Leukemia Research Vol. 12, No. 5, pp. 419-422, 1988. 0145-2126/88 $3.00 + .00 Printed in Great Britain. Pergamon Press plc

PRODUCTION OF AN ACTIVE UROKINASE BY LEUKEMIA CELLS: A NOVEL DISTINCTION FROM CELL LINES OF SOLID

TUMORS

Ross STEPHENS, RIITI'A ALITALO,* HANNELE TAPIOVAARA and ANTFI VAHERI

Department of Virology and * Transplantation Laboratory, University of Helsinki, Helsinki, Finland

(Received 15 December 1987. Revision accepted 13 February 1988)

Abstraet--A new screening test is described which enabled rapid determination of the proportion of single-chain and two-chain urokinase produced in the culture supernatants of 18 human cell lines. A clear distinction was found between two groups of cell lines: cells derived from ten solid tumors produced almost exclusively single-chain proenzyme, while the majority of the enzyme found in cultures of eight leukemia cell lines was in the active, two-chain form.

Key words: Two-chain urokinase, proenzyme, leukemia, solid tumors.

I N T R O D U C T I O N

THE MALIGNANT transformation of animal cells has been shown to lead to increased secretion of plasm- inogen-activating enzymes [1, 2]. Cells expressing plasminogen activators can produce plasmin f rom the plasminogen present in plasma and other body fluids, so that they possess a high potential for extracellular proteolytic activity. The plasmin formed by these means can have direct [3] as well as indirect [4] actions which lead to destruction of extracellular matrix and detachment of cells. Evidence is now available which implicates plasminogen activator and plasmin as initiators of a proteolytic cascade which facilitates the invasive propert ies of tumor cells [5- 8].

It has been shown by several authors [9-12] that cultured tumor cells secrete the single-chain pro- enzyme form of urokinase (scu-PA), which has very little, if any, action on plasminogen. Efficient acti- vation of plasminogen requires the conversion of scu- PA to its two-chain, active form (tcu-PA) [9, 10]. This reaction can be catalysed by plasmin [9-12].

However , it may also be possible for tumor cells to convert scu-PA to tcu-PA by using other, as yet unidentified, cell-associated proteases. In this report

Abbreviations: u-PA, urokinase type plasminogen acti- vator; scu-PA, single-chain (proenzyme) form of uro- kinase-type plasminogen activator; tcu-PA, two-chain (active) form of urokinase; NPGB, para-nitrophenyl guanidinobenzoate.

Correspondence to: Dr R. W. Stephens, Department of Virology, University of Helsinki, Haartmaninkatu 3, SF- 00290 Helsinki, Finland.

we explore such a possibility by testing culture super- natants of 18 human tumor cell lines for the presence of tcu-PA. It was found that while cells from solid tumors produced almost exclusively scu-PA, tcu-PA was found to be the major form of urokinase pro- duced by all the leukemic lines studied.

419

M A T E R I A L S A N D M E T H O D S

All cells used were of human origin. The following lines were obtained from the American Type Culture Collection (ATCC): HT-1080, fibrosarcoma, CCL 121; RD, embry- onal rhabdomyosarcoma, CCL 136; SV40 virus-trans- formed embryonic lung fibroblast WI-26 VA4, CCL 95.1; MRC-5, embryonic lung fibroblast, CCL 171; WI-38, embryonic lung fibroblast, CCL 75 and A549, lung carcinoma, CCL 185. The epidermoid carcinoma A-431 and the fibrosarcoma 8387 were obtained from Dr G. Todaro and Dr J. De Larco at the Laboratory of Viral Carcinogenesis, National Cancer Institute, Bethesda, MD. The double transformed (Rous sarcoma virus, SV40) WI- 38 fibroblast lines RSA and RSB were a gift from Dr T. Kuwata [13]. The following human leukemic cell lines were used in the study: K-562, an erythroblastoid stem cell line derived from the pleural effusion of a patient with chronic myeloid leukemia in blast crisis [14]; HEL, an erythro- leukemic line [15]; HL-60, a promyelocytic leukemic line with a bipotential differentiation capacity (granulocyte/ macrophage) [16]; KG-1 and ML-2, myeloid leukemia lines [17]; U-937, a promonoblast leukemia line [18]; MOLT- 4, a T-cell leukemia line [19] and RC2A, a myelomonocytic leukemic line [20]. The leukemia cell lines were obtained either through the ATCC or from Professor Leif Andersson, Department of Pathology, University of Helsinki.

