Keratinocytes Modify Fibroblast Metabolism in Hereditary Gingival Fibromatosis

archives of oral biology 53 (2008) 1050–1057

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/arob

Keratinocytes modify ﬁbroblast metabolism in hereditary gingival ﬁbromatosis

Liuyan Meng a, Xiaoqian Ye a, Mingwen Fan a, Xuepeng Xiong a, Johannes W. Von den Hoff b,**, Zhuan Bian a,* a Key Laboratory for Oral Biomedical Engineering Ministry of Education, School and Hospital of Stomatology, Wuhan University, Luoyu Road 237, 430079 Wuhan, PR China b Department of Orthodontics and Oral Biology, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands article info abstract

Article history: Objectives: Hereditary gingival fibromatosis (HGF) is a rare benign disorder characterized by Accepted 15 May 2008 progressive fibrous overgrowth of the gingiva. The proliferation and expression of growth factors of HGF keratinocytes are abnormal. However, the exact role of keratinocytes in HGF Keywords: pathogenesis is still unknown. The present study aimed to clarify the interactions between Hereditary gingival fibromatosis HGF keratinocytes and underlying fibroblasts in the pathogenesis of HGF. Keratinocyte–fibroblast interaction Design: Gingival tissues, fibroblasts and keratinocytes from three Chinese HGF patients and Collagen production three healthy subjects were collected. Histological analyses were performed by histochemical and immunohistochemical staining (Ki-67). Gingival fibroblasts were cocultured with gingival keratinocytes in an in vitro coculture system. The mRNA levels of type I collagen, MMP-1, MMP-3, and TIMP-1 were analysed in the cocultured gingival fibroblasts by reverse- transcription polymerase chain reaction (RT-PCR). The production of type I collagen and TIMP-1 was examined by ELISA. Results: The number of Ki-67-positive keratinocytes in tissue sections from patients was higher than in those from controls. HGF fibroblasts cocultured with HGF keratinocytes showed an increased expression of type I collagen and TIMP-1. Transmission electron microscopy showed increased rough endoplasmic reticulum and ribosomes in cocultured HGF fibroblasts. Conclusions: These results suggest that HGF keratinocytes have an important role in HGF pathogenesis by inducing extracellular matrix (ECM) accumulation by fibroblasts. # 2008 Elsevier Ltd. All rights reserved.

1. Introduction which cause both aesthetic and functional problems. HGF is characterized by excess accumulation of extracellular matrix Hereditary gingival ﬁbromatosis (HGF) is a heterogeneous (ECM). Fibroblasts seem to play a central role in the gingival disorder with a benign, slowly progressive ﬁbrous remodelling of ECM,2 and previous studies have mainly hyperplasia.1 The clinical features of HGF include gingival concentrated on their role in HGF pathogenesis.3–12 These enlargement, malocclusion, and delayed tooth eruption, studies suggest that gingival overgrowth is the result of a

* Corresponding author. Tel.: +86 27 87646337; fax: +86 27 87873260. ** Corresponding author. Tel.: +31 24 3614084; fax: +31 24 3540631. E-mail addresses: [email protected] (J.W. Von den Hoff), [email protected] (Z. Bian). 0003–9969/$ – see front matter # 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2008.05.011 archives of oral biology 53 (2008) 1050–1057 1051

