A novel gene ZNF862 causes hereditary gingival fibromatosis.

Hereditary gingival fibromatosis (HGF) is the most common genetic form of gingival fibromatosis which is featured as a localized or generalized overgrowth of gingivae. Currently two genes (SOS1 and REST), as well as four loci (2p22.1, 2p23.3-p22.3, 5q13-q22, and 11p15), have been identified as associated with HGF in a dominant inheritance pattern. Here, we report 13 individuals with autosomal-dominant HGF from a four-generation Chinese family. Whole-exome sequencing followed by further genetic co-segregation analysis was performed for the family members across three generations. A novel heterozygous missense mutation (c.2812G > A) in zinc finger protein 862 gene (ZNF862) was identified, and it is absent among the population as per the Genome Aggregation Database. The functional study supports a biological role of ZNF862 for increasing the profibrotic factors particularly COL1A1 synthesis and hence resulting in HGF. Here, for the first time we identify the physiological role of ZNF862 for the association with the HGF.

Introduction

Gingival fibromatosis (GF) is a rare disorder which is characterized by a benign, nonhemorrhagic, localized, or generalized fibrous enlargement of free and attached gingivae with slow progression. GF could be complicated by epilepsy, hypertrichosis, and mental retardation (Balaji and Balaji, 2017; Snyder, 1965) or it can develop as a part of syndromes like Cowden’s syndrome (Witkop, 1971), Zimmerman-Laband syndrome (Guglielmi et al., 2019), Cross syndrome (Poulopoulos et al., 2011), Rutherford syndrome (Häkkinen and Csiszar, 2007), Ramon syndrome (Suhanya et al., 2010), Jones syndrome (Gita et al., 2014), Costello syndrome (Hennekam, 2003), Ectro-amelia syndrome (Morey and Higgins, 1990), and hyaline fibromatosis syndrome (Hamada et al., 1980); it may also be caused by poor oral hygiene-related gingival inflammation, puberty, pregnancy, or a side effect of the common medications such as calcium channel blockers (nifedipine and verapamil) (Livada and Shiloah, 2012); usually it betrays as an isolated condition as non-syndromic hereditary gingival fibromatosis (HGF) (Jorgenson and Cocker, 1974). HGF is the most common genetic form of GF and usually transmitted as an autosomal-dominant inheritance pattern. Nevertheless, sporadic cases and autosomal-recessive inheritance pedigrees have also been previously described (Majumder et al., 2013).

As reported previously, to date, four loci (2p22.1 [MIM: 135,300], 5q13–q22 [MIM: 605,544], 2p23.3–p22.3 [MIM: 609,955], and 11p15 [MIM: 611,010]) have been identified associated with HGF (Xiao et al., 2001; Ye et al., 2005; Zhu et al., 2007; Xiao et al., 2000); besides a heterozygous frameshift mutation in SOS1 (MIM: 182530) has been identified as causative for the autosomal-dominant HGF in a Brazilian family (Hart et al., 2002); and protein truncating mutations in REST (MIM: 600571) have been reported as the genetic cause of an autosomal-dominant inheritance pattern HGF across three independent families (Bayram et al., 2017). The estimated incidence of HGF is 1 per 175,000 of the population (Ahmed and Ali, 2015), and equal between males and females (Gawron et al., 2016; Odessey et al., 2006). Periodontal surgery including gingivectomy, gingivoplasty, and flap surgery can be applied to patients with HGF due to the aesthetic and functional concern, whereas the recurrence of hyperplasia is relatively high potentially owing to the underlying genetic predisposition (Chaurasia, 2014; Zhou et al., 2011). Therefore, it is important to explore the genetic causes and etiology accounting for HGF with expectations of the precise genetic diagnosis and of the potential gene therapy development, to help individuals who desire to circumvent the sufferings and improve their lives. We report a large Chinese family with 13 affected individuals clinically diagnosed with non-syndromic HGF, and with 12 unaffected family members. Genetic analysis of this large family led to the identification of a heterozygous mutations in a novel gene ZNF862.

Results and discussion

This four-generation pedigree with HGF was shown in Figure 1A. Visual inspection indicated severe or mild generalized GF in all affected members, whereas not in any unaffected member, from this family. No individuals in this family was ever exposed to the medication that may result in GF according to our investigation. No other congenital abnormality has been found for all family members. The clinical data of all individuals in this family are summarized in , and clinical photos of one control and three patients from each generation are available in . From a clinical perspective, the proband (IV-2) had the mild recurrence of hyperplasia during the second year after the periodontal surgery of gingivectomy, gingivoplasty, and flap surgery; while the patient II-2 had an additional clinical finding (hypertension, with no medication of calcium channel blockers treated).

Figure 1.

Pedigree and co-segregation analysis.

