This article was published in an Elsevier journal. The attached copy is furnished to the author for non-commercial research and education use, including for instruction at the author’s institution, sharing with colleagues and providing to institution administration. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
w w w. e l s e v i e r. c o m / l o c a t e / y e x c r
Research Article
Intracellular co-localization of trypsin-2 and matrix metalloprotease-9: Possible proteolytic cascade of trypsin-2, MMP-9 and enterokinase in carcinoma Suvi-Tuuli Vilena , Pia Nybergb , Mika Hukkanenc , Meeri Sutinenb , Merja Ylipalosaarib , Anders Bjartelld , Annukka Pajue , Virpi Haaparantaa , Ulf-Håkan Stenmane , Timo Sorsaa , Tuula Salob,f,⁎ a
Institute of Dentistry, University of Helsinki, Department of Oral and Maxillofacial Diseases, Helsinki University Hospital, Finland Department of Diagnostics and Oral Medicine, Institute of Dentistry, University of Oulu, Finland c Institute of Biomedicine/Anatomy, University of Helsinki, Finland d Department of Clinical Sciences, Division of Urological Research, Malmö University Hospital, Lund University, Sweden e Department of Clinical Chemistry in Biomedicum, Helsinki University Central Hospital, Finland f Oulu University Hospital, Finland b
ARTICLE INFORMATION
ABS T R AC T
Article Chronology:
Tumor-associated trypsin-2 and matrix metalloprotease-9 (MMP-9) are associated with
Received 13 August 2007
cancer, particularly with invasive squamous cell carcinomas. They require activation for
Revised version received
catalytical competence via proteolytic cascades. One cascade is formed by enterokinase,
25 October 2007
trypsin-2 and MMP-9; enterokinase activates trypsinogen-2 to trypsin-2, which is an efficient
Accepted 25 October 2007
proMMP-9 activator. We describe here that oral squamous cell carcinomas express all
Available online 12 November 2007
members of this cascade: MMP-9, trypsin-2 and enterokinase. The expression of enterokinase in a carcinoma cell line not derived from the duodenum was shown here for
Keywords:
the first time. Enterokinase directly cleaved proMMP-9 at the Lys65-Ser66 site, but failed to
Enterokinase
activate it in vitro. We demonstrated by confocal microscopy that MMP-9 and trypsin-2 co-
Intracellular vesicle
localized in intracellular vesicles of the carcinoma cells. This co-localization of trypsin-2 and
MMP-9
MMP-9 resulted in intracellular proMMP-9 processing that represented fully or partially
Oral carcinoma
activated MMP-9. However, although both proteases were present also in various bone tumor
Proteolysis
tissues, MMP-9 and trypsin-2 never co-localized at the cellular level in these tissues. This
Trypsin-2
suggests that the intracellular vesicular co-localization, storage and possible activation of these proteases may be a unique feature for aggressive epithelial tumors, such as squamous cell carcinomas, but not for tumors of mesenchymal origin. © 2007 Elsevier Inc. All rights reserved.
⁎ Corresponding author. Department of Diagnostics and Oral Medicine, Institute of Dentistry, PO BOX 5281, University of Oulu, FIN-90014 Oulu, Finland. Fax: +358 8 537 5560. E-mail address:
[email protected] (T. Salo). Abbreviations: AEC, 3-amino-9-ethylcarbazole; APMA, p-aminophenyl mercuric acetate; BSA, bovine serum albumin; DAB, 3,3diaminobenzidine tetrahydrochloride; ECL, enhanced chemilumisence; ECM, extracellular matrix; mAb, monoclonal antibody; MMP, matrix metalloprotease; OSCC, oral squamous cell carcinoma; pAb, polyclonal antibody; PBS, phosphate buffered saline; RT-PCR, reverse transcriptase polymerase chain reaction; SDS-PAGE, sodium dodecyl sulphate polyacrylamide gel electrophoresis; TAT, tumor associated trypsinogen 0014-4827/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2007.10.025
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
Introduction Oral squamous cell carcinomas (OSCCs) are aggressive tumors with high potential for metastasis; the overall five-year survival rate of patients with oral cancer is only approximately 50% and has not changed substantially over the last decades [1,2]. A better understanding of the biological characteristics of oral cancer is important for the development of novel diagnostic and curative strategies. One of the critical steps in all tumor growth is the proteolytic processing of the extracellular matrix (ECM) environment. Serine proteases and matrix metalloproteases (MMPs) are families of proteolytic enzymes involved in physiological and pathological processing of ECM and basement membranes. Trypsinogens are effective serine proteases that play a significant role in tumor progression although originally characterized as digestive enzymes in the pancreas [3,4]. La Bombardi et al. identified a trypsin-like protease in the cell membrane of Walker-256 carcinoma cells [5]. Tumor-associated trypsinogens were first identified in ovarian cancer [6] and have been described to be expressed in several other carcinomas as well, e.g. pancreatic [7], hepatocellular, cholangio [8] and colorectal carcinomas [9] in addition to various cancer cell lines, such as colon carcinoma, fibrosarcoma and erythroleukemia [10]. Two isoforms of tumorassociated trypsinogens, trypsin-1 and trypsin-2 have been well characterized [11]. Trypsin-2 is the predominant form in tumors [3], and its levels correlate with the malignancy and metastatic potential [11]. Trypsin-2 expression has been shown to be strongly associated with recurrence and poor prognosis in esophageal squamous cell carcinoma [12] and colorectal cancer [13]. Furthermore, locally produced trypsin-2 may act in a paracrine way to promote angiogenesis and tumor invasion by activating and degrading other proteases [4]. Matrix metalloproteases (MMPs) are a family of zincdependent endopeptidases capable of cleaving and degrading almost all extracellular matrix and basement membrane components [14]. They are classified into collagenases, gelatinases, stromelysins and membrane type (MT)-MMPs based on their structure and substrate specificity [15]. Particularly MMP9 (gelatinase B) has an important role in tumor growth, invasion and metastasis [16–19]. In addition to tumor growth, MMPs play a role in various biological events including bone morphogenesis and turnover [20,21]. Both bone destruction and formation is characteristic for malignant bone tumors. Therefore it is not surprising that several types of bone tumors, including osteosarcomas, giant cell tumor of bone and other osteolytic lesions show high expression of MMP-9 [22,23]. Both trypsinogens and MMPs are secreted as latent proenzymes that need activation to be catalytically competent. The activation process transforming trypsinogens into active trypsins involve a serine protease, enterokinase / enteropeptidase [24]. Activated trypsin can activate other latent proteases to form proteolytic cascades [4]. Our previous work described proMMP-9 activation by trypsin-2 in vitro at an extremely low molar ratio, and less efficiently proMMP-2 activation [25]. In addition to the gelatinases, trypsin-2 can in vitro activate also MMP-1, -3, -8 and -13 [26]. The proteolytic cascades can form complicated networks. It has been found that enterokinase is also capable of activating proMMP-9 in vitro [27] and thus can be
915
involved in MMP-9 activation both directly and via trypsin-2. This MMP-9 – trypsin-2 -activation cascade seems to be significant also in vivo. In trypsin-2 transfected OSCC cell lines the activation of proMMP-9 increased significantly leading to increased intravasation of the OSCC cells [28]. The levels of MMP-9 and trypsin-2 are significantly higher in malignant than in benign ovarian cyst fluids which suggest that trypsin may play a role in the regulation of the MMP-dependent proteolysis associated with invasion and metastasis of cancer [29]. The proteolytic cascades are complex, and potentially the activation of one protease can lead to increased activity of several downstream proteases leading to massive uncontrolled proteolysis. Therefore the cells need to strictly control the proteolytic activities. One way of achieving that is to control the location of proteases as well as the protease activation. Effective proteolysis will only occur when the activators and target proteases are in close proximity to each other. In human melanoma cells, MMP-9 is transported and stored in small cytoplasmic vesicles that are associated with microtubules and molecular motor protein kinesin indicating that these vesicles are actively propelled along microtubules towards the plasma membrane [30]. In this study, we describe that OSCCs express both trypsin-2 and enterokinase in addition to MMP-9. Enterokinase from OSCCs was further demonstrated to directly process proMMP-9 by cleaving the Lys65-Ser66 bond. Trypsin-2 and MMP-9 are compartmentalized in vesicular structures within the carcinoma cells. On the other hand, in osteosarcoma tissues trypsin-2 and MMP-9 showed differential localization at the cellular level. Based on our previous findings the proteolytic cascades involving trypsin-2 and MMP-9 seem to be very important for the invasive capacity of the cancers. The proteolytic cascade involving enterokinase/trypsin-2 activation of MMP-9 seems to be biologically relevant in carcinomas but possibly not in mesenchymal malignancies. This cascade may partially explain the aggressive nature of oral squamous cell carcinomas.
