differential expression of matrix metalloproteinase...

Differential Expression of Matrix Metalloproteinase (MMP)-2,MMP-9, and Membrane Type 1-MMP in Hepatocellular andPancreatic Adenocarcinoma: Implications for TumorProgression and Clinical Prognosis1

Marko Ma atta,2 Ylermi Soini, Annikki Liakka,and Helena Autio-HarmainenDepartment of Pathology, University of Oulu, 90014 Oulu, Finland

ABSTRACTIn the present study, we usedin situ hybridization to

study 36 primary hepatocellular carcinomas (HCCs) and 35pancreatic adenocarcinomas to analyze the expressions ofmembrane-type 1 matrix metalloproteinase (MT1-MMP),MMP-2, and MMP-9 mRNAs. In HCCs, MT1-MMP mRNAwas mainly expressed by cancer cells and to a lesser extentby stromal cells. MMP-2 mRNA was expressed predomi-nantly by cells of tumor stroma, whereas MMP-9 mRNAwas seen mainly in neoplastic epithelial cells. In pancreaticadenocarcinomas, MT1-MMP and MMP-9 mRNA wereseen at moderate levels both in cancer and in stromal cells,whereas MMP-2 mRNA was predominantly expressed bythe tumor stroma. Antigens of MMP-2, MMP-9, and MT1-MMP immunolocalized to the neoplastic epithelium and tothe stromal cells in both tumor types. In gelatin zymogra-phy, increased amounts of latent and active MMP-2 werefound in tumor samples of HCC as compared with adjacentnontumorous liver tissue. On the other hand, the latent formof MMP-9 was found in almost equal amounts both in tumorand normal liver samples, but its active form was presentonly in HCC.

Expression of MT1-MMP mRNA had a tendency to beassociated with a lower degree of differentiation in HCC, butsuch association was not noticed in pancreatic tumors. Cor-relation to the clinical data showed that MT1-MMP expres-sion had a strong statistical association with a poor outcomeof patients (P< 0.01). A similar tendency was also observedin pancreatic adenocarcinomas, but the association did notreach statistical significance. MMP-2 and MMP-9 mRNAexpression did not have significant correlation with progno-sis. The results of this study support the previous suggestions

of the importance of MT1-MMP for malignant growth andindicate that increased MT1-MMP mRNA expression bytumor cells in HCCs and pancreatic adenocarcinomas mayhave prognostic significance.

INTRODUCTIONMMPs3 are a group of zinc-dependent endopeptidases with

different substrate specificities (1). They share many structuraland functional properties and are known to take part in extra-cellular matrix remodeling and can be inhibited by TIMPs.

In cancer research, much interest has been devoted recentlyto a gelatinase subgroup of MMPs that include MMP-2 (gela-tinase A,Mr 72,000 type IV collagenase) and MMP-9 (gelatin-ase B,Mr 92,000 type IV collagenase). This is partly because oftheir ability to degrade type IV collagen, the major structuralcomponent of BMs and because of their assumed important rolein cellular invasion (2). MT1-MMP, the first member of a morerecently established group of MMPs containing a membrane-spanning sequence, has been shown to have an important role inMMP-2 activation in cell membranes (3, 4). By immunohisto-chemistry, MMP-2 has been shown to be localized in carcino-mas to the neoplastic epithelial cells (5–7), butin situ hybrid-ization, on the other hand, has shown that most of the expressionof mRNA for MMP-2 occurs in fibroblasts and endothelial cellsof the tumor stroma (6, 7). In numerous studies, expression ofMMP-2 has been shown to be related to invasive phenotype andmetastatic potential of tumor cells (8, 9).

HCCs are relatively uncommon in Western countries butcommon in parts of Africa and the Far East. The most importantetiological factors implicated in HCC are hepatitis B and C,abuse of alcohol, and exposure to aflatoxin. Increased preva-lence of HCC is also associated with cirrhosis, and generally theclinical prognosis of HCC is very poor.

Pancreatic adenocarcinoma is another aggressive tumortype that frequently develops early regional and distant metas-tases and therefore has an unfavorable prognosis. Previous stud-ies have shown that many MMPs and TIMPs are overexpressedin these two tumor types (10–14). However, results from thesestudies have not provided a systematic concept of how MMP-2,MMP-9, and MT1-MMP are expressed in mRNA as well asantigen levels, and moreover, what the clinical aspect of MMPsin these malignancies with poor prognosis is. Therefore, wecollected a tumor series that comprised primary HCCs and

Received 8/30/99; revised 3/17/00; accepted 3/27/00.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 This study is financially supported by the Cancer Foundation ofNorthern Finland.2 To whom requests for reprints should be addressed, at Department ofPathology, P. O. Box 5000 (Aapistie 5), University of Oulu, 90014Oulu, Finland. Fax: 358-8-537-5953.

3 The abbreviations used are: MMP, matrix metalloproteinase; MT1,membrane type 1; HCC, hepatocellular carcinoma; TIMP, tissue inhib-itor of matrix metalloproteinase; BM, basement membrane.

2726Vol. 6, 2726–2734, July 2000 Clinical Cancer Research

Research. on July 12, 2018. © 2000 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

pancreatic adenocarcinomas to study systematically MMP-2,MMP-9, and MT1-MMP expressions in these tumors. Becauseof some discrepancies in previous results, we wanted to clarifywhich cells are responsible for the production of these mRNAsand corresponding antigens and whether there are any differ-ences in expression between tumors of high and poor differen-tiation. Because there are some previous reports with othermalignancies that indicate the association of MMPs, especiallythat of MT1-MMP to poor prognosis of patients (15, 16), wealso correlated our findings with the clinical outcome of thepatients.