disturbed collagen turnover, or a change in the fibroblast The sections were incubated with a mouse anti-human Ki-67 proliferation. However, little is known about the role of the antibody (Clone: MIB1) (1:100, Dako, CA, USA) at 4 8C overnight. overlying keratinocytes. Normal serum was used as a negative control. Subsequently, Epithelial homeostasis, growth, and differentiation are the standard streptavidin–biotin–peroxidase complex method controlled by epithelial-mesenchyme interactions.13,14 Like- was performed using a SP kit (Zhongshan Goldenbridge wise, the underlying connective tissues may also be controlled Biotechnology, Beijing, China). Reaction products were visua- by these interactions. It has been reported that the expression lized by immersing the sections for 5 min in diaminobenzidine of ECM proteins, and the metabolic activity of fibroblasts can solution. Nuclei were lightly counterstained with hematox- be modulated by oral keratinocytes.15 Keloid keratinocytes ylin. Each step was followed by thorough washes with PBS. promote proliferation and collagen production of the under- Finally, the sections were viewed by light microscopy. lying fibroblasts, and inhibit their apoptosis.16–19 This indi- For the analysis of Ki-67 staining, three fields of a gingival cates a role of the epithelium in the physiology of the section were chosen randomly. Nine sections were used for underlying connective tissue in normal or diseased condi- each subject, so that a total of 27 microscopic fields were tions. Besides excessive ECM accumulation, HGF gingiva analysed in a double-blinded manner by two experimenters in exhibits markedly extended rete pegs in the epithelium.20 each case. The Ki-67 labelling index was determined as the The expression of growth factors and their receptors is percentage of positive cells per 200 cells counted. abnormal in the gingival keratinocytes of HGF and drug- induced gingival overgrowth, which is associated with an 2.3. Isolation and culture of cells increase in their proliferation.21–24 These data suggest that the HGF epithelium plays an important role in the pathogenesis of Six strains of keratinocytes were cultured from the gingival gingival overgrowth. The present study therefore aimed to test tissues of three HGF patients and three controls according to the hypothesis that interactions between HGF keratinocytes the method previously described.16 In brief, the specimens and the underlying fibroblasts contribute to the pathogenesis were cut into pieces, and incubated in 5 mg/ml dispase of HGF. overnight at 4 8C. The epithelium had detached from the connective tissue the next day and was incubated in a solution of 0.25% trypsin, 0.1% glucose, and 0.02% EDTA for 10 min. The 2. Materials and methods cells were suspended in Dulbecco’s modified Eagle medium (DMEM, Gibco, NY, USA) supplemented with 10% foetal bovine 2.1. Tissue collection serum (FBS, Gibco), centrifuged, and seeded in keratinocyte growth medium (KGM, Clonetics, CA, USA). Cell strains were Gingival tissue samples without clinical signs of periodontal stored in liquid nitrogen. Cells from the second passage were inflammation were obtained from three unrelated HGF used for all experiments. patients (two males and one female, aged 16–28 yrs) by The remaining gingival connective tissue from the samples gingivectomy, and from three healthy individuals undergoing were minced, put in culture flasks, and incubated in DMEM routine surgical crown lengthening (two males and one containing 10% FBS and antibiotics. When fibroblasts had female, aged 15–24 yrs). Diagnosis of HGF was based on the grown out of the explants, they were trypsinized (0.2% trypsin) previously described criteria.3,25 In brief, inclusion criteria and passaged. Fibroblasts were used between passages 3 and were: enlarged gingiva covering at least one third of the 6. clinical dental crowns of five or more teeth; no use of inducing The purity of the cells was evaluated by their morphology drugs including nifedipine, phenytoin, and cyclosporine; no and immunohistochemical staining for vimentin and keratin. other systemic diseases. The study was approved by the For immunohistochemical stainings, gingival fibroblasts and Institutional Review Board of Hospital and School of Stoma- keratinocytes were cultured in chamberslides. Cells were fixed tology, Wuhan University. Before sampling, formal written in 2% paraformaldehyde and immersed in 0.3% hydrogen consent was obtained from each subject. According to clinical peroxide. They were incubated with mouse anti-human and family histories, all three patients had the nonsyndromic vimentin (1:50, Zhongshan Goldenbridge Biotechnology) or form of HGF with autosomal dominant inheritance. pan-keratin (1:25, Abcam, Cambridge, UK) at room tempera- ture for 1 h. Subsequently, the standard streptavidin–biotin– 2.2. Histological analysis peroxidase complex method and DAB staining were performed as described above. The percentages of pan-keratin- or Specimens were separated into two pieces. One piece was vimentin-positive cells were over 90% for keratinocytes and fixed in 3% glutaraldehyde for transmission electronic micro- fibroblasts respectively. scopy (TEM), postfixed in 1% osmium tetroxide, and dehy- drated in a graded series of ethanol. Specimens were then 2.4. Coculture of keratinocytes and fibroblasts embedded in epoxy resin. Sections were stained with uranyl acetate and lead citrate and viewed in TEM (H-600, Hitachi, Keratinocytes were seeded at a density of 105 or Japan). 0.5 Â 104 cells/well in 6- or 24-well Transwell membranes The other piece was fixed in 4% paraformaldehyde and respectively (Corning, NY, USA). The cells were cultured in embedded in paraffin. Sections of 5 mm were stained with KGM until 100% confluence. The medium was changed to hematoxylin eosin and for Ki-67. For Ki-67 staining, sections DMEM containing 10% FBS, and the cultures were raised to the were deparaffinized and immersed in 3% hydrogen peroxide. air–liquid interface for another 3 days. Fibroblasts were 1052 archives of oral biology 53 (2008) 1050–1057