Materials and methods Tissue samples Representative human mobile tongue squamous cell carcinoma (n = 10), osteosarcoma (n = 12), giant cell lesion (also called reparative or central giant cell granuloma) from the mandible (n = 2) and giant cell tumor (also called osteoclastoma) from vertebra (n = 3) samples were collected from the files of the Department of Pathology, Oulu University Hospital. The classification and nomenclature of the tumors were done according to the WHO Classification of Tumors [31]. The tumors were used for in situ hybridization and immunohistochemistry. All samples were fixed in 10% (v/v) buffered formalin, routinely processed, embedded in paraffin and haematoxylin-eosin-stained for histopathology. Fresh giant cell lesion, obtained by surgery at the Oral and Maxillofacial Department, Oulu University Hospital, was used for measurement of gelatinases by zymography, and trypsin-2 by Western blotting (see below). The study was approved by The Ethical Committee of the Northern Ostrobothnia Hospital District.
Author's personal copy 916
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
Cell cultures The human tongue squamous cell carcinoma cell line HSC-3 was purchased from JCRB Cell Bank (JCRB 0623, Osaka National Institute of Health Sciences, Japan). The cells were cultured in a humid atmosphere of 5% CO2 and 95% air at 37 °C in 1:1 DMEM (Cambrex, Verviers, Belgium) and Ham's Nutritient Mixture F-12 supplemented with 10% heat-inactivated fetal calf serum, 1 mM sodium pyruvate, 250 ng/ml fungizone (Invitrogen, Carlsbad, CA), 100 units/ml penicillin, 100 μg/ml streptomycin, 2 mM L-Glutamine and 0.4 ng/ml hydrocortisone (Sigma-Aldrich, Ayrshire, United Kingdom).
Labs) was used according to the product protocol to intensify the stain end-product in some staining instances. For double immunohistochemical stainings, the giant cell lesion and tumor samples were first immunostained for trypsin-2 using the AEC method. After AEC staining the sections were washed for 10 min and processed for a second immunostaining by blocking nonspecific binding sites and incubating with the second primary antibody. After incubation with the avidinbiotin-peroxidase complex the samples were stained with SGcolor (Vector Labs). For morphologic examination routine haematoxylin was used.
Tartrate-resistant acid phosphatase staining Antibodies Monoclonal (mAb) and polyclonal (pAb) trypsin-2 antibodies have been previously characterized by Itkonen et al. [32] and Koivunen et al. [10,11], respectively. Monoclonal MMP-9 antibody recognizing both the latent and active forms of MMP-9 was described in Nikkari et al. [33], and a monoclonal MMP-9 antibody recognizing only the active form of the enzyme (antibody kindly provided by Dr. John E. Fothergill) was described previously [34]. The other antibodies were commercially available: polyclonal anti-MMP-9 recognizing both the active and latent forms was from Calbiochem (Darmstadt, Germany), monoclonal anti-enterokinase from R&D Systems (Minneapolis, MN), Alexa Fluor goat anti-rabbit 568 and Alexa Fluor goat anti-mouse 488 were from Molecular Probes/ Invitrogen (Eugene, OR). Monoclonal antibodies to the lymphocyte marker Leucocyte Common Antigen (LCA) and the macrophage-marker CD68 (clone KP1) were purchased from Dako (Glostrup, Denmark).
Immunohistochemical staining of tongue SCC and bone tissue sections 5 μm paraffin sections were used for immunohistochemical stainings. After pepsin treatment and blocking of the endogenous peroxidase activity by H2O2 in methanol, nonspecific binding of IgGs was blocked using fetal calf serum (1:5 in PBS) or normal goat or horse serum (1:50). The sections were incubated with primary antibodies overnight at +4 °C. The sections from giant cell lesions and tumors were incubated with polyclonal rabbit trypsinogen antibody [10,11] (dilution 1:1300), monoclonal MMP-9 antibody [33] (dilution 1:200), antibody recognizing only the active form of MMP-9 [34] (undiluted tissue culture supernatant), or monoclonal antibodies to LCA, a lymphocyte marker (dilution 1:50) or CD68, a macrophage marker (dilution 1:50). OSCC samples were subjected to trypsin-2 immunohistochemistry using a monoclonal antibody (clone 6D11, dilution 1:500) [32]. Negative controls were included (PBS or nonimmune rabbit serum (dilution 1:200) instead of the primary antibody). The sections were incubated (30 min) with biotinylated anti-mouse or anti-rabbit IgG secondary antibody (Dako and Vector Labs, Burlingame, CA), the avidin-biotin-peroxidase complex (30 min), and the substrate (0.05% 3,3-diaminobenzidine tetrahydrochloride (DAB) and H2O2 in Tris-buffer, pH 7.4, (Sigma) or 3-amino-9ethylcarbazole (AEC; Sigma) 0.3 mg/ml in 0.05 M sodium acetate, pH 5.5) for 10 min. DAB enhancing solution (Vector
As a marker for osteoclasts, tartrate-resistant acid phosphatase (TRAP) staining was performed to some sections according to Thompson [35] and using a TRAP kit (Sigma). Briefly, the deparaffinized slides were incubated in 50 mM tartrate in acetate buffer (pH 5.2) at 37 °C for 2 h. Then the sections were incubated in an acid phosphatase substrate buffer (25% Michaelis acetate buffer, 0.16% paraorsaniline, 0.16% NaNO2, 0.05% naphtol AS-BI phosphate), pH 5.0 with 20 mM tartrate at 37 °C for 60 min, washed and mounted.
Analysis of gelatinases by zymography A giant cell lesion tissue was homogenized in sample buffer (4x loading buffer: 0.25 M Tris-HCl, pH 6.8, 8% SDS, 40% glycerol, 0.0096% bromophenol blue). When HSC-3 cell extracts were analyzed, total protein was extracted from the cells, after washing the cells twice with PBS, by lysis buffer (50 mM Tris, 10 mM CaCl2 2H2O, 150 mM NaCl, 0.05% Brij-35, pH 7.5). 10 μg of the cell extract per lane was loaded for zymography. The gelatinases from the tissue and cell extracts were analyzed by the zymography using 1.5 mm 10% polyacrylamide slab gels containing 1 mg/ml fluorescently labeled (2-methoxy-2,4-diphenyl-3(2H)-furanone (Fluka, Ronkonkoma, NY) gelatin as a substrate [36]. Prestained low molecular weight marker (Bio-Rad Laboratories, Hercules, CA) was used.
Western blot Total proteins were extracted from the HSC-3 cells cultured in 250 mm2 dishes and from snap-frozen giant cell lesion samples by the Trizol® method (Invitrogen, Carlsbad, CA). Protein samples of 15 μg and 30 μg were separated on 11% SDSPAGE gels and electrophoretically transferred to a nitrocellulose membrane (Bioscience, Dassel Germany). Nonspecific binding was blocked with 5% non-fat dry milk for 1 h at 37 °C. The membranes were incubated with anti-trypsin-2 mAb (dilution 1:25) [32] and anti-enterokinase mAb (dilution 1:200) (R&D Systems) overnight at RT. After washing, the membranes were incubated with the secondary peroxidase-conjugated anti-mouse IgG antibody (dilution 1:800) (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom) for 1 h at RT. An ECL Western blotting detection kit (Amersham Pharmacia Biotech) was used to visualize the proteins as described by the manufacturer. In Western blotting experiments of the giant cell lesion tissue extracts, the membranes were incubated
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
with polyclonal trypsin antibody (dilution 1:1300) [10,11] at 4 °C overnight, and the secondary antibody was biotinylated anti-rabbit IgG (diluted 1:500, Dako) followed by incubation of avidin-biotin-peroxidase complex (Dako).