Solid tumor cells were grown to form confluent layers in Linbro plastic wells (2 cm2; Flow Laboratories) in Eagle's minimal essential medium (MEM) and leukemia cells were

420 R o s s STEPHENS et al.

grown to a density of 0.5-1.0 x 106/ml in RPMI 1640 in Falcon bottles, both supplemented with 10% heat-inac- tivated fetal calf serum (Gibco), 100 IU/ml penicillin and 50 p,g/ml streptomycin.

Twenty hours before collection of media for assay, the cells were changed to serum-free medium supplemented with 0.1% bovine serum albumin (fraction V, Boehringer Mannheim GmbH).

The conditioned medium was tested for scu-PA and tcu- PA by the following immunocapture method [21]. Micro- titre wells of polystyrene immunoplates (type 269620, A/S Nunc, Roskilde, Denmark) were coated overnight at 37°C with 50 ~tl of a solution of goat IgG antibodies to human urokinase (cat, no. 398, American Diagnostica, Greenwich, CT). The solution contained 2.5 ~tg IgG per ml of 0.1 M sodium carbonate pH 9.8. After washing the wells were treated with conditioned medium (50 ~tl) for 2 h at 23°C, then washed again. Half the wells were then treated with 50~tl of 2~tM p-nitrophenyl guanidino- benzoate (NPGB) (Sigma) for 20 min at 37°C. The other half (controls) received 50 ~tl washing buffer (0.05% Tween 20 in phosphate-buffered saline). After washing, the resid- ual plasminogen activator activity was assayed in all the wells, by addition of 40 ~tl plasminogen solution [100 ~tg/ ml in assay buffer consisting of 50 mM sodium glycinate pH 7.8, 0.1% Triton X-100, 0.1% gelatin and 10mM 6- aminocaproic acid (this was added as a stimulator of the activation of glu-plasminogen, to improve sensitivity)] and incubation for 30 min at 37°C. The plasminogen was puri- fied from fresh human plasma by affinity chromatography on lysine-Sepharose [22]. The plasmin produced was assayed by its thioesterase activity [23] by the addition of 150 ~tl of a solution containing 200 mM potassium phos- phate (pH 7.5), 200 mM KC1, 0.1% Triton X-100,220 ~tM Z-lysine thiobenzyl ester (Peninsula Laboratories, San Carlos, CA) and 220~tM 5,5'-dithiobis(2-nitrobenzoic acid) (Sigma). This mixture was incubated for 30 min at 37°C, and the absorbancies of the wells read at 405 nm. Urokinase (tcu-PA, 60,000 I.U./mg) was purchased from Calbiochem-Behring Corp., La Jolla, CA). Urokinase pro- enzyme (scu-PA) was a gift of Dr K. Dan0, Finsen Insti- tute, Copenhagen, Denmark.

o 2 8 :51 125 500

[ N P G B ] (/./. M )

FIG. 1. Effect of NPGB on bound two-chain urokinase (O---O) and single-chain urokinase (O--Q) in the micro-

plate imrnunocapture assay.

mation of tcu-PA in the assay. The necessity of this plasmin in proenzyme assays has been previously shown with similar preparat ions of plasminogen by the use of inhibitory monoclonal antibodies to plas- min [12]. From the experiments with NPGB treat- ments, it was therefore possible to determine conditions (2 p~M NPGB) which could be used in a standardised assay enabling clear distinction between tcu-PA and scu-PA.

