ICANCERRESEARCH54,3611-3616,July1. 1994) Autocrine Factor Enhancing the Secretion of Mr 95,000 (Matrix 9) in Serum.free Medium Conditioned with Murine Metastatic Colon Carcinoma Cells'

Sumiko Hyuga, Yohko Nishikawa, Keita Sakata, Hidekazu Tanaka, Sadako Yamagata, Kenji Sugita, Shinsuke Saga, Mutsushi Matsuyama, and Satoru Shimizu2

Pathophysiology Unit, Aichi Cancer Center Research institute, 1—1Kanokoden. Chikusa-ku, Nagoya 464 fS. H., Y. N.. S. Y.. S. 5.1; Second Department of Pathology, Nagoya University School ofMedicine, Showa-ku, Nagoya 466 [K. S., S. Sa., M. MI; and Shionogi Research Laboratory, Shionogi and Company, bd. Sagisu. Fukushima-ku, Osaka 533 (H. T.. K. Su.J. Japan

ABSTRACT can; and (MMP2 and MMP9) on type IV (3). Increased expression of stromelysins, matrilysin, or gelatinases has We previously reported that murine tumor cells with a high spontane been reported in tumor cells with an invasive or metastatic phenotype oils metastatic potential to the lung secrete higher amounts of M@95,000 (4—12).The observed association between secretion of MMPs and gelatinase (matrix metalloprOtelnase 9, MMP9) than do poorly metastatic cells. The present study, conducted to clarify the mechanisms underlying invasive or metastatic properties of tumor cells implies that their the increase In MMP9, revealed an autocrine factor that enhances the expression could be crucial for progression of a tumor to malignancy. secretion of M@95,000 gelatinase (MMP9). The secretion of MMP9 by Elucidation of the mechanisms of how the secretion of these MMPs highly metastatic colon cardnoma LuM1 cells, detected by zyinography, is regulated is therefore of prime importance to understanding of was augmented 10-fold when cultured In medium supplemented with malignant progression. serum-free medium conditioned with LuM1 cells. Because the secretion of We previously reported that cultured murine tumor cells secrete Mr Mr 60,000 gelatinase (MMP2), as well as total protein, by the same cells 60,000 and 95,000 gelatinases/type IV (MMP2 and was not affected under these conditions, the augmentation appears specific MMP9, respectively) and that those with high spontaneous metastatic for MMP9. The steady-state level of MMP9 mRNA was elevated In LuM1 potential to the lung produce remarkably higher amounts of MMP9 cells CUltUredIn the presence ofthe supernatant. The amount ofthe factor than their counterparts with poorly metastatic potential (6). The in the culture medium Increased with time In culture, indicating that it present study was undertaken to elucidate mechanisms underlying the was produced by the LuM1 cells. It was found to be heat stable but sensitive to try@n digestion. Conditioned medium from poorly metastatic augmented secretion of MMP9 by murine metastatic colon carcinoma NM11 cells did not stimulate the secretion of gelatinases by NM11 cells, colon 26 cells. As a result, an autocrine factor enhancing secretion of suggesting that autocrine StImUlatiOnof MMP9 secretion Is a character MMP9 was detected in medium conditioned with metastatic cells. lade of metastatic cells. This factor could account for the augmented secretion of MMl@ by marine tumor cells with spontaneous metastatic MATERIALS AND METHODS potential to the lung. Cells. Highly and poorly metastatic LuMI and NMI I, respectively. cells INTRODUCTION were isolated by in vivo selection from mumine colon carcinoma colon 26 cells according to the method described by Tsuruo ci a!. (13). Briefly, cultured colon The process of tumor metastasis consists of many complex steps, 26 colon cells (106cells) were implanted intrademmallyintothe abdominal skin each involving specific interaction of tumor cells with host cells or the of a 5-week-old BALB/Cfemale mouse. After 4 weeks, the lung of the mouse . Migration of tumor cells across the basement bearing colon 26 tumors was dissected out and transplanted to another mouse. membrane is considered to be essential for the tumor cells to achieve This was then repeated, and cloned metastatic cells in culture (LuM1) were successful metastasis (1). Thus, the secretion of that can obtained from metastatic lung nodules after 7 cycles of lung transplantation. Poorly metastatic cells (NM) I) were cloned from cultured cells obtained from degrade basement membrane components would clearly favor metas theprimarytumor.Averagenumbersof metastaticnodulesin thelungsof mice tasis.Thebasementmembraneextracellularmatrixseparatingepithe 5 weeks after intrademmal implantation of 5 X l0@ LuMI or NM1I cells to lium, endothelium, and methothelium from underlying stroma is pre 6-week-old BALB/c female mice were 98 and 4, respectively. Cell lines were dominantly made up of type IV collagen, laminin, and heparan sulfate maintained in RPM! 1640 medium with 10% FCS. . These components of basement membranes can all be Detection and Quantitatlon of Gelatinase. Zymography with gelatin as a degraded by MMPs,3 a family of proteinases acting specifically on substrate was used to detect gelatinases in serum-free culture medium (6). extracellular matrix proteins and therefore playing important roles in Quantitation was performed by densitometric tracing of zymograms using a normal and pathological processes involving their degradation. Nine Shimadzu CS-930 densitometer. Gelatinase activity, detected as a band of separate members of the MMP family have been identified thus far, reduced Coomassie blue staining, showed a peak in the negative direction from each having a distinct substrate specificity (2). Some members of the background staining, and the peak area of the detected band was proportional to the amount of gelatinase. Calculated values calibrated with the numbers of MMP family act on specific components of basement membranes: cells in culture were used for relative amounts of gelatinase. In all experiments, stromelysins (MMP3 and MMP1O) on type IV collagen, laminin, and two independent culture supernatants were obtained and used for measurement ; matrilysin (MMP7) on type IV collagen and proteogly of gelatinase activity. All data are means ±rangeof points and represent the results of at least three independent experiments. For immunochemical detec Received l2/2@W93;accepted4129/94. tion of Mr 95,000 gelatinase in the culture, 100 p.1 of supematant were Thecostsof publicationofthisarticleweredefrayedinpartby thepaymentofpage concentrated using an ultrafree membrane (Millipore) and subjected to SDS charges. This article must therefore be hereby marked advertisement in accordance with PAGE.Fractionatedproteinswereblottedontonitrocellulosepaperandprobed 18U.S.C.Section1734solelyto indicatethisfact. I This study was supported in part by Special Coordination Funds of the Science and with a monoclonal antibody raised in a rat to mouse [email protected] gelatinase Technology Agency of the Japanese Government. (MMP9)from LuM1cells. 2 To whom requests for reprints should be addressed. Preparation ofSerum-free Medium COnditiOned with LuM1 Cells. The 3 The abbreviations used are: MMP, ; BSA, bovine serum medium of confluent culture LuM1 cells was changed to serum-free RPM! albumin;EGF,epidermalgrowthfactor,FCS,fetalcalfserum;IL-l,interleukin-l;PAGE, polyacrylamide gel electmophoresis; SDS. sodium dodecyl sulfate; TGF, transforming 1640 medium, and cells were cultured overnight. Then, medium was again growth factor, liMP, tissue inhibitor of metallopmoteinases; cDNA, complementary DNA. changed, and after 48 h, culture supematant was obtained and centrifuged to 3611