Immunolocalisation of MMP-9 and trypsin-2 in oral squamous cell carcinoma cells in vitro 1 × 105 HSC-3 carcinoma cells were plated onto coverslips and grown for 24 h in complete media at 37 °C in 5% CO2. To block the protein secretion the cells were incubated in media containing 1 μM monensin for 1 h and 3 h. The cells were rinsed with PBS, fixed in methanol at -20 °C for 10 min, washed and stored in PBS containing 0.02% sodium-azide at +4 °C until analyzed. The coverslips were rinsed three times with PBS. To block nonspecific binding sites the cells were incubated in 5% normal goat serum (Dako) in PBS containing 3% BSA for 1 h at room temperature. Primary mAb against trypsin-2 (1:40 dilution in PBS-3% BSA) and pAb against MMP-9 (1:60 dilution, Calbiochem) were incubated for 30 min at RT and rinsed 3 times in PBS. Secondary antibody conjugates Alexa Fluor 488 and 568 (Molecular Probes/Invitrogen) were diluted in a PBS0.1% BSA (1:400) and incubated 30 min at RT. The coverslips were rinsed in PBS and in some instances incubated for 10 min with the DNA-specific probe TO-PRO-3 (Molecular Probes/ Invitrogen) at RT, washed and mounted with Vectashield H1000 (Vector Labs).
Laser scanning confocal microscopy, deconvolution and co-localization analysis Confocal microscopy was carried out using a Leica TCS SP2 system (Leica Microsystems AG, Mannheim, Germany) with HCX PL APO CS 63/ 1.40 NA objective, and 488, 568 and 633 nm laser excitation for Alexa Fluor 488 and 568 conjugates, and TO-PRO-3, respectively. Stacks including six images each were collected using sequential scanning to avoid channel crosstalk, and with 1024x1024 lateral resolution and optimized zsampling density. In order to facilitate 3-D co-localization analysis, the image stacks were deconvolved with 10 iterations using theoretical point spread function and maximum likelihood estimation algorithms of Huygens Essential (Scientific Volume Imaging BV, Hilversum, The Netherlands). 3-D colocalization analyses of trypsin-2 and MMP-9 were performed from approximately 50 cells per group using automatic threshold co-localization algorithm by Costes and Locket [37] implemented in Bitplane Imaris suite (Bitplane AG, Zurich, Switzerland). The results are expressed as a percentage of channel volume co-localization (co-compartmentalization) for trypsin-2 and MMP-9 immunoreactivities, and as a channel correlation in co-localized volume using Pearson's correlation coefficiencies [37].
Riboprobe synthesis for in situ hybridization In vitro transcriptions of sense and antisense trypsinogen probes were made by fluorescein-UTP riboprobe synthesis using the RNA color kit (Amersham Pharmacia Biotech) as described [38]. As a template, a 627-bp long trypsinogen-2 cDNA fragment (corresponding to nucleotides 42-688, acces-
917
sion number M27602 [39]) was cloned from COLO 205 cells by RT-PCR using the TA Cloning Kit (Invitrogen) and the following primers: 5′ – TGC TGT TGC TGC CCC CTT TG – 3′ (sense) and 5′ – GCA CAG CCA TAG CCC CAG GAG – 3′ (antisense). The integrity and length of the probes was determined by gel electrophoresis. The antisense probe is presumed to recognize both trypsinogen-1 and -2 transcripts in in situ hybridization experiments as concluded from extensive similarity of their nucleotide sequences. For the MMP-9 probe, a 574 bp HindIII – EcoRI fragment of human MMP-9 cDNA [40] ligated into a pGEM 4Z vector was used. The probe was tested by Northern hybridization and sequencing.
In situ hybridization All reagents for trypsinogen in situ hybridization were purchased from Sigma-Aldrich and Amersham Pharmacia Biotech. Tissue specimens were fixed, paraffin embedded, sectioned (4 μm), dried for 2 h at 65 °C and mounted on SuperFrost™ plus slides (Menzel-Gläser, Braunschweig, Germany) under RNase free conditions. The sections were deparaffinized in xylene, rehydrated, and processed as described [38]. Briefly, sections were pretreated with 0.2 M HCl to abolish endogenous enzyme activity, digested with proteinase K (20 μg/ml in 20 mM Tris-HCl, 2 mM CaCl2, pH 7.5) for 25 min at 37 °C. After prehybridization with 40 μl of hybridization buffer containing 50% (v/v) formamide, 10 mM Tris-HCl pH 7.6, 1x Denhardt's solution (BSA, polyvinylpyrrolidone and Ficoll, all at 0.2 mg/ml), 2x SSC, and 0.4 μg/ml salmon sperm DNA at 55°C for 1 h, the slides were hybridized with 40 μl of 250 ng/ml antisense or sense probe in hybridization buffer first for 8 min at 85 °C and then for 16 h at 55°C. After hybridization, the slides were washed at a stringency of 0.1x SSC at 60 °C (4 × 15 min) and then equilibrated in TBS (100 mM Tris-HCl, 0.4 M NaCl, pH 7.5). For detection of hybridization signals, tissue sections were first incubated in blocking reagent and subsequently incubated with antifluorescein alkaline phosphatase conjugate (Amersham Pharmacia Biotech) diluted 1:1000 in TBS containing 0.5% (w/v) BSA for 2 h at RT. Hybridization signals were visualized with levamisol, NBT (nitro blue tetrazolium chloride) and BCIP (5bromo-4-chloro-3-indolyl phosphate). The color reaction was stopped after 2-8 h and the slides were coverslipped using Faramount™ mounting medium (DakoCytomation, Dako). The MMP-9 in situ hybridizations in the bone tissues were carried out as previously described [41].
Extraction of RNA from oral carcinoma cells for RT-PCR detection of enterokinase and trypsin-4 For nested RT-PCR analysis of enterokinase expression, total RNA was extracted from the HSC-3 cells. Purification of RNA was done according to the instructions accompanying the Trizol® kit (Invitrogen). The cDNA-reaction was done from 2 μg of total RNA with primers from SuperScriptIII-kit (Invitrogen) according to the manufacturer's instructions. Specific primers for amplifying enterokinase were 5′-TTGTTGTTCGTTGTGCCATT and 5′-ACCAACTTTGGTGCCAACTC (outer-primers), and5′-GAAAATGGTCTGCCTTGCAT and 5′- TCTGGCTTTCTGTGTTTGGA (inner-primers). The PCR reaction was performed
Author's personal copy 918
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
with 2U Dynazyme polymerase (Finnzymes, Espoo, Finland), 15 pmol of each primer and 200 μM dNTP in a final volume of 50 μl of 1 × PCR buffer. After initial denaturation (2 min at 94 °C) 35 cycles were performed (20 sec 94 °C, 20 sec 63 °C and 1 min 72 °C). The PCR products were run on 1.25% agarose gel. The enterokinase PCR product was purified according to the instructions for the E.Z.N.A. ® Microspin Gel Extraction Kit (Omega Bio-tek, Doraville, GA). Sequencing was performed using cycle sequencing with the Big Dye Terminator kit (version 3.1; Applied Biosystems) using the ABI 3130 capillary sequencer. For trypsin-4 PCR analysis 100 ng of total RNA and 2 pmol of gene specific primers were used and the RT reaction was performed according to the instructions of SuperScriptIIIkit (Invitrogen). The trypsin-4 primers were 5′-CTT CTG GGT GGA CGC ACT TGG (sense) and 5′-GGG GGC TTT AGC TGT TGG CA (antisense). The PCR was performed with 0.6 U of AmpliTaq Gold (Applied Biosystems, Branchburg, NJ), 200 μM of dNTP and 20 pmol of each primer in total volume of 25 μl 1 × buffer, and performing 35 cycles (1 min 95 °C, 1 min at annealing temperature 60°C, 1 min 72 °C).
