Comprehensive Analysis of Matrix Metalloproteinase and Tissue Inhibitor Expression in Pancreatic Cancer

2832 Vol. 10, 2832–2845, April 15, 2004 Clinical Cancer Research

Comprehensive Analysis of Matrix Metalloproteinase and Tissue Inhibitor Expression in Pancreatic Cancer: Increased Expression of Matrix Metalloproteinase-7 Predicts Poor Survival

Lucie E. Jones,1 Michelle J. Humphreys,1 Conclusions: We propose that the principle specificity Fiona Campbell,2 John P. Neoptolemos,1 and for effective inhibitors of MMPs in pancreatic cancer should be for MMP-7 with secondary specificity against MMP-11. Mark T. Boyd1 Moreover, these studies indicate that MMP-7 expression is a 1 Department of Surgery, University of Liverpool, Liverpool, United powerful independent prognostic indicator and potentially Kingdom, and 2Department of Pathology, Royal Liverpool University Hospital, Liverpool, United Kingdom of considerable clinical value.

INTRODUCTION ABSTRACT Pancreatic cancer is a major cause of cancer death in North Purpose: To enable the design of improved inhibitors of America and Europe, and its incidence has appeared to increase matrix metalloproteinases (MMPs) for the treatment of pan- dramatically in the last 70 years (1–4). Disappointingly, the creatic cancer, the expression profiles of a range of MMPs 5-year survival rate of Ͻ5% has shown little improvement in the and tissue inhibitors of MMPs (TIMPs) were determined. last 30 years (2, 5, 6). In those patients with resectable tumors, Experimental Design: Nine MMPs (MMPs 1–3, 7–9, 11, the 5-year survival is of the order of 10–24% and may be 12, and 14) and three TIMPs (TIMPs 1–3) were examined in improved by adjuvant therapies (7–9). The low survival rates are up to 75 pancreatic ductal adenocarcinomas and 10 normal attributed to an aggressive biological phenotype that is charac- pancreata by immunohistochemistry. Eighteen additional terized by early local invasion and metastasis (10). At a molec- pancreatic ductal adenocarcinomas and an additional eight ular level, pancreatic cancer is characterized by a high frequency normal pancreata were also analyzed by real-time reverse of KRAS mutations in combination with the loss of function transcription-PCR and additionally for MMP-15. of the TP53, p16INK4A, and SMAD4 tumor suppressor genes Results: There was increased expression by immunohis- (11–17). tochemistry for MMPs 7, 8, 9, and 11 and TIMP-3 in pan- Pancreatic cancer progression is associated with increased creatic cancer compared with normal pancreas (P < 0.0001, expression of growth factor receptors and their ligands (18–21), 0.04, 0.0009, 0.005, and 0.0001, respectively). Real-time re- alterations in the level of expression of matrix metalloprotein- verse transcription-PCR showed a significant increase in ases (MMPs) and tissue inhibitors of MMPs (TIMPs), which mRNA levels for MMP-11 in tumor tissue compared with may result in an imbalance between these (22–24), alterations in and also significantly antiapoptotic pathways (25–27), the loss of additional tumor (0.0005 ؍ normal pancreatic tissue (P -Univariate analysis suppressor gene function (28–30), and the acquisition of neo .(0.0026 ؍ reduced levels of MMP-15 (P revealed that survival was reduced by lymph node involve- angiogenic properties (31, 32). Early invasion means that most patients have lymph node ؍ and increased expression of MMP-7 (P (0.0007 ؍ ment (P but not metastases and both neural and vascular invasion at the time of (0.02 ؍ and (for the first time) MMP-11 (P (0.005 reduced by tumor grade, tumor diameter, positive resection presentation (33–35). Indeed, it seems very likely that the cancer margins, adjuvant treatment, or expression of the remaining is disseminated at the time of diagnosis because virtually all MMPs and TIMPs. On multivariate analysis, only MMP-7 patients die of pancreatic cancer irrespective of their treatment predicted shortened survival (P < 0.05); however, increased (2, 5–7, 36, 37). MMP-11 expression was strongly associated with lymph Given the limited response to standard forms of chemotherapy, there has been considerable interest in treatments based .(0.0073 ؍ node involvement (P on biologically relevant target molecules (38–40). Synthetically developed inhibitors of MMPs (MPIs) have been of particular interest (41–45). The MMPs are a family of zinc-containing proteolytic Received 7/29/04; revised 1/16/04; accepted 1/27/04. enzymes that break down proteins in the extracellular matrix The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked in both physiological and pathological conditions (46–49). advertisement in accordance with 18 U.S.C. Section 1734 solely to The four main subgroups, organized in part according to indicate this fact. substrate specificity in vitro and in part by association with Note: L. Jones and M. Humphreys contributed equally to this work. the cell membrane, are the collagenases (MMPs 1, 8, and 13), Requests for reprints: Mark T. Boyd, Department of Surgery, Univer- sity of Liverpool, UCD Building, Daulby Street, Liverpool, L69 3GA, the gelatinases (MMPs 2 and 9), stromelysins (MMPs 3, 10, United Kingdom. Phone: 44-151-706-4185; Fax: 44-151-706-5826; E- 11, and 18), and the membrane-bound MMPs (MMPs 14–17; mail: [email protected]. Ref. 40). MMPs can regulate the activity of several classes of

molecules including cytokines, growth factors, and growth Table 1 Details of the 75 patients with pancreatic ductal factor receptors such as interleukin-1␤, insulin growth factor- adenocarcinoma used for immunohistochemistry binding proteins, heparin-binding epidermal growth factor- Correlation like growth factor, and fibroblast growth factor receptor-1 Number Prognostic with MMPa ϭ (reviewed in Ref. 50). In addition proteolytic modification of (total 75) significance expression extracellular matrix components such as laminin-5, decorin, Sex NS Nil Men 43 entactin, and fibronectin by MMPs have been connected to Women 32 alterations in a broad range of cellular events including cell Median age (range) 63 NS Nil migration, apoptosis, and proliferation as well as growth (40–77) years factor sequestration. Loss of the tight control of MMP activ- Tumor grade NS Nil Well differentiated 8 ity in cancer is thought to contribute to excessive destruction Moderately differentiated 37 of the extracellular matrix, neo-vascularization, tumor Poorly differentiated 30 spread, and metastases. Tumor stage NS Nil pT 5 There have been reports of increased MMP expression in 1 pT2 28 several human malignancies with reported prognostic signif- pT3 39 icance. In gastric cancer, the levels of MMPs 2 and 9 have pT4 3 ϭ been found to be elevated along with MMPs 7 and MMP 14 Lymph node status P 0.007 MMP-11 (P ϭ 0.0073) (51–53). The expression of MMP-2 was higher in patients Negative 20 with gastric cancer that had a poor prognosis, although the Positive 53 difference was not significant (54). Murray et al. (55) showed Unknown 2 that MMP-1 expression was associated with a poor prognosis Resection margins NS Nil (P ϭ 0.06) in colorectal cancer. Increased expression of MMPs 2 and 9 Negative (R0) 35 and TIMPs 1 and 2 has also been shown to correlate with a Positive (R1) 40 b poor prognosis in renal cell carcinoma (56). Other studies Adjuvant therapy NS Nil Yes 18 using various techniques have reported a prognostic signifi- No 57 cance of increased MMP expression in carcinomas of the a MMP, matrix metalloproteinases; NS, not significant. colon (57), ovary (58), breast (59), prostate (60), and lung b As part of the European Study Group for Pancreatic Cancer-1 trial (61). Although early studies suggested that TIMPs may have (8, 9). an antitumor or antimetastatic activity, it is now thought that they may have a more complex function. For example there is a positive correlation between increased TIMP-2 levels and MATERIALS AND METHODS poor outcome in breast cancer (62). Patients and Specimens. The pancreatic cancer and con- In pancreatic cancer, there have been several studies of trol tissue samples (described in Table 1) were obtained from MMP and TIMP expression, and these have often involved the Pancreas Tissue Bank of the Department of Surgery, Liver- either limited numbers of patient samples (22, 63) and/or the pool Candis Tissue Bank and the Department of Pathology, analysis of only a few MMPs or TIMPs (64, 65). Perhaps not Royal Liverpool University Hospital. Formalin-fixed, paraffin- surprisingly, given the variety of techniques used in these embedded material was available for immunohistochemistry studies, some conflicting results have been generated. More from between 45 and 75 (depending on the MMP or TIMP recently studies by Yamamoto et al. (66) have suggested that studied) consecutive tissue sections from patients who had un- MMP-7 might be an independent prognostic indicator and the dergone pancreas resection and for whom clinical and follow-up potential involvement of MMP-7 has been strongly supported data were available (note however that two of these patients by another recent investigation (67). To further investigate were lost to follow-up). Tissue from 18 additional freshly col- the involvement of MMPs in pancreatic cancer and to expand lected patient samples were rapidly frozen and stored in liquid the range of individual MMPs and TIMPs studied for this nitrogen for reverse transcription (RT)-PCR analysis. All of the disease, we have sought to perform the most comprehensive tissues were examined by an experienced pancreas pathologist evaluation of MMPs and TIMPs that has been performed to (F. C.) using H&E-stained slides, and only pancreatic ductal date. This is a high priority because the poor or inappropriate carcinomas (ICD-0 C25.0–2, 8) were included in the study. specificity of therapeutic MPIs tested thus far may account Adenocarcinomas arising from the intra-pancreatic bile duct, for the rather disappointing results in recent pancreatic cancer ampulla of Vater, or duodenum and other malignant neoplasms trials (44, 45, 68). such as acinar cell carcinomas or cystadenocarcinoma were We have used immunohistochemistry and real-time reverse specifically excluded. Pathological staging was undertaken ac- transcription-PCR analysis of pancreatic ductal adenocarcinoma cording to the Union Internationale Contre le Cancer TNM and normal tissue to examine the level and pattern of expression classification (69). Clinical and staging details are described in of MMPs and TIMPs. Using this data, we then determined Table 1. The ten “normal pancreas” tissue specimens were whether any of these molecules may be implicated in the pro- derived from (a) four patients with common bile duct tumors gression of pancreatic cancer and also whether any of these with histologically normal pancreas removed, (b) two patients molecules might act as prognostic indicators. with pancreatic cancer with normal pancreas removed distant