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. AUTOCRINE FACIOR ENHANCEMENT OF GELA11NASE SECREI1ON remove cell debris. To eliminate gelatinases, the centrifuged supernatant (1.5 with that OfMr 60'000 gelatinase, which increased linearly with time. liters) was subjected to column chromatography using gelatin-cellulofine (bed Secretion of both Mr 60,000 and 95,000 gelatinases by NM11 cells volume of 90 ml; Seikagaku Kogyo, Tokyo), and the fraction passing through also occurred linearly. These results suggested the presence of some the column was obtained. This procedure was repeated three times to remove factor(s) that enhanced the secretion of Mr 95,000 gelatinase in the any trace amounts of remaining gelatinase, and the resulting passing-through culture medium of LuM1 cells. fraction was concentrated to 50 ml. The concentrated fraction (fraction 0), In order to detect the factor enhancing M@95,000 gelatinase secre after testing for the absence of gelatinase activity, was used for detection of lion, the effect of serum-free culture medium conditioned with LuM1 factor(s) enhancing gelatinase secretion. Amounts of protein were determined with a protein assay reagent from Bio-Rad Laboratories using BSA as a cells, after included gelatinases were eliminated (fraction 0), was standard. examined on gelatinase secretion by LuM1 cells. When cultured in Detection of a Factor Enhancing the Secretion OfMr 95,000 GCintiIthSC medium supplemented with fraction G, LuM1 cells showed enhanced (MMP9). Trypsinized cells were washed once with 10% FCS-RPMI 1640 secretion of Mr 95,000 gelatinase (MMP9) as shown in Fig. 2, A and medium and twice with serum-free RMPI 1640 medium. They were then B, the amount being about 10-fold higher than in control culture suspended in serum-free RPM! 1640 medium at the concentration of 10@ without fraction 0. Enhancement of gelatinase secretion by fraction G cells/mI, and 1 ml of cell suspension was seeded in a culture dish of 30 mm was specific for Mr 95,000 gelatinase, and the secretion of Mr 60'000 diameter. After culture overnight, medium was changed to RPM! 1640 me gelatinase was not affected by its addition to the culture medium. dium containing 100 @g/mlofBSA with or without fraction G, and culture Immunochemical detection of Mr 95,000 gelatinase using a monoclo supernatant was collected after 24 h. The gelatinase activities in the supema tant with fraction 0 were measured and compared with those in the culture nal antibody raised against mouse M1 95,000 gelatinase also showed supernatant without fraction 0. To examine the effect of fraction 0 on the a remarkable increase of the amount of secreted Mr 95,000 gelatinase secretion of total protein, 10 @Ciof[35S]methionine(Trans 35S-label;ICN in the culture supernatant with fraction 0 (Fig. 2C); total protein Biochemicals, Inc.) were added to the culture, and after 24 h, protein in the secreted by LuM1 cells, however, was not seriously affected by the culture supernatant was precipitated with 5% trichloroacetic acid. The amount addition of fraction G (Table 1). The effect of fraction 0 on the of secreted protein was expressed as total radioactivity in the precipitate, and secretion of Mr 95,000 gelatinase was dose dependent up to a maxi the radioactivity from the culture supernatant with fraction G was compared mum of 20—50 @gprotein/mI(Fig. 3). The amount of the enhancing with that without fraction 0. The heat stability of the factor enhancing activity of Mr 95,000 gelatinase secretion in the culture supernatant gelatinase secretion was determined by measuring the activity of the fraction increased with time in culture (Fig. 4), specific activity (enhancing 0 after heating for 10 mm in boiling water. To test the sensitivity of the factor activity/mg protein) being almost the