Vol. 112 No. 4 October 2011

ORAL AND MAXILLOFACIAL PATHOLOGY Editor: Paul C. Edwards

Comparative analysis of the immunohistochemical expression ofcollagen IV, MMP-9, and TIMP-2 in odontogenic cystsand tumorsÁguida Cristina Gomes Henriques, DDS, MSc,a Marcelo Gadelha Vasconcelos, DDS, MSc,a

Hébel Cavalcanti Galvão, DDS, MSc, PhD,b Lélia Batista de Souza, DDS, MSc, PhD,b andRoseana de Almeida Freitas, DDS, MSc, PhD,b Natal, BrazilPOSTGRADUATE PROGRAM IN ORAL PATHOLOGY, DEPARTMENT OF DENTISTRY, FEDERALUNIVERSITY OF RIO GRANDE DO NORTE

Objective. The aim of this study was to evaluate the immunohistochemical expression of collagen IV, matrixmetalloproteinase (MMP) 9 and tissue inhibitor of MMP (TIMP) 2 in dentigerous cysts (DCs), radicular cysts (RCs),keratocystic odontogenic tumors (KOTs), and ameloblastomas.Study design. Twenty cases of DCs, 20 RCs, 20 KOTs, and 20 ameloblastomas were selected and analyzed byimmunohistochemistry.Results. Most DCs and RCs showed continuous and �50% staining for collagen IV in the basement membrane of theepithelium, whereas predominantly discontinuous thin and �50% staining was observed in KOTs andameloblastomas, with a significant difference in staining percentage (P � .001). MMP-9 was diffusely distributed andlocalized in both epithelial and mesenchymal cells of all of the lesions analyzed. The staining percentage was higherin the epithelium (P � .058) and mesenchyme (P � .005) of KOTs and ameloblastomas. Moreover, the distributionpattern, location, and percentage of expression of TIMP-2 were similar in the lesions studied, except forameloblastoma, with a significant difference in staining percentage (P � .001).Conclusion. These results demonstrate that the interaction between collagen IV, MMP-9, and TIMP-2 is an importantfactor for the establishment of differences in the biologic behavior of the odontogenic cysts and tumors studied. (Oral

Surg Oral Med Oral Pathol Oral Radiol Endod 2011;112:468-475)

Factors related to the epithelial and mesenchymal com-ponents participate in the regulation of the growth ofodontogenic cystic lesions and tumors.1 The alteredexpression of specific proteins of the extracellular ma-trix (ECM), associated with the exuberant presence ofmatrix metalloproteinases (MMPs) and the absence ofexpression of metalloproteinase inhibitors (TIMPs),may influence the behavior of these lesions. In the case

Supported by the Brazilian Research Council for Science and Tech-nology Development (CNPq).aPhD Student.bProfessor.Received for publication Mar 24, 2011; returned for revision May 17,2011; accepted for publication May 23, 2011.1079-2104/$ - see front matter© 2011 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2011.05.033

468

of tumors, this situation contributes to the growth andhigher aggressiveness of the tumor.2

The odontogenic keratocyst, recently reclassified askeratocystic odontogenic tumor (KOT),3 is known forits aggressive nature and high rate of recurrence, espe-cially compared with other odontogenic cysts.4 Amelo-blastoma is a locally aggressive benign epithelial odon-togenic tumor with a marked invasion potential thatresults in multiple recurrences after enucleation andcurettage.5 In contrast, dentigerous (DC) and radicular(RC) cysts show an indolent behavior and rarely recurafter surgical removal. KOT presents a cystic structuresimilar to that of DC and RC, but its invasive anddestructive growth is similar to that of ameloblastoma.6

KOT, ameloblastoma, DC, and RC show distinctevolutions and biologic behaviors. In view of this fact,

a growing number of studies have tried to identify

pH 6.0

OOOOEVolume 112, Number 4 Henriques et al. 469

epithelial and mesenchymal factors that determine thedifferences in behavior between these diseases. Theselesions arise from the remnants of odontogenesis,which is controlled by the interaction between epithe-lial and mesenchymal components.

Because odontogenic lesions sometimes mimic theevents related to the morpho- and histodifferentiationnecessary for tooth development, it is thought thatduring tumorigenesis, epithelial-mesenchymal interac-tions are also determinant for the evolution of theselesions. However, according to Oliveira et al.,7 manystudies have investigated only the epithelial compo-nents of odontogenic cysts and tumors.

