Identification of a rare missense variant in the gene SPATA31H1 associated with hereditary gingival fibromatosis in a large Chinese family

Abstract

Background

Hereditary gingival fibromatosis (HGF), a benign hereditary oral disease, is characterized by significant gingival hyperplasia. The disease typically manifests as autosomal dominant inheritance, with great genetic heterogeneity. In this study, we studied a four-generation Chinese kindred with nonsyndromic HGF, including 19 patients. Given the significant genetic heterogeneity of HGF, identifying novel pathogenic genes is crucial for advancing our understanding of the disease. This study offers a novel theoretical foundation for HGF research.

Methods

Family members underwent clinical examination and data collection. Pathogenic genes and mutation sites were identified using Whole Exome Sequencing (WES), SNP chip whole-genome linkage analysis, and Sanger sequencing. The impact of mutations on gene function was assessed via reverse transcription quantitative polymerase chain reaction (RT-qPCR), Western blot and subcellular localization experiments. The potential pathogenic mechanism of HGF was explored through transcriptome sequencing of affected gingival tissues.

Results

Exome sequencing identified a novel heterozygous missense variant (c.14366G > T) in SPATA31H1 gene of the patient, resulting in a Tryptophan to Leucine substitution at the conserved amino acid position 4789 of the protein. Protein structure prediction, RT-qPCR, and western blotting analysis revealed that the missense variant in this gene was likely pathogenic. Transcriptome sequencing analysis of the family samples suggested that the variant affected pathways related to gingival fibrosis.

Conclusions

In this study, it was found that a missense mutation in the gene SPATA31H1 is closely associated with non-syndromic HGF disease.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-08507-z.

Background

Hereditary gingival fibromatosis (HGF) is a hereditary oral disease, and its most notable characteristic is the benign hyperplasia of the gingival tissue. The primary manifestation of gum tissue is usually extensive, progressive hyperplasia, which can affect the marginal gingiva, attached gingiva, and gingival papilla of the entire mouth [1]. Enlarged gingival tissue is usually pink, smooth, nodular, tough, and difficult to bleed; it may cover part or even the entire clinical crown. This can cause primary tooth retention, delayed permanent tooth eruption, malocclusion, and other symptoms, leading to the abnormal morphology of dentition and the maxillofacial region, severely affecting the masticatory function, aesthetics, and pronunciation of patients, and making patients feel uncomfortable [1–3]. According to the clinical phenotype, HGF is usually divided into nonsyndromic and syndromic HGF. Nonsyndromic HGF manifests as a single symptom of gingival hyperplasia [4]. Syndromic HGF is not only characterized by diffuse and extensive gingival overgrowth but can also be associated with other systemic diseases. For example: Cantú Syndrome (causative gene ABCC9, OMIM: 239850), which can manifest as hypertrichosis, mental retardation, and epilepsy [5]; Rutherfurd Syndrome (causative gene unknown, OMIM: 180900), which can present with abnormal tooth eruption and corneal dysfunction [6]; Ramon Syndrome (causative gene unknown, OMIM: 266270), characterized by cherubism, epilepsy, mental retardation, and slow growth [7]; Zimmermann-Laband Syndrome Type 1 (causative gene KCNH1, OMIM: 135500), which can manifest with osteogenesis imperfecta of the fingers and toes with hepatosplenomegaly [8].

Studies have demonstrated that HGF exhibits obvious genetic heterogeneity [9]. So far, numerous scholars both domestically and internationally have mapped several chromosome regions that are associated with possible disease-causing mutations, including 2p21-p22 (GINGF1, OMIM: 135300) [10]; 5q13-q22 (GINGF2, OMIM: 605544) [11]; 2p22.3-p23.3 (GINGF3, OMIM: 609955) [12]; 11p15 (GINGF4, OMIM: 611010) [13]; and 4q12 (GINGF5, OMIM: 617626) [14]. SOS1 (Son of Sevenless 1) in the GINGF1 locus (OMIM: 182530) [10, 15] and REST (RE1-silencing transcription factor) in the GINGF5 locus (OMIM: 600571) [14] are currently recognized as HGF-related pathogenic genes. In addition, the ZNF513 (zinc finger protein 513) and KIF3C (kinesin family member 3 C) genes on chromosome 2 [16], as well as the ZNF862 (zinc finger protein 862) gene on chromosome 7 [17] have been considered to be closely related to nonsyndromic HGF. The clinical phenotypes of HGF are rich and diverse, with significant differences between individuals with different genetic backgrounds. Additionally, the significant genetic heterogeneity of HGF complicates the study of its pathogenesis. Currently, gingivectomy, gingivoplasty, and flap surgery are the primary methods used to treat HGF [2, 18], however, the postoperative recurrence rate is high, causing significant pain in patients. Therefore, it is of great significance for clinical work to explore the genetic basis and pathogenesis of HGF.

