The pedigree of the Chinese SEDT family is shown in Figure 1, demonstrating a characteristic X-linked recessive inheritance pattern. The proband, a 16-year-old Chinese male, sought evaluation for short stature at Shanghai Children’s Hospital. Peripheral blood samples and clinical information were collected for available family members.

To elucidate the genetic cause of the disorder, we performed trio whole-exome sequencing (TWES) on the proband and his parents as previously described (Xiaozhen et al., 2021). With participant consent, we extracted genomic DNA for TWES and subsequent Sanger sequencing from peripheral blood samples, employing the DNA Blood Mini Kit (QIAGEN, Germany). An IDTxGen® Exome Research Panel (IDT, USA) was used to capture the exons, and HiseqX10 (Illumina, USA) was used to sequence the DNA fragments. The acquired sequencing data were further annotated and filtered. Minor allele frequency (MAF) < 1% in the gnomAD (https://gnomad.broadin-stitute.org/), pathogenicity prediction algorithms, disease-association database (ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) and HGMD (http://www.hgmd.cf.ac.uk/ac/index.php)), and disease phenotypes were used to search for all possible disease-cause variants. The potentially pathogenic variants were verified through Sanger sequencing. The pathogenicity of the identified variations was determined following the guidelines set forth by American College of Medical Genetics and genomics and association for molecular pathology (ACMG/AMP) (Richards et al., 2015).

To validate the effect of c.94–11C > G variation on function, we performed in vitro experiments involving a minigene splicing study and protein expression assay. Amplicons containing either wild type or variant allele [intron 3 (211bp)-Exon 4 (145 bp)] were generated by standard PCR and cloned into pcMINI-C vector. We verified the successful construction of the expression vectors through bacterial liquid PCR and Sanger sequencing.

The recombinant vectors (pcMINI-C-TRAPPC2-wt and pcMINI-C-TRAPPC2-mut) were then transiently transfected into human embryonic kidney cells (HEK-293T) and human cervical cancer cells (HeLa), according to the provided instructions. After 48 h of cell culture, total RNA was extracted from both HEK-293T and HeLa cells using the TRIzol method (Thermo Fisher Scientific, Waltham, MA, USA). This RNA was then reverse transcribed into cDNA using a reverse transcription kit (TransGen Biotech, Beijing, China), following the manufacturer’s instructions. Subsequently, standard PCR was performed with primers flanking the minigene. The PCR products were identified by 2% agarose gel electrophoresis and verified by Sanger sequencing. The used primers are listed in Supplementary Table S1.

We synthesized human TRAPPC2 cDNA (transcript NM_001011658.4), followed by amplification to obtain PHAGE-TRAPPC2-wt (using primers PHAGE-TRAPPC2-SalI-F and PHAGE-TRAPPC2-NotI-R) and pEGFP-TRAPPC2-wt (using primers pEGFP-C1-TRAPPC2-HindIII-F and pEGFP-C1-TRAPPC2-BamHI-R) fragments. Mutant fragments were obtained with TRAPPC2-MUT-F and TRAFPPC2-MUT-R primers. SalI and NotI were used to construct wt or mutant TRAPPC2 onto the PHAGE vector, and HindIII and BamHI were used for construction onto the pEGFP-C1 vector. We assessed the successful construction of eukaryotic expression plasmids through PCR and Sanger sequencing.

Then we transfected HEK-293T cells with the different expression plasmids 1 μg when the cell density reached 60%. After 48 h, total RNA was isolated from cultured cells and 1 μg of RNA was reverse transcribed to cDNA. The expression of wild type and mutant TRAPPC2 was detected by qPCR. Total protein was extracted using a Protein Extraction Kit (Sangon Biotech, Shanghai, China). Subsequently, 25 μg of total protein was separated using 8% sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) and then transferred to nitrocellulose membranes (Merck Millipore, Darmstadt, Germany). After blocking with 5% skim milk, the membranes were incubated overnight at 4 °C with specific primary antibodies. Finally, the bands were detected using ECL Prime Western blotting Reagent (GE Healthcare, Buckinghamshire, UK). The primary antibodies used were rabbit anti-GAPDH (Cell signaling technology; 1:1000), mouse anti-HA (Dia-an; 1:2000), and mouse anti-GFP (Dia-an; 1:1000).