same during the same time, to trypsin digestion, fraction 0 was incubated with one-twentieth of its amount by weight of trypsin for 20 h at 37°C.Aftersoybean trypsin inhibitor was suggesting that LuM1 cells continuously secrete the autocrine factor added, digested fraction 0 was examined for enhancing activity. involved. Results of Northern analysis are illustrated in Fig. 5. LuM1 Northern Analysl& To analyze the expression of gelatinases, total cellular cells cultured in serum-free medium expressed mRNAs of both RNA was prepared by acid guanidine thiocyanate-phenol-chloroform extrac MMP9 and MMP2. Elevated steady-state levels of both 3.5- and tion (14). RNA (20 @g)was fractionated on a 1% agarose gel containing 2.2 2.7-kilobase pair mRNAs of MMP9 were observed in LuM1 cells M formaldehyde, transferred to Hybond N (Amersham), and hybridized with cultured in the presence of fraction 0, which is consistent with the DNA probe randomly labeled with [a-32PJdCT'P. cDNA for murine MMP9 enhanced secretion of Mr 95,000 gelatinase (MMP9). In contrast, the (15) and human MMP2 (16) were used as DNA probes. steady-state level of MMP2 mRNA was not effected by the presence of fraction G in the culture medium. RESULTS Effects of heating or trypsin digestion of fraction 0 on the enhance ment of the gelatinase secretion are shown in Fig. 6. The activity in Time courses of secretion of Mr 60,000 and 95,000 gelatinases fraction 0 was relatively heat stable, 80% being retained after heating (MMP2 and MMP9, respectively) by LuM1 and NM11 cells into in boiling water for 10 min. After digestion with trypsin, however, serum-free RPMI 1640 medium are shown in Fig. 1. The amount of fraction 0 lost a considerable proportion of its enhancing activity on M1 95,000 gelatinase in the serum-free culture medium of LuM1 cells gelatinase secretion. In the process ofpreparing fraction 0, serum-free increased exponentially, not linearly, with time in culture, contrasting culture supernatant was concentrated using a membrane filter with a molecular sieve size of 30,000. These results suggest that the factor having an enhancing activity on gelatinase secretion is a heat-stable protein with a molecular weight of >30,000. To compare autocrine enhancement of Mr 95,000 gelatinase secre @0 tion between LuM1 and NM11 cells, 0 fractions were prepared from 03 both, and enhancing activity was assessed in the two cell types (Fig. 7). The Mr 95,000 gelatinase secretion of LuM1 cells was enhanced by fraction 0 from either LuM1 or NM11 cells, the extent being lower in the latter case. With NM11 cells, only the LuM1 cell fraction 0 enhanced gelatinase secretion, and no stimulation by frac I: tion G from NM11 cells was evident, indicating that the autocrine enhancing mechanism of MMP9 secretion observed in LuM1 cells does not operate in the NM11 case. 0 20 40 60 80 0 20 40 60 80 Time (h) Time (h) DISCUSSION Fig. 1. Gelatinase secretion into serum-free culture medium by LuM1 and NM11 Cells. Samples of 2.5 x 1O@cellsin culture dishes of 60 mm diameter were cultured overnight In the present study, aimed at clarifying the mechanism underlying in 4 ml of RPMI 1640 medium with 10% FCS, the medium was changed to serum-free higher amounts of secretion of MMP9 (Mr 95,000 gelatinase) by RPM) 1640, and the culture supernatants obtained at 24-h intervals were analyzed for metastatic LuM1 cells than poorly metastatic NM1 1 cells, both de gelatinases by zymography. M, 95,000 and 60,000 gelatinase activities detected were quantitated by densitometrictracing ofthe zymograms. Points(bars), means(±ranges of rived from murine colon carcinoma colon 26 cells (6), there was clear points) from duplicate determinations. Insets, zymograms of 48-h culture supematants. evidence that an autocrine factor might be responsible. 3612