from the tumor, (c) two patients with focal pancreatitis with different antibodies, each antibody was tested using a range of adjacent normal pancreas, (d) one patient with duodenal tumor, dilutions, with or without pretreatment, on these sandwich and (e) one patient with a benign pancreatic cyst. blocks. Table 2 lists the antibodies used and the control tissues. For RNA analysis, there were 18 pancreatic ductal adeno- Note that for MMP-7, MMP-11 commercially available anti- carcinomas, and 8 “normal” pancreatic tissues were obtained bodies were used to confirm the staining patterns observed using from histologically normal areas of the parenchyma from two the British Biotech antibodies. For trypsinization pretreatment, patients with focal chronic pancreatitis, four patients with pan- slides were soaked in 0.05% trypsin, in Tris-buffered saline creatic tumors, and two transplant donors. [TBS; 50 mM Tris-HCl, 150 mM NaCl (pH 7.6)] for 20 min at This study was undertaken with the formal approval of the 37°C. For microwave pretreatment, slides were immersed in 10 Ϫ1 Liverpool Research Ethics Committee and written informed mM EDTA and irradiated with 850 joules/second for 15 min. consent was obtained for all samples. Slides were washed with TBS and then blocked with 5% BSA Antibodies. Mono- and polyclonal antibodies were sup- in TBS. Primary antibodies were diluted in 5% BSA in TBS, plied by the research and development section of British Biotech and 100 ␮l of each antibody solution was added to each slide for (Oxford, UK) and are not commercially available. These in- 1 h at room temperature. Slides were washed twice in TBS for cluded antibodies to MMP-1 (interstitial collagenase; 36632), 4 min. Secondary antibody staining was performed using the MMP-2 (gelatinase A; 36033, 1A10, SE592), MMP-3 (strome- DAKO EnVisionϩ kit (DAKO, Ely, UK) according to the lysin 1; 10D6, RP58), MMP-7 (matrilysin; RP21), MMP-8 manufacturers instructions. Slides were then washed twice with (neutrophil collagenase; SE603), MMP-9 (gelatinase B; 36074, TBS for 4 min, incubated with 3,3-diaminobenzidine HCl so- 4H3, 2C10, SE601), MMP-12 (metalloellastase; 10D10, RP54), lution for a further 20 min, finally rinsed with distilled water, and MMP-14 (MT1-MMP; RP64, 5H2). Antibodies from Brit- and counterstained with hematoxylin. Sections were dehydrated ish Biotech were later validated by comparing the staining with ethanol followed by xylene and mounted using DPX pattern to that obtained with commercially available antibodies (VWR International Ltd., Poole, UK) mountant. Immuno- on a subset of samples and examining these for consistency. In staining specificity was established using a negative control for addition, the Ab-1 antibody to MMP-7 and Ab-5 antibody to each section, in which the primary antibody was omitted. Also MMP-11 were obtained from Labvision (Newmarket, UK). In included in each batch of samples was a known positive control. all cases, antibodies recognize both latent zymogens and active Assessment of Immunohistochemical Staining. Two species. Monoclonal antibodies to TIMP-1 (147-6D11), TIMP-2 observers scored all sections independently. The intensity of the (67-4H11) and TIMP-3 (136-13H4) were obtained from Chemi- immunostaining of the tumor epithelium, tumor stroma, and con Europe Ltd. (Chandlers Ford, UK). We observed no evi- pancreatitis (if present) and of normal ducts, acini, and islets dence of cross-reactivity between antibodies to different MMPs was graded as negative (0), weak (1), moderate (2), and strong or TIMPs. (3). To facilitate interpretation and comparison of heterogeneous Immunohistochemical Staining. Tissue specimens tissue staining, whole sections were assigned scores according were fixed for 48 h in 10% formalin and embedded in paraffin to the strongest intensity staining on a section, providing that wax before sectioning. Four-␮m sections of each specimen were Ն10% of the specified cell type (i.e., adenocarcinoma/stroma/ prepared on poly-L-lysine-coated slides and kept in an oven at epithelia) on the section stained. In cases where Ͻ 10% of cells 37°C overnight before further processing. Positive controls for of a specific type were stained the section was assigned a score each antibody were obtained from a bank of control slides in use of 0. Note that there were no cases in which Ͻ10% of cells in the Immunohistochemistry Laboratory, Department of Path- stained predominantly strongly or moderately. Differences in ology, Royal Liverpool University Hospital. To establish which assigned scores were resolved by simultaneous re-examination tissues were reproducible and consistent positive controls for by both scorers to achieve a consensus. There were no cases each antibody, “sandwich” blocks (containing small tissue sam- where the scores assigned differed by Ͼ1 grade. ples from multiple different sites) were stained with each anti- RNA Isolation. Total RNA was isolated from approxi- body. To establish the optimum staining conditions for the mately 100 mg of tissue using RNAzol-B (Biogenesis, Poole,

Table 2 Details of the antibodies used MMPa Antibody name Pretreatment Dilution Control MMP1: 36632 monoclonal Interstitial collagenase None 1:100 Duodenum MMP2: 36033 monoclonal Gelatinase A None 1:100 Smooth muscle MMP3: RP58 monoclonal Stromelysin 1 Trypsin (20Ј) 1:50 Kidney MMP7: RP21 polyclonal Matrilysin None 1:100 Skeletal muscle MMP8: SE603 polyclonal Neutrophil collagenase None 1:200 to 1:400 Skeletal muscle MMP9: 4H3 polyclonal Gelatinase B Trypsin (20Ј) 1:100 Appendix MMP11 monoclonal Stromelysin 3 None 1:100 Breast cancer MMP12: RP54 monoclonal Metalloellastase None 1:400 Kidney MMP14: RP63 monoclonal MT-MMP1 None 1:200 Smooth muscle TIMP 1 monoclonal 147-6D11 Microwave 1:200 Pancreas TIMP 2 monoclonal 67-4H11 Trypsin (20Ј) 1:200 Colon TIMP 3 monoclonal 136-13H4 Microwave 1:1000 Colon a MMP, matrix metalloproteinases; TIMP, tissue inhibitor of the metalloproteinases.

UK) as described by the manufacturer. RNA was quantified by were as follows: (5Ј-3Ј) MMP-1 forward AATGTGCTACACG- spectrometry at 260 nm and integrity was determined by run- GATACCC reverse CTTTGTGGCCAATTCCAGGA; MMP-2 ning 500 ng of total RNA on a 1% Tris-acetate-EDTA gel. To forward CCGCCTTTAACTGGAGCAAA, reverse TTTGGTT- remove any contaminating DNA, the samples were treated with CTCCAGCTTCAGG; MMP-3 forward GAGGAAAATCGAT- DNase I. Approximately 10 ␮g of total RNA was digested with GCAGCCA, reverse CTCCAACTGTGAAGATCCAG; MMP- DNase I (RNase free; Sigma, Poole, UK) in the presence of 40 7 forward TTTGATGGGCCAGGAAACAC, reverse GGGG- units rRNasin (Promega, Southampton, UK), in 100 mM sodium ATCTCCATTTCCATAG; MMP-9 forward GAAGATGCT- acetate (pH 5) and 5 mM MgCl2 at 25°C for 1 h. The integrity GCTGTTCAGCG, reverse ACTTGGTCCACCTGGTTCAA; of the RNA was checked before and after digestion by gel MMP-11 forward AGATCTACTTCTTCCGAGGC, reverse electrophoresis. DNase was then removed by phenol/chloro- TTCCAGAGCCTTCACCTTCA3; MMP-13 forward ATA- form/isoamyl alcohol extraction, and RNA was precipitated CAGGCAAGACTCTCCTG, reverse GCGAACAATACGGG- with absolute ethanol. TTACTCC; MMP-14 forward CAGAAGCTGAAGGTAGA- cDNA Synthesis. cDNA was synthesized from 5 ␮gof ACC, reverse CATGGCGTCTGAAGAAGAAG; MMP-15 for- total RNA using 200 ng of oligonucleotide-dT primer (Life ward GGACAGCGCAGAGGGCG, reverse TTGCAGTA- Technologies, Inc., Paisley, UK) annealed in 10 mM HEPES (pH AAGCAGGACACG; TIMP-1 forward CGGGGCTTCACC- 7.0), 1 mM EDTA in a final volume of 10 ␮l, heated to 90°C for AAGACC, reverse TCAGGCTATCTGGGACCGC; TIMP-2 2 min, and cooled on ice. First-stand synthesis was performed forward GAAGAAGAGCCTGAACCACA, reverse GTCCTC- by adding 1ϫ first-strand buffer (5ϫ buffer 250 mM Tris-HCl GATGTCGAGAAACT; TIMP-3 forward TCGGTATCACCT- (pH 8.3), 375 mM KCl, 15 mM MgCl2), 1 mM DTT and 0.5 mM GGGTTGTA, reverse GTCTGTGGCATTGATGATGC. Prim- dNTPs, and 300 units Moloney murine leukemia virus reverse ers were supplied by either Roche or Helena Biosciences (Tyne transcriptase (all from Life Technologies, Inc.) to a final volume and Wear, UK). of 20 ␮l and incubating at 37°C for 90 min. Control samples Statistical Analysis. Categorical data were analyzed by were prepared without reverse transcriptase. To verify the the ␹2 test and Fisher’s exact probability test. Continuous data cDNA and the absence of contaminating DNA, PCR was per- were analyzed by the Mann-Whitney U test, and correlation formed on both test and control samples for glyceraldehyde analysis was by Spearman’s rank correlation. Survival curves 3-phosphate dehydrogenase (GAPDH). The PCR reaction con- were calculated using the Kaplan-Meier method. Cox propor- tained 0.5 ␮l of cDNA product, 200 ␮M dNTPs, 0.5 ␮M of each tional hazards modeling was used for multivariate analysis. The primer forward 5ЈTGCCGTCTAGAAAAACCTGC 3Ј and re- level of significance was set at 0.05. verse 5ЈACCCTGTTGCTGTAGCCAAA 3Ј and 1 unit of