Gelatin degradation assay The activity of the cleaved MMP-9 was measured with the gelatin degradation assay [42], where the amount of degraded 125 I-labelled gelatin reflected the activity of the gelatinases. Briefly, 50 ng of recombinant human proMMP-9 (Invitek) was incubated with 0.15 IU human enterokinase (Prospec-Tany TechnoGene) at 37 °C for 90 min, 2 mM APMA at 37 °C for 60 min, or 25 mM EDTA at room temperature for 10 min. As controls, MMP-9 or enterokinase alone was used. To determine whether enterokinase cleavage makes MMP-9 more susceptible for further cleavage with APMA, proMMP-9 was first incubated with enterokinase as previously, and then incubated with APMA either 20 min or 60 min. The incubations were done in total volume of 20 μl of 50 mM Tris-HCl, pH 7.8, 200 mM NaCl, 1 mM CaCl2. Thereafter, all the samples were incubated with soluble 125I-labeled gelatin (1.5 μM) at 37 °C for 60 min. Undegraded gelatin was precipitated with 20% trichloroacetic acid. The radioactivity in the supernatants containing the degraded gelatin and reflecting gelatinase activity was counted with a γ-counter.
Cleavage of proMMP-9 by recombinant human enterokinase N-terminal sequencing of human recombinant MMP-9 1 μg of purified human recombinant proMMP-9 (Invitek, Berlin, Germany) was treated with 3 IU of recombinant human enterokinase (Prospec-Tany TechnoGene Ltd, Rehovot, Israel) for 0, 20, 40 and 90 minutes at 37 °C. As a control, proMMP-9 was incubated without enterokinase for 90 min at 37 °C. The reactions were terminated by adding SDS-sample buffer and boiling the samples for 5 min. Proteins were run to 11% SDSPAGE and stained with Coomassie Brilliant Blue.
Recombinant human MMP-9 cleaved by enterokinase was run on 8% SDS-PAGE and the cleavage product corresponding to the 77-82 kDa activated species of MMP-9 was identified. After SDS-PAGE the gel was washed 30 min in blotting buffer. The ProBlot (Applied Biosystems) membrane was soaked in 100% methanol for 5 min. Proteins were electrophoretically transferred to Problot at 40 mA for 3 h. The membrane was washed
Fig. 1 – Immunohistochemical localization of trypsin-2 in tongue SCC tissue samples. In all of the SCC samples (n = 10) trypsin-2 localized mainly in the peripheral cancer cells (A and B), but also throughout carcinoma cell islands and in some fibroblast-like cells of the stroma (see arrows). Macrophages scattered within the stroma were also positive (C, arrows). Pancreas was used as a positive control (D). Original magnification:× 100 (A),×200 (B and D) and×400 (C).
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
with distilled water and stained for 5 min in 0.1% Coomassie Brilliant Blue. The membrane was destained with 50% methanol, rinsed with distilled water, dried at RT and stored at -20 °C. The N-terminal sequence was analyzed with a Procise 492 protein sequencer (Applied Biosystems) using Edman chemistry at the Protein Sequencing Core Facility at the University of Oulu.
Statistical Analysis Results are expressed as the mean ± SEM of n experiments. Statistical significances were calculated using Student's
919
unpaired t-test and analysis of one-way variance (ANOVA) with Bonferroni's post-tests for multiple comparisons.
Results Expression of trypsin-2 in oral squamous cell carcinomas Tongue squamous cell carcinoma (SCC) tissue samples were analyzed for trypsin-2 immunoreactivity. Trypsin-2 positive cells were localized particularly in the peripheral carcinoma cells, but also throughout the cancer islands in all tongue SCC
Fig. 2 – In situ hybridization analysis for trypsin-2 in tongue SCC tissue samples and HSC-3 cells. In situ hybridization showed trypsin-2 mRNA expression in carcinoma cell islands (white arrows) (B), macrophage-like cells (black arrow) and fibroblast-like cells (black arrowheads) in all of the SCC samples analyzed (n = 10) (A and C), HSC-3 cells (black arrows) (n = 2) (E), and pancreas (positive control) (G). No signal was detected in control hybridizations with trypsin-2 sense probe in OSCC tissue (D), HSC-3 cells (F) or pancreas (H).
Author's personal copy 920
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
samples analyzed (n = 10) (Figs. 1A and B). Similar distribution has previously been revealed for MMP-9: positive staining in malignant epithelial cells and in inflammatory cells surrounding the tumor islands [43]. Moreover, positive staining for trypsin-2 was sporadically found in fibroblasts and macrophages of the surrounding tumor stroma (Figs. 1B-C). In situ hybridization showed that trypsin-2 mRNA was expressed in carcinoma cells (Fig. 2B), stromal fibroblasts and macrophages in all the samples (n = 10) (Figs. 2A and C). Trypsin-2 mRNA was also detected in cultured HSC-3 cells embedded in paraffin (n = 2) (Fig. 2E). Pancreas tissue served as a positive control for trypsin-2 expression analysed by in situ hybridization and immunohistochemistry (Figs. 1D and 2G). No staining was observed in control hybridizations with trypsin-2 sense probe in OSCC tissue (Fig. 2D), HSC-3 cells (Fig. 2F) or pancreas (Fig. 2H).
Co-compartmentalization of trypsin-2 and MMP-9 We have previously shown that tumor-associated trypsin-2 can activate MMP-9 in vitro [25], and that HSC-3 carcinoma cells transfected with trypsin-2 secrete increased amounts of the active form of MMP-9, but not other MMPs analyzed, into the culture media, resulting in increased cell invasion in vivo [26,28]. With this background, we wished to examine by confocal microscopy and 3D co-localization analysis whether trypsin-2 and MMP-9 are co-compartmentalized within the HSC-3 carcinoma cells. This would give new insights how protease activation is regulated in the trypsin-2 – MMP-9 cascade. HSC-3 cells were analyzed for the presence and volume co-localization of trypsin-2 and MMP-9. We used cells that were incubated in the absence or presence (for 1 h or 3 h) of 1 μM monensin, an inhibitor that blocks intracellular
Fig. 3 – Co-compartmentalization of trypsin-2 and MMP-9 in HSC-3 cells. Immunocytochemical detection of trypsin-2 (A, green) and MMP-9 (B, red) in HSC-3 cells without monensin treatment. In the merged image the subjective impression of trypsin-2 and MMP-9 co-localization in intracellular vesicles is seen in yellow (subjective; C). When the degree of co-localization and the correlation coefficiencies are calculated on a voxel-to-voxel basis a separate co-localization channel is built and shown here in white (D). The effect of monensin, a blocker of intracellular transport of proteins and vesicle production, on co-compartmentalization of trypsin-2 and MMP-9 (E-H). HSC-3 cells were treated with monensin for 1 or 3 h, and double immunostained for trypsin-2 and MMP-9. The re-distribution of trypsin-2 (E) and MMP-9 (F) into separate intracellular vesicle after 3 h monensin treatment: subjective co-localization shown in yellow (G) and calculated co-localization channel is shown in white (H). Arrows in (G-H) indicate perinuclear accumulation of MMP-9 after 3 hours of monensin treatment. Pearson's correlation coefficiencies indicate the degree of co-localization and time-dependent redistribution of trypsin-2 and MMP-9 as a result of monensin blockage, ***P < 0.001 by ANOVA for intergroup analyses of correlation coefficiencies (I). The HSC-3 carcinoma cell extract contained both latent and cleaved MMP-9, but only latent MMP-2 by zymography (J, lane 1). APMA completely activated both MMP-9 and -2 (J, lane 2).