seeded in 6- or 24-well plates at a density of 104 or 2.4.3. TEM 0.5 Â 103 cells/well respectively in DMEM containing 10% NFs and GFs were seeded on polyester membranes and then FBS for 1 day. The same number of fibroblasts was seeded cocultured with NKs or GKs as described above. At day 2, in the Transwell membrane as described above for keratino- fibroblasts were fixed on the membranes. TEM was performed cytes, for the purposes of a fibroblast–fibroblast coculture to provide detailed cross-sections of cells on the membrane. control. Both the keratinocytes on the membranes and the For the analysis of TEM, three fields of a gingival section fibroblasts in the plates were washed twice with PBS, and were chosen randomly and at least three cells were analysed cultured separately in DMEM containing 1% FBS for 1 day. in each field. Nine sections were used for each subject, so that Then the membranes and plates were combined for coculture more than 81 cells were analysed in a double-blinded manner in DMEM containing 1% FBS. Four study groups were by two experimenters in each case. composed of normal (NF) or HGF fibroblasts (GF) cocultured with HGF keratinocytes (GK) or normal keratinocytes (NK) 2.5. Statistical analysis (NK/NF,NK/GF,GK/NF,andGK/GF).ControlswereNForGF without keratinocytes, NF with NF, and GF with GF (NF, GF, Data were evaluated by SPSS software (11.0, SPSS Inc., NF/NF, and GF/GF). Chicago, IL, USA). Data were compared by one-way analysis of variance followed by Tukey test. p 0.05 was considered to 2.4.1. Reverse-transcription polymerase chain reaction (RT- be statistically significant. PCR) Total RNA was extracted from fibroblasts after 2 days coculture with keratinocytes according to a standard protocol 3. Results (Total RNA isolation kit, Takara, Tokyo, Japan). RNA concen- tration and purity were determined by a spectrophotometric 3.1. Histology method. Two microgram of total RNA was used as a template for reverse-transcription, and PCR was performed using A well-structured squamous epithelium was present in both previously described primers.3 PCR products were run on a HGF specimens and controls, but the rete ridges were elongated gel. The gels were digitized with a Syngene Gene Genius in HGF (Fig. 1A and D). Under TEM, a basal lamina was observed system (Synoptics, Cambridge, UK). The intensities of the between the epithelium and the connective tissue both in HGF bands corresponding to type I collagen, MMP-1, MMP-3, TIMP- and control specimens (Fig. 1B and E). Hemi-desmosomes and 1, and b-actin mRNA were quantified by image analysis with desmosomes were present in the epithelium of both tissues. Gene Tools software (Synoptics). The relative amounts of There were no Ki-67-positive cells in the connective tissue of mRNA transcripts present in each sample were calculated as both controls and HGF. Ki-67-positive cells were only observed the ratio of the bands to the b-actin band. The experiment was in the basal and suprabasal cell layers of the epithelium in both repeated three times. tissues (Fig. 1C and F). The percentage of Ki-67-positive cells in HGF epithelium was 18.02 Æ 7.11%, which was higher than that 2.4.2. Type I collagen and TIMP-1 protein of controls (10.40 Æ 1.92%) ( p < 0.05). For the determination of TIMP-1 production, culture medium was collected after 2 days coculture of keratinocytes and 3.2. Expression of type I collagen, MMP-1, MMP-3, and fibroblasts, and stored at À20 8C. Fibroblasts were collected for TIMP-1 in cocultured fibroblasts the determination of type I collagen production. In brief, fibroblasts were cultured in DMEM containing 10% FBS and The mRNA expression of type I collagen and TIMP-1 was higher 1 mM ascorbic acid 2-phosphate (Asc 2-P, Sigma, MO, USA) for in GK/GF than in all the other groups ( p < 0.01) (Fig. 2A and D). 5 days. The cultured keratinocytes on the membranes and the There were no significant differences in MMP-1 and MMP-3 gene cultured fibroblasts in the plates were washed twice with PBS expression between the experimental groups (Fig. 2B and C). and cultured separately in DMEM containing 1% FBS for 1 day. Type I collagen protein expression in all GF cocultures was Then the membranes and plates were combined for coculture approximately 2-fold higher than in all the NF cocultures in DMEM containing 1% FBS. After 2 days coculture, the ( p < 0.01) (Fig. 3E). GK/GF cultures produced 279.9 Æ 8.3 ng/ml fibroblasts were collected in 0.5 ml 0.05 M acetic acid (pH 3.0) type I collagen and 181.2 Æ 33.4 ng/ml TIMP-1, which was and stored at À20 8C. significantly higher than that of GK/NF, NK/GF, GF/GF and GF The proteins were measured using a human type I collagen cultures ( p < 0.01) (Fig. 2E and F). There were no significant detection kit (Chondrex, WA, USA) and a human TIMP-1 differences between the other groups. immunoassay kit (R&D Inc., Minneapolis, USA). In brief, 100 ml sample was added to each well, and type I collagen and TIMP-1 3.3. TEM of fibroblasts were detected by specific antibodies. The colour reaction was stopped by the addition of an acid solution and the intensity of In vivo, the nucleus, mitochondria, ribosome and rough the colour was read at 450 nm on a spectrophotometer endoplasmic reticulum (rER) were clearly discernible in both (ELX800, Bio-Tek Instruments, Inc., USA). The protein con- NF and GF. Dilatation of the cisternae of the rER was often centration was determined by interpolation from a standard observed in GF (Fig. 3A and C). These structures were also curve using known concentrations of type I collagen and detected in cultured fibroblasts from NK/NF and GK/GF (Fig. 3B TIMP-1 standards as supplied. The experiment was repeated and D), but the number of these organelles was less than in three times. vivo. Many secretory vesicles were detected at the cell surface archives of oral biology 53 (2008) 1050–1057 1053

