Int. J . Cancer: 46, 32&329 (1990) 0 1990 Wiley-Liss, Inc.
Publication of the International Union Against Cancer Publication de I'Union Internationale Contre 1e Cancer
EFFECTS OF TUMOR NECROSIS FACTOR, ALONE OR IN COMBINATION WITH TOPOISOMERASE-11-TARGETED DRUGS, ON HUMAN LUNG CANCER CELL LINES Giuseppe GIACCONE, Chikabumi KADOYAMA', Rhoda MANECKJEE, David VENZON,Richard B. ALEXANDER and Adi F. GAZDAR~ NU-Navy Medical Oncology and Surgery Branches, and the Biostatistics and Data Management Section, National Cancer Institute, and Naval Hospital, Bethesda MD 20814, USA. Tumor necrosis factor a (TNF) exhibits cytotoxic activity on some solid tumors and has been reported to be synergistic with topoisomerase-ll-targeted antineoplastic agents. A wide range of T N F concentrations (from 10 to 10,000 U/ml) was tested in 9 human lung cancer cell lines (5 small-cell and 4 non-small-cell carcinomas) using a semi-automated M l T assay. TNF was not cytotoxic in 8 cell lines, while an adenocarcinoma cell line war marginally sensitive to the cytokine. Usi?g Iz51-TNFwe were able to show the presence of specific binding sites for TNF in 4/9 human lung cancer cell lines. Scatchard analysis of the marginally sensitive cell line showed high-affinity,saturable binding. With 5 cell lines we also tested whether TNF affected the cytotoxicity of doxorubicin and etoposide, 2 topoisomerase Il-targeted drugs which are widely used in the therapy of lung cancer. N o significant increase in cytotoxicity was seen when TNF was added to the 2 anti-neoplastic agents. In contrast to certain other human and mouse lines, human lung cancer cell lines appear to be resistant to TNF, despite the presence of the receptor in some of them; moreover, no synergistic effect of TNF and 2 topoisomerase-11-targeted drugs was evident in these human cell lines.
A number of human cytokines, including recombinant tumor necrosis factor (Y (TNF), have recently entered clinical trials for the treatment of human cancers. Naturally occurring TNF is produced by activated macrophages (Old, 1985). In murine tumor models TNF causes acute hemorrhagic necrosis of tumor nodules 24 to 48 hr after local or systemic injection. In vivo anti-tumor effects have been demonstrated both in syngeneic murine tumor models and in human tumor xenografts in nude mice (Carswell et al., 19756; Oettgen et al., 1980; Haranaka et al., 1984; Asher et al., 1987). In vitro studies indicate both cytostatic and cytotoxic effects against a wide range of human tumor cells (Carswell et al., 1975a,b; Oettgen et al., 1980; Williamson et al., 1983; Spriggs et al., 1988; Sugarman et al., 1985). However, TNF is cytotoxic in less than one-third of the cell lines tested (Frei and Spriggs, 1989). Marked enhancement of cytotoxicity has been reported when TNF is added to topoisomerase-11-targeted anti-neoplastic agents on murine L929 tumor cells in vino (Alexander et al., 19876) and on MBT-2 cells both in vitro and in vivo (Alexander et al., 1 9 8 7 ~ ) . Relatively little is known about the effects of TNF on lung cancer, in vivo and in vitro. In the present investigation, TNF cytotoxic activity and receptor studies were evaluated on a panel of human lung cancer cell lines. In addition, we studied the combination of TNF with doxorubicin or etoposide (VP16), 2 anti-neoplastic agents known to act through interactions with topoisomerase II (Ross, 1985). These 2 drugs are among the most active and widely used agents in the chemotherapy of lung cancer (Minna et al., 1985) and a study of the effects of these drugs combined with TNF is of particular interest due to their different mechanisms of action and different toxicities. MATERIAL AND METHODS
Cell lines A list of the cell lines used in this study, their types and culture media are listed in Table I. Establishment, character-
ization and maintenance of the 9 human lung cancer cell lines used in this study have been described previously (Carney ei al., 1985). Cells were maintained in a humidified atmosphere of 95% air and 5% COz at 37°C. Receptor binding assay I2%TNF (600 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). The binding assay was performed by a modification of the methods described by Ishikura et al. 1989). Cell membranes were prepared from all cell lines and '251-TNF binding to intact membranes was measured as follows. In a final volume of 1 ml, aliquots of membrane preparations in 50 m~ Tris-HC1 buffer PH 7.4 were incubated in triplicate for 60 min at 4°C in the presence or absence of excess unlabelled TNF. After centrifugation and removal of supernatant, pellets were counted using a liquid scintillation gammacounter and the specific binding was calculated as the difference between total binding and binding in the presence of excess unlabelled TNF. Scatchard plots of the data were evaluated using the computer program Ligand. Drugs and drug sensitivity testing Drug sensitivity testing was performed using the semiautomated MTT test (Carmichael et al., 1987). Recombinant human TNF was kindly provided by Cetus, Emeryville, CA. TNF had a bioactivity of 24 X lo6 U/mg and was dissolved in phosphate-buffered saline (PBS), aliquoted and stored at -70°C. Bioactivity of the TNF preparations was assayed as the cytotoxic effect obtained by TNF on the L929 cell line in the presence of 1 kg/ml actinomycin D. TNF alone was tested at concentrations ranging from 10 to 10,OOO U/ml. TNF activity was evaluated after 4-day treatments in all the cell lines and also after 6-day treatments in 5 cell lines. Doxorubicin and VP16 were formulated for clinical use. When combined with anti-neoplastic agents, TNF was tested at concentrations of 10, 100 and 1,OOO U/ml. TNF was added 30 min after cytotoxic drugs, because increased cytotoxic effects were observed when TNF was given after topoisomerase-11-targeted agents (Alexander et al., 19876). Cell exposure to the combination of TNF and doxorubicin or VP16 was continued for 4 days. Each experiment was repeated at least 3 times. RESULTS
Expression of TNF receptors in humn lung cancer cell lines We screened cell membranes prepared from 5 small-cell and 4 non-small-cell lung cancer cell lines for '251-TNF binding activity and found specific binding in 4 out of 9 lines (Table I). *To whom requests for reprints should be addressed, at: NCI-Navy Medical Oncology Branch, National Cancer Institute, Naval Hospital, Bethesda, MD 20814, USA. 'Present address: Dept. Surgery, Chiba University, School of Medicine, 1-8-1 Inohana, Chiba 280, Japan. Received April 26, 1990.