The growth mechanism of odontogenic cysts, as wellas the invasion and destructive potential of some odon-togenic tumors, might be influenced by the secretion ofMMPs, proteins that can be produced by both epithelialand mesenchymal cells.8,9 Alaeddini et al.10 suggestedthat for a better understanding of the biologic behaviorof these lesions, it is necessary to study the componentsof the extracellular matrix (ECM) to identify differ-ences in the distribution and expression of these pro-teins between indolent cystic lesions and more aggres-sive tumors.

In view of the above considerations, the objective ofthe present study was to evaluate the immunohisto-chemical expression of collagen IV, MMP-9, andTIMP-2 in a series of cases of RC, DC, KOT, andameloblastoma in order to contribute to a better under-standing of the role of these proteins in the biologicbehavior of these lesions.

MATERIAL AND METHODSThe sample consisted of 80 paraffin-embedded tissue

specimens comprising 20 cases of solid ameloblastoma,20 KOTs, 20 DCs, and 20 RCs, obtained from thePathological Anatomy Service of the Discipline of OralPathology, Department of Dentistry, Federal Universityof Rio Grande do Norte (UFRN). The diagnoses ofodontogenic cysts and tumors were made using the1992 and 2005 World Health Organization classifica-tion criteria, respectively. The study was approved bythe Research Ethics Committee of UFRN.

Immunohistochemical methodsFor immunohistochemistry, 3 �m thick sections

Table I. Specifications of the antibodies usedAntibody Manufacturer Clone

Collagen IV Dako CIV22 PepsinMMP-9 NovoCastra 2C3 Citrate,TIMP-2 NovoCastra 3A4 Citrate,

were mounted on glass slides previously prepared with

organosilane adhesive (3-aminopropyltrithoxy-silane;Sigma Chemical Co.) and submitted to the streptavidin-biotin method. The histologic sections were deparaf-finized in xylene and rehydrated in a decreasing alcoholseries. The sections were submitted to antigen retrieval(Table I) and blockade of endogenous peroxidase with10 volumes of hydrogen peroxide, washed in water, andincubated with Tris-HCl, pH 7.4, for 10 minutes. Next,the sections were incubated with the primary antibodies(Table I) diluted in 1% bovine serum albumin (BSA)/Tris-HCl, pH 7.4. The sections were then washed twicein phosphate-buffered saline solution (PBS) and treatedwith the labeled streptavidin biotin complex (LSAB;Dako, Carpinteria, CA) at room temperature to bind theprimary antibodies. The reactions were developed with0.03% diaminobenzidine (Liquid DAB� Substrate;Dako) as chromogen, and the slides were counter-stained with Mayer hematoxylin for 10 minutes. Fi-nally, the sections were dehydrated in alcohol andcleared in xylene for mounting in Permount resin(Fisher Scientific) under a coverslip. Liver and bladdersections were used as positive control samples forMMP-9 and TIMP-2, respectively. Immunostaining ofblood vessels in the basement membrane (BM) wasused as internal positive control for collagen IV. Sam-ples treated as described above, except that the primaryantibody was replaced with a solution of BSA in PBS,served as negative control samples.

Immunostaining was evaluated by 2 examiners atdifferent times under a light microscope. For collagenIV, immunoreactivity was analyzed at the BM at theepithelial/mesenchymal interface. Staining percentagewas analyzed semiquantitatively, as �50% or �50% ofthe BM extension. The pattern of expression was eval-uated according to continuity (continuous, discontinu-ous) and thickness (thin, thick, and mixed thin and thickareas) of staining at the BM.

The expression of MMP-9 and TIMP-2 was ana-lyzed separately in the epithelial and mesenchymalcomponents. In the mesenchymal component, stain-ing was detected in fibroblasts and endothelial andinflammatory cells. Staining percentage was ana-lyzed semiquantitatively according to the method ofFregnani et al.11: �50% stained cells and �50%stained cells. The distribution of staining was clas-sified as focal and diffuse. The localization of stain-

ntigen retrieval Dilution Incubation

a 1%, 60 min, 37°C, steamer 1:50 120 min, 30 min, steamer 1:20 Overnight, 30 min, steamer 1:50 Overnight

A

pH 1.8pH 6.0

ing in the epithelial component was classified as

OOOOE470 Henriques et al. October 2011

follows: basal, suprabasal, and basal/suprabasallayer.

