The present study recruited a Chinese Han family in which two family members (the father and the son) presented with GS in August, 2016. The father was 52-years-old, and the son were 26-years-old. The diagnosis of GS was based on clinical features, family history, cone-beam computed tomography (CBCT), colonoscopy and pathological examinations. Genomic DNA was extracted from peripheral venous blood from the two affected individuals using a QIAamp DNA Blood Mini kit (Qiagen GmbH, Hilden, Germany) following the manufacturer's protocol. Informed consent was obtained from each patient involved in the study, and the study protocol was approved by the Ethics Committee of Guangxi Key Laboratory of Metabolic Diseases Research (Guilin, China).

Genomic DNA was isolated from the peripheral venous blood of the patients using the QIAamp DNA Blood Mini kit (Qiagen GmbH) and then used for exome capture using a NimbleGen SeqCap EZ Human Exome Library v2.0 kit (Roche NimbleGen Inc., Wisconsin, USA) following the manufacturer's protocol. Random DNA fragmentation was performed with a Covaris Ultrasonicator system (Covaris Inc., Woburn, MA, USA), after which the sizes of the library fragments were mainly distributed between 150 and 250 bp. An ‘A’ base was added at the 3′-end of each strand, then ligated to sequencing adapters, which was followed by ligation-mediated polymerase chain reaction (PCR) with probe hybridization, amplification and purification to enrich for targets to sequence. Primers used were included in the NimbleGen SeqCap EZ Human Exome Library v2.0 kit. Each resulting qualified captured library then was sequenced on a BGISEQ-500 sequencing platform (Beijing Genomics Institute, Guangdong, China) and the desired average sequencing coverage for each sample was obtained. Raw image files were processed to produce pair-end reads for each individual using default parameters of base calling software developed for BGISEQ-500.

Clean data was obtained by raw data filtering, and the clean data of each sample was mapped to the human reference genome (GRCh37/HG19) using a Burrows-Wheeler Aligner (BWA V0.7.15) (13,14). All genomic variations, including single-nucleotide polymorphisms (SNPs) and insertions/deletions (InDels) were identified using HaplotypeCaller of GATK (v3.3.0) (Broad Institute, Cambridge, MA, USA) with the proper filtering parameters (15,16). Subsequently, the SnpEff tool (http://snpeff.sourceforge.net/SnpEff\\_manual.html) was used to perform a series of annotations for variants. Potential disease-causing mutations were predicted using the sorting intolerant from tolerant (SIFT) algorithm (17). If the SIFT score was ≤0.05, the present study predicted this variant to be a deleterious variant. Data were filtered with several variant databases, including dbSNP (https://www.ncbi.nlm.nih.gov/projects/SNP/), the 1000 Genomes Project (ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/release), and the NHLBI-ESP6500 database (http://evs.gs.washington.edu/EVS/). Candidate mutations were expected to be absent from these databases. The conservation analysis of amino acid sequences were aligned using ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/).

Sanger sequencing was used to confirm candidate mutations in the MLH1 gene identified by exome sequencing. The PCR primers used were as follows: Forward, 5′-CTTAGTACTGCTCCATTTGGGGA-3′ and reverse, 5′-TTGTTGTATCCCCCTCCAAGC-3′. PCR amplification was performed using a HEMA 9600 PCR thermo cycler (Technological Innovation Beach, Zhu Hai, Guangdong, China) with 35 cycles of denaturation at 98°C for 10 sec, annealing at 55°C for 30 sec and extension at 72°C for 30 sec. The PCR reagents, including Taq polymerase, Taq polymerase buffer, MgCl₂ and dNTP mixture were purchased from Takara Biotechnology, Co., Ltd. (Dalian, China). The PCR products were sequenced using an ABI Prism 3730 DNA analyzer (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA). Each read was compared with the genomic DNA sequence of MLH1, and nucleotide alterations were numbered according to their position in MLH1.

Clinical data was collected for the two patients through patient interviews, medical record extraction, physical examinations and CBCT imaging.

