A Thiol Protease Inhibitor Released from Cultured Human Malignant Melanoma Cells

(CANCER RESEARCH 44, 3324-3329, August 1984]

Yasuhiro Nishida,1 Hiroyuki SumÃ,and Hisashi Mihara

Department ol Physiology, Miyazaki Medical College, 5200 Kihara, Kiyotake-cho, Miyazaki-gun, Miyazaki-ken 889-16, Japan

ABSTRACT human breast carcinomas and that the thiol protease activity was highest in the region where tumor invasion had occurred. A thiol protease inhibitor (TPI) was found in culture media of Sloane ef a/. (30, 31) indicated that cathepsin B activity was human malignant melanoma cells (Bowes) at 1.5 to 2.3 units/ higher in high-metastatic murine melanoma than in low-meta- day/flask (full sheet, 75 sq cm). This amount well exceeded that static melanoma. Thus, cathepsin B-like activity has been sug for cultured nonmalignant cells (human fetal lung fibroblasts). In gested to be correlated with tumor invasion or metastasis. the Â¡ntracellularregion of the melanoma cells, TPI activity was In the present study, using one particular established cell line localized mainly in the cytosol fraction. The difference in specific of human malignant melanoma, we examined the TPI activity in activities between the intracellular and extracellular TPI and the the culture media and compared it with that of nonmalignant TPI accumulation in the culture media indicated that cultured fibroblasts. The melanoma cell line was found to release far more melanoma cells release TPI. Partial purification and characteri TPI than did the fibroblasts. We therefore investigated the re zation of the TPI by column chromatography using Sephadex G- leased TPI and cathepsin B activity in the culture media of this 150, papain-Sepharose, and Sephadex G-50, followed by sodium cell line and attempted to evaluate the possible pathophysiolog dodecyl sulfate-polyacrylamide gel electrophoresis, revealed two ical significance of the TPI. distinct TPIs with molecular weights of 56,000 and 9,800 to 10,800. The latter (main) TPI had a high specificity for thiol proteases and was heat stable (60Â°for 60 min), like previously MATERIALS AND METHODS reported normal human TPIs. The inhibitor, however, differed Congo red-elastin (200 to 400 mesh) was purchased from United from normal human TPIs in that it had a lower molecular weight States Biochemical Corporation, Cleveland, OH. BAPA was obtained than any normal TPI, was unable to inhibit bromelain, and exhib from the Protein Research Foundation, Osaka, Japan. H-o-Valylleucyl- ited a mosaic pattern; namely, the low-molecular-weight TPI lysyl-p-nitroanilide was from AB Kabi Diagnostica, Sweden. BAÃ‘A was resembled liver-type TPI but the pH stability curve resembled from Koch-Light Laboratories, Colnbrook, Berkshire, England. Stem serum-type TPI. The thiol protease, cathepsin B, was not de bromelain (EC 3.4.22.4) was further purified from a commercial bromelain tected in culture media of this human melanoma cell line. (Wako Pure Chemical Company) with the use of the alternative purifica tion method of Murachi (15). Cathepsin B (EC 3.4.22.1) was purified from human liver by the modified method of Towatari ef a/. (37). Instead INTRODUCTION of acetone treatment, lyophilization was performed. Elution on carboxy- Many kinds of TPIs2 have been separated from microorga methyl-Sephadex C-50 was carried out with 0.2 M NaCI in the buffer. Ficin (EC 3.4.22.3), bovine trypsin (EC 3.4.21.4), bovine Â«-chymotrypsin nisms (leupeptin, antipain, chymostatin, and E-64) and from (EC 3.4.21.1), and porcine stomach pepsin (EC 3.4.23.1) were products various specimens of several animals [hen (4, 9), rat (6, 7, 10), of Sigma Chemical Company, St. Louis, MO. Papain (EC 3.4.22.2) was rabbit (38), pig (11,13,16), bovine (24)] as well as human beings. obtained from P-L Biochemicals Inc., Milwaukee, Wl, porcine elastase Human TPIs have been purified or partially purified from normal (EC 3.4.21.11) was from Boehringer/Mannheim GmbH, Mannheim, West serum (25, 26), urine (34, 35), skin (8), and other organs (13). Germany, and human plasmin (EC 3.4.21.7) was from Green Cross However, thus far, there has been no report of TPI release from Corporation, Osaka, Japan. All other chemicals were obtained from malignant cells. commercial sources and were of the best grade available. Human TPIs have been shown to have the physicochemical Cell Culture, Cell Extraction, and Subcellular Fractionation. An function of inhibiting specifically the thiol protease group (ficin, established human malignant melanoma cell line (Bowes) and a human fetal lung fibroblast line were kindly supplied by Professor O. Matsuo, papain, bromelain, cathepsin B, etc.). Udaka and Hayashi (38), Kinki University, Osaka, Japan. These cells were cultured in Coming's Hayashi (5), Ooyama ef al. (16), and Toki ef al. (36) suggested tissue culture flasks (75 sq cm) as described by Rijkin and Collen (23). that TPI is correlated with inflammatory reactions. Sasaki ef al. To collect large pools of conditioned media, confluent flasks were rinsed (27) suggested that one probable role of Â«,,Â«2-TPIwasto neu once with DPBS and twice with serum-free Eagle's minimum essential tralize the proteolytic activity of calcium-activated neutral pro medium (NISSUI 2; Nissui Seiyaku, Co., Ltd., Tokyo, Japan), replenished tease derived from tissues. However, the precise physiological with 7.5 ml of the serum-free medium, and incubated under the same or pathophysiological significance of TPI still remains virtually conditions. The collected cell culture supernatant (serum free) was undetermined. centrifuged (2800 x g for 10 min), and the supernatant was stored at -20Â° until use. On the other hand, Poole ef al. (19) reported that an excess amount of a cathepsin B-like thiol protease was released from Monolayer cultures were washed twice with ice-cold DPBS and re moved from the surface with 1 ITIMEDTA in Ca2+- and Mg2+-free DPBS 1To whom requests for reprints should be addressed. by pipeting gently. The cell volume was determined after centrifugaron 2 The abbreviations used are: TPI, thiol protease inhibitor; LTPI, low-molecular- (800 x g for 5 min). The cell pellets were frozen (-20Â°) and rewarmed weight thiol protease inhibitor; BAPA, N-Â«-benzoyl-L-arginine-p-nitroanilide; BAÃ‘A, at 37Â°for 10 min in an equal volume of 0.5% Triton X-100, sonicated W-