327
T N F A N D L U N G CANCER C E L L LINES
TABLE I - BINDING OF %TNF TO HUMAN SMALL-CELL AND NON-SMALLCELL LUNG CANCER CELL LINES Culture medium’
Subtype
Cell line
N-SCLC
”’I-TNF bound’ (fmovmg orotein)
H-23 H43
NSCLC NCI-H23 NCI-H460 NCI-H596 NCI-676 SCLC NCI-H82 NCI-H146 NCI-H 187 NCI-H660 NCI-H1045 Positive control
Adenocarcinoma Large-cell ca. Adenosquamous Adenocarcinoma
SSM SSM SSM SFM
55
Variant Classic Classic Classic Classic Mouse fibrosarcoma
SSM SSM SSM SFM SFM SSM
33 0 48 0 0 162
’ H-596 H-676
C
8- 25
36 0 0
0
L
2
0’
I
I
I
I
I
I
I
SCLC
A H-660 0 H-146
c
’Culture medium: SSM = serum-supplemented medium (RPMI-1640 medium plus 10%fetal bovine serum). The murine fibrosarcoma cell line L929 was grown in DMEM medium plus 10% fetal calf serum. SFM = serum-free medium (ACL-4 for NSCLC lines and HlTES for SCLC) (Gazdar and Oie, 1986; Simms et al., 1980).-*A 2 nM concentration of radioactive ligand was used. Binding conditions were as described in “Material and Methods”. Values represent specific binding to cell membranes determined in the presence of 100 mi non-labelled TNF.
a
0
10 50 100 500 1000 50001oooo TNF Concentration (U/rnl) FIGURE 2 - Cytotoxicity curves of TNF alone in 4 non-small-cell (N-SCLC) and 5 small-cell (SCLC) human lung cancer cell lines. Treatments were for 4 days. See “Material and Methods” for details. Points are means of at least 3 independent experiments.
Specific binding constituted an average of 15% to 40% of the total binding activity at the 2-nM concentration of ‘251-TNF used in our study. growth being 41% at 10,OOO U/ml TNF concentration (data not Scatchard analysis of the 12’I-TNF binding to the non- shown). No increase in cytotoxicity was observed on the other small-cell line NCI-H23 (Fig. 1) revealed an apparent single cell lines exposed to a 6-day treatment with TNF (data not class of high-affinity saturable binding sites (Bmu = 130 fmoll shown). The fibrosarcoma cell line L929 exhibited a 38% inmg protein; K, = 1.25 X 1 0 - 9 ~ r; = 0.9). hibition of growth after 4-days’ incubation with 10,OOO U/ml TNF concentration, which corresponds to previous experience Cytotoxic effects of TNF alone (Alexander et al., 1987b). TNF alone had very little cytotoxic activity in the 9 lung cancer cell lines, regardless of histologic type (Fig. 2). No Effects of TNF upon cytotoxicity of doxorubicin and VP16 dose-response relationship was seen in most of the cell lines, The effects of TNF on cytotoxicity of doxorubicin and VP16 despite the wide range of TNF concentrations tested. A mar- were studied in 5 cell lines, including 3 non-small-cell and 2 ginal inhibition of cell growth was observed only for NCI-H23, small-cell lung cancer cell lines. Table 11 depicts the IC,, vala lung adenocarcinoma cell line: in this case cell growth was ues obtained with doxorubicin or VP16 with and without TNF. inhibited by 22% at the highest TNF concentrations used Cytotoxicities of doxorubicin and VP16 alone varied widely in (5,oOCrlO,OOO U/ml). When exposure of cells to TNF was the 5 cell lines. No significant effect of TNF on the cytotoxicity prolonged for 6 days, an increase in cytotoxicity was observed for the NCI-H23 cell line, the maximum inhibition of cell TABLE U - CYTOTOXICITY OF DOXOKUBICM AND ETOPOSIDE WITH AND WITHOUT TNF ON 5 LUNG CANCER CELL LINES
30000 -0
c
5 d
Line
Drug
VP16 NCI-H23 I
DX
I
VP16 NCI-H596 DX VP16 NCI-H187 DX VP16
0
10 1251-TNF(nM)
20
FIGURE1 - Saturable binding and Scatchard analysis of the specific ‘2SI-TNFbinding to membranes of the non-small cell lung cancer cell line NCI-H23. Binding conditions were as described in “Material and Methods”. Scatchard analysis of the binding data is shown in the inset.
1Mn ~~
NCI-H460 DX
150,
TNF (U/ml) in 1M
n
NCI-H146 DX VP16
0.061 0.0838 0.0819 0.0327 (0.061) (0.084) (0.046) (0.033) 0.5971 0.5385 0.6396 0.8078 (0.58) (0.872) (0.305) (0.59) 0.2197 0.2165 0.1683 0.1264 (0.129) (0.077) (0.071) (0.127) 0.620 1.592 1.491 0.968 (1.03) (0.694) (0.377) (0.263) 1.108 1.695 1.246 1.292 (0.105) (0.029) (0.193) (0.059) 138.7 165.3 144.3 155.9 (30.6) (48.6) (30) (40.5) 0.406 0.326 0.281 0.388 (0.119) (0.127) (0.021) (0.096) 0.3557 0.406 0.3671 0.337 (0.111) (0.216) (0.127) (0.081) 0.1241 0.2083 0.2606 0.1526 (0.051) (0.11) (0.178) (0.051) 0.3004 0.5362 0.442 0.5926 (0.18) (0.358) (0.484) 10.475)
‘
‘Values are IC,, (SD) of doxorubicin (DX) and etoposide (VP16) expressed as p ~ . Cell survival was determined by the semi-automated M l T assay; cells were treated with drugs in the presence or absence of TNF for 4 days.