(A) The black arrow indicates the proband. The affected individuals are indicated with black-filled boxes in this family. Whole-exome sequencing (WES) was performed for the proband and nine other members (numbers indicated in red color). (B) The photographs of gingival overgrowth are revealed by intraoral examination in affected members compared with in the unaffected member in this family. (C) In this family, all affected individuals harbor the heterozygous variant (c.2812G > A) whereas the unaffected individuals are wild-type. (D) Schematic structure with putative domains of ZNF862 protein and the localization of the novel variant (red arrows). Green rectangles indicate KRAB (krueppel-associated box) domains; yellow rectangles indicate ZnF_TTF (zinc finger in transposases and transcription factors) domains; blue rectangle indicates Dimer_Tnp_hAT (hAT family C-terminal dimerization region) domain.

Table 1.

Clinical Characteristics of the individuals in this family.

NA = Not available.

Members	Identified variant	Age at last exam	Gender	Gingival fibromatosis	Age of onset	Exposure to medication	Gingivectomy/recurrence	Other clinical finding
Ⅱ–1	Wild type	67 years	Male	−	–	No	-/-	No
Ⅱ–2	p.A938T (ZNF862)	70 years	Female	+	NA	No	-/-	Hypertension
II-3	p.A938T (ZNF862)	66 years	Male	+	NA	No	-/-	No
II-4	Wild type	66 years	Female	−	–	No	-/-	No
II-5	Wild type	64 years	Male	−	–	No	-/-	No
II-6	Wild type	64 years	Female	−	–	No	-/-	No
Ⅱ–7	Wild type	62 years	Male	−	–	No	-/-	No
Ⅱ–8	p.A938T (ZNF862)	60 years	Female	+	NA	No	-/-	No
Ⅱ–9	Wild type	54 years	Female	−	–	No	-/-	No
Ⅲ–1	Wild type	49 years	Female	−	–	No	-/-	No
Ⅲ–2	p.A938T (ZNF862)	47 years	Male	+	NA	No	-/-	No
Ⅲ–3	p.A938T (ZNF862)	45 years	Female	+ (mild)	NA	No	-/-	No
Ⅲ–4	p.A938T (ZNF862)	41 years	Male	+	NA	No	-/-	No
Ⅲ–5	Wild type	36 years	Female	−	–	No	-/-	No
Ⅲ–6	p.A938T (ZNF862)	36 years	Male	+	NA	No	-/-	No
III-7	p.A938T (ZNF862)	34 years	Female	+	NA	No	-/-	No
III-8	Wild type	32 years	Female	−	–	No	-/-	No
Ⅲ–9	Wild type	30 years	Male	−	–	No	-/-	No
Ⅳ–1	p.A938T (ZNF862)	26 years	Male	+	6–7 years	No	+/-	No
Ⅳ–2	p.A938T (ZNF862)	22 years	Female	+	6–7 years	No	+/+ (mild)	No
Ⅳ–3	p.A938T (ZNF862)	13 years	Female	+	2–3 years	No	-/-	No
Ⅳ–4	Wild type	11 years	Female	−	–	No	-/-	No
Ⅳ–5	p.A938T (ZNF862)	7 years	Male	+	2–3 years	No	-/-	No

Whole-exome sequencing (WES) was performed for the proband and nine other members (numbers indicated in red color in Figure 1A) in the family according to previously described protocols (Zhang et al., 2015). After analyzing all single nucleotide variants (SNVs) and indels for each gene including the promoter region, exons, splicing sites, introns, and untranslated region (UTR), only three variants ATP7B (c.3403G > A), CDADC1 (c.83–13G > T), ZNF862 (c.2812G > A) were identified to be co-segregated with the phenotype among the 10 family members who underwent WES. To further screen and validate the potential disease-causing variants co-segregation with the phenotype in this pedigree, conventional PCR was performed for sanger sequencing evaluation. Eventually, only ZNF862 (c.2812G > A), no any other variant, was validated to be co-segregated with the phenotype in all 23 alive members in this family, as is shown in Figure 1C.

ZNF862 gene harbors eight exons, encodes a 1169-amino acid protein, and is located on chromosome 7q36.1. It does not escape our notice that ZNF862 does not map to any one of the loci that were reported previously as be associated with HGF. ZNF862 is a predicted intracellular protein which function, to the best of our knowledge, is not yet clearly identified, as a zinc finger protein it may be involved in transcriptional regulation. ZNF862 is expressed ubiquitously across tissues (Fagerberg et al., 2014), it may play various roles under different physiological condition. Schwartz et al., 2019, suggested a plausible role of ZNF862 in matrix-producing metaplastic carcinoma. Peng et al., 2018, reported ZNF862 associated with IgE-mediated type-I hypersensitivity in children. The WES-identified heterozygous variant (p.A938T) in our study located close to and upstream of the Dimer_Tnp_hAT (hAT family C-terminal dimerization region) domain of ZNF862 protein (Figure 1).