3324 CANCER RESEARCH VOL. 44

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1984 American Association for Cancer Research. Human Melanoma Thiol Protease Inhibitor was carried out as described by De Duve ef al. (3) and Quigley (21). The removed and centrifuged cell pellets were washed again by centrifugation (800 x g for 5 min) in ice-cold homogenizing medium and were homog enized until 90 to 98% of the cells appeared broken. The homogenates were then subjected to differential centrifugation to provide a nuclear pellet (600 x g for 10 min), a heavy mitochondrial pellet (10,000 x g for 5 min), a light paniculate pellet (100,000 x g for 90 min), and a soluble fraction (supernatant from the 100,000 x g step). The pellets were finally resuspended in the homogenizing medium. Inhibition and Enzyme Assay. Ficin, papain, bromelain, trypsin, chymotrypsin, and plasmin activities were assayed by the caseinolytic (ca sein; Hammarsten) technique described previously (32, 33), except that cysteine (base free) was included at 5 HIM in the cases of ficin, papain, and bromelain. Cathepsin B activity was assayed by the method of Barrett (1) using 5 mw BAÃ‘A as substrate and a total assay volume of 1.0 ml. Neutral thiol protease activity was assayed by the modified method of Otto and Bhakdi (17) using 4.6 mM BAPA and the incubation buffer, pH 7.4. Pepsin activity was determined by the method of Ume- zawa (40). Elastase activity was assayed by the method of Shotton (28) as modified by Umezawa (39) using Congo red-elastin as substrate. Ficin activity was also estimated with a synthetic substrate H-o-valyl- leucyllysyl-p-nitroanilide by the modified method of Claeson ef a/. (2). We had confirmed that this synthetic substrate was suitable for sensitive 20 40 60 80 100 colorimetrie assay. Trypsin-like activity was also determined with the Concentrated media lull synthetic substrate by almost the same method as for caseinolysis, but Chart 1. Inhibitory effect of human melanoma cell-conditioned media on ficin using 0.46 mM BAPA. activity. Concentrated media were prepared by lyophilization of crude conditioned To estimate the inhibitory activities, enzyme inhibitor mixtures were media for 24 hr. a, caseinolysis with 30-fold-concentrated media; b, amidolysis with preincubated at 37Â°for 5 min, and the residual enzyme activities were 6-fold-concentrated media (the reaction mixture was 0.6 ml with 0.1 M9 ficin, 0.5 then determined as described above. One unit of TPI was defined as the mM H-D-valylleucyllycyl-p-nitroanilide, and 5 mw cysteine, pH 7.4). amount required to inhibit 1 ^g of ficin as a representative thiol protease and was usually calculated from the amount causing 50% inhibition 0.5 under standard assay conditions. The inhibition curves became almost linear at the center of the residual ficin activity plots (Chart 1), and the inhibitory activities calculated from the curves for caseinolysis (Chart 1a) 0.4 and amidolysis (Chart 1b) agreed well with each other. Miscellaneous Techniques. Papain-Sepharose was prepared as de scribed by JÃ¤rvinen (7). Protein concentration was measured by the method of Lowry ef al. (14), using bovine albumin (Grand Island Biological Company, Grand Island, NY) as the reference. SDS-PAGE was per formed according to the modified method of Laemmli (12). 0.2