TaqDNA polymerase in 50 mM KCl, 3 mM MgCl2, and 10 mM Tris-HCl (pH 9; Amersham Pharmacia Biotech, Little Chalfont, RESULTS UK) in a volume of 20 ␮l. The PCR cycles were 95°C for 30 s, For immunohistochemical analysis, either 45 or, for some 57°C for 30 s, and 72°C for 1 min. This produced a 200-bp MMPs and TIMPs depending on availability of the tissue and/or GAPDH product. Only samples that had this product in their antibody, 75 cancer and 10 normal specimens were stained with ϩRT and not in their ϪRT controls were used in experiments. each antibody to assess patterns in the differential expression of all Real-Time RT-PCR. RT-PCR was carried out in a of the MMPs and TIMPs. Where results suggested significant Roche Lightcycler (Roche Diagnostics, Lewes, UK) and quan- differences, antibody staining was confirmed with an additional tified by measuring SYBR green. PCR was performed using the antibody or antiserum to the MMP or TIMP on a subset of these DNA Master SYBR green I PCR reaction mix that contains Taq samples. These results are summarized in Table 3. For the meas- polymerase, dNTP mix (with dUTP instead of dTTP; Roche), 3 urement of mRNA levels, real-time RT-PCR was performed on 18 ␮ ␮ mM MgCl2 with 0.5 M of each primer, and either 2 lof cancer and 8 normal specimens. These results are summarized in diluted cDNA or 2 ␮l of water in a total volume of 20 ␮l. The Table 4. PCR cycles were 95°C for 30 s, followed by 45 cycles of 95°C MMP-1. MMP-1 was expressed in 29 of 45 (64%) can- for 10 s, 57°C for 10 s, and 72°C for 12 s. Quantification was cers and in 3 of 10 (30%) normals. Immunostaining was weak achieved by comparison with an internal standard curve con- to moderate in the majority of cancers with only 3 to 29 (10%) taining 10-fold dilutions of a plasmid containing the target cases staining strongly. Staining in all cancers was restricted to sequence (108-102 copies). Two of these plasmids were pro- the tumor cells, with no staining of the stroma. Staining oc- vided by British Biotech, and these contained the priming sites curred in the acini, ducts, and islets of adjacent pancreatitis in 7 for GAPDH and all of the MMPs and TIMPs studied. The PCR of 45 (16%) cancers. Weak staining was observed in the acini of performed on the plasmid with each specific primer pair pro- 3 to 10 normals. There was no significant difference in the duced a product of approximately 300 bp. As the efficiency of expression of MMP-1 between the cancer and normal samples. PCR amplification from the plasmid was comparable for each There was no difference in survival between the two groups. product (with detection of each product occurring within one Also, no correlation was observed between the intensity of cycle of all of the other products), all MMPs and TIMPs were MMP-1 staining and tumor differentiation, tumor size, lymph- quantified against the internal GAPDH standard curve. The node status, or patient survival. Real-time RT-PCR demon- quantity of each MMP and TIMP was then expressed as a strated that MMP-1 mRNA was expressed at significantly fraction of GAPDH level for each sample. The efficacy of all higher levels (P ϭ 0.006) in tumor tissue than in normal tissue, primer pairs for MMPs and TIMPs was validated on RNA and this is compatible with the immunohistochemical staining obtained from appropriate positive control tissues. The primers pattern albeit that the latter did not reach statistical significance.

Table 3 Immunohistochemical staining of pancreatic tissue MMP or TIMPa Pancreatic ductal adenocarcinoma Normal pancreas Staining Adjacent Prognostic Antibody for: intensity Carcinoma Stroma pancreatitisa Epithelia Stroma Cancer vs. normal significance MMP-1 0 16 (36%) Nil 38 (85%) 7 (70%) Nil NS NS N ϭ 45 PDAC, 1 11 (24%) 2 (4%) N ϭ 10 normal 2 15 (33%) 5 (11%) 3 (30%) 3 3 (7%) MMP-2 0 25 (33%)b Nil 35 (47%) 3 (30%) Nil NS NS N ϭ 75 PDAC, 1 21 (28%) 9 (12%) 4 (40%) N ϭ 10 normal 2 13 (17%) 17 (23%) 3 (30%) 3 16 (21%) 14 (19%) MMP-3 0 5 (11%) 18 (40%) 13 (29%) stroma vs. stroma, NS P Ͻ 0.0001 N ϭ 45 PDAC, 1 16 (36%) 22 (49%) 1 (2%) 10 (100%) N ϭ 10 normal 2 12 (27%) 5 (11%) 31 (69%) 10 (100%) 3 12 (27%) 0 MMP-7 0 4 (9%)b Nil 19 (42%) 8 (80%) Nil P Ͻ 0.0001 P Ͻ 0.005 N ϭ 45 PDAC, 1 13 (29%) 6 (13%) 2 (20%) N ϭ 10 normal 2 19 (42%) 16 (35%) 3 9 (20%) 4 (11%) MMP-8 0 12 (27%)d Nil 2 (8%) 4 (40%) Nil P ϭ 0.04 NS N ϭ 45 PDAC, 1 18 (40%) 9 (36%) 6 (60%) N ϭ 10 normal 2 13 (29%) 13 (52%) 3 2 (4%) 1 (4%) MMP-9 0 26 (58%) Nil Nil 10 (100%) Nil P ϭ 0.0009 NS N ϭ 45 PDAC, 1 13 (29%) N ϭ 10 normal 2 3 (7%) 3 3 (7%) MMP-11 0 23 (31%) 53 (71%) 35 (47%) 8 (80%) Nil P ϭ 0.004 P ϭ 0.02 N ϭ 75 PDAC, 1 20 (27%) 18 (24%) 9 (12%) 2 (20%) (RT-PCR, P ϭ 0.0005) N ϭ 10 normal 2 18 (24%) 3 (4%) 17 (23%) 3 14 (19%) 1 (1%) 14 (19%) MMP-12 0 25 (56%) Nil 20 (44%) Nil NS NS N ϭ 45 PDAC, 1 19 (42%) 20 (44%) 8 (80%) N ϭ 10 normal 2 5 (11%) 2 (20%) 3 1 (2%) MMP-14 0 31 (41%)b,c,d Nil 37 (49%) 6 (60%) Nil NS NS N ϭ 75 PDAC, 1 26 (35%) 22 (29%) 4 (40%) N ϭ 10 normal 2 12 (16%) 15 (20%) 3 6 (8%) 1 (1%) TIMP-1 0 42 (93%) Nil Nil 10 (100%) Nil NS (RT-PCR, NS P ϭ 0.03) N ϭ 45 PDAC, 1 N ϭ 10 normal 2 3 (7%) 3 TIMP-2 0 25 (56%) Nil 17 (49%) 2 (20%) Nil NS NS N ϭ 45 PDAC, 1 17 (38%) 14 (40%) 8 (80%) N ϭ 10 normal 2 1 (2%) 4 (11%) 3 2 (4%) TIMP-3 0 7 (9%) 21 (28%) 8 (11%) 2 (20%) Nil P Ͻ 0.0001, NS (P ϭ 0.06) stroma vs. stroma, P Ͻ 0.0003 N ϭ 75 PDAC, 1 13 (17%) 9 (12%) 13 (17%) 8 (80%) N ϭ 10 normal 2 26 (35%) 17 (23%) 26 (35%) 3 29 (39%) 28 (37%) 28 (37%) a MMP, matrix metalloproteinases; TIMP, tissue inhibitor of the metalloproteinases; NS, not significant; RT, reverse transcription; PDAC, pancreatic durtal adenocarcinoma. b MMP-2 correlated with MMP-14 (P Ͻ 0.0001). c MMP-7 correlated with MMP-14 (P ϭ 0.0005). d MMP-8 correlated with MMP-14 (P Ͻ 0.0001). NS ϭ not significant. e Not all tumor sections had adjacent pancreatitis, percentage expressed as fraction of scored adjacent pancreatitis.

Table 4 Real-time RT-PCR for MMPs and TIMPs from pancreatic tissue RT-PCR for Pancreatic ductal adenocarcinoma Normal pancreas Cancer vs. the following: (N ϭ 18) (N ϭ 8) normal MMP-1 0.027 (0.011–0.052) 0.004 (0.001–0.005) P ϭ 0.006 MMP-2 0.032 (0.016–0.073) 0.029 (0.017–0.122) NS MMP-3 0.00096 (0.00066–0.0017) 0.0037 (0.0005–0.011) NS MMP-7 0.032 (0.012–0.10) 0.044 (0.038–0.072) NS MMP-9 0.008 (0.004–0.02) 0.004 (0.001–0.02) NS MMP-11 0.057 (0.016–0.102) 0.0001 (0.0001–0.001) P ϭ 0.0005 MMP-14 2.4 ϫ 10Ϫ5 (4.3 ϫ 10Ϫ7–2.1 ϫ 10Ϫ4) 6.7 ϫ 10Ϫ5 (3.2 ϫ 10Ϫ5–2.0 ϫ 10Ϫ3) NS MMP-15 9.48 ϫ 10Ϫ5 (1.3 ϫ 10Ϫ5–2.0 ϫ 10Ϫ4) 1 ϫ 10Ϫ3 (3.6 ϫ 10Ϫ4–2 ϫ 10Ϫ3) P ϭ 0.0026 TIMP-1 0.76 (0.42–1.24) 0.338 (0.081–0.648) P ϭ 0.030 TIMP-2 0.002 (0.0004–0.004) 0.003 (0.001–0.006) NS TIMP-3 0.035 (0.01–0.06) 0.031 (0.006–0.075) NS Values are the median (inter-quartile range) expression ratio of the sample, glyceraldehyde-3-phosphate dehydrogenase NS, not significant. RT, reverse transcription; MMP, matrix metalloproteinases; TIMP, tissue inhibitor of the metalloproteinases; NS, not significant

