The Veterinary Journal 190 (2011) 396–402

Contents lists available at ScienceDirect

The Veterinary Journal

journal homepage: www.elsevier .com/ locate / tv j l

Immunohistochemical evaluation of MMP-2 and TIMP-2 in canine mammarytumours: A survival study

Andreia Santos a,b, Célia Lopes a, Carlos Frias a, Irina Amorim a, Corália Vicente a, Fátima Gärtner a,c,Augusto de Matos a,b,⇑a Institute of Biomedical Sciences Abel Salazar (ICBAS), University of Porto, Largo Professor Abel Salazar, 2, 4099-003 Porto, Portugalb Multidisciplinary Unit for Biomedical Research (UMIB), University of Porto, Largo Professor Abel Salazar, 2, 4099-003 Porto, Portugalc The Institute of Molecular Pathology and Immunology of the University of Porto (IPATIMUP), R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history:Accepted 3 December 2010

Keywords:CanineNeoplasiaMMP-2TIMP-2Prognosis

1090-0233/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.tvjl.2010.12.003

⇑ Corresponding author. Tel.: +351 222 062 260.E-mail address: ajmatos@icbas.up.pt (A.de Matos)

Canine mammary tumours (CMTs) are a very heterogeneous group of neoplasms with variable prognosis.Their aggressiveness is mainly due to their ability to invade locally and to metastasize. The degradation ofextracellular matrix components is an important determinant of the invasive phenotype. The aims of thisstudy were to analyse by immunohistochemistry and double immunofluorescence the expression ofmetalloproteinase 2 (MMP-2) and tissue inhibitor of metalloproteinase 2 (TIMP-2) in eight normal caninemammary glands and 118 CMTs (24 benign, 94 malignant) and to investigate relationships with meta-static disease and survival.

MMP-2 and TIMP-2 expression was higher in malignant tumours than in normal canine mammary tis-sue and benign tumours. The main difference between benign and malignant CMTs was the pattern ofexpression of both molecules: benign tumours presented TIMP-2 and MMP-2 immunoreactivity in themyoepithelial cells lining the basement membrane of tubuloalveolar structures, while malignanttumours showed mainly diffuse expression in neoplastic cells. In malignant tumours, increased TIMP-2expression was significantly associated with the development of distant metastases, lower overall sur-vival and lower disease-free survival. MMP-2 expression was not significantly associated to any of theseparameters. These results suggest that the immunohistochemical expression of TIMP-2 is a useful prog-nostic factor in CMTs.

� 2010 Elsevier Ltd. All rights reserved.

Introduction

Tumour metastasis involves a multi-step process that includesdisruption of the interstitial connective tissue and basement mem-branes (BM) allowing tumour cells to escape from the primary tu-mour and to invade surrounding tissues (Hojilla et al., 2008).Matrix metalloproteinases (MMPs) are among the most importantproteolytic enzymes implicated in the degradation of variousextracellular matrix (ECM) components (Pupa et al., 2002).

MMPs are a family of zinc-dependent endopeptidases(Köhrmann et al., 2009), that are secreted as inactive zymogens(pro-MMPs) and that require extracellular activation by otherMMPs (such as membrane-type MMPs) and serine proteases (Tothet al., 2000). According to their substrate specificity, MMPs are di-vided into several subfamilies, namely, collagenases, gelatinases,stromalysins, matrilysins, membrane-type MMPs, elastases andothers (Somerville et al., 2003). Many human breast cancer inves-tigators have focused on the gelatinases (MMP-2 and MMP-9),

ll rights reserved.

.

since they are able to degrade type IV collagen, an important com-ponent of BM (Köhrmann et al., 2009).

Evidence has recently emerged showing that the role of MMPsgoes beyond digesting ECM components (Somerville et al., 2003).It is believed that MMPs influence tumour behaviour through theirability to cleave (and activate) growth factors, cell surface recep-tors, cell adhesion molecules and pro-apoptotic factors, therebypromoting tumour cell growth, migration and dissemination(Köhrmann et al., 2009).