RESULTS 0.1 Presence and Accumulation of Human Melanoma TPI The thiol protease inhibitory activity in the melanoma cell- conditioned media (serum free) was examined by the general method using ficin. As shown in Chart 1, the conditioned media 12 24 48 inhibited ficin under both assay conditions of caseinolysis (Chart 1a) and amidolysis (Chart 1b). These results indicated that the Incubation time (hr) culture supernatant of melanoma cells contained TPI. Chart 2. Accumulation of TPI in culture media of human malignant melanoma cells and human fetal lung fibroblasts. â€¢,melanoma cells; O, fibroblasts. Points, We investigated the accumulation of TPI in the culture media mean accumulated TPI activity; oars, S.D. (N = 4). Each group of melanoma cells of human melanoma cells as compared to nonmalignant cells and fibroblasts formed a confluent monolayer on the 75-sq cm surface of the (human fetal lung fibroblasts). Each group of melanoma cells and culture flask. U, units. fibroblasts formed a confluent monÃ³layer on the 75-sq cm sur face in the culture flask. In the case of the human melanoma 6-, 12-, 24-, and 48-hr-incubated media were 1.60, 1.90, 2.79, cells, the TPI activity in the culture supernatant increased pro and 1.64 units/mg protein, respectively (data not shown in Chart gressively with increasing incubation time (Chart 2). On the other 2). The specific activity of TPI in the culture media thus increased hand, in the case of the fibroblasts, TPI activity was barely up to 24 hr incubation. detectable and scarcely increased. The fibroblasts could not be Cell extraction and fractionation studies were then performed stably cultured with serum-free medium; therefore, we were to confirm the presence of TPI in the cells and the localization of unable to investigate them after 24 hr. We clearly recognized the TPI in the subcellular components. These techniques were that human melanoma cells accumulated TPI activity in the carried out twice, and the 2 sets of results were found to be culture supernatant, and the mean specific activities of TPI in the similar. One set is shown in Table 1. TPI activity was also

AUGUST 1984 3325

Tabtei double peaks, corresponding to molecular weights of about Intracellularactivity and subcellular distribution of human melanoma TPI 56,000 and less than 12,500, respectively. However, the LTPI activityml) TPI activity formed the main peak. We searched all fractions by %0.40 (mg)" Units" passage through a Sephadex G-150 column for several activities: neutral thiol protease activity; cathepsin B activity; trypsin-inhib 100SubcellularCell extractionveCell 30.0 8.1 itory activity; and trypsin-like activity. However, none of these 4 fractionation activities could be detected in any fraction in spite of 114-fold Nuclearfraction (600 x g for 10 min) 0 0 Heavy mitochondrialfraction (10,000 8.1514.06.91 1.5 concentration of the applied sample under such assay conditions x g for 5 min) (Chart 3). The results indicated that both types of TPI were Light particulate fraction (100,000 x 0 0 unable to inhibit trypsin and revealed no other biochemical func g for 90 min) Solublefraction (100,000 x g super 86.00.50 12.15 9.2 tions. natant)Total'Total Partial Purification of Human Melanoma TPI 10023456i 36.56 10.7 a"Casein$.0I The whole-purification procedure was carried out at 4Â°. A olysis.14/-9.35 volume of 12,664 ml of serum-free melanoma cell-conditioned medium for 24 hr was used as starting material. This contained 1,242 mg of protein and 3,604 units of TPI activity. The specific 1HiÂ«i i i activity of the crude TPI was thus 2.9 units/mg protein. Papain-Sepharose Affinity Chromatography. As shown in 4.0a* Chart 4, TPI activity was detected in only 20 mw NaaPO4, pH 3.0i- > 12.1, containing 0.1 M NaCI elution fractions but was partially ^ai0 separated from the main protein peak. These active fractions 2.00.1.000.210i"* were collected and dialyzed against 0.15 M NH4HCO3 and then \A lyophilized. The same procedures were repeated several times, and the lyophilized samples were pooled at -80Â°. M^J We also examined the cathepsin B activity and neutral thiol vyi â€¢â€¢1â€”-oâ€”ooâ€”^~-_0 protease activity in other fractions in which TPI could not be eluted. Conditioned media contained TPI, but such fractions did not contain the inhibitor. The 2 thiol protease activities should rÃ¯â€” _l_1 II thus be more detectable in these fractions than in the crude media. However, we failed to detect cathepsin B activity and