MATERIALS AND METHODSMaterials. Samples from 36 primary HCCs from pa-

tients who underwent partial hepatectomy during 1985–1997were retrieved from the files of the Department of Pathology,Oulu University Central Hospital. There were 18 females and 18males with a mean age of 64 years (range, 7–83 years). Theaverage tumor size was 5.1 cm (range, 2–15 cm). There were 13tumors of differentiation grade I, 13 of grade II, and 10 of gradeIII. Similarly, samples from 35 pancreatic ductal adenocarcino-mas were collected from years 1985–1994. Of the tumors, 13were of grade I, 15 of grade II, and 7 of grade III. The averagetumor size was 3.4 cm (range, 1–7 cm). The mean age of thepatients was 64.6 years (range, 53–73 years). Of the patients, 25were females and 10 were males. Clinical data, including age,sex, tumor size, survival time after surgical operation, and causeof death, are presented in Tables 1 and 2. The tumor materialwas fixed in 10% neutral formalin and embedded in paraffin.The diagnoses of tumors were based on a light microscopicexamination of H&E-stained sections. Grading was done ac-cording to the criteria of the WHO (17, 18).

RNA Probes for Hybridization. A 635-bp ScaI-SacIfragment of the K-191 cDNA clone coding for the humanMMP-2 (19) was ligated into the M13 polylinker site of pSP64and pSP65 vectors (Promega Corp., Madison, WI). A 574-bpEcoRI-HindIII fragment of human MMP-9 cDNA cloneK-174.1 (20) was subcloned into the pGEM 4Z vector (Pro-mega). A 420-bp fragment from nucleotide 218 to nucleotide638 of the human MT1-MMP (MMP-14) cDNA was amplifiedby PCR from human placental cDNA (21) and ligated intoBluescriptSK1vector (Stratagene, La Jolla, CA) to generatepBL420.

In Situ Hybridization. For in situ hybridization, thesections were treated with 0.2M hydrochloric acid for 20 min atroom temperature and washed in diethylpyrocarbonate-treatedwater for 5 min. Proteinase K (1 mg/ml; Roche Diagnostics,Indianapolis, IN) treatment was carried out for 30 min at 37°C,and the reaction was terminated with 0.2% glycine in PBS.Sections were fixed with 4% paraformaldehyde in PBS for 20min and washed in PBS. Acetylation was done in 0.25–0.5%acetic anhydride in 0.1M triethanolamine for 10 min. Sectionswere washed in PBS and air-dried for 1–2 h at room tempera-ture. Incubation in prehybridization mixture [10 mM DTT (Sig-ma, St. Louis, MO), 10 mM Tris-HCl, 10 mM NaPO4, 5 mM

EDTA (Sigma), 0.3M NaCl, 1 mg/ml yeast tRNA, deionizedformamide 50%, and dextran sulfate 10% (w/v); 0.02% (w/v)Ficoll (Amersham Pharmacia Biotech, Uppsala, Sweden),

0.02% (w/v) polyvinylpyrrolidone, and 0.02 mg/ml BSA] wascarried out for 2 h. Thereafter, the sections were washed in PBSand dehydrated. For hybridization, the probes were first dena-tured by boiling them for 1 min and then placed on ice. Three3106 cpm of the35S-labeled antisense or sense probe in 40ml ofprehybridization buffer was applied on each section, and thehybridization was carried out at 50°C overnight. The posthy-bridization washes were as follows: twice at 50°C for 1 h inprehybridization mixture except for dextran sulfate and tRNA,15 min in 0.5M NaCl in 10 mM Tris-HCl, 1 mM EDTA at 37°C,30 min in 0.5M NaCl containing 40mg/ml RNase A (Sigma) at37°C, 15 min in 0.5M NaCl at 37°C, 15 min twice in 23 SSC,and 15 min twice in 13SSC at 50°C. The sections were thendehydrated in ethanol containing 300 mM ammonium acetateand air-dried at room temperature for 1 h. For autoradiography,the slides were dipped in NTB-2 film emulsion (Kodak, New

Table 1 Tumor grade; patient’s age, sex, tumor size, and follow-up(months); and MMP-2, MMP-9, and MT1-MMP mRNA expression

in individual cases of HCC

CaseaMMP-2st/cab

MMP-9st/cab

MT1-MMPst/cab

1. I/65/F/3/18c 2/2 2/2 2/22. I/53/M/2/27c 2/2 2/11 2/23. I/69/M/15/5c 11/1 2/2 2/24. I/58/F/3/15c 2/2 2/2 2/25. I/72/F/5/2c 11/2 11/11 2/16. I/69/M/3/6c 2/2 2/1 1/27. I/68/F/9/20c 2/2 1/11 1/28. I/54/F/2/2c 2/2 2/2 2/29. I/74/M/5/8c 1/1 2/2 2/2

10. I/73/F/5/28c 2/1 2/2 2/211. I/63/F/5/45d 2/2 2/2 2/212. I/65/M/2/40c 2/2 2/2 2/213. I/58/F/4/108d 2/2 2/2 2/214. II/60/M/3/1e 2/2 2/2 2/215. II/67/F/6/12c 2/2 2/2 2/116. II/59/M/3/55e 2/2 2/2 2/217. II/55/M/7/6c 1/1 2/111 2/218. II/75/F/5/120d 1/2 2/2 2/219. II/69/F/5/102c 2/2 2/2 2/220. II/61/M/5/12c 11/1 2/2 1/221. II/53/F/4/48d 1/2 2/1 1/222. II/75/F/4/54d 11/2 1/1 2/223. II/83/M/5/23d 2/2 2/11 2/224. II/70/M/8/7c 2/2 2/2 2/125. II/62/F/4/15c 2/2 2/1 2/226. II/58/M/2/36d 2/2 2/2 2/227. III/76/F/4/6c 2/2 2/1 2/11128. III/79/M/8/2.5e 2/2 2/1 2/11129. III/55/M/3/8c 1/1 1/1 11/230. III/64/M/5/108d 1/2 2/2 2/231. III/66/M/7/3c 1/2 2/2 2/232. III/45/F/8/2c 111/1 2/2 2/133. III/7/M/11/12c 1/1 2/2 2/234. III/66/F/5/1e 2/2 2/2 2/235. III/72/M/4/14d 2/2 2/2 2/236. III/77/F/5/1c 2/2 2/2 2/11

a The case number is followed by grade, age, sex, tumor size, andfollow-up in months.

b st, stromal cell; ca, carcinoma cell.c Died of cancer.d Still alive.e Died of other disease.