The expression and activity of MMPs are tightly regulated atmultiple levels, including gene transcription, pro-enzyme activa-tion and the action of specific physiological inhibitors, known asthe tissue inhibitors of metalloproteinases (TIMPs) (Somervilleet al., 2003; Peterson et al., 2009). The TIMP family comprises fourdistinct members: TIMP-1, TIMP-2, TIMP-3 and TIMP-4 (Lambertet al., 2004). TIMP-2 plays a dual paradoxical function in regulatingMMP-2 activity (Duffy et al., 2000; Lambert et al., 2004), andwhereas TIMP-2 is an endogenous inhibitor of MMP-2, it is also in-volved in pro-MMP-2 activation (Toth et al., 2000).

CMTs have a very heterogeneous behaviour and histologicalevidence of malignancy does not invariably imply a malignant

A. Santos et al. / The Veterinary Journal 190 (2011) 396–402 397

clinical course (Lana et al., 2007). Therefore, it is important tosearch for prognostic factors other than histopathology to predictindividual tumour behaviour. Currently, there are no studiesregarding the possible associations of MMP-2 and TIMP-2 expres-sion and clinical outcome. Hence, the aims of this study were toevaluate the expression of MMP-2 and TIMP-2 by immunohisto-chemistry in benign and malignant CMTs and to investigate asso-ciation between their expression and clinical outcome.

Fig. 1. Canine benign mixed mammary tumour. Strong MMP-2 (a) and TIMP-2 (b)immunoreactivity in myoepithelial cells lining the BM of tubuloalveolar structures.IHC, bar 50 lm (a) and 100 lm (b).

Materials and methods

Specimens

Tumours (n = 118) were surgically removed from 118 female dogs, aged 5–16 years. Animals with malignant mammary tumours (MMTs) were enrolled in a2-year post-operative follow-up study with no adjuvant therapy. Samples of normalmammary tissue were obtained from eight bitches that were humanely euthanasedas part of the national stray dog control programme.

All specimens were fixed in 10% neutral buffered formalin for 48 h. Tumours61 cm were paraffin-embedded in one block, while larger tumours were cutsequentially at 5 mm intervals to provide a series of tissue blocks representativeof the entire lesion. After dehydration and embedding in paraffin, 3 lm sectionswere cut from each block. One section was stained with haematoxylin and eosin(HE) for diagnostic purposes and two representative blocks from tumours >1 cmwere selected for immunohistochemical studies. Tumours were evaluated indepen-dently by two observers (F.G. and I.A.) according to the criteria of the World HealthOrganization (WHO) for the histological classification of mammary tumours ofdomestic animals (Misdorp et al., 1999).

MMP-2 and TIMP-2 immunohistochemistry

Selected tumour sections adjacent to those used for HE staining were analysedby immunohistochemistry (IHC) using a polymer based system (Novocastra Novo-Link Max Polymer Detection System) according to the manufacturer’s instructions.Sections (3 lm) were de-waxed in xylene and rehydrated in graded alcohols. Anti-gen retrieval was performed by immersion in citrate buffer (10 mM, pH 6.0) for30 min at 100 �C in a water bath. Endogenous peroxidase activity was blocked with3% hydrogen peroxide in methanol for 10 min. Slides were incubated overnight at4 �C in a wet chamber with the anti-MMP-2 mouse monoclonal (clone A-Gel VC2,NeoMarkers) and anti-TIMP-2 rabbit polyclonal (H-140, Santa Cruz Biotechnology)antibodies, diluted 1:75 and 1:40, respectively, in Tris-buffered saline (TBS) solutionwith 5% bovine serum albumin (BSA). Colour was developed with a solution of 3,30-diaminobenzidine for 2 min at room temperature and the sections were counter-stained with haematoxylin.

For the negative controls, the anti-MMP-2 antibody was replaced with mouseIgG and the anti-TIMP-2 antibody with non-immune rabbit immunoglobulin.Negative controls did not show any staining. Positive controls consisted of sec-tions of human breast cancer known to express MMP-2 and TIMP-2 and of nor-mal canine mammary tissue (Papparella et al., 1997). The epidermis was usedas internal positive control. Positive controls showed diffuse granular cytoplasmiclabelling with strong intensity. Consulting GenBank indicated that canine MMP-2and TIMP-2 proteins are conserved among several species including human anddog. A BLAST analysis revealed strong homologies (>95%) between human anddog proteins as previously reported (Lana et al., 2000; Hirayama et al., 2002).Therefore, the antibodies were considered specifically recognizing as canineMMP-2 and TIMP-2.