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
transport of proteins from the Golgi complex resulting in inhibition of transit of secretory vesicles, swelling of Golgi and accumulation of proteins into the endoplasmic reticulum and Golgi. In the absence of monensin (Figs. 3A-D), both trypsin-2 and MMP-9 were found predominantly to localize in small cytoplasmic vesicles. Co-localization analyses from approximately 50 random cells per group showed that in HSC-3 cells (in the absence of monensin) 58% (±2.0%) of MMP-9 immunoreactivity was co-compartmentalized with trypsin-2 and 36% (±4.8%) of trypsin-2 was co-compartmentalized with MMP-9
921
(Figs. 3A-D). Confocal microscopy and calculation of correlation coefficients showed that monensin treatment resulted in a time-dependent redistribution MMP-9 into separate vesicles (p < 0.001 at 3 h; Figs. 3E-I) that predominantly accumulated within the perinuclear area (Figs. 3G-H, white arrows). HSC-3 carcinoma cell extracts were used to determine whether there are processed forms of MMP-9 already within the cells. By gelatin zymography, we found that there was both pro- and cleaved, possibly active MMP-9 in the cells (Fig. 3J, lane 1). On the other hand, MMP-2 existed only in the latent form (Fig. 3J,
Fig. 4 – Differential localization of trypsin-2 and MMP-9 in bone lesions. Immunohistochemistry stained MMP-9 strongly in low-grade osteoblastic osteosarcoma (n = 12 ) (A, arrows) and in osteoclastic tumor cells (B, arrowheads) and some individual mononuclear tumor cells in giant cell lesion (n = 2 ) (B, arrows). Demonstration of MMP-9 and MMP-2 by zymography (inverted image) (C) and trypsin-2 expression by Western blot with monoclonal trypsin-2 antibody DB-102B1 (D) from homogenized giant cell lesion tissue extract. Molecular weight markers (kDa) are shown on the left. Trypsin-2 positive immunostaining distributed to inflammatory cells (arrows) and negative staining in mononuclear (arrowheads) and osteoclastic (open arrows) tumor cells (E). In (F) a negative osteoclastic tumor cell (open arrow), in (G) a polymorphonuclear (arrow) and in (H) a lymphocyte-like (arrow) appearance of trypsin-2 positive cells are shown with higher magnification. Trypsin-2 (red) and MMP-9 (blue) double immunostaining showing different cellular localization for trypsin-2 and MMP-9 (I). Inflammatory cells (arrows) were trypsin-2 positive and mononuclear (arrowheads) and osteoclastic (open arrows) tumor cells were MMP-9 positive (I). E, F, and H are from giant cell lesion (n = 2), and G and I from giant cell tumor (n = 3). All osteosarcoma, giant cell lesion and giant cell tumor samples were examined for MMP-9, and half of them for trypsin-2. Magnifications in A 20×, in B, E and I 140×and in F-H 700×.
Author's personal copy 922
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
lane 1). APMA caused a nearly complete activation of both MMP-9 and -2 (Fig. 3J, lane 2).
Localization of MMP-9 and trypsin-2 in bone tumors MMP-9 has been found to be very highly expressed and present both in active and latent forms in several benign and malignant bone tumors [22,23], especially in osteosarcomas (n = 12) (Fig. 4A), mono- and multinucleated osteoclast-like cells of giant cell lesions (n = 2) and tumors (n = 3) (Fig. 4B). Therefore, we further investigated the possible presence and co-localization of trypsin-2 with MMP-9 in benign and malignant bone tumors; giant cell lesion, giant cell tumor and osteosarcoma by in situ hybridization, zymography, Western blotting and immunohistochemistry. In giant cell lesion tissue extracts, active forms of MMP-9 and MMP-2 were detected by zymography (Fig. 4C). In addition, an antibody recognizing only the active form of MMP-9 [34] stained some osteoclastic cells in a giant cell tumor (not shown). All bone tumor samples examined showed positive staining for MMP-9 either with in situ hybridization (not shown) or immunohistochemistry. We selected approximately half of the samples with the most intense MMP-9 staining to examine whether trypsin-2 could be detected in the same tissues as well. By Western blotting we found that the giant cell lesion tissue extracts contained approximately 27-28 kDa trypsin-2 immunoreactivities and some higher molecular weight bands likely to represent trypsin-2 associated with protease inhibitors (Fig. 4D). However, in immunohistochemistry trypsin-2 was detected only in some lymphocyte-like and polymorphonuclear cells (Figs. 4E-H), whereas mononuclear and osteoclastic tumor cells and their macrophage precursors were consistently negative (Figs. 4F-H). To further identify trypsin-2 -containing cells within these tumors, double immunostaining for trypsin-2 and lymphocyte (LCA) and macrophage (CD68) markers were performed; trypsin2 immunoreactive cells were partially identified based on LCA expression, however, they were negative for CD68 indicating that they were not macrophages. In giant cell lesions, antibodies for MMP-9 labelled osteoclastic and tartrate resistant acid phosphatase positive cells (not shown). In addition, MMP-9 double immunostainings for trypsin-2 and MMP-9 provided further evidence that the two proteases localized in separate cell populations in the giant cell tumor tissue (Fig. 4I). For this reason, we also explored whether trypsin-2 and MMP-9 were differentially expressed in osteosarcomas. Unlike MMP-9 (Fig. 4A), trypsin-2 was not immunohistochemically detected in the tumor tissue in any of the osteosarcoma samples analyzed (not shown). In summary, this localization of MMP-9 and trypsin-2 in bone tumors of mesenchymal origin was different from the co-localization in oral squamous cell carcinoma samples.
Expression of enterokinase in HSC-3 oral carcinoma cells Enterokinase is a physiological activator of trypsinogens in the intestine [3]. Therefore we wished to elucidate whether the HSC-3 carcinoma cells express enterokinase. We used nested RT-PCR with specific enterokinase primers spanning an area of four exons. The PCR resulted in a 302-bp product (Fig. 5A). The identity of the product was confirmed by sequencing and
Fig. 5 – The expression of enterokinase in oral carcinoma cells and enterokinase cleavage on proMMP-9. The production of enterokinase in HSC-3 tongue carcinoma cells was demonstrated by nested RT-PCR (A); the 302-bp enterokinase PCR product (lane1) and 100-bp ladder standard (lane 2), and Western blotting (B); approximately 80 kDa heavy chain and about 40 kDa light chain of enterokinase were strongly stained with monoclonal enterokinase antibody. Molecular weight stardards are on the left. (C) Purified MMP-9 (1 μg) was treated with recombinant enterokinase (3 IU) for 0 min (lane 1), 20 min (lane 2), 40 min (lane 3) and 90 min (lane 4) and the resulting MMP-9-cleavage products were separated on a SDS-PAGE and visualized with Coomassie blue staining. (D) Protein sequencing showed that enterokinase cleaves proMMP-9 at the Lys65-Ser66 site resulting in an active form (aMMP-9) as seen in (C) after 40 and 90 minutes. (E) Enzyme activity assay results with proMMP-9 (50 ng) alone, proMMP9 treated with enterokinase (E) (0.15 IU), APMA (2 mM), both enterokinase and APMA, or EDTA (25 mM), and finally enterokinase alone. The results represent relative MMP-9 activity (±SEM) measured by counting the amount of degraded 125I-labeled gelatin (n = 4).
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
comparisons using the Clustalw multiple alignment program. In the negative control (no template), there was no PCRproduct (not shown). By Western blot we observed the presence of both enterokinase (Fig. 5B) and trypsin-2 (not shown) from total proteins extracted from HSC-3 cells. Attempts to demonstrate the localization of enterokinase in HSC-3 cells with confocal microscopy were not successful because the available antibodies were found to be unsuitable for immunofluorescence staining applications.
Processing of proMMP-9 by enterokinase To examine whether enterokinase could activate MMP-9 also directly independent of trypsin-2 activation, purified proMMP9 was treated with recombinant enterokinase for 0, 20, 40 and 90 min. After termination of the reactions, the cleavage products were analyzed with SDS-PAGE. After the 90-minute incubation, 92 kDa proMMP-9 conversion to 77-82 kDa forms was clearly visible, and the cleavage process started already after 40 minutes of incubation. No intermediate forms or further fragmentation of MMP-9 was observed (Fig. 5C). In the absence of enterokinase, no degradation of proMMP-9 was observed after the 90 minute incubation ruling out any natural degradation of proMMP-9 under these conditions (not shown). Sequence analysis revealed that enterokinase cleaved the Lys65-Ser66 site of proMMP-9 (Fig. 5D). To determine if this cleavage resulted in functionally active MMP-9, enzyme activity was measured by gelatin degradation assay (Fig. 5E). Enterokinase cleavage did not increase the activity of MMP-9 (103% ± 11%) compared to control, MMP-9 alone (100% ± 8%). To rule out the possibility that enterokinase cleavage might make MMP-9 more susceptible to APMA activation, we treated MMP9 first with enterokinase and then with APMA for 20 min (not shown) or 1 h and compared the activity to MMP-9 treated only with APMA (Fig. 5E). In both time points APMA caused an approximately double increase in the MMP-9 activity regardless of the presence or absence of enterokinase. Enterokinase alone and MMP-9 inhibited with EDTA served as negative controls and did not cleave gelatin.