Fig. 1 – Morphological analysis of gingival tissues. (A and D) hematoxylin–eosin-stained sections of control (A) and HGF (D) gingival tissues. (B and E) Transmission electron microscopy of gingival tissues showing the ultrastructure of epithelial cells in control (B) and HGF (E). Basal laminae were observed between the epithelial layer and connective tissue. Hemi- desmosomes (white arrow) and desmosomes (black arrow) were present in control and HGF epithelium. (C and F) Ki-67 immunohistochemically stained sections of gingival tissues showing positive cells in control (n = 3) (C) and HGF (n = 3) (F) epithelial layers. The brown staining indicates the mitotic activity of epithelial cells. Scale bar: A and D = 100 mm, B and E=1mm, C and F = 50 mm. N, nucleus. Black arrow, desmosome; white arrow, hemi-desmosome and basal lamina.

of cultured fibroblasts in GK/GF cultures, and cellular 4. Discussion organelles related to protein synthesis such as rER and ribosome were obvious (Fig. 3D). Secretory vesicles were less HGF is a heterogeneous gingival overgrowth condition obvious in fibroblasts in NK/NF cultures (Fig. 4B). Mature characterized by a benign, slowly progressive, nonhaemor- collagenous fibrils were observed around fibroblasts in vivo, rhagic fibrous hyperplasia. The pathology leading to HGF is but not in vitro. poorly understood, but previous studies indicate an increased 1054 archives of oral biology 53 (2008) 1050–1057

Fig. 2 – RNA and protein expression. (A–D) Experiments showing RT-PCR analysis of gingival fibroblasts from coculture. RNA obtained from gingival fibroblasts was reverse-transcribed and amplified with the use of type I collagen (A), MMP-1 (B), MMP-3 (C), TIMP-1 (D) and b-actin specific oligonucleotide primers. The relative amounts of mRNA transcripts were calculated from the ratio of the bands to the b-actin band. (E and F) experiments showing type I collagen (E) and TIMP-1 (F) protein ELISA analysis of gingival fibroblasts in coculture. *Difference between GF (n = 3) and NF (n =3)(p < 0.01). #Difference between GK/GF and NK/GF ( p < 0.01). $Difference between GK/GF and GF/GF or GF ( p < 0.01). &Difference between GK/GF and GK/NF ( p < 0.01). Values represent the means W S.D. from 3 independent experiments with all cell lines.