This study examined a four-generation family diagnosed with nonsyndromic HGF, identified a heterozygous pathogenic mutation (c.14366G > T) in the SPATA31H1 gene associated with HGF disease, and explored HGF pathogenesis.

Methods

Clinical diagnosis and sample collection

The proband in this study was a 36-year-old female who visited the Department of Stomatology at the First Affiliated Hospital of Zhengzhou University. Specialized examination revealed diffuse hyperplasia of the gingiva, and most teeth were displaced. The patient reported a history of four gingivectomies. We further conducted oral specialty examinations and general health surveys on other members of the family to rule out drug-induced gingival hyperplasia and syndromic HGF, confirming that this family has nonsyndromic HGF. The family consisted of 48 members from 4 generations, including 19 patients. This study was approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University. After obtaining informed consent from patients and their families and signing informed consent forms, we collected peripheral venous blood from the proband (Ⅲ-11) and 24 other family members (13 patients: Ⅱ-2, Ⅱ-3, Ⅱ-6, Ⅱ-8, Ⅱ-10, Ⅱ-12, Ⅲ-2, Ⅲ-5, Ⅲ-15, Ⅲ-18, Ⅳ-2, Ⅳ-3, and Ⅳ-9, and 11 non-patients: Ⅱ-4, Ⅱ-7, Ⅱ-14, Ⅲ-1, Ⅲ-3, Ⅲ-6, Ⅲ-9, Ⅲ-16, Ⅲ-17, Ⅳ-6, and Ⅳ-8). We also collected gingival tissue samples from four of the patients (Ⅲ-2, Ⅲ-11, Ⅲ-18, and Ⅲ-19). The tissue samples from these four patients were used as follows: the sample from individual Ⅲ-11 was used for HE staining and Masson staining; the samples from individuals Ⅲ-2, Ⅲ-18, and Ⅲ-19 (designated as H1, H2, and H3, respectively) were used for RNA-seq, Western blotting, and RT-qPCR experiments. Patient Ⅲ-11 is the proband, a 36-year-old female. Patient Ⅲ-2 is a 32-year-old male, Patient Ⅲ-18 is a 27-year-old female, and Patient Ⅲ-19 is a 25-year-old male. Healthy control group samples were obtained from healthy, non-inflammatory gingival tissue resected during wisdom tooth extraction by five volunteers aged between 20 and 35 years, all of whom provided written informed consent.

Histological staining

Hematoxylin and eosin (H&E) staining

Human gingival biopsy tissues were fixed in 4% paraformaldehyde for 24 h at room temperature. Following fixation, tissues were dehydrated through a graded ethanol series, cleared in xylene, and embedded in paraffin. Serial Sect.  (2 μm) were prepared and baked at 65 °C for 30 min. Sections were deparaffinized and rehydrated through xylene and ethanol gradients before staining with hematoxylin for 10–15 min. After differentiation in 1% hydrochloric acid alcohol for 5 s and rinsing in running water for 15 min, eosin staining was performed for 3 min. Sections were dehydrated, cleared in xylene, mounted with neutral gum, and examined under a light microscope.

Masson trichrome staining

Masson trichrome staining was performed using a commercial staining kit according to the manufacturer’s instructions. Briefly, paraffin-embedded tissue Sect.  (2 μm) were deparaffinized, rehydrated, and stained with Masson’s trichrome solution for 5 min. After rinsing, sections were treated with phosphomolybdic acid for 5 min, followed by aniline blue staining for 5 min. Differentiation was performed using a differentiating solution for 1 min. Sections were washed, dehydrated, mounted with neutral gum, and observed under a light microscope.

Detailed protocols for both staining methods is detailed in the Supplementary Material 1.

Exome sequencing

We selected whole blood genomic DNA from two patients (II-6 and IV-10) and one normal control (III-6) from this family and performed exome sequencing through Novogene Beijing (China). Agilent’s liquid chip capture system was used to efficiently enrich the DNA, followed by high-throughput, high-depth sequencing on the Illumina HiSeq2000 platform with a sequencing depth of 100×. All variants detected by exome sequencing were screened using the following criteria: (1) variants with a frequency higher than 1% in at least one of the following databases were removed: the 1000 Genomes Project (1000g_all), the ESP6500 database (esp6500siv2_all), and the gnomAD database (gnomAD_ALL and gnomAD_EAS); (2) variations in the exon or splice site regions were retained. (3) synonymous SNP variants that are not predicted by software to affect splicing and are located in non-highly conserved regions were removed; small segment (< 10 bp) non-frameshift InDel variants located in repeated region were also removed; (4) variant sites were selected based on the scoring predictions from SIFT, Polyphen, MutationTaster, and CADD, with the requirement that at least half of these four software tools indicate that a site is potentially harmful for it to be retained; splice site region mutations within 2 base pairs (± 1 to 2 bp) of the exon region were retained; variants predicted by dbscSNV to have a splicing impact were also retained. The screened variants were classified according to the American College of Medical Genetics and Genomics (ACMG) variant criteria, and the final variant sites were evaluated based on the principle of co-segregation within the family. Gene-specific primers were designed (forward: 5ʹ-GGATTGGAGCAACTCAGACAAG-3ʹ and reverse: 5ʹ -CCCACAATATGTCCTCCTGCT-3ʹ) to amplify the candidate gene and confirm the variant by Sanger sequencing.