The evaluation of MMP-2 and TIMP-2 expression was semi-quantitative andbased on the percentage of positively stained cancer cells (<10%; 10–25%; 25–50%; >50%) and staining intensity (weak; moderate; strong). The level ofexpression was considered low when tumours showed <25% of positive tumourcells with strong intensity, <50% with moderate intensity, or any percentage ofcells with weak intensity. Tumours were classified as demonstrating high levelof TIMP-2 and MMP-2 expression when they had >25% of strongly labelled cancercells or >50% of cells with moderate labelling intensity. Tumours were alsodescribed according to the pattern of expression of both enzymes: basal expres-sion (when the immunoreactivity was observed in normal myoepithelial cells lin-ing BM) vs. diffuse expression in neoplastic cells. Slides were examinedindependently by two observers (A.S. and A.M.) and when there was disagree-ment (<5% of the slides), a consensus was obtained using a multi-headmicroscope.

MMP-2 and TIMP-2 double-labelling immunofluorescence (DIF)

Six benign and 13 malignant CMTs were also analysed by DIF. Sections(3 lm) were de-waxed in xylene and rehydrated in graded alcohols. Antigen re-trieval was performed for both antibodies by immersion in boiling citrate buffer(30 min in water bath at 100 �C). Slides were then incubated for 20 min at room

temperature with monkey non-immune serum (Invitrogen) at a 1:5 dilution inphosphate buffered saline (PBS) containing 1% BSA. Slides were then incubatedovernight at 4 �C in a wet chamber with the first primary antibody, mouseanti-MMP-2 (diluted 1:15 in PBS with 1% BSA). After washing in PBS, slides wereincubated with Alexa Fluor 568 donkey anti-mouse IgG (diluted 1:100 in PBSwith 1% BSA; Invitrogen) for 60 min. Sections were washed three times in PBSand blocked with non-immune goat serum (diluted 1:5 in PBS with 1% BSA)for 20 min at room temperature. After incubation with the anti-TIMP-2 antibody(diluted 1:10 in PBS with 1% BSA; overnight at 4 �C in a wet chamber), slideswere washed in PBS and then incubated for 60 min at room temperature withFITC-conjugated goat anti-rabbit IgG (diluted 1:100 in PBS with 1% BSA; SouthernBiotech). Finally, slides were rinsed in PBS and mounted in Vectashield with DAPI(Vector).

Immunostained tissue sections were examined under a Laser ScanningConfocal Microscope (Olympus FV1000/IX81) with appropriate filters. Aftersequential scanning to assure the absence of cross-talk between channels, separateimages for Alexa Fluor 568 and FITC were captured at a 400� magnification andthen were merged to allow the visualization of MMP-2–TIMP-2 doubleimmunostaining.

Follow-up study

All animals were examined prior to surgery, 3 weeks after surgery and every3 months subsequently for a 2-year period. Each examination consisted of a com-plete physical examination, thoracic radiographs (three views) and completeabdominal ultrasound. Whenever necessary, additional examinations (e.g. aspira-tion cytology, excisional biopsy, skeletal radiography, etc.) were performed withthe owner’s consent. Complete necropsies were performed on all dogs that diedspontaneously or were euthanased. Suspected metastases were confirmedhistologically.

Table 1Patterns of metalloproteinase 2 (MMP-2) and tissue inhibitor of metalloproteinase 2 (TIMP-2) expression in benign and malignant canine mammary tumours.

Total number Number (%) of tumours with MMP-2 expression Number (%) of tumours with TIMP-2 expression

Diffuse Basal P Diffuse Basal P

Tumour type <0.001 <0.001Benign 24 8 (36.4) 14 (63.6) 11 (45.8) 13 (54.2)Malignant 94 85 (90.4) 9 (9.6) 79 (84.0) 15 (16.0)

398 A. Santos et al. / The Veterinary Journal 190 (2011) 396–402

Statistical analysis

Statistical analysis was performed using the PASW Statistics 18.0 statisticalpackage (IBM SPSS Statistics). Fisher’s exact test was used to evaluate the relation-ship between MMP-2 and TIMP-2 expression in CMTs and tumour type and distantmetastases. Overall survival (OS) was calculated from the date of tumour removalto the date of animal death/euthanasia due to tumour metastasis. Disease-free sur-vival (DFS) was calculated from the date of surgery to the date of detection of thefirst local recurrence or distant metastases. The Kaplan–Meier method was usedto compute OS and DFS times and to construct survival curves. The log-rank testwas used to analyse the significance of the differences between the groups. In theOS analysis, dogs were censored if and when they died for causes unrelated tomammary tumours, or were lost to follow up, or were still alive 2 years after sur-gery. In the DFS analysis, dogs were censored if and when they were lost during fol-low up, or died for causes unrelated to mammary tumours before developing signsof metastatic disease, or were free of distant metastases at 2 years after surgery. Thesignificance level was set at P < 0.05.