proliferation rate of both epithelial cells and fibroblasts.5–8,23,24 HGF patients and controls,3 which is consistent with the It has also been reported that elongated rete ridges, increased results of others.4,26 cell proliferation, and increased expression of epidermal Since interstitial collagen accumulation is the main feature growth factor and epidermal growth factor receptor occur of HGF, and the epithelium likely plays a role in the in HGF epithelium.23,24 In our study, sections of HGF tissues pathogenesis of HGF, we evaluated the expression of factors also showed elongated rete ridges, and increased mitotic related to collagen synthesis and degradation in a coculture activity of epithelial cells. These findings suggest that the system. Increased type I collagen and TIMP-1 synthesis in GK/ epithelium plays a role in the pathogenesis of HGF. Our GF culture compared with NF/NF, NK/GF, and GK/NF cultures previous studies demonstrated no significant difference in cell suggests that HGF keratinocytes promote collagen accumula- proliferation between gingival fibroblasts of three Chinese tion by HGF fibroblasts. In addition, our TEM observations archives of oral biology 53 (2008) 1050–1057 1055

Fig. 3 – TEM examination of gingival fibroblasts. (A and C) TEM of gingival tissues showing the ultrastructure of fibroblasts in control (A) and HGF (C) tissues. Abundant cellular organelles were discernible in fibroblasts. Dilatation of the cisternae of rER was observed in HGF fibroblasts (C). Collagenous fibres were observed in control and HGF tissues. (B and D) TEM of gingival fibroblasts showing the ultrastructure of fibroblasts in coculture. Secretory vesicles were detected at the surface of cultured fibroblasts in GK/GF and cellular organelles related to protein synthesis such as rER and ribosomes were obvious (D). Secretory vesicles were less obvious in cultured fibroblasts in NK/NF (B). Scale bar: A–D = 1 mm, Co, collagen fibrils; Mi, mitochondria; N, nucleus; rER, rough endoplasmic reticulum; SV, secretory vesicle.

suggest that HGF fibroblasts cocultured with HGF keratino- normal adherens junctions and desmosomes. This suggests cytes form more secretory vesicles, rER, and ribosomes that epithelial–mesenchymal transition does not occur in compared to controls. HGF. Keratinocytes are known to produce a series of soluble MMPs are a family of Zn-containing proteases that degrade proteins that modulate both their own growth and that ECM proteins. The balance between ECM synthesis and its of other cell types. Increased expression of connective degradation by MMPs, as well as the level of TIMPs, regulates tissue growth factor, keratinocyte growth factor and ECM remodelling. Its disturbance may lead to overgrowth. keratinocyte growth factor receptor has been reported There is ample evidence for a disturbance of MMP and TIMP in fibrotic gingival tissues.21,27 This might stimulate kera- levels in the destructive process of periodontal disease.28,29 In tinocyte or fibroblast proliferation, and the formation of epithelial inflammation and cancer, epithelial cells them- fibrotic tissue. However, these interactions were only selves secrete MMPs and TIMPs, or secrete factors such as investigated in independent cultures of epithelial cells TNF-a and TGF-b to stimulate local fibroblasts to produce and fibroblasts but never in coculture. Some authors MMPs.30–32 In the present study, GF cocultured with GK hypothesized that epithelial–mesenchymal transition might showed similar MMP expression levels as the controls, but be involved in the pathogenesis of HGF.27 In our TEM results, an upregulation of TIMP-1. The lack of normal epithelial the basement membrane in the HGF tissues was always signals might result in excessive ECM accumulation through continuous and intact, and the epithelium possessed the inhibition of MMPs by TIMPs. 1056 archives of oral biology 53 (2008) 1050–1057