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. @ @@!;

AUTOCRINE FACFOR ENHANCEMENT OF GELATINASE SECREI1ON AB C Fig. 2. Effect of fraction G on gelatinase kDa 40 kDa secretion by LuM1 cells. Samples of 1 X 10' trypsinized LuM1 cells were cultured overnight in culture dishes of 30 mm diameter with 1 ml of 95@ . -95 serum-free RPM! 1640 medium, the medium was changed to the RPMI 1640 medium containing 60-i BSA (100 g.@g/ml)with or without fraction 0 (Fr.G), and after 24 h, culture supernatants were analyzed for gelatinase secretion. A, zymograms of culturesupernatantswith(+) andwithout(—) fraction G. B, relative gelatinase activities quan titated by densitometric tracing of the zymogram. Columns (bars), means (±ranges of points) from duplicate determinations. C, immunostaining of M, 95,000 gelatinase in culture supernatant with (+) or without(—)fraction0, after concentra tion, fractionationby SDS-PAGE,and blotting onto nitrocellulose,using monoclonalantibody raisedin a rat againstmouseM@95,000gelatinase (MMP9). 95kDa 60 kDa

cellsLuM1Table1 EffectoffractionG onproteinsecretionbyLuMJ processes of secretion, activation, and inhibition. This means that cells (5 x 10') were cultured overnight in 1 ml of serum-free RPMI 1640 there are various difficulties in estimating the activity of metallopro mediumpresenceand containing 100 pg/mi of BSA and 2 pCi/mi of [35S]methionine in the teinases in crude samples, because they must be activated at the same precipitationwithabsenceofFraction0. Proteinintheculturemediumwascollectedby time as inhibitors are eliminated. Zymography overcomes these dii expressedas5% trichloroacetic acid. Amount of prutcin secreted into the medium was twoindependentradiOactiVityinthisprecipitatedfraction.Dataaremeans±therangeofvaluesof ficulties and, therefore, is considered to be the best method for determinations.Protein detection of gelatinases, with proteins in the samples being first secretedCulture separated by SDS-PAGE so that the liMPs and the molecular species cells)Withoutsupernatant (10@ X cpm/10' of gelatinases are located according to their molecular size. The 0.32Withfraction0 3.91± inactive form of gelatinases, zymogens, are activated by SDS in the fraction G 4.53 ±0.05 process. After samples are renatured with Triton X-100, all of the molecular species of gelatinases can be easily detected without any special treatment to activate gelatinases or to eliminate the inhibitors 20 in the crude samples. Quantitative measurements reflecting secretion of Mr 60,000 and 95,000 gelatinases (MMP2 and MMP9, respectively) by metastatic :@ LuM1 cells showed that the accumulation of the latter in serum-free medium was not linearly proportional to time in culture but rather w 10 exponential, in contrast to the secretion of MMP2 by LuM1 cells and

C both MMP2 and MMP9 by poorly metastatic NM11 cells. Because the amount of MMP9 secreted in 24 h by LuM1 cells, tested after daily C 5 exchange of the culture medium, did not change for several days (data

0 60 3 0.01 0.1 1 10 100 C 0 Fraction G (@sg/mIprotein) Fig. 3. Effect of the amount of fraction 0 on gelatinase secretion of LuM1 cells. Cells were cultured as described for Fig. 2 with varying amounts of added fraction 0, and the culture supernatant was analyzed for gelatinase secretion. Points (bars), means (±ranges of points) from duplicate determinations.

Matrix are secreted as zymogens that have no activity without activation in vivo by proteolytic cleavage of the @2t@mn'nalpeptideor in vitro by treatment with various reagents such as amino phenyl mercuric acetate and SDS. Activated metallo 24 48 72 proteinases degrade the components of extracellular matrix according Time (h) to their substrate specificities: interstitial by interstitial Fig. 4. Amount of factor enhancing the secretion OfM, 95,000 gelatinase in serum-free (MMP1); type N collagen by gelatinases (MMP2 and culture supernatant of LuM1 cells. LuMI cells, after reaching approximate confluence, MMP9), stromelysins (MMP3 and MMP1O), and matrilysin (MMP7); were cultured in serum-free RPM! 1640, and culture supernatants were obtained every and proteoglycans by MMP3, MMP1O, and MMP7 (2). However, the 24 h after changing to further serum-free medium. Enhancing activity of Mr 95,000 gelatinase in culture supernatants was estimated and expressed as units/ml (1 unit of activity of these metalloproteinases is inhibited by inhibitors such as enhancing activity equals that giving half-maximum enhancement of gelatinase secretion a@-macroglobulin in serum or liMP in the body fluid (2). The liMPs in the standard assay). Specific activity (units/mg protein) was calculated from units of enhancing activity, and the amount of protein in the supernatant was determined with a are secreted by the same cells that secrete the metalloproteinases. The protein assay reagent from Bio-Rad Laboratories. Columns and points (bars), means activity of metalloproteinases is thus strictly controlled throughout the (±ranges of points) from duplicate determinations. 3613

