A naturally conceiving couple comprising a 23-year-old male and a 23-year-old female was recruited for the present study in September 2023 at the Department of Maternity of The First Hospital of Changsha (Changsha, China). Early pregnancy tests, including Down syndrome screening, nuchal translucency screening and noninvasive prenatal testing returned normal results, as did tests for the infectious diseases (HIV, hepatitis B and Treponema pallidum). At 24 weeks and 1 day of pregnancy, an ultrasound examination revealed abnormal sonographic images of the fetal heart, including dextroversion of the heart, complete atrioventricular septal defect, subcardiac total pulmonary venous drainage, pulmonary atresia, arterial catheter blood supply and C-type collateral circulation (Fig. 1), raising the suspicion of complex CHD. The positions of the stomach, liver and spleen of the fetus, a male, were all normal. The pregnant woman later underwent an ultrasound examination at the Obstetrics Department of The Second Xiangya Hospital of Central South University (Changsha, China), and the results were consistent with those obtained at The First Hospital of Changsha. Medical professionals in the Department of Cardiac Surgery of The Second Xiangya Hospital of Central South University (Changsha, China) assessed the situation and recommended early surgical intervention after birth, despite the high surgical risks and uncertain outcomes associated with the procedure. After careful consideration, the pregnant woman and her family chose to terminate the pregnancy and sought to identify a genetic cause. Following an induced abortion, tissue samples from the fetus were collected at The First Hospital of Changsha for copy number variation sequencing and whole exome sequencing (WES) at BGI Genomics Co., Ltd. No chromosomal aneuploidy variants or microdeletions/microduplications known to be clearly pathogenic or suspected to be pathogenic were detected. The present study was approved by the Ethics Committee of The First Hospital of Changsha and the family members signed an informed consent form.

The genomic DNA of the male proband was extracted from fetal tissue obtained following pregnancy termination using a MagPure Buffy Coat DNA Midi KF Kit according to the instructions provided by the manufacturer (Magen Biotechnology Co., Ltd.). Shearing enzyme (Enzymatics) and magnetic beads (VAHTS DNA Clean Beads; cat. no. N411; Vazyme) were used to fragment and purify genomic DNA to obtain fragments of 200–300 bp. Pre-PCR amplification was then performed to complete library construction. The loading concentration of the final library was 29.2 ng/ml. The DNA of target gene exons and adjacent splicing regions was captured and enriched using the Roche KAPA HyperExome probe set (cat. no. 9718630001; Roche Diagnostics, Ltd.). Gel electrophoresis was used to assess the integrity of DNA, with high-quality DNA displaying a single, clear band. Sequencing Reaction General Kit (T7 SM FCL PE100) v2.0 (cat. no. 1000028455; MGI Tech Co., Ltd.) was used on the MGISEQ-2000 platform (MGI Tech Co., Ltd.) for sequencing. The quality of the raw sequencing data was assessed using SOAPnuke software v2.1.2 (18), and the clean reads were aligned to the hg19 reference genome using Burrows-Wheeler Aligner software BWA-0.7.17 (r1188) (19). Single nucleotide variants and insertions/deletions were identified using the Genome Analysis Toolkit (20), generating results for base polymorphisms in the target region. Subsequently, 1,000 GEA, 1,000 Genomes (East Asian; www.internationalgenome.org/home), gnomAD_exome (East Asian; http://gnomad.broadinstitute.org/) and the Exome Aggregation Consortium (ExAC; East Asian; http://exac.broadinstitute.org), were used to identify the frequency and context of the identified mutations. The BGI-varanno algorithm (21,22) (BGICG_ANNO 0.39; BGI Genomics Co., Ltd) was employed for variant screening and annotation, with the integration of disease data from Clin-Var (https://www.ncbi.nlm.nih.gov/clinvar/), Online Mendelian Inheritance in Man (https://www.omim.org/) and the Human Gene Mutation Database (http://www.hgmd.cf.ac.uk/). Variant filtering was performed as described in our previous study (20), and the American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology guidelines for mutation pathogenicity were referred to for interpretation (23,24). PCR was performed to amplify the target regions containing the putative variants in the fetus and the parents: DNA was extracted using the QIAamp DNA Blood Mini Kit (Qiagen GmbH), PCR amplification using GoTaq Green Master Mix (Promega Corporation) under the following conditions: 95°C for 30 sec (denaturation), 59°C for 30 sec (annealing) and 72°C for 30 sec (extension), for a total of 35 cycles, the PCR products were analyzed via gel (2% agarose) electrophoresis to confirm the successful amplification of the target regions. Finally, the PCR products were directly sequenced using Sanger sequencing. The primers used are listed in Table SI.