The pathophysiological mechanisms underlying HGF remain largely elusive. Nevertheless, overproduction of extracellular matrix, particularly the major component collagen type I (COL1A1), may account for the gingival fibroblast overgrowth phenotype; meanwhile increased synthesis of TIMP-1 seems to be associated with COL1A1 excessive accumulation in HGF gingival fibroblasts (Gawron et al., 2018; Roman-Malo et al., 2019). Besides, Martelli-Junior et al. reported that overexpression of TGF-β1 and IL-6 in gingival fibroblasts play pivotal roles in increasing the synthesis of the COL1A1 along with other specific growth factors (Martelli-Junior et al., 2003). Systematically, transcriptomic analysis of HGF patients and controls performed by Han et al., 2019, demonstrated that regulatory network connection of TGF-β/SMAD signaling pathway and craniofacial development processes contributes to the molecular mechanism of clinical-pathological manifestations.

Aiming to explore the underlying mechanism related to the ZNF862 mutation, RNA sequencing was performed according to previously described protocols (Liang et al., 2019). The primary fibroblasts were isolated from fibromatic gingival specimens obtained from two patients who underwent gingivectomy (IV-1 and IV-2) using standard explant culture as described previously (de Andrade et al., 2001), while control cells were obtained from four independent age- and gender-matched controls who underwent crown lengthening surgery for the restorative purpose. The genes expression profiling and changes in patients over controls were summarized in Supplementary file 1, the samples average counts per million mapped reads (CPM) for each gene is shown, the expression fold-change (FC) and their statistical significance false discovery rate (FDR) are also provided. The expression profiling of attested profibrotic genes (Gao et al., 2019), which may be involved in HGF, was interrogated (Figure 2A). A part of such genes, including COL1A1, TGFB1/2, and SMAD1, was up-regulated in HGF fibroblasts compared with controls; whereas another part was down-regulated, suggesting that the special HGF in our study was attributed to a corresponding part of profibrotic factors, including TGF-β/SMAD1 signaling pathway and COL1A1. Besides RNA sequencing, the in vitro cultured gingival fibroblasts from control 1 underwent adenovirus-mediated delivery of short hairpin RNA (shRNA) targeting ZNF862. The expression of four most significant genes ACTA2, FOS, SMAD1, AGT, as well as COL1A1, under shRNA treatment were investigated (Figure 2B), the knockdown of ZNF862 in control gingival fibroblasts leads to the profibrotic genes up-regulation, reminiscent of the expression in the patients. Here, we speculate that the ZNF862 mutation in patients attenuate the protein function, similar with the ZNF862 shRNA, and promote special profibrotic genes expression to result in the HGF trait. Meanwhile, it did not escape our notice that the causative gene SOS1 was mildly down-regulated, which may be associated with the phenotype (Figure 2 and Figure 2—figure supplement 3). Nevertheless, the proliferation in control gingival fibroblasts was not significantly stimulated by the knockdown of ZNF862 during 5 days’ culture period (Figure 2—figure supplement 2), which may be partially resulting from the too short time course to observe the sophisticated physiological effect.

Figure 2.

RNA sequencing and knockdown analysis.

(A) Heatmap of the 70 attested profibrotic genes expression profiling in RNA sequencing. Rows represent genes and columns represent samples. The heatmap is color-coded based on an average normalization; red represents high expression value and green represents low expression value. (B) Real-time PCR analysis of ACTA2, FOS, SMAD1, AGT, COL1A1, and ZNF862 mRNA abundance in untreated (Un), adenovirus delivered scrambler (Sc), adenovirus delivered ZNF862 short hairpin RNA (shRNA) 1 (S1), and adenovirus delivered ZNF862 shRNA 2 (S2) groups, respectively. Results are given as means ± SEM, dots indicate the relative values of gene expression levels. *p < 0.05, **p < 0.01 in comparison with Sc group. N.S. means not significant. (C) Heatmap of the 100 most up-regulated differentially expressed (DE) genes and 100 most down-regulated DE genes expression profiling in RNA sequencing. The color key is the same with (A). (D) Functional annotation chart of DE genes, the 20 most significantly enriched functions of DE genes were plotted. Functional pathways enrichment tests were based on KEGG database. Enrichment denotes the proportion of DE genes among the indicated pathway.

(A) The transcriptomic comparison of HGF patients and controls yielded 597 differentially expressed (DE) transcripts (red dots) with a flexible cut-off (|log2 fold-change (FC)| ≥ 1.0 and false discovery rate (FDR) < 0.05), of which 355 up-regulated while 242 down-regulated DE transcripts were posited. The FC correlation and statistical significance is shown in the volcano plot. (B) The FC correlation and log2 counts per million mapped reads (CPM) is shown in the MA plot.