328
GIACCONE ET AL.
of the 2 antineoplastic agents was observed in any cell line, as determined by the Wilcoxon test. Moreover, the nonparametric tests of trend did not allow any statement of significance in the joint consideration of all 10 trends. While TNF alone was slightly cytotoxic to cell line NCI-H23, a modest dose-related enhancement of cytotoxicity was noted when TNF was added to doxorubicin or VP16. However, thep, values for the non-parametric test of trends in this cell line, not corrected for the multiplicity of tests, were not significant, suggesting that combined effects were additive. DISCUSSION
Despite the effectiveness of initial chemotherapy in some forms of lung cancer, especially small-cell lung cancer, longterm survivals are achieved in only a minority of patients. The search for new therapeutic modalities has included cytokines combined with conventional cytotoxic agents (Frei and Spriggs, 1989). It is of interest that TNF demonstrates synergism in vitro and in vivo with topoisomerase I1 inhibitors (Alexander et al., 1987a,b; Burgers et al., 1989; Coffman et al., 1989). In addition, treatment with doxorubicin or the epipodophyllotoxins VP16 and VM26 can reverse the TNF resistance of human cell lines in vitro and in xenografts (Fruehauf et al., 1989; Coffrnan et al., 1989; Das et al., 1989). Phase-I studies of TNF in combination with doxorubicin or VP16 are currently in progress (Taylor et al., 1989; Orr et al., 1989). Thus, we studied the interaction of TNF and the topoisomerase-11-targeteddrugs doxorubicin and VP16 in lung cancer cell lines. In a panel of 9 lung cancer cell lines of several histological types, TNF alone in concentrations up to 10,OOO U/ml produced very little cytotoxicity, despite the demonstration of TNF receptors on 4/9 cell lines. Scatchard analysis of the cell line most sensitive to the effect of TNF (NCI-H23) showed high-affinity saturable binding. The presence of receptors, however, appears to be necessary but not sufficient for the cytotoxic effects of TNF in v i m (Baglioni et al., 1985). L929, a mouse fibrosarcoma line exhibiting synergism between TNF
and topoisomerase-11-targeted drugs, expressed relatively higher binding. Our results confirm and extend previous reports of the lack of cytotoxic effects of TNF alone in a total of 7 other lung cancer cell lines tested (Spriggs et al., 1988; Sugarman et al., 1985). In contrast, a recent in vitro phase-11 trial performed on fresh tumor biopsies taken from patients with different malignancies using the clonogenic assay showed a significant activity of TNF alone and in combination with y-interferon on lung cancers (Salmon et al., 1987). The activity of TNF in fresh lung tumor biopsy material might be explained by interactions of this cytokine with immune system cells remaining in the samples. Anti-tumor effects of TNF include a direct cytotoxicity as well as indirect mechanisms via the immune system or via an action on the vascular epithelium (Darzynkiewicz et al., 1984; Sato et al., 1986; Krosnick et al., 1989). Our findings indicate that TNF did not significantly increase the cytotoxic activity of doxorubicin and etoposide in the 5 lung cancer cell lines studied. In only one of these cell lines (NCI-H23) was even a modest (but not significant) increase in the cytotoxicity of these 2 drugs observed, at concentrations of TNF that are not clinically achievable (5000-10,OOO U/ml) (Salmon et al., 1987). Thus, an unequivocal synergistic effect with either of 2 cytotoxic agents could not be demonstrated. In conclusion, unless TNF exerts its in vivo effects by a mechanism that cannot readily be reproduced in vitro, our studies suggest that TNF is unlikely to be active against human lung cancers of either the small-cell or the non-small-cell type. In addition, TNF is not directly synergistic with 2 other cytotoxic agents commonly used in the treatment of lung cancer. ACKNOWLEDGEMENTS
Dr. G. Giaccone was supported by awards from the EORTC Research Training Program and from the Associazione Italiana per la Ricerca sul Cancro. We thank Dr. G. Capranico for helpful discussions.
REFERENCES E.A., OLD, L.J., KASSEL,R.L., GREEN,S., FIORE,N. and ALEXANDER, R.B., ISAACS, J.T. and COFFEY, D.S., Tumor necrosis factor CARSWELL, B., An endotoxin-induced serum factor that causes necrosis enhances the in vino and in vivo efficacy of chemotherapeutic drugs tar- WILLIAMSON, geted at DNA TopoisomeraseI1 in the treatment of murine bladder cancer. of tumors. Proc. nut. Acud. Sci. (Wash.), 72, 3666-3670 (1975b). J. Urol., 138, 421429 (1987~). COFFMAN, F.D., GREEN,L.M., GODWIN,A. and WARE,C.F., CytotoxALEXANDER, R.B., NELSON,W.G. and COFFEY,D.S., Synergistic en- icity mediated by tumor necrosis factor in variant subclones of ME-180 hancement by tumor necrosis factor of in vino cytotoxicity from chemo- cervical carcinoma line: modulation by specific inhibitors of DNA topoitherapeutic drugs targeted at DNA Topoisomerase 11. Cancer Res., 47, somerase II. J . cell. Biochem., 39, 95-105 (1989). 