MMP-2. MMP-2 was expressed in 50 of 75 (66%) can- ison, in normal pancreas. The identity of the staining pattern was cers and in 7 of 10 (70%) normals. Staining was moderate to confirmed using a commercial antibody for MMP-7. No corre- strong in 29 of 50 (58%) cancers and was weak in 21 of 50 lation was observed among MMP-7 expression and tumor dif- (42%). Staining in all cancers was restricted to the tumor cells, ferentiation, tumor size, or lymph node status. However, there with no staining of the stroma. Adjacent pancreatitis similarly was a correlation between MMP-7 expression and the expres- expressed MMP-2. Expression in the normals was weak or sion of MMP-14 (P ϭ 0.005). Most significantly tumors moderate predominantly in acini and islets. There was no sig- strongly positive for MMP-7 had a significantly worse progno- nificant difference in the intensity of staining between the can- sis than MMP-7 negative/weakly staining tumors (P ϭ 0.005) as cer and normal specimens. There was no correlation among shown in Fig. 2A. Real-time RT-PCR demonstrated that MMP-7 MMP-2 staining and tumor differentiation, tumor size, tumor mRNA was expressed to comparable levels. lymph node status, or patient survival. Nevertheless, there was MMP-8. MMP-8 was expressed in 33 of 45 (73%) of the a correlation between the staining of MMPs 2 and 14 (P Ͻ cancers and in 6 of 10 (60%) of normals. Expression in tumor 0.0001). Real-time RT-PCR demonstrated that MMP-2 mRNA cells was moderate to strong in 15 of 33 (45%) cancers. There was expressed to comparable levels in both tumor and normal was no staining of the tumor stroma. Expression in the normal tissues, and this is compatible with the immunohistochemistry. pancreas was weak within ducts, and occasionally islet staining MMP-3. MMP-3 was expressed in 40 of 45 (89%) can- was seen. The intensity of staining was increased in cancers cers and in 10 of 10 (100%) normals. In cancers, staining was compared with normals and was significant (P ϭ 0.04). There generally weak or moderate in both carcinoma and stroma, was no correlation among MMP-8 staining and tumor differen- although staining was strong in 12 of 40 (30%) cancers. Normal tiation, tumor size, tumor lymph node status, or patient survival. stroma stained strongly in 10 of 10 samples, including the Real-time RT-PCR failed to detect the mRNA for MMP-8. periductal and inter-lobular connective tissue, and also the walls MMP-9. MMP-9 was expressed by 19 of 45 (43%) of of vessels. Interestingly, the staining within the normal stroma cancers and was not detectable in normals (P ϭ 0.0009). Six of was significantly stronger than the staining of the malignant 19 (32%) cancers stained moderately or strongly, and this stain- stroma (P Ͻ 0.0001). Islets, ducts, and acinar cells stained ing was focal. There was no staining of the adjacent pancreatitis weakly in all normal pancreas and weak to moderately in and no staining of stroma in either cancers or normals. Neutro- adjacent pancreatitis. There was no correlation between the phils stained positively in both the cancer and normals. No tumor cell expression and tumor stroma expression. Nor was a correlation was observed between MMP-9 expression and tumor correlation observed among MMP-3 expression and tumor dif- differentiation, tumor size, lymph node status, or patient sur- ferentiation, tumor size, lymph node status, or patient survival. vival. Real-time RT-PCR detected low-level expression of Real-time RT-PCR demonstrated that MMP-3 mRNA was ex- MMP-9 in normal and cancer samples at comparable levels. pressed to comparable levels in both tumor and normal tissues, MMP-11. MMP-11 was expressed by 52 of 75 (69%) of and this is compatible with the immunohistochemistry. cancer cases and in 2 of 10 (20%) normals (P ϭ 0.004). Fig. 1B MMP-7. MMP-7 was expressed in 41 of 45 (91%) of illustrates a typical staining pattern for pancreatic adenocarci- cancers and in 2 of 10 (20%) normals (P Ͻ 0.0001). Tumor cells noma and for comparison, normal pancreas. There was staining in 28 of 41 (68%) cancers stained moderately to strongly. The of stroma in 22 of 75 (29%) cancers, but only one of these proportion of positive cells in tumors was variable; in some stained strongly. Only stromal cells adjacent to cancer stained cases, focal staining was observed. Expression in adjacent pan- positively, and the staining was weaker or not detectable in creatitis was predominantly moderate with 26 of 45 (58%) stromal cells further from the primary tumor. Forty of 75 (53%) staining positively. Staining in the two normals was weak with of adjacent pancreatitis stained for MMP-11 with 31 of 40 only a few islets or ducts being positive. There was no staining (78%) staining moderate to strong. The two normals that were in the stroma of cancers or normals. Fig. 1A illustrates a typical positive had weak staining in the acini, islets, and ducts. The staining pattern in pancreatic adenocarcinoma and, for compar- identity of the staining pattern was confirmed using a commer-

Fig. 1 Immunohistochemistry of tumor and normal pancreas. A, immunohistochemical staining for matrix metalloproteinase (MMP)-7 demonstrates differences in expression between normal (3 and 4) and tumorous (1 and 2) pancreas tissue. Original magnification 1 and 3;40ϫ and 2 and 4, 400ϫ. B, immunohistochemical staining for MMP-11. 1, 2, 3, and 4 as above.

cial antibody for MMP-11. There was no correlation among MMP-12. MMP-12 was expressed by 20 of 45 (44%) MMP-11 expression, tumor differentiation, and tumor size. cancers and in 10 of 10 (100%) normals. Nineteen of 45 However, there was a correlation between MMP-11 expression (42%) cancers stained weakly. There was no stromal staining. and lymph node status, with those tumors having a higher level There was weak to moderate staining in 25 of 45 (55%) of expression correlating with positive lymph nodes (P ϭ adjacent pancreatitis. In normals ducts and islets stained 0.0073). In one case in which there was an adjacent lymph node, more strongly than acini, there was no significant difference there was staining of MMP-11 in the metastatic deposit in the in the expression of MMP-12 between cancer and normal and lymph node. Patients with MMP-11-positive carcinoma had a no correlation was observed among MMP-12 expression and significantly shorter overall survival time than did those with tumor differentiation, tumor size, lymph node status, or pa- MMP-11-negative carcinoma (P ϭ 0.02) as shown in Fig. 2B. tient survival. Real-time RT-PCR failed to detect the mRNA Real-time RT-PCR demonstrated that MMP-11 mRNA was of MMP-12. expressed at significantly higher levels (P ϭ 0.0005) in tumor MMP-14. MMP-14 was expressed by 44 of 75 (59%) of tissue than in normal tissue, and this is compatible with the cancers and 4 of 10 (40%) normals. Thirty-eight of 44 (86%) immunohistochemical staining pattern. cancers stained weakly to moderately and only 6 of 44 (14%)

Fig. 2 Kaplan-Meier survival curve data showing the correlation between levels of matrix metalloproteinase (MMP) expression with survival for MMP-7 (see Fig. 1A1) and MMP-11 (see Fig. 1B2). 〈, survival data for patients with either negative or weak expression of MMP-7 versus patients with moderate to strong expression of MMP-7. B, survival data for patients with either negative or weak expression of MMP-11 versus patients with moderate to strong expression of MMP-11.

stained strongly. There was no staining of stroma. Thirty-seven TIMP-2. TIMP-2 was expressed by 20 of 45 (44%) of 75 (49%) adjacent pancreatitis stained weak to moderate. In cancers and in 8 of 10 (80%) normals. The staining was weak in normals, occasional acini and ducts stained weakly. There was 17 of 20 (85%) positive cancers with only 2 of 20 (10%) no significant difference in expression of MMP-14 in the can- staining strongly. Adjacent pancreatitis stained moderately or cers compared with normals. There was no staining of any weakly in 18 of 35 (51%) cases, and this did not correlate with stroma. No correlation was observed among MMP-14 expres- the level of tumor staining. There was no stromal staining. Eight sion and tumor differentiation, tumor size, or lymph node status. of 10 normals stained weakly with some positivity in ducts and However, there was a strong correlation with MMP-14 expres- the occasional acinar cell. There was no significant difference in sion and MMP-2 (P Ͻ 0.0001), MMP-7 (P ϭ 0.005), and also expression of TIMP-2 between cancers and normals. No corre- MMP-8 (P Ͻ 0.0001) expression. lation was observed among TIMP-2 expression and tumor dif- MMP-15. In the absence of an antibody at the time of the ferentiation, tumor size, lymph node status, or patient survival. study, MMP-15 was only studied by real-time RT-PCR. Nev- Real-time RT-PCR detected low levels of TIMP-2 mRNA in ertheless, MMP-15 expression was found to be significantly both cancers and normals. lower (P ϭ 0.0026) in tumor tissue compared with normal tissue TIMP-3. TIMP-3 was expressed by 68 of 75 (91%) by real-time RT-PCR (see Table 4). cancers and in 8 of 10 (100%) normals. Staining was moderate TIMP-1. TIMP-1 was expressed in 3 of 45 (7%) cancer to strong in 55 of 68 (81%) cancers. There was moderate to cases and in 0 of 10 (0%) normals. In the three cancer cases, strong staining of stroma in 45 of 75 (60%) cancers. Adjacent there were only occasional foci of TIMP-1 staining. There was pancreatitis stained in 67 of 75 cancers and 54 of 67 (81%) of no staining of adjacent pancreatitis or stroma. No correlation these were stained moderate to strong. Eight of 10 (80%) nor- was observed among TIMP-1 expression and tumor differenti- mals stained weakly for TIMP-3 in ductal elements. However, ation, tumor size, lymph node status, or patient survival. RT- acini and islets stained more strongly. There was no stromal PCR demonstrated that TIMP-1 mRNA was expressed at higher staining in normals. There was an increased intensity and sta- levels in cancers than in normals, and this difference was sig- tistically significant difference in expression of TIMP-3 be- nificant (P ϭ 0.03). tween cancers and normals (P Ͻ 0.0001). There was also a