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. AUTOCRINEFACFORENHANCEMENTOFGELAT1NASESECREI1ON MMP9 MMP2 collagen, and Matrigel has been reported (21—23),and perturbation of cell-extracellular matrix contacts with antibodies to cellular receptors is known to influence the secretion of type IV collagenase (24, 25). Soluble factors such as growth factors and cytokines have been shown to stimulate the secretion of matrix metalloproteinases: collagenase (MMP1) and stromelysin (MMP3) by IL-1f3, tumor necrosis factor-a, EGF, basic fibroblasts growth factor, and platelet-derived growth factor (26—30),and Mr 92,000 gelatinase (MMP9) by EGF, IL-1@3, 28S— and tumor necrosis factor-a (31—33).TGF-f3 suppresses the secretion @ •0 *@• of MMP1 and MMP3 (30, 34) but augments that of MMP2 and MMP9 (35—37).Thus, the expression of MMPs by tumor cells ap 18S— pears to depend on what kinds of receptors for growth factors or cytokines they possess or whether the soluble factors enhancing or inhibiting secretion of MMPs are supplied from outside. For example, compounds from fibroblasts in the stroma surrounding tumor tissues augment their secretion of MMP2 and MMP9, influencing the meta static potential of tumor cells (38, 39). Because it is well known that tumor cells can also, themselves, secrete growth factors or cytokines, the possibility of autocrine stirs —+ —+ ulation must be taken into account as demonstrated by the present Fig. 5. Northern analysis of mRNA levels of MMP9 and MMP2. Total cellular RNA was preparedby acid guanidinethincyanate-phenol-chlomformextractionfromLuM! study. Paracrine stimulation by factors derived from tumor cells has cells cultured in the presence (+) or absence (—)of fraction 0. RNA (20 @g)was been reported for production of MMPs by fibroblasts. The paracrine fractionated on a 1% agarose gel containing 2.2 Mformaldehyde, transferred to Hybond factor, tumor cell-derived collagenase stimulatory factor, was origi N (Amersham), and hybridized with DNA probes randomly labeled with [a-32P]dCTP. cDNAsusedas probeswerefor murineMMP9andhumanMMP2. nally found in conditioned medium of human tumor cells and, when purified from the cell surface, strongly stimulated the secretion of MMP1, MMP3, and MMP2 (40). While tumor cell-derived collage nase stimulatory factor does not act on the tumor cells themselves, autocrine factors stimulating MMP secretion have been reported. Rabbit synovial fibroblasts, when stimulated by 12-O-tetrade canoylphorbol-13-acetate or IL-1@3,synthesize and secrete autocrine factors, which in turn induce secretion of MMP1. Purification of the factors revealed them to be serum amyloid A and @32-microglobulin (41, 42). In addition, autocrine intermediate stimulating factors have been reported in SPARC (osteonectin)-treated fibroblasts (43). Auto

80 Control FractionG Heated TIYPSinIZOd FraCtionG FractionG Fig. 6. Stability of enhancing activity of gelatinase secretion after heating or trypsin digestion. Fraction G (20 @gproteinin 200 @d)wasincubated with ! @.tgtrypsinfor 24 h at 37°Cand then tested for activity after 5 @gsoybeantrypsin inhibitor was added. Fortheheatstabilitytest,fractionGwaskeptinboilingwaterfor10mm.Columns(bars), I means (±ranges of points) from duplicate determinations.

0 not shown), the presence of a factor enhancing its production was .E suggested. A similar pattern was recently reported for 12-O-tetrade 0 C canoylphorbol-13-acetate-induced secretion of MMP1 and liMP by U937 cells (17) and for interleukin-1@3- and TGF-a/EGF-induced expression of MMP9 and MMP1 by rat mucosal keratinocytes (18). A factor(s) enhancing MMP9 secretion could actually be demonstrated in culture medium conditioned with LuM1 cells for 48 h, without any effects on secretion of MMP2. The fact that the amount of the factor in the culture medium increased with time in culture directly impli cated the LuM1 cells in its production, thereby indicating autocrine enhancement. 0 20 40 60 80 100 Fracdon G Added (pg/mi protein) Microenvironmental conditions have been shown to regulate the expression of MMPs: cell-cell or cell-extracellular matrix interactions Fig.7. ComparisonofautocrineenhancementofMr95,000gelatinasesecretionbe and transport of soluble factors such as growth factors and cytokines tweenLuM1andNM11cells.Gfractions(Fr.G)werepreparedfromLuM1andNM!! cells, and enhancmg activity of M 95,000 gelatmase secretion of each was examined. are involved. Signal transduction through the receptor has SccretionofM,95,000gelatinaseofLuMicellswasstimulatedbyfraction0preparations been reported to induce collagenase and stromelysin expression frombothLuM1andNM!! cells,butthatof NM!! cellswasenhancedonlyby the . . fraction G from LuMI cells. Fraction G from NM! 1 cells did not stimulate M 95,000 (19, 20). Augmented secretlon of type IV collagenase due to interac- gelatinasesecretionbyNM!1cells,incontrasttotheclearstimulationintheLuM1case. tion of tumor cells with extracellular matrix proteins such as laminin, Points(bars),means(±rangesof points)fromduplicatedeterminations. 3614