CCK8 proliferation assay in untreated control (UN), adenovirus delivered scrambler (SC), adenovirus delivered ZNF862 short hairpin RNA (shRNA) 1 (S1), and adenovirus delivered ZNF862 shRNA 2 (S2) fibroblasts, respectively. Results are given as means ± SEM. *p < 0.05, **p < 0.01 in comparison with control group.

Real-time PCR analysis of SOS1 mRNA abundance in untreated (Un), adenovirus delivered scrambler (Sc), adenovirus delivered ZNF862 short hairpin RNA (shRNA) 1 (S1), and adenovirus delivered ZNF862 shRNA 2 (S2) groups, respectively. Results are given as means ± SEM, dots indicate the relative values of gene expression levels. *p < 0.05, **p < 0.01 in comparison with Sc group. N.S. means not significant.

Figure 2—figure supplement 3.

SOS1 expression in gingival fibroblasts.

Real-time PCR analysis of SOS1 mRNA abundance in untreated (Un), adenovirus delivered scrambler (Sc), adenovirus delivered ZNF862 short hairpin RNA (shRNA) 1 (S1), and adenovirus delivered ZNF862 shRNA 2 (S2) groups, respectively. Results are given as means ± SEM, dots indicate the relative values of gene expression levels. *p < 0.05, **p < 0.01 in comparison with Sc group. N.S. means not significant.

Figure 2—figure supplement 2.

The proliferation of gingival fibroblasts.

CCK8 proliferation assay in untreated control (UN), adenovirus delivered scrambler (SC), adenovirus delivered ZNF862 short hairpin RNA (shRNA) 1 (S1), and adenovirus delivered ZNF862 shRNA 2 (S2) fibroblasts, respectively. Results are given as means ± SEM. *p < 0.05, **p < 0.01 in comparison with control group.

Transcriptome comparison between hereditary gingival fibromatosis (HGF) patients and controls.

To screen all the genes differentially expressed (DE) between patients and controls, we quantified genes whose expression changed over twofold (|log2 FC| ≥ 1.0), and we employed a cut-off of FDR < 0.05 due to the statistical power limitation of the small number of entire samples (N = 6). Finally, the transcriptomic comparison of HGF patients and controls yielded 597 DE transcripts, 355 up-regulated while 242 down-regulated (Figure S1). Of which, the 100 most up-regulated DE genes and 100 most down-regulated DE genes (Supplementary file 2) were shown in Figure 2C, furthermore the top 10 genes were illustrated accordingly. These DE genes probably correlated with the HGF trait. The 20 most significantly enriched functional pathways of all DE genes were shown (Figure 2D). Thereinto, lipoic acid metabolism is the most enriched cluster; notably the MAPK signaling pathway is associated with IL-6 pathway, besides several infection-related clusters in this chart is related with IL-6-induced autophagy pathway.

Taken together we can speculate that ZNF862 may execute as transcriptional repressors since ZNF862 has the zinc-finger DNA-binding domain, meanwhile this missense mutation may attenuate biological function of ZNF862 and hereby enhance the TGF-β/SMAD1 signaling pathway that increases profibrotic factors, particularly COL1A1 accumulation in gingival tissue and results in HGF. Nevertheless, the underlying mechanism of this novel mutation leading to HGF remains enigmatic, with no more specific experiment designed for exploring the biological function of ZNF862. Further studies are expected to expand the mutational spectrum of ZNF862 and the associations with phenotypes, and to investigate the physiological role of ZNF862 in the process of HGF.

In this study, we identify a missense variant (c.2812G > A) in a novel gene ZNF862 that causes autosomal-dominant HGF in a four-generation Chinese family. The functional study supports a biological role of ZNF862 for increasing the profibrotic factors, particularly COL1A1 synthesis in gingiva and hence resulting in HGF. This variant has been absent among the large population according to the report of Exome Aggregation Consortium (ExAC) database, 1000 Genomes, and Genome Aggregation Database. The log of the odds (LOD) score is calculated as 3.6 based on the dominant segregations. According to the combining criteria set guidelines by the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen) (Oza et al., 2018; Richards et al., 2015), the missense variant is rated as ‘Pathogenic’ with the following evidence rules: PP1_Strong (LOD > 3), PS3 (in vitro functional study support), and PM2 (absent from controls). We can speculate that this finding may shed lights on the precise genetic diagnosis and on the potential therapy development, helping individuals continue their lives better.