Single nucleotide polymorphism (SNP) chip linkage analysis

Peripheral blood samples were collected from 25 family members (15 patients and 10 normal controls) and genomic DNA was extracted. The Affymetrix Axiom PMRA chip containing 900,000 SNPs, was used to scan the whole genome of 25 family members. The raw scanning results of the SNP chip were genotyped using Affymetrix’s GTC or Illumina’s GenomeStudio. After conducting genome-wide SNP genotyping, quality control was performed on the samples and SNPs to obtain high-quality samples and SNPs, with a final total of 18,455 chromosomal SNPs remaining. A genome-wide non-parametric linkage analysis was conducted on these 18,455 SNPs, and regions with an ExLOD value > 2.0 were selected. For the significant regions identified, haplotype analysis was performed using Merlin 1.1.2.

Bioinformatics

Predicted the functional effect of this variation through Provean and SIFT online websites. Utilizing the I-TASSER software, the three-dimensional models of the wild-type (WT) and mutant-type (MUT) SPATA31H1 proteins were predicted. Analyzed the conservation of the variant site among different species using the ESPript3. The ExPASy database was utilized to predict the physicochemical properties of proteins following mutation.

Cell transfection and subcellular localization

First, using total RNA from normal human gingival tissue as a template, the coding sequence (CDS) of the SPATA31H1 gene was amplified (forward: 5ʹ-AAATCTGTGGCCTTTGCATCG-3ʹand reverse: 5ʹ-AATGGAAGCCTTTCCACGCT-3ʹ). This CDS sequence was then cloned into the XhoI and BamHI sites of the pEGFP-N1 plasmid, and site-directed mutagenesis was performed using this as a template to obtain the SPATA31H1 mutant (forward: 5ʹ-GGCCAGACAATTTGTtGGCAAGCAAAAACTA-3ʹ and reverse: 5ʹ-TAGTTTTTGCTTGCCaACAAATTGTCTGGCC-3ʹ). Human embryonic kidney (HEK) 293T cells were cultured in 12-well dishes, and when the cell confluence reached 60%, transfection with WT or MUT SPATA31H1 plasmid was carried out using Lipofectamine™ 3000 transfection reagent (Invitrogen). After transfection for 48 h, cells were washed three times with phosphate-buffered saline (PBS, Sigma-Aldrich) and fixed with 4% paraformaldehyde (Sigma-Aldrich) for 30 min. Nuclear staining was performed using 4’,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich). Following this, the samples were examined under a confocal fluorescence microscope (model LSM 880).

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

RNA was isolated from gingival tissue samples with Trizol reagent (Invitrogen), followed by reverse transcription to cDNA using the PrimeScript™ RT reagent kit (Takara). The relative mRNA levels of the genes were detected by real-time quantitative PCR using a 2 × RealStar Green Fast Mixture (GenStar). Normalize the expression of the target gene using GAPDH as the reference gene. (The SPATA31H1 amplification primers: forward: 5ʹ-CAGAGCCAACTAGGAAGTTCCC-3ʹ and reverse: 5ʹ-ACCCTGGTGCTAATTCCACAG-3ʹ. The GAPDH amplification primers: forward: 5ʹ-GTGAAGGTCGGAGTCAACG-3ʹ and reverse: 5ʹ-TGAGGTCAATGAAGGGGTC-3ʹ). The expression levels of the gene were calculated using the (2^−ΔΔCT) method.

