Journal of Dermatological Science 61 (2011) 94–100

Differential expression of matrix metalloproteinases and proteoglycansin Juvenile Hyaline Fibromatosis

Thrasivoulos G. Tzellos a,1, Spyros P. Batzios a,b,1, Alexander Dionyssopoulos c,1,George Karakiulakis a, Eleni Papakonstantinou a,*a 2nd Department of Pharmacology, School of Medicine, Aristotle University of Thessaloniki, Greeceb 1st Department of Pediatrics, Aristotle University of Thessaloniki, Hippokration General Hospital, Thessaloniki, Greecec Department of Plastic Surgery, School of Medicine, Aristotle University of Thessaloniki, Greece

A R T I C L E I N F O

Article history:

Received 8 August 2010

Received in revised form 1 December 2010

Accepted 7 December 2010

Keywords:

Juvenile Hyaline Fibromatosis

Matrix metalloproteinases

Gelatinases

Proteoglycans

Tissue inhibitors of matrix

metalloproteinases

Extracellular matrix

A B S T R A C T

Background: Juvenile Hyaline Fibromatosis (JHF) is a rare autosomal recessive disorder, histologically

characterized by the production and deposition of an unidentified hyaline material in the skin and other

organs. Extracellular matrix molecules are implicated in the development of skin lesion which is

debilitating and recurrent and, so far, no treatment is satisfactory.

Objective: To investigate the expression of matrix metalloproteinases (MMPs), their tissue inhibitors

(TIMPs) and proteoglycans in lesional as compared to site-matched lesion-free skin tissue specimens of a

JHF patient, aiming to elucidate the aetiopathological mechanisms involved in the development of JHF

skin lesions.

Methods: Gelatinase activity of MMP-2 and MMP-9 was investigated by gelatine zymography. Protein

levels of MMP-2, MMP-9, TIMP-1 and TIMP-2 in skin tissue extracts were measured by ELISA. Gene

expression of MMPs, TIMPs and proteoglycans was examined by quantitative RT-PCR.

Results: JHF lesions exhibited significantly higher activity as well as elevated protein and gene

expression of MMP-2 and MMP-9, as compared to lesion-free skin tissue specimens. Decorin was

downregulated and aggrecan was upregulated in lesional skin, as compared to normal skin.

Conclusion: The results presented in this study indicate that MMPs and proteoglycans may be involved

in the pathogenesis of JHF and therefore these molecules may offer alternative targets for

pharmacological intervention to achieve more radical and effective treatment.

� 2010 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights

reserved.

Contents lists available at ScienceDirect

Journal of Dermatological Science

journa l homepage: www.e lsev ier .com/ jds

1. Introduction

Juvenile Hyaline Fibromatosis (JHF) and Infantile SystemicHyalinosis are rare mesenchymal dysplasias inherited in anautosomal recessive manner and histologically characterized bythe production and deposition of an unidentified hyalinematerial in the skin and other organs [1]. Clinical features ofJHF disease include multiple skin lesions, gingival hypertrophy,joint contractures, osteolysis, and osteoporosis [2]. Onset of JHFoccurs within the first few years of life, and most describedpatients have survived into adulthood. The skin lesions mainlyconsist of multiple large tumors, located mostly on the scalp andaround the neck, as well as small pearly, pink papules andplaques on the back, ears, scalp, hands, paranasal, and perianalregions [1].

* Corresponding author. Tel.: +30 2310 999367; fax: +30 2310 999367.

E-mail address: epap@med.auth.gr (E. Papakonstantinou).1 The first three authors have equal contribution to this study.

0923-1811/$36.00 � 2010 Japanese Society for Investigative Dermatology. Published b

doi:10.1016/j.jdermsci.2010.12.002

In JHF, histopathology of skin lesions is diagnostic [3]. Theselesions are characterized by deposition of an amorphous hyaline,eosinophilic substance, in the dermis, where spindle-shaped cellsare embedded. Older lesions are characterized by excessiveextracellular matrix (ECM) deposition, as opposed to recent lesionswhich tend to be more cellular [3]. The amorphous hyalinematerial is mainly composed of glycoproteins and glycosamino-glycans [4], as a result of aberrant synthesis by fibroblasts [4,5] andabnormal collagen metabolism [3,6]. It has been shown thatmutations in the gene encoding capillary morphogenesis protein 2,a type 1 transmembrane protein which shows strong binding tolaminin and collagen IV may be the cause for the above describedabnormalities [7]. Recent studies suggested that mutations inAnthrax Toxin Receptor 2 gene (ANTRX2), abrogate normal cellinteractions with the ECM [1], and that such mutations can confirmthe diagnosis of JHF [8]. However, the exact pathophysiology bywhich mutations in Anthrax Toxin Receptor 2 gene result in thephenotype of systemic hyalinosis is unknown.

It has been postulated that JHF may be a disorder of collagenmetabolism since different authors have suggested a defect in type

y Elsevier Ireland Ltd. All rights reserved.