0i 00--000 -ooÃiw-ntwiK neutral thiol protease activity. We further concentrated these i<> 1 t 1as0.40.3 io fractions to about 10-fold using solid polyethylene glycol 20,000 *io0.21 in order to enhance the detectability of the 2 activities. However, â€¢â€”â€¢â€¢ â€¢ â€¢â€”<â€¢Aif4^\=- still no cathepsin B activity or neutral thiol protease activity was detected. Sephadex G-50 Chromatography. Next, pooled TPI obtained â€¢Â» nmount, from papain-Sepharose affinity Chromatography was gel filtered 10 20 30 40 50 60 70 80 90 on Sephadex G-50 (superfine) in order to separate the main Fraction number Chart 3. Gel filtration of human melanomacell-conditionedmedia on Sephadex G-150. Pooledmediaof 250 ml were concentrated to 2.2 ml by a solid polyethylene glycol 20,000 and lyophilization technique and were centrifuged (1,500 x g for 10 6.0 - 2.0 min). The supernatant (2.0 ml) was applied to a column (1.6 x 70 cm) and equilibrated with 10 mw phosphate buffer-0.2 M NaCI(pH 7.4), at 4Â°ata flow rate 5.0 - of 9.7 ml/hr. Several bioactive proteins were measured in all fractions: O, TPI activity; A, neutral thiol protease activity (500 Â¿ilofeachfraction as enzyme solution, incubation for 2 hr at 37Â°);D,cathepsin B activity (500 ^l as enzyme solution, for = 4.0 r- 2 hr at 40Â°);â€¢,trypsin-inhibitoryactivity (expressed as percentage inhibition; 100 (il as inhibitor solution, using 1 /Â¿gtrypsin);A, trypsin-like activity (100 pi as enzyme Â£ 1 3.0 1.0 solution, for 2 hr); 9, absorbance at 280 nm. Arrows, elution positions of the following standard proteins: 7, Blue Dextran; 2, aldolase (M, 158,000); 3, bovine serum albumin (M,67,000); 4, egg albumin (M,45,000); 5, chymotrypsinogenA (M, 2.0 25,000); 6, cytochrome c (M, 12,500). U, units. 1.0 observed in the cells at about 21 units/ml of cell volume. Eighty- 0 six% of the total intracellular TPI activity was found in the soluble I i l L. 10 20 30 40 50 60 70 90 100 110 fraction (cytosol). Another 14% was present in the heavy mito chondrial fraction. The specific activity of the intracellular TPI Fraction number Chart 4. Affinity chromatogram of human melanomacell-conditionedmedia on was about 0.28 unit/mg protein, and that of the cytosol TPI was papain-Sepharose.The column (1.4 x 3.5 cm) was equibrated with 10 mM phos about 0.76 unit/mg protein. phate buffer-0.1 M NaCI, pH 6.0. Eluents: 7, the equilibration buffer; 2, 10 mM The molecular distribution of melanoma TPI was investigated phosphate buffer-3 M KCI,pH 6.0; 3, the equilibration buffer; 4, 20 HIMNajPCvO.I MNaCI,pH 12.1. Each unadsorbedfraction was 10 ml, and all other fractions were by Sephadex G-150 (fine) gel filtration of crude conditioned 2.0 ml in volume. O, TPI activity; â€¢,absorbanceat 280 nm; I 1,samplecollected media. As shown in Chart 3, the TPI activity was eluted as (Fractions 79 to 92). U, units.