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. AtTFOCRINE FACrOR ENHANtEMENT OF GELATINASE SECRETION

crine stimulation of MMP3 has been reported with culture medium sis Cancer Metastasis Rev., 9: 305-319, !990. conditioned with metastatic murine mammary carcinoma cells, al 12. McDonnell, S., Navre, M., Coffey, R. J., and Matrisian, L M. Expression and localization of the matrix metalloprotcinase pump-! (MMP-7) in human gastric and though the nature of the factor has not yet been characterized (44). A colon carcinomas. Mol. Carcinog., 4: 527—533,1991. unique autocrmnefactor with an Mr of 70,000, the invasion-stimulating 13. Tsuruo, T., Yamori, T., Naganuma, K., Tsukagoshi, S., and Sakurai, Y. Character factor that stimulates secretion of MMP2 and invasion of tumor cells, izationof metastaticclones derivedfrom a metastaticvariantof mouse colon adenocarcinoma 26. Cancer Rca., 43: 5437-5442, 1983. has also been purifiedfrom culturemediumconditionedwith meta !4. Chomczynski, P., and Sacehi, N. Single-step method of RNA isolation by acid static human prostatic cancer cells (45). Invasion-stimulating factor guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162: 156—159,1987. did not directly stimulate an increase in mRNA levels of MMP2 but 15. Tanaka, H., Hojo, K., Yoshida, H., Yoshioka, T., and Sugita, K. Molecular cloning was suggested to regulate the receptor-mediated activation of pro and expression of the mouse 105-kDa gelatinase cDNA. Biochem. Biophys. Res. cesses controlling release (46). The description of an auto Commun., 190: 732—740,1993. crine stimulation of MMP9 secretion by TGF-f3 released from fibro 16. Collier, I. E., Wilhelm, S. M., Eiscn, A. Z., Manner, B. L, Grant, G. A., Seltzer, J. L, Kronberger, A.. He, C., Bauer, E. A., and Goldberg, G. I. H-ms oncogene sarcoma cells transfected with a plasmid containing a porcine TGF-@1 transformed human bronchial epithelial cells (TBE-!)secrete a single metalloprotease regulated by a metallothionein promoter has been reported (47), and capable ofdegrading basement membrane collagen. J. Biol. Chem., 263: 6579—6587, 1988. the present results further suggest that autocrine control of MMPs is 17. ShapirO,S. D., Doyle, 0. A. R., Ley, T. J., Parks, W. C., and Webs, H. 0. Molecular a widespread phenomenon, occurring in a variety of cells. mechanisms regulating the production of collagenase and TIMP in U937 cells: The factor detected in the culture medium of metastatic murine evidence for involvement of delayed transcriptional activation and enhanced mRNA stability. Biochemistry, 32: 4286—4292, 1993. colon carcinoma LuM1 cells in the present study demonstrated a 18. Lyons, J. G., Birkedal-Hansen, B., Pierson, M. C., \Vhitelock, 1. M., and specificity for MMP9 with no stimulation of MMP2. The steady-state Birkedal-Hansen, H. Interleukin-1@ and transforming growth factor-a/epidermal level of MMP9 mRNA was elevated, indicating that transcription or growth factor induce expression of M, 95,000 type IV collagenase/gelatinase and interstitial fibroblast-type collagenasc by rat mucosal keratinocytes. 1. Biol. Chem., stabilization of mRNA is switched on in contrast to the prostatic 268: 19143—19151,1993. cancer cell case. The comparison of autocrine enhancement of MMP9 19. Werb, Z., Tremble, P. M., Behrendtsen, 0., @rowley,E., and Damsky, C. H. Signal transduction through the fibronectin receptor induces collagenase and stromelysin secretion between LuM1 and NM11 cells revealed a complex nature geneexpression.J.CellBiol.,109:877—889,1989. for the factor stimulation. The fact that the MMP9 secretion of LuM1 20. Werb, Z., Tremble, P., and Damsky, C. H. Regulation of extracellular matrix degra cells was stimulated by fraction 0 from both sources, but in the NM11 dationbycdll-extracellularmatrixinteractions.CellDiffer.Dcv.,32:299-306,1990. 21. Turpeenniemi-Hujanen, T., Thorgeirsson, U. P., Rao, C. N., and Uotta, L A. Laminin case, only that from LuM1 cells demonstrated activity, precludes a increases the release oftype N collagenase from malignantcells. 1. Biol. Chem., 261: single factor and receptor explanation and suggests that multiple 1883—1889,1986. factors or steps such as secretion and activation are involved in the 22. Kanemoto, T., Reich, R., Royce, L, Greatorex, D., Adler, S. H., Shiraishi, N., Martin, G. R., Yamada, Y., and Kleinman, H. K. Identification of an amino acid sequence enhanced secretion of MMP9 by LuM1 cells. However, the finding from the laminin A chain that stimulates metastasis and collagenasc IV production. that culture supernatant conditioned with poorly metastatic NM11 Proc. Nat!. Acad. Sci. USA, 87: 2279-2283, 1990. 23. Emonard, H., Christiane, Y., Smet, M., Grimaud, J. A., and Foidart, 1. M. Type IV cells did not give any enhancement of MMP9 secretion by the same and interstitial collagenolytic activities in normal and malignant trophoblast cells are cells suggests that the autocrine function is directly related to the specifically regulated by the extracellular matrix. Invasion Metastasis, 10: !70—l77, metastatic nature of LuM1 cells. Thus, expression of the factor(s) by 1990. 