Ethical statement

These studies involving human participants were approved by the Medical Ethics Committee of Xinhua Hospital (NO. XHEC-C-2012–018) and Shanghai Children’s Medical Center (NO. SCMC-201015). Human peripheral blood samples were extracted voluntarily when donors signed the informed consent (or their parents/guardian if the donors were too young to consent). Zebrafish experiments were conducted under the approval of the Animal Ethics Committee of Xinhua Hospital, affiliated to Shanghai Jiao Tong University School of Medicine.

Study population

From January 2012 to December 2016, we recruited 81 HTX patients with CHDs and 89 healthy children from Shanghai Children’s Medical Center and Xinhua Hospital. All patients were diagnosed using an echocardiogram, computed tomography, or magnetic resonance imaging. The patient cohort included 50 males and 31 females, with a mean age of 3.01 ± 2.60 years (range 12 days–15 years). All patients had cardiac defects including dextrocardia, mesocardia, left atrium isomerism, right atrium isomerism (RAI), conotruncal defects, and abnormal superior/inferior vena cava. Extracardiac malformations included bronchial inversus, polysplenia/asplenia, and dextrogastria (Table 1). The control cohort included 58 males and 31 females, with an average age of 4.75 ± 3.75 years (range 3 months–13 years). The controls were children presenting for routine health checkups. We informed the children and their parents about the background, purpose, and significance of our study. After obtaining the signed written informed consent of the children and their legal guardians, we included the children in the control group and extracted peripheral blood for exome sequencing. All study participants were Han Chinese and not related to one another. The study design conformed to the guidelines of the Declaration of Helsinki.

Table 1 Cardiac and extracardiac abnormalities of 81 HTX patients with CHDs in this studyExome sequencing identification and Sanger sequencing validation

After obtaining informed consent, we strictly collected whole peripheral blood samples from all participants and stored them in EDTA tubes individually without sample mixup. The QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) was used to isolate genomic DNA. Exome sequencing was performed in 81 HTX patients and 89 control children. Specifically, exome sequencing was performed using a commercial provider (Shanghai Biotechnology Co., Ltd.). Briefly, the SureSelect Human All Exon V6 kit (Agilent Technologies, Santa Clara, CA, USA) and Illumina HiSeq 2500 platform (Shanghai Biotechnology Co., Ltd., Shanghai, China) were used, and raw sequencing reads were compared to the reference human genome (hg19). The mapping ratio of the 170 samples included in this study (case and control) was above 99%, and the average sequencing depth was > 25X. Among all bases in the captured target area, the percentage of coverage ≥ 10 was 98%–99.99%. Exome sequencing data were filtered based on the following criteria: (1) the mutation genotype included exonic non-synonymous, coding indels, splice-site variants, or frameshift; (2) the average depth of sequencing was > 10X; (3) minor allele frequency was < 1% or not found in population-based databases including SNP database at NCBI (http://www.ncbi.nlm.nih.gov/), 1000 Genomes (http://www.1000genomes.org/), ESP6500 (http://evs.gs.washington.edu/EVS/), ExAC (http://exac.broadinstitute.org/), and gnomAD (http://gnomad.broadinstitute.org/).

Sanger sequencing was performed for validating the candidate variants of MMP21 and confirming that all variants were de novo. The PCR primers were designed to amplify the coding regions, including the candidate variants, using Primer premier5 software (Additional file 1: Table S1). We added 250 ng of each genomic DNA sample and 1 μM of each primer to 20 μl of 1 × MyTaq™Mix (Bioline USA Inc.). The PCR conditions were as follows: 98 °C for 2 min; 30 cycles at 98 °C for 20 s, Tm for 30 s, 72 °C for 45 s, and 72 °C for 10 min. We used 1% agarose gel staining with GelRed (Biotium, USA) to analyze 2 μl of PCR products. The residual PCR products were sequenced using an ABI 3730XL sequencer (Applied Biosystems, USA), and the results were compared to reference MMP21 cDNA sequences from NCBI (#NM_147191.1) using the GenBank BLAST program (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

In silico analysis

Different in silico pathogenicity prediction tools include MutationTaster (http://www.mutationtaster.org/), PROVEAN (http://provean.jcvi.org/), and Polyphen2 (http://genetics.bwh.harvard.edu/pph2/) were used to evaluate possible pathogenic effects of the identified variants (Table 2). Three-dimensional structural models of wild-type (WT) and mutant MMP21 proteins were constructed using SWISS-MODEL (https://www.swissmodel.expasy.org), and the effects of mutant protein configuration changes were analyzed using HOPE (http://www.cmbi.ru.nl/hope/).