The male proband (IV-1) (Figure 1), a 16-year-old Chinese boy, first presented to our clinic with a concern of short stature. He measured 137 cm in height, with an arm span of 150 cm, displaying disproportionate short stature, a short neck, and a barrel-shaped chest. He did not exhibit any other systemic complications. The frontal and lateral radiograph of the spine showed mild kyphoscoliosis (Figures 2A,B). The lateral radiograph of the lumbosacral spine showed platyspondyly and a hump-shaped appearance in the central and posterior portions of vertebral endplates, with no significant narrowing of disc spaces (Figure 2B). Knee radiographs showed flattened tibial plateaus (Figure 2C). Mild flattened femoral heads, coxa vara, irregular hip-joint surfaces and short femoral neck were observed in the radiographs of the pelvis (Figure 2D). The pedigree of the SEDT family was shown in Figure 1. In the family, three other living adult males (II-1, III-13, III-18) experienced similar symptoms and manifestations. Their heights measured 134, 138, and 134 cm, respectively, and they also exhibited disproportionate short stature, short neck, a barrel-shaped chest. Their arm spans exceed their height measurements by 13–14 cm. Case II-1, who was 65 years old, reported severe back, knee and hip pain since the age of 17, with limited knee joint motion. Case III-13, the maternal uncle of the proband, was 32 years old. X-ray radiographs of III-13 showed more typical characteristics of SEDT, including flattening of the vertebral bodies, significantly narrow intervertebral disc spaces, end plate sclerosis, narrow hip-joint surfaces and short femoral necks (Figures 2E–H). Detailed clinical and radiographic findings of the affected males are summarized in Table 1. There were no affected women in the family (the average height was 156.4 cm). The pattern of inheritance supported an X-linked recessive disorder through four generations.

TWES and Sanger sequencing revealed a novel hemizygous variant, c.94–11C>G, situated in intron three of the TRAPPC2 gene (NM_001011658.4), which was inherited from the proband’s mother (Figure 1). Notably, this variant is novel and absent from both the HGMD and the gnomAD. Further Sanger sequencing analysis of this family showed that the same variant was detected as hemizygous in affected individuals (II-1, III-13 and III-18), and multiple maternal female relatives of the proband (II-3, II-9, II-11and IV-2) were carriers of the variant (Figure 1). The remaining males with a normal phenotype tested were negative for this variant. In silico tools, Varseak (https://varseak.bio) and SpliceAI (https://github.com/Illumina/SpliceAI), predicted that the c.94–11C>G variant affects normal splicing and generates an aberrant splicing site. No other likely pathogenic or pathogenic (LP/P) variants associated with short stature were found.

The intronic variant (c.94–11C > G) detected in TRAPPC2 exhibited a co-segregation feature within this pedigree, indicating its pathogenicity for the disease. Since we could not obtain sufficient blood samples from proband to validate the variation in vivo, in vitro minigene assays were then performed to determine the effect of the variant on mRNA splicing. Recombinant vector sequencing showed the successful insertion of both wt and mutant TRAPPC2 minigenes into the pcMINI-C vector (Figure 3A). The minigene assays showed that pcMINI-C-TRAPPC2-wt produced an expected band of approximately 500 bps in both 293T cells and Hela cells (Figure 3B). In contrast, pcMINI-C-TRAPPC2-mut produced a larger band than pcMINI-C-TRAPPC2-wt (Figure 3B), suggesting that c.94–11C > G variant may have caused the abnormal splicing shown in Figure 3C. Sequencing and online alignment of the PCR product showed the retention of an additional 10 bps in intron three of TRAPPC2 (Figure 3D). These results confirm that the intronic variant c.94–11C > G caused abnormal TRAPPC2 mRNA splicing, with the retained 10 bps resulted in a frameshift and finally lead to the creation of a premature termination codon (PTC).

Given the abnormal TRAPPC2 mRNA splicing and the appearance of a PTC, we postulated two potential outcomes: the encoding of a C-terminal truncated TRAPPC2 protein or the activation of nonsense-mediated mRNA decay. To test this hypothesis, we constructed two sets of expression plasmids to verify the effect of aberrant splicing on TRAPPC2 expression.

The RT-PCR product sequencing showed that the mut-vector (PHAGE and pEGFP-C1) produced c.93_94ins tgtcatctag (p.Asp32Cysfs*23), indicating that the eukaryotic expression vectors were successfully constructed (Figure 4A). The RT-PCR results also showed that the mRNA expression level of the splicing variant (mut) was indeed significantly lower than those of the wt in both sets of vectors (Figure 4B).

Then we transfected four TRAPPC2 expression constructs into 293T cells to evaluate the effect of variant on protein expression. Finally, we found that the c.94–11C>G variant led to the production of a truncated TRAPPC2 protein with significantly reduced expression compared to the wild type when using the pEGFP-C1 vector. In the case of the PHAGE vector, the mutant construct exhibited no discernible band, indicating that the c.94–11C>G variant activated nonsense-mediated mRNA decay (Figure 4C). These results underscore that the c.94–11C>G variant results in the absence of TRAPPC2 protein, which is required for the export of procollagen from the endoplasmic reticulum to the Golgi apparatus.