Genomic DNA was extracted from peripheral blood from the patient and his healthy parents. Written informed consent was obtained from the patient’s parents. The present study was ethically approved by the Ethics Committee of The First Affiliated Hospital of Dalian Medical University.

Whole exome sequencing (WES) was performed using the IDT xGen Exome Research Panel v2.0. High-throughput sequencing was performed by BGI DNBSEQ-T7 (PE150). The sequencing process was performed by Beijing Chigene Translational Medicine Research Center Co. Ltd. The reliability of the whole exome sequencing data was confirmed by the following quality control metrics: the average sequencing depth ≥100×, and the 20× coverage rate of the target region ≥95%, Q20 percentage (indicating a ≤1% error probability) ≥95%, Q30 percentage (indicating a ≤0.1% error probability) ≥85% and GC content 48%–51%.

Raw Data Preprocessing: Adapters and low-quality reads were trimmed using fastp. Read Alignment: Clean reads were aligned to the Ensembl GRCh37/hg19 reference genome with the Burrows-Wheeler Aligner (BWA). Variant Calling & Filtering: Single nucleotide polymorphisms (SNPs) and short insertions/deletions (indels) were called using the Genome Analysis Toolkit (GATK). High-confidence variants were filtered with criteria: minimum sequencing depth (DP) ≥5×, minimum genotype quality (GQ) ≥35, and allele balance (for heterozygous variants) between 10% and 85%. Minor Allele Frequency (MAF) Annotation: Variants were annotated for MAFs using Chigene’s in-house online system (www.chigene.org), leveraging databases such as the 1000 Genomes Project, dbSNP, ESP, ExAC, and Chigene in-house MAFs database. Protein Structural Impact Prediction: Functional alterations in protein structure were predicted using PROVEAN, SIFT, PolyPhen-2 (HDIV/HVAR), MutationTaster, M-Cap, and REVEL. Splice Site Functional Impact Prediction: Effects on splice sites were assessed via MaxEntScan, dbscSNV, and GTAG. Pathogenicity Annotation: Variants were evaluated for pathogenicity following the ACMG guidelines, with annotations cross-referenced against OMIM, HGMD, and ClinVar databases. Co-segregation & Phenotype Correlation Analysis: Finally, co-segregation analysis (to validate inheritance patterns) and phenotype correlation analysis (to link genetic variants with clinical manifestations) were conducted.

Sanger sequencing was employed to validate the presence of the c.3184G>A and c.2704-18C>G mutations in the proband and his parents. The primers used for PCR amplification are detailed in Supplementary Table S1. The PCR amplicons were subjected to Sanger sequencing on an ABI 3730XL sequencer, followed by sequence analysis utilizing DNASTAR software and alignment against the reference sequence of the PNPLA6 gene (NM_006702.5).

This experiment was designed to verify whether the c.2704-18C>G variant of the PNPLA6 gene (a variant near the splice site of intron 25) disrupts normal mRNA splicing, detect the presence of aberrant transcripts in the patient (e.g., exon skipping, intron retention), and further clarify the impact of this variant on gene function (e.g., whether it causes frameshift or protein truncation). Total RNA was extracted from both patient and control samples using the Beyo Tech Blood RNA Mini Extraction Kit, followed by reverse transcription into cDNA. Specific RT-PCR amplification was performed simultaneously on patient and control cDNA samples targeting the PNPLA6 gene (888 bp product) and the reference gene WDR45 (508 bp product). The primers used for RT-PCR amplification are detailed in Supplementary Table S2.

The patient is a 31-year-old young male with the following clinical features: height 165 cm, weight 52.5 kg, absence of an Adam’s apple, thick eyebrows and eyelashes, sparse facial and axillary hair, genital Tanner stage I, pubic hair Tanner stage III-IV and pes cavus. The patient was born full-term via vaginal delivery, but exhibited slow growth. At the age of 9, he experienced bilateral visual decline and was suspected to exist growth hormone deficiency. He received growth hormone therapy for 2–3 months but discontinued treatment on his own, with no significant improvement in vision. At 21 years old, he sustained an ankle fracture in a car accident and after orthopedic treatment, he still had gait instability. By the age of 31, he developed involuntary head tremors, which worsened with emotional stress or postural changes. The patient has been treated sequentially with Almarl 5 mg twice daily to reduce involuntary tremors and Idebenone 30 mg three times daily to improve cerebral metabolism and alleviate psychiatric symptoms, both of which have been confirmed to be effective. There is no family history of similar conditions in his extended relatives.