[()TD$FIG]

Fig. 1. Histological analysis of JHF skin lesions. Area of Juvenile Hyaline Fibromatosis

with characteristic spindle shaped cells deposited in densely hyalinized stroma under

the epidermis. High-power view of fibroblasts associated with extracellular hyaline

material. (Hematoxylin–eosin stain; original magnification: �400).

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–100 95

IV [4,9,10], type VI [11,12] and type III collagen [13,14]. However,there are no consistent defects of collagen, and many affectedindividuals have normal collagen profiles.

Matrix metalloproteinases (MMPs) are zinc-dependent endo-peptidases classified according to domain structure into collage-nases, gelatinases, stromelysins, matrilysines, membrane-type andothers [15]. They represent key enzymes involved in thedissolution of ECM [16], and have been implicated in variousnormal and pathological processes, usually related to inflamma-tion and cell apoptosis [17,18]. Most MMPs are secreted aszymogens and require proteolytic activation, whereas theirtranscription, translation and proenzyme activity is regulated bygrowth factors, cytokines and tissue inhibitors of metalloprotei-nases (TIMPs) [19,20]. Gelatinase A (MMP-2) and gelatinase B(MMP-9) digest the denatured collagens, gelatines. MMP-2, but notMMP-9, digests type I, II, and III collagens.

Main components of ECM are the proteoglycans, which arecomposed of a core protein bearing one or more glycosaminogly-can chains [21]. Proteoglycans are implicated in a number ofbiological activities like assembly of ECM components, dermalhydration, cell growth, haemostasis, wound healing, as well as incollagen and elastic fiber formation [22].

In the present study, we examined the enzyme activity, as wellas, protein and gene expression of MMP-2 and MMP-9, in JHF skinlesions from various body regions, as compared to macroscopicallynormal skin obtained from matched sites of the same patient.Furthermore, we examined gene expression of certain importantproteoglycans such as biglycan, decorin, aggrecan and perlecan.

We demonstrate that the activity, as well as gene and proteinexpression of MMP-2 and MMP-9, are significantly increased inlesional as compared to normal skin tissue specimens. In addition,skin lesions are associated with upregulation of aggrecan anddownregulation of decorin gene expression. Our results providenew insights into the role of ECM components in JHF.

2. Materials and methods

2.1. Brief case report

A 9-year-old Greek female patient presented to us with JHFdiagnosed 7 years ago. She was born to unrelated parents after anormal pregnancy and caesarean delivery with a birthweight of2500 g. Her APGAR score was 5 out of 10. Although normal at birth,at 13 months of age she began to develop stiffness and progressivelimitation of motion mainly of the knees and elbows. Concurrently,a progressive, popular eruption developed on her face, ears,dorsum of the neck, and perianal region. Nodules began appearingand rapidly enlarged.

At the time of presentation, she had multiple, pearly, groupedpapules on the dorsum of the neck and behind the ears. There werealso several fleshy nodules in the perianal region and multiple,nontender, firm, large nodules on her ears, scalp, lumbosacral areaand left elbow. Similar nodules were also present on the upper andlower distal extremities, particularly the tips of the digits. Skin wasthickened over the joints, marked gingival hypertrophy affectingthe upper and lower dentition was present and flexuralcontractures of knee and hip joints resulted in severe immobility.Histological examination revealed hyaline depositions and scat-tered cells in densely hyalinized stroma under the epidermis(Fig. 1). All biochemical tests were within normal limits. Thepatient was of normal intelligence.

2.2. Skin tissue specimens

Eight excisional biopsies from papulo-nodular skin lesions fromscalp (n = 2), left elbow (n = 2), left ear lobe (n = 2) and lumbosacral

area (n = 2) were obtained at the Department of Plastic Surgery ofAristotle University. Four punch biopsies (4 mm) from macroscop-ically normal skin from matched sites were also obtained ascontrols. Parents gave written consent for the participation of theirchild in the study, which conformed fully to the Declaration ofHelsinki for biomedical research on human subjects.

2.3. Gelatin zymography

The gelatinolytic activity of MMPs was determined by gelatinzymography analysis using sodium dodecyl sulfate-polyacryl-amide gel electrophoresis (SDS-PAGE) under denaturing but nonreducing conditions [23,24]. In brief, skin specimens wereprocessed accordingly to contain the same amount of protein(2 mg) and applied on an 8% SDS/polyacrylamide gel containing0.1% gelatin (25 mA, 2 h, at room temperature). Gels were thenequilibrated in 2.5% Triton X-100 buffer for 1 h and subsequentlyincubated in 50 mM Tris–HCI, pH 7.3 buffer containing 200 mMNaCI, 5 mM CaCl2 and 0.1% Triton X-100 for 18 h, at 37 8C. Bands ofenzymatic activity were visualized by negative staining withstandard Coomassie brilliant blue R-250 dye solution. Molecularsize of bands displaying enzymatic activity were estimated bycomparison to purified proMMP-2 (72.0 kDa), active MMP-2(64.0 kDa), proMMP-9 (92.0 kDa) and active MMP-9 (78.0 kDa)(Anawa Trading, Wangen). Prestained standard protein molecularweight markers used were: myosin (250 kDa), phosphorylase(148 kDa), bovine serum albumin (98 kDa), L-glutamic dehydro-genase (64 kDa), alcohol dehydrogenase (50 kDa), carbonicanhydrase (36 kDa), myoglobin red (22 kDa), lysozyme(16 kDa), aprotinin (6 kDa) and insulin, B chain (4 kDa) (SeeBluePlus2 Prestained, Invitrogen, USA). Gelatinolytic activity wasquantified using a computer-assisted image analysis program (1DImage Analysis Software, Kodak Digital Science v.3.0, EastmanKodak, Rochester, NY, USA). All experiments were performed induplicate.

