Ⅰ.INTRODUCTION
White sponge nevus (WSN) is a rare autosomal dominant disorder characterised by benign, painless, white plaques of the mucous membranes. It mainly affects the stratified squamous epithelium of the oral mucosa. This disease was first described by Hyde in 1909, but the term “white sponge nevus” was coined by Cannon in 1935. The buccal mucosa is the most commonly affected site, followed by the tongue, labial mucosa, floor of the mouth, and the palate1). Extraoral sites, including the mucosa of the nose, pharynx, esophagus, vagina, and rectum can also be affected. The symptoms are typically present during childhood or adolescence. Most lesions are painless and do not require treatment. Differential diagnosis of this lesion is important, because some whitish lesions are premalignant or are manifestations of specific systemic disease. Histologically, WSN shows epithelial thickening, parakeratosis, and extensive vacuolization of the suprabasal keratinocytes.
We examined a familial case of WSN in a Korean family and identified the clinical and histological characteristics. We also analysed the keratin gene mutations to determine the familial pattern.
Ⅱ.Methods
1.Description of patients
The proband in this family was 19-year-old female Korean patient (Figs. 1-A and 2-II:3). She and her older sister (23 years, Figs. 1-B and 2-II:1) presented complaining about the esthetic problems associated with white, asymptomatic, corrugated, spongy mucosal plaques with thick, fissured surfaces. The lesions were distributed bilaterally in the buccal mucosa, labial mucosa, gingiva, palate, tongue, and floor of the mouth (Figs. 1-A, B). The patients’ histories revealed that disease onset occurred several years previously. A review of the family history revealed that the proband’s mother (45 years, Figs. 1-C and 2-I:2) had similar oral lesions. The patients were healthy non-smokers with no relevant previous medical history. The patients stated that no other mucosal lesions were present. Incisional biopsies of the buccal mucosa were obtained from the proband (II:3) and her sister (II:1). Histological analysis revealed the presence of hyperparakeratosis and acanthosis with vacuolization in the spinous cell layer. Intracellular edema and perinuclear keratin condensation were present in both cases (Fig. 1-D, E).
2.Genetic analysis
Ethical approval for the genetic studies was obtained from the Review Board at the Yongin Severance Hospital, Yongin, Korea (IRB number: 8-2013-0020). Written consent was obtained from the members of the family. Genomic DNA was extracted from oral mucosal samples from the two patients and from one unaffected family member (Fig. 2-I:1, I:2, II:3). The samples were taken and processed using the swab protocol of the QIAamp® DNA Mini Kit (Qiagen, Hilden, Germany).
The KRT4 and KRT13 genes were amplified using the polymerase chain reaction (PCR) method and primers described in previous studies2). The primers were added to a mixture of Gold ST*R 10X Buffer (Promega, Madison, WI, USA), 1.0 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA, USA), and template DNA. The PCR conditions were: (95°C for 7 min) × 1; (94°C for 1 min, 62°C for 1 min, 72°C for 1 min) × 35; and (72°C for 7 min) × 1. The products were stored at 4°C until direct sequencing (Macrogen, Seoul, Korea).
3.Immunohistochemical staining
Cytokeratin 4 (mouse monoclonal IgG, working dilution: 1/100; abcam, USA) and Cytokeratin 13 (mouse monoclonal IgG, working dilution: 1/100, abcam) were used as primary antibody in this study. Formalin fixed, paraffin embedded 2 normal oral mucosa and 2 white sponge nevus tissue samples were cut into 4uM tissue sections and deparaffinised with xylene. After hydrated with graded ethanol, the endogenous peroxidise activity was blocked with endogenous block solution (Dako, USA). Antigen retrieval was then performed by pressure-cooking for 3min at full pressure in antigen retrieval buffer (Dako). Tissue sections were incubated with primary antibody at room temperature for 1h and were then sequentially incubated with the Real Envision HRP Rabbit/Mouse detection system (Dako) at room temperature for 30 min. The sections were developed with 3,3’-diaminobenzidin chromogen and were then counterstained with haematoxylin.