3326 CANCER RESEARCH VOL. 44

melanoma TPI, LTPI. The LTPI activity was eluted as a symmet rical single peak which coincided with a very small protein peak (Chart 5), and its molecular weight was 10,800 as calculated using the calibration proteins. The high-molecular-weight TPI was also eluted in fractions of the void volume as a very small peak. LTPI fractions were collected and dialyzed against 0.15 M NH4HCO3 and then lyophilized. The specific activity was 3,215 units/mg protein. The same procedures were repeated, and 87 tig (278.9 units) of partially purified melanoma TPI were finally obtained from 12,664 ml of serum-free conditioned medium. PAGE. SDS-PAGE of lyophilized LTPI obtained from Sepha dex G-50 was carried out. As shown in Fig. 1, 2 bands appeared. One (Band A) had a molecular weight of 9,800, and the other (Band B) had a molecular weight of 3,100, as calibrated from the 10 20 30 40 60 70 80 90 100 marker proteins. The former corresponded to the value indicated by gel filtration on Sephadex G-50. Fraction number Chart 5. Gel filtration of papain-Sepharose eluate on Sephadex G-50. Pooled Characterization of Melanoma TPI TPI obtained from a papain-Sepharose column was dialyzed, lyophilized, and then dissolved in 5 ml of the equilibration buffer and centrifuged at 1500 x g for 10 min. The resultant supernatant (1.0 ml) was applied to a Sephadex G-50 column (1.6 x Inhibition Spectrum. The inhibition spectrum of human mela 40 cm) and equilibrated with 10 mw phosphate buffer, pH 7.4, containing 0.2 M noma TPI was investigated using partially purified LTPI. The NaCI. Fractions of 1.0 ml were collected at a flow rate of 21 ml/hr. O, TPI activity; amount of each enzyme used is indicated in //g in Chart 6. The â€¢,absorbance at 280 nm; I 1,sample collected (Fractions 41 to 49). Arrows, elution positions of the following standard proteins: T, Blue Dextran; 2, chymotryp- incubation times were all 1 hr, except in the case of the ficin sinogen A (M, 25,000); 3, cytochrome c (M, 12,500); 4, aprotinin (M, 6,512); 5, inhibition assay where it was 45 min. As shown in Chart 6, bacitracin (M, 1,450). U, units. melanoma TPI strongly inhibited human cathepsin B, ficin, and papain but not bromelain. The TPI did not inhibit the serine protease group (trypsin, chymotrypsin, and plasmin), pepsin, or elastase using the standard assay methods. Human melanoma LTPI was found to have a high specificity for thiol proteases, although among the thiol proteases bromelain was not inhibited by melanoma LTPI.

100

ficin bromelain plasmin 50 113.3pgI Chymotrypsinogen A (25,000)

- 100 Cytochrome c (12,500) Â£

papain trypsin pepsin A 50 lljjg) is- Aprotinin (6,512) 0 Insulin B-chain 100' (3,495)

cathepsin6 chymotrypsin elastase 50 llp-g) ( lOpg )

Fig. 1. SDS-PAGE of partially purified melanoma TPI. Inhibitor (0.2 Â¿ig)was 0.3 0.6 0.3 0.6 1.2 0.3 0.6 1.2 dissolved in 0.1 ml of buffer and applied to the gel. The separation gel was 15% gel. The electrophoresis was carried out at a current of 60 ma/gel for 2.5 hr. Inhibitor (ug) Proteins were stained with a Bio-Rad Silver Stain Kit (Bio-Rad Laboratories, Richmond, CA). The migration positions of 4 marker proteins are indicated on the Chart 6. Inhibitory effect of melanoma LTPI on several proteases. The assay conditions were as described in "Materials and Methods." right. Numbers in parentheses, molecular weight.