significant increase in TIMP-3 expression in cancer stroma FasL has been most clearly implicated in events likely to lead to versus normals (P Ͻ 0.0003). No correlation was observed pancreatic ductal adenocarcinoma. Crawford et al. (67) demon- among TIMP-3 expression and tumor differentiation, tumor strated that FasL was cleaved to a soluble form, soluble (s)FasL, size, lymph node status. Real-time RT-PCR demonstrated that in an MMP-7-dependent manner during acinar to ductal meta- TIMP-3 mRNA was expressed to comparable levels in tumor plasia in a mouse model. One explanation for the metaplasia and normal tissues. could be that there is a selective advantage as a result of Survival Analysis. Of the tumor staging characteristics, insensitivity to apoptosis in the emerging ductal population (67, tumor grade, diameter, and differentiation were not significant 85). This interpretation requires either that ductal cells no longer as summarized in Table 1. Lymph node status correlated to respond to proapoptotic signaling by sFasL or that sFasL has survival (P ϭ 0.007). Univariate analysis revealed a relationship more potent antiapoptotic activity on the trans-differentiated between overexpression of MMPs 7 and 11 (P ϭ 0.005 and P ϭ ductal cells. Which of these possibilities is correct is unclear. 0.02, respectively) and an unfavourable outcome in operable Certainly, sFasL has been shown to exhibit both pro- and pancreatic carcinoma. Kaplan-Meier plots for MMPs 7 and 11 antiapoptotic activities (81, 86–88), and this might reflect dif- are shown in Fig. 2. Multivariate analysis was then conducted ferential cleavage (producing different peptides with various that included all significant and marginally significant values activities) or differential activity of sFasL on cells at different from the univariate analysis to identify any independent predic- stages of either tumor development or differentiation, perhaps as tors of survival. Cox proportional hazard modeling demon- a result of changes in the protein composition of the extracel- strated that MMP-7 overexpression was the only significant lular milieu (86). independent indicator for survival (P Ͻ 0.05). Osteopontin has been identified as a target of MMP-7 (and of MMP-3; Ref. 82) and has been implicated in wound healing, inflammation, and tumorigenesis. In vitro, cleavage of os- DISCUSSION teopontin by MMP-7 was observed to lead to an increase in the We have found that several MMPs and TIMPs may have a migratory and adhesive activity of osteopontin in an integrin- role in the evolution of human pancreatic adenocarcinoma. In dependent manner (82), although whether this contributes to particular, MMPs 7, 8, 9, 11 and TIMPs 1 and 3 were expressed tumor invasiveness in vivo is not known. at significantly higher levels in the tumor as shown by immu- MMP-7 expression is regulated by the wnt/␤-catenin/tcf nohistochemistry and/or real-time RT-PCR. Also, we found that signaling pathway, and it is therefore not surprising to find other MMP-15 was expressed at significantly lower levels in the connections between MMP-7 and this pathway, as can be illus- tumor by real-time RT-PCR, but we were unable to confirm this trated by ␤4-integrin (79) and E-cadherin (89). Interestingly, by immunohistochemistry. Of most significance, both MMPs 7 ␤4-integrin has been implicated recently in pancreatic ductal and 11 predicted survival. In the case of MMP-7, this was of adenocarcinoma by virtue of its expression being up-regulated independent prognostic significance. (90). Several members of the integrin family have been shown Studies of MMP-7 in pancreatic cancer have shown that it to contribute to repression of metastasis by promoting adhesion is present at higher levels in cancers than in the normal pancreas to the extracellular matrix. Thus cleavage by MMPs may com- (24, 66, 67), and our results support this conclusion. Yamamoto promise the activity of other ␤4-integrin-containing complexes, et al. (66) found that MMP-7 expression in pancreatic adeno- albeit this does not explain why there appears to be selective carcinoma correlated with a poor prognosis and was a signifi- advantage during tumor evolution for up-regulation of expres- cant independent prognostic factor for overall survival. Our data sion of ␤4-integrin. One potential explanation for this may be also support this finding. that the proteolytic release of the integrin ectodomains leads to High levels of MMP-7 expression have been associated these domains competing with membrane-associated integrins with a poor prognosis in other tumors: for example, colon and for target binding, thus reducing the strength of interaction of gastric (70, 71), pancreas (66, 67), prostate (72), and brain (73). the latter with the extracellular matrix. It is noteworthy that in MMP-7 up-regulation is often identified as an early event in studies of an oral carcinoma cell line, it was found that in ad- tumor development (66, 74). Although the mechanisms under- dition to E and N-cadherins, ␤1 integrins also played a role in lying these observations are likely to involve a number of activating expression of MMP-7 (89). Thus a pathway that leads molecules and pathways, several important parts of the jigsaw to amplification of the effects of increased MMP-7 expression have been identified (see below). MMP-7 is unusual among the and activity may connect the activity of MMPs with expression MMPs in that expression is mainly restricted to epithelial cells, and feedback regulation of specific adhesion molecules. rather than the stroma (71, 75). MMP-7 is produced mainly by MMP-7 has been shown to cleave E-cadherin at the cell tumor cells themselves rather than as a result of a desmoplastic surface and release soluble E-cadherin (80, 91). The soluble reaction and can increase the invasive potential of the tumor E-cadherin fragment thus released has been shown to inhibit cells (76). Of particular note in this regard, MMP-7 transcription E-cadherin function in a paracrine fashion (80, 91). Loss of is regulated by the wnt/␤-catenin/tcf signaling pathway (77, 78). E-cadherin expression in a transgenic mouse model of pancre- Taken together, these observations suggest a link between early atic ␤-cell carcinogenesis (Rip1Tag2) coincides with the tran- events in tumor evolution such as loss of normal signaling sition from well-differentiated adenoma to invasive carcinoma pathway regulation and MMP-7 activity. To what is this activity (92). Curiously, it has also been demonstrated that E-cadherin directed? MMP-7 has been shown, inter alia, to proteolytically stimulates expression of MMP-7 through the wnt-signaling cleave ␤4 integrin (79), E-cadherin (80), Fas ligand (FasL; Refs. pathway (93). 67 and 81), osteopontin (82), and TNF-␣ (83, 84). Of these, Our study is the first to identify a correlation between

MMP-11 expression and outcome. MMP-11 was found previ- other studies (63). Forty-two percent of our cancer cases stained ously to be expressed more highly in pancreatic ductal adeno- positively for MMP-9 with individual cells and cellular clusters carcinoma than in normal pancreata by immunohistochemistry staining, whereas the adjacent stroma did not. In situ hybridiza- with the highest expression observed in stromal regions adjacent tion studies have also detected higher levels of MMP-9 expres- to the tumor (64). MMP-11 expression has been observed in sion in tumor epithelium than in the adjacent stroma (63, 65). breast tumors and tumors of the head and neck (94, 95). In these Membrane-bound MMP-14 has previously been found ex- reports, MMP-11 has also been detected only in stromal fibro- pressed exclusively localized in the tumor epithelial cells (24). blasts surrounding the malignant cells and not in the malignant In contrast to our study, studies using in situ hybridization found cells themselves or in fibroblasts distant from the tumor (94). that MMP-14 mRNA was not localized to the tumor epithelial Others have also found that the expression of MMP-11 was cells, but was present in cellular elements within the tumor limited to stromal fibroblasts adjacent to breast cancer cells (96). stroma (102, 103). It has been suggested that MMP-14 may This is not always the case however. In one series, it was activate pro-MMP-2 (104). Interestingly, we have found that the reported that in 15 of 54 (28%) breast cancer samples, the tumor pancreatic cancers that expressed MMP-14 also expressed cells themselves were positive, and also in the adjacent stroma, MMPs 2 and 7. Others have found that the presence of MMP-14 34 of 54 (63%) samples exhibited positive staining (97). In mRNA correlated with that of MMP-2 mRNA (24). Although it another study, it was reported that 6 of 21 (29%) cases of is clear that the activation and inhibition of MMPs and TIMPs pancreatic cancer expressed MMP-11 in the epithelial tumor is a complex issue, it has been suggested previously that there is cells with 17 of 21 (81%) cases displaying expression in the a link between MMP-14 and pro-MMP-2 activation (24). The adjacent stromal cells (64). Data from our study suggests that correlation that we and others have found may reflect such a the majority of MMP-11 expression is derived from the tumor link. epithelial cells themselves, with 52 of 75 (69%) cases staining There are currently only two studies in which expression of these cells. We have also observed adjacent staining of malig- MMP-15 in cancer has been described, and no doubt this reflects nant stromal cells in 22 of 75 (29%) of cases. In vitro studies in the past shortage of available reagents. In recent studies, fibroblasts have demonstrated that basic fibroblast growth factor MMP-15 mRNA was found at higher levels in hepatocellular and epidermal growth factor can act as potent inducers of carcinoma cell lines (105) and by microarray analysis in lung MMP-11 expression (18, 97). Pancreatic cancer cells are known cancer tissues from patients who had received chemotherapy to over-express these growth factors; therefore, these may stim- (106). By contrast, we have found that MMP-15 mRNA levels ulate the expression of MMP-11 in pancreatic cancer (18, 19). were significantly lower (P ϭ 0.0026) in tumor tissue compared In breast carcinoma, the level of MMP-11 mRNA has been with normal tissue. Clearly, none of these studies determine the found to be highly predictive for the presence or absence of source of the expressing cell type or even whether the changes distant metastases (98). This has also been related to invasive- in expression are the result of changes in tissue composition. ness, with only one in ten ductal-carcinoma in situ cases ex- Nonetheless, the loss of expression in the tumor that we have pressing MMP-11 compared with 62% of invasive cancers (99). observed warrants additional investigation. A correlation has been also shown in breast cancer between Early studies suggested that the TIMPs exerted an antitu- expression of MMP-11 and survival, with those tumors express- mor or antimetastatic effect, and this was supported by in vitro ing a high level of the enzyme being associated with fatal studies which showed that increased TIMP expression was metastatic disease (98). Multivariate analysis has identified that associated with a decrease in tumor cell growth and spread (107, MMP-11 may be a strong independent prognostic parameter for 108). However, more recent work has suggested a positive disease-free survival (P ϭ 0.005; Ref. 96). Those patients with correlation between TIMP levels and poor outcome. High ex- a moderate to strong MMP-11 level had significantly shorter pression levels of TIMP-1 have been linked with a poor prog- disease-free survival than those with no or low expression. In nosis in gastric tumors, with those tumors that expressed the our study, MMP-11 expression in pancreatic cancer also ap- highest levels of TIMP1 having a poorer prognosis (57). Higher peared to correlate with invasive potential, with high levels of levels of TIMP mRNA have also been shown to correlate with expression being related to the extent of nodal disease (P ϭ lymph node metastases and reduced 5-year survival in patients 0.0073). Also in our study, those patients with moderate or high with colorectal carcinoma (109). In breast carcinoma TIMP MMP-11 expression had a significantly shorter survival than mRNA expression was found to be higher in the carcinoma than those patients with negative or low expression (P ϭ 0.02). As the surrounding normal breast tissue (24). In this latter study, the has been found previously in breast and head and neck cancer, TIMP-1 signal was substantially lower in the tumor cells than in we found that there was no obvious correlation between the the stroma. Our own data show that TIMP-1 expression can only degree of pancreatic tumor differentiation and the extent of be detected by immunohistochemistry in 7% (3 of 45) of pan- MMP-11 expression (95, 98). creatic adenocarcinoma samples. In these three cases, the stain- Previous studies have suggested that MMP-9 expression ing was confined to individual tumor cell clusters, and there was correlates with increased metastatic potential in colorectal can- no staining of the stroma. This contrasts with the study by cer (100). It has been demonstrated that latent MMP-9 was Bramhall et al. (24) where by Northern blot analysis, seven (7 of present in both normal pancreas and pancreatic adenocarcinoma 7) pancreatic cancer cell lines and all (17 of 17) of the pancreatic (101). However, there was an increase in activity shown by tumor samples expressed detectable TIMP-1 mRNA. However, zymography, with 21.2% of tumors having MMP-9 activity in using in situ hybridization, these authors found that the expres- comparison to 15.4% of normal pancreas specimens. Neutrophil sion was epithelial in origin. staining was a consistent feature, and this was supported by TIMP-3 has been less well studied than TIMP-1 and unlike