Table 2 Clinical information and MMP21 variant characteristics in HTX patients with complex cardiac malformationsAlignment of multiple MMP21 protein sequences

The MMP21 amino acid sequences of various species, including Homo sapiens (human), Xenopus tropicalis (frog), Canis lupus familiaris (dog), Danio rerio (zebrafish), Mus musculus (house mouse), Macaca mulatta (rhesus monkey), Rattus norvegicus (rat), and Pan troglodytes (chimpanzee) were downloaded from the UniProt database (http://www.uniprot.org). The Clustal X software (http://www.clustal.org) was used for sequence alignment.

Plasmids and mutagenesis

The pCMV3-Human MMP21 expression vector and pGEM-T-Human MMP21 clone vector containing human MMP21 cDNA (NCBI RefSeq NM_147191.1) were purchased from Sino Biological (Sino Biological, China). The variants p.G244E (NM_147191.1: c.G731A), p.L277F (NM_147191.1: c.C829T), and p.K487E (NM_147191.1: c.A1459G) were introduced into the WT pCMV3-Human MMP21 vector and pGEM-T-Human MMP21 vector using site-mutagenesis primers (Additional file 1: Table S1) via PCR. The accuracy and integrity of all plasmids were verified by Sanger sequencing.

Cell culture and transfection

Human embryonic kidney 293 T (HEK-293 T) cells obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) were cultured in Dulbecco’s modified Eagle’s medium (HyClone, USA) supplemented with 1% penicillin–streptomycin (Gibco, USA) and 10% fetal bovine serum (MP Biomedicals, USA) under a humidified atmosphere (95% air and 5% carbon dioxide) at 37 °C. When cell density in the 6/12-well plate reached 80%, plasmids were transfected into HEK-293 T cells using FuGene HD (Promega, USA) according to the manufacturer’s protocol. Homozygous and heterozygous variants of MMP21 were simulated using the following two transfection schemes: (1) 1 µg of blank, WT, or mutant MMP21 vectors were separately transfected into HEK-293 T cells; (2) 1 µg of blank, WT, or WT and mutant MMP21 vectors (1:1) were transfected into HEK-293 T cells.

Quantitative real-time polymerase chain reaction

Total RNA was extracted from HEK-293 T cells using TRIzol reagent (Invitrogen, USA) after transfection for 36 h. Zebrafish embryos were crushed using a magnetic bead homogenate pulverizer and total RNA was extracted using TRIzol reagent after microinjection of control MO and splice-blocking Morpholino oligo (SB-MO) for 30 h. Subsequently, PrimeScript RT Master Mix (TaKaRa, Japan) was used for reverse transcription. Quantitative real-time amplification was performed using SYBR Premix Ex Taq (TaKaRa, Japan) on an Applied Biosystems 7500 system. Primers were synthesized by Shanghai Sangon Biotechnology Co., Ltd. (Additional file 1: Table S1). The 2−ΔΔCt method was used to calculate the relative expression of MMP21, and human 18sRNA and zebrafish actin were used as internal references [33].

Western blot

After transfection for 48 h, proteins from HEK-293 T cells were extracted using RIPA lysis buffer (Beyotime, China) supplemented with PMSF (1:100). Proteins were then separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred electrophoretically onto polyvinylidene fluoride membranes (Millipore, USA). The membranes were subsequently incubated with 5% skim milk for 1.5 h at room temperature and then with anti-MMP21 (1:650, 55,289-1-AP, Proteintech) and anti–actin antibodies (1:2000, AF5001, Beyotime) overnight at 4 °C. Next, the membranes were incubated with horseradish peroxidase-labeled goat anti-rabbit secondary antibody (1:5000) and goat anti-mouse secondary antibody (1:5000) for 1.5 h at room temperature and detected with Immobilon Western Chemiluminescent HRP Substrate (Millipore, USA) using a chemiluminescence system (BioRad).