Discussion In this study we found that trypsin-2 mRNA and protein are expressed in OSCCs and cultured invasive tongue carcinoma cell line (HSC-3), and that trypsin-2 is intracellularly cocompartmentalized with MMP-9. This suggests that MMP-9 activation by trypsin-2 may take place already in intracellular vesicles within the carcinoma cells. In contrast, in giant cell tumor and osteosarcoma MMP-9 and trypsin-2 were localized in different cells. Thus, the intracellular MMP-9 co-localization with trypsin-2 might be a characteristic feature for aggressive cancers of epithelial, but not of mesenchymal origin. Schnaeker et al. [30] demonstrated that MMP-9 is transported from melanoma cells by microtubule-dependent traffic. Our results support the existence of corresponding vesicular intracellular transportation of MMP-9 in OSCCs, and we found that also trypsin-2 may be transported in these same vesicles. We show by laser scanning confocal microscopy that in human tongue SCC cells almost 40% of trypsin-2 is co-compartmenta-
923
lized with MMP-9, and nearly 60% of MMP-9 co-compartmentalized with trypsin-2. Further evidence for their spatial relationship was demonstrated by the disruption of the intracellular protein transport from the Golgi complex by monensin treatment. As the transit of trypsin-2 and MMP-9 from the Golgi to the secretory vesicles was impaired by monensin, it resulted in vesicular dissociation and differential localization of trypsin-2 and MMP-9. Small fractions of trypsin-2 and MMP-9 were still co-localized even in the presence of monensin, but presumably that is due to the abnormal accumulation of proteins into Golgi. The vesicular traffic in tongue SCCs containing activated MMP-9 and trypsin-2 may thus play an important role in the directional proteolysis of the extracellular matrix during cellular migration and invasion [3,26,28]. Ovarian carcinoma ascites-derived membrane vesicles have been shown to contain activated MMP-9 and MMP-2 [44], and the contents of vesicles in ascites were generally higher in women with malignancies than in patients with more benign diseases [45]. This kind of intracellular proteinase activation may thus contribute to directional extracellular matrix degradation, and may facilitate cancer invasion and metastasis formation. Our results show a possible mechanism of how MMP-9 can be activated within the vesicles. Taraboletti et al. [46] have shown that endothelial cell -derived vesicles also contain MMP2 and MMP-9 in both active and proenzyme forms. It has been speculated that the presence and activation of MMP-2 and -9 in endothelial cells may have a role in the mechanisms of cancer growth -related angiogenesis. The intracellular activation of proteases is a controversial issue, but our observations of the existence of both latent and cleaved MMP-9 within the carcinoma cells supports the above mentioned previous works [44–46]. The intracellular cleaved form of MMP-9 seemed to be somewhat larger in size than cleaved MMP-9 obtained by the APMA-activation (Fig. 3J) or trypsin-2 activated MMP-9 found in cell culture media (about 77 kDa) [28]. Therefore we speculate that it may be possible that the cells could control the intracellular proteolytic activity by partially processed intermediate forms that might be more susceptible for full activation when they are trafficked outside the cells. Intense MMP-9 production is characteristic for osteoclastic cells both in normal bone and bone tumors, as evidenced by this and many previous studies [20,22,23]. Furthermore, MMP-9 expression has been found to correlate with the aggressiveness of head and neck carcinomas [47,48]. The expression of trypsin2 also correlates with the malignant phenotype of cancers [11]. We have previously observed that a highly invasive tongue carcinoma cell line (HSC-3) produced trypsin-2, but less invasive OSCC cell lines did not. Furthermore, increased trypsin-2 production by transfection correlated with increased cell intravasation and invasion [28]. In addition to MMP-9, trypsin2 can activate in vitro several other MMPs, e.g. MMP-2, MMP-8 and MMP-13 [26] making the role of trypsin-2 more complex. However, in trypsin-2 transfected HSC-3 cells only MMP-9 existed in the active form in the culture media suggesting that trypsin-2 might not be a biologically relevant activator of other MMPs in these cells. Interestingly, the amount of MMP-8 protein even decreased in trypsin-2 transfected HSC-3 cells [26]. MMP-8 has been shown to have a dual effect on cancer: on one hand it is involved in tissue destruction facilitating cancer progression and inflammation, but on the other hand it exhibits cancer-
Author's personal copy 924
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
protective and anti-inflammatory properties due to its ability to process certain anti-inflammatory cytokines and chemokines [49,50]. In fact, our unpublished results suggest that MMP-8 seems to be a protective factor in OSCC (Korpi et al. unpublished data). Our previous observation showing that the transfection of trypsin-2 decreases the amount of MMP-8 [26] is thus in line with the data demonstrating that trypsin-2 can promote aggressive cancer growth [28]. We also demonstrate by RT-PCR that another trypsin isoform, trypsin-4, was not expressed by HSC3 carcinoma cells (not shown). Interestingly, trypsin-4 has been reported to possess a tumor-suppressive role in cancer progression [51]. Our finding that the HSC-3 cells do not produce trypsin-4, which possesses tumor suppressor activity, may also in part explain the aggressive nature of these cells. Proteases often form complex proteolytic activation cascades. We show here by RT-PCR and Western analysis that enterokinase is synthesized by oral squamous cell carcinoma cells, and not just within the digestive track as previously thought. Tumors originating from enterocytes and goblet cells in the duodenum are known to express enterokinase [52], but our work demonstrated for the first time the existence of enterokinase in carcinoma cells outside the duodenum. Enterokinase is a physiological activator of trypsin-2 which in turn can activate proMMP-9 [24,25]. As described in this study, enterokinase also directly converted the 92 kDa proMMP-9 to about 80 kDa form without formation of activation intermediates or further fragmentation. Enterokinase cleaved proMMP-9 at Lys65-Ser66 site, but this cleavage did not result in activated MMP-9. Trypsin-2 causes superactivation by cleaving proMMP9 at the Arg106-Phe107 site leading to complete one-step removal of the latency motif PRCGVPD resulting in structural stability of MMP-9 with full activity [25]. In this regard, the enterokinase-induced Lys65-Ser66 split occurs N-terminally of the latency motif and results in a non-active form. However, this enterokinase cleavage might make proMMP-9 more susceptible to other downstream activators. To test this hypothesis, we treated proMMP-9 first with enterokinase followed by incubation with APMA. In these experiments proMMP-9 was not more readily or rapidly activated by APMA, but as APMA is an artificial MMP-activator, the situation might be different with the natural MMP-9 activators. Multistep activation has in fact been observed with other proteases. Ogata et al. [53] have shown that MMP-3 induces multiple cleavages and activates proMMP-9; the first cleavage site is Glu59-Met60 and the second one is Arg106-Phe107 representing the superactive form of MMP-9. In this two-step model, the first cleavage happens within seconds, and the cleavage product serves as the substrate for the second activation event. For the second cleavage to take place, a relaxation of the MMP-9 structure is required. This entails dissociation of Cys99 from coordination from the active site zinc [54]. Interestingly, this Cys-switch [55] does not always seem to be absolutely required for proMMP-9 activation. ProMMP-9 is cleaved by MMP-26 at the Ala93-Met94 site resulting in active MMP-9. These results indicate that the latent form of MMP-9 can be transiently activated without the proteolytic loss of Cys99-residue [56]. This activated form can be further activated to produce lower molecular mass active species. Likewise, the combined action of enterokinase, trypsin-2 and MMP-9 may form an activation cascade in OSCC.
In conclusion, we found that the localization and compartmentalization of MMP-9 and its activator, trypsin-2 is different in OSCCs compared to bone tumors. All of these tumors produce both MMP-9 and trypsin-2, but only in OSCCs MMP-9 and trypsin-2 are co-localized within the carcinoma cells and co-compartmentalized within the same intracellular vesicles. Furthermore, observed here for the first time, enterokinase, the physiological activator of trypsin-2, was present in OSCC. Enterokinase cleaved also directly proMMP-9, but this processing did not result in activated MMP-9. These and our previous results therefore suggest that this proteolytic activation cascade may be important for OSCCs enabling effective local proteolysis. Thus it may represent one of the mechanisms for the aggressive and invasive behavior of OSCCs.