2727Clinical Cancer Research



York, NY) and placed in light-tight boxes for 10–14 days. Theslides were developed in D-19 developer (Kodak) fixed inAgefix (Kodak) and counterstained with H&E.

Evaluation of the in Situ Hybridization Results. Theresults of thein situ hybridization experiment were scored byusing a semiquantitative scale as follows; negative (2), nosignal distinguishable from the background labeling; weak (1),signal was slightly increased or only a few cells were observedto be labeled; moderate (11), distinct labeling was observed ina moderate number of cells; and strong (111), distinct andabundant labeling observed in the majority of the cells. Thescoring was done independently by two observers.

Immunohistochemistry. Sections, 4mm thick, were in-cubated overnight at 4°C with a monoclonal antibody againstMMP-2, MMP-9 (gifts from Dr. Matti Hoyhtya, BiocenterOulu, Finland), and MT1-MMP (21). They were then incubatedwith a biotinylated antimouse antibody (Dako, Glostrup A/S,Copenhagen, Denmark) for 30 min, followed by the avidin-

biotin-peroxidase complex, and the substrate solution (3,3-diaminobenzidine tetrahydrochloride in H2O2 in Tris buffer, pH7.4; Sigma) for 10 min. For control stains, PBS was used insteadof the primary antibody.

Gelatin Zymography. To assess the MMP-2 andMMP-9 activity in HCC tissues, gelatin zymography modifiedfrom Heussen and Dowdle (22) was performed. Fiftymg oftumor tissue obtained from HCC and an equal amount of adja-cent nontumor tissue were manually homogenized. Thirty-mlaliquots of both tumor and nontumor samples were dissolved inthe SDS sample buffer in the absence of reducing agents withoutboiling. The samples were then run in SDS-PAGE in gels thatalso contained copolymerized gelatin (1.5 mg/ml). After elec-trophoresis, the gels were washed twice for 15 min in buffercontaining 5 mM calcium chloride, 1mM zinc chloride, 50mM-Tris-HCl (pH 7.6), supplemented with 2.5% Triton X-100.Finally, the gel was incubated for 16 h in same buffer alsocontaining 0.2M NaCl, 0.02% sodium trinitrate, and 10 mM

calcium dichloride at 37°C. The enzyme activity was visualizedby staining the gels with a solution containing 50% methanol,10% acetic acid, and 0.1% Coomassie Brilliant Blue anddestaining in 10% methanol and 10% acetic acid.

Statistical Analysis. Clinical data were collected fromthe patients’ clinical records, and survival times were counted(months) from the date of partial hepatectomy or pancreatec-tomy. Statistical analysis was performed by the Kaplan-Meiermethod with log-rank analysis. HCCs were analyzed comparinggroups negative (2) for mRNA expression to positively ex-pressing (1through111) groups. Because there were onlyfew cases of pancreas adenocarcinomas that did not express anymRNAs investigated, we combined negative (2) and low ex-pression (1) groups and compared those with moderate (11)and strong (111) expressions. The results from survival anal-ysis were expressed by reportingPs, and values, 0.05 wereregarded as significant.

RESULTSPositive signals inin situ hybridization experiments were

scored by using a semiquantitative scale. The results for HCCsare presented in Table 1 and for pancreatic adenocarcinomas inTable 2.

MT1-MMP mRNA Expression. mRNA for MT1-MMPwas detected in 12 of 36 (33%) HCCs. Of these, the expressionwas seen exclusively in carcinoma cells in seven tumors. In fivecases, there was also expression in fibroblasts and, more weakly,in endothelial cells. The stromal signal was generally weak andmost distinct in areas of abundant fibrosis and was often local-ized to cells adjacent to islands of tumor cells. In some areas,positive labeling was also seen in epithelial cells of proliferat-ing, but not of normal, bile ducts. In carcinoma cells of HCCs,the MT1-MMP mRNA expression was independent from that instroma, and no clear coexpression with fibroblasts was noticed.The expression level in carcinoma cells varied from negative tostrong positivity and had a tendency to be associated with a lowdegree of differentiation; 5 of 10 of grade III tumors showedpositive labeling (Fig. 1a).

In pancreatic adenocarcinomas, MT1-MMP mRNA ex-pression was seen in 34 of 35 (97%) cases. There was only one

Table 2 Tumor grade; patient’s age, sex, tumor size, and follow-up(months); and MMP-2, MMP-9, and MT1-MMP mRNA expression

in individual cases of pancreatic adenocarcinoma

CaseaMMP-2st/cab

MMP-9st/ca

MT1-MMPst/ca

1. I/67/F/3/52c 1/2 2/2 1/112. I/60/F/2.5/14c 2/2 11/11 1/13. I/66/F/3/6c 11/1 11/11 1/114. I/63/M/5/20c 1/2 1/1 11/15. I/68/F/3.5/38c ND ND 1/16. I/59/F/2/15c 11/1 1/1 1/117. I/63/F/3/11c ND ND 1/118. I/57/M/2.5/5c 11/1 11/1 2/19. I/67/F/1.5/2c ND 1/1 1/11

