Skeletal dysplasia is a group of rare genetic disorders associated with abnormalities of cartilage and bone (Victoria et al., 2018). Genetic skeletal disorders are clinically and genetically heterogeneous with more than 350 genes accounting for the diversity of disease phenotypes (Bonafe et al., 2015). Despite great advances in imaging technology, fetal skeletal dysplasia is difficult to diagnose in utero because of its large number and overlapping phenotypic features. Prenatal diagnosis of skeletal dysplasia is particularly challenging due to highly variable but limited prenatal phenotypes.

Currently, short-rib polydactyly syndromes is classified as short-rib thoracic dysplasia with or without polydactyly types 1–17 (Loo et al., 2021). Short-rib thoracic dysplasia 3 with or without polydactyly (SRTD3, OMIM: 613091) covers a range of autosomal recessive or digenic recessive skeletal dysplasia characterized by shortened limbs, a narrow trunk, and associated visceral abnormalities with or without polydactyly (Chen et al., 2018; Cheng et al., 2022). Homozygous or compound heterozygous mutations of dynein cytoplasmic 2 heavy chain 1 (DYNC2H1) gene usually lead to SRTD3 (Mei et al., 2015; Chen et al., 2016; Xia et al., 2021; Zhao et al., 2022).

To date, more than 140 variants in the DYNC2H1 gene have been found in SRTD3 (Chen et al., 2016). Mutations in DYNC2H1 will cause the dysfunction of primary cilia, causing a heterogeneous spectrum of conditions such as skeletal dysplasia (Okamoto et al., 2015; Chen et al., 2016). Whole exome sequencing (WES) has been shown to be a powerful mean for prenatal diagnosis. Nevertheless, the different etiology of ciliopathies and extensive genetic variation lead to phenotypic variability. Thence, it is essential to fully understand the prenatal phenotypes of different ciliopathic syndromes before performing WES assays.

In our study, we analyzed gene mutations in four SRTD3 fetuses by WES and determined that the compound heterozygous mutation of DYNC2H1 was the causative mutation of SRTD3. In this study, we found typical and atypical features of four fetuses with pathogenic variants in DYNC2H1 identified by WES. The discovery of new clinical manifestations and genetic variants may facilitate the prenatal diagnosis of skeletal dysplasia. The possible causes of the affected fetuses in four families were preliminarily identified from the perspective of genetics, which provided strong evidence for their clinical diagnosis and reliable molecular basis for genetic counseling and prenatal diagnosis.

Despite the emergence of skeletal dysplasia in each individual is relatively rare, skeletal dysplasia accounts for a considerable number of congenital malformations in neonates (Lord et al., 2019; He et al., 2021). About 40% of skeletal dysplasia manifests at birth or before birth, and ultrasound remains an essential first-line screening way (Mary et al., 2019; Tang et al., 2020). Due to the diversity and variability of skeletal dysplasia manifestations and their widely overlapping phenotypes, ultrasound alone cannot effectively distinguish between various skeletal dysplasia. It has been reported that prenatal WES can help increase the detection rate of fetal skeletal abnormalities (Lord et al., 2019; Zhao et al., 2021). Therefore, more accurate gene diagnosis is of great value for obstetricians to evaluate fetal prognosis, provide genetic counseling and determine the optimal treatment. In the current study, by using trio-WES and proband-WES study with comprehensive analysis, eight mutations (4 pairs of compound heterozygous sites) of DYNC2H1 were identified in four Chinese families with fetal skeletal dysplasia, supporting a diagnosis of SRTD3 (OMIM; 613091).

At present, approximately twenty genes have been found to be involved in SRTD3 (Badiner et al., 2017). Among them, DYNC2H1 gene is more common and has been identified as the causative gene of SRTD3. DYNC2H1, located on chromosome 11q22.3 and contained 90 exons, is a cytoplasmic dynein involved in the retrograde transport of cilia, which is an important component of endochondral bone development (Vig et al., 2020; Zhang et al., 2020). Fetuses with SRTD3 usually displays with extremely shortened long bones, round metaphyseal ends and lateral spikes, a small and narrow thorax, and preaxial and postaxial polydactyly, sometimes exhibiting with bowing femora, genitals and multiple deformities. In our study, four affected fetuses showed different SRTD3 phenotypes. SRTD3 disease has been found to have clinical and genetic heterogeneity, such as significant differences in the course and severity of the thoracic phenotype between individuals with different alleles, indicating that the DYNC2H1 phenotype may be affected by allele modification (Zhang et al., 2020).