Transcriptional synthesis of mRNA in vitro

The WT and mutant pGEM-T-human MMP21 vectors were used as templates. NotI restriction enzymes were selected based on the location of the T7 promoter in the plasmid map to perform single enzyme digestion of the vector and linearize the plasmid. We generated WT and mutant human MMP21 mRNAs by in vitro transcription, mRNA capping, poly-A tail addition, and mRNA purification, using the mMESSAGE mMACHINE™ T7 Transcription Kit (Thermo Fisher Scientific, USA) according to the manufacturer’s protocol. Human MMP21 mRNA was verified using 1% agarose gel electrophoresis.

Zebrafish maintenance

WT AB strains of zebrafish were obtained from China Zebrafish Resource Center and bred under standard laboratory conditions as previously reported [34]. The water in the feeding system was purified circularly using water filtration equipment and maintained at 28 ± 1 °C with a 14 h/10 h light/dark cycle.

Morpholinos and microinjection in zebrafish

SB-MO was designed and synthesized by Gene Tools (OR, USA) to suppress mmp21 (NM_001317753.1) splicing and translation by targeting the exon 3–intron 3 junction (sequences: 5ʹ-AAATGTGCGATTTAAAACCTGTGCA-3ʹ). MO (5ʹ-CCTCTTACCTCAGTTACAATTTATA-3ʹ) was used as a negative control. Sexually mature male and female zebrafish were separately placed in the same mating box at a 1:1 ratio in the evening and mated the following morning. Zebrafish eggs were collected and arranged in the microinjection mold groove and a working solution was injected into the yolk sac of embryos using a pressure microinjection apparatus (Warner PLI-100A, USA) when the embryos were at the 1–2 cell stage. SB-MO efficiency was verified by RT-qPCR of mmp21 cDNA (NM_001317753.1) obtained from whole embryos 30 h after microinjection. A preliminary experiment was then conducted to determine the optimal interventional concentration of SB-MO. Human MMP21 mRNA (300 ng/μL) was used to rescue the phenotype of SB-MO. Since variants found in the patients were heterozygous, we mixed WT and mutant MMP21 mRNAs (1:1) for microinjection to achieve a heterozygous state. Each microinjection had a volume of approximately 2–5 nL, and 470–570 embryos were injected in each group. Embryos were then regularly observed, and their phenotypes within each group were recorded at 48 h post-fertilization (hpf). The phenotypes recorded include death, cardiac location, looping abnormality, pericardial edema, tail deformity, and spinal curvature. Zebrafish embryos were imaged using an SMZ25 microscope (Nikon, Tokyo, Japan) equipped with a digital camera.

Whole-mount in situ hybridization (WISH)

Primers (Additional file 1: Table S1) were designed to amplify zebrafish cmlc2 (NM_131329.3) by PCR via 2 × TransTaq® High Fidelity (HiFi) PCR SuperMix II (TransGen Inc., China). The pGEM®-T Easy Vector System (Promega, USA) was used to construct the cmlc2 antisense probe plasmid. NcoI restriction enzyme was selected to perform single-enzyme digestion of the vector and linearize the plasmid. Digoxigenin (DIG)-labeled cmlc2 antisense RNA probes were transcriptionally synthesized in vitro using an SP6 RNA Polymerase (Promega, USA). The 48 hpf embryos preserved by methanol dehydration were gradually rehydrated, fixed with 4% PFA, and digested with proteinase K (10 μg/mL). After pre-incubation with hybridization buffer at 65 °C for 4 h, the embryos were incubated with Dig-labeled cmlc2 RNA antisense probes at 65 °C overnight. Next, after gradient washing, the 48 hpf embryos were immersed in anti-DIG antibody (Roche, Germany) and shaken slowly at room temperature for 1 h and then at 4 °C overnight. Finally, immunopositive alkaline phosphatase signals of the 48 hpf embryos were visualized using BM Purple (Sigma, Japan) after washing. An SMZ25 microscope (Nikon, Japan) equipped with a digital camera was used to capture images.

Statistical analysis

All measurement data are presented as the mean ± standard error (SEM). The experiments were independently repeated at least three times. Statistical analyses were performed using GraphPad Prism 8. The results of RT-qPCR and western blotting were analyzed using one-way ANOVA with Tukey’s multiple comparison test. The percentage of zebrafish embryonic phenotypes was compared between groups using the chi-square test in zebrafish MO rescue experiments (and Fisher’s exact test). Statistical significance was set at p < 0.05.