Evolutionary conservation analysis was conducted by aligning the amino acid sequences of DNAH9 proteins across various species using the BLAST tool available on the NCBI website (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The potential pathogenicity of the DNAH9 mutations was assessed through in silico analysis utilizing four online tools: Polyphen-2 (http://genetics.bwh.harvard.edu/pph2), MutationTaster (v2021 for GRCh37; http://www.mutationtaster.org/), Sorting Intolerant from Tolerant (SIFT; http://sift.bii.a-star.edu.sg/) and Combined Annotation Dependent Depletion (CADD; v1.7; http://cadd.gs.washington.edu/). The RNA splicing prediction model from the Rare Disease Data Center (RDDC) tool of the Guangzhou Rare Disease Gene Therapy Alliance (https://rddc.tsinghua-gd.org/tool) was used to forecast splicing variations. The mutated sequence was derived from the prediction results. Protein structural analysis of the DNAH9 (NM_001372.3) mutant was performed using the online SWISS-MODEL tool (https://swissmodel.expasy.org), using sequences identified through bioinformatics and minigene experimental products of the splice mutation. Protein structure visualization was performed using PyMol (v3.1.0a0 Open-Source; http://github.com/cgohlke/pymol-open-source-wheels/releases).

Due to the absence or low levels of DNAH9 in peripheral blood, obtaining DNAH9 cDNA from this source for splicing pattern analysis was not feasible. Therefore, minigene technology was used to analyze the splicing pattern of the c.3743+1G>T mutation, which is located in intron 19. Bioinformatics analysis using the RDDC tool predicted that this mutation could potentially affect the splicing of exon 19. To investigate the splicing pattern, the target fragment containing exons 18, 19, and 20, along with a partial sequence (150–200 bp) of intron 20, was integrated into a pCMV-MYC vector (Takara Biotechnology Co., Ltd.) using homologous recombination technology. Briefly, primers with homologous arms were designed to amplify the target fragment, which then carried overlapping sequences homologous to the ends of the linearized vector. The integration was achieved through a homologous recombination reaction (ClonExpress MultiS One Step Cloning Kit; Vazyme). Genomic DNA from a control individual and the proband were used as templates to amplify the target sequence, with the necessary primers listed in Table SI. PrimerSTAR MAX DNA Polymerase (Takara Biotechnology Co., Ltd.) was used for amplification under the following thermal cycling conditions: 98°C for 10 sec (denaturation), 58°C for 5 sec (annealing) and 72°C for 30 sec (extension), for a total of 30 cycles. After the plasmid construction was completed, Sanger sequencing was performed on the plasmid. The cells were cultured in a 6-well cell culture plate (CellPro) containing medium composed of 90% DMEM, 10% fetal bovine serum (FBS), and 1% penicillin-streptomycin mixture. They were then incubated in a cell culture incubator at 37°C with 5% CO₂. When the cell density reached 60–70%, 2 µg of the plasmid DNA was transfected into 293T cells using Neofect DNA transfection reagent (Neofect). At two days following transfection, cellular RNA was extracted using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and reverse transcribed into cDNA under the following thermal cycling conditions: 37°C for 15 min, 85°C for 5 sec (PrimeScript RT reagent Kit with gDNA Eraser, Takara Biotechnology Co., Ltd.). The resulting cDNA was then subjected to PCR amplification using GoTaq Green Master Mix (Promega Corporation) under the following conditions: 95°C for 30 sec (denaturation), 59°C for 30 sec (annealing) and 72°C for 30 sec (extension), for a total of 35 cycles. The PCR products were subsequently analyzed by Sanger sequencing. The primers used for amplification are listed in Table SI.