Western blotting analysis

Fresh gingival tissue was collected and immediately placed it in a mortar, added liquid nitrogen and grinded thoroughly. Tissue was lysed with lysis buffer (Beyotime Biotechnology)containing 1% phenylmethylsulfonyl fluoride (PMSF, Beyotime Biotechnology) to prevent protein degradation. After adding protein loading buffer to each protein sample, they were placed it in a metal bath at 95℃ for 5 min. Protein samples were separated by 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore). After blocking with 5% skim milk at room temperature for 2 h, the membranes were incubated overnight at 4 °C with rabbit monoclonal antibodies against SPATA31H1 (Sigma) and HRP-conjugated antibodies against GAPDH (Sigma). On the subsequent day, the membranes were incubated with goat anti-rabbit IgG (Proteintech) at room temperature for a duration of 2 h, following three washes with Tris-Buffered Saline-Tween 20 (TBST). Protein expression was detected using an automated chemiluminescent image analysis system (Tanon) and SuperSignal™ West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific). The intensity of protein bands was quantified using ImageJ software.

RNA-seq

Transcriptome sequencing was performed on patient samples (Ⅲ-2, Ⅲ-18, and Ⅲ-19), and control samples from two healthy 26-year-old women. From each of the five participants, three tissue samples were collected, comprising a total of 15 samples for sequencing. Patient samples were collected during gingivectomy, and control samples were collected from healthy, non-inflamed gingival tissue cut during wisdom tooth extraction. Total RNA from gingival tissue was extracted using the TRIZOL method, and sequencing was commissioned to Beijing Tsingke Biotech Corporation with the Illumina HiSeq platform. Following the base calling analysis by CASAVA, the initial data files from high-throughput sequencing were converted into sequence data, resulting in a total of 94.24 gigabytes of Cleandata. The error rate of single-base sequencing for all the samples was < 0.0005, with each sample’s Q30 base percentage being no less than 94.64%. The distribution of the base content per position of reads was stable, with no segregation of AT or GC. Using Hierarchical Indexing for Spliced Alignment of Transcripts (HISAT2), Clean Reads were aligned against the reference genome (Hg38) to obtain gene expression levels.

The significance p-values obtained from hypothesis testing were adjusted using the Benjamini-Hochberg correction method, with the adjusted p-values referred to as False Discovery Rate (FDR). The final criteria for screening differentially expressed genes (DEGs) were set as Fold Change ≥ 2 and FDR < 0.05.

Gene set enrichment analysis (GSEA)

All expressed genes were sorted by Fold change and then subjected to enrichment analysis using the GSEA software provided by the Broad Institute (http://software.broadinstitute.org/gsea/index.jsp). The database used for this analysis was the HALLMARK database within MSigDB.

Results

Clinical manifestation

The proband (III-11), a 36-year-old female, presented to our hospital with the chief complaint of recurrent gingival hypertrophy affecting chewing for more than 30 years.

Upon inquiring about her medical history, it was found that she had undergone multiple gum surgery procedures and had a family history of the condition. The patient’s anterior teeth showed a deep overbite, and the gingiva exhibited extensive and diffuse nodular hyperplasia, involving the entire gingiva and even the mucogingival junction. The gums throughout the entire mouth, especially in the posterior teeth area, showed significant hyperplasia, covering more than 1/3 of the clinical crown. Due to the hyperplastic gums, some teeth were displaced, resulting in misaligned teeth. The gums appeared pink, tough, and exhibited prominent stippling.

Other family members also exhibited varying degrees of gingival hyperplasia, with no reported history of other systemic diseases or use of medications known to cause gingival hyperplasia. The family pedigree is shown in Fig. 1c, and the corresponding oral photographs are shown in Fig. 1a.

Fig. 1.

Clinical characteristics and pathological findings and pedigree. a Intraoral photos of some family members. Individuals I-2, Ⅱ-8, Ⅱ-12, Ⅲ-5, Ⅲ-11, and Ⅳ-10 are affected, with Ⅲ-11 being probands; Ⅱ-14, Ⅲ-9, and Ⅳ-6 are unaffected. b Paraffin sections of gingival tissue from unaffected individuals and patients stained with HE and Masson’s. The upper row shows HE staining, and the lower row shows Masson staining. “Normal” indicates the gingival tissue from a healthy individual, “Ⅲ-11” represents the gingival tissue from the patient in the family. HE staining and Masson’s staining: 10x. Scale bar = 100 μm. c Pedigree. Arrow indicates the proband, black symbols represent affected individuals, and slashes denote deceased family members. Generations are labeled I-IV. The haplotypes are localized to the 2p23.3-p24.1 region, with black indicating the disease-associated haplotype

Histopathological examination

HE staining results revealed that under light microscopy, there was abnormal hyperplasia of the gingival squamous epithelium, with significant elongation of the epithelial pegs into the lamina propria, thickening of the spinous layer, loose connective tissue filled with a large number of dense and hypertrophic collagen fiber bundles, and fewer fibroblasts, which are consistent with the histological characteristics of gingival fibromatosis. Masson staining revealed that collagen fibers appeared blue, and under the microscope, there was a significant enrichment of fibrous components in the connective tissue. Many neatly arranged thick collagen fiber bundles filled the basal and deep parts of the connective tissue, with few blood vessels and cellular components (Fig. 1b).