10. I/62/F/1/28c 1/2 ND 11/111. I/59/M/4/50c 11/2 1/1 1/212. I/53/M/3/62d ND ND 1/1113. I/65/F/3/52c 11/2 1/2 2/114. II/73/F/4/75d 11/2 ND 1/1115. II/67/F/4/4c 1/2 1/1 1/116. II/61/M/5/7c 11/1 11/11 1/1117. II/64/F/3/29c ND ND 1/1118. II/68/F/2/1c 111/11 1/1 11/1119. II/72/M/2/5c ND ND 1/1120. II/64/F/4/14c 111/1 11/1 1/1121. II/61/F/3/19c 1/2 1/1 11/122. II/64/F/3/9c 11/1 2/1 11/1123. II/58/M/1/3c 111/2 2/11 111/1124. II/60/F/2.5/27c ND ND 1/125. II/64/M/2.5/36c 111/1 1/11 11/126. II/67/F/7/29c ND ND 1/127. II/68/F/1.5/156d 1/1 2/1 1/128. II/62/F/3/14c 2/2 11/1 1/129. III/61/M/6/3c 111/1 2/2 1/130. III/73/F/7/4c 111/2 2/2 11/1131. III/70/F/7/13c 11/1 1/1 11/1132. III/64/M/5/13d 11/2 11/11 11/233. III/69/F/1.5/23c 2/2 11/11 1/134. III/70/F/4.5/16c 11/1 111/1 1/135. III/64/F/5/15d 2/2 1/1 1/1

a The case number is followed by grade, age, sex, tumor size, andfollow-up in months.

b st, stromal cell; ca, carcinoma cell; ND, not done.c Died of cancer.d Still alive.

2728MMP-2, MMP-9, and MT1-MMP in HCC and Pancreatic Adenocarcinoma



case showing no signals for MT1-MMP mRNA in neoplasticepithelium, and the expression was weak in 18, moderate in 16,but not strong in any of the cases. Thirty-three of 35 tumorsshowed MT1-MMP mRNAs also in the tumor stroma, where theexpression was weak in 23 cases and moderate in nine cases,and in one case the labeling was strong. Strongest stromalMT1-MMP mRNA signal was seen in the fibroblasts of newlyformed fibrous tissue in the vicinity of invasive epithelial ductsor islands (Fig. 1c). Normal acinar or ductal structures werenegative.

MMP-2 mRNA Expression. MMP-2 mRNAs were de-tected in 14 of 36 (38%) of HCCs. The expression occurred instromal cells of 13 of 36 tumors. The mRNA level varied fromweak to moderate. In eight tumors, mRNAs were detected alsoin carcinoma cells. Of stromal cells both fibroblasts and, to alesser extent, endothelial cells showed positive signals, usuallyin areas of abundant fibrosis either around tumor islands or morediffusely in the tumor stroma (Fig. 2a).

Because of exhausted tumor material,in situhybridization forMMP-2 mRNA was performed for 27 pancreatic adenocarcino-mas. MMP-2 mRNAs were detected in 23 of 27 (85%) tumors.Stromal expression clearly predominated over that in neoplasticepithelium; it was weak in 6, moderate in 10, and strong in 6tumors. Positive transcripts were localized to fibroblasts and endo-thelial cells, especially in areas of abundant fibrous desmoplasia.mRNA expression for MMP-2 by neoplastic epithelium was seenin 12 of 27 tumors, and the labeling was weak in 11 and moderatein 1 of them. Epithelial expression never occurred without simul-taneous stromal positivity (Fig. 2b).

MMP-9 mRNA Expression. MMP-9 mRNAs were ob-served in 12 of 36 (33%) HCCs. Twelve of them showedexpression in carcinoma cells and four also in the stromal cells.The intensity of mRNA signal varied from weak to moderateand was generally not associated with the amount of fibrousstroma. Within tumor tissue, the expression was more pro-nounced in the more dysplastic nodules. In stroma, both fibro-blasts and endothelial cells contained the mRNA. Occasionally,mRNAs were also seen in nontumorous hepatocytes of cirrhoticareas and in cells of proliferating bile ducts (Fig. 3a).

In situhybridization for MMP-9 mRNA was performed for26 pancreatic adenocarcinomas. Twenty-three of 26 (91%) tu-mors expressed the mRNA. Generally, carcinoma cells andstromal cells were equally labeled for the mRNA. Twenty-twotumors showed mRNA signal in the carcinoma cells, of which15 cases had weak and 7 moderate expression patterns. Twentycases showed stromal labeling, and of these 11 had weak, 8moderate, and 1 strong expression patterns. mRNAs in thetumor stroma were mainly detected in widely spread fibroblastsand to a lesser extent also in vascular endothelium (Fig. 3b).

Immunohistochemical Staining. Immunohistochemicalstaining was performed to show the corresponding MMP-2,MMP-9, and MT1-MMP protein expression in HCCs and pan-creatic adenocarcinomas. Because of the shortage of antibodies,four cases of HCCs were stained. Carcinoma cells in all caseswere positive for MMP-2, and a clear staining reaction was alsoseen in stromal fibroblasts and endothelial cells (Fig. 2c).

MMP-9 staining reaction was seen in HCCs in individualcarcinoma cells scattered over the tumor area. Occasionally,stromal fibroblasts and endothelial cells were seen to express

Fig. 1 a, in situhybridization for MT1-MMP mRNA in HCC. Carci-noma cells of HCC show diffusely localized positive signals for MT1-MMP mRNA. b, as a control, corresponding sense probe shows nospecific labeling.c, in situ hybridization for MT1-MMP mRNA inpancreatic adenocarcinoma. Positive labeling can be abundantly seen inthe carcinoma cells (arrowheads).d, immunohistochemical staining forMT1-MMP in pancreatic adenocarcinoma. Carcinoma cells containstrong intracytoplasmic positivity for MT1-MMP antigen (arrowheads).3265.




Fig. 2 a, in situhybridization for MMP-2 in HCC. Positive signals arelocalized to the stromal fibroblasts showing elongated cell shape (ar-rows).b, in situhybridization for MMP-2 mRNA in pancreatic adeno-carcinoma. Positive grains are localized abundantly to the stromalfibroblasts (arrows), whereas the neoplastic gland (G) does not containlabeling (arrowhead).c, immunohistochemical staining for MMP-2 inHCC. Cancer cells contains intracytoplasmic positivity for MMP-2antigen. d, immunohistochemical staining for MMP-2 in pancreaticadenocarcinoma. Cancer cells are strongly stained, but the stroma re-mains negative.a, b, andd, 3265; c, 3410.