A compound heterozygous mutation in DNAH9 [NM_001372.3: c.11176C>T (p.Arg3726Trp); c.3743+1G>T] was identified in the fetus by WES and Sanger sequencing. No pathogenic mutations were identified in other genes known to be associated with primary ciliary dyskinesia (PCD). The mother of the proband was a carrier of the c.3743+ 1G>T mutation and the father was a carrier of the c.11176C>T mutation (Fig. 2). The 1000 Genomes Project (East Asian), ESP6500, gnomAD exome and ExAC databases indicate that the c.11176C>T (p.Arg3726Trp) and c.3743+1G>T mutations are extremely rare or absent, respectively, in the general population. Bioinformatics analysis performed using PolyPhen-2, MutationTaster, SIFT and CADD tools predicted both mutations to be pathogenic (Table SII). The c.11176C>T mutation occurs in a residue that is highly conserved across species (Fig. 2C). In addition, this mutation has been previously reported as pathogenic, causing the downregulation of DNAH9 mRNA expression, ultrastructural defects in ciliary outer dynein arms and reduced ciliary beating frequency (11). The c.3743+1G>T mutation is predicted to result in two aberrant splicing events: Skipping of exon 19 (167 bp) and the insertion of a 691-bp segment of intron 19. Compared with the structure of the wild-type DNAH9 protein (Fig. 3A), each aberrant splicing event is predicted to lead to a frameshift and the production of truncated DNAH9 proteins (Fig. 3B). In addition, 3-dimensional (3D) structure prediction indicates that the c.11176C>T (p.Arg3726Trp) mutation will disrupt the hydrogen bond between the arginine residue at position 3,726 and the glutamic acid residue at position 3,594 (Fig. 3C, red arrow). This disruption is expected to lead to the formation of an α-helical structural segment (Fig. 3C, dashed red oval) not present in the wild-type protein (Fig. 3C, dashed blue oval).

Minigene analysis revealed that the c.3743+1G>T mutation produced two expression products while the wild-type gene produced only one (Fig. 3D), which is consistent with the results of the bioinformatics analysis. Sanger sequencing confirmed that the product with the higher molecular weight contained a portion of intron 19 (Fig. 3D), while the product with the lower molecular weight resulted from the skipping of exon 19 (Fig. 3D). The 3D structure predictions were consistent with the bioinformatics forecasts, indicating that the two abnormal splicing events lead to the production of two truncated DNAH9 proteins, one lacking a tail and the other missing the spherical head domain (Fig. 3B). According to ACMG guidelines, both the c.3743+1G>T and c.11176C>T (p.Arg3726Trp) mutations are classified as likely pathogenic.

The 9 previous publications on DNAH9 mutations (9–17), describe findings for 24 patients in total (Table SIII; Fig. S1). These include 10 patients who presented with congenital heart disease (P3, P7-11, P13, P14, P18 and P20), 7 patients with respiratory diseases (P3-6, P21-P23), 3 patients with asthenozoospermia (P1, P2 and P6), and 5 patients who exhibited only situs anomalies (P12, P15-17 and P19). In addition, 1 patient exhibited heterotaxy of abdominal organs and intrauterine fetal death (P24); this patient also carried the RSPH1, c.121G>A (p.G41R) mutation. P6 presented with both respiratory diseases and asthenozoospermia, and P3 had both congenital heart disease and respiratory diseases. Mutation sites associated with different phenotypes were widely distributed across the DNAH9 gene locus, with no obvious pattern discernible.