Acknowledgments We thank Annikki Huhtela, Riitta Vuento, Tuula Lujala, Paula Hakso, Ritva Keva, and Erkki Hänninen for expert technical assistance, Hongmin Tu for protein sequencing and Juha Risteli for 125I-labelled gelatin. This work has been supported by grants from Academy of Finland, Finnish Dental Society, Helsinki Biomedical Graduate School, Cancer Society of Northern Finland, the EVO-HUCH-grants [TI 020 Y0002, TYH 6104, 7114, TYH 5306], and the KEVO-OUCH-grants.
REFERENCES
[1] J.N. Myers, T. Elkins, D. Roberts, R.M. Byers, Squamous cell carcinoma of the tongue in young adults: increasing incidence and factors that predict treatment outcomes, Otolaryng. Head Neck 122 (2000) 44–51. [2] A.S. Kashyap, S. Kashyap, Head and neck cancer, New Engl. J. Med 346 (2002) 1416–1417. [3] P. Nyberg, M. Ylipalosaari, T. Sorsa, T. Salo, Trypsins and their role in carcinoma growth, Exp. Cell Res 312 (2006) 1219–1228. [4] A. Paju, U.H. Stenman, Biochemistry and clinical role of trypsinogens and pancreatic secretory trypsin inhibitor, Crit. Rev. Clin. Lab. Sci 43 (2006) 103–142. [5] V.J. LaBombardi, E. Shaw, J.F. DiStefano, G. Beck, F. Brown, S. Zucker, Isolation and characterization of a trypsin-like serine proteinase from the membranes of Walker 256 carcino-sarcoma cells, Biochem. J 211 (1983) 695–700. [6] E. Koivunen, M.L. Huhtala, U.H. Stenman, Human ovarian tumor-associated trypsin. Its purification and characterization from mucinous cyst fluid and identification as an activator of pro-urokinase, J. Biol. Chem 264 (1989) 14095–14099. [7] T. Ohta, T. Terada, T. Nagakawa, H. Tajima, H. Itoh, L. Fonseca, I. Miyazaki, Pancreatic trypsinogen and cathepsin B in human pancreatic carcinomas and associated metastatic lesions, Brit. J. Cancer 69 (1994) 152–156. [8] T. Terada, T. Ohta, H. Minato, Y. Nakanuma, Expression of pancreatic trypsinogen/trypsin and cathepsin B in human cholangiocarcinomas and hepatocellular carcinomas, Hum. Pathol 26 (1995) 746–752. [9] K. Oyama, T. Ohta, G.I. Nishimura, A. Elnemr, T. Yasui, T. Fujimura, S. Fushida, H. Kitagawa, M. Kayahara, T. Terada, K. Miwa, Trypsinogen expression in colorectal cancers, Int. J. Mol. Med 6 (2000) 543–548. [10] E. Koivunen, O. Saksela, O. Itkonen, S. Osman, M.L. Huhtala, U.H. Stenman, Human colon carcinoma, fibrosarcoma and
Author's personal copy E XP E RI ME N TA L CE LL RE S EA RCH 3 14 ( 20 0 8 ) 9 1 4 –92 6
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
leukemia cell lines produce tumor-associated trypsinogen, Int. J. Cancer 47 (1991) 592–596. E. Koivunen, O. Itkonen, H. Halila, U.H. Stenman, Cyst fluid of ovarian cancer patients contains high concentrations of trypsinogen-2, Cancer Res 50 (1990) 2375–2378. H. Yamamoto, S. Iku, F. Itoh, X. Tang, M. Hosokawa, K. Imai, Association of trypsin expression with recurrence and poor prognosis in human esophageal squamous cell carcinoma, Cancer 91 (2001) 1324–1331. H. Yamamoto, S. Iku, Y. Adachi, A. Imsumran, H. Taniguchi, K. Nosho, Y. Min, S. Horiuchi, M. Yoshida, F. Itoh, K. Imai, Association of trypsin expression with tumour progression and matrilysin expression in human colorectal cancer, J. Pathol 199 (2003) 176–184. M.D. Sternlicht, Z. Werb, How matrix metalloproteinases regulate cell behaviour, Ann. Rev. Cell Dev. Biol 17 (2001) 463–516. G.T. Thomas, M. Lewis, P.M. Speight, Matrix metalloproteinases and oral cancer, Oral Oncol 35 (1999) 227–233. W.G. Stetler-Stevenson, Type IV collagenases in tumor invasion and metastasis, Cancer Metastasis Rev 9 (1990) 289–303. T.A. Giambernardi, G.M. Grant, G.P. Taylor, R.J. Hay, V.M. Maher, J.J. McCormick, R.J. Klebe, Overview of matrix metalloproteinase expression in cultured human cells, Matrix Biol 16 (1998) 483–496. P. Thomas, R. Khokha, F.A. Shepherd, R. Feld, M.S. Tsao, Differential expression of matrix metalloproteinases and their inhibitors in non-small cell lung cancer, J. Pathol 190 (2000) 150–156. P. Vihinen, V.M. Kähäri, Matrix metalloproteinases in cancer: prognostic markers and therapeutic targets, Int. J. Cancer 99 (2002) 157–166. J.M. Delaisse, M.T. Engsig, V. Everts, M. del Carmen Ovejero, M. Ferreras, L. Lund, T.H. Vu, Z. Werb, B. Winding, A. Lochter, M.A. Karsdal, T. Troen, T. Kirkegaard, T. Lenhard, A.M. Heegaard, L. Neff, R. Baron, N.T. Foged, Proteinases in bone resorption: obvious and less obvious roles, Clin. Chim. Acta 291 (2000) 223–234. N. Ortega, D. Behonick, D. Stickens, Z. Werb, How proteases regulate bone morphogenesis, Ann. N.Y. Acad. Sci 995 (2003) 109–116. S.M. Kumta, L. Huang, Y.Y. Cheng, L.T. Chow, K.M. Lee, M.H. Zheng, Expression of VEGF and MMP-9 in giant cell tumor of bone and other osteolytic lesions, Life Sci 73 (2003) 1427–1436. K. Bjørnland, K. Flatmark, S. Pettersen, A.O. Aasen, O. Fodstad, G.M. Maelandsmo, Matrix metalloproteinases participate in osteosarcoma invasion, J. Surg. Res 127 (2005) 151–156. D. Lu, K. Futterer, S. Korolev, X. Zheng, K. Tan, G. Waksman, J.E. Sadler, Crystal structure of enteropeptidase light chain complexed with an analog of the trypsinogen activation peptide, J. Mol. Biol 292 (1999) 361–373. T. Sorsa, T. Salo, E. Koivunen, J. Tyynela, Y.T. Konttinen, U. Bergmann, A. Tuuttila, E. Niemi, O. Teronen, P. Heikkilä, H. Tschesche, J. Leinonen, S. Osman, U.H. Stenman, Activation of type IV procollagenases by human tumor-associated trypsin-2, J. Biol. Chem 272 (1997) 21067–21074. M. Moilanen, T. Sorsa, M. Stenman, P. Nyberg, O. Lindy, J. Vesterinen, A. Paju, Y.T. Konttinen, U.H. Stenman, T. Salo, Tumor-associated trypsinogen-2 (trypsinogen-2) activates procollagenases (MMP-1, -8, -13) and stromelysin-1 (MMP-3) and degrades type I collagen, Biochemistry 42 (2003) 5414–5420. A. Lukkonen, T. Sorsa, T. Salo, T. Tervahartiala, E. Koivunen, L. Golub, S. Simon, U.H. Stenman, Down-regulation of trypsinogen-2 expression by chemically modified tetracyclines: association with reduced cancer cell migration, Int. J. Cancer 86 (2000) 577–581.
