Fig. 3b, c.

HEK293T and C2C12 cell lines were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) and cultured in the DMEM medium (Invitrogen, California, USA) supplemented with 10% foetal calf serum (Invitrogen, California, USA), penicillin (100 units/ml) and streptomycin (100 μg/ml). Cells were incubated at 37 °C with 5% CO₂. All the transient transfections were performed with Fugene HD transfection reagent (Promega, Madison, Wisconsin, USA) according to the manufacturer’s protocol for adherent cells.

HEK293T or C2C12 cells were seeded into 48-well plates 24 h before transfection. Plasmids including the wild-type or variant RIPPLY3 vector, the TBX1 vector, the luciferase reporters and the corresponding renilla luciferase reporter were co-transfected into HEK293T or C2C12 cellsusing Fugene HD following the manufacturer’s protocol. Transfected cells were incubated for 36 h, then washed using DPBS and lysed using 1× passive lysis buffer for 15 min at room temperature (RT) provided by the Dual Luciferase Reporter Assay Kit (Promega). Firefly luciferase and renilla luciferase activities were measured with the Dual-Glo luciferase assay system (Promega) and the Centro XS³ LB 960 Microplate Luminometer (Berthold, Bad Wildbad, Germany) according to the manufacturer’s recommended protocol. The transfection efficiency was normalized to paired renilla luciferase activity. The results are the means of three independent experiments, each done in duplicate.

HEK293T cells were seeded into 12-well plates 24 h before transfection, and cells were transiently transfected with a total 800 ng pCMV6-Entry-RIPPLY3 or variant RIPPLY3 constructs, and pcDNA3.1-TBX1 was also transfected as positive control. Cells were harvested 24 h after transfection, washed three times with PBS, fixed using 4% paraformaldehyde/PBS for 10 min and washed again in PBS. Cells were permeabilized using 0.3% Triton X-100/PBS for 10 min. After washing with PBS, the cells were blocked with 5% BSA/PBS solution for 1 h at RT. Then, the cells were incubated with a 1:100 solution of the rabbit primary anti-TBX1 antibody (ab18530, abcam) or 1:200 solution of the goat primary anti-Myc tag antibody (ab9132, abcam) in 1% BSA/PBS overnight at 4 °C. Next,the cells were washed with PBS and incubated with a 1:500 solution of the secondary anti-Rabbit IgG H&L (Cy3^®) (ab6939, abcam) or 1:200 solution of the secondary Anti-Goat IgG H&L (Alexa Fluor^® 488) preadsorbed (ab150129, abcam) in 1% BSA/PBS for 2 h at RT. After washing with PBS, the cells were incubated with DAPI for 6–8 min at RT, then washed with PBS and mounted on a rectangular slide containing an anti-fading agent. The slides were examined using the Olympus BX43 microscope (Olympus, Shinjuku-ku, Tokyo, Japan).

To confirm whether these variants affect the expression of RIPPLY3 protein, western blot analysis was carried out. HEK293T cells were transfected with a total of 500 ng of pCMV6-Entry-RIPPLY3 or variant RIPPLY3 constructs in 24-well plates. Cells were harvested 40–48 h after transfection and lysed in 90 μl cell lysis buffer. After boiling for 10 min, cell lysates were analyzed by sodium dodecyl sulfatepolyacrylamide gel electrophoresis and electrotransferred to Nitrocellulose Blotting Membrane (GE Healthcare). Then, we block the membranes with 5% skim milk for 2 h at RT, and incubated the membranes with anti-Myc tag antibody (ab9132, abcam) and anti-actin antibody (ab3280, abcam) overnight at 4 °C. Next, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (Proteintech, Chicago, IL, USA) for 2 h at RT. The membranes were washed, and detected using enhanced chemiluminescence (ECL) reagents (Millipore, Billerica, MA, USA). Bands were visualized with a chemiluminescence detection system (BioRad, Philadelphia, PA, USA) and analyzed using the Imagelab program (BioRad, Philadelphia, PA, USA).

For co-immunoprecipitation experiment, HEK293T cells were cotransfected with a total 7.2 μg pcDNA3.1-TBX1 and pCMV6-Entry-RIPPLY3 or variant RIPPLY3 constructs. The washed Protein A/G Sepharose (GE Healthcare) was mixed with 6 μg of rabbit anti-TBX1 antibody (ab18530, abcam) and placed on a shaker at 4 °C for 3 h. Then, 1 mg protein from HEK 293T cell lysates were added and left at 4 °C for 8 h or overnight. After washing 4 times with PBS, the pellets were resuspended in 30 μl 1× Protein Loading buffer (Sangon Biotech), boiled for 10 min at 100 °C, and the products were analyzed by sodium dodecyl sulfatepolyacrylamide gel electrophoresis, next, a regular Western blot was run as previously described. The grayscale value of protein bands was calculated using Image J software.

Four heterozygous missense variants in RIPPLY3 were identified in 4 unrelated CTD patients out of the 577 CTD patients without 22q11.2 deletion/duplication, all the variants were absent from the 391 controls in our cohort: p.P30L in TOF, p.T52S in TOF, p.D113N in TGA/PA/VSD, and p.V179D in PA/VSD and patent ductus arteriosus (PDA) (Table 2).