the soluble TIMPs 1 and 2, TIMP-3 is unusual because it is a ACKNOWLEDGMENTS component of the extracellular matrix (110). A study of breast We thank Steve Taylor, Department of Mathematics, University of carcinoma and malignant melanoma cell lines showed that in- Liverpool for generous assistance with statistical analysis and Steve creased expression of TIMP-3 resulted in a significant reduction Bradburn and Andy Dodson, Department of Pathology, Royal Liverpool in tumor cell growth. Adenovirus-mediated gene delivery of University Hospital for assistance with immunohistochemistry. TIMPs-1, -2, and -3 revealed that the overproduction of TIMP-3 inhibited invasion of malignant melanoma cells (111). In human pancreatic cancer, TIMP-3 expression has not been investigated REFERENCES previously. Our findings suggest that increased TIMP-3 expres- 1. Fernandez E, La Vecchia C, Porta M, Negri E, Lucchini F, Levi F. sion could be associated with a survival disadvantage although Trends in pancreatic cancer mortality in Europe, 1955–1989. Int J Cancer 1994;57:786–92. this link did not attain statistical significance (P ϭ 0.06). This 2. Bramhall SDJ, Neoptolemos JP. Epidemiology of pancreatic cancer, observation may warrant additional investigation. vol. 2. Boston, MA: Blackwell Scientific; 1998. p. 889–906. Ultimately, the expression profiles of MMPs and TIMPs 3. Parkin DM, Bray FI, and Devesa, SS. Cancer burden in the year are of interest for two reasons. First, they may inform us about 2000. The global picture. Eur J Cancer 2001;37(Suppl 8):S4–66. the changes that take place in the cells (and also in the tissues) 4. Jemal A, Thomas A, Murray T, Thun M. Cancer statistics, 2002. CA that ultimately become tumors. Second, this information can be Cancer J Clin 2002;52:23–47. used to guide the design of novel therapeutics. 5. Bramhall SR, Allum WH, Jones AG, Allwood A, Cummins C, The potential for MPI therapy in pancreatic cancer was Neoptolemos JP. Treatment and survival in 13,560 patients with pan- demonstrated recently in a phase III trial of 414 patients with creatic cancer, and incidence of the disease, in the West Midlands: an advanced disease who were randomized to an MPI (marimastat epidemiological study. Br J Surg 1995;82:111–5. at 5, 10, or 25 mg twice daily) or gemcitabine (1000 mg/m2; 6. Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for 100,313 patients Ref. 44). There appeared to be a dose response in terms of diagnosed from 1985–1995, using the National Cancer Database. J Am patient survival for marimastat. The survival rate of patients Coll Surg 1999;189:1–7. treated with gemcitabine was significantly greater than for pa- 7. Sohn TA, Yeo CJ, Cameron JL, et al. Resected adenocarcinoma of tients who received 5 and 10 mg of marimastat b.i.d but was not the pancreas-616 patients: results, outcomes, and prognostic indicators. different from the survival of patients who received 25 mg of J Gastrointest Surg 2000;4:567–79. marimastat b.i.d. The major toxicity of marimastat was muscu- 8. Neoptolemos JP, Dunn JA, Stocken DD, et al. Adjuvant chemora- loskeletal affecting 44% of the patients so treated (44). More- diotherapy and chemotherapy in resectable pancreatic cancer: a random- over, the combination of gemcitabine and marimastat was found ised controlled trial. Lancet 2001;358:1576–85. not to be superior to gemcitabine and placebo in a phase III of 9. Neoptolemos JP, Stocken DD, Dunn JA, et al. Influence of resection margins on survival for patients with pancreatic cancer treated by advanced pancreatic cancer (45). In the dose-determining phase adjuvant chemoradiation and/or chemotherapy in the ESPAC-1 random- II trial, prolonged usage of the higher dose of 25 mg of mari- ized controlled trial. Ann Surg 2001;234:758–68. mastat at b.i.d caused musculoskeletal side effects in 59% of 10. Magee CJ, Greenhalf W, Howes N, Ghaneh P, Neoptolemos JP. patients (43). Thus the efficacy of MPI therapy could be im- Molecular pathogenesis of pancreatic ductal adenocarcinoma and clin- proved by higher dosing and by using more selective inhibitors ical implications. Surg Oncol 2001;10:1–23. that minimize the musculoskeletal side effects. 11. Barton CM, Staddon SL, Hughes CM, et al. Abnormalities of the Marimastat is a powerful general MPI with an IC to p53 tumour suppressor gene in human pancreatic cancer. Br J Cancer 50 1991;64:1076–82. MMPs 1, 2, 3, 7, 8, 9, and 14 of 5, 6, 200, 20, 2, 3, and 1.8 nM, respectively (40). We would conclude from our study that the 12. Caldas C, Hahn SA, da Costa LT, et al. Frequent somatic mutations and homozygous deletions of the p16 (MTS1) gene in pancreatic ade- principle specificity for an effective MPI in pancreatic cancer nocarcinoma. Nat Genet 1994;8:27–32. should be for MMP-7, but MMP-7 is also heavily expressed by 13. Rozenblum E, Schutte M, Goggins M, et al. Tumor-suppressive skeletal muscle (Table 2). Whether the musculoskeletal side pathways in pancreatic carcinoma. Cancer Res 1997;57:1731–4. effects of marimastat are attributable to specific MMP inhibition 14. Wilentz RE, Iacobuzio-Donahue CA, Argani P, et al. Loss of or whether there is another mechanism responsible is unclear. A expression of Dpc4 in pancreatic intraepithelial neoplasia: evidence that recent phase I study of the MMP inhibitor MMI270 resulted in DPC4 inactivation occurs late in neoplastic progression. Cancer Res musculoskeletal toxicity in 42% of patients (112). Of interest is 2000;60:2002–6. that this MPI has IC to MMPs 1, 2, 3, 7, 9, and 13 of 33–43, 15. Kawesha A, Ghaneh P, Andren-Sandberg A, et al. K-ras onco- 50 gene subtype mutations are associated with survival but not expres- 11, 13–50, 8, and 6 nM, respectively (40). The agent BAY sion of p53, p16(INK4A), p21(WAF-1), cyclin D1, erbB-2 and 12-9566 is an example of a “deep-pocket” MMP inhibitor with erbB-3 in resected pancreatic ductal adenocarcinoma. Int J Cancer Ͼ Ki to MMPs 1, 2, 3, 9, and 13 of 5000, 11, 143, 301, and 1470 2000;89:469–74. nM, respectively (40), that has no musculoskeletal side effects 16. Hruban RH, Iacobuzio-Donahue C, Wilentz RE, Goggins M, Kern (113) but also lacks clinical efficacy (114). It has been proposed SE. Molecular pathology of pancreatic cancer. Cancer J 2001;7:251–8. that dose-limiting musculoskeletal toxicity may result from the 17. Hahn SA, Schutte M, Hoque AT, et al. DPC4, a candidate tumor inhibition of MMP-1 or of sheddases, which are closely related suppressor gene at human chromosome 18q21.1. Science (Wash D C) metalloproteases, but at present there are insufficient clinical 1996;271:350–3. data for a clear conclusion (68, 114). 18. Korc M, Chandrasekar B, Yamanaka Y, Friess H, Buchier M, Beger HG. Overexpression of the epidermal growth factor receptor in human We believe that the findings of this investigation may aid in pancreatic cancer is associated with concomitant increases in the levels the development of MPIs that are more specifically targeted of epidermal growth factor and transforming growth factor ␣. J Clin toward pancreatic cancer than heretofore possible. Investig 1992;90:1352–60.