Currently, a large number of different variants of DYNC2H1 gene have been reported in many databases, but there are no hot spot variants. Most of these variants are compound heterozygous variants, generally one is missense variant, and the other is frameshift or truncated protein variant (Chen et al., 2018; Thakur et al., 2021; Xia et al., 2021). To the best of our knowledge, only four were homozygous mutations of DYNC2H1 gene in all cases (Geng et al., 2020). The low reported rate of homozygous mutations in DYNC2H1 may be due to the early death of homozygous embryos. The DYNC2H1 compound heterozygous variant fits the autosomal recessive inheritance pattern of SRTD3. Cheng et al. report a Chinese foetus with SRTD3 carrying a c.11483T>G and c.2106+3A > T mutations who exhibited severe SRTD3 phenotypes (Cheng et al., 2022). Zhang et al. identify two novel compound heterozygous variants c.358G>T and c.928A>T in DYNC2H1 in two Chinese fetus, but the phenotypes of the two fetuses showed different SRTD3 phenotypes (Zhang et al., 2020). Deng et al. find two novel compound heterozygous mutations c.2992C>T and c.12836G>C of the DYNC2H1 gene in a fetus with polyhydramnios in China (Deng et al., 2018). In our study, WES also identified a compound heterozygous variation c.8617A>G (p.Met2873Val) and c.7053_7054del (p.Cys2351Ter) of DYNC2H1 gene in fetus 2, which also presented with polyhydramnios. Therefore, we hypothesized that polyhydramnios may also be one of the prenatal manifestations of SRTD3 fetuses.

In our study, eight mutations, 4 pairs of compound heterozygous variants including c.3842A>C (p.Tyr1281Ser) and c.8833-1G>A, c.8617A>G (p.Met2873Val) and c.7053_7054del (p.Cys2351Ter), c.5984C>T (p.Ala1995Val) and c.10219C>T (p.Arg3407Ter), c.5256del (p.Ala1753GlnfsTer13) and c.9737C>T (p.Thr3246Ile), were found in four Chinese fetus. According to the ACMG guidelines, c.8617A>G (p.Met2873Val), c.7053_7054del (p.Cys2351Ter), c.5984C>T (p.Ala1995Val), c.10219C>T (p.Arg3407Ter) and c.5256del (p.Ala1753GlnfsTer13) were rated as pathogenic or likely pathogenic variants, others variants were predicted to be variants of uncertain significance mutations. Four of the eight variants, including c.3842A>C (p.Tyr1281Ser), c.8833-1G>A, c.7053_7054del (p.Cys2351Ter), c.5256del (p.Ala1753GlnfsTer13), were not reported in HGMD and ClinVar databases. Among which, c.10219C>T (p.Arg3407Ter), c.5984C>T (p.Ala1995Val) and c.9737C>T (p.Thr3246Ile) were reported in ClinVar databases, and c.8617A>G (p.Met2873Val), c.10219C>T (p.Arg3407Ter), c.5984C>T (p.Ala1995Val) were found in HGMD databases. Herein, we reported two novel compound heterozygous variants c.3842A>C (p.Tyr1281Ser) and c.8833-1G>Ain DYNC2H1 in fetus 1. Moreover, c.7053_7054del (p.Cys2351Ter) and c.5256del (p.Ala1753GlnfsTer13) were novel variants found in fetus 2 and fetus 4, respectively. In addition, the parents in all cases were without any minor symptoms. Moreover, parental consanguinity is presented in each case. As a whole, the discovery of these novel mutations not only enriched the mutation database, but also provided a basis for genetic counseling and further prenatal diagnosis of families.