Fig. 2. Canine mammary tubulopapillary carcinomas. Neoplastic cells with strongMMP-2 (a) and TIMP-2 (b) expression. IHC, bar 20 lm (a) and 50 lm (b).

Fig. 3. Double-labelling immunofluorescence of a canine benign mixed mammarytumour. Myoepithelial cells demonstrating co-expression (yellow) of MMP-2 (red)and TIMP-2 (green). Magnification 400�.

Fig. 4. Double-labelling immunofluorescence of a canine micropapillary mammarycarcinoma. Large neoplastic epithelial cells showing co-expression (yellow) ofMMP-2 (red) and TIMP-2 (green). Magnification 400�.

Table 2TIMP-2 and MMP-2 expression in benign and malignant canine mammary tumours.

Total number Number (%) of tumours with MMP-2 expression Number (%) of tumours with TIMP-2 expression

Low High P Low High P

Tumour type NS <0.001Benign 24 14 (58.3) 10 (41.7) 21 (87.5) 3 (12.5)Malignant 94 40 (42.6) 54 (57.4) 44 (46.8) 55 (53.2)

NS, not significant.

Table 3Relationship between TIMP-2 and MMP-2 expression in malignant canine mammary tumours and presence of distant metastases.

Total number Number (%) of tumours with MMP-2 expression in cancer cells Number (%) of tumours with TIMP-2 expression in cancer cells

Low High P Low High P

Distant metastasesa NS 0.015Absent 51 18 (35.3) 33 (64.7) 30 (58.8) 21 (41.2)Present 24 13 (54.2) 11 (45.8) 7 (29.2) 17 (70.8)

NS, not significant.a Twelve dogs are still being followed; seven dogs were lost to follow up.

Table 4Relationship between TIMP-2 and MMP-2 expression in canine mammary tumours.

TIMP-2 expression Total number Number (%) of tumours with MMP-2expression

Low High P

Low 65 30 (46.2) 35 (53.8) NSHigh 53 24 (45.3) 29 (54.7)

NS, not significant.

A. Santos et al. / The Veterinary Journal 190 (2011) 396–402 399

Results

Benign tumours consisted of 6 simple adenomas, 11 complexadenomas and 7 benign mixed tumours. The group of MMT in-cluded 31 solid, 21 complex, 20 tubulopapillary, 2 micropapillary,2 mucinous, 2 anaplastic and 1 spindle cell carcinomas, 13 carcino-sarcomas and 2 carcinomas in benign tumours.

Normal mammary glands exhibited immunoreactivity for bothenzymes only in myoepithelial cells. Neoplastic cells showed dif-fuse cytoplasmic immunoreactivity with variations in the numberof positive cells (epithelial, myoepithelial and sarcomatous) and inlabelling intensity. Tumour-associated macrophages demonstratedconsistently strong reactivity.

Benign tumours more often presented basal patterns of expres-sion of MMP-2 (Fig. 1a and Table 1) and TIMP-2 (Fig. 1b and Table1) in contrast to malignant tumours, which showed mainly diffuseexpression in neoplastic cells (Fig. 2a and b). The DIF analysis dem-onstrated co-localization of TIMP-2 and MMP-2 with strong inten-sity in >90% of cells lining BM in benign tumours (Fig. 3). Largecancer cells with atypical shapes showed strong MMP-2 andTIMP-2 expression. The DIF analysis demonstrated co-expressionof MMP-2 and TIMP-2 in all undifferentiated cancer cells in 10/13 analysed MMTs (Fig. 4).

Increased TIMP-2 expression in neoplastic cells was signifi-cantly associated with tumour type (Table 2) and distant metasta-ses (Table 3) in contrast with MMP-2 expression. There was nosignificant association between the neoplastic cells expression ofMMP-2 and TIMP-2 in CMTs (P = 0.537; Table 4).