AUGUST 1984 3327

Stability. Three stability tests were performed using the par teins flowed out into the culture media and the extracellular tially purified melanoma TPI. After 5 freezings and thawings, proteins increased. The specific activity of TPI in the 48-hr about 96% of the activity remained. At neutral and alkaline pH cultured media thus decreased in spite of the TPI activity in values, the TPI was stable, but it was labile below pH 5.0 (Chart crease. The specific activity of TPI in the 24-hr-incubated media 7). As shown in Chart 8, it was relatively stable below 60Â°but was about 10-fold higher than that of the intracellular TPI or was gradually inactivated above 70Â°. about 3.7-fold higher than that of the cytosol TPI. These results demonstrated that the cultured human melanoma cells were releasing TPI. Approximately 1.5 to 2.3 units of TPI per day was DISCUSSION released per culture flask (full sheet, 75 sq cm; 1.0 Â±0.3 x 107 TPI activity was first detected in the culture media of one (S.D.) cells; about 0.067 ml of cell volume). On the other hand, particular established cell line (Bowes) of malignant melanoma. in culture media of nonmalignant cells (human fetal lung fibro- We then confirmed the presence of TPI activity also in the blasts), TPI activity was neither accumulated nor detected. intracellular parts of the melanoma cells. Extracellular TPI activity Preliminary examinations by Sephadex G-150 column chro- was accumulated progressively in the culture media as the matography and partial purification of human melanoma TPI were incubation time increased (Chart 2). The specific activity of TPI carried out. The latter involved 2 column Chromatographie steps, in the culture media also increased until 24 hr of incubation. with papain-Sepharose and with Sephadex G-50. The human However, melanoma cells cultured for 48 hr without fetal calf melanoma TPI consisted of 2 types, high-molecular-weight TPI serum became more broken, so that abundant intracellular pro- and LTPI, of which LTPI was the main type. LTPI was therefore partially purified, and the LTPI thus obtained had a specific activity of 3215 units/mg protein. 100 oâ€”oâ€”oâ€”o Based on our investigations of the properties of the partially purified melanoma LTPI, the inhibitor was found to have a high specificity for thiol proteases and to be heat stable (60Â°for 60 min) like normal human TPIs. However, it displayed 3 distinct characteristics, (a) The melanoma LTPI had a lower molecular Â¡a- weight than did any of the normal human TPIs previously re ported. Normal human serum TPI is reported to consist of 2 broad types. One is full-component TPI with a molecular weight of about 170,000, and the other is its subunit-like groups with molecular weights of approximately 90,000 and 95,000 (25, 26). Urinary TPIs have been reported to have molecular weights of 76,000 and 22,500 (35). Epidermal TPI has a small molecular weight (from 12,100 to 13,100) (8). The present melanoma LTPI revealed molecular weights of 10,800 (gel filtration) or 9,800 (SDS-PAGE). (b) The melanoma LTPI was unable to inhibit bromelain, although normal human TPIs have been reported to inhibit bromelain with slightly lower activity than in their inhibition Chart 7. pH stability curve of melanoma LTPI. The TPI was incubated in Britton- Robinson buffer of various pH values as indicated on the abscissa for 18 hr at 4Â°. of other thiol proteases, (c) Concerning the combination of It was then neutralized by 30-fold dilution with 0.1 M phosphate buffer, pH 7.4, and individual properties, recently, Wakamatsu ef al. (41) reported the residual inhibitory activity was estimated. that TPIs could be divided into 2 types, the serum type and the liver type, based on differences in their molecular weights, anti- 100 genicity, inhibition spectrum, and affinity to concanavalin A- Sepharose. The serum type had a large molecular weight (more than 90,000) (26, 41), and its inhibitory activity was stable at neutral pH but unstable at acid pH (26). The liver type had a small molecular weight (about 12,500) (8,10, 41), and its inhibi tory activity was unstable at neutral pH but stable at acid pH (10, 13). The present melanoma LTPI, however, cannot be assigned to either type. It is similar to the liver type from the viewpoint of its molecular weight but resembles the serum type from the viewpoint of its pH stability. This distinction from normal human TPIs is significant. Clearly, it is necessary to determine the detailed properties of melanoma TPI in order to assess its usefulness as a possible diagnostic tool for human melanomas. We plan to undertake immunological and enzymologial investi gations in the future. O 20 40 60 100 Poole and Recklies ef al. (20, 22) reported that human breast Temperature O carcinoma secreted a thiol protease, cathepsin B, at significantly Charts. Heat stability curve of melanoma LTPI. The TPI was incubated at higher levels than found in nonmalignant tissue. Sloane ef al. various temperatures as indicated on the abscissa for 1 hr in 0.1 M phosphate buffer, pH 7.4. It was then rapidly cooled at 4Â°.and the residual inhibitory activity found that the cathepsin B activity was 3-fold higher in homog- was estimated. enates of B16 F10 tumor (a murine melanoma cell line of high