24. Seftor, R. E. B., Seftor, E. A., Stetler-Stevenson, W. G., and Hendrix, M. J. C. The colon tumor cells enhancing MMP9 secretion might be intimately 72 kDa type IV collagenase is modulated via differential expression aj3, and a5@1 linked with tumor progression. integrins during human melanoma cell invasion. Cancer Rca., 53: 3411—3415,1993. 25. Latjava, H., Lyons, J. 0., Salo, T., Macla, M., Koivisto, L, Birkedal-Hansen, H., Akiyama, S. K., Yamada, K. M., and Heino, J. Anti-integrin antibodies induce type IV collagenase expression in keratinocytes. J. Cell. Physiol., 157: 190—200,1993. ACKNOWLEDGMENTS 26. Chua, C. C., Geiman, D. E., Keller, 0. H., and Ladda, R. L Induction of collagenase secretion in human fibroblast cultures by growth promoting factors. J. Biol. (1cm., We gratefully acknowledge the kind gift of humanMMP2 cDNA from 260: 5213—5216,1985. Dr. G. I. Goldberg, Washington University School of Medicine. 27. Dayer, J-M., Beutler, B., and Cerami, A. Cachectin/tumor necrosis factor stimulates collagenase and prostaglandin E@production by human synovial cells and dermal fibroblasts. J. Exp. Med., 162: 2163—2168, 1985. REFERENCES 28. Matrisian, L M., Leroy, P., Ruhlmann, C., Gesnel, M-C., and Breathnach, R. Isolation of the oncogene and epidermal growth factor-induced transin : complex 1. Liotta, L A., Rao, C. N., and Wewer, U. M. Biochemical interactions of tumor cells control in rat fibroblasts. Mol. Cell. Biol., 6: 1679—1686,1986. with the basement membrane. Annu. Rev. Biochem., 55: 1037—!057,1986. 29. Saus, J., Quinones, S., Otani, Y., Nagase, H., Harris, E. D., Jr., and Kurkinen, M. The 2. Matrisian, L M. The matrix-degrading metalloproteinases. Bioessays, 14: 455-463, complete primary structure of human matrix metalloproteinase-3. Identity with 1992 stromelysin.J.Biol.(1cm., 263:6742—6745,1988. 3. Woessner, J. F., Jr. Matrix metalloproteinases and their inhibitors in connective tissue 30. Edwards, D. R., Murphy, 0., Reynolds, J. J., Vlhitham, S. E., Docherty, A. J. P., remodeling. FASEB J., 5: 2145—2154,1991. Angel, P., and Heath, J. K. Transforming growth factor beta modulates the expression 4. Liotta, L A., Tryggvason, K., GarbiSa, S., Hart, I., Foltz, C. M., and Shafle, S. of collagenase and metalloproteinase inhibitor. EMBO J., 6: 1899—1904,1987. Metastatic potential correlates with enzymatic degradation of basement membrane 31. Wilhelm, S. M., Collier, I. E., Manner, B. L, Eisen, A. Z., Grant, 0. A., and collagen. Nature (Lund.), 284: 67—68,1980. Goldberg,0. I. SV4O-transformedhumanlungfibroblastssecretea 92-Wa typeIV 5. Garbisa,S., Poziatti,R@,Muschel,R.J., Saffiotti,U.,Ballin,M.,Goldfarb,R.H., collagenase which is identical to that secreted by normal human macrophages. J. Biol. Khoury, 0., and Uotta, L A. Secretion of type IV collagenolytic protease and (1cm.,264:17213—17221,1989. metastaticphenotype:inductionbytransfectionwithc-Ha-msbutnotc-Ha-rasplus 32. Okada, Y., Tsuchiya, H., Shimizu, H., Tomita, K., Nakanishi, I., Sato, H., Sciki, M., Ad2-Ela. Cancer Res., 47: 1523—1528,1987. Yamashita, K., and Hayakawa, T. Induction and stimulation of 92-kDa gelatinasc/ 6. Yamagata, S., Ito, Y., Tanaka, R., and Shimizu, S. Gelatinases of metastatic cell lines type IV collagenase production in osteosarcoma and fibrosarcoma cell lines by tumor of murinecoloniccarcinomaasdetectedbysubstrate-gelelectrophoresis.Biochem. necrosis factor a. Biochem. Biophys. Rca. Commun., 171: 610—617, 1990. Biophys.Res.Commun.,151:158—162.,1988. 33. Shima, I., Sasagun, Y., Kusukawa, J., Nakano, R, Yamana, H., Fujita, H., Kakegawa, 7. Yamagata, S., Tanaka, R., Ito, Y., and Shimizu, S. Gelatinases of murine metastatic T.,andMorimatsu,M.Productionofmatrixmetalloproteinase9(92-kDagelatinase) tumor cells. Biochem. Biophys. Res. Commun., 158: 228—234, 1989. by humanoesophagealsquamouscellcarcinomain responseto epidermalgrowth 8. Bernhard, E. J., Muschel, R. J., and Hughes, E. N. M, 92,000 gclatinasc release factor. Br. J. Cancer, 67: 721—727,1993. correlates with the metastatic phenotype in transformed rat embryo cells. Cancer Rca., 34. Machida, C. M., Muldoon, L L, Rodland, K. D., and Magun, B. E. Transcriptional 50: 3872—3877,1990. modulation of transin by epidermal growth factor and transforming 9. Margulies, I. M. K.. Hoyhtya, M., Evans, C., Stracke, M. L, Uotta, L A., and growth factor beta. Mol. Cell. Biol., 8: 2479—2483,1988. Stctler-Stevenson, W. 0. Urinary type N collagcnasc: elevated levels are associated 35. Overall, C. M., Wrana, J. L., and Sodek, J. Independent regulation of collagenase, with bladder transitional cell carcinoma. Cancer EpidemioL Biomarkers Prey., 1: 72-kDa progelatinase, and metalloendoproteinase inhibitor expression in human 467—474,1992. fibroblasts by transforming growth factor-a. J. Biol. Chem., 264: 1860—!869,1989. @ 10. Srcenath, T., Matrisian, L M., Stetler-Stevenson, W., Gattom-Celli, S., and Pozzatti, 36. Welch, D. R., Fabra, A., and Nakajima, M. Transforming growth factor stimulates R. 0. Expressionofmatrixmetallopmteinasegenesin transformedratcelllinesof mammary adenocarcinoma cell invasion and metastatic potential. Proc. Nat!. Acad. high and low metastatic potential. Cancer Res., 52: 4942—4947,1992. Sd. USA, 87: 7678—7682,1990. 11. McDonnell, S., and Matrisian, L M. Stromelysin in tumor progression and metasta 37. Salo, T., Lyons, J. G., Rahemtulla, F., Birkedal-Hansen, H., and Larjava, H. Trans 3615