Fig. 3 a, in situhybridization for MMP-9 mRNA in HCCs. Labeling isstrongly localized to carcinoma cells (arrows).b, in situ hybridizationfor MMP-9 mRNA in pancreatic adenocarcinoma. Positive grains arelocalized especially to the carcinoma cells of neoplastic gland (G) butare also present in stromal fibroblasts (arrows).c, immunohistochemicalstaining for MMP-9 in HCC. Individual carcinoma cells contain intra-cytoplasmic positivity for MMP-9 antigen (arrowheads).d, immuno-histochemical staining for MMP-9 in pancreatic adenocarcinoma. Theneoplastic epithelium contains strong intracytoplasmic positivity.3265.




MMP-9 antigen (Fig. 3c). Because of a high content of endog-enous peroxidases and biotin in liver cells, resulting in increasedpossibilities for nonspecific binding, the immunohistochemicalstainings were not as successful for MT1-MMP in HCCs. Insome grade III carcinomas, there was, however, intracytoplas-mic or cell membrane staining of the carcinoma cells for MT1-MMP.

Immunohistochemical staining for the presence of MMP-2was performed on four pancreatic adenocarcinomas. They allshowed a clear positive staining reaction in carcinoma cells butalso in fibroblasts and endothelial cells (Fig. 2d). Occasionalstaining could also be seen in cells of normal epithelial ducts ofthe adjacent pancreatic tissue.

MMP-9 staining was performed for eight pancreatic ade-nocarcinomas. They all showed a clear intracytoplasmic reac-tion in neoplastic epithelial cells. Overlying tumor stroma re-mained, in most of the cases, negative, and only occasionaltumor fibroblasts and endothelial cells were seen to expressMMP-9 antigen (Fig. 3d).

MT1-MMP staining was performed on eight pancreaticadenocarcinomas. In all cases, intracellular staining was seen inboth carcinomas and also to a lesser extent in stromal cells. Insome areas, a positive staining reaction could also be seen in cellmembranes of the neoplastic epithelial cells (Fig. 1d).

Gelatin Zymography. Gelatinolytic activity has beenpreviously well documented in pancreatic adenocarcinomas(11, 14). Therefore, gelatin zymography was performed onHCCs (Fig. 4).

In normal liver tissue, there was a faintMr 72,000 gelat-inolytic band, corresponding to the latent form of MMP-2. Noevidence of the molecular size of its active form was present.Tumor tissues of HCC contained the clearMr 72,000 band, butthere was also aMr 62,000 band corresponding the molecularsize of the active form of MMP-2.

Tumor and nontumor samples showed almost comparablebands for the latentMr 92,000 MMP-9, but the active,Mr 82,000form was seen only in HCC samples.

Statistical Analysis. To analyze whether the expressionof MMPs was associated with the outcome of the patients, thesemiquantitative results ofin situ hybridization were comparedwith the survival data of the patients. A significant associationwas seen between the expression of MT1-MMP mRNA by

carcinoma cells of HCC and shortened survival of patients (P 5, 0.01, log rank; Fig. 5). No such association was seen in caseswhere the mRNA expression was seen present only in stromalcells (P5 0.81, log rank). Generally, there was a tendency ofincreased MT1-MMP mRNA expression in grade III tumorswhen compared with grade I tumors, but the difference was notstatistically significant (P5 0.07).

In HCCs, there was no statistically significant difference insurvival between patients showing negative or positive resultsfor the presence of MMP-2 or MMP-9 mRNA in carcinomacells (P5 0.24, log rank, andP 5 0.97, log rank, respectively)or in stromal cells (P5 0.94, log rank, andP 5 0.70, log rank,respectively). There was, however, a tendency that positiveexpression was associated with a poorer prognosis of patients.

In pancreatic adenocarcinomas when compared with prog-nosis MT1-MMP mRNA expression by neoplastic cells did notassociate with survival (P5 0.32, log rank), but the Breslow testgave a significant value (P5 0.04). No significant associationwas seen when correlated to the mRNA expression by thestroma (P5 0.12, log rank). Neither was there any difference inthe mRNA expression between various differentiation grades(I–III) of pancreatic adenocarcinomas.

Because only one of the pancreatic tumors expressed mod-erately or strongly mRNA for MMP-2 by carcinoma cells,statistical analysis was not performed. The association of stro-mal expression to the patients’ outcome did not show statisticalsignificance (P5 0.27, log rank). Nor was there any significantassociation between the patients’ prognosis and the expressionof mRNA for MMP-9 by cancer cells (P5 0.65, log rank) or bystromal cells (P5 0.35, log rank). However, according to lifetables, there was a tendency for a poorer prognosis of patientswith increased MMP-9 mRNA expression by tumor cells.

DISCUSSIONIn the present study, we investigated the expressions of

MMP-2, MMP-9, and MT1-MMP mRNA in two different

Fig. 4 Gelatin zymography on normal liver tissue and HCCs. Normalliver (Lane 1) shows clear reaction only forMr 92,000 MMP-9, but onlya faint reaction band can be seen forMr 72,000 MMP-2. Tumor samplesof HCC (Lanes 2–6) contain increased amounts of bothMr 72,000 andMr 62,000 gelatinases, corresponding the latent and active forms ofMMP-2, respectively. In tumor samples, theMr 92,000 latent form ofMMP-9 can be seen and the activeMr 82,000 MMP-9 is also present(Lanes 5and6). Lane 7presents molecular markers:upper arrowhead,Mr 97,000 (dark blue); andlower arrowhead, Mr 66,000 (faint yellow).