3328 CANCER RESEARCH VOL. 44

10. Kominami, E., Wakamatsu, N., and Katanuma, N. Purification and characteri metastatic potential) than in those of B16 F1 tumor (the same zation of thiol proteinase inhibitor from rat liver. J. Biol. Chem., 257: 14648- with a low metastatic potential) (30) and indicated that cathepsin 14652,1982. B activity could represent a possible marker for metastatic 11. Kopitar, M., Brzin, J., Zvonar, T., LÃ²cnikar, P., Kregar, I., and Turk, V. Inhibition studies of an intracellular inhibitor on thiol proteinases. FEBS Lett., 91: 355- potential (31). Pietras ef al. (18) reported that the serum cathepsin 359, 1978. B level was high in women with clear cell adenocarcinoma of the 12. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head genital tract. Thus, increase in cathepsin B activity has been of bacteriophage T4. Nature (Lond.), 227: 680-685,1970. 13. Lenney, J. F., Tolan, J. R., Sugai, W. J., and Lee, A. G. Thermostable correlated with the ability of primary malignant neoplasms to endogenous inhibitors of cathepsin Band H. Eur. J. Biochem., 707:153-161, invade or metastasize normal tissues. In the present study, 1979. however, human melanoma cells appeared not to secrete ca 14. Lowry, O. H., Rosebrough, N. J., Fair, A. L., and Randall, R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275, thepsin B and neutral thiol protease under the conditions of 1951. tissue culture used. Human melanoma cells were found to re 15. Murachi, T. Bromelain enzymes. Methods Enzymol., 79: 273-284,1970. 16. Ooyama, T., Sakamoto, H., and Mayumi, M. Studies on the role of proteases lease dominantly TPI which strongly inhibits cathepsin B activity. in the biochemical mechanisms of tissue injury. Med. Biol., 53:462-468,1975. We cannot make a comparison with TPI release from other 17. Otto, K., and Bhakdi, S. Zur Kenntnis des Kathepsins B': SpezifitÃ¤t und Eigenschaften. Hoppe-Seyler's Z. Physiol. Chem., 350: 1577-1588, 1969. human melanoma cell lines or other cell lines, since there are as 18. Pietras, R. J., Szego, C. M., Mangan, C. E., Seeler, B. J., Burtnett, M. M., and yet no further reports of such cells releasing TPI. We need to Orevi, M. Elevated serum cathepsin B, and vaginal pathology after prenatal investigate the TPI activity of other malignant cell lines and also DES exposure. Obstet. Gynecol., 52: 321 -327, 1978. of tissues removed from patients with cancer. Further, it is clear 19. Poole, A. R., Recklies, A. D., and Mort, J. S. Secretion of proteinases from human breast tumors: excessive release from carcinomas of a thiol proteinase. that we need to consider the inhibitor as well as cathepsin B In: P. StrÃ¤uli,A. J. Barrett, and A. Baici (eds.), Proteinases and Tumor Invasion, when discussing the mechanisms of invasion and metastasis of pp. 81-95. New York: Raven Press, 1980. human melanoma cells and other malignant tumor cells in terms 20. Poole, A. R., Tiltman, K. J., Recklies, A. D., and Stoker, T. A. M. Differences in secretion of the proteinase cathepsin B at the edges of human breast of thiol proteases. carcinomas and fibroadenomas. Nature (Lond.), 273: 545-547,1978. There are several reports suggesting that TPI may play a 21. Quigley, J. P. Association of a protease (plasminogen activator) with a specific membrane fraction isolated from transformed cells. J. Cell Biol., 77: 472-486, significant role in regulating the inflammatory process (5,16, 38). 1976. Moreover, Toki ef al. (36) reported that purified urinary TPI 22. Recklies, A. D., Mort, J. S., and Poole, A. R. Secretion of a thiol proteinase from mouse mammary carcinomas and its characterization. Cancer Res., 42: suppressed the inflammatory reaction in Arthus skin lesions in 1026-1032,1982. rats. Based on these facts, it could also be speculated that the 23. Rijken, D. C., and Collen, D. Purification and characterization of the plasmin released TPI may in some way be related to the inflammatory ogen activator secreted by human melanoma cells in culture. J. Biol. Chem., 256:7035-7041,1981. reaction of hosts with a melanoma such as that used in the 24. Roughley, P. J., Murphy, G., and Barrett, A. J. Proteinase inhibitors of bovine present experiment. nasal cartilage. Biochem. J., 769: 721-724, 1978. 