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. AUTOCRINE FACI'OR ENHANCEMENT OF GELATINASE SECRETION

forminggrowthfactor-a! up-regulatestypeIV collagenaseexpressionincultured associated with morphogenesis and tissue remodeling, induces expression of metal humankeratinocytes.J.Biol.(1cm., 266: 11436—1144!,199!. loproteinasesinfibroblaststhrougha novelextracellularmatrix-dependentpathway. 38. Nakajima, M., Morikawa, K., Fabra, A., Bucana, C. D., and Fidler, I. J. Influence of J. Cell Biol., 121: 1433—1444,1993. organ environment on extracellular matrix degradative activity and metastasis of 44. Korczak, B., Kerbel, R. S., and Dennis, J. W. Autocrine and paracrine regulation of human colon carcinoma cells. J. Nati. Cancer Inst., 82: 1890—1898, 1990. tissue inhibitor of metalloproteinases, transin, and urokinasc gene expression in 39. Fabra,A.,Nakajima,M.,Bucana,C.D.,andFidler,I.J. Modulationoftheinvasive metastatic and nonmetastatic mammary carcinoma cells. Cell Growth Differ., phenotypeofhumancoloncarcinomacellsbyorganspecificfibroblastsofnudemice. 2: 335—341,1991. Differentiation, 52: 101—110,1992. 45. Stearns, M. E., and Wang, M. Regulation of kinesin expression and type IV collag 40. Kataoka, H., DeCastro, R., Zucker, S., and Biswas, C. Tumor cell-derived collage enase secretion in invasive human prostate PC-3 tumor sublines. Cancer Rca., 51: nase-stimulatory factor increases expression of , stromelysin, and 72-kDa gelatinase. Cancer Rca., 53: 3154—3!58,1993. 5866—5875,199!. 41. Brinckerhoff,C.E.,andMitchell,T.I.Autocrinecontrolofcollagenasesynthesisby 46. Stearns, M. E., and Stearns, M. Autocrine factors, type IV collagenase secretion and synovial fibroblasts. J. Cell. Physiol., 136: 72—80,1988. prostaticcancercellinvasion.CancerMetastasisRev.,12: 39—52,1993. 42. Brinckerhoff, C. E., Mitchell, 1. 1., Karmilowicz, M. J., Kluve-Beckerman, B., and 47. Samuel, S. K., Hurts, R. A. R., Kondaiah, P., Khalil, N., Turley, E. A., Wright, J. A.. Benson, M. D. Autocrine induction of collagenase by serum amyloid A-like and andGreenberg,A.H.Autocrineinductionoftumorproteaseproductionandinvasion @2-microglobulin-like proteins. Science (Washington DC), 243: 655—657, 1989. by a metallothionein-regulatedTGF-@1(Ser223,225).EMBOJ., 11: 1599—1605, 43. Tremble, P. M., Lane, T. F., Sage, E. H., and Werb, Z. SPARC, a secreted protein 1992.

3616

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. Autocrine Factor Enhancing the Secretion of Mr 95,000 Gelatinase (Matrix Metalloproteinase 9) in Serum-free Medium Conditioned with Murine Metastatic Colon Carcinoma Cells

Sumiko Hyuga, Yohko Nishikawa, Keita Sakata, et al.

Cancer Res 1994;54:3611-3616.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/54/13/3611

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://cancerres.aacrjournals.org/content/54/13/3611. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research.