Materials and methods

Samples collection and WES

The participants from this family were recruited from Nanjing Stomatological Hospital. The genomic DNA was extracted from the peripheral blood. WES was performed for probands and nine other members (numbers indicated in red color in Figure 1A) in the family. Average WES sequencing depth of target region for each sample was higher than 100, and 20× coverage for more than 95% targeted bases was achieved for each sample. Sequencing reads were aligned to the human genome reference sequence (hg19) through the in-house programs. Genome Analysis Toolkit (GATK) was used to call SNVs and indels; and Annotation of Genetic Variants (ANNOVAR) was used for the annotation. A custom Perl script (Source code 1) was used for retrieving the variants corresponding to the inheritance model of this family and that do not appear in unaffected individuals. All the WES data underwent the same quality control filtering and pruning procedures to maximize parity. The following primers are used to amplify products for Sanger sequencing: ZNF862 (NM_001099220.3) forward: 5'-TCATGCGCGGCTTCCACTTTGT; reverse: 5'-AGCACTCGAAATACCTGGCCAG; ATP7B (NM_000053.4) forward: 5'-TCCAGTCGGTAACCTGTTCAC; reverse: 5'-GCACAGCAGAGGCAATCAC; CDADC1 (NM_030911.4) forward: 5'-CGACTAAGGCTGAAGCGTCT; reverse: 5'-AATTGGGGGAAATTATGTGG.

Transcriptome sequencing

The primary fibroblasts were isolated from fibromatic gingival specimens obtained from the patients who underwent gingivectomy (IV-1 as patient 1 and IV-2 as patient 2) using standard explant culture, while control cells were obtained from independent age- and gender-matched controls (two males at 25 and 26 years of age as controls 1 and 2, respectively, and two females at 21 and 23 years of age as controls 3 and 4, respectively) who underwent crown lengthening surgery for the restorative purpose.

Total RNA samples were extracted from the cultured fibroblast cells using Trizol, RNA integrity was assessed using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA). A total amount of 1 μg RNA per sample was used as input material for the RNA sample preparations. Briefly, mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. Fragmentation was carried out using divalent cations under elevated temperature in First Strand Synthesis Reaction Buffer. First strand cDNA was synthesized using random hexamer primer and M-MuLV Reverse Transcriptase (RNase H-). Second strand cDNA synthesis was subsequently performed using DNA Polymerase I and RNase H. Remaining overhangs were converted into blunt ends via exonuclease/polymerase activities. After adenylation of 3’ ends of DNA fragments, adaptor with hairpin loop structure was ligated to prepare for hybridization. In order to select cDNA fragments of preferentially 370–420 bp in length, the library fragments were purified with AMPure XP system (Beckman Coulter, Beverly, MA). Then PCR was performed with Phusion High-Fidelity DNA polymerase, Universal PCR primers, and Index Primer. At last, PCR products were purified (AMPure XP system) and library quality was assessed on the Agilent Bioanalyzer 2100 system. The library preparations were sequenced on an Illumina Novaseq platform and 150 bp paired-end reads were generated.

Raw data were processed through an in-house bioinformatics workflow. Paired-end clean reads were aligned to the reference genome ensembl release-97 (http://asia.ensembl.org/Homo_sapiens/Info/Index) using Hisat2 v2.0.5, featureCounts v1.5.0-p3 was used to count the reads numbers mapped to each gene. And then the CPM and fragments per kilobase of transcript per million mapped reads (FPKM) of each gene was calculated based on the length of the gene and reads count mapped to this gene. At least 45 million clean mapped reads per sample were yielded. The clean bases after filtration for each sample are higher than 6.7 Giga-base, with the effective rate higher than 99.7%, the Q20 bases is higher than 94% for each sample. The GC content for each sample is ranging from 51% to 55%. DE analysis of two groups (patients and controls) was performed using the edgeR R package (3.22.5). The resulting p-values were adjusted using the Benjamini and Hochberg’s approach for controlling the FDR. Corrected p-value of 0.05 and absolute FC of 2 were set as the threshold for significantly DE. We used clusterProfiler R package to test the statistical enrichment of DE genes in KEGG pathways, corrected p-value less than 0.05 was considered significantly enriched. KEGG is a database resource for understanding high-level functions and utilities of the biological system (http://www.genome.jp/kegg/).