When this test was applied to the culture media from 18 cell lines which produced significant levels of u-PA, it was found that these lines fell into two groups (Fig. 2). Ten cell lines derived from solid tumors all produced more than 80% single-chain u- PA, which was not inhibited by t reatment with 2 ~tM

RESULTS

The solid-phase screening assay employed in this study made use of the fact that active two-chain urokinase (tcu-PA) could be absorbed from culture media with an immobilised polyclonal antibody and still retain activity towards plasminogen substrate [21] and the active-site inhibitor, p-nitrophenyl guan- idinobenzoate (NPGB, see Fig. 1). When the con- centration of N P G B used to treat absorbed u-PA was varied, it was found that tcu-PA was still fully inactivated by very low concentrations of NPGB, down to less than 1 ~M. On the contrary, single- chain urokinase proenzyme (scu-PA), which could also be absorbed by immobilised antibody, did not react with such low concentrations of NPGB. The potential plasminogen activator activity of the scu- PA could be assayed by addition of plasminogen which contained traces of plasmin, to initiate for-

o, g o

CeLL Lines

FIG. 2. Immunocapture assay of single-chain urokinase in culture supernatants of cells derived from solid tissues (left panel) and leukemias (right panel). The percentage activity which was resistant to NPGB inhibition represents the proportion of urokinase which was present as single-chain

proenzyme (scu-PA).

Two-chain urokinase from leukemia cells 421

NPGB. This was independent of the total enzyme produced by each cell line, which varied over a five- fold range. However, eight leukemic cell lines all produced more than 75% two-chain enzyme. The total enzyme activity produced by leukemia cells varied over a six-fold range. Note, however, that the ranges of activity in cultures from solid tumors and leukemia cultures could not be compared directly in this study, due to the difficulty in comparing cell numbers of adherent and non-adherent cells.

Thus, while the total level of enzyme produced in these cultures varied widely, the percentage of enzyme present as the single-chain form was clearly related to their tumor of origin. Cells derived from solid tumors produced predominantly single-chain enzyme, while cells of leukemia origin produced mainly two-chain enzyme.

The difference between the two groups could not be related to the method of culture; the cells in both groups had a high percentage viability and conversion by intracellular proteases from dying cells appeared unlikely. The type of culture media used did not affect the form of the secreted enzyme. RC2A cells secreted tcu-PA irrespective of whether they were in RPMI-1640 or MEM (as used for the cells from solid tumors). The finding of active enzyme in leukemia cultures was also unrelated to the presence of serum albumin, which was a supplement in serum-free media.

cells and leukemia cells may produce inhibitors [32], it appears unlikely that removal of two-chain enzyme by inhibitors can explain the lack of two-chain enzyme in cultures of cells from solid tumors.

The results obtained are consistent with the possi- bility that leukemic cells, unlike cells from solid tumors, possess an enzyme activity with appropriate specificity to substitute for plasmin in the known proteolytic activation of urokinase proenzyme [12]. Leukemic cells are thus able to continuously express active urokinase, probably bound to their surface through specific receptors [33, 34]. Why the origin of the cells, or their adherence/non-adherence in culture should so sharply define this property is as yet unclear, but may possibly be related to the fact that immature blood cells must have an efficient mechanism for escape from the bone marrow, which could conceivably involve cell-surface proteases. By contrast, the cells of tissues which give rise to solid tumors do not normally leave their site of origin. Further studies are needed to provide evidence for such a distinction, and to identify candidate protease(s).

Acknowledgements--The authors thank Anja Virtanen, Pirjo Sarjakivi, Taija Ky6sti6-Renvall and Hilkka Toivonen for excellent technical assistance. This work was supported by the Sigrid Juselius Foundation, Helsinki, and the Medical Research Council of the Academy of Finland.

DISCUSSION

The results obtained here for cells derived from solid tumors are consistent with several published observations. Single-chain urokinase has been found in culture supernatants of many human cell lines derived from solid tumors, including glioblastoma UCT/gI-1 [10], epidermoid carcinoma HEp3 [24], transformed kidney [25], epidermoid carcinoma A- 431 [26], lung adenocarcinoma CALU-3 [27] and colon carcinoma COLO-394 [28]. Moreover, single- chain enzyme has been shown to be the predominant form of urokinase present extracellularly in murine Lewis lung tumors in v i vo [29].