Fig. 5 Statistical life table. Patients with HCCs showing strong MT1-MMP mRNA expression had a significantly shorter survival whencompared with other cases (P, 0.01, log rank).




tumor types, HCC and pancreatic adenocarcinoma, byin situhybridization. The results show that carcinoma cells of HCCsare the main producers of MT1-MMP mRNA, and only a lowlevel of expression could be detected in stromal cells. Thispoints out that carcinoma cells themselves are active modu-lators of extracellular matrix space, especially in high-gradetumors. This result is similar to the recent study by Ogataetal. (11), who also showed correlation between increasedMT1-MMP mRNA expression and tumor dedifferentiation.Our result differs from the result of Theretet al. (23), whodemonstrated mRNA synthesis of MT1-MMP solely by stro-mal cells of HCCs, with no clear neoplastic epithelial expres-sion. In pancreatic adenocarcinoma, MT1-MMP mRNA wasalmost equally expressed by neoplastic epithelium and stro-mal cells. Interestingly, MT1-MMP mRNA expression wasequal in high- and low-grade tumors, showing fundamentaldifferences to HCCs. The finding agrees well with the resultby Imamuraet al. (13) but differs from that of Bramhallet al.(12), who demonstrated mRNA synthesis predominantly inthe epithelial cancer cells. In both tumor types, MT1-MMPimmunolocalized both to the neoplastic epithelial cell popu-lation and occasionally to the stromal cells, indicating thatcancer cells and stromal cells are producing both the mRNAand corresponding protein.

There was a statistically significant association with a poorprognosis of patients in HCCs when mRNA synthesis of MT1-MMP was present in cancer cells themselves but not whenstromal cells alone expressed the mRNA. A similar tendencycould also be observed in pancreatic adenocarcinomas but withno statistical significance. The association of increased MT1-MMP mRNA expression with shortened lifetime in HCCs iseasily understandable by its connection to the poor differentia-tion grade of tumors, although it is generally difficult to assessthe prognosis of patients with HCC, because some patients withlarge and progressive tumors have late recurrences and longsurvival periods. Our result agrees also with Haradaet al. (24),who showed increased MT1-MMP mRNA expression in HCCto be associated with capsular infiltration. Our results alsosupport the previous reports of the importance of MT1-MMP inmalignant growth but also add evidence for its association withpoor prognosis of patients with a malignant disease of the liver.MT1-MMP is one of the key enzymes among MMPs, and itsoverexpression seems to have a significant effect on tumorgrowth. To date, there is at least four mechanisms by whichMT1-MMP can enhance tumor progression: it can activateMMP-2 on tumor cell membranes (3, 4, 24); it is a very effectivedegradative enzyme by itself, having substrates such as fi-bronectin, tenascin, nidogen, aggregan, and perlecan (25); MT1-MMP is a very potent regulator for neovascularization (26), aphenomenon that is critical for malignant growth (reviewed inRef. 27); and recently MT1-MMP was shown to process directlylaminin-5 g2 chain, which caused a strong migration effect bytumor cells over laminin-5 surfaces (28). Previously, we haveshown that pancreatic adenocarcinomas strongly express lami-nin-5 g2 chain by tumor cells and also contain laminin-5-bearing BM structures around tumor islands (29), and in apreliminary experiments, we noticed that laminin-5g2 chain is

abundantly expressed also by HCCs.4 MT1-MMP and laminin-5can trigger an important migratory effect in carcinoma cells,leading to their spreading. How these mechanisms together areoperating in malignant growth remains to be clarified. It seems,however, that the association of MT1-MMP with poor prognosisof patients is rather connected to tumor type and general growthpattern than differentiation grade. Unraveling the regulatorymechanisms among the functions of MT1-MMP, MMP-2,TIMP-2, and laminin-5 and possibly some other factors mayprove to be pivotal to our understanding of malignant behaviorin general.

In their study, Theretet al. (23) also showed a correlationbetween MT1-MMP mRNA levels and MMP-2 and TIMP-2mRNA levels, as well as with MMP-2 activation in HCCs. Inthe present study, simultaneously increased MMP-2 and MT1-MMP expression by carcinoma cells was not observed. Gelatinzymography showed that although latent MMP-2 was presentboth in tumor and adjacent nontumor liver tissues, activeMMP-2 was only seen in tumor tissue of HCC. This finding isin agreement with the result by Ogataet al. (11), who alsoshowed increased MMP-2 activation in tissue samples of HCC.

MMP-2 has been suggested to have a very important role incontrolling tumor cell invasion because it is very commonlyoverexpressed in tumor masses and has a high ability to degradeBM macromolecules. In HCCs and pancreatic adenocarcino-mas, MMP-2 mRNA synthesis occurred mainly by stromalfibroblasts and endothelial cells and only to a lesser extent bycarcinoma cells. This result is similar to what has been reportedpreviously in many tumor types such as breast, ovary, and coloncarcinomas (5, 6, 7). Bramhallet al. (12) also reported MMP-2mRNA synthesis to be more abundantly present in the tumorstroma of pancreatic adenocarcinoma. It seems likely that thetype or amount of tumor stroma is an important modulator forMMP-2 expression, because there was higher MMP-2 produc-tion in pancreatic tumors when compared with liver tumors,which are not usually accompanied by fibrosis so abundant as inpancreatic adenocarcinomas. Tumor stromal cells are alsoknown to express many different types of regulatory factors thatup-regulate MMP expression (30). Although MMP-2 was abun-dantly synthesized by stromal cells, it immunolocalized mainlyto tumor cells in HCCs and pancreatic adenocarcinoma.

In HCCs, MMP-9 mRNA was found to be expressedmainly by tumor cells. There was no clear association with theprognosis in mRNA expressions of MMP-9 or MMP-2. How-ever, there was a tendency to have a strong expression for bothmRNAs in tumors with a poorer survival of patients. In aprevious study on HCCs, it was shown that increased MMP-9mRNA expression, as studied by Northern hybridization, wasassociated with capsular infiltration (10). Elevated plasma levelsof MMP-9 in patients with HCC have also been observed,especially in patients having macroscopic portal vein invasion(31). It is notable that in some cases of this study, MMP-9mRNA expression varied from nodule to nodule, being strongerin the more dysplastic ones. Gelatin zymography showed thepresence ofMr 92,000-sized MMP-9 in almost equal amounts in

4 Unpublished data.




tumor samples of HCC and in adjacent nontumor liver tissue,but the activeMr 82,000 form was present only in tumorsamples. All of this evidence indicates that MMP-9 productionand functional activity are increased in HCCs. It is also aninteresting notion that activated MMP-2 can activate proMMP-9(32). One could speculate that MT1-MMP/TIMP-2 complex incell membranes starts the activation cascade by activatingproMMP-2, which in turn is able to convert proMMP-9 to itsactive form.