2403-2406 (19876). DARZYNKIEWICZ, Z., WILLIAMSON, B., CARSWELL, E.A. and OLD,L.J., ASHER,A,, MULE,J.J., REICHERT, C.M., SHILONI,E. ~ ~ ~ R O S E N E E Cell R G ,cycle-specific effects of tumor necrosis factor. Cancer Res., 44, 83S.A., Studies on the anti-tumor efficacy of systemically administered re- 90 (1984). combinant tumor necrosis factor against several murine tumors in vivo. J. DAS,A.K., WALTHER, P.J., BUCKLEY, N.J. and POULTON, S.H.M., ReImmunol., 138, 96S914 (1987). combinant human tumor necrosis factor alone and with chemotherapeutic BAGLIONI. C.. MCCANDLESS. S.. TAVERNIER. J. and FIERS.W.. Bindine agents. Arch. Surg., 124, 107-110 (1989). FREI, 111, E. and SPRIGGS,D., Tumor necrosis factor: still a promising of human’&or necrosis factor ’to high affinity receptors ’on HeLa an; lymphoblastoid cells sensitive to growth inhibition. J . biol. Chem., 260, agent. J. clin. Oncol., 7, 291-294 (1989). 13395-1 3397 (1 985). FRUEHAUF, J.P., MIMNAUGH, E., FAIRCHILD, C. and SINHA,B.K.,InBURGERS, J.K., MARSHALL, F.F. and ISAACS, J.T., Enhanced anti-tumor duction of doxorubicin resistance in the MCM breast cancer cell line effects of recombinant human tumor necrosis factor plus VP-16 on meta- confers cross-resistance to tumor necrosis factor. Proceedings Amer. Ass. static renal cell carcinoma in a xenograft model. J. Urol., 142, 16CL164 Cancer Res., Vol. 30, pp. 769, H-28 (1989). (1989). GAZDAR, A.F. and OIE, H.K., Cell culture methods for human lung canCARMICHAEL, J., DEGRAAF,W.G., GAZDAR,A.F., MINNA,J.D. and cer. Cancer Gener. Cyrogener., 19, 5-10 (1986). MITCHELL,J.B., Evaluation of a tetrazolium-based semi-automated col- HARANAKA, K., SATOMI,N. and SAKURAI, A., Antitumor activity of orimetric assay: assessment of chemosensitivitytesting. Cancer Res., 47, murine tumor necrosis factor (TNF) against transplanted murine tumors 936-942 (1987). and heterotransplanted human tumors in nude mice. Inr. J. Cancer, 34, CARNEY, D.N., GAZDAR, A.F., BEPLER,G.,GUCCION, J.G., MARANGOS, 263-267 (1984). P.J., MOODY, T.W., ZWIG, M.H. and MINNA,J.D., Establishment and ISHIKURA, H., HORI, K. and BLOCH,A., Differential biologic effects identification of small cell lung cancer cell lines having classic and variant resulting from bimodal binding of recombinant human tumor necrosis facfeatures. Cancer Res., 45, 2913-2923 (1985). tor to myeloid leukemia cells. Blood, 73, 419424 (1989). J.A., MULE,J.J., MCINTOSH, J.K. and ROSENBERG, S.A., CARSWELL, E.A., GREEN,S., EVERSON, T.C., NATHANSON, T., KROSNICK, BIEDLER, J.L., HELSON, L. and SPENGLER, B.A., Effect of tumor necrosis Augmentation of antitumor efficacy by combination of recombinant tumor in vivo. Cancer Res., 49, necrosis factor and chemotherapeutic agents factor on cultured human melanoma cells. Nature (Lond.),258, 731-732 3729-3733 (1989). ( 19754.
TNF AND LUNG CANCER CELL LINES
MINNA,J.D., HIGGINS,G.E. and GLATSTEIN, E., Cancer of the lung. In: V.T. De Vita Jr., S. Hellman and S.A. Rosenberg (eds.), Cancerprinciples and practice of oncology. 2nd ed., pp. 507-597, Lippincott, Philadelphia (1985). OETTGEN, H.F., CARSWELL, E. A., KASSEL,R.L., FIORE,N., WILLIAMSON, B., HOFFMANN, M.K., HARANAKA, K. and OLD, L.J., Endotoxininduced tumor necrosis factor. Rec. Res. Cancer Res., 75, 207-212 (1980). OLD,L., Tumor necrosis factor (rTNF). Science, 230, 630-632 (1985). ORR,D., OLDHAM, R., LEWIS, M., BERTOLI,L. and BICH,R., Phase I study of the sequential administration of etoposide (VP-16) and recombinant tumor necrosis factor (rTNF; Cetus) in patients with advanced malignancy. Proc. Amer. SOC.clin. Oncol., 8, 190 (1989). Ross, W.E., DNA topoisomerases I1 as targets for cancer therapy. Biochern. Phurmucol., 34, 4191-4195 (1985). SALMON, S.E., YOUNG,L., SCUDERI,P. and CLARK,B., Antineoplastic effects of tumor necrosis factor alone and in combination with gammainterferon on tumor biopsies in clonogenic assay. J . d i n . Oncol., 5 , 18161821 (1987). SATO,N., GOTO,T., HARANAKA, K., SATOMI, N., NARIUCHI, H., MANOY., Actions of tumor necrosis factor on culHIRANO,Y. and SAWASAKI,
329
tured vascular endothelial cells: morphologic modulation, growth inhibition, and cytotoxicity. J. nut. Cancer Znst.. 76, 1113-1 121 (1986). SIMMS,E., GAZDAR, A.F., ABRAMS, P.G. and MINNA,J.D., Growth of human small-cell (oat-cell) carcinoma of the lung in serum-free growth factor supplemented medium. Cancer Res., 40, 4356-4363 (1980). SPRIGGS,D.R., IMAMURA, K., RODRIGUES, C. and KUFE,D.W., Tumor necrosis factor expression in human epithelial tumor cell lines. J . clin. Invest., 81, 455-460 (1988). SUGARMAN, B.J., AGGARWAL, B.B., HASS,P.E., FIGARI,IS., PALLADINO, M.A. and SHEPARD, M.A., Recombinant human tumor necrosis factor-a: effects on proliferation of normal and transformed cells in virro. Science, 230, 943-945 (1985). TAYLOR, C., HERSH,E., PLEZIA, P., ALBERTS, D., MCCLOSKEY, T.M., WIGGINS,C., AHMANN, F., KING, D.and RUDOLPH,A., Phase I study and pharmacokinetic study of recombinant tumor necrosis factor (rTNF) and doxorubicin (DOX). Proc. Amer. SOC.clin. Oncol., 8, 183 (1989). WILLIAMSON, B.D., CARSWELL, E.A., RUBIN,B.Y., PRENDERGAST, J.S. and OLD, L.J., Human tumor necrosis factor produced by human B-cell lines: synergistic cytotoxic interaction with human interferon. Proc. mt. A c u ~Sci. . (Wash.), 80, 5297-5401 (1983).