However, relatively little has been reported on the form of urokinase produced by leukemia cell lines. In the studies of phorbol ester stimulated U-937 reported by Genton et al. [30], it was found that a large proportion of extracellular urokinase antigen appeared to be present as a complex with an inhibi- tor. Since complex formation with inhibitors only occurs with two-chain enzyme [31], this result would imply that U-937 cells are able to produce two-chain urokinase, consistent with our findings here with unstimulated cells. Moreover, since both solid tumor

REFERENCES

1. Dan0, K., Andreasen P. A., Gr0ndahl-Hansen J., Kristensen P., Nielsen L. S. & Skriver L. (1985) Plas- minogen activators, tissue degradation and cancer. Ado. Cancer Res. 44, 139.

2. Mullins D. E. & Rohrlich S. T. (1983) The role of proteinases in cellular invasion. Biochim. biophys. Acta 695, 177.

3. Liotta L. A., Goldfarb R. H., Brundale R., Siegal G. P., Terranova V. & Garbiza S. (1981) Effect of plasminogen activator (urokinase), plasmin, and thrombin on glycoprotein and collagenous components of basement membrane. Cancer Res. 41, 4629.

4. O'Grady R. L., Upfold L. I. & Stephens R. W. (1981) Rat mammary carcinoma cells secrete active col- lagenase and activate latent enzyme in the stroma via plasminogen activator. Int. J. Cancer 28, 509.

5. Ossowski L. & Reich E. (1983) Antibodies to plas- minogen activator inhibit human tumor metastasis. Cell 35, 611.

6. Bergman B. L., Scott R. W., Bajpai A., Watts S. & Baker J. B. (1986) Inhibition of tumor-cell-mediated extracellular matrix destruction by a fibroblast pro- teinase inhibitor, protease nexin 1. Proc. natn. Acad. Sci. U.S.A. 83, 996.

7. Mignatti P., Robbins E. & Rifkin D. B. (1986) Tumor invasion through the human amniotic membrane: Requirement for a proteinase cascade. Cell 47, 487.

422 Ross STEPHENS et al.

8. Sullivan L. M. & Quigley J. P. (1986) An anticatalytic monocional antibody to avian plasminogen activator: Its effect on behavior of RSV-transformed chick fibro- blasts. Cell 43, 905.

9. Skriver L., Nielsen L. S., Stephens R. W. & Dan~ K. (1982) Plasminogen activator released as inactive proenzyme from murine cells transformed by sarcoma virus. Eur. J. Biochem. 124, 409.

10. Nielsen L. S., Hansen J. G., Skriver L., Wilson E. L., Kaltoft K., Zeuthen J. & Dan~ K. (1982) Purification of zymogen to plasminogen activator from human glioblastoma cells by affinity chromatography with monoclonal antibody. Biochemistry 21, 6410.

11. Eaton D. L., Scott R. W. & Baker J. B. (1984) Puri- fication of human fibroblast urokinase proenzyme and analysis of its regulation by proteases and protease nexin. J. biol. Chem. 259, 6241.

12. Sim P-S., Fayle D. R. H., Doe W. F. & Stephens R. W. (1986) Monoclonal antibodies inhibitory to human plasmin: Definitive demonstration of a role for plasmin in activating the proenzyme of urokinase-type plas- minogen activator. Eur. J. Biochem. 158, 537.

13. Kuwata T., Oda T., Sekiya S. & Morinage N. (1976) Characterisation of a human cell line successively trans- formed by Rous sarcoma virus and Simian virus 40. J. natn. Cancer Inst. 56, 919.

14. Lozzio C. B. & Lozzio B. B. (1975) Human chronic myelogenous leukaemia cell-line with positive Phi- ladelphia chromosome. Blood 45, 321.

15. Martin P. & Papyannopoulou T. (1982) HEL cells: A new human erythroleukemia cell line with spontaneous and induced globin expression. Science. N.Y. 216, 1233.

16. Collins S. J., Gallo R. C. & Gallagher R. E. (1977) Continuous growth and differentiation of human myeloid leukemic cells in suspension culture. Nature, Lond. 270, 347.