925
[28] P. Nyberg, M. Moilanen, A. Paju, A. Sarin, U.H. Stenman, T. Sorsa, T. Salo, MMP-9 activation by tumor trypsin-2 enhances in vivo invasion of human tongue carcinoma cells, J. Dent. Res 81 (2002) 831–835. [29] A. Paju, T. Sorsa, T. Tervahartiala, E. Koivunen, C. Haglund, A. Leminen, T. Wahlström, T. Salo, U.H. Stenman, The levels of trypsinogen isoenzymes in ovarian tumour cyst fluids are associated with promatrix metalloproteinase-9 but not pro matrix metalloproteinase-2 activation, Brit. J. Cancer 84 (2001) 1363–1371. [30] E.M. Schnaeker, R. Ossig, T. Ludwig, R. Dreier, H. Oberleithner, M. Wilhelmi, S.W. Schneider, Microtubule-dependent matrix metalloproteinase-2/matrix metalloproteinase-9 exocytosis: prerequisite in human melanoma cell invasion, Cancer Res 64 (2004) 8924–8931. [31] L. Barnes, J.F. Eveson, P. Reichart, D. Sidransky (Eds.), World Health Organization Classification of Tumors. Pathology and Genetics of Head and Neck Tumours, IARC Press, Lyon, 2005. [32] O. Itkonen, U.H. Stenman, S. Osman, E. Koivunen, H. Halila, T. Schroder, Serum samples from pancreatectomized patients contain trypsinogen immunoreactivity, J. Lab. Clin. Med 128 (1996) 98–102. [33] S.T. Nikkari, M. Höyhtyä, J. Isola, T. Nikari, Macrophages contain 92-kDa gelatinase (MMP-9) at the site of degenerated internal elastic lamina in temporal arthritis, Am. J. Pathol 149 (1996) 1427–1433. [34] M.E. Duncan, J.P. Richardson, G.I. Murray, W.T. Melvin, J.E. Fothergill, Human matrix metalloproteinase-9: activation by limited trypsin treatment and generation of monoclonal antibodies specific for the activated form, Eur. J. Biochem 258 (1998) 37–43. [35] S.W. Thompson (Ed.), Selected histochemical and histopathological methods, Charles C. Thomas Publisher, Springfield, Illinois, 1966, pp. 644–646. [36] R.L. O’Grady, A. Nethery, N. Hunter, A fluorescent screening assay for collagenase using collagen labelled with 2-methoxy-2,4-diphenyl-3(2H)-furanone, Anal. Biochem 140 (1984) 490–494. [37] S.V. Costes, D. Daelemans, E.H. Cho, Z. Dobbin, G. Pavlakis, S. Lockett, Automatic and quantitative measurement of protein-protein colocalization in live cells, Biochem. J 86 (2004) 3993–4003. [38] A. Paju, A. Bjartell, W.M. Zhang, S. Nordling, A. Borgström, J. Hansson, U.H. Stenman, Expression and characterization of trypsinogen produced in the male genital tract, Am. J. Pathol 157 (2000) 2011–2021. [39] M. Emi, Y. Nakamura, M. Ogawa, T. Yamamoto, T. Nishide, T. Mori, K. Matsubara, Cloning, characterization and nucleotide sequences of two cDNAs encoding human pancreatic trypsinogens, Gene 1 (1986) 305–310. [40] P. Huhtala, A. Tuuttila, L.T. Chow, J. Lohi, J. Keski-Oja, K. Tryggvason, Complete structure of the human gene for 92-kDa type IV collagenase. Divergent regulation of expression for the 92- and 72-kilodalton enzyme genes in HT-1080 cells, J. Biol. Chem 266 (1991) 16485–16490. [41] H. Autio-Harmainen, T. Hurskainen, K. Niskasaari, M. Höyhtyä, K. Tryggvason, Simuultaneous expression of 70 kilodalton type IV collagenase and type IV collagen α1(IV) chain genes by cells of early human placenta and gestational endometrium, Lab. Invest 67 (1992) 191–200. [42] P. Nyberg, P. Heikkilä, T. Sorsa, J. Luostarinen, R. Heljasvaara, U.H. Stenman, T. Pihlajaniemi, T. Salo, Endostatin inhibits human tongue carcinoma cell invasion and intravasation and blocks the activation of matrix metalloprotease-2, -9 and -13, J. Biol. Chem. 278 (2003) 22404–22411. [43] U. Impola, V.J. Uitto, J. Hietanen, L. Häkkinen, L. Zhang, H. Larjava, K. Isaka, U. Saarialho-Kere, Differential expression of matrilysin-1 (MMP-7), 92 kD gelatinase (MMP-9), and
Author's personal copy 926
[44]
[45]
[46]
[47]
[48]
[49] [50]
E XP ER I ME NT A L C EL L RE S EA R CH 3 14 ( 20 0 8 ) 9 1 4 –92 6
metalloelastase (MMP-12) in oral verrucous and squamous cell cancer, J. Pathol 02 (2004) 14–22. L.E. Graves, E.V. Ariztia, J.R. Navari, H.J. Matzel, M.S. Stack, D.A. Fishman, Proinvasive properties of ovarian cancer ascites-derived membrane vesicles, Cancer Res 64 (2004) 7045–7049. A. Ginestra, D. Miceli, V. Dolo, F.M. Romano, M.L. Vittorelli, Membrane vesicles in ovarian cancer fluids: a new potential marker, Anticancer Res 19 (1999) 3439–3445. G. Taraboletti, S. D'Ascenzo, P. Borsotti, R. Giavazzi, A. Pavan, V. Dolo, Shedding of the matrix metalloproteinases MMP-2, MMP-9, and MT1-MMP as membrane vesicle-associated components by endothelial cells, Am. J. Pathol 160 (2002) 673–680. H. Ruokolainen, P. Pääkkö, T. Turpeenniemi-Hujanen, Expression of matrix metalloproteinase-9 in head and neck squamous cell carcinoma: a potential marker for prognosis, Clin. Cancer Res 10 (2004) 3110–3116. R.C. Jordan, M. Macabeo-Ong, C.H. Shiboski, N. Dekker, D.G. Ginzinger, D.T. Wong, B.L. Schmidt, Overexpression of matrix metalloproteinase-1 and -9 mRNA is associated with progression of oral dysplasia to cancer, Clin. Cancer Res 10 (2004) 6460–6465. R. Ala-aho, V.M. Kähäri, Collagenases in cancer, Biochimie 87 (2005) 273–286. M. Balbin, A. Fueyo, A.M. Tester, A.M. Pendas, A.S. Pitiot, A. Astudillo, C.M. Overall, S.D. Shapiro, C. Lopez-Otin, Loss of
[51]
[52]
[53]
[54]
[55]
[56]
collagenase-2 confers increased skin tumor susceptibility to male mice, Nat. Genet 35 (2003) 252–257. K. Yamashita, K. Mimori, H. Inoue, M. Mori, D. Sidransky, A tumor-suppressive role for trypsin in human cancer progression, Cancer Res 63 (2003) 6575–6578. T. Imamura, Y. Kitamoto, Expression of enteropeptidase in differentiated enterocytes, goblet cells, and the tumor cells in human duodenum, Am. J. Physiol. Gastrointest. Physiol 285 (2003) G1235–G1241. Y. Ogata, J.J. Enghild, H. Nagase, Matrix metalloproteinase 3 (stromelysin) activates the precursor for the human matrix metalloproteinase 9, J. Biol. Chem 267 (1992) 3581–3584. M.W. Olson, M.M. Bernardo, M. Pietila, D.C. Gervasi, M. Toth, L.P. Kotra, I. Massova, S. Mobashery, R. Fridman, Characterization of the monomeric and dimeric forms of latent and active matrix metalloproteinase-9. Differential rates for activation by stromelysin, J. Biol. Chem 275 (2000) 2661–2668. H.E. Van Wart, H. Birkedal-Hansen, The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family, Proc. Natl. Acad. Sci. U. S. A 87 (1990) 5578–5582. Y.G. Zhao, A.Z. Xiao, R.G. Newcomer, H. Park, T. Kang, L.W.K. Chumg, M.G. Swanson, H.E. Zhau, J. Kurhanewicz, Q.X.A. Sang, Activation of progelatinase B by endometase/matrilysin-2 promotes invasion of human prostate cancer cells, J. Biol. Chem 278 (2003) 15056–15064.