Mutation analysis

We performed exome sequencing on the whole blood samples of two patients (II-6 and IV-10) and one normal control (III-6) from this family, identifying 262,664 variants. Next, these variants were screened and filtered according to the aforementioned criteria, resulting in 985 variants. A novel missense variant c.14366G > T (p.W4789L) in the SPATA31H1 (NM_032266.5) gene was identified as the candidate molecular cause of the disease in this family through family co-segregation analysis (Fig. 2a). Subsequently, Sanger sequencing was used to verify in all affected individuals and normal controls within the family that this variant met the criteria for family co-segregation with the disease (Fig. 2c). For the SNP genome-wide linkage analysis results, we selected regions with ExLOD values greater than 2.0. We then performed haplotype analysis on these significant regions using Merlin 1.1.2 and imported the haplotype information into HaploPainter 1.043 software for visualization. Ultimately, the candidate region for the pathogenic gene was precisely localized between rs12998454 and rs79391131, within the 2p23.3-p24.1 region. Following this, 25 individuals in the family were genotyped, and the results showed that all patients in the family shared the same haplotype (Supplementary Fig. 1). The variant site identified by exome sequencing, SPATA31H1 (c.14366G > T), was also located within the 2p23.3-p24.1 region (Fig. 2b).

Fig. 2.

Mutation screening and analysis. a Candidate gene screening by exome sequencing. b Chromosomal localization of the SPATA31H1 gene and distribution of six SNP sites. c Sanger sequencing results from normal individuals and patients. All patients in the family have the c.14366G > T variant in the SPATA31H1 gene, with the mutated base indicated in the box. d Analysis of amino acid conservation at the mutation site. e Prediction of protein tertiary structure before and after the mutation using I-TASSER

Functional analysis

In the SPATA31H1 gene, the base at position 14,366 changed from G to T, resulting in the amino acid at position 4789 of the encoded protein changing from tryptophan (Trp) to leucine (Leu). The online prediction tools SIFT and PROVEAN revealed that the missense variant in this gene was harmful, with predicted scores of 0.015 and − 4.16, respectively (Supplementary Table). ESPript3 was used to analyze the conservation of SPATA31H1 p.W4789 amino acid residues in different species. The results revealed that the tryptophan at the mutation site is highly conserved across various animals (Fig. 2d). The ExPASy database was used to compare the physicochemical properties of the proteins before and after the mutation. The analysis results exhibited that the relative molecular mass, isoelectric point, stability coefficient, hydrophilicity, and other physicochemical properties of the MUT protein did not change significantly. However, the extinction coefficient was decreased compared to the WT protein. These results indicate that this gene variant affects the physical and chemical properties of the corresponding protein (Table 1). Besides, I-TASSER revealed that the mutation caused a change in the protein’s tertiary structure (Fig. 2e). The gingival tissues of patients and normal control individuals were collected to understand the effect of the variant on tissue expression. The results demonstrated a marked elevation in both mRNA and protein expression of the gene within the patient’s gingival tissue (Fig. 3).

Table 1.

Analysis of the physicochemical properties of the SPATA31H1 protein before and after mutation

SPATA31H1	WT	MUT
amino acid composition	P.4789 W	P.4789 L
relative molecular mass	224,321	224,248
isoelectric point	10.09	10.09
extinction coefficient	125,140	119,640
instability index	80.89	80.85
half-life (h)	30	30
fat coefficient	64.24	64.44
Grand Average of Hydropathicity	-0.926	-0.924

Fig. 3.

Expression analysis of SPATA31H1 in gingival tissue from 3 patients and 5 healthy volunteers. a Analysis of SPATA31H1 mRNA expression levels in gingival tissue, showing significantly higher mRNA expression in the patient group compared to the normal group (****p < 0.0001). b-c Analysis of SPATA31H1 protein expression levels, with significantly increased protein expression in the patient group (***p < 0.001)

To assess the impact of the variant, the cellular localization of SPATA31H1 was analyzed in HEK293T cells using both WT and MUT vectors. The results indicated that the mutant SPATA31H1 was localized to the cytoplasm and nucleus, similar to the wild-type SPATA31H1 (Supplementary Fig. 2).