Brain Magnetic Resonance Imaging (MRI) revealed cerebellar atrophy with mild thinning of the pons and medulla oblongata (Figures 1A,B). Coronal MRI reveals atrophy of the anterior pituitary lobe, revealing diffuse volume loss and symmetric thinning to a height of 1–2 mm, and an anatomically normal posterior pituitary and stalk (Figure 1C). These features were compared with normal control MRIs (Figures 1D–F). Fundoscopic examination revealed pale optic discs with blurred margins, tortuous retinal arteries and veins, extensive atrophy of the posterior pole, and peripheral retinal depigmentation with pigmentary changes. The macular foveal reflex was absent. Optical coherence tomography (OCT) of both eyes demonstrated thinning and atrophy of the macular retina (Supplementary Figure S1). The hormone profile of the pituitary-thyroid axis revealed a markedly suppressed thyroid-stimulating hormone (TSH) level of <0.005  mIU/L, elevated free triiodothyronine (FT3) at 11.37 pmol/L, elevated free thyroxine (FT4) at 36.25 pmol/L, and a thyrotropin receptor antibody (TRAb) level of 5.97 IU/L. These findings collectively indicated thyrotoxicosis. Hormonal assessment of the pituitary-gonadal axis revealed decreased levels of testosterone (TES), luteinizing hormone (LH), follicle-stimulating hormone (FSH), and androstenedione (AND) establishing the diagnosis of hypogonadotropic hypogonadism.

To further investigate the genetic basis of the patient’s condition, WES was performed. The analysis identified two heterozygous variants in the PNPLA6 (NM_006702.5) gene: c.3184G>A (p.Val1062Met) and c.2704–18C>G (Figures 2A,B). Segregation analysis further revealed that the c.3184G>A variant was inherited from the father, whereas the c.2704–18C>G variant was derived from the mother (Figure 2C). The first variant, c.3184G>A, results in a nucleotide substitution from guanine (G) to adenine (A) at position 3,184, leading to an amino acid change from valine (Val) to methionine (Met) at position 1,062 (a missense mutation). This variant has been documented in the dbSNP database under the accession number rs587777182. According to the ACMG classification framework, this variant was classified as likely pathogenic (PM1+PM2_Supporting + PM3+PP2+PP3) (Table 1). Conservation analysis was performed using MEGA X software (http://www.megasoftware.net), which revealed that this site is highly conserved across multiple species (Figure 2D). Pathogenicity prediction tools, including Provean, SIFT, Polyphen2_HDIV, Polyphen2_HVAR and MutationTaster, consistently indicated that this variant is likely to affect protein function. The substitution of Val with Met disrupts the stability of the α-helical scaffold, causing structural disorganization of the catalytic funnel and misalignment of the Ser1014/Asp1144 dyad, which significantly impairs or abolishes the esterase activity of PNPLA6 (Synofzik et al., 2014). The second variant, c.2704–18C>G, is located within intron 25 and involves a substitution of cytosine (C) to guanine (G) at the −18 site of nucleotide 2,704 of the PNPLA6 coding sequence representing a splice site-proximal intronic mutation. Splicing prediction tools, such as MaxEntScan and GTAG suggested that this variant may disrupt normal mRNA splicing. Sanger sequencing of the proband’s sample illustrates wild-type (WT) and mutant PNPLA6 cDNA spanning exons 25–26, demonstrating the aberrant splicing at this locus. Specifically, the mutant trace reveals a heterozygous 29-bp deletion upstream of exon 26 (marked by the red bracket), which disrupts the canonical splice junction between exons 25 and 26—in contrast to the WT trace, which exhibits seamless splicing of exons 25 and 26 with no deletions. These sequencing data confirm the presence of abnormal splicing (Figure 3A). The predicted amino acid alteration was identified as NM_006702.5:c.2704_2732del,p. (His902Alafs*108). This variant results a 29-nucleotide deletion (c.2704_2732del) in the coding region, causing a frameshift mutation. The original histidine at position 902 is replaced by alanine, followed by premature termination of protein synthesis after 108 amino acids downstream due to encounter with a stop codon (Figures 3B,C). Such a frameshift alteration is likely to severely impact protein structure and function. According to the ACMG classification framework, this variant was classified as likely pathogenic (PM2_Supporting + PM3+PP3+PS3 _Moderate).