Publication of the International Union Against Cancer Publication de I'Union Internationale Contre 1e Cancer
EFFECTS OF TUMOR NECROSIS FACTOR, ALONE OR IN COMBINATION WITH TOPOISOMERASE-11-TARGETED DRUGS, ON HUMAN LUNG CANCER CELL LINES Giuseppe GIACCONE, Chikabumi KADOYAMA', Rhoda MANECKJEE, David VENZON,Richard B. ALEXANDER and Adi F. GAZDAR~ NU-Navy Medical Oncology and Surgery Branches, and the Biostatistics and Data Management Section, National Cancer Institute, and Naval Hospital, Bethesda MD 20814, USA. Tumor necrosis factor a (TNF) exhibits cytotoxic activity on some solid tumors and has been reported to be synergistic with topoisomerase-ll-targeted antineoplastic agents. A wide range of T N F concentrations (from 10 to 10,000 U/ml) was tested in 9 human lung cancer cell lines (5 small-cell and 4 non-small-cell carcinomas) using a semi-automated M l T assay. TNF was not cytotoxic in 8 cell lines, while an adenocarcinoma cell line war marginally sensitive to the cytokine. Usi?g Iz51-TNFwe were able to show the presence of specific binding sites for TNF in 4/9 human lung cancer cell lines. Scatchard analysis of the marginally sensitive cell line showed high-affinity,saturable binding. With 5 cell lines we also tested whether TNF affected the cytotoxicity of doxorubicin and etoposide, 2 topoisomerase Il-targeted drugs which are widely used in the therapy of lung cancer. N o significant increase in cytotoxicity was seen when TNF was added to the 2 anti-neoplastic agents. In contrast to certain other human and mouse lines, human lung cancer cell lines appear to be resistant to TNF, despite the presence of the receptor in some of them; moreover, no synergistic effect of TNF and 2 topoisomerase-11-targeted drugs was evident in these human cell lines.
A number of human cytokines, including recombinant tumor necrosis factor (Y (TNF), have recently entered clinical trials for the treatment of human cancers. Naturally occurring TNF is produced by activated macrophages (Old, 1985). In murine tumor models TNF causes acute hemorrhagic necrosis of tumor nodules 24 to 48 hr after local or systemic injection. In vivo anti-tumor effects have been demonstrated both in syngeneic murine tumor models and in human tumor xenografts in nude mice (Carswell et al., 19756; Oettgen et al., 1980; Haranaka et al., 1984; Asher et al., 1987). In vitro studies indicate both cytostatic and cytotoxic effects against a wide range of human tumor cells (Carswell et al., 1975a,b; Oettgen et al., 1980; Williamson et al., 1983; Spriggs et al., 1988; Sugarman et al., 1985). However, TNF is cytotoxic in less than one-third of the cell lines tested (Frei and Spriggs, 1989). Marked enhancement of cytotoxicity has been reported when TNF is added to topoisomerase-11-targeted anti-neoplastic agents on murine L929 tumor cells in vino (Alexander et al., 19876) and on MBT-2 cells both in vitro and in vivo (Alexander et al., 1 9 8 7 ~ ) . Relatively little is known about the effects of TNF on lung cancer, in vivo and in vitro. In the present investigation, TNF cytotoxic activity and receptor studies were evaluated on a panel of human lung cancer cell lines. In addition, we studied the combination of TNF with doxorubicin or etoposide (VP16), 2 anti-neoplastic agents known to act through interactions with topoisomerase II (Ross, 1985). These 2 drugs are among the most active and widely used agents in the chemotherapy of lung cancer (Minna et al., 1985) and a study of the effects of these drugs combined with TNF is of particular interest due to their different mechanisms of action and different toxicities. MATERIAL AND METHODS
Cell lines A list of the cell lines used in this study, their types and culture media are listed in Table I. Establishment, character-
ization and maintenance of the 9 human lung cancer cell lines used in this study have been described previously (Carney ei al., 1985). Cells were maintained in a humidified atmosphere of 95% air and 5% COz at 37°C. Receptor binding assay I2%TNF (600 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). The binding assay was performed by a modification of the methods described by Ishikura et al. 1989). Cell membranes were prepared from all cell lines and '251-TNF binding to intact membranes was measured as follows. In a final volume of 1 ml, aliquots of membrane preparations in 50 m~ Tris-HC1 buffer PH 7.4 were incubated in triplicate for 60 min at 4°C in the presence or absence of excess unlabelled TNF. After centrifugation and removal of supernatant, pellets were counted using a liquid scintillation gammacounter and the specific binding was calculated as the difference between total binding and binding in the presence of excess unlabelled TNF. Scatchard plots of the data were evaluated using the computer program Ligand. Drugs and drug sensitivity testing Drug sensitivity testing was performed using the semiautomated MTT test (Carmichael et al., 1987). Recombinant human TNF was kindly provided by Cetus, Emeryville, CA. TNF had a bioactivity of 24 X lo6 U/mg and was dissolved in phosphate-buffered saline (PBS), aliquoted and stored at -70°C. Bioactivity of the TNF preparations was assayed as the cytotoxic effect obtained by TNF on the L929 cell line in the presence of 1 kg/ml actinomycin D. TNF alone was tested at concentrations ranging from 10 to 10,OOO U/ml. TNF activity was evaluated after 4-day treatments in all the cell lines and also after 6-day treatments in 5 cell lines. Doxorubicin and VP16 were formulated for clinical use. When combined with anti-neoplastic agents, TNF was tested at concentrations of 10, 100 and 1,OOO U/ml. TNF was added 30 min after cytotoxic drugs, because increased cytotoxic effects were observed when TNF was given after topoisomerase-11-targeted agents (Alexander et al., 19876). Cell exposure to the combination of TNF and doxorubicin or VP16 was continued for 4 days. Each experiment was repeated at least 3 times. RESULTS
Expression of TNF receptors in humn lung cancer cell lines We screened cell membranes prepared from 5 small-cell and 4 non-small-cell lung cancer cell lines for '251-TNF binding activity and found specific binding in 4 out of 9 lines (Table I). *To whom requests for reprints should be addressed, at: NCI-Navy Medical Oncology Branch, National Cancer Institute, Naval Hospital, Bethesda, MD 20814, USA. 'Present address: Dept. Surgery, Chiba University, School of Medicine, 1-8-1 Inohana, Chiba 280, Japan. Received April 26, 1990.