In summary, our data show that mRNA synthesis of allMMPs investigated is up-regulated in HCCs and pancreaticadenocarcinomas but are differentially expressed by intensityand a cellular origin. Furthermore, MT1-MMP produced bycarcinoma cells of HCCs was associated with a poor prognosisof patients, and a similar tendency was also present in pancreaticadenocarcinomas. The results support previous suggestions thatMT1-MMP could be one of the key enzymes among MMPs inthe process of invasion. Additionally, the uncontrolled produc-tion of MT1-MMP alone could be of significance for malignantgrowth and spread.

ACKNOWLEDGMENTSWe thank Dr. Suneel Apte, Department of Biomedical Engineer-

ing, Cleveland, OH, for providing us with the MT1-MMP probe; Dr.Matti Hoyhtya for providing us with MMP-2 and MMP-9 antibodies;and Professor Veli-Pekka Lehto for valuable comments during prepa-ration of the manuscript. The skillful technical assistance of Heli Auno,Annikki Huhtela, Kaisu Jarvenpa¨a, Tapio Leinonen, and HannuVaananen is greatly appreciated.

REFERENCES1. Docherty, A. J. P., O’Connell, J., Crabbe, T., Angal, S., and Murphy,G. The matrix metalloproteinases and their natural inhibitors: prospectsfor treating degenerative tissue diseases. Trends Biotechnol.,10: 200–207, 1992.2. Liotta, L. A., Tryggvason, K., Garbisa, S., Hart, I., Foltz, C. M.,and Shafie, S. Metastatic potential correlates with enzymatic degra-dation of basement membrane collagen. Nature (Lond.),284: 67– 68,1980.3. Cao, J., Sato, H., Takino, T., and Seiki, M. The C-terminal region ofmembrane type matrix metalloproteinase is a functional transmembranedomain required for pro-gelatinase A activation. J. Biol. Chem.,270:801–805, 1995.4. Strongin, A. Y., Collier, I., Bannikov, G., Marmer, B. L., Grant,G. A., and Goldberg, G. I. Mechanism of cell surface activation of72-kDa type IV collagenase: isolation of the activated form of themembrane metalloproteinase. J. Biol. Chem.,270: 5331–5338, 1995.5. Monteagudo, C., Merino, M. J., San-Juan, J., Liotta, L. A., andStetler-Stevenson, W. G. Immunohistochemical distribution of type IVcollagenase in normal, benign, and malignant breast tissue. Am. J.Pathol.,136: 585–592, 1990.6. Autio-Harmainen, H., Karttunen, T., Hurskainen, T., Hoyhtya, M.,Kauppila, A., and Tryggvason, K. Expression of 72 kilodalton type IVcollagenase (gelatinase A) in benign and malignant ovarian tumors. Lab.Invest.,69: 312–321, 1993.7. Poulsom, R., Pignatelli, M., Stetler-Stevenson, W. G., Liotta, L. A.,Wright, P. A., Jeffrey, R. E., Longcroft, J. M., Rogers, L., and Stamp,G. W. H. Stromal expression of 72 kDa type IV collagenase (MMP-2)and TIMP-2 mRNAs in colorectal neoplasia. Am. J. Pathol.,141:389–396, 1992.8. Nakajima, M., Welch, D. R., Belloni, P. N., and Nicolson, G. L.Degradation of basement membrane type IV collagen and lungsubendothelial matrix by rat mammary adenocarcinoma cell clones

of differing metastatic potentials. Cancer Res.,47: 4869 – 4876,1987.

9. Sreenath, T., Matrisian, L. M., Stetler-Stevenson, W. G., Gattoni-Celli, S., and Pozzatti, R. O. Expression of matrix metalloproteinasegenes in transformed rat cell lines of high and low metastatic potential.Cancer Res.,52: 4942–4947, 1992.

10. Arii, S., Mise, M., Harada, T., Furutani, M., Ishigami, S., Niwano,M., Mizumoto, M., Fukumato, M., and Imamura, M. Overexpression ofmatrix metalloproteinase 9 gene in hepatocellular carcinoma with inva-sive potential. J. Hepatol.,24: 316–322, 1996.

11. Ogata, R., Torimura, T., Motoaki, K., Ueno, T., Tateishi, Y., Ku-romatsu, R., Shimauchi, Y., Sakamoto, M., Tamaki, S., Sata, M., andTanikawa, K. Increased expression of membrane type 1 matrix metal-loproteinase and matrix metalloproteinase-2 with tumor dedifferentia-tion in hepatocellular carcinomas. Hum. Pathol.,30: 443–450, 1999.

12. Bramhall, S. R., Neoptolemos, J. P., Stamp, G. W. H., and Lemoine,N. R. Imbalance of expression of matrix metalloproteinases (MMPs)and tissue inhibitors of the matrix metalloproteinases (TIMPs) in humanpancreatic carcinoma. J. Pathol.,182: 347–355, 1997.

13. Imamura, T., Ohshio, G., Mise, M., Harada, T., Okada, N., Wang,Z., Yoshitomi, S., Tanaka, T., Sato, H., Arii, S., Seiki, M., and Imamura,M. Expression of membrane-type matrix metalloproteinase-1 in humanpancreatic adenocarcinomas. J. Cancer Res. Clin. Oncol.,124: 65–72,1998.