Ⅲ.RESULTS
Direct sequencing of the PCR products derived from the patients revealed the presence of two new mutations. One was heterozygous missense mutation 344 T to G in exon 1A of the KRT13 gene from the proband (II:3) and her mother (I:2). This mutation is predicted to change codon 115 of the KRT13 coding sequence from leucine to arginine. This mutation was not found in the unaffected father (I:1) (Fig. 3). The other mutation was 1583 A to G, which was located in the non-coding region of KRT4 of the proband (II:3), her mother(I:2) and her father (I:1). Neither the N160del, 153-154insQ mutation in the KRT4-1A region nor the E449K, E520K mutations in the KRT4-2B region were present in any of the three family members.
1.Immunohistochemical staining
Keratin 4 and keratin 13 expression were comparatively investigated in 2 normal oral mucosa and 2 white sponge nevus tissue samples. No difference was found between normal oral mucosa (Fig.4 a) and white sponge nevus tissues (Fig.4 b,c) in expression pattern of keratin 13. Diffuse positive for keratin 13 was found in suprabasal layer of 2 normal oral mucosa and 2 white sponge nevus tissues. By contrast, keratin 4 expression was remarkably decreased in white sponge nevus tissues than normal oral mucosa tissues. Diffuse positive pattern for keratin 4 expression was found in the suprabasal layer of 2 normal oral mucosa tissues (Fig.4 d). In contrast, keratin 4 showed no expression in one white sponge nevus patient (Fig.4 e) and other one showed focal positive reaction for keratin 4 in parakeratotic layer (Fig.4 f).
Ⅳ.DISCUSSION
The KRT4 and KRT13 gene mutations were found to be associated with WSN. Keratins are structural proteins of epithelial cells that can be divided into two subgroups, type 1 and type 2, based on biochemical properties. They polymerize to form hetero-polymeric intermediate filaments (IF). IF are abundant in stratified epithelia, in particular in the suprabasal layers of the epidermis and key components of the cytoskeleton in cells. Keratin polypeptides consist of four central α-helical rod domains (1A, 1B, 2A, and 2B). These domains show a remarkable degree of conservation. Keratin missense mutations lead to a collapse of the IF network that causes cytolysis in response to even minor mechanical stress and induce excessive basal cell proliferation. The suprabasal keratinocytes of oral mucosa express KRT4 and KRT13, and the mutation leads to reactive change manifesting as mucosal hyperkeratosis3-6).
To date, four pathogenic mutations in the KRT4 gene2,5-10,) and seven pathogenic mutations in the KRT13 gene9,11-15) have been identified. To the best of our knowledge, our findings have revealed the presence of a novel KRT13 gene mutation2,5-15). We confirmed the presence of a heterozygous missense mutation 344 T to G in the KRT13-1A region, and the predicted amino acid change was leucine to arginine. Many researchers have reported T to C changes in the same or a similar region of KRT13, which results in the replacement of leucine by proline9,11-13). Our results revealed a T to G change that predicted the amino acid change of leucine to arginine, instead of proline. However, because leucine changed to a different amino acid, it was considered that leucine might have an important role in WSN. Since leucine is the non-polar hydrophobic amino acid, and arginine is a polar hydrophilic alkali amino acid, the changes in polarity could affect the structure of the protein that causes damage to the cytokeratin frame work. A 1583 A to G mutation in non-coding area of KRT4 gene was observed in the proband and her mother, but it seemed that this single nucleotide phenomenon (SNP) was not related to WSN, because the same mutation was observed in the unaffected father. Zhang et al. reported a 2324 A to G mutation in non-coding region of the KRT4 gene, which might be a neutral polymorphism.2) Even when compared to our results, it is possible that this SNP might not be associated with the disease. Shimizu el al.2,5,10) reported that mutation of the KRT4-2B gene might be a characteristic of populations in East Asian countries, but this mutation was not found in the present study.
In contrast to genetic analysis, the patients expressed keratin 13 and loss of keratin 4 immunohistochemically. We can extrapolate that genetic mutation of keratin 13 does not seem to affect its expression, Instead, the loss of keratin 4 may be due to unproven genetic mutation or epigenetic dysregulation. Further study to investigate genetic and epigenetic mechanism of the loss of keratin expression should be required.