Differences in gene expression and GSEA analysis

To reveal the molecular mechanisms associated with gingival fibrosis, we conducted DEG analysis to determine the differences in gene expression between the patient group and the control group in gingival tissues. We used the criteria of an expression change more than twofold (|log2 FC| ≥ 1.0) and FDR < 0.05 to screen for DEGs between patients and controls. Ultimately, 427 DEGs were identified, with 178 being upregulated and 249 being downregulated (Fig. 4a; Supplementary Material 2). Compared with the control group, fibrosis-related genes such as COL1A1, COL10A1, COL11A1, and KCNK3 were significantly upregulated in the patient group (Fig. 4b). General differential analysis usually focuses on a few significantly upregulated or downregulated genes, potentially missing some biologically important genes with insignificant differential expressions. GSEA analysis can be conducted without prior experience to perform enrichment analysis on all genes based on their expression profiles. GSEA does not set a threshold for differential expression, allowing it to detect subtle but consistent trends. We performed GSEA analysis on all genes detected using transcriptome sequencing and screened 33 gene sets using a nominal p-value < 0.05 and FDR q-value < 0.25 as the criteria. Among them, 32 gene sets were downregulated, and one was upregulated in the MUT group (relative to the WT group) (Supplementary Material 3). The most significantly enriched gene set was TNFA_SIGNALING_VIA_NFKB, with an enrichment score of − 0.75284463. The normalized enrichment score was − 2.8238971, indicating that this pathway was significantly inhibited in gingival fibrosis (Supplementary Fig. 3). We used RT-qPCR to examine the expression levels of four upregulated candidate genes (COL1A1, COL10A1, COL11A1, and KCNK3) (Fig. 4c) and four downregulated candidate genes (CCL20, CXCL1, CXCL10, and MYC) (Fig. 4d) and verified the reliability of the gene expression profiles obtained using RNA-Seq and GSEA analyses. The trends shown by RT-qPCR results are essentially consistent with those from RNA-seq, indicating that the molecules related to gingival fibrosis identified in RNA-seq are reliable.

Fig. 4.

Analysis of RNA sequencing results. a Volcano plot of differentially expressed genes. The volcano plot displays all differential genes from the RNASeq database using -log10(P-value) and log2(Fold Change). The horizontal axis represents the logarithmic value of the fold change in expression levels of a particular gene between two samples; the y-axis represents the negative logarithmic value of the statistical significance of gene expression changes. Green dots indicate down-regulated genes, red dots indicate up-regulated genes, and black dots represent genes that are not differentially expressed. b Heatmap of differentially expressed genes. The horizontal axis represents sample names and clustering results, with the first 9 columns for the patient group and the last 6 columns for the healthy control group. The vertical axis represents differential genes and their clustering results. Columns denote samples, and rows denote genes. Colors indicate the log10-transformed expression levels (FPKM + 0.000001) of genes in samples. c RT-qPCR validation of selected differentially expressed genes. d RT-qPCR validation of some genes within this gene set

Discussion

HGF has high genetic heterogeneity. Since Hart et al. identified the SOS1 gene in 2002 in a Brazilian family with nonsyndromic HGF, identifying the associated genes and molecular pathogenesis of HGF has been challenging in this field [10, 15]. Currently, the genes SOS1 and REST [14] are internationally recognized as pathogenic genes associated with nonsyndromic hereditary gingival fibromatosis. Studies on the mechanisms underlying nonsyndromic HGF primarily concentrate on Transforming Growth Factor-Beta (TGF-β) signaling pathway, imbalanced synthesis and degradation of the extracellular matrix (ECM), abnormal ion channels, and oxidative stress. The pathological changes of HGF are mainly characterized by the excessive accumulation of ECM, particularly the increase of COL1A1 and COL1A2 [19]. The general consensus is that the imbalance of ECM synthesis and degradation is pivotal to the development of HGF. Scholars both domestically and internationally have conducted long-term, unremitting research on the regulation of ECM homeostasis. Cytokines such as HSP47, TGF-β1 [20, 21], MMPs (matrix metalloproteinases), and TIMPs (tissue inhibitors of metalloproteinases) [22] have very close connections with the synthesis and degradation of ECM [23]. The truncated protein produced by the mutation of the SOS1 gene can mediate the continuous activation of Ras/mitogen-activated protein kinase signaling pathway, increasing the regulatory factors that control the cell cycle, thereby promoting the cell cycle transition from the G1 phase to the S phase of gingival fibroblasts, ultimately leading to gingival fibroblast proliferation in patients with HGF [24, 25]. The zinc finger protein encoded by the REST gene is a transcriptional repressor. The research results from Kong et al. suggest that truncation mutations in the REST gene may weaken the gene’s own repressive function, thereby upregulating the TGF-β signaling pathway to promote the synthesis of collagen fibers, leading to extensive accumulation of collagen fibers [26]. Additionally, it has been reported that the Ras/MAPK/AP-1 signaling cascade is involved in the pro-fibrotic response of HGF-derived gingival fibroblasts after potassium channel activation [27].