14. Koshiba, T., Hosotani, R., Wada, M., Miyamoto, Y., Fujimoto, K.,Lee, J-U., Doi, R., Arii, S., and Imamura, M. Involvement of matrixmetalloproteinase-2 activity in invasion and metastasis of pancreaticcarcinoma. Cancer (Phila.),82: 642–650, 1998.

15. Caenazzo, C., Onisto, M., Sartor, L., Scalerta, R., Giraldo, A., Nitti,D., and Garbisa, S. Augmented membrane type 1 matrix metallopro-teinase (MT1-MMP):MMP-2 messenger RNA ratio in gastric carcino-mas with poor prognosis. Clin. Cancer Res.,4: 2179–2186, 1998.

16. Sakakibara, M., Koizumi, S., Saikawa, Y., Wada, H., Ichihara, T.,Sato, H., Horita, S., Mugishima, H., Kaneko, Y., and Koike, K. Mem-brane-type matrix metalloproteinase-1 expression and activation of gel-atinase A as prognostic markers in advanced pediatric neuroblastoma.Cancer (Phila.),85: 231–239, 1999.

17. Ishak, K. G., Anthony, P. P., and Sobin, L. H. Histological Typingof Tumours of the Liver: International Histological Classification of theTumours. Geneva: WHO, 1994.

18. WHO. Histological Typing of Tumours of the Liver, Biliary Tractand Pancreas: International Histological Classification of Tumours, No.20. Geneva: WHO, 1978.

19. Huhtala, P., Chow, L. T., and Tryggvason, K. Structure of thehuman type IV collagenase gene. J. Biol. Chem.,265: 11077–11082,1990.

20. Wilhelm, S. M., Collier, I. E., Marmer, B. L., Eisen, A. Z., Grant,G. A., and Goldberg, G. I. SV40-transformed human lung fibroblastssecrete a 92-kDa type IV collagenase which is identical to that secretedby normal human macrophages. J. Biol. Chem.,264: 17213–17221,1989.21. Sato, H., Takino, T., Okada, Y., Cao, J., Shinagawa, A., Yamamoto,E., and Seiki, M. A matrix metalloproteinase expressed on the surface ofinvasive tumor cells. Nature (Lond.),370: 61–65, 1994.22. Heussen, C., and Dowdle, E. B. Electrophoretic analysis of plas-minogen activators in polyacrylamide gels containing sodium dodecylsulphate and copolymerized substrates. Anal. Biochem.,102: 196–202,1980.23. Theret, N., Musso, O., L’Helgoualc’h, A., Campion, J-P., andClement, B. Differential expression and origin of membrane-type 1 and2 matrix metalloproteinases (MT-MMPs) in association with MMP-2activation in injured human livers. Am. J. Pathol.,153: 945–954, 1998.24. Harada, T., Arii, S., Mise, M., Imamura, T., Higashitsuji, H.,Furutani, M., Niwano, M., Ishigami, S., Fukumoto, M., Seiki, M., Sato,H., and Imamura, M. Membrane-type matrix metalloproteinase-1 (MT1-MMP) gene is overexpressed in highly invasive hepatocellular carcino-mas. J. Hepatol.,28: 231–239, 1998.




25. D’ortho, M-P., Will, H., Atkinson, S., Butler, G., Messent, A.,Gavrilovic, J., Smith, B., Timpl, R., Zardi, L., and Murphy, G. Mem-brane-type matrix metalloproteinases 1 and 2 exhibit broad-spectrumproteolytic capacities comparable to many matrix metalloproteinases.Eur. J. Biochem.,250: 751–757, 1997.26. Hiraoka, N., Allen, E., Apel, I. J., Gyetko, M. R., and Weiss, S. J.Matrix metalloproteinase regulate neovascularization by acting as peri-cellular fibrinolysins. Cell,95: 365–377, 1998.27. Weidner, N. Tumoural vascularity as a prognostic factor in cancerpatients: the evidence continues to grow. J. Pathol.,184: 119–122, 1998.28. Koshikawa, N., Giannelli, G., Cirulli, V., Miyazaki, K., andQuaranta, V. Role of cell surface metalloprotease MT1-MMP inepithelial cell migration over laminin-5. J. Cell Biol.,148: 615– 624,2000.

29. Soini, Y., Maatta, M., Salo, S., Tryggvason, K., and Autio-Har-mainen, H. Expression of the lamining2 chain in pancreatic adenocar-cinoma. J. Pathol.,180: 290–294, 1996.30. Watari, M., Watari, H., DiSanto, M. E., Chacko, S., Shi, G-P., andStrauss, J. F. Pro-inflammatory cytokines induce expression of matrix-metabolizing enzymes in human cervical smooth muscle cells. Am. J.Pathol.,154: 1755–1762, 1999.31. Hayasaka, A., Suzuki, N., Fujimoto, N., Iwama, S., Fukuyama, E.,Kanda, Y., and Saisho, H. Elevated plasma levels of matrix metallopro-teinase-9 (92-kD type IV collagenase/gelatinase B) in hepatocellularcarcinoma. J. Hepatol.,24: 1058–1062, 1996.32. Fridman, R., Toth, M., Pena, D., and Mobashery, S. Activation ofprogelatinase B (MMP-9) by gelatinase A (MMP-2). Cancer Res.,55:2548–2555, 1995.




2000;6:2726-2734. Clin Cancer Res Marko Määttä, Ylermi Soini, Annikki Liakka, et al. Progression and Clinical PrognosisPancreatic Adenocarcinoma: Implications for TumorMMP-9, and Membrane Type 1-MMP in Hepatocellular and Differential Expression of Matrix Metalloproteinase (MMP)-2,

Updated version

http://clincancerres.aacrjournals.org/content/6/7/2726

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/6/7/2726.full#ref-list-1

This article cites 29 articles, 9 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/6/7/2726.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/6/7/2726To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/6/7/2726.full#ref-list-1

http://clincancerres.aacrjournals.org/content/6/7/2726.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



differential expression of matrix metalloproteinase...

Documents