17. Pelicci P. G., Lanfrancone L., Brathwaite M. D., Wolman S. R. & Dalla-Favera R. (1984) Amplification of the c-myb oncogene in the case of human acute myelogenous leukemia. Science, N.Y. 224, 1117.

18. Sundstr6m C. & Nillson K. (1976) Establishment and characterisation of a human histiocytic lymphoma cell line (U-937). Int. J. Cancer 17, 565.

19. Minowada J., Ohnuma T. & Moore G. E. (1972) Rosette-forming human lymphoid cell lines. 1. Estab- lishment and evidence for origin of thymus-derived lymphocytes. J. natn. Cancer Inst. 49, 891.

20. Bradley T. R., Pilkington G. R., Garsson M., Hodgson G. S. & Kraft N. (1982) Cell lines derived from a human myelomonocytic leukemia. Br. J. Haemat. 51, 595.

21. Stephens R. W., Leung K-C., P611~nen J., Salonen E- M. & Vaheri A. (1987) Microplate immunocapture assay for plasminogen activators and their specific inhibitors. J. Immun. Meth. 105, 245.

22. Deutsch D. G. & Mertz E. T. (1970) Plasminogen: Purification from human plasma by affinity chroma- tography. Science, N.Y. 170, 1095.

23. Green G. D. G. & Shaw E. (1979) Thiobenzyl benzyloxycarbonyl-L-lysinate, substrate for a sensitive colorimetric assay for trypsin-like enzymes. Analyt. Biochem. 93, 223.

24. Wun T. C., Ossowski L. & Reich E. (1982) A pro- enzyme form of human urokinase. J. biol. Chem. 257, 7262.

25. Gurewich V., Pannell R., Louie S., Kelley P., Suddith R. L. & Greenlee R. (1984) Effective and fibrin-specific clot Iysis by a zymogen precursor form of urokinase (pro-urokinase): A study in vitro and in two animal species. J. clin. Invest. 73, 1731.

26. Corti A., Nolli M. L., Soffientini A. & Cassani G. (1986) Purification and characterization of single-chain urokinase-type plasminogen activator (pro-urokinase) from human A431 cells. Thromb. Haemat. 56, 219.

27. Stump D. C., Lijnen H. R. & Collen D. (1986) Puri- fication and characterization of single-chain urokinase- type plasminogen activator from human cell cultures. J. biol. Chem. 261, 1274.

28. Stephens R. W., Fordham C. J. & Doe W. F. (1987) Proenzyme to urokinase-type plasminogen activator in human colon cancer: in-vitro inhibition by monocyte minactivin after proteolytic activation. Eur. J. Cancer clin. Oncol. 23, 213.

29. Skriver L., Larsson L-I., Kielberg V., Nielsen L. S., Andreasen P. A., Kristensen P. & Dane K. (1984) Immunocytochemical localization of urokinase-type plasminogen activator in Lewis lung carcinoma. J. Cell Biol. 99, 753.

30. Genton C., Kruithof E. K. O. & Schleuning W-D. (1987) Phorbol ester induces the biosynthesis of gly- cosylated and non-glycosylated plasminogen activator inhibitor 2 in high excess over urokinase-type plas- minogen activator in human U-937 lymphoma cells. J. Cell. Biol. 104, 705.

31. Wun T-C. & Reich E. (1987) An inhibitor of plas- minogen activator from human placenta. J. biol. Chem. 262, 3646.

32. Saksela O., Vaheri A., Schleuning W-D., Mignatti P. & Barlati S. (1984) Plasminogen activators, activation inhibitors and alpha2-macroglobulin produced by cul- tured normal and malignant human cells. Int. J. Cancer 33, 609.

33. Blasi F., Stoppelli M. P. & Cubellis M. V. (1986) The receptor for urokinase plasminogen activator. J. Cell Biochem. 32, 179.

34. Blasi F., Vassalli J-D. & Dano K. (1987) Urokinase- type plasminogen activator: proenzyme, receptor and inhibitors. J. Cell Biol. 104, 801.