Immunohistochemical staining of collagen IV,MMP-9, and TIMP-2 was evaluated by descriptive andsemiquantitative methods. Statistical analysis was per-formed using the SPSS for Windows program, version7.0 (SPSS, Chicago, IL). The �2 test was used tocompare epithelial and mesenchymal staining. A Pvalue of �.05 was considered to indicate statisticalsignificance.

RESULTSMost DCs (65%) and RCs (55%) exhibited continu-

ous staining for collagen IV in the BM of the epithe-lium (Fig. 1, a and b), whereas discontinuous stainingpredominated in KOTs (60%) and ameloblastomas(75%) (Fig. 1, c and d). Staining appeared thick ormixed in 70% of DCs and thin in 75% of RCs, 60% ofKOTs, and 75% of ameloblastomas. The percentage ofcollagen IV expression differed significantly betweenlesions (P � .001). Immunostaining �50% was ob-served in 70% of DCs and immunostaining �50% in55% of RCs. The lowest staining percentage was ob-

Fig. 1. Immunohistochemical staining of collagen IV at the b(RC), (c) keratocystic odontogenic tumor (KOT), and (d) ameobserved in DC and RC. c, Note linear, discontinuous, thcharacteristic of ameloblastoma.

served for KOTs and ameloblastomas, with immuno-

staining �50% in 80% and 100% of the lesions, re-spectively (Table II).

Expression of MMP-9 was mainly diffuse in theepithelial and mesenchymal component of 100% ofthe lesions studied. There was minimal variation in thepattern of localization of stained cells. In the epithe-lium, staining was observed in all layers of 95% of DCsand 90% of RCs (Fig. 2, a and b), whereas in KOTspositive staining was detected throughout the epithe-lium in 100% of cases (Fig. 2, c). In ameloblastomas,staining was detected in both peripheral cells and in thecenter of islands of the neoplastic epithelium (Fig. 2, d).A staining percentage �50% predominated in all le-sions studied, with a trend to higher expression inKOTs (90%) and ameloblastomas (95%) comparedwith DCs (65%) and RCs (75%). However, there wasno statistically significant difference (P � .058; TableII). Mesenchymal staining was observed in fibroblastsand inflammatory and endothelial cells in all casesstudied. Regarding the expression of MMP-9 in themesenchyme, staining percentage �50% was higher inameloblastomas (70%) and KOTs (40%), whereas only25% of RCs and 20% of DCs exhibited �50% stained

nt membrane in (a) dentigerous cyst (DC), (b) radicular cystoma (�400). Strong, thick continuous and linear deposit wasosit in KOT. d, Linear discontinuous and thin deposit is

asemeloblastin dep

cells (P � .05; Table II).

tic odo

OOOOEVolume 112, Number 4 Henriques et al. 471

The expression of TIMP-2 was similar in all lesions,except for ameloblastoma. The distribution of stainingwas diffuse in the epithelium and connective tissue of100% of DCs, RCs and KOTs (Fig. 3, a-c), whereas in

Fig. 2. Immunohistochemical expression of MMP-9 (�400).dentigerous cyst (DC), (b) radicular cyst (RC), (c) keratocys

Table II. Immunohistochemical expression of collagencyst (DC), keratocystic odontogenic tumor (KOT), and

Protein

Le

RC DC K

Collagen IV1%-50% 11 (55%) 6 (30%) 16 (�50% 9 (45%) 14 (70%) 4 (

MMP-9 (epithelium)1%-50% 7 (35%) 5 (25%) 2 (�50% 13 (65%) 15 (75%) 18 (

MMP-9 (mesenchyme)1%-50% 15 (75%) 16 (80%) 12 (�50% 5 (25%) 4 (20%) 8 (

TIMP-2 (epithelium)1%-50% 1 (5%) 0 (0%) 0 (�50% 19 (95%) 20 (100%) 20 (

TIMP-2 (mesenchyme)1%-50% 0 (0%) 1 (5%) 6 (�50% 20 (100%) 19 (95%) 14 (

ameloblastomas focal staining was observed in the ep-

ithelium of 55% of cases and diffuse staining in theconnective tissue of 80% (Fig. 3, d). Positive stainingof all epithelial cell layers was observed in 100% ofDCs and RCs and in 95% of KOTs. In ameloblastomas,

the diffuse staining in the epithelium and mesenchyme in (a)ntogenic tumor (KOT), and (d) ameloblastoma.