327
T N F A N D L U N G CANCER C E L L LINES
TABLE I - BINDING OF %TNF TO HUMAN SMALL-CELL AND NON-SMALLCELL LUNG CANCER CELL LINES Culture medium’
Subtype
Cell line
N-SCLC
”’I-TNF bound’ (fmovmg orotein)
H-23 H43
NSCLC NCI-H23 NCI-H460 NCI-H596 NCI-676 SCLC NCI-H82 NCI-H146 NCI-H 187 NCI-H660 NCI-H1045 Positive control
Adenocarcinoma Large-cell ca. Adenosquamous Adenocarcinoma
SSM SSM SSM SFM
55
Variant Classic Classic Classic Classic Mouse fibrosarcoma
SSM SSM SSM SFM SFM SSM
33 0 48 0 0 162
’ H-596 H-676
C
8- 25
36 0 0
0
L
2
0’
I
I
I
I
I
I
I
SCLC
A H-660 0 H-146
c
’Culture medium: SSM = serum-supplemented medium (RPMI-1640 medium plus 10%fetal bovine serum). The murine fibrosarcoma cell line L929 was grown in DMEM medium plus 10% fetal calf serum. SFM = serum-free medium (ACL-4 for NSCLC lines and HlTES for SCLC) (Gazdar and Oie, 1986; Simms et al., 1980).-*A 2 nM concentration of radioactive ligand was used. Binding conditions were as described in “Material and Methods”. Values represent specific binding to cell membranes determined in the presence of 100 mi non-labelled TNF.
a
0
10 50 100 500 1000 50001oooo TNF Concentration (U/rnl) FIGURE 2 - Cytotoxicity curves of TNF alone in 4 non-small-cell (N-SCLC) and 5 small-cell (SCLC) human lung cancer cell lines. Treatments were for 4 days. See “Material and Methods” for details. Points are means of at least 3 independent experiments.
Specific binding constituted an average of 15% to 40% of the total binding activity at the 2-nM concentration of ‘251-TNF used in our study. growth being 41% at 10,OOO U/ml TNF concentration (data not Scatchard analysis of the 12’I-TNF binding to the non- shown). No increase in cytotoxicity was observed on the other small-cell line NCI-H23 (Fig. 1) revealed an apparent single cell lines exposed to a 6-day treatment with TNF (data not class of high-affinity saturable binding sites (Bmu = 130 fmoll shown). The fibrosarcoma cell line L929 exhibited a 38% inmg protein; K, = 1.25 X 1 0 - 9 ~ r; = 0.9). hibition of growth after 4-days’ incubation with 10,OOO U/ml TNF concentration, which corresponds to previous experience Cytotoxic effects of TNF alone (Alexander et al., 1987b). TNF alone had very little cytotoxic activity in the 9 lung cancer cell lines, regardless of histologic type (Fig. 2). No Effects of TNF upon cytotoxicity of doxorubicin and VP16 dose-response relationship was seen in most of the cell lines, The effects of TNF on cytotoxicity of doxorubicin and VP16 despite the wide range of TNF concentrations tested. A mar- were studied in 5 cell lines, including 3 non-small-cell and 2 ginal inhibition of cell growth was observed only for NCI-H23, small-cell lung cancer cell lines. Table 11 depicts the IC,, vala lung adenocarcinoma cell line: in this case cell growth was ues obtained with doxorubicin or VP16 with and without TNF. inhibited by 22% at the highest TNF concentrations used Cytotoxicities of doxorubicin and VP16 alone varied widely in (5,oOCrlO,OOO U/ml). When exposure of cells to TNF was the 5 cell lines. No significant effect of TNF on the cytotoxicity prolonged for 6 days, an increase in cytotoxicity was observed for the NCI-H23 cell line, the maximum inhibition of cell TABLE U - CYTOTOXICITY OF DOXOKUBICM AND ETOPOSIDE WITH AND WITHOUT TNF ON 5 LUNG CANCER CELL LINES
30000 -0
c
5 d
Line
Drug
VP16 NCI-H23 I
DX
I
VP16 NCI-H596 DX VP16 NCI-H187 DX VP16
0
10 1251-TNF(nM)
20
FIGURE1 - Saturable binding and Scatchard analysis of the specific ‘2SI-TNFbinding to membranes of the non-small cell lung cancer cell line NCI-H23. Binding conditions were as described in “Material and Methods”. Scatchard analysis of the binding data is shown in the inset.
1Mn ~~
NCI-H460 DX
150,
TNF (U/ml) in 1M
n
NCI-H146 DX VP16
0.061 0.0838 0.0819 0.0327 (0.061) (0.084) (0.046) (0.033) 0.5971 0.5385 0.6396 0.8078 (0.58) (0.872) (0.305) (0.59) 0.2197 0.2165 0.1683 0.1264 (0.129) (0.077) (0.071) (0.127) 0.620 1.592 1.491 0.968 (1.03) (0.694) (0.377) (0.263) 1.108 1.695 1.246 1.292 (0.105) (0.029) (0.193) (0.059) 138.7 165.3 144.3 155.9 (30.6) (48.6) (30) (40.5) 0.406 0.326 0.281 0.388 (0.119) (0.127) (0.021) (0.096) 0.3557 0.406 0.3671 0.337 (0.111) (0.216) (0.127) (0.081) 0.1241 0.2083 0.2606 0.1526 (0.051) (0.11) (0.178) (0.051) 0.3004 0.5362 0.442 0.5926 (0.18) (0.358) (0.484) 10.475)
‘
‘Values are IC,, (SD) of doxorubicin (DX) and etoposide (VP16) expressed as p ~ . Cell survival was determined by the semi-automated M l T assay; cells were treated with drugs in the presence or absence of TNF for 4 days.