Within the group of 4 missense RIPPLY3 variants, p.T52S (NM_018962.2:c.155C>G, rs745539198) and p.D113N (NM_018962.2:c.337G>A, rs747419773) were found in ExAC database, the allelic frequencies of these two variants were 3.32e−05 and 1.654e−05 respectively, and the p.T52S variant was found only in Asian population. Notably, variants p.P30L and p.V179D were not identified in the NCBI’s SNP database, 1000 Genomes database, ExAC database or Ensembl database. The chromatograms showing the identified heterozygous RIPPLY3 variants in contrast to corresponding control sequences are shown in Fig. 1a–d, and the positions of these variants in the RIPPLY3 protein are illustrated in Fig. 1e.

As shown in Fig. 1f, a cross-species alignment of multiple RIPPLY3 protein sequences showed that both the p.P30L and p.T52S occur in the highly evolutionarily conserved residues, indicating that the amino acids are functionally important. However, the valine 179 of the protein was just conserved in humans and gorillas, and the aspartate 113 of the protein was not conserved.

Several known CHD pathogenic genes (TBX1, GATA4, NKX2.5 and TBX5) were also screened in these four patients. There were no nonsynonymous variants identified in GATA4, NKX2.5 and TBX5, but two known nonsynonymous single nucleotide polymorphisms (SNPs) were identified in TBX1C (Table 3). The nonsynonymous SNP rs72646967 (TBX1C NM_080647.1:c.1189A>C) appeared commonly in population (MAF > 0.05), and the allele frequency was similar between CHD patients and controls [29]. The another nonsynonymous SNP rs41298838 (TBX1C NM_080647.1:c.928G>A) was previously predicted to be a tolerant polymorphism [30], and was found in both the non-del22q11 CTD patients and healthy controls at similar frequencies [31]. The allele frequency of rs72646967 and rs41298838 was similar between CTD patients and controls in this study (p > 0.05) (Table 3). Therefore, there were no deleterious variants identified in these four CHD pathogenic genes (TBX1, GATA4, NKX2.5 and TBX5) in patients harboring RIPPLY3 variants.

The generated plasmids were transfected into HEK293T cells to investigate the effect of the variants on the trafficking of the RIPPLY3 protein. Immunofluorescence staining showed that both the wild-type RIPPLY3 and variant RIPPLY3 proteins (p.P30L, p.T52S, p.D113N and p.V179D) were located in both the cytoplasm and nucleus (Fig. 2). The TBX1 protein was located only in the nucleus (Additional file 1: Figure S1).

Wild-type and variant RIPPLY3 expression plasmids were separately transfected into HEK293T cells. Western blot results revealed that there was no significant change in dosage between wild-type and variant RIPPLY3 proteins (Fig. 3a).

RIPPLY3 is able to repress the transcriptional activity of TBX1 in vitro. To further validate the effects of these variants, we performed dual luciferase reporter assays using the Wnt5a-luc and FGF10-luc in HEK293T cells and C2C12 cells. All of the RIPPLY3 variants are heterozygous in patients, so we also detected whether the variants would affect function of the wild-type RIPPLY3 protein by co-transfected the variant and wild-type RIPPLY3 protein together. Compared to the wild-type RIPPLY3 protein, the p.D113N variant showed no significant change in inhibition of TBX1 transcriptional activity in our assay. The others, namely, p.P30L, p.T52S and p.V179D, showed impaired inhibition of TBX1 transcriptional activity in both 293T cells for Wnt5a-luc (p.P30L: p = 0.0144 for homozygote and p = 0.0258 for heterozygote; p.T52S: p = 0.0031 for homozygote and p = 0.0167 for heterozygote; p.V179D: p = 0.0109 for homozygote and p = 0.0105 for heterozygote, respectively) and C2C12 cells for FGF10-luc. (p.P30L: p = 0.0032 for homozygote and p = 0.0102 for heterozygote; p.T52S: p = 0.0036 for homozygote and p = 0.0098 for heterozygote; p.V179D: p = 0.016 for homozygote and p = 0.0102 for heterozygote, respectively) (Fig. 3b, c). Therefore, the three variants (p.P30L, p.T52S and p.V179D) showed impaired inhibition of TBX1 transcriptional activity in both heterozygote and homozygote, although it seems that RIPPLY3 variants do not have significant dominant negative effect on the wild-type protein.

Evidence suggested that RIPPLY3 inhibited TBX1 transcriptional activity by physically interacting with TBX1 in a T-domain-dependent manner [15]. To characterize the effect of the variants on the protein–protein interaction, co-immunoprecipitation assays were performed using equal amounts of protein extracts from HEK293T cells overexpressing the wild-type and variant RIPPLY3 proteins. The results showed that the p.T52S variant significantly disrupted the physical interaction between RIPPLY3 and TBX1 (p = 0.0059). However, the p.P30L, p.D113N and p.V179D variants had no effect on the interaction between RIPPLY3 and TBX1 (Fig. 4).