MP-9, and TIMP-2 in radicular cyst (RC), dentigerousoblastoma(%)

�2 P valueAmeloblastoma Total

20 (100%) 53 (66.3%) 24.766 �.0010 (0)%) 27 (33.8%)

1 (5%) 15 (18.8%) 7.467 .05819 (75%) 65 (81.3%)

6 (30%) 49 (61.3%) 12.798 .00514 (70%) 31 (38.8%)

9 (45%) 10 (12.5%) 26.057 �.00111 (55%) 70 (87.5%)

11 (55%) 11 (55%) 22.079 �.0019 (45%) 9 (45%)

Note

IV, Mamel

sion, n

OT

80%)20%)

10%)90%)

60%)40%)

0%)100%)

30%)70%)

staining of peripheral epithelial cells and cells in the

OOOOE472 Henriques et al. October 2011

center of tumor islands was observed in 95% of cases.Mesenchymal staining was observed in fibroblasts andinflammatory and endothelial cells in all cases studied.The percentage of TIMP-2 staining in the epitheliumwas �50% in most lesions (100% DCs, 95% RCs, and100% KOTs), whereas �50% stained cells were ob-served in only 55% of ameloblastomas (P � .001;Table II). In the mesenchyme, �50% stained cells wereobserved in 95% of DCs and in 100% of RCs andKOTs, whereas 55% of cases of ameloblastoma pre-sented �50% stained cells (P � .001; Table II).

DISCUSSIONKOT, ameloblastoma, DC, and RC present distinct

evolutions and biological behaviors. Despite the cysticstructure of KOT, some studies emphasize the similar-ity in the behavior of these cysts to the ameloblastomain view of their invasive growth and destructive poten-tial.6

Several factors have been associated with the aggres-sive behavior of ameloblastomas, such as an increasedproliferation potential,12 alterations in the expression oftumor suppressor genes, and the aberrant expression ofcell cycle–regulating proteins,13 adhesion molecules,14

Fig. 3. Immunohistochemical expression of TIMP-2 (�400).dentigerous cyst (DC), (b) radicular cyst (RC), and (c) keratoand stroma of ameloblastoma were focal.

and MMPs and their inhibitors (TIMPs).15 In addition,

Silveira et al.8 suggested that the mechanism of expan-sion of KOTs, DCs, and RCs may also be influenced, oreven be conducted, by the secretion of MMPs. Thus, anexuberant expression of MMPs might be related to ahigher aggressiveness of cystic lesions.

MMP-9 is known to play an important role in BMdegradation owing to its ability to degrade collagen IV,the main BM component. However, a balanced actionof activators and inhibitors of MMPs is necessary sothat these proteins can exert their function. In thisrespect, TIMPs inhibit the active forms of MMPs.16 Onthe basis of these observations it could be suggestedthat the degradation of ECM is not dependent on theexpression of a single MMP, but rather that the com-bined action of several MMPs and TIMPs is essentialfor the efficient degradation of the BM and interstitialstroma. Kumamoto et al.15 performed an immunohis-tochemical study and found that increased expressionof TIMP-1 and TIMP-2 was related to weak immuno-reactivity for MMP-9. According to those authors, it ispossible that the TIMPs contribute to suppress tumorprogression in ameloblastomas through the inhibitionof MMP-9 activity.

Within this context, the evaluation of the expression

the staining diffuse in the epithelium and mesenchyme in (a)odontogenic tumor (KOT). d, The staining in the epithelium

Notecystic

of collagen IV, MMP-9, and TIMP-2 may contribute to

OOOOEVolume 112, Number 4 Henriques et al. 473

the understanding of the distinct biologic behavior ofthese odontogenic lesions. ECM proteins have beendetected in several odontogenic cysts and tu-mors,7,8,17-19 but no investigation has explored the re-lationship between collagen IV, MMPs, and TIMPs inodontogenic cysts and tumors.

In the present study, a lower expression of collagenIV was observed in KOT and ameloblastoma. A pos-sible explanation is that a higher degradation or areduced synthesis of this protein occurred in theselesions. We speculate that both increased degradationand reduced synthesis of collagen may influence theaggressiveness of the lesions. A higher capacity todegrade the BM and other ECM components has beenassociated that with a higher destructive potential of thelesions.20 According to Nagatsuka et al.,17 the expres-sion of collagen IV may be correlated with the growthand aggressiveness of odontogenic cysts and tumors.