It is noteworthy that GF cultured with or without 10. de Andrade CR, Cotrin P, Graner E, Almeida OP, Sauk JJ, keratinocytes had a higher collagen production compared Coletta RD. Transforming growth factor-beta1 autocrine stimulation regulates fibroblast proliferation in hereditary with NF. This indicates that fibroblasts in HGF inherently gingival fibromatosis. J Periodontol 2001;72:1726–33. produce elevated amounts of collagen, which was also 11. Martelli-Junior H, Cotrim P, Graner E, Sauk JJ, Coletta RD. 3,6,11 reported previously. The increased amount of collagen Effect of transforming growth factor-beta1, interleukin-6, is the result of increased synthesis and/or decreased degrada- and interferon-gamma on the expression of type I collagen, tion. Increased collagen synthesis in HGF might result from a heat shock protein 47, matrix metalloproteinase (MMP)-1 higher expression of prolyl 4-hydroxylase, a rate-limiting and MMP-2 by fibroblasts from normal gingiva and enzyme in the synthesis of collagens, and a higher exogenous hereditary gingival fibromatosis. J Periodontol 2003;74:296– 306. or autocrine production of transforming growth factor beta- 3,10,11 12. Lee EJ, Jang SI, Pallos D, Kather J, Hart TC. Characterization 1. Decreased collagen degradation could be induced by a of fibroblasts with Son of Sevenless-1 Mutation. J Dent Res lower expression of MMPs, and/or increased collagen cross- 2006;85:1050–5. linking.9,11,33 13. Mackenzie IC, Binnie WH. Recent advances in oral mucosal It remains to be evaluated, whether all fibroblasts and research. J Oral Pathol 1983;12:389–415. keratinocytes in HGF behave in this way, or only a particular 14. Boukamp P, Breitkreutz D, Stark HJ, Fusenig NE. subpopulation of the cells. Similarly, the reason why the HGF Mesenchyme-mediated and endogenous regulation of growth and differentiation of human skin keratinocytes phenotype only occurs in the gingival tissue is unknown, derived from different body sites. Differentiation 1990;44:150– although the keratinized masticatory mucosa is recognized to 61. 34 be developmentally unique. 15. Hou LT, Kollar EJ, Yaeger JA. Epithelial cell–fibroblast In conclusion, our results strongly suggest that keratino- interactions: modulation of extracellular matrix proteins in cyte–fibroblast interactions contribute to the pathogenesis of cultured oral cells. J Periodontal Res 1993;28:102–14. HGF. 16. Lim IJ, Phan TT, Song C, Tan WT, Longaker MT. Investigation of the influence of keloid-derived keratinocytes on fibroblast growth and proliferation in vitro. Plast Reconstr Surg 2001;107:797–808. Acknowledgements 17. Lim IJ, Phan TT, Bay BH, Qi R, Huynh H, Tan WT, et al. Fibroblasts cocultured with keloid keratinocytes: normal This study was supported by a grant (No. 30670115) from the fibroblasts secrete collagen in a keloidlike manner. Am J National Natural Science Foundation of China, China. Physiol Cell Physiol 2002;283:212–22. 18.PhanTT,LimIJ,BayBH,QiR,HuynhHT,LeeST,etal. Differences in collagen production between normal and references keloid-derived fibroblasts in serum-media co-culture with keloid-derived keratinocytes. JDermatolSci 2002;29:26–34. 19. Funayama E, Chodon T, Oyama A, Sugihara T. Keratinocytes 1. Ha¨kkinen L, Csiszar A. Hereditary gingival fibromatosis: promote proliferation and inhibit apoptosis of the characteristics and novel putative pathogenic mechanisms. underlying fibroblasts: an important role in the J Dent Res 2007;86:25–34. pathogenesis of keloid. J Invest Dermatol 2003;121:1326–31. 2. Bartold PM, Walsh LJ, Narayanan AS. Molecular and cell 20. Sakamoto R, Nitta T, Kamikawa Y, Kono S, Kamikawa Y, biology of the gingiva. Periodontol 2000 2000;24:28–55. Sugihara K, et al. Histochemical, immunohistochemical, 3. Meng L, Huang M, Ye X, Fan M, Bian Z. Increased expression and ultrastructural studies of gingival fibromatosis: a case of collagen prolyl 4-hydroxylases in Chinese patients with report. Med Electron Microsc 2002;35:248–54. hereditary gingival fibromatosis. Arch Oral Biol 2007;52:1209– 21. Das SJ, Parkar MH, Olsen I. Upregulation of keratinocyte 14. growth factor in cyclosporin A-induced gingival overgrowth. 4. Johnson BD, el-Guindy M, Ammons WF, Narayanan AS, Page J Periodontol 2001;72:745–52. RC. A defect in fibroblasts from an unidentified syndrome 22. Das SJ, Newman HN, Olsen I. Keratinocyte growth factor with gingival hyperplasia as the predominant feature. J receptor is up-regulated in cyclosporin A-induced gingival Periodontal Res 1986;21:403–13. hyperplasia. J Dent Res 2002;81:683–7. 5. Shirasuna K, Okura M, Watatami I, Hayashido Y, Saka M, 23. Araujo CS, Graner E, Almeida OP, Sauk JJ, Coletta RD. Matsuya T. Abnormal cellular property of fibroblasts Histomorphometric characteristics and expression of from congenital gingival fibromatosis. J Oral Pathol epidermal growth factor and its receptor by epithelial cells 1988;7:381–5. of normal gingiva and hereditary gingival fibromatosis. J 6. Tipton DA, Howell KJ, Dabbous MK. Increased proliferation, Periodontal Res 2003;38:237–41. collagen, and fibronectin production by hereditary gingival 24. Martelli-Junior H, Lemos DP, Silva CO, Graner E, Coletta RD. fibromatosis fibroblasts. J Periodontol 1997;68:524–30. Hereditary gingival fibromatosis: report of a five-generation 7. Tipton DA, Woodard 3rd ES, Baber MA, Dabbous M. Role of family using cellular proliferation analysis. J Periodontol the c-myc proto-oncogene in the proliferation of hereditary 2005;76:2299–305. gingival fibromatosis fibroblasts. J Periodontol 2004;75:360–9. 25. Ye X, Shi L, Cheng Y, Peng Q, Huang S, Liu J, et al. A novel 8. Coletta RD, Almeida OP, Graner E, Page RC, Bozzo L. locus for autosomal dominant hereditary gingival Differential proliferation of fibroblasts cultured from fibromatosis, GINGF3, maps to chromosome 2p22.3–p23.3. hereditary gingival fibromatosis and normal gingiva. J Clin Genet 2005;68:239–44. Periodontal Res 1998;33:469–75. 26. Saygun I, Ozdemir A, Gunhan O, Aydintug YS, Karslioglu Y. 9. Coletta RD, Almeida OP, Reynolds MA, Sauk JJ. Alteration in Hereditary gingival fibromatosis and expression of Ki-67 expression of MMP-1 and MMP-2 but not TIMP-1 and TIMP-2 antigen: a case report. J Periodontol 2003;74:873–8. in hereditary gingival fibromatosis is mediated by TGF-beta 27. Kantarci A, Black SA, Xydas CE, Murawel P, Uchida Y, 1 autocrine stimulation. J Periodontal Res 1999;34:457–63. Yucekal-Tuncer B, et al. Epithelial and connective tissue cell archives of oral biology 53 (2008) 1050–1057 1057