The OS analysis was performed in 70 patients and the DFS studyin 74 patients with MMTs. Increased expression of TIMP-2 wassignificantly associated with worse OS (Fig. 5 and Table 5) and

reduced DFS (Fig. 6 and Table 6). There were no statistical signifi-cant relationships between MMP-2 expression and survival param-eters (Tables 5 and 6).

Discussion

The ability to degrade ECM components is an important deter-minant of the invasive phenotype of malignant tumours (Pupaet al., 2002). Therefore, this work intended to study the patternsof MMP-2 and TIMP-2 expression in normal and neoplastic caninemammary tissues and to investigate any relationship betweenexpression and survival.

In normal mammary glands and in benign tumours, TIMP-2 andMMP-2 were mainly expressed in normal myoepithelial cells liningthe BM of tubuloalveolar glands and ducts, in contrast to malignanttumours where diffuse cytoplasm immunoreactivity was predom-inantly observed in epithelial cells, as previously described forMMP-2 in CMTs by Papparella et al. (1997). The basal expressionof both molecules may reflect the MMP-TIMP equilibrium neces-sary for normal tissue turnover (Papparella et al., 1997). In fact,in normal cells, the role of TIMP-2 is likely to regulate MMPs activ-ity during tissue turnover, growth and development (Bee et al.,2000). However, this balance may be lost during carcinogenesis,which may explain the increased expression in a diffuse patternof both enzymes in MMTs.

In the present study, there was no significant association be-tween MMP-2 expression and prognostic factors (malignant tu-mour type, development of distant metastases, OS and DFS).Papparella et al. (2002) reported higher MMP-2 expression inmalignant CMTs when compared to normal glands, which is inagreement with our results. Malignant CMTs tended to show high-er levels of MMP-2 expression than benign tumours, as previouslydescribed (Hirayama et al., 2002); however this association wasnot statistically significant.

Studies of MMP-2 activity in canine (Kawai et al., 2006) and hu-man (Figueira et al., 2009) mammary neoplasia demonstrated a sig-nificant direct correlation with tumour malignancy. However, thisis a controversial topic, since other studies in canine (Papparellaet al., 2002) and human breast cancer could not find any associationwith tumour aggressive features, such as histological grade(Nakopoulou et al., 2003; Talvensaari-Mattila et al., 2003; Jobimet al., 2009; Köhrmann et al., 2009), development of metastases

Table 5Association between TIMP-2 and MMP-2 expression in canine mammary tumoursand overall survival (OS in months).

Level ofexpression

Number ofdogsa

Dead/euthanased

Alive Mean OS(±SEM)

P

TIMP-2 0.005Low 31 4 27 21.91 (1,00)High 39 17 22 17.97 (1.35)

MMP-2 NSLow 28 11 17 19.04 (1.39)High 42 10 32 20.24 (1.19)

NS, not significant.a Twelve dogs are still being followed; seven dogs were lost to follow up before

detection of distant metastases; five dogs were lost to follow up after developmentof distant metastases.

Fig. 5. Kaplan–Meier overall survival (OS) curves following surgical removal of malignant mammary tumours grouped based on TIMP-2 expression.

400 A. Santos et al. / The Veterinary Journal 190 (2011) 396–402

(Nakopoulou et al., 2003; Vizoso et al., 2007; González et al., 2009)and reduced DFS (Pellikainen et al., 2004; Vizoso et al., 2007;González et al., 2009).

There are several possible reasons for the absence of relation-ships between MMP-2 expression and tumour aggressiveness.One is the inability of the anti-MMP-2 antibody to distinguish be-tween the active and latent forms of the enzyme, although theexpression of active MMP-2 (rather than total MMP-2 expression)may better relate to invasive characteristics. This could also ex-plain why our study failed to demonstrate a relationship betweenMMP-2 and TIMP-2 levels. Furthermore, MMP-2 expression bycancer cells may occur mainly at the early stages of carcinogenesis(Liotta and Kohn, 2001). It is possible that CMTs, when clinicallydetected, are already at an advanced stage of the carcinogenic pro-cess, explaining the absence of significant associations betweenMMP-2 expression in cancer cells and clinical outcome.

An interesting finding was the demonstration of MMP-2 andTIMP-2 co-localization in highly undifferentiated cancer cells,which led us to hypothesize that those cells could be expressingthe active form of MMP-2, since MMP-2 needs to establish a linkwith TIMP-2 to become activated (Bee et al., 2000).