In normal and benign tissue, collagen IV has beenreported to appear in a continuous, thick and stronglystained linear pattern.21,22 In the present study, RCs,and DCs generally exhibited a linear and continuousexpression of collagen IV, in agreement with the stain-ing pattern described for normal tissues. On the otherhand, KOTs and ameloblastomas presented a highlydiscontinuous thin linear staining pattern, with �50%extension of the BM in 80% and 100% of cases, re-spectively. This finding might be related to the pro-liferative potential of these odontogenic tumors. Theloss of BM integrity is known to favor alterations inepithelial-mesenchymal interactions that contributeto the mechanisms of cell differentiation and tumorproliferation.7,23

Regarding the thickness of collagen IV staining, RCspresented an irregularity compared with the stainingseen in DCs. Most RCs (75%) exhibited a thin line,whereas staining in DCs was characterized by thick andthin areas in 70% of specimens. It is therefore thoughtthat in RCs the inflammatory infiltrate, which is adeterminant component of the lesion, contributed to thelower expression of collagen IV, because those cellsshowed high staining percentage for MMP-9. On theother hand, the presence of an inflammatory infiltrate isa common finding in DCs, and MMP-9 secretion waspossibly lower, a fact explaining the regular and thickcollagen IV staining.

The percentage of collagen IV staining was morepronounced in DCs than in RCs, KOTs and ameloblas-tomas. However, 45% of RC cases showed a stainingpercentage similar to that of DCs. Therefore, the pres-ent results suggest that the weak expression of collagenIV, mainly observed in KOTs and ameloblastomas,

may induce changes in the interaction between the

epithelium and mesenchyme, contributing to the moreaggressive growth of these lesions.

In agreement with the findings of the present study,Oliveira et al.7 observed differences in the compositionof the BM of RCs, DCs, and KOTs. Those authorsobserved continuous and intense expression of collagenIV in RCs. On the other hand, KOTs exhibited thin,linear, markedly discontinuous and weak staining. Ac-cording to those authors, the epithelial component ofKOTs has a higher proliferative capacity than that ofDCs and RCs, a fact supporting the distinct biologicbehavior of those lesions. In addition, it is possible thatthis BM irregularity shown by KOTs favors the con-stant shedding of epithelial cells from the capsule andmay be related to the higher recurrence rates of thesetumors. Although consensus exists regarding the rela-tionship between BM integrity and aggressiveness ofthe lesion, Poomsawat et al.19 found no associationbetween the expression of BM components (laminin 1and 5, collagen IV, and fibronectin) in RCs, DCs, andKOTs and the biologic behavior of these lesions.Therefore, in view of the functions of the BM, espe-cially its involvement in the control of cell prolifera-tion, differentiation, and invasion, it is evident howmuch these processes are altered in KOTs and amelo-blastomas, a fact that can explain the more aggressivebehavior of these tumors compared with RCs and DCs.

Considerable research effort has been devoted to theMMP subfamily of gelatinases/type IV collagenases,including MMP-2, referred to as gelatinase-A, andMMP-9, referred to as gelatinase-B, because of theirability to degrade type IV collagen and gelatins.15 Ex-pression of gelatinases/type IV collagenases has beeninvestigated in numerous tumors and has been shown tobe associated with an invasive phenotype and meta-static potential of tumor cells.24,25 Preclinical trialshave shown that squamous cell carcinomas expresshigh levels of MMPs in vivo and that the inhibition ofthese enzymes in vitro and in animal models reducesthe invasion potential and metastatic capacity of thetumor.2

Regarding the expression of MMP-9 in odontogeniccysts and tumors, in the present study epithelial andmesenchymal cells of the lesions studied diffusely se-creted this protein. This finding suggests a role ofMMP-9 in tumor growth, because the protein is knownto act on the remodeling of ECM and the degradation ofbone matrix.26,27 However, the staining percentage ofMMP-9 in the epithelial and mesenchymal componentwas markedly higher in KOTs and ameloblastomasthan in RCs and DCs, with a staining percentage �50%being observed in 90% and 95% of cases, respectively.These findings suggest that the growth of these lesions

is influenced by the secretion of MMPs. Therefore, the

OOOOE474 Henriques et al. October 2011

more exuberant expression of MMP-9 in the odonto-genic tumors analyzed suggest a possible participationof this protein in cell proliferation and differentiationand may explain the greater bone resorption, higherinvasion potential, and poorer prognosis compared withDC and RC.11 The immunoreactivity to MMP-9 seen inboth the epithelium and the mesenchyme of these le-sions can be explained by the fact that the protein isproduced by epithelial cells, fibroblasts, macrophages,plasma cells, lymphocytes, and neutrophils.28