CTGF/CCN2 expression in gingival fibrosis. J Pathol controlled by distinct components of different extracellular 2006;210:59–66. matrix compartments. J Biol Chem 1997;272:22103–10. 28. Birkedal-Hansen H. Matrix metalloproteinases in human 32. Stuelten CH, DaCosta Byfield S, Arany PR, Karpova TS, periodontal diseases. J Periodontol 1993;64:474–84. Stetler-Stevenson WG, et al. Breast cancer cells induce 29. Nagase H, Woessner Jr JF. Matrix metalloproteinases. J Biol stromal fibroblasts to express MMP-9 via secretion of TNF- Chem 1999;274:21491–4. alpha and TGF-beta. J Cell Sci 2005;118:2143–53. 30. Saarialho-Kere UK, Chang ES, Welgus HG, Parks WC. 33. Pallos D, Hart TC, Cortelli JR, Yamauchi M. Altered post- Distinct localization of collagenase and tissue inhibitor of translational modification of collagen in hereditary gingival metalloproteinases expression in wound healing associated fibromatosis. J Dent Res 1997;76:32. with ulcerative pyogenic granuloma. J Clin Invest 34. Hart TC, Zhang Y, Gorry MC, Hart PS, Cooper M, Marazita 1992;90:1952–7. ML, et al. A mutation in the SOS1 gene causes hereditary 31. Sudbeck BD, Pilcher BK, Welgus HG, Parks WC. Induction gingival fibromatosis type 1. Am J Hum Genet and repression of collagenase-1 by keratinocytes is 2002;70:943–54.