2.4. Determination of MMP-2, MMP-9, TIMP-1 and TIMP-2

MMP-2, MMP-9, TIMP-1 and TIMP-2 were measured inaliquots of tissue extracts, corresponding to 4 mg of protein,using ELISA (R&D Systems Europe, Abingdon, UK), which wasperformed according to manufacturer’s instructions. The MMP-2and MMP-9 assays measure total MMP-2 and MMP-9 (proen-zymes and activated forms). All experiments were performed induplicates.

[()TD$FIG]

Fig. 2. Gelatin zymography analysis. (A) Representative gelatin zymography of

aliquots (2 mg of protein) from homogenized JHF skin lesions and normal appearing

skin obtained from matched sites. Bands of enzymatic activity were visualized by

negative staining with standard Coomassie brilliant blue dye solution. (B)

Quantitative analysis of gelatinase activity of JHF lesional (n = 8) and normal

(n = 4) skin tissue specimens. Values represent mean � SD. Statistical significance:**0.001 < p < 0.01, ***p < 0.001, as compared to normal skin.

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–10096

2.5. Protein determination

Protein content was determined in aliquots of tissue extractswith the standard Bradford assay (Bio-Rad, Glattbrugg, Switzerland)[25], using bovine serum albumin (Sigma) as standard.

2.6. Isolation of RNA

Skin tissue specimens from both lesional and normal regionswere disrupted and homogenized and total RNA was isolated by theuse of RNeasy Protect Mini Kit (Qiagen, Basel, Switzerland) followingmanufacturer’s instructions. The concentration and purity of RNAwere determined by measuring the absorbance at 260 and 280 nm.

2.7. Quantitative RT-PCR analysis

The cDNA was synthesized by using 1 mg of total RNA in thepresence of random primers, dNTPs, RNase inhibitor and reversetranscriptase (M-MLV Reverse Transcriptase, Promega, Madison,WI, USA), according to manufacturer’ s protocol. The PCR wasperformed in a 25 ml reaction volume, while for the analysis of geneexpression SYBR GreenER qPCR SuperMix Universal (InvitrogenGmbH, Life Technologies, Karlsruhe, Germany) was employed. Thereaction mixtures were amplified for 40 cycles at an annealingtemperature of 60 8C using a Step One Plus Real-Time PCR Systemsinstrument (Applied Biosystems). The primers used were designedusing the Primer ExpressTM software (Version 2.0, AppliedBiosystems) (supplementary Table). The primer concentrationoptimization and the absence of nonspecific products wereconfirmed by performing dissociation curve analysis, which resultedin single products at specific melting temperatures. The relativeexpression of mRNA for the target genes was performed by thecomparative Ct (threshold cycle), DDCt method, by using GAPDH(Glyceraldehyde-3-phosphate dehydrogenase) as an endogenouscontrol for the relative quantification for the target messages [26].The normalized Ct was obtained by subtraction of the Ct value forGAPDH from the Ct value for the gene of interest (DCt). The differencebetween the DCt values for lesional and normal skin tissuespecimens gave the DDCt value, which was used for the calculationof the relative mRNA expression using the formula 2�

DDCt. Therelative mRNA levels were expressed as percentage change in JHFlesions over normal skin obtained from matched sites.

2.8. Statistical analysis

Normal distribution was checked using Kolmogorov Smirnovfor all dependent variables. Paired t-test was used for variablesfollowing normal distribution, whereas Wilcoxon matched pairssigned-rank test was used for variables violating the assumption ofnormality. Two-tailed levels of significance were used in allstatistical calculations. All data are expressed as mean � SD.Difference was considered to be statistically significant at a level ofp < 0.05. The computer software SPSS 15.0 (SPSS Inc., Chicago, IL,USA) was used for all statistical calculations and analyses. Graphswere produced with MS Excel 2000 (Microsoft Corporation, USA).

3. Results

3.1. Increased gelatinase activity in JHF

Gelatin zymography analysis of tissue extracts correspondingto 2 mg of protein, revealed that skin tissue specimens from bothnormal and lesional regions express gelatinase activity of variablemolecular mass (Fig. 2A). The two gelatine lysis bands with thelower molecular mass co-migrated as purified proMMP-2 (72 kDa)and MMP-2 (64 kDa) and were detected in lesional as well as in

normal skin tissue specimens (Fig. 2A). Quantification of gelati-nolytic activity using a computer-assisted image analysis pro-gramme showed a statistically significant increase in the activity ofproMMP-2 and MMP-2 in lesional as compared to normal skin(p = 0.007 and 0.006 respectively) (Fig. 2B).