Based on our current experimental evidence, we propose the following mechanistic model by which the SPATA31H1 missense variant leads to increased expression of fibrosis-related genes. Our subcellular localization experiments demonstrated that the missense variant does not alter the localization of SPATA31H1 protein, suggesting that the mutation affects protein function rather than intracellular trafficking. Transcriptome sequencing of patient gingival tissues revealed significant upregulation of fibrosis-associated genes, including COL1A1, COL10A1, COL11A1, and KCNK3. Furthermore, GSEA enrichment analysis identified 33 altered gene sets in patients, with the TNFα-NF-κB signaling pathway being significantly downregulated. Previous studies have established that Tumor Necrosis Factor alpha (TNFα) suppresses COL1A1 expression through NF-κB-dependent mechanisms, thereby limiting excessive extracellular matrix deposition [28]. Therefore, we hypothesize that the SPATA31H1 missense variant results in an altered protein that downregulates the TNFα-NF-κB signaling pathway in gingival tissues, which subsequently relieves the TNFα-mediated transcriptional inhibition of COL1A1 and other extracellular matrix-related genes, ultimately promoting fibrotic responses and leading to the gingival overgrowth characteristic of hereditary gingival fibromatosis. Our preliminary validation of the transcriptomic findings supports this hypothesis. Nevertheless, the precise molecular mechanism remains to be fully elucidated. Since the mutation does not alter the subcellular localization of SPATA31H1, several alternative or complementary mechanisms may contribute to the observed phenotype. First, the missense variant may alter the transcriptional regulatory activity of SPATA31H1, affecting its ability to modulate target gene expression. Second, the mutation may be located at a protein-protein interaction interface, thereby changing the binding affinity between SPATA31H1 and its interacting partners, such as components of the NF-κB signaling cascade or other transcriptional regulators. Third, the mutation may affect post-translational modification states of SPATA31H1, such as phosphorylation or acetylation, which could in turn influence its functional activity without altering its localization. Further mechanistic validation, such as ChIP-seq, rescue experiments, or CRISPR-Cas9 gene editing, is required to definitively establish causality. Unfortunately, such experiments are currently constrained by several factors. First, this study relies on precious clinical samples from a single large Chinese family, and the invasive nature of gingival tissue collection limits repeated sampling. Moreover, the severe fibrosis in the patient’s gingiva has prevented us from successfully culturing primary fibroblasts from the patient’s gingiva for molecular function research. Second, the SPATA31H1 gene contains 5 exons and spans 16,403 base pairs. Its coding sequence is 5,952 base pairs, encoding a protein of 1,984 amino acids with a calculated molecular weight of 224 kDa. Due to the large size of the SPATA31H1 coding region, we have been unable to clone the full-length cDNA for functional studies, which poses a significant technical challenge. These experimental constraints, combined with the fact that research on the SPATA31H1 gene’s function remains nascent, have limited our ability to perform comprehensive mechanistic validation. To date, no studies have detailed the structure and function of the SPATA31H1 protein, nor have any documented its association with gum-related diseases, although preliminary findings suggest its potential significance in non-alcoholic fatty liver disease, prostate cancer, and metabolic syndrome [29–31].

Looking ahead, we will conduct in-depth research in the following directions. First, we will expand the sample size by screening for SPATA31H1 gene mutations in additional unrelated hereditary gingival fibromatosis families and sporadic patients, aiming to validate the generalizability of the association between this gene and HGF, and to clarify its mutation frequency and clinical phenotypic spectrum. Second, we will attempt to construct a full-length SPATA31H1 expression vector using segmented cloning or synthetic biology strategies, and overexpress wild-type and mutant proteins in in vitro cell models. Through NF-κB reporter gene assays, co-immunoprecipitation, and proteomics techniques, we will systematically dissect the regulatory mechanisms by which the mutant protein affects the TNFα-NF-κB signaling pathway and its interaction network. Meanwhile, we will utilize CRISPR-Cas9 gene editing technology to introduce this missense mutation into healthy gingival fibroblasts to observe whether it can recapitulate the transcriptomic characteristics and fibrotic phenotype of patient-derived cells, thereby establishing more direct causal evidence. Furthermore, since mice lack an orthologous gene for SPATA31H1, we plan to explore the construction of alternative animal models to validate the effects of this mutation on gingival tissue development, inflammatory response, and fibrotic progression at the organismal level, and to explore potential therapeutic intervention targets.