SRTD3 has been reported to have a variety of phenotypes such as shortened limbs, a narrow trunk, and associated visceral abnormalities with or without polydactyly and so on. Chen et al. report a Chinese fetus with SRTD3 carrying c.8077G > T and c.11741_11742delTT mutations in DYNC2H1 who exhibited short limbs, a narrow chest and bilateral polydactyly of the hands and feet phenotypes (Chen et al., 2018). Cheng et al. identify novel compound variants of DYNC2H1 (c.11483T > G and c.2106 + 3A > T), which presented with abnormal rib curvature, narrow thorax, bilateral hypoplastic lungs, bilateral polydactyly, syndactyly, and foetal visceral situs inversus with mirror-image dextrocardia (Cheng et al., 2022). Xia et al. report novel compound heterozygous variants of DYNC2H1 (c.6591_6593delTGG and c.7883T>C) in a fetus, which was exhibited a long, narrow thorax with short ribs, shortened long bones, spurs at the metaphysis of the long bones and congenital bowing of the femurs (Xia et al., 2021). Mei et al. find compound heterozygous mutations of DYNC2H1 gene (c.1151 C>T and c.4351 C>T) in the fetus with SRTD3, which were showed short limbs, a narrow thorax, short ribs with marginal spurs, and polydactyly (Mei et al., 2015). He et al. find a Chinese fetus carrying compound heterozygous variants of DYNC2H1 gene (c.2225T > G and c.10219C > T) in the fetus with phenotypes including an increased nuchal translucency, a narrow thorax, severely short limbs, postaxial polydactyly of right hand, endocardial cushion defect, megacystis and omphalocele (He et al., 2020). In our study, two compound heterozygous mutations (c.5984C>T (p.Ala1995Val) and c.10219C>T (p.Arg3407Ter)) in the DYNC2H1 gene were observed by WES from fetus 3, and fetus 3 phenotypes containing severe short limbs, abnormal posture of hand and foot, cardiac malformation and single umbilical artery. Both the case reported by He et al. and case 3 in this study carried the c.10219C>T (p.Arg3407Ter) mutation and both fetuses had severely short limb and cardiac malformations. The c.10219C>T (p.Arg3407Ter) mutation of DYNC2H1 may be associated with severe short limbs and heart malformations in SRTD3. In our study, the phenotype of the four fetuses were also inconsistent. For example, the phenotype of fetus 1 included multiple anomalies of the fetus, including short long bones and narrow thorax. The phenotype of fetus 2 contained short long bones, seroperitoneum and polyhydramnios. The phenotype of fetus 3 involved severe short limbs, abnormal posture of hand and foot, cardiac malformation and single umbilical artery. The phenotype of fetus 4 concerned short femur length and slightly bending. In addition to fetus 3, the SRTD3 phenotype of the other three cases was relatively lighter. In conclusion, we speculate that different mutation sites of DYNC2H1 may produce different SRTD3 phenotypes. Therefore, it is important to further explore the effects of different sites on SRTD3 phenotypes.

Rapid advances in high-throughput sequencing and bioinformatics technology have allowed it possible to reveal how disease mutations disrupt protein-protein interaction (PPI) networks in human cells (Cheng et al., 2021). It is well known that protein-protein interaction network plays a crucial role in all biological processes (Sneha et al., 2018). Protein-protein interaction networks have been used to study the pathogenesis and potential pathogenic molecules of many diseases (Röhl et al., 2022; Vagiona et al., 2022). In our study, the protein-protein interaction network was consisted of 11 nodes and 51 edges. Then, these 11 genes were utilized to perform the GO enrichment analysis. GO terms enrichment analysis showed that ciliary tip, intraciliary transport, intraciliary retrograde transport, cytoplasmic dynein complex, intraciliary transport particle, ciliary plasm, centrosome, ciliary basal body, appendage development and motile cilium were significantly enriched GO terms. All these GO processes are biological processes related to cilia function. SRTD3 has been widely recognized as a typical ciliary disease, and its characteristics overlap with the molecular mechanisms of ciliary dysfunction (Taylor et al., 2015). Therefore, we speculated that these 11 genes may participate in the SRTD3 progression by regulating these GO processes.

Several studies have found that genes splicing mutations can affect phenotype and disease risk (Manning and Cooper, 2017; Park et al., 2018). These studies have broad implications as they can shed light on disease pathogenesis and inform treatment options for the disease (Matesanz et al., 2015; Chen et al., 2021). In regard to the c.8833-1G > A variant, the vitro minigene splice assay confirmed that the variant caused exon 56 skipping, suggesting that this variant could increase the use of PS3 evidence. On the basis of the ACMG/AMP guidelines, c.8833-1G > A variant was classified as likely pathogenic with the criteria of PS3+PVS1_S + PM2_Supporting.

The present study built the genetic diagnosis on four families affected with SRTD3 by WES analysis, and identified eight DYNC2H1 variants, which expanded its mutation spectrum. The results of our study offer a strong theoretical foundation for genetic counseling in affected families and provide a basis for elucidating importance of DYNC2H1 function in the pathogenesis of SRTD3. The next functional studies will contribute to uncover the potential pathogenesis of SRTD3 induced by DYNC2H1 gene variants and offer potential therapeutic ways in the future.