19. Yamanaka Y, Friess H, Buchler M, et al. Overexpression of acidic 41. Jones L, Ghaneh P, Humphreys M, Neoptolemos JP. The matrix and basic fibroblast growth factors in human pancreatic cancer corre- metalloproteinases and their inhibitors in the treatment of pancreatic lates with advanced tumor stage. Cancer Res 1993;53:5289–96. cancer. Ann N Y Acad Sci 1999;880:288–307. 20. Friess H, Yamanaka Y, Buchler M, et al. Enhanced expression of 42. Rosemurgy A, Harris J, Langleben A, Casper E, Goode S, Rasmus- transforming growth factor ␤ isoforms in pancreatic cancer correlates sen H. Marimastat in patients with advanced pancreatic cancer: a dose- with decreased survival. Gastroenterology 1993;105:1846–56. finding study. Am J Clin Oncol 1999;22:247–52. 21. Kleeff J, Ishiwata T, Maruyama H, et al. The TGF-␤ signaling 43. Evans JD, Stark A, Johnson CD, et al. A phase II trial of marimastat inhibitor Smad7 enhances tumorigenicity in pancreatic cancer. Onco- in advanced pancreatic cancer. Br J Cancer 2001;85:1865–70. gene 1999;18:5363–72. 44. Bramhall SR, Rosemurgy A, Brown PD, Bowry C, Buckels JA. 22. Satoh K, Ohtani H, Shimosegawa T, Koizumi M, Sawai T, Toyota Marimastat as first-line therapy for patients with unresectable pancreatic T. Infrequent stromal expression of gelatinase A and intact basement cancer: a randomized trial. J Clin Oncol 2001;19:3447–55. membrane in intraductal neoplasms of the pancreas. Gastroenterology 45. Bramhall SR, Schulz J, Nemunaitis J, Brown PD, Baillet M, Buck- 1994;107:1488–95. els JA. A double-blind placebo-controlled, randomised study comparing 23. Bramhall SR, Stamp GW, Dunn J, Lemoine NR, Neoptolemos JP. gemcitabine and marimastat with gemcitabine and placebo as first line Expression of collagenase (MMP2), stromelysin (MMP3) and tissue therapy in patients with advanced pancreatic cancer. Br J Cancer 2002; inhibitor of the metalloproteinases (TIMP1) in pancreatic and ampullary 87:161–7. disease. Br J Cancer 1996;73:972–8. 46. Brenner CA, Adler RR, Rappolee DA, Pedersen RA, Werb Z. 24. Bramhall SR, Neoptolemos JP, Stamp GW, Lemoine NR. Imbal- Genes for extracellular-matrix-degrading metalloproteinases and their ance of expression of matrix metalloproteinases (MMPs) and tissue inhibitor, TIMP, are expressed during early mammalian development. inhibitors of the matrix metalloproteinases (TIMPs) in human pancreatic Genes Dev 1989;3:848–59. carcinoma. J Pathol 1997;182:347–55. 47. Wolf C, Chenard MP, Durand de Grossouvre P, Bellocq JP, 25. Friess H, Lu Z, Andren-Sandberg A, et al. Moderate activation of Chambon P, Basset P. Breast-cancer-associated stromelysin-3 gene is the apoptosis inhibitor bcl-xL worsens the prognosis in pancreatic expressed in basal cell carcinoma and during cutaneous wound healing. cancer. Ann Surg 1998;228:780–7. J Investig Dermatol 1992;99:870–2. 26. Evans JD, Cornford PA, Dodson A, Greenhalf W, Foster CS, 48. Nelson AR, Fingleton B, Rothenberg ML, Matrisian LM. Matrix Neoptolemos JP. Detailed tissue expression of bcl-2, bax, bak and bcl-x metalloproteinases: biologic activity and clinical implications. J Clin in the normal human pancreas and in chronic pancreatitis, ampullary and Oncol 2000;18:1135–49. pancreatic ductal adenocarcinomas. Pancreatology 2001;1:254–62. 49. McCawley LJ, Matrisian LM. Matrix metalloproteinases: they’re 27. Greten FR, Weber CK, Greten TF, et al. Stat3 and NF-␬B activation not just for matrix anymore! Curr Opin Cell Biol 2001;13:534–40. prevents apoptosis in pancreatic carcinogenesis. Gastroenterology 2002; 50. Sternlicht MD, Werb Z. How matrix metalloproteinases regulate 123:2052–63. cell behavior. Annu Rev Cell Dev Biol 2001;17:463–516. 28. Goggins M, Hruban RH, Kern SE. BRCA2 is inactivated late in the 51. David L, Nesland JM, Holm R, Sobrinho-Simoes M. Expression of development of pancreatic intraepithelial neoplasia: evidence and im- laminin, collagen IV, fibronectin, and type IV collagenase in gastric plications. Am J Pathol 2000;156:1767–71. carcinoma. An immunohistochemical study of 87 patients. Cancer 29. Hruban RH, Wilentz RE, Kern SE. Genetic progression in the (Phila) 1994;73:518–27. pancreatic ducts. Am J Pathol 2000;156:1821–5. 52. Honda M, Mori M, Ueo H, Sugimachi K, Akiyoshi T. Matrix metal- 30. Schneider G, Schmid RM. Genetic alterations in pancreatic carci- loproteinase-7 expression in gastric carcinoma. Gut 1996;39:444–8. noma. Mol Cancer 2003;2:15. 53. Nomura H, Sato H, Seiki M, Mai M, Okada Y. Expression of 31. von Marschall Z, Cramer T, Hocker M, et al. De novo expression of membrane-type matrix metalloproteinase in human gastric carcinomas. vascular endothelial growth factor in human pancreatic cancer: evidence Cancer Res 1995;55:3263–6. for an autocrine mitogenic loop. Gastroenterology 2000;119:1358–72. 54. Allgayer H, Babic R, Beyer BC, et al. Prognostic relevance of 32. Schwarte-Waldhoff I, Volpert OV, Bouck NP, et al. Smad4/DPC4- MMP-2 (72-kD collagenase IV) in gastric cancer. Oncology 1998;55: mediated tumor suppression through suppression of angiogenesis. Proc 152–60. Natl Acad Sci USA 2000;97:9624–9. 55. Murray GI, Duncan ME, O’Neil P, Melvin WT, Fothergill JE. 33. Nagakawa T, Konishi I, Higashino Y, et al. The spread and prog- Matrix metalloproteinase-1 is associated with poor prognosis in colo- nosis of carcinoma in the region of the pancreatic head. Jpn J Surg rectal cancer. Nat Med 1996;2:461–2. 1989;19:510–8. 56. Kallakury BV, Karikehalli S, Haholu A, Sheehan CE, Azumi N, 34. Nakao A, Harada A, Nonami T, et al. Lymph node metastases in Ross JS. Increased expression of matrix metalloproteinases 2 and 9 and carcinoma of the head of the pancreas region. Br J Surg 1995;82:399– tissue inhibitors of metalloproteinases 1 and 2 correlate with poor 402. prognostic variables in renal cell carcinoma. Clin Cancer Res 2001;7: 35. Cusack JC, Jr, Fuhrman GM, Lee JE, Evans DB. Managing unsus- 3113–9. pected tumor invasion of the superior mesenteric-portal venous conflu- 57. Murashige M, Miyahara M, Shiraishi N, Saito T, Kohno K, Koba- ence during pancreaticoduodenectomy. Am J Surg 1994;168:352–4. yashi M. Enhanced expression of tissue inhibitors of metalloproteinases 36. Trede M, Schwall G, Saeger HD. Survival after pancreatoduode- in human colorectal tumors. Jpn J Clin Oncol 1996;26:303–9. nectomy. 118 consecutive resections without an operative mortality. 58. Young TN, Rodriguez GC, Rinehart AR, Bast RC, Jr, Pizzo SV, Ann Surg 1990;211:447–58. Stack MS. Characterization of gelatinases linked to extracellular matrix 37. Yeo CJ, Sohn TA, Cameron JL, Hruban RH, Lillemoe KD, Pitt HA. invasion in ovarian adenocarcinoma: purification of matrix metallopro- Periampullary adenocarcinoma: analysis of 5-year survivors. Ann Surg teinase 2. Gynecol Oncol 1996;62:89–99. 1998;227:821–31. 59. Talvensaari-Mattila A, Paakko P, Hoyhtya M, Blanco-Sequeiros G, 38. Neoptolemos JP, Ghaneh P, Hartley M, Sutton R, Magee CJ. The Turpeenniemi-Hujanen T. Matrix metalloproteinase-2 immunoreactive role of adjuvant therapy for pancreatic cancer. Expert Opin Investig protein: a marker of aggressiveness in breast carcinoma. Cancer (Phila) Drugs 2002;11:87–107. 1998;83:1153–62. 39. Ko AH, Tempero MA. Current and future strategies for combined- 60. Stearns M, Stearns ME. Evidence for increased activated metallo- modality therapy in pancreatic cancer. Curr Oncol Rep 2002;4:202–12. proteinase 2 (MMP-2a) expression associated with human prostate 40. Whittaker M, Floyd CD, Brown P, Gearing AJ. Design and thera- cancer progression. Oncol Res 1996;8:69–75. peutic application of matrix metalloproteinase inhibitors. Chem Rev 61. Kawano N, Osawa H, Ito T, et al. Expression of gelatinase A, tissue 1999;99:2735–76. inhibitor of metalloproteinases-2, matrilysin, and trypsin(ogen) in lung