ShRNA interference

Besides RNA sequencing, the in vitro cultured fibroblasts (from control 1, control 2, and control 3, respectively) underwent adenovirus-mediated delivery of shRNA targeting ZNF862, the following sequences were used for the ZNF862 shRNA interference experiments, ZNF862 shRNA 1: GCTCTGTTCTGCTCTGCTTGC; ZNF862 shRNA 2: GGATTTACATCCGCTACTTCA; scrambler: GCACCCAGTCCGCCCTGAGCAAA, at 72 hr after the adenovirus-mediated U6 promoter-driven shRNA delivered into the cells (MOI = 70) real-time PCR was performed using 18s rRNA as an internal control. Cells between the third and sixth passages were used for all abovementioned experiments. The following primers are used to amplify products for the real-time PCR: FOS (NM_005252) forward: 5'-GCCGGGGATAGCCTCTCTTACT; reverse: 5'-CCAGGTCCGTGCAGAAGTC; SMAD1 (NM_001003688) forward: 5'-AGAGACTTCTTGGGTGGAAACA; reverse: 5'-ATGGTGACACAGTTACTCGGT; AGT (NM_000029) forward: 5'-GCTGACAGGCTACAGGCAATC; reverse: 5'-TGTGAACACGCCCACCACC; CCL2 (NM_002982) forward: 5'-CAGCCAGATGCAATCAATGCC; reverse: 5'-TGGAATCCTGAACCCACTTCT; COL1A1 (NM_000088.3) forward: 5'-GAGTCTACATGTCTAGGGTCT; reverse: 5'-CACGTCATCGCACAACACCT; ZNF862 (NM_001099220.3) forward: 5'-CTTACTCCAGGAGGAATGGGT; reverse: 5'-CTCCCATGTAGCCCATCTGT; ACTA2 (NM_001613.3) forward: 5'-CCAGACATCAGGGGGTGAT; reverse: 5'-TGGTGCCAGATCTTTTCCAT; FN1 (NM_001365524.2) forward: 5'-CCATAAAGGGCAACCAAGAG; reverse: 5'-AAACCAATTCTTGGAGCAGG; SOS1 (NM_005633.4) forward: 5'- GAAACCCTTTATCTCTCCCAGT; reverse: 5'- CTTGTCAGCACACATTGCCACT; 18srRNA (NR_145822.1) forward: 5'-CGAACGTCTGCCCTATCAACT; reverse: 5'-CAGACTTGCCCTCCAATGGATCCTCGTT.

Proliferation assay

Cell Counting Kit 8 (CCK8, Dojindo Molecular Technologies) proliferation assay was performed with the above-mentioned in vitro cultured fibroblasts. The cells in the logarithmic growth phase were plated overnight in a 96-well plate at an initial density of 2000 per well. Starting with adenovirus infection (MOI = 70), 10 μl CCK8 per well was added to the culture medium on the days 1, 2, 3, 4, and 5. The fluorescent optical density value was measured using a microplate reader. The in vitro cultured fibroblasts underwent adenovirus-mediated delivery of ZNF862 siRNA (NM_001099220.3).

Statistical analysis

All data from the semi-quantitative real-time PCR analyses are representative of at least three independent experiments. Data are presented as the mean ± SEM of at least three independent experiments and differences (in comparison with Sc group) are considered statistically significant at p < 0.05 using Student’s t-test. *p < 0.05, **p < 0.01. N.S. means not significant. All analyses were performed using R v4.0.2.

1000 Genomes, http://www.internationalgenome.org/

ClinVar, https://www.ncbi.nlm.nih.gov/clinvar/

ExAC, http://exac.broadinstitute.org

Genome Aggregation Database, https://gnomad.broadinstitute.org/

GenBank, http://www.ncbi.nlm.nih.gov/genbank/

OMIM, http://www.omim.org/

UniProt, http://www.uniprot.org/

This study is of clinical relevance to those interested in the etiology and pathology of hereditary gingival fibromatosis (HGF). The paper discusses two novel findings: identification of a causative role of a missense mutation in the gene encoding the zinc finger protein 862 (ZNF862) that leads to hereditary gingival fibromatosis (HGF), a rare disease characterized by overgrowth of gingivae, in an examined family, and a suggestion of the molecular consequences of that mutation that leads to the disease.

Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "A Novel Gene ZNF862 Causes Hereditary Gingival Fibromatosis" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Beate Maria Lichtenberger as Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Wafik El-Deiry as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Charles Swanton (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) The study presents novel findings and interesting results despite the small sample size of the experimental cohorts. However, few aspects of data visualization should be improved such as the graphical representation of differentially expressed genes. In addition, since this is a 4-generation HGF family and the sample size is large enough, genome-wide linkage analysis is essential

2) Several conclusions should be backed up by further experimental data or referencing previously published studies or removed:

– The suggestion that top down-regulated and up-regulated genes from RNA-seq lie downstream of ZNF862 and play a role in HGF pathophysiology.

– The conclusion that REST and ZNF862 may execute similar transcriptional functions

3) Please, acknowledge a previous body of work on ZNF862 in the manuscript

4) Data should be deposited so that readers can access them. The provided accession number the data cannot be found or accessed at the China Genebank Nucleotide Sequence Archive.

Reviewer #1 (Recommendations for the authors):

According to the authors the sequencing data set has been deposited in the China Genebank Nucleotide Sequence Archive. However, with the provided accession number the data cannot be found or accessed. The data should be deposited so that readers can access them.

It is not clear whether the data from 3 independent experiments shown in Figure 2B have been done with fibroblasts from the same donor or from different donors. If fibroblasts from only 1 donor were used, the number of biological replicates in figure 2B should be increased.

Could you provide data if shRNA knock-down of ZNF862 affects fibroblast proliferation?

Bulk RNA sequencing in Figure 2A indicates a down-regulation of SOS1 in HGF tissue. Since mutations in SOS1 have been associated with HGF, can you confirm that ZNF862 KD in gingival fibroblasts affects SOS1 expression?