The proband (II:1; Fig. 1) of the GS family, a 52-year-old male patient, presented at the hospital with swelling and pain in the mandibular region, and was diagnosed with GS. From his medical history, it was discovered that the patient had already undergone intestinal polypectomy 3 years ago with recurrence. He had presented with swelling and pain in the mandibular region 2 years previously, and had undergone surgical debridement following anti-inflammatory treatment. A pathological diagnosis of ‘ossified fibromatosis with infection’ was made. The patient continued to present with swelling and discomfort repeatedly, and underwent curative jaw osteomyelitis surgery in the other hospital a year prior to the most recent presentation. Family history revealed that the patient's father had died from gastrointestinal cancer. At the clinical examination, the patient demonstrated no abnormal skin pigmentation, skin mass or swelling in the liver and spleen. On intraoral examination, it was observed that the patient's lower teeth on the right side were missing since surgery, a swelling was present in the right mandibular buccal side, a swelling with pus was present in the mucosal layer, the deciduous teeth 54, 55, 65 and 73 were retained and tooth 34 was unerupted (Fig. 2A). The panoramic radiograph (Fig. 2B) and CBCT revealed that there was an increase in irregular density in the upper and lower jaw regions, and the retained deciduous teeth corresponding to permanent teeth 14, 15, 25, 33 and 34 were affected, and that multiple convex hyperplastic bone lesions were present in the right upper and lower jaw and ethmoidal sinus (Fig. 2C).

The son of the proband (III:1; Fig. 1), a 26-year-old male, requested a clinical assessment due to his father being diagnosed with GS. Medical history revealed that the son had repeatedly abdominal fibrous tumors lasting for 10 years and intestinal polyps lasting for 3 years. At clinical examination, the patient exhibited a semicircular elevated region due to a tough, fixed mass with clear boundaries on the right abdominal skin, ~15 cm in diameter (Fig. 3A). On intraoral examination, it was revealed that deciduous teeth 53, 63, 73 and 83 were retained, whereas teeth 25, 35 and 46 were erupted (Fig. 3B). The panoramic radiograph (Fig. 3C) and CBCT indicated that permanent teeth 13, 23, 25, 33, 35 and 46 were impacted, that supernumerary teeth were not present, that there was an increase in irregular density in the upper and lower jaw regions, and that multiple convex hyperplastic bone lesions were present in the ethmoid sinus (Fig. 3D).

The present study performed whole-exome sequencing on both affected individuals (II:1 and III:1) in the GS Chinese family, for which an average of 31,766.99 Mb raw bases were screened. Following removal of low-quality reads, an average of 31,753.08 Mb clean reads were obtained. The average GC content was 54.26%. Total clean reads per sample were aligned to the human reference genome (GRCh37/HG19) using BWA. On average, 99.05% mapped successfully. The duplicate reads were removed, resulting in an average of 20,694.12 Mb effective reads. The mean sequencing depth of target regions was 232.41-fold. On average, per individual sequenced, 99.86% of targeted bases were covered by at least 1× coverage while 95.09% of the targeted bases had at least 10× coverage (Table I). Overall, 112,819 single nucleotide polymorphisms (SNPs) across both individuals were identified. Of these variants, 96.79% were represented in dbSNP and 94.31% were annotated in the 1000 Genomes Project database. The number of novel SNPs was 2,967. The ratio of transition to transversion was 2.33. Of the total SNPs, 12,736 were synonymous, 11,716 were missense, 37 were stop-loss, 96 were stop-gain, 24 were start-loss and 105 were splice site mutations. In total, 16,829 InDels were detected across all samples. Of these variants, 76.89% were represented in dbSNP and 57.26% were annotated in the 1000 Genomes Project database. The number of novel InDels was 3,422. Of the overall InDels, 344 were frameshift, 4 were stop-loss, 3 were start-loss and 58 were splice site mutations (Table II).

Through read mapping and variant analysis, the presence of previously known mutations in APC and MYH in the two patients was ruled out. It was observed that a novel missense mutation in the MLH1 gene (NM_000249.3:p.Tyr379Ser/c.1136A>C) was present in both affected individuals, and absent in the dbSNP, 1000 Genomes Project, and NHLBI-ESP6500 databases.

To verify the mutation in the MLH1 gene that was identified using whole-exome sequencing, Sanger sequencing was used to examine the MLH1 gene of the two patients (II:1 and III:1). Using method, it was revealed that a missense mutation in the MLH1 gene (NM_000249.3:p.Tyr379Ser/c.1136A>C) was present in both affected individuals (Fig. 4). Therefore, combining the clinical data, whole-exome sequencing data, and the literature on GS, it was suggested that the MLH1 gene mutation may be responsible for GS.

MLH1 amino acid sequences from the species Homo sapiens, Pan troglodytes, Pongo abelii, Mus musculus, Danio rerio, Solanum lycopersicum and Saccharomyces cerevisiae were aligned, and found that p.Tyr379 was located within a highly conserved region of MLH1, and was revealed to be highly conserved among the different species, which indicated the structural and functional importance (Fig. 5).