According to Vicente et al.,16 MMP-2 and MMP-9are involved in angiogenesis and tumor growth, a factsuggesting an association of these gelatinases with theaggressive behavior and unpredictable clinical courseof some human neoplasms. Similarly to the presentstudy, Silveira et al.8 observed more pronounced ex-pression of MMP-9 in the mesenchymal component ofKOTs compared with RCs and DCs, a finding support-ing the hypothesis of a more aggressive behavior of thistumor compared with other cysts. In the study by Ku-mamoto et al.,15 immunoreactivity to MMP-9 wasstronger in the stroma of ameloblastomas than in themesenchymal component of tooth germs, suggestingthat an increased production of this protein by neoplas-tic cells is related to the transformation of odontogenictissues and aggressiveness of this tumor. Ribeiro et al.29

also found intense immunostaining for this gelatinase inthe parenchyma and stroma of ameloblastomas, indi-cating its possible participation in the growth of thesetumors.

Curiously, combined analysis of the present resultsshowed that KOTs and ameloblastomas, which pre-sented higher expression of MMP-9 in epithelial andmesenchymal components, also exhibited low expres-sion of collagen IV, which may result from higher BMdegradation or reduced synthesis of collagen IV. Wethink that these findings suggest higher BM degradationin consequence of high expression of MMP-9. Somereports in the literature show that MMP-9 plays animportant role in the degradation of the BM as a resultof its ability to degrade collagen IV.16

Nagase et al.30 and Verma and Hansch31 demon-strated that MMPs are weakly expressed by tissuesunder physiologic conditions, whereas high expressionof these proteins is observed in neoplastic tissues, ow-ing to an imbalance in the activity of MMPs and theirinhibitors (TIMPs). In the present study, high percent-age and diffuse expression of TIMP-2 was observed inepithelial and mesenchymal cells of DCs, RCs, andKOTs. In contrast, focal staining was detected inameloblastomas, a finding possibly implying lower se-cretion of this protein which inhibits the activity ofMMP-9. Similarly, the staining percentage was less

pronounced in the epithelial and stromal components of

ameloblastoma compared with the other lesions, a find-ing suggesting higher production of MMP-9, increasedBM degradation and the consequent higher invasionpotential of this tumor. Siqueira et al.32 observed in-tense and diffuse TIMP-2 staining in ameloblastomas,which was associated with lower aggressiveness of thetumor. In an in vitro study, Wang et al.5 showed thatoverexpression of TIMP-2 in ameloblastoma cells wasassociated with the suppression of MMP-2 activity anda consequently lower invasion potential. It is believedthat TIMP-2 suppresses tumor invasion by inhibitingMMPs.33 Taken together, these studies suggest that inthe present series, the lower expression of TIMP-2 inameloblastomas might be an indicator of the moreaggressive behavior of this tumor compared with theother lesions studied, a fact supported by the higherexpression of MMP-9.

The higher expression of TIMP-2 in DCs and RCsseems to be the result of a lower activity of MMP-9 andlower BM degradation, which is compatible with themore indolent behavior of these lesions. The balancebetween MMPs and TIMPs ultimately determines theextent of ECM degradation under physiologic andpathologic conditions.34 These features suggest thatTIMP-2 may suppress tumor progression in ameloblas-tomas.

Immunoreactivity to MMP-9 and TIMP-2 was ob-served in the epithelial and mesenchymal componentsof all odontogenic lesions studied herein. This findingsuggests that, similar to odontogenesis, epithelial-mes-enchymal interactions are determinant for the regula-tion of the growth of cystic lesions and for tumorprogression.

In conclusion, the present results suggest an interac-tion between the production of MMP-9 and TIMP-2and the degradation of BM components, which maycontribute to the distinct biologic behavior of amelo-blastomas and KOTs compared with DCs and RCs.Ameloblastomas and KOTs demonstrate higher pro-duction of MMP-9 and increased degradation or re-duced synthesis of collagen IV. The expression of thesemolecules may serve as an indicator of the degree oflocal aggressiveness of these odontogenic cysts andtumors.