The authors speculate that the top 100 up- or down-regulated genes lie downstream of ZNF862. However, they do not provide any evidence for this conclusion. If you cannot provide any data to support this conclusion, the statement should be down-tuned or removed.

Furthermore, pathway enrichment analysis revealed an association to several infection-related pathways. The authors conclude that this is due to an association with IL-6 signaling. However, differential expression of IL-6 or its downstream targets is not shown in the bulk sequencing, neither in the KD experiments. If the authors cannot provide any evidence, the statement should be removed.

Reviewer #2 (Recommendations for the authors):

Since this is a 4-generation HGF family and the sample size is large enough, genome-wide linkage analysis is essential. If the ZNF862 gene is located within the final linkage region, it would be a very important genetic evidence. Detecting the presence of the c.2812G>A variant in a certain number of normal controls is required.

Reviewer #3 (Recommendations for the authors):

Because of the novelty of the findings the work contributes to advancing the understanding of the genetic components of the disease. We recommend this article for publication in eLife after addressing revisions suggested below:

– Please acknowledge previously published work suggesting a plausible role of ZNF862 in matrix-producing metaplastic carcinoma (Schwartz et al., 2019) and association with IgE-mediated type-I hypersensitivity in children (Peng et al., 2018).

– Line 121 – the statement on ZNF862 ubiquitous expression should include a reference.

– Figure 2D – Please clarify what is meant by "gene number" in the legend of the graph and check if the decimal point is in the correct place. Currently it reads as though enrichment pathway is based on only 1 or 2 genes.

– Figure 2D – Please include two colours in the p-value scale to help the reader identify categories with statistical significance.

– Figure 2D – Please order categories either according to the p-value or to the enrichment for easier interpretation.

– Line 138 – Please specify in the text how many patients and controls were included in the RNA-seq experiment

– Line 175-181 – the statement that REST and ZNF862 might have similar mechanisms of action in causing HGF only because both have zinc finger domains in their structures is too speculative. Without referenced studies on REST downstream gene targets, suggested mechanism, comparison between REST and ZNF862 zinc finger DNA binding domains structures, suggestions of which protein complexes REST and ZNF862 might be part of, or any other experimental data this statement is not justified. This section requires more experimental evidence to support the connection between REST and ZNF862.

– Line 168 – The statement "These DE genes probably lie in the downstream of the ZNF862 and correlated with the HGF trait" is not supported by experimental data in the paper. To make this claim a ChIP-seq showing ZNF862 binding to those target genes, or qPCR validation of the finding in a larger cohort of patients with HGF, or functional studies demonstrating up and down regulations of those genes following ZNF862 knock down/knock out, or discussion of the genes function is required before publication.

– Line 277 – please change shRNA interfering for shRNA interference and include information about shRNA delivery into the cells in this section.

Reviewer #1 (Recommendations for the authors):

According to the authors the sequencing data set has been deposited in the China Genebank Nucleotide Sequence Archive. However, with the provided accession number the data cannot be found or accessed. The data should be deposited so that readers can access them.

The WES and RNA sequencing data supporting this study have been deposited in the China Genebank Nucleotide Sequence Archive (https://db.cngb.org/cnsa, accession number CNP0000995), and the data now are totally publicly accessed. Thanks for the suggestion.

It is not clear whether the data from 3 independent experiments shown in Figure 2B have been done with fibroblasts from the same donor or from different donors. If fibroblasts from only 1 donor were used, the number of biological replicates in figure 2B should be increased.

We used the control cells from three different donors, and we added this information in the ShRNA interference section of the Materials and methods part of the revised manuscript. Thanks for the suggestion.

Could you provide data if shRNA knock-down of ZNF862 affects fibroblast proliferation?

We added this information in the Proliferation assay section in the Materials and methods part of the revised manuscript, and the figure is provided as Supplementary Figure 2—figure supplement 2. Thanks for the suggestion.

Actually, we did not observe a significant change of proliferation between the scrambler and shRNA groups during 5 days cell culture, which may be resulting from the too short time course to observe the sophisticated physiological effect.

Bulk RNA sequencing in Figure 2A indicates a down-regulation of SOS1 in HGF tissue. Since mutations in SOS1 have been associated with HGF, can you confirm that ZNF862 KD in gingival fibroblasts affects SOS1 expression?

We added this information as Supplementary Figure S3 of the revised manuscript. Thanks for the suggestion.

The authors speculate that the top 100 up- or down-regulated genes lie downstream of ZNF862. However, they do not provide any evidence for this conclusion. If you cannot provide any data to support this conclusion, the statement should be down-tuned or removed.

We changed the description “These DE genes probably lie in the downstream of the ZNF862 and correlated with the HGF trait.” into “These DE genes probably correlated with the HGF trait.” in the revised manuscript. Thanks for the suggestion.