The most remarkable observation in this study was that MMTsshowed significantly higher expression of TIMP-2 than benign tu-mours, suggesting that TIMP-2 is involved in the acquisition ofmalignant phenotype. To the best of our knowledge, this is the firstIHC study of TIMP-2 expression in CMTs. Consistent with our re-sults, Kawai et al. (2006), using reverse zymography, reported asignificant increased activity of TIMP-2 in canine mammary carci-nomas when compared to adenomas.

Traditionally considered as suppressors of cancer growth andmetastasis, TIMPs have been found to have paradoxical effectsin tumorigenesis (Toth et al., 2000; Lambert et al., 2004). Evi-dence has emerged showing that TIMPs are multifunctional pro-teins involved in the induction of proliferation and inhibition ofapoptosis, contributing to tumour initiation and growth (Jianget al., 2002; Lambert et al., 2004). Moreover, TIMPs may also pro-mote angiogenesis either directly by up-regulating VEGF (Yoshiji

et al., 1998; Hajitou et al., 2001) or indirectly by inhibitingMMPs that are involved in the production of angiogenic inhibi-tors such as endostatin and angiostatin (reviewed by Duffy etal. (2000)).

Our study demonstrates a significant association betweenTIMP-2 over-expression and the development of distant metasta-ses, a finding also described in human breast cancer (Ree et al.,1997; Vizoso et al., 2007; Gonzalez et al., 2010). Figueira et al.(2009) also demonstrated significantly higher TIMP-2 mRNA levelsin highly invasive and metastatic breast cancer cell lines whencompared to less aggressive ones, supporting a contribution ofTIMP-2 in aggressive tumour behaviour. Our results support thehypothesis that high levels of TIMP-2 in canine MMTs facilitate tu-mour invasion. Mechanisms for this effect may be reduction ofapoptosis, stimulation of angiogenesis (Jiang et al., 2002) and acti-vation of pro-MMP-2 (Toth et al., 2000; Somerville et al., 2003).Furthermore, several studies suggested that the growth modulat-ing properties of TIMP-2 are independent of any MMP inhibitoryeffect; rather, they seem to be a direct concentration-dependentcellular effect, mediated by specific receptors that transduce intra-cellular signalling to induce tumour proliferation (Lambert et al.,2004).

Fig. 6. Kaplan–Meier disease-free survival (DFS) curves following surgical removal of malignant mammary tumours grouped based on TIMP-2 expression.

Table 6Association between TIMP-2 and MMP-2 expressions in canine mammary tumoursand disease-free survival (DFS in months).

Level ofexpression

Number ofdogsa

With distantmetastases

Mean DFS(±SEM)

P

TIMP-2 0.017Low 33 6 20.27 (1.41)High 41 18 15.40 (1.58)

MMP-2 NSLow 31 13 15.87 (1.77)High 43 11 18.99 (1.37)

NS, not significant.a Twelve dogs are still being followed; seven dogs were lost to follow up before

detection of distant metastases; it was not possible to determine the time fordevelopment of distant metastases in one dog.

A. Santos et al. / The Veterinary Journal 190 (2011) 396–402 401

Conclusions

The high level of TIMP-2 expression observed in MMTs wasassociated with increased incidence of distant metastases and poorclinical outcome, suggesting that TIMP-2 may be a valuable prog-nostic factor in dogs with mammary tumours.

Conflict of interest statement

None of the authors of this paper has a financial or personalrelationship with other people or organisations that could inappro-priately influence or bias the content of the paper.

Acknowledgments

The authors would like to express their gratitude to ProfessorPaulo Correia de Sá and Doctor Fátima Ferreirinha for the assis-tance in the manipulation of the laser confocal microscope. Grantsupport: Fundação para a Ciência e Tecnologia, Project SFRH/BD/42363/2007.

References

Bee, A., Barnes, A., Jones, M.D., Robertson, D.H.L., Clegg, P.D., Carter, S.D., 2000.Canine TIMP-2: purification, characterization and molecular detection. TheVeterinary Journal 160, 126–134.

Duffy, M.J., Maguire, T.M., Hill, A., McDermott, E., O’Higgins, N., 2000.Metalloproteinases: role in breast carcinogenesis, invasion and metastasis.Breast Cancer Research 2, 252–257.