Furthermore, we have observed that patients within this family exhibit faster healing following gingivectomy compared to non-HGF patients. The underlying cause of this phenomenon, whether attributed to an increased presence of collagen fibers in the gingival tissue of HGF patients or a consequence of the pathogenic gene variant, remains unknown and necessitates further investigation. The distinction between syndromic and nonsyndromic Hereditary Gingival Fibromatosis is typically based on the presence or absence of systemic diseases. We observed that, although the patients in this family did not exhibit systemic diseases beyond the progressive hyperplasia of gingival tissue, they presented with rough facial skin and a broad nasal bridge. Therefore, we suspected that the variant in this gene may not only affect the gingival tissue but also induce corresponding changes in the skin, hypothesizing that the gene may influence shared molecular mechanisms in the development and growth of tissues such as the gums and skin. With the patients’ consent in subsequent studies, we hope to investigate the skin of these patients. Although the main phenotype of this family is gingival fibromatosis, the rough facial skin and wide nasal bridge features present in some of the affected individuals suggest possible mild ectodermal involvement. These accompanying features have not been reported in other HGF families, and this family lacks other clear syndromic features (such as intellectual disability, growth and development abnormalities, etc.). Therefore, we currently classify this family as a non-syndromic HGF. Regarding this issue, we need to expand the sample size and conduct long-term follow-up in the future to further clarify the relationship between this genotype and the phenotypic spectrum.

Conclusions

In summary, we identified a missense variant (c. 14366G > T, p.W4789L) in the SPATA31H1 gene segregating within a family with 19 cases of Hereditary Gingival Fibromatosis (HGF). To date, no studies have associated SPATA31H1 gene variants with hereditary gingival fibromatosis. In vitro experiments indicated that the levels of mRNA and protein expression of the altered gene were higher than those of the wild type. Bioinformatics analysis predicted that the variant was deleterious and changed the three-dimensional structure of the protein, suggesting the pathogenicity of this missense variant. Transcriptome sequencing revealed that the variant altered the molecules and pathways associated with fibrosis in gingival tissue. We have for the first time linked the variation of the SPATA31H1 gene to the non-syndromic HGF, expanding the mutation spectrum of HGF-related diseases and providing new clues for the study of the pathogenic mechanism of non-syndromic HGF.

Abbreviations

HGF

Hereditary gingival fibromatosis

WES

Whole Exome Sequencing

RT-qPCR

Reverse transcription quantitative polymerase chain reaction

SOS1

Son of Sevenless 1

REST

RE1-silencing transcription factor

ZNF513

Zinc finger protein 513

KIF3C

Kinesin family member 3 C

ZNF862

Zinc finger protein 862

H&E

Hematoxylin and Eosin

ACMG

American College of Medical Genetics and Genomics

SNP

Single nucleotide polymorphism

WT

Wild-type

MUT

Mutant-type

CDS

Coding sequence

HEK

Human embryonic kidney

PBS

Phosphate-buffered saline

DAPI

4’,6-diamidino-2-phenylindole

PMSF

Phenylmethylsulfonyl fluoride

SDS-PAGE

SDS-polyacrylamide gel electrophoresis

PVDF

Polyvinylidene fluoride

TBST

Tris-Buffered Saline-Tween 20

HISAT2

Hierarchical Indexing for Spliced Alignment of Transcripts

FDR

False Discovery Rate

DEGs

Differentially expressed genes

GSEA

Gene set enrichment analysis

Trp

Tryptophan

Leu

Leucine

TGF-β

Transforming Growth Factor-Beta

ECM

Extracellular matrix

MMPs

Matrix metalloproteinases

TIMPs

Tissue inhibitors of metalloproteinases

TNFα

Tumor Necrosis Factor alpha

Authors’ contributions

All authors contributed to the study conception and design. D.C. and F.X. initiated the study. D.C., F.X., J.Z., G.-F.W. and F.H. designed and oversaw the entire project. D.C., W.-Q.W., L.-T.Z., J.-L.X., S.-N.D., Z.-L.Y.,Y.-H.P., G.-T.W. and Z.-H.Y. recruited patients and performed clinical evaluation and sample collection. W.-Q.W. and J.Z. conducted the transcriptome analysis. W.-Q.W., J.-F.C. and Y.-C.Z. performed function analysis. W.-Q.W., D.C. and F.X. prepared the manuscript.

Funding

This work was supported by National Natural Science Foundation of China (82302078, 82170920, 32370649), National Key S&T Special Projects (2022YFC2703303), Guangdong Basic and Applied Basic Research Foundation (2023A1515110129), and Undergraduate Innovation and Entrepreneurship Training Program Project (202212121041).

Data availability

The datasets generated and analysed during the current study are available in the [GSA for human] repository, [https://ngdc.cncb.ac.cn/gsa-human/browse, HRA009216].

Declarations

Ethics approval and consent to participate

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University (2021-KY-0039-004). Informed consent was obtained from all individual participants included in the study.

Consent for publication

The authors affirm that human research participants provided informed consent.

Competing interests

The authors declare no competing interests.