328
GIACCONE ET AL.
of the 2 antineoplastic agents was observed in any cell line, as determined by the Wilcoxon test. Moreover, the nonparametric tests of trend did not allow any statement of significance in the joint consideration of all 10 trends. While TNF alone was slightly cytotoxic to cell line NCI-H23, a modest dose-related enhancement of cytotoxicity was noted when TNF was added to doxorubicin or VP16. However, thep, values for the non-parametric test of trends in this cell line, not corrected for the multiplicity of tests, were not significant, suggesting that combined effects were additive. DISCUSSION
Despite the effectiveness of initial chemotherapy in some forms of lung cancer, especially small-cell lung cancer, longterm survivals are achieved in only a minority of patients. The search for new therapeutic modalities has included cytokines combined with conventional cytotoxic agents (Frei and Spriggs, 1989). It is of interest that TNF demonstrates synergism in vitro and in vivo with topoisomerase I1 inhibitors (Alexander et al., 1987a,b; Burgers et al., 1989; Coffman et al., 1989). In addition, treatment with doxorubicin or the epipodophyllotoxins VP16 and VM26 can reverse the TNF resistance of human cell lines in vitro and in xenografts (Fruehauf et al., 1989; Coffrnan et al., 1989; Das et al., 1989). Phase-I studies of TNF in combination with doxorubicin or VP16 are currently in progress (Taylor et al., 1989; Orr et al., 1989). Thus, we studied the interaction of TNF and the topoisomerase-11-targeteddrugs doxorubicin and VP16 in lung cancer cell lines. In a panel of 9 lung cancer cell lines of several histological types, TNF alone in concentrations up to 10,OOO U/ml produced very little cytotoxicity, despite the demonstration of TNF receptors on 4/9 cell lines. Scatchard analysis of the cell line most sensitive to the effect of TNF (NCI-H23) showed high-affinity saturable binding. The presence of receptors, however, appears to be necessary but not sufficient for the cytotoxic effects of TNF in v i m (Baglioni et al., 1985). L929, a mouse fibrosarcoma line exhibiting synergism between TNF
and topoisomerase-11-targeted drugs, expressed relatively higher binding. Our results confirm and extend previous reports of the lack of cytotoxic effects of TNF alone in a total of 7 other lung cancer cell lines tested (Spriggs et al., 1988; Sugarman et al., 1985). In contrast, a recent in vitro phase-11 trial performed on fresh tumor biopsies taken from patients with different malignancies using the clonogenic assay showed a significant activity of TNF alone and in combination with y-interferon on lung cancers (Salmon et al., 1987). The activity of TNF in fresh lung tumor biopsy material might be explained by interactions of this cytokine with immune system cells remaining in the samples. Anti-tumor effects of TNF include a direct cytotoxicity as well as indirect mechanisms via the immune system or via an action on the vascular epithelium (Darzynkiewicz et al., 1984; Sato et al., 1986; Krosnick et al., 1989). Our findings indicate that TNF did not significantly increase the cytotoxic activity of doxorubicin and etoposide in the 5 lung cancer cell lines studied. In only one of these cell lines (NCI-H23) was even a modest (but not significant) increase in the cytotoxicity of these 2 drugs observed, at concentrations of TNF that are not clinically achievable (5000-10,OOO U/ml) (Salmon et al., 1987). Thus, an unequivocal synergistic effect with either of 2 cytotoxic agents could not be demonstrated. In conclusion, unless TNF exerts its in vivo effects by a mechanism that cannot readily be reproduced in vitro, our studies suggest that TNF is unlikely to be active against human lung cancers of either the small-cell or the non-small-cell type. In addition, TNF is not directly synergistic with 2 other cytotoxic agents commonly used in the treatment of lung cancer. ACKNOWLEDGEMENTS
Dr. G. Giaccone was supported by awards from the EORTC Research Training Program and from the Associazione Italiana per la Ricerca sul Cancro. We thank Dr. G. Capranico for helpful discussions.
REFERENCES E.A., OLD, L.J., KASSEL,R.L., GREEN,S., FIORE,N. and ALEXANDER, R.B., ISAACS, J.T. and COFFEY, D.S., Tumor necrosis factor CARSWELL, B., An endotoxin-induced serum factor that causes necrosis enhances the in vino and in vivo efficacy of chemotherapeutic drugs tar- WILLIAMSON, geted at DNA TopoisomeraseI1 in the treatment of murine bladder cancer. of tumors. Proc. nut. Acud. Sci. (Wash.), 72, 3666-3670 (1975b). J. Urol., 138, 421429 (1987~). COFFMAN, F.D., GREEN,L.M., GODWIN,A. and WARE,C.F., CytotoxALEXANDER, R.B., NELSON,W.G. and COFFEY,D.S., Synergistic en- icity mediated by tumor necrosis factor in variant subclones of ME-180 hancement by tumor necrosis factor of in vino cytotoxicity from chemo- cervical carcinoma line: modulation by specific inhibitors of DNA topoitherapeutic drugs targeted at DNA Topoisomerase 11. Cancer Res., 47, somerase II. J . cell. Biochem., 39, 95-105 (1989). 