Furthermore, pathway enrichment analysis revealed an association to several infection-related pathways. The authors conclude that this is due to an association with IL-6 signaling. However, differential expression of IL-6 or its downstream targets is not shown in the bulk sequencing, neither in the KD experiments. If the authors cannot provide any evidence, the statement should be removed.

We removed this description in the revised manuscript. Thanks for the suggestion.

Reviewer #2 (Recommendations for the authors):

Since this is a 4-generation HGF family and the sample size is large enough, genome-wide linkage analysis is essential. If the ZNF862 gene is located within the final linkage region, it would be a very important genetic evidence. Detecting the presence of the c.2812G>A variant in a certain number of normal controls is required.

We really appreciate your valuable comments. However, we have some difficulties to achieve this. Since we only performed WES on 10 individuals in this family (the red color indicated in Figure 1A), the other members do not agree to provide additional DNA for WES sequencing at this moment, as a result, the genome-wide linkage analysis seems not easily achieved in our current situation.

Meanwhile, we calculated the LOD score of the ZNF862 mutation, LOD=3.6, that may be a statistical significance for a linkage evidence; considering the meiotic recombination may happen freely between haplotypes, this ZNF862 mutation may not be ruled out as the causative mutation in this family even if the ZNF862 mutation be not located in the significant region of genome-wide linkage analysis based on SNP markers. Our opinion is just for your reference. Thanks for the suggestion.

Reviewer #3 (Recommendations for the authors):

Because of the novelty of the findings the work contributes to advancing the understanding of the genetic components of the disease. We recommend this article for publication in eLife after addressing revisions suggested below:

– Please acknowledge previously published work suggesting a plausible role of ZNF862 in matrix-producing metaplastic carcinoma (Schwartz et al., 2019) and association with IgE-mediated type-I hypersensitivity in children (Peng et al., 2018).

We added this information in line 124 of the revised manuscript. Thanks for the suggestion.

– Line 121 – the statement on ZNF862 ubiquitous expression should include a reference.

We added the reference in the revised manuscript. Thanks for the suggestion.

– Figure 2D – Please clarify what is meant by "gene number" in the legend of the graph and check if the decimal point is in the correct place. Currently it reads as though enrichment pathway is based on only 1 or 2 genes.

We carefully revised the Figure 2D, and the gene number indicate the differentially expressed genes among the relevant pathway, we revised the legend of the graph in the Figure 2D. Thanks for the suggestion.

– Figure 2D – Please include two colours in the p-value scale to help the reader identify categories with statistical significance.

We carefully revised the Figure 2D. Thanks for the suggestion.

– Figure 2D – Please order categories either according to the p-value or to the enrichment for easier interpretation.

We carefully revised the Figure 2D. We ordered categories according to the enrichment. Thanks for the suggestion.

– Line 138 – Please specify in the text how many patients and controls were included in the RNA-seq experiment

We added this information in the revised manuscript. Totally two patients (IV-1 and IV-2) and four controls were included in the RNA-seq experiments. Thanks for the suggestion.

Additionally, in the Transcriptome sequencing section of the Materials and methods part of the revised manuscript. The details of the controls were described as following:

“The primary fibroblasts were isolated from fibromatic gingival specimens obtained from the patients who underwent gingivectomy (IV-1 as patient 1 and IV-2 as patient 2) using standard explant culture, while control cells were obtained from independent age- and gender-matched controls (two males at 25 and 26 years old as control 1 and 2 respectively, and two females at 21 and 23 years old as control 3 and 4 respectively) who underwent crown lengthening surgery for the restorative purpose.”

– Line 175-181 – the statement that REST and ZNF862 might have similar mechanisms of action in causing HGF only because both have zinc finger domains in their structures is too speculative. Without referenced studies on REST downstream gene targets, suggested mechanism, comparison between REST and ZNF862 zinc finger DNA binding domains structures, suggestions of which protein complexes REST and ZNF862 might be part of, or any other experimental data this statement is not justified. This section requires more experimental evidence to support the connection between REST and ZNF862.

We removed this REST mechanisms description in line 180-182 of the revised manuscript. Thanks for the suggestion.

– Line 168 – The statement "These DE genes probably lie in the downstream of the ZNF862 and correlated with the HGF trait" is not supported by experimental data in the paper. To make this claim a ChIP-seq showing ZNF862 binding to those target genes, or qPCR validation of the finding in a larger cohort of patients with HGF, or functional studies demonstrating up and down regulations of those genes following ZNF862 knock down/knock out, or discussion of the genes function is required before publication.

We changed the description “These DE genes probably lie in the downstream of the ZNF862 and correlated with the HGF trait.” into “These DE genes probably correlated with the HGF trait.” in the revised manuscript. Thanks for the suggestion.

– Line 277 – please change shRNA interfering for shRNA interference and include information about shRNA delivery into the cells in this section.

We revised the description in this section of the revised manuscript. Thanks for the suggestion.