Figueira, R.C.S., Gomes, L.R., Neto, J.S., Silva, F.C., Silva, I.D.C.G., Sogayar, M.C., 2009.Correlation between MMPs and their inhibitors in breast cancer tumor tissuespecimens and in cell lines with different metastatic potential. BMC Cancer 9,20.

González, L.O., Corte, M.D., Junquera, S., González-Fernández, R., del Casar, J.M.,García, C., Andicoechea, A., Vázquez, J., Pérez-Fernández, R., Vizoso, F.J., 2009.Expression and prognostic significance of metalloproteases and their inhibitorsin luminal A and basal-like phenotypes of breast carcinoma. Human Pathology40, 1224–1233.

Gonzalez, L.O., Junquera, S., Del Casar, J.M., González, L., Marín, L., González-Reyes,S., Andicoechea, A., González-Fernández, R., González, J.M., Pérez-Fernández, R.,Vizoso, F.J., 2010. Immunohistochemical study of matrix metalloproteinasesand their inhibitors in pure and mixed invasive and in situ ductal carcinomas ofthe breast. Human Pathology 41, 980–989.

Hajitou, A., Sounni, N., Devy, L., Grignet-Debrus, C., Lewalle, J., Li, H., Deroanne, C.F.,Lu, H., Colige, A., Nusgens, B.V., Frankenne, F., Maron, A., Yeh, P., Perricaudet, M.,Chang, Y., Soria, C., Calberg-Bacq, C.M., Foidart, J.M., Noël, A., 2001. Down-regulation of vascular endothelial growth factor by tissue inhibitor ofmetalloproteinase-2: effect on in vivo mammary tumor growth andangiogenesis. Cancer Research 61, 3450–3457.

Hirayama, K., Yokota, H., Onai, R., Kobayashi, T., Kumata, T., Kihara, K., Okomoto, M.,Sako, T., Nakade, T., Izumisawa, Y., Taniyama, H., 2002. Detection of matrixmetalloproteinases in canine mammary tumours: analysis by immuno-histochemistry and zymography. Journal of Comparative Pathology 127, 249–256.

Hojilla, C.V., Wood, G.A., Khokha, R., 2008. Inflammation and breast cancer:metalloproteinases as common effectors of inflammation and extracellularmatrix breakdown in breast cancer. Breast Cancer Research 10, 205.

Jiang, Y., Goldberg, I.D., Shi, Y.E., 2002. Complex roles of tissue inhibitors ofmetalloproteinases in cancer. Oncogene 21, 2245–2252.

Jobim, F.C., Xavier, N.L., Uchoa, D.M., Cruz, D.B., Saciloto, M., Chemello, N.,Schwartsmann, G., 2009. Prevalence of vascular endothelial growth factor,matrix metalloproteinases and tissue inhibitors of metalloproteinases inprimary breast cancer. Brazilian Journal of Medical and Biological Research42, 979–987.

Kawai, K., Uetsuka, K., Doi, K., Nakayama, H., 2006. The activity of matrixmetalloproteinases (MMPs) and tissue inhibitors of metalloproteinases(TIMPs) in mammary tumors of dogs and rats. Journal of Veterinary MedicalScience 68, 105–111.

Köhrmann, A., Kammerer, U., Kapp, M., Dietl, J., Anacker, J., 2009. Expression ofmatrix metalloproteinases (MMPs) in primary human breast cancer and breastcancer cell lines: new findings and review of the literature. BMC Cancer 9, 188.

402 A. Santos et al. / The Veterinary Journal 190 (2011) 396–402

Lambert, E., Dassé, E., Haye, B., Petitfrère, E., 2004. TIMPs as multifacial proteins.Critical Reviews in Oncology/Hematology 49, 187–198.

Lana, S.E., Ogilvie, G.K., Hansen, R.A., Powers, B.E., Dernell, H.W., Withrow, S.J., 2000.Identification of matrix metalloproteinases in canine neoplastic tissue.American Journal of Veterinary Research 61, 111–114.

Lana, S.E., Rutteman, G.R., Withrow, S.J., 2007. Tumors of the mammary gland. In:Withrow, S.J., MacEwen, E.G. (Eds.), Small Animal Clinical Oncology. SaundersElsevier, St. Louis, pp. 619–636.