REFERENCES1. Pinheiro JJ, Freitas VM, Moretti AI, Jorge AG, Jaeger RG. Local

invasiveness of ameloblastoma. Role played by matrix metallo-proteinases and proliferative activity. Histopathology 2004;45:65-72.

2. Rosenthal EL, Matrisian LM. Matrix metalloproteases in headand neck cancer. Head Neck 2006;28:639-48.

3. Philipsen HP. Keratocystic odontogenic tumour. Barnes L, Eve-son JW, Reichart P, Sidransky D, editors. World Health Orga-nization classification of tumors. Pathology and genetics of head

and neck tumours. Lyon, France: IARC Press, pp. 306-7.

OOOOEVolume 112, Number 4 Henriques et al. 475

4. Kolár Z, Geierová M, Bouchal J, Pazdera J, Zboril V, Tvrdý P.Immunohistochemical analysis of the biological potential ofodontogenic keratocysts. J Oral Pathol Med 2006;35:75-80.

5. Wang A, Zhang B, Huang H, Zhang L, Zeng D, Tao Q, et al.Suppression of local invasion of ameloblastoma by inhibition ofmatrix metalloproteinase-2 in vitro. BMC Cancer 2008;8:182.

6. da Silva TA, Batista AC, Mendonça EF, Leles CR, Fukada S,Cunha FQ. Comparative expression of RANK, RANKL, andOPG in keratocystic odontogenic tumors, ameloblastomas, anddentigerous cysts. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2008;105:333-41.

7. Oliveira MDC, Souza LB, Pinto LP, Freitas RA. Immunohisto-chemical study of components of the basement membrane inodontogenic cysts. Pesqui Odontol Bras 2002;16:157-62.

8. Silveira EJD, Piva MR, Galvão HC, Souza LB, Freitas RA. Roleof matrix metalloproteinases in the etiopathogeny of odontogeniccysts. J Bras Patol Med Lab 2007;43:203-9.

9. Freitas VS, Araújo CRF, Alves PM, Souza LB, Galvão HC,Freitas RA. Immunohistochemical expression of matrilysins(MMP-7 and MMP-26) in ameloblastomas and adenomatoidodontogenic tumors. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2009;108:417-24.

10. Alaeddini M, Salah S, Dehghan F, Eshghyar N, Etemad-Mogh-adam S. Comparison of angiogenesis in keratocystic odontogenictumours, dentigerous cysts and ameloblastomas. Oral Dis2009;15:422-7.

11. Fregnani ER, Sobral LM, Alves FA, Soares AS, Kowalski LP,Coletta RD. Presence of myofibroblasts and expression of matrixmetalloproteinase-2 (MMP-2) in ameloblastoma correlate withrupture of the osseous cortical. Pathol Oncol Res 2009;15:231-40.

12. Meer S, Galpin JS, Altini M, Ali N, Ali H. Proliferating cellnuclear antigen Ki67 immunoreactivity in ameloblastomas. OralSurg Oral Med Oral Pathol Oral Radiol Endod 2003;95:213-21.

13. Kumamoto H, Izutsu T, Ohki K, Takahashi N, Ooya K. p53 genestatus and expression of p53, MDM2, and p14ARF proteins inameloblastoma. J Oral Pathol Med 2004;33:292-9.

14. Souza ESA, Miguel MC, Freitas RA, Pinto LP, Souza LB.Immunoexpression of integrins in ameloblastoma, adenomatoidodontogenic tumor, and human tooth germs. Int J Surg Pathol2008;16:277-85.

15. Kumamoto H, Yamauchi K, Yoshida M, Ooya K. Immunohis-tochemical detection of matrix metalloproteinases (MMPs) andtissue inhibitors of metalloproteinases (TIMPs) in ameloblasto-mas. J Oral Pathol Med 2003;32:114-20.

16. Vicente JC, Fresno MF, Villalain L, Vega JA, Hernández-Vallejo G. Expression and clinical significance of matrix metal-loproteinase-2 and matrix metalloproteinase-9 in oral squamouscell carcinoma. Oral Oncol 2005;41:283-93.

17. Nagatsuka H, Siar CH, Nakano K, Tsujigiwa H, Gunduz M,Choufuku H, et al. Differential expression of collagen IV alpha1to alpha6 chains in basement membranes of benign and malig-nant odontogenic tumors. Virchows Arch 2002;441:392-9.