25. Ryley, H. C. Isolation and partial characterization of a thiol proteinase inhibitor from human plasma. Biochem. Biophys. Res. Commun., 89: 871-878, 1979. ACKNOWLEDGMENTS 26. Sasaki, M., Taniguchi, K., and Minakata, K. Multimolecular forms of thiol proteinase inhibitor in human plasma. J. Biochem. (Tokyo), 89:169-177,1981. The authors wish to thank Professor O. Matsuo, Department of Physiology, 27. Sasaki, M., Taniguchi, K., Suzuki, K., and Iwahori, K. Human plasma Â«,-and Â«Â¡,-thiolproteinaseinhibitors strongly inhibit Ca-activated neutral protease from Kinki University School of Medicine, Sayama 589, Osaka, Japan, for his kind supply muscle. Biochem. Biophys. Res. Commun., 770: 256-261, 1983. of the melanoma cell line. They also thank Dr. S. Hashida, Department of Biochem 28. Shotton, D. M. Elastase. Methods Enzymol., 79: 113-140,1970. istry, Miyazaki Medical College, for his kind suggestions regarding the purification 29. Singh, H., and kalnitsky, G. Separation of a new a-N-benzoyl-arginine-/3- of human cathepsin B, and Dr. T. Nakayama, Depratment of Biochemistry, Miyazaki Medical College, for his kind support in carrying out the SDS-PAGE. naphthylamide hydrolase from cathepsin B,. J. Biol. Chem., 253: 4319-4326, 1978. 30. Sloane, B. F., Dunn, J. R., and Honn, K. V. Lysosomal cathepsin B: Correlation with metastatic potential. Science (Wash. DC), 272: 1151 -1153,1981. REFERENCES 31. Sloane, B. F., Honn, K. V., Sadler, J. G., Turner, W. A., Kimpson, J. J., and Taylor, J. D. Cathepsin B activity in B16 melanoma cells: a possible marker 1. Barrett, A. J. Cathepsin B and other thiol proteinases. In: A. J. Barrett (Ed.), for metastatic potential. Cancer Res., 42: 980-986,1982. Proteinases in Mammalian Cells and Tissues, pp. 181-208. Proteinases in 32. Sumi, H., and Fujii, S. Purification of a new plasma inhibitor of kallikrein. J. Mammalian Cells and Tissues, pp. 181-208. Amsterdam: Elsevier/North- Biochem. (Tokyo), 75: 541-551,1974. Holland BiomÃ©dicalPress, Inc., 1977. 33. Sumi, H., and Muramatu, M. Purification and partial characterization of human 2. Claeson, G., Friberger, P., KnÃ¶s,M., and Eriksson, E. Method for determination C,-esterase. Agr. Biol. Chem., 38: 605-611,1974. of prekallikrein in plasma, grandular kallikrein and urokinase. Haemostasis, 7: 34. Sumi, H., and Toki, N. Inhibitors of acrosin and SH-protease in normal human 76-78, 1978. urine. Proc. Soc. Exp. Biol. Med., 767: 530-535,1981. 3. De Duve, C., Pressman, B. C., Gianetto, R., Wattiaux, R., and Appelmans, F. 35. Taniguchi, K., Ito, J., and Sasaki, M. Partial purification and properties of Tissue fractionation studies: 6. Intracellular distribution patterns of enzymes in urinary thiol proteinase inhibitors. J. Biochem. (Tokyo), 89: 179-184,1981. rat-liver tissue. Biochem. J., 60: 604-617,1955. 36. Toki, N., Hamanaka, K., Ishizu, K., Takasugi, S., Sumi, H., and Sato, H. Effects 4. Possum, K., and Whitaker, J. R. Ficin and papain inhibitor from chicken egg of human urinary thiol protease inhibitor on Arthus-type reaction in rabbit skin white. Arch. Biochem. Biophys. 725: 367-375, 1968. (in Japanese) Jpn. J. Allergol., 37: 764,1982. 5. Hayashi, H. Mechanisms of generation and inhibition of inflammation (in Japa 37. Towatari, T., Kawabata, Y., and Katunuma, N. Crystallization and properties nese). Trans. Soc. Pathol. Jpn., 56: 37-63, 1967. of cathepsin B from rat liver. Eur. J. Biochem., 702: 279-289,1979. 6. Iwata, D., Mirado, M., Niinobe, M., and Fujii, S. Existence of two kinds of thiol 38. Udaka, K., and Hayashi, H. Further purification of a protease inhibitor from protease inhibitors in rat serum. Biochem. Biophys. Res. Commun., 704: rabbit skin with healing inflammation. Biochim. Biophys. Acta, 97: 251-261, 1525-1531,1982. 1965. 7. JÃ¤rvinen,M. Purification and properties of two protease inhibitors from rat skin 39. Umezawa, H. Elastatinal-inhibiting elastase. Methods Enzymol., 45: 687-689, inhibiting papain and other SH-proteases. Acta Chem. Scand. B, 30:933-940, 1976. 1976. 40. Umezawa, H. Pepstatin-inhibiting acid proteases. Methods Enzymol., 45:689- 8. JÃ¤rvinen,M. Purification and some characteristics of the human epidermal SH- 693, 1976. protease inhibitor. J. Invest. Dermatol., 77: 114-118,1978. 41. Wakamatsu, N., Kominami, E., and Katunuma, N. Comparison of properties of 9. Keilova, B., and TomÃ¡sek, V. Effect of papain inhibitor from chicken egg white thiol proteinase inhibitors from rat serum and liver. J. Biol. Chem., 257:14653- on cathepsin Bi. Biochim. Biophys. Acta, 334: 179-186,1974. 14656,1982.

AUGUST 1984 3329

Yasuhiro Nishida, Hiroyuki Sumi and Hisashi Mihara

Cancer Res 1984;44:3324-3329.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/44/8/3324

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/44/8/3324. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.