Liotta, L.A., Kohn, E.C., 2001. The microenvironment of tumour–host interface.Nature 411, 375–379.

Misdorp, W., Else, R.W., Helmén, E., Lipscomb, T.P., 1999. Histological classificationof the mammary tumours of the dog and the cat. In: Shulman, Fl. (Ed.), WorldHealth Organization International Histological Classification of Tumours ofDomestic Animals, Second Series, vol. 7. Armed Forces Institute of Pathology,Washington, DC, pp. 16–29.

Nakopoulou, L., Tsirmpa, I., Alexandrou, P., Louvrou, A., Ampela, C., Markaki, S.,Davaris, P.S., 2003. MMP-2 protein in invasive breast cancer and the impact ofMMP-2/TIMP-2 phenotype on overall survival. Breast Cancer Research andTreatment 77, 145–155.

Papparella, S., Restucci, B., Maiolino, P., De Vico, G., 1997. Immunohistochemicaldistribution of type IV collagenase in normal, dysplastic and neoplastic caninemammary gland. Journal of Comparative Pathology 117, 277–282.

Papparella, S., Restucci, B., Paciello, O., Maiolino, P., 2002. Expression of matrixmetalloproteinase-2 (MMP-2) and the activator membrane type 1 (MT1-MMP)in canine mammary carcinomas. Journal of Comparative Pathology 126, 271–276.

Pellikainen, J.M., Ropponen, K.M., Kataja, V.V., Kellokoski, J.K., Eskelinen, M.J.,Kosma, V.M., 2004. Expression of matrix metalloproteinase (MMP)-2 and MMP-9 in breast cancer with special reference to activator protein-2, HER2, andprognosis. Clinical Cancer Research 10, 7621–7628.

Peterson, N.B., Beeghly-Fadiel, A., Gao, Y., Long, J., Cai, Q., Shu, X., Zheng, W., 2009.Polymorphisms in tissue inhibitors of metalloproteinases-2 and -3 and breastcancer susceptibility and survival. International Journal of Cancer 15, 844–850.

Pupa, S.M., Ménard, S., Forti, S., Tagliabue, E., 2002. New insights into the role ofextracellular matrix during tumour onset and progression. Journal of CellularPhysiology 192, 259–267.

Ree, A.H., Florenes, V.A., Berg, J.P., Maelandsmo, G.M., Nesland, J.M., Fodstad, O.,1997. High levels of messenger RNAs for tissue inhibitors of metalloproteinases(TIMP-1 and TIMP-2) of primary breast carcinomas are associated withdevelopment of distant metastases. Clinical Cancer Research 3, 1623–1628.

Somerville, R.P.T., Oblander, S.A., Apte, S.S., 2003. Matrix metalloproteinases: olddogs and new tricks. Genome Biology 4, 216.

Talvensaari-Mattila, A., Pääkkö, P., Turpeenniemi-Hujanen, T., 2003. Matrixmetalloproteinase-2 (MMP-2) is associated with survival in breast carcinoma.British Journal of Cancer 89, 1270–1275.

Toth, M., Bernardo, M., Gervasi, D.C., Soloway, P.D., Wang, Z., Bigg, H.F., Overall, C.M.,DeClerck, Y.A., Tschesche, H., Cher, M.L., Brown, S., Mobashery, S., Fridman, R.,2000. Tissue inhibitor of metalloproteinase (TIMP)-2 acts synergistically withsynthetic matrix metalloproteinase (MMP) inhibitors but not with TIMP-4 toenhance the membrane type 1–MMP-dependent activation of pro-MMP-2.Journal of Biological Chemistry 52, 41415–41423.

Vizoso, F.J., González, L.O., Corte, M.D., Rodríguez, J.C., Vázquez, J., Lamelas, M.L.,Junquera, S., Merino, A.M., García-Muñiz, J.L., 2007. Study of matrixmetalloproteinases and their inhibitors in breast cancer. British Journal ofCancer 96, 903–911.

Yoshiji, H., Harris, S.R., Raso, E., Gomez, D.E., Lindsay, C.K., Shibuya, M., Sinha, C.C.,Thorgeirsson, U.P., 1998. Mammary carcinoma cells over-expressing tissueinhibitor of metalloproteinases-1 show enhanced vascular endothelial growthfactor expression. International Journal of Cancer 75, 81–87.