18. Oliveira MDC, Miranda JL, Amorim RFB, Souza LB, FreitasRA. Tenascin and fibronectin expression in odontogenic cysts.J Oral Pathol Med 2004;33:354-9.

19. Poomsawat S, Punyasingh J, Weerapradist W. Expression ofbasement membrane components in odontogenic cysts. Oral Dis2006;12:290-6.

20. Pereira, AL, Veras SSL, Silveira EJD, Seabra FRG, Pinto LP,

Souza LB, Freitas RA. The role of matrix extracellular proteinsand metalloproteinases in head and neck carcinomas: an updatedreview. Rev Bras Otorrinolaringol 2005;71:81-6.

21. Kannan S, Balaram P, Chandran GJ, Pillai MR, Mathew B,Nalinakumari KR, et al. Alterations in expression of basementmembrane proteins during tumor progression in oral mucosa.Histopathology 1994;24:531-7.

22. Tsubura A, Shikata N, Inui T, Morii S, Hatano T, Oikawa T, etal. Immunohistochemical localization of myoepithelial cells andbasement membrane in normal, benign and malignant humanbreast lesions. Virchows Arch 1998;413:133-9.

23. Nakano K, Siar CH, Nagai N, Naito I, Sado Y, Nagatsuka H, etal. Distribution of basement membrane type IV collagen alphachains in ameloblastoma: an immunofluorescence study. J OralPathol Med 2002;31:494-9.

24. Duffy MJ, Maguire TM, Hill A, McDermott E, O’Higgins N.Metalloproteinases: role in breast carcinogenesis, invasion andmetastasis. Breast Cancer Res 2000;2:252-7.

25. Werner JA, Rathcke IO, Mandic R. The role of matrix metallo-proteinases in squamous cell carcinomas of the head and neck.Clin Exp Metastasis 2002;19:275-82.

26. Delaissé JM, Engsig MT, Everts V, del Carmen, MO, FerrerasM, Lund L, et al. Proteinases in bone resorption: obvious and lessobvious roles. Clin Chim Acta 2000;291:223-34.

27. Chirgwin JM, Guise TA. Molecular mechanisms of tumor-boneinteractions in osteolytic metastases. Crit Rev Eukaryot GeneExpr 2000;10:159-78.

28. Heidinger M, Kolb H, Krell HW, Jochum M, Ries C. Modulationof autocrine TNF-alpha–stimulated matrix metalloproteinase 9(MMP-9) expression by mitogen-activated protein kinases inTHP-1 monocytic cells. Biol Chem 2006;387:69-78.

29. Ribeiro BF, Iglesias DPP, Nascimento GJF, Galvão HC, Medei-ros AMC. Immunoexpression of MMPs-1, -2, and -9 in amelo-blastoma and odontogenic adenomatoid tumor. Oral Dis 2009;15:472-7.

30. Nagase H, Visse R, Murphy G. Structure and function of matrixmetalloproteinases and TIMPs. Cardiovasc Res 2006;69:562-73.

31. Verma RP, Hansch C. Matrix metalloproteinases (MMPs):chemical-biological functions and (Q)SARs. Bioorg Med Chem2007;15:2223-68.

32. Siqueira AS, Carvalho MR, Monteiro AC, Freitas VM, JaegerRG, Pinheiro JJ. Matrix metalloproteinases, TIMPs and growthfactors regulating ameloblastoma behavior. Histopathology2010;57:128-37.

33. Ikebe T, Shinohara M, Takeuchi H, Beppu M, Kurahara S,Nakamura S et al. Gelatinolytic activity of matrix metalloprotei-nase in tumor tissues correlates with the invasiveness of oralcancer. Clin Exp Metastasis 1999;17:315-23.

34. Meneses-García A, Betancourt AM, Abarca JH, Montes AB,Roa LS, Ruíz-Godoy L. Expression of the metalloproteasesMMP-1, MMP-2, MMP-3, MMP-9, MMP-11, TIMP-1 andTIMP-2 in angiocentric midfacial lymphomas. World J SurgOncol 2008;6:114.

Reprint requests:

Roseana de Almeida FreitasAv. Senador Salgado Filho, 1787Lagoa Nova, Natal, Rio Grande do NorteBrazil

roseanafreitas@hotmail.com