The twogelatine lysis bands ofhigher molecular mass correspondto proMMP-9 (92 kDa) and to activated MMP-9 (78 kDa) andwere detected only in lesional skin tissue specimens (Fig. 2A).

3.2. Increased protein expression of MMPs in JHF

The concentration of MMP-2, MMP-9, TIMP-1 and TIMP-2 wasassessed in tissue extracts using ELISA. As shown in Fig. 3A, MMP-2was significantly higher in lesional (118.28 � 24.18 pg/mg protein),as compared to normal skin tissue specimens (6.92 � 2.8 pg/mgprotein), (p = 0.007). The same results were also obtained for MMP-9(27.98 � 7.02 and 0.63 � 0.4 pg/mg protein, for lesional and normalskin tissue specimens respectively), (p = 0.007), (Fig. 3B).

No significant changes were observed in the concentration ofTIMP-1 (p = 0.174) and TIMP-2 (p = 1.000) between lesional andnormal skin tissue specimens (Fig. 3C and D).

3.3. Increased gene expression of MMPs in JHF

Gene expression of MMP-2, MMP-9, TIMP-1 and TIMP-2 wasinvestigated by quantitative Real-Time RT-PCR analysis usingglyceraldehyde-3-phosphate dehydrogenase (GAPDH), as a refer-ence gene. All four genes are expressed in JHF skin lesions, as wellas in normal skin tissue specimens. However, gene expression ofMMP-2, MMP-9, TIMP-1 and TIMP-2 was significantly upregulatedin lesional, as compared to normal skin tissue specimens (Fig. 4).Gene expression of MMP-2 was increased by 12 fold (p = 0.006)(Fig. 4A), MMP-9 showed a 309-fold increase (p = 0.000) (Fig. 4B),TIMP-1 upregulated by 1.5 fold (p = 0.006) (Fig. 4C) and TIMP-2 by85-fold (p = 0.000) (Fig. 4D).

3.4. Differential gene expression of proteoglycans in JHF

Gene expression of biglycan, decorin, perlecan and aggrecanwas investigated by quantitative Real-Time RT-PCR analysis, using

[()TD$FIG]

Fig. 3. Concentration of gelatinases and their tissue inhibitors in JHF skin tissue specimens. Aliquots (4 mg of protein) from skin tissue extracts of JHF lesional (n = 8) and

normal (n = 4) skin obtained from matched sites, were assessed for their concentration of MMP-2 (A), MMP-9 (B), TIMP-1 (C) and TIMP-2 (D) by ELISA. Values represent

mean � SD. Statistical significance: **0.001 < p < 0.01, as compared to normal skin.

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–100 97

glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a refer-ence gene. All four genes are expressed in JHF skin lesions, as wellas in normal skin tissue specimens.

Gene expression of biglycan, which is a small leucine-rich proteoglycan and of perlecan, which is the most abundantproteoglycan in the basement membranes, was not significantlyaltered in lesional, as compared to normal skin tissuespecimens (p = 0.15 and p = 0.066 respectively) (Fig. 5A and B,Table 1).

On the other hand, gene expression of aggrecan, which is a largechondroitin and keratan sulfate proteoglycan, was significantlyupregulated in lesional, as compared to normal skin tissuespecimens (p = 0.005) (Fig. 5C, Table 1).

[()TD$FIG]

Fig. 4. Gene expression of gelatinases and their tissue inhibitors in JHF skin tissue specime

from matched sites, were assessed for gene expression of MMP-2 (A), MMP-9 (B), T

percentage � SD of the value of normal skin tissue specimens which was set to 100%. Sta

Gene expression of the small proteoglycan decorin wassignificantly downregulated in JHF lesional skin, as compared tonormal skin tissue specimens (p = 0.001) (Fig. 5D, Table 1).

4. Discussion

JHF is a connective tissue disorder, and abnormal collagenmetabolism represents a distinct feature of its skin lesions.Numerous studies have shown the existence of defects in differenttypes of collagen [4,9–14], and it has been suggested that thisabnormal formation of collagen fibers is indicative of an underlyingdefective turnover of glycosaminoglycans [4]. We have previouslyshown that JHF skin lesions are characterized by altered expression

ns. Skin tissue specimens from JHF lesional (n = 8) and normal (n = 4) skin, obtained

IMP-1 (C) and TIMP-2 (D) by quantitative RT-PCR. Results are presented as the

tistical significance: **0.001 < p < 0.01, ***p < 0.001.

[()TD$FIG]

Fig. 5. Gene expression of proteoglycans in JHF skin tissue specimens. Skin tissue specimens from JHF lesional (n = 8) and normal (n = 4) skin, obtained from matched sites,

were assessed for gene expression of Biglycan (A), Perlecan (B), Decorin (C) and Aggrecan (D) by quantitative RT-PCR. Results are presented as the percentage � SD of the value

of normal skin tissue specimens which was set to 100%. Statistical significance: *0.01 < p < 0.05; **0.001 < p < 0.01.

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–10098

of glycosaminoglycans, especially of hyaluronic acid, as compared tonormal-appearing skin [27]. Since hyaluronic acid is involved in theaggregation process of collagen fibers it may be postulated that JHFrepresent a disease where several components of the ECM might bederegulated due to abnormal cell–matrix and cell–cell interactions.

The ECM is composed of 50 different gene products, but fibrilsof collagen I and III are almost completely responsible for thecohesion and structural stability of interstitial connective tissue.The degradation of ECM macromolecules is primarily regulated byMMPs, a family of zinc-dependent proteases, whereas TIMPs arethe molecules responsible for the regulation of the activity ofMMPs and the blockage of matrix degradation [19,20].

In this study, we investigated the involvement of gelatinasesand TIMPs in lesional as compared to macroscopically normal-appearing skin tissue specimens obtained from matched sites.Healthy skin tissue samples of relative skin areas (scalp, elbow, earlobe, lumbosacral area) from age and sex matched controls, wouldoffer additional information, provided that there are no ethicalissues raised. However, the use of site-matched, lesion-free JHFskin, serve the main objective of this study which was to elucidatethe aetiopathological mechanisms involved in the development ofJHF skin lesions, in an effort to identify novel pharmacologicaltargets. In this sense, comparing JHF lesional and non-lesional skinis appropriate and provides important information regarding theECM biology of skin lesions of JHF patients. Furthermore, thedifferential expression of the molecules studied between affectedand normal skin in JHF was consistently observed in each of thesamples obtained from different anatomical areas, indicating thatthe same changes in ECM molecules are associated to lesions,irrespective the anatomical site.

Table 1Gene expression of proteoglycans in skin tissue specimens (n = 8) from a patient

with Juvenile Hyaline Fibromatosis (JHF).

Proteoglycan Gene expression (% of normal) p-value

Perlecan 108.0�3 0.066

Decorin 63.4�2 0.001*

Aggrecan 158.0�8 0.005*

Biglycan 110.0�9 0.150

Results are presented as mean� standard deviation, * statistical significance.

To our knowledge, this is the first attempt to comprehensivelyreport differences in enzyme activity and expression of thesemolecules in JHF skin specimens, using a variety of methods, suchas gelatin zymography, ELISA, and quantitative RT-PCR. We reportsignificant alteration in enzyme activity of both MMP-2 and MMP-9, with normal skin exhibiting no MMP-9 activity. This is inagreement with previous studies in human fibroblasts fromnormal skin where MMP-9 was not present [28]. Increasedgelatinolytic activity plays a central role in the homeostasis ofECM and processes such as angiogenesis [29]. Furthermore, ELISAanalysis revealed significant increase in the content of gelatinases,whereas TIMPs did not show alterations between lesional and non-lesional JHF skin tissue. In addition, we demonstrate significantalterations in gene expression of both gelatinases and TIMPs. Thehighest increase (309-fold) in gelatinase levels in lesional skinspecimens was observed for MMP-9, followed by a significant butyet lower increase of TIMP-2 expression (85-fold). TIMP-1 andTIMP-2 have been shown to possess cell growth promoting activity[30]. Thus, we can conclude that the above constellation of findingsis indicative of a collagen-degrading microenvironment, where aprimarily increase in gelatinase expression is followed by asecondary upregulation of TIMP-1 and TIMP-2 genes. However, inJHF lesions, the increased gene expression of TIMP-1 and TIMP-2 isnot translated to increased protein formation, indicating that thebalance is moved towards increased gelatinase activity andenhanced ECM degradation.

The apparent divergence of increased expression of gelatinasesin lesional skin tissue specimens may be explained by the fact thatJHF skin lesions are characterized by abnormal collagen metabo-lism (probably due to increased expression of gelatinases) and inaddition, by deposition of an amorphous hyaline, eosinophilicsubstance, in the dermis. The amorphous hyaline eosinophilicmaterial is mainly composed of glycoproteins and glycosamino-glycans, as a result of aberrant synthesis by fibroblasts [3,6]. In thisrespect, many other skin lesions that have been characterizedclinically and histologically by overgrowth mechanisms are alsoassociated with increased MMP-2 and MMP-9 expression.Increased expression of gelatinases has also been reported ineosinophilic granuloma, where a highly significant relationshipbetween the expression of MMP-9 and macrophage count with

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–100 99

recurrent and more aggressive lesions was demonstrated [31].Similarly, it is well documented that non-melanoma skin cancers,like basal cell carcinoma, exhibit an increased expression of MMP-2 and MMP-9, that has been correlated with depth of lesion,inflammation and microvessel density [32]. Most interestingly,keloids, skin lesions that are highly characterized by overgrowthmechanisms, are associated with activation of MMP-2 and MMP-9[33,34].

The above mentioned data suggest that MMP-2 and MMP-9play a significant role in the development of skin lesionscharacterized by overgrowth mechanisms, which may be attrib-uted to their effect in the remodeling of the ECM.

MMPs and TIMPs have not been sufficiently investigated inpatients with JHF, with the exception of two studies. Senzaki et al.examined skin tissue specimens immunohistochemically andfound that MMP-2 was present both in spindle-shaped cells andhyaline matrix, whereas MMP-9 expression was seen only in somespindle-shaped cells [35]. They have also shown a universalexpression of TIMP-2, while on the other hand TIMP-1 wasnegative. In agreement to our findings, Hakki et al. [36], havedemonstrated a higher TIMP-2 expression in JHF gingivalfibroblasts, as compared to normal gingival tissue. In contrast,these authors reported that there were no significant differencesregarding the expression of MMP-2 and TIMP-1 and there wasdecreased expression of MMP-1 and MMP-3, concluding thatgingival overgrowth that is observed in JHF patients may beexplained by a decreased level of MMPs and increased inhibition ofMMPs by TIMPs. However, these results were obtained usingfibroblasts from normal and JHF gingival tissue, which is mainlycomposed of collagen I, and is different from skin. Furthermore, atthe molecular level, there are baseline differences regardingproduction of fibroblasts due to anatomically dependent diversityand that genes expressed by fibroblasts demonstrate site-specificdistinct programmes for ECM synthesis [37,38].

In this study, we also present data indicating that JHF lesions arecharacterized by altered expression of proteoglycans compared tomacroscopically normal skin obtained from matched sites.Biglycan and decorin are small leucine-rich proteoglycans.Biglycan is a proteoglycan with two dermatan/chondroitin sulfatechains. Although it is present in low amount in normal human skin,it is required for endothelial cell migration. In our study biglycanremained unaltered in JHF lesions as compared to normal skintissue specimens. Decorin is the main proteoglycan in the skin withone chondroitin sulfate chain, which has been shown to bind tocollagen fibers and regulate the collagen matrix assembly [39].Thus, decorin seems to play a crucial role in collagen fibberformation and remodeling. In JHF, fibroblasts have shown evidenceof defective synthesis of collagen, which was deposited asfibrillogranular material in the matrix [3]. Our results showsignificant decrease of decorin expression in JHF, thus indicatingthat decorin could be responsible for collagen alterationsassociated with JHF.

Perlecan is a heparan sulfate proteoglycan. Perlecan is propablythe most abundant basement membrane proteoglycan, whichinteracts with other basement membrane components (laminin-1,b1-integrin) to form the matrix, and mediate selective filtrationand interactions with cell surface integrin receptors [40].Furthermore, Sher et al. recently demonstrated that perlecan isessential for the formation of the epidermis and that there is astrong correlation between the presence of perlecan in thebasement membrane zone and the normal formation of theepidermis [40]. In our study, perlecan gene expression showed nosignificant alteration in JHF lesions as compared to normal skintissue specimens.

Aggrecan is a large chondroitin and keratan sulfate proteogly-can that is the most abundant proteoglycan in cartilaginous tissue.

In the present study, we provide evidence that gene expression ofaggrecan is significantly increased in JHF lesions as compared tonormal skin tissue specimens obtained from matched sites. Thesedata are in agreement with our previous observation thatchondroitin sulfate is significantly increased in JHF lesions ascompared to normal skin [27]. Furthermore, it has been shown thatgene expression of aggrecan was dramatically up regulated indermal skin fibroblasts of Hutchinson-Gilford Progeria Syndrome,as compared to chronologically age-matched controls [41].Hutchinson-Gilford Progeria Syndrome is considered as a disorderof accelerated ageing characterized by connective tissue defects,alopecia and loss of subcutaneous tissue [41]. The increasedexpression of aggrecan, as well as the increased expression andactivity of MMP-2 and MMP-9 in JHF lesions that we report in thepresent study, indicate that JHF lesions are associated withalterations in the skin similar to those observed in skin ageing.

The present findings suggest that JHF skin lesions and normalskin differ, not only morphologically but also in the expressionlevel of ECM components and genes involving connective tissueturnover and remodeling. Results from these analyses, may behelpful to clarify the nature and degree of the involvement ofMMPs, TIMPs and proteoglycans in the etiology of JHF skin lesionformation, and thus to provide new insights into the aetiopatho-logical mechanisms of this disease. JHF is a rare, progressive andfatal disease with no treatment. To date, many therapeuticregimens have been attempted with limited or no response, whilethe management of skin lesions mainly consists of surgicalremoval, however with frequent local recurrences. The datapresented in this study, may offer novel targets for pharmacologicintervention towards a more effective strategy for the manage-ment of JHF, and indicate the possibility of a potential therapeuticintervention which could be feasible with the use of specificnaturally occurring or synthetic MMP inhibitors [42,43]. Finally,further research is required to clarify the role of all types of MMPsas well as other genes, regulating the expression of gelatinases,while there seems to be much ground to be covered in order tohave a complete overview of the effects of excessive gelatinasesand TIMPs production in JHF.

Acknowledgements

We thank Dr. I. Efstratiou from the Department of Pathology,General Hospital Papageorgiou, Thessaloniki, Greece for providingthe histological analysis.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in

the online version, at doi:10.1016/j.jdermsci.2010.12.002.

References

[1] Dowling O, Difeo A, Ramirez MC, Tukel T, Narla G, Bonafe L, et al. Mutations incapillary morphogenesis gene-2 result in the allelic disorders juvenile hyalinefibromatosis and infantile systemic hyalinosis. Am J Hum Genet 2003;73:957–66.

[2] Tanaka K, Ebihara T, Kusubata M, Adachi E, Arai M, Kawaguchi N, et al.Abnormal collagen deposition in fibromas from patient with juvenile hyalinefibromatosis. J Dermatol Sci 2009;55:197–200.

[3] Haleem A, Al-Hindi HN, Juboury MA, Husseini HA, Ajlan AA. Juvenile hyalinefibromatosis. Morphologic, immunohistochemical and ultrastructural study ofthree siblings. Am J Dermatopathol 2002;24:218–24.

[4] Iwata S, Horiuchi R, Maeda H, Ishikawa H. Systemic hyalinosis or juvenilehyaline fibromatosis. Ultrastructural and biochemical study of cultured skinfibroblasts. Arch Dermatol Res 1980;267:115–21.

[5] Mayer-da-Silva A, Poiares-Baptista A, Guerra Rodrigo F, Teresa-Lopes M.Juvenile hyaline fibromatosis. A histologic and histochemical study. ArchPathol Lab Med 1988;12. 928-311.

[6] Winik BC, Boente MC, Asial R. Juvenile hyaline fibromatosis: ultrastructuralstudy. Am J Dermatopathol 1998;20:373–8.

T.G. Tzellos et al. / Journal of Dermatological Science 61 (2011) 94–100100

[7] Hanks S, Adams S, Douglas J, Arbour L, Atherton DJ, Balci S, et al. Mutations inthe gene encoding capillary morphogenesis protein 2 cause juvenile hyalinefibromatosis and infantile systemic hyalinosis. Am J Hum Genet 2003;73:791–800.

[8] Shieh JT, Swidler P, Martignetti JA, Ramirez MC, Balboni I, Kaplan J, et al.Systemic hyalinosis: a distinctive early childhood–onset disorder character-ized by mutations in the anthrax toxin receptor 2 gene (ANTRX2). Pediatrics2006;118:1485–92.

[9] DeRosa G, Tornillo L, Orabona P, D’Antonio A, Staibano S, Boscaino A. Juvenilehyaline fibromatosis: a case report of a localized form? Am J Dermatopathol1994;16:624–7.

[10] Remberger K, Krieg T, Kunze D, Weinmann HM, Hubner G. Fibromatosishyalinica multiplex (juvenile hyaline fibromatosis). Light microscopic, elec-tron microscopic, immunohistochemical and biochemical findings. Cancer1985;56:614–24.

[11] Katagiri K, Takasaki S, Fujiwara S, Kayashima K, Ono T, Shinkai H. Purificationand structural analysis of extracellular matrix of a skin tumor from a patientwith juvenile hyaline fibromatosis. J Dermatol Sci 1996;13:37–48.

[12] Larralde M, Santos-Munoz A, Calb I, Magarinos C. Juvenile hyaline fibroma-tosis. Pediatr Dermatol 2001;18:400–2.

[13] Lubec B, Steinert I, Breier F, Jurecka W, Pillwein K, Fang-Kircher S. Skin collagendefects in a patient with juvenile hyaline fibromatosis. Arch Dis Child1995;73:246–8.

[14] Breier F, Fang-Kircher S, Wolff K, Jurecka W. Juvenile hyaline fibromatosis:impaired collagen metabolism in culture skin fibroblasts. Arch Dis Child1997;77:436–40.

[15] Nagase H, Woessner Jr JF. Matrix metalloproteinases. J Biol Chem 1999;274:21491–4.

[16] Woessner Jr JF. Matrix metalloproteinases and their inhibitors in connectivetissue remodeling. FASEB J 1991;5:2145–21454.

[17] Ohnishi Y, Ito Y, Tajima S, Ishibashi A, Arai K. Immunohistochemical study ofmembrane type-matrix metalloproteinases (MT-MMPs) and matrix metallo-proteinase-2 (MMP-2) in dermatofibroma and malignant fibrous histiocy-toma. J Dermatol Sci 2002;28:119–25.

[18] Rydlova M, Holubec Jr L, Ludvikova Jr M, Kalfert D, Franekova J, Povysil C, et al.Biological activity and clinical implications of the matrix metalloproteinases.Anticancer Res 2008;28:1389–97.

[19] Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases:evolution, structure and function. Biochim Biophys Acta 2000;1477:267–83.

[20] Clark IM, Swingler TE, Sampieri CL, Edwards DR. The regulation of matrixmetalloproteinases and their inhibitors. Int J Biochem Cell Biol 2008;40:1362–78.

[21] Iozzo RV. Matrix proteoglycans from molecular design to cellular function.Ann Rev Biochem 1998;67:609–52.

[22] Hardingham TE, Fosang AJ. Proteoglycans: many forms and many functions.FASEB J 1992;6:861–70.

[23] Papakonstantinou E, Aletras AJ, Glass E, Tsogas P, Dionyssopoulos A, Adjaye J,et al. Matrix metalloproteinases of epithelial origin in facial sebum of patientswith acne and their regulation by isotretinoin. J Invest Dermatol 2005;125:673–84.