2403-2406 (19876). DARZYNKIEWICZ, Z., WILLIAMSON, B., CARSWELL, E.A. and OLD,L.J., ASHER,A,, MULE,J.J., REICHERT, C.M., SHILONI,E. ~ ~ ~ R O S E N E E Cell R G ,cycle-specific effects of tumor necrosis factor. Cancer Res., 44, 83S.A., Studies on the anti-tumor efficacy of systemically administered re- 90 (1984). combinant tumor necrosis factor against several murine tumors in vivo. J. DAS,A.K., WALTHER, P.J., BUCKLEY, N.J. and POULTON, S.H.M., ReImmunol., 138, 96S914 (1987). combinant human tumor necrosis factor alone and with chemotherapeutic BAGLIONI. C.. MCCANDLESS. S.. TAVERNIER. J. and FIERS.W.. Bindine agents. Arch. Surg., 124, 107-110 (1989). FREI, 111, E. and SPRIGGS,D., Tumor necrosis factor: still a promising of human’&or necrosis factor ’to high affinity receptors ’on HeLa an; lymphoblastoid cells sensitive to growth inhibition. J . biol. Chem., 260, agent. J. clin. Oncol., 7, 291-294 (1989). 13395-1 3397 (1 985). FRUEHAUF, J.P., MIMNAUGH, E., FAIRCHILD, C. and SINHA,B.K.,InBURGERS, J.K., MARSHALL, F.F. and ISAACS, J.T., Enhanced anti-tumor duction of doxorubicin resistance in the MCM breast cancer cell line effects of recombinant human tumor necrosis factor plus VP-16 on meta- confers cross-resistance to tumor necrosis factor. Proceedings Amer. Ass. static renal cell carcinoma in a xenograft model. J. Urol., 142, 16CL164 Cancer Res., Vol. 30, pp. 769, H-28 (1989). (1989). GAZDAR, A.F. and OIE, H.K., Cell culture methods for human lung canCARMICHAEL, J., DEGRAAF,W.G., GAZDAR,A.F., MINNA,J.D. and cer. Cancer Gener. Cyrogener., 19, 5-10 (1986). MITCHELL,J.B., Evaluation of a tetrazolium-based semi-automated col- HARANAKA, K., SATOMI,N. and SAKURAI, A., Antitumor activity of orimetric assay: assessment of chemosensitivitytesting. Cancer Res., 47, murine tumor necrosis factor (TNF) against transplanted murine tumors 936-942 (1987). and heterotransplanted human tumors in nude mice. Inr. J. Cancer, 34, CARNEY, D.N., GAZDAR, A.F., BEPLER,G.,GUCCION, J.G., MARANGOS, 263-267 (1984). P.J., MOODY, T.W., ZWIG, M.H. and MINNA,J.D., Establishment and ISHIKURA, H., HORI, K. and BLOCH,A., Differential biologic effects identification of small cell lung cancer cell lines having classic and variant resulting from bimodal binding of recombinant human tumor necrosis facfeatures. Cancer Res., 45, 2913-2923 (1985). tor to myeloid leukemia cells. Blood, 73, 419424 (1989). J.A., MULE,J.J., MCINTOSH, J.K. and ROSENBERG, S.A., CARSWELL, E.A., GREEN,S., EVERSON, T.C., NATHANSON, T., KROSNICK, BIEDLER, J.L., HELSON, L. and SPENGLER, B.A., Effect of tumor necrosis Augmentation of antitumor efficacy by combination of recombinant tumor in vivo. Cancer Res., 49, necrosis factor and chemotherapeutic agents factor on cultured human melanoma cells. Nature (Lond.),258, 731-732 3729-3733 (1989). ( 19754.
TNF AND LUNG CANCER CELL LINES
MINNA,J.D., HIGGINS,G.E. and GLATSTEIN, E., Cancer of the lung. In: V.T. De Vita Jr., S. Hellman and S.A. Rosenberg (eds.), Cancerprinciples and practice of oncology. 2nd ed., pp. 507-597, Lippincott, Philadelphia (1985). OETTGEN, H.F., CARSWELL, E. A., KASSEL,R.L., FIORE,N., WILLIAMSON, B., HOFFMANN, M.K., HARANAKA, K. and OLD, L.J., Endotoxininduced tumor necrosis factor. Rec. Res. Cancer Res., 75, 207-212 (1980). OLD,L., Tumor necrosis factor (rTNF). Science, 230, 630-632 (1985). ORR,D., OLDHAM, R., LEWIS, M., BERTOLI,L. and BICH,R., Phase I study of the sequential administration of etoposide (VP-16) and recombinant tumor necrosis factor (rTNF; Cetus) in patients with advanced malignancy. Proc. Amer. SOC.clin. Oncol., 8, 190 (1989). Ross, W.E., DNA topoisomerases I1 as targets for cancer therapy. Biochern. Phurmucol., 34, 4191-4195 (1985). SALMON, S.E., YOUNG,L., SCUDERI,P. and CLARK,B., Antineoplastic effects of tumor necrosis factor alone and in combination with gammainterferon on tumor biopsies in clonogenic assay. J . d i n . Oncol., 5 , 18161821 (1987). SATO,N., GOTO,T., HARANAKA, K., SATOMI, N., NARIUCHI, H., MANOY., Actions of tumor necrosis factor on culHIRANO,Y. and SAWASAKI,
329
tured vascular endothelial cells: morphologic modulation, growth inhibition, and cytotoxicity. J. nut. Cancer Znst.. 76, 1113-1 121 (1986). SIMMS,E., GAZDAR, A.F., ABRAMS, P.G. and MINNA,J.D., Growth of human small-cell (oat-cell) carcinoma of the lung in serum-free growth factor supplemented medium. Cancer Res., 40, 4356-4363 (1980). SPRIGGS,D.R., IMAMURA, K., RODRIGUES, C. and KUFE,D.W., Tumor necrosis factor expression in human epithelial tumor cell lines. J . clin. Invest., 81, 455-460 (1988). SUGARMAN, B.J., AGGARWAL, B.B., HASS,P.E., FIGARI,IS., PALLADINO, M.A. and SHEPARD, M.A., Recombinant human tumor necrosis factor-a: effects on proliferation of normal and transformed cells in virro. Science, 230, 943-945 (1985). TAYLOR, C., HERSH,E., PLEZIA, P., ALBERTS, D., MCCLOSKEY, T.M., WIGGINS,C., AHMANN, F., KING, D.and RUDOLPH,A., Phase I study and pharmacokinetic study of recombinant tumor necrosis factor (rTNF) and doxorubicin (DOX). Proc. Amer. SOC.clin. Oncol., 8, 183 (1989). WILLIAMSON, B.D., CARSWELL, E.A., RUBIN,B.Y., PRENDERGAST, J.S. and OLD, L.J., Human tumor necrosis factor produced by human B-cell lines: synergistic cytotoxic interaction with human interferon. Proc. mt. A c u ~Sci. . (Wash.), 80, 5297-5401 (1983).