neoplasms: an immunohistochemical study. Hum Pathol 1997;28: ble Fas ligand and potentiates epithelial cell apoptosis. Curr Biol 1999; 613–22. 9:1441–7. 62. Remacle A, McCarthy K, Noel A, et al. High levels of TIMP-2 82. Agnihotri R, Crawford HC, Haro H, Matrisian LM, Havrda MC, correlate with adverse prognosis in breast cancer. Int J Cancer 2000;89: Liaw L. Osteopontin, a novel substrate for matrix metalloproteinase-3 118–21. (stromelysin-1) and matrix metalloproteinase-7 (matrilysin). J Biol 63. Gress TM, Muller-Pillasch F, Lerch MM, Friess H, Buchler M, Chem 2001;276:28261–7. Adler G. Expression and in-situ localization of genes coding for extra- 83. Gearing AJ, Beckett P, Christodoulou M, et al. Processing of cellular matrix proteins and extracellular matrix degrading proteases in tumour necrosis factor-␣ precursor by metalloproteinases. Nature pancreatic cancer. Int J Cancer 1995;62:407–13. (Lond) 1994;370:555–7. 64. von Marschall Z, Riecken EO, Rosewicz S. Stromelysin 3 is over- 84. Haro H, Crawford HC, Fingleton B, Shinomiya K, Spengler DM, expressed in human pancreatic carcinoma and regulated by retinoic acid Matrisian LM. Matrix metalloproteinase-7-dependent release of tumor in pancreatic carcinoma cell lines. Gut 1998;43:692–8. necrosis factor-␣ in a model of herniated disc resorption. J Clin Investig 65. Maatta M, Soini Y, Liakka A, Autio-Harmainen H. Differential 2000;105:143–50. expression of matrix metalloproteinase (MMP)-2, MMP-9, and mem- 85. Schmid RM. Acinar-to-ductal metaplasia in pancreatic cancer de- brane type 1-MMP in hepatocellular and pancreatic adenocarcinoma: velopment. J Clin Investig 2002;109:1403–4. implications for tumor progression and clinical prognosis. Clin Cancer 86. Vargo-Gogola T, Crawford HC, Fingleton B, Matrisian LM. Iden- Res 2000;6:2726–34. tification of novel matrix metalloproteinase-7 (matrilysin) cleavage sites 66. Yamamoto H, Itoh F, Iku S, et al. Expression of matrix metallo- in murine and human Fas ligand. Arch Biochem Biophys 2002;408: proteinases and tissue inhibitors of metalloproteinases in human pan- 155–61. creatic adenocarcinomas: clinicopathologic and prognostic significance 87. Tanaka M, Itai T, Adachi M, Nagata S. Downregulation of Fas of matrilysin expression. J Clin Oncol 2001;19:1118–27. ligand by shedding. Nat Med 1998;4:31–6. 67. Crawford HC, Scoggins CR, Washington MK, Matrisian LM, 88. Mitsiades N, Yu WH, Poulaki V, Tsokos M, Stamenkovic I. Matrix Leach SD. Matrix metalloproteinase-7 is expressed by pancreatic cancer metalloproteinase-7-mediated cleavage of Fas ligand protects tumor precursors and regulates acinar-to-ductal metaplasia in exocrine pan- cells from chemotherapeutic drug cytotoxicity. Cancer Res 2001;61: creas. J Clin Investig 2002;109:1437–44. 577–81. 68. Coussens LM, Fingleton B, Matrisian LM. Matrix metalloprotein- 89. Bair EL, Massey CP, Tran NL, et al. Integrin- and cadherin- ase inhibitors and cancer: trials and tribulations. Science (Wash D C) mediated induction of the matrix metalloprotease matrilysin in cocul- 2002;295:2387–92. tures of malignant oral squamous cell carcinoma cells and dermal 69. Fleming ID, Greene, FL, Balch CM, et al. AJCC Cancer staging fibroblasts. Exp Cell Res 2001;270:259–67. handbook: TNM classification of malignant tumours, 6th edition. New 90. Logsdon CD, Simeone DM, Binkley C, et al. Molecular profiling of York: Springer-Verlag, 2002. pancreatic adenocarcinoma and chronic pancreatitis identifies multiple 70. Newell KJ, Matrisian LM, Driman DK. Matrilysin (matrix metal- genes differentially regulated in pancreatic cancer. Cancer Res 2003;63: loproteinase-7) expression in ulcerative colitis-related tumorigenesis. 2649–57. Mol Carcinog 2002;34:59–63. 91. Davies G, Jiang WG, Mason MD. Matrilysin mediates extracellular 71. McDonnell S, Navre M, Coffey RJ Jr, Matrisian LM. Expression cleavage of E-cadherin from prostate cancer cells: a key mechanism in and localization of the matrix metalloproteinase pump-1 (MMP-7) in hepatocyte growth factor/scatter factor-induced cell-cell dissociation human gastric and colon carcinomas. Mol Carcinog 1991;4:527–33. and in vitro invasion. Clin Cancer Res 2001;7:3289–97. 72. Pajouh MS, Nagle RB, Breathnach R, Finch JS, Brawer MK, 92. Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. A causal Bowden GT. Expression of metalloproteinase genes in human prostate role for E-cadherin in the transition from adenoma to carcinoma. Nature cancer. J Cancer Res Clin Oncol 1991;117:144–50. (Lond) 1998;392:190–3. 73. Nakano A, Tani E, Miyazaki K, Yamamoto Y, Furuyama J. Matrix 93. Borchers AH, Sanders LA, Bowden GT. Regulation of matrilysin metalloproteinases and tissue inhibitors of metalloproteinases in human expression in cells of squamous cell carcinoma by E-cadherin-mediated gliomas. J Neurosurg 1995;83:298–307. cell-cell contact. J Cancer Res Clin Oncol 1997;123:13–20. 74. Ougolkov AV, Yamashita K, Mai M, Minamoto T. Oncogenic 94. Basset P, Wolf C, Rouyer N, Bellocq JP, Rio MC, Chambon P. ␤-catenin and MMP-7 (matrilysin) cosegregate in late-stage clinical Stromelysin-3 in stromal tissue as a control factor in breast cancer colon cancer. Gastroenterology 2002;122:60–71. behavior. Cancer (Phila) 1994;74:1045–9. 75. Wilson CL, Matrisian LM. Matrilysin: an epithelial matrix metal- 95. Muller D, Wolf C, Abecassis J, et al. Increased stromelysin 3 gene loproteinase with potentially novel functions. Int J Biochem Cell Biol expression is associated with increased local invasiveness in head and 1996;28:123–36. neck squamous cell carcinomas. Cancer Res 1993;53:165–9. 76. Powell WC, Knox JD, Navre M, et al. Expression of the metallo- 96. Ahmad A, Hanby A, Dublin E, et al. Stromelysin 3: an independent proteinase matrilysin in DU-145 cells increases their invasive potential prognostic factor for relapse-free survival in node-positive breast cancer in severe combined immunodeficient mice. Cancer Res 1993;53: and demonstration of novel breast carcinoma cell expression. Am J 417–22. Pathol 1998;152:721–8. 77. Kolligs FT, Bommer G, Goke B. Wnt/␤-catenin/tcf signaling: a 97. Linder C, Bystrom P, Engel G, et al. Correlation between basic critical pathway in gastrointestinal tumorigenesis. Digestion 2002;66: fibroblast growth factor immunostaining of stromal cells and stromely- 131–44. sin-3 mRNA expression in human breast carcinoma. Br J Cancer 1998; 78. Crawford HC, Fingleton BM, Rudolph-Owen LA, et al. The met- 77:941–5. alloproteinase matrilysin is a target of ␤-catenin transactivation in 98. Engel G, Heselmeyer K, Auer G, Backdahl M, Eriksson E, Linder intestinal tumors. Oncogene 1999;18:2883–91. S. Correlation between stromelysin-3 mRNA level and outcome of 79. von Bredow DC, Nagle RB, Bowden GT, Cress AE. Cleavage of ␤ human breast cancer. Int J Cancer 1994;58:830–5. 4 integrin by matrilysin. Exp Cell Res 1997;236:341–5. 99. Hahnel E, Harvey JM, Joyce R, Robbins PD, Sterrett GF, Hahnel R. 80. Noe V, Fingleton B, Jacobs K, et al. Release of an invasion Stromelysin-3 expression in breast cancer biopsies: clinico-pathological promoter E-cadherin fragment by matrilysin and stromelysin-1. J Cell correlations. Int J Cancer 1993;55:771–4. Sci 2001;114:111–8. 100. van der Stappen JW, Hendriks T, Wobbes T. Correlation between 81. Powell WC, Fingleton B, Wilson CL, Boothby M, Matrisian LM. collagenolytic activity and grade of histological differentiation in colo- The metalloproteinase matrilysin proteolytically generates active solu- rectal tumors. Int J Cancer 1990;45:1071–8.

101. Koshiba T, Hosotani R, Wada M, et al. Involvement of matrix 108. Mimori K, Mori M, Shiraishi T, et al. Analysis of ornithine metalloproteinase-2 activity in invasion and metastasis of pancreatic decarboxylase messenger ribonucleic acid expression in colorectal car- carcinoma. Cancer (Phila) 1998;82:642–50. cinoma. Dis Colon Rectum 1997;40:1095–100. 102. Imamura T, Ohshio G, Mise M, et al. Expression of membrane- 109. Yoshiji H, Gomez DE, Thorgeirsson UP. Enhanced RNA expres- type matrix metalloproteinase-1 in human pancreatic adenocarcinomas. sion of tissue inhibitor of metalloproteinases-1 (TIMP-1) in human J Cancer Res Clin Oncol 1998;124:65–72. breast cancer. Int J Cancer 1996;69:131–4. 103. Ellenrieder V, Alber B, Lacher U, et al. Role of MT-MMPs and 110. Yang TT, Hawkes SP. Role of the 21-kDa protein TIMP-3 in MMP-2 in pancreatic cancer progression. Int J Cancer 2000;85:14–20. oncogenic transformation of cultured chicken embryo fibroblasts. Proc 104. Hernandez-Barrantes S, Toth M, Bernardo MM, et al. Binding of Natl Acad Sci USA 1992;89:10676–80. active (57 kDa) membrane type 1-matrix metalloproteinase (MT1-MMP) to 111. Ahonen M, Baker AH, Kahari VM. Adenovirus-mediated gene tissue inhibitor of metalloproteinase (TIMP)-2 regulates MT1-MMP processing and pro-MMP-2 activation. J Biol Chem 2000;275:12080–9. delivery of tissue inhibitor of metalloproteinases-3 inhibits invasion and induces apoptosis in melanoma cells. Cancer Res 1998;58:2310–5. 105. Kim JH, Kim TH, Jang JW, Jang YJ, Lee KH, Lee ST. Analysis of matrix metalloproteinase mRNAs expressed in hepatocellular carci- 112. Levitt NC, Eskens FA, O’Byrne KJ, et al. Phase I and pharmaco- noma cell lines. Mol Cells 2001;12:32–40. logical study of the oral matrix metalloproteinase inhibitor, MMI270 106. Ohira T, Akutagawa S, Usuda J, et al. Up-regulated gene expres- (CGS27023A), in patients with advanced solid cancer. Clin Cancer Res sion of angiogenesis factors in post-chemotherapeutic lung cancer tis- 2001;7:1912–22. sues determined by cDNA macroarray. Oncol Rep 2002;9:723–8. 113. Heath EI, O’Reilly S, Humphrey R, et al. Phase I trial of the matrix 107. Watanabe M, Takahashi Y, Ohta T, Mai M, Sasaki T, Seiki M. metalloproteinase inhibitor BAY12–9566 in patients with advanced Inhibition of metastasis in human gastric cancer cells transfected with solid tumors. Cancer Chemother Pharmacol 2001;48:269–74. tissue inhibitor of metalloproteinase 1 gene in nude mice. Cancer (Phila) 114. Brown PD. Ongoing trials with matrix metalloproteinase inhibi- 1996;77:1676–80. tors. Expert Opin Investig Drugs 2000;9:2167–77.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2004 American Association for Cancer Research. Comprehensive Analysis of Matrix Metalloproteinase and Tissue Inhibitor Expression in Pancreatic Cancer: Increased Expression of Matrix Metalloproteinase-7 Predicts Poor Survival

Lucie E. Jones, Michelle J. Humphreys, Fiona Campbell, et al.

Clin Cancer Res 2004;10:2832-2845.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/10/8/2832

Cited articles This article cites 106 articles, 24 of which you can access for free at: http://clincancerres.aacrjournals.org/content/10/8/2832.full#ref-list-1

Citing articles This article has been cited by 24 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/10/8/2832.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/10/8/2832. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.