[24] Papakonstantinou E, Dionyssopoulos A, Aletras AJ, Pesintzaki C, Minas A,Karakiulakis G. Expression of matrix metalloproteinases and their endogenoustissue inhibitors in skin lesions from patients with tuberous sclerosis. J AmAcad Dermatol 2004;51:526–33.

[25] Bradford MM. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. AnalBiochem 1976;72:248–54.

[26] Barber RD, Harmer DW, Coleman RA, Clark BJ. GAPDH as a housekeeping gene:analysis of GAPDH mRNA expression in a panel of 72 human tissues. PhysiolGenomics 2005;21:389–95.

[27] Tzellos TG, Dionyssopoulos A, Klagas I, Karakiulakis G, Lazaridis L, Papakon-stantinou E. Differential glycosaminoglycan expression and hyaluronan ho-meostasis in juvenile hyaline fibromatosis. J Am Acad Dermatol 2009;61:629–38.

[28] Maioli E, Fortino V, Torricelli C, Arezzini B, Gardi C. Effect of parathyroidhormone-related protein on fibroblast proliferation and collagen metabolismin human skin. Exp Dermatol 2002;11:302–10.

[29] Kerkela E, Saarialho-Kere U. Matrix metalloproteinases in tumor progression:focus on basal and squamous cell skin cancer. Exp Dermatol 2003;12:109–25.

[30] Hayakawa T, Yamashita K, Shinagawa A. Cell growth promoting activity oftissue inhibitor of metalloproteinases-2 (TIMP-2). J Cell Sci 1994;107:2372–9.

[31] Zyada MM. Expression of matrix metalloproteinase-9 and significance of amacrophage assay in eosinophilic granuloma. Ann Diagn Pathol 2009;13:367–72.

[32] O’Grady A, Dunne C, O’Kelly P, Murphy GM, Leader M, Kay E. Differentialexpression of matrix metalloproteinase (MMP)-2. MMP-9 and tissue inhibitorof metalloproteinase (TIMP)-1 and TIMP-2 in non-melanoma skin cancer:implications for tumour progression. Histopathology 2007;51:793–804.

[33] Imaizumi R, Akasaka Y, Inomata N, Okada E, Ito K, Ishikawa Y, et al. Promotedactivation of matrix metalloproteinase (MMP)-2 in keloid fibroblasts andincreased expression of MMP-2 in collagen bundle regions: implications formechanisms of keloid progression. Histopathology 2009;54:722–30.

[34] Bran GM, Goessler UR, Baftiri A, Hormann K, Riedel F, Sadick H. Effect oftransforming growth factor-beta1 antisense oligonucleotides on matrixmetalloproteinases and their inhibitors in keloid fibroblasts. Otolaryngol HeadNeck Surg 2010;143:66–71.

[35] Senzaki H, Kiyozuka Y, Uemura Y, Shikata N, Ueda S, Tsubura A. Juvenilehyaline fibromatosis: a report of two unrelated adult sibling cases and aliterature review. Pathol Int 1998;48:230–6.

[36] Hakki SS, Balci B, Hakki EE, Yilmaz E, Nohutcu RM. Identification of thedifference in ECM and adhesion molecules of cultured human gingival fibro-blasts versus juvenile hyaline fibromatosis gingival fibroblasts using cDNAmicroarray analysis. J Periodontol 2005;76:2244–53.

[37] Rinn JL, Wang JK, Liu H, Montgomery K, van de Rijn M, Chang HY. A systemsbiology approach to anatomic diversity of skin. J Invest Dermatol 2008;128:776–82.

[38] Chang HY, Chi JT, Dudoit S, Bondre C, van de Rijn M, Botstein D, et al. Diversity,topographic differentiation, and positional memory in human fibroblasts. ProcNatl Acad Sci USA 2002;99:12877–82.

[39] Kuwaba K, Kobayashi M, Nomura Y, Irie S, Koyama Y. Size control of decorindermatan sulfate during remodeling of collagen fibrils in healing skin. JDermatol Sci 2002;29:185–94.

[40] Sher I, Zisman-Rozen S, Eliahu L, Whitelock JM, Maas-Szabowski N, Yamada Y,et al. Targeting perlecan in human keratinocytes reveals novel roles forperlecan in epidermal formation. J Biol Chem 2006;281:5178–87.

[41] Lemire JM, Patis C, Gordon LB, Sandy JD, Toole BP, Weiss AS. Aggrecanexpression is substantially and abnormally upregulated in Hutchinson-GilfordProgeria Syndrome dermal fibroblasts. Mech Ageing Dev 2006;127:660–9.

[42] El Bradey M, Cheng L, Bartsch DU, Appelt K, Rodanant N, Bergeron-Lynn G,et al. Preventive versus treatment effect of AG3340, a potent matrix metallo-proteinase inhibitor in a rat model of choroidal neovascularization. J OculPharmacol Ther 2004;20:217–36.

[43] Skiles JW, Gonella NC, Jeng AY. The design, structure and therapeutic applica-tion of matrix metalloproteinase inhibitors